diff --git a/.gitignore b/.gitignore
index e4e47aa4b705c9035394f5ea15c0d5e653859206..3b236ea5bdd793ed34603010c520fa1f1a43e34b 100644
--- a/.gitignore
+++ b/.gitignore
@@ -9,8 +9,10 @@ buildGCC
_skbuild/
dist/
*.egg-info/
-__pycache__/
+**/__pycache__/
.venv/
+pythonbindings/pyfluids/bindings*
+pythonbindings/pymuparser/bindings*
# IDE
.vscode/
@@ -38,4 +40,7 @@ stl/
.DS_Store
# Settings
-.gitconfig
\ No newline at end of file
+.gitconfig
+
+# User Settings
+CMakeUserPresets.json
\ No newline at end of file
diff --git a/.gitlab-ci.yml b/.gitlab-ci.yml
index b34c5a8f66c1340670b6acd80ea6a9901b2760d1..e171e2e7fbe1984588355f5a833a21160024da32 100644
--- a/.gitlab-ci.yml
+++ b/.gitlab-ci.yml
@@ -1,7 +1,7 @@
###############################################################################
## VirtualFluids CI Pipeline ##
###############################################################################
-image: git.rz.tu-bs.de:4567/irmb/virtualfluids/ubuntu20_04:1.3
+image: git.rz.tu-bs.de:4567/irmb/virtualfluids/ubuntu20_04:1.4
stages:
- build
@@ -49,7 +49,7 @@ stages:
- cd $CI_PROJECT_DIR/$BUILD_FOLDER
- rm -r -f ./*
- cmake .. -LAH
- --preset=all_make
+ --preset=make_all
-DBUILD_WARNINGS_AS_ERRORS=ON
-DCMAKE_CUDA_ARCHITECTURES=60
- make -j4
@@ -75,7 +75,7 @@ clang_10:
- export CXX=clang++
###############################################################################
-msvc_16:
+msvc_17:
stage: build
tags:
@@ -92,14 +92,14 @@ msvc_16:
- git --version
- $env:Path += ";C:\Program Files\CMake\bin\"
- cmake --version
- - $env:Path += ";C:\Program Files (x86)\Microsoft Visual Studio\2019\Enterprise\MSBuild\Current\Bin"
+ - $env:Path += ";C:\Program Files\Microsoft Visual Studio\2022\Enterprise\MSBuild\Current\Bin"
- MSBuild.exe -version
script:
- cd $CI_PROJECT_DIR
- md -force $env:BUILD_FOLDER
- cd $env:BUILD_FOLDER
- - cmake .. --preset=all_msvc -DCMAKE_CUDA_ARCHITECTURES=61 -DBUILD_WARNINGS_AS_ERRORS=ON
+ - cmake .. --preset=msvc_all -DCMAKE_CUDA_ARCHITECTURES=61 -DBUILD_WARNINGS_AS_ERRORS=ON
- MSBuild.exe VirtualFluids.sln /property:Configuration=$env:BUILD_CONFIGURATION /verbosity:minimal /maxcpucount:4
artifacts:
@@ -126,33 +126,44 @@ gcc_9_python:
paths:
- build/
- dist/
+ - _skbuild/
before_script:
- export CCACHE_BASEDIR=$CI_PROJECT_DIR
- export CCACHE_DIR=$CI_PROJECT_DIR/cache
script:
- - python3 setup.py bdist_wheel build_ext --build-temp=build
+ - python3 setup.py bdist_wheel build_ext --build-temp=_skbuild -- -DBUILD_VF_CPU=ON -DBUILD_VF_DOUBLE_ACCURACY=ON -DCMAKE_CXX_COMPILER_LAUNCHER=ccache -DCMAKE_CUDA_COMPILER_LAUNCHER=ccache -DCMAKE_C_COMPILER_LAUNCHER=ccache
###############################################################################
## Container Upload ##
###############################################################################
-build_singularity_image:
+build_poiseuille_test_container:
+ image:
+ name: quay.io/singularity/singularity:v3.10.2
+ entrypoint: [""]
+
stage: container_upload
- needs:
- - gcc_9_python
+ rules:
+ - if: $REMOTE_USER && $REMOTE_HOST && $PRIVATE_KEY && $CI_PIPELINE_SOURCE == "schedule"
+ when: always
+ - if: $CI_PIPELINE_SOURCE == "merge_request_event"
+ when: never
+ - when: manual
+ allow_failure: true
tags:
- linux
- privileged
- rules:
- - if: $CI_COMMIT_TAG
+ artifacts:
+ expire_in: 1 hrs
+ paths:
+ - Containers/PoiseuilleTestContainer.sif
script:
- - singularity build Containers/VirtualFluidsPython.sif Containers/VirtualFluidsPython.def
- - singularity push --docker-username "${CI_REGISTRY_USER}" --docker-password "${CI_REGISTRY_PASSWORD}" Containers/VirtualFluidsPython.sif oras://"$CI_REGISTRY_IMAGE"/"$CI_PROJECT_NAME":"$CI_COMMIT_TAG"
+ - singularity build "Containers/PoiseuilleTestContainer.sif" "Python/SlurmTests/poiseuille/PoiseuilleTestContainer.def"
###############################################################################
## Tests ##
@@ -169,14 +180,14 @@ gcc_9_unit_tests:
- ctest
###############################################################################
-msvc_16_unit_tests:
+msvc_17_unit_tests:
stage: test
tags:
- win
- gpu
- needs: ["msvc_16"]
+ needs: ["msvc_17"]
before_script:
- $env:Path += ";C:\Program Files\CMake\bin\"
@@ -202,6 +213,52 @@ gcc_9_python_bindings_test:
- python3 -m unittest discover -s Python -v
+###############################################################################
+gcc_9_python_hpc_test:
+ image: python:latest
+ stage: test
+
+ needs: ["build_poiseuille_test_container"]
+
+ rules:
+ - if: $REMOTE_USER && $REMOTE_HOST && $PRIVATE_KEY && $CI_PIPELINE_SOURCE == "schedule"
+ when: always
+ - if: $CI_PIPELINE_SOURCE == "merge_request_event"
+ when: never
+ - when: manual
+ allow_failure: true
+
+ before_script:
+ - pip install hpc-rocket
+
+ script:
+ - hpc-rocket launch --watch Python/SlurmTests/poiseuille/rocket.yml
+
+###############################################################################
+multigpu_hpc_test:
+ image: python:latest
+ stage: test
+
+ rules:
+ - if: $REMOTE_USER && $REMOTE_HOST && $PRIVATE_KEY && $CI_PIPELINE_SOURCE == "schedule"
+ when: always
+ - if: $CI_PIPELINE_SOURCE == "merge_request_event"
+ when: never
+ - when: manual
+ allow_failure: true
+
+ before_script:
+ - pip install hpc-rocket
+ - pip install "fieldcompare[all]"
+
+ script:
+ - hpc-rocket launch --watch regression-tests/multigpu_test/rocket.yml
+ - git clone --depth 1 --filter=blob:none --sparse https://github.com/irmb/test_data
+ - cd test_data
+ - git sparse-checkout set regression_tests/gpu/DrivenCavity_4GPU_2Levels
+ - cd ..
+ - fieldcompare dir output/results test_data/regression_tests/gpu/DrivenCavity_4GPU_2Levels --include-files "*.vtu"
+
###############################################################################
## Benchmark ##
###############################################################################
@@ -248,7 +305,7 @@ gpu_numerical_tests:
- cd $CI_PROJECT_DIR/build
- rm -r -f ./*
- cmake ..
- --preset=gpu_numerical_tests_make
+ --preset=make_numerical_tests_gpu
-DCMAKE_CUDA_ARCHITECTURES=60
-DPATH_NUMERICAL_TESTS=/tmp/test_data/numerical_tests_gpu
- make -j4
@@ -319,8 +376,7 @@ clang_build_analyzer_clang_10:
- mkdir -p $CI_PROJECT_DIR/build
- cd $CI_PROJECT_DIR/build
- cmake ..
- -DBUILD_VF_CPU=ON
- -DBUILD_VF_GPU=ON
+ --preset=make_all
-DCMAKE_CUDA_ARCHITECTURES=60
-DCMAKE_CXX_FLAGS=-ftime-trace
- ClangBuildAnalyzer --start .
@@ -352,8 +408,7 @@ include_what_you_use_clang_10:
- mkdir -p $CI_PROJECT_DIR/build
- cd $CI_PROJECT_DIR/build
- cmake ..
- -DBUILD_VF_CPU=ON
- -DBUILD_VF_GPU=ON
+ --preset=make_all
-DCMAKE_CUDA_ARCHITECTURES=60
-DBUILD_VF_INCLUDE_WHAT_YOU_USE=ON
- make
@@ -430,7 +485,7 @@ gcov_gcc_9:
- mkdir -p $CI_PROJECT_DIR/build
- cd $CI_PROJECT_DIR/build
- cmake ..
- --preset=all_make
+ --preset=make_all
-DCMAKE_CUDA_ARCHITECTURES=60
-DBUILD_VF_COVERAGE=ON
- make -j4
@@ -473,6 +528,7 @@ clang-tidy:
- cd $CI_PROJECT_DIR/build
- cmake ..
-DBUILD_VF_CPU=ON
+ -DBUILD_VF_DOUBLE_ACCURACY=ON
-DBUILD_VF_GPU=OFF
- python3 ../utilities/filterCompileCommands.py compile_commands.json
- run-clang-tidy -quiet > clangtidy.txt
@@ -604,26 +660,3 @@ sonar-scanner:
script:
- cd $CI_PROJECT_DIR
- sonar-scanner -X -Dsonar.verbose=true -Dsonar.login=$SONAR_SECURITY_TOKEN
-
-###############################################################################
-## Release ##
-###############################################################################
-create_release:
- stage: release
-
- image: registry.gitlab.com/gitlab-org/release-cli:latest
-
- needs: ["build_singularity_image"]
-
- rules:
- - if: $CI_COMMIT_TAG
-
- script:
- - echo "Creating release with tag $CI_COMMIT_TAG"
- - release-cli create --name "VirtualFluids $CI_COMMIT_TAG" \
- --description "VirtualFluids CFD Simulator" \
- --tag-name "$CI_COMMIT_TAG" \
- --ref "$CI_COMMIT_SHA" \
- --job-token "$CI_JOB_TOKEN" \
- --assets-link="{'name':'VirtualFluidsSingularityImage_OpenMPI','url':'','type':'other','filepath':'Containers/VirtualFluidsOpenMPI.sif'}"
- - build/bin/basicsTests
diff --git a/CITATION.cff b/CITATION.cff
new file mode 100644
index 0000000000000000000000000000000000000000..50d4989d5c269521392644515d716fa93b3cf6e3
--- /dev/null
+++ b/CITATION.cff
@@ -0,0 +1,40 @@
+cff-version: 1.2.0
+message: "If you use this software, please cite it as below."
+type: software
+authors:
+ - family-names: Kutscher
+ given-names: Konstantin
+ orcid: https://orcid.org/0000-0002-1099-1608
+ - family-names: Schönherr
+ given-names: Martin
+ orcid: https://orcid.org/0000-0002-4774-1776
+ - family-names: Geier
+ given-names: Martin
+ orcid: https://orcid.org/0000-0002-8367-9412
+ - family-names: Krafczyk
+ given-names: Manfred
+ orcid: https://orcid.org/0000-0002-8509-0871
+ - family-names: Alihussein
+ given-names: Hussein
+ orcid: https://orcid.org/0000-0003-3656-7028
+ - family-names: Linxweiler
+ given-names: Jan
+ orcid: https://orcid.org/0000-0002-2755-5087
+ - family-names: Peters
+ given-names: Sören
+ orcid: https://orcid.org/0000-0001-5236-3776
+ - family-names: Wellmann
+ given-names: Anna
+ orcid: https://orcid.org/0000-0002-8825-2995
+ - family-names: Safari
+ given-names: Hesameddin
+ orcid: https://orcid.org/0000-0002-2755-5087
+ - family-names: Marcus
+ given-names: Sven
+ orcid: https://orcid.org/0000-0003-3689-2162
+title: "VirtualFluids"
+version: 0.1.0
+license: GPL-3.0-or-later
+repository-code: "https://git.rz.tu-bs.de/irmb/VirtualFluids"
+date-released: "XXXXXXX"
+
diff --git a/CMake/FileUtilities.cmake b/CMake/FileUtilities.cmake
index 151000a681795923d4e31ed8c5f06dfd1e7af7fd..13057ef832b5aa2d7ce303fe55e95a91284f5f56 100644
--- a/CMake/FileUtilities.cmake
+++ b/CMake/FileUtilities.cmake
@@ -5,7 +5,7 @@
## After function call the files are stored in: MY_SRCS
#################################################################################
-macro(includeAllFiles targetName file_path)
+macro(includeAllFiles folderName targetName file_path)
if(NOT DEFINED collectTestFiles)
set(collectTestFiles ON)
endif()
@@ -14,11 +14,11 @@ macro(includeAllFiles targetName file_path)
set(collectProductionFiles ON)
endif()
- includeFiles(${targetName} "${file_path}")
+ includeFiles(${folderName} ${targetName} "${file_path}")
endmacro(includeAllFiles)
-macro(includeProductionFiles targetName file_path)
+macro(includeProductionFiles folderName targetName file_path)
if(NOT DEFINED collectTestFiles)
set(collectTestFiles OFF)
endif()
@@ -27,12 +27,12 @@ macro(includeProductionFiles targetName file_path)
set(collectProductionFiles ON)
endif()
- includeFiles(${targetName} "${file_path}")
+ includeFiles(${folderName} ${targetName} "${file_path}")
endmacro(includeProductionFiles)
-macro(includeTestFiles targetName file_paths)
+macro(includeTestFiles folderName file_paths)
if(NOT DEFINED collectTestFiles)
set(collectTestFiles ON)
endif()
@@ -41,13 +41,13 @@ macro(includeTestFiles targetName file_paths)
set(collectProductionFiles OFF)
endif()
- includeFiles(${targetName} "${file_paths}")
+ includeFiles(${folderName} ${folderName} "${file_paths}")
endmacro(includeTestFiles)
-macro(includeFiles targetName file_paths)
+macro(includeFiles folderName targetName file_paths)
foreach(file ${file_paths})
@@ -57,7 +57,7 @@ macro(includeFiles targetName file_paths)
collectFilesFrom(${file})
if (package_dir)
- setSourceGroupForFilesIn(${file} ${package_dir} ${targetName})
+ setSourceGroupForFilesIn(${file} ${package_dir} ${targetName} ${folderName})
endif()
endforeach()
@@ -90,9 +90,9 @@ endmacro()
-macro(setSourceGroupForFilesIn file package_dir targetName)
+macro(setSourceGroupForFilesIn file package_dir targetName folderName)
#input: target_name PACKAGE_SRCS
- buildSourceGroup(${targetName} ${package_dir})
+ buildSourceGroup(${folderName} ${package_dir})
if(isAllTestSuite)
source_group(${targetName}\\${SOURCE_GROUP} FILES ${file})
@@ -105,20 +105,20 @@ endmacro(setSourceGroupForFilesIn)
-macro(buildSourceGroup targetName path)
-#input: targetName (e.g. lib name, exe name)
+macro(buildSourceGroup folderName path)
+#input: folderName (e.g. name of folder after src/)
unset(SOURCE_GROUP)
string(REPLACE "/" ";" folderListFromPath ${path})
- set(findTargetName 0)
+ set(findFolderName 0)
foreach(folder ${folderListFromPath})
- if(findTargetName)
+ if(findFolderName)
set(SOURCE_GROUP ${SOURCE_GROUP}\\${folder})
endif()
- if(${folder} STREQUAL ${targetName})
- SET(findTargetName 1)
+ if(${folder} STREQUAL ${folderName})
+ SET(findFolderName 1)
endif()
endforeach()
diff --git a/CMake/VirtualFluidsMacros.cmake b/CMake/VirtualFluidsMacros.cmake
index 63503f5f14221bb8cec7670dbdda6aa92497d327..4fd163b2cc1b53fe461ef482d906f4cb1255a76c 100644
--- a/CMake/VirtualFluidsMacros.cmake
+++ b/CMake/VirtualFluidsMacros.cmake
@@ -105,14 +105,15 @@ function(vf_add_library)
set( options )
set( oneValueArgs NAME BUILDTYPE)
- set( multiValueArgs PUBLIC_LINK PRIVATE_LINK FILES FOLDER EXCLUDE)
+ set( multiValueArgs PUBLIC_LINK PRIVATE_LINK FILES FOLDER EXCLUDE MODULEFOLDER)
cmake_parse_arguments( ARG "${options}" "${oneValueArgs}" "${multiValueArgs}" ${ARGN} )
- if(DEFINED ARG_NAME)
+ if(DEFINED ARG_NAME)
set(library_name ${ARG_NAME})
else()
vf_get_library_name (library_name)
endif()
+ vf_get_library_name (folder_name) # folder_name is not equal to library_name when ARG_NAME was set
if(NOT DEFINED ARG_BUILDTYPE)
if(BUILD_SHARED_LIBS)
@@ -122,12 +123,16 @@ function(vf_add_library)
endif()
endif()
+ if(DEFINED ARG_MODULEFOLDER)
+ set(folder_name ${ARG_MODULEFOLDER})
+ endif()
+
status("Configuring the target: ${library_name} (type=${ARG_BUILDTYPE})...")
collectFiles(sourceFiles "${ARG_FILES}" "${ARG_FOLDER}" "${ARG_EXCLUDE}")
- includeProductionFiles (${library_name} "${sourceFiles}")
+ includeProductionFiles (${folder_name} ${library_name} "${sourceFiles}")
#################################################################
### ADD TARGET ###
@@ -325,4 +330,4 @@ function(vf_load_user_apps)
foreach(app IN LISTS USER_APPS)
add_subdirectory(${app})
endforeach()
-endfunction()
\ No newline at end of file
+endfunction()
diff --git a/CMake/cmake_config_files/MOLLOK.config.cmake b/CMake/cmake_config_files/MOLLOK.config.cmake
index f700f3cd7a4b5669ef6ffee9436a1528e50e9dc9..72470da1bc52a242cb8e3c341e0e7f87bb06ab26 100644
--- a/CMake/cmake_config_files/MOLLOK.config.cmake
+++ b/CMake/cmake_config_files/MOLLOK.config.cmake
@@ -12,4 +12,5 @@ set(PATH_NUMERICAL_TESTS "D:/out/numericalTests/")
list(APPEND VF_COMPILER_DEFINITION "PATH_NUMERICAL_TESTS=${PATH_NUMERICAL_TESTS}")
# add invidual apps here
-list(APPEND USER_APPS "apps/gpu/LBM/WTG_RUB")
\ No newline at end of file
+list(APPEND USER_APPS "apps/gpu/LBM/WTG_RUB")
+list(APPEND USER_APPS "apps/gpu/LBM/TGV_3D_GridRef")
diff --git a/CMake/cmake_config_files/MULE.config.cmake b/CMake/cmake_config_files/MULE.config.cmake
index 02f61b7988c5b3af9cd58bc52e46b1b2edfe8aae..2afbce6cc257fa0b8ff4dd7de580cb50c01369f1 100644
--- a/CMake/cmake_config_files/MULE.config.cmake
+++ b/CMake/cmake_config_files/MULE.config.cmake
@@ -1 +1,4 @@
-SET(CMAKE_CUDA_ARCHITECTURES "75")
\ No newline at end of file
+SET(CMAKE_CUDA_ARCHITECTURES "75")
+
+list(APPEND USER_APPS "apps/gpu/LBM/ActuatorLine")
+list(APPEND USER_APPS "apps/gpu/LBM/SphereScaling")
diff --git a/CMake/cmake_config_files/PHOENIX.config.cmake b/CMake/cmake_config_files/PHOENIX.config.cmake
index d31d8684a53a769e48408ad5febe7d2c6b22c623..5ca4d9821d918f66745fc27363975811dc278440 100644
--- a/CMake/cmake_config_files/PHOENIX.config.cmake
+++ b/CMake/cmake_config_files/PHOENIX.config.cmake
@@ -28,7 +28,7 @@ set(CMAKE_CUDA_ARCHITECTURES 60) # NVIDIA Tesla P100
set(GPU_APP "apps/gpu/LBM/")
list(APPEND USER_APPS
- # "${GPU_APP}DrivenCavityMultiGPU"
+ "${GPU_APP}DrivenCavityMultiGPU"
# "${GPU_APP}SphereScaling"
# "${GPU_APP}MusselOyster"
)
diff --git a/CMakeLists.txt b/CMakeLists.txt
index b3af407acd66ec3223f55de7753df879786ce561..c6498bf19bb021f3ae19d69c4131aa56476149be 100644
--- a/CMakeLists.txt
+++ b/CMakeLists.txt
@@ -9,7 +9,7 @@
cmake_minimum_required(VERSION 3.15..3.20 FATAL_ERROR)
project(VirtualFluids
- VERSION 1.0.0
+ VERSION 0.1.0
DESCRIPTION "CFD code based on the Lattice Boltzmann Method"
HOMEPAGE_URL "https://www.tu-braunschweig.de/irmb/forschung/virtualfluids"
LANGUAGES CXX)
diff --git a/CMakePresets.json b/CMakePresets.json
index 0f360fd303cdcad923b01d56df5c6d48ad62ca2c..6e2658d148bddf55950e5849adcf10709a8b8caf 100644
--- a/CMakePresets.json
+++ b/CMakePresets.json
@@ -1,34 +1,47 @@
{
- "version": 2,
+ "version": 3,
"cmakeMinimumRequired": {
"major": 3,
- "minor": 20,
+ "minor": 21,
"patch": 0
},
"configurePresets": [
{
"name": "default",
+ "binaryDir": "build",
+ "hidden": true
+ },
+ {
+ "name": "msvc",
"hidden": true,
- "binaryDir": "${sourceDir}/build/",
- "cacheVariables": {
- "BUILD_VF_UNIT_TESTS": "ON"
- }
+ "generator": "Visual Studio 17 2022",
+ "architecture": "x64",
+ "condition": {
+ "type": "equals",
+ "lhs": "${hostSystemName}",
+ "rhs": "Windows"
+ }
},
{
- "name": "default_make",
- "inherits": "default",
+ "name": "make",
"hidden": true,
- "generator": "Unix Makefiles"
+ "generator": "Unix Makefiles",
+ "condition": {
+ "type": "notEquals",
+ "lhs": "${hostSystemName}",
+ "rhs": "Windows"
+ }
},
{
- "name": "default_msvc",
- "inherits": "default",
+ "name": "unit_tests",
"hidden": true,
- "generator": "Visual Studio 16 2019",
- "architecture": "x64"
+ "cacheVariables": {
+ "BUILD_VF_UNIT_TESTS": "ON"
+ }
},
{
- "name": "default_cpu",
+ "name": "cpu",
+ "inherits": "default",
"hidden": true,
"description": "CPU build of VirtualFluids",
"cacheVariables": {
@@ -37,7 +50,8 @@
}
},
{
- "name": "default_gpu",
+ "name": "gpu",
+ "inherits": "default",
"hidden": true,
"description": "GPU build of VirtualFluids",
"cacheVariables": {
@@ -46,9 +60,10 @@
}
},
{
- "name": "default_gpu_numerical_tests",
+ "name": "gpu_numerical_tests",
"inherits": [
- "default_gpu"
+ "gpu",
+ "unit_tests"
],
"hidden": true,
"description": "GPU numerical tests of VirtualFluids",
@@ -58,78 +73,74 @@
}
},
{
- "name": "default_all",
- "hidden": true,
- "description": "All build of VirtualFluids",
+ "name": "make_all",
"inherits": [
- "default_cpu",
- "default_gpu"
+ "cpu",
+ "gpu",
+ "unit_tests",
+ "make"
],
- "cacheVariables": {
- "BUILD_VF_DOUBLE_ACCURACY": "ON"
- }
+ "displayName": "all make configuration"
},
{
- "name": "cpu_make",
+ "name": "make_cpu",
"inherits": [
- "default_make",
- "default_cpu"
+ "cpu",
+ "unit_tests",
+ "make"
],
"displayName": "cpu make configuration"
},
{
- "name": "cpu_msvc",
- "inherits": [
- "default_msvc",
- "default_cpu"
- ],
- "displayName": "cpu msvc configuration"
- },
- {
- "name": "gpu_make",
+ "name": "make_gpu",
"inherits": [
- "default_make",
- "default_gpu"
+ "gpu",
+ "unit_tests",
+ "make"
],
"displayName": "gpu make configuration"
},
{
- "name": "gpu_msvc",
+ "name": "msvc_all",
"inherits": [
- "default_msvc",
- "default_gpu"
+ "cpu",
+ "gpu",
+ "unit_tests",
+ "msvc"
],
- "displayName": "gpu msvc configuration"
+ "displayName": "all msvc configuration"
},
{
- "name": "all_make",
+ "name": "msvc_cpu",
"inherits": [
- "default_make",
- "default_all"
+ "cpu",
+ "unit_tests",
+ "msvc"
],
- "displayName": "all make configuration"
+ "displayName": "cpu msvc configuration"
},
{
- "name": "all_msvc",
+ "name": "msvc_gpu",
"inherits": [
- "default_msvc",
- "default_all"
+ "gpu",
+ "unit_tests",
+ "msvc"
],
- "displayName": "all msvc configuration"
+ "displayName": "gpu msvc configuration"
},
{
- "name": "gpu_numerical_tests_make",
+ "name": "make_numerical_tests_gpu",
"inherits": [
- "default_make",
- "default_gpu_numerical_tests"
+ "gpu_numerical_tests",
+ "make"
],
"displayName": "gpu numerical tests make configuration"
},
{
- "name": "gpu_numerical_tests_msvc",
+ "name": "msvc_numerical_tests_gpu",
"inherits": [
- "default_msvc",
- "default_gpu_numerical_tests"
+ "msvc",
+ "gpu_numerical_tests"
],
"displayName": "gpu numerical tests msvc configuration"
}
diff --git a/MANIFEST.in b/MANIFEST.in
new file mode 100644
index 0000000000000000000000000000000000000000..adafcf99560acd9da79aa060194df8263b6e77e0
--- /dev/null
+++ b/MANIFEST.in
@@ -0,0 +1 @@
+include pythonbindings/*/bindings*
\ No newline at end of file
diff --git a/Python/SlurmTests/poiseuille/PoiseuilleTestContainer.def b/Python/SlurmTests/poiseuille/PoiseuilleTestContainer.def
index a3836e7906b9be66ec79f68bf53ccc079db9d9ef..d31a7b82a4e9e988f815139fb46318d231d450f8 100644
--- a/Python/SlurmTests/poiseuille/PoiseuilleTestContainer.def
+++ b/Python/SlurmTests/poiseuille/PoiseuilleTestContainer.def
@@ -1,11 +1,13 @@
BootStrap: docker
From: ubuntu:20.04
+Stage: build
%files
3rdParty 3rdParty
apps apps
CMake CMake
Python Python
+ pythonbindings pythonbindings
src src
CMakeLists.txt CMakeLists.txt
cpu.cmake cpu.cmake
@@ -19,7 +21,8 @@ From: ubuntu:20.04
apt-get update && \
apt-get install -y \
build-essential \
- cmake=3.16.3-1ubuntu1 \
+ ccache \
+ git \
python3 \
python3-dev \
python3-pip \
@@ -27,10 +30,32 @@ From: ubuntu:20.04
libomp-dev \
libgl1
- pip3 install setuptools wheel numpy scipy pyvista
+ pip3 install setuptools wheel cmake numpy scipy pyvista scikit-build
export PYTHONPATH=Python
- python3 /setup.py install
+ python3 /setup.py bdist_wheel build_ext --build-temp=_skbuild -- -DBUILD_VF_CPU=ON -DBUILD_VF_DOUBLE_ACCURACY=ON
+
+ pip3 install $(find dist/*.whl)
+
+
+BootStrap: docker
+From: ubuntu:20.04
+Stage: runtime
+
+%files from build
+ Python Python
+ dist dist
+
+%post
+ export DEBIAN_FRONTEND=noninteractive
+ apt-get update && \
+ apt-get install -y \
+ python3 \
+ python3-pip \
+ mpich \
+ libgl1
+
+ pip3 install $(find dist/*.whl)
%environment
export PYTHONPATH=/Python
diff --git a/Python/SlurmTests/poiseuille/rocket.yml b/Python/SlurmTests/poiseuille/rocket.yml
new file mode 100644
index 0000000000000000000000000000000000000000..b186469a4d3fd4b8edfafa4fc3f6dcd64e311d70
--- /dev/null
+++ b/Python/SlurmTests/poiseuille/rocket.yml
@@ -0,0 +1,23 @@
+host: $REMOTE_HOST
+user: $REMOTE_USER
+private_keyfile: $PRIVATE_KEY
+
+copy:
+ - from: Python/SlurmTests/poiseuille/slurm.job
+ to: poiseuille_test/slurm.job
+ overwrite: true
+
+ - from: Containers/PoiseuilleTestContainer.sif
+ to: poiseuille_test/PoiseuilleTestContainer.sif
+ overwrite: true
+
+collect:
+ - from: poiseuille_test/POISEUILLE_TEST.out
+ to: POISEUILLE_TEST.out
+ overwrite: true
+
+#clean:
+# - poiseuille_test/PoiseuilleTestContainer.sif
+
+sbatch: poiseuille_test/slurm.job
+continue_if_job_fails: true
diff --git a/Python/SlurmTests/poiseuille/settings.py b/Python/SlurmTests/poiseuille/settings.py
index 4b4a1e4e9cc7f6118a60c22a40c70b027e3ac4e2..a3cdc5dc8b627612c2d57a58db36c9fbaa72efac 100644
--- a/Python/SlurmTests/poiseuille/settings.py
+++ b/Python/SlurmTests/poiseuille/settings.py
@@ -1,25 +1,58 @@
+r"""
+=======================================================================================
+ ____ ____ __ ______ __________ __ __ __ __
+ \ \ | | | | | _ \ |___ ___| | | | | / \ | |
+ \ \ | | | | | |_) | | | | | | | / \ | |
+ \ \ | | | | | _ / | | | | | | / /\ \ | |
+ \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
+ \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
+ \ \ | | ________________________________________________________________
+ \ \ | | | ______________________________________________________________|
+ \ \| | | | __ __ __ __ ______ _______
+ \ | | |_____ | | | | | | | | | _ \ / _____)
+ \ | | _____| | | | | | | | | | | \ \ \_______
+ \ | | | | |_____ | \_/ | | | | |_/ / _____ |
+ \ _____| |__| |________| \_______/ |__| |______/ (_______/
+
+ This file is part of VirtualFluids. VirtualFluids is free software: you can
+ redistribute it and/or modify it under the terms of the GNU General Public
+ License as published by the Free Software Foundation, either version 3 of
+ the License, or (at your option) any later version.
+
+ VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
+ ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
+ for more details.
+
+ You should have received a copy of the GNU General Public License along
+ with VirtualFluids (see COPYING.txt). If not, see <http://www.gnu.org/licenses/>.
+
+! \file settings.py
+! \ingroup Poiseuille
+! \author Sven Marcus, Henry Korb
+=======================================================================================
+"""
import os
from acousticscaling import OneDirectionalAcousticScaling
-from pyfluids.cpu.kernel import LBMKernel, KernelType
-from pyfluids.cpu.parameters import RuntimeParameters, GridParameters, PhysicalParameters
+from pyfluids import cpu
-grid_params = GridParameters()
+grid_params = cpu.parameters.GridParameters()
grid_params.node_distance = 1
grid_params.number_of_nodes_per_direction = [1, 1, 16]
grid_params.blocks_per_direction = [1, 1, 4]
grid_params.periodic_boundary_in_x1 = True
grid_params.periodic_boundary_in_x2 = True
-physical_params = PhysicalParameters()
+physical_params = cpu.parameters.PhysicalParameters()
physical_params.lattice_viscosity = 1e-4
-runtime_params = RuntimeParameters()
+runtime_params = cpu.parameters.RuntimeParameters()
runtime_params.number_of_threads = int(os.environ["PYFLUIDS_NUM_THREADS"])
runtime_params.number_of_timesteps = 4_000_000
runtime_params.timestep_log_interval = 1_000_000
-kernel = LBMKernel(KernelType.CompressibleCumulantFourthOrderViscosity)
+kernel = cpu.kernel.LBMKernel(cpu.kernel.KernelType.CompressibleCumulantFourthOrderViscosity)
kernel.use_forcing = True
kernel.forcing_in_x1 = 5e-10
diff --git a/Python/SlurmTests/poiseuille/simulation_runner.py b/Python/SlurmTests/poiseuille/simulation_runner.py
index 03fb24be7ea1a6468ae25ec3aa40ab59962ef91e..d54a35e72b298562f8ccec82677089f3898eec9b 100644
--- a/Python/SlurmTests/poiseuille/simulation_runner.py
+++ b/Python/SlurmTests/poiseuille/simulation_runner.py
@@ -5,7 +5,7 @@ from poiseuille.simulation import run_simulation
from pyfluids.cpu.writer import Writer, OutputFormat
-scale_level = int(os.environ["PYFLUIDS_SCALE_LEVEL"])
+scale_level = int(os.environ.get("PYFLUIDS_SCALE_LEVEL", 1))
grid_params, physical_params, runtime_params, kernel = Scaling.configuration_for_scale_level(scale_level)
writer = Writer()
diff --git a/Python/SlurmTests/poiseuille/slurm.job b/Python/SlurmTests/poiseuille/slurm.job
index 488fc9a42f261d69a8212cff389721fdfb9cbf6e..b4e4da271920479ade008b28d4d2e6ce6343c3d3 100644
--- a/Python/SlurmTests/poiseuille/slurm.job
+++ b/Python/SlurmTests/poiseuille/slurm.job
@@ -1,5 +1,6 @@
#!/bin/bash
#SBATCH -J PyFluidsTest
+#SBATCH -o poiseuille_test/POISEUILLE_TEST.out
#SBATCH --nodes=1
#SBATCH --ntasks-per-node=1
#SBATCH --cpus-per-task=20
@@ -9,6 +10,9 @@
#SBATCH --partition=standard
source $HOME/.bashrc
+module load singularity/3.9.9
+
+cd poiseuille_test
echo "PyFluids Poiseuille Test Case"
echo "Number of tasks: ${SLURM_NTASKS}"
diff --git a/Python/acousticscaling.py b/Python/acousticscaling.py
index a664b8e924d648b680562b9aef11bee87b3562b1..7e71fed9fdd9f86415261ef4e22797021581f60c 100644
--- a/Python/acousticscaling.py
+++ b/Python/acousticscaling.py
@@ -1,22 +1,55 @@
-from pyfluids.cpu.kernel import LBMKernel
-from pyfluids.cpu.parameters import GridParameters, PhysicalParameters, RuntimeParameters
+r"""
+=======================================================================================
+ ____ ____ __ ______ __________ __ __ __ __
+ \ \ | | | | | _ \ |___ ___| | | | | / \ | |
+ \ \ | | | | | |_) | | | | | | | / \ | |
+ \ \ | | | | | _ / | | | | | | / /\ \ | |
+ \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
+ \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
+ \ \ | | ________________________________________________________________
+ \ \ | | | ______________________________________________________________|
+ \ \| | | | __ __ __ __ ______ _______
+ \ | | |_____ | | | | | | | | | _ \ / _____)
+ \ | | _____| | | | | | | | | | | \ \ \_______
+ \ | | | | |_____ | \_/ | | | | |_/ / _____ |
+ \ _____| |__| |________| \_______/ |__| |______/ (_______/
+
+ This file is part of VirtualFluids. VirtualFluids is free software: you can
+ redistribute it and/or modify it under the terms of the GNU General Public
+ License as published by the Free Software Foundation, either version 3 of
+ the License, or (at your option) any later version.
+
+ VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
+ ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
+ for more details.
+
+ You should have received a copy of the GNU General Public License along
+ with VirtualFluids (see COPYING.txt). If not, see <http://www.gnu.org/licenses/>.
+
+! \file acousticscaling.py
+! \ingroup tests
+! \author Sven Marcus, Henry Korb
+=======================================================================================
+"""
+from pyfluids import cpu
class OneDirectionalAcousticScaling:
- def __init__(self, grid_parameters: GridParameters,
- physical_parameters: PhysicalParameters,
- runtime_parameters: RuntimeParameters,
- kernel: LBMKernel):
+ def __init__(self, grid_parameters: cpu.parameters.GridParameters,
+ physical_parameters: cpu.parameters.PhysicalParameters,
+ runtime_parameters: cpu.parameters.RuntimeParameters,
+ kernel: cpu.kernel.LBMKernel):
self._grid_params = grid_parameters
self._physical_params = physical_parameters
self._runtime_params = runtime_parameters
self._kernel = kernel
- def configuration_for_scale_level(self, level: int = 1) -> tuple[GridParameters,
- PhysicalParameters,
- RuntimeParameters,
- LBMKernel]:
+ def configuration_for_scale_level(self, level: int = 1) -> tuple[cpu.parameters.GridParameters,
+ cpu.parameters.PhysicalParameters,
+ cpu.parameters.RuntimeParameters,
+ cpu.kernel.LBMKernel]:
if level < 0:
raise ValueError("level must be >= 0")
@@ -27,8 +60,8 @@ class OneDirectionalAcousticScaling:
return grid_params, physical_params, runtime_params, kernel
- def clone_grid_params_for_level(self, level) -> GridParameters:
- grid_params = GridParameters()
+ def clone_grid_params_for_level(self, level) -> cpu.parameters.GridParameters:
+ grid_params = cpu.parameters.GridParameters()
grid_params.reference_direction_index = self._grid_params.reference_direction_index
grid_params.periodic_boundary_in_x1 = self._grid_params.periodic_boundary_in_x1
grid_params.periodic_boundary_in_x2 = self._grid_params.periodic_boundary_in_x2
@@ -51,7 +84,7 @@ class OneDirectionalAcousticScaling:
return grid_params
def clone_physical_parameters(self, level):
- physical_params = PhysicalParameters()
+ physical_params = cpu.parameters.PhysicalParameters()
physical_params.lattice_viscosity = self._physical_params.lattice_viscosity
if level > 0:
@@ -60,7 +93,7 @@ class OneDirectionalAcousticScaling:
return physical_params
def clone_runtime_params_for_level(self, level):
- runtime_params = RuntimeParameters()
+ runtime_params = cpu.parameters.RuntimeParameters()
runtime_params.number_of_timesteps = self._runtime_params.number_of_timesteps
runtime_params.number_of_threads = self._runtime_params.number_of_threads
runtime_params.timestep_log_interval = self._runtime_params.timestep_log_interval
@@ -71,7 +104,7 @@ class OneDirectionalAcousticScaling:
return runtime_params
def clone_kernel_for_level(self, level):
- kernel = LBMKernel(self._kernel.type)
+ kernel = cpu.kernel.LBMKernel(self._kernel.type)
kernel.use_forcing = self._kernel.use_forcing
kernel.forcing_in_x1 = self._kernel.forcing_in_x1
kernel.forcing_in_x2 = self._kernel.forcing_in_x2
diff --git a/Python/actuator_line/actuator_line.py b/Python/actuator_line/actuator_line.py
index 6e3c8608617df1267535984d53307dea9184c6ab..721af737ff6ef3340c3c2f6204aa6a7824cd1d2f 100644
--- a/Python/actuator_line/actuator_line.py
+++ b/Python/actuator_line/actuator_line.py
@@ -1,23 +1,48 @@
+r"""
+=======================================================================================
+ ____ ____ __ ______ __________ __ __ __ __
+ \ \ | | | | | _ \ |___ ___| | | | | / \ | |
+ \ \ | | | | | |_) | | | | | | | / \ | |
+ \ \ | | | | | _ / | | | | | | / /\ \ | |
+ \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
+ \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
+ \ \ | | ________________________________________________________________
+ \ \ | | | ______________________________________________________________|
+ \ \| | | | __ __ __ __ ______ _______
+ \ | | |_____ | | | | | | | | | _ \ / _____)
+ \ | | _____| | | | | | | | | | | \ \ \_______
+ \ | | | | |_____ | \_/ | | | | |_/ / _____ |
+ \ _____| |__| |________| \_______/ |__| |______/ (_______/
+
+ This file is part of VirtualFluids. VirtualFluids is free software: you can
+ redistribute it and/or modify it under the terms of the GNU General Public
+ License as published by the Free Software Foundation, either version 3 of
+ the License, or (at your option) any later version.
+
+ VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
+ ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
+ for more details.
+
+ You should have received a copy of the GNU General Public License along
+ with VirtualFluids (see COPYING.txt). If not, see <http://www.gnu.org/licenses/>.
+
+! \file actuator_line.py
+! \ingroup actuator_line
+! \author Henry Korb, Henrik Asmuth
+=======================================================================================
+"""
#%%
import numpy as np
from pathlib import Path
from mpi4py import MPI
-from pyfluids import basics, gpu, logger
+from pyfluids.bindings import basics, gpu, logger
#%%
-reference_diameter = 126
-
-length = np.array([29,6,6])*reference_diameter
-viscosity = 1.56e-5
-velocity = 9
-mach = 0.1
-nodes_per_diameter = 32
-
-sim_name = "ActuatorLine"
-config_file = Path(__file__).parent/Path("config.txt")
+sim_name = "ABL"
+config_file = Path(__file__).parent/"configActuatorLine.txt"
output_path = Path(__file__).parent/Path("output")
output_path.mkdir(exist_ok=True)
-t_out = 100.
-t_end = 500.
+
#%%
logger.Logger.initialize_logger()
@@ -25,87 +50,175 @@ basics.logger.Logger.add_stdout()
basics.logger.Logger.set_debug_level(basics.logger.Level.INFO_LOW)
basics.logger.Logger.time_stamp(basics.logger.TimeStamp.ENABLE)
basics.logger.Logger.enable_printed_rank_numbers(True)
-# %%
-comm = gpu.Communicator.get_instance()
#%%
grid_factory = gpu.grid_generator.GridFactory.make()
grid_builder = gpu.grid_generator.MultipleGridBuilder.make_shared(grid_factory)
+communicator = gpu.Communicator.get_instance()
-#%%
-dx = reference_diameter/nodes_per_diameter
-
-grid_builder.add_coarse_grid(0.0, 0.0, 0.0, *length, dx)
-grid_builder.set_periodic_boundary_condition(False, False, False)
-grid_builder.build_grids(basics.LbmOrGks.LBM, False)
-#%%
config = basics.ConfigurationFile()
config.load(str(config_file))
+
+para = gpu.Parameter(communicator.get_number_of_process(), communicator.get_pid(), config)
+bc_factory = gpu.BoundaryConditionFactory()
+
#%%
-para = gpu.Parameter(config, comm.get_number_of_process(), comm.get_pid())
+turbine_diameter = config.get_float_value("turbineDiameter", 126)
+boundary_layer_height = config.get_float_value("boundaryLayerHeight", 1000)
+z0 = config.get_float_value("z0", 0.1)
+u_star = config.get_float_value("u_star", 0.4)
+
+kappa = config.get_float_value("vonKarmanConstant", 0.4) # von Karman constant
+
+viscosity = config.get_float_value("viscosity", 1.56e-5)
+
+velocity = 0.5*u_star/kappa*np.log(boundary_layer_height/z0+1) #0.5 times max mean velocity at the top in m/s
+
+mach = config.get_float_value("Ma", 0.1)
+nodes_per_height = config.get_uint_value("nz", 64)
+
+
+turb_pos = np.array([3,3,3])*turbine_diameter
+epsilon = config.get_float_value("SmearingWidth", 5)
+density = config.get_float_value("Density", 1.225)
+level = 0
+n_blades = 3
+n_blade_nodes = config.get_int_value("NumberOfNodesPerAL", 32)
+
+read_precursor = config.get_bool_value("readPrecursor", False)
+
+if read_precursor:
+ nTReadPrecursor = config.get_int_value("nTimestepsReadPrecursor")
+ use_distributions = config.get_bool_value("useDistributions", False)
+ precursor_directory = config.get_string_value("precursorDirectory")
+
+# all in s
+t_start_out = config.get_float_value("tStartOut")
+t_out = config.get_float_value("tOut")
+t_end = config.get_float_value("tEnd") # total time of simulation
+t_start_averaging = config.get_float_value("tStartAveraging")
+t_start_tmp_averaging = config.get_float_value("tStartTmpAveraging")
+t_averaging = config.get_float_value("tAveraging")
+t_start_out_probe = config.get_float_value("tStartOutProbe")
+t_out_probe = config.get_float_value("tOutProbe")
+
+#%%
+length = np.array([6,4,1])*boundary_layer_height
+dx = boundary_layer_height/nodes_per_height
dt = dx * mach / (np.sqrt(3) * velocity)
-velocity_lb = velocity * dt / dx # LB units
-viscosity_lb = viscosity * dt / (dx * dx) # LB units
+velocity_ratio = dx/dt
+velocity_LB = velocity / velocity_ratio # LB units
+viscosity_LB = viscosity / (velocity_ratio * dx) # LB units
+pressure_gradient = u_star * u_star / boundary_layer_height
+pressure_gradient_LB = pressure_gradient * (dt*dt)/dx
+
+logger.vf_log_info(f"velocity [dx/dt] = {velocity_LB}")
+logger.vf_log_info(f"dt = {dt}")
+logger.vf_log_info(f"dx = {dx}")
+logger.vf_log_info(f"viscosity [10^8 dx^2/dt] = {viscosity_LB*1e8}")
+logger.vf_log_info(f"u* /(dx/dt) = {u_star*dt/dx}")
+logger.vf_log_info(f"dpdx = {pressure_gradient}")
+logger.vf_log_info(f"dpdx /(dx/dt^2) = {pressure_gradient_LB}")
+
#%%
-para.set_devices([0])
para.set_output_prefix(sim_name)
-para.set_output_path(str(output_path))
-para.set_f_name(para.get_output_path() + "/" + para.get_output_prefix())
para.set_print_files(True)
-para.set_max_level(1)
-#%%
-para.set_velocity(velocity_lb)
-para.set_viscosity(viscosity_lb)
+
+para.set_forcing(pressure_gradient_LB, 0, 0)
+para.set_velocity_LB(velocity_LB)
+para.set_viscosity_LB(viscosity_LB)
para.set_velocity_ratio(dx/dt)
para.set_viscosity_ratio(dx*dx/dt)
-para.set_main_kernel("TurbulentViscosityCumulantK17CompChim")
-para.set_use_AMD(True)
-para.set_SGS_constant(0.083)
+para.set_density_ratio(1.0)
-def init_func(coord_x, coord_y, coord_z):
- return [0.0, velocity_lb, 0.0, 0.0]
+para.set_main_kernel("CumulantK17")
-para.set_initial_condition(init_func)
-para.set_t_out(int(t_out/dt))
-para.set_t_end(int(t_end/dt))
+para.set_timestep_start_out(int(t_start_out/dt))
+para.set_timestep_out(int(t_out/dt))
+para.set_timestep_end(int(t_end/dt))
para.set_is_body_force(True)
-
#%%
-grid_builder.set_velocity_boundary_condition(gpu.SideType.MX, velocity_lb, 0.0, 0.0)
-
-grid_builder.set_velocity_boundary_condition(gpu.SideType.MY, velocity_lb, 0.0, 0.0)
-grid_builder.set_velocity_boundary_condition(gpu.SideType.PY, velocity_lb, 0.0, 0.0)
-
-grid_builder.set_velocity_boundary_condition(gpu.SideType.MZ, velocity_lb, 0.0, 0.0)
-grid_builder.set_velocity_boundary_condition(gpu.SideType.PZ, velocity_lb, 0.0, 0.0)
+tm_factory = gpu.TurbulenceModelFactory(para)
+tm_factory.read_config_file(config)
+#%%
+grid_scaling_factory = gpu.GridScalingFactory()
+grid_scaling_factory.set_scaling_factory(gpu.GridScaling.ScaleCompressible)
-grid_builder.set_pressure_boundary_condition(gpu.SideType.PX, 0.0)
+grid_builder.add_coarse_grid(0.0, 0.0, 0.0, *length, dx)
+grid_builder.set_periodic_boundary_condition(not read_precursor, True, False)
+grid_builder.build_grids(basics.LbmOrGks.LBM, False)
+sampling_offset = 2
+if read_precursor:
+ precursor = gpu.create_file_collection(precursor_directory + "/precursor", gpu.FileType.VTK)
+ grid_builder.set_precursor_boundary_condition(gpu.SideType.MX, precursor, nTReadPrecursor, 0, 0, 0)
+
+grid_builder.set_stress_boundary_condition(gpu.SideType.MZ, 0, 0, 1, sampling_offset, z0, dx)
+para.set_has_wall_model_monitor(True)
+grid_builder.set_slip_boundary_condition(gpu.SideType.PZ, 0, 0, -1)
+
+if read_precursor:
+ grid_builder.set_pressure_boundary_condition(gpu.SideType.PX, 0)
+bc_factory.set_stress_boundary_condition(gpu.StressBC.StressPressureBounceBack)
+bc_factory.set_slip_boundary_condition(gpu.SlipBC.SlipBounceBack)
+bc_factory.set_pressure_boundary_condition(gpu.PressureBC.OutflowNonReflective)
+if read_precursor:
+ bc_factory.set_precursor_boundary_condition(gpu.PrecursorBC.DistributionsPrecursor if use_distributions else gpu.PrecursorBC.VelocityPrecursor)
+para.set_outflow_pressure_correction_factor(0.0);
#%%
-cuda_memory_manager = gpu.CudaMemoryManager(para)
-grid_generator = gpu.GridProvider.make_grid_generator(grid_builder, para, cuda_memory_manager, comm)
+# don't use python init functions, they are very slow! Just kept as an example.
+# Define lambda in bindings and set it here.
+# def init_func(coord_x, coord_y, coord_z):
+# return [
+# 0.0,
+# (u_star/0.4 * np.log(np.maximum(coord_z,z0)/z0) + 2.0*np.sin(np.pi*16*coord_x/length[0])*np.sin(np.pi*8*coord_z/boundary_layer_height)/(np.square(coord_z/boundary_layer_height)+1)) * dt / dx,
+# 2.0*np.sin(np.pi*16.*coord_x/length[0])*np.sin(np.pi*8.*coord_z/boundary_layer_height)/(np.square(coord_z/boundary_layer_height)+1.) * dt / dx,
+# 8.0*u_star/0.4*(np.sin(np.pi*8.0*coord_y/boundary_layer_height)*np.sin(np.pi*8.0*coord_z/boundary_layer_height)+np.sin(np.pi*8.0*coord_x/length[0]))/(np.square(length[2]/2.0-coord_z)+1.) * dt / dx]
+# para.set_initial_condition(init_func)
+para.set_initial_condition_perturbed_log_law(u_star, z0, length[0], length[2], boundary_layer_height, velocity_ratio)
+
#%%
-turb_pos = np.array([3,3,3])*reference_diameter
-epsilon = 5
+turb_pos = np.array([3,3,3])*turbine_diameter
+epsilon = 1.5*dx
density = 1.225
level = 0
n_blades = 3
n_blade_nodes = 32
-alm = gpu.ActuatorLine(n_blades, density, n_blade_nodes, epsilon, *turb_pos, reference_diameter, level, dt, dx)
+omega = 1
+blade_radii = np.arange(n_blade_nodes, dtype=np.float32)/(0.5*turbine_diameter)
+alm = gpu.ActuatorFarm(n_blades, density, n_blade_nodes, epsilon, level, dt, dx, True)
+alm.add_turbine(turb_pos[0],turb_pos[1],turb_pos[2], turbine_diameter, omega, 0, 0, blade_radii)
para.add_actuator(alm)
#%%
-point_probe = gpu.probes.PointProbe("pointProbe", str(output_path), 100, 1, 500, 100)
-point_probe.add_probe_points_from_list(np.array([1,2,5])*reference_diameter, np.array([3,3,3])*reference_diameter, np.array([3,3,3])*reference_diameter)
-point_probe.add_statistic(gpu.probes.Statistic.Means)
-
-para.add_probe(point_probe)
-
-plane_probe = gpu.probes.PlaneProbe("planeProbe", str(output_path), 100, 1, 500, 100)
-plane_probe.set_probe_plane(5*reference_diameter, 0, 0, dx, length[1], length[2])
-para.add_probe(plane_probe)
+planar_average_probe = gpu.probes.PlanarAverageProbe("horizontalPlanes", para.get_output_path(), 0, int(t_start_tmp_averaging/dt), int(t_averaging/dt) , int(t_start_out_probe/dt), int(t_out_probe/dt), 'z')
+planar_average_probe.add_all_available_statistics()
+planar_average_probe.set_file_name_to_n_out()
+para.add_probe(planar_average_probe)
#%%
-sim = gpu.Simulation(para, cuda_memory_manager, comm, grid_generator)
+wall_model_probe = gpu.probes.WallModelProbe("wallModelProbe", para.get_output_path(), 0, int(t_start_tmp_averaging/dt), int(t_averaging/dt/4), int(t_start_out_probe/dt), int(t_out_probe/dt))
+wall_model_probe.add_all_available_statistics()
+wall_model_probe.set_file_name_to_n_out()
+wall_model_probe.set_force_output_to_stress(True)
+if para.get_is_body_force():
+ wall_model_probe.set_evaluate_pressure_gradient(True)
+para.add_probe(wall_model_probe)
+
+plane_locs = [100,]
+if read_precursor: plane_locs.extend([1000, 1500, 2000, 2500, 0])
+
+for n_probe, probe_pos in enumerate(plane_locs):
+ plane_probe = gpu.probes.PlaneProbe(f"planeProbe_{n_probe+1}", para.get_output_path(), int(t_start_averaging/dt), 10, int(t_start_out_probe/dt), int(t_out_probe/dt))
+ plane_probe.set_probe_plane(probe_pos, 0, 0, dx, length[1], length[2])
+ plane_probe.add_all_available_statistics()
+ para.add_probe(plane_probe)
+#%%
+cuda_memory_manager = gpu.CudaMemoryManager(para)
+grid_generator = gpu.GridProvider.make_grid_generator(grid_builder, para, cuda_memory_manager, communicator)
+#%%
+#%%
+sim = gpu.Simulation(para, cuda_memory_manager, communicator, grid_generator, bc_factory, tm_factory, grid_scaling_factory)
#%%
sim.run()
-MPI.Finalize()
\ No newline at end of file
+MPI.Finalize()
+
diff --git a/Python/actuator_line/config.txt b/Python/actuator_line/config.txt
deleted file mode 100644
index e4c778c4cc048f54c0a32310e6bf4a7343a263fa..0000000000000000000000000000000000000000
--- a/Python/actuator_line/config.txt
+++ /dev/null
@@ -1,2 +0,0 @@
-Path = .
-GridPath = .
diff --git a/Python/actuator_line/configActuatorLine.txt b/Python/actuator_line/configActuatorLine.txt
new file mode 100644
index 0000000000000000000000000000000000000000..c45d170f039274ab355f3fe1dc044536f1f29e6f
--- /dev/null
+++ b/Python/actuator_line/configActuatorLine.txt
@@ -0,0 +1,39 @@
+##################################################
+#informations for Writing
+##################################################
+Path = .
+##################################################
+#informations for reading
+##################################################
+GridPath = .
+##################################################
+Devices = 0
+##################################################
+tStartOut = 0
+tOut = 100000
+tEnd = 300000
+##################################################
+tStartAveraging = 0
+tStartTmpAveraging = 100000
+tAveraging = 200
+tStartOutProbe = 0
+tOutProbe = 1000
+##################################################
+Ma = 0.1
+nz = 96
+
+bodyForce = true
+SGSconstant = 0.333
+TurbulenceModel = QR
+
+QuadricLimiterP = 100000.0
+QuadricLimiterM = 100000.0
+QuadricLimiterD = 100000.0
+
+##################################################
+readPrecursor = false
+nTimestepsReadPrecursor = 1
+precursorFile = precursor/Precursor
+
+##################################################
+turbineDiameter = 126.0
diff --git a/Python/boundary_layer/boundary_layer.py b/Python/boundary_layer/boundary_layer.py
index 1c01f50946b49bc0ddab7e50065a24aab4ae869f..6f6c64bc072d3afbb8aa5febbec209c26af2deee 100644
--- a/Python/boundary_layer/boundary_layer.py
+++ b/Python/boundary_layer/boundary_layer.py
@@ -1,37 +1,48 @@
+r"""
+=======================================================================================
+ ____ ____ __ ______ __________ __ __ __ __
+ \ \ | | | | | _ \ |___ ___| | | | | / \ | |
+ \ \ | | | | | |_) | | | | | | | / \ | |
+ \ \ | | | | | _ / | | | | | | / /\ \ | |
+ \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
+ \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
+ \ \ | | ________________________________________________________________
+ \ \ | | | ______________________________________________________________|
+ \ \| | | | __ __ __ __ ______ _______
+ \ | | |_____ | | | | | | | | | _ \ / _____)
+ \ | | _____| | | | | | | | | | | \ \ \_______
+ \ | | | | |_____ | \_/ | | | | |_/ / _____ |
+ \ _____| |__| |________| \_______/ |__| |______/ (_______/
+
+ This file is part of VirtualFluids. VirtualFluids is free software: you can
+ redistribute it and/or modify it under the terms of the GNU General Public
+ License as published by the Free Software Foundation, either version 3 of
+ the License, or (at your option) any later version.
+
+ VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
+ ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
+ for more details.
+
+ You should have received a copy of the GNU General Public License along
+ with VirtualFluids (see COPYING.txt). If not, see <http://www.gnu.org/licenses/>.
+
+! \file boundary_layer.py
+! \ingroup boundary_layer
+! \author Henry Korb, Henrik Asmuth
+=======================================================================================
+"""
#%%
import numpy as np
from pathlib import Path
from mpi4py import MPI
from pyfluids import basics, gpu, logger
#%%
-reference_height = 1000 # boundary layer height in m
-
-length = np.array([6,4,1])*reference_height
-viscosity = 1.56e-5
-mach = 0.1
-nodes_per_height = 32
-
-z_0 = 0.1
-u_star = 0.4
-kappa = 0.4
-
-velocity = 0.5*u_star/kappa*np.log(length[2]/z_0+1)
-flow_through_time = length[0]/velocity
-use_AMD = True
-
-
-sim_name = "BoundaryLayer"
-config_file = Path(__file__).parent/Path("config.txt")
+sim_name = "ABL"
+config_file = Path(__file__).parent/"configBoundaryLayer.txt"
output_path = Path(__file__).parent/Path("output")
output_path.mkdir(exist_ok=True)
-t_out = 1000.
-t_end = 5000.
-t_start_averaging = 0
-t_start_tmp_averaging = 100_000
-t_averaging = 200
-t_start_out_probe = 0
-t_out_probe = 1000
#%%
logger.Logger.initialize_logger()
@@ -39,95 +50,161 @@ basics.logger.Logger.add_stdout()
basics.logger.Logger.set_debug_level(basics.logger.Level.INFO_LOW)
basics.logger.Logger.time_stamp(basics.logger.TimeStamp.ENABLE)
basics.logger.Logger.enable_printed_rank_numbers(True)
-# %%
-comm = gpu.Communicator.get_instance()
#%%
grid_factory = gpu.grid_generator.GridFactory.make()
grid_builder = gpu.grid_generator.MultipleGridBuilder.make_shared(grid_factory)
+communicator = gpu.Communicator.get_instance()
+
+config = basics.ConfigurationFile()
+config.load(str(config_file))
+
+para = gpu.Parameter(communicator.get_number_of_process(), communicator.get_pid(), config)
+bc_factory = gpu.BoundaryConditionFactory()
#%%
-dx = reference_height/nodes_per_height
-dt = dx * mach / (np.sqrt(3) * velocity)
-velocity_lb = velocity * dt / dx # LB units
-viscosity_lb = viscosity * dt / (dx * dx) # LB units
+boundary_layer_height = config.get_float_value("boundaryLayerHeight", 1000)
+z0 = config.get_float_value("z0", 0.1)
+u_star = config.get_float_value("u_star", 0.4)
-pressure_gradient = u_star**2 / reference_height
-pressure_gradient_lb = pressure_gradient * dt**2 / dx
+kappa = config.get_float_value("vonKarmanConstant", 0.4) # von Karman constant
-logger.vf_log_info(f"velocity = {velocity_lb:1.6} dx/dt")
-logger.vf_log_info(f"dt = {dt:1.6}")
-logger.vf_log_info(f"dx = {dx:1.6}")
-logger.vf_log_info(f"u* = {u_star:1.6}")
-logger.vf_log_info(f"dpdx = {pressure_gradient:1.6}")
-logger.vf_log_info(f"dpdx = {pressure_gradient_lb:1.6} dx/dt^2")
-logger.vf_log_info(f"viscosity = {viscosity_lb:1.6} dx^2/dt")
+viscosity = config.get_float_value("viscosity", 1.56e-5)
+velocity = 0.5*u_star/kappa*np.log(boundary_layer_height/z0+1) #0.5 times max mean velocity at the top in m/s
-#%%
-config = basics.ConfigurationFile()
-config.load(str(config_file))
-#%%
-para = gpu.Parameter(config, comm.get_number_of_process(), comm.get_pid())
+mach = config.get_float_value("Ma", 0.1)
+nodes_per_height = config.get_uint_value("nz", 64)
+
+
+
+write_precursor = config.get_bool_value("_p", False)
+read_precursor = config.get_bool_value("readPrecursor", False)
+
+if write_precursor:
+ nTWritePrecursor = config.get_int_value("nTimestepsWritePrecursor")
+ t_start_precursor = config.get_float_value("tStartPrecursor")
+ pos_x_precursor = config.get_float_value("posXPrecursor")
+if read_precursor:
+ nTReadPrecursor = config.get_int_value("nTimestepsReadPrecursor")
+if write_precursor or read_precursor:
+ use_distributions = config.get_bool_value("useDistributions", False)
+ precursor_directory = config.get_string_value("precursorDirectory")
+
+# all in s
+t_start_out = config.get_float_value("tStartOut")
+t_out = config.get_float_value("tOut")
+t_end = config.get_float_value("tEnd") # total time of simulation
+
+t_start_averaging = config.get_float_value("tStartAveraging")
+t_start_tmp_averaging = config.get_float_value("tStartTmpAveraging")
+t_averaging = config.get_float_value("tAveraging")
+t_start_out_probe = config.get_float_value("tStartOutProbe")
+t_out_probe = config.get_float_value("tOutProbe")
+
+#%%
+length = np.array([6,4,1])*boundary_layer_height
+dx = boundary_layer_height/nodes_per_height
+dt = dx * mach / (np.sqrt(3) * velocity)
+velocity_LB = velocity * dt / dx # LB units
+viscosity_LB = viscosity * dt / (dx * dx) # LB units
+pressure_gradient = u_star * u_star / boundary_layer_height
+pressure_gradient_LB = pressure_gradient * (dt*dt)/dx
+
+logger.vf_log_info(f"velocity [dx/dt] = {velocity_LB}")
+logger.vf_log_info(f"dt = {dt}")
+logger.vf_log_info(f"dx = {dx}")
+logger.vf_log_info(f"viscosity [10^8 dx^2/dt] = {viscosity_LB*1e8}")
+logger.vf_log_info(f"u* /(dx/dt) = {u_star*dt/dx}")
+logger.vf_log_info(f"dpdx = {pressure_gradient}")
+logger.vf_log_info(f"dpdx /(dx/dt^2) = {pressure_gradient_LB}")
+
+#%%
#%%
-para.set_devices([0])
para.set_output_prefix(sim_name)
-para.set_output_path(str(output_path))
-para.set_f_name(para.get_output_path() + "/" + para.get_output_prefix())
para.set_print_files(True)
-para.set_max_level(1)
-#%%
-para.set_velocity(velocity_lb)
-para.set_viscosity(viscosity_lb)
+
+para.set_forcing(pressure_gradient_LB, 0, 0)
+para.set_velocity_LB(velocity_LB)
+para.set_viscosity_LB(viscosity_LB)
para.set_velocity_ratio(dx/dt)
para.set_viscosity_ratio(dx*dx/dt)
-para.set_use_AMD(use_AMD)
+para.set_density_ratio(1.0)
+
+para.set_main_kernel("CumulantK17")
-para.set_main_kernel("TurbulentViscosityCumulantK17CompChim" if para.get_use_AMD() else "CummulantK17CompChim")
+para.set_timestep_start_out(int(t_start_out/dt))
+para.set_timestep_out(int(t_out/dt))
+para.set_timestep_end(int(t_end/dt))
+para.set_is_body_force(config.get_bool_value("bodyForce"))
+#%%
+tm_factory = gpu.TurbulenceModelFactory(para)
+tm_factory.read_config_file(config)
+#%%
+grid_builder.add_coarse_grid(0.0, 0.0, 0.0, *length, dx)
+grid_builder.set_periodic_boundary_condition(not read_precursor, True, False)
+grid_builder.build_grids(basics.LbmOrGks.LBM, False)
-para.set_SGS_constant(0.083)
+sampling_offset = 2
+if read_precursor:
+ precursor = gpu.create_file_collection(precursor_directory + "/precursor", gpu.FileType.VTK)
+ grid_builder.set_precursor_boundary_condition(gpu.SideType.MX, precursor, nTReadPrecursor, 0, 0, 0)
+grid_builder.set_stress_boundary_condition(gpu.SideType.MZ, 0, 0, 1, sampling_offset, z0/dx)
+para.set_has_wall_model_monitor(True)
+grid_builder.set_slip_boundary_condition(gpu.SideType.PZ, 0, 0, -1)
+
+if read_precursor:
+ grid_builder.set_pressure_boundary_condition(gpu.SideType.PX, 0)
+bc_factory.set_stress_boundary_condition(gpu.StressBC.StressPressureBounceBack)
+bc_factory.set_slip_boundary_condition(gpu.SlipBC.SlipBounceBack)
+bc_factory.set_pressure_boundary_condition(gpu.PressureBC.OutflowNonReflective)
+bc_factory.set_precursor_boundary_condition(gpu.PrecursorBC.DistributionsPrecursor if use_distributions else gpu.PrecursorBC.VelocityPrecursor)
+para.set_outflow_pressure_correction_factor(0.0);
+#%%
def init_func(coord_x, coord_y, coord_z):
return [
0.0,
- (u_star/kappa*np.log(max(coord_z/z_0,0)+1) + 2*np.sin(np.pi*16*coord_x/length[0])*np.sin(np.pi*8*coord_z/length[2]))/((coord_z/reference_height)**2+0.1)*dt/dx,
- 2*np.sin(np.pi*16*coord_x/length[0])*np.sin(np.pi*8*coord_z/length[2])/((coord_z/reference_height)**2+0.1)*dt/dx,
- 8*u_star/kappa*(np.sin(np.pi*8*coord_y/reference_height)*np.sin(np.pi*8*coord_z/reference_height)+np.sin(np.pi*8*coord_x/length[0]))/((length[2]/2-coord_z)**2+0.1)*dt/dx
- ]
-
+ (u_star/0.4 * np.log(np.maximum(coord_z,z0)/z0) + 2.0*np.sin(np.pi*16*coord_x/length[0])*np.sin(np.pi*8*coord_z/boundary_layer_height)/(np.square(coord_z/boundary_layer_height)+1)) * dt / dx,
+ 2.0*np.sin(np.pi*16.*coord_x/length[0])*np.sin(np.pi*8.*coord_z/boundary_layer_height)/(np.square(coord_z/boundary_layer_height)+1.) * dt / dx,
+ 8.0*u_star/0.4*(np.sin(np.pi*8.0*coord_y/boundary_layer_height)*np.sin(np.pi*8.0*coord_z/boundary_layer_height)+np.sin(np.pi*8.0*coord_x/length[0]))/(np.square(length[2]/2.0-coord_z)+1.) * dt / dx]
para.set_initial_condition(init_func)
-para.set_t_out(int(t_out/dt))
-para.set_t_end(int(t_end/dt))
-para.set_is_body_force(True)
-para.set_has_wall_model_monitor(True)
-
-grid_builder.add_coarse_grid(0.0, 0.0, 0.0, *length, dx)
-grid_builder.set_periodic_boundary_condition(True, True, False)
-grid_builder.build_grids(basics.LbmOrGks.LBM, False)
#%%
-sampling_offset = 2
-grid_builder.set_stress_boundary_condition(gpu.SideType.MZ, 0.0, 0.0, 1.0, sampling_offset, z_0/dx)
-grid_builder.set_slip_boundary_condition(gpu.SideType.PZ, 0.0, 0.0, 0.0)
+planar_average_probe = gpu.probes.PlanarAverageProbe("horizontalPlanes", para.get_output_path(), 0, int(t_start_tmp_averaging/dt), int(t_averaging/dt) , int(t_start_out_probe/dt), int(t_out_probe/dt), 'z')
+planar_average_probe.add_all_available_statistics()
+planar_average_probe.set_file_name_to_n_out()
+para.add_probe(planar_average_probe)
+#%%
+wall_model_probe = gpu.probes.WallModelProbe("wallModelProbe", para.get_output_path(), 0, int(t_start_tmp_averaging/dt), int(t_averaging/dt/4), int(t_start_out_probe/dt), int(t_out_probe/dt))
+wall_model_probe.add_all_available_statistics()
+wall_model_probe.set_file_name_to_n_out()
+wall_model_probe.set_force_output_to_stress(True)
+if para.get_is_body_force():
+ wall_model_probe.set_evaluate_pressure_gradient(True)
+para.add_probe(wall_model_probe)
+
+plane_locs = [100,]
+if read_precursor: plane_locs.extend([1000, 1500, 2000, 2500, 0])
+
+for n_probe, probe_pos in enumerate(plane_locs):
+ plane_probe = gpu.probes.PlaneProbe(f"planeProbe_{n_probe+1}", para.get_output_path(), int(t_start_averaging/dt), 10, int(t_start_out_probe/dt), int(t_out_probe/dt))
+ plane_probe.set_probe_plane(probe_pos, 0, 0, dx, length[1], length[2])
+ plane_probe.add_all_available_statistics()
+ para.add_probe(plane_probe)
+
+if write_precursor:
+ precursor_writer = gpu.PrecursorWriter("precursor", para.get_output_path() + precursor_directory, pos_x_precursor, 0,length[1], 0, length[2], t_start_precursor/dt, nTWritePrecursor, gpu.OutputVariable.Distributions if use_distributions else gpu.OutputVariable.Velocities)
+ para.add_probe(precursor_writer)
#%%
cuda_memory_manager = gpu.CudaMemoryManager(para)
-grid_generator = gpu.GridProvider.make_grid_generator(grid_builder, para, cuda_memory_manager, comm)
-
+grid_generator = gpu.GridProvider.make_grid_generator(grid_builder, para, cuda_memory_manager, communicator)
#%%
-wall_probe = gpu.probes.WallModelProbe("wallModelProbe", str(output_path), int(t_start_averaging/dt), int(t_start_tmp_averaging/dt), int(t_averaging/dt/4), int(t_start_out_probe/dt), int(t_out_probe/dt))
-wall_probe.add_all_available_statistics()
-wall_probe.set_file_name_to_n_out()
-wall_probe.set_force_output_to_stress(True)
-if para.get_is_body_force():
- wall_probe.set_evaluate_pressure_gradient(True)
-planar_probe = gpu.probes.PlanarAverageProbe("planarAverageProbe", str(output_path), int(t_start_averaging/dt), int(t_start_tmp_averaging/dt), int(t_averaging/dt), int(t_start_out_probe/dt), int(t_out_probe/dt), "z")
-para.add_probe(wall_probe)
-
#%%
-sim = gpu.Simulation(para, cuda_memory_manager, comm, grid_generator)
+sim = gpu.Simulation(para, cuda_memory_manager, communicator, grid_generator, bc_factory, tm_factory)
#%%
sim.run()
MPI.Finalize()
\ No newline at end of file
diff --git a/Python/boundary_layer/config.txt b/Python/boundary_layer/config.txt
deleted file mode 100644
index e4c778c4cc048f54c0a32310e6bf4a7343a263fa..0000000000000000000000000000000000000000
--- a/Python/boundary_layer/config.txt
+++ /dev/null
@@ -1,2 +0,0 @@
-Path = .
-GridPath = .
diff --git a/Python/boundary_layer/configBoundaryLayer.txt b/Python/boundary_layer/configBoundaryLayer.txt
new file mode 100644
index 0000000000000000000000000000000000000000..83e7861a5fb85ea800d187699f1c6c1409422f0a
--- /dev/null
+++ b/Python/boundary_layer/configBoundaryLayer.txt
@@ -0,0 +1,42 @@
+##################################################
+#informations for Writing
+##################################################
+Path = .
+##################################################
+#informations for reading
+##################################################
+GridPath = .
+##################################################
+Devices = 0
+##################################################
+tStartOut = 0
+tOut = 100000
+tEnd = 300000
+##################################################
+tStartAveraging = 0
+tStartTmpAveraging = 100000
+tAveraging = 200
+tStartOutProbe = 0
+tOutProbe = 1000
+##################################################
+Ma = 0.1
+nz = 96
+
+bodyForce = true
+UseAMD = true
+SGSconstant = 0.2
+QuadricLimiterP = 100000.0
+QuadricLimiterM = 100000.0
+QuadricLimiterD = 100000.0
+
+##################################################
+readPrecursor = false
+nTimestepsReadPrecursor = 10
+precursorFile = precursor/Precursor
+
+##################################################
+writePrecursor = false
+nTimestepsWritePrecursor = 10
+
+tStartPrecursor = 100
+posXPrecursor = 3000
\ No newline at end of file
diff --git a/Python/cubeflow/simulation.py b/Python/cubeflow/simulation.py
index 9e77e8d747c072188d8d81150afa8e2ccb76a792..deb0411963aec65522af45cc48d7367f103232c6 100644
--- a/Python/cubeflow/simulation.py
+++ b/Python/cubeflow/simulation.py
@@ -1,13 +1,42 @@
-from pyfluids.cpu import Simulation
-from pyfluids.cpu.boundaryconditions import NoSlipBoundaryCondition, VelocityBoundaryCondition, DensityBoundaryCondition
-from pyfluids.cpu.geometry import GbCuboid3D
-from pyfluids.cpu.kernel import LBMKernel, KernelType
-from pyfluids.cpu.parameters import PhysicalParameters, RuntimeParameters, GridParameters
-from pyfluids.cpu.writer import Writer, OutputFormat
-from pymuparser import Parser
-
+r"""
+=======================================================================================
+ ____ ____ __ ______ __________ __ __ __ __
+ \ \ | | | | | _ \ |___ ___| | | | | / \ | |
+ \ \ | | | | | |_) | | | | | | | / \ | |
+ \ \ | | | | | _ / | | | | | | / /\ \ | |
+ \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
+ \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
+ \ \ | | ________________________________________________________________
+ \ \ | | | ______________________________________________________________|
+ \ \| | | | __ __ __ __ ______ _______
+ \ | | |_____ | | | | | | | | | _ \ / _____)
+ \ | | _____| | | | | | | | | | | \ \ \_______
+ \ | | | | |_____ | \_/ | | | | |_/ / _____ |
+ \ _____| |__| |________| \_______/ |__| |______/ (_______/
+
+ This file is part of VirtualFluids. VirtualFluids is free software: you can
+ redistribute it and/or modify it under the terms of the GNU General Public
+ License as published by the Free Software Foundation, either version 3 of
+ the License, or (at your option) any later version.
+
+ VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
+ ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
+ for more details.
+
+ You should have received a copy of the GNU General Public License along
+ with VirtualFluids (see COPYING.txt). If not, see <http://www.gnu.org/licenses/>.
+
+! \file simulation.py
+! \ingroup cubeflow
+! \author Sven Marcus, Henry Korb
+=======================================================================================
+"""
import os
+from pyfluids import cpu
+from pymuparser import Parser
+
def get_max_length(number_of_nodes_per_direction, delta_x):
return (number_of_nodes_per_direction[0] * delta_x,
@@ -15,10 +44,10 @@ def get_max_length(number_of_nodes_per_direction, delta_x):
number_of_nodes_per_direction[2] * delta_x)
-physical_params = PhysicalParameters()
+physical_params = cpu.parameters.PhysicalParameters()
physical_params.lattice_viscosity = 0.005
-grid_params = GridParameters()
+grid_params = cpu.parameters.GridParameters()
grid_params.number_of_nodes_per_direction = [200, 120, 120]
grid_params.blocks_per_direction = [2, 2, 2]
grid_params.node_distance = 0.125
@@ -26,7 +55,7 @@ grid_params.periodic_boundary_in_x1 = False
grid_params.periodic_boundary_in_x2 = True
grid_params.periodic_boundary_in_x3 = True
-runtime_params = RuntimeParameters()
+runtime_params = cpu.parameters.RuntimeParameters()
runtime_params.timestep_log_interval = 1000
runtime_params.number_of_timesteps = 50000
runtime_params.number_of_threads = int(os.environ.get("OMP_NUM_THREADS", 4))
@@ -39,46 +68,46 @@ def run_simulation(physical_parameters=physical_params, grid_parameters=grid_par
min_x, min_y, min_z = 0, 0, 0
max_x, max_y, max_z = get_max_length(grid_parameters.number_of_nodes_per_direction, grid_parameters.node_distance)
- bottom_wall = GbCuboid3D(min_x - wall_thickness, min_y - wall_thickness, min_z, max_x + wall_thickness,
+ bottom_wall = cpu.geometry.GbCuboid3D(min_x - wall_thickness, min_y - wall_thickness, min_z, max_x + wall_thickness,
max_y + wall_thickness, min_z - wall_thickness)
- top_wall = GbCuboid3D(min_x - wall_thickness, min_y - wall_thickness, max_z, max_x + wall_thickness,
+ top_wall = cpu.geometry.GbCuboid3D(min_x - wall_thickness, min_y - wall_thickness, max_z, max_x + wall_thickness,
max_y + wall_thickness,
max_z + wall_thickness)
- left_wall = GbCuboid3D(min_x - wall_thickness, min_y, min_z - wall_thickness, max_x + wall_thickness,
+ left_wall = cpu.geometry.GbCuboid3D(min_x - wall_thickness, min_y, min_z - wall_thickness, max_x + wall_thickness,
min_y - wall_thickness,
max_z + wall_thickness)
- right_wall = GbCuboid3D(min_x - wall_thickness, max_y, min_z - wall_thickness, max_x + wall_thickness,
+ right_wall = cpu.geometry.GbCuboid3D(min_x - wall_thickness, max_y, min_z - wall_thickness, max_x + wall_thickness,
max_y + wall_thickness, max_z + wall_thickness)
- obstacle = GbCuboid3D(7, 7, 7, 8, 8, 8)
+ obstacle = cpu.geometry.GbCuboid3D(7, 7, 7, 8, 8, 8)
- velocity_boundary = GbCuboid3D(min_x - wall_thickness, min_y - wall_thickness, min_z - wall_thickness, min_x,
+ velocity_boundary = cpu.geometry.GbCuboid3D(min_x - wall_thickness, min_y - wall_thickness, min_z - wall_thickness, min_x,
max_y + wall_thickness, max_z + wall_thickness)
- outflow_boundary = GbCuboid3D(max_x, min_y - wall_thickness, min_z - wall_thickness, max_x + wall_thickness,
+ outflow_boundary = cpu.geometry.GbCuboid3D(max_x, min_y - wall_thickness, min_z - wall_thickness, max_x + wall_thickness,
max_y + wall_thickness, max_z + wall_thickness)
- no_slip_bc = NoSlipBoundaryCondition()
+ no_slip_bc = cpu.boundaryconditions.NoSlipBoundaryCondition()
- outflow_bc = DensityBoundaryCondition()
+ outflow_bc = cpu.boundaryconditions.DensityBoundaryCondition()
velocity_function = Parser()
velocity_function.define_constant("u", 0.07)
velocity_function.expression = "u"
- velocity_bc = VelocityBoundaryCondition(True, False, False, velocity_function, 0, -10)
+ velocity_bc = cpu.boundaryconditions.VelocityBoundaryCondition(True, False, False, velocity_function, 0, -10)
- kernel = LBMKernel(KernelType.CompressibleCumulantFourthOrderViscosity)
+ kernel = cpu.kernel.LBMKernel(cpu.kernel.KernelType.CompressibleCumulantFourthOrderViscosity)
# kernel.use_forcing = True
# kernel.forcing_in_x1 = 3e-6
- writer = Writer()
+ writer = cpu.writer.Writer()
writer.output_path = "./output"
- writer.output_format = OutputFormat.BINARY
+ writer.output_format = cpu.writer.OutputFormat.BINARY
- simulation = Simulation()
+ simulation = cpu.Simulation()
simulation.set_writer(writer)
simulation.set_physical_parameters(physical_parameters)
diff --git a/Python/liddrivencavity/simulation.py b/Python/liddrivencavity/simulation.py
index 155fad2f6f8aade0368c8a7006b88f7985f8822c..3c247b87a102e3c5a720f20748acc9f9f50bb178 100644
--- a/Python/liddrivencavity/simulation.py
+++ b/Python/liddrivencavity/simulation.py
@@ -1,32 +1,61 @@
-from pyfluids.cpu import Simulation
-from pyfluids.cpu.boundaryconditions import NoSlipBoundaryCondition, VelocityBoundaryCondition
-from pyfluids.cpu.geometry import GbCuboid3D
-from pyfluids.cpu.kernel import LBMKernel, KernelType
-from pyfluids.cpu.parameters import GridParameters, PhysicalParameters, RuntimeParameters
-from pyfluids.cpu.writer import Writer, OutputFormat
+r"""
+=======================================================================================
+ ____ ____ __ ______ __________ __ __ __ __
+ \ \ | | | | | _ \ |___ ___| | | | | / \ | |
+ \ \ | | | | | |_) | | | | | | | / \ | |
+ \ \ | | | | | _ / | | | | | | / /\ \ | |
+ \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
+ \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
+ \ \ | | ________________________________________________________________
+ \ \ | | | ______________________________________________________________|
+ \ \| | | | __ __ __ __ ______ _______
+ \ | | |_____ | | | | | | | | | _ \ / _____)
+ \ | | _____| | | | | | | | | | | \ \ \_______
+ \ | | | | |_____ | \_/ | | | | |_/ / _____ |
+ \ _____| |__| |________| \_______/ |__| |______/ (_______/
+
+ This file is part of VirtualFluids. VirtualFluids is free software: you can
+ redistribute it and/or modify it under the terms of the GNU General Public
+ License as published by the Free Software Foundation, either version 3 of
+ the License, or (at your option) any later version.
+
+ VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
+ ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
+ for more details.
+
+ You should have received a copy of the GNU General Public License along
+ with VirtualFluids (see COPYING.txt). If not, see <http://www.gnu.org/licenses/>.
+
+! \file simulation.py
+! \ingroup liddrivencavity
+! \author Sven Marcus, Henry Korb
+=======================================================================================
+"""
+from pyfluids import cpu
from pymuparser import Parser
-runtime_params = RuntimeParameters()
+runtime_params = cpu.parameters.RuntimeParameters()
runtime_params.number_of_threads = 4
runtime_params.number_of_timesteps = 10000
runtime_params.timestep_log_interval = 1000
-physical_params = PhysicalParameters()
+physical_params = cpu.parameters.PhysicalParameters()
physical_params.lattice_viscosity = 0.005
-grid_params = GridParameters()
+grid_params = cpu.parameters.GridParameters()
grid_params.number_of_nodes_per_direction = [64, 64, 64]
grid_params.blocks_per_direction = [2, 2, 2]
grid_params.node_distance = 1 / 10
def run_simulation(physical_params=physical_params, grid_params=grid_params, runtime_params=runtime_params):
- simulation = Simulation()
- kernel = LBMKernel(KernelType.CompressibleCumulantFourthOrderViscosity)
+ simulation = cpu.Simulation()
+ kernel = cpu.kernel.LBMKernel(cpu.kernel.KernelType.CompressibleCumulantFourthOrderViscosity)
- writer = Writer()
+ writer = cpu.writer.Writer()
writer.output_path = "./output"
- writer.output_format = OutputFormat.BINARY
+ writer.output_format = cpu.writer.OutputFormat.BINARY
simulation.set_grid_parameters(grid_params)
simulation.set_physical_parameters(physical_params)
@@ -34,12 +63,12 @@ def run_simulation(physical_params=physical_params, grid_params=grid_params, run
simulation.set_kernel_config(kernel)
simulation.set_writer(writer)
- no_slip_bc_adapter = NoSlipBoundaryCondition()
+ no_slip_bc_adapter = cpu.boundaryconditions.NoSlipBoundaryCondition()
fct = Parser()
fct.expression = "u"
fct.define_constant("u", 0.005)
- velocity_bc_adapter = VelocityBoundaryCondition(True, True, False, fct, 0, -10.0)
+ velocity_bc_adapter = cpu.boundaryconditions.VelocityBoundaryCondition(True, True, False, fct, 0, -10.0)
g_min_x1, g_min_x2, g_min_x3 = 0, 0, 0
g_max_x1 = grid_params.number_of_nodes_per_direction[0] * grid_params.node_distance
@@ -48,12 +77,12 @@ def run_simulation(physical_params=physical_params, grid_params=grid_params, run
dx = grid_params.node_distance
- wall_x_min = GbCuboid3D(g_min_x1 - dx, g_min_x2 - dx, g_min_x3 - dx, g_min_x1, g_max_x2 + dx, g_max_x3)
- wall_x_max = GbCuboid3D(g_max_x1, g_min_x2 - dx, g_min_x3 - dx, g_max_x1 + dx, g_max_x2 + dx, g_max_x3)
- wall_y_min = GbCuboid3D(g_min_x1 - dx, g_min_x2 - dx, g_min_x3 - dx, g_max_x1 + dx, g_min_x2, g_max_x3)
- wall_y_max = GbCuboid3D(g_min_x1 - dx, g_max_x2, g_min_x3 - dx, g_max_x1 + dx, g_max_x2 + dx, g_max_x3)
- wall_z_min = GbCuboid3D(g_min_x1 - dx, g_min_x2 - dx, g_min_x3 - dx, g_max_x1 + dx, g_max_x2 + dx, g_min_x3)
- wall_z_max = GbCuboid3D(g_min_x1 - dx, g_min_x2 - dx, g_max_x3, g_max_x1 + dx, g_max_x2 + dx, g_max_x3 + dx)
+ wall_x_min = cpu.geometry.GbCuboid3D(g_min_x1 - dx, g_min_x2 - dx, g_min_x3 - dx, g_min_x1, g_max_x2 + dx, g_max_x3)
+ wall_x_max = cpu.geometry.GbCuboid3D(g_max_x1, g_min_x2 - dx, g_min_x3 - dx, g_max_x1 + dx, g_max_x2 + dx, g_max_x3)
+ wall_y_min = cpu.geometry.GbCuboid3D(g_min_x1 - dx, g_min_x2 - dx, g_min_x3 - dx, g_max_x1 + dx, g_min_x2, g_max_x3)
+ wall_y_max = cpu.geometry.GbCuboid3D(g_min_x1 - dx, g_max_x2, g_min_x3 - dx, g_max_x1 + dx, g_max_x2 + dx, g_max_x3)
+ wall_z_min = cpu.geometry.GbCuboid3D(g_min_x1 - dx, g_min_x2 - dx, g_min_x3 - dx, g_max_x1 + dx, g_max_x2 + dx, g_min_x3)
+ wall_z_max = cpu.geometry.GbCuboid3D(g_min_x1 - dx, g_min_x2 - dx, g_max_x3, g_max_x1 + dx, g_max_x2 + dx, g_max_x3 + dx)
simulation.add_object(wall_x_min, no_slip_bc_adapter, 1, "/geo/wallXmin")
simulation.add_object(wall_x_max, no_slip_bc_adapter, 1, "/geo/wallXmax")
diff --git a/Python/poiseuille/poiseuille_hpc.py b/Python/poiseuille/poiseuille_hpc.py
index f5f5a1387c9fe234abae0c6f979cc7d5b283d1a4..b108f34445a71a686c4e22f685e26e10204113b3 100644
--- a/Python/poiseuille/poiseuille_hpc.py
+++ b/Python/poiseuille/poiseuille_hpc.py
@@ -1,15 +1,49 @@
+r"""
+=======================================================================================
+ ____ ____ __ ______ __________ __ __ __ __
+ \ \ | | | | | _ \ |___ ___| | | | | / \ | |
+ \ \ | | | | | |_) | | | | | | | / \ | |
+ \ \ | | | | | _ / | | | | | | / /\ \ | |
+ \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
+ \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
+ \ \ | | ________________________________________________________________
+ \ \ | | | ______________________________________________________________|
+ \ \| | | | __ __ __ __ ______ _______
+ \ | | |_____ | | | | | | | | | _ \ / _____)
+ \ | | _____| | | | | | | | | | | \ \ \_______
+ \ | | | | |_____ | \_/ | | | | |_/ / _____ |
+ \ _____| |__| |________| \_______/ |__| |______/ (_______/
+
+ This file is part of VirtualFluids. VirtualFluids is free software: you can
+ redistribute it and/or modify it under the terms of the GNU General Public
+ License as published by the Free Software Foundation, either version 3 of
+ the License, or (at your option) any later version.
+
+ VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
+ ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
+ for more details.
+
+ You should have received a copy of the GNU General Public License along
+ with VirtualFluids (see COPYING.txt). If not, see <http://www.gnu.org/licenses/>.
+
+! \file poiseuille_hpc.py
+! \ingroup poiseuille
+! \author Sven Marcus, Henry Korb
+=======================================================================================
+"""
from poiseuille.simulation import run_simulation
-from pyfluids.cpu.parameters import *
+from pyfluids import cpu
-grid_parameters = GridParameters()
+grid_parameters = cpu.prameters.GridParameters()
grid_parameters.number_of_nodes_per_direction = [64, 64, 512]
grid_parameters.node_distance = 1
grid_parameters.blocks_per_direction = [1, 2, 2]
-physical_parameters = PhysicalParameters()
+physical_parameters = cpu.prameters.PhysicalParameters()
physical_parameters.lattice_viscosity = 0.0005
-runtime_parameters = RuntimeParameters()
+runtime_parameters = cpu.prameters.RuntimeParameters()
runtime_parameters.number_of_threads = 4
runtime_parameters.number_of_timesteps = 1000
runtime_parameters.timestep_log_interval = 100
diff --git a/Python/poiseuille/simulation.py b/Python/poiseuille/simulation.py
index d107801fa84cfe16d1d7e91d31dc3ff4b8671f02..a6f12e59fbd0a0ccad9a4db9ccde69b828cf90bf 100644
--- a/Python/poiseuille/simulation.py
+++ b/Python/poiseuille/simulation.py
@@ -1,35 +1,65 @@
-from pyfluids.cpu import Simulation
-from pyfluids.cpu.boundaryconditions import NoSlipBoundaryCondition
-from pyfluids.cpu.geometry import GbCuboid3D, State
-from pyfluids.cpu.kernel import LBMKernel, KernelType
-from pyfluids.cpu.parameters import RuntimeParameters, GridParameters, PhysicalParameters
-from pyfluids.cpu.writer import Writer, OutputFormat
-
-default_grid_params = GridParameters()
+r"""
+=======================================================================================
+ ____ ____ __ ______ __________ __ __ __ __
+ \ \ | | | | | _ \ |___ ___| | | | | / \ | |
+ \ \ | | | | | |_) | | | | | | | / \ | |
+ \ \ | | | | | _ / | | | | | | / /\ \ | |
+ \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
+ \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
+ \ \ | | ________________________________________________________________
+ \ \ | | | ______________________________________________________________|
+ \ \| | | | __ __ __ __ ______ _______
+ \ | | |_____ | | | | | | | | | _ \ / _____)
+ \ | | _____| | | | | | | | | | | \ \ \_______
+ \ | | | | |_____ | \_/ | | | | |_/ / _____ |
+ \ _____| |__| |________| \_______/ |__| |______/ (_______/
+
+ This file is part of VirtualFluids. VirtualFluids is free software: you can
+ redistribute it and/or modify it under the terms of the GNU General Public
+ License as published by the Free Software Foundation, either version 3 of
+ the License, or (at your option) any later version.
+
+ VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
+ ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
+ for more details.
+
+ You should have received a copy of the GNU General Public License along
+ with VirtualFluids (see COPYING.txt). If not, see <http://www.gnu.org/licenses/>.
+
+! \file simulation.py
+! \ingroup poiseuille
+! \author Sven Marcus, Henry Korb
+=======================================================================================
+"""
+from pyfluids import cpu
+
+
+default_grid_params = cpu.parameters.GridParameters()
default_grid_params.node_distance = 10 / 32
default_grid_params.number_of_nodes_per_direction = [8, 8, 32]
default_grid_params.blocks_per_direction = [1, 1, 4]
default_grid_params.periodic_boundary_in_x1 = True
default_grid_params.periodic_boundary_in_x2 = True
-default_physical_params = PhysicalParameters()
+default_physical_params = cpu.parameters.PhysicalParameters()
default_physical_params.lattice_viscosity = 0.005
-default_runtime_params = RuntimeParameters()
+default_runtime_params = cpu.parameters.RuntimeParameters()
default_runtime_params.number_of_threads = 4
default_runtime_params.number_of_timesteps = 10000
default_runtime_params.timestep_log_interval = 1000
-default_kernel = LBMKernel(KernelType.CompressibleCumulantFourthOrderViscosity)
+default_kernel = cpu.kernel.LBMKernel(cpu.kernel.KernelType.CompressibleCumulantFourthOrderViscosity)
default_kernel.use_forcing = True
default_kernel.forcing_in_x1 = 1e-8
-default_writer = Writer()
+default_writer = cpu.writer.Writer()
default_writer.output_path = "./output"
-default_writer.output_format = OutputFormat.BINARY
+default_writer.output_format = cpu.writer.OutputFormat.BINARY
-default_kernel = LBMKernel(KernelType.CompressibleCumulantFourthOrderViscosity)
+default_kernel = cpu.kernel.LBMKernel(cpu.kernel.KernelType.CompressibleCumulantFourthOrderViscosity)
default_kernel.use_forcing = True
default_kernel.forcing_in_x1 = 1e-8
@@ -39,7 +69,7 @@ def run_simulation(physical_params=default_physical_params,
runtime_params=default_runtime_params,
kernel=default_kernel,
writer=default_writer):
- simulation = Simulation()
+ simulation = cpu.Simulation()
simulation.set_kernel_config(kernel)
simulation.set_physical_parameters(physical_params)
@@ -47,11 +77,11 @@ def run_simulation(physical_params=default_physical_params,
simulation.set_runtime_parameters(runtime_params)
simulation.set_writer(writer)
- no_slip_bc = NoSlipBoundaryCondition()
+ no_slip_bc = cpu.boundaryconditions.NoSlipBoundaryCondition()
block_thickness = 3 * grid_params.node_distance
simulation.add_object(
- GbCuboid3D(
+ cpu.geometry.GbCuboid3D(
grid_params.bounding_box.min_x1 - block_thickness,
grid_params.bounding_box.min_x2 - block_thickness,
grid_params.bounding_box.min_x3 - block_thickness,
@@ -59,10 +89,10 @@ def run_simulation(physical_params=default_physical_params,
grid_params.bounding_box.max_x2 + block_thickness,
grid_params.bounding_box.min_x3),
no_slip_bc,
- State.SOLID, "/geo/addWallZMin")
+ cpu.geometry.State.SOLID, "/geo/addWallZMin")
simulation.add_object(
- GbCuboid3D(
+ cpu.geometry.GbCuboid3D(
grid_params.bounding_box.min_x1 - block_thickness,
grid_params.bounding_box.min_x2 - block_thickness,
grid_params.bounding_box.max_x3,
@@ -70,7 +100,7 @@ def run_simulation(physical_params=default_physical_params,
grid_params.bounding_box.max_x2 + block_thickness,
grid_params.bounding_box.max_x3 + block_thickness),
no_slip_bc,
- State.SOLID, "/geo/addWallZMax")
+ cpu.geometry.State.SOLID, "/geo/addWallZMax")
simulation.run_simulation()
diff --git a/Python/poiseuille/test_poiseuille_l2.py b/Python/poiseuille/test_poiseuille_l2.py
index 93aa2600d5260dea7e72f3aa98db7334fe5285c6..818cba40e115945c60e4fa2ac96b3b6b5ab0bba8 100644
--- a/Python/poiseuille/test_poiseuille_l2.py
+++ b/Python/poiseuille/test_poiseuille_l2.py
@@ -1,3 +1,37 @@
+r"""
+=======================================================================================
+ ____ ____ __ ______ __________ __ __ __ __
+ \ \ | | | | | _ \ |___ ___| | | | | / \ | |
+ \ \ | | | | | |_) | | | | | | | / \ | |
+ \ \ | | | | | _ / | | | | | | / /\ \ | |
+ \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
+ \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
+ \ \ | | ________________________________________________________________
+ \ \ | | | ______________________________________________________________|
+ \ \| | | | __ __ __ __ ______ _______
+ \ | | |_____ | | | | | | | | | _ \ / _____)
+ \ | | _____| | | | | | | | | | | \ \ \_______
+ \ | | | | |_____ | \_/ | | | | |_/ / _____ |
+ \ _____| |__| |________| \_______/ |__| |______/ (_______/
+
+ This file is part of VirtualFluids. VirtualFluids is free software: you can
+ redistribute it and/or modify it under the terms of the GNU General Public
+ License as published by the Free Software Foundation, either version 3 of
+ the License, or (at your option) any later version.
+
+ VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
+ ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
+ for more details.
+
+ You should have received a copy of the GNU General Public License along
+ with VirtualFluids (see COPYING.txt). If not, see <http://www.gnu.org/licenses/>.
+
+! \file test_poiseuille_l2.py
+! \ingroup poiseuille
+! \author Sven Marcus, Henry Korb
+=======================================================================================
+"""
import os
import shutil
import unittest
@@ -5,8 +39,7 @@ import unittest
import matplotlib.pyplot as plt
import numpy as np
import pyvista as pv
-from pyfluids.cpu.kernel import LBMKernel, KernelType
-from pyfluids.cpu.parameters import GridParameters, PhysicalParameters, RuntimeParameters
+from pyfluids import cpu
from scipy import stats
from errors import normalized_l2_error
@@ -33,13 +66,13 @@ class TestPoiseuilleFlow(unittest.TestCase):
self.skipTest("This test is not implemented correctly yet")
plt.ion()
- physical_params = PhysicalParameters()
+ physical_params = cpu.parameters.PhysicalParameters()
- runtime_params = RuntimeParameters()
+ runtime_params = cpu.parameters.RuntimeParameters()
runtime_params.number_of_threads = os.cpu_count()
runtime_params.timestep_log_interval = 10000
- kernel = LBMKernel(KernelType.CompressibleCumulantFourthOrderViscosity)
+ kernel = cpu.kernel.LBMKernel(cpu.kernel.KernelType.CompressibleCumulantFourthOrderViscosity)
kernel.use_forcing = True
normalized_l2_errors = []
@@ -140,7 +173,7 @@ def get_heights_from_indices(mesh, indices):
def create_grid_params_with_nodes_in_column(nodes_in_column, delta_x):
- grid_params = GridParameters()
+ grid_params = cpu.parameters.GridParameters()
grid_params.node_distance = delta_x
grid_params.number_of_nodes_per_direction = [1, 1, nodes_in_column]
grid_params.blocks_per_direction = [1, 1, 8]
diff --git a/Python/tests/test_acousticscaling.py b/Python/tests/test_acousticscaling.py
index 6413123a80db8c5882fcf1dbe6f72a1f5438736c..02454b935e3a147e045f45c273392646aeca6b8c 100644
--- a/Python/tests/test_acousticscaling.py
+++ b/Python/tests/test_acousticscaling.py
@@ -1,9 +1,41 @@
+r"""
+=======================================================================================
+ ____ ____ __ ______ __________ __ __ __ __
+ \ \ | | | | | _ \ |___ ___| | | | | / \ | |
+ \ \ | | | | | |_) | | | | | | | / \ | |
+ \ \ | | | | | _ / | | | | | | / /\ \ | |
+ \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
+ \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
+ \ \ | | ________________________________________________________________
+ \ \ | | | ______________________________________________________________|
+ \ \| | | | __ __ __ __ ______ _______
+ \ | | |_____ | | | | | | | | | _ \ / _____)
+ \ | | _____| | | | | | | | | | | \ \ \_______
+ \ | | | | |_____ | \_/ | | | | |_/ / _____ |
+ \ _____| |__| |________| \_______/ |__| |______/ (_______/
+
+ This file is part of VirtualFluids. VirtualFluids is free software: you can
+ redistribute it and/or modify it under the terms of the GNU General Public
+ License as published by the Free Software Foundation, either version 3 of
+ the License, or (at your option) any later version.
+
+ VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
+ ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
+ for more details.
+
+ You should have received a copy of the GNU General Public License along
+ with VirtualFluids (see COPYING.txt). If not, see <http://www.gnu.org/licenses/>.
+
+! \file test_acousticscaling.py
+! \ingroup tests
+! \author Sven Marcus, Henry Korb
+=======================================================================================
+"""
import unittest
from typing import List
-from pyfluids.cpu.kernel import LBMKernel, KernelType
-from pyfluids.cpu.parameters import GridParameters, PhysicalParameters, RuntimeParameters
-
+from pyfluids import cpu
from acousticscaling import OneDirectionalAcousticScaling
@@ -58,18 +90,18 @@ class OneDirectionalAcousticScalingTest(unittest.TestCase):
self.assertEqual(self.grid_params.periodic_boundary_in_x2, actual_grid_params.periodic_boundary_in_x2)
self.assertEqual(self.grid_params.periodic_boundary_in_x3, actual_grid_params.periodic_boundary_in_x3)
- def assert_physical_params_scaled_by_factor(self, actual_params: PhysicalParameters, factor: int):
+ def assert_physical_params_scaled_by_factor(self, actual_params: cpu.parameters.PhysicalParameters, factor: int):
self.assertEqual(self.physical_params.lattice_viscosity * factor, actual_params.lattice_viscosity)
self.assertEqual(self.physical_params.bulk_viscosity_factor, actual_params.bulk_viscosity_factor)
- def assert_runtime_params_scaled_by_factor(self, actual_params: RuntimeParameters, factor: int):
+ def assert_runtime_params_scaled_by_factor(self, actual_params: cpu.parameters.RuntimeParameters, factor: int):
self.assertEqual(self.runtime_params.number_of_timesteps * factor, actual_params.number_of_timesteps)
self.assertEqual(self.runtime_params.number_of_threads, actual_params.number_of_threads)
self.assertEqual(self.runtime_params.timestep_log_interval, actual_params.timestep_log_interval)
- def assert_kernel_forcing_scaled_by_factor(self, actual_kernel: LBMKernel, factor: int):
+ def assert_kernel_forcing_scaled_by_factor(self, actual_kernel: cpu.kernel.LBMKernel, factor: int):
self.assertEqual(self.kernel.type, actual_kernel.type)
- self.assertEqual(self.kernel.use_forcing, actual_kernel.use_forcing)
+ self.assertEqual(self.kernel.use_forcing, actual_kernel.cpu.parameters.use_forcing)
self.assertAlmostEqual(self.kernel.forcing_in_x1 / factor, actual_kernel.forcing_in_x1)
self.assertAlmostEqual(self.kernel.forcing_in_x2, actual_kernel.forcing_in_x2)
self.assertAlmostEqual(self.kernel.forcing_in_x3, actual_kernel.forcing_in_x3)
@@ -80,14 +112,14 @@ class OneDirectionalAcousticScalingTest(unittest.TestCase):
@staticmethod
def make_kernel():
- kernel = LBMKernel(KernelType.CompressibleCumulantFourthOrderViscosity)
+ kernel = cpu.kernel.LBMKernel(cpu.kernel.KernelType.CompressibleCumulantFourthOrderViscosity)
kernel.use_forcing = True
kernel.forcing_in_x1 = 5e-10
return kernel
@staticmethod
def make_runtime_params():
- runtime_params = RuntimeParameters()
+ runtime_params = cpu.parameters.RuntimeParameters()
runtime_params.number_of_threads = 4
runtime_params.number_of_timesteps = 4_000_000
runtime_params.timestep_log_interval = 1_000_000
@@ -95,13 +127,13 @@ class OneDirectionalAcousticScalingTest(unittest.TestCase):
@staticmethod
def make_physical_params():
- physical_params = PhysicalParameters()
+ physical_params = cpu.parameters.PhysicalParameters()
physical_params.lattice_viscosity = 1e-4
return physical_params
@staticmethod
def make_grid_params():
- grid_params = GridParameters()
+ grid_params = cpu.parameters.GridParameters()
grid_params.node_distance = 1
grid_params.number_of_nodes_per_direction = [1, 1, 16]
grid_params.blocks_per_direction = [1, 1, 16]
diff --git a/Python/tests/test_boundaryconditions.py b/Python/tests/test_boundaryconditions.py
index e004ddfa21c78ea3d63a89f5dbc3bd7438a18ff1..d914c50cad2051188331b2efe604907091fa731e 100644
--- a/Python/tests/test_boundaryconditions.py
+++ b/Python/tests/test_boundaryconditions.py
@@ -1,5 +1,39 @@
+r"""
+=======================================================================================
+ ____ ____ __ ______ __________ __ __ __ __
+ \ \ | | | | | _ \ |___ ___| | | | | / \ | |
+ \ \ | | | | | |_) | | | | | | | / \ | |
+ \ \ | | | | | _ / | | | | | | / /\ \ | |
+ \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
+ \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
+ \ \ | | ________________________________________________________________
+ \ \ | | | ______________________________________________________________|
+ \ \| | | | __ __ __ __ ______ _______
+ \ | | |_____ | | | | | | | | | _ \ / _____)
+ \ | | _____| | | | | | | | | | | \ \ \_______
+ \ | | | | |_____ | \_/ | | | | |_/ / _____ |
+ \ _____| |__| |________| \_______/ |__| |______/ (_______/
+
+ This file is part of VirtualFluids. VirtualFluids is free software: you can
+ redistribute it and/or modify it under the terms of the GNU General Public
+ License as published by the Free Software Foundation, either version 3 of
+ the License, or (at your option) any later version.
+
+ VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
+ ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
+ for more details.
+
+ You should have received a copy of the GNU General Public License along
+ with VirtualFluids (see COPYING.txt). If not, see <http://www.gnu.org/licenses/>.
+
+! \file test_boundaryconditions.py
+! \ingroup tests
+! \author Sven Marcus, Henry Korb
+=======================================================================================
+"""
import unittest
-from pyfluids.cpu.boundaryconditions import *
+from pyfluids import cpu
class BoundaryConditionsTest(unittest.TestCase):
@@ -8,13 +42,13 @@ class BoundaryConditionsTest(unittest.TestCase):
"""
Should be able to create NoSlipBoundaryCondition
"""
- sut = NoSlipBoundaryCondition()
+ sut = cpu.boundaryconditions.NoSlipBoundaryCondition()
def test__can_create_velocity_bc(self):
"""
Should be able to create VelocityBoundaryCondition
"""
- sut = VelocityBoundaryCondition()
+ sut = cpu.boundaryconditions.VelocityBoundaryCondition()
def test__can_create_velocity_bc_with_directions_function_and_time(self):
"""
@@ -24,7 +58,7 @@ class BoundaryConditionsTest(unittest.TestCase):
parser = Parser()
parser.expression = "1"
- sut = VelocityBoundaryCondition(True, True, True, parser, 0, 1)
+ sut = cpu.boundaryconditions.VelocityBoundaryCondition(True, True, True, parser, 0, 1)
def test__can_create_velocity_bc_with_directions__function_per_direction__and__time(self):
"""
@@ -40,7 +74,7 @@ class BoundaryConditionsTest(unittest.TestCase):
f3 = Parser()
f3.expression = "1"
- sut = VelocityBoundaryCondition(True, True, True, f1, f2, f3, 0, 1)
+ sut = cpu.boundaryconditions.VelocityBoundaryCondition(True, True, True, f1, f2, f3, 0, 1)
def test__can_create_velocity_bc_with_speeds_and_times_per_direction(self):
"""
@@ -51,11 +85,11 @@ class BoundaryConditionsTest(unittest.TestCase):
start2, end2 = 1, 2
start3, end3 = 2, 3
- sut = VelocityBoundaryCondition(vx1, start1, end1, vx2, start2, end2, vx3, start3, end3)
+ sut = cpu.boundaryconditions.VelocityBoundaryCondition(vx1, start1, end1, vx2, start2, end2, vx3, start3, end3)
def test__can_create_non_reflecting_outflow(self):
"""
Should be able to create NonReflectingOutflow
"""
- sut = NonReflectingOutflow()
+ sut = cpu.boundaryconditions.NonReflectingOutflow()
diff --git a/Python/tests/test_geometry.py b/Python/tests/test_geometry.py
index 5bb89eb245b6055653b78fde381da050d402b0cc..3d297f5c176cd99f7969adf37333588d86b77627 100644
--- a/Python/tests/test_geometry.py
+++ b/Python/tests/test_geometry.py
@@ -1,6 +1,40 @@
+r"""
+=======================================================================================
+ ____ ____ __ ______ __________ __ __ __ __
+ \ \ | | | | | _ \ |___ ___| | | | | / \ | |
+ \ \ | | | | | |_) | | | | | | | / \ | |
+ \ \ | | | | | _ / | | | | | | / /\ \ | |
+ \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
+ \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
+ \ \ | | ________________________________________________________________
+ \ \ | | | ______________________________________________________________|
+ \ \| | | | __ __ __ __ ______ _______
+ \ | | |_____ | | | | | | | | | _ \ / _____)
+ \ | | _____| | | | | | | | | | | \ \ \_______
+ \ | | | | |_____ | \_/ | | | | |_/ / _____ |
+ \ _____| |__| |________| \_______/ |__| |______/ (_______/
+
+ This file is part of VirtualFluids. VirtualFluids is free software: you can
+ redistribute it and/or modify it under the terms of the GNU General Public
+ License as published by the Free Software Foundation, either version 3 of
+ the License, or (at your option) any later version.
+
+ VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
+ ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
+ for more details.
+
+ You should have received a copy of the GNU General Public License along
+ with VirtualFluids (see COPYING.txt). If not, see <http://www.gnu.org/licenses/>.
+
+! \file test_geometry.py
+! \ingroup tests
+! \author Sven Marcus, Henry Korb
+=======================================================================================
+"""
import unittest
-from pyfluids.cpu.geometry import *
+from pyfluids import cpu
class TestGeometry(unittest.TestCase):
@@ -9,7 +43,7 @@ class TestGeometry(unittest.TestCase):
"""
WHEN setting point coordinates in constructor THEN point should have coordinates
"""
- sut = GbPoint3D(4, 8, 3)
+ sut = cpu.geometry.GbPoint3D(4, 8, 3)
self.assertEqual(sut.x1, 4)
self.assertEqual(sut.x2, 8)
@@ -19,7 +53,7 @@ class TestGeometry(unittest.TestCase):
"""
WHEN setting point coordinates THEN point should have coordinates
"""
- sut = GbPoint3D()
+ sut = cpu.geometry.GbPoint3D()
sut.x1 = 4
sut.x2 = 8
@@ -33,10 +67,10 @@ class TestGeometry(unittest.TestCase):
"""
WHEN setting line points THEN line should have points
"""
- sut = GbLine3D()
+ sut = cpu.geometry.GbLine3D()
- point1 = GbPoint3D()
- point2 = GbPoint3D()
+ point1 = cpu.geometry.GbPoint3D()
+ point2 = cpu.geometry.GbPoint3D()
sut.point1 = point1
sut.point2 = point2
diff --git a/Python/tests/test_kernel.py b/Python/tests/test_kernel.py
index 8f58a1c869f9e292856268d43245a75f1dcfe213..e0159bec6802cb08d73214038b177091879fee46 100644
--- a/Python/tests/test_kernel.py
+++ b/Python/tests/test_kernel.py
@@ -1,12 +1,46 @@
+r"""
+=======================================================================================
+ ____ ____ __ ______ __________ __ __ __ __
+ \ \ | | | | | _ \ |___ ___| | | | | / \ | |
+ \ \ | | | | | |_) | | | | | | | / \ | |
+ \ \ | | | | | _ / | | | | | | / /\ \ | |
+ \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
+ \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
+ \ \ | | ________________________________________________________________
+ \ \ | | | ______________________________________________________________|
+ \ \| | | | __ __ __ __ ______ _______
+ \ | | |_____ | | | | | | | | | _ \ / _____)
+ \ | | _____| | | | | | | | | | | \ \ \_______
+ \ | | | | |_____ | \_/ | | | | |_/ / _____ |
+ \ _____| |__| |________| \_______/ |__| |______/ (_______/
+
+ This file is part of VirtualFluids. VirtualFluids is free software: you can
+ redistribute it and/or modify it under the terms of the GNU General Public
+ License as published by the Free Software Foundation, either version 3 of
+ the License, or (at your option) any later version.
+
+ VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
+ ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
+ for more details.
+
+ You should have received a copy of the GNU General Public License along
+ with VirtualFluids (see COPYING.txt). If not, see <http://www.gnu.org/licenses/>.
+
+! \file test_kernel.py
+! \ingroup tests
+! \author Sven Marcus, Henry Korb
+=======================================================================================
+"""
import unittest
-from pyfluids.cpu.kernel import LBMKernel, KernelType
+from pyfluids import cpu
class TestLBMKernel(unittest.TestCase):
def setUp(self) -> None:
- self.sut = LBMKernel(KernelType.BGK)
+ self.sut = cpu.kernel.LBMKernel(cpu.kernel.KernelType.BGK)
def test_lbm_kernel__when_use_forcing_set_to_true__use_forcing_should_be_true(self) -> None:
"""
@@ -57,4 +91,4 @@ class TestLBMKernel(unittest.TestCase):
"""
actual = self.sut.type
- self.assertEqual(KernelType.BGK, actual)
+ self.assertEqual(cpu.kernel.KernelType.BGK, actual)
diff --git a/apps/cpu/ConvectionOfVortex/CMakeLists.txt b/apps/cpu/ConvectionOfVortex/CMakeLists.txt
index de3034c04bb2f2f16edd9b4bf48db81c83d15b3e..33d60676c7e0dfdde411c3c5b92a2534ea54fbfe 100644
--- a/apps/cpu/ConvectionOfVortex/CMakeLists.txt
+++ b/apps/cpu/ConvectionOfVortex/CMakeLists.txt
@@ -1,3 +1,6 @@
+########################################################
+## C++ PROJECT ###
+########################################################
PROJECT(ConvectionOfVortex)
vf_add_library(BUILDTYPE binary PRIVATE_LINK VirtualFluidsCore basics ${MPI_CXX_LIBRARIES} FILES cov.cpp )
diff --git a/apps/gpu/LBM/ActuatorLine/ActuatorLine.cpp b/apps/gpu/LBM/ActuatorLine/ActuatorLine.cpp
index 58e5aede18b9c4197b4d21b129c6347023b9390e..9d982ebac0059b4512041194100f6e1fdfa61924 100644
--- a/apps/gpu/LBM/ActuatorLine/ActuatorLine.cpp
+++ b/apps/gpu/LBM/ActuatorLine/ActuatorLine.cpp
@@ -1,4 +1,35 @@
-
+//=======================================================================================
+// ____ ____ __ ______ __________ __ __ __ __
+// \ \ | | | | | _ \ |___ ___| | | | | / \ | |
+// \ \ | | | | | |_) | | | | | | | / \ | |
+// \ \ | | | | | _ / | | | | | | / /\ \ | |
+// \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
+// \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
+// \ \ | | ________________________________________________________________
+// \ \ | | | ______________________________________________________________|
+// \ \| | | | __ __ __ __ ______ _______
+// \ | | |_____ | | | | | | | | | _ \ / _____)
+// \ | | _____| | | | | | | | | | | \ \ \_______
+// \ | | | | |_____ | \_/ | | | | |_/ / _____ |
+// \ _____| |__| |________| \_______/ |__| |______/ (_______/
+//
+// This file is part of VirtualFluids. VirtualFluids is free software: you can
+// redistribute it and/or modify it under the terms of the GNU General Public
+// License as published by the Free Software Foundation, either version 3 of
+// the License, or (at your option) any later version.
+//
+// VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
+// ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+// FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
+// for more details.
+//
+// You should have received a copy of the GNU General Public License along
+// with VirtualFluids (see COPYING.txt). If not, see <http://www.gnu.org/licenses/>.
+//
+//! \file ActuatorLine.cpp
+//! \ingroup ActuatorLine
+//! \author Henry Korb, Henrik Asmuth
+//=======================================================================================
#define _USE_MATH_DEFINES
#include <math.h>
#include <string>
@@ -28,12 +59,14 @@
#include "GridGenerator/grid/GridBuilder/LevelGridBuilder.h"
#include "GridGenerator/grid/GridBuilder/MultipleGridBuilder.h"
#include "GridGenerator/grid/BoundaryConditions/Side.h"
+#include "GridGenerator/grid/BoundaryConditions/BoundaryCondition.h"
+
#include "GridGenerator/grid/GridFactory.h"
#include "GridGenerator/io/SimulationFileWriter/SimulationFileWriter.h"
#include "GridGenerator/io/GridVTKWriter/GridVTKWriter.h"
-#include "GridGenerator/io/STLReaderWriter/STLReader.h"
-#include "GridGenerator/io/STLReaderWriter/STLWriter.h"
+#include "GridGenerator/TransientBCSetter/TransientBCSetter.h"
+
//////////////////////////////////////////////////////////////////////////
@@ -44,10 +77,12 @@
#include "VirtualFluids_GPU/DataStructureInitializer/GridReaderFiles/GridReader.h"
#include "VirtualFluids_GPU/Parameter/Parameter.h"
#include "VirtualFluids_GPU/Output/FileWriter.h"
-#include "VirtualFluids_GPU/PreCollisionInteractor/ActuatorLine.h"
+#include "VirtualFluids_GPU/PreCollisionInteractor/ActuatorFarm.h"
#include "VirtualFluids_GPU/PreCollisionInteractor/Probes/PointProbe.h"
#include "VirtualFluids_GPU/PreCollisionInteractor/Probes/PlaneProbe.h"
#include "VirtualFluids_GPU/Factories/BoundaryConditionFactory.h"
+#include "VirtualFluids_GPU/TurbulenceModels/TurbulenceModelFactory.h"
+#include "VirtualFluids_GPU/Factories/GridScalingFactory.h"
#include "VirtualFluids_GPU/GPU/CudaMemoryManager.h"
@@ -63,26 +98,10 @@
LbmOrGks lbmOrGks = LBM;
-const real reference_diameter = 126.0; // diameter in m
-
-const real L_x = 10*reference_diameter;
-const real L_y = 6*reference_diameter;
-const real L_z = 6*reference_diameter;
-
-const real viscosity = 1.56e-5;
-
-const real velocity = 9.0;
-
-const real mach = 0.1;
-
-const uint nodes_per_diameter = 16;
-
std::string path(".");
std::string simulationName("ActuatorLine");
-const float tOut = 100;
-const float tEnd = 280; // total time of simulation in s
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
@@ -98,30 +117,59 @@ void multipleLevel(const std::string& configPath)
vf::gpu::Communicator& communicator = vf::gpu::Communicator::getInstance();
auto gridFactory = GridFactory::make();
- gridFactory->setTriangularMeshDiscretizationMethod(TriangularMeshDiscretizationMethod::POINT_IN_OBJECT);
auto gridBuilder = MultipleGridBuilder::makeShared(gridFactory);
- ////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
- ////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
+ vf::basics::ConfigurationFile config;
+ config.load(configPath);
+
+ const real reference_diameter = config.getValue<real>("ReferenceDiameter");
+ const uint nodes_per_diameter = config.getValue<uint>("NodesPerDiameter");
+ const real velocity = config.getValue<real>("Velocity");
+
+
+ const real L_x = 24*reference_diameter;
+ const real L_y = 6*reference_diameter;
+ const real L_z = 6*reference_diameter;
+
+ const real viscosity = 1.56e-5;
+
+ const real mach = 0.1;
+
+
+ const float tStartOut = config.getValue<real>("tStartOut");
+ const float tOut = config.getValue<real>("tOut");
+ const float tEnd = config.getValue<real>("tEnd"); // total time of simulation
+
+ const float tStartAveraging = config.getValue<real>("tStartAveraging");
+ const float tStartTmpAveraging = config.getValue<real>("tStartTmpAveraging");
+ const float tAveraging = config.getValue<real>("tAveraging");
+ const float tStartOutProbe = config.getValue<real>("tStartOutProbe");
+ const float tOutProbe = config.getValue<real>("tOutProbe");
+
+ SPtr<Parameter> para = std::make_shared<Parameter>(communicator.getNummberOfProcess(), communicator.getPID(), &config);
+ BoundaryConditionFactory bcFactory = BoundaryConditionFactory();
+ GridScalingFactory scalingFactory = GridScalingFactory();
+
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
const real dx = reference_diameter/real(nodes_per_diameter);
+ real turbPos[3] = {3*reference_diameter, 3*reference_diameter, 3*reference_diameter};
+
gridBuilder->addCoarseGrid(0.0, 0.0, 0.0,
L_x, L_y, L_z, dx);
+ gridBuilder->setNumberOfLayers(4,0);
+ gridBuilder->addGrid( new Cuboid( turbPos[0]-1.5*reference_diameter, turbPos[1]-1.5*reference_diameter, turbPos[2]-1.5*reference_diameter,
+ turbPos[0]+10.0*reference_diameter, turbPos[1]+1.5*reference_diameter, turbPos[2]+1.5*reference_diameter) , 1 );
+ para->setMaxLevel(2);
+ scalingFactory.setScalingFactory(GridScalingFactory::GridScaling::ScaleCompressible);
+
gridBuilder->setPeriodicBoundaryCondition(false, false, false);
gridBuilder->buildGrids(lbmOrGks, false); // buildGrids() has to be called before setting the BCs!!!!
- ////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
-
- vf::basics::ConfigurationFile config;
- config.load(configPath);
- ////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////^
- SPtr<Parameter> para = std::make_shared<Parameter>(communicator.getNummberOfProcess(), communicator.getPID(), &config);
- BoundaryConditionFactory bcFactory = BoundaryConditionFactory();
- ////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
+ ////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
const real dt = dx * mach / (sqrt(3) * velocity);
@@ -140,14 +188,11 @@ void multipleLevel(const std::string& configPath)
para->setPrintFiles(true);
- para->setMaxLevel(1);
-
-
para->setVelocityLB(velocityLB);
para->setViscosityLB(viscosityLB);
para->setVelocityRatio( dx / dt );
para->setViscosityRatio( dx*dx/dt );
- para->setMainKernel("CumulantK17CompChim");
+ para->setMainKernel("CumulantK17");
para->setInitialCondition([&](real coordX, real coordY, real coordZ, real &rho, real &vx, real &vy, real &vz) {
rho = (real)0.0;
@@ -156,13 +201,15 @@ void multipleLevel(const std::string& configPath)
vz = (real)0.0;
});
+ para->setTimestepStartOut( uint(tStartOut/dt) );
para->setTimestepOut( uint(tOut/dt) );
para->setTimestepEnd( uint(tEnd/dt) );
para->setIsBodyForce( true );
-
+ para->setUseStreams( true );
/////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
+
gridBuilder->setVelocityBoundaryCondition(SideType::MX, velocityLB, 0.0, 0.0);
gridBuilder->setVelocityBoundaryCondition(SideType::MY, velocityLB, 0.0, 0.0);
@@ -172,42 +219,52 @@ void multipleLevel(const std::string& configPath)
gridBuilder->setPressureBoundaryCondition(SideType::PX, 0.0);
bcFactory.setVelocityBoundaryCondition(BoundaryConditionFactory::VelocityBC::VelocityAndPressureCompressible);
- bcFactory.setPressureBoundaryCondition(BoundaryConditionFactory::PressureBC::PressureNonEquilibriumCompressible);
+ bcFactory.setPressureBoundaryCondition(BoundaryConditionFactory::PressureBC::OutflowNonReflective);
+
+ SPtr<TurbulenceModelFactory> tmFactory = std::make_shared<TurbulenceModelFactory>(para);
+ tmFactory->readConfigFile(config);
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
- real turbPos[3] = {3*reference_diameter, 3*reference_diameter, 3*reference_diameter};
- real epsilon = 5.f; // width of gaussian smearing
- real density = 1.225f;
- int level = 0;
- uint nBlades = 3;
- uint nBladeNodes = 32;
+ int level = 1; // grid level at which the turbine samples velocities and distributes forces
+ const real epsilon = dx*exp2(-level)*1.5; // width of gaussian smearing
+ const real density = 1.225f;
+ const uint nBlades = 3;
+ const uint nBladeNodes = 32;
+ const real tipspeed_ratio = 7.5f; // tipspeed ratio = angular vel * radius / inflow vel
+ const real omega = 2*tipspeed_ratio*velocity/reference_diameter;
+
+
+ SPtr<ActuatorFarm> actuator_farm = std::make_shared<ActuatorFarm>(nBlades, density, nBladeNodes, epsilon, level, dt, dx, true);
+ std::vector<real> bladeRadii;
+ real dr = reference_diameter/(nBladeNodes*2);
+ for(uint node=0; node<nBladeNodes; node++){ bladeRadii.emplace_back(dr*(node+1)); }
+ actuator_farm->addTurbine(turbPos[0], turbPos[1], turbPos[2], reference_diameter, omega, 0, 0, bladeRadii);
+ para->addActuator( actuator_farm );
- SPtr<ActuatorLine> actuator_line =SPtr<ActuatorLine>( new ActuatorLine(nBlades, density, nBladeNodes, epsilon, turbPos[0], turbPos[1], turbPos[2], reference_diameter, level, dt, dx) );
- para->addActuator( actuator_line );
- SPtr<PointProbe> pointProbe = SPtr<PointProbe>( new PointProbe("pointProbe", para->getOutputPath(), 100, 1, 500, 100) );
- std::vector<real> probeCoordsX = {reference_diameter,2*reference_diameter,5*reference_diameter};
- std::vector<real> probeCoordsY = {3*reference_diameter,3*reference_diameter,3*reference_diameter};
- std::vector<real> probeCoordsZ = {3*reference_diameter,3*reference_diameter,3*reference_diameter};
- pointProbe->addProbePointsFromList(probeCoordsX, probeCoordsY, probeCoordsZ);
- // pointProbe->addProbePointsFromXNormalPlane(2*D, 0.0, 0.0, L_y, L_z, (uint)L_y/dx, (uint)L_z/dx);
+ // SPtr<PointProbe> pointProbe = std::make_shared<PointProbe>("pointProbe", para->getOutputPath(), 100, 1, 500, 100);
+ // std::vector<real> probeCoordsX = {reference_diameter,2*reference_diameter,5*reference_diameter};
+ // std::vector<real> probeCoordsY = {3*reference_diameter,3*reference_diameter,3*reference_diameter};
+ // std::vector<real> probeCoordsZ = {3*reference_diameter,3*reference_diameter,3*reference_diameter};
+ // pointProbe->addProbePointsFromList(probeCoordsX, probeCoordsY, probeCoordsZ);
+ // // pointProbe->addProbePointsFromXNormalPlane(2*D, 0.0, 0.0, L_y, L_z, (uint)L_y/dx, (uint)L_z/dx);
- pointProbe->addStatistic(Statistic::Means);
- pointProbe->addStatistic(Statistic::Variances);
- para->addProbe( pointProbe );
+ // pointProbe->addStatistic(Statistic::Means);
+ // pointProbe->addStatistic(Statistic::Variances);
+ // para->addProbe( pointProbe );
- SPtr<PlaneProbe> planeProbe = SPtr<PlaneProbe>( new PlaneProbe("planeProbe", para->getOutputPath(), 100, 500, 100, 100) );
- planeProbe->setProbePlane(5*reference_diameter, 0, 0, dx, L_y, L_z);
- planeProbe->addStatistic(Statistic::Means);
- para->addProbe( planeProbe );
+ // SPtr<PlaneProbe> planeProbe = std::make_shared<PlaneProbe>("planeProbe", para->getOutputPath(), 100, 500, 100, 100);
+ // planeProbe->setProbePlane(5*reference_diameter, 0, 0, dx, L_y, L_z);
+ // planeProbe->addStatistic(Statistic::Means);
+ // para->addProbe( planeProbe );
auto cudaMemoryManager = std::make_shared<CudaMemoryManager>(para);
auto gridGenerator = GridProvider::makeGridGenerator(gridBuilder, para, cudaMemoryManager, communicator);
- Simulation sim(para, cudaMemoryManager, communicator, *gridGenerator, &bcFactory);
+ Simulation sim(para, cudaMemoryManager, communicator, *gridGenerator, &bcFactory, tmFactory, &scalingFactory);
sim.run();
}
diff --git a/apps/gpu/LBM/ActuatorLine/configActuatorLine.txt b/apps/gpu/LBM/ActuatorLine/configActuatorLine.txt
index 233994f0d32a48190d84f7044500e24b06b926a9..5799f24716777295b2f835ab00561ff767ba87b9 100644
--- a/apps/gpu/LBM/ActuatorLine/configActuatorLine.txt
+++ b/apps/gpu/LBM/ActuatorLine/configActuatorLine.txt
@@ -6,3 +6,29 @@ Path = .
#informations for reading
##################################################
GridPath=.
+##################################################
+ReferenceDiameter=126
+NodesPerDiameter=32
+Velocity=9
+##################################################
+tStartOut=100
+tOut=100
+tEnd=1000
+##################################################
+
+tStartTmpAveraging=100
+tStartAveraging=100
+tAveraging=100
+tTmpAveraging=100
+tStartOutProbe=100
+tOutProbe=100
+
+##################################################
+#TurbulenceModel = QR
+#SGSconstant = 0.3333333
+#
+#QuadricLimiterP = 100000.0
+#QuadricLimiterM = 100000.0
+#QuadricLimiterD = 100000.0
+##################################################
+
diff --git a/apps/gpu/LBM/BoundaryLayer/BoundaryLayer.cpp b/apps/gpu/LBM/BoundaryLayer/BoundaryLayer.cpp
index 991025b649d69305c030fe2f1dd1763a2137af9b..5fc31904433bfe2df0722ab1c63f574d3fcb9a35 100644
--- a/apps/gpu/LBM/BoundaryLayer/BoundaryLayer.cpp
+++ b/apps/gpu/LBM/BoundaryLayer/BoundaryLayer.cpp
@@ -1,4 +1,35 @@
-
+//=======================================================================================
+// ____ ____ __ ______ __________ __ __ __ __
+// \ \ | | | | | _ \ |___ ___| | | | | / \ | |
+// \ \ | | | | | |_) | | | | | | | / \ | |
+// \ \ | | | | | _ / | | | | | | / /\ \ | |
+// \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
+// \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
+// \ \ | | ________________________________________________________________
+// \ \ | | | ______________________________________________________________|
+// \ \| | | | __ __ __ __ ______ _______
+// \ | | |_____ | | | | | | | | | _ \ / _____)
+// \ | | _____| | | | | | | | | | | \ \ \_______
+// \ | | | | |_____ | \_/ | | | | |_/ / _____ |
+// \ _____| |__| |________| \_______/ |__| |______/ (_______/
+//
+// This file is part of VirtualFluids. VirtualFluids is free software: you can
+// redistribute it and/or modify it under the terms of the GNU General Public
+// License as published by the Free Software Foundation, either version 3 of
+// the License, or (at your option) any later version.
+//
+// VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
+// ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+// FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
+// for more details.
+//
+// You should have received a copy of the GNU General Public License along
+// with VirtualFluids (see COPYING.txt). If not, see <http://www.gnu.org/licenses/>.
+//
+//! \file BoundaryLayer.cpp
+//! \ingroup BoundaryLayer
+//! \author Henry Korb, Henrik Asmuth
+//=======================================================================================
#define _USE_MATH_DEFINES
#include <math.h>
#include <string>
@@ -8,6 +39,7 @@
#include <fstream>
#include <exception>
#include <memory>
+#include <numeric>
//////////////////////////////////////////////////////////////////////////
@@ -19,6 +51,7 @@
#include "Core/VectorTypes.h"
#include <basics/config/ConfigurationFile.h>
+#include "lbm/constants/NumericConstants.h"
#include <logger/Logger.h>
@@ -28,12 +61,16 @@
#include "GridGenerator/grid/GridBuilder/LevelGridBuilder.h"
#include "GridGenerator/grid/GridBuilder/MultipleGridBuilder.h"
#include "GridGenerator/grid/BoundaryConditions/Side.h"
+#include "GridGenerator/grid/BoundaryConditions/BoundaryCondition.h"
+
#include "GridGenerator/grid/GridFactory.h"
+#include "geometries/Cuboid/Cuboid.h"
+#include "geometries/TriangularMesh/TriangularMesh.h"
+
#include "GridGenerator/io/SimulationFileWriter/SimulationFileWriter.h"
#include "GridGenerator/io/GridVTKWriter/GridVTKWriter.h"
-#include "GridGenerator/io/STLReaderWriter/STLReader.h"
-#include "GridGenerator/io/STLReaderWriter/STLWriter.h"
+#include "GridGenerator/TransientBCSetter/TransientBCSetter.h"
//////////////////////////////////////////////////////////////////////////
@@ -44,24 +81,28 @@
#include "VirtualFluids_GPU/DataStructureInitializer/GridReaderFiles/GridReader.h"
#include "VirtualFluids_GPU/Parameter/Parameter.h"
#include "VirtualFluids_GPU/Output/FileWriter.h"
-#include "VirtualFluids_GPU/PreCollisionInteractor/ActuatorLine.h"
#include "VirtualFluids_GPU/PreCollisionInteractor/Probes/PointProbe.h"
#include "VirtualFluids_GPU/PreCollisionInteractor/Probes/PlaneProbe.h"
#include "VirtualFluids_GPU/PreCollisionInteractor/Probes/PlanarAverageProbe.h"
#include "VirtualFluids_GPU/PreCollisionInteractor/Probes/WallModelProbe.h"
+#include "VirtualFluids_GPU/PreCollisionInteractor/PrecursorWriter.h"
#include "VirtualFluids_GPU/Factories/BoundaryConditionFactory.h"
+#include "VirtualFluids_GPU/Factories/GridScalingFactory.h"
#include "VirtualFluids_GPU/TurbulenceModels/TurbulenceModelFactory.h"
#include "VirtualFluids_GPU/GPU/CudaMemoryManager.h"
+#include "utilities/communication.h"
+
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
std::string path(".");
-std::string simulationName("BoundayLayer");
+std::string simulationName("BoundaryLayer");
+using namespace vf::lbm::constant;
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
@@ -87,8 +128,16 @@ void multipleLevel(const std::string& configPath)
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////^
SPtr<Parameter> para = std::make_shared<Parameter>(communicator.getNummberOfProcess(), communicator.getPID(), &config);
BoundaryConditionFactory bcFactory = BoundaryConditionFactory();
-
+ GridScalingFactory scalingFactory = GridScalingFactory();
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
+
+ const int nProcs = communicator.getNummberOfProcess();
+ const uint procID = vf::gpu::Communicator::getInstance().getPID();
+ std::vector<uint> devices(10);
+ std::iota(devices.begin(), devices.end(), 0);
+ para->setDevices(devices);
+ para->setMaxDev(nProcs);
+
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
//
@@ -100,23 +149,45 @@ void multipleLevel(const std::string& configPath)
LbmOrGks lbmOrGks = LBM;
- const real H = 1000.0; // boundary layer height in m
+ const real H = config.getValue("boundaryLayerHeight", 1000.0); // boundary layer height in m
const real L_x = 6*H;
const real L_y = 4*H;
- const real L_z = 1*H;
+ const real L_z = H;
+
+ const real z0 = config.getValue("z0", 0.1f); // roughness length in m
+ const real u_star = config.getValue("u_star", 0.4f); //friction velocity in m/s
+ const real kappa = config.getValue("vonKarmanConstant", 0.4f); // von Karman constant
- const real z0 = 0.1; // roughness length in m
- const real u_star = 0.4; //friction velocity in m/s
- const real kappa = 0.4; // von Karman constant
+ const real viscosity = config.getValue("viscosity", 1.56e-5f);
- const real viscosity = 1.56e-5;
+ const real velocity = 0.5f*u_star/kappa*log(H/z0+1.f); //0.5 times max mean velocity at the top in m/s
- const real velocity = 0.5*u_star/kappa*log(L_z/z0); //0.5 times max mean velocity at the top in m/s
+ const real mach = config.getValue<real>("Ma", 0.1);
- const real mach = config.contains("Ma")? config.getValue<real>("Ma"): 0.1;
+ const uint nodes_per_H = config.getValue<uint>("nz", 64);
- const uint nodes_per_H = config.contains("nz")? config.getValue<uint>("nz"): 64;
+ const bool writePrecursor = config.getValue("writePrecursor", false);
+ bool useDistributions;
+ std::string precursorDirectory;
+ int nTWritePrecursor; real tStartPrecursor, posXPrecursor;
+ if(writePrecursor)
+ {
+ nTWritePrecursor = config.getValue<int>("nTimestepsWritePrecursor");
+ tStartPrecursor = config.getValue<real>("tStartPrecursor");
+ posXPrecursor = config.getValue<real>("posXPrecursor");
+ useDistributions = config.getValue<bool>("useDistributions", false);
+ precursorDirectory = config.getValue<std::string>("precursorDirectory");
+ }
+
+ const bool readPrecursor = config.getValue("readPrecursor", false);
+ int timestepsBetweenReadsPrecursor;
+ if(readPrecursor)
+ {
+ timestepsBetweenReadsPrecursor = config.getValue<int>("nTimestepsReadPrecursor");
+ precursorDirectory = config.getValue<std::string>("precursorDirectory");
+ useDistributions = config.getValue<bool>("useDistributions", false);
+ }
// all in s
const float tStartOut = config.getValue<real>("tStartOut");
@@ -130,7 +201,7 @@ void multipleLevel(const std::string& configPath)
const float tOutProbe = config.getValue<real>("tOutProbe");
- const real dx = L_z/real(nodes_per_H);
+ const real dx = H/real(nodes_per_H);
const real dt = dx * mach / (sqrt(3) * velocity);
@@ -155,15 +226,17 @@ void multipleLevel(const std::string& configPath)
para->setPrintFiles(true);
- para->setForcing(pressureGradientLB, 0, 0);
+ if(!readPrecursor) para->setForcing(pressureGradientLB, 0, 0);
para->setVelocityLB(velocityLB);
para->setViscosityLB(viscosityLB);
para->setVelocityRatio( dx / dt );
para->setViscosityRatio( dx*dx/dt );
para->setDensityRatio( 1.0 );
- para->setMainKernel("TurbulentViscosityCumulantK17CompChim");
-
+ bool useStreams = (nProcs > 1 ? true: false);
+ // useStreams=false;
+ para->setUseStreams(useStreams);
+ para->setMainKernel("CumulantK17");
para->setIsBodyForce( config.getValue<bool>("bodyForce") );
para->setTimestepStartOut(uint(tStartOut/dt) );
@@ -172,64 +245,206 @@ void multipleLevel(const std::string& configPath)
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
- SPtr<TurbulenceModelFactory> tmFactory = SPtr<TurbulenceModelFactory>( new TurbulenceModelFactory(para) );
+ SPtr<TurbulenceModelFactory> tmFactory = std::make_shared<TurbulenceModelFactory>(para);
tmFactory->readConfigFile( config );
+ /////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
+
+ const real xSplit = L_x/nProcs;
+ const real overlap = 8.0*dx;
+
+ real xMin = procID * xSplit;
+ real xMax = (procID+1) * xSplit;
+ real xGridMin = procID * xSplit;
+ real xGridMax = (procID+1) * xSplit;
+
+ real yMin = 0.0;
+ real yMax = L_y;
+ real zMin = 0.0;
+ real zMax = L_z;
+
+ bool isFirstSubDomain = (procID == 0 && nProcs > 1)? true: false;
+ bool isLastSubDomain = (procID == nProcs-1 && nProcs > 1)? true: false;
+ bool isMidSubDomain = (!isFirstSubDomain && !isLastSubDomain && nProcs > 1)? true: false;
- // tmFactory->setTurbulenceModel(TurbulenceModel::AMD);
- // tmFactory->setModelConstant(config.getValue<real>("SGSconstant"));
+ if(isFirstSubDomain)
+ {
+ xGridMax += overlap;
+ if(!readPrecursor) xGridMin -= overlap;
+ }
+ if(isLastSubDomain)
+ {
+ xGridMin -= overlap;
+ if(!readPrecursor) xGridMax += overlap;
+ }
+ if(isMidSubDomain)
+ {
+ xGridMax += overlap;
+ xGridMin -= overlap;
+ }
- /////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
+ gridBuilder->addCoarseGrid( xGridMin, 0.0, 0.0,
+ xGridMax, L_y, L_z, dx);
+ if(true)// Add refinement
+ {
+ gridBuilder->setNumberOfLayers(4,0);
+ real xMaxRefinement = readPrecursor? xGridMax-H: xGridMax; //Stop refinement some distance before outlet if domain ist not periodic
+ gridBuilder->addGrid( new Cuboid( xGridMin, 0.f, 0.f, xMaxRefinement, L_y, 0.5*L_z) , 1 );
+ para->setMaxLevel(2);
+ scalingFactory.setScalingFactory(GridScalingFactory::GridScaling::ScaleCompressible);
+ }
+
+ if(nProcs > 1)
+ {
+ gridBuilder->setSubDomainBox(
+ std::make_shared<BoundingBox>(xMin, xMax, yMin, yMax, zMin, zMax));
+ gridBuilder->setPeriodicBoundaryCondition(false, true, false);
+ }
+ else
+ {
+ gridBuilder->setPeriodicBoundaryCondition(!readPrecursor, true, false);
+ }
- gridBuilder->addCoarseGrid(0.0, 0.0, 0.0,
- L_x, L_y, L_z, dx);
- // gridBuilder->setNumberOfLayers(12, 8);
+ gridBuilder->buildGrids(lbmOrGks, true); // buildGrids() has to be called before setting the BCs!!!!
- // gridBuilder->addGrid( new Cuboid( 0.0, 0.0, 0.0, L_x, L_y, 0.3*L_z) , 1 );
- // para->setMaxLevel(2);
+ std::cout << "nProcs: "<< nProcs << "Proc: " << procID << " isFirstSubDomain: " << isFirstSubDomain << " isLastSubDomain: " << isLastSubDomain << " isMidSubDomain: " << isMidSubDomain << std::endl;
+
+ if(nProcs > 1){
+ if (isFirstSubDomain || isMidSubDomain) {
+ gridBuilder->findCommunicationIndices(CommunicationDirections::PX, lbmOrGks);
+ gridBuilder->setCommunicationProcess(CommunicationDirections::PX, procID+1);
+ }
- gridBuilder->setPeriodicBoundaryCondition(true, true, false);
+ if (isLastSubDomain || isMidSubDomain) {
+ gridBuilder->findCommunicationIndices(CommunicationDirections::MX, lbmOrGks);
+ gridBuilder->setCommunicationProcess(CommunicationDirections::MX, procID-1);
+ }
- gridBuilder->buildGrids(lbmOrGks, false); // buildGrids() has to be called before setting the BCs!!!!
+ if (isFirstSubDomain && !readPrecursor) {
+ gridBuilder->findCommunicationIndices(CommunicationDirections::MX, lbmOrGks);
+ gridBuilder->setCommunicationProcess(CommunicationDirections::MX, nProcs-1);
+ }
+ if (isLastSubDomain && !readPrecursor) {
+ gridBuilder->findCommunicationIndices(CommunicationDirections::PX, lbmOrGks);
+ gridBuilder->setCommunicationProcess(CommunicationDirections::PX, 0);
+ }
+ }
uint samplingOffset = 2;
- // gridBuilder->setVelocityBoundaryCondition(SideType::MZ, 0.0, 0.0, 0.0);
+
+ std::cout << " precursorDirectory " << precursorDirectory << std::endl;
+
+ if(readPrecursor)
+ {
+ if(isFirstSubDomain || nProcs == 1)
+ {
+ auto precursor = createFileCollection(precursorDirectory + "/precursor", FileType::VTK);
+ gridBuilder->setPrecursorBoundaryCondition(SideType::MX, precursor, timestepsBetweenReadsPrecursor);
+ // gridBuilder->setVelocityBoundaryCondition(SideType::MX, velocityLB, 0.0, 0.0);
+ }
+
+ if(isLastSubDomain || nProcs == 1)
+ {
+ gridBuilder->setPressureBoundaryCondition(SideType::PX, 0.f);
+ }
+ }
+
gridBuilder->setStressBoundaryCondition(SideType::MZ,
0.0, 0.0, 1.0, // wall normals
- samplingOffset, z0/dx); // wall model settinng
- para->setHasWallModelMonitor(true);
- bcFactory.setStressBoundaryCondition(BoundaryConditionFactory::StressBC::StressPressureBounceBack);
+ samplingOffset, z0, dx); // wall model settinng
+ para->setHasWallModelMonitor(true);
+ gridBuilder->setSlipBoundaryCondition(SideType::PZ, 0.0f, 0.0f, -1.0f);
- gridBuilder->setSlipBoundaryCondition(SideType::PZ, 0.0, 0.0, 0.0);
+ bcFactory.setVelocityBoundaryCondition(BoundaryConditionFactory::VelocityBC::VelocityCompressible);
+ bcFactory.setStressBoundaryCondition(BoundaryConditionFactory::StressBC::StressPressureBounceBack);
bcFactory.setSlipBoundaryCondition(BoundaryConditionFactory::SlipBC::SlipBounceBack);
-
+ bcFactory.setPressureBoundaryCondition(BoundaryConditionFactory::PressureBC::OutflowNonReflective);
+ bcFactory.setPrecursorBoundaryCondition(useDistributions ? BoundaryConditionFactory::PrecursorBC::DistributionsPrecursor : BoundaryConditionFactory::PrecursorBC::VelocityPrecursor);
+ para->setOutflowPressureCorrectionFactor(0.0);
- real cPi = 3.1415926535897932384626433832795;
- para->setInitialCondition([&](real coordX, real coordY, real coordZ, real &rho, real &vx, real &vy, real &vz) {
+ if(readPrecursor)
+ {
+ para->setInitialCondition([&](real coordX, real coordY, real coordZ, real &rho, real &vx, real &vy, real &vz) {
rho = (real)0.0;
- vx = (u_star/0.4 * log(coordZ/z0) + 2.0*sin(cPi*16.0f*coordX/L_x)*sin(cPi*8.0f*coordZ/H)/(pow(coordZ/H,c2o1)+c1o1)) * dt / dx;
- vy = 2.0*sin(cPi*16.0f*coordX/L_x)*sin(cPi*8.0f*coordZ/H)/(pow(coordZ/H,c2o1)+c1o1) * dt / dx;
- vz = 8.0*u_star/0.4*(sin(cPi*8.0*coordY/H)*sin(cPi*8.0*coordZ/H)+sin(cPi*8.0*coordX/L_x))/(pow(L_z/2.0-coordZ, c2o1)+c1o1) * dt / dx;
- });
- ////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
+ vx = rho = c0o1;
+ vx = u_star/c4o10*(u_star/c4o10 * log(coordZ/z0+c1o1)) * dt/dx;
+ vy = c0o1;
+ vz = c0o1;
+ });
+ }
+ else
+ {
+ para->setInitialCondition([&](real coordX, real coordY, real coordZ, real &rho, real &vx, real &vy, real &vz) {
+ rho = (real)0.0;
+ vx = rho = c0o1;
+ vx = (u_star/c4o10 * log(coordZ/z0+c1o1) + c2o1*sin(cPi*c16o1*coordX/L_x)*sin(cPi*c8o1*coordZ/H)/(pow(coordZ/H,c2o1)+c1o1)) * dt/dx;
+ vy = c2o1*sin(cPi*c16o1*coordX/L_x)*sin(cPi*c8o1*coordZ/H)/(pow(coordZ/H,c2o1)+c1o1) * dt/dx;
+ vz = c8o1*u_star/c4o10*(sin(cPi*c8o1*coordY/H)*sin(cPi*c8o1*coordZ/H)+sin(cPi*c8o1*coordX/L_x))/(pow(c1o2*L_z-coordZ, c2o1)+c1o1) * dt/dx;
+ });
+ }
+
+
- SPtr<PlanarAverageProbe> planarAverageProbe = SPtr<PlanarAverageProbe>( new PlanarAverageProbe("planeProbe", para->getOutputPath(), tStartAveraging/dt, tStartTmpAveraging/dt, tAveraging/dt , tStartOutProbe/dt, tOutProbe/dt, 'z') );
- planarAverageProbe->addAllAvailableStatistics();
- planarAverageProbe->setFileNameToNOut();
- para->addProbe( planarAverageProbe );
+ //////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
+ if(!readPrecursor && (isFirstSubDomain || nProcs == 1))
+ {
+ SPtr<PlanarAverageProbe> planarAverageProbe = SPtr<PlanarAverageProbe>( new PlanarAverageProbe("planeProbe", para->getOutputPath(), tStartAveraging/dt, tStartTmpAveraging/dt, tAveraging/dt , tStartOutProbe/dt, tOutProbe/dt, 'z') );
+ planarAverageProbe->addAllAvailableStatistics();
+ planarAverageProbe->setFileNameToNOut();
+ para->addProbe( planarAverageProbe );
+
+ para->setHasWallModelMonitor(true);
+ SPtr<WallModelProbe> wallModelProbe = SPtr<WallModelProbe>( new WallModelProbe("wallModelProbe", para->getOutputPath(), tStartAveraging/dt, tStartTmpAveraging/dt, tAveraging/dt/4.0 , tStartOutProbe/dt, tOutProbe/dt) );
+ wallModelProbe->addAllAvailableStatistics();
+ wallModelProbe->setFileNameToNOut();
+ wallModelProbe->setForceOutputToStress(true);
+ if(para->getIsBodyForce())
+ wallModelProbe->setEvaluatePressureGradient(true);
+ para->addProbe( wallModelProbe );
+ }
+
+ SPtr<PlaneProbe> planeProbe1 = SPtr<PlaneProbe>( new PlaneProbe("planeProbe_1", para->getOutputPath(), tStartAveraging/dt, 10, tStartOutProbe/dt, tOutProbe/dt) );
+ planeProbe1->setProbePlane(100.0, 0.0, 0, dx, L_y, L_z);
+ planeProbe1->addAllAvailableStatistics();
+ para->addProbe( planeProbe1 );
+
+ if(readPrecursor)
+ {
+ SPtr<PlaneProbe> planeProbe2 = SPtr<PlaneProbe>( new PlaneProbe("planeProbe_2", para->getOutputPath(), tStartAveraging/dt, 10, tStartOutProbe/dt, tOutProbe/dt) );
+ planeProbe2->setProbePlane(1000.0, 0.0, 0, dx, L_y, L_z);
+ planeProbe2->addAllAvailableStatistics();
+ para->addProbe( planeProbe2 );
+
+ SPtr<PlaneProbe> planeProbe3 = SPtr<PlaneProbe>( new PlaneProbe("planeProbe_3", para->getOutputPath(), tStartAveraging/dt, 10, tStartOutProbe/dt, tOutProbe/dt) );
+ planeProbe3->setProbePlane(1500.0, 0.0, 0, dx, L_y, L_z);
+ planeProbe3->addAllAvailableStatistics();
+ para->addProbe( planeProbe3 );
+
+ SPtr<PlaneProbe> planeProbe4 = SPtr<PlaneProbe>( new PlaneProbe("planeProbe_4", para->getOutputPath(), tStartAveraging/dt, 10, tStartOutProbe/dt, tOutProbe/dt) );
+ planeProbe4->setProbePlane(2000.0, 0.0, 0, dx, L_y, L_z);
+ planeProbe4->addAllAvailableStatistics();
+ para->addProbe( planeProbe4 );
+
+ SPtr<PlaneProbe> planeProbe5 = SPtr<PlaneProbe>( new PlaneProbe("planeProbe_5", para->getOutputPath(), tStartAveraging/dt, 10, tStartOutProbe/dt, tOutProbe/dt) );
+ planeProbe5->setProbePlane(2500.0, 0.0, 0, dx, L_y, L_z);
+ planeProbe5->addAllAvailableStatistics();
+ para->addProbe( planeProbe5 );
+
+ SPtr<PlaneProbe> planeProbe6 = SPtr<PlaneProbe>( new PlaneProbe("planeProbe_6", para->getOutputPath(), tStartAveraging/dt, 10, tStartOutProbe/dt, tOutProbe/dt) );
+ planeProbe6->setProbePlane(0.0, L_y/2.0, 0, L_x, dx, L_z);
+ planeProbe6->addAllAvailableStatistics();
+ para->addProbe( planeProbe6 );
+ }
- para->setHasWallModelMonitor(true);
- SPtr<WallModelProbe> wallModelProbe = SPtr<WallModelProbe>( new WallModelProbe("wallModelProbe", para->getOutputPath(), tStartAveraging/dt, tStartTmpAveraging/dt, tAveraging/dt/4.0 , tStartOutProbe/dt, tOutProbe/dt) );
- wallModelProbe->addAllAvailableStatistics();
- wallModelProbe->setFileNameToNOut();
- wallModelProbe->setForceOutputToStress(true);
- if(para->getIsBodyForce())
- wallModelProbe->setEvaluatePressureGradient(true);
- para->addProbe( wallModelProbe );
+ if(writePrecursor)
+ {
+ SPtr<PrecursorWriter> precursorWriter = std::make_shared<PrecursorWriter>("precursor", para->getOutputPath()+precursorDirectory, posXPrecursor, 0, L_y, 0, L_z, tStartPrecursor/dt, nTWritePrecursor, useDistributions? OutputVariable::Distributions: OutputVariable::Velocities, 1000);
+ para->addProbe(precursorWriter);
+ }
auto cudaMemoryManager = std::make_shared<CudaMemoryManager>(para);
auto gridGenerator = GridProvider::makeGridGenerator(gridBuilder, para, cudaMemoryManager, communicator);
- Simulation sim(para, cudaMemoryManager, communicator, *gridGenerator, &bcFactory, tmFactory);
+ Simulation sim(para, cudaMemoryManager, communicator, *gridGenerator, &bcFactory, tmFactory, &scalingFactory);
sim.run();
}
diff --git a/apps/gpu/LBM/BoundaryLayer/configBoundaryLayer.txt b/apps/gpu/LBM/BoundaryLayer/configBoundaryLayer.txt
index a489f0ab89738a193b16fee41c212a5943f6525d..83e7861a5fb85ea800d187699f1c6c1409422f0a 100644
--- a/apps/gpu/LBM/BoundaryLayer/configBoundaryLayer.txt
+++ b/apps/gpu/LBM/BoundaryLayer/configBoundaryLayer.txt
@@ -7,7 +7,7 @@ Path = .
##################################################
GridPath = .
##################################################
-Devices = 1
+Devices = 0
##################################################
tStartOut = 0
tOut = 100000
@@ -28,3 +28,15 @@ SGSconstant = 0.2
QuadricLimiterP = 100000.0
QuadricLimiterM = 100000.0
QuadricLimiterD = 100000.0
+
+##################################################
+readPrecursor = false
+nTimestepsReadPrecursor = 10
+precursorFile = precursor/Precursor
+
+##################################################
+writePrecursor = false
+nTimestepsWritePrecursor = 10
+
+tStartPrecursor = 100
+posXPrecursor = 3000
\ No newline at end of file
diff --git a/apps/gpu/LBM/DrivenCavity/DrivenCavity.cpp b/apps/gpu/LBM/DrivenCavity/DrivenCavity.cpp
index 69ecb3d8cbd45a8a7419437e934a57bd20b0bc9f..5e1cab7f48f7fb672c85f0decee4bcc2d4ac158f 100644
--- a/apps/gpu/LBM/DrivenCavity/DrivenCavity.cpp
+++ b/apps/gpu/LBM/DrivenCavity/DrivenCavity.cpp
@@ -85,7 +85,7 @@ int main()
const real L = 1.0;
const real Re = 1000.0;
const real velocity = 1.0;
- const real dt = (real)0.5e-3;
+ const real velocityLB = 0.05; // LB units
const uint nx = 64;
const uint timeStepOut = 1000;
@@ -109,10 +109,20 @@ int main()
auto gridBuilder = MultipleGridBuilder::makeShared(gridFactory);
//////////////////////////////////////////////////////////////////////////
- // create grid
+ // compute parameters in lattice units
//////////////////////////////////////////////////////////////////////////
- real dx = L / real(nx);
+ const real dx = L / real(nx);
+ const real dt = velocityLB / velocity * dx;
+
+ const real vxLB = velocityLB / sqrt(2.0); // LB units
+ const real vyLB = velocityLB / sqrt(2.0); // LB units
+
+ const real viscosityLB = nx * velocityLB / Re; // LB units
+
+ //////////////////////////////////////////////////////////////////////////
+ // create grid
+ //////////////////////////////////////////////////////////////////////////
gridBuilder->addCoarseGrid(-0.5 * L, -0.5 * L, -0.5 * L, 0.5 * L, 0.5 * L, 0.5 * L, dx);
@@ -124,17 +134,6 @@ int main()
gridBuilder->buildGrids(LbmOrGks::LBM, false);
- //////////////////////////////////////////////////////////////////////////
- // compute parameters in lattice units
- //////////////////////////////////////////////////////////////////////////
-
- const real velocityLB = velocity * dt / dx; // LB units
-
- const real vxLB = velocityLB / sqrt(2.0); // LB units
- const real vyLB = velocityLB / sqrt(2.0); // LB units
-
- const real viscosityLB = nx * velocityLB / Re; // LB units
-
//////////////////////////////////////////////////////////////////////////
// set parameters
//////////////////////////////////////////////////////////////////////////
@@ -154,7 +153,7 @@ int main()
para->setTimestepOut(timeStepOut);
para->setTimestepEnd(timeStepEnd);
- para->setMainKernel("CumulantK17CompChimRedesigned");
+ para->setMainKernel("CumulantK17");
//////////////////////////////////////////////////////////////////////////
// set boundary conditions
@@ -164,8 +163,8 @@ int main()
gridBuilder->setNoSlipBoundaryCondition(SideType::MX);
gridBuilder->setNoSlipBoundaryCondition(SideType::PY);
gridBuilder->setNoSlipBoundaryCondition(SideType::MY);
- gridBuilder->setVelocityBoundaryCondition(SideType::PZ, vxLB, vyLB, 0.0);
gridBuilder->setNoSlipBoundaryCondition(SideType::MZ);
+ gridBuilder->setVelocityBoundaryCondition(SideType::PZ, vxLB, vyLB, 0.0);
BoundaryConditionFactory bcFactory;
diff --git a/apps/gpu/LBM/DrivenCavityMultiGPU/DrivenCavityMultiGPU.cpp b/apps/gpu/LBM/DrivenCavityMultiGPU/DrivenCavityMultiGPU.cpp
index 8ca6939924fcfba22c8b96f000b9d8d05a3f7f43..ed6b4da7a3218e4d89ac90b053d9c054e4dd8205 100644
--- a/apps/gpu/LBM/DrivenCavityMultiGPU/DrivenCavityMultiGPU.cpp
+++ b/apps/gpu/LBM/DrivenCavityMultiGPU/DrivenCavityMultiGPU.cpp
@@ -50,6 +50,7 @@
#include "VirtualFluids_GPU/Kernel/Utilities/KernelFactory/KernelFactoryImp.h"
#include "VirtualFluids_GPU/PreProcessor/PreProcessorFactory/PreProcessorFactoryImp.h"
#include "VirtualFluids_GPU/Factories/BoundaryConditionFactory.h"
+#include "VirtualFluids_GPU/Factories/GridScalingFactory.h"
#include "VirtualFluids_GPU/GPU/CudaMemoryManager.h"
@@ -57,19 +58,6 @@
#include "utilities/communication.h"
-////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
-////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
-////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
-//
-// U s e r s e t t i n g s
-//
-////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
-////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
-////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
-
-const std::string outPath("output/DrivenCavity_Results/");
-const std::string gridPath = "output/DrivenCavity_Results/grid/";
-
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
@@ -92,15 +80,12 @@ void multipleLevel(std::filesystem::path& configPath)
config.load(configPath.string());
SPtr<Parameter> para = std::make_shared<Parameter>(communicator.getNummberOfProcess(), communicator.getPID(), &config);
BoundaryConditionFactory bcFactory = BoundaryConditionFactory();
-
+ GridScalingFactory scalingFactory = GridScalingFactory();
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
bool useGridGenerator = true;
bool useLevels = true;
- // para->setUseStreams(useStreams); // set in config
- // para->useReducedCommunicationAfterFtoC = true; // set in config
- para->setCalcTurbulenceIntensity(false);
if (para->getNumprocs() == 1) {
para->useReducedCommunicationAfterFtoC = false;
@@ -108,47 +93,40 @@ void multipleLevel(std::filesystem::path& configPath)
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
- const real L = 1.0;
- const real Re = 1000.0; // 1000
+ const std::string outPath("output/");
+ const std::string gridPath = "output/";
+ std::string simulationName("DrivenCavityMultiGPU");
+
+ const real L = 1.0;
+ const real Re = 1000.0;
const real velocity = 1.0;
- const real dt = (real)1.0e-3; // 0.5e-3;
- const uint nx = 64;
- std::string simulationName("DrivenCavityChimMultiGPU");
+ const real velocityLB = 0.05; // LB units
+ const uint nx = 64;
// para->setTimestepOut(10000); // set in config
// para->setTimestepEnd(10000); // set in config
const real dxGrid = L / real(nx);
- const real velocityLB = velocity * dt / dxGrid; // LB units
+ const real dt = velocityLB / velocity * dxGrid;
const real vxLB = velocityLB / (real)sqrt(2.0); // LB units
const real vyLB = velocityLB / (real)sqrt(2.0); // LB units
const real viscosityLB = nx * velocityLB / Re; // LB units
- para->setInitialCondition([&](real coordX, real coordY, real coordZ, real &rho, real &vx, real &vy, real &vz) {
- rho = (real)1.0;
- vx = (real)(coordX * velocityLB);
- vy = (real)(coordY * velocityLB);
- vz = (real)(coordZ * velocityLB);
- });
-
para->setVelocityLB(velocityLB);
para->setViscosityLB(viscosityLB);
para->setVelocityRatio(velocity / velocityLB);
- para->setDensityRatio((real)1.0); // correct value?
+ para->setDensityRatio((real)1.0);
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
- para->setCalcDragLift(false);
- para->setUseWale(false);
-
if (para->getOutputPath() == "output/") {para->setOutputPath(outPath);}
para->setOutputPrefix(simulationName);
para->setPrintFiles(true);
std::cout << "Write result files to " << para->getFName() << std::endl;
- // para->setMainKernel("CumulantK17CompChim");
- para->setMainKernel("CumulantK17CompChimStream");
+ para->setMainKernel("CumulantK17");
+ scalingFactory.setScalingFactory(GridScalingFactory::GridScaling::ScaleCompressible);
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
@@ -157,7 +135,7 @@ void multipleLevel(std::filesystem::path& configPath)
VF_LOG_INFO("velocity LB [dx/dt] = {}", vxLB);
VF_LOG_INFO("viscosity LB [dx/dt] = {}", viscosityLB);
VF_LOG_INFO("dxGrid [-] = {}\n", dxGrid);
-
+ VF_LOG_INFO("dt [s] = {}", dt);
VF_LOG_INFO("simulation parameters:");
VF_LOG_INFO("mainKernel = {}\n", para->getMainKernel());
@@ -226,7 +204,7 @@ void multipleLevel(std::filesystem::path& configPath)
if (generatePart == 0)
gridBuilder->setVelocityBoundaryCondition(SideType::MZ, 0.0, 0.0, 0.0);
if (generatePart == 1)
- gridBuilder->setVelocityBoundaryCondition(SideType::PZ, vxLB, 0.0, 0.0);
+ gridBuilder->setVelocityBoundaryCondition(SideType::PZ, vxLB, vyLB, 0.0);
gridBuilder->setVelocityBoundaryCondition(SideType::MX, 0.0, 0.0, 0.0);
gridBuilder->setVelocityBoundaryCondition(SideType::MY, 0.0, 0.0, 0.0);
gridBuilder->setVelocityBoundaryCondition(SideType::PX, 0.0, 0.0, 0.0);
@@ -303,13 +281,13 @@ void multipleLevel(std::filesystem::path& configPath)
}
if (generatePart == 2) {
gridBuilder->setVelocityBoundaryCondition(SideType::MX, 0.0, 0.0, 0.0);
- gridBuilder->setVelocityBoundaryCondition(SideType::PZ, vxLB, 0.0, 0.0);
+ gridBuilder->setVelocityBoundaryCondition(SideType::PZ, vxLB, vyLB, 0.0);
}
gridBuilder->setVelocityBoundaryCondition(SideType::MY, 0.0, 0.0, 0.0);
gridBuilder->setVelocityBoundaryCondition(SideType::PY, 0.0, 0.0, 0.0);
if (generatePart == 3) {
gridBuilder->setVelocityBoundaryCondition(SideType::PX, 0.0, 0.0, 0.0);
- gridBuilder->setVelocityBoundaryCondition(SideType::PZ, vxLB, 0.0, 0.0);
+ gridBuilder->setVelocityBoundaryCondition(SideType::PZ, vxLB, vyLB, 0.0);
}
if (generatePart == 1) {
gridBuilder->setVelocityBoundaryCondition(SideType::MZ, 0.0, 0.0, 0.0);
@@ -472,22 +450,22 @@ void multipleLevel(std::filesystem::path& configPath)
if (generatePart == 4) {
gridBuilder->setVelocityBoundaryCondition(SideType::MX, 0.0, 0.0, 0.0);
gridBuilder->setVelocityBoundaryCondition(SideType::MY, 0.0, 0.0, 0.0);
- gridBuilder->setVelocityBoundaryCondition(SideType::PZ, vxLB, 0.0, 0.0);
+ gridBuilder->setVelocityBoundaryCondition(SideType::PZ, vxLB, vyLB, 0.0);
}
if (generatePart == 5) {
gridBuilder->setVelocityBoundaryCondition(SideType::MX, 0.0, 0.0, 0.0);
gridBuilder->setVelocityBoundaryCondition(SideType::PY, 0.0, 0.0, 0.0);
- gridBuilder->setVelocityBoundaryCondition(SideType::PZ, vxLB, 0.0, 0.0);
+ gridBuilder->setVelocityBoundaryCondition(SideType::PZ, vxLB, vyLB, 0.0);
}
if (generatePart == 6) {
gridBuilder->setVelocityBoundaryCondition(SideType::MY, 0.0, 0.0, 0.0);
gridBuilder->setVelocityBoundaryCondition(SideType::PX, 0.0, 0.0, 0.0);
- gridBuilder->setVelocityBoundaryCondition(SideType::PZ, vxLB, 0.0, 0.0);
+ gridBuilder->setVelocityBoundaryCondition(SideType::PZ, vxLB, vyLB, 0.0);
}
if (generatePart == 7) {
gridBuilder->setVelocityBoundaryCondition(SideType::PY, 0.0, 0.0, 0.0);
gridBuilder->setVelocityBoundaryCondition(SideType::PX, 0.0, 0.0, 0.0);
- gridBuilder->setVelocityBoundaryCondition(SideType::PZ, vxLB, 0.0, 0.0);
+ gridBuilder->setVelocityBoundaryCondition(SideType::PZ, vxLB, vyLB, 0.0);
}
//////////////////////////////////////////////////////////////////////////
}
@@ -513,7 +491,7 @@ void multipleLevel(std::filesystem::path& configPath)
gridBuilder->setVelocityBoundaryCondition(SideType::MY, 0.0, 0.0, 0.0);
gridBuilder->setVelocityBoundaryCondition(SideType::PY, 0.0, 0.0, 0.0);
gridBuilder->setVelocityBoundaryCondition(SideType::MZ, 0.0, 0.0, 0.0);
- gridBuilder->setVelocityBoundaryCondition(SideType::PZ, vxLB, 0.0, 0.0);
+ gridBuilder->setVelocityBoundaryCondition(SideType::PZ, vxLB, vyLB, 0.0);
//////////////////////////////////////////////////////////////////////////
gridBuilder->writeGridsToVtk(outPath + "/grid/");
@@ -534,7 +512,7 @@ void multipleLevel(std::filesystem::path& configPath)
gridGenerator = GridProvider::makeGridReader(FILEFORMAT::BINARY, para, cudaMemoryManager);
}
- Simulation sim(para, cudaMemoryManager, communicator, *gridGenerator, &bcFactory);
+ Simulation sim(para, cudaMemoryManager, communicator, *gridGenerator, &bcFactory, &scalingFactory);
sim.run();
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
diff --git a/apps/gpu/LBM/DrivenCavityMultiGPU/configPhoenix4GPU.txt b/apps/gpu/LBM/DrivenCavityMultiGPU/configPhoenix4GPU.txt
index c710922b9fc82ac7680f5f7daade4faa235bc957..c5789cdf96049b7c0a31ce693c29cd2db4952a58 100644
--- a/apps/gpu/LBM/DrivenCavityMultiGPU/configPhoenix4GPU.txt
+++ b/apps/gpu/LBM/DrivenCavityMultiGPU/configPhoenix4GPU.txt
@@ -4,35 +4,11 @@
Devices="0 1 2 3"
NumberOfDevices=4
-##################################################
-#informations for Writing
-##################################################
-Path=/work/y0078217/Results/DrivenCavityMultiGPUResults/4GPU/
-#Prefix="DrivenCavityMultiGPU"
-#WriteGrid=true
-##################################################
-#informations for reading
-##################################################
-GridPath=/work/y0078217/Grids/GridDrivenCavityMultiGPU/4GPU/
-#GridPath="C:"
-
-##################################################
-#number of grid levels
-##################################################
-#NOGL=1
-
-##################################################
-#LBM Version
-##################################################
-#D3Qxx=27
-#MainKernelName=CumulantK17CompChim
-
##################################################
#simulation parameter
##################################################
-TimeEnd=1
-TimeOut=1
-#TimeStartOut=0
+TimeEnd=10000
+TimeOut=10000
##################################################
# CUDA Streams and optimized communication (only used for multiple GPUs)
diff --git a/apps/gpu/LBM/DrivenCavityUniform/CMakeLists.txt b/apps/gpu/LBM/DrivenCavityUniform/CMakeLists.txt
new file mode 100644
index 0000000000000000000000000000000000000000..40b4f08d7500c56efae7378df6398d065e4ecbfb
--- /dev/null
+++ b/apps/gpu/LBM/DrivenCavityUniform/CMakeLists.txt
@@ -0,0 +1,10 @@
+PROJECT(DrivenCavityUniform LANGUAGES CUDA CXX)
+
+#LIST(APPEND CS_COMPILER_FLAGS_CXX "-DOMPI_SKIP_MPICXX" )
+
+vf_add_library(BUILDTYPE binary PRIVATE_LINK basics VirtualFluids_GPU GridGenerator MPI::MPI_CXX FILES DrivenCavity.cpp)
+
+set_source_files_properties(DrivenCavity.cpp PROPERTIES LANGUAGE CUDA)
+
+set_target_properties(DrivenCavityUniform PROPERTIES CUDA_SEPARABLE_COMPILATION ON)
+
diff --git a/apps/gpu/LBM/DrivenCavityUniform/DrivenCavity.cpp b/apps/gpu/LBM/DrivenCavityUniform/DrivenCavity.cpp
new file mode 100644
index 0000000000000000000000000000000000000000..958ef4714118aac34b8cfb0bec3aab97b108b01d
--- /dev/null
+++ b/apps/gpu/LBM/DrivenCavityUniform/DrivenCavity.cpp
@@ -0,0 +1,231 @@
+//=======================================================================================
+// ____ ____ __ ______ __________ __ __ __ __
+// \ \ | | | | | _ \ |___ ___| | | | | / \ | |
+// \ \ | | | | | |_) | | | | | | | / \ | |
+// \ \ | | | | | _ / | | | | | | / /\ \ | |
+// \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
+// \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
+// \ \ | | ________________________________________________________________
+// \ \ | | | ______________________________________________________________|
+// \ \| | | | __ __ __ __ ______ _______
+// \ | | |_____ | | | | | | | | | _ \ / _____)
+// \ | | _____| | | | | | | | | | | \ \ \_______
+// \ | | | | |_____ | \_/ | | | | |_/ / _____ |
+// \ _____| |__| |________| \_______/ |__| |______/ (_______/
+//
+// This file is part of VirtualFluids. VirtualFluids is free software: you can
+// redistribute it and/or modify it under the terms of the GNU General Public
+// License as published by the Free Software Foundation, either version 3 of
+// the License, or (at your option) any later version.
+//
+// VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
+// ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+// FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
+// for more details.
+//
+// You should have received a copy of the GNU General Public License along
+// with VirtualFluids (see COPYING.txt). If not, see <http://www.gnu.org/licenses/>.
+//
+//! \file LidDrivenCavity.cpp
+//! \ingroup Applications
+//! \author Martin Schoenherr, Stephan Lenz
+//=======================================================================================
+#define _USE_MATH_DEFINES
+#include <exception>
+#include <fstream>
+#include <iostream>
+#include <memory>
+#include <sstream>
+#include <stdexcept>
+#include <string>
+
+//////////////////////////////////////////////////////////////////////////
+
+#include "Core/DataTypes.h"
+#include "Core/LbmOrGks.h"
+#include "Core/Logger/Logger.h"
+#include "Core/VectorTypes.h"
+#include "PointerDefinitions.h"
+
+#include <logger/Logger.h>
+
+//////////////////////////////////////////////////////////////////////////
+
+#include "GridGenerator/grid/BoundaryConditions/Side.h"
+#include "GridGenerator/grid/GridBuilder/LevelGridBuilder.h"
+#include "GridGenerator/grid/GridBuilder/MultipleGridBuilder.h"
+#include "GridGenerator/grid/GridFactory.h"
+#include "GridGenerator/geometries/Cuboid/Cuboid.h"
+
+//////////////////////////////////////////////////////////////////////////
+
+#include "VirtualFluids_GPU/Factories/BoundaryConditionFactory.h"
+#include "VirtualFluids_GPU/Factories/GridScalingFactory.h"
+#include "VirtualFluids_GPU/Communication/Communicator.h"
+#include "VirtualFluids_GPU/DataStructureInitializer/GridProvider.h"
+#include "VirtualFluids_GPU/DataStructureInitializer/GridReaderGenerator/GridGenerator.h"
+#include "VirtualFluids_GPU/GPU/CudaMemoryManager.h"
+#include "VirtualFluids_GPU/LBM/Simulation.h"
+#include "VirtualFluids_GPU/Output/FileWriter.h"
+#include "VirtualFluids_GPU/Parameter/Parameter.h"
+#include "VirtualFluids_GPU/Factories/GridScalingFactory.h"
+
+//////////////////////////////////////////////////////////////////////////
+
+int main()
+{
+ try {
+ vf::logging::Logger::initalizeLogger();
+ //////////////////////////////////////////////////////////////////////////
+ // Simulation parameters
+ //////////////////////////////////////////////////////////////////////////
+ std::string path("./output/DrivenCavity_uniform");
+ std::string simulationName("LidDrivenCavity");
+
+ const real L = 1.0;
+ const real Re = 1000.0;
+ const real velocity = 1.0;
+ const real dt = (real)0.5e-3;
+ const uint nx = 64;
+
+ const uint timeStepOut = 1000;
+ const uint timeStepEnd = 10000;
+
+ //////////////////////////////////////////////////////////////////////////
+ // setup logger
+ //////////////////////////////////////////////////////////////////////////
+
+ logging::Logger::addStream(&std::cout);
+ logging::Logger::setDebugLevel(logging::Logger::Level::INFO_LOW);
+ logging::Logger::timeStamp(logging::Logger::ENABLE);
+ logging::Logger::enablePrintedRankNumbers(logging::Logger::ENABLE);
+
+ //////////////////////////////////////////////////////////////////////////
+ // setup gridGenerator
+ //////////////////////////////////////////////////////////////////////////
+
+ auto gridFactory = GridFactory::make();
+ gridFactory->setTriangularMeshDiscretizationMethod(TriangularMeshDiscretizationMethod::POINT_IN_OBJECT);
+ auto gridBuilder = MultipleGridBuilder::makeShared(gridFactory);
+
+ //////////////////////////////////////////////////////////////////////////
+ // create grid
+ //////////////////////////////////////////////////////////////////////////
+
+ real dx = L / real(nx);
+
+ gridBuilder->addCoarseGrid(-0.5 * L, -0.5 * L, -0.5 * L, 0.5 * L, 0.5 * L, 0.5 * L, dx);
+
+ // gridBuilder->addGrid(new Cuboid(-0.25, -0.25, -0.25, 0.25, 0.25, 0.25), 1); // add fine grid
+ GridScalingFactory scalingFactory = GridScalingFactory();
+ scalingFactory.setScalingFactory(GridScalingFactory::GridScaling::ScaleCompressible);
+
+ gridBuilder->setPeriodicBoundaryCondition(false, false, false);
+
+ gridBuilder->buildGrids(LbmOrGks::LBM, false);
+
+ //////////////////////////////////////////////////////////////////////////
+ // compute parameters in lattice units
+ //////////////////////////////////////////////////////////////////////////
+
+ const real velocityLB = velocity * dt / dx; // LB units
+
+ const real vxLB = velocityLB / sqrt(2.0); // LB units
+ const real vyLB = velocityLB / sqrt(2.0); // LB units
+
+ const real viscosityLB = nx * velocityLB / Re; // LB units
+
+ //////////////////////////////////////////////////////////////////////////
+ // set parameters
+ //////////////////////////////////////////////////////////////////////////
+ SPtr<Parameter> para = std::make_shared<Parameter>();
+
+ para->setOutputPath(path);
+ para->setOutputPrefix(simulationName);
+
+ para->setPrintFiles(true);
+
+ para->setVelocityLB(velocityLB);
+ para->setViscosityLB(viscosityLB);
+
+ para->setVelocityRatio(velocity / velocityLB);
+ para->setDensityRatio(1.0);
+
+ para->setTimestepOut(timeStepOut);
+ para->setTimestepEnd(timeStepEnd);
+
+ para->setMainKernel("CumulantK17");
+
+ //////////////////////////////////////////////////////////////////////////
+ // set boundary conditions
+ //////////////////////////////////////////////////////////////////////////
+
+ gridBuilder->setNoSlipBoundaryCondition(SideType::PX);
+ gridBuilder->setNoSlipBoundaryCondition(SideType::MX);
+ gridBuilder->setNoSlipBoundaryCondition(SideType::PY);
+ gridBuilder->setNoSlipBoundaryCondition(SideType::MY);
+ gridBuilder->setVelocityBoundaryCondition(SideType::PZ, vxLB, vyLB, 0.0);
+ gridBuilder->setNoSlipBoundaryCondition(SideType::MZ);
+
+ BoundaryConditionFactory bcFactory;
+
+ bcFactory.setNoSlipBoundaryCondition(BoundaryConditionFactory::NoSlipBC::NoSlipBounceBack);
+ bcFactory.setVelocityBoundaryCondition(BoundaryConditionFactory::VelocityBC::VelocitySimpleBounceBackCompressible);
+
+ //////////////////////////////////////////////////////////////////////////
+ // set copy mesh to simulation
+ //////////////////////////////////////////////////////////////////////////
+
+ vf::gpu::Communicator &communicator = vf::gpu::Communicator::getInstance();
+
+ auto cudaMemoryManager = std::make_shared<CudaMemoryManager>(para);
+ SPtr<GridProvider> gridGenerator =
+ GridProvider::makeGridGenerator(gridBuilder, para, cudaMemoryManager, communicator);
+
+
+ //////////////////////////////////////////////////////////////////////////
+ // run simulation
+ //////////////////////////////////////////////////////////////////////////
+
+ VF_LOG_INFO("Start Running DrivenCavity Showcase...");
+ printf("\n");
+ VF_LOG_INFO("world parameter:");
+ VF_LOG_INFO("--------------");
+ VF_LOG_INFO("dt [s] = {}", dt);
+ VF_LOG_INFO("world_length [m] = {}", L);
+ VF_LOG_INFO("world_velocity [m/s] = {}", velocity);
+ VF_LOG_INFO("dx [m] = {}", dx);
+ printf("\n");
+ VF_LOG_INFO("LB parameter:");
+ VF_LOG_INFO("--------------");
+ VF_LOG_INFO("Re = {}", Re);
+ VF_LOG_INFO("lb_velocity [dx/dt] = {}", velocityLB);
+ VF_LOG_INFO("lb_viscosity [dx^2/dt] = {}", viscosityLB);
+ VF_LOG_INFO("lb_vx [dx/dt] (lb_velocity/sqrt(2)) = {}", vxLB);
+ VF_LOG_INFO("lb_vy [dx/dt] (lb_velocity/sqrt(2)) = {}", vyLB);
+ printf("\n");
+ VF_LOG_INFO("simulation parameter:");
+ VF_LOG_INFO("--------------");
+ VF_LOG_INFO("nx = {}", nx);
+ VF_LOG_INFO("ny = {}", nx);
+ VF_LOG_INFO("nz = {}", nx);
+ VF_LOG_INFO("number of nodes = {}", nx * nx * nx);
+ VF_LOG_INFO("n timesteps = {}", timeStepOut);
+ VF_LOG_INFO("write_nth_timestep = {}", timeStepEnd);
+ VF_LOG_INFO("output_path = {}", path);
+
+ Simulation sim(para, cudaMemoryManager, communicator, *gridGenerator, &bcFactory, &scalingFactory);
+ sim.run();
+
+ } catch (const spdlog::spdlog_ex &ex) {
+ std::cout << "Log initialization failed: " << ex.what() << std::endl;
+ } catch (const std::bad_alloc &e) {
+ VF_LOG_CRITICAL("Bad Alloc: {}", e.what());
+ } catch (const std::exception &e) {
+ VF_LOG_CRITICAL("exception: {}", e.what());
+ } catch (...) {
+ VF_LOG_CRITICAL("Unknown exception!");
+ }
+
+ return 0;
+}
diff --git a/apps/gpu/LBM/DrivenCavityUniform/configDrivenCavity.txt b/apps/gpu/LBM/DrivenCavityUniform/configDrivenCavity.txt
new file mode 100644
index 0000000000000000000000000000000000000000..458346a67c7f001580494af1dc9262034613be68
--- /dev/null
+++ b/apps/gpu/LBM/DrivenCavityUniform/configDrivenCavity.txt
@@ -0,0 +1,34 @@
+##################################################
+#GPU Mapping
+##################################################
+#Devices="0 1 2 3"
+#NumberOfDevices=4
+
+##################################################
+#informations for Writing
+##################################################
+#Path = "output/"
+#Prefix="DrivenCavity"
+#WriteGrid=true
+##################################################
+#informations for reading
+##################################################
+#GridPath="grid/"
+
+##################################################
+#number of grid levels
+##################################################
+#NOGL=1
+
+##################################################
+#LBM Version
+##################################################
+#D3Qxx=27
+#MainKernelName=CumulantAA2016CompSP27
+
+##################################################
+#simulation parameter
+##################################################
+#TimeEnd=100000
+#TimeOut=1000
+#TimeStartOut=0
\ No newline at end of file
diff --git a/apps/gpu/LBM/MusselOyster/MusselOyster.cpp b/apps/gpu/LBM/MusselOyster/MusselOyster.cpp
index efac863fc9efd446e5f266648ad4fa74c954634f..dc5eaf58aff9b4a1b87d70c187b81461330ee3da 100644
--- a/apps/gpu/LBM/MusselOyster/MusselOyster.cpp
+++ b/apps/gpu/LBM/MusselOyster/MusselOyster.cpp
@@ -40,7 +40,6 @@
//////////////////////////////////////////////////////////////////////////
-#include "VirtualFluids_GPU/BoundaryConditions/BoundaryConditionFactory.h"
#include "VirtualFluids_GPU/Communication/Communicator.h"
#include "VirtualFluids_GPU/DataStructureInitializer/GridProvider.h"
#include "VirtualFluids_GPU/DataStructureInitializer/GridReaderFiles/GridReader.h"
@@ -49,7 +48,7 @@
#include "VirtualFluids_GPU/LBM/Simulation.h"
#include "VirtualFluids_GPU/Output/FileWriter.h"
#include "VirtualFluids_GPU/Parameter/Parameter.h"
-#include "VirtualFluids_GPU/BoundaryConditions/BoundaryConditionFactory.h"
+#include "VirtualFluids_GPU/Factories/BoundaryConditionFactory.h"
//////////////////////////////////////////////////////////////////////////
diff --git a/apps/gpu/LBM/MusselOyster/configPhoenix8GPU.txt b/apps/gpu/LBM/MusselOyster/configPhoenix8GPU.txt
index 4e2b0c91482b6a650ff28a210673cac097cb8c2d..2bf6955062da5c98f6a7b931c19821c52eaf15ea 100644
--- a/apps/gpu/LBM/MusselOyster/configPhoenix8GPU.txt
+++ b/apps/gpu/LBM/MusselOyster/configPhoenix8GPU.txt
@@ -7,14 +7,14 @@ NumberOfDevices=4
##################################################
#informations for Writing
##################################################
-Path=/work/y0078217/Results/MusselOysterResults/8GPUOyster05/
+#Path=/work/y0078217/Results/MusselOysterResults/8GPUOyster05/
#Path="F:/Work/Computations/out/MusselOyster/"
#Prefix="MusselOyster"
#WriteGrid=true
##################################################
#informations for reading
##################################################
-GridPath=/work/y0078217/Grids/GridMusselOyster/Oyster8GPU/
+#GridPath=/work/y0078217/Grids/GridMusselOyster/Oyster8GPU/
#GridPath="C:"
##################################################
@@ -31,8 +31,8 @@ GridPath=/work/y0078217/Grids/GridMusselOyster/Oyster8GPU/
##################################################
#simulation parameter
##################################################
-TimeEnd=400000 # 800000
-TimeOut=100000 # 400000
+TimeEnd=100000 # 800000
+TimeOut=10000 # 400000
#TimeStartOut=0
##################################################
diff --git a/apps/gpu/LBM/TGV_3D/TGV_3D.cpp b/apps/gpu/LBM/TGV_3D/TGV_3D.cpp
index d8642c7b267bcad6c58ab2a9c178c2d9394ecf2a..7514c2b273bf60d6e2523f132911dde8839d296a 100644
--- a/apps/gpu/LBM/TGV_3D/TGV_3D.cpp
+++ b/apps/gpu/LBM/TGV_3D/TGV_3D.cpp
@@ -1,63 +1,95 @@
-//#define MPI_LOGGING
+//=======================================================================================
+// ____ ____ __ ______ __________ __ __ __ __
+// \ \ | | | | | _ \ |___ ___| | | | | / \ | |
+// \ \ | | | | | |_) | | | | | | | / \ | |
+// \ \ | | | | | _ / | | | | | | / /\ \ | |
+// \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
+// \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
+// \ \ | | ________________________________________________________________
+// \ \ | | | ______________________________________________________________|
+// \ \| | | | __ __ __ __ ______ _______
+// \ | | |_____ | | | | | | | | | _ \ / _____)
+// \ | | _____| | | | | | | | | | | \ \ \_______
+// \ | | | | |_____ | \_/ | | | | |_/ / _____ |
+// \ _____| |__| |________| \_______/ |__| |______/ (_______/
+//
+// This file is part of VirtualFluids. VirtualFluids is free software: you can
+// redistribute it and/or modify it under the terms of the GNU General Public
+// License as published by the Free Software Foundation, either version 3 of
+// the License, or (at your option) any later version.
+//
+// VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
+// ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+// FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
+// for more details.
+//
+// You should have received a copy of the GNU General Public License along
+// with VirtualFluids (see COPYING.txt). If not, see <http://www.gnu.org/licenses/>.
+//
+//! \file TGV_3D.cpp
+//! \ingroup Applications
+//! \author Martin Schoenherr
+//=======================================================================================
+#define _USE_MATH_DEFINES
+#include <exception>
+#include <filesystem>
+#include <fstream>
+#include <iostream>
+#include <math.h>
+#include <memory>
+#include <sstream>
+#include <stdexcept>
+#include <string>
-//Martin Branch
+#include "mpi.h"
-#include <mpi.h>
-#if defined( MPI_LOGGING )
- #include <mpe.h>
-#endif
+//////////////////////////////////////////////////////////////////////////
-#include <string>
-#include <sstream>
-#include <iostream>
-#include <stdexcept>
-#include <fstream>
-#define _USE_MATH_DEFINES
-#include <math.h>
+#include "Core/DataTypes.h"
+#include "Core/LbmOrGks.h"
+#include "Core/Logger/Logger.h"
+#include "Core/VectorTypes.h"
+#include "PointerDefinitions.h"
-//#include "metis.h"
+//////////////////////////////////////////////////////////////////////////
-#include "basics/Core/LbmOrGks.h"
-#include "basics/Core/StringUtilities/StringUtil.h"
-#include <basics/config/ConfigurationFile.h>
+#include "GridGenerator/geometries/Conglomerate/Conglomerate.h"
+#include "GridGenerator/geometries/TriangularMesh/TriangularMesh.h"
+#include "GridGenerator/grid/BoundaryConditions/BoundaryCondition.h"
+#include "GridGenerator/grid/BoundaryConditions/Side.h"
+#include "GridGenerator/grid/GridBuilder/LevelGridBuilder.h"
+#include "GridGenerator/grid/GridBuilder/MultipleGridBuilder.h"
+#include "GridGenerator/grid/GridFactory.h"
+
+#include "GridGenerator/io/GridVTKWriter/GridVTKWriter.h"
+#include "GridGenerator/io/STLReaderWriter/STLReader.h"
+#include "GridGenerator/io/STLReaderWriter/STLWriter.h"
+#include "GridGenerator/io/SimulationFileWriter/SimulationFileWriter.h"
+
+//////////////////////////////////////////////////////////////////////////
-#include "VirtualFluids_GPU/LBM/Simulation.h"
#include "VirtualFluids_GPU/Communication/Communicator.h"
-#include "VirtualFluids_GPU/DataStructureInitializer/GridReaderGenerator/GridGenerator.h"
#include "VirtualFluids_GPU/DataStructureInitializer/GridProvider.h"
#include "VirtualFluids_GPU/DataStructureInitializer/GridReaderFiles/GridReader.h"
-#include "VirtualFluids_GPU/Parameter/Parameter.h"
-#include "VirtualFluids_GPU/Output/FileWriter.h"
-
-#include "VirtualFluids_GPU/Kernel/Utilities/KernelFactory/KernelFactoryImp.h"
-#include "VirtualFluids_GPU/PreProcessor/PreProcessorFactory/PreProcessorFactoryImp.h"
+#include "VirtualFluids_GPU/DataStructureInitializer/GridReaderGenerator/GridGenerator.h"
#include "VirtualFluids_GPU/Factories/BoundaryConditionFactory.h"
-
#include "VirtualFluids_GPU/GPU/CudaMemoryManager.h"
+#include "VirtualFluids_GPU/LBM/Simulation.h"
+#include "VirtualFluids_GPU/Output/FileWriter.h"
+#include "VirtualFluids_GPU/Parameter/Parameter.h"
-#include "global.h"
-
-#include "geometries/Sphere/Sphere.h"
-#include "geometries/VerticalCylinder/VerticalCylinder.h"
-#include "geometries/Cuboid/Cuboid.h"
-#include "geometries/TriangularMesh/TriangularMesh.h"
-#include "geometries/Conglomerate/Conglomerate.h"
-#include "geometries/TriangularMesh/TriangularMeshStrategy.h"
-
-#include "grid/GridBuilder/LevelGridBuilder.h"
-#include "grid/GridBuilder/MultipleGridBuilder.h"
-#include "grid/BoundaryConditions/Side.h"
-#include "grid/BoundaryConditions/BoundaryCondition.h"
-#include "grid/GridFactory.h"
+#include <logger/Logger.h>
-#include "io/SimulationFileWriter/SimulationFileWriter.h"
-#include "io/GridVTKWriter/GridVTKWriter.h"
-#include "io/STLReaderWriter/STLReader.h"
-#include "io/STLReaderWriter/STLWriter.h"
+////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
+////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
+////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
+//
+// U s e r s e t t i n g s
+//
+////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
+////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
+////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
-#include "utilities/math/Math.h"
-#include "utilities/communication.h"
-#include "utilities/transformator/TransformatorImp.h"
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
// from https://stackoverflow.com/questions/865668/how-to-parse-command-line-arguments-in-c
@@ -94,8 +126,8 @@ bool useWale = false;
std::string kernel( "CumulantK17Comp" );
-std::string path("F:/Work/Computations/out/TaylorGreen3DNew/"); //LEGOLAS
-//std::string path("E:/DrivenCavity/results/"); //TESLA03
+//std::string path("F:/Work/Computations/out/TaylorGreen3DNew/"); //LEGOLAS
+std::string path("D:/out/TGV_3D/"); //TESLA03
std::string simulationName("TGV_3D");
//////////////////////////////////////////////////////////////////////////
diff --git a/apps/gpu/LBM/TGV_3D_GridRef/CMakeLists.txt b/apps/gpu/LBM/TGV_3D_GridRef/CMakeLists.txt
new file mode 100644
index 0000000000000000000000000000000000000000..efb4310669f9c0de7aa5cf3f1e4dffa00bd66cbf
--- /dev/null
+++ b/apps/gpu/LBM/TGV_3D_GridRef/CMakeLists.txt
@@ -0,0 +1,7 @@
+PROJECT(TGV_3D_GridRef LANGUAGES CUDA CXX)
+
+vf_add_library(BUILDTYPE binary PRIVATE_LINK basics VirtualFluids_GPU GridGenerator FILES TGV_3D_GridRef.cpp)
+
+set_source_files_properties(TGV_3D_GridRef.cpp PROPERTIES LANGUAGE CUDA)
+
+set_target_properties(TGV_3D_GridRef PROPERTIES CUDA_SEPARABLE_COMPILATION ON)
\ No newline at end of file
diff --git a/apps/gpu/LBM/TGV_3D_GridRef/TGV_3D_GridRef.cpp b/apps/gpu/LBM/TGV_3D_GridRef/TGV_3D_GridRef.cpp
new file mode 100644
index 0000000000000000000000000000000000000000..a88fee2e583a7cb227702ff19ada7daced1b1708
--- /dev/null
+++ b/apps/gpu/LBM/TGV_3D_GridRef/TGV_3D_GridRef.cpp
@@ -0,0 +1,399 @@
+//=======================================================================================
+// ____ ____ __ ______ __________ __ __ __ __
+// \ \ | | | | | _ \ |___ ___| | | | | / \ | |
+// \ \ | | | | | |_) | | | | | | | / \ | |
+// \ \ | | | | | _ / | | | | | | / /\ \ | |
+// \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
+// \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
+// \ \ | | ________________________________________________________________
+// \ \ | | | ______________________________________________________________|
+// \ \| | | | __ __ __ __ ______ _______
+// \ | | |_____ | | | | | | | | | _ \ / _____)
+// \ | | _____| | | | | | | | | | | \ \ \_______
+// \ | | | | |_____ | \_/ | | | | |_/ / _____ |
+// \ _____| |__| |________| \_______/ |__| |______/ (_______/
+//
+// This file is part of VirtualFluids. VirtualFluids is free software: you can
+// redistribute it and/or modify it under the terms of the GNU General Public
+// License as published by the Free Software Foundation, either version 3 of
+// the License, or (at your option) any later version.
+//
+// VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
+// ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+// FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
+// for more details.
+//
+// You should have received a copy of the GNU General Public License along
+// with VirtualFluids (see COPYING.txt). If not, see <http://www.gnu.org/licenses/>.
+//
+//! \file TGV_3D.cpp
+//! \ingroup Applications
+//! \author Martin Schoenherr
+//=======================================================================================
+#define _USE_MATH_DEFINES
+#include <exception>
+#include <filesystem>
+#include <fstream>
+#include <iostream>
+#include <math.h>
+#include <memory>
+#include <sstream>
+#include <stdexcept>
+#include <string>
+
+#include "mpi.h"
+
+//////////////////////////////////////////////////////////////////////////
+
+#include "Core/DataTypes.h"
+#include "Core/LbmOrGks.h"
+#include "Core/Logger/Logger.h"
+#include "Core/VectorTypes.h"
+#include "PointerDefinitions.h"
+
+//////////////////////////////////////////////////////////////////////////
+
+#include "GridGenerator/geometries/Conglomerate/Conglomerate.h"
+#include "GridGenerator/geometries/TriangularMesh/TriangularMesh.h"
+#include "GridGenerator/grid/BoundaryConditions/BoundaryCondition.h"
+#include "GridGenerator/grid/BoundaryConditions/Side.h"
+#include "GridGenerator/grid/GridBuilder/LevelGridBuilder.h"
+#include "GridGenerator/grid/GridBuilder/MultipleGridBuilder.h"
+#include "GridGenerator/grid/GridFactory.h"
+
+#include "GridGenerator/io/GridVTKWriter/GridVTKWriter.h"
+#include "GridGenerator/io/STLReaderWriter/STLReader.h"
+#include "GridGenerator/io/STLReaderWriter/STLWriter.h"
+#include "GridGenerator/io/SimulationFileWriter/SimulationFileWriter.h"
+
+//////////////////////////////////////////////////////////////////////////
+
+#include "VirtualFluids_GPU/Communication/Communicator.h"
+#include "VirtualFluids_GPU/DataStructureInitializer/GridProvider.h"
+#include "VirtualFluids_GPU/DataStructureInitializer/GridReaderFiles/GridReader.h"
+#include "VirtualFluids_GPU/DataStructureInitializer/GridReaderGenerator/GridGenerator.h"
+#include "VirtualFluids_GPU/Factories/BoundaryConditionFactory.h"
+#include "VirtualFluids_GPU/Factories/GridScalingFactory.h"
+#include "VirtualFluids_GPU/GPU/CudaMemoryManager.h"
+#include "VirtualFluids_GPU/LBM/Simulation.h"
+#include "VirtualFluids_GPU/Output/FileWriter.h"
+#include "VirtualFluids_GPU/Parameter/Parameter.h"
+
+#include <logger/Logger.h>
+
+////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
+////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
+////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
+//
+// U s e r s e t t i n g s
+//
+////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
+////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
+////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
+
+
+////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
+// from https://stackoverflow.com/questions/865668/how-to-parse-command-line-arguments-in-c
+////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
+
+
+char* getCmdOption(char ** begin, char ** end, const std::string & option)
+{
+ char ** itr = std::find(begin, end, option);
+ if (itr != end && ++itr != end)
+ {
+ return *itr;
+ }
+ return 0;
+}
+
+bool cmdOptionExists(char** begin, char** end, const std::string& option)
+{
+ return std::find(begin, end, option) != end;
+}
+
+////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
+
+//////////////////////////////////////////////////////////////////////////
+real Re = 1600.0;
+
+uint dtPerL = 500;
+
+uint nx = 64;
+uint gpuIndex = 0;
+
+bool useLimiter = false;
+bool useWale = false;
+
+std::string kernel( "CumulantK17CompChimRedesigned" );
+
+std::string path("D:/out/TGV_3D/"); //MOLLOK
+
+std::string simulationName("TGV_3D_Gridref_noSqPress");
+//////////////////////////////////////////////////////////////////////////
+
+void multipleLevel(const std::string& configPath)
+{
+ logging::Logger::addStream(&std::cout);
+ logging::Logger::setDebugLevel(logging::Logger::Level::INFO_LOW);
+ logging::Logger::timeStamp(logging::Logger::ENABLE);
+ logging::Logger::enablePrintedRankNumbers(logging::Logger::ENABLE);
+
+ vf::gpu::Communicator& communicator = vf::gpu::Communicator::getInstance();
+
+ auto gridFactory = GridFactory::make();
+ //gridFactory->setTriangularMeshDiscretizationMethod(TriangularMeshDiscretizationMethod::RAYCASTING);
+ gridFactory->setTriangularMeshDiscretizationMethod(TriangularMeshDiscretizationMethod::POINT_IN_OBJECT);
+ //gridFactory->setTriangularMeshDiscretizationMethod(TriangularMeshDiscretizationMethod::POINT_UNDER_TRIANGLE);
+
+ auto gridBuilder = MultipleGridBuilder::makeShared(gridFactory);
+
+ vf::basics::ConfigurationFile config;
+ config.load(configPath);
+ SPtr<Parameter> para = std::make_shared<Parameter>(communicator.getNummberOfProcess(), communicator.getPID(), &config);
+ BoundaryConditionFactory bcFactory = BoundaryConditionFactory();
+ GridScalingFactory scalingFactory = GridScalingFactory();
+
+ ////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
+ ////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
+ ////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
+
+ const real PI = 3.141592653589793238462643383279;
+
+ real L = nx / ( 2.0 * PI );
+
+ const real velocity = 64.0 / ( dtPerL * 2.0 * PI );
+
+ const real viscosity = nx / ( 2.0 * PI ) * velocity / Re;
+
+ *logging::out << logging::Logger::INFO_HIGH << "velocity = " << velocity << " s\n";
+
+ *logging::out << logging::Logger::INFO_HIGH << "viscosity = " << viscosity << "\n";
+
+ ////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
+
+ real dx = 2.0 * PI / real(nx);
+
+ gridBuilder->addCoarseGrid(-PI, -PI, -PI,
+ PI, PI, PI, dx);
+
+ gridBuilder->setNumberOfLayers(0, 0);
+
+ auto fineGrid = new Cuboid(-PI * 0.5, -PI * 0.5, -PI * 0.5,
+ 0.0, PI * 0.5, 0.0);
+
+ gridBuilder->addGrid(fineGrid, 1);
+
+ gridBuilder->setPeriodicBoundaryCondition(true, true, true);
+
+ gridBuilder->buildGrids(LBM, true); // buildGrids() has to be called before setting the BCs!!!!
+
+ ////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
+ ////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
+ ////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
+
+ scalingFactory.setScalingFactory(GridScalingFactory::GridScaling::ScaleCompressible);
+
+ //std::stringstream _path;
+ // std::stringstream _prefix;
+
+ // //_path << "F:/Work/Computations/TaylorGreenVortex_3D/TGV_LBM/" << nx << "_Re_1.6e4";
+ // //_path << "F:/Work/Computations/TaylorGreenVortex_3D/TGV_LBM/" << nx << "_neqInit";
+ // _path << "F:/Work/Computations/TaylorGreenVortex_3D/TGV_LBM/Re_1600/AA2016/" << nx << "_FD_O8";
+
+ // //_path << "./results/AA2016/" << nx;
+ // //_path << "./results/CumOne/" << nx;
+ // //_path << "./results/F3_2018/" << nx;
+
+ // _prefix << "TGV_3D_" << nx << "_" ;
+
+ // para->setOutputPath(_path.str());
+ // para->setOutputPrefix(_prefix.str());
+ // para->setPathAndFilename(_path.str() + "/" + _prefix.str());
+
+ //////////////////////////////////////////////////////////////////////////
+
+ {
+ std::stringstream _path;
+
+ _path << path;
+ _path << kernel;
+ _path << "SingleGPU";
+
+ if (useLimiter) _path << "_Limiter";
+
+ path = _path.str();
+ }
+
+ //////////////////////////////////////////////////////////////////////////
+
+ {
+ std::stringstream _simulationName;
+
+ _simulationName << simulationName;
+ _simulationName << "_nx_" << nx;
+ _simulationName << "_dtPerL_" << dtPerL << "_";
+
+ simulationName = _simulationName.str();
+ }
+
+ //////////////////////////////////////////////////////////////////////////
+
+ para->setDevices(std::vector<uint>{gpuIndex});
+
+ //////////////////////////////////////////////////////////////////////////
+
+ para->setOutputPath( path );
+ para->setOutputPrefix( simulationName );
+
+ para->setPrintFiles(true);
+
+ para->setTimestepEnd(40 * lround(L / velocity));
+ para->setTimestepOut(5 * lround(L / velocity));
+ //para->setTimestepOut(lround(L / velocity));
+ // para->setTimestepEnd(2048);
+ //para->setTimestepOut(512);
+ // para->setTimestepStartOut(500);
+
+ para->setVelocityLB( velocity );
+
+ para->setViscosityLB( viscosity );
+
+ para->setVelocityRatio( 1.0 / velocity );
+
+ para->setDensityRatio(1.0);
+
+ para->setInitialCondition( [&]( real coordX, real coordY, real coordZ, real& rho, real& vx, real& vy, real& vz){
+
+ real a = 1.0;
+ real b = 1.0;
+ real c = 1.0;
+
+ rho = 3.0 * ((velocity * velocity) / 16.0 * ( cos( 2.0 * a * coordX ) + cos( 2.0 * b * coordY ) ) * ( cos( 2.0 * c * coordZ ) + 2.0 ) );
+ vx = velocity * sin( a * coordX ) * cos( b * coordY ) * cos( c * coordZ );
+ vy = -velocity * cos( a * coordX ) * sin( b * coordY ) * cos( c * coordZ );
+ vz = 0.0;
+
+ } );
+
+ para->setMainKernel( kernel );
+
+ if( !useLimiter )
+ para->setQuadricLimiters( 1000000.0, 1000000.0, 1000000.0 );
+
+ if( useWale )
+ para->setUseWale( true );
+
+ para->setUseInitNeq( true );
+
+ ////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
+ ////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
+ ////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
+
+ auto cudaMemoryManager = std::make_shared<CudaMemoryManager>(para);
+ SPtr<GridProvider> gridGenerator = GridProvider::makeGridGenerator(gridBuilder, para, cudaMemoryManager, communicator);
+ //SPtr<GridProvider> gridGenerator = GridProvider::makeGridReader(FILEFORMAT::BINARY, para, cudaMemoryManager);
+
+ SPtr<FileWriter> fileWriter = SPtr<FileWriter>(new FileWriter());
+ Simulation sim(para, cudaMemoryManager, communicator, *gridGenerator, &bcFactory, &scalingFactory);
+ sim.run();
+
+ //sim.addKineticEnergyAnalyzer( 10 );
+ //sim.addEnstrophyAnalyzer( 10 );
+
+ //sim.run();
+}
+
+
+int main( int argc, char* argv[])
+{
+ MPI_Init(&argc, &argv);
+ std::string str, str2;
+ if ( argv != NULL )
+ {
+ //str = static_cast<std::string>(argv[0]);
+
+ try
+ {
+ //////////////////////////////////////////////////////////////////////////
+ std::string targetPath( __FILE__ );
+
+#ifdef _WIN32
+ targetPath = targetPath.substr(0, targetPath.find_last_of('\\') + 1);
+#else
+ targetPath = targetPath.substr(0, targetPath.find_last_of('/') + 1);
+#endif
+
+ //////////////////////////////////////////////////////////////////////////
+
+ if( cmdOptionExists( argv, argv+argc, "--Re" ) )
+ Re = atof( getCmdOption( argv, argv+argc, "--Re" ) );
+
+ if( cmdOptionExists( argv, argv+argc, "--nx" ) )
+ nx = atoi( getCmdOption( argv, argv+argc, "--nx" ) );
+
+ if( cmdOptionExists( argv, argv+argc, "--dtPerL" ) )
+ dtPerL = atoi( getCmdOption( argv, argv+argc, "--dtPerL" ) );
+
+ if( cmdOptionExists( argv, argv+argc, "--kernel" ) )
+ kernel = getCmdOption( argv, argv+argc, "--kernel" );
+
+ if( cmdOptionExists( argv, argv+argc, "--gpu" ) )
+ gpuIndex = atoi( getCmdOption( argv, argv+argc, "--gpu" ) );
+
+ if( cmdOptionExists( argv, argv+argc, "--useLimiter" ) )
+ useLimiter = true;
+
+ if( cmdOptionExists( argv, argv+argc, "--useWale" ) )
+ useWale = true;
+
+ multipleLevel(targetPath + "config.txt");
+
+ //////////////////////////////////////////////////////////////////////////
+ }
+ catch (const std::bad_alloc& e)
+ {
+
+ *logging::out << logging::Logger::LOGGER_ERROR << "Bad Alloc:" << e.what() << "\n";
+ //std::cout << e.what() << std::flush;
+ //MPI_Abort(MPI_COMM_WORLD, -1);
+ }
+ catch (const std::exception& e)
+ {
+
+ *logging::out << logging::Logger::LOGGER_ERROR << e.what() << "\n";
+ //std::cout << e.what() << std::flush;
+ //MPI_Abort(MPI_COMM_WORLD, -1);
+ }
+ catch (...)
+ {
+ *logging::out << logging::Logger::LOGGER_ERROR << "Unknown exception!\n";
+ //std::cout << "unknown exeption" << std::endl;
+ }
+
+ //std::cout << "\nConfiguration file must be set!: lbmgm <config file>" << std::endl << std::flush;
+ //MPI_Abort(MPI_COMM_WORLD, -1);
+ }
+
+
+ /*
+ MPE_Init_log() & MPE_Finish_log() are NOT needed when
+ liblmpe.a is linked with this program. In that case,
+ MPI_Init() would have called MPE_Init_log() already.
+ */
+#if defined( MPI_LOGGING )
+ MPE_Init_log();
+#endif
+
+#if defined( MPI_LOGGING )
+ if ( argv != NULL )
+ MPE_Finish_log( argv[0] );
+ if ( str != "" )
+ MPE_Finish_log( str.c_str() );
+ else
+ MPE_Finish_log( "TestLog" );
+#endif
+
+ MPI_Finalize();
+ return 0;
+}
diff --git a/apps/gpu/LBM/TGV_3D_GridRef/config.txt b/apps/gpu/LBM/TGV_3D_GridRef/config.txt
new file mode 100644
index 0000000000000000000000000000000000000000..ae6d3e9bc4be5403d151f3d59ffb13af7164abf0
--- /dev/null
+++ b/apps/gpu/LBM/TGV_3D_GridRef/config.txt
@@ -0,0 +1,36 @@
+##################################################
+#GPU Mapping
+##################################################
+#Devices="0 1 2 3"
+#NumberOfDevices=4
+
+##################################################
+#informations for Writing
+##################################################
+#Path="E:/DrivenCavity/results"
+#Path="F:/Work/Computations/out/DrivenCavity/"
+#Prefix="DrivenCavity"
+#WriteGrid=true
+##################################################
+#informations for reading
+##################################################
+#GridPath="E:/DrivenCavity/dummy"
+GridPath="F:/Work/Computations/out/TaylorGreen3DNew/grid"
+
+##################################################
+#number of grid levels
+##################################################
+NOGL=2
+
+##################################################
+#LBM Version
+##################################################
+#D3Qxx=27
+#MainKernelName=CumulantAA2016CompSP27
+
+##################################################
+#simulation parameter
+##################################################
+#TimeEnd=100000
+#TimeOut=1000
+#TimeStartOut=0
\ No newline at end of file
diff --git a/apps/gpu/LBM/TGV_3D_MultiGPU/TGV_3D_MultiGPU.cpp b/apps/gpu/LBM/TGV_3D_MultiGPU/TGV_3D_MultiGPU.cpp
index 8c303dc07c911c363e892ce53f7bfe7f48e284d6..045c208274bc6bc216d25e8c2fa905916a52f87b 100644
--- a/apps/gpu/LBM/TGV_3D_MultiGPU/TGV_3D_MultiGPU.cpp
+++ b/apps/gpu/LBM/TGV_3D_MultiGPU/TGV_3D_MultiGPU.cpp
@@ -1,7 +1,38 @@
+//=======================================================================================
+// ____ ____ __ ______ __________ __ __ __ __
+// \ \ | | | | | _ \ |___ ___| | | | | / \ | |
+// \ \ | | | | | |_) | | | | | | | / \ | |
+// \ \ | | | | | _ / | | | | | | / /\ \ | |
+// \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
+// \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
+// \ \ | | ________________________________________________________________
+// \ \ | | | ______________________________________________________________|
+// \ \| | | | __ __ __ __ ______ _______
+// \ | | |_____ | | | | | | | | | _ \ / _____)
+// \ | | _____| | | | | | | | | | | \ \ \_______
+// \ | | | | |_____ | \_/ | | | | |_/ / _____ |
+// \ _____| |__| |________| \_______/ |__| |______/ (_______/
+//
+// This file is part of VirtualFluids. VirtualFluids is free software: you can
+// redistribute it and/or modify it under the terms of the GNU General Public
+// License as published by the Free Software Foundation, either version 3 of
+// the License, or (at your option) any later version.
+//
+// VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
+// ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+// FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
+// for more details.
+//
+// You should have received a copy of the GNU General Public License along
+// with VirtualFluids (see COPYING.txt). If not, see <http://www.gnu.org/licenses/>.
+//
+//! \file TGV_3d_MuitiGPU.cpp
+//! \ingroup TGV_3D_MultiGPU
+//! \author Martin Schoenherr
+//=======================================================================================
//#define MPI_LOGGING
//Martin Branch
-
#include <mpi.h>
#if defined( MPI_LOGGING )
#include <mpe.h>
@@ -97,7 +128,7 @@ bool useWale = false;
int mpirank;
int mpiWorldSize;
-std::string kernel( "CumulantK20Comp" );
+std::string kernel( "CumulantK17CompChim" );
//std::string path("F:/Work/Computations/out/TaylorGreen3DNew/"); //LEGOLAS
//std::string path("results/"); //PHOENIX
diff --git a/apps/gpu/LBM/WTG_RUB/WTG_RUB.cpp b/apps/gpu/LBM/WTG_RUB/WTG_RUB.cpp
index 06b3678d7c8ddd236c26a69686356fbe87c31db2..3e083afd690632dbaabdde5d00f2ab454d86032b 100644
--- a/apps/gpu/LBM/WTG_RUB/WTG_RUB.cpp
+++ b/apps/gpu/LBM/WTG_RUB/WTG_RUB.cpp
@@ -1,3 +1,35 @@
+//=======================================================================================
+// ____ ____ __ ______ __________ __ __ __ __
+// \ \ | | | | | _ \ |___ ___| | | | | / \ | |
+// \ \ | | | | | |_) | | | | | | | / \ | |
+// \ \ | | | | | _ / | | | | | | / /\ \ | |
+// \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
+// \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
+// \ \ | | ________________________________________________________________
+// \ \ | | | ______________________________________________________________|
+// \ \| | | | __ __ __ __ ______ _______
+// \ | | |_____ | | | | | | | | | _ \ / _____)
+// \ | | _____| | | | | | | | | | | \ \ \_______
+// \ | | | | |_____ | \_/ | | | | |_/ / _____ |
+// \ _____| |__| |________| \_______/ |__| |______/ (_______/
+//
+// This file is part of VirtualFluids. VirtualFluids is free software: you can
+// redistribute it and/or modify it under the terms of the GNU General Public
+// License as published by the Free Software Foundation, either version 3 of
+// the License, or (at your option) any later version.
+//
+// VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
+// ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+// FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
+// for more details.
+//
+// You should have received a copy of the GNU General Public License along
+// with VirtualFluids (see COPYING.txt). If not, see <http://www.gnu.org/licenses/>.
+//
+//! \file WTG_RUB.cpp
+//! \ingroup Applications
+//! \author Martin Schoenherr
+//=======================================================================================
#define _USE_MATH_DEFINES
#include <math.h>
#include <string>
@@ -15,15 +47,10 @@
#include "Core/DataTypes.h"
#include "PointerDefinitions.h"
-
#include "Core/LbmOrGks.h"
-#include "Core/StringUtilities/StringUtil.h"
-
#include "Core/VectorTypes.h"
#include "Core/Logger/Logger.h"
-#include <basics/config/ConfigurationFile.h>
-
//////////////////////////////////////////////////////////////////////////
#include "GridGenerator/grid/GridBuilder/LevelGridBuilder.h"
diff --git a/metadata.xml b/metadata.xml
deleted file mode 100644
index 7cbae3ae7e1d5d7d48af2f0e5577253a89f953f5..0000000000000000000000000000000000000000
--- a/metadata.xml
+++ /dev/null
@@ -1,204 +0,0 @@
-<?xml version="1.0" encoding="UTF-8"?>
-<resource xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xmlns="http://datacite.org/schema/kernel-4" xsi:schemaLocation="http://datacite.org/schema/kernel-4 http://schema.datacite.org/meta/kernel-4.3/metadata.xsd">
- <identifier identifierType="DOI">PLACEHOLDER</identifier>
- <titles>
- <title xml:lang="en">VirtualFluids</title>
- </titles>
- <language>en</language>
- <creators>
- <creator>
- <creatorName nameType="Personal">Krafczyk, Manfred</creatorName>
- <givenName>Manfred</givenName>
- <familyName>Krafczyk</familyName>
- <nameIdentifier nameIdentifierScheme="ORCID">0000-0002-8509-0871</nameIdentifier>
- <affiliation xml:lang="de">TU Braunschweig</affiliation>
- <affiliation xml:lang="de">Institut für rechnergestützte Modellierung im Bauingenieurwesen</affiliation>
- </creator>
- <creator>
- <creatorName nameType="Organizational">Institut für rechnergestützte Modellierung im Bauingenieurwesen</creatorName>
- <affiliation xml:lang="de">TU Braunschweig</affiliation>
- </creator>
- </creators>
- <publisher xml:lang="de">Institut für rechnergestützte Modellierung im Bauingenieurwesen</publisher>
- <publicationYear>2021</publicationYear>
- <resourceType resourceTypeGeneral="Software">Computational Fluid Dynamics Solver</resourceType>
- <subjects>
- <subject subjectScheme="DDC" schemeURI="https://www.oclc.org/en/dewey.html">532 Fluid Mechanics, liquid mechanics</subject>
- </subjects>
- <contributors>
- <contributor contributorType="Researcher">
- <contributorName>Ahrenholz, Benjamin</contributorName>
- <givenName>Benjamin</givenName>
- <familyName>Ahrenholz</familyName>
- </contributor>
-
- <contributor contributorType="Researcher">
- <contributorName>Alihussein, Hussein</contributorName>
- <givenName>Hussein</givenName>
- <familyName>Alihussein</familyName>
- <nameIdentifier nameIdentifierScheme="ORCID" schemeURI="http://orcid.org/">0000-0003-3656-7028</nameIdentifier>
- <affiliation xml:lang="de">TU Braunschweig</affiliation>
- <affiliation xml:lang="en">Institut für rechnergestützte Modellierung im Bauingenieurwesen</affiliation>
- </contributor>
-
- <contributor contributorType="Researcher">
- <contributorName>Bindick, Sebastian</contributorName>
- <givenName>Sebastian</givenName>
- <familyName>Bindick</familyName>
- </contributor>
-
- <contributor contributorType="Researcher">
- <contributorName>Brendel, Aileen</contributorName>
- <givenName>Aileen</givenName>
- <familyName>Brendel</familyName>
- </contributor>
-
- <contributor contributorType="Researcher">
- <contributorName>Geier, Martin</contributorName>
- <givenName>Martin</givenName>
- <familyName>Geier</familyName>
- <nameIdentifier nameIdentifierScheme="ORCID" schemeURI="http://orcid.org/">0000-0002-8367-9412</nameIdentifier>
- <affiliation xml:lang="de">TU Braunschweig</affiliation>
- <affiliation xml:lang="en">Institut für rechnergestützte Modellierung im Bauingenieurwesen</affiliation>
- </contributor>
-
- <contributor contributorType="Researcher">
- <contributorName>Geller, Sebastian</contributorName>
- <givenName>Sebastian</givenName>
- <familyName>Geller</familyName>
- </contributor>
-
- <contributor contributorType="Researcher">
- <contributorName>Goraki Fard, Ehsan</contributorName>
- <givenName>Ehsan</givenName>
- <familyName>Goraki Fard</familyName>
- </contributor>
-
- <contributor contributorType="Researcher">
- <contributorName>Hegewald, Jan</contributorName>
- <givenName>Jan</givenName>
- <familyName>Hegewald</familyName>
- </contributor>
-
- <contributor contributorType="Researcher">
- <contributorName>Janßen, Christian</contributorName>
- <givenName>Christian</givenName>
- <familyName>Janßen</familyName>
- </contributor>
-
- <contributor contributorType="Researcher">
- <contributorName>Kutscher, Konstantin</contributorName>
- <givenName>Konstantin</givenName>
- <familyName>Kutscher</familyName>
- <nameIdentifier nameIdentifierScheme="ORCID" schemeURI="http://orcid.org/">0000-0002-1099-1608</nameIdentifier>
- <affiliation xml:lang="de">TU Braunschweig</affiliation>
- <affiliation xml:lang="en">Institut für rechnergestützte Modellierung im Bauingenieurwesen</affiliation>
- </contributor>
-
- <contributor contributorType="Researcher">
- <contributorName>Lenz, Stephan</contributorName>
- <givenName>Stephan</givenName>
- <familyName>Lenz</familyName>
- </contributor>
-
- <contributor contributorType="Researcher">
- <contributorName>Linxweiler, Jan</contributorName>
- <givenName>Jan</givenName>
- <familyName>Linxweiler</familyName>
- <nameIdentifier nameIdentifierScheme="ORCID" schemeURI="http://orcid.org/">0000-0002-2755-5087</nameIdentifier>
- <affiliation xml:lang="de">TU Braunschweig</affiliation>
- <affiliation xml:lang="en">Institut für rechnergestützte Modellierung im Bauingenieurwesen</affiliation>
- </contributor>
-
- <contributor contributorType="Researcher">
- <contributorName>Lux, Lennard</contributorName>
- <givenName>Lennard</givenName>
- <familyName>Lux</familyName>
- </contributor>
-
- <contributor contributorType="Researcher">
- <contributorName>Marcus, Sven</contributorName>
- <givenName>Sven</givenName>
- <familyName>Marcus</familyName>
- <nameIdentifier nameIdentifierScheme="ORCID" schemeURI="http://orcid.org/">0000-0003-3689-2162</nameIdentifier>
- <affiliation xml:lang="de">TU Braunschweig</affiliation>
- <affiliation xml:lang="en">Universitätsbibliothek Braunschweig</affiliation>
- </contributor>
-
- <contributor contributorType="Researcher">
- <contributorName>Peters, Sören</contributorName>
- <givenName>Sören</givenName>
- <familyName>Peters</familyName>
- <affiliation xml:lang="de">TU Braunschweig</affiliation>
- <affiliation xml:lang="en">Institut für rechnergestützte Modellierung im Bauingenieurwesen</affiliation>
- </contributor>
-
- <contributor contributorType="Researcher">
- <contributorName>Safari, Hesameddin</contributorName>
- <givenName>Hesameddin</givenName>
- <familyName>Safari</familyName>
- </contributor>
-
- <contributor contributorType="Researcher">
- <contributorName>Schönherr, Martin</contributorName>
- <givenName>Martin</givenName>
- <familyName>Schönherr</familyName>
- <nameIdentifier nameIdentifierScheme="ORCID" schemeURI="http://orcid.org/">0000-0002-4774-1776</nameIdentifier>
- <affiliation xml:lang="de">TU Braunschweig</affiliation>
- <affiliation xml:lang="en">Institut für rechnergestützte Modellierung im Bauingenieurwesen</affiliation>
- </contributor>
-
- <contributor contributorType="Researcher">
- <contributorName>Stiebler, Maik</contributorName>
- <givenName>Maik</givenName>
- <familyName>Stiebler</familyName>
- </contributor>
-
- <contributor contributorType="Researcher">
- <contributorName>Textor, Sören</contributorName>
- <givenName>Sören</givenName>
- <familyName>Textor</familyName>
- </contributor>
-
- <contributor contributorType="Researcher">
- <contributorName>Tölke, Jonas</contributorName>
- <givenName>Jonas</givenName>
- <familyName>Tölke</familyName>
- </contributor>
-
- <contributor contributorType="Researcher">
- <contributorName>Uphoff, Sonja</contributorName>
- <givenName>Sonja</givenName>
- <familyName>Uphoff</familyName>
- </contributor>
-
- <contributor contributorType="Researcher">
- <contributorName>Wellmann, Anna</contributorName>
- <givenName>Anna</givenName>
- <familyName>Wellmann</familyName>
- </contributor>
- </contributors>
- <dates>
- <date dateType="Created">2000</date>
- </dates>
- <formats>
- <format>text/x-c</format>
- <format>text/x-h</format>
- <format>text/x-script.python</format>
- </formats>
- <relatedIdentifiers>
- <relatedIdentifier relatedIdentifierType="URL" relationType="Requires" resourceTypeGeneral="Software">https://www.open-mpi.org/software/ompi/v4.1/</relatedIdentifier>
- <relatedIdentifier relatedIdentifierType="URL" relationType="IsCompiledBy" resourceTypeGeneral="Software">https://cmake.org</relatedIdentifier>
- <relatedIdentifier relatedIdentifierType="URL" relationType="IsCompiledBy" resourceTypeGeneral="Software">https://gcc.gnu.org</relatedIdentifier>
- <relatedIdentifier relatedIdentifierType="URL" relationType="IsCompiledBy" resourceTypeGeneral="Software">https://clang.llvm.org</relatedIdentifier>
- <relatedIdentifier relatedIdentifierType="URL" relationType="IsCompiledBy" resourceTypeGeneral="Software">https://visualstudio.microsoft.com/vs/features/cplusplus/</relatedIdentifier>
- </relatedIdentifiers>
- <rightsList>
- <rights xml:lang="en" schemeURI="https://spdx.org/licenses/" rightsIdentifierScheme="SPDX" rightsIdentifier="GPL-3.0-only" rightsURI="https://www.gnu.org/licenses/gpl-3.0-standalone.html">GNU General Public License Version 3</rights>
- </rightsList>
- <descriptions>
- <description descriptionType="Abstract">
- VirtualFluids (VF) is a research code developed at the Institute for Computational Modeling in Civil Engineering (iRMB). The code is a Computational Fluid Dynamics (CFD) solver based on the Lattice Boltzmann Method (LBM) for turbulent, thermal, multiphase and multicomponent flow problems as well as for multi-field problems such as Fluid-Structure-interaction including distributed pre- and postprocessing capabilities for simulations with more than 100 billion degrees of freedom.
- </description>
- </descriptions>
-</resource>
diff --git a/pyproject.toml b/pyproject.toml
index 8fcb7926102d188b44d8c74084235b6f175edf80..257da6fd95d683081dbff865c864079eae9c675d 100644
--- a/pyproject.toml
+++ b/pyproject.toml
@@ -1,2 +1,9 @@
[build-system]
-requires = ["setuptools", "wheel", "scikit-build"]
\ No newline at end of file
+requires = [
+ "setuptools>=42",
+ "scikit-build",
+ "cmake",
+ "ninja; platform_system!='Windows'"
+]
+build-backend = "setup_builder"
+backend-path = ["utilities"]
\ No newline at end of file
diff --git a/pythonbindings/CMakeLists.txt b/pythonbindings/CMakeLists.txt
index 5a84adef027fdfa2953e016693bb64570e48c1ef..815a4b59cf6c3e4e5ac4a7a72a5bd4e374d64c96 100644
--- a/pythonbindings/CMakeLists.txt
+++ b/pythonbindings/CMakeLists.txt
@@ -1,24 +1,45 @@
-project(VirtualFluidsPython LANGUAGES CUDA CXX)
+set(PYFLUIDS_LANGUAGES CXX)
+
+if(BUILD_VF_GPU)
+ set(PYFLUIDS_LANGUAGES CUDA CXX)
+endif()
+
+project(VirtualFluidsPython LANGUAGES ${PYFLUIDS_LANGUAGES})
+
+pybind11_add_module(python_bindings MODULE src/VirtualFluids.cpp)
+
+set_target_properties( python_bindings PROPERTIES
+ LIBRARY_OUTPUT_DIRECTORY ${CMAKE_SOURCE_DIR}/pythonbindings/pyfluids
+ OUTPUT_NAME "bindings")
+
+target_link_libraries(python_bindings PRIVATE basics logger mpi)
+
IF(BUILD_VF_GPU)
- pybind11_add_module(pyfluids src/VirtualFluidsModulesGPU.cpp)
- set_source_files_properties(src/VirtualFluidsModulesGPU.cpp PROPERTIES LANGUAGE CUDA)
+ set_source_files_properties(src/VirtualFluids.cpp PROPERTIES LANGUAGE CUDA)
- target_link_libraries(pyfluids PRIVATE GridGenerator VirtualFluids_GPU basics lbmCuda logger)
- target_include_directories(pyfluids PRIVATE ${VF_THIRD_DIR}/cuda_samples/)
+ target_include_directories(python_bindings PRIVATE ${VF_THIRD_DIR}/cuda_samples/)
+ target_compile_definitions(python_bindings PRIVATE VF_GPU_PYTHONBINDINGS)
+ target_link_libraries(python_bindings PRIVATE GridGenerator VirtualFluids_GPU lbm)
ENDIF()
+
IF(BUILD_VF_CPU)
- pybind11_add_module(pyfluids src/VirtualFluidsModulesCPU.cpp)
- pybind11_add_module(pymuparser src/muParser.cpp)
+ target_compile_definitions(python_bindings PRIVATE VF_METIS VF_MPI VF_CPU_PYTHONBINDINGS)
+ target_link_libraries(python_bindings PRIVATE simulationconfig VirtualFluidsCore muparser lbm)
+
+ # include bindings for muparsers
+ pybind11_add_module(pymuparser MODULE src/muParser.cpp)
# TODO: Move this to MuParser CMakeLists.txt
set_target_properties(muparser PROPERTIES POSITION_INDEPENDENT_CODE ON)
- target_compile_definitions(pyfluids PRIVATE VF_METIS VF_MPI)
+ set_target_properties( pymuparser PROPERTIES
+ LIBRARY_OUTPUT_DIRECTORY ${CMAKE_SOURCE_DIR}/pythonbindings/pymuparser
+ OUTPUT_NAME "bindings")
target_compile_definitions(pymuparser PRIVATE VF_METIS VF_MPI)
-
- target_link_libraries(pyfluids PRIVATE simulationconfig VirtualFluidsCore muparser basics)
target_link_libraries(pymuparser PRIVATE muparser)
ENDIF()
-target_include_directories(pyfluids PRIVATE ${CMAKE_SOURCE_DIR}/src/)
-target_include_directories(pyfluids PRIVATE ${CMAKE_BINARY_DIR})
\ No newline at end of file
+
+
+target_include_directories(python_bindings PRIVATE ${CMAKE_SOURCE_DIR}/src/)
+target_include_directories(python_bindings PRIVATE ${CMAKE_BINARY_DIR})
\ No newline at end of file
diff --git a/Python/boundary_layer/__init__.py b/pythonbindings/pyfluids-stubs/__init__.pyi
similarity index 100%
rename from Python/boundary_layer/__init__.py
rename to pythonbindings/pyfluids-stubs/__init__.pyi
diff --git a/pythonbindings/pyfluids-stubs/bindings/__init__.pyi b/pythonbindings/pyfluids-stubs/bindings/__init__.pyi
new file mode 100644
index 0000000000000000000000000000000000000000..4e7f353eab97cc536f8f18e72319af1cd7a1916a
--- /dev/null
+++ b/pythonbindings/pyfluids-stubs/bindings/__init__.pyi
@@ -0,0 +1,38 @@
+r"""
+=======================================================================================
+ ____ ____ __ ______ __________ __ __ __ __
+ \ \ | | | | | _ \ |___ ___| | | | | / \ | |
+ \ \ | | | | | |_) | | | | | | | / \ | |
+ \ \ | | | | | _ / | | | | | | / /\ \ | |
+ \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
+ \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
+ \ \ | | ________________________________________________________________
+ \ \ | | | ______________________________________________________________|
+ \ \| | | | __ __ __ __ ______ _______
+ \ | | |_____ | | | | | | | | | _ \ / _____)
+ \ | | _____| | | | | | | | | | | \ \ \_______
+ \ | | | | |_____ | \_/ | | | | |_/ / _____ |
+ \ _____| |__| |________| \_______/ |__| |______/ (_______/
+
+ This file is part of VirtualFluids. VirtualFluids is free software: you can
+ redistribute it and/or modify it under the terms of the GNU General Public
+ License as published by the Free Software Foundation, either version 3 of
+ the License, or (at your option) any later version.
+
+ VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
+ ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
+ for more details.
+
+ You should have received a copy of the GNU General Public License along
+ with VirtualFluids (see COPYING.txt). If not, see <http://www.gnu.org/licenses/>.
+
+! \file __init__.pyi
+! \ingroup bindings
+! \author Henry Korb
+=======================================================================================
+"""
+class ostream_redirect:
+ def __init__(self, stdout: bool = ..., stderr: bool = ...) -> None: ...
+ def __enter__(self) -> None: ...
+ def __exit__(self, *args) -> None: ...
diff --git a/pythonbindings/pyfluids-stubs/bindings/basics/__init__.pyi b/pythonbindings/pyfluids-stubs/bindings/basics/__init__.pyi
new file mode 100644
index 0000000000000000000000000000000000000000..a41b7934ca706dc0db5bd6188fee3150456e0cd9
--- /dev/null
+++ b/pythonbindings/pyfluids-stubs/bindings/basics/__init__.pyi
@@ -0,0 +1,82 @@
+r"""
+=======================================================================================
+ ____ ____ __ ______ __________ __ __ __ __
+ \ \ | | | | | _ \ |___ ___| | | | | / \ | |
+ \ \ | | | | | |_) | | | | | | | / \ | |
+ \ \ | | | | | _ / | | | | | | / /\ \ | |
+ \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
+ \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
+ \ \ | | ________________________________________________________________
+ \ \ | | | ______________________________________________________________|
+ \ \| | | | __ __ __ __ ______ _______
+ \ | | |_____ | | | | | | | | | _ \ / _____)
+ \ | | _____| | | | | | | | | | | \ \ \_______
+ \ | | | | |_____ | \_/ | | | | |_/ / _____ |
+ \ _____| |__| |________| \_______/ |__| |______/ (_______/
+
+ This file is part of VirtualFluids. VirtualFluids is free software: you can
+ redistribute it and/or modify it under the terms of the GNU General Public
+ License as published by the Free Software Foundation, either version 3 of
+ the License, or (at your option) any later version.
+
+ VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
+ ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
+ for more details.
+
+ You should have received a copy of the GNU General Public License along
+ with VirtualFluids (see COPYING.txt). If not, see <http://www.gnu.org/licenses/>.
+
+! \file __init__.py
+! \ingroup basics
+! \author Henry Korb
+=======================================================================================
+"""
+from typing import ClassVar
+
+from typing import overload
+
+class ConfigurationFile:
+ def __init__(self) -> None: ...
+ def contains(self, key: str) -> bool: ...
+ @overload
+ def get_bool_value(self, key: str) -> bool: ...
+ @overload
+ def get_bool_value(self, key: str, default_value: bool) -> bool: ...
+ @overload
+ def get_double_value(self, key: str) -> float: ...
+ @overload
+ def get_double_value(self, key: str, default_value: float) -> float: ...
+ @overload
+ def get_float_value(self, key: str) -> float: ...
+ @overload
+ def get_float_value(self, key: str, default_value: float) -> float: ...
+ @overload
+ def get_int_value(self, key: str) -> int: ...
+ @overload
+ def get_int_value(self, key: str, default_value: int) -> int: ...
+ @overload
+ def get_string_value(self, key: str) -> str: ...
+ @overload
+ def get_string_value(self, key: str, default_value: str) -> str: ...
+ @overload
+ def get_uint_value(self, key: str) -> int: ...
+ @overload
+ def get_uint_value(self, key: str, default_value: int) -> int: ...
+ def load(self, file: str) -> bool: ...
+
+class LbmOrGks:
+ __members__: ClassVar[dict] = ... # read-only
+ GKS: ClassVar[LbmOrGks] = ...
+ LBM: ClassVar[LbmOrGks] = ...
+ __entries: ClassVar[dict] = ...
+ def __init__(self, arg0: int) -> None: ...
+ def __eq__(self, arg0: object) -> bool: ...
+ def __getstate__(self) -> int: ...
+ def __hash__(self) -> int: ...
+ def __index__(self) -> int: ...
+ def __int__(self) -> int: ...
+ def __ne__(self, arg0: object) -> bool: ...
+ def __setstate__(self, arg0: int) -> None: ...
+ @property
+ def name(self) -> str: ...
diff --git a/pythonbindings/pyfluids-stubs/bindings/basics/logger.pyi b/pythonbindings/pyfluids-stubs/bindings/basics/logger.pyi
new file mode 100644
index 0000000000000000000000000000000000000000..43938ff7646efd3c596ae29971cce39fed865fa6
--- /dev/null
+++ b/pythonbindings/pyfluids-stubs/bindings/basics/logger.pyi
@@ -0,0 +1,83 @@
+r"""
+=======================================================================================
+ ____ ____ __ ______ __________ __ __ __ __
+ \ \ | | | | | _ \ |___ ___| | | | | / \ | |
+ \ \ | | | | | |_) | | | | | | | / \ | |
+ \ \ | | | | | _ / | | | | | | / /\ \ | |
+ \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
+ \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
+ \ \ | | ________________________________________________________________
+ \ \ | | | ______________________________________________________________|
+ \ \| | | | __ __ __ __ ______ _______
+ \ | | |_____ | | | | | | | | | _ \ / _____)
+ \ | | _____| | | | | | | | | | | \ \ \_______
+ \ | | | | |_____ | \_/ | | | | |_/ / _____ |
+ \ _____| |__| |________| \_______/ |__| |______/ (_______/
+
+ This file is part of VirtualFluids. VirtualFluids is free software: you can
+ redistribute it and/or modify it under the terms of the GNU General Public
+ License as published by the Free Software Foundation, either version 3 of
+ the License, or (at your option) any later version.
+
+ VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
+ ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
+ for more details.
+
+ You should have received a copy of the GNU General Public License along
+ with VirtualFluids (see COPYING.txt). If not, see <http://www.gnu.org/licenses/>.
+
+! \file logger.pyi
+! \ingroup basics
+! \author Henry Korb
+=======================================================================================
+"""
+from typing import Any, ClassVar
+
+log: None
+
+class Level:
+ __members__: ClassVar[dict] = ... # read-only
+ INFO_HIGH: ClassVar[Level] = ...
+ INFO_INTERMEDIATE: ClassVar[Level] = ...
+ INFO_LOW: ClassVar[Level] = ...
+ LOGGER_ERROR: ClassVar[Level] = ...
+ WARNING: ClassVar[Level] = ...
+ __entries: ClassVar[dict] = ...
+ def __init__(self, arg0: int) -> None: ...
+ def __eq__(self, arg0: object) -> bool: ...
+ def __getstate__(self) -> int: ...
+ def __hash__(self) -> int: ...
+ def __index__(self) -> int: ...
+ def __int__(self) -> int: ...
+ def __ne__(self, arg0: object) -> bool: ...
+ def __setstate__(self, arg0: int) -> None: ...
+ @property
+ def name(self) -> str: ...
+
+class Logger:
+ def __init__(self, *args, **kwargs) -> None: ...
+ @staticmethod
+ def add_stdout() -> None: ...
+ @staticmethod
+ def enable_printed_rank_numbers(print: bool) -> None: ...
+ @staticmethod
+ def set_debug_level(level: int) -> None: ...
+ @staticmethod
+ def time_stamp(time_stemp: TimeStamp) -> None: ...
+
+class TimeStamp:
+ __members__: ClassVar[dict] = ... # read-only
+ DISABLE: ClassVar[TimeStamp] = ...
+ ENABLE: ClassVar[TimeStamp] = ...
+ __entries: ClassVar[dict] = ...
+ def __init__(self, arg0: int) -> None: ...
+ def __eq__(self, arg0: object) -> bool: ...
+ def __getstate__(self) -> int: ...
+ def __hash__(self) -> int: ...
+ def __index__(self) -> int: ...
+ def __int__(self) -> int: ...
+ def __ne__(self, arg0: object) -> bool: ...
+ def __setstate__(self, arg0: int) -> None: ...
+ @property
+ def name(self) -> str: ...
diff --git a/pythonbindings/pyfluids-stubs/bindings/gpu/__init__.pyi b/pythonbindings/pyfluids-stubs/bindings/gpu/__init__.pyi
new file mode 100644
index 0000000000000000000000000000000000000000..36c2fea76713e980bb95eb6726d778de8c9a6583
--- /dev/null
+++ b/pythonbindings/pyfluids-stubs/bindings/gpu/__init__.pyi
@@ -0,0 +1,436 @@
+r"""
+=======================================================================================
+ ____ ____ __ ______ __________ __ __ __ __
+ \ \ | | | | | _ \ |___ ___| | | | | / \ | |
+ \ \ | | | | | |_) | | | | | | | / \ | |
+ \ \ | | | | | _ / | | | | | | / /\ \ | |
+ \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
+ \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
+ \ \ | | ________________________________________________________________
+ \ \ | | | ______________________________________________________________|
+ \ \| | | | __ __ __ __ ______ _______
+ \ | | |_____ | | | | | | | | | _ \ / _____)
+ \ | | _____| | | | | | | | | | | \ \ \_______
+ \ | | | | |_____ | \_/ | | | | |_/ / _____ |
+ \ _____| |__| |________| \_______/ |__| |______/ (_______/
+
+ This file is part of VirtualFluids. VirtualFluids is free software: you can
+ redistribute it and/or modify it under the terms of the GNU General Public
+ License as published by the Free Software Foundation, either version 3 of
+ the License, or (at your option) any later version.
+
+ VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
+ ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
+ for more details.
+
+ You should have received a copy of the GNU General Public License along
+ with VirtualFluids (see COPYING.txt). If not, see <http://www.gnu.org/licenses/>.
+
+! \file __init__.pyi
+! \ingroup gpu
+! \author Henry Korb
+=======================================================================================
+"""
+from typing import Any, Callable, ClassVar, List, Optional
+
+from typing import overload
+import numpy
+import pyfluids.bindings.basics
+import pyfluids.bindings.gpu.grid_generator as grid_generator
+
+class ActuatorFarm(PreCollisionInteractor):
+ def __init__(self, number_of_blades_per_turbine: int, density: float, number_of_nodes_per_blade: int, epsilon: float, level: int, delta_t: float, delta_x: float, use_host_arrays: bool) -> None: ...
+ def add_turbine(self, posX: float, posY: float, posZ: float, diameter: float, omega: float, azimuth: float, yaw: float, bladeRadii: List[float]) -> None: ...
+ def calc_blade_forces(self) -> None: ...
+ def get_all_azimuths(self) -> numpy.ndarray[numpy.float32]: ...
+ def get_all_blade_coords_x(self) -> numpy.ndarray[numpy.float32]: ...
+ def get_all_blade_coords_x_device(self) -> int: ...
+ def get_all_blade_coords_y(self) -> numpy.ndarray[numpy.float32]: ...
+ def get_all_blade_coords_y_device(self) -> int: ...
+ def get_all_blade_coords_z(self) -> numpy.ndarray[numpy.float32]: ...
+ def get_all_blade_coords_z_device(self) -> int: ...
+ def get_all_blade_forces_x(self) -> numpy.ndarray[numpy.float32]: ...
+ def get_all_blade_forces_x_device(self) -> int: ...
+ def get_all_blade_forces_y(self) -> numpy.ndarray[numpy.float32]: ...
+ def get_all_blade_forces_y_device(self) -> int: ...
+ def get_all_blade_forces_z(self) -> numpy.ndarray[numpy.float32]: ...
+ def get_all_blade_forces_z_device(self) -> int: ...
+ def get_all_blade_radii(self) -> numpy.ndarray[numpy.float32]: ...
+ def get_all_blade_radii_device(self) -> int: ...
+ def get_all_blade_velocities_x(self) -> numpy.ndarray[numpy.float32]: ...
+ def get_all_blade_velocities_x_device(self) -> int: ...
+ def get_all_blade_velocities_y(self) -> numpy.ndarray[numpy.float32]: ...
+ def get_all_blade_velocities_y_device(self) -> int: ...
+ def get_all_blade_velocities_z(self) -> numpy.ndarray[numpy.float32]: ...
+ def get_all_blade_velocities_z_device(self) -> int: ...
+ def get_all_omegas(self) -> numpy.ndarray[numpy.float32]: ...
+ def get_all_turbine_pos_x(self) -> numpy.ndarray[numpy.float32]: ...
+ def get_all_turbine_pos_y(self) -> numpy.ndarray[numpy.float32]: ...
+ def get_all_turbine_pos_z(self) -> numpy.ndarray[numpy.float32]: ...
+ def get_all_yaws(self) -> numpy.ndarray[numpy.float32]: ...
+ def get_turbine_azimuth(self, turbine: int) -> float: ...
+ def get_turbine_blade_coords_x(self, turbine: int) -> numpy.ndarray[numpy.float32]: ...
+ def get_turbine_blade_coords_x_device(self, turbine: int) -> int: ...
+ def get_turbine_blade_coords_y(self, turbine: int) -> numpy.ndarray[numpy.float32]: ...
+ def get_turbine_blade_coords_y_device(self, turbine: int) -> int: ...
+ def get_turbine_blade_coords_z(self, turbine: int) -> numpy.ndarray[numpy.float32]: ...
+ def get_turbine_blade_coords_z_device(self, turbine: int) -> int: ...
+ def get_turbine_blade_forces_x(self, turbine: int) -> numpy.ndarray[numpy.float32]: ...
+ def get_turbine_blade_forces_x_device(self, turbine: int) -> int: ...
+ def get_turbine_blade_forces_y(self, turbine: int) -> numpy.ndarray[numpy.float32]: ...
+ def get_turbine_blade_forces_y_device(self, turbine: int) -> int: ...
+ def get_turbine_blade_forces_z(self, turbine: int) -> numpy.ndarray[numpy.float32]: ...
+ def get_turbine_blade_forces_z_device(self, turbine: int) -> int: ...
+ def get_turbine_blade_radii(self, turbine: int) -> numpy.ndarray[numpy.float32]: ...
+ def get_turbine_blade_radii_device(self, turbine: int) -> int: ...
+ def get_turbine_blade_velocities_x(self, turbine: int) -> numpy.ndarray[numpy.float32]: ...
+ def get_turbine_blade_velocities_x_device(self, turbine: int) -> int: ...
+ def get_turbine_blade_velocities_y(self, turbine: int) -> numpy.ndarray[numpy.float32]: ...
+ def get_turbine_blade_velocities_y_device(self, turbine: int) -> int: ...
+ def get_turbine_blade_velocities_z(self, turbine: int) -> numpy.ndarray[numpy.float32]: ...
+ def get_turbine_blade_velocities_z_device(self, turbine: int) -> int: ...
+ def get_turbine_omega(self, turbine: int) -> float: ...
+ def get_turbine_pos(self, turbine: int) -> numpy.ndarray[numpy.float32]: ...
+ def get_turbine_yaw(self, turbine: int) -> float: ...
+ def set_all_azimuths(self, azimuths: numpy.ndarray[numpy.float32]) -> None: ...
+ def set_all_blade_coords(self, blade_coords_x: numpy.ndarray[numpy.float32], blade_coords_y: numpy.ndarray[numpy.float32], blade_coords_z: numpy.ndarray[numpy.float32]) -> None: ...
+ def set_all_blade_forces(self, blade_forces_x: numpy.ndarray[numpy.float32], blade_forces_y: numpy.ndarray[numpy.float32], blade_forces_z: numpy.ndarray[numpy.float32]) -> None: ...
+ def set_all_blade_velocities(self, blade_velocities_x: numpy.ndarray[numpy.float32], blade_velocities_y: numpy.ndarray[numpy.float32], blade_velocities_z: numpy.ndarray[numpy.float32]) -> None: ...
+ def set_all_omegas(self, omegas: numpy.ndarray[numpy.float32]) -> None: ...
+ def set_all_yaws(self, yaws: numpy.ndarray[numpy.float32]) -> None: ...
+ def set_turbine_azimuth(self, turbine: int, azimuth: float) -> None: ...
+ def set_turbine_blade_coords(self, turbine: int, blade_coords_x: numpy.ndarray[numpy.float32], blade_coords_y: numpy.ndarray[numpy.float32], blade_coords_z: numpy.ndarray[numpy.float32]) -> None: ...
+ def set_turbine_blade_forces(self, turbine: int, blade_forces_x: numpy.ndarray[numpy.float32], blade_forces_y: numpy.ndarray[numpy.float32], blade_forces_z: numpy.ndarray[numpy.float32]) -> None: ...
+ def set_turbine_blade_velocities(self, turbine: int, blade_velocities_x: numpy.ndarray[numpy.float32], blade_velocities_y: numpy.ndarray[numpy.float32], blade_velocities_z: numpy.ndarray[numpy.float32]) -> None: ...
+ def set_turbine_omega(self, turbine: int, omega: float) -> None: ...
+ def set_turbine_yaw(self, turbine: int, yaw: float) -> None: ...
+ @property
+ def delta_t(self) -> float: ...
+ @property
+ def delta_x(self) -> float: ...
+ @property
+ def density(self) -> float: ...
+ @property
+ def number_of_blades_per_turbine(self) -> int: ...
+ @property
+ def number_of_indices(self) -> int: ...
+ @property
+ def number_of_nodes(self) -> int: ...
+ @property
+ def number_of_nodes_per_blade(self) -> int: ...
+ @property
+ def number_of_turbines(self) -> int: ...
+
+class BoundaryConditionFactory:
+ def __init__(self) -> None: ...
+ def set_geometry_boundary_condition(self, boundary_condition_type) -> None: ...
+ def set_no_slip_boundary_condition(self, boundary_condition_type) -> None: ...
+ def set_precursor_boundary_condition(self, boundary_condition_type) -> None: ...
+ def set_pressure_boundary_condition(self, boundary_condition_type) -> None: ...
+ def set_slip_boundary_condition(self, boundary_condition_type) -> None: ...
+ def set_stress_boundary_condition(self, boundary_condition_type) -> None: ...
+ def set_velocity_boundary_condition(self, boundary_condition_type) -> None: ...
+
+class Communicator:
+ def __init__(self, *args, **kwargs) -> None: ...
+ @staticmethod
+ def get_instance() -> Communicator: ...
+ def get_number_of_process(self) -> int: ...
+ def get_pid(self) -> int: ...
+
+class CudaMemoryManager:
+ def __init__(self, parameter: Parameter) -> None: ...
+
+class FileType:
+ __members__: ClassVar[dict] = ... # read-only
+ VTK: ClassVar[FileType] = ...
+ __entries: ClassVar[dict] = ...
+ def __init__(self, arg0: int) -> None: ...
+ def __eq__(self, arg0: object) -> bool: ...
+ def __getstate__(self) -> int: ...
+ def __hash__(self) -> int: ...
+ def __index__(self) -> int: ...
+ def __int__(self) -> int: ...
+ def __ne__(self, arg0: object) -> bool: ...
+ def __setstate__(self, arg0: int) -> None: ...
+ @property
+ def name(self) -> str: ...
+
+class GridProvider:
+ def __init__(self, *args, **kwargs) -> None: ...
+ @staticmethod
+ def make_grid_generator(builder: grid_generator.GridBuilder, para: Parameter, cuda_memory_manager: CudaMemoryManager, communicator: Communicator) -> GridProvider: ...
+
+class GridScaling:
+ __members__: ClassVar[dict] = ... # read-only
+ NotSpecified: ClassVar[GridScaling] = ...
+ ScaleCompressible: ClassVar[GridScaling] = ...
+ ScaleRhoSq: ClassVar[GridScaling] = ...
+ __entries: ClassVar[dict] = ...
+ def __init__(self, arg0: int) -> None: ...
+ def __eq__(self, arg0: object) -> bool: ...
+ def __getstate__(self) -> int: ...
+ def __hash__(self) -> int: ...
+ def __index__(self) -> int: ...
+ def __int__(self) -> int: ...
+ def __ne__(self, arg0: object) -> bool: ...
+ def __setstate__(self, arg0: int) -> None: ...
+ @property
+ def name(self) -> str: ...
+
+class GridScalingFactory:
+ def __init__(self) -> None: ...
+ def set_scaling_factory(self, scaling_type) -> None: ...
+
+class NoSlipBC:
+ __members__: ClassVar[dict] = ... # read-only
+ NoSlip3rdMomentsCompressible: ClassVar[NoSlipBC] = ...
+ NoSlipBounceBack: ClassVar[NoSlipBC] = ...
+ NoSlipCompressible: ClassVar[NoSlipBC] = ...
+ NoSlipImplicitBounceBack: ClassVar[NoSlipBC] = ...
+ NoSlipIncompressible: ClassVar[NoSlipBC] = ...
+ __entries: ClassVar[dict] = ...
+ def __init__(self, arg0: int) -> None: ...
+ def __eq__(self, arg0: object) -> bool: ...
+ def __getstate__(self) -> int: ...
+ def __hash__(self) -> int: ...
+ def __index__(self) -> int: ...
+ def __int__(self) -> int: ...
+ def __ne__(self, arg0: object) -> bool: ...
+ def __setstate__(self, arg0: int) -> None: ...
+ @property
+ def name(self) -> str: ...
+
+class OutputVariable:
+ __members__: ClassVar[dict] = ... # read-only
+ Distributions: ClassVar[OutputVariable] = ...
+ Velocities: ClassVar[OutputVariable] = ...
+ __entries: ClassVar[dict] = ...
+ def __init__(self, arg0: int) -> None: ...
+ def __eq__(self, arg0: object) -> bool: ...
+ def __getstate__(self) -> int: ...
+ def __hash__(self) -> int: ...
+ def __index__(self) -> int: ...
+ def __int__(self) -> int: ...
+ def __ne__(self, arg0: object) -> bool: ...
+ def __setstate__(self, arg0: int) -> None: ...
+ @property
+ def name(self) -> str: ...
+
+class Parameter:
+ @overload
+ def __init__(self, number_of_processes: int, my_ID: int, config_data: Optional[pyfluids.bindings.basics.ConfigurationFile]) -> None: ...
+ @overload
+ def __init__(self, number_of_processes: int, my_ID: int) -> None: ...
+ @overload
+ def __init__(self, config_data: pyfluids.bindings.basics.ConfigurationFile) -> None: ...
+ def add_actuator(self, actuator: PreCollisionInteractor) -> None: ...
+ def add_probe(self, probe: PreCollisionInteractor) -> None: ...
+ def get_SGS_constant(self) -> float: ...
+ def get_density_ratio(self) -> float: ...
+ def get_force_ratio(self) -> float: ...
+ def get_is_body_force(self) -> bool: ...
+ def get_output_path(self) -> str: ...
+ def get_output_prefix(self) -> str: ...
+ def get_velocity(self) -> float: ...
+ def get_velocity_ratio(self) -> float: ...
+ def get_viscosity(self) -> float: ...
+ def get_viscosity_ratio(self) -> float: ...
+ def set_AD_kernel(self, ad_kernel: str) -> None: ...
+ def set_calc_turbulence_intensity(self, calc_velocity_and_fluctuations: bool) -> None: ...
+ def set_comp_on(self, is_comp: bool) -> None: ...
+ def set_density_ratio(self, density_ratio: float) -> None: ...
+ def set_devices(self, devices: List[int]) -> None: ...
+ def set_diff_on(self, is_diff: bool) -> None: ...
+ def set_forcing(self, forcing_x: float, forcing_y: float, forcing_z: float) -> None: ...
+ def set_has_wall_model_monitor(self, has_wall_monitor: bool) -> None: ...
+ def set_initial_condition(self, init_func: Callable[[float,float,float],List[float]]) -> None: ...
+ def set_initial_condition_log_law(self, u_star: float, z0: float, velocity_ratio: float) -> None: ...
+ def set_initial_condition_perturbed_log_law(self, u_star: float, z0: float, length_x: float, length_z: float, height: float, velocity_ratio: float) -> None: ...
+ def set_initial_condition_uniform(self, velocity_x: float, velocity_y: float, velocity_z: float) -> None: ...
+ def set_is_body_force(self, is_body_force: bool) -> None: ...
+ def set_main_kernel(self, kernel: str) -> None: ...
+ def set_max_dev(self, max_dev: int) -> None: ...
+ def set_max_level(self, number_of_levels: int) -> None: ...
+ def set_outflow_pressure_correction_factor(self, correction_factor: float) -> None: ...
+ def set_output_path(self, o_path: str) -> None: ...
+ def set_output_prefix(self, o_prefix: str) -> None: ...
+ def set_print_files(self, print_files: bool) -> None: ...
+ def set_quadric_limiters(self, quadric_limiter_p: float, quadric_limiter_m: float, quadric_limiter_d: float) -> None: ...
+ def set_temperature_BC(self, temp_bc: float) -> None: ...
+ def set_temperature_init(self, temp: float) -> None: ...
+ def set_timestep_end(self, tend: int) -> None: ...
+ def set_timestep_of_coarse_level(self, timestep: int) -> None: ...
+ def set_timestep_out(self, tout: int) -> None: ...
+ def set_timestep_start_out(self, t_start_out: int) -> None: ...
+ def set_use_streams(self, use_streams: bool) -> None: ...
+ def set_velocity_LB(self, velocity: float) -> None: ...
+ def set_velocity_ratio(self, velocity_ratio: float) -> None: ...
+ def set_viscosity_LB(self, viscosity: float) -> None: ...
+ def set_viscosity_ratio(self, viscosity_ratio: float) -> None: ...
+
+class PreCollisionInteractor:
+ def __init__(self, *args, **kwargs) -> None: ...
+
+class PrecursorBC:
+ __members__: ClassVar[dict] = ... # read-only
+ DistributionsPrecursor: ClassVar[PrecursorBC] = ...
+ NotSpecified: ClassVar[PrecursorBC] = ...
+ VelocityPrecursor: ClassVar[PrecursorBC] = ...
+ __entries: ClassVar[dict] = ...
+ def __init__(self, arg0: int) -> None: ...
+ def __eq__(self, arg0: object) -> bool: ...
+ def __getstate__(self) -> int: ...
+ def __hash__(self) -> int: ...
+ def __index__(self) -> int: ...
+ def __int__(self) -> int: ...
+ def __ne__(self, arg0: object) -> bool: ...
+ def __setstate__(self, arg0: int) -> None: ...
+ @property
+ def name(self) -> str: ...
+
+class PrecursorWriter(PreCollisionInteractor):
+ def __init__(self, filename: str, output_path: str, x_pos: float, y_min: float, y_max: float, z_min: float, z_max: float, t_start_out: int, t_save: int, output_variable: OutputVariable, max_timesteps_per_file: int) -> None: ...
+
+class PressureBC:
+ __members__: ClassVar[dict] = ... # read-only
+ NotSpecified: ClassVar[PressureBC] = ...
+ OutflowNonReflective: ClassVar[PressureBC] = ...
+ OutflowNonReflectivePressureCorrection: ClassVar[PressureBC] = ...
+ PressureEquilibrium: ClassVar[PressureBC] = ...
+ PressureEquilibrium2: ClassVar[PressureBC] = ...
+ PressureNonEquilibriumCompressible: ClassVar[PressureBC] = ...
+ PressureNonEquilibriumIncompressible: ClassVar[PressureBC] = ...
+ __entries: ClassVar[dict] = ...
+ def __init__(self, arg0: int) -> None: ...
+ def __eq__(self, arg0: object) -> bool: ...
+ def __getstate__(self) -> int: ...
+ def __hash__(self) -> int: ...
+ def __index__(self) -> int: ...
+ def __int__(self) -> int: ...
+ def __ne__(self, arg0: object) -> bool: ...
+ def __setstate__(self, arg0: int) -> None: ...
+ @property
+ def name(self) -> str: ...
+
+class SideType:
+ __members__: ClassVar[dict] = ... # read-only
+ GEOMETRY: ClassVar[SideType] = ...
+ MX: ClassVar[SideType] = ...
+ MY: ClassVar[SideType] = ...
+ MZ: ClassVar[SideType] = ...
+ PX: ClassVar[SideType] = ...
+ PY: ClassVar[SideType] = ...
+ PZ: ClassVar[SideType] = ...
+ __entries: ClassVar[dict] = ...
+ def __init__(self, arg0: int) -> None: ...
+ def __eq__(self, arg0: object) -> bool: ...
+ def __getstate__(self) -> int: ...
+ def __hash__(self) -> int: ...
+ def __index__(self) -> int: ...
+ def __int__(self) -> int: ...
+ def __ne__(self, arg0: object) -> bool: ...
+ def __setstate__(self, arg0: int) -> None: ...
+ @property
+ def name(self) -> str: ...
+
+class Simulation:
+ @overload
+ def __init__(self, parameter: Parameter, memoryManager: CudaMemoryManager, communicator, gridProvider: GridProvider, bcFactory: BoundaryConditionFactory, gridScalingFactory: GridScalingFactory) -> None: ...
+ @overload
+ def __init__(self, parameter: Parameter, memoryManager: CudaMemoryManager, communicator, gridProvider: GridProvider, bcFactory: BoundaryConditionFactory) -> None: ...
+ @overload
+ def __init__(self, parameter: Parameter, memoryManager: CudaMemoryManager, communicator, gridProvider: GridProvider, bcFactory: BoundaryConditionFactory, tmFactory: TurbulenceModelFactory, gridScalingFactory: GridScalingFactory) -> None: ...
+ def addEnstrophyAnalyzer(self, t_analyse: int) -> None: ...
+ def addKineticEnergyAnalyzer(self, t_analyse: int) -> None: ...
+ def run(self) -> None: ...
+
+class SlipBC:
+ __members__: ClassVar[dict] = ... # read-only
+ NotSpecified: ClassVar[SlipBC] = ...
+ SlipBounceBack: ClassVar[SlipBC] = ...
+ SlipCompressible: ClassVar[SlipBC] = ...
+ SlipCompressibleTurbulentViscosity: ClassVar[SlipBC] = ...
+ SlipIncompressible: ClassVar[SlipBC] = ...
+ SlipPressureCompressibleTurbulentViscosity: ClassVar[SlipBC] = ...
+ __entries: ClassVar[dict] = ...
+ def __init__(self, arg0: int) -> None: ...
+ def __eq__(self, arg0: object) -> bool: ...
+ def __getstate__(self) -> int: ...
+ def __hash__(self) -> int: ...
+ def __index__(self) -> int: ...
+ def __int__(self) -> int: ...
+ def __ne__(self, arg0: object) -> bool: ...
+ def __setstate__(self, arg0: int) -> None: ...
+ @property
+ def name(self) -> str: ...
+
+class StressBC:
+ __members__: ClassVar[dict] = ... # read-only
+ NotSpecified: ClassVar[StressBC] = ...
+ StressBounceBack: ClassVar[StressBC] = ...
+ StressCompressible: ClassVar[StressBC] = ...
+ StressPressureBounceBack: ClassVar[StressBC] = ...
+ __entries: ClassVar[dict] = ...
+ def __init__(self, arg0: int) -> None: ...
+ def __eq__(self, arg0: object) -> bool: ...
+ def __getstate__(self) -> int: ...
+ def __hash__(self) -> int: ...
+ def __index__(self) -> int: ...
+ def __int__(self) -> int: ...
+ def __ne__(self, arg0: object) -> bool: ...
+ def __setstate__(self, arg0: int) -> None: ...
+ @property
+ def name(self) -> str: ...
+
+class TurbulenceModel:
+ __members__: ClassVar[dict] = ... # read-only
+ AMD: ClassVar[TurbulenceModel] = ...
+ NONE: ClassVar[TurbulenceModel] = ...
+ QR: ClassVar[TurbulenceModel] = ...
+ Smagorinsky: ClassVar[TurbulenceModel] = ...
+ __entries: ClassVar[dict] = ...
+ def __init__(self, arg0: int) -> None: ...
+ def __eq__(self, arg0: object) -> bool: ...
+ def __getstate__(self) -> int: ...
+ def __hash__(self) -> int: ...
+ def __index__(self) -> int: ...
+ def __int__(self) -> int: ...
+ def __ne__(self, arg0: object) -> bool: ...
+ def __setstate__(self, arg0: int) -> None: ...
+ @property
+ def name(self) -> str: ...
+
+class TurbulenceModelFactory:
+ def __init__(self, para: Parameter) -> None: ...
+ def read_config_file(self, config_data: pyfluids.bindings.basics.ConfigurationFile) -> None: ...
+ def set_model_constant(self, model_constant: float) -> None: ...
+ def set_turbulence_model(self, turbulence_model: TurbulenceModel) -> None: ...
+
+class VTKFileCollection(FileCollection):
+ def __init__(self, prefix: str) -> None: ...
+
+class VelocityBC:
+ __members__: ClassVar[dict] = ... # read-only
+ NotSpecified: ClassVar[VelocityBC] = ...
+ VelocityAndPressureCompressible: ClassVar[VelocityBC] = ...
+ VelocityCompressible: ClassVar[VelocityBC] = ...
+ VelocityIncompressible: ClassVar[VelocityBC] = ...
+ VelocitySimpleBounceBackCompressible: ClassVar[VelocityBC] = ...
+ __entries: ClassVar[dict] = ...
+ def __init__(self, arg0: int) -> None: ...
+ def __eq__(self, arg0: object) -> bool: ...
+ def __getstate__(self) -> int: ...
+ def __hash__(self) -> int: ...
+ def __index__(self) -> int: ...
+ def __int__(self) -> int: ...
+ def __ne__(self, arg0: object) -> bool: ...
+ def __setstate__(self, arg0: int) -> None: ...
+ @property
+ def name(self) -> str: ...
+
+class FileCollection:
+ def __init__(self, *args, **kwargs) -> None: ...
+
+def create_file_collection(prefix: str, type: FileType) -> FileCollection: ...
diff --git a/pythonbindings/pyfluids-stubs/bindings/gpu/grid_generator.pyi b/pythonbindings/pyfluids-stubs/bindings/gpu/grid_generator.pyi
new file mode 100644
index 0000000000000000000000000000000000000000..8d715e4b4cd49e6dbf92da3aedddbc4b869067c4
--- /dev/null
+++ b/pythonbindings/pyfluids-stubs/bindings/gpu/grid_generator.pyi
@@ -0,0 +1,100 @@
+r"""
+=======================================================================================
+ ____ ____ __ ______ __________ __ __ __ __
+ \ \ | | | | | _ \ |___ ___| | | | | / \ | |
+ \ \ | | | | | |_) | | | | | | | / \ | |
+ \ \ | | | | | _ / | | | | | | / /\ \ | |
+ \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
+ \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
+ \ \ | | ________________________________________________________________
+ \ \ | | | ______________________________________________________________|
+ \ \| | | | __ __ __ __ ______ _______
+ \ | | |_____ | | | | | | | | | _ \ / _____)
+ \ | | _____| | | | | | | | | | | \ \ \_______
+ \ | | | | |_____ | \_/ | | | | |_/ / _____ |
+ \ _____| |__| |________| \_______/ |__| |______/ (_______/
+
+ This file is part of VirtualFluids. VirtualFluids is free software: you can
+ redistribute it and/or modify it under the terms of the GNU General Public
+ License as published by the Free Software Foundation, either version 3 of
+ the License, or (at your option) any later version.
+
+ VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
+ ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
+ for more details.
+
+ You should have received a copy of the GNU General Public License along
+ with VirtualFluids (see COPYING.txt). If not, see <http://www.gnu.org/licenses/>.
+
+! \file grid_generator.pyi
+! \ingroup gpu
+! \author Henry Korb
+=======================================================================================
+"""
+from typing import Any, List
+
+from typing import overload
+import pyfluids.bindings.basics
+import pyfluids.bindings.gpu
+
+class BoundingBox:
+ def __init__(self, min_x: float, max_x: float, min_y: float, max_y: float, min_z: float, max_z: float) -> None: ...
+
+class Conglomerate(Object):
+ def __init__(self, *args, **kwargs) -> None: ...
+ def add(self, object: Object) -> None: ...
+ @staticmethod
+ def make_shared() -> Conglomerate: ...
+ def subtract(self, object: Object) -> None: ...
+
+class Cuboid(Object):
+ def __init__(self, min_x1: float, min_x2: float, min_x3: float, max_x1: float, max_x2: float, max_x3: float) -> None: ...
+
+class GridBuilder:
+ def __init__(self, *args, **kwargs) -> None: ...
+ def get_number_of_grid_levels(self) -> int: ...
+
+class GridFactory:
+ def __init__(self, *args, **kwargs) -> None: ...
+ @staticmethod
+ def make() -> GridFactory: ...
+
+class LevelGridBuilder(GridBuilder):
+ def __init__(self, *args, **kwargs) -> None: ...
+ def set_no_slip_boundary_condition(self, side_type: pyfluids.bindings.gpu.SideType) -> None: ...
+ def set_periodic_boundary_condition(self, periodic_x: bool, periodic_y: bool, periodic_z: bool) -> None: ...
+ def set_precursor_boundary_condition(self, side_type: pyfluids.bindings.gpu.SideType, file_collection: pyfluids.bindings.gpu.VelocityFileCollection, n_t_read: int, velocity_x: float = ..., velocity_y: float = ..., velocity_z: float = ..., file_level_to_grid_level_map: List[int] = ...) -> None: ...
+ def set_pressure_boundary_condition(self, side_type: pyfluids.bindings.gpu.SideType, rho: float) -> None: ...
+ def set_slip_boundary_condition(self, side_type: pyfluids.bindings.gpu.SideType, normal_x: float, normal_y: float, normal_z: float) -> None: ...
+ def set_stress_boundary_condition(self, side_type: pyfluids.bindings.gpu.SideType, normal_x: float, normal_y: float, normal_z: float, sampling_offset: int, z0: float, dx: float) -> None: ...
+ def set_velocity_boundary_condition(self, side_type: pyfluids.bindings.gpu.SideType, vx: float, vy: float, vz: float) -> None: ...
+
+class MultipleGridBuilder(LevelGridBuilder):
+ def __init__(self, *args, **kwargs) -> None: ...
+ def add_coarse_grid(self, start_x: float, start_y: float, start_z: float, end_x: float, end_y: float, end_z: float, delta: float) -> None: ...
+ @overload
+ def add_geometry(self, solid_object: Object) -> None: ...
+ @overload
+ def add_geometry(self, solid_object: Object, level: int) -> None: ...
+ @overload
+ def add_grid(self, grid_shape: Object) -> None: ...
+ @overload
+ def add_grid(self, grid_shape: Object, level_fine: int) -> None: ...
+ def build_grids(self, lbm_or_gks: pyfluids.bindings.basics.LbmOrGks, enable_thin_walls: bool) -> None: ...
+ def get_number_of_levels(self) -> int: ...
+ @staticmethod
+ def make_shared(grid_factory: GridFactory) -> MultipleGridBuilder: ...
+
+class Object:
+ def __init__(self, *args, **kwargs) -> None: ...
+
+class Sphere(Object):
+ def __init__(self, *args, **kwargs) -> None: ...
+ @staticmethod
+ def make_shared() -> Sphere: ...
+
+class TriangularMesh(Object):
+ def __init__(self, *args, **kwargs) -> None: ...
+ @staticmethod
+ def make() -> TriangularMesh: ...
diff --git a/pythonbindings/pyfluids-stubs/bindings/gpu/probes.pyi b/pythonbindings/pyfluids-stubs/bindings/gpu/probes.pyi
new file mode 100644
index 0000000000000000000000000000000000000000..af9c40078e6009efebda4450b5c5e23586aa1e83
--- /dev/null
+++ b/pythonbindings/pyfluids-stubs/bindings/gpu/probes.pyi
@@ -0,0 +1,85 @@
+r"""
+=======================================================================================
+ ____ ____ __ ______ __________ __ __ __ __
+ \ \ | | | | | _ \ |___ ___| | | | | / \ | |
+ \ \ | | | | | |_) | | | | | | | / \ | |
+ \ \ | | | | | _ / | | | | | | / /\ \ | |
+ \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
+ \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
+ \ \ | | ________________________________________________________________
+ \ \ | | | ______________________________________________________________|
+ \ \| | | | __ __ __ __ ______ _______
+ \ | | |_____ | | | | | | | | | _ \ / _____)
+ \ | | _____| | | | | | | | | | | \ \ \_______
+ \ | | | | |_____ | \_/ | | | | |_/ / _____ |
+ \ _____| |__| |________| \_______/ |__| |______/ (_______/
+
+ This file is part of VirtualFluids. VirtualFluids is free software: you can
+ redistribute it and/or modify it under the terms of the GNU General Public
+ License as published by the Free Software Foundation, either version 3 of
+ the License, or (at your option) any later version.
+
+ VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
+ ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
+ for more details.
+
+ You should have received a copy of the GNU General Public License along
+ with VirtualFluids (see COPYING.txt). If not, see <http://www.gnu.org/licenses/>.
+
+! \file probes.pyi
+! \ingroup gpu
+! \author Henry Korb
+=======================================================================================
+"""
+from typing import ClassVar, List
+
+import pyfluids.bindings.gpu
+
+class PlanarAverageProbe(Probe):
+ def __init__(self, probe_name: str, output_path: str, t_start_avg: int, t_start_tmp_avg: int, t_avg: int, t_start_out: int, t_out: int, plane_normal: str) -> None: ...
+
+class PlaneProbe(Probe):
+ def __init__(self, probe_name: str, output_path: str, t_start_avg: int, t_avg: int, t_start_out: int, t_out: int) -> None: ...
+ def set_probe_plane(self, pos_x: float, pos_y: float, pos_z: float, delta_x: float, delta_y: float, delta_z: float) -> None: ...
+
+class PointProbe(Probe):
+ def __init__(self, probe_name: str, output_path: str, t_start_avg: int, t_avg: int, t_start_out: int, t_out: int, output_timeseries: bool) -> None: ...
+ def add_probe_points_from_list(self, point_coords_x: List[float], point_coords_y: List[float], point_coords_z: List[float]) -> None: ...
+ def add_probe_points_from_x_normal_plane(self, pos_x: float, pos0_y: float, pos0_z: float, pos1_y: float, pos1_z: float, n_y: int, n_z: int) -> None: ...
+
+class Probe(pyfluids.bindings.gpu.PreCollisionInteractor):
+ def __init__(self, *args, **kwargs) -> None: ...
+ def add_all_available_statistics(self) -> None: ...
+ def add_statistic(self, variable: Statistic) -> None: ...
+ def set_file_name_to_n_out(self) -> None: ...
+
+class Statistic:
+ __members__: ClassVar[dict] = ... # read-only
+ Instantaneous: ClassVar[Statistic] = ...
+ Means: ClassVar[Statistic] = ...
+ SpatialCovariances: ClassVar[Statistic] = ...
+ SpatialFlatness: ClassVar[Statistic] = ...
+ SpatialMeans: ClassVar[Statistic] = ...
+ SpatialSkewness: ClassVar[Statistic] = ...
+ SpatioTemporalCovariances: ClassVar[Statistic] = ...
+ SpatioTemporalFlatness: ClassVar[Statistic] = ...
+ SpatioTemporalMeans: ClassVar[Statistic] = ...
+ SpatioTemporalSkewness: ClassVar[Statistic] = ...
+ Variances: ClassVar[Statistic] = ...
+ __entries: ClassVar[dict] = ...
+ def __init__(self, arg0: int) -> None: ...
+ def __eq__(self, arg0: object) -> bool: ...
+ def __getstate__(self) -> int: ...
+ def __hash__(self) -> int: ...
+ def __index__(self) -> int: ...
+ def __int__(self) -> int: ...
+ def __ne__(self, arg0: object) -> bool: ...
+ def __setstate__(self, arg0: int) -> None: ...
+ @property
+ def name(self) -> str: ...
+
+class WallModelProbe(Probe):
+ def __init__(self, probe_name: str, output_path: str, t_start_avg: int, t_start_tmp_avg: int, t_avg: int, t_start_out: int, t_out: int) -> None: ...
+ def set_evaluate_pressure_gradient(self, eval_press_grad: bool) -> None: ...
+ def set_force_output_to_stress(self, output_stress: bool) -> None: ...
diff --git a/pythonbindings/pyfluids-stubs/bindings/lbm.pyi b/pythonbindings/pyfluids-stubs/bindings/lbm.pyi
new file mode 100644
index 0000000000000000000000000000000000000000..e69de29bb2d1d6434b8b29ae775ad8c2e48c5391
diff --git a/pythonbindings/pyfluids-stubs/bindings/logger.pyi b/pythonbindings/pyfluids-stubs/bindings/logger.pyi
new file mode 100644
index 0000000000000000000000000000000000000000..fe84eeb18f3245ef72ed023b2de9db7b9131d144
--- /dev/null
+++ b/pythonbindings/pyfluids-stubs/bindings/logger.pyi
@@ -0,0 +1,45 @@
+r"""
+=======================================================================================
+ ____ ____ __ ______ __________ __ __ __ __
+ \ \ | | | | | _ \ |___ ___| | | | | / \ | |
+ \ \ | | | | | |_) | | | | | | | / \ | |
+ \ \ | | | | | _ / | | | | | | / /\ \ | |
+ \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
+ \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
+ \ \ | | ________________________________________________________________
+ \ \ | | | ______________________________________________________________|
+ \ \| | | | __ __ __ __ ______ _______
+ \ | | |_____ | | | | | | | | | _ \ / _____)
+ \ | | _____| | | | | | | | | | | \ \ \_______
+ \ | | | | |_____ | \_/ | | | | |_/ / _____ |
+ \ _____| |__| |________| \_______/ |__| |______/ (_______/
+
+ This file is part of VirtualFluids. VirtualFluids is free software: you can
+ redistribute it and/or modify it under the terms of the GNU General Public
+ License as published by the Free Software Foundation, either version 3 of
+ the License, or (at your option) any later version.
+
+ VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
+ ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
+ for more details.
+
+ You should have received a copy of the GNU General Public License along
+ with VirtualFluids (see COPYING.txt). If not, see <http://www.gnu.org/licenses/>.
+
+! \file logger.pyi
+! \ingroup bindings
+! \author Henry Korb
+=======================================================================================
+"""
+class Logger:
+ @staticmethod
+ def change_log_path(path: str) -> None: ...
+ @staticmethod
+ def initialize_logger() -> None: ...
+
+def vf_log_critical(message: str) -> None: ...
+def vf_log_debug(message: str) -> None: ...
+def vf_log_info(message: str) -> None: ...
+def vf_log_trace(message: str) -> None: ...
+def vf_log_warning(message: str) -> None: ...
diff --git a/pythonbindings/pyfluids/__init__.py b/pythonbindings/pyfluids/__init__.py
new file mode 100644
index 0000000000000000000000000000000000000000..f0537b758267e22a72e5030340de7b87d52f35c3
--- /dev/null
+++ b/pythonbindings/pyfluids/__init__.py
@@ -0,0 +1,54 @@
+r"""
+=======================================================================================
+ ____ ____ __ ______ __________ __ __ __ __
+ \ \ | | | | | _ \ |___ ___| | | | | / \ | |
+ \ \ | | | | | |_) | | | | | | | / \ | |
+ \ \ | | | | | _ / | | | | | | / /\ \ | |
+ \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
+ \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
+ \ \ | | ________________________________________________________________
+ \ \ | | | ______________________________________________________________|
+ \ \| | | | __ __ __ __ ______ _______
+ \ | | |_____ | | | | | | | | | _ \ / _____)
+ \ | | _____| | | | | | | | | | | \ \ \_______
+ \ | | | | |_____ | \_/ | | | | |_/ / _____ |
+ \ _____| |__| |________| \_______/ |__| |______/ (_______/
+
+ This file is part of VirtualFluids. VirtualFluids is free software: you can
+ redistribute it and/or modify it under the terms of the GNU General Public
+ License as published by the Free Software Foundation, either version 3 of
+ the License, or (at your option) any later version.
+
+ VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
+ ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
+ for more details.
+
+ You should have received a copy of the GNU General Public License along
+ with VirtualFluids (see COPYING.txt). If not, see <http://www.gnu.org/licenses/>.
+
+! \file __init__.py
+! \ingroup pyfluids
+! \author Henry Korb
+=======================================================================================
+"""
+try:
+ from .bindings import basics
+except ImportError:
+ print("Basics bindings not included")
+try:
+ from .bindings import logger
+except ImportError:
+ print("Logger bindings not included")
+try:
+ from .bindings import lbm
+except ImportError:
+ print("LBM bindings not included")
+try:
+ from .bindings import gpu
+except ImportError:
+ print("GPU bindings not included")
+try:
+ from .bindings import cpu
+except ImportError:
+ print("CPU bindings not included")
\ No newline at end of file
diff --git a/pythonbindings/pyfluids/py.typed b/pythonbindings/pyfluids/py.typed
new file mode 100644
index 0000000000000000000000000000000000000000..e69de29bb2d1d6434b8b29ae775ad8c2e48c5391
diff --git a/pythonbindings/pymuparser/__init__.py b/pythonbindings/pymuparser/__init__.py
new file mode 100644
index 0000000000000000000000000000000000000000..398069bcba03b3fe710d9d9a6398e9c530b19ee9
--- /dev/null
+++ b/pythonbindings/pymuparser/__init__.py
@@ -0,0 +1,38 @@
+r"""
+=======================================================================================
+ ____ ____ __ ______ __________ __ __ __ __
+ \ \ | | | | | _ \ |___ ___| | | | | / \ | |
+ \ \ | | | | | |_) | | | | | | | / \ | |
+ \ \ | | | | | _ / | | | | | | / /\ \ | |
+ \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
+ \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
+ \ \ | | ________________________________________________________________
+ \ \ | | | ______________________________________________________________|
+ \ \| | | | __ __ __ __ ______ _______
+ \ | | |_____ | | | | | | | | | _ \ / _____)
+ \ | | _____| | | | | | | | | | | \ \ \_______
+ \ | | | | |_____ | \_/ | | | | |_/ / _____ |
+ \ _____| |__| |________| \_______/ |__| |______/ (_______/
+
+ This file is part of VirtualFluids. VirtualFluids is free software: you can
+ redistribute it and/or modify it under the terms of the GNU General Public
+ License as published by the Free Software Foundation, either version 3 of
+ the License, or (at your option) any later version.
+
+ VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
+ ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
+ for more details.
+
+ You should have received a copy of the GNU General Public License along
+ with VirtualFluids (see COPYING.txt). If not, see <http://www.gnu.org/licenses/>.
+
+! \file __init__.pyi
+! \ingroup pymuparser
+! \author Henry Korb
+=======================================================================================
+"""
+try:
+ from .bindings import Parser
+except ImportError as e:
+ raise ImportError("Pymuparser bindings were not built. Only included if VirtualFluids is built with VF_BUILD_CPU=ON.")
\ No newline at end of file
diff --git a/pythonbindings/src/VirtualFluids.cpp b/pythonbindings/src/VirtualFluids.cpp
new file mode 100644
index 0000000000000000000000000000000000000000..20e5012e0af325440e502c704d6f372100306ab1
--- /dev/null
+++ b/pythonbindings/src/VirtualFluids.cpp
@@ -0,0 +1,63 @@
+//=======================================================================================
+// ____ ____ __ ______ __________ __ __ __ __
+// \ \ | | | | | _ \ |___ ___| | | | | / \ | |
+// \ \ | | | | | |_) | | | | | | | / \ | |
+// \ \ | | | | | _ / | | | | | | / /\ \ | |
+// \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
+// \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
+// \ \ | | ________________________________________________________________
+// \ \ | | | ______________________________________________________________|
+// \ \| | | | __ __ __ __ ______ _______
+// \ | | |_____ | | | | | | | | | _ \ / _____)
+// \ | | _____| | | | | | | | | | | \ \ \_______
+// \ | | | | |_____ | \_/ | | | | |_/ / _____ |
+// \ _____| |__| |________| \_______/ |__| |______/ (_______/
+//
+// This file is part of VirtualFluids. VirtualFluids is free software: you can
+// redistribute it and/or modify it under the terms of the GNU General Public
+// License as published by the Free Software Foundation, either version 3 of
+// the License, or (at your option) any later version.
+//
+// VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
+// ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+// FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
+// for more details.
+//
+// You should have received a copy of the GNU General Public License along
+// with VirtualFluids (see COPYING.txt). If not, see <http://www.gnu.org/licenses/>.
+//
+//! \file VirtualFluids.cpp
+//! \ingroup src
+//! \author Henry Korb
+//=======================================================================================
+#include <pybind11/pybind11.h>
+#include "basics/basics.cpp"
+#include "lbm/lbm.cpp"
+#include "logger/logger.cpp"
+
+#ifdef VF_GPU_PYTHONBINDINGS
+#include "gpu/gpu.cpp"
+#endif
+#ifdef VF_CPU_PYTHONBINDINGS
+#include "cpu/cpu.cpp"
+#endif
+
+
+namespace py_bindings
+{
+ namespace py = pybind11;
+
+ PYBIND11_MODULE(bindings, m)
+ {
+ py::add_ostream_redirect(m, "ostream_redirect");
+ basics::makeModule(m);
+ lbm::makeModule(m);
+ logging::makeModule(m);
+#ifdef VF_GPU_PYTHONBINDINGS
+ gpu::makeModule(m);
+#endif
+#ifdef VF_CPU_PYTHONBINDINGS
+ cpu::makeModule(m);
+#endif
+ }
+}
\ No newline at end of file
diff --git a/pythonbindings/src/VirtualFluidsModulesCPU.cpp b/pythonbindings/src/VirtualFluidsModulesCPU.cpp
deleted file mode 100644
index 2fba3da494f568f7d0d0a117a579a45c9c1b9245..0000000000000000000000000000000000000000
--- a/pythonbindings/src/VirtualFluidsModulesCPU.cpp
+++ /dev/null
@@ -1,12 +0,0 @@
-#include <pybind11/pybind11.h>
-#include "cpu/cpu.cpp"
-
-namespace py_bindings
-{
- namespace py = pybind11;
-
- PYBIND11_MODULE(pyfluids, m)
- {
- cpu::makeModule(m);
- }
-}
\ No newline at end of file
diff --git a/pythonbindings/src/VirtualFluidsModulesGPU.cpp b/pythonbindings/src/VirtualFluidsModulesGPU.cpp
deleted file mode 100644
index b96971caf381faada76ee676cf60469492d055c2..0000000000000000000000000000000000000000
--- a/pythonbindings/src/VirtualFluidsModulesGPU.cpp
+++ /dev/null
@@ -1,19 +0,0 @@
-#include <pybind11/pybind11.h>
-#include "basics/basics.cpp"
-#include "lbm/lbm.cpp"
-#include "gpu/gpu.cpp"
-#include "logger/logger.cpp"
-
-namespace py_bindings
-{
- namespace py = pybind11;
-
- PYBIND11_MODULE(pyfluids, m)
- {
- basics::makeModule(m);
- gpu::makeModule(m);
- lbm::makeModule(m);
- logging::makeModule(m);
- py::add_ostream_redirect(m, "ostream_redirect");
- }
-}
\ No newline at end of file
diff --git a/pythonbindings/src/basics/basics.cpp b/pythonbindings/src/basics/basics.cpp
index 381e345d78226b25ec3a77a14340d2ef1171c8c9..e67dfb05308511c8bf79d7e860299f062f317194 100644
--- a/pythonbindings/src/basics/basics.cpp
+++ b/pythonbindings/src/basics/basics.cpp
@@ -1,3 +1,35 @@
+//=======================================================================================
+// ____ ____ __ ______ __________ __ __ __ __
+// \ \ | | | | | _ \ |___ ___| | | | | / \ | |
+// \ \ | | | | | |_) | | | | | | | / \ | |
+// \ \ | | | | | _ / | | | | | | / /\ \ | |
+// \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
+// \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
+// \ \ | | ________________________________________________________________
+// \ \ | | | ______________________________________________________________|
+// \ \| | | | __ __ __ __ ______ _______
+// \ | | |_____ | | | | | | | | | _ \ / _____)
+// \ | | _____| | | | | | | | | | | \ \ \_______
+// \ | | | | |_____ | \_/ | | | | |_/ / _____ |
+// \ _____| |__| |________| \_______/ |__| |______/ (_______/
+//
+// This file is part of VirtualFluids. VirtualFluids is free software: you can
+// redistribute it and/or modify it under the terms of the GNU General Public
+// License as published by the Free Software Foundation, either version 3 of
+// the License, or (at your option) any later version.
+//
+// VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
+// ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+// FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
+// for more details.
+//
+// You should have received a copy of the GNU General Public License along
+// with VirtualFluids (see COPYING.txt). If not, see <http://www.gnu.org/licenses/>.
+//
+//! \file basics.cpp
+//! \ingroup basics
+//! \author Henry Korb
+//=======================================================================================
#include <pybind11/pybind11.h>
#include "submodules/logger.cpp"
#include "submodules/configuration_file.cpp"
diff --git a/pythonbindings/src/basics/submodules/configuration_file.cpp b/pythonbindings/src/basics/submodules/configuration_file.cpp
index f5a2f87135a17f5eda34a7467d95f9db6b1c21d1..7fcd48c34824b9370eeac1872c899bf980176a52 100644
--- a/pythonbindings/src/basics/submodules/configuration_file.cpp
+++ b/pythonbindings/src/basics/submodules/configuration_file.cpp
@@ -1,5 +1,37 @@
+//=======================================================================================
+// ____ ____ __ ______ __________ __ __ __ __
+// \ \ | | | | | _ \ |___ ___| | | | | / \ | |
+// \ \ | | | | | |_) | | | | | | | / \ | |
+// \ \ | | | | | _ / | | | | | | / /\ \ | |
+// \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
+// \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
+// \ \ | | ________________________________________________________________
+// \ \ | | | ______________________________________________________________|
+// \ \| | | | __ __ __ __ ______ _______
+// \ | | |_____ | | | | | | | | | _ \ / _____)
+// \ | | _____| | | | | | | | | | | \ \ \_______
+// \ | | | | |_____ | \_/ | | | | |_/ / _____ |
+// \ _____| |__| |________| \_______/ |__| |______/ (_______/
+//
+// This file is part of VirtualFluids. VirtualFluids is free software: you can
+// redistribute it and/or modify it under the terms of the GNU General Public
+// License as published by the Free Software Foundation, either version 3 of
+// the License, or (at your option) any later version.
+//
+// VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
+// ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+// FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
+// for more details.
+//
+// You should have received a copy of the GNU General Public License along
+// with VirtualFluids (see COPYING.txt). If not, see <http://www.gnu.org/licenses/>.
+//
+//! \file configuration_file.cpp
+//! \ingroup submodules
+//! \author Henry Korb
+//=======================================================================================
#include <pybind11/pybind11.h>
-#include <basics/config/ConfigurationFile.h>
+#include "basics/config/ConfigurationFile.h"
namespace configuration
{
@@ -9,6 +41,19 @@ namespace configuration
{
py::class_<vf::basics::ConfigurationFile>(parentModule, "ConfigurationFile")
.def(py::init<>())
- .def("load", &vf::basics::ConfigurationFile::load);
+ .def("load", &vf::basics::ConfigurationFile::load, py::arg("file"))
+ .def("contains", &vf::basics::ConfigurationFile::contains, py::arg("key"))
+ .def("get_int_value" , static_cast<int (vf::basics::ConfigurationFile::*)(const std::string&) const>(&vf::basics::ConfigurationFile::getValue), py::arg("key"))
+ .def("get_int_value" , static_cast<int (vf::basics::ConfigurationFile::*)(const std::string&, int ) const>(&vf::basics::ConfigurationFile::getValue), py::arg("key"), py::arg("default_value"))
+ .def("get_uint_value" , static_cast<uint (vf::basics::ConfigurationFile::*)(const std::string&) const>(&vf::basics::ConfigurationFile::getValue), py::arg("key"))
+ .def("get_uint_value" , static_cast<uint (vf::basics::ConfigurationFile::*)(const std::string&, uint ) const>(&vf::basics::ConfigurationFile::getValue), py::arg("key"), py::arg("default_value"))
+ .def("get_float_value" , static_cast<float (vf::basics::ConfigurationFile::*)(const std::string&) const>(&vf::basics::ConfigurationFile::getValue), py::arg("key"))
+ .def("get_float_value" , static_cast<float (vf::basics::ConfigurationFile::*)(const std::string&, float ) const>(&vf::basics::ConfigurationFile::getValue), py::arg("key"), py::arg("default_value"))
+ .def("get_double_value", static_cast<double (vf::basics::ConfigurationFile::*)(const std::string&) const>(&vf::basics::ConfigurationFile::getValue), py::arg("key"))
+ .def("get_double_value", static_cast<double (vf::basics::ConfigurationFile::*)(const std::string&, double ) const>(&vf::basics::ConfigurationFile::getValue), py::arg("key"), py::arg("default_value"))
+ .def("get_bool_value" , static_cast<bool (vf::basics::ConfigurationFile::*)(const std::string&) const>(&vf::basics::ConfigurationFile::getValue), py::arg("key"))
+ .def("get_bool_value" , static_cast<bool (vf::basics::ConfigurationFile::*)(const std::string&, bool ) const>(&vf::basics::ConfigurationFile::getValue), py::arg("key"), py::arg("default_value"))
+ .def("get_string_value", static_cast<std::string (vf::basics::ConfigurationFile::*)(const std::string&) const>(&vf::basics::ConfigurationFile::getValue), py::arg("key"))
+ .def("get_string_value", static_cast<std::string (vf::basics::ConfigurationFile::*)(const std::string&, std::string) const>(&vf::basics::ConfigurationFile::getValue), py::arg("key"), py::arg("default_value"));
}
}
\ No newline at end of file
diff --git a/pythonbindings/src/basics/submodules/lbm_or_gks.cpp b/pythonbindings/src/basics/submodules/lbm_or_gks.cpp
index ed1deeca62fc57b7f44499b306e9f99b7f990604..d20cf2d1f631f6d36a80c36f1fb6c9c59d192090 100644
--- a/pythonbindings/src/basics/submodules/lbm_or_gks.cpp
+++ b/pythonbindings/src/basics/submodules/lbm_or_gks.cpp
@@ -1,3 +1,35 @@
+//=======================================================================================
+// ____ ____ __ ______ __________ __ __ __ __
+// \ \ | | | | | _ \ |___ ___| | | | | / \ | |
+// \ \ | | | | | |_) | | | | | | | / \ | |
+// \ \ | | | | | _ / | | | | | | / /\ \ | |
+// \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
+// \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
+// \ \ | | ________________________________________________________________
+// \ \ | | | ______________________________________________________________|
+// \ \| | | | __ __ __ __ ______ _______
+// \ | | |_____ | | | | | | | | | _ \ / _____)
+// \ | | _____| | | | | | | | | | | \ \ \_______
+// \ | | | | |_____ | \_/ | | | | |_/ / _____ |
+// \ _____| |__| |________| \_______/ |__| |______/ (_______/
+//
+// This file is part of VirtualFluids. VirtualFluids is free software: you can
+// redistribute it and/or modify it under the terms of the GNU General Public
+// License as published by the Free Software Foundation, either version 3 of
+// the License, or (at your option) any later version.
+//
+// VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
+// ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+// FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
+// for more details.
+//
+// You should have received a copy of the GNU General Public License along
+// with VirtualFluids (see COPYING.txt). If not, see <http://www.gnu.org/licenses/>.
+//
+//! \file lbm_or_gks.cpp
+//! \ingroup submodules
+//! \author Henry Korb
+//=======================================================================================
#include <pybind11/pybind11.h>
#include "basics/Core/LbmOrGks.h"
diff --git a/pythonbindings/src/basics/submodules/logger.cpp b/pythonbindings/src/basics/submodules/logger.cpp
index d46648e349b44243581e083f3561e8a13648f3b2..fa7e00e4dca06581b7a14d2bcf2628ed6af60001 100644
--- a/pythonbindings/src/basics/submodules/logger.cpp
+++ b/pythonbindings/src/basics/submodules/logger.cpp
@@ -1,3 +1,35 @@
+//=======================================================================================
+// ____ ____ __ ______ __________ __ __ __ __
+// \ \ | | | | | _ \ |___ ___| | | | | / \ | |
+// \ \ | | | | | |_) | | | | | | | / \ | |
+// \ \ | | | | | _ / | | | | | | / /\ \ | |
+// \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
+// \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
+// \ \ | | ________________________________________________________________
+// \ \ | | | ______________________________________________________________|
+// \ \| | | | __ __ __ __ ______ _______
+// \ | | |_____ | | | | | | | | | _ \ / _____)
+// \ | | _____| | | | | | | | | | | \ \ \_______
+// \ | | | | |_____ | \_/ | | | | |_/ / _____ |
+// \ _____| |__| |________| \_______/ |__| |______/ (_______/
+//
+// This file is part of VirtualFluids. VirtualFluids is free software: you can
+// redistribute it and/or modify it under the terms of the GNU General Public
+// License as published by the Free Software Foundation, either version 3 of
+// the License, or (at your option) any later version.
+//
+// VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
+// ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+// FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
+// for more details.
+//
+// You should have received a copy of the GNU General Public License along
+// with VirtualFluids (see COPYING.txt). If not, see <http://www.gnu.org/licenses/>.
+//
+//! \file logger.cpp
+//! \ingroup submodules
+//! \author Henry Korb
+//=======================================================================================
#include <pybind11/pybind11.h>
#include <pybind11/iostream.h>
#include <basics/Core/Logger/Logger.h>
@@ -12,12 +44,12 @@ namespace logger
py::module loggerModule = parentModule.def_submodule("logger");
py::class_<logging::Logger>(loggerModule, "Logger")
- .def("add_stdout", [](){
+ .def_static("add_stdout", [](){
logging::Logger::addStream(&std::cout);
})
- .def("set_debug_level", &logging::Logger::setDebugLevel)
- .def("time_stamp", &logging::Logger::timeStamp)
- .def("enable_printed_rank_numbers", &logging::Logger::enablePrintedRankNumbers);
+ .def_static("set_debug_level", &logging::Logger::setDebugLevel)
+ .def_static("time_stamp", &logging::Logger::timeStamp, py::arg("time_stamp"))
+ .def_static("enable_printed_rank_numbers", &logging::Logger::enablePrintedRankNumbers, py::arg("print"));
loggerModule.attr("log") = logging::out;
py::enum_<logging::Logger::Level>(loggerModule, "Level")
diff --git a/pythonbindings/src/cpu/cpu.cpp b/pythonbindings/src/cpu/cpu.cpp
index 554de53b47446366693aed31d534f6145ebea8ba..75143d913596c74a26f25ce64f1e6d214a442e34 100644
--- a/pythonbindings/src/cpu/cpu.cpp
+++ b/pythonbindings/src/cpu/cpu.cpp
@@ -1,3 +1,35 @@
+//=======================================================================================
+// ____ ____ __ ______ __________ __ __ __ __
+// \ \ | | | | | _ \ |___ ___| | | | | / \ | |
+// \ \ | | | | | |_) | | | | | | | / \ | |
+// \ \ | | | | | _ / | | | | | | / /\ \ | |
+// \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
+// \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
+// \ \ | | ________________________________________________________________
+// \ \ | | | ______________________________________________________________|
+// \ \| | | | __ __ __ __ ______ _______
+// \ | | |_____ | | | | | | | | | _ \ / _____)
+// \ | | _____| | | | | | | | | | | \ \ \_______
+// \ | | | | |_____ | \_/ | | | | |_/ / _____ |
+// \ _____| |__| |________| \_______/ |__| |______/ (_______/
+//
+// This file is part of VirtualFluids. VirtualFluids is free software: you can
+// redistribute it and/or modify it under the terms of the GNU General Public
+// License as published by the Free Software Foundation, either version 3 of
+// the License, or (at your option) any later version.
+//
+// VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
+// ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+// FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
+// for more details.
+//
+// You should have received a copy of the GNU General Public License along
+// with VirtualFluids (see COPYING.txt). If not, see <http://www.gnu.org/licenses/>.
+//
+//! \file cpu.cpp
+//! \ingroup cpu
+//! \author Sven Marcus, Henry Korb
+//=======================================================================================
#include <pybind11/pybind11.h>
#include "submodules/boundaryconditions.cpp"
#include "submodules/simulationconfig.cpp"
diff --git a/pythonbindings/src/cpu/submodules/boundaryconditions.cpp b/pythonbindings/src/cpu/submodules/boundaryconditions.cpp
index 3bff7bc069ca20fe1c0cf3d1847b9714e0381505..ac9ec8605dec51e8374c850b1c1b58314674c426 100644
--- a/pythonbindings/src/cpu/submodules/boundaryconditions.cpp
+++ b/pythonbindings/src/cpu/submodules/boundaryconditions.cpp
@@ -1,3 +1,35 @@
+//=======================================================================================
+// ____ ____ __ ______ __________ __ __ __ __
+// \ \ | | | | | _ \ |___ ___| | | | | / \ | |
+// \ \ | | | | | |_) | | | | | | | / \ | |
+// \ \ | | | | | _ / | | | | | | / /\ \ | |
+// \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
+// \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
+// \ \ | | ________________________________________________________________
+// \ \ | | | ______________________________________________________________|
+// \ \| | | | __ __ __ __ ______ _______
+// \ | | |_____ | | | | | | | | | _ \ / _____)
+// \ | | _____| | | | | | | | | | | \ \ \_______
+// \ | | | | |_____ | \_/ | | | | |_/ / _____ |
+// \ _____| |__| |________| \_______/ |__| |______/ (_______/
+//
+// This file is part of VirtualFluids. VirtualFluids is free software: you can
+// redistribute it and/or modify it under the terms of the GNU General Public
+// License as published by the Free Software Foundation, either version 3 of
+// the License, or (at your option) any later version.
+//
+// VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
+// ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+// FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
+// for more details.
+//
+// You should have received a copy of the GNU General Public License along
+// with VirtualFluids (see COPYING.txt). If not, see <http://www.gnu.org/licenses/>.
+//
+//! \file boundaryconditions.cpp
+//! \ingroup submodules
+//! \author Sven Marcus, Henry Korb
+//=======================================================================================
#include <pybind11/pybind11.h>
#include <pybind11/stl.h>
#include <BoundaryConditions/DensityBCAdapter.h>
diff --git a/pythonbindings/src/cpu/submodules/geometry.cpp b/pythonbindings/src/cpu/submodules/geometry.cpp
index b7ff4dd761258d41687589d2dd89c3479093753e..4c4c47b002b9c7451a8d788ba82c4a19b78ca96f 100644
--- a/pythonbindings/src/cpu/submodules/geometry.cpp
+++ b/pythonbindings/src/cpu/submodules/geometry.cpp
@@ -1,3 +1,35 @@
+//=======================================================================================
+// ____ ____ __ ______ __________ __ __ __ __
+// \ \ | | | | | _ \ |___ ___| | | | | / \ | |
+// \ \ | | | | | |_) | | | | | | | / \ | |
+// \ \ | | | | | _ / | | | | | | / /\ \ | |
+// \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
+// \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
+// \ \ | | ________________________________________________________________
+// \ \ | | | ______________________________________________________________|
+// \ \| | | | __ __ __ __ ______ _______
+// \ | | |_____ | | | | | | | | | _ \ / _____)
+// \ | | _____| | | | | | | | | | | \ \ \_______
+// \ | | | | |_____ | \_/ | | | | |_/ / _____ |
+// \ _____| |__| |________| \_______/ |__| |______/ (_______/
+//
+// This file is part of VirtualFluids. VirtualFluids is free software: you can
+// redistribute it and/or modify it under the terms of the GNU General Public
+// License as published by the Free Software Foundation, either version 3 of
+// the License, or (at your option) any later version.
+//
+// VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
+// ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+// FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
+// for more details.
+//
+// You should have received a copy of the GNU General Public License along
+// with VirtualFluids (see COPYING.txt). If not, see <http://www.gnu.org/licenses/>.
+//
+//! \file geometry.cpp
+//! \ingroup submodules
+//! \author Sven Marcus, Henry Korb
+//=======================================================================================
#include <pybind11/pybind11.h>
#include <geometry3d/GbPoint3D.h>
#include <geometry3d/GbObject3D.h>
diff --git a/pythonbindings/src/cpu/submodules/kernel.cpp b/pythonbindings/src/cpu/submodules/kernel.cpp
index fb291790632cc2041410f60a14fca8d966283343..b00d86579540a299e4bf3ed47bc09d4386f420a2 100644
--- a/pythonbindings/src/cpu/submodules/kernel.cpp
+++ b/pythonbindings/src/cpu/submodules/kernel.cpp
@@ -1,3 +1,35 @@
+//=======================================================================================
+// ____ ____ __ ______ __________ __ __ __ __
+// \ \ | | | | | _ \ |___ ___| | | | | / \ | |
+// \ \ | | | | | |_) | | | | | | | / \ | |
+// \ \ | | | | | _ / | | | | | | / /\ \ | |
+// \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
+// \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
+// \ \ | | ________________________________________________________________
+// \ \ | | | ______________________________________________________________|
+// \ \| | | | __ __ __ __ ______ _______
+// \ | | |_____ | | | | | | | | | _ \ / _____)
+// \ | | _____| | | | | | | | | | | \ \ \_______
+// \ | | | | |_____ | \_/ | | | | |_/ / _____ |
+// \ _____| |__| |________| \_______/ |__| |______/ (_______/
+//
+// This file is part of VirtualFluids. VirtualFluids is free software: you can
+// redistribute it and/or modify it under the terms of the GNU General Public
+// License as published by the Free Software Foundation, either version 3 of
+// the License, or (at your option) any later version.
+//
+// VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
+// ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+// FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
+// for more details.
+//
+// You should have received a copy of the GNU General Public License along
+// with VirtualFluids (see COPYING.txt). If not, see <http://www.gnu.org/licenses/>.
+//
+//! \file kernel.cpp
+//! \ingroup submodules
+//! \author Sven Marcus, Henry Korb
+//=======================================================================================
#include <memory>
#include <pybind11/pybind11.h>
#include <simulationconfig/KernelFactory.h>
diff --git a/pythonbindings/src/cpu/submodules/simulationconfig.cpp b/pythonbindings/src/cpu/submodules/simulationconfig.cpp
index 60af4e36af4dca67e9262dd9f5ee1f46d5b7bb58..09d91f44e85f03c6150c56ce5762e7629212fba0 100644
--- a/pythonbindings/src/cpu/submodules/simulationconfig.cpp
+++ b/pythonbindings/src/cpu/submodules/simulationconfig.cpp
@@ -1,3 +1,35 @@
+//=======================================================================================
+// ____ ____ __ ______ __________ __ __ __ __
+// \ \ | | | | | _ \ |___ ___| | | | | / \ | |
+// \ \ | | | | | |_) | | | | | | | / \ | |
+// \ \ | | | | | _ / | | | | | | / /\ \ | |
+// \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
+// \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
+// \ \ | | ________________________________________________________________
+// \ \ | | | ______________________________________________________________|
+// \ \| | | | __ __ __ __ ______ _______
+// \ | | |_____ | | | | | | | | | _ \ / _____)
+// \ | | _____| | | | | | | | | | | \ \ \_______
+// \ | | | | |_____ | \_/ | | | | |_/ / _____ |
+// \ _____| |__| |________| \_______/ |__| |______/ (_______/
+//
+// This file is part of VirtualFluids. VirtualFluids is free software: you can
+// redistribute it and/or modify it under the terms of the GNU General Public
+// License as published by the Free Software Foundation, either version 3 of
+// the License, or (at your option) any later version.
+//
+// VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
+// ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+// FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
+// for more details.
+//
+// You should have received a copy of the GNU General Public License along
+// with VirtualFluids (see COPYING.txt). If not, see <http://www.gnu.org/licenses/>.
+//
+//! \file simulationconfig.cpp
+//! \ingroup submodules
+//! \author Sven Marcus, Henry Korb
+//=======================================================================================
#include <pybind11/pybind11.h>
#include <simulationconfig/Simulation.h>
diff --git a/pythonbindings/src/cpu/submodules/simulationparameters.cpp b/pythonbindings/src/cpu/submodules/simulationparameters.cpp
index acc272f2ee412cfbafd9007b4b18610cfd0a1e9b..b33d20f9e5d335a0ed381faf8786d88cc7642738 100644
--- a/pythonbindings/src/cpu/submodules/simulationparameters.cpp
+++ b/pythonbindings/src/cpu/submodules/simulationparameters.cpp
@@ -1,3 +1,35 @@
+//=======================================================================================
+// ____ ____ __ ______ __________ __ __ __ __
+// \ \ | | | | | _ \ |___ ___| | | | | / \ | |
+// \ \ | | | | | |_) | | | | | | | / \ | |
+// \ \ | | | | | _ / | | | | | | / /\ \ | |
+// \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
+// \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
+// \ \ | | ________________________________________________________________
+// \ \ | | | ______________________________________________________________|
+// \ \| | | | __ __ __ __ ______ _______
+// \ | | |_____ | | | | | | | | | _ \ / _____)
+// \ | | _____| | | | | | | | | | | \ \ \_______
+// \ | | | | |_____ | \_/ | | | | |_/ / _____ |
+// \ _____| |__| |________| \_______/ |__| |______/ (_______/
+//
+// This file is part of VirtualFluids. VirtualFluids is free software: you can
+// redistribute it and/or modify it under the terms of the GNU General Public
+// License as published by the Free Software Foundation, either version 3 of
+// the License, or (at your option) any later version.
+//
+// VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
+// ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+// FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
+// for more details.
+//
+// You should have received a copy of the GNU General Public License along
+// with VirtualFluids (see COPYING.txt). If not, see <http://www.gnu.org/licenses/>.
+//
+//! \file simulationparameters.cpp
+//! \ingroup submodules
+//! \author Sven Marcus, Henry Korb
+//=======================================================================================
#include <pybind11/pybind11.h>
#include <pybind11/stl.h>
#include <complex>
diff --git a/pythonbindings/src/cpu/submodules/writer.cpp b/pythonbindings/src/cpu/submodules/writer.cpp
index d5ec527a27caf63d9a3066c51e1f675b307fe0b2..f1cfd8934c2da84266a93d5bcd91eb26f5f69d3f 100644
--- a/pythonbindings/src/cpu/submodules/writer.cpp
+++ b/pythonbindings/src/cpu/submodules/writer.cpp
@@ -1,3 +1,35 @@
+//=======================================================================================
+// ____ ____ __ ______ __________ __ __ __ __
+// \ \ | | | | | _ \ |___ ___| | | | | / \ | |
+// \ \ | | | | | |_) | | | | | | | / \ | |
+// \ \ | | | | | _ / | | | | | | / /\ \ | |
+// \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
+// \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
+// \ \ | | ________________________________________________________________
+// \ \ | | | ______________________________________________________________|
+// \ \| | | | __ __ __ __ ______ _______
+// \ | | |_____ | | | | | | | | | _ \ / _____)
+// \ | | _____| | | | | | | | | | | \ \ \_______
+// \ | | | | |_____ | \_/ | | | | |_/ / _____ |
+// \ _____| |__| |________| \_______/ |__| |______/ (_______/
+//
+// This file is part of VirtualFluids. VirtualFluids is free software: you can
+// redistribute it and/or modify it under the terms of the GNU General Public
+// License as published by the Free Software Foundation, either version 3 of
+// the License, or (at your option) any later version.
+//
+// VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
+// ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+// FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
+// for more details.
+//
+// You should have received a copy of the GNU General Public License along
+// with VirtualFluids (see COPYING.txt). If not, see <http://www.gnu.org/licenses/>.
+//
+//! \file writer.cpp
+//! \ingroup submodules
+//! \author Sven Marcus, Henry Korb
+//=======================================================================================
#include <pybind11/pybind11.h>
#include <simulationconfig/WriterConfiguration.h>
diff --git a/pythonbindings/src/gpu/gpu.cpp b/pythonbindings/src/gpu/gpu.cpp
index dc110cd5e19a9aad4937f9c2133ddf74c0ddf9bf..9eb160ae7765f16a6437e343cb878bb4b80877bf 100644
--- a/pythonbindings/src/gpu/gpu.cpp
+++ b/pythonbindings/src/gpu/gpu.cpp
@@ -1,14 +1,50 @@
+//=======================================================================================
+// ____ ____ __ ______ __________ __ __ __ __
+// \ \ | | | | | _ \ |___ ___| | | | | / \ | |
+// \ \ | | | | | |_) | | | | | | | / \ | |
+// \ \ | | | | | _ / | | | | | | / /\ \ | |
+// \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
+// \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
+// \ \ | | ________________________________________________________________
+// \ \ | | | ______________________________________________________________|
+// \ \| | | | __ __ __ __ ______ _______
+// \ | | |_____ | | | | | | | | | _ \ / _____)
+// \ | | _____| | | | | | | | | | | \ \ \_______
+// \ | | | | |_____ | \_/ | | | | |_/ / _____ |
+// \ _____| |__| |________| \_______/ |__| |______/ (_______/
+//
+// This file is part of VirtualFluids. VirtualFluids is free software: you can
+// redistribute it and/or modify it under the terms of the GNU General Public
+// License as published by the Free Software Foundation, either version 3 of
+// the License, or (at your option) any later version.
+//
+// VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
+// ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+// FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
+// for more details.
+//
+// You should have received a copy of the GNU General Public License along
+// with VirtualFluids (see COPYING.txt). If not, see <http://www.gnu.org/licenses/>.
+//
+//! \file gpu.cpp
+//! \ingroup gpu
+//! \author Henry Korb
+//=======================================================================================
#include <pybind11/pybind11.h>
-#include "submodules/actuator_line.cpp"
#include "submodules/pre_collision_interactor.cpp"
#include "submodules/simulation.cpp"
#include "submodules/parameter.cpp"
#include "submodules/boundary_conditions.cpp"
#include "submodules/communicator.cpp"
#include "submodules/cuda_memory_manager.cpp"
+#include "submodules/probes.cpp"
+#include "submodules/precursor_writer.cpp"
#include "submodules/grid_provider.cpp"
#include "submodules/grid_generator.cpp"
-#include "submodules/probes.cpp"
+#include "submodules/turbulence_models.cpp"
+#include "submodules/transient_bc_setter.cpp"
+#include "submodules/actuator_farm.cpp"
+#include "submodules/grid_scaling_factory.cpp"
namespace gpu
{
@@ -20,13 +56,17 @@ namespace gpu
simulation::makeModule(gpuModule);
parameter::makeModule(gpuModule);
pre_collision_interactor::makeModule(gpuModule);
- actuator_line::makeModule(gpuModule);
+ actuator_farm::makeModule(gpuModule);
boundary_conditions::makeModule(gpuModule);
+ transient_bc_setter::makeModule(gpuModule);
communicator::makeModule(gpuModule);
cuda_memory_manager::makeModule(gpuModule);
- grid_provider::makeModule(gpuModule);
probes::makeModule(gpuModule);
+ precursor_writer::makeModule(gpuModule);
grid_generator::makeModule(gpuModule);
+ grid_provider::makeModule(gpuModule);
+ turbulence_model::makeModule(gpuModule);
+ grid_scaling_factory::makeModule(gpuModule);
return gpuModule;
}
}
\ No newline at end of file
diff --git a/pythonbindings/src/gpu/submodules/actuator_farm.cpp b/pythonbindings/src/gpu/submodules/actuator_farm.cpp
new file mode 100644
index 0000000000000000000000000000000000000000..a930616db3e0d0713bdf57157387d75d171603de
--- /dev/null
+++ b/pythonbindings/src/gpu/submodules/actuator_farm.cpp
@@ -0,0 +1,171 @@
+//=======================================================================================
+// ____ ____ __ ______ __________ __ __ __ __
+// \ \ | | | | | _ \ |___ ___| | | | | / \ | |
+// \ \ | | | | | |_) | | | | | | | / \ | |
+// \ \ | | | | | _ / | | | | | | / /\ \ | |
+// \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
+// \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
+// \ \ | | ________________________________________________________________
+// \ \ | | | ______________________________________________________________|
+// \ \| | | | __ __ __ __ ______ _______
+// \ | | |_____ | | | | | | | | | _ \ / _____)
+// \ | | _____| | | | | | | | | | | \ \ \_______
+// \ | | | | |_____ | \_/ | | | | |_/ / _____ |
+// \ _____| |__| |________| \_______/ |__| |______/ (_______/
+//
+// This file is part of VirtualFluids. VirtualFluids is free software: you can
+// redistribute it and/or modify it under the terms of the GNU General Public
+// License as published by the Free Software Foundation, either version 3 of
+// the License, or (at your option) any later version.
+//
+// VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
+// ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+// FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
+// for more details.
+//
+// You should have received a copy of the GNU General Public License along
+// with VirtualFluids (see COPYING.txt). If not, see <http://www.gnu.org/licenses/>.
+//
+//! \file actuator_farm.cpp
+//! \ingroup submodules
+//! \author Henry Korb
+//=======================================================================================
+#include <pybind11/pybind11.h>
+#include <pybind11/numpy.h>
+#include <gpu/VirtualFluids_GPU/PreCollisionInteractor/ActuatorFarm.h>
+#include <gpu/VirtualFluids_GPU/PreCollisionInteractor/PreCollisionInteractor.h>
+class PyActuatorFarm : public ActuatorFarm
+{
+public:
+ using ActuatorFarm::ActuatorFarm; // Inherit constructors
+ void calcBladeForces() override
+ {
+ PYBIND11_OVERRIDE_NAME(void, ActuatorFarm, "calc_blade_forces", calcBladeForces);
+ }
+};
+namespace actuator_farm
+{
+ namespace py = pybind11;
+
+ void makeModule(py::module_ &parentModule)
+ {
+ using arr = py::array_t<float, py::array::c_style>;
+
+ py::class_<ActuatorFarm, PreCollisionInteractor, PyActuatorFarm, std::shared_ptr<ActuatorFarm>>(parentModule, "ActuatorFarm", py::dynamic_attr())
+ .def(py::init< const uint,
+ const real,
+ const uint,
+ const real,
+ int,
+ const real,
+ const real,
+ const bool>(),
+ py::arg("number_of_blades_per_turbine"),
+ py::arg("density"),
+ py::arg("number_of_nodes_per_blade"),
+ py::arg("epsilon"),
+ py::arg("level"),
+ py::arg("delta_t"),
+ py::arg("delta_x"),
+ py::arg("use_host_arrays"))
+ .def_property_readonly("number_of_turbines", &ActuatorFarm::getNumberOfTurbines)
+ .def_property_readonly("number_of_nodes_per_blade", &ActuatorFarm::getNumberOfNodesPerBlade)
+ .def_property_readonly("number_of_blades_per_turbine", &ActuatorFarm::getNumberOfBladesPerTurbine)
+ .def_property_readonly("number_of_nodes", &ActuatorFarm::getNumberOfNodes)
+ .def_property_readonly("number_of_indices", &ActuatorFarm::getNumberOfIndices)
+ .def_property_readonly("density", &ActuatorFarm::getDensity)
+ .def_property_readonly("delta_t", &ActuatorFarm::getDeltaT)
+ .def_property_readonly("delta_x", &ActuatorFarm::getDeltaX)
+
+ .def("add_turbine", &ActuatorFarm::addTurbine, py::arg("posX"), py::arg("posY"), py::arg("posZ"), py::arg("diameter"), py::arg("omega"), py::arg("azimuth"), py::arg("yaw"), py::arg("bladeRadii"))
+
+ .def("get_turbine_pos", [](ActuatorFarm& al, uint turbine){ real position[3] = {al.getTurbinePosX(turbine), al.getTurbinePosY(turbine), al.getTurbinePosZ(turbine)}; return arr(3, position); }, py::arg("turbine"))
+ .def("get_turbine_azimuth", &ActuatorFarm::getTurbineAzimuth, py::arg("turbine"))
+ .def("get_turbine_yaw", &ActuatorFarm::getTurbineYaw, py::arg("turbine"))
+ .def("get_turbine_omega", &ActuatorFarm::getTurbineOmega, py::arg("turbine"))
+ .def("get_all_azimuths", [](ActuatorFarm& al){ return arr(al.getNumberOfTurbines(), al.getAllAzimuths()); } )
+ .def("get_all_yaws", [](ActuatorFarm& al){ return arr(al.getNumberOfTurbines(), al.getAllYaws()); } )
+ .def("get_all_omegas", [](ActuatorFarm& al){ return arr(al.getNumberOfTurbines(), al.getAllOmegas()); } )
+ .def("get_all_turbine_pos_x", [](ActuatorFarm& al){ return arr(al.getNumberOfTurbines(), al.getAllTurbinePosX()); } )
+ .def("get_all_turbine_pos_y", [](ActuatorFarm& al){ return arr(al.getNumberOfTurbines(), al.getAllTurbinePosY()); } )
+ .def("get_all_turbine_pos_z", [](ActuatorFarm& al){ return arr(al.getNumberOfTurbines(), al.getAllTurbinePosZ()); } )
+
+ .def("get_all_blade_radii", [](ActuatorFarm& al){ return arr({al.getNumberOfTurbines(), al.getNumberOfNodesPerBlade()}, al.getAllBladeRadii()); } )
+ .def("get_all_blade_coords_x", [](ActuatorFarm& al){ return arr({al.getNumberOfTurbines(), al.getNumberOfBladesPerTurbine(), al.getNumberOfNodesPerBlade()}, al.getAllBladeCoordsX()); } )
+ .def("get_all_blade_coords_y", [](ActuatorFarm& al){ return arr({al.getNumberOfTurbines(), al.getNumberOfBladesPerTurbine(), al.getNumberOfNodesPerBlade()}, al.getAllBladeCoordsY()); } )
+ .def("get_all_blade_coords_z", [](ActuatorFarm& al){ return arr({al.getNumberOfTurbines(), al.getNumberOfBladesPerTurbine(), al.getNumberOfNodesPerBlade()}, al.getAllBladeCoordsZ()); } )
+ .def("get_all_blade_velocities_x", [](ActuatorFarm& al){ return arr({al.getNumberOfTurbines(), al.getNumberOfBladesPerTurbine(), al.getNumberOfNodesPerBlade()}, al.getAllBladeVelocitiesX()); } )
+ .def("get_all_blade_velocities_y", [](ActuatorFarm& al){ return arr({al.getNumberOfTurbines(), al.getNumberOfBladesPerTurbine(), al.getNumberOfNodesPerBlade()}, al.getAllBladeVelocitiesY()); } )
+ .def("get_all_blade_velocities_z", [](ActuatorFarm& al){ return arr({al.getNumberOfTurbines(), al.getNumberOfBladesPerTurbine(), al.getNumberOfNodesPerBlade()}, al.getAllBladeVelocitiesZ()); } )
+ .def("get_all_blade_forces_x", [](ActuatorFarm& al){ return arr({al.getNumberOfTurbines(), al.getNumberOfBladesPerTurbine(), al.getNumberOfNodesPerBlade()}, al.getAllBladeForcesX()); } )
+ .def("get_all_blade_forces_y", [](ActuatorFarm& al){ return arr({al.getNumberOfTurbines(), al.getNumberOfBladesPerTurbine(), al.getNumberOfNodesPerBlade()}, al.getAllBladeForcesY()); } )
+ .def("get_all_blade_forces_z", [](ActuatorFarm& al){ return arr({al.getNumberOfTurbines(), al.getNumberOfBladesPerTurbine(), al.getNumberOfNodesPerBlade()}, al.getAllBladeForcesZ()); } )
+
+ .def("get_turbine_blade_radii", [](ActuatorFarm& al, uint turbine){ return arr(al.getNumberOfNodesPerBlade(), al.getTurbineBladeRadiiDevice(turbine)); } , py::arg("turbine"))
+ .def("get_turbine_blade_coords_x", [](ActuatorFarm& al, uint turbine){ return arr({al.getNumberOfBladesPerTurbine(), al.getNumberOfNodesPerBlade()}, al.getTurbineBladeCoordsXDevice(turbine)); }, py::arg("turbine") )
+ .def("get_turbine_blade_coords_y", [](ActuatorFarm& al, uint turbine){ return arr({al.getNumberOfBladesPerTurbine(), al.getNumberOfNodesPerBlade()}, al.getTurbineBladeCoordsYDevice(turbine)); }, py::arg("turbine") )
+ .def("get_turbine_blade_coords_z", [](ActuatorFarm& al, uint turbine){ return arr({al.getNumberOfBladesPerTurbine(), al.getNumberOfNodesPerBlade()}, al.getTurbineBladeCoordsZDevice(turbine)); }, py::arg("turbine") )
+ .def("get_turbine_blade_velocities_x", [](ActuatorFarm& al, uint turbine){ return arr({al.getNumberOfBladesPerTurbine(), al.getNumberOfNodesPerBlade()}, al.getTurbineBladeVelocitiesXDevice(turbine)); }, py::arg("turbine") )
+ .def("get_turbine_blade_velocities_y", [](ActuatorFarm& al, uint turbine){ return arr({al.getNumberOfBladesPerTurbine(), al.getNumberOfNodesPerBlade()}, al.getTurbineBladeVelocitiesYDevice(turbine)); }, py::arg("turbine") )
+ .def("get_turbine_blade_velocities_z", [](ActuatorFarm& al, uint turbine){ return arr({al.getNumberOfBladesPerTurbine(), al.getNumberOfNodesPerBlade()}, al.getTurbineBladeVelocitiesZDevice(turbine)); }, py::arg("turbine") )
+ .def("get_turbine_blade_forces_x", [](ActuatorFarm& al, uint turbine){ return arr({al.getNumberOfBladesPerTurbine(), al.getNumberOfNodesPerBlade()}, al.getTurbineBladeForcesXDevice(turbine)); }, py::arg("turbine") )
+ .def("get_turbine_blade_forces_y", [](ActuatorFarm& al, uint turbine){ return arr({al.getNumberOfBladesPerTurbine(), al.getNumberOfNodesPerBlade()}, al.getTurbineBladeForcesYDevice(turbine)); }, py::arg("turbine") )
+ .def("get_turbine_blade_forces_z", [](ActuatorFarm& al, uint turbine){ return arr({al.getNumberOfBladesPerTurbine(), al.getNumberOfNodesPerBlade()}, al.getTurbineBladeForcesZDevice(turbine)); }, py::arg("turbine") )
+
+ .def("get_all_blade_radii_device", [](ActuatorFarm& al) -> intptr_t { return reinterpret_cast<intptr_t>(al.getAllBladeRadiiDevice()); } )
+ .def("get_all_blade_coords_x_device", [](ActuatorFarm& al) -> intptr_t { return reinterpret_cast<intptr_t> (al.getAllBladeCoordsXDevice()); } )
+ .def("get_all_blade_coords_y_device", [](ActuatorFarm& al) -> intptr_t { return reinterpret_cast<intptr_t> (al.getAllBladeCoordsYDevice()); } )
+ .def("get_all_blade_coords_z_device", [](ActuatorFarm& al) -> intptr_t { return reinterpret_cast<intptr_t> (al.getAllBladeCoordsZDevice()); } )
+ .def("get_all_blade_velocities_x_device", [](ActuatorFarm& al) -> intptr_t { return reinterpret_cast<intptr_t> (al.getAllBladeVelocitiesXDevice()); } )
+ .def("get_all_blade_velocities_y_device", [](ActuatorFarm& al) -> intptr_t { return reinterpret_cast<intptr_t> (al.getAllBladeVelocitiesYDevice()); } )
+ .def("get_all_blade_velocities_z_device", [](ActuatorFarm& al) -> intptr_t { return reinterpret_cast<intptr_t> (al.getAllBladeVelocitiesZDevice()); } )
+ .def("get_all_blade_forces_x_device", [](ActuatorFarm& al) -> intptr_t { return reinterpret_cast<intptr_t> (al.getAllBladeForcesXDevice()); } )
+ .def("get_all_blade_forces_y_device", [](ActuatorFarm& al) -> intptr_t { return reinterpret_cast<intptr_t> (al.getAllBladeForcesYDevice()); } )
+ .def("get_all_blade_forces_z_device", [](ActuatorFarm& al) -> intptr_t { return reinterpret_cast<intptr_t> (al.getAllBladeForcesZDevice()); } )
+
+ .def("get_turbine_blade_radii_device", [](ActuatorFarm& al, uint turbine) -> intptr_t { return reinterpret_cast<intptr_t>(al.getTurbineBladeRadiiDevice(turbine)); }, py::arg("turbine") )
+ .def("get_turbine_blade_coords_x_device", [](ActuatorFarm& al, uint turbine) -> intptr_t { return reinterpret_cast<intptr_t>(al.getTurbineBladeCoordsXDevice(turbine)); }, py::arg("turbine") )
+ .def("get_turbine_blade_coords_y_device", [](ActuatorFarm& al, uint turbine) -> intptr_t { return reinterpret_cast<intptr_t>(al.getTurbineBladeCoordsYDevice(turbine)); }, py::arg("turbine") )
+ .def("get_turbine_blade_coords_z_device", [](ActuatorFarm& al, uint turbine) -> intptr_t { return reinterpret_cast<intptr_t>(al.getTurbineBladeCoordsZDevice(turbine)); }, py::arg("turbine") )
+ .def("get_turbine_blade_velocities_x_device", [](ActuatorFarm& al, uint turbine) -> intptr_t { return reinterpret_cast<intptr_t>(al.getTurbineBladeVelocitiesXDevice(turbine)); }, py::arg("turbine") )
+ .def("get_turbine_blade_velocities_y_device", [](ActuatorFarm& al, uint turbine) -> intptr_t { return reinterpret_cast<intptr_t>(al.getTurbineBladeVelocitiesYDevice(turbine)); }, py::arg("turbine") )
+ .def("get_turbine_blade_velocities_z_device", [](ActuatorFarm& al, uint turbine) -> intptr_t { return reinterpret_cast<intptr_t>(al.getTurbineBladeVelocitiesZDevice(turbine)); }, py::arg("turbine") )
+ .def("get_turbine_blade_forces_x_device", [](ActuatorFarm& al, uint turbine) -> intptr_t { return reinterpret_cast<intptr_t>(al.getTurbineBladeForcesXDevice(turbine)); }, py::arg("turbine") )
+ .def("get_turbine_blade_forces_y_device", [](ActuatorFarm& al, uint turbine) -> intptr_t { return reinterpret_cast<intptr_t>(al.getTurbineBladeForcesYDevice(turbine)); }, py::arg("turbine") )
+ .def("get_turbine_blade_forces_z_device", [](ActuatorFarm& al, uint turbine) -> intptr_t { return reinterpret_cast<intptr_t>(al.getTurbineBladeForcesZDevice(turbine)); }, py::arg("turbine") )
+
+ .def("set_all_azimuths", [](ActuatorFarm& al, arr azimuths){ al.setAllAzimuths(static_cast<float *>(azimuths.request().ptr)); }, py::arg("azimuths"))
+ .def("set_all_yaws", [](ActuatorFarm& al, arr yaws){ al.setAllYaws(static_cast<float *>(yaws.request().ptr)); }, py::arg("yaws"))
+ .def("set_all_omegas", [](ActuatorFarm& al, arr omegas){ al.setAllOmegas(static_cast<float *>(omegas.request().ptr)); }, py::arg("omegas"))
+
+ .def("set_turbine_azimuth", &ActuatorFarm::setTurbineAzimuth, py::arg("turbine"), py::arg("azimuth"))
+ .def("set_turbine_yaw", &ActuatorFarm::setTurbineYaw, py::arg("turbine"), py::arg("yaw"))
+ .def("set_turbine_omega", &ActuatorFarm::setTurbineOmega, py::arg("turbine"), py::arg("omega"))
+
+ .def("set_all_blade_coords", [](ActuatorFarm& al, arr coordsX, arr coordsY, arr coordsZ)
+ {
+ al.setAllBladeCoords(static_cast<float *>(coordsX.request().ptr), static_cast<float *>(coordsY.request().ptr), static_cast<float *>(coordsZ.request().ptr));
+ }, py::arg("blade_coords_x"), py::arg("blade_coords_y"), py::arg("blade_coords_z") )
+ .def("set_all_blade_velocities", [](ActuatorFarm& al, arr velocitiesX, arr velocitiesY, arr velocitiesZ)
+ {
+ al.setAllBladeVelocities(static_cast<float *>(velocitiesX.request().ptr), static_cast<float *>(velocitiesY.request().ptr), static_cast<float *>(velocitiesZ.request().ptr));
+ }, py::arg("blade_velocities_x"), py::arg("blade_velocities_y"), py::arg("blade_velocities_z") )
+ .def("set_all_blade_forces", [](ActuatorFarm& al, arr forcesX, arr forcesY, arr forcesZ)
+ {
+ al.setAllBladeForces(static_cast<float *>(forcesX.request().ptr), static_cast<float *>(forcesY.request().ptr), static_cast<float *>(forcesZ.request().ptr));
+ }, py::arg("blade_forces_x"), py::arg("blade_forces_y"), py::arg("blade_forces_z") )
+ .def("set_turbine_blade_coords", [](ActuatorFarm& al, uint turbine, arr coordsX, arr coordsY, arr coordsZ)
+ {
+ al.setTurbineBladeCoords(turbine, static_cast<float *>(coordsX.request().ptr), static_cast<float *>(coordsY.request().ptr), static_cast<float *>(coordsZ.request().ptr));
+ }, py::arg("turbine"), py::arg("blade_coords_x"), py::arg("blade_coords_y"), py::arg("blade_coords_z") )
+ .def("set_turbine_blade_velocities", [](ActuatorFarm& al, uint turbine, arr velocitiesX, arr velocitiesY, arr velocitiesZ)
+ {
+ al.setTurbineBladeVelocities(turbine, static_cast<float *>(velocitiesX.request().ptr), static_cast<float *>(velocitiesY.request().ptr), static_cast<float *>(velocitiesZ.request().ptr));
+ }, py::arg("turbine"), py::arg("blade_velocities_x"), py::arg("blade_velocities_y"), py::arg("blade_velocities_z") )
+ .def("set_turbine_blade_forces", [](ActuatorFarm& al, uint turbine, arr forcesX, arr forcesY, arr forcesZ)
+ {
+ al.setTurbineBladeForces(turbine, static_cast<float *>(forcesX.request().ptr), static_cast<float *>(forcesY.request().ptr), static_cast<float *>(forcesZ.request().ptr));
+ }, py::arg("turbine"), py::arg("blade_forces_x"), py::arg("blade_forces_y"), py::arg("blade_forces_z") )
+ .def("calc_blade_forces", &ActuatorFarm::calcBladeForces);
+ }
+}
\ No newline at end of file
diff --git a/pythonbindings/src/gpu/submodules/actuator_line.cpp b/pythonbindings/src/gpu/submodules/actuator_line.cpp
deleted file mode 100644
index 3207fadbc37df38e53e00adcb9a86f0b8e82ba98..0000000000000000000000000000000000000000
--- a/pythonbindings/src/gpu/submodules/actuator_line.cpp
+++ /dev/null
@@ -1,72 +0,0 @@
-#include <pybind11/pybind11.h>
-#include <pybind11/stl.h>
-#include <pybind11/numpy.h>
-#include <gpu/VirtualFluids_GPU/PreCollisionInteractor/ActuatorLine.h>
-#include <gpu/VirtualFluids_GPU/PreCollisionInteractor/PreCollisionInteractor.h>
-class PyActuatorLine : public ActuatorLine
-{
-public:
- using ActuatorLine::ActuatorLine; // Inherit constructors
- void calcBladeForces() override
- {
- PYBIND11_OVERRIDE_NAME(void, ActuatorLine, "calc_blade_forces", calcBladeForces,);
- }
-};
-namespace actuator_line
-{
- namespace py = pybind11;
-
- void makeModule(py::module_ &parentModule)
- {
- using arr = py::array_t<float, py::array::c_style>;
-
- py::class_<ActuatorLine, PreCollisionInteractor, PyActuatorLine, std::shared_ptr<ActuatorLine>>(parentModule, "ActuatorLine", py::dynamic_attr())
- .def(py::init< const uint,
- const real,
- const uint,
- const real,
- real, real, real,
- const real,
- int,
- const real,
- const real>(),
- "n_blades",
- "density",
- "n_blade_nodes",
- "epsilon",
- "turbine_pos_x", "turbine_pos_y", "turbine_pos_z",
- "diameter",
- "level",
- "delta_t",
- "delta_x")
- .def_property("omega", &ActuatorLine::getOmega, &ActuatorLine::setOmega)
- .def_property("azimuth", &ActuatorLine::getAzimuth, &ActuatorLine::setAzimuth)
- .def_property("yaw", &ActuatorLine::getYaw, &ActuatorLine::setYaw)
- .def_property_readonly("n_blades", &ActuatorLine::getNBlades)
- .def_property_readonly("n_blade_nodes", &ActuatorLine::getNBladeNodes)
- .def_property_readonly("n_nodes", &ActuatorLine::getNNodes)
- .def_property_readonly("n_indices", &ActuatorLine::getNIndices)
- .def_property_readonly("density", &ActuatorLine::getDensity)
- .def_property_readonly("position_x", &ActuatorLine::getPositionX)
- .def_property_readonly("position_y", &ActuatorLine::getPositionY)
- .def_property_readonly("position_z", &ActuatorLine::getPositionZ)
- .def_property_readonly("position", [](ActuatorLine& al){ real position[3] = {al.getPositionX(), al.getPositionY(), al.getPositionZ()}; return arr(3, position); } )
- .def("get_radii", [](ActuatorLine& al){ return arr(al.getNBladeNodes(), al.getBladeRadii()); } )
- .def("get_blade_coords_x", [](ActuatorLine& al){ return arr({al.getNBlades(), al.getNBladeNodes()}, al.getBladeCoordsX()); } )
- .def("get_blade_coords_y", [](ActuatorLine& al){ return arr({al.getNBlades(), al.getNBladeNodes()}, al.getBladeCoordsY()); } )
- .def("get_blade_coords_z", [](ActuatorLine& al){ return arr({al.getNBlades(), al.getNBladeNodes()}, al.getBladeCoordsZ()); } )
- .def("get_blade_velocities_x", [](ActuatorLine& al){ return arr({al.getNBlades(), al.getNBladeNodes()}, al.getBladeVelocitiesX()); } )
- .def("get_blade_velocities_y", [](ActuatorLine& al){ return arr({al.getNBlades(), al.getNBladeNodes()}, al.getBladeVelocitiesY()); } )
- .def("get_blade_velocities_z", [](ActuatorLine& al){ return arr({al.getNBlades(), al.getNBladeNodes()}, al.getBladeVelocitiesZ()); } )
- .def("get_blade_forces_x", [](ActuatorLine& al){ return arr({al.getNBlades(), al.getNBladeNodes()}, al.getBladeForcesX()); } )
- .def("get_blade_forces_y", [](ActuatorLine& al){ return arr({al.getNBlades(), al.getNBladeNodes()}, al.getBladeForcesY()); } )
- .def("get_blade_forces_z", [](ActuatorLine& al){ return arr({al.getNBlades(), al.getNBladeNodes()}, al.getBladeForcesZ()); } )
- .def("set_blade_coords", [](ActuatorLine& al, arr coordsX, arr coordsY, arr coordsZ){
- al.setBladeCoords(static_cast<float *>(coordsX.request().ptr), static_cast<float *>(coordsY.request().ptr), static_cast<float *>(coordsZ.request().ptr)); } )
- .def("set_blade_velocities", [](ActuatorLine& al, arr velocitiesX, arr velocitiesY, arr velocitiesZ){
- al.setBladeVelocities(static_cast<float *>(velocitiesX.request().ptr), static_cast<float *>(velocitiesY.request().ptr), static_cast<float *>(velocitiesZ.request().ptr)); } )
- .def("set_blade_forces", [](ActuatorLine& al, arr forcesX, arr forcesY, arr forcesZ){
- al.setBladeForces(static_cast<float *>(forcesX.request().ptr), static_cast<float *>(forcesY.request().ptr), static_cast<float *>(forcesZ.request().ptr)); } )
- .def("calc_blade_forces", &ActuatorLine::calcBladeForces);
- }
-}
\ No newline at end of file
diff --git a/pythonbindings/src/gpu/submodules/boundary_conditions.cpp b/pythonbindings/src/gpu/submodules/boundary_conditions.cpp
index 8f941a8705c225275d25291205ebdaeef8de5c9e..865817bb16f7b164c40bdc066645fb2e1f1c842e 100644
--- a/pythonbindings/src/gpu/submodules/boundary_conditions.cpp
+++ b/pythonbindings/src/gpu/submodules/boundary_conditions.cpp
@@ -1,5 +1,38 @@
+//=======================================================================================
+// ____ ____ __ ______ __________ __ __ __ __
+// \ \ | | | | | _ \ |___ ___| | | | | / \ | |
+// \ \ | | | | | |_) | | | | | | | / \ | |
+// \ \ | | | | | _ / | | | | | | / /\ \ | |
+// \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
+// \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
+// \ \ | | ________________________________________________________________
+// \ \ | | | ______________________________________________________________|
+// \ \| | | | __ __ __ __ ______ _______
+// \ | | |_____ | | | | | | | | | _ \ / _____)
+// \ | | _____| | | | | | | | | | | \ \ \_______
+// \ | | | | |_____ | \_/ | | | | |_/ / _____ |
+// \ _____| |__| |________| \_______/ |__| |______/ (_______/
+//
+// This file is part of VirtualFluids. VirtualFluids is free software: you can
+// redistribute it and/or modify it under the terms of the GNU General Public
+// License as published by the Free Software Foundation, either version 3 of
+// the License, or (at your option) any later version.
+//
+// VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
+// ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+// FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
+// for more details.
+//
+// You should have received a copy of the GNU General Public License along
+// with VirtualFluids (see COPYING.txt). If not, see <http://www.gnu.org/licenses/>.
+//
+//! \file boindary_conditions.cpp
+//! \ingroup submodules
+//! \author Henry Korb
+//=======================================================================================
#include <pybind11/pybind11.h>
#include <gpu/GridGenerator/grid/BoundaryConditions/Side.h>
+#include "gpu/VirtualFluids_GPU/Factories/BoundaryConditionFactory.h"
namespace boundary_conditions
{
@@ -14,7 +47,59 @@ namespace boundary_conditions
.value("PY", SideType::PY)
.value("MZ", SideType::MZ)
.value("PZ", SideType::PZ)
- .value("GEOMETRY", SideType::GEOMETRY)
- .export_values();
+ .value("GEOMETRY", SideType::GEOMETRY);
+
+ py::class_<BoundaryConditionFactory>(parentModule, "BoundaryConditionFactory")
+ .def(py::init<>())
+ .def("set_velocity_boundary_condition", &BoundaryConditionFactory::setVelocityBoundaryCondition, py::arg("boundary_condition_type"))
+ .def("set_no_slip_boundary_condition", &BoundaryConditionFactory::setNoSlipBoundaryCondition, py::arg("boundary_condition_type"))
+ .def("set_slip_boundary_condition", &BoundaryConditionFactory::setSlipBoundaryCondition, py::arg("boundary_condition_type"))
+ .def("set_pressure_boundary_condition", &BoundaryConditionFactory::setPressureBoundaryCondition, py::arg("boundary_condition_type"))
+ .def("set_stress_boundary_condition", &BoundaryConditionFactory::setStressBoundaryCondition, py::arg("boundary_condition_type"))
+ .def("set_precursor_boundary_condition", &BoundaryConditionFactory::setPrecursorBoundaryCondition, py::arg("boundary_condition_type"))
+ .def("set_geometry_boundary_condition", &BoundaryConditionFactory::setGeometryBoundaryCondition, py::arg("boundary_condition_type"));
+
+ py::enum_<BoundaryConditionFactory::VelocityBC>(parentModule, "VelocityBC")
+ .value("VelocitySimpleBounceBackCompressible", BoundaryConditionFactory::VelocityBC::VelocitySimpleBounceBackCompressible)
+ .value("VelocityIncompressible", BoundaryConditionFactory::VelocityBC::VelocityIncompressible)
+ .value("VelocityCompressible", BoundaryConditionFactory::VelocityBC::VelocityCompressible)
+ .value("VelocityAndPressureCompressible", BoundaryConditionFactory::VelocityBC::VelocityAndPressureCompressible)
+ .value("NotSpecified", BoundaryConditionFactory::VelocityBC::NotSpecified);
+
+
+ py::enum_<BoundaryConditionFactory::NoSlipBC>(parentModule, "NoSlipBC")
+ .value("NoSlipImplicitBounceBack", BoundaryConditionFactory::NoSlipBC::NoSlipImplicitBounceBack)
+ .value("NoSlipBounceBack", BoundaryConditionFactory::NoSlipBC::NoSlipBounceBack)
+ .value("NoSlipIncompressible", BoundaryConditionFactory::NoSlipBC::NoSlipIncompressible)
+ .value("NoSlipCompressible", BoundaryConditionFactory::NoSlipBC::NoSlipCompressible)
+ .value("NoSlip3rdMomentsCompressible", BoundaryConditionFactory::NoSlipBC::NoSlip3rdMomentsCompressible);
+
+ py::enum_<BoundaryConditionFactory::SlipBC>(parentModule, "SlipBC")
+ .value("SlipIncompressible", BoundaryConditionFactory::SlipBC::SlipIncompressible)
+ .value("SlipCompressible", BoundaryConditionFactory::SlipBC::SlipCompressible)
+ .value("SlipBounceBack", BoundaryConditionFactory::SlipBC::SlipBounceBack)
+ .value("SlipCompressibleTurbulentViscosity", BoundaryConditionFactory::SlipBC::SlipCompressibleTurbulentViscosity)
+ .value("SlipPressureCompressibleTurbulentViscosity", BoundaryConditionFactory::SlipBC::SlipPressureCompressibleTurbulentViscosity)
+ .value("NotSpecified", BoundaryConditionFactory::SlipBC::NotSpecified);
+
+ py::enum_<BoundaryConditionFactory::PressureBC>(parentModule, "PressureBC")
+ .value("PressureEquilibrium", BoundaryConditionFactory::PressureBC::PressureEquilibrium)
+ .value("PressureEquilibrium2", BoundaryConditionFactory::PressureBC::PressureEquilibrium2)
+ .value("PressureNonEquilibriumIncompressible", BoundaryConditionFactory::PressureBC::PressureNonEquilibriumIncompressible)
+ .value("PressureNonEquilibriumCompressible", BoundaryConditionFactory::PressureBC::PressureNonEquilibriumCompressible)
+ .value("OutflowNonReflective", BoundaryConditionFactory::PressureBC::OutflowNonReflective)
+ .value("OutflowNonReflectivePressureCorrection", BoundaryConditionFactory::PressureBC::OutflowNonReflectivePressureCorrection)
+ .value("NotSpecified", BoundaryConditionFactory::PressureBC::NotSpecified);
+
+ py::enum_<BoundaryConditionFactory::StressBC>(parentModule, "StressBC")
+ .value("StressCompressible", BoundaryConditionFactory::StressBC::StressCompressible)
+ .value("StressBounceBack", BoundaryConditionFactory::StressBC::StressBounceBack)
+ .value("StressPressureBounceBack", BoundaryConditionFactory::StressBC::StressPressureBounceBack)
+ .value("NotSpecified", BoundaryConditionFactory::StressBC::NotSpecified);
+
+ py::enum_<BoundaryConditionFactory::PrecursorBC>(parentModule, "PrecursorBC")
+ .value("VelocityPrecursor", BoundaryConditionFactory::PrecursorBC::VelocityPrecursor)
+ .value("DistributionsPrecursor", BoundaryConditionFactory::PrecursorBC::DistributionsPrecursor)
+ .value("NotSpecified", BoundaryConditionFactory::PrecursorBC::NotSpecified);
}
}
\ No newline at end of file
diff --git a/pythonbindings/src/gpu/submodules/communicator.cpp b/pythonbindings/src/gpu/submodules/communicator.cpp
index edb36e2c2f774903590a16a0b406c721662827b1..26a57061933fbdbfe3447ec89eeb07116a9b974b 100644
--- a/pythonbindings/src/gpu/submodules/communicator.cpp
+++ b/pythonbindings/src/gpu/submodules/communicator.cpp
@@ -1,3 +1,35 @@
+//=======================================================================================
+// ____ ____ __ ______ __________ __ __ __ __
+// \ \ | | | | | _ \ |___ ___| | | | | / \ | |
+// \ \ | | | | | |_) | | | | | | | / \ | |
+// \ \ | | | | | _ / | | | | | | / /\ \ | |
+// \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
+// \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
+// \ \ | | ________________________________________________________________
+// \ \ | | | ______________________________________________________________|
+// \ \| | | | __ __ __ __ ______ _______
+// \ | | |_____ | | | | | | | | | _ \ / _____)
+// \ | | _____| | | | | | | | | | | \ \ \_______
+// \ | | | | |_____ | \_/ | | | | |_/ / _____ |
+// \ _____| |__| |________| \_______/ |__| |______/ (_______/
+//
+// This file is part of VirtualFluids. VirtualFluids is free software: you can
+// redistribute it and/or modify it under the terms of the GNU General Public
+// License as published by the Free Software Foundation, either version 3 of
+// the License, or (at your option) any later version.
+//
+// VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
+// ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+// FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
+// for more details.
+//
+// You should have received a copy of the GNU General Public License along
+// with VirtualFluids (see COPYING.txt). If not, see <http://www.gnu.org/licenses/>.
+//
+//! \file communicator.cpp
+//! \ingroup submodules
+//! \author Henry Korb
+//=======================================================================================
#include <pybind11/pybind11.h>
#include <gpu/VirtualFluids_GPU/Communication/Communicator.h>
@@ -8,7 +40,7 @@ namespace communicator
void makeModule(py::module_ &parentModule)
{
py::class_<vf::gpu::Communicator, std::unique_ptr<vf::gpu::Communicator, py::nodelete>>(parentModule, "Communicator")
- .def("get_instance", &vf::gpu::Communicator::getInstance, py::return_value_policy::reference)
+ .def_static("get_instance", &vf::gpu::Communicator::getInstance, py::return_value_policy::reference)
.def("get_number_of_process", &vf::gpu::Communicator::getNummberOfProcess)
.def("get_pid", &vf::gpu::Communicator::getPID);
}
diff --git a/pythonbindings/src/gpu/submodules/cuda_memory_manager.cpp b/pythonbindings/src/gpu/submodules/cuda_memory_manager.cpp
index bf27080cb3cd050343ba42b0571827ed58870cfd..bbff4832cb73f47e3d1a5a6abd78e21da2473deb 100644
--- a/pythonbindings/src/gpu/submodules/cuda_memory_manager.cpp
+++ b/pythonbindings/src/gpu/submodules/cuda_memory_manager.cpp
@@ -1,3 +1,35 @@
+//=======================================================================================
+// ____ ____ __ ______ __________ __ __ __ __
+// \ \ | | | | | _ \ |___ ___| | | | | / \ | |
+// \ \ | | | | | |_) | | | | | | | / \ | |
+// \ \ | | | | | _ / | | | | | | / /\ \ | |
+// \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
+// \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
+// \ \ | | ________________________________________________________________
+// \ \ | | | ______________________________________________________________|
+// \ \| | | | __ __ __ __ ______ _______
+// \ | | |_____ | | | | | | | | | _ \ / _____)
+// \ | | _____| | | | | | | | | | | \ \ \_______
+// \ | | | | |_____ | \_/ | | | | |_/ / _____ |
+// \ _____| |__| |________| \_______/ |__| |______/ (_______/
+//
+// This file is part of VirtualFluids. VirtualFluids is free software: you can
+// redistribute it and/or modify it under the terms of the GNU General Public
+// License as published by the Free Software Foundation, either version 3 of
+// the License, or (at your option) any later version.
+//
+// VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
+// ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+// FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
+// for more details.
+//
+// You should have received a copy of the GNU General Public License along
+// with VirtualFluids (see COPYING.txt). If not, see <http://www.gnu.org/licenses/>.
+//
+//! \file cuda_memory_manager.cpp
+//! \ingroup submodules
+//! \author Henry Korb
+//=======================================================================================
#include <pybind11/pybind11.h>
#include <gpu/VirtualFluids_GPU/GPU/CudaMemoryManager.h>
#include <gpu/VirtualFluids_GPU/Parameter/Parameter.h>
@@ -10,6 +42,6 @@ namespace cuda_memory_manager
void makeModule(py::module_ &parentModule)
{
py::class_<CudaMemoryManager, std::shared_ptr<CudaMemoryManager>>(parentModule, "CudaMemoryManager")
- .def(py::init<std::shared_ptr<Parameter>>(), "parameter");
+ .def(py::init<std::shared_ptr<Parameter>>(), py::arg("parameter"));
}
}
\ No newline at end of file
diff --git a/pythonbindings/src/gpu/submodules/grid_generator.cpp b/pythonbindings/src/gpu/submodules/grid_generator.cpp
index 579c06c4e00cae9646ced8b554d71631eeb7e793..3e9fb5655e26ffa6053a205da5a3e3f0f2ecd49f 100644
--- a/pythonbindings/src/gpu/submodules/grid_generator.cpp
+++ b/pythonbindings/src/gpu/submodules/grid_generator.cpp
@@ -1,4 +1,37 @@
+//=======================================================================================
+// ____ ____ __ ______ __________ __ __ __ __
+// \ \ | | | | | _ \ |___ ___| | | | | / \ | |
+// \ \ | | | | | |_) | | | | | | | / \ | |
+// \ \ | | | | | _ / | | | | | | / /\ \ | |
+// \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
+// \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
+// \ \ | | ________________________________________________________________
+// \ \ | | | ______________________________________________________________|
+// \ \| | | | __ __ __ __ ______ _______
+// \ | | |_____ | | | | | | | | | _ \ / _____)
+// \ | | _____| | | | | | | | | | | \ \ \_______
+// \ | | | | |_____ | \_/ | | | | |_/ / _____ |
+// \ _____| |__| |________| \_______/ |__| |______/ (_______/
+//
+// This file is part of VirtualFluids. VirtualFluids is free software: you can
+// redistribute it and/or modify it under the terms of the GNU General Public
+// License as published by the Free Software Foundation, either version 3 of
+// the License, or (at your option) any later version.
+//
+// VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
+// ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+// FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
+// for more details.
+//
+// You should have received a copy of the GNU General Public License along
+// with VirtualFluids (see COPYING.txt). If not, see <http://www.gnu.org/licenses/>.
+//
+//! \file grid_generator.cpp
+//! \ingroup submodules
+//! \author Henry Korb, Henrik Asmuth
+//=======================================================================================
#include <pybind11/pybind11.h>
+#include "gpu/GridGenerator/utilities/communication.h"
#include "gpu/GridGenerator/geometries/Object.h"
#include "gpu/GridGenerator/geometries/BoundingBox/BoundingBox.h"
#include "gpu/GridGenerator/geometries/Conglomerate/Conglomerate.h"
@@ -17,51 +50,63 @@ namespace grid_generator
{
py::module gridGeneratorModule = parentModule.def_submodule("grid_generator");
+ //TODO:
+ // py::enum_<CommunicationDirections>(gridGeneratorModule, "CommunicationDirections")
+ // .value("MX", CommunicationDirections::MX)
+ // .value("PX", CommunicationDirections::PX)
+ // .value("MY", CommunicationDirections::MY)
+ // .value("PY", CommunicationDirections::PY)
+ // .value("MZ", CommunicationDirections::MZ)
+ // .value("PZ", CommunicationDirections::PZ);
+
py::class_<GridFactory, std::shared_ptr<GridFactory>>(gridGeneratorModule, "GridFactory")
- .def("make", &GridFactory::make, py::return_value_policy::reference);
+ .def_static("make", &GridFactory::make, py::return_value_policy::reference);
- py::class_<BoundingBox>(gridGeneratorModule, "BoundingBox")
- .def(py::init<real, real, real, real, real, real>(),"min_x","max_x","min_y","max_y","min_z","max_z");
+ py::class_<BoundingBox, std::shared_ptr<BoundingBox>>(gridGeneratorModule, "BoundingBox")
+ .def(py::init<real, real, real, real, real, real>(), py::arg("min_x"), py::arg("max_x"), py::arg("min_y"), py::arg("max_y"), py::arg("min_z"), py::arg("max_z"));
py::class_<Object, std::shared_ptr<Object>>(gridGeneratorModule, "Object");
py::class_<Conglomerate, Object, std::shared_ptr<Conglomerate>>(gridGeneratorModule, "Conglomerate")
- .def("make_shared", &Conglomerate::makeShared, py::return_value_policy::reference)
- .def("add", &Conglomerate::add)
- .def("subtract", &Conglomerate::subtract);
+ .def_static("make_shared", &Conglomerate::makeShared, py::return_value_policy::reference)
+ .def("add", &Conglomerate::add, py::arg("object"))
+ .def("subtract", &Conglomerate::subtract, py::arg("object"));
py::class_<Cuboid, Object, std::shared_ptr<Cuboid>>(gridGeneratorModule, "Cuboid")
.def(py::init<const double&, const double&, const double&, const double&, const double&, const double&>(),
- "min_x1", "min_x2", "min_x3", "max_x1", "max_x2", "max_x3");
+ py::arg("min_x1"), py::arg("min_x2"), py::arg("min_x3"), py::arg("max_x1"), py::arg("max_x2"), py::arg("max_x3"));
py::class_<Sphere, Object, std::shared_ptr<Sphere>>(gridGeneratorModule, "Sphere")
- .def("make_shared", &Sphere::makeShared, py::return_value_policy::reference);
+ .def_static("make_shared", &Sphere::makeShared, py::return_value_policy::reference);
py::class_<TriangularMesh, Object, std::shared_ptr<TriangularMesh>>(gridGeneratorModule, "TriangularMesh")
- .def("make", &TriangularMesh::make, py::return_value_policy::reference);
+ .def_static("make", &TriangularMesh::make, py::return_value_policy::reference);
py::class_<GridBuilder, std::shared_ptr<GridBuilder>>(gridGeneratorModule, "GridBuilder")
- .def("get_number_of_grid_levels", &GridBuilder::getNumberOfGridLevels)
- .def("get_grid", &GridBuilder::getGrid);
+ .def("get_number_of_grid_levels", &GridBuilder::getNumberOfGridLevels);
py::class_<LevelGridBuilder, GridBuilder, std::shared_ptr<LevelGridBuilder>>(gridGeneratorModule, "LevelGridBuilder")
- .def("get_grid", py::overload_cast<int, int>(&LevelGridBuilder::getGrid))
- .def("set_slip_boundary_condition", &LevelGridBuilder::setSlipBoundaryCondition)
- .def("set_velocity_boundary_condition", &LevelGridBuilder::setVelocityBoundaryCondition)
- .def("set_pressure_boundary_condition", &LevelGridBuilder::setPressureBoundaryCondition)
- .def("set_periodic_boundary_condition", &LevelGridBuilder::setPeriodicBoundaryCondition)
- .def("set_no_slip_boundary_condition", &LevelGridBuilder::setNoSlipBoundaryCondition)
- .def("set_stress_boundary_condition", &LevelGridBuilder::setStressBoundaryCondition);
+ .def("set_slip_boundary_condition", &LevelGridBuilder::setSlipBoundaryCondition, py::arg("side_type"), py::arg("normal_x"), py::arg("normal_y"), py::arg("normal_z"))
+ .def("set_velocity_boundary_condition", &LevelGridBuilder::setVelocityBoundaryCondition, py::arg("side_type"), py::arg("vx"), py::arg("vy"), py::arg("vz"))
+ .def("set_pressure_boundary_condition", &LevelGridBuilder::setPressureBoundaryCondition, py::arg("side_type"), py::arg("rho"))
+ .def("set_periodic_boundary_condition", &LevelGridBuilder::setPeriodicBoundaryCondition, py::arg("periodic_x"), py::arg("periodic_y"), py::arg("periodic_z"))
+ .def("set_no_slip_boundary_condition", &LevelGridBuilder::setNoSlipBoundaryCondition, py::arg("side_type"))
+ .def("set_precursor_boundary_condition", &LevelGridBuilder::setPrecursorBoundaryCondition, py::arg("side_type"), py::arg("file_collection"), py::arg("n_t_read"), py::arg("velocity_x")=0.0f, py::arg("velocity_y")=0.0f, py::arg("velocity_z")=0.0f, py::arg("file_level_to_grid_level_map")=std::vector<uint>())
+ .def("set_stress_boundary_condition", &LevelGridBuilder::setStressBoundaryCondition, py::arg("side_type"), py::arg("normal_x"), py::arg("normal_y"), py::arg("normal_z"), py::arg("sampling_offset"), py::arg("z0"), py::arg("dx"));
py::class_<MultipleGridBuilder, LevelGridBuilder, std::shared_ptr<MultipleGridBuilder>>(gridGeneratorModule, "MultipleGridBuilder")
- .def("make_shared", &MultipleGridBuilder::makeShared, py::return_value_policy::reference)
- .def("add_coarse_grid", &MultipleGridBuilder::addCoarseGrid)
- .def("add_grid", py::overload_cast<Object*>(&MultipleGridBuilder::addGrid))
- .def("add_grid", py::overload_cast<Object*, uint>(&MultipleGridBuilder::addGrid))
- .def("add_geometry", py::overload_cast<Object*>(&MultipleGridBuilder::addGeometry))
- .def("add_geometry", py::overload_cast<Object*, uint>(&MultipleGridBuilder::addGeometry))
+ .def_static("make_shared", &MultipleGridBuilder::makeShared, py::return_value_policy::reference, py::arg("grid_factory"))
+ .def("add_coarse_grid", &MultipleGridBuilder::addCoarseGrid, py::arg("start_x"), py::arg("start_y"), py::arg("start_z"), py::arg("end_x"), py::arg("end_y"), py::arg("end_z"), py::arg("delta"))
+ .def("add_grid", py::overload_cast<Object*>(&MultipleGridBuilder::addGrid), py::arg("grid_shape"))
+ .def("add_grid", py::overload_cast<Object*, uint>(&MultipleGridBuilder::addGrid), py::arg("grid_shape"), py::arg("level_fine"))
+ .def("add_geometry", py::overload_cast<Object*>(&MultipleGridBuilder::addGeometry), py::arg("solid_object"))
+ .def("add_geometry", py::overload_cast<Object*, uint>(&MultipleGridBuilder::addGeometry), py::arg("solid_object"), py::arg("level"))
.def("get_number_of_levels", &MultipleGridBuilder::getNumberOfLevels)
- .def("build_grids", &MultipleGridBuilder::buildGrids);
+ .def("build_grids", &MultipleGridBuilder::buildGrids, py::arg("lbm_or_gks"), py::arg("enable_thin_walls"))
+ .def("set_subdomain_box", &MultipleGridBuilder::setSubDomainBox, py::arg("bounding_box"))
+ .def("find_communication_indices", &MultipleGridBuilder::findCommunicationIndices)
+ .def("set_communication_process", &MultipleGridBuilder::setCommunicationProcess)
+ .def("set_number_of_layers", &MultipleGridBuilder::setNumberOfLayers, py::arg("number_of_layers_fine"), py::arg("number_of_layers_between_levels"));
return gridGeneratorModule;
}
diff --git a/pythonbindings/src/gpu/submodules/grid_provider.cpp b/pythonbindings/src/gpu/submodules/grid_provider.cpp
index 02ff273e2cd1a2022943e19c9a48a447d9dfe54b..717e9d5cd82100636a5398c09662a0895ce8fb56 100644
--- a/pythonbindings/src/gpu/submodules/grid_provider.cpp
+++ b/pythonbindings/src/gpu/submodules/grid_provider.cpp
@@ -1,8 +1,36 @@
+//=======================================================================================
+// ____ ____ __ ______ __________ __ __ __ __
+// \ \ | | | | | _ \ |___ ___| | | | | / \ | |
+// \ \ | | | | | |_) | | | | | | | / \ | |
+// \ \ | | | | | _ / | | | | | | / /\ \ | |
+// \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
+// \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
+// \ \ | | ________________________________________________________________
+// \ \ | | | ______________________________________________________________|
+// \ \| | | | __ __ __ __ ______ _______
+// \ | | |_____ | | | | | | | | | _ \ / _____)
+// \ | | _____| | | | | | | | | | | \ \ \_______
+// \ | | | | |_____ | \_/ | | | | |_/ / _____ |
+// \ _____| |__| |________| \_______/ |__| |______/ (_______/
+//
+// This file is part of VirtualFluids. VirtualFluids is free software: you can
+// redistribute it and/or modify it under the terms of the GNU General Public
+// License as published by the Free Software Foundation, either version 3 of
+// the License, or (at your option) any later version.
+//
+// VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
+// ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+// FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
+// for more details.
+//
+// You should have received a copy of the GNU General Public License along
+// with VirtualFluids (see COPYING.txt). If not, see <http://www.gnu.org/licenses/>.
+//
+//! \file grid_provider
+//! \author Henry Korb
+//=======================================================================================
#include <pybind11/pybind11.h>
#include "gpu/VirtualFluids_GPU/DataStructureInitializer/GridProvider.h"
-// #include <gpu/VirtualFluids_GPU/GPU/CudaMemoryManager.h>
-// #include <gpu/VirtualFluids_GPU/Parameter/Parameter.h>
-// #include "gpu/GridGenerator/grid/GridBuilder/GridBuilder.h"
namespace grid_provider
{
@@ -11,6 +39,6 @@ namespace grid_provider
void makeModule(py::module_ &parentModule)
{
py::class_<GridProvider, std::shared_ptr<GridProvider>>(parentModule, "GridProvider")
- .def("make_grid_generator", &GridProvider::makeGridGenerator, py::return_value_policy::reference);
+ .def_static("make_grid_generator", &GridProvider::makeGridGenerator, py::return_value_policy::reference, py::arg("builder"), py::arg("para"), py::arg("cuda_memory_manager"), py::arg("communicator"));
}
}
\ No newline at end of file
diff --git a/pythonbindings/src/gpu/submodules/grid_scaling_factory.cpp b/pythonbindings/src/gpu/submodules/grid_scaling_factory.cpp
new file mode 100644
index 0000000000000000000000000000000000000000..a3a572875a4695871c482a4308acab4214dbb481
--- /dev/null
+++ b/pythonbindings/src/gpu/submodules/grid_scaling_factory.cpp
@@ -0,0 +1,52 @@
+//=======================================================================================
+// ____ ____ __ ______ __________ __ __ __ __
+// \ \ | | | | | _ \ |___ ___| | | | | / \ | |
+// \ \ | | | | | |_) | | | | | | | / \ | |
+// \ \ | | | | | _ / | | | | | | / /\ \ | |
+// \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
+// \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
+// \ \ | | ________________________________________________________________
+// \ \ | | | ______________________________________________________________|
+// \ \| | | | __ __ __ __ ______ _______
+// \ | | |_____ | | | | | | | | | _ \ / _____)
+// \ | | _____| | | | | | | | | | | \ \ \_______
+// \ | | | | |_____ | \_/ | | | | |_/ / _____ |
+// \ _____| |__| |________| \_______/ |__| |______/ (_______/
+//
+// This file is part of VirtualFluids. VirtualFluids is free software: you can
+// redistribute it and/or modify it under the terms of the GNU General Public
+// License as published by the Free Software Foundation, either version 3 of
+// the License, or (at your option) any later version.
+//
+// VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
+// ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+// FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
+// for more details.
+//
+// You should have received a copy of the GNU General Public License along
+// with VirtualFluids (see COPYING.txt). If not, see <http://www.gnu.org/licenses/>.
+//
+//! \file grid_scaling_factory.cpp
+//! \ingroup submodules
+//! \author Henry Korb
+//=======================================================================================
+#include <pybind11/pybind11.h>
+#include <gpu/VirtualFluids_GPU/Factories/GridScalingFactory.h>
+
+namespace grid_scaling_factory
+{
+ namespace py = pybind11;
+
+ void makeModule(py::module_ &parentModule)
+ {
+
+ py::class_<GridScalingFactory, std::shared_ptr<GridScalingFactory>>(parentModule, "GridScalingFactory")
+ .def(py::init<>())
+ .def("set_scaling_factory", &GridScalingFactory::setScalingFactory, py::arg("scaling_type"));
+
+ py::enum_<GridScalingFactory::GridScaling>(parentModule, "GridScaling")
+ .value("ScaleCompressible", GridScalingFactory::GridScaling::ScaleCompressible)
+ .value("ScaleRhoSq", GridScalingFactory::GridScaling::ScaleRhoSq)
+ .value("NotSpecified", GridScalingFactory::GridScaling::NotSpecified);
+ }
+}
\ No newline at end of file
diff --git a/pythonbindings/src/gpu/submodules/parameter.cpp b/pythonbindings/src/gpu/submodules/parameter.cpp
index 7b4e67f101e3928abbd4262557864ea1d0f45b02..a7c42223e6a5bfa3caa89c0879e4133fc4123ad0 100644
--- a/pythonbindings/src/gpu/submodules/parameter.cpp
+++ b/pythonbindings/src/gpu/submodules/parameter.cpp
@@ -1,10 +1,46 @@
+//=======================================================================================
+// ____ ____ __ ______ __________ __ __ __ __
+// \ \ | | | | | _ \ |___ ___| | | | | / \ | |
+// \ \ | | | | | |_) | | | | | | | / \ | |
+// \ \ | | | | | _ / | | | | | | / /\ \ | |
+// \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
+// \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
+// \ \ | | ________________________________________________________________
+// \ \ | | | ______________________________________________________________|
+// \ \| | | | __ __ __ __ ______ _______
+// \ | | |_____ | | | | | | | | | _ \ / _____)
+// \ | | _____| | | | | | | | | | | \ \ \_______
+// \ | | | | |_____ | \_/ | | | | |_/ / _____ |
+// \ _____| |__| |________| \_______/ |__| |______/ (_______/
+//
+// This file is part of VirtualFluids. VirtualFluids is free software: you can
+// redistribute it and/or modify it under the terms of the GNU General Public
+// License as published by the Free Software Foundation, either version 3 of
+// the License, or (at your option) any later version.
+//
+// VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
+// ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+// FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
+// for more details.
+//
+// You should have received a copy of the GNU General Public License along
+// with VirtualFluids (see COPYING.txt). If not, see <http://www.gnu.org/licenses/>.
+//
+//! \file parameter.cpp
+//! \ingroup submodules
+//! \author Henry Korb
+//=======================================================================================
#include <pybind11/pybind11.h>
#include <pybind11/functional.h>
#include <pybind11/stl.h>
#include <gpu/VirtualFluids_GPU/Parameter/Parameter.h>
+#include "lbm/constants/NumericConstants.h"
#include <basics/config/ConfigurationFile.h>
#include <gpu/VirtualFluids_GPU/PreCollisionInteractor/PreCollisionInteractor.h>
+
+using namespace vf::lbm::constant;
+
namespace parameter
{
namespace py = pybind11;
@@ -13,42 +49,44 @@ namespace parameter
{
py::class_<Parameter, std::shared_ptr<Parameter>>(parentModule, "Parameter")
.def(py::init<
- const vf::basics::ConfigurationFile&,
int,
- int
- >(),
- "config_data",
- "number_of_processes",
- "my_ID")
- .def("set_forcing", &Parameter::setForcing)
- .def("set_diff_on", &Parameter::setDiffOn)
- .def("set_comp_on", &Parameter::setCompOn)
- .def("set_max_level", &Parameter::setMaxLevel)
- .def("set_t_end", &Parameter::setTEnd)
- .def("set_t_out", &Parameter::setTOut)
- .def("set_t_start_out", &Parameter::setTStartOut)
- .def("set_timestep_of_coarse_level", &Parameter::setTimestepOfCoarseLevel)
- .def("set_output_path", &Parameter::setOutputPath)
- .def("set_output_prefix", &Parameter::setOutputPrefix)
- .def("set_f_name", &Parameter::setFName)
- .def("set_print_files", &Parameter::setPrintFiles)
- .def("set_temperature_init", &Parameter::setTemperatureInit)
- .def("set_temperature_BC", &Parameter::setTemperatureBC)
- .def("set_viscosity", &Parameter::setViscosity)
- .def("set_velocity", &Parameter::setVelocity)
- .def("set_viscosity_ratio", &Parameter::setViscosityRatio)
- .def("set_velocity_ratio", &Parameter::setVelocityRatio)
- .def("set_density_ratio", &Parameter::setDensityRatio)
- .def("set_devices", &Parameter::setDevices)
- .def("set_is_body_force", &Parameter::setIsBodyForce)
- .def("set_use_AMD", &Parameter::setUseAMD)
- .def("set_use_Wale", &Parameter::setUseWale)
- .def("set_SGS_constant", &Parameter::setSGSConstant)
- .def("set_main_kernel", &Parameter::setMainKernel)
- .def("set_AD_kernel", &Parameter::setADKernel)
- .def("set_use_AMD", &Parameter::setUseAMD)
- .def("set_use_Wale", &Parameter::setUseWale)
- .def("set_SGS_constant", &Parameter::setSGSConstant)
+ int,
+ std::optional<const vf::basics::ConfigurationFile*>>(),
+ py::arg("number_of_processes"),
+ py::arg("my_ID"),
+ py::arg("config_data"))
+ .def(py::init<int, int>(),
+ py::arg("number_of_processes"),
+ py::arg("my_ID"))
+ .def(py::init<const vf::basics::ConfigurationFile*>(), py::arg("config_data"))
+ .def("set_forcing", &Parameter::setForcing, py::arg("forcing_x"), py::arg("forcing_y"), py::arg("forcing_z"))
+ .def("set_quadric_limiters", &Parameter::setQuadricLimiters, py::arg("quadric_limiter_p"), py::arg("quadric_limiter_m"), py::arg("quadric_limiter_d"))
+ .def("set_diff_on", &Parameter::setDiffOn, py::arg("is_diff"))
+ .def("set_comp_on", &Parameter::setCompOn, py::arg("is_comp"))
+ .def("set_max_level", &Parameter::setMaxLevel, py::arg("number_of_levels"))
+ .def("set_timestep_end", &Parameter::setTimestepEnd, py::arg("tend"))
+ .def("set_timestep_out", &Parameter::setTimestepOut, py::arg("tout"))
+ .def("set_timestep_start_out", &Parameter::setTimestepStartOut, py::arg("t_start_out"))
+ .def("set_timestep_of_coarse_level", &Parameter::setTimestepOfCoarseLevel, py::arg("timestep"))
+ .def("set_calc_turbulence_intensity", &Parameter::setCalcTurbulenceIntensity, py::arg("calc_velocity_and_fluctuations"))
+ .def("set_output_path", &Parameter::setOutputPath, py::arg("o_path"))
+ .def("set_output_prefix", &Parameter::setOutputPrefix, py::arg("o_prefix"))
+ .def("set_print_files", &Parameter::setPrintFiles, py::arg("print_files"))
+ .def("set_temperature_init", &Parameter::setTemperatureInit, py::arg("temp"))
+ .def("set_temperature_BC", &Parameter::setTemperatureBC, py::arg("temp_bc"))
+ .def("set_viscosity_LB", &Parameter::setViscosityLB, py::arg("viscosity"))
+ .def("set_velocity_LB", &Parameter::setVelocityLB, py::arg("velocity"))
+ .def("set_viscosity_ratio", &Parameter::setViscosityRatio, py::arg("viscosity_ratio"))
+ .def("set_velocity_ratio", &Parameter::setVelocityRatio, py::arg("velocity_ratio"))
+ .def("set_density_ratio", &Parameter::setDensityRatio, py::arg("density_ratio"))
+ .def("set_devices", &Parameter::setDevices, py::arg("devices"))
+ .def("set_max_dev", &Parameter::setMaxDev, py::arg("max_dev"))
+ .def("set_is_body_force", &Parameter::setIsBodyForce, py::arg("is_body_force"))
+ .def("set_use_streams", &Parameter::setUseStreams, py::arg("use_streams"))
+ .def("set_main_kernel", &Parameter::setMainKernel, py::arg("kernel"))
+ .def("set_AD_kernel", &Parameter::setADKernel, py::arg("ad_kernel"))
+ .def("set_has_wall_model_monitor", &Parameter::setHasWallModelMonitor, py::arg("has_wall_monitor"))
+ .def("set_outflow_pressure_correction_factor", &Parameter::setOutflowPressureCorrectionFactor, py::arg("correction_factor"))
.def("set_initial_condition", [](Parameter ¶, std::function<std::vector<float>(real, real, real)> &init_func)
{
para.setInitialCondition([init_func](real coordX, real coordY, real coordZ, real& rho, real& vx, real& vy, real& vz)
@@ -59,9 +97,46 @@ namespace parameter
vy = values[2];
vz = values[3];
});
- })
- .def("add_actuator", &Parameter::addActuator)
- .def("add_probe", &Parameter::addProbe)
+ }, py::arg("init_func"))
+ .def("set_initial_condition_uniform", [](Parameter ¶, real velocity_x, real velocity_y, real velocity_z)
+ {
+ para.setInitialCondition([velocity_x, velocity_y, velocity_z](real coordX, real coordY, real coordZ, real& rho, real& vx, real& vy, real& vz) // must capture values explicitly!
+ {
+ rho = c0o1;
+ vx = velocity_x;
+ vy = velocity_y;
+ vz = velocity_z;
+ });
+ }, py::arg("velocity_x"), py::arg("velocity_y"), py::arg("velocity_z"))
+ .def("set_initial_condition_log_law", [](Parameter ¶, real u_star, real z0, real velocityRatio)
+ {
+ para.setInitialCondition(
+ [u_star, z0, velocityRatio](real coordX, real coordY, real coordZ, real& rho, real& vx, real& vy, real& vz)
+ {
+ coordZ = coordZ > c0o1 ? coordZ : c0o1;
+
+ rho = c0o1;
+ vx = u_star/c4o10 * log(coordZ/z0+c1o1) / velocityRatio;
+ vy = c0o1;
+ vz = c0o1;
+ }
+ );
+ }, py::arg("u_star"), py::arg("z0"), py::arg("velocity_ratio"))
+ .def("set_initial_condition_perturbed_log_law", [](Parameter ¶, real u_star, real z0, real L_x, real L_z, real H, real velocityRatio)
+ {
+ para.setInitialCondition(
+ [u_star, z0, L_x, L_z, H, velocityRatio](real coordX, real coordY, real coordZ, real& rho, real& vx, real& vy, real& vz)
+ {
+ coordZ = coordZ > c0o1 ? coordZ : c0o1;
+ rho = c0o1;
+ vx = (u_star/c4o10 * log(coordZ/z0+c1o1) + c2o1*sin(cPi*c16o1*coordX/L_x)*sin(cPi*c8o1*coordZ/H)/(pow(coordZ/H,c2o1)+c1o1)) / velocityRatio;
+ vy = c2o1*sin(cPi*c16o1*coordX/L_x)*sin(cPi*c8o1*coordZ/H)/(pow(coordZ/H,c2o1)+c1o1) / velocityRatio;
+ vz = c8o1*u_star/c4o10*(sin(cPi*c8o1*coordY/H)*sin(cPi*c8o1*coordZ/H)+sin(cPi*c8o1*coordX/L_x))/(pow(c1o2*L_z-coordZ, c2o1)+c1o1) / velocityRatio;
+ }
+ );
+ }, py::arg("u_star"), py::arg("z0"), py::arg("length_x"), py::arg("length_z"), py::arg("height"), py::arg("velocity_ratio"))
+ .def("add_actuator", &Parameter::addActuator, py::arg("actuator"))
+ .def("add_probe", &Parameter::addProbe, py::arg("probe"))
.def("get_output_path", &Parameter::getOutputPath)
.def("get_output_prefix", &Parameter::getOutputPrefix)
.def("get_velocity", &Parameter::getVelocity)
@@ -70,11 +145,9 @@ namespace parameter
.def("get_viscosity_ratio", &Parameter::getViscosityRatio)
.def("get_density_ratio", &Parameter::getDensityRatio)
.def("get_force_ratio", &Parameter::getForceRatio)
- .def("get_use_AMD", &Parameter::getUseAMD)
- .def("get_use_Wale", &Parameter::getUseWale)
.def("get_SGS_constant", &Parameter::getSGSConstant)
.def("get_is_body_force", &Parameter::getIsBodyForce)
- .def("set_has_wall_model_monitor", &Parameter::setHasWallModelMonitor)
;
+
}
}
\ No newline at end of file
diff --git a/pythonbindings/src/gpu/submodules/pre_collision_interactor.cpp b/pythonbindings/src/gpu/submodules/pre_collision_interactor.cpp
index 362ee1a8ce6112cfa9543f1b254e10f3e35822a1..308f6c37aada14c8c25c69245f603274ae2f18d8 100644
--- a/pythonbindings/src/gpu/submodules/pre_collision_interactor.cpp
+++ b/pythonbindings/src/gpu/submodules/pre_collision_interactor.cpp
@@ -1,3 +1,36 @@
+
+//=======================================================================================
+// ____ ____ __ ______ __________ __ __ __ __
+// \ \ | | | | | _ \ |___ ___| | | | | / \ | |
+// \ \ | | | | | |_) | | | | | | | / \ | |
+// \ \ | | | | | _ / | | | | | | / /\ \ | |
+// \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
+// \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
+// \ \ | | ________________________________________________________________
+// \ \ | | | ______________________________________________________________|
+// \ \| | | | __ __ __ __ ______ _______
+// \ | | |_____ | | | | | | | | | _ \ / _____)
+// \ | | _____| | | | | | | | | | | \ \ \_______
+// \ | | | | |_____ | \_/ | | | | |_/ / _____ |
+// \ _____| |__| |________| \_______/ |__| |______/ (_______/
+//
+// This file is part of VirtualFluids. VirtualFluids is free software: you can
+// redistribute it and/or modify it under the terms of the GNU General Public
+// License as published by the Free Software Foundation, either version 3 of
+// the License, or (at your option) any later version.
+//
+// VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
+// ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+// FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
+// for more details.
+//
+// You should have received a copy of the GNU General Public License along
+// with VirtualFluids (see COPYING.txt). If not, see <http://www.gnu.org/licenses/>.
+//
+//! \file pre_collision_interactor.cpp
+//! \ingroup submodules
+//! \author Henry Korb
+//=======================================================================================
#include <pybind11/pybind11.h>
#include <gpu/VirtualFluids_GPU/PreCollisionInteractor/PreCollisionInteractor.h>
diff --git a/pythonbindings/src/gpu/submodules/precursor_writer.cpp b/pythonbindings/src/gpu/submodules/precursor_writer.cpp
new file mode 100644
index 0000000000000000000000000000000000000000..64164ef9993d7b4f22bff2390b418718f7c3208f
--- /dev/null
+++ b/pythonbindings/src/gpu/submodules/precursor_writer.cpp
@@ -0,0 +1,67 @@
+//=======================================================================================
+// ____ ____ __ ______ __________ __ __ __ __
+// \ \ | | | | | _ \ |___ ___| | | | | / \ | |
+// \ \ | | | | | |_) | | | | | | | / \ | |
+// \ \ | | | | | _ / | | | | | | / /\ \ | |
+// \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
+// \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
+// \ \ | | ________________________________________________________________
+// \ \ | | | ______________________________________________________________|
+// \ \| | | | __ __ __ __ ______ _______
+// \ | | |_____ | | | | | | | | | _ \ / _____)
+// \ | | _____| | | | | | | | | | | \ \ \_______
+// \ | | | | |_____ | \_/ | | | | |_/ / _____ |
+// \ _____| |__| |________| \_______/ |__| |______/ (_______/
+//
+// This file is part of VirtualFluids. VirtualFluids is free software: you can
+// redistribute it and/or modify it under the terms of the GNU General Public
+// License as published by the Free Software Foundation, either version 3 of
+// the License, or (at your option) any later version.
+//
+// VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
+// ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+// FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
+// for more details.
+//
+// You should have received a copy of the GNU General Public License along
+// with VirtualFluids (see COPYING.txt). If not, see <http://www.gnu.org/licenses/>.
+//
+//! \file precursor_writer.cpp
+//! \ingroup submodules
+//! \author Henry Korb
+//=======================================================================================
+#include <pybind11/pybind11.h>
+#include <pybind11/stl.h>
+#include <pybind11/numpy.h>
+#include <gpu/VirtualFluids_GPU/PreCollisionInteractor/PreCollisionInteractor.h>
+#include <gpu/VirtualFluids_GPU/PreCollisionInteractor/PrecursorWriter.h>
+
+namespace precursor_writer
+{
+ namespace py = pybind11;
+
+ void makeModule(py::module_ &parentModule)
+ {
+ py::enum_<OutputVariable>(parentModule, "OutputVariable")
+ .value("Velocities", OutputVariable::Velocities)
+ .value("Distributions", OutputVariable::Distributions);
+
+ py::class_<PrecursorWriter, PreCollisionInteractor, std::shared_ptr<PrecursorWriter>>(parentModule, "PrecursorWriter")
+ .def(py::init < std::string,
+ std::string,
+ real,
+ real, real,
+ real, real,
+ uint, uint,
+ OutputVariable,
+ uint>(),
+ py::arg("filename"),
+ py::arg("output_path"),
+ py::arg("x_pos"),
+ py::arg("y_min"), py::arg("y_max"),
+ py::arg("z_min"), py::arg("z_max"),
+ py::arg("t_start_out"), py::arg("t_save"),
+ py::arg("output_variable"),
+ py::arg("max_timesteps_per_file"));
+ }
+}
\ No newline at end of file
diff --git a/pythonbindings/src/gpu/submodules/probes.cpp b/pythonbindings/src/gpu/submodules/probes.cpp
index 6993d9617d870922d7ed90ed9ecbebb8a797be25..7c26958df81a60f00c9909a91f5576a5931652d4 100644
--- a/pythonbindings/src/gpu/submodules/probes.cpp
+++ b/pythonbindings/src/gpu/submodules/probes.cpp
@@ -1,3 +1,35 @@
+//=======================================================================================
+// ____ ____ __ ______ __________ __ __ __ __
+// \ \ | | | | | _ \ |___ ___| | | | | / \ | |
+// \ \ | | | | | |_) | | | | | | | / \ | |
+// \ \ | | | | | _ / | | | | | | / /\ \ | |
+// \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
+// \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
+// \ \ | | ________________________________________________________________
+// \ \ | | | ______________________________________________________________|
+// \ \| | | | __ __ __ __ ______ _______
+// \ | | |_____ | | | | | | | | | _ \ / _____)
+// \ | | _____| | | | | | | | | | | \ \ \_______
+// \ | | | | |_____ | \_/ | | | | |_/ / _____ |
+// \ _____| |__| |________| \_______/ |__| |______/ (_______/
+//
+// This file is part of VirtualFluids. VirtualFluids is free software: you can
+// redistribute it and/or modify it under the terms of the GNU General Public
+// License as published by the Free Software Foundation, either version 3 of
+// the License, or (at your option) any later version.
+//
+// VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
+// ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+// FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
+// for more details.
+//
+// You should have received a copy of the GNU General Public License along
+// with VirtualFluids (see COPYING.txt). If not, see <http://www.gnu.org/licenses/>.
+//
+//! \file probes.cpp
+//! \ingroup submodules
+//! \author Henry Korb
+//=======================================================================================
#include <pybind11/pybind11.h>
#include <pybind11/stl.h>
#include <gpu/VirtualFluids_GPU/PreCollisionInteractor/Probes/Probe.h>
@@ -29,7 +61,7 @@ namespace probes
.value("SpatioTemporalFlatness", Statistic::SpatioTemporalFlatness);
py::class_<Probe, PreCollisionInteractor, std::shared_ptr<Probe>>(probeModule, "Probe")
- .def("add_statistic", &Probe::addStatistic)
+ .def("add_statistic", &Probe::addStatistic, py::arg("variable"))
.def("set_file_name_to_n_out", &Probe::setFileNameToNOut)
.def("add_all_available_statistics", &Probe::addAllAvailableStatistics);
@@ -41,14 +73,14 @@ namespace probes
uint,
uint,
uint>(),
- "probe_name",
- "output_path"
- "t_start_avg",
- "t_avg",
- "t_start_out",
- "t_out")
- .def("add_probe_points_from_list", &PointProbe::addProbePointsFromList)
- .def("add_probe_points_from_x_normal_plane", &PointProbe::addProbePointsFromXNormalPlane);
+ py::arg("probe_name"),
+ py::arg("output_path"),
+ py::arg("t_start_avg"),
+ py::arg("t_avg"),
+ py::arg("t_start_out"),
+ py::arg("t_out"))
+ .def("add_probe_points_from_list", &PointProbe::addProbePointsFromList, py::arg("point_coords_x"), py::arg("point_coords_y"), py::arg("point_coords_z"))
+ .def("add_probe_points_from_x_normal_plane", &PointProbe::addProbePointsFromXNormalPlane, py::arg("pos_x"), py::arg("pos0_y"), py::arg("pos0_z"), py::arg("pos1_y"), py::arg("pos1_z"), py::arg("n_y"), py::arg("n_z"));
py::class_<PlaneProbe, Probe, std::shared_ptr<PlaneProbe>>(probeModule, "PlaneProbe")
.def(py::init<
@@ -58,13 +90,13 @@ namespace probes
uint,
uint,
uint>(),
- "probe_name",
- "output_path"
- "t_start_avg",
- "t_avg",
- "t_start_out",
- "t_out")
- .def("set_probe_plane", &PlaneProbe::setProbePlane);
+ py::arg("probe_name"),
+ py::arg("output_path"),
+ py::arg("t_start_avg"),
+ py::arg("t_avg"),
+ py::arg("t_start_out"),
+ py::arg("t_out"))
+ .def("set_probe_plane", &PlaneProbe::setProbePlane, py::arg("pos_x"), py::arg("pos_y"), py::arg("pos_z"), py::arg("delta_x"), py::arg("delta_y"), py::arg("delta_z"));
py::class_<PlanarAverageProbe, Probe, std::shared_ptr<PlanarAverageProbe>>(probeModule, "PlanarAverageProbe")
.def(py::init<
@@ -76,14 +108,14 @@ namespace probes
uint,
uint,
char>(),
- "probe_name",
- "output_path",
- "t_start_avg",
- "t_start_tmp_avg",
- "t_avg",
- "t_start_out",
- "t_out",
- "plane_normal");
+ py::arg("probe_name"),
+ py::arg("output_path"),
+ py::arg("t_start_avg"),
+ py::arg("t_start_tmp_avg"),
+ py::arg("t_avg"),
+ py::arg("t_start_out"),
+ py::arg("t_out"),
+ py::arg("plane_normal"));
py::class_<WallModelProbe, Probe, std::shared_ptr<WallModelProbe>>(probeModule, "WallModelProbe")
@@ -95,15 +127,15 @@ namespace probes
uint,
uint,
uint>(),
- "probe_name",
- "output_path"
- "t_start_avg",
- "t_start_tmp_avg",
- "t_avg",
- "t_start_out",
- "t_out")
- .def("set_force_output_to_stress", &WallModelProbe::setForceOutputToStress)
- .def("set_evaluate_pressure_gradient", &WallModelProbe::setEvaluatePressureGradient);
+ py::arg("probe_name"),
+ py::arg("output_path"),
+ py::arg("t_start_avg"),
+ py::arg("t_start_tmp_avg"),
+ py::arg("t_avg"),
+ py::arg("t_start_out"),
+ py::arg("t_out"))
+ .def("set_force_output_to_stress", &WallModelProbe::setForceOutputToStress, py::arg("output_stress"))
+ .def("set_evaluate_pressure_gradient", &WallModelProbe::setEvaluatePressureGradient, py::arg("eval_press_grad"));
return probeModule;
}
diff --git a/pythonbindings/src/gpu/submodules/simulation.cpp b/pythonbindings/src/gpu/submodules/simulation.cpp
index b775d604ba41530223f22738c72785b2c15348b3..d32ef272a1fd26510439dde6ab3a9438d68009a7 100644
--- a/pythonbindings/src/gpu/submodules/simulation.cpp
+++ b/pythonbindings/src/gpu/submodules/simulation.cpp
@@ -1,3 +1,35 @@
+//=======================================================================================
+// ____ ____ __ ______ __________ __ __ __ __
+// \ \ | | | | | _ \ |___ ___| | | | | / \ | |
+// \ \ | | | | | |_) | | | | | | | / \ | |
+// \ \ | | | | | _ / | | | | | | / /\ \ | |
+// \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
+// \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
+// \ \ | | ________________________________________________________________
+// \ \ | | | ______________________________________________________________|
+// \ \| | | | __ __ __ __ ______ _______
+// \ | | |_____ | | | | | | | | | _ \ / _____)
+// \ | | _____| | | | | | | | | | | \ \ \_______
+// \ | | | | |_____ | \_/ | | | | |_/ / _____ |
+// \ _____| |__| |________| \_______/ |__| |______/ (_______/
+//
+// This file is part of VirtualFluids. VirtualFluids is free software: you can
+// redistribute it and/or modify it under the terms of the GNU General Public
+// License as published by the Free Software Foundation, either version 3 of
+// the License, or (at your option) any later version.
+//
+// VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
+// ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+// FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
+// for more details.
+//
+// You should have received a copy of the GNU General Public License along
+// with VirtualFluids (see COPYING.txt). If not, see <http://www.gnu.org/licenses/>.
+//
+//! \file simulation.cpp
+//! \ingroup submodules
+//! \author Henry Korb
+//=======================================================================================
#include <pybind11/pybind11.h>
#include <gpu/VirtualFluids_GPU/LBM/Simulation.h>
#include <gpu/VirtualFluids_GPU/Communication/Communicator.h>
@@ -8,6 +40,9 @@
#include <gpu/VirtualFluids_GPU/GPU/CudaMemoryManager.h>
#include <gpu/VirtualFluids_GPU/DataStructureInitializer/GridProvider.h>
#include <gpu/VirtualFluids_GPU/Output/DataWriter.h>
+#include "gpu/VirtualFluids_GPU/Factories/BoundaryConditionFactory.h"
+#include "gpu/VirtualFluids_GPU/TurbulenceModels/TurbulenceModelFactory.h"
+#include "gpu/VirtualFluids_GPU/Factories/GridScalingFactory.h"
namespace simulation
{
@@ -20,13 +55,41 @@ namespace simulation
.def(py::init< std::shared_ptr<Parameter>,
std::shared_ptr<CudaMemoryManager>,
vf::gpu::Communicator &,
- GridProvider &>(),
- "parameter",
- "memoryManager",
- "communicator",
- "gridProvider")
+ GridProvider &,
+ BoundaryConditionFactory*,
+ GridScalingFactory*>(),
+ py::arg("parameter"),
+ py::arg("memoryManager"),
+ py::arg("communicator"),
+ py::arg("gridProvider"),
+ py::arg("bcFactory"),
+ py::arg("gridScalingFactory"))
+ .def(py::init< std::shared_ptr<Parameter>,
+ std::shared_ptr<CudaMemoryManager>,
+ vf::gpu::Communicator &,
+ GridProvider &,
+ BoundaryConditionFactory*>(),
+ py::arg("parameter"),
+ py::arg("memoryManager"),
+ py::arg("communicator"),
+ py::arg("gridProvider"),
+ py::arg("bcFactory"))
+ .def(py::init< std::shared_ptr<Parameter>,
+ std::shared_ptr<CudaMemoryManager>,
+ vf::gpu::Communicator &,
+ GridProvider &,
+ BoundaryConditionFactory*,
+ std::shared_ptr<TurbulenceModelFactory>,
+ GridScalingFactory*>(),
+ py::arg("parameter"),
+ py::arg("memoryManager"),
+ py::arg("communicator"),
+ py::arg("gridProvider"),
+ py::arg("bcFactory"),
+ py::arg("tmFactory"),
+ py::arg("gridScalingFactory"))
.def("run", &Simulation::run)
- .def("addKineticEnergyAnalyzer", &Simulation::addKineticEnergyAnalyzer)
- .def("addEnstrophyAnalyzer", &Simulation::addEnstrophyAnalyzer);
+ .def("addKineticEnergyAnalyzer", &Simulation::addKineticEnergyAnalyzer, py::arg("t_analyse"))
+ .def("addEnstrophyAnalyzer", &Simulation::addEnstrophyAnalyzer, py::arg("t_analyse"));
}
}
\ No newline at end of file
diff --git a/pythonbindings/src/gpu/submodules/transient_bc_setter.cpp b/pythonbindings/src/gpu/submodules/transient_bc_setter.cpp
new file mode 100644
index 0000000000000000000000000000000000000000..89370ef4c1b91a0c8e480e968a1df3bd4fe540ca
--- /dev/null
+++ b/pythonbindings/src/gpu/submodules/transient_bc_setter.cpp
@@ -0,0 +1,52 @@
+//=======================================================================================
+// ____ ____ __ ______ __________ __ __ __ __
+// \ \ | | | | | _ \ |___ ___| | | | | / \ | |
+// \ \ | | | | | |_) | | | | | | | / \ | |
+// \ \ | | | | | _ / | | | | | | / /\ \ | |
+// \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
+// \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
+// \ \ | | ________________________________________________________________
+// \ \ | | | ______________________________________________________________|
+// \ \| | | | __ __ __ __ ______ _______
+// \ | | |_____ | | | | | | | | | _ \ / _____)
+// \ | | _____| | | | | | | | | | | \ \ \_______
+// \ | | | | |_____ | \_/ | | | | |_/ / _____ |
+// \ _____| |__| |________| \_______/ |__| |______/ (_______/
+//
+// This file is part of VirtualFluids. VirtualFluids is free software: you can
+// redistribute it and/or modify it under the terms of the GNU General Public
+// License as published by the Free Software Foundation, either version 3 of
+// the License, or (at your option) any later version.
+//
+// VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
+// ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+// FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
+// for more details.
+//
+// You should have received a copy of the GNU General Public License along
+// with VirtualFluids (see COPYING.txt). If not, see <http://www.gnu.org/licenses/>.
+//
+//! \file transient_bc_setter.cpp
+//! \ingroup submodules
+//! \author Henry Korb
+//=======================================================================================
+#include <pybind11/pybind11.h>
+#include <gpu/GridGenerator/TransientBCSetter/TransientBCSetter.h>
+
+namespace transient_bc_setter
+{
+ namespace py = pybind11;
+
+ void makeModule(py::module_ &parentModule)
+ {
+ py::enum_<FileType>(parentModule, "FileType")
+ .value("VTK", FileType::VTK);
+
+ parentModule.def("create_file_collection", &createFileCollection, py::arg("prefix"), py::arg("type"));
+
+ py::class_<FileCollection, std::shared_ptr<FileCollection>>(parentModule, "FileCollection");
+
+ py::class_<VTKFileCollection, FileCollection, std::shared_ptr<VTKFileCollection>>(parentModule, "VTKFileCollection")
+ .def(py::init <std::string>(), py::arg("prefix"));
+ }
+}
\ No newline at end of file
diff --git a/pythonbindings/src/gpu/submodules/turbulence_models.cpp b/pythonbindings/src/gpu/submodules/turbulence_models.cpp
new file mode 100644
index 0000000000000000000000000000000000000000..cfbb9e56127fee0cd90a482dde258d8b96389989
--- /dev/null
+++ b/pythonbindings/src/gpu/submodules/turbulence_models.cpp
@@ -0,0 +1,56 @@
+//=======================================================================================
+// ____ ____ __ ______ __________ __ __ __ __
+// \ \ | | | | | _ \ |___ ___| | | | | / \ | |
+// \ \ | | | | | |_) | | | | | | | / \ | |
+// \ \ | | | | | _ / | | | | | | / /\ \ | |
+// \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
+// \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
+// \ \ | | ________________________________________________________________
+// \ \ | | | ______________________________________________________________|
+// \ \| | | | __ __ __ __ ______ _______
+// \ | | |_____ | | | | | | | | | _ \ / _____)
+// \ | | _____| | | | | | | | | | | \ \ \_______
+// \ | | | | |_____ | \_/ | | | | |_/ / _____ |
+// \ _____| |__| |________| \_______/ |__| |______/ (_______/
+//
+// This file is part of VirtualFluids. VirtualFluids is free software: you can
+// redistribute it and/or modify it under the terms of the GNU General Public
+// License as published by the Free Software Foundation, either version 3 of
+// the License, or (at your option) any later version.
+//
+// VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
+// ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+// FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
+// for more details.
+//
+// You should have received a copy of the GNU General Public License along
+// with VirtualFluids (see COPYING.txt). If not, see <http://www.gnu.org/licenses/>.
+//
+//! \file turbulence_models.cpp
+//! \ingroup submodules
+//! \author Henry Korb
+//=======================================================================================
+#include "pybind11/pybind11.h"
+#include "gpu/VirtualFluids_GPU/TurbulenceModels/TurbulenceModelFactory.h"
+#include "gpu/VirtualFluids_GPU/LBM/LB.h"
+
+namespace turbulence_model
+{
+ namespace py = pybind11;
+
+ void makeModule(py::module_ &parentModule)
+ {
+ py::enum_<TurbulenceModel>(parentModule, "TurbulenceModel")
+ .value("Smagorinsky", TurbulenceModel::Smagorinsky)
+ .value("AMD", TurbulenceModel::AMD)
+ .value("QR", TurbulenceModel::QR)
+ .value("None", TurbulenceModel::None);
+
+ py::class_<TurbulenceModelFactory, std::shared_ptr<TurbulenceModelFactory>>(parentModule, "TurbulenceModelFactory")
+ .def(py::init< std::shared_ptr<Parameter>>(), py::arg("para"))
+ .def("set_turbulence_model", &TurbulenceModelFactory::setTurbulenceModel, py::arg("turbulence_model"))
+ .def("set_model_constant", &TurbulenceModelFactory::setModelConstant, py::arg("model_constant"))
+ .def("read_config_file", &TurbulenceModelFactory::readConfigFile, py::arg("config_data"));
+
+ }
+}
\ No newline at end of file
diff --git a/pythonbindings/src/lbm/lbm.cpp b/pythonbindings/src/lbm/lbm.cpp
index 441b9ff372f4e4513fee58c4a8a1cd78d38582dd..90fd4a71b0101469666936c89974de316e0e2b18 100644
--- a/pythonbindings/src/lbm/lbm.cpp
+++ b/pythonbindings/src/lbm/lbm.cpp
@@ -1,3 +1,35 @@
+//=======================================================================================
+// ____ ____ __ ______ __________ __ __ __ __
+// \ \ | | | | | _ \ |___ ___| | | | | / \ | |
+// \ \ | | | | | |_) | | | | | | | / \ | |
+// \ \ | | | | | _ / | | | | | | / /\ \ | |
+// \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
+// \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
+// \ \ | | ________________________________________________________________
+// \ \ | | | ______________________________________________________________|
+// \ \| | | | __ __ __ __ ______ _______
+// \ | | |_____ | | | | | | | | | _ \ / _____)
+// \ | | _____| | | | | | | | | | | \ \ \_______
+// \ | | | | |_____ | \_/ | | | | |_/ / _____ |
+// \ _____| |__| |________| \_______/ |__| |______/ (_______/
+//
+// This file is part of VirtualFluids. VirtualFluids is free software: you can
+// redistribute it and/or modify it under the terms of the GNU General Public
+// License as published by the Free Software Foundation, either version 3 of
+// the License, or (at your option) any later version.
+//
+// VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
+// ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+// FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
+// for more details.
+//
+// You should have received a copy of the GNU General Public License along
+// with VirtualFluids (see COPYING.txt). If not, see <http://www.gnu.org/licenses/>.
+//
+//! \file lbm.cpp
+//! \ingroup lbm
+//! \author Henry Korb
+//=======================================================================================
#include <pybind11/pybind11.h>
namespace lbm
diff --git a/pythonbindings/src/logger/logger.cpp b/pythonbindings/src/logger/logger.cpp
index 82ad3d92760ae38c0eb62b16be726e4eeaca08ac..555b502fa9a56299895de0fa6dd6cfeb66c15024 100644
--- a/pythonbindings/src/logger/logger.cpp
+++ b/pythonbindings/src/logger/logger.cpp
@@ -1,3 +1,35 @@
+//=======================================================================================
+// ____ ____ __ ______ __________ __ __ __ __
+// \ \ | | | | | _ \ |___ ___| | | | | / \ | |
+// \ \ | | | | | |_) | | | | | | | / \ | |
+// \ \ | | | | | _ / | | | | | | / /\ \ | |
+// \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
+// \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
+// \ \ | | ________________________________________________________________
+// \ \ | | | ______________________________________________________________|
+// \ \| | | | __ __ __ __ ______ _______
+// \ | | |_____ | | | | | | | | | _ \ / _____)
+// \ | | _____| | | | | | | | | | | \ \ \_______
+// \ | | | | |_____ | \_/ | | | | |_/ / _____ |
+// \ _____| |__| |________| \_______/ |__| |______/ (_______/
+//
+// This file is part of VirtualFluids. VirtualFluids is free software: you can
+// redistribute it and/or modify it under the terms of the GNU General Public
+// License as published by the Free Software Foundation, either version 3 of
+// the License, or (at your option) any later version.
+//
+// VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
+// ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+// FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
+// for more details.
+//
+// You should have received a copy of the GNU General Public License along
+// with VirtualFluids (see COPYING.txt). If not, see <http://www.gnu.org/licenses/>.
+//
+//! \file logging.cpp
+//! \ingroup logger
+//! \author Henry Korb
+//=======================================================================================
#include <pybind11/pybind11.h>
#include <logger/Logger.h>
@@ -10,15 +42,15 @@ namespace logging
py::module loggerModule = parentModule.def_submodule("logger");
py::class_<vf::logging::Logger>(loggerModule, "Logger")
- .def("initialize_logger", &vf::logging::Logger::initalizeLogger)
- .def("change_log_path", &vf::logging::Logger::changeLogPath);
+ .def_static("initialize_logger", &vf::logging::Logger::initalizeLogger)
+ .def_static("change_log_path", &vf::logging::Logger::changeLogPath, py::arg("path"));
// use f-strings (f"text {float}") in python for compounded messages
- loggerModule.def("vf_log_trace", [](std::string arg){ VF_LOG_TRACE(arg); });
- loggerModule.def("vf_log_debug", [](std::string arg){ VF_LOG_DEBUG(arg); });
- loggerModule.def("vf_log_info", [](std::string arg){ VF_LOG_INFO(arg); });
- loggerModule.def("vf_log_warning", [](std::string arg){ VF_LOG_WARNING(arg); });
- loggerModule.def("vf_log_critical", [](std::string arg){ VF_LOG_CRITICAL(arg); });
+ loggerModule.def("vf_log_trace", [](std::string message){ VF_LOG_TRACE(message); }, py::arg("message"));
+ loggerModule.def("vf_log_debug", [](std::string message){ VF_LOG_DEBUG(message); }, py::arg("message"));
+ loggerModule.def("vf_log_info", [](std::string message){ VF_LOG_INFO(message); }, py::arg("message"));
+ loggerModule.def("vf_log_warning", [](std::string message){ VF_LOG_WARNING(message); }, py::arg("message"));
+ loggerModule.def("vf_log_critical", [](std::string message){ VF_LOG_CRITICAL(message); }, py::arg("message"));
return loggerModule;
}
diff --git a/pythonbindings/src/muParser.cpp b/pythonbindings/src/muParser.cpp
index 47408c2758fc92991f1be3113d78b8741215b152..eec39de0b72c21aaa924ea805414847aa9de4492 100644
--- a/pythonbindings/src/muParser.cpp
+++ b/pythonbindings/src/muParser.cpp
@@ -1,9 +1,41 @@
+//=======================================================================================
+// ____ ____ __ ______ __________ __ __ __ __
+// \ \ | | | | | _ \ |___ ___| | | | | / \ | |
+// \ \ | | | | | |_) | | | | | | | / \ | |
+// \ \ | | | | | _ / | | | | | | / /\ \ | |
+// \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
+// \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
+// \ \ | | ________________________________________________________________
+// \ \ | | | ______________________________________________________________|
+// \ \| | | | __ __ __ __ ______ _______
+// \ | | |_____ | | | | | | | | | _ \ / _____)
+// \ | | _____| | | | | | | | | | | \ \ \_______
+// \ | | | | |_____ | \_/ | | | | |_/ / _____ |
+// \ _____| |__| |________| \_______/ |__| |______/ (_______/
+//
+// This file is part of VirtualFluids. VirtualFluids is free software: you can
+// redistribute it and/or modify it under the terms of the GNU General Public
+// License as published by the Free Software Foundation, either version 3 of
+// the License, or (at your option) any later version.
+//
+// VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
+// ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+// FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
+// for more details.
+//
+// You should have received a copy of the GNU General Public License along
+// with VirtualFluids (see COPYING.txt). If not, see <http://www.gnu.org/licenses/>.
+//
+//! \file muParser.cpp
+//! \ingroup src
+//! \author Henry Korb
+//=======================================================================================
#include <pybind11/pybind11.h>
#include <muParser.h>
namespace py = pybind11;
-PYBIND11_MODULE(pymuparser, m) {
+PYBIND11_MODULE(bindings, m) {
py::class_<mu::ParserBase>(m, "_ParserBase");
py::class_<mu::Parser, mu::ParserBase>(m, "Parser")
diff --git a/regression-tests/driven_cavity_test.sh b/regression-tests/driven_cavity_test.sh
index e10a829d2680ab647ba0f66e0f2e85a70186007e..7f799facb4459ddafcd8b210a5477954af1444cb 100755
--- a/regression-tests/driven_cavity_test.sh
+++ b/regression-tests/driven_cavity_test.sh
@@ -7,8 +7,8 @@
# build VirtualFluids accordingly to our specific test scenario.
# in this case adding -DUSER_APPS="apps/gpu/LBM/DrivenCavity to the cmake command is not necessary, because the DrivenCavity is added to VirtualFluids by default.
mkdir -p build
-cmake -B build --preset=gpu_make -DCMAKE_CUDA_ARCHITECTURES=75 #-DUSER_APPS="apps/gpu/LBM/DrivenCavity"
-cd build && make -j 8 && cd ..
+cmake -B build --preset=make_gpu -DCMAKE_BUILD_TYPE=Release -DCMAKE_CUDA_ARCHITECTURES=75 #-DUSER_APPS="apps/gpu/LBM/DrivenCavity"
+cmake --build build --parallel 8
# execute VirtualFluids
./build/bin/DrivenCavity
@@ -21,4 +21,4 @@ PATH_TO_DIR=output/DrivenCavity
PATH_TO_REFERENCE_DIR=regression-tests/reference_data/regression_tests/gpu/DrivenCavity_2Levels
# execute fieldcompare (A more comprehensive manual can be found here https://gitlab.com/dglaeser/fieldcompare)
-fieldcompare dir $PATH_TO_DIR --reference $PATH_TO_REFERENCE_DIR --include-files "*.vtu"
\ No newline at end of file
+fieldcompare dir $PATH_TO_DIR $PATH_TO_REFERENCE_DIR --include-files "*.vtu"
diff --git a/regression-tests/driven_cavity_uniform_test.sh b/regression-tests/driven_cavity_uniform_test.sh
new file mode 100755
index 0000000000000000000000000000000000000000..95e2bab635d3a6a73fb514a1f67902083c98e5d3
--- /dev/null
+++ b/regression-tests/driven_cavity_uniform_test.sh
@@ -0,0 +1,25 @@
+#!/bin/bash
+
+#################################
+# Driven Cavity Regression Test
+#################################
+
+# build VirtualFluids accordingly to our specific test scenario.
+# in this case adding -DUSER_APPS="apps/gpu/LBM/DrivenCavity to the cmake command is not necessary, because the DrivenCavity is added to VirtualFluids by default.
+mkdir -p build
+cmake -B build --preset=make_gpu -DCMAKE_BUILD_TYPE=Release -DCMAKE_CUDA_ARCHITECTURES=75 -DUSER_APPS="apps/gpu/LBM/DrivenCavityUniform"
+cmake --build build --parallel 8
+
+# execute VirtualFluids
+./build/bin/DrivenCavityUniform
+
+
+# set the path to the produced data
+PATH_TO_DIR=output/DrivenCavity_uniform
+
+# set the path to the reference data.
+# `regression-tests/reference_data` is fix `regression_tests/gpu/DrivenCavity_uniform_2022_12_16` must match the structure in https://github.com/irmb/test_data:
+PATH_TO_REFERENCE_DIR=regression-tests/reference_data/regression_tests/gpu/DrivenCavity_uniform
+
+# execute fieldcompare (A more comprehensive manual can be found here https://gitlab.com/dglaeser/fieldcompare)
+fieldcompare dir $PATH_TO_DIR $PATH_TO_REFERENCE_DIR --include-files "*.vtu"
diff --git a/regression-tests/multigpu_test/rocket.yml b/regression-tests/multigpu_test/rocket.yml
new file mode 100755
index 0000000000000000000000000000000000000000..f621b1349c042e02f2e834e697147da0822ffe1f
--- /dev/null
+++ b/regression-tests/multigpu_test/rocket.yml
@@ -0,0 +1,48 @@
+host: $PHOENIX_REMOTE_HOST
+user: $PHOENIX_REMOTE_USER
+private_keyfile: $PHOENIX_PRIVATE_KEY
+
+copy:
+ - from: regression-tests/multigpu_test/slurm.job
+ to: multigpu_test/slurm.job
+ overwrite: true
+
+ - from: "CMake/"
+ to: "multigpu_test/CMake/"
+ overwrite: true
+
+ - from: "3rdParty/"
+ to: "multigpu_test/3rdParty/"
+ overwrite: true
+
+ - from: "CMakeLists.txt"
+ to: "multigpu_test/CMakeLists.txt"
+ overwrite: true
+
+ - from: "gpu.cmake"
+ to: "multigpu_test/gpu.cmake"
+ overwrite: true
+
+ - from: "src/"
+ to: "multigpu_test/src/"
+ overwrite: true
+
+ - from: "CMakePresets.json"
+ to: "multigpu_test/CMakePresets.json"
+ overwrite: true
+
+ - from: "apps/gpu/LBM/"
+ to: "multigpu_test/apps/gpu/LBM/"
+ overwrite: true
+
+collect:
+ - from: multigpu_test/output/
+ to: output/results/
+ overwrite: true
+
+ - from: multigpu_test/slurmMultiGPU.out
+ to: output/slurmMultiGPU.out
+ overwrite: true
+
+sbatch: multigpu_test/slurm.job
+continue_if_job_fails: true
diff --git a/regression-tests/multigpu_test/slurm.job b/regression-tests/multigpu_test/slurm.job
new file mode 100755
index 0000000000000000000000000000000000000000..0ee0df46ab64bab6520f9f46fc939d5b3186fae7
--- /dev/null
+++ b/regression-tests/multigpu_test/slurm.job
@@ -0,0 +1,29 @@
+#!/bin/bash -l
+
+#SBATCH --partition=gpu01_queue
+#SBATCH --nodes=1
+#SBATCH --time=10:00:00
+#SBATCH --job-name=Cavity4GPU
+#SBATCH --ntasks-per-node=4
+#SBATCH --gres=gpu:4
+#SBATCH --output=multigpu_test/slurmMultiGPU.out
+##SBATCH --exclusive
+
+module purge
+module load comp/ccache/4.1 # loads comp/gcc/9.3.0
+module load mpi/openmpi/4.0.5_gcc_9.3/openmpi
+module load cuda/11.3
+module load comp/git/2.27.0
+PATH=/home/irmb/tools/cmake-3.20.3-linux-x86_64/bin:$PATH
+
+module list
+
+cd multigpu_test
+mkdir -p build
+cd build
+cmake .. -DBUILD_VF_GPU=ON -DCMAKE_CUDA_ARCHITECTURES=60 -DUSER_APPS="apps/gpu/LBM/DrivenCavityMultiGPU"
+make -j 16
+cd ..
+mkdir -p output
+
+mpirun -np 4 "./build/bin/DrivenCavityMultiGPU" "configPhoenix4GPU.txt"
\ No newline at end of file
diff --git a/regression-tests/regression-tests.sh b/regression-tests/regression-tests.sh
index 5b7d227907594b727103be91d2382c05a07b9c6f..9f5dc8cf758b380709fcc9ad8020d1335f760f64 100755
--- a/regression-tests/regression-tests.sh
+++ b/regression-tests/regression-tests.sh
@@ -13,11 +13,10 @@ git clone https://github.com/irmb/test_data regression-tests/reference_data
# by cloning our meshio patch and fieldcompare into a venv
python3 -m venv .venv
source .venv/bin/activate
-pip install rich
-pip install git+https://github.com/soerenPeters/meshio@update-pyproject-version
-pip install git+https://gitlab.com/dglaeser/fieldcompare
+pip install fieldcompare
# 3. Running the specific tests
+./regression-tests/driven_cavity_uniform_test.sh
./regression-tests/driven_cavity_test.sh
diff --git a/setup.cfg b/setup.cfg
new file mode 100644
index 0000000000000000000000000000000000000000..5894f9dec06953c3eeb909af96db9cb19d202d65
--- /dev/null
+++ b/setup.cfg
@@ -0,0 +1,11 @@
+[metadata]
+name = pyfluids
+description = Python binding for VirtualFluids
+long_description = file: README.md
+long_description_content_type = text/markdown
+platforms = any
+url = https://git.rz.tu-bs.de/irmb/virtualfluids
+version = 0.1.0
+
+[options]
+python_requires = >=3.6
diff --git a/setup.py b/setup.py
index b26e1c13d09447d17f8e9fd6e2cd0d0671595bf3..530431b3775970b5222bc87d32bfb407363f95d6 100644
--- a/setup.py
+++ b/setup.py
@@ -1,137 +1,72 @@
-import os
-import re
import sys
-import platform
-import subprocess
+from pathlib import Path
+from typing import List
-from setuptools import setup, Extension
-from setuptools.command.build_ext import build_ext
-from setuptools.command.install import install
-from setuptools.command.develop import develop
-from distutils.version import LooseVersion
+import skbuild
"""
-Install python wrapper of virtual fluids
-Install GPU backend with option --GPU
-(pass to pip via --install-option="--GPU")
+Install python wrapper of Virtual Fluids
+install via python:
+ python setup.py install
+ set CMAKE Flags via -DBUILD_VF_GPU:BOOL=ON
+ CMAKE flags have to be separated by --
+ example: python setup.py install -- -DBUILD_VF_CPU:BOOL=ON
+or install via pip:
+ pip install .
+ for pip>21:
+ set CMAKE Flags via --config-settings "-DBUILD_VF_GPU=ON"
+ example: pip install . --config-settings="-DBUILD_VF_GPU=ON"
+ each option has to be passed in individually i.e --config-settings="-DOPT1=ON" --config-settings="-DOPT2=OFF"
+ for pip <21:
+ set CMAKE Flags via --global-option ="-DBUILD_VF_GPU=ON"
+ example: pip install . --global-option="-DBUILD_VF_GPU=ON"
"""
-vf_cmake_args = [
- "-DBUILD_VF_PYTHON_BINDINGS=ON",
- "-DCMAKE_CXX_COMPILER_LAUNCHER=ccache",
- "-DCMAKE_CUDA_COMPILER_LAUNCHER=ccache",
- "-DCMAKE_C_COMPILER_LAUNCHER=ccache",
- "-DBUILD_SHARED_LIBS=OFF",
- "-DBUILD_WARNINGS_AS_ERRORS=OFF"
-]
-
-vf_cpu_cmake_args = [
- "-DBUILD_VF_DOUBLE_ACCURACY=ON",
- "-DBUILD_VF_CPU:BOOL=ON",
- "-DBUILD_VF_UNIT_TESTS:BOOL=ON",
- "-DUSE_METIS=ON",
- "-DUSE_MPI=ON"
-]
-
-vf_gpu_cmake_args = [
- "-DBUILD_VF_DOUBLE_ACCURACY=OFF",
- "-DBUILD_VF_GPU:BOOL=ON",
- "-DBUILD_VF_UNIT_TESTS:BOOL=OFF",
-]
-
-GPU = False
-
-class CommandMixin:
- user_options = [
- ('GPU', None, 'compile pyfluids with GPU backend'),
+package_name = "pyfluids"
+target = "python_bindings"
+src_dir = "pythonbindings"
+stub_package = package_name+"-stubs"
+
+stub_dir = Path(src_dir)/stub_package
+
+
+def add_subfiles(dir_path: Path, suffix: str, root_dir: Path) -> List[str]:
+ files = []
+ for f in dir_path.iterdir():
+ if f.is_dir():
+ files.extend(add_subfiles(f, suffix, root_dir))
+ if f.is_file():
+ if f.suffix != suffix:
+ continue
+ files.append(str(f.relative_to(root_dir)))
+ return files
+
+def add_directory(dir_path: Path, suffix: str):
+ return add_subfiles(dir_path, suffix, dir_path)
+
+stub_files = add_directory(stub_dir, ".pyi")
+
+# hack to get config-args for installation with pip>21
+cmake_args = []
+if "config_args" in locals():
+ cmake_args.extend([f"{k}={v}" for k, v in locals()["config_args"].items()])
+
+cmake_args += [
+ f"-DPython3_ROOT_DIR={Path(sys.prefix)}",
+ "-DBUILD_VF_PYTHON_BINDINGS=ON",
+ "-DBUILD_SHARED_LIBS=OFF",
+ "-DBUILD_VF_DOUBLE_ACCURACY=OFF",
+ "-DBUILD_VF_UNIT_TESTS:BOOL=OFF",
+ "-DBUILD_WARNINGS_AS_ERRORS=OFF",
]
- def initialize_options(self):
- super().initialize_options()
- self.GPU = False
-
- def finalize_options(self):
- super().finalize_options()
-
- def run(self):
- global GPU
- GPU = GPU or self.GPU
- super().run()
-
-
-class InstallCommand(CommandMixin, install):
- user_options = getattr(install, 'user_options', []) + CommandMixin.user_options
-
-
-class DevelopCommand(CommandMixin, develop):
- user_options = getattr(develop, 'user_options', []) + CommandMixin.user_options
-
-
-class CMakeExtension(Extension):
- def __init__(self, name, sourcedir=''):
- Extension.__init__(self, name, sources=[])
- self.sourcedir = os.path.abspath(sourcedir)
-
-
-class CMakeBuild(CommandMixin, build_ext):
- user_options = getattr(build_ext, 'user_options', []) + CommandMixin.user_options
-
- def run(self):
- super().run()
- try:
- out = subprocess.check_output(['cmake', '--version'])
- except OSError:
- raise RuntimeError("CMake must be installed to build the following extensions: " +
- ", ".join(e.name for e in self.extensions))
-
- if platform.system() == "Windows":
- cmake_version = LooseVersion(re.search(r'version\s*([\d.]+)', out.decode()).group(1))
- if cmake_version < '3.1.0':
- raise RuntimeError("CMake >= 3.1.0 is required on Windows")
-
- for ext in self.extensions:
- self.build_extension(ext)
-
- def build_extension(self, ext):
- extdir = os.path.abspath(os.path.dirname(self.get_ext_fullpath(ext.name)))
- # required for auto-detection of auxiliary "native" libs
- if not extdir.endswith(os.path.sep):
- extdir += os.path.sep
-
- cmake_args = ['-DCMAKE_LIBRARY_OUTPUT_DIRECTORY=' + extdir,
- '-DPYTHON_EXECUTABLE=' + sys.executable]
-
- cfg = 'Debug' if self.debug else 'Release'
- build_args = ['--config', cfg]
-
- if platform.system() == "Windows":
- cmake_args += ['-DCMAKE_LIBRARY_OUTPUT_DIRECTORY_{}={}'.format(cfg.upper(), extdir)]
- if sys.maxsize > 2**32:
- cmake_args += ['-A', 'x64']
- build_args += ['--', '/m']
- else:
- cmake_args += ['-DCMAKE_BUILD_TYPE=' + cfg]
- build_args += ['--', '-j2']
-
- cmake_args.extend(vf_cmake_args)
- cmake_args.extend(vf_gpu_cmake_args if GPU else vf_cpu_cmake_args)
-
- env = os.environ.copy()
- env['CXXFLAGS'] = '{} -DVERSION_INFO=\\"{}\\"'.format(env.get('CXXFLAGS', ''),
- self.distribution.get_version())
- if not os.path.exists(self.build_temp):
- os.makedirs(self.build_temp)
- cmake_cache_file = self.build_temp+"/CMakeCache.txt"
- if os.path.exists(cmake_cache_file):
- os.remove(cmake_cache_file)
- subprocess.check_call(['cmake', ext.sourcedir] + cmake_args, cwd=self.build_temp, env=env)
- subprocess.check_call(['cmake', '--build', '.'] + build_args, cwd=self.build_temp)
-
-
-setup(
- name='pyfluids',
- version='0.0.1',
- ext_modules=[CMakeExtension('pyfluids')],
- cmdclass={"install": InstallCommand, "develop": DevelopCommand, "build_ext": CMakeBuild},
- zip_safe=False,
+skbuild.setup(
+ name=package_name,
+ packages=[package_name, "pymuparser", "pyfluids-stubs"],
+ package_dir={"": src_dir},
+ cmake_args=cmake_args,
+ cmake_install_target=target,
+ package_data={ "pyfluids": ["py.typed"],
+ "pyfluids-stubs": stub_files},
+ include_package_data=True,
)
diff --git a/src/basics/basics/utilities/UbTuple.h b/src/basics/basics/utilities/UbTuple.h
index fe9c787cead38621beafab3d082122277bdcff73..228ab48898e5e61777d2fcc0061eb6f0434d5cad 100644
--- a/src/basics/basics/utilities/UbTuple.h
+++ b/src/basics/basics/utilities/UbTuple.h
@@ -597,6 +597,8 @@ inline UbTuple<T1, T2, T3, T4, T5, T6, T7, T8> makeUbTuple(T1 const &a1, T2 cons
// some typedefs
using UbTupleFloat2 = UbTuple<float, float>;
using UbTupleFloat3 = UbTuple<float, float, float>;
+using UbTupleFloat4 = UbTuple<float, float, float, float>;
+using UbTupleFloat6 = UbTuple<float, float, float,float, float, float>;
using UbTupleInt2 = UbTuple<int, int>;
using UbTupleInt3 = UbTuple<int, int, int>;
using UbTupleInt4 = UbTuple<int, int, int, int>;
diff --git a/src/basics/basics/writer/WbWriter.h b/src/basics/basics/writer/WbWriter.h
index 26d43464c03311a2cbc14cd4fc9fe717d4b01531..55dceb7cb4a64dc90f0677796cab52135b726f56 100644
--- a/src/basics/basics/writer/WbWriter.h
+++ b/src/basics/basics/writer/WbWriter.h
@@ -88,7 +88,12 @@ public:
{
throw UbException(UB_EXARGS, "not implemented for " + (std::string) typeid(*this).name());
}
-
+ virtual std::string writeLinesWithLineData(const std::string & /*filename*/, std::vector<UbTupleFloat3> & /*nodes*/,
+ std::vector<UbTupleInt2> & /*lines*/, std::vector<std::string> & /*datanames*/,
+ std::vector<std::vector<float>> & /*celldata*/)
+ {
+ throw UbException(UB_EXARGS, "not implemented for " + (std::string) typeid(*this).name());
+ }
//////////////////////////////////////////////////////////////////////////
// triangles
// cell numbering:
diff --git a/src/basics/basics/writer/WbWriterVtkXmlBinary.cpp b/src/basics/basics/writer/WbWriterVtkXmlBinary.cpp
index 6731fa56026ca284ad671cb6ce59000a609bbb8c..55c3541983ea4248512508146792832a34a1c563 100644
--- a/src/basics/basics/writer/WbWriterVtkXmlBinary.cpp
+++ b/src/basics/basics/writer/WbWriterVtkXmlBinary.cpp
@@ -34,6 +34,8 @@
#include <basics/writer/WbWriterVtkXmlASCII.h>
#include <basics/writer/WbWriterVtkXmlBinary.h>
#include <cstring>
+#include <fstream>
+#include <string>
using namespace std;
@@ -154,12 +156,13 @@ string WbWriterVtkXmlBinary::writeParallelFile(const string &filename, vector<st
return vtkfilename;
}
+
/*===============================================================================*/
-string WbWriterVtkXmlBinary::writeLines(const string &filename, vector<UbTupleFloat3> &nodes,
- vector<UbTupleInt2> &lines)
+
+// helper functions
+
+ofstream createFileStream(std::string vtkfilename)
{
- string vtkfilename = filename + getFileExtension();
- UBLOG(logDEBUG1, "WbWriterVtkXmlBinary::writeLines to " << vtkfilename << " - start");
ofstream out(vtkfilename.c_str(), ios::out | ios::binary);
if (!out) {
@@ -172,89 +175,199 @@ string WbWriterVtkXmlBinary::writeLines(const string &filename, vector<UbTupleFl
if (!out)
throw UbException(UB_EXARGS, "couldn't open file " + vtkfilename);
}
+ return out;
+}
- int nofNodes = (int)nodes.size();
- int nofCells = (int)lines.size();
-
- int bytesPerByteVal = 4; //==sizeof(int)
- int bytesPoints = 3 /*x1/x2/x3 */ * nofNodes * sizeof(float);
- int bytesCellConnectivty = 2 /*nodes per line */ * nofCells * sizeof(int);
- int bytesCellOffsets = 1 /*offset per line */ * nofCells * sizeof(int);
- int bytesCellTypes = 1 /*type of line */ * nofCells * sizeof(unsigned char);
-
- int offset = 0;
- // VTK FILE
+void writeVtkHeader(ofstream &out, int numberOfNodes, int numberOfCells)
+{
out << "<?xml version=\"1.0\"?>\n";
out << "<VTKFile type=\"UnstructuredGrid\" version=\"0.1\" byte_order=\"LittleEndian\" >"
<< "\n";
out << " <UnstructuredGrid>"
<< "\n";
- out << " <Piece NumberOfPoints=\"" << nofNodes << "\" NumberOfCells=\"" << nofCells << "\">\n";
+ out << " <Piece NumberOfPoints=\"" << numberOfNodes << "\" NumberOfCells=\"" << numberOfCells << "\">\n";
+}
- // POINTS SECTION
+int writePointHeader(ofstream &out, int offset, int bytesPerByteVal, int bytesPoints)
+{
out << " <Points>\n";
out << " <DataArray type=\"Float32\" NumberOfComponents=\"3\" format=\"appended\" offset=\"" << offset
<< "\" />\n";
out << " </Points>\n";
offset += (bytesPerByteVal + bytesPoints);
+ return offset;
+}
- // CELLS SECTION
+int writeCellHeader(ofstream &out, int offset, int bytesPerByteVal, int bytesCellConnectivity, int bytesCellOffsets,
+ int bytesCellTypes)
+{
out << " <Cells>\n";
out << " <DataArray type=\"Int32\" Name=\"connectivity\" format=\"appended\" offset=\"" << offset
<< "\" />\n";
- offset += (bytesPerByteVal + bytesCellConnectivty);
+ offset += (bytesPerByteVal + bytesCellConnectivity);
out << " <DataArray type=\"Int32\" Name=\"offsets\" format=\"appended\" offset=\"" << offset
<< "\" />\n";
offset += (bytesPerByteVal + bytesCellOffsets);
out << " <DataArray type=\"UInt8\" Name=\"types\" format=\"appended\" offset=\"" << offset << "\" />\n ";
offset += (bytesPerByteVal + bytesCellTypes);
out << " </Cells>\n";
+ return offset;
+}
+int writeDataHeader(ofstream &out, vector<string> &datanames, int offset, int bytesPerByteVal, int bytesScalarData)
+{
+ out << " <CellData>\n";
+ for (size_t s = 0; s < datanames.size(); ++s) {
+ out << " <DataArray type=\"Float32\" Name=\"" << datanames[s] << "\" format=\"appended\" offset=\""
+ << offset << "\" /> \n";
+ offset += (bytesPerByteVal + bytesScalarData);
+ }
+ out << " </CellData>\n";
+ return offset;
+}
+
+void writeAppendDataHeader(ofstream &out)
+{
out << " </Piece>\n";
out << " </UnstructuredGrid>\n";
-
- // AppendedData SECTION
out << " <AppendedData encoding=\"raw\">\n";
out << "_";
+}
- // POINTS SECTION
+void writePoints(ofstream &out, int bytesPerByteVal, int bytesPoints, vector<UbTupleFloat3> &nodes)
+{
out.write((char *)&bytesPoints, bytesPerByteVal);
- for (int n = 0; n < nofNodes; n++) {
+ for (int n = 0; n < (int)nodes.size(); n++) {
out.write((char *)&val<1>(nodes[n]), sizeof(float));
out.write((char *)&val<2>(nodes[n]), sizeof(float));
out.write((char *)&val<3>(nodes[n]), sizeof(float));
}
+}
- // CELLS SECTION
- // cellConnectivity
- out.write((char *)&bytesCellConnectivty, bytesPerByteVal);
- for (int c = 0; c < nofCells; c++) {
- out.write((char *)&val<1>(lines[c]), sizeof(int));
- out.write((char *)&val<2>(lines[c]), sizeof(int));
+void writeCellConnectivity(ofstream &out, int bytesPerByteVal, int bytesCellConnectivity, vector<UbTupleInt2> &cells)
+{
+ out.write((char *)&bytesCellConnectivity, bytesPerByteVal);
+ for (int c = 0; c < (int)cells.size(); c++) {
+ out.write((char *)&val<1>(cells[c]), sizeof(int));
+ out.write((char *)&val<2>(cells[c]), sizeof(int));
}
+}
- // cellOffsets
+void writeCellOffsets(ofstream &out, int bytesPerByteVal, int bytesCellOffsets, int numberOfCells)
+{
out.write((char *)&bytesCellOffsets, bytesPerByteVal);
int itmp;
- for (int c = 1; c <= nofCells; c++) {
+ for (int c = 1; c <= numberOfCells; c++) {
itmp = 2 * c;
out.write((char *)&itmp, sizeof(int));
}
+}
- // cellTypes
+void writeCellTypes(ofstream &out, int bytesPerByteVal, int bytesCellTypes, int numberOfCells)
+{
out.write((char *)&bytesCellTypes, bytesPerByteVal);
unsigned char vtkCellType = 3;
- for (int c = 0; c < nofCells; c++) {
+ for (int c = 0; c < numberOfCells; c++) {
out.write((char *)&vtkCellType, sizeof(unsigned char));
}
+}
+
+void writeCellData(ofstream &out, int bytesPerByteVal, int bytesScalarData, vector<string> &datanames,
+ vector<vector<float>> &celldata)
+{
+ for (size_t s = 0; s < datanames.size(); ++s) {
+ out.write((char *)&bytesScalarData, bytesPerByteVal);
+ for (size_t d = 0; d < celldata[s].size(); ++d) {
+ // loake kopie machen, da in celldata "doubles" sind
+ float tmp = (float)celldata[s][d];
+ out.write((char *)&tmp, sizeof(float));
+ }
+ }
+}
+
+void writeEndOfFile(ofstream &out)
+{
out << "\n</AppendedData>\n";
out << "</VTKFile>";
out << endl;
out.close();
+}
+
+/*===============================================================================*/
+string WbWriterVtkXmlBinary::writeLines(const string &filename, vector<UbTupleFloat3> &nodes,
+ vector<UbTupleInt2> &lines)
+{
+ string vtkfilename = filename + getFileExtension();
+ UBLOG(logDEBUG1, "WbWriterVtkXmlBinary::writeLines to " << vtkfilename << " - start");
+
+ ofstream out = createFileStream(vtkfilename);
+
+ int nofNodes = (int)nodes.size();
+ int nofCells = (int)lines.size();
+
+ int bytesPerByteVal = 4; //==sizeof(int)
+ int bytesPoints = 3 /*x1/x2/x3 */ * nofNodes * sizeof(float);
+ int bytesCellConnectivity = 2 /*nodes per line */ * nofCells * sizeof(int);
+ int bytesCellOffsets = 1 /*offset per line */ * nofCells * sizeof(int);
+ int bytesCellTypes = 1 /*type of line */ * nofCells * sizeof(unsigned char);
+
+ int offset = 0;
+
+ writeVtkHeader(out, nofNodes, nofCells);
+ offset = writePointHeader(out, offset, bytesPerByteVal, bytesPoints);
+ writeCellHeader(out, offset, bytesPerByteVal, bytesCellConnectivity, bytesCellOffsets, bytesCellTypes);
+ writeAppendDataHeader(out);
+
+ writePoints(out, bytesPerByteVal, bytesPoints, nodes);
+ writeCellConnectivity(out, bytesPerByteVal, bytesCellConnectivity, lines);
+ writeCellOffsets(out, bytesPerByteVal, bytesCellOffsets, nofCells);
+ writeCellTypes(out, bytesPerByteVal, bytesCellTypes, nofCells);
+ writeEndOfFile(out);
UBLOG(logDEBUG1, "WbWriterVtkXmlBinary::writeLines to " << vtkfilename << " - end");
return vtkfilename;
}
+
+/*===============================================================================*/
+string WbWriterVtkXmlBinary::writeLinesWithLineData(const string &filename, vector<UbTupleFloat3> &nodes,
+ vector<UbTupleInt2> &lines, vector<string> &datanames,
+ vector<vector<float>> &celldata)
+{
+ string vtkfilename = filename + getFileExtension();
+ UBLOG(logDEBUG1, "WbWriterVtkXmlBinary::writeLinesWithLineData to " << vtkfilename << " - start");
+
+ ofstream out = createFileStream(vtkfilename);
+
+ int nofNodes = (int)nodes.size();
+ int nofCells = (int)lines.size();
+
+ int bytesPerByteVal = 4; //==sizeof(int)
+ int bytesPoints = 3 /*x1/x2/x3 */ * nofNodes * sizeof(float);
+ int bytesCellConnectivity = 2 /*nodes per line */ * nofCells * sizeof(int);
+ int bytesCellOffsets = 1 /*offset per line */ * nofCells * sizeof(int);
+ int bytesCellTypes = 1 /*type of line */ * nofCells * sizeof(unsigned char);
+ int bytesScalarData = 1 /*scalar */ * nofCells * sizeof(float);
+
+ int offset = 0;
+
+ writeVtkHeader(out, nofNodes, nofCells);
+ offset = writePointHeader(out, offset, bytesPerByteVal, bytesPoints);
+ offset = writeCellHeader(out, offset, bytesPerByteVal, bytesCellConnectivity, bytesCellOffsets, bytesCellTypes);
+ writeDataHeader(out, datanames, offset, bytesPerByteVal, bytesScalarData);
+ writeAppendDataHeader(out);
+
+ writePoints(out, bytesPerByteVal, bytesPoints, nodes);
+ writeCellConnectivity(out, bytesPerByteVal, bytesCellConnectivity, lines);
+ writeCellOffsets(out, bytesPerByteVal, bytesCellOffsets, nofCells);
+ writeCellTypes(out, bytesPerByteVal, bytesCellTypes, nofCells);
+ writeCellData(out, bytesPerByteVal, bytesScalarData, datanames, celldata);
+ writeEndOfFile(out);
+
+ UBLOG(logDEBUG1, "WbWriterVtkXmlBinary::writeLinesWithLineData to " << vtkfilename << " - end");
+
+ return vtkfilename;
+}
+
/*===============================================================================*/
// std::string WbWriterVtkXmlBinary::writeLinesWithNodeData(const string& filename,vector<UbTupleFloat3 >& nodes,
// vector<UbTupleInt2 >& lines, std::vector< std::string >& datanames, std::vector< std::vector< double > >& nodedata)
@@ -276,7 +389,7 @@ string WbWriterVtkXmlBinary::writeLines(const string &filename, vector<UbTupleFl
//
// int bytesPerByteVal = 4; //==sizeof(int)
// int bytesPoints = 3 /*x1/x2/x3 */ * nofNodes * sizeof(float);
-// int bytesCellConnectivty = 2 /*nodes per line */ * nofCells * sizeof(int );
+// int bytesCellConnectivity = 2 /*nodes per line */ * nofCells * sizeof(int );
// int bytesCellOffsets = 1 /*offset per line */ * nofCells * sizeof(int );
// int bytesCellTypes = 1 /*type of line */ * nofCells * sizeof(unsigned char);
// int bytesScalarData = 1 /*scalar */ * nofNodes * sizeof(float);
@@ -296,7 +409,7 @@ string WbWriterVtkXmlBinary::writeLines(const string &filename, vector<UbTupleFl
// //CELLS SECTION
// out<<" <Cells>\n";
// out<<" <DataArray type=\"Int32\" Name=\"connectivity\" format=\"appended\" offset=\""<< offset <<"\"
-// />\n"; offset += (bytesPerByteVal + bytesCellConnectivty); out<<" <DataArray type=\"Int32\"
+// />\n"; offset += (bytesPerByteVal + bytesCellConnectivity); out<<" <DataArray type=\"Int32\"
// Name=\"offsets\" format=\"appended\" offset=\""<< offset <<"\" />\n"; offset += (bytesPerByteVal +
// bytesCellOffsets); out<<" <DataArray type=\"UInt8\" Name=\"types\" format=\"appended\" offset=\""<<
// offset <<"\" />\n "; offset += (bytesPerByteVal + bytesCellTypes); out<<" </Cells>\n";
@@ -328,7 +441,7 @@ string WbWriterVtkXmlBinary::writeLines(const string &filename, vector<UbTupleFl
//
// //CELLS SECTION
// //cellConnectivity
-// out.write( (char*)&bytesCellConnectivty, bytesPerByteVal );
+// out.write( (char*)&bytesCellConnectivity, bytesPerByteVal );
// for(int c=0; c<nofCells; c++)
// {
// out.write( (char*)&val<1>(lines[c]), sizeof(int) );
@@ -397,7 +510,7 @@ string WbWriterVtkXmlBinary::writeTriangles(const string &filename, vector<UbTup
int bytesPerByteVal = 4; //==sizeof(int)
int bytesPoints = 3 /*x1/x2/x3 - coord */ * nofNodes * sizeof(float);
- int bytesCellConnectivty = 3 /*nodes per triangle */ * nofCells * sizeof(int);
+ int bytesCellConnectivity = 3 /*nodes per triangle */ * nofCells * sizeof(int);
int bytesCellOffsets = 1 /*offset per triangle */ * nofCells * sizeof(int);
int bytesCellTypes = 1 /*type of triangle */ * nofCells * sizeof(unsigned char);
@@ -421,7 +534,7 @@ string WbWriterVtkXmlBinary::writeTriangles(const string &filename, vector<UbTup
out << " <Cells>\n";
out << " <DataArray type=\"Int32\" Name=\"connectivity\" format=\"appended\" offset=\"" << offset
<< "\" />\n";
- offset += (bytesPerByteVal + bytesCellConnectivty);
+ offset += (bytesPerByteVal + bytesCellConnectivity);
out << " <DataArray type=\"Int32\" Name=\"offsets\" format=\"appended\" offset=\"" << offset
<< "\" />\n";
offset += (bytesPerByteVal + bytesCellOffsets);
@@ -446,7 +559,7 @@ string WbWriterVtkXmlBinary::writeTriangles(const string &filename, vector<UbTup
// CELLS SECTION
// cellConnectivity
- out.write((char *)&bytesCellConnectivty, bytesPerByteVal);
+ out.write((char *)&bytesCellConnectivity, bytesPerByteVal);
for (int c = 0; c < nofCells; c++) {
out.write((char *)&val<1>(triangles[c]), sizeof(int));
out.write((char *)&val<2>(triangles[c]), sizeof(int));
@@ -502,7 +615,7 @@ string WbWriterVtkXmlBinary::writeTrianglesWithNodeData(const string &filename,
int bytesPerByteVal = 4; //==sizeof(int)
int bytesPoints = 3 /*x1/x2/x3 */ * nofNodes * sizeof(float);
- int bytesCellConnectivty = 3 /*nodes per tri */ * nofCells * sizeof(int);
+ int bytesCellConnectivity = 3 /*nodes per tri */ * nofCells * sizeof(int);
int bytesCellOffsets = 1 /*offset per tri */ * nofCells * sizeof(int);
int bytesCellTypes = 1 /*type of tri */ * nofCells * sizeof(unsigned char);
int bytesScalarData = 1 /*scalar */ * nofNodes * sizeof(float);
@@ -527,7 +640,7 @@ string WbWriterVtkXmlBinary::writeTrianglesWithNodeData(const string &filename,
out << " <Cells>\n";
out << " <DataArray type=\"Int32\" Name=\"connectivity\" format=\"appended\" offset=\"" << offset
<< "\" />\n";
- offset += (bytesPerByteVal + bytesCellConnectivty);
+ offset += (bytesPerByteVal + bytesCellConnectivity);
out << " <DataArray type=\"Int32\" Name=\"offsets\" format=\"appended\" offset=\"" << offset
<< "\" />\n";
offset += (bytesPerByteVal + bytesCellOffsets);
@@ -561,7 +674,7 @@ string WbWriterVtkXmlBinary::writeTrianglesWithNodeData(const string &filename,
// CELLS SECTION
// cellConnectivity
- out.write((char *)&bytesCellConnectivty, bytesPerByteVal);
+ out.write((char *)&bytesCellConnectivity, bytesPerByteVal);
for (int c = 0; c < nofCells; c++) {
out.write((char *)&val<1>(cells[c]), sizeof(int));
out.write((char *)&val<2>(cells[c]), sizeof(int));
@@ -625,7 +738,7 @@ string WbWriterVtkXmlBinary::writeQuads(const string &filename, vector<UbTupleFl
int bytesPerByteVal = 4; //==sizeof(int)
int bytesPoints = 3 /*x1/x2/x3 */ * nofNodes * sizeof(float);
- int bytesCellConnectivty = 4 /*nodes per quad */ * nofCells * sizeof(int);
+ int bytesCellConnectivity = 4 /*nodes per quad */ * nofCells * sizeof(int);
int bytesCellOffsets = 1 /*offset per quad */ * nofCells * sizeof(int);
int bytesCellTypes = 1 /*type of quad */ * nofCells * sizeof(unsigned char);
@@ -649,7 +762,7 @@ string WbWriterVtkXmlBinary::writeQuads(const string &filename, vector<UbTupleFl
out << " <Cells>\n";
out << " <DataArray type=\"Int32\" Name=\"connectivity\" format=\"appended\" offset=\"" << offset
<< "\" />\n";
- offset += (bytesPerByteVal + bytesCellConnectivty);
+ offset += (bytesPerByteVal + bytesCellConnectivity);
out << " <DataArray type=\"Int32\" Name=\"offsets\" format=\"appended\" offset=\"" << offset
<< "\" />\n";
offset += (bytesPerByteVal + bytesCellOffsets);
@@ -674,7 +787,7 @@ string WbWriterVtkXmlBinary::writeQuads(const string &filename, vector<UbTupleFl
// CELLS SECTION
// cellConnectivity
- out.write((char *)&bytesCellConnectivty, bytesPerByteVal);
+ out.write((char *)&bytesCellConnectivity, bytesPerByteVal);
for (int c = 0; c < nofCells; c++) {
out.write((char *)&val<1>(cells[c]), sizeof(int));
out.write((char *)&val<2>(cells[c]), sizeof(int));
@@ -730,7 +843,7 @@ string WbWriterVtkXmlBinary::writeQuadsWithNodeData(const string &filename, vect
int bytesPerByteVal = 4; //==sizeof(int)
int bytesPoints = 3 /*x1/x2/x3 */ * nofNodes * sizeof(float);
- int bytesCellConnectivty = 4 /*nodes per quad */ * nofCells * sizeof(int);
+ int bytesCellConnectivity = 4 /*nodes per quad */ * nofCells * sizeof(int);
int bytesCellOffsets = 1 /*offset per quad */ * nofCells * sizeof(int);
int bytesCellTypes = 1 /*type of quad */ * nofCells * sizeof(unsigned char);
int bytesScalarData = 1 /*scalar */ * nofNodes * sizeof(float);
@@ -755,7 +868,7 @@ string WbWriterVtkXmlBinary::writeQuadsWithNodeData(const string &filename, vect
out << " <Cells>\n";
out << " <DataArray type=\"Int32\" Name=\"connectivity\" format=\"appended\" offset=\"" << offset
<< "\" />\n";
- offset += (bytesPerByteVal + bytesCellConnectivty);
+ offset += (bytesPerByteVal + bytesCellConnectivity);
out << " <DataArray type=\"Int32\" Name=\"offsets\" format=\"appended\" offset=\"" << offset
<< "\" />\n";
offset += (bytesPerByteVal + bytesCellOffsets);
@@ -789,7 +902,7 @@ string WbWriterVtkXmlBinary::writeQuadsWithNodeData(const string &filename, vect
// CELLS SECTION
// cellConnectivity
- out.write((char *)&bytesCellConnectivty, bytesPerByteVal);
+ out.write((char *)&bytesCellConnectivity, bytesPerByteVal);
for (int c = 0; c < nofCells; c++) {
out.write((char *)&val<1>(cells[c]), sizeof(int));
out.write((char *)&val<2>(cells[c]), sizeof(int));
@@ -855,7 +968,7 @@ string WbWriterVtkXmlBinary::writeQuadsWithCellData(const string &filename, vect
int bytesPerByteVal = 4; //==sizeof(int)
int bytesPoints = 3 /*x1/x2/x3 */ * nofNodes * sizeof(float);
- int bytesCellConnectivty = 4 /*nodes per quad */ * nofCells * sizeof(int);
+ int bytesCellConnectivity = 4 /*nodes per quad */ * nofCells * sizeof(int);
int bytesCellOffsets = 1 /*offset per quad */ * nofCells * sizeof(int);
int bytesCellTypes = 1 /*type of quad */ * nofCells * sizeof(unsigned char);
int bytesScalarData = 1 /*scalar */ * nofCells * sizeof(float);
@@ -880,7 +993,7 @@ string WbWriterVtkXmlBinary::writeQuadsWithCellData(const string &filename, vect
out << " <Cells>\n";
out << " <DataArray type=\"Int32\" Name=\"connectivity\" format=\"appended\" offset=\"" << offset
<< "\" />\n";
- offset += (bytesPerByteVal + bytesCellConnectivty);
+ offset += (bytesPerByteVal + bytesCellConnectivity);
out << " <DataArray type=\"Int32\" Name=\"offsets\" format=\"appended\" offset=\"" << offset
<< "\" />\n";
offset += (bytesPerByteVal + bytesCellOffsets);
@@ -914,7 +1027,7 @@ string WbWriterVtkXmlBinary::writeQuadsWithCellData(const string &filename, vect
// CELLS SECTION
// cellConnectivity
- out.write((char *)&bytesCellConnectivty, bytesPerByteVal);
+ out.write((char *)&bytesCellConnectivity, bytesPerByteVal);
for (int c = 0; c < nofCells; c++) {
out.write((char *)&val<1>(cells[c]), sizeof(int));
out.write((char *)&val<2>(cells[c]), sizeof(int));
@@ -984,7 +1097,7 @@ string WbWriterVtkXmlBinary::writeQuadsWithNodeAndCellData(const string &filenam
int bytesPerByteVal = 4; //==sizeof(int)
int bytesPoints = 3 /*x1/x2/x3 */ * nofNodes * sizeof(float);
- int bytesCellConnectivty = 4 /*nodes per quad */ * nofCells * sizeof(int);
+ int bytesCellConnectivity = 4 /*nodes per quad */ * nofCells * sizeof(int);
int bytesCellOffsets = 1 /*offset per quad */ * nofCells * sizeof(int);
int bytesCellTypes = 1 /*type of quad */ * nofCells * sizeof(unsigned char);
int bytesScalarDataPoint = 1 /*scalar */ * nofNodes * sizeof(float);
@@ -1010,7 +1123,7 @@ string WbWriterVtkXmlBinary::writeQuadsWithNodeAndCellData(const string &filenam
out << " <Cells>\n";
out << " <DataArray type=\"Int32\" Name=\"connectivity\" format=\"appended\" offset=\"" << offset
<< "\" />\n";
- offset += (bytesPerByteVal + bytesCellConnectivty);
+ offset += (bytesPerByteVal + bytesCellConnectivity);
out << " <DataArray type=\"Int32\" Name=\"offsets\" format=\"appended\" offset=\"" << offset
<< "\" />\n";
offset += (bytesPerByteVal + bytesCellOffsets);
@@ -1052,7 +1165,7 @@ string WbWriterVtkXmlBinary::writeQuadsWithNodeAndCellData(const string &filenam
// CELLS SECTION
// cellConnectivity
- out.write((char *)&bytesCellConnectivty, bytesPerByteVal);
+ out.write((char *)&bytesCellConnectivity, bytesPerByteVal);
for (int c = 0; c < nofCells; c++) {
out.write((char *)&val<1>(cells[c]), sizeof(int));
out.write((char *)&val<2>(cells[c]), sizeof(int));
@@ -1128,7 +1241,7 @@ string WbWriterVtkXmlBinary::writeOctsWithCellData(const string &filename, vecto
int bytesPerByteVal = 4; //==sizeof(int)
int bytesPoints = 3 /*x1/x2/x3 */ * nofNodes * sizeof(float);
- int bytesCellConnectivty = 8 /*nodes per oct */ * nofCells * sizeof(int);
+ int bytesCellConnectivity = 8 /*nodes per oct */ * nofCells * sizeof(int);
int bytesCellOffsets = 1 /*offset per oct*/ * nofCells * sizeof(int);
int bytesCellTypes = 1 /*type of oct */ * nofCells * sizeof(unsigned char);
int bytesScalarData = 1 /*scalar */ * nofCells * sizeof(float);
@@ -1153,7 +1266,7 @@ string WbWriterVtkXmlBinary::writeOctsWithCellData(const string &filename, vecto
out << " <Cells>\n";
out << " <DataArray type=\"Int32\" Name=\"connectivity\" format=\"appended\" offset=\"" << offset
<< "\" />\n";
- offset += (bytesPerByteVal + bytesCellConnectivty);
+ offset += (bytesPerByteVal + bytesCellConnectivity);
out << " <DataArray type=\"Int32\" Name=\"offsets\" format=\"appended\" offset=\"" << offset
<< "\" />\n";
offset += (bytesPerByteVal + bytesCellOffsets);
@@ -1187,7 +1300,7 @@ string WbWriterVtkXmlBinary::writeOctsWithCellData(const string &filename, vecto
// CELLS SECTION
// cellConnectivity
- out.write((char *)&bytesCellConnectivty, bytesPerByteVal);
+ out.write((char *)&bytesCellConnectivity, bytesPerByteVal);
for (int c = 0; c < nofCells; c++) {
out.write((char *)&val<1>(cells[c]), sizeof(int));
out.write((char *)&val<2>(cells[c]), sizeof(int));
@@ -1257,7 +1370,7 @@ string WbWriterVtkXmlBinary::writeOctsWithNodeData(const string &filename, vecto
int bytesPerByteVal = 4; //==sizeof(int)
int bytesPoints = 3 /*x1/x2/x3 */ * nofNodes * sizeof(float);
- int bytesCellConnectivty = 8 /*nodes per oct */ * nofCells * sizeof(int);
+ int bytesCellConnectivity = 8 /*nodes per oct */ * nofCells * sizeof(int);
int bytesCellOffsets = 1 /*offset per oct*/ * nofCells * sizeof(int);
int bytesCellTypes = 1 /*type of oct */ * nofCells * sizeof(unsigned char);
int bytesScalarData = 1 /*scalar */ * nofNodes * sizeof(double);
@@ -1282,7 +1395,7 @@ string WbWriterVtkXmlBinary::writeOctsWithNodeData(const string &filename, vecto
out << " <Cells>\n";
out << " <DataArray type=\"Int32\" Name=\"connectivity\" format=\"appended\" offset=\"" << offset
<< "\" />\n";
- offset += (bytesPerByteVal + bytesCellConnectivty);
+ offset += (bytesPerByteVal + bytesCellConnectivity);
out << " <DataArray type=\"Int32\" Name=\"offsets\" format=\"appended\" offset=\"" << offset
<< "\" />\n";
offset += (bytesPerByteVal + bytesCellOffsets);
@@ -1316,7 +1429,7 @@ string WbWriterVtkXmlBinary::writeOctsWithNodeData(const string &filename, vecto
// CELLS SECTION
// cellConnectivity
- out.write((char *)&bytesCellConnectivty, bytesPerByteVal);
+ out.write((char *)&bytesCellConnectivity, bytesPerByteVal);
for (int c = 0; c < nofCells; c++) {
out.write((char *)&val<1>(cells[c]), sizeof(int));
out.write((char *)&val<2>(cells[c]), sizeof(int));
@@ -1386,7 +1499,7 @@ string WbWriterVtkXmlBinary::writeOcts(const string &filename, vector<UbTupleFlo
int bytesPerByteVal = 4; //==sizeof(int)
int bytesPoints = 3 /*x1/x2/x3 */ * nofNodes * sizeof(float);
- int bytesCellConnectivty = 8 /*nodes per oct */ * nofCells * sizeof(int);
+ int bytesCellConnectivity = 8 /*nodes per oct */ * nofCells * sizeof(int);
int bytesCellOffsets = 1 /*offset per oct*/ * nofCells * sizeof(int);
int bytesCellTypes = 1 /*type of oct */ * nofCells * sizeof(unsigned char);
// int bytesScalarData = 1 /*scalar */ * nofNodes * sizeof(float);
@@ -1411,7 +1524,7 @@ string WbWriterVtkXmlBinary::writeOcts(const string &filename, vector<UbTupleFlo
out << " <Cells>\n";
out << " <DataArray type=\"Int32\" Name=\"connectivity\" format=\"appended\" offset=\"" << offset
<< "\" />\n";
- offset += (bytesPerByteVal + bytesCellConnectivty);
+ offset += (bytesPerByteVal + bytesCellConnectivity);
out << " <DataArray type=\"Int32\" Name=\"offsets\" format=\"appended\" offset=\"" << offset
<< "\" />\n";
offset += (bytesPerByteVal + bytesCellOffsets);
@@ -1436,7 +1549,7 @@ string WbWriterVtkXmlBinary::writeOcts(const string &filename, vector<UbTupleFlo
// CELLS SECTION
// cellConnectivity
- out.write((char *)&bytesCellConnectivty, bytesPerByteVal);
+ out.write((char *)&bytesCellConnectivity, bytesPerByteVal);
for (int c = 0; c < nofCells; c++) {
out.write((char *)&val<1>(cells[c]), sizeof(int));
out.write((char *)&val<2>(cells[c]), sizeof(int));
@@ -1491,7 +1604,7 @@ std::string WbWriterVtkXmlBinary::writeNodes(const std::string &filename, std::v
int bytesPerByteVal = 4; //==sizeof(int)
int bytesPoints = 3 /*x1/x2/x3 */ * nofNodes * sizeof(float);
- int bytesCellConnectivty = 1 /*nodes per cell */ * nofNodes * sizeof(int);
+ int bytesCellConnectivity = 1 /*nodes per cell */ * nofNodes * sizeof(int);
int bytesCellOffsets = 1 /*offset per cell */ * nofNodes * sizeof(int);
int bytesCellTypes = 1 /*type of line */ * nofNodes * sizeof(unsigned char);
@@ -1515,7 +1628,7 @@ std::string WbWriterVtkXmlBinary::writeNodes(const std::string &filename, std::v
out << " <Cells>\n";
out << " <DataArray type=\"Int32\" Name=\"connectivity\" format=\"appended\" offset=\"" << offset
<< "\" />\n";
- offset += (bytesPerByteVal + bytesCellConnectivty);
+ offset += (bytesPerByteVal + bytesCellConnectivity);
out << " <DataArray type=\"Int32\" Name=\"offsets\" format=\"appended\" offset=\"" << offset
<< "\" />\n";
offset += (bytesPerByteVal + bytesCellOffsets);
@@ -1540,7 +1653,7 @@ std::string WbWriterVtkXmlBinary::writeNodes(const std::string &filename, std::v
// CELLS SECTION
// cellConnectivity
- out.write((char *)&bytesCellConnectivty, bytesPerByteVal);
+ out.write((char *)&bytesCellConnectivity, bytesPerByteVal);
for (int c = 0; c < nofNodes; c++)
out.write((char *)&c, sizeof(int));
@@ -1586,7 +1699,7 @@ std::string WbWriterVtkXmlBinary::writeNodesWithNodeData(const std::string &file
int bytesPerByteVal = 4; //==sizeof(int)
int bytesPoints = 3 /*x1/x2/x3 */ * nofNodes * sizeof(float);
- int bytesCellConnectivty = 1 /*nodes per cell */ * nofNodes * sizeof(int);
+ int bytesCellConnectivity = 1 /*nodes per cell */ * nofNodes * sizeof(int);
int bytesCellOffsets = 1 /*offset per cell*/ * nofNodes * sizeof(int);
int bytesCellTypes = 1 /*type of oct */ * nofNodes * sizeof(unsigned char);
int bytesScalarData = 1 /*scalar */ * nofNodes * sizeof(double);
@@ -1611,7 +1724,7 @@ std::string WbWriterVtkXmlBinary::writeNodesWithNodeData(const std::string &file
out << " <Cells>\n";
out << " <DataArray type=\"Int32\" Name=\"connectivity\" format=\"appended\" offset=\"" << offset
<< "\" />\n";
- offset += (bytesPerByteVal + bytesCellConnectivty);
+ offset += (bytesPerByteVal + bytesCellConnectivity);
out << " <DataArray type=\"Int32\" Name=\"offsets\" format=\"appended\" offset=\"" << offset
<< "\" />\n";
offset += (bytesPerByteVal + bytesCellOffsets);
@@ -1645,7 +1758,7 @@ std::string WbWriterVtkXmlBinary::writeNodesWithNodeData(const std::string &file
// CELLS SECTION
// cellConnectivity
- out.write((char *)&bytesCellConnectivty, bytesPerByteVal);
+ out.write((char *)&bytesCellConnectivity, bytesPerByteVal);
for (int c = 0; c < nofNodes; c++)
out.write((char *)&c, sizeof(int));
diff --git a/src/basics/basics/writer/WbWriterVtkXmlBinary.h b/src/basics/basics/writer/WbWriterVtkXmlBinary.h
index 421148d90497e3628ed274439c0b2fd7636b7fd2..0f2c31eda81ad0c1975c9715ac1b7fb37a06339b 100644
--- a/src/basics/basics/writer/WbWriterVtkXmlBinary.h
+++ b/src/basics/basics/writer/WbWriterVtkXmlBinary.h
@@ -93,6 +93,9 @@ public:
// nodedata);
// FIXME: hides function in base class
+ std::string writeLinesWithLineData(const std::string &filename, std::vector<UbTupleFloat3> &nodes, std::vector<UbTupleInt2> &lines,
+ std::vector<std::string> &datanames, std::vector<std::vector<float>> &celldata) override;
+
//////////////////////////////////////////////////////////////////////////
// triangles
// 2
diff --git a/src/basics/basics/writer/WbWriterVtkXmlImageBinary.cpp b/src/basics/basics/writer/WbWriterVtkXmlImageBinary.cpp
new file mode 100644
index 0000000000000000000000000000000000000000..798b55919df9e24dbc71ecfded5fb8a913cff8cf
--- /dev/null
+++ b/src/basics/basics/writer/WbWriterVtkXmlImageBinary.cpp
@@ -0,0 +1,360 @@
+//=======================================================================================
+// ____ ____ __ ______ __________ __ __ __ __
+// \ \ | | | | | _ \ |___ ___| | | | | / \ | |
+// \ \ | | | | | |_) | | | | | | | / \ | |
+// \ \ | | | | | _ / | | | | | | / /\ \ | |
+// \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
+// \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
+// \ \ | | ________________________________________________________________
+// \ \ | | | ______________________________________________________________|
+// \ \| | | | __ __ __ __ ______ _______
+// \ | | |_____ | | | | | | | | | _ \ / _____)
+// \ | | _____| | | | | | | | | | | \ \ \_______
+// \ | | | | |_____ | \_/ | | | | |_/ / _____ |
+// \ _____| |__| |________| \_______/ |__| |______/ (_______/
+//
+// This file is part of VirtualFluids. VirtualFluids is free software: you can
+// redistribute it and/or modify it under the terms of the GNU General Public
+// License as published by the Free Software Foundation, either version 3 of
+// the License, or (at your option) any later version.
+//
+// VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
+// ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+// FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
+// for more details.
+//
+// You should have received a copy of the GNU General Public License along
+// with VirtualFluids (see COPYING.txt). If not, see <http://www.gnu.org/licenses/>.
+//
+//! \file WbWriterVtkXmlImageBinary.cpp
+//! \ingroup writer
+//! \author Soeren Freudiger, Sebastian Geller, Henry Korb, Henrik Asmuth
+//=======================================================================================
+#include <basics/utilities/UbLogger.h>
+#include <basics/writer/WbWriterVtkXmlImageBinary.h>
+#include <cstring>
+
+using namespace std;
+
+/*===============================================================================*/
+const std::string WbWriterVtkXmlImageBinary::pvdEndTag = " </Collection>\n</VTKFile>";
+/*===============================================================================*/
+string WbWriterVtkXmlImageBinary::writeCollection(const string &filename, const vector<string> &filenames,
+ const double &timeStep, const bool &sepGroups)
+{
+ string vtkfilename = filename + ".pvd";
+ ofstream out(vtkfilename.c_str());
+ if (!out) {
+ out.clear(); // flags ruecksetzen (ansonsten liefert utern if(!out) weiterhin true!!!
+ string path = UbSystem::getPathFromString(vtkfilename);
+ if (path.size() > 0) {
+ UbSystem::makeDirectory(path);
+ out.open(vtkfilename.c_str());
+ }
+ if (!out)
+ throw UbException(UB_EXARGS, "couldn't open file " + vtkfilename);
+ }
+
+ string endian;
+ if (UbSystem::isLittleEndian())
+ endian = "LittleEndian";
+ else
+ endian = "BigEndian";
+ out << "<VTKFile type=\"Collection\" version=\"0.1\" byte_order=\"" << endian << "\" >" << endl;
+ out << " <Collection>" << endl;
+
+ int group = 0, part = 0;
+ for (size_t i = 0; i < filenames.size(); i++) {
+ out << " <DataSet timestep=\"" << timeStep << "\" group=\"" << group << "\" part=\"" << part
+ << "\" file=\"" << filenames[i] << "\"/>" << endl;
+ if (sepGroups)
+ group++;
+ else
+ part++;
+ }
+ out << pvdEndTag;
+ out.close();
+
+ return vtkfilename;
+}
+/*===============================================================================*/
+string WbWriterVtkXmlImageBinary::addFilesToCollection(const string &filename, const vector<string> &filenames,
+ const double &timeStep, const bool &sepGroups)
+{
+ string vtkfilename = filename;
+ fstream test(vtkfilename.c_str(), ios::in);
+ if (!test) {
+ test.clear();
+ vtkfilename += ".pvd";
+ test.open(vtkfilename.c_str(), ios::in);
+ if (!test)
+ return this->writeCollection(filename, filenames, timeStep, sepGroups);
+ }
+
+ fstream out(vtkfilename.c_str(), ios::in | ios::out);
+ out.seekp(-(int)pvdEndTag.size() - 1, ios_base::end);
+
+ int group = 0;
+ for (size_t i = 0; i < filenames.size(); i++) {
+ out << " <DataSet timestep=\"" << timeStep << "\" group=\"" << group << "\" part=\"" << i << "\" file=\""
+ << filenames[i] << "\"/>" << endl;
+ if (sepGroups)
+ group++;
+ }
+ out << pvdEndTag;
+
+ return vtkfilename;
+}
+/*===============================================================================*/
+string WbWriterVtkXmlImageBinary::writeParallelFile(const string &filename, const UbTupleInt6 &wholeExtent,
+ const UbTupleFloat3 &origin, const UbTupleFloat3 &spacing,
+ vector<string> &pieceSources, vector<UbTupleInt6> &pieceExtents,
+ vector<string> &pointDataNames, vector<string> &cellDataNames)
+{
+ string vtkfilename = filename + ".pvti";
+ UBLOG(logDEBUG1, "WbWriterVtkXmlImageBinary::writeParallelFile to " << vtkfilename << " - start");
+
+ ofstream out(vtkfilename.c_str());
+ if (!out) {
+ out.clear(); // flags ruecksetzen (ansonsten liefert utern if(!out) weiterhin true!!!
+ string path = UbSystem::getPathFromString(vtkfilename);
+ if (path.size() > 0) {
+ UbSystem::makeDirectory(path);
+ out.open(vtkfilename.c_str());
+ }
+ if (!out)
+ throw UbException(UB_EXARGS, "couldn't open file " + vtkfilename);
+ }
+
+ // VTK FILE
+ out << "<VTKFile type=\"PImageData\" version=\"0.1\" byte_order=\"LittleEndian\">"
+ << "\n";
+ out << " <PImageData "
+ << "WholeExtent=\"" << val<1>(wholeExtent) << " "
+ << val<2>(wholeExtent) << " "
+ << val<3>(wholeExtent) << " "
+ << val<4>(wholeExtent) << " "
+ << val<5>(wholeExtent) << " "
+ << val<6>(wholeExtent) << "\" "
+ << "GhostLevel=\"0\" "
+ << "Origin=\"" << val<1>(origin) << " "
+ << val<2>(origin) << " "
+ << val<3>(origin) << "\" "
+ << "Spacing=\"" << val<1>(spacing) << " "
+ << val<2>(spacing) << " "
+ << val<3>(spacing) << "\" "
+ << "> \n";
+ out << " <PPointData>\n";
+ for (size_t s = 0; s < pointDataNames.size(); s++)
+ out << " <PDataArray type=\"Float32\" Name=\"" << pointDataNames[s] << "\"/>\n";
+ out << " </PPointData>\n";
+ if (cellDataNames.size() > 0) {
+ out << " <PCellData>\n";
+ for (size_t s = 0; s < cellDataNames.size(); s++)
+ out << " <PDataArray type=\"Float32\" Name=\"" << cellDataNames[s] << "\"/>\n";
+ out << " </PCellData>\n";
+ }
+ for (size_t s = 0; s < pieceSources.size(); s++)
+ out << " <Piece Extent=\"" << val<1>(pieceExtents[s]) << " "
+ << val<2>(pieceExtents[s]) << " "
+ << val<3>(pieceExtents[s]) << " "
+ << val<4>(pieceExtents[s]) << " "
+ << val<5>(pieceExtents[s]) << " "
+ << val<6>(pieceExtents[s]) << "\" Source=\"" << pieceSources[s] << "\"/>\n";
+ out << " </PImageData>\n";
+ out << "</VTKFile>";
+ out << endl;
+ out.close();
+ UBLOG(logDEBUG1, "WbWriterVtkXmlImageBinary::writeParallelFile to " << vtkfilename << " - end");
+
+ return vtkfilename;
+}
+/*===============================================================================*/
+string WbWriterVtkXmlImageBinary::writeOctsWithCellData(const string &filename, vector<UbTupleFloat3> &nodes,
+ vector<UbTupleInt8> & /*cells*/, vector<string> &datanames,
+ vector<vector<double>> &celldata)
+{
+ string vtkfilename = filename + getFileExtension();
+ UBLOG(logDEBUG1, "WbWriterVtkXmlImageBinary::writeOctsWithCellData to " << vtkfilename << " - start");
+
+ vector<string> nodeDataNames;
+ vector<vector<double>> nodedata;
+
+ UbTupleFloat3 origin, spacing;
+ UbTupleInt6 extent;
+
+ getMetaDataOfImage(nodes, origin, spacing, extent);
+
+ this->writeData(vtkfilename, nodeDataNames, datanames, nodedata, celldata, extent, origin, spacing, extent);
+ UBLOG(logDEBUG1, "WbWriterVtkXmlImageBinary::writeOctsWithCellData to " << vtkfilename << " - end");
+
+ return vtkfilename;
+}
+/*===============================================================================*/
+string WbWriterVtkXmlImageBinary::writeOctsWithNodeData(const string &filename, vector<UbTupleFloat3> &nodes,
+ vector<UbTupleUInt8> & /*cells*/, vector<string> &datanames,
+ vector<vector<double>> &nodedata)
+{
+ string vtkfilename = filename + getFileExtension();
+ UBLOG(logDEBUG1, "WbWriterVtkXmlImageBinary::writeOctsWithNodeData to " << vtkfilename << " - start");
+
+ vector<string> cellDataNames;
+ vector<vector<double>> cellData;
+
+ UbTupleFloat3 origin, spacing;
+ UbTupleInt6 extent;
+
+ getMetaDataOfImage(nodes, origin, spacing, extent);
+
+ this->writeData(vtkfilename, datanames, cellDataNames, nodedata, cellData, extent, origin, spacing, extent);
+
+ UBLOG(logDEBUG1, "WbWriterVtkXmlImageBinary::writeOctsWithNodeData to " << vtkfilename << " - end");
+
+ return vtkfilename;
+}
+/*===============================================================================*/
+string WbWriterVtkXmlImageBinary::writeNodesWithNodeData(const string &filename, vector<UbTupleFloat3> &nodes,
+ vector<string> &datanames, vector<vector<double>> &nodedata)
+{
+ string vtkfilename = filename + getFileExtension();
+ UBLOG(logDEBUG1, "WbWriterVtkXmlImageBinary::writeNodesWithNodeData to " << vtkfilename << " - start");
+
+ vector<string> cellDataNames;
+ vector<vector<double>> cellData;
+
+ UbTupleFloat3 origin, spacing;
+ UbTupleInt6 extent;
+
+ getMetaDataOfImage(nodes, origin, spacing, extent);
+ this->writeData(vtkfilename, datanames, cellDataNames, nodedata, cellData, extent, origin, spacing, extent);
+
+ return vtkfilename;
+}
+
+void WbWriterVtkXmlImageBinary::getMetaDataOfImage(vector<UbTupleFloat3> &nodes, UbTupleFloat3 &origin,
+ UbTupleFloat3 &spacing, UbTupleInt6 &extent)
+{
+ int nofNodes = (int)nodes.size();
+ val<1>(origin) = val<1>(nodes[0]);
+ val<2>(origin) = val<2>(nodes[0]);
+ val<3>(origin) = val<3>(nodes[0]);
+
+ float l_x = val<1>(nodes[nofNodes-1])-val<1>(origin);
+ float l_y = val<2>(nodes[nofNodes-1])-val<2>(origin);
+
+ val<1>(spacing) = val<1>(nodes[1])-val<1>(nodes[0]);
+ int nx = (l_x) / val<1>(spacing);
+ val<2>(spacing) = val<2>(nodes[nx])-val<2>(nodes[0]);
+ int ny = (l_y) / val<2>(spacing);
+ val<3>(spacing) = val<3>(nodes[nx*ny])-val<3>(nodes[0]);
+
+ val<1>(extent) = val<1>(origin) / val<1>(spacing); val<2>(extent) = val<1>(nodes[nofNodes - 1]) / val<1>(spacing);
+ val<3>(extent) = val<2>(origin) / val<2>(spacing); val<4>(extent) = val<2>(nodes[nofNodes - 1]) / val<2>(spacing);
+ val<5>(extent) = val<3>(origin) / val<3>(spacing); val<6>(extent) = val<3>(nodes[nofNodes - 1]) / val<3>(spacing);
+
+}
+
+void WbWriterVtkXmlImageBinary::writeData(const string &vtkfilename, vector<string> &pointDataNames,
+ vector<string> &cellDataNames, vector<vector<double>> &nodedata,
+ vector<vector<double>> &celldata, UbTupleInt6 &wholeExtent,
+ UbTupleFloat3 &origin, UbTupleFloat3 &spacing, UbTupleInt6 &extent,
+ unsigned int precision)
+{
+ ofstream out(vtkfilename.c_str(), ios::out | ios::binary);
+ out.precision(precision);
+
+ if (!out) {
+ out.clear(); // flags ruecksetzen (ansonsten liefert utern if(!out) weiterhin true!!!
+ string path = UbSystem::getPathFromString(vtkfilename);
+ if (path.size() > 0) {
+ UbSystem::makeDirectory(path);
+ out.open(vtkfilename.c_str(), ios::out | ios::binary);
+ }
+ if (!out)
+ throw UbException(UB_EXARGS, "couldn't open file " + vtkfilename);
+ }
+
+ size_t nPoints = pointDataNames.size() > 0 ? nodedata[0].size() : celldata[0].size();
+
+ int bytesPerByteVal = 4; //==sizeof(int)
+
+ int bytesScalarData = 1 /*scalar */ * (int)nPoints * sizeof(double);
+
+ int offset = 0;
+
+ // VTK FILE
+ out << "<?xml version=\"1.0\"?>\n";
+ out << "<VTKFile type=\"ImageData\" version=\"0.1\" byte_order=\"LittleEndian\" >"
+ << "\n";
+ out << " <ImageData "
+ << "WholeExtent=\"" << val<1>(wholeExtent) << " "
+ << val<2>(wholeExtent) << " "
+ << val<3>(wholeExtent) << " "
+ << val<4>(wholeExtent) << " "
+ << val<5>(wholeExtent) << " "
+ << val<6>(wholeExtent) << "\" "
+ << "Origin=\"" << val<1>(origin) << " "
+ << val<2>(origin) << " "
+ << val<3>(origin) << "\" "
+ << "Spacing=\"" << val<1>(spacing) << " "
+ << val<2>(spacing) << " "
+ << val<3>(spacing) << "\""
+ << "> \n";
+ out << " <Piece Extent=\"" << val<1>(extent) << " "
+ << val<2>(extent) << " "
+ << val<3>(extent) << " "
+ << val<4>(extent) << " "
+ << val<5>(extent) << " "
+ << val<6>(extent) << "\">\n";
+
+ // DATA SECTION
+ if (pointDataNames.size() > 0) {
+ out << " <PointData>\n";
+ for (size_t s = 0; s < pointDataNames.size(); ++s) {
+ out << " <DataArray type=\"Float64\" Name=\"" << pointDataNames[s]
+ << "\" format=\"appended\" offset=\"" << offset << "\" /> \n";
+ offset += (bytesPerByteVal + bytesScalarData);
+ }
+ out << " </PointData>\n";
+ }
+
+ if (cellDataNames.size() > 0) {
+ out << " <CellData>\n";
+ for (size_t s = 0; s < cellDataNames.size(); ++s) {
+ out << " <DataArray type=\"Float64\" Name=\"" << cellDataNames[s]
+ << "\" format=\"appended\" offset=\"" << offset << "\" /> \n";
+ offset += (bytesPerByteVal + bytesScalarData);
+ }
+ out << " </CellData>\n";
+ }
+
+ out << " </Piece>\n";
+ out << " </ImageData>\n";
+
+ // AppendedData SECTION
+ out << " <AppendedData encoding=\"raw\">\n";
+ out << "_";
+
+ // DATA SECTION
+ // pointData
+ for (size_t s = 0; s < pointDataNames.size(); ++s) {
+ out.write((char *)&bytesScalarData, bytesPerByteVal);
+ for (size_t d = 0; d < nodedata[s].size(); ++d) {
+ double tmp = nodedata[s][d];
+ out.write((char *)&tmp, sizeof(double));
+ }
+ }
+
+ // cellData
+ for (size_t s = 0; s < cellDataNames.size(); ++s) {
+ out.write((char *)&bytesScalarData, bytesPerByteVal);
+ for (size_t d = 0; d < celldata[s].size(); ++d) {
+ double tmp = celldata[s][d];
+ out.write((char *)&tmp, sizeof(double));
+ }
+ }
+ out << "\n </AppendedData>\n";
+ out << "</VTKFile>";
+ out << endl;
+ out.close();
+}
diff --git a/src/basics/basics/writer/WbWriterVtkXmlImageBinary.h b/src/basics/basics/writer/WbWriterVtkXmlImageBinary.h
new file mode 100644
index 0000000000000000000000000000000000000000..c41ff442732e5f65db0f1dd1ec63e5c3ffca1486
--- /dev/null
+++ b/src/basics/basics/writer/WbWriterVtkXmlImageBinary.h
@@ -0,0 +1,110 @@
+//=======================================================================================
+// ____ ____ __ ______ __________ __ __ __ __
+// \ \ | | | | | _ \ |___ ___| | | | | / \ | |
+// \ \ | | | | | |_) | | | | | | | / \ | |
+// \ \ | | | | | _ / | | | | | | / /\ \ | |
+// \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
+// \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
+// \ \ | | ________________________________________________________________
+// \ \ | | | ______________________________________________________________|
+// \ \| | | | __ __ __ __ ______ _______
+// \ | | |_____ | | | | | | | | | _ \ / _____)
+// \ | | _____| | | | | | | | | | | \ \ \_______
+// \ | | | | |_____ | \_/ | | | | |_/ / _____ |
+// \ _____| |__| |________| \_______/ |__| |______/ (_______/
+//
+// This file is part of VirtualFluids. VirtualFluids is free software: you can
+// redistribute it and/or modify it under the terms of the GNU General Public
+// License as published by the Free Software Foundation, either version 3 of
+// the License, or (at your option) any later version.
+//
+// VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
+// ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+// FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
+// for more details.
+//
+// You should have received a copy of the GNU General Public License along
+// with VirtualFluids (see COPYING.txt). If not, see <http://www.gnu.org/licenses/>.
+//
+//! \file WbWriterVtkXmlBinary.h
+//! \ingroup writer
+//! \author Soeren Freudiger, Sebastian Geller
+//=======================================================================================
+#ifndef WBWRITERVTKXMLIMAGEBINARY_H
+#define WBWRITERVTKXMLIMAGEBINARY_H
+
+#include <string>
+
+#include <basics/writer/WbWriter.h>
+
+#include "basics_export.h"
+
+class BASICS_EXPORT WbWriterVtkXmlImageBinary : public WbWriter
+{
+public:
+ static WbWriterVtkXmlImageBinary *getInstance()
+ {
+ static WbWriterVtkXmlImageBinary instance;
+ return &instance;
+ }
+
+ WbWriterVtkXmlImageBinary(const WbWriterVtkXmlImageBinary &) = delete;
+ const WbWriterVtkXmlImageBinary &operator=(const WbWriterVtkXmlImageBinary &) = delete;
+
+private:
+ WbWriterVtkXmlImageBinary() : WbWriter()
+ {
+ if (sizeof(unsigned char) != 1)
+ throw UbException(UB_EXARGS, "machine error char type mismatch");
+ if (sizeof(int) != 4)
+ throw UbException(UB_EXARGS, "machine error int type mismatch");
+ if (sizeof(float) != 4)
+ throw UbException(UB_EXARGS, "machine error float type mismatch");
+ }
+
+ static const std::string pvdEndTag;
+
+public:
+ std::string getFileExtension() override { return ".bin.vti"; }
+
+ // write a metafile
+ std::string writeCollection(const std::string &filename, const std::vector<std::string> &filenames,
+ const double ×tep, const bool &sepGroups);
+ std::string addFilesToCollection(const std::string &filename, const std::vector<std::string> &filenames,
+ const double ×tep, const bool &sepGroups);
+ std::string writeParallelFile(const std::string &filename, const UbTupleInt6 &wholeExtent, const UbTupleFloat3 &origin, const UbTupleFloat3 &spacing,
+ std::vector<std::string> &pieceSources, std::vector<UbTupleInt6> &pieceExtents,
+ std::vector<std::string> &pointDataNames, std::vector<std::string> &cellDataNames);
+
+ //////////////////////////////////////////////////////////////////////////
+ // nodes
+ std::string writeNodesWithNodeData(const std::string &filename, std::vector<UbTupleFloat3> &nodes,
+ std::vector<std::string> &datanames,
+ std::vector<std::vector<double>> &nodedata) override;
+
+ //////////////////////////////////////////////////////////////////////////
+ // octs
+ // 7 ---- 6
+ // /| /|
+ // 4 +--- 5 |
+ // | | | |
+ // | 3 ---+ 2
+ // |/ |/
+ // 0 ---- 1
+ std::string writeOctsWithCellData(const std::string &filename, std::vector<UbTupleFloat3> &nodes,
+ std::vector<UbTupleInt8> &cells, std::vector<std::string> &datanames,
+ std::vector<std::vector<double>> &celldata) override;
+ std::string writeOctsWithNodeData(const std::string &filename, std::vector<UbTupleFloat3> &nodes,
+ std::vector<UbTupleUInt8> &cells, std::vector<std::string> &datanames,
+ std::vector<std::vector<double>> &nodedata) override;
+ void writeData(const std::string &vtkfilename,
+ std::vector<std::string> &pointDataNames, std::vector<std::string> &cellDataNames,
+ std::vector<std::vector<double>> &nodedata, std::vector<std::vector<double>> &celldata,
+ UbTupleInt6 &wholeExtent,
+ UbTupleFloat3 &origin, UbTupleFloat3 &spacing, UbTupleInt6 &extent, unsigned int precision=6);
+
+private:
+ void getMetaDataOfImage(std::vector<UbTupleFloat3> &nodes, UbTupleFloat3& origin, UbTupleFloat3& spacing, UbTupleInt6& extent);
+};
+
+#endif // WBWRITERVTKXMLIMAGEBINARY_H
diff --git a/src/basics/config/ConfigurationFile.h b/src/basics/config/ConfigurationFile.h
index ef7e7c9f06f94cabb3ba9cbefe95c8ee75736958..4a53f7add85b9c6461fda0bab20fa6656eebc5d3 100644
--- a/src/basics/config/ConfigurationFile.h
+++ b/src/basics/config/ConfigurationFile.h
@@ -64,6 +64,10 @@ public:
template<class T>
T getValue(const std::string& key) const;
+ //! get value with key and default value
+ template<class T>
+ T getValue(const std::string& key, T defaultValue) const;
+
private:
//! the container
std::map<std::string, std::string> data;
@@ -138,6 +142,19 @@ T ConfigurationFile::getValue(const std::string& key) const
return x;
}
+template<class T>
+T ConfigurationFile::getValue(const std::string& key, T defaultValue) const
+{
+ if (contains(key))
+ {
+ return getValue<T>(key);
+ }
+ else
+ {
+ return defaultValue;
+ }
+}
+
}
#endif
diff --git a/src/basics/tests/testUtilities.h b/src/basics/tests/testUtilities.h
index c70d9cc5c11633ded6b696d92692e3d4edf8d2ca..57606edc130b0471b957202420cb12859a9cde84 100644
--- a/src/basics/tests/testUtilities.h
+++ b/src/basics/tests/testUtilities.h
@@ -1,6 +1,8 @@
#ifndef TESTUTILITIES_H
#define TESTUTILITIES_H
+#include <gmock/gmock.h>
+
inline auto RealEq = [](auto value) {
#ifdef VF_DOUBLE_ACCURACY
return testing::DoubleEq(value);
diff --git a/src/gpu/GksGpu/BoundaryConditions/BoundaryCondition.h b/src/gpu/GksGpu/BoundaryConditions/BoundaryCondition.h
index fe4078af95904fa5e1580b54f3aa2edbb006bd3d..9c3bac9c3e2795fa99f339461c6a7f2d16448696 100644
--- a/src/gpu/GksGpu/BoundaryConditions/BoundaryCondition.h
+++ b/src/gpu/GksGpu/BoundaryConditions/BoundaryCondition.h
@@ -47,13 +47,13 @@ struct GKSGPU_EXPORT BoundaryCondition : virtual public BoundaryConditionStruct,
virtual bool isWall() = 0;
virtual bool isFluxBC();
-
+
virtual bool isInsulated();
virtual bool secondCellsNeeded();
virtual void runBoundaryConditionKernel( const SPtr<DataBase> dataBase,
- const Parameters parameters,
+ const Parameters parameters,
const uint level ) = 0;
BoundaryConditionStruct toStruct()
diff --git a/src/gpu/GksGpu/CMakeLists.txt b/src/gpu/GksGpu/CMakeLists.txt
index 5dbc533cc5f45c006c29a12242350f0433518bbf..6db6cbac1ff60c76986c3c22cc8017300d4f71ea 100644
--- a/src/gpu/GksGpu/CMakeLists.txt
+++ b/src/gpu/GksGpu/CMakeLists.txt
@@ -1,6 +1,6 @@
project(GksGpu LANGUAGES CUDA CXX)
-vf_add_library(PRIVATE_LINK basics lbmCuda GksMeshAdapter OpenMP::OpenMP_CXX MPI::MPI_CXX)
+vf_add_library(PRIVATE_LINK basics lbm GksMeshAdapter OpenMP::OpenMP_CXX MPI::MPI_CXX)
target_include_directories(GksGpu PRIVATE "${VF_THIRD_DIR}/cuda_samples/")
diff --git a/src/gpu/GksMeshAdapter/CMakeLists.txt b/src/gpu/GksMeshAdapter/CMakeLists.txt
index b9a2d12df4d0bee9396a706c6636b5f4056b2d3a..8ac5e69513eca94710797db1f971b2461336b769 100644
--- a/src/gpu/GksMeshAdapter/CMakeLists.txt
+++ b/src/gpu/GksMeshAdapter/CMakeLists.txt
@@ -1,3 +1,3 @@
project(GksMeshAdapter LANGUAGES CUDA CXX)
-vf_add_library(PRIVATE_LINK basics GridGenerator lbmCuda)
+vf_add_library(PRIVATE_LINK basics GridGenerator lbm)
diff --git a/src/gpu/GridGenerator/TransientBCSetter/TransientBCSetter.cpp b/src/gpu/GridGenerator/TransientBCSetter/TransientBCSetter.cpp
new file mode 100644
index 0000000000000000000000000000000000000000..5f3c4ad492b16c09b26acd00a624a54ad65dffda
--- /dev/null
+++ b/src/gpu/GridGenerator/TransientBCSetter/TransientBCSetter.cpp
@@ -0,0 +1,444 @@
+#include "TransientBCSetter.h"
+#include "GridGenerator/grid/Grid.h"
+#include "GridGenerator/grid/BoundaryConditions/BoundaryCondition.h"
+#include <logger/Logger.h>
+
+
+#include <math.h>
+#include <sstream>
+#include <fstream>
+#include <iostream>
+#include <algorithm>
+
+SPtr<FileCollection> createFileCollection(std::string prefix, FileType type)
+{
+ switch(type)
+ {
+ case FileType::VTK:
+ return std::make_shared<VTKFileCollection>(prefix);
+ break;
+ default:
+ return nullptr;
+ }
+}
+
+SPtr<TransientBCInputFileReader> createReaderForCollection(SPtr<FileCollection> fileCollection, uint readLevel)
+{
+ switch(fileCollection->getFileType())
+ {
+ case FileType::VTK:
+ return std::make_shared<VTKReader>(std::static_pointer_cast<VTKFileCollection>(fileCollection), readLevel);
+ break;
+ default:
+ return nullptr;
+ }
+}
+
+template<typename T>
+std::vector<T> readStringToVector(std::string s)
+{
+ std::vector<T> out;
+ std::stringstream input(s);
+ float num;
+ while(input >> num)
+ {
+ out.push_back(num);
+ }
+ return out;
+}
+
+std::string readElement(std::string line)
+{
+ size_t elemStart = line.find("<")+1;
+ // size_t elemEnd = line.find("/>", elemStart);
+ size_t nameLen = line.find(" ", elemStart)-elemStart;
+ return line.substr(elemStart, nameLen);
+}
+
+std::string readAttribute(std::string line, std::string attributeName)
+{
+ size_t attributeStart = line.find(attributeName)+attributeName.size() + 2; // add 2 for '="'
+ size_t attributeLen = line.find("\"", attributeStart)-attributeStart;
+ return line.substr(attributeStart, attributeLen);
+}
+
+void VTKFile::readHeader()
+{
+ //TODO make this more flexible
+ std::ifstream file(this->fileName);
+
+ std::string line;
+
+ getline(file, line); // VTKFile
+ if(line[1]=='?') getline(file, line); // ignore first line if xml version
+
+ getline(file, line); // ImageData
+ std::vector<int> wholeExtent = readStringToVector<int>(readAttribute(line, "WholeExtent"));
+ std::vector<float> origin = readStringToVector<float>(readAttribute(line, "Origin"));
+ std::vector<float> spacing = readStringToVector<float>(readAttribute(line, "Spacing"));
+
+ getline(file, line); // Piece
+ std::vector<int> pieceExtent = readStringToVector<int>(readAttribute(line, "Extent"));
+ getline(file, line); // PointData
+
+ getline(file, line);
+ while(strcmp(readElement(line).c_str(), "DataArray")==0)
+ {
+ Quantity quant = Quantity();
+ quant.name = readAttribute(line, "Name");
+ quant.offset = std::stoi(readAttribute(line, "offset"));
+ this->quantities.push_back( quant );
+ getline(file, line);
+ }
+ getline(file, line); // </Piece
+ getline(file, line); // </ImageData
+ getline(file, line); // AppendedData
+
+ int offset = int(file.tellg())+sizeof(char)+4; // skip underscore and bytesPerVal
+
+ for(auto& quantity: this->quantities)
+ {
+ quantity.offset += offset;
+ }
+
+ file.close();
+
+ this->deltaX = spacing[0];
+ this->deltaY = spacing[1];
+ this->deltaZ = spacing[2];
+
+ this->nx = pieceExtent[1]-pieceExtent[0]+1;
+ this->ny = pieceExtent[3]-pieceExtent[2]+1;
+ this->nz = pieceExtent[5]-pieceExtent[4]+1;
+
+ this->minX = origin[0]+this->deltaX*pieceExtent[0]; this->maxX = (this->nx-1)*this->deltaX+this->minX;
+ this->minY = origin[1]+this->deltaY*pieceExtent[2]; this->maxY = (this->ny-1)*this->deltaY+this->minY;
+ this->minZ = origin[2]+this->deltaZ*pieceExtent[4]; this->maxZ = (this->nz-1)*this->deltaZ+this->minZ;
+ // printFileInfo();
+
+}
+
+bool VTKFile::markNANs(std::vector<uint> readIndices)
+{
+ std::ifstream buf(fileName.c_str(), std::ios::in | std::ios::binary);
+
+ std::vector<double> tmp;
+ tmp.reserve(readIndices.size());
+ buf.seekg(this->quantities[0].offset);
+ buf.read((char*) tmp.data(), sizeof(double)*readIndices.size());
+ auto firstNAN = std::find_if(tmp.begin(), tmp.end(), [](auto it){ return isnan(it); });
+
+ return firstNAN != tmp.end();
+}
+
+void VTKFile::loadFile()
+{
+ std::ifstream buf(this->fileName.c_str(), std::ios::in | std::ios::binary);
+ for(auto& quantity: this->quantities)
+ {
+ quantity.values.resize(getNumberOfPoints());
+ buf.seekg(quantity.offset);
+ buf.read(reinterpret_cast<char*>(quantity.values.data()), this->getNumberOfPoints()*sizeof(double));
+ }
+
+ buf.close();
+
+ this->loaded = true;
+}
+
+void VTKFile::unloadFile()
+{
+ for(auto& quantity : this->quantities)
+ {
+ std::vector<double> replacement;
+ quantity.values.swap(replacement);
+ }
+ this->loaded = false;
+}
+
+void VTKFile::getData(real *data, uint numberOfNodes, const std::vector<uint> &readIndices,
+ const std::vector<uint> &writeIndices, uint offsetRead, uint offsetWrite)
+{
+ if(!this->loaded) loadFile();
+
+ size_t nPoints = writeIndices.size();
+
+ for(size_t j=0; j<this->quantities.size(); j++)
+ {
+ real* quant = &data[j*numberOfNodes];
+ for(size_t i=0; i<nPoints; i++)
+ {
+ quant[offsetWrite+writeIndices[i]] = this->quantities[j].values[readIndices[i]+offsetRead];
+ }
+ }
+}
+
+void VTKFile::printFileInfo()
+{
+ printf("file %s with \n nx %i ny %i nz %i \n origin %f %f %f \n spacing %f %f %f \n",
+ fileName.c_str(), nx, ny, nz, minX, minY, minZ, deltaX, deltaY, deltaZ);
+ for(auto quantity: this->quantities)
+ {
+ printf("\t quantity %s offset %i \n", quantity.name.c_str(), quantity.offset);
+ }
+
+}
+
+
+void VTKFileCollection::findFiles()
+{
+ bool foundLastLevel = false;
+
+ while(!foundLastLevel)
+ {
+ bool foundLastID = false;
+ std::vector<std::vector<VTKFile>> filesOnThisLevel;
+ while(!foundLastID)
+ {
+ bool foundLastPart = false;
+ std::vector<VTKFile> filesWithThisId;
+ while (!foundLastPart)
+ {
+ std::string fname = makeFileName((int)files.size(), (int)filesOnThisLevel.size(), (int)filesWithThisId.size());
+ std::ifstream f(fname);
+ if(f.good())
+ filesWithThisId.emplace_back(fname);
+ else
+ foundLastPart = true;
+ }
+ if(!filesWithThisId.empty())
+ {
+ VF_LOG_INFO("VTKFileCollection found {} files with ID {} level {}", filesWithThisId.size(), filesOnThisLevel.size(), files.size() );
+ filesOnThisLevel.push_back(filesWithThisId);
+ }
+ else foundLastID = true;
+ }
+
+
+ if(!filesOnThisLevel.empty())
+ files.push_back(filesOnThisLevel);
+ else
+ foundLastLevel = true;
+
+ }
+
+ if(files.empty())
+ VF_LOG_CRITICAL("VTKFileCollection found no files!");
+}
+
+void TransientBCInputFileReader::getNeighbors(uint* neighbor0PP, uint* neighbor0PM, uint* neighbor0MP, uint* neighbor0MM)
+{
+ std::copy(planeNeighbor0PP.begin(), planeNeighbor0PP.end(), &neighbor0PP[writingOffset]);
+ std::copy(planeNeighbor0PM.begin(), planeNeighbor0PM.end(), &neighbor0PM[writingOffset]);
+ std::copy(planeNeighbor0MP.begin(), planeNeighbor0MP.end(), &neighbor0MP[writingOffset]);
+ std::copy(planeNeighbor0MM.begin(), planeNeighbor0MM.end(), &neighbor0MM[writingOffset]);
+}
+
+void TransientBCInputFileReader::getWeights(real* _weights0PP, real* _weights0PM, real* _weights0MP, real* _weights0MM)
+{
+ std::copy(weights0PP.begin(), weights0PP.end(), &_weights0PP[writingOffset]);
+ std::copy(weights0PM.begin(), weights0PM.end(), &_weights0PM[writingOffset]);
+ std::copy(weights0MP.begin(), weights0MP.end(), &_weights0MP[writingOffset]);
+ std::copy(weights0MM.begin(), weights0MM.end(), &_weights0MM[writingOffset]);
+}
+
+
+void VTKReader::initializeIndexVectors()
+{
+ this->readIndices.resize(this->fileCollection->files.size());
+ this->writeIndices.resize(this->fileCollection->files.size());
+ this->nFile.resize(this->fileCollection->files.size());
+ for(size_t lev=0; lev<this->fileCollection->files.size(); lev++)
+ {
+ this->readIndices[lev].resize(this->fileCollection->files[lev].size());
+ this->writeIndices[lev].resize(this->fileCollection->files[lev].size());
+ this->nFile[lev].resize(this->fileCollection->files[lev].size());
+ }
+}
+
+void VTKReader::fillArrays(std::vector<real>& coordsY, std::vector<real>& coordsZ)
+{
+ this->nPoints = (uint)coordsY.size();
+ this->initializeIndexVectors();
+ real max_diff = 1e-4; // maximum distance between point on grid and precursor plane to count as exact match
+ real eps = 1e-7; // small number to avoid division by zero
+ bool perfect_match = true;
+
+ this->weights0PP.reserve(this->nPoints);
+ this->weights0PM.reserve(this->nPoints);
+ this->weights0MP.reserve(this->nPoints);
+ this->weights0MM.reserve(this->nPoints);
+
+ this->planeNeighbor0PP.reserve(this->nPoints);
+ this->planeNeighbor0PM.reserve(this->nPoints);
+ this->planeNeighbor0MP.reserve(this->nPoints);
+ this->planeNeighbor0MM.reserve(this->nPoints);
+
+ for(uint i=0; i<nPoints; i++)
+ {
+
+ real posY = coordsY[i];
+ real posZ = coordsZ[i];
+ bool found0PP = false, found0PM = false, found0MP = false, found0MM = false, foundAll = false;
+
+ uint level = this->readLevel;
+
+ for(int fileId=0; fileId<(int)this->fileCollection->files[level].size(); fileId++)
+ {
+ VTKFile &file = this->fileCollection->files[level][fileId][0];
+ if(!file.inBoundingBox(posY, posZ, 0.0f)) continue;
+
+ // y in simulation is x in precursor/file, z in simulation is y in precursor/file
+ // simulation -> file: N -> E, S -> W, T -> N, B -> S
+ int idx = file.findNeighborMMM(posY, posZ, 0.f); //!> index of nearest WSB neighbor on precursor file
+
+ if(idx!=-1)
+ {
+ // Filter for exact matches
+ if(abs(posY-file.getX(idx)) < max_diff && abs(posZ-file.getY(idx)) < max_diff)
+ {
+ this->weights0PP.emplace_back(1e6f);
+ this->weights0PM.emplace_back(0.f);
+ this->weights0MP.emplace_back(0.f);
+ this->weights0MM.emplace_back(0.f);
+ uint writeIdx = this->getWriteIndex(level, fileId, idx); //!> writeIdx: index on host/device array where precursor value will be written to after loading from file
+ this->planeNeighbor0PP.push_back(writeIdx); //!> neighbor lists mapping where BC kernel should read from on host/device array
+ this->planeNeighbor0PM.push_back(writeIdx);
+ this->planeNeighbor0MP.push_back(writeIdx);
+ this->planeNeighbor0MM.push_back(writeIdx);
+ found0PP = true;
+ found0PM = true;
+ found0MM = true;
+ found0MP = true;
+ }
+ else
+ {
+ perfect_match = false;
+ }
+
+ if(!found0MM)
+ {
+ found0MM = true;
+ real dy = file.getX(idx)-posY;
+ real dz = file.getY(idx)-posZ;
+ this->weights0MM.emplace_back(1.f/(dy*dy+dz*dz+eps));
+ this->planeNeighbor0MM.emplace_back(getWriteIndex(level, fileId, idx));
+ }
+
+ }
+
+ if(!found0PP) //NT in simulation is EN in precursor
+ {
+ int index = file.findNeighborPPM(posY, posZ, 0.f);
+ if(index!=-1)
+ {
+ found0PP = true;
+ real dy = file.getX(index)-posY;
+ real dz = file.getY(index)-posZ;
+ this->weights0PP.emplace_back(1.f/(dy*dy+dz*dz+eps));
+ this->planeNeighbor0PP.emplace_back(getWriteIndex(level, fileId, index));
+ }
+ }
+
+ if(!found0PM) //NB in simulation is ES in precursor
+ {
+ int index = file.findNeighborPMM(posY, posZ, 0.f);
+ if(index!=-1)
+ {
+ found0PM = true;
+ real dy = file.getX(index)-posY;
+ real dz = file.getY(index)-posZ;
+ this->weights0PM.emplace_back(1.f/(dy*dy+dz*dz+eps));
+ this->planeNeighbor0PP.emplace_back(getWriteIndex(level, fileId, index));
+ }
+ }
+
+ if(!found0MP) //ST in simulation is WN in precursor
+ {
+ int index = file.findNeighborMPM(posY, posZ, 0.f);
+ if(index!=-1)
+ {
+ found0MP = true;
+ real dy = file.getX(index)-posY;
+ real dz = file.getY(index)-posZ;
+ this->weights0MP.emplace_back(1.f/(dy*dy+dz*dz+eps));
+ this->planeNeighbor0MP.emplace_back(getWriteIndex(level, fileId, index));
+ }
+ }
+
+ foundAll = found0PP && found0PM && found0MP && found0MM;
+
+ if(foundAll) break;
+ }
+
+ if(!foundAll)
+ {
+ VF_LOG_CRITICAL("Found no matching precursor neighbors for grid point at y={}, z={} \n", posY, posZ);
+ throw std::runtime_error("VTKReader::fillArrays(): Did not find neighbors in the FileCollection for all points");
+ }
+ }
+
+ if(perfect_match)
+ printf("Precursor was a perfect match \n");
+
+
+ for(size_t level=0; level<this->fileCollection->files.size(); level++){
+ for(size_t id=0; id<this->fileCollection->files[level].size(); id++){
+ if(this->fileCollection->files[level][id][0].markNANs(this->readIndices[level][id]))
+ throw std::runtime_error("Found a NAN in the precursor where a velocity is needed");
+ }}
+}
+
+uint VTKReader::getWriteIndex(int level, int id, int linearIndex)
+{
+ auto it = std::find(this->writeIndices[level][id].begin(), this->writeIndices[level][id].end(), linearIndex);
+ uint idx = it-this->writeIndices[level][id].begin();
+ if(it==this->writeIndices[level][id].end())
+ {
+ this->writeIndices[level][id].push_back(this->nPointsRead); //!> index on host/device array where value from file will be written to
+ this->readIndices[level][id].push_back(linearIndex); //!> index in file that will be read from
+ this->nPointsRead++;
+ }
+ return idx;
+}
+
+
+void VTKReader::getNextData(real* data, uint numberOfNodes, real time)
+{
+ // for(size_t level=0; level<this->fileCollection->files.size(); level++)
+ // {
+ uint level = this->readLevel;
+ for(size_t id=0; id<this->fileCollection->files[level].size(); id++)
+ {
+ size_t numberOfFiles = this->nFile[level][id];
+
+
+ if(!this->fileCollection->files[level][id][numberOfFiles].inZBounds(time))
+ {
+ numberOfFiles++;
+
+ printf("switching to precursor file no. %zu\n", numberOfFiles);
+ if(numberOfFiles == this->fileCollection->files[level][id].size())
+ throw std::runtime_error("Not enough Precursor Files to read");
+
+ this->fileCollection->files[level][id][numberOfFiles-1].unloadFile();
+ if(numberOfFiles+1<this->fileCollection->files[level][id].size())
+ {
+ VTKFile* nextFile = &this->fileCollection->files[level][id][numberOfFiles+1];
+ if(! nextFile->isLoaded())
+ {
+ read.wait();
+ read = std::async(std::launch::async, [](VTKFile* file){ file->loadFile(); }, &this->fileCollection->files[level][id][numberOfFiles+1]);
+ }
+ }
+ }
+
+
+ VTKFile* file = &this->fileCollection->files[level][id][numberOfFiles];
+
+ int off = file->getClosestIdxZ(time)*file->getNumberOfPointsInXYPlane();
+ file->getData(data, numberOfNodes, this->readIndices[level][id], this->writeIndices[level][id], off, this->writingOffset);
+ this->nFile[level][id] = numberOfFiles;
+ }
+ // }
+}
diff --git a/src/gpu/GridGenerator/TransientBCSetter/TransientBCSetter.h b/src/gpu/GridGenerator/TransientBCSetter/TransientBCSetter.h
new file mode 100644
index 0000000000000000000000000000000000000000..1663a3ff37ba1bb062647847462d4e364baed93b
--- /dev/null
+++ b/src/gpu/GridGenerator/TransientBCSetter/TransientBCSetter.h
@@ -0,0 +1,201 @@
+#ifndef TRANSIENTBCSETTER_H_
+#define TRANSIENTBCSETTER_H_
+
+#include "Core/DataTypes.h"
+#include <Core/StringUtilities/StringUtil.h>
+#include "PointerDefinitions.h"
+
+#include <string>
+#include <vector>
+#include <math.h>
+#include <sstream>
+#include <future>
+class Grid;
+namespace gg
+{
+ class BoundaryCondition;
+}
+
+
+enum class FileType
+{
+ VTK
+};
+
+struct Quantity
+{
+ std::string name;
+ int offset;
+ std::vector<double> values;
+};
+
+class VTKFile
+{
+public:
+ explicit VTKFile(std::string _fileName):
+ fileName(_fileName)
+ {
+ readHeader();
+ this->loaded = false;
+ // printFileInfo();
+ };
+
+ void getData(real* data, uint numberOfNodes, const std::vector<uint>& readIndices, const std::vector<uint>& writeIndices, uint offsetRead, uint offsetWrite);
+ bool markNANs(std::vector<uint> readIndices);
+ bool inBoundingBox(real posX, real posY, real posZ){return inXBounds(posX) && inYBounds(posY) && inZBounds(posZ); };
+ bool inXBounds(real posX){ return posX<=maxX && posX>=minX; };
+ bool inYBounds(real posY){ return posY<=maxY && posY>=minY; };
+ bool inZBounds(real posZ){ return posZ<=maxZ && posZ>=minZ; };
+ int findNeighborMMM(real posX, real posY, real posZ){ int idx = getLinearIndex(getIdxM00(posX) , getIdx0M0(posY) , getIdx00M(posZ) ); return (idx>=0) && (idx<nx*ny*nz) ? idx : -1; };
+ int findNeighborMMP(real posX, real posY, real posZ){ int idx = getLinearIndex(getIdxM00(posX) , getIdx0M0(posY) , getIdx00M(posZ)+1); return (idx>=0) && (idx<nx*ny*nz) ? idx : -1; };
+ int findNeighborMPM(real posX, real posY, real posZ){ int idx = getLinearIndex(getIdxM00(posX) , getIdx0M0(posY)+1, getIdx00M(posZ) ); return (idx>=0) && (idx<nx*ny*nz) ? idx : -1; };
+ int findNeighborMPP(real posX, real posY, real posZ){ int idx = getLinearIndex(getIdxM00(posX) , getIdx0M0(posY)+1, getIdx00M(posZ)+1); return (idx>=0) && (idx<nx*ny*nz) ? idx : -1; };
+ int findNeighborPMM(real posX, real posY, real posZ){ int idx = getLinearIndex(getIdxM00(posX)+1, getIdx0M0(posY) , getIdx00M(posZ) ); return (idx>=0) && (idx<nx*ny*nz) ? idx : -1; };
+ int findNeighborPMP(real posX, real posY, real posZ){ int idx = getLinearIndex(getIdxM00(posX)+1, getIdx0M0(posY) , getIdx00M(posZ)+1); return (idx>=0) && (idx<nx*ny*nz) ? idx : -1; };
+ int findNeighborPPM(real posX, real posY, real posZ){ int idx = getLinearIndex(getIdxM00(posX)+1, getIdx0M0(posY)+1, getIdx00M(posZ) ); return (idx>=0) && (idx<nx*ny*nz) ? idx : -1; };
+ int findNeighborPPP(real posX, real posY, real posZ){ int idx = getLinearIndex(getIdxM00(posX)+1, getIdx0M0(posY)+1, getIdx00M(posZ)+1); return (idx>=0) && (idx<nx*ny*nz) ? idx : -1; };
+ int getIdxX(int linearIdx){ return linearIdx%nx;};
+ int getIdxY(int linearIdx){ return (linearIdx/nx)%ny;};
+ int getIdxZ(int linearIdx){ return linearIdx/(nx*ny); };
+ real getX(int linearIdx){ return getIdxX(linearIdx)*deltaX+minX; };
+ real getY(int linearIdx){ return getIdxY(linearIdx)*deltaY+minY; };
+ real getZ(int linearIdx){ return getIdxZ(linearIdx)*deltaZ+minZ; };
+ int getIdxM00(real posX){ return (posX-minX)/deltaX; };
+ int getIdx0M0(real posY){ return (posY-minY)/deltaY; };
+ int getIdx00M(real posZ){ return (posZ-minZ)/deltaZ; };
+ int getClosestIdxX(real posX){ int x = round((posX-minX)/deltaX); return x>nx ? nx : (x<0 ? 0 : x);};
+ int getClosestIdxY(real posY){ int y = round((posY-minY)/deltaY); return y>ny ? ny : (y<0 ? 0 : y);};
+ int getClosestIdxZ(real posZ){ int z = round((posZ-minZ)/deltaZ); return z>nz ? nz : (z<0 ? 0 : z);};
+ int getLinearIndex(int idxX, int idxY, int idxZ){ return idxX + nx*(idxY+ny*idxZ); };
+ int getNumberOfPointsInXYPlane(){ return nx*ny; }
+ int getNumberOfPointsInYZPlane(){ return ny*nz; }
+ int getNumberOfPointsInXZPlane(){ return nx*nz; }
+ int getNumberOfPoints(){ return nx*ny*nz; }
+ size_t getNumberOfQuantities(){ return quantities.size(); }
+ void loadFile();
+ void unloadFile();
+ bool isLoaded(){return loaded;};
+
+
+private:
+ void readHeader();
+ void printFileInfo();
+
+public:
+
+private:
+ std::string fileName;
+ real minX, maxX, minY, maxY, minZ, maxZ;
+ real deltaX, deltaY, deltaZ;
+ int nx, ny, nz;
+ std::vector<Quantity> quantities;
+ bool loaded;
+};
+
+class FileCollection
+{
+public:
+ FileCollection(std::string _prefix):
+ prefix(_prefix){};
+
+ virtual ~FileCollection() = default;
+
+ virtual size_t getNumberOfQuantities() = 0;
+
+ virtual FileType getFileType() = 0;
+
+protected:
+ std::string prefix;
+};
+
+
+class VTKFileCollection : public FileCollection
+{
+public:
+ VTKFileCollection(std::string _prefix):
+ FileCollection(_prefix)
+ {
+ findFiles();
+ };
+
+ FileType getFileType() override{ return FileType::VTK; };
+ size_t getNumberOfQuantities() override{ return files[0][0][0].getNumberOfQuantities(); }
+
+
+private:
+ void findFiles();
+ std::string makeFileName(int level, int id, int part)
+ {
+ return prefix + "_lev_" + StringUtil::toString<int>(level)
+ + "_ID_" + StringUtil::toString<int>(id)
+ + "_File_" + StringUtil::toString<int>(part)
+ + ".bin." + suffix;
+ };
+
+
+public:
+ static const inline std::string suffix = "vti";
+ std::vector<std::vector<std::vector<VTKFile>>> files;
+};
+
+
+class TransientBCInputFileReader
+{
+public:
+ TransientBCInputFileReader()
+ {
+ this->nPoints = 0;
+ this->nPointsRead = 0;
+ this->writingOffset = 0;
+ };
+ virtual ~TransientBCInputFileReader() = default;
+
+ virtual void getNextData(real* data, uint numberOfNodes, real time)=0;
+ virtual void fillArrays(std::vector<real>& coordsY, std::vector<real>& coordsZ)=0;
+ uint getNPoints(){return nPoints; };
+ uint getNPointsRead(){return nPointsRead; };
+ size_t getNumberOfQuantities(){ return nQuantities; };
+ void setWritingOffset(uint offset){ this->writingOffset = offset; }
+ void getNeighbors(uint* neighbor0PP, uint* neighbor0PM, uint* neighbor0MP, uint* neighbor0MM);
+ void getWeights(real* _weights0PP, real* _weights0PM, real* _weights0MP, real* _weights0MM);
+
+public:
+ std::vector<uint> planeNeighbor0PP, planeNeighbor0PM, planeNeighbor0MP, planeNeighbor0MM;
+ std::vector<real> weights0PP, weights0PM, weights0MP, weights0MM;
+
+protected:
+ uint nPoints, nPointsRead, writingOffset;
+ uint nReads=0;
+ size_t nQuantities=0;
+};
+
+
+class VTKReader : public TransientBCInputFileReader
+{
+public:
+ VTKReader(SPtr<VTKFileCollection> _fileCollection, uint _readLevel):
+ fileCollection(_fileCollection),
+ readLevel(_readLevel)
+ {
+ this->nQuantities = fileCollection->getNumberOfQuantities();
+ read = std::async([](){});
+ };
+ void getNextData(real* data, uint numberOfNodes, real time) override;
+ void fillArrays(std::vector<real>& coordsY, std::vector<real>& coordsZ) override;
+private:
+ uint getWriteIndex(int level, int id, int linearIdx);
+ void initializeIndexVectors();
+
+private:
+ std::vector<std::vector<std::vector<uint>>> readIndices, writeIndices;
+ std::vector<std::vector<size_t>> nFile;
+ SPtr<VTKFileCollection> fileCollection;
+ uint readLevel;
+ std::future<void> read;
+};
+
+
+SPtr<FileCollection> createFileCollection(std::string prefix, FileType type);
+SPtr<TransientBCInputFileReader> createReaderForCollection(SPtr<FileCollection> fileCollection, uint readLevel);
+
+#endif //TRANSIENTBCSETTER_H_
diff --git a/src/gpu/GridGenerator/grid/BoundaryConditions/BoundaryCondition.cpp b/src/gpu/GridGenerator/grid/BoundaryConditions/BoundaryCondition.cpp
index 5102f60fc295aadf4323a4b332bf3dd8f7f21dbf..b0fb2604946b83ead45c30adabbcfe8dc26fa656 100644
--- a/src/gpu/GridGenerator/grid/BoundaryConditions/BoundaryCondition.cpp
+++ b/src/gpu/GridGenerator/grid/BoundaryConditions/BoundaryCondition.cpp
@@ -36,12 +36,12 @@
#include "grid/BoundaryConditions/Side.h"
#include "grid/Grid.h"
+#include "GridGenerator/TransientBCSetter/TransientBCSetter.h"
bool gg::BoundaryCondition::isSide( SideType side ) const
{
return this->side->whoAmI() == side;
}
-
//////////////////////////////////////////////////////////////////////////
void VelocityBoundaryCondition::setVelocityProfile(
@@ -124,5 +124,4 @@ void StressBoundaryCondition::fillSamplingIndices(std::vector<SPtr<Grid> > grid,
this->velocitySamplingIndices.push_back( grid[level]->transCoordToIndex(x_sampling, y_sampling, z_sampling) );
}
-}
-
+}
\ No newline at end of file
diff --git a/src/gpu/GridGenerator/grid/BoundaryConditions/BoundaryCondition.h b/src/gpu/GridGenerator/grid/BoundaryConditions/BoundaryCondition.h
index 4a3990d9f815042297be76ae83a61268c8ad6815..22342aec9839afad9bb37b1b11812f6d1750ed7b 100644
--- a/src/gpu/GridGenerator/grid/BoundaryConditions/BoundaryCondition.h
+++ b/src/gpu/GridGenerator/grid/BoundaryConditions/BoundaryCondition.h
@@ -45,6 +45,8 @@ class Grid;
class Side;
enum class SideType;
+class TransientBCInputFileReader;
+
namespace gg
{
class BoundaryCondition
@@ -63,6 +65,8 @@ public:
bool isSide(SideType side) const;
real getQ(uint index, uint dir) { return this->qs[index][dir]; }
+
+ void getCoords( SPtr<Grid> grid, std::vector<real>& x, std::vector<real>& y, std::vector<real>& z);
};
}
@@ -246,6 +250,7 @@ public:
real getVy(uint index) { return this->vyList[index]; }
real getVz(uint index) { return this->vzList[index]; }
+
void setVelocityProfile( SPtr<Grid> grid, std::function<void(real,real,real,real&,real&,real&)> velocityProfile );
};
@@ -329,5 +334,32 @@ public:
real getNormalz(uint index) { return this->normalZList[index]; }
};
+class PrecursorBoundaryCondition : public gg::BoundaryCondition
+{
+public:
+ static SPtr<PrecursorBoundaryCondition> make(SPtr<TransientBCInputFileReader> reader, int timeStepsBetweenReads, real velocityX, real velocityY, real velocityZ)
+ {
+ return SPtr<PrecursorBoundaryCondition>(new PrecursorBoundaryCondition(reader, timeStepsBetweenReads, velocityX, velocityY, velocityZ));
+ }
+ SPtr<TransientBCInputFileReader> getReader(){ return reader; }
+ real getVelocityX() { return velocityX; }
+ real getVelocityY() { return velocityY; }
+ real getVelocityZ() { return velocityZ; }
+
+private:
+ PrecursorBoundaryCondition(SPtr<TransientBCInputFileReader> _reader, uint _timeStepsBetweenReads, real vx, real vy, real vz) : reader(_reader), timeStepsBetweenReads(_timeStepsBetweenReads), velocityX(vx), velocityY(vy), velocityZ(vz) { };
+ virtual char getType() const override
+ {
+ return vf::gpu::BC_VELOCITY;
+ }
+public:
+ uint timeStepsBetweenReads; //!> read data every nth timestep
+
+private:
+ real velocityX = 0.0;
+ real velocityY = 0.0;
+ real velocityZ = 0.0;
+ SPtr<TransientBCInputFileReader> reader;
+};
#endif
\ No newline at end of file
diff --git a/src/gpu/GridGenerator/grid/BoundaryConditions/Side.cpp b/src/gpu/GridGenerator/grid/BoundaryConditions/Side.cpp
index 6c7bf8ca1853826d83fb6a713ffe03716bd2cf9a..ba4eea50ffb6bc136528db31207274d626fe9b15 100644
--- a/src/gpu/GridGenerator/grid/BoundaryConditions/Side.cpp
+++ b/src/gpu/GridGenerator/grid/BoundaryConditions/Side.cpp
@@ -1,28 +1,28 @@
//=======================================================================================
-// ____ ____ __ ______ __________ __ __ __ __
-// \ \ | | | | | _ \ |___ ___| | | | | / \ | |
-// \ \ | | | | | |_) | | | | | | | / \ | |
-// \ \ | | | | | _ / | | | | | | / /\ \ | |
-// \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
-// \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
-// \ \ | | ________________________________________________________________
-// \ \ | | | ______________________________________________________________|
-// \ \| | | | __ __ __ __ ______ _______
-// \ | | |_____ | | | | | | | | | _ \ / _____)
-// \ | | _____| | | | | | | | | | | \ \ \_______
+// ____ ____ __ ______ __________ __ __ __ __
+// \ \ | | | | | _ \ |___ ___| | | | | / \ | |
+// \ \ | | | | | |_) | | | | | | | / \ | |
+// \ \ | | | | | _ / | | | | | | / /\ \ | |
+// \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
+// \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
+// \ \ | | ________________________________________________________________
+// \ \ | | | ______________________________________________________________|
+// \ \| | | | __ __ __ __ ______ _______
+// \ | | |_____ | | | | | | | | | _ \ / _____)
+// \ | | _____| | | | | | | | | | | \ \ \_______
// \ | | | | |_____ | \_/ | | | | |_/ / _____ |
-// \ _____| |__| |________| \_______/ |__| |______/ (_______/
+// \ _____| |__| |________| \_______/ |__| |______/ (_______/
//
-// This file is part of VirtualFluids. VirtualFluids is free software: you can
+// This file is part of VirtualFluids. VirtualFluids is free software: you can
// redistribute it and/or modify it under the terms of the GNU General Public
-// License as published by the Free Software Foundation, either version 3 of
+// License as published by the Free Software Foundation, either version 3 of
// the License, or (at your option) any later version.
-//
-// VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
-// ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
-// FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
+//
+// VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
+// ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+// FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
// for more details.
-//
+//
// You should have received a copy of the GNU General Public License along
// with VirtualFluids (see COPYING.txt). If not, see <http://www.gnu.org/licenses/>.
//
@@ -37,9 +37,24 @@
#include "grid/NodeValues.h"
#include "utilities/math/Math.h"
+#include <array>
+#include <cstddef>
+#include <vector>
using namespace gg;
+std::array<real, 3> Side::getNormal() const
+{
+ std::array<real, 3> normal;
+ if(this->getCoordinate()==X_INDEX)
+ normal = {(real)this->getDirection(), 0.0, 0.0};
+ if(this->getCoordinate()==Y_INDEX)
+ normal = {0.0, (real)this->getDirection(), 0.0};
+ if(this->getCoordinate()==Z_INDEX)
+ normal = {0.0, 0.0, (real)this->getDirection()};
+ return normal;
+}
+
void Side::addIndices(SPtr<Grid> grid, SPtr<BoundaryCondition> boundaryCondition, std::string coord, real constant,
real startInner, real endInner, real startOuter, real endOuter)
{
@@ -49,11 +64,17 @@ void Side::addIndices(SPtr<Grid> grid, SPtr<BoundaryCondition> boundaryCondition
{
const uint index = getIndex(grid, coord, constant, v1, v2);
- if ((index != INVALID_INDEX) && ( grid->getFieldEntry(index) == vf::gpu::FLUID
- || grid->getFieldEntry(index) == vf::gpu::FLUID_CFC
- || grid->getFieldEntry(index) == vf::gpu::FLUID_CFF
- || grid->getFieldEntry(index) == vf::gpu::FLUID_FCC
- || grid->getFieldEntry(index) == vf::gpu::FLUID_FCF ))
+ if(index == INVALID_INDEX)
+ continue;
+
+ if ( grid->getFieldEntry(index) == vf::gpu::FLUID
+ || grid->getFieldEntry(index) == vf::gpu::FLUID_CFC
+ || grid->getFieldEntry(index) == vf::gpu::FLUID_CFF
+ || grid->getFieldEntry(index) == vf::gpu::FLUID_FCC
+ || grid->getFieldEntry(index) == vf::gpu::FLUID_FCF
+ || grid->getFieldEntry(index) == vf::gpu::FLUID_FCF
+ // Overlap of BCs on edge nodes
+ || grid->nodeHasBC(index) )
{
grid->setFieldEntry(index, boundaryCondition->getType());
boundaryCondition->indices.push_back(index);
@@ -64,9 +85,12 @@ void Side::addIndices(SPtr<Grid> grid, SPtr<BoundaryCondition> boundaryCondition
boundaryCondition->patches.push_back(0);
}
-
}
}
+
+ const auto currentBCSide = this->whoAmI();
+ if(currentBCSide != SideType::GEOMETRY)
+ grid->addBCalreadySet(currentBCSide);
}
void Side::setPressureNeighborIndices(SPtr<BoundaryCondition> boundaryCondition, SPtr<Grid> grid, const uint index)
@@ -119,50 +143,111 @@ void Side::setStressSamplingIndices(SPtr<BoundaryCondition> boundaryCondition, S
void Side::setQs(SPtr<Grid> grid, SPtr<BoundaryCondition> boundaryCondition, uint index)
{
-
std::vector<real> qNode(grid->getEndDirection() + 1);
- for (int dir = 0; dir <= grid->getEndDirection(); dir++)
- {
- real x,y,z;
- grid->transIndexToCoords( index, x, y, z );
+ for (int dir = 0; dir <= grid->getEndDirection(); dir++) {
+ real x, y, z;
+ grid->transIndexToCoords(index, x, y, z);
- real coords[3] = {x,y,z};
+ std::array<real, 3> coords = { x, y, z };
+ std::array<real, 3> neighborCoords = getNeighborCoordinates(grid.get(), coords, (size_t)dir);
- real neighborX = x + grid->getDirection()[dir * DIMENSION + 0] * grid->getDelta();
- real neighborY = y + grid->getDirection()[dir * DIMENSION + 1] * grid->getDelta();
- real neighborZ = z + grid->getDirection()[dir * DIMENSION + 2] * grid->getDelta();
+ correctNeighborForPeriodicBoundaries(grid.get(), coords, neighborCoords);
- // correct neighbor coordinates in case of periodic boundaries
- if( grid->getPeriodicityX() && grid->getFieldEntry( grid->transCoordToIndex( neighborX, y, z ) ) == vf::gpu::STOPPER_OUT_OF_GRID_BOUNDARY )
- {
- if( neighborX > x ) neighborX = grid->getFirstFluidNode( coords, 0, grid->getStartX() );
- else neighborX = grid->getLastFluidNode ( coords, 0, grid->getEndX() );
- }
+ const uint neighborIndex = grid->transCoordToIndex(neighborCoords[0], neighborCoords[1], neighborCoords[2]);
- if( grid->getPeriodicityY() && grid->getFieldEntry( grid->transCoordToIndex( x, neighborY, z ) ) == vf::gpu::STOPPER_OUT_OF_GRID_BOUNDARY )
- {
- if( neighborY > y ) neighborY = grid->getFirstFluidNode( coords, 1, grid->getStartY() );
- else neighborY = grid->getLastFluidNode ( coords, 1, grid->getEndY() );
+ //! Only setting q's that partially point in the Side-normal direction
+ const bool alignedWithNormal = this->isAlignedWithMyNormal(grid.get(), dir);
+ if (grid->isStopperForBC(neighborIndex) && alignedWithNormal) {
+ qNode[dir] = 0.5;
+ } else {
+ qNode[dir] = -1.0;
}
- if( grid->getPeriodicityZ() && grid->getFieldEntry( grid->transCoordToIndex( x, y, neighborZ ) ) == vf::gpu::STOPPER_OUT_OF_GRID_BOUNDARY )
- {
- if( neighborZ > z ) neighborZ = grid->getFirstFluidNode( coords, 2, grid->getStartZ() );
- else neighborZ = grid->getLastFluidNode ( coords, 2, grid->getEndZ() );
+ // reset diagonals in case they were set by another bc
+ resetDiagonalsInCaseOfOtherBC(grid.get(), qNode, dir, coords);
+ }
+
+ boundaryCondition->qs.push_back(qNode);
+}
+
+std::array<real, 3> Side::getNeighborCoordinates(Grid *grid, const std::array<real, 3> &coordinates, size_t direction) const
+{
+ return { coordinates[0] + grid->getDirection()[direction * DIMENSION + 0] * grid->getDelta(),
+ coordinates[1] + grid->getDirection()[direction * DIMENSION + 1] * grid->getDelta(),
+ coordinates[2] + grid->getDirection()[direction * DIMENSION + 2] * grid->getDelta() };
+}
+
+bool Side::neighborNormalToSideIsAStopper(Grid *grid, const std::array<real, 3> &coordinates, SideType side) const
+{
+ const auto neighborCoords = getNeighborCoordinates(grid, coordinates, sideToD3Q27.at(side));
+ const auto neighborIndex = grid->transCoordToIndex(neighborCoords[0], neighborCoords[1], neighborCoords[2]);
+ return grid->isStopperForBC(neighborIndex);
+}
+
+void Side::resetDiagonalsInCaseOfOtherBC(Grid *grid, std::vector<real> &qNode, int dir,
+ const std::array<real, 3> &coordinates) const
+{
+ // When to reset a diagonal q to -1:
+ // - it is normal to another boundary condition which was already set
+ // - and it actually is influenced by the other bc:
+ // We check if its neighbor in the regular direction to the other bc is a stopper. If it is a stopper, it is influenced by the other bc.
+
+ if (qNode[dir] == 0.5 && grid->getBCAlreadySet().size() > 0) {
+ for (int i = 0; i < (int)grid->getBCAlreadySet().size(); i++) {
+ SideType otherDir = grid->getBCAlreadySet()[i];
+
+ // only reset normals for nodes on edges and corners, not on faces
+ if (!neighborNormalToSideIsAStopper(grid, coordinates, otherDir))
+ continue;
+
+ const auto otherNormal = normals.at(otherDir);
+ if (isAlignedWithNormal(grid, dir, otherNormal)) {
+ qNode[dir] = -1.0;
+ }
}
+ }
+}
- uint neighborIndex = grid->transCoordToIndex( neighborX, neighborY, neighborZ );
- if( grid->getFieldEntry(neighborIndex) == vf::gpu::STOPPER_OUT_OF_GRID_BOUNDARY ||
- grid->getFieldEntry(neighborIndex) == vf::gpu::STOPPER_OUT_OF_GRID ||
- grid->getFieldEntry(neighborIndex) == vf::gpu::STOPPER_SOLID )
- qNode[dir] = 0.5;
+bool Side::isAlignedWithMyNormal(const Grid *grid, int dir) const
+{
+ std::array<real, 3> normal = this->getNormal();
+ return isAlignedWithNormal(grid, dir, normal);
+}
+
+bool Side::isAlignedWithNormal(const Grid *grid, int dir, const std::array<real, 3> &normal) const
+{
+ return (normal[0] * grid->getDirection()[dir * DIMENSION + 0] +
+ normal[1] * grid->getDirection()[dir * DIMENSION + 1] +
+ normal[2] * grid->getDirection()[dir * DIMENSION + 2]) > 0;
+}
+
+void Side::correctNeighborForPeriodicBoundaries(const Grid *grid, std::array<real, 3>& coords, std::array<real, 3>& neighborCoords) const
+{
+ // correct neighbor coordinates in case of periodic boundaries
+ if (grid->getPeriodicityX() &&
+ grid->getFieldEntry(grid->transCoordToIndex(neighborCoords[0], coords[1], coords[2])) == vf::gpu::STOPPER_OUT_OF_GRID_BOUNDARY) {
+ if (neighborCoords[0] > coords[0])
+ neighborCoords[0] = grid->getFirstFluidNode(coords.data(), 0, grid->getStartX());
else
- qNode[dir] = -1.0;
+ neighborCoords[0] = grid->getLastFluidNode(coords.data(), 0, grid->getEndX());
+ }
+ if (grid->getPeriodicityY() &&
+ grid->getFieldEntry(grid->transCoordToIndex(coords[0], neighborCoords[1], coords[2])) == vf::gpu::STOPPER_OUT_OF_GRID_BOUNDARY) {
+ if (neighborCoords[1] > coords[1])
+ neighborCoords[1] = grid->getFirstFluidNode(coords.data(), 1, grid->getStartY());
+ else
+ neighborCoords[1] = grid->getLastFluidNode(coords.data(), 1, grid->getEndY());
}
- boundaryCondition->qs.push_back(qNode);
+ if (grid->getPeriodicityZ() &&
+ grid->getFieldEntry(grid->transCoordToIndex(coords[0], coords[1], neighborCoords[2])) == vf::gpu::STOPPER_OUT_OF_GRID_BOUNDARY) {
+ if (neighborCoords[2] > coords[2])
+ neighborCoords[2] = grid->getFirstFluidNode(coords.data(), 2, grid->getStartZ());
+ else
+ neighborCoords[2] = grid->getLastFluidNode(coords.data(), 2, grid->getEndZ());
+ }
}
uint Side::getIndex(SPtr<Grid> grid, std::string coord, real constant, real v1, real v2)
@@ -177,7 +262,7 @@ uint Side::getIndex(SPtr<Grid> grid, std::string coord, real constant, real v1,
}
-void Geometry::addIndices(std::vector<SPtr<Grid> > grids, uint level, SPtr<BoundaryCondition> boundaryCondition)
+void Geometry::addIndices(const std::vector<SPtr<Grid>> &grids, uint level, SPtr<BoundaryCondition> boundaryCondition)
{
auto geometryBoundaryCondition = std::dynamic_pointer_cast<GeometryBoundaryCondition>(boundaryCondition);
@@ -190,7 +275,7 @@ void Geometry::addIndices(std::vector<SPtr<Grid> > grids, uint level, SPtr<Bound
for (int dir = 0; dir <= grids[level]->getEndDirection(); dir++)
{
- const real q = grids[level]->getQValue(index, dir);
+ const real q = grids[level]->getQValue(index, dir);
qNode[dir] = q;
@@ -218,7 +303,7 @@ void Geometry::addIndices(std::vector<SPtr<Grid> > grids, uint level, SPtr<Bound
-void MX::addIndices(std::vector<SPtr<Grid> > grid, uint level, SPtr<BoundaryCondition> boundaryCondition)
+void MX::addIndices(const std::vector<SPtr<Grid>> &grid, uint level, SPtr<BoundaryCondition> boundaryCondition)
{
real startInner = grid[level]->getStartY();
real endInner = grid[level]->getEndY();
@@ -234,7 +319,7 @@ void MX::addIndices(std::vector<SPtr<Grid> > grid, uint level, SPtr<BoundaryCond
}
-void PX::addIndices(std::vector<SPtr<Grid> > grid, uint level, SPtr<BoundaryCondition> boundaryCondition)
+void PX::addIndices(const std::vector<SPtr<Grid>> &grid, uint level, SPtr<BoundaryCondition> boundaryCondition)
{
real startInner = grid[level]->getStartY();
real endInner = grid[level]->getEndY();
@@ -249,7 +334,7 @@ void PX::addIndices(std::vector<SPtr<Grid> > grid, uint level, SPtr<BoundaryCond
Side::addIndices(grid[level], boundaryCondition, "x", coordinateNormal, startInner, endInner, startOuter, endOuter);
}
-void MY::addIndices(std::vector<SPtr<Grid> > grid, uint level, SPtr<BoundaryCondition> boundaryCondition)
+void MY::addIndices(const std::vector<SPtr<Grid>> &grid, uint level, SPtr<BoundaryCondition> boundaryCondition)
{
real startInner = grid[level]->getStartX();
real endInner = grid[level]->getEndX();
@@ -265,7 +350,7 @@ void MY::addIndices(std::vector<SPtr<Grid> > grid, uint level, SPtr<BoundaryCond
}
-void PY::addIndices(std::vector<SPtr<Grid> > grid, uint level, SPtr<BoundaryCondition> boundaryCondition)
+void PY::addIndices(const std::vector<SPtr<Grid>> &grid, uint level, SPtr<BoundaryCondition> boundaryCondition)
{
real startInner = grid[level]->getStartX();
real endInner = grid[level]->getEndX();
@@ -281,7 +366,7 @@ void PY::addIndices(std::vector<SPtr<Grid> > grid, uint level, SPtr<BoundaryCond
}
-void MZ::addIndices(std::vector<SPtr<Grid> > grid, uint level, SPtr<BoundaryCondition> boundaryCondition)
+void MZ::addIndices(const std::vector<SPtr<Grid>> &grid, uint level, SPtr<BoundaryCondition> boundaryCondition)
{
real startInner = grid[level]->getStartX();
real endInner = grid[level]->getEndX();
@@ -296,7 +381,7 @@ void MZ::addIndices(std::vector<SPtr<Grid> > grid, uint level, SPtr<BoundaryCond
Side::addIndices(grid[level], boundaryCondition, "z", coordinateNormal, startInner, endInner, startOuter, endOuter);
}
-void PZ::addIndices(std::vector<SPtr<Grid> > grid, uint level, SPtr<BoundaryCondition> boundaryCondition)
+void PZ::addIndices(const std::vector<SPtr<Grid>> &grid, uint level, SPtr<BoundaryCondition> boundaryCondition)
{
real startInner = grid[level]->getStartX();
real endInner = grid[level]->getEndX();
@@ -307,6 +392,6 @@ void PZ::addIndices(std::vector<SPtr<Grid> > grid, uint level, SPtr<BoundaryCond
real coordinateNormal = grid[level]->getEndZ() - grid[level]->getDelta();
if( coordinateNormal < grid[0]->getEndZ() - grid[0]->getDelta() ) return;
-
+
Side::addIndices(grid[level], boundaryCondition, "z", coordinateNormal, startInner, endInner, startOuter, endOuter);
}
diff --git a/src/gpu/GridGenerator/grid/BoundaryConditions/Side.h b/src/gpu/GridGenerator/grid/BoundaryConditions/Side.h
index 6df6bfccc9a39b80de3ac43d057a03945d035b34..624b3722a1c909ba26063b49565779b924d34adc 100644
--- a/src/gpu/GridGenerator/grid/BoundaryConditions/Side.h
+++ b/src/gpu/GridGenerator/grid/BoundaryConditions/Side.h
@@ -33,10 +33,14 @@
#ifndef SIDE_H
#define SIDE_H
+#include <cstddef>
#include <string>
#include <vector>
+#include <map>
+#include <array>
#include "gpu/GridGenerator/global.h"
+#include "lbm/constants/D3Q27.h"
#define X_INDEX 0
#define Y_INDEX 1
@@ -59,37 +63,59 @@ enum class SideType
MX, PX, MY, PY, MZ, PZ, GEOMETRY
};
-
-
class Side
{
public:
virtual ~Side() = default;
- virtual void addIndices(std::vector<SPtr<Grid> > grid, uint level, SPtr<gg::BoundaryCondition> boundaryCondition) = 0;
+ virtual void addIndices(const std::vector<SPtr<Grid>> &grid, uint level,
+ SPtr<gg::BoundaryCondition> boundaryCondition) = 0;
virtual int getCoordinate() const = 0;
virtual int getDirection() const = 0;
virtual SideType whoAmI() const = 0;
+ std::array<real, 3> getNormal() const;
+
protected:
- static void addIndices(SPtr<Grid> grid, SPtr<gg::BoundaryCondition> boundaryCondition, std::string coord, real constant,
+ void addIndices(SPtr<Grid> grid, SPtr<gg::BoundaryCondition> boundaryCondition, std::string coord, real constant,
real startInner, real endInner, real startOuter, real endOuter);
static void setPressureNeighborIndices(SPtr<gg::BoundaryCondition> boundaryCondition, SPtr<Grid> grid, const uint index);
static void setStressSamplingIndices(SPtr<gg::BoundaryCondition> boundaryCondition, SPtr<Grid> grid, const uint index);
- static void setQs(SPtr<Grid> grid, SPtr<gg::BoundaryCondition> boundaryCondition, uint index);
+ void setQs(SPtr<Grid> grid, SPtr<gg::BoundaryCondition> boundaryCondition, uint index);
+
+ virtual void correctNeighborForPeriodicBoundaries(const Grid *grid, std::array<real, 3>& coords, std::array<real, 3>& neighbors) const;
+
+ virtual bool isAlignedWithMyNormal(const Grid *grid, int dir) const;
+ bool isAlignedWithNormal(const Grid *grid, int dir, const std::array<real, 3>& normal) const;
private:
static uint getIndex(SPtr<Grid> grid, std::string coord, real constant, real v1, real v2);
+ void resetDiagonalsInCaseOfOtherBC(Grid *grid, std::vector<real>& qNode, int dir, const std::array<real, 3> &coordinates) const;
+ std::array<real, 3> getNeighborCoordinates(Grid *grid, const std::array<real, 3> &coordinates,
+ size_t direction) const;
+ bool neighborNormalToSideIsAStopper(Grid *grid, const std::array<real, 3> &coordinates, SideType side) const;
+
+protected:
+ const std::map<SideType, const std::array<real, 3>> normals = {
+ { SideType::MX, { NEGATIVE_DIR, 0.0, 0.0 } }, { SideType::PX, { POSITIVE_DIR, 0.0, 0.0 } },
+ { SideType::MY, { 0.0, NEGATIVE_DIR, 0.0 } }, { SideType::PY, { 0.0, POSITIVE_DIR, 0.0 } },
+ { SideType::MZ, { 0.0, 0.0, NEGATIVE_DIR } }, { SideType::PZ, { 0.0, 0.0, POSITIVE_DIR } }
+ };
+ const std::map<SideType, size_t> sideToD3Q27 = {
+ { SideType::MX, vf::lbm::dir::DIR_M00 }, { SideType::PX, vf::lbm::dir::DIR_P00 },
+ { SideType::MY, vf::lbm::dir::DIR_0M0 }, { SideType::PY, vf::lbm::dir::DIR_0P0 },
+ { SideType::MZ, vf::lbm::dir::DIR_00M }, { SideType::PZ, vf::lbm::dir::DIR_00P }
+ };
};
class Geometry : public Side
{
public:
- void addIndices(std::vector<SPtr<Grid> > grid, uint level, SPtr<gg::BoundaryCondition> boundaryCondition) override;
+ void addIndices(const std::vector<SPtr<Grid>> &grid, uint level, SPtr<gg::BoundaryCondition> boundaryCondition) override;
int getCoordinate() const override
{
@@ -110,7 +136,7 @@ public:
class MX : public Side
{
public:
- void addIndices(std::vector<SPtr<Grid> > grid, uint level, SPtr<gg::BoundaryCondition> boundaryCondition) override;
+ void addIndices(const std::vector<SPtr<Grid>> &grid, uint level, SPtr<gg::BoundaryCondition> boundaryCondition) override;
int getCoordinate() const override
{
@@ -131,7 +157,7 @@ public:
class PX : public Side
{
public:
- void addIndices(std::vector<SPtr<Grid> > grid, uint level, SPtr<gg::BoundaryCondition> boundaryCondition) override;
+ void addIndices(const std::vector<SPtr<Grid>> &grid, uint level, SPtr<gg::BoundaryCondition> boundaryCondition) override;
int getCoordinate() const override
{
@@ -153,7 +179,7 @@ public:
class MY : public Side
{
public:
- void addIndices(std::vector<SPtr<Grid> > grid, uint level, SPtr<gg::BoundaryCondition> boundaryCondition) override;
+ void addIndices(const std::vector<SPtr<Grid>> &grid, uint level, SPtr<gg::BoundaryCondition> boundaryCondition) override;
int getCoordinate() const override
{
@@ -174,7 +200,7 @@ public:
class PY : public Side
{
public:
- void addIndices(std::vector<SPtr<Grid> > grid, uint level, SPtr<gg::BoundaryCondition> boundaryCondition) override;
+ void addIndices(const std::vector<SPtr<Grid>> &grid, uint level, SPtr<gg::BoundaryCondition> boundaryCondition) override;
int getCoordinate() const override
{
@@ -196,7 +222,7 @@ public:
class MZ : public Side
{
public:
- void addIndices(std::vector<SPtr<Grid> > grid, uint level, SPtr<gg::BoundaryCondition> boundaryCondition) override;
+ void addIndices(const std::vector<SPtr<Grid>> &grid, uint level, SPtr<gg::BoundaryCondition> boundaryCondition) override;
int getCoordinate() const override
{
@@ -217,7 +243,7 @@ public:
class PZ : public Side
{
public:
- void addIndices(std::vector<SPtr<Grid> > grid, uint level, SPtr<gg::BoundaryCondition> boundaryCondition) override;
+ void addIndices(const std::vector<SPtr<Grid>> &grid, uint level, SPtr<gg::BoundaryCondition> boundaryCondition) override;
int getCoordinate() const override
{
diff --git a/src/gpu/GridGenerator/grid/BoundaryConditions/SideTest.cpp b/src/gpu/GridGenerator/grid/BoundaryConditions/SideTest.cpp
new file mode 100644
index 0000000000000000000000000000000000000000..36a286a8766db4af7e109eb3f8d47add401779f9
--- /dev/null
+++ b/src/gpu/GridGenerator/grid/BoundaryConditions/SideTest.cpp
@@ -0,0 +1,873 @@
+#include "Side.h"
+#include "PointerDefinitions.h"
+#include "gpu/GridGenerator/grid/BoundaryConditions/BoundaryCondition.h"
+#include "grid/GridImp.h"
+#include "grid/NodeValues.h"
+#include "lbm/constants/D3Q27.h"
+#include "gmock/gmock.h"
+#include <algorithm>
+#include <gtest/gtest.h>
+#include <iostream>
+#include <memory>
+#include <stdexcept>
+#include <vector>
+
+using namespace vf::gpu;
+using namespace vf::lbm::dir;
+
+class SideTestSpecificSubclass : public Side
+{
+
+public:
+ void setQs(SPtr<Grid> grid, SPtr<gg::BoundaryCondition> boundaryCondition, uint index)
+ {
+ Side::setQs(grid, boundaryCondition, index);
+ };
+ int sideDirection = POSITIVE_DIR;
+ int coordinateDirection = X_INDEX;
+ SideType mySide = SideType::PX;
+
+private:
+ void correctNeighborForPeriodicBoundaries(const Grid *grid, std::array<real, 3>& coords, std::array<real, 3>& neighbors) const override
+ {
+ }
+
+ int getDirection() const override
+ {
+ return sideDirection;
+ }
+
+ void addIndices(const std::vector<SPtr<Grid>> &grid, uint level, SPtr<gg::BoundaryCondition> boundaryCondition) override
+ {
+ }
+
+ int getCoordinate() const override
+ {
+ return coordinateDirection;
+ }
+
+ SideType whoAmI() const override
+ {
+ return mySide;
+ }
+};
+
+class GridDouble : public GridImp
+{
+
+public:
+ int endDirection = -1;
+
+ GridDouble()
+ {
+ this->distribution = DistributionHelper::getDistribution27();
+ }
+
+ void transIndexToCoords(uint index, real &x, real &y, real &z) const override
+ {
+ x = 0;
+ y = 0;
+ z = 0;
+ }
+
+ real getDelta() const override
+ {
+ return 1.0;
+ }
+
+ uint transCoordToIndex(const real &x, const real &y, const real &z) const override
+ {
+ return 0;
+ }
+
+ char getFieldEntry(uint /*matrixIndex*/) const override
+ {
+ return STOPPER_OUT_OF_GRID_BOUNDARY;
+ }
+
+ int getEndDirection() const override
+ {
+ return endDirection;
+ }
+};
+
+class BoundaryConditionSpy : public gg::BoundaryCondition
+{
+public:
+ char getType() const override
+ {
+ return 't';
+ };
+ const std::vector<std::vector<real>> &getQs()
+ {
+ return this->qs;
+ }
+ void resetQVector()
+ {
+ this->qs.clear();
+ }
+};
+
+class SideTestBC : public testing::Test
+{
+protected:
+ SideTestSpecificSubclass side;
+ SPtr<GridDouble> grid = std::make_shared<GridDouble>();
+ SPtr<BoundaryConditionSpy> bc = std::make_shared<BoundaryConditionSpy>();
+ uint index = 0;
+
+ std::vector<real> noBC;
+
+ void SetUp() override
+ {
+ grid->endDirection = 26;
+ }
+};
+
+TEST_F(SideTestBC, setQs2D_whenSettingPX_setAllQsNormalToBC)
+{
+ grid->endDirection = 10;
+ side.coordinateDirection = X_INDEX;
+ side.sideDirection = POSITIVE_DIR;
+
+ side.setQs(grid, bc, index);
+ auto actualQs = bc->getQs()[0];
+
+ std::vector<real> expectedQs(11, -1);
+ expectedQs[DIR_P00] = 0.5;
+ expectedQs[DIR_PP0] = 0.5;
+ expectedQs[DIR_PM0] = 0.5;
+ EXPECT_THAT(actualQs, testing::Eq(expectedQs));
+}
+
+TEST_F(SideTestBC, setQs2D_givenPYhasBeenSet_thenSetPX_doNotSetSameQsAgain)
+{
+ grid->endDirection = 10;
+ side.coordinateDirection = X_INDEX;
+ side.sideDirection = POSITIVE_DIR;
+ grid->addBCalreadySet(SideType::PY);
+
+ side.setQs(grid, bc, index);
+ auto actualQs = bc->getQs()[0];
+
+ std::vector<real> expectedQs(11, -1);
+ expectedQs[DIR_P00] = 0.5;
+ expectedQs[DIR_PM0] = 0.5;
+ EXPECT_THAT(actualQs, testing::Eq(expectedQs));
+}
+
+TEST_F(SideTestBC, setQs3D_givenMXhasBeenSet_thenSetPX_setAllQsNormalToPX)
+{
+ side.coordinateDirection = X_INDEX;
+ side.sideDirection = POSITIVE_DIR;
+
+ // no previous BC on this node
+
+ side.setQs(grid, bc, index);
+ auto actualQs = bc->getQs()[0];
+
+ std::vector<real> expectedQs(27, -1);
+ expectedQs[DIR_P00] = 0.5;
+ expectedQs[DIR_PP0] = 0.5;
+ expectedQs[DIR_PM0] = 0.5;
+ expectedQs[DIR_P0P] = 0.5;
+ expectedQs[DIR_P0M] = 0.5;
+ expectedQs[DIR_PPP] = 0.5;
+ expectedQs[DIR_PMP] = 0.5;
+ expectedQs[DIR_PPM] = 0.5;
+ expectedQs[DIR_PMM] = 0.5;
+ EXPECT_THAT(actualQs, testing::Eq(expectedQs));
+
+ // node already has BC in MX direction, but this does not change anything
+
+ grid->addBCalreadySet(SideType::MX);
+
+ side.setQs(grid, bc, index);
+ actualQs = bc->getQs()[0];
+
+ EXPECT_THAT(actualQs, testing::Eq(expectedQs));
+}
+
+TEST_F(SideTestBC, setQs3D_givenGeometryBCInVector_thenSetPX_throws)
+{
+ // do not add Geometry BC to this vector, as it has an invalid normal
+ grid->addBCalreadySet(SideType::GEOMETRY);
+
+ EXPECT_THROW(side.setQs(grid, bc, index), std::out_of_range);
+}
+
+TEST_F(SideTestBC, setQs3D_whenSettingPX_setAllQsNormalToBC)
+{
+ side.coordinateDirection = X_INDEX;
+ side.sideDirection = POSITIVE_DIR;
+
+ side.setQs(grid, bc, index);
+ auto actualQs = bc->getQs()[0];
+
+ std::vector<real> expectedQs(27, -1);
+ expectedQs[DIR_P00] = 0.5;
+ expectedQs[DIR_PP0] = 0.5;
+ expectedQs[DIR_PM0] = 0.5;
+ expectedQs[DIR_P0P] = 0.5;
+ expectedQs[DIR_P0M] = 0.5;
+ expectedQs[DIR_PPP] = 0.5;
+ expectedQs[DIR_PMP] = 0.5;
+ expectedQs[DIR_PPM] = 0.5;
+ expectedQs[DIR_PMM] = 0.5;
+ EXPECT_THAT(actualQs, testing::Eq(expectedQs));
+}
+
+TEST_F(SideTestBC, setQs3D_givenPYhasBeenSet_thenSetPX_doNotSetSameQsAgain)
+{
+ side.coordinateDirection = X_INDEX;
+ side.sideDirection = POSITIVE_DIR;
+ grid->addBCalreadySet(SideType::PY);
+
+ side.setQs(grid, bc, index);
+ auto actualQs = bc->getQs()[0];
+
+ std::vector<real> expectedQs(27, -1);
+ expectedQs[DIR_P00] = 0.5;
+ expectedQs[DIR_PM0] = 0.5;
+ expectedQs[DIR_P0P] = 0.5;
+ expectedQs[DIR_P0M] = 0.5;
+ expectedQs[DIR_PMP] = 0.5;
+ expectedQs[DIR_PMM] = 0.5;
+ EXPECT_THAT(actualQs, testing::Eq(expectedQs));
+}
+
+TEST_F(SideTestBC, setQs3D_givenMYhasBeenSet_thenSetPX_doNotSetSameQsAgain)
+{
+ side.coordinateDirection = X_INDEX;
+ side.sideDirection = POSITIVE_DIR;
+ grid->addBCalreadySet(SideType::MY);
+
+ side.setQs(grid, bc, index);
+ auto actualQs = bc->getQs()[0];
+
+ std::vector<real> expectedQs(27, -1);
+ expectedQs[DIR_P00] = 0.5;
+ expectedQs[DIR_PP0] = 0.5;
+ expectedQs[DIR_P0P] = 0.5;
+ expectedQs[DIR_P0M] = 0.5;
+ expectedQs[DIR_PPP] = 0.5;
+ expectedQs[DIR_PPM] = 0.5;
+ EXPECT_THAT(actualQs, testing::Eq(expectedQs));
+}
+
+TEST_F(SideTestBC, setQs3D_givenPZhasBeenSet_thenSetPX_doNotSetSameQsAgain)
+{
+ side.coordinateDirection = X_INDEX;
+ side.sideDirection = POSITIVE_DIR;
+ grid->addBCalreadySet(SideType::PZ);
+
+ side.setQs(grid, bc, index);
+ auto actualQs = bc->getQs()[0];
+
+ std::vector<real> expectedQs(27, -1);
+ expectedQs[DIR_P00] = 0.5;
+ expectedQs[DIR_PP0] = 0.5;
+ expectedQs[DIR_PM0] = 0.5;
+ expectedQs[DIR_P0M] = 0.5;
+ expectedQs[DIR_PPM] = 0.5;
+ expectedQs[DIR_PMM] = 0.5;
+ EXPECT_THAT(actualQs, testing::Eq(expectedQs));
+}
+
+TEST_F(SideTestBC, setQs3D_givenMZhasBeenSet_thenSetPX_doNotSetSameQsAgain)
+{
+ side.coordinateDirection = X_INDEX;
+ side.sideDirection = POSITIVE_DIR;
+ grid->addBCalreadySet(SideType::MZ);
+
+ side.setQs(grid, bc, index);
+ auto actualQs = bc->getQs()[0];
+
+ std::vector<real> expectedQs(27, -1);
+ expectedQs[DIR_P00] = 0.5;
+ expectedQs[DIR_PP0] = 0.5;
+ expectedQs[DIR_PM0] = 0.5;
+ expectedQs[DIR_P0P] = 0.5;
+ expectedQs[DIR_PPP] = 0.5;
+ expectedQs[DIR_PMP] = 0.5;
+ EXPECT_THAT(actualQs, testing::Eq(expectedQs));
+}
+
+TEST_F(SideTestBC, setQs3D_givenPYandMZhaveBeenSet_thenSetPX_doNotSetSameQsAgain)
+{
+ side.coordinateDirection = X_INDEX;
+ side.sideDirection = POSITIVE_DIR;
+ grid->addBCalreadySet(SideType::PY);
+ grid->addBCalreadySet(SideType::MZ);
+
+ side.setQs(grid, bc, index);
+ auto actualQs = bc->getQs()[0];
+
+ std::vector<real> expectedQsForTwoPreviousBCs(27, -1);
+ expectedQsForTwoPreviousBCs[DIR_P00] = 0.5;
+ expectedQsForTwoPreviousBCs[DIR_PM0] = 0.5;
+ expectedQsForTwoPreviousBCs[DIR_P0P] = 0.5;
+ expectedQsForTwoPreviousBCs[DIR_PMP] = 0.5;
+ EXPECT_THAT(actualQs, testing::Eq(expectedQsForTwoPreviousBCs));
+}
+
+TEST_F(SideTestBC, setQs3D_givenPYandPZhaveBeenSet_thenSetPX_doNotSetSameQsAgain)
+{
+ side.coordinateDirection = X_INDEX;
+ side.sideDirection = POSITIVE_DIR;
+ grid->addBCalreadySet(SideType::PY);
+ grid->addBCalreadySet(SideType::PZ);
+
+ side.setQs(grid, bc, index);
+ auto actualQs = bc->getQs()[0];
+
+ std::vector<real> expectedQsForTwoPreviousBCs(27, -1);
+ expectedQsForTwoPreviousBCs[DIR_P00] = 0.5;
+ expectedQsForTwoPreviousBCs[DIR_PM0] = 0.5;
+ expectedQsForTwoPreviousBCs[DIR_P0M] = 0.5;
+ expectedQsForTwoPreviousBCs[DIR_PMM] = 0.5;
+ EXPECT_THAT(actualQs, testing::Eq(expectedQsForTwoPreviousBCs));
+}
+
+TEST_F(SideTestBC, setQs3D_givenMYandPZhaveBeenSet_thenSetPX_doNotSetSameQsAgain)
+{
+ side.coordinateDirection = X_INDEX;
+ side.sideDirection = POSITIVE_DIR;
+ grid->addBCalreadySet(SideType::MY);
+ grid->addBCalreadySet(SideType::PZ);
+
+ side.setQs(grid, bc, index);
+ auto actualQs = bc->getQs()[0];
+
+ std::vector<real> expectedQsForTwoPreviousBCs(27, -1);
+ expectedQsForTwoPreviousBCs[DIR_P00] = 0.5;
+ expectedQsForTwoPreviousBCs[DIR_PP0] = 0.5;
+ expectedQsForTwoPreviousBCs[DIR_P0M] = 0.5;
+ expectedQsForTwoPreviousBCs[DIR_PPM] = 0.5;
+ EXPECT_THAT(actualQs, testing::Eq(expectedQsForTwoPreviousBCs));
+}
+
+TEST_F(SideTestBC, setQs3D_givenMYandMZhaveBeenSet_thenSetPX_doNotSetSameQsAgain)
+{
+ side.coordinateDirection = X_INDEX;
+ side.sideDirection = POSITIVE_DIR;
+ grid->addBCalreadySet(SideType::MY);
+ grid->addBCalreadySet(SideType::MZ);
+
+ side.setQs(grid, bc, index);
+ auto actualQs = bc->getQs()[0];
+
+ std::vector<real> expectedQsForTwoPreviousBCs(27, -1);
+ expectedQsForTwoPreviousBCs[DIR_P00] = 0.5;
+ expectedQsForTwoPreviousBCs[DIR_PP0] = 0.5;
+ expectedQsForTwoPreviousBCs[DIR_P0P] = 0.5;
+ expectedQsForTwoPreviousBCs[DIR_PPP] = 0.5;
+ EXPECT_THAT(actualQs, testing::Eq(expectedQsForTwoPreviousBCs));
+}
+
+TEST_F(SideTestBC, setQs3D_whenSettingMX_setAllQsNormalToBC)
+{
+ side.coordinateDirection = X_INDEX;
+ side.sideDirection = NEGATIVE_DIR;
+
+ side.setQs(grid, bc, index);
+ auto actualQs = bc->getQs()[0];
+
+ std::vector<real> expectedQs(27, -1);
+ expectedQs[DIR_M00] = 0.5;
+ expectedQs[DIR_MP0] = 0.5;
+ expectedQs[DIR_MM0] = 0.5;
+ expectedQs[DIR_M0P] = 0.5;
+ expectedQs[DIR_M0M] = 0.5;
+ expectedQs[DIR_MPP] = 0.5;
+ expectedQs[DIR_MMP] = 0.5;
+ expectedQs[DIR_MPM] = 0.5;
+ expectedQs[DIR_MMM] = 0.5;
+ EXPECT_THAT(actualQs, testing::Eq(expectedQs));
+}
+
+TEST_F(SideTestBC, setQs3D_givenPYhasBeenSet_thenSetMX_doNotSetSameQsAgain)
+{
+ side.coordinateDirection = X_INDEX;
+ side.sideDirection = NEGATIVE_DIR;
+ grid->addBCalreadySet(SideType::PY);
+
+ side.setQs(grid, bc, index);
+ auto actualQs = bc->getQs()[0];
+
+ std::vector<real> expectedQs(27, -1);
+ expectedQs[DIR_M00] = 0.5;
+ expectedQs[DIR_MM0] = 0.5;
+ expectedQs[DIR_M0P] = 0.5;
+ expectedQs[DIR_M0M] = 0.5;
+ expectedQs[DIR_MMP] = 0.5;
+ expectedQs[DIR_MMM] = 0.5;
+ EXPECT_THAT(actualQs, testing::Eq(expectedQs));
+}
+
+TEST_F(SideTestBC, setQs3D_givenMYhasBeenSet_thenSetMX_doNotSetSameQsAgain)
+{
+ side.coordinateDirection = X_INDEX;
+ side.sideDirection = NEGATIVE_DIR;
+ grid->addBCalreadySet(SideType::MY);
+
+ side.setQs(grid, bc, index);
+ auto actualQs = bc->getQs()[0];
+
+ std::vector<real> expectedQs(27, -1);
+ expectedQs[DIR_M00] = 0.5;
+ expectedQs[DIR_MP0] = 0.5;
+ expectedQs[DIR_M0P] = 0.5;
+ expectedQs[DIR_M0M] = 0.5;
+ expectedQs[DIR_MPP] = 0.5;
+ expectedQs[DIR_MPM] = 0.5;
+ EXPECT_THAT(actualQs, testing::Eq(expectedQs));
+}
+
+TEST_F(SideTestBC, setQs3D_givenPZhasBeenSet_thenSetMX_doNotSetSameQsAgain)
+{
+ side.coordinateDirection = X_INDEX;
+ side.sideDirection = NEGATIVE_DIR;
+ grid->addBCalreadySet(SideType::PZ);
+
+ side.setQs(grid, bc, index);
+ auto actualQs = bc->getQs()[0];
+
+ std::vector<real> expectedQs(27, -1);
+ expectedQs[DIR_M00] = 0.5;
+ expectedQs[DIR_MP0] = 0.5;
+ expectedQs[DIR_MM0] = 0.5;
+ expectedQs[DIR_M0M] = 0.5;
+ expectedQs[DIR_MPM] = 0.5;
+ expectedQs[DIR_MMM] = 0.5;
+ EXPECT_THAT(actualQs, testing::Eq(expectedQs));
+}
+
+TEST_F(SideTestBC, setQs3D_givenMZhasBeenSet_thenSetMX_doNotSetSameQsAgain)
+{
+ side.coordinateDirection = X_INDEX;
+ side.sideDirection = NEGATIVE_DIR;
+ grid->addBCalreadySet(SideType::MZ);
+
+ side.setQs(grid, bc, index);
+ auto actualQs = bc->getQs()[0];
+
+ std::vector<real> expectedQs(27, -1);
+ expectedQs[DIR_M00] = 0.5;
+ expectedQs[DIR_MP0] = 0.5;
+ expectedQs[DIR_MM0] = 0.5;
+ expectedQs[DIR_M0P] = 0.5;
+ expectedQs[DIR_MPP] = 0.5;
+ expectedQs[DIR_MMP] = 0.5;
+ EXPECT_THAT(actualQs, testing::Eq(expectedQs));
+}
+
+TEST_F(SideTestBC, setQs3D_givenPYandMZhaveBeenSet_thenSetMX_doNotSetSameQsAgain)
+{
+ side.coordinateDirection = X_INDEX;
+ side.sideDirection = NEGATIVE_DIR;
+ grid->addBCalreadySet(SideType::PY);
+ grid->addBCalreadySet(SideType::MZ);
+
+ side.setQs(grid, bc, index);
+ auto actualQs = bc->getQs()[0];
+
+ std::vector<real> expectedQsForTwoPreviousBCs(27, -1);
+ expectedQsForTwoPreviousBCs[DIR_M00] = 0.5;
+ expectedQsForTwoPreviousBCs[DIR_MM0] = 0.5;
+ expectedQsForTwoPreviousBCs[DIR_M0P] = 0.5;
+ expectedQsForTwoPreviousBCs[DIR_MMP] = 0.5;
+ EXPECT_THAT(actualQs, testing::Eq(expectedQsForTwoPreviousBCs));
+}
+
+TEST_F(SideTestBC, setQs3D_givenPYandPZhaveBeenSet_thenSetMX_doNotSetSameQsAgain)
+{
+ side.coordinateDirection = X_INDEX;
+ side.sideDirection = NEGATIVE_DIR;
+ grid->addBCalreadySet(SideType::PY);
+ grid->addBCalreadySet(SideType::PZ);
+
+ side.setQs(grid, bc, index);
+ auto actualQs = bc->getQs()[0];
+
+ std::vector<real> expectedQsForTwoPreviousBCs(27, -1);
+ expectedQsForTwoPreviousBCs[DIR_M00] = 0.5;
+ expectedQsForTwoPreviousBCs[DIR_MM0] = 0.5;
+ expectedQsForTwoPreviousBCs[DIR_M0M] = 0.5;
+ expectedQsForTwoPreviousBCs[DIR_MMM] = 0.5;
+ EXPECT_THAT(actualQs, testing::Eq(expectedQsForTwoPreviousBCs));
+}
+
+TEST_F(SideTestBC, setQs3D_givenMYandPZhaveBeenSet_thenSetMX_doNotSetSameQsAgain)
+{
+ side.coordinateDirection = X_INDEX;
+ side.sideDirection = NEGATIVE_DIR;
+ grid->addBCalreadySet(SideType::MY);
+ grid->addBCalreadySet(SideType::PZ);
+
+ side.setQs(grid, bc, index);
+ auto actualQs = bc->getQs()[0];
+
+ std::vector<real> expectedQsForTwoPreviousBCs(27, -1);
+ expectedQsForTwoPreviousBCs[DIR_M00] = 0.5;
+ expectedQsForTwoPreviousBCs[DIR_MP0] = 0.5;
+ expectedQsForTwoPreviousBCs[DIR_M0M] = 0.5;
+ expectedQsForTwoPreviousBCs[DIR_MPM] = 0.5;
+ EXPECT_THAT(actualQs, testing::Eq(expectedQsForTwoPreviousBCs));
+}
+
+TEST_F(SideTestBC, setQs3D_givenMYandMZhaveBeenSet_thenSetMX_doNotSetSameQsAgain)
+{
+ side.coordinateDirection = X_INDEX;
+ side.sideDirection = NEGATIVE_DIR;
+ grid->addBCalreadySet(SideType::MY);
+ grid->addBCalreadySet(SideType::MZ);
+
+ side.setQs(grid, bc, index);
+ auto actualQs = bc->getQs()[0];
+
+ std::vector<real> expectedQsForTwoPreviousBCs(27, -1);
+ expectedQsForTwoPreviousBCs[DIR_M00] = 0.5;
+ expectedQsForTwoPreviousBCs[DIR_MP0] = 0.5;
+ expectedQsForTwoPreviousBCs[DIR_M0P] = 0.5;
+ expectedQsForTwoPreviousBCs[DIR_MPP] = 0.5;
+ EXPECT_THAT(actualQs, testing::Eq(expectedQsForTwoPreviousBCs));
+}
+
+TEST_F(SideTestBC, setQs3D_whenSettingMZ_setAllQsNormalToBC)
+{
+ side.coordinateDirection = Z_INDEX;
+ side.sideDirection = NEGATIVE_DIR;
+
+ side.setQs(grid, bc, index);
+ auto actualQs = bc->getQs()[0];
+
+ std::vector<real> expectedQs(27, -1);
+ expectedQs[DIR_00M] = 0.5;
+ expectedQs[DIR_P0M] = 0.5;
+ expectedQs[DIR_M0M] = 0.5;
+ expectedQs[DIR_0PM] = 0.5;
+ expectedQs[DIR_0MM] = 0.5;
+ expectedQs[DIR_PPM] = 0.5;
+ expectedQs[DIR_MPM] = 0.5;
+ expectedQs[DIR_PMM] = 0.5;
+ expectedQs[DIR_MMM] = 0.5;
+ EXPECT_THAT(actualQs, testing::Eq(expectedQs));
+}
+
+TEST_F(SideTestBC, setQs3D_givenMYhasBeenSet_thenSetMZ_doNotSetSameQsAgain)
+{
+ side.coordinateDirection = Z_INDEX;
+ side.sideDirection = NEGATIVE_DIR;
+ grid->addBCalreadySet(SideType::MY);
+
+ side.setQs(grid, bc, index);
+ auto actualQs = bc->getQs()[0];
+
+ std::vector<real> expectedQs(27, -1);
+ expectedQs[DIR_00M] = 0.5;
+ expectedQs[DIR_P0M] = 0.5;
+ expectedQs[DIR_M0M] = 0.5;
+ expectedQs[DIR_0PM] = 0.5;
+ expectedQs[DIR_PPM] = 0.5;
+ expectedQs[DIR_MPM] = 0.5;
+ EXPECT_THAT(actualQs, testing::Eq(expectedQs));
+}
+
+TEST_F(SideTestBC, setQs3D_givenPYhasBeenSet_thenSetMZ_doNotSetSameQsAgain)
+{
+ side.coordinateDirection = Z_INDEX;
+ side.sideDirection = NEGATIVE_DIR;
+ grid->addBCalreadySet(SideType::PY);
+
+ side.setQs(grid, bc, index);
+ auto actualQs = bc->getQs()[0];
+
+ std::vector<real> expectedQs(27, -1);
+ expectedQs[DIR_00M] = 0.5;
+ expectedQs[DIR_P0M] = 0.5;
+ expectedQs[DIR_M0M] = 0.5;
+ expectedQs[DIR_0MM] = 0.5;
+ expectedQs[DIR_PMM] = 0.5;
+ expectedQs[DIR_MMM] = 0.5;
+ EXPECT_THAT(actualQs, testing::Eq(expectedQs));
+}
+
+TEST_F(SideTestBC, setQs3D_givenPXhasBeenSet_thenSetMZ_doNotSetSameQsAgain)
+{
+ side.coordinateDirection = Z_INDEX;
+ side.sideDirection = NEGATIVE_DIR;
+ grid->addBCalreadySet(SideType::PX);
+
+ side.setQs(grid, bc, index);
+ auto actualQs = bc->getQs()[0];
+
+ std::vector<real> expectedQs(27, -1);
+ expectedQs[DIR_00M] = 0.5;
+ expectedQs[DIR_M0M] = 0.5;
+ expectedQs[DIR_0PM] = 0.5;
+ expectedQs[DIR_0MM] = 0.5;
+ expectedQs[DIR_MPM] = 0.5;
+ expectedQs[DIR_MMM] = 0.5;
+ EXPECT_THAT(actualQs, testing::Eq(expectedQs));
+}
+
+TEST_F(SideTestBC, setQs3D_givenMXhasBeenSet_thenSetMZ_doNotSetSameQsAgain)
+{
+ side.coordinateDirection = Z_INDEX;
+ side.sideDirection = NEGATIVE_DIR;
+ grid->addBCalreadySet(SideType::MX);
+
+ side.setQs(grid, bc, index);
+ auto actualQs = bc->getQs()[0];
+
+ std::vector<real> expectedQs(27, -1);
+ expectedQs[DIR_00M] = 0.5;
+ expectedQs[DIR_P0M] = 0.5;
+ expectedQs[DIR_0PM] = 0.5;
+ expectedQs[DIR_0MM] = 0.5;
+ expectedQs[DIR_PPM] = 0.5;
+ expectedQs[DIR_PMM] = 0.5;
+ EXPECT_THAT(actualQs, testing::Eq(expectedQs));
+}
+
+TEST_F(SideTestBC, setQs3D_givenMYandPXhaveBeenSet_thenSetMZ_doNotSetSameQsAgain)
+{
+ side.coordinateDirection = Z_INDEX;
+ side.sideDirection = NEGATIVE_DIR;
+ grid->addBCalreadySet(SideType::MY);
+ grid->addBCalreadySet(SideType::PX);
+
+ side.setQs(grid, bc, index);
+ auto actualQs = bc->getQs()[0];
+
+ std::vector<real> expectedQsForTwoPreviousBCs(27, -1);
+ expectedQsForTwoPreviousBCs[DIR_00M] = 0.5;
+ expectedQsForTwoPreviousBCs[DIR_M0M] = 0.5;
+ expectedQsForTwoPreviousBCs[DIR_0PM] = 0.5;
+ expectedQsForTwoPreviousBCs[DIR_MPM] = 0.5;
+ EXPECT_THAT(actualQs, testing::Eq(expectedQsForTwoPreviousBCs));
+}
+
+TEST_F(SideTestBC, setQs3D_givenMYandMXhaveBeenSet_thenSetMZ_doNotSetSameQsAgain)
+{
+ side.coordinateDirection = Z_INDEX;
+ side.sideDirection = NEGATIVE_DIR;
+ grid->addBCalreadySet(SideType::MY);
+ grid->addBCalreadySet(SideType::MX);
+
+ side.setQs(grid, bc, index);
+ auto actualQs = bc->getQs()[0];
+
+ std::vector<real> expectedQsForTwoPreviousBCs(27, -1);
+ expectedQsForTwoPreviousBCs[DIR_00M] = 0.5;
+ expectedQsForTwoPreviousBCs[DIR_P0M] = 0.5;
+ expectedQsForTwoPreviousBCs[DIR_0PM] = 0.5;
+ expectedQsForTwoPreviousBCs[DIR_PPM] = 0.5;
+ EXPECT_THAT(actualQs, testing::Eq(expectedQsForTwoPreviousBCs));
+}
+
+TEST_F(SideTestBC, setQs3D_givenPYandPXhaveBeenSet_thenSetMZ_doNotSetSameQsAgain)
+{
+ side.coordinateDirection = Z_INDEX;
+ side.sideDirection = NEGATIVE_DIR;
+ grid->addBCalreadySet(SideType::PY);
+ grid->addBCalreadySet(SideType::PX);
+
+ side.setQs(grid, bc, index);
+ auto actualQs = bc->getQs()[0];
+
+ std::vector<real> expectedQsForTwoPreviousBCs(27, -1);
+ expectedQsForTwoPreviousBCs[DIR_00M] = 0.5;
+ expectedQsForTwoPreviousBCs[DIR_M0M] = 0.5;
+ expectedQsForTwoPreviousBCs[DIR_0MM] = 0.5;
+ expectedQsForTwoPreviousBCs[DIR_MMM] = 0.5;
+ EXPECT_THAT(actualQs, testing::Eq(expectedQsForTwoPreviousBCs));
+}
+
+TEST_F(SideTestBC, setQs3D_givenPYandMXhaveBeenSet_thenSetMZ_doNotSetSameQsAgain)
+{
+ side.coordinateDirection = Z_INDEX;
+ side.sideDirection = NEGATIVE_DIR;
+ grid->addBCalreadySet(SideType::PY);
+ grid->addBCalreadySet(SideType::MX);
+
+ side.setQs(grid, bc, index);
+ auto actualQs = bc->getQs()[0];
+
+ std::vector<real> expectedQsForTwoPreviousBCs(27, -1);
+ expectedQsForTwoPreviousBCs[DIR_00M] = 0.5;
+ expectedQsForTwoPreviousBCs[DIR_P0M] = 0.5;
+ expectedQsForTwoPreviousBCs[DIR_0MM] = 0.5;
+ expectedQsForTwoPreviousBCs[DIR_PMM] = 0.5;
+ EXPECT_THAT(actualQs, testing::Eq(expectedQsForTwoPreviousBCs));
+}
+
+TEST_F(SideTestBC, setQs3D_whenSettingPZ_setAllQsNormalToBC)
+{
+ side.coordinateDirection = Z_INDEX;
+ side.sideDirection = POSITIVE_DIR;
+
+ side.setQs(grid, bc, index);
+ auto actualQs = bc->getQs()[0];
+
+ std::vector<real> expectedQs(27, -1);
+ expectedQs[DIR_00P] = 0.5;
+ expectedQs[DIR_P0P] = 0.5;
+ expectedQs[DIR_M0P] = 0.5;
+ expectedQs[DIR_0PP] = 0.5;
+ expectedQs[DIR_0MP] = 0.5;
+ expectedQs[DIR_PPP] = 0.5;
+ expectedQs[DIR_MPP] = 0.5;
+ expectedQs[DIR_PMP] = 0.5;
+ expectedQs[DIR_MMP] = 0.5;
+ EXPECT_THAT(actualQs, testing::Eq(expectedQs));
+}
+
+TEST_F(SideTestBC, setQs3D_givenMYhasBeenSet_thenSetPZ_doNotSetSameQsAgain)
+{
+ side.coordinateDirection = Z_INDEX;
+ side.sideDirection = POSITIVE_DIR;
+ grid->addBCalreadySet(SideType::MY);
+
+ side.setQs(grid, bc, index);
+ auto actualQs = bc->getQs()[0];
+
+ std::vector<real> expectedQs(27, -1);
+ expectedQs[DIR_00P] = 0.5;
+ expectedQs[DIR_P0P] = 0.5;
+ expectedQs[DIR_M0P] = 0.5;
+ expectedQs[DIR_0PP] = 0.5;
+ expectedQs[DIR_PPP] = 0.5;
+ expectedQs[DIR_MPP] = 0.5;
+ EXPECT_THAT(actualQs, testing::Eq(expectedQs));
+}
+
+TEST_F(SideTestBC, setQs3D_givenPYhasBeenSet_thenSetPZ_doNotSetSameQsAgain)
+{
+ side.coordinateDirection = Z_INDEX;
+ side.sideDirection = POSITIVE_DIR;
+ grid->addBCalreadySet(SideType::PY);
+
+ side.setQs(grid, bc, index);
+ auto actualQs = bc->getQs()[0];
+
+ std::vector<real> expectedQs(27, -1);
+ expectedQs[DIR_00P] = 0.5;
+ expectedQs[DIR_P0P] = 0.5;
+ expectedQs[DIR_M0P] = 0.5;
+ expectedQs[DIR_0MP] = 0.5;
+ expectedQs[DIR_PMP] = 0.5;
+ expectedQs[DIR_MMP] = 0.5;
+ EXPECT_THAT(actualQs, testing::Eq(expectedQs));
+}
+
+TEST_F(SideTestBC, setQs3D_givenPXhasBeenSet_thenSetPZ_doNotSetSameQsAgain)
+{
+ side.coordinateDirection = Z_INDEX;
+ side.sideDirection = POSITIVE_DIR;
+ grid->addBCalreadySet(SideType::PX);
+
+ side.setQs(grid, bc, index);
+ auto actualQs = bc->getQs()[0];
+
+ std::vector<real> expectedQs(27, -1);
+ expectedQs[DIR_00P] = 0.5;
+ expectedQs[DIR_M0P] = 0.5;
+ expectedQs[DIR_0PP] = 0.5;
+ expectedQs[DIR_0MP] = 0.5;
+ expectedQs[DIR_MPP] = 0.5;
+ expectedQs[DIR_MMP] = 0.5;
+ EXPECT_THAT(actualQs, testing::Eq(expectedQs));
+}
+
+TEST_F(SideTestBC, setQs3D_givenMXhasBeenSet_thenSetPZ_doNotSetSameQsAgain)
+{
+ side.coordinateDirection = Z_INDEX;
+ side.sideDirection = POSITIVE_DIR;
+ grid->addBCalreadySet(SideType::MX);
+
+ side.setQs(grid, bc, index);
+ auto actualQs = bc->getQs()[0];
+
+ std::vector<real> expectedQs(27, -1);
+ expectedQs[DIR_00P] = 0.5;
+ expectedQs[DIR_P0P] = 0.5;
+ expectedQs[DIR_0PP] = 0.5;
+ expectedQs[DIR_0MP] = 0.5;
+ expectedQs[DIR_PPP] = 0.5;
+ expectedQs[DIR_PMP] = 0.5;
+ EXPECT_THAT(actualQs, testing::Eq(expectedQs));
+}
+
+TEST_F(SideTestBC, setQs3D_givenMYandPXhaveBeenSet_thenSetPZ_doNotSetSameQsAgain)
+{
+ side.coordinateDirection = Z_INDEX;
+ side.sideDirection = POSITIVE_DIR;
+ grid->addBCalreadySet(SideType::MY);
+ grid->addBCalreadySet(SideType::PX);
+
+ side.setQs(grid, bc, index);
+ auto actualQs = bc->getQs()[0];
+
+ std::vector<real> expectedQsForTwoPreviousBCs(27, -1);
+ expectedQsForTwoPreviousBCs[DIR_00P] = 0.5;
+ expectedQsForTwoPreviousBCs[DIR_M0P] = 0.5;
+ expectedQsForTwoPreviousBCs[DIR_0PP] = 0.5;
+ expectedQsForTwoPreviousBCs[DIR_MPP] = 0.5;
+ EXPECT_THAT(actualQs, testing::Eq(expectedQsForTwoPreviousBCs));
+}
+
+TEST_F(SideTestBC, setQs3D_givenMYandMXhaveBeenSet_thenSetPZ_doNotSetSameQsAgain)
+{
+ side.coordinateDirection = Z_INDEX;
+ side.sideDirection = POSITIVE_DIR;
+ grid->addBCalreadySet(SideType::MY);
+ grid->addBCalreadySet(SideType::MX);
+
+ side.setQs(grid, bc, index);
+ auto actualQs = bc->getQs()[0];
+
+ std::vector<real> expectedQsForTwoPreviousBCs(27, -1);
+ expectedQsForTwoPreviousBCs[DIR_00P] = 0.5;
+ expectedQsForTwoPreviousBCs[DIR_P0P] = 0.5;
+ expectedQsForTwoPreviousBCs[DIR_0PP] = 0.5;
+ expectedQsForTwoPreviousBCs[DIR_PPP] = 0.5;
+ EXPECT_THAT(actualQs, testing::Eq(expectedQsForTwoPreviousBCs));
+}
+
+TEST_F(SideTestBC, setQs3D_givenPYandPXhaveBeenSet_thenSetPZ_doNotSetSameQsAgain)
+{
+ side.coordinateDirection = Z_INDEX;
+ side.sideDirection = POSITIVE_DIR;
+ grid->addBCalreadySet(SideType::PY);
+ grid->addBCalreadySet(SideType::PX);
+
+ side.setQs(grid, bc, index);
+ auto actualQs = bc->getQs()[0];
+
+ std::vector<real> expectedQsForTwoPreviousBCs(27, -1);
+ expectedQsForTwoPreviousBCs[DIR_00P] = 0.5;
+ expectedQsForTwoPreviousBCs[DIR_M0P] = 0.5;
+ expectedQsForTwoPreviousBCs[DIR_0MP] = 0.5;
+ expectedQsForTwoPreviousBCs[DIR_MMP] = 0.5;
+ EXPECT_THAT(actualQs, testing::Eq(expectedQsForTwoPreviousBCs));
+}
+
+TEST_F(SideTestBC, setQs3D_givenPYandMXhaveBeenSet_thenSetPZ_doNotSetSameQsAgain)
+{
+ side.coordinateDirection = Z_INDEX;
+ side.sideDirection = POSITIVE_DIR;
+ grid->addBCalreadySet(SideType::PY);
+ grid->addBCalreadySet(SideType::MX);
+
+ side.setQs(grid, bc, index);
+ auto actualQs = bc->getQs()[0];
+
+ std::vector<real> expectedQsForTwoPreviousBCs(27, -1);
+ expectedQsForTwoPreviousBCs[DIR_00P] = 0.5;
+ expectedQsForTwoPreviousBCs[DIR_P0P] = 0.5;
+ expectedQsForTwoPreviousBCs[DIR_0MP] = 0.5;
+ expectedQsForTwoPreviousBCs[DIR_PMP] = 0.5;
+ EXPECT_THAT(actualQs, testing::Eq(expectedQsForTwoPreviousBCs));
+}
diff --git a/src/gpu/GridGenerator/grid/Grid.h b/src/gpu/GridGenerator/grid/Grid.h
index 3f28120a5d969fcc5d7b2a3402a2169ff97c0cc3..ad2ce473fb65fe4414f6da5c4caf0d3e140b7e02 100644
--- a/src/gpu/GridGenerator/grid/Grid.h
+++ b/src/gpu/GridGenerator/grid/Grid.h
@@ -47,6 +47,7 @@ struct Triangle;
class GridInterface;
class Object;
class BoundingBox;
+enum class SideType;
class GRIDGENERATOR_EXPORT Grid
{
@@ -84,6 +85,8 @@ public:
virtual void getGridInterfaceIndices(uint* iCellCfc, uint* iCellCff, uint* iCellFcc, uint* iCellFcf) const = 0;
virtual bool isSparseIndexInFluidNodeIndicesBorder(uint &sparseIndex) const = 0;
+ virtual bool isStopperForBC(uint index) const = 0;
+
virtual int *getNeighborsX() const = 0;
virtual int *getNeighborsY() const = 0;
virtual int *getNeighborsZ() const = 0;
@@ -133,9 +136,9 @@ public:
virtual void setPeriodicityY(bool periodicity) = 0;
virtual void setPeriodicityZ(bool periodicity) = 0;
- virtual bool getPeriodicityX() = 0;
- virtual bool getPeriodicityY() = 0;
- virtual bool getPeriodicityZ() = 0;
+ virtual bool getPeriodicityX() const = 0;
+ virtual bool getPeriodicityY() const = 0;
+ virtual bool getPeriodicityZ() const = 0;
virtual void setEnableFixRefinementIntoTheWall(bool enableFixRefinementIntoTheWall) = 0;
@@ -170,6 +173,11 @@ public:
virtual void repairCommunicationIndices(int direction) = 0;
+ virtual bool nodeHasBC(uint index) const = 0;
+
+ virtual std::vector<SideType> getBCAlreadySet() = 0;
+ virtual void addBCalreadySet(SideType side) = 0;
+
// needed for CUDA Streams
virtual void findFluidNodeIndices(bool onlyBulk) = 0;
virtual uint getNumberOfFluidNodes() const = 0;
@@ -178,6 +186,20 @@ public:
virtual void findFluidNodeIndicesBorder() = 0;
virtual uint getNumberOfFluidNodesBorder() const = 0;
virtual void getFluidNodeIndicesBorder(uint *fluidNodeIndicesBorder) const = 0;
+
+ virtual void addFluidNodeIndicesMacroVars(std::vector<uint> _fluidNodeIndicesMacroVars) = 0;
+ virtual void addFluidNodeIndicesApplyBodyForce(std::vector<uint> _fluidNodeIndicesApplyBodyForce) = 0;
+ virtual void addFluidNodeIndicesAllFeatures(std::vector<uint> _fluidNodeIndicesAllFeatures) = 0;
+ virtual void sortFluidNodeIndicesMacroVars() = 0;
+ virtual void sortFluidNodeIndicesApplyBodyForce() = 0;
+ virtual void sortFluidNodeIndicesAllFeatures() = 0;
+
+ virtual uint getNumberOfFluidNodeIndicesMacroVars() const = 0;
+ virtual uint getNumberOfFluidNodeIndicesApplyBodyForce() const = 0;
+ virtual uint getNumberOfFluidNodeIndicesAllFeatures() const = 0;
+ virtual void getFluidNodeIndicesMacroVars(uint *fluidNodeIndicesMacroVars) const = 0;
+ virtual void getFluidNodeIndicesApplyBodyForce(uint *fluidNodeIndicesApplyBodyForce) const = 0;
+ virtual void getFluidNodeIndicesAllFeatures(uint *fluidNodeIndicesAllFeatures) const = 0;
};
#endif
diff --git a/src/gpu/GridGenerator/grid/GridBuilder/GridBuilder.h b/src/gpu/GridGenerator/grid/GridBuilder/GridBuilder.h
index 739aef59f76a33fa67d472a77ef258469f5e411c..f3d850384816f6690e5ffc158bbdc5e1df0ab328 100644
--- a/src/gpu/GridGenerator/grid/GridBuilder/GridBuilder.h
+++ b/src/gpu/GridGenerator/grid/GridBuilder/GridBuilder.h
@@ -54,6 +54,7 @@ class GridWrapper;
class Transformator;
class ArrowTransformator;
class PolyDataWriterWrapper;
+class TransientBCInputFileReader;
class BoundingBox;
class Grid;
@@ -113,6 +114,15 @@ public:
virtual void getPressureValues(real *rho, int *indices, int *neighborIndices, int level) const = 0;
virtual void getPressureQs(real *qs[27], int level) const = 0;
+ virtual uint getPrecursorSize(int level) const = 0;
+ virtual void getPrecursorValues(uint* neighbor0PP, uint* neighbor0PM, uint* neighbor0MP, uint* neighbor0MM,
+ real* weights0PP, real* weights0PM, real* weights0MP, real* weights0MM,
+ int* indices, std::vector<SPtr<TransientBCInputFileReader>>& reader,
+ int& numberOfPrecursorNodes, size_t& numberOfQuantities, uint& timeStepsBetweenReads,
+ real& velocityX, real& velocityY, real& velocityZ, int level) const = 0;
+
+ virtual void getPrecursorQs(real* qs[27], int level) const = 0;
+
virtual uint getGeometrySize(int level) const = 0;
virtual void getGeometryIndices(int *indices, int level) const = 0;
virtual void getGeometryQs(real *qs[27], int level) const = 0;
@@ -136,6 +146,21 @@ public:
virtual void getReceiveIndices(int *sendIndices, int direction, int level) = 0;
virtual void findFluidNodes(bool splitDomain) = 0;
+
+ virtual void addFluidNodeIndicesMacroVars(const std::vector<uint>& fluidNodeIndicesMacroVars, uint level) = 0;
+ virtual void addFluidNodeIndicesApplyBodyForce(const std::vector<uint>& fluidNodeIndicesApplyBodyForce, uint level) = 0;
+ virtual void addFluidNodeIndicesAllFeatures(const std::vector<uint>& fluidNodeIndicesAllFeatures, uint level) = 0;
+ virtual void sortFluidNodeIndicesMacroVars(uint level) = 0;
+ virtual void sortFluidNodeIndicesApplyBodyForce(uint level) = 0;
+ virtual void sortFluidNodeIndicesAllFeatures(uint level) = 0;
+ virtual uint getNumberOfFluidNodesMacroVars(uint level) const = 0;
+ virtual void getFluidNodeIndicesMacroVars(uint *fluidNodeIndicesMacroVars, int level) const = 0;
+ virtual uint getNumberOfFluidNodesApplyBodyForce(uint level) const = 0;
+ virtual void getFluidNodeIndicesApplyBodyForce(uint *fluidNodeIndicesApplyBodyForce, int level) const = 0;
+ virtual uint getNumberOfFluidNodesAllFeatures(uint level) const = 0;
+ virtual void getFluidNodeIndicesAllFeatures(uint *fluidNodeIndicesAllFeatures, int level) const = 0;
+
+
};
-#endif
\ No newline at end of file
+#endif
diff --git a/src/gpu/GridGenerator/grid/GridBuilder/LevelGridBuilder.cpp b/src/gpu/GridGenerator/grid/GridBuilder/LevelGridBuilder.cpp
index 083b9a51e0b151f49922df456e968c4b204e4af7..003e6dcd223d2bf019c83f71349a9a7bec84efdc 100644
--- a/src/gpu/GridGenerator/grid/GridBuilder/LevelGridBuilder.cpp
+++ b/src/gpu/GridGenerator/grid/GridBuilder/LevelGridBuilder.cpp
@@ -1,28 +1,28 @@
//=======================================================================================
-// ____ ____ __ ______ __________ __ __ __ __
-// \ \ | | | | | _ \ |___ ___| | | | | / \ | |
-// \ \ | | | | | |_) | | | | | | | / \ | |
-// \ \ | | | | | _ / | | | | | | / /\ \ | |
-// \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
-// \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
-// \ \ | | ________________________________________________________________
-// \ \ | | | ______________________________________________________________|
-// \ \| | | | __ __ __ __ ______ _______
-// \ | | |_____ | | | | | | | | | _ \ / _____)
-// \ | | _____| | | | | | | | | | | \ \ \_______
+// ____ ____ __ ______ __________ __ __ __ __
+// \ \ | | | | | _ \ |___ ___| | | | | / \ | |
+// \ \ | | | | | |_) | | | | | | | / \ | |
+// \ \ | | | | | _ / | | | | | | / /\ \ | |
+// \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
+// \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
+// \ \ | | ________________________________________________________________
+// \ \ | | | ______________________________________________________________|
+// \ \| | | | __ __ __ __ ______ _______
+// \ | | |_____ | | | | | | | | | _ \ / _____)
+// \ | | _____| | | | | | | | | | | \ \ \_______
// \ | | | | |_____ | \_/ | | | | |_/ / _____ |
-// \ _____| |__| |________| \_______/ |__| |______/ (_______/
+// \ _____| |__| |________| \_______/ |__| |______/ (_______/
//
-// This file is part of VirtualFluids. VirtualFluids is free software: you can
+// This file is part of VirtualFluids. VirtualFluids is free software: you can
// redistribute it and/or modify it under the terms of the GNU General Public
-// License as published by the Free Software Foundation, either version 3 of
+// License as published by the Free Software Foundation, either version 3 of
// the License, or (at your option) any later version.
-//
-// VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
-// ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
-// FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
+//
+// VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
+// ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+// FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
// for more details.
-//
+//
// You should have received a copy of the GNU General Public License along
// with VirtualFluids (see COPYING.txt). If not, see <http://www.gnu.org/licenses/>.
//
@@ -52,6 +52,8 @@
#include "io/QLineWriter.h"
#include "io/SimulationFileWriter/SimulationFileWriter.h"
+#include "TransientBCSetter/TransientBCSetter.h"
+
#include "utilities/communication.h"
#include "utilities/transformator/ArrowTransformator.h"
@@ -103,28 +105,33 @@ void LevelGridBuilder::setSlipGeometryBoundaryCondition(real normalX, real norma
for (uint level = 0; level < getNumberOfGridLevels(); level++)
{
- if (boundaryConditions[level]->geometryBoundaryCondition != nullptr)
- {
- boundaryConditions[level]->geometryBoundaryCondition->normalX = normalX;
- boundaryConditions[level]->geometryBoundaryCondition->normalY = normalY;
- boundaryConditions[level]->geometryBoundaryCondition->normalZ = normalZ;
- boundaryConditions[level]->geometryBoundaryCondition->side->addIndices(grids, level, boundaryConditions[level]->geometryBoundaryCondition);
+ if (boundaryConditions[level]->geometryBoundaryCondition != nullptr)
+ {
+ boundaryConditions[level]->geometryBoundaryCondition->normalX = normalX;
+ boundaryConditions[level]->geometryBoundaryCondition->normalY = normalY;
+ boundaryConditions[level]->geometryBoundaryCondition->normalZ = normalZ;
+ boundaryConditions[level]->geometryBoundaryCondition->side->addIndices(grids, level, boundaryConditions[level]->geometryBoundaryCondition);
boundaryConditions[level]->geometryBoundaryCondition->fillSlipNormalLists();
*logging::out << logging::Logger::INFO_INTERMEDIATE << "Set Geometry Slip BC on level " << level << " with " << (int)boundaryConditions[level]->geometryBoundaryCondition->indices.size() <<"\n";
- }
+ }
}
}
-void LevelGridBuilder::setStressBoundaryCondition( SideType sideType,
- real nomalX, real normalY, real normalZ,
- uint samplingOffset, real z0)
+//=======================================================================================
+//! \brief Set stress boundary concdition using iMEM
+//! \param samplingOffset number of grid points above boundary where velocity for wall model is sampled
+//! \param z0 roughness length [m]
+//! \param dx dx of level 0 [m]
+//!
+void LevelGridBuilder::setStressBoundaryCondition( SideType sideType,
+ real nomalX, real normalY, real normalZ,
+ uint samplingOffset, real z0, real dx)
{
for (uint level = 0; level < getNumberOfGridLevels(); level++)
{
- SPtr<StressBoundaryCondition> stressBoundaryCondition = StressBoundaryCondition::make(nomalX, normalY, normalZ, samplingOffset, z0);
-
+ SPtr<StressBoundaryCondition> stressBoundaryCondition = StressBoundaryCondition::make(nomalX, normalY, normalZ, samplingOffset, z0*pow(2.0f,level)/dx);
auto side = SideFactory::make(sideType);
stressBoundaryCondition->side = side;
@@ -171,17 +178,17 @@ void LevelGridBuilder::setVelocityGeometryBoundaryCondition(real vx, real vy, re
for (uint level = 0; level < getNumberOfGridLevels(); level++)
{
- if (boundaryConditions[level]->geometryBoundaryCondition != nullptr)
- {
- boundaryConditions[level]->geometryBoundaryCondition->vx = vx;
- boundaryConditions[level]->geometryBoundaryCondition->vy = vy;
- boundaryConditions[level]->geometryBoundaryCondition->vz = vz;
- boundaryConditions[level]->geometryBoundaryCondition->side->addIndices(grids, level, boundaryConditions[level]->geometryBoundaryCondition);
+ if (boundaryConditions[level]->geometryBoundaryCondition != nullptr)
+ {
+ boundaryConditions[level]->geometryBoundaryCondition->vx = vx;
+ boundaryConditions[level]->geometryBoundaryCondition->vy = vy;
+ boundaryConditions[level]->geometryBoundaryCondition->vz = vz;
+ boundaryConditions[level]->geometryBoundaryCondition->side->addIndices(grids, level, boundaryConditions[level]->geometryBoundaryCondition);
boundaryConditions[level]->geometryBoundaryCondition->fillVelocityLists();
*logging::out << logging::Logger::INFO_INTERMEDIATE << "Set Geometry Velocity BC on level " << level << " with " << (int)boundaryConditions[level]->geometryBoundaryCondition->indices.size() <<"\n";
- }
+ }
}
}
@@ -223,7 +230,7 @@ void LevelGridBuilder::setNoSlipBoundaryCondition(SideType sideType)
noSlipBoundaryCondition->fillVelocityLists();
// now effectively just a wrapper for velocityBC with zero velocity. No distinction in Gridgenerator.
- boundaryConditions[level]->velocityBoundaryConditions.push_back(noSlipBoundaryCondition);
+ boundaryConditions[level]->velocityBoundaryConditions.push_back(noSlipBoundaryCondition);
}
}
}
@@ -234,12 +241,45 @@ void LevelGridBuilder::setNoSlipGeometryBoundaryCondition()
for (uint level = 0; level < getNumberOfGridLevels(); level++)
{
- if (boundaryConditions[level]->geometryBoundaryCondition != nullptr)
- {
- boundaryConditions[level]->geometryBoundaryCondition->side->addIndices(grids, level, boundaryConditions[level]->geometryBoundaryCondition);
+ if (boundaryConditions[level]->geometryBoundaryCondition != nullptr)
+ {
+ boundaryConditions[level]->geometryBoundaryCondition->side->addIndices(grids, level, boundaryConditions[level]->geometryBoundaryCondition);
*logging::out << logging::Logger::INFO_INTERMEDIATE << "Set Geometry No-Slip BC on level " << level << " with " << (int)boundaryConditions[level]->geometryBoundaryCondition->indices.size() <<"\n";
- }
+ }
+ }
+}
+
+void LevelGridBuilder::setPrecursorBoundaryCondition(SideType sideType, SPtr<FileCollection> fileCollection, int timeStepsBetweenReads,
+ real velocityX, real velocityY, real velocityZ, std::vector<uint> fileLevelToGridLevelMap)
+{
+ if(fileLevelToGridLevelMap.empty())
+ {
+ *logging::out << logging::Logger::INFO_INTERMEDIATE << "Mapping precursor file levels to the corresponding grid levels" << "\n";
+
+ for (uint level = 0; level < getNumberOfGridLevels(); level++)
+ fileLevelToGridLevelMap.push_back(level);
+ }
+ else
+ {
+ if(fileLevelToGridLevelMap.size()!=getNumberOfGridLevels())
+ throw std::runtime_error("In setPrecursorBoundaryCondition: fileLevelToGridLevelMap does not match with the number of levels");
+ *logging::out << logging::Logger::INFO_INTERMEDIATE << "Using user defined file to grid level mapping" << "\n";
+ }
+
+ for (uint level = 0; level < getNumberOfGridLevels(); level++)
+ {
+ auto reader = createReaderForCollection(fileCollection, fileLevelToGridLevelMap[level]);
+ SPtr<PrecursorBoundaryCondition> precursorBoundaryCondition = PrecursorBoundaryCondition::make( reader, timeStepsBetweenReads, velocityX, velocityY, velocityZ);
+
+ auto side = SideFactory::make(sideType);
+
+ precursorBoundaryCondition->side = side;
+ precursorBoundaryCondition->side->addIndices(grids, level, precursorBoundaryCondition);
+
+ boundaryConditions[level]->precursorBoundaryConditions.push_back(precursorBoundaryCondition);
+
+ *logging::out << logging::Logger::INFO_INTERMEDIATE << "Set Precursor BC on level " << level << " with " << (int)precursorBoundaryCondition->indices.size() << "\n";
}
}
@@ -373,9 +413,9 @@ std::shared_ptr<Grid> LevelGridBuilder::getGrid(int level, int box)
void LevelGridBuilder::checkLevel(int level)
{
if (level >= (int)grids.size())
- {
+ {
std::cout << "wrong level input... return to caller\n";
- return;
+ return;
}
}
@@ -386,16 +426,16 @@ void LevelGridBuilder::getDimensions(int &nx, int &ny, int &nz, const int level)
nz = grids[level]->getNumberOfNodesZ();
}
-void LevelGridBuilder::getNodeValues(real *xCoords, real *yCoords, real *zCoords,
- uint *neighborX, uint *neighborY, uint *neighborZ, uint *neighborNegative,
+void LevelGridBuilder::getNodeValues(real *xCoords, real *yCoords, real *zCoords,
+ uint *neighborX, uint *neighborY, uint *neighborZ, uint *neighborNegative,
uint *geo, const int level) const
{
grids[level]->getNodeValues(xCoords, yCoords, zCoords, neighborX, neighborY, neighborZ, neighborNegative, geo);
}
-GRIDGENERATOR_EXPORT void LevelGridBuilder::getFluidNodeIndices(uint *fluidNodeIndices, const int level) const
-{
+GRIDGENERATOR_EXPORT void LevelGridBuilder::getFluidNodeIndices(uint *fluidNodeIndices, const int level) const
+{
grids[level]->getFluidNodeIndices(fluidNodeIndices);
}
@@ -404,9 +444,9 @@ GRIDGENERATOR_EXPORT void LevelGridBuilder::getFluidNodeIndicesBorder(uint *flui
grids[level]->getFluidNodeIndicesBorder(fluidNodeIndices);
}
-uint LevelGridBuilder::getNumberOfFluidNodes(unsigned int level) const
+uint LevelGridBuilder::getNumberOfFluidNodes(unsigned int level) const
{
- return grids[level]->getNumberOfFluidNodes();
+ return grids[level]->getNumberOfFluidNodes();
}
GRIDGENERATOR_EXPORT uint LevelGridBuilder::getNumberOfFluidNodesBorder(unsigned int level) const
@@ -432,7 +472,7 @@ void LevelGridBuilder::getSlipValues(real* normalX, real* normalY, real* normalZ
for (uint index = 0; index < boundaryCondition->indices.size(); index++)
{
indices[allIndicesCounter] = grids[level]->getSparseIndex(boundaryCondition->indices[index]) + 1;
-
+
normalX[allIndicesCounter] = boundaryCondition->getNormalx(index);
normalY[allIndicesCounter] = boundaryCondition->getNormaly(index);
normalZ[allIndicesCounter] = boundaryCondition->getNormalz(index);
@@ -467,9 +507,9 @@ uint LevelGridBuilder::getStressSize(int level) const
return size;
}
-void LevelGridBuilder::getStressValues( real* normalX, real* normalY, real* normalZ,
- real* vx, real* vy, real* vz,
- real* vx1, real* vy1, real* vz1,
+void LevelGridBuilder::getStressValues( real* normalX, real* normalY, real* normalZ,
+ real* vx, real* vy, real* vz,
+ real* vx1, real* vy1, real* vz1,
int* indices, int* samplingIndices, int* samplingOffset, real* z0, int level) const
{
@@ -525,7 +565,7 @@ void LevelGridBuilder::getVelocityValues(real* vx, real* vy, real* vz, int* indi
{
for (uint i = 0; i < (uint)boundaryCondition->indices.size(); i++)
{
- indices[allIndicesCounter] = grids[level]->getSparseIndex(boundaryCondition->indices[i]) +1;
+ indices[allIndicesCounter] = grids[level]->getSparseIndex(boundaryCondition->indices[i]) +1;
vx[allIndicesCounter] = boundaryCondition->getVx(i);
vy[allIndicesCounter] = boundaryCondition->getVy(i);
@@ -594,11 +634,91 @@ void LevelGridBuilder::getPressureQs(real* qs[27], int level) const
}
}
+uint LevelGridBuilder::getPrecursorSize(int level) const
+{
+ uint size = 0;
+ for (auto boundaryCondition : boundaryConditions[level]->precursorBoundaryConditions)
+ {
+ size += uint(boundaryCondition->indices.size());
+ }
+ return size;
+}
+
+void LevelGridBuilder::getPrecursorValues( uint* neighbor0PP, uint* neighbor0PM, uint* neighbor0MP, uint* neighbor0MM,
+ real* weights0PP, real* weights0PM, real* weights0MP, real* weights0MM,
+ int* indices, std::vector<SPtr<TransientBCInputFileReader>>& reader,
+ int& numberOfPrecursorNodes, size_t& numberOfQuantities, uint& timeStepsBetweenReads,
+ real& velocityX, real& velocityY, real& velocityZ, int level) const
+{
+ int allIndicesCounter = 0;
+ int allNodesCounter = 0;
+ uint tmpTimeStepsBetweenReads = 0;
+ size_t tmpNumberOfQuantities = 0;
+
+ for (auto boundaryCondition : boundaryConditions[level]->precursorBoundaryConditions)
+ {
+ if( tmpTimeStepsBetweenReads == 0 )
+ tmpTimeStepsBetweenReads = boundaryCondition->timeStepsBetweenReads;
+ if( tmpTimeStepsBetweenReads != boundaryCondition->timeStepsBetweenReads )
+ throw std::runtime_error("All precursor boundary conditions must have the same timeStepsBetweenReads value");
+ auto BCreader = boundaryCondition->getReader();
+ BCreader->setWritingOffset(allIndicesCounter);
+ reader.push_back(BCreader);
+
+ std::vector<real> y, z;
+ real xTmp, yTmp, zTmp;
+ for(uint i = 0; i<boundaryCondition->indices.size(); i++)
+ {
+ indices[allIndicesCounter] = grids[level]->getSparseIndex(boundaryCondition->indices[i]) + 1;
+ grids[level]->transIndexToCoords(boundaryCondition->indices[i], xTmp, yTmp, zTmp);
+ y.push_back(yTmp);
+ z.push_back(zTmp);
+ allIndicesCounter++;
+ }
+ BCreader->fillArrays(y, z);
+ BCreader->getNeighbors(neighbor0PP, neighbor0PM, neighbor0MP, neighbor0MM);
+ BCreader->getWeights(weights0PP, weights0PM, weights0MP, weights0MM);
+ if(tmpNumberOfQuantities == 0)
+ tmpNumberOfQuantities = BCreader->getNumberOfQuantities();
+ if(tmpNumberOfQuantities != BCreader->getNumberOfQuantities())
+ throw std::runtime_error("All precursor files must have the same quantities.");
+ allNodesCounter += BCreader->getNPointsRead();
+ velocityX = boundaryCondition->getVelocityX();
+ velocityY = boundaryCondition->getVelocityY();
+ velocityZ = boundaryCondition->getVelocityZ();
+ }
+ numberOfPrecursorNodes = allNodesCounter;
+
+ if (tmpTimeStepsBetweenReads == 0)
+ throw std::runtime_error("timeStepsBetweenReads of precursor needs to be larger than 0.");
+ timeStepsBetweenReads = tmpTimeStepsBetweenReads;
+
+ if (tmpNumberOfQuantities == 0)
+ throw std::runtime_error("Number of quantities in precursor needs to be larger than 0.");
+ numberOfQuantities = tmpNumberOfQuantities;
+}
+
+void LevelGridBuilder::getPrecursorQs(real* qs[27], int level) const
+{
+ int allIndicesCounter = 0;
+ for (auto boundaryCondition : boundaryConditions[level]->precursorBoundaryConditions)
+ {
+ for ( uint index = 0; index < boundaryCondition->indices.size(); index++ )
+ {
+ for (int dir = 0; dir <= grids[level]->getEndDirection(); dir++)
+ {
+ qs[dir][allIndicesCounter] = boundaryCondition->qs[index][dir];
+ }
+ allIndicesCounter++;
+ }
+ }
+}
+
uint LevelGridBuilder::getGeometrySize(int level) const
{
if (boundaryConditions[level]->geometryBoundaryCondition)
return (uint)boundaryConditions[level]->geometryBoundaryCondition->indices.size();
-
+
return 0;
}
@@ -619,9 +739,9 @@ void LevelGridBuilder::getGeometryValues(real* vx, real* vy, real* vz, int level
{
for (uint i = 0; i < boundaryConditions[level]->geometryBoundaryCondition->indices.size(); i++)
{
- vx[i] = boundaryConditions[level]->geometryBoundaryCondition->getVx(i);
- vy[i] = boundaryConditions[level]->geometryBoundaryCondition->getVy(i);
- vz[i] = boundaryConditions[level]->geometryBoundaryCondition->getVz(i);
+ vx[i] = boundaryConditions[level]->geometryBoundaryCondition->getVx(i);
+ vy[i] = boundaryConditions[level]->geometryBoundaryCondition->getVy(i);
+ vz[i] = boundaryConditions[level]->geometryBoundaryCondition->getVz(i);
}
}
@@ -636,7 +756,7 @@ void LevelGridBuilder::getGeometryQs(real* qs[27], int level) const
}
}
-void LevelGridBuilder::writeArrows(std::string fileName) const
+void LevelGridBuilder::writeArrows(std::string fileName) const
{
QLineWriter::writeArrows(fileName, boundaryConditions[getNumberOfGridLevels() - 1]->geometryBoundaryCondition, grids[getNumberOfGridLevels() - 1]);
}
@@ -674,4 +794,65 @@ void LevelGridBuilder::findFluidNodes(bool splitDomain)
for (uint i = 0; i < grids.size(); i++)
grids[i]->findFluidNodeIndices(splitDomain);
*logging::out << logging::Logger::INFO_HIGH << "Done with findFluidNodes()\n";
-}
\ No newline at end of file
+}
+
+
+void LevelGridBuilder::addFluidNodeIndicesMacroVars(const std::vector<uint>& fluidNodeIndicesMacroVars, uint level)
+{
+ grids[level]->addFluidNodeIndicesMacroVars(fluidNodeIndicesMacroVars);
+}
+
+void LevelGridBuilder::addFluidNodeIndicesApplyBodyForce(const std::vector<uint>& fluidNodeIndicesApplyBodyForce, uint level)
+{
+ grids[level]->addFluidNodeIndicesApplyBodyForce(fluidNodeIndicesApplyBodyForce);
+}
+
+void LevelGridBuilder::addFluidNodeIndicesAllFeatures(const std::vector<uint>& fluidNodeIndicesAllFeatures, uint level)
+{
+ grids[level]->addFluidNodeIndicesAllFeatures(fluidNodeIndicesAllFeatures);
+}
+
+void LevelGridBuilder::sortFluidNodeIndicesMacroVars(uint level)
+{
+ grids[level]->sortFluidNodeIndicesMacroVars();
+}
+
+void LevelGridBuilder::sortFluidNodeIndicesApplyBodyForce(uint level)
+{
+ grids[level]->sortFluidNodeIndicesApplyBodyForce();
+}
+
+void LevelGridBuilder::sortFluidNodeIndicesAllFeatures(uint level)
+{
+ grids[level]->sortFluidNodeIndicesAllFeatures();
+}
+
+uint LevelGridBuilder::getNumberOfFluidNodesMacroVars(unsigned int level) const
+{
+ return grids[level]->getNumberOfFluidNodeIndicesMacroVars();
+}
+
+void LevelGridBuilder::getFluidNodeIndicesMacroVars(uint *fluidNodeIndicesMacroVars, const int level) const
+{
+ grids[level]->getFluidNodeIndicesMacroVars(fluidNodeIndicesMacroVars);
+}
+
+uint LevelGridBuilder::getNumberOfFluidNodesApplyBodyForce(unsigned int level) const
+{
+ return grids[level]->getNumberOfFluidNodeIndicesApplyBodyForce();
+}
+
+void LevelGridBuilder::getFluidNodeIndicesApplyBodyForce(uint *fluidNodeIndicesApplyBodyForce, const int level) const
+{
+ grids[level]->getFluidNodeIndicesApplyBodyForce(fluidNodeIndicesApplyBodyForce);
+}
+
+uint LevelGridBuilder::getNumberOfFluidNodesAllFeatures(unsigned int level) const
+{
+ return grids[level]->getNumberOfFluidNodeIndicesAllFeatures();
+}
+
+void LevelGridBuilder::getFluidNodeIndicesAllFeatures(uint *fluidNodeIndicesAllFeatures, const int level) const
+{
+ grids[level]->getFluidNodeIndicesAllFeatures(fluidNodeIndicesAllFeatures);
+}
diff --git a/src/gpu/GridGenerator/grid/GridBuilder/LevelGridBuilder.h b/src/gpu/GridGenerator/grid/GridBuilder/LevelGridBuilder.h
index afb027fc1665ab874523bf39ec2a05518d28f7a1..2e0eaf13080c46260de2a0c845fbf784a2cc3e09 100644
--- a/src/gpu/GridGenerator/grid/GridBuilder/LevelGridBuilder.h
+++ b/src/gpu/GridGenerator/grid/GridBuilder/LevelGridBuilder.h
@@ -1,28 +1,28 @@
//=======================================================================================
-// ____ ____ __ ______ __________ __ __ __ __
-// \ \ | | | | | _ \ |___ ___| | | | | / \ | |
-// \ \ | | | | | |_) | | | | | | | / \ | |
-// \ \ | | | | | _ / | | | | | | / /\ \ | |
-// \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
-// \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
-// \ \ | | ________________________________________________________________
-// \ \ | | | ______________________________________________________________|
-// \ \| | | | __ __ __ __ ______ _______
-// \ | | |_____ | | | | | | | | | _ \ / _____)
-// \ | | _____| | | | | | | | | | | \ \ \_______
+// ____ ____ __ ______ __________ __ __ __ __
+// \ \ | | | | | _ \ |___ ___| | | | | / \ | |
+// \ \ | | | | | |_) | | | | | | | / \ | |
+// \ \ | | | | | _ / | | | | | | / /\ \ | |
+// \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
+// \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
+// \ \ | | ________________________________________________________________
+// \ \ | | | ______________________________________________________________|
+// \ \| | | | __ __ __ __ ______ _______
+// \ | | |_____ | | | | | | | | | _ \ / _____)
+// \ | | _____| | | | | | | | | | | \ \ \_______
// \ | | | | |_____ | \_/ | | | | |_/ / _____ |
-// \ _____| |__| |________| \_______/ |__| |______/ (_______/
+// \ _____| |__| |________| \_______/ |__| |______/ (_______/
//
-// This file is part of VirtualFluids. VirtualFluids is free software: you can
+// This file is part of VirtualFluids. VirtualFluids is free software: you can
// redistribute it and/or modify it under the terms of the GNU General Public
-// License as published by the Free Software Foundation, either version 3 of
+// License as published by the Free Software Foundation, either version 3 of
// the License, or (at your option) any later version.
-//
-// VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
-// ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
-// FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
+//
+// VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
+// ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+// FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
// for more details.
-//
+//
// You should have received a copy of the GNU General Public License along
// with VirtualFluids (see COPYING.txt). If not, see <http://www.gnu.org/licenses/>.
//
@@ -38,6 +38,8 @@
#include <memory>
#include <array>
+#include <lbm/constants/NumericConstants.h>
+
#include "gpu/GridGenerator/global.h"
#include "gpu/GridGenerator/grid/GridBuilder/GridBuilder.h"
@@ -45,6 +47,8 @@
#include "gpu/GridGenerator/grid/GridInterface.h"
#include "gpu/GridGenerator/grid/NodeValues.h"
+using namespace vf::lbm::constant;
+
struct Vertex;
class Grid;
class Transformator;
@@ -58,9 +62,11 @@ class SlipBoundaryCondition;
class StressBoundaryCondition;
class PressureBoundaryCondition;
class GeometryBoundaryCondition;
+class PrecursorBoundaryCondition;
enum class SideType;
-
+class TransientBCInputFileReader;
+class FileCollection;
class LevelGridBuilder : public GridBuilder
{
@@ -75,11 +81,14 @@ public:
GRIDGENERATOR_EXPORT ~LevelGridBuilder() override;
GRIDGENERATOR_EXPORT void setSlipBoundaryCondition(SideType sideType, real nomalX, real normalY, real normalZ);
- GRIDGENERATOR_EXPORT void setStressBoundaryCondition(SideType sideType, real nomalX, real normalY, real normalZ, uint samplingOffset, real z0);
+ GRIDGENERATOR_EXPORT void setStressBoundaryCondition(SideType sideType, real nomalX, real normalY, real normalZ, uint samplingOffset, real z0, real dx);
GRIDGENERATOR_EXPORT void setVelocityBoundaryCondition(SideType sideType, real vx, real vy, real vz);
GRIDGENERATOR_EXPORT void setPressureBoundaryCondition(SideType sideType, real rho);
GRIDGENERATOR_EXPORT void setPeriodicBoundaryCondition(bool periodic_X, bool periodic_Y, bool periodic_Z);
GRIDGENERATOR_EXPORT void setNoSlipBoundaryCondition(SideType sideType);
+ GRIDGENERATOR_EXPORT void setPrecursorBoundaryCondition(SideType sideType, SPtr<FileCollection> fileCollection, int timeStepsBetweenReads,
+ real velocityX=c0o1, real velocityY=c0o1, real velocityZ=c0o1,
+ std::vector<uint> fileLevelToGridLevelMap = {});
GRIDGENERATOR_EXPORT void setEnableFixRefinementIntoTheWall(bool enableFixRefinementIntoTheWall);
@@ -97,7 +106,7 @@ public:
GRIDGENERATOR_EXPORT virtual void getFluidNodeIndicesBorder(uint *fluidNodeIndices, const int level) const override;
GRIDGENERATOR_EXPORT virtual void getNodeValues(real *xCoords, real *yCoords, real *zCoords,
- uint *neighborX, uint *neighborY, uint *neighborZ, uint *neighborNegative,
+ uint *neighborX, uint *neighborY, uint *neighborZ, uint *neighborNegative,
uint *geo, const int level) const override;
GRIDGENERATOR_EXPORT virtual void getDimensions(int &nx, int &ny, int &nz, const int level) const override;
@@ -107,12 +116,12 @@ public:
GRIDGENERATOR_EXPORT virtual void getSlipQs(real* qs[27], int level) const override;
GRIDGENERATOR_EXPORT uint getStressSize(int level) const override;
- GRIDGENERATOR_EXPORT virtual void getStressValues( real* normalX, real* normalY, real* normalZ,
- real* vx, real* vy, real* vz,
- real* vx1, real* vy1, real* vz1,
+ GRIDGENERATOR_EXPORT virtual void getStressValues( real* normalX, real* normalY, real* normalZ,
+ real* vx, real* vy, real* vz,
+ real* vx1, real* vy1, real* vz1,
int* indices, int* samplingIndices, int* samplingOffsets, real* z0, int level) const override;
GRIDGENERATOR_EXPORT virtual void getStressQs(real* qs[27], int level) const override;
-
+
GRIDGENERATOR_EXPORT uint getVelocitySize(int level) const override;
GRIDGENERATOR_EXPORT virtual void getVelocityValues(real* vx, real* vy, real* vz, int* indices, int level) const override;
GRIDGENERATOR_EXPORT virtual void getVelocityQs(real* qs[27], int level) const override;
@@ -121,6 +130,14 @@ public:
GRIDGENERATOR_EXPORT void getPressureValues(real* rho, int* indices, int* neighborIndices, int level) const override;
GRIDGENERATOR_EXPORT virtual void getPressureQs(real* qs[27], int level) const override;
+ GRIDGENERATOR_EXPORT uint getPrecursorSize(int level) const override;
+ GRIDGENERATOR_EXPORT void getPrecursorValues( uint* neighbor0PP, uint* neighbor0PM, uint* neighbor0MP, uint* neighbor0MM,
+ real* weights0PP, real* weights0PM, real* weights0MP, real* weights0MM,
+ int* indices, std::vector<SPtr<TransientBCInputFileReader>>& reader,
+ int& numberOfPrecursorNodes, size_t& numberOfQuantities, uint& timeStepsBetweenReads,
+ real& velocityX, real& velocityY, real& velocityZ, int level) const override;
+ GRIDGENERATOR_EXPORT virtual void getPrecursorQs(real* qs[27], int level) const override;
+
GRIDGENERATOR_EXPORT virtual void getGeometryQs(real *qs[27], int level) const override;
GRIDGENERATOR_EXPORT virtual uint getGeometrySize(int level) const override;
GRIDGENERATOR_EXPORT virtual void getGeometryIndices(int *indices, int level) const override;
@@ -133,11 +150,11 @@ public:
GRIDGENERATOR_EXPORT SPtr<GeometryBoundaryCondition> getGeometryBoundaryCondition(uint level) const override;
protected:
-
+
struct BoundaryConditions
{
- BoundaryConditions() = default;
+ BoundaryConditions() = default;
std::vector<SPtr<SlipBoundaryCondition>> slipBoundaryConditions;
@@ -149,13 +166,15 @@ protected:
std::vector<SPtr<VelocityBoundaryCondition>> noSlipBoundaryConditions;
+ std::vector<SPtr<PrecursorBoundaryCondition>> precursorBoundaryConditions;
+
SPtr<GeometryBoundaryCondition> geometryBoundaryCondition;
};
bool geometryHasValues = false;
std::vector<std::shared_ptr<Grid> > grids;
std::vector<SPtr<BoundaryConditions> > boundaryConditions;
-
+
std::array<uint, 6> communicationProcesses;
void checkLevel(int level);
@@ -194,7 +213,21 @@ public:
// needed for CUDA Streams MultiGPU (Communication Hiding)
void findFluidNodes(bool splitDomain) override;
+
+ void addFluidNodeIndicesMacroVars(const std::vector<uint>& fluidNodeIndicesMacroVars, uint level) override;
+ void addFluidNodeIndicesApplyBodyForce(const std::vector<uint>& fluidNodeIndicesApplyBodyForce, uint level) override;
+ void addFluidNodeIndicesAllFeatures(const std::vector<uint>& fluidNodeIndicesAllFeatures, uint level) override;
+
+ void sortFluidNodeIndicesMacroVars(uint level) override;
+ void sortFluidNodeIndicesApplyBodyForce(uint level) override;
+ void sortFluidNodeIndicesAllFeatures(uint level) override;
+
+ uint getNumberOfFluidNodesMacroVars(unsigned int level) const override;
+ void getFluidNodeIndicesMacroVars(uint *fluidNodeIndicesMacroVars, const int level) const override;
+ uint getNumberOfFluidNodesApplyBodyForce(unsigned int level) const override;
+ void getFluidNodeIndicesApplyBodyForce(uint *fluidNodeIndicesApplyBodyForce, const int level) const override;
+ uint getNumberOfFluidNodesAllFeatures(unsigned int level) const override;
+ void getFluidNodeIndicesAllFeatures(uint *fluidNodeIndicesAllFeatures, const int level) const override;
};
#endif
-
diff --git a/src/gpu/GridGenerator/grid/GridImp.cpp b/src/gpu/GridGenerator/grid/GridImp.cpp
index 31bbf3ddc87184846fcb01a3e6631358b6a6f864..32cf9d07da87149695a5bf548ed357be2b2f71b4 100644
--- a/src/gpu/GridGenerator/grid/GridImp.cpp
+++ b/src/gpu/GridGenerator/grid/GridImp.cpp
@@ -1,28 +1,28 @@
//=======================================================================================
-// ____ ____ __ ______ __________ __ __ __ __
-// \ \ | | | | | _ \ |___ ___| | | | | / \ | |
-// \ \ | | | | | |_) | | | | | | | / \ | |
-// \ \ | | | | | _ / | | | | | | / /\ \ | |
-// \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
-// \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
-// \ \ | | ________________________________________________________________
-// \ \ | | | ______________________________________________________________|
-// \ \| | | | __ __ __ __ ______ _______
-// \ | | |_____ | | | | | | | | | _ \ / _____)
-// \ | | _____| | | | | | | | | | | \ \ \_______
+// ____ ____ __ ______ __________ __ __ __ __
+// \ \ | | | | | _ \ |___ ___| | | | | / \ | |
+// \ \ | | | | | |_) | | | | | | | / \ | |
+// \ \ | | | | | _ / | | | | | | / /\ \ | |
+// \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
+// \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
+// \ \ | | ________________________________________________________________
+// \ \ | | | ______________________________________________________________|
+// \ \| | | | __ __ __ __ ______ _______
+// \ | | |_____ | | | | | | | | | _ \ / _____)
+// \ | | _____| | | | | | | | | | | \ \ \_______
// \ | | | | |_____ | \_/ | | | | |_/ / _____ |
-// \ _____| |__| |________| \_______/ |__| |______/ (_______/
+// \ _____| |__| |________| \_______/ |__| |______/ (_______/
//
-// This file is part of VirtualFluids. VirtualFluids is free software: you can
+// This file is part of VirtualFluids. VirtualFluids is free software: you can
// redistribute it and/or modify it under the terms of the GNU General Public
-// License as published by the Free Software Foundation, either version 3 of
+// License as published by the Free Software Foundation, either version 3 of
// the License, or (at your option) any later version.
-//
-// VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
-// ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
-// FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
+//
+// VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
+// ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+// FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
// for more details.
-//
+//
// You should have received a copy of the GNU General Public License along
// with VirtualFluids (see COPYING.txt). If not, see <http://www.gnu.org/licenses/>.
//
@@ -33,7 +33,6 @@
#include "GridImp.h"
#include <iostream>
-#include <omp.h>
#include <sstream>
# include <algorithm>
#include <cmath>
@@ -61,8 +60,8 @@ int DIRECTIONS[DIR_END_MAX][DIMENSION];
using namespace vf::gpu;
-GridImp::GridImp(Object* object, real startX, real startY, real startZ, real endX, real endY, real endZ, real delta, Distribution distribution, uint level)
- : object(object),
+GridImp::GridImp(Object* object, real startX, real startY, real startZ, real endX, real endY, real endZ, real delta, Distribution distribution, uint level)
+ : object(object),
startX(startX),
startY(startY),
startZ(startZ),
@@ -135,7 +134,7 @@ void GridImp::inital(const SPtr<Grid> fineGrid, uint numberOfLayers)
#pragma omp parallel for
for (int index = 0; index < (int)this->size; index++)
this->initalNodeToOutOfGrid(index);
-
+
if( this->innerRegionFromFinerGrid ){
*logging::out << logging::Logger::INFO_INTERMEDIATE << "Start setInnerBasedOnFinerGrid()\n";
this->setInnerBasedOnFinerGrid(fineGrid);
@@ -147,12 +146,12 @@ void GridImp::inital(const SPtr<Grid> fineGrid, uint numberOfLayers)
*logging::out << logging::Logger::INFO_INTERMEDIATE << "Start addOverlap()\n";
this->addOverlap();
-
+
*logging::out << logging::Logger::INFO_INTERMEDIATE << "Start fixOddCells()\n";
#pragma omp parallel for
for (int index = 0; index < (int)this->size; index++)
this->fixOddCell(index);
-
+
if( enableFixRefinementIntoTheWall )
{
*logging::out << logging::Logger::INFO_INTERMEDIATE << "Start fixRefinementIntoWall()\n";
@@ -180,12 +179,12 @@ void GridImp::inital(const SPtr<Grid> fineGrid, uint numberOfLayers)
}
}
}
-
+
*logging::out << logging::Logger::INFO_INTERMEDIATE << "Start findEndOfGridStopperNodes()\n";
#pragma omp parallel for
for (int index = 0; index < (int)this->size; index++)
this->findEndOfGridStopperNode(index);
-
+
*logging::out << logging::Logger::INFO_INTERMEDIATE
<< "Grid created: " << "from (" << this->startX << ", " << this->startY << ", " << this->startZ << ") to (" << this->endX << ", " << this->endY << ", " << this->endZ << ")\n"
<< "nodes: " << this->nx << " x " << this->ny << " x " << this->nz << " = " << this->size << "\n";
@@ -209,9 +208,9 @@ void GridImp::freeMemory()
if( this->neighborIndexZ != nullptr ) { delete[] this->neighborIndexZ; this->neighborIndexZ = nullptr; }
if( this->neighborIndexNegative != nullptr ) { delete[] this->neighborIndexNegative; this->neighborIndexNegative = nullptr; }
if( this->sparseIndices != nullptr ) { delete[] this->sparseIndices; this->sparseIndices = nullptr; }
- if( this->qIndices != nullptr ) { delete[] this->qIndices; this->qIndices = nullptr; }
- if( this->qValues != nullptr ) { delete[] this->qValues; this->qValues = nullptr; }
- if( this->qPatches != nullptr ) { delete[] this->qPatches; this->qPatches = nullptr; }
+ if( this->qIndices != nullptr ) { delete[] this->qIndices; this->qIndices = nullptr; }
+ if( this->qValues != nullptr ) { delete[] this->qValues; this->qValues = nullptr; }
+ if( this->qPatches != nullptr ) { delete[] this->qPatches; this->qPatches = nullptr; }
field.freeMemory();
}
@@ -254,7 +253,7 @@ void GridImp::discretize(Object* solidObject, char innerType, char outerType)
this->sparseIndices[index] = index;
if( this->getFieldEntry(index) == innerType ) continue;
-
+
real x, y, z;
this->transIndexToCoords(index, x, y, z);
@@ -279,7 +278,7 @@ bool GridImp::isInside(const Cell& cell) const
// | +-----+-----+-----+ | +-----+-----+-----+
// +---------+ +---------+
// 0 1 2 0 1 2
-// even even even
+// even even even
// odd odd odd
//
Cell GridImp::getOddCellFromIndex(uint index) const
@@ -349,7 +348,7 @@ void GridImp::addOverlap()
void GridImp::setOverlapTmp( uint index )
{
if( this->field.is( index, INVALID_OUT_OF_GRID ) ){
-
+
if( this->hasNeighborOfType(index, FLUID) ){
this->field.setFieldEntry( index, OVERLAP_TMP );
}
@@ -380,7 +379,7 @@ void GridImp::fixRefinementIntoWall(uint xIndex, uint yIndex, uint zIndex, int d
if( this->xOddStart && ( dir == 1 || dir == -1 ) && ( xIndex % 2 == 0 && xIndex != 0 ) ) return;
if( this->yOddStart && ( dir == 2 || dir == -2 ) && ( yIndex % 2 == 0 && yIndex != 0 ) ) return;
if( this->zOddStart && ( dir == 3 || dir == -3 ) && ( zIndex % 2 == 0 && zIndex != 0 ) ) return;
-
+
//////////////////////////////////////////////////////////////////////////
real dx{ 0.0 }, dy{ 0.0 }, dz{ 0.0 };
@@ -433,31 +432,31 @@ void GridImp::findStopperNode(uint index) // deprecated
void GridImp::findEndOfGridStopperNode(uint index)
{
- if (isValidEndOfGridStopper(index)){
+ if (isValidEndOfGridStopper(index)){
if( this->level != 0 )
- this->field.setFieldEntryToStopperOutOfGrid(index);
+ this->field.setFieldEntryToStopperOutOfGrid(index);
else
this->field.setFieldEntryToStopperOutOfGridBoundary(index);
}
-
- if (isValidEndOfGridBoundaryStopper(index))
- this->field.setFieldEntryToStopperOutOfGridBoundary(index);
+
+ if (isValidEndOfGridBoundaryStopper(index))
+ this->field.setFieldEntryToStopperOutOfGridBoundary(index);
}
void GridImp::findSolidStopperNode(uint index)
{
- if (isValidSolidStopper(index))
- this->field.setFieldEntry(index, STOPPER_SOLID);
+ if (isValidSolidStopper(index))
+ this->field.setFieldEntry(index, STOPPER_SOLID);
}
void GridImp::findBoundarySolidNode(uint index)
{
- if (shouldBeBoundarySolidNode(index))
- {
- this->field.setFieldEntry(index, BC_SOLID);
- this->qIndices[index] = this->numberOfSolidBoundaryNodes++;
- //grid->setNumberOfSolidBoundaryNodes(grid->getNumberOfSolidBoundaryNodes() + 1);
- }
+ if (shouldBeBoundarySolidNode(index))
+ {
+ this->field.setFieldEntry(index, BC_SOLID);
+ this->qIndices[index] = this->numberOfSolidBoundaryNodes++;
+ //grid->setNumberOfSolidBoundaryNodes(grid->getNumberOfSolidBoundaryNodes() + 1);
+ }
}
void GridImp::fixOddCell(uint index)
@@ -483,9 +482,9 @@ bool GridImp::isOutSideOfGrid(Cell &cell) const
bool GridImp::contains(Cell &cell, char type) const
{
for (const auto point : cell) {
- uint index = transCoordToIndex(point.x, point.y, point.z);
- if (index == INVALID_INDEX)
- continue;
+ uint index = transCoordToIndex(point.x, point.y, point.z);
+ if (index == INVALID_INDEX)
+ continue;
if (field.is(index, type))
return true;
}
@@ -495,8 +494,8 @@ bool GridImp::contains(Cell &cell, char type) const
bool GridImp::cellContainsOnly(Cell &cell, char type) const
{
for (const auto point : cell) {
- uint index = transCoordToIndex(point.x, point.y, point.z);
- if (index == INVALID_INDEX)
+ uint index = transCoordToIndex(point.x, point.y, point.z);
+ if (index == INVALID_INDEX)
return false;
if (!field.is(index, type))
return false;
@@ -507,8 +506,8 @@ bool GridImp::cellContainsOnly(Cell &cell, char type) const
bool GridImp::cellContainsOnly(Cell &cell, char typeA, char typeB) const
{
for (const auto point : cell) {
- uint index = transCoordToIndex(point.x, point.y, point.z);
- if (index == INVALID_INDEX)
+ uint index = transCoordToIndex(point.x, point.y, point.z);
+ if (index == INVALID_INDEX)
return false;
if (!field.is(index, typeA) && !field.is(index, typeB))
return false;
@@ -524,91 +523,91 @@ const Object * GridImp::getObject() const
void GridImp::setNodeTo(Cell &cell, char type)
{
for (const auto point : cell) {
- uint index = transCoordToIndex(point.x, point.y, point.z);
- if (index == INVALID_INDEX)
- continue;
- field.setFieldEntry(index, type);
+ uint index = transCoordToIndex(point.x, point.y, point.z);
+ if (index == INVALID_INDEX)
+ continue;
+ field.setFieldEntry(index, type);
}
}
void GridImp::setNodeTo(uint index, char type)
{
- if( index != INVALID_INDEX )
- field.setFieldEntry(index, type);
+ if( index != INVALID_INDEX )
+ field.setFieldEntry(index, type);
}
bool GridImp::isNode(uint index, char type) const
{
if( index != INVALID_INDEX )
- return field.is(index, type);
+ return field.is(index, type);
throw std::runtime_error("GridImp::isNode() -> index == INVALID_INDEX not supported.");
}
bool GridImp::isValidEndOfGridStopper(uint index) const
{
- // Lenz: also includes corner stopper nodes
- if (!this->field.is(index, INVALID_OUT_OF_GRID))
- return false;
+ // Lenz: also includes corner stopper nodes
+ if (!this->field.is(index, INVALID_OUT_OF_GRID))
+ return false;
- return hasNeighborOfType(index, FLUID);
+ return hasNeighborOfType(index, FLUID);
}
bool GridImp::isValidEndOfGridBoundaryStopper(uint index) const
{
- // Lenz: also includes corner stopper nodes
- if (!this->field.is(index, FLUID))
- return false;
+ // Lenz: also includes corner stopper nodes
+ if (!this->field.is(index, FLUID))
+ return false;
- return ! hasAllNeighbors(index);
+ return ! hasAllNeighbors(index);
}
bool GridImp::isValidSolidStopper(uint index) const
{
- // Lenz: also includes corner stopper nodes
- if (!this->field.is(index, INVALID_SOLID))
- return false;
+ // Lenz: also includes corner stopper nodes
+ if (!this->field.is(index, INVALID_SOLID))
+ return false;
- return hasNeighborOfType(index, FLUID);
+ return hasNeighborOfType(index, FLUID);
}
bool GridImp::shouldBeBoundarySolidNode(uint index) const
{
- if (!this->field.is(index, FLUID))
- return false;
+ if (!this->field.is(index, FLUID))
+ return false;
- return hasNeighborOfType(index, STOPPER_SOLID);
+ return hasNeighborOfType(index, STOPPER_SOLID);
}
bool GridImp::hasAllNeighbors(uint index) const
{
- // new version by Lenz, utilizes the range based for loop for all directions
- real x, y, z;
- this->transIndexToCoords(index, x, y, z);
- for (const auto dir : this->distribution) {
- const uint neighborIndex = this->transCoordToIndex(x + dir[0] * this->getDelta(), y + dir[1] * this->getDelta(), z + dir[2] * this->getDelta());
+ // new version by Lenz, utilizes the range based for loop for all directions
+ real x, y, z;
+ this->transIndexToCoords(index, x, y, z);
+ for (const auto dir : this->distribution) {
+ const uint neighborIndex = this->transCoordToIndex(x + dir[0] * this->getDelta(), y + dir[1] * this->getDelta(), z + dir[2] * this->getDelta());
- if (neighborIndex == INVALID_INDEX) return false;
- }
+ if (neighborIndex == INVALID_INDEX) return false;
+ }
- return true;
+ return true;
}
bool GridImp::hasNeighborOfType(uint index, char type) const
{
- // new version by Lenz, utilizes the range based for loop for all directions
- real x, y, z;
- this->transIndexToCoords(index, x, y, z);
- for (const auto dir : this->distribution) {
- const uint neighborIndex = this->transCoordToIndex(x + dir[0] * this->getDelta(), y + dir[1] * this->getDelta(), z + dir[2] * this->getDelta());
+ // new version by Lenz, utilizes the range based for loop for all directions
+ real x, y, z;
+ this->transIndexToCoords(index, x, y, z);
+ for (const auto dir : this->distribution) {
+ const uint neighborIndex = this->transCoordToIndex(x + dir[0] * this->getDelta(), y + dir[1] * this->getDelta(), z + dir[2] * this->getDelta());
- if (neighborIndex == INVALID_INDEX) continue;
+ if (neighborIndex == INVALID_INDEX) continue;
- if (this->field.is(neighborIndex, type))
- return true;
- }
+ if (this->field.is(neighborIndex, type))
+ return true;
+ }
- return false;
+ return false;
}
bool GridImp::nodeInNextCellIs(int index, char type) const
@@ -630,13 +629,13 @@ bool GridImp::nodeInNextCellIs(int index, char type) const
const uint indexXYZ = transCoordToIndex(neighborX, neighborY, neighborZ);
- const bool typeX = indexX == INVALID_INDEX ? false : this->field.is(indexX, type);
- const bool typeY = indexY == INVALID_INDEX ? false : this->field.is(indexY, type);
- const bool typeXY = indexXY == INVALID_INDEX ? false : this->field.is(indexXY, type);
- const bool typeZ = indexZ == INVALID_INDEX ? false : this->field.is(indexZ, type);
- const bool typeYZ = indexYZ == INVALID_INDEX ? false : this->field.is(indexYZ, type);
- const bool typeXZ = indexXZ == INVALID_INDEX ? false : this->field.is(indexXZ, type);
- const bool typeXYZ = indexXYZ == INVALID_INDEX ? false : this->field.is(indexXYZ, type);
+ const bool typeX = indexX == INVALID_INDEX ? false : this->field.is(indexX, type);
+ const bool typeY = indexY == INVALID_INDEX ? false : this->field.is(indexY, type);
+ const bool typeXY = indexXY == INVALID_INDEX ? false : this->field.is(indexXY, type);
+ const bool typeZ = indexZ == INVALID_INDEX ? false : this->field.is(indexZ, type);
+ const bool typeYZ = indexYZ == INVALID_INDEX ? false : this->field.is(indexYZ, type);
+ const bool typeXZ = indexXZ == INVALID_INDEX ? false : this->field.is(indexXZ, type);
+ const bool typeXYZ = indexXYZ == INVALID_INDEX ? false : this->field.is(indexXYZ, type);
return typeX || typeY || typeXY || typeZ || typeYZ
|| typeXZ || typeXYZ;
@@ -661,13 +660,13 @@ bool GridImp::nodeInPreviousCellIs(int index, char type) const
const uint indexXYZ = transCoordToIndex(neighborX, neighborY, neighborZ);
- const bool typeX = indexX == INVALID_INDEX ? false : this->field.is(indexX , type);
- const bool typeY = indexY == INVALID_INDEX ? false : this->field.is(indexY , type);
- const bool typeXY = indexXY == INVALID_INDEX ? false : this->field.is(indexXY , type);
- const bool typeZ = indexZ == INVALID_INDEX ? false : this->field.is(indexZ , type);
- const bool typeYZ = indexYZ == INVALID_INDEX ? false : this->field.is(indexYZ , type);
- const bool typeXZ = indexXZ == INVALID_INDEX ? false : this->field.is(indexXZ , type);
- const bool typeXYZ = indexXYZ == INVALID_INDEX ? false : this->field.is(indexXYZ, type);
+ const bool typeX = indexX == INVALID_INDEX ? false : this->field.is(indexX , type);
+ const bool typeY = indexY == INVALID_INDEX ? false : this->field.is(indexY , type);
+ const bool typeXY = indexXY == INVALID_INDEX ? false : this->field.is(indexXY , type);
+ const bool typeZ = indexZ == INVALID_INDEX ? false : this->field.is(indexZ , type);
+ const bool typeYZ = indexYZ == INVALID_INDEX ? false : this->field.is(indexYZ , type);
+ const bool typeXZ = indexXZ == INVALID_INDEX ? false : this->field.is(indexXZ , type);
+ const bool typeXYZ = indexXYZ == INVALID_INDEX ? false : this->field.is(indexXYZ, type);
return typeX || typeY || typeXY || typeZ || typeYZ
|| typeXZ || typeXYZ;
@@ -678,8 +677,8 @@ bool GridImp::nodeInCellIs(Cell& cell, char type) const
for (const auto node : cell)
{
const uint index = transCoordToIndex(node.x, node.y, node.z);
- if (index == INVALID_INDEX)
- continue;
+ if (index == INVALID_INDEX)
+ continue;
if (field.is(index, type))
return true;
}
@@ -696,9 +695,9 @@ void GridImp::setCellTo(uint index, char type)
for (const auto node : cell)
{
const uint nodeIndex = transCoordToIndex(node.x, node.y, node.z);
- if (nodeIndex == INVALID_INDEX)
- continue;
- this->field.setFieldEntry(nodeIndex, type);
+ if (nodeIndex == INVALID_INDEX)
+ continue;
+ this->field.setFieldEntry(nodeIndex, type);
}
}
@@ -712,15 +711,21 @@ void GridImp::setNonStopperOutOfGridCellTo(uint index, char type)
for (const auto node : cell)
{
const uint nodeIndex = transCoordToIndex(node.x, node.y, node.z);
- if (nodeIndex == INVALID_INDEX)
- continue;
+ if (nodeIndex == INVALID_INDEX)
+ continue;
- if( this->getFieldEntry( nodeIndex ) != STOPPER_OUT_OF_GRID &&
+ if( this->getFieldEntry( nodeIndex ) != STOPPER_OUT_OF_GRID &&
this->getFieldEntry( nodeIndex ) != STOPPER_OUT_OF_GRID_BOUNDARY )
this->field.setFieldEntry(nodeIndex, type);
}
}
+bool GridImp::nodeHasBC(uint index) const
+{
+ return (getFieldEntry(index) == vf::gpu::BC_PRESSURE || getFieldEntry(index) == vf::gpu::BC_VELOCITY ||
+ getFieldEntry(index) == vf::gpu::BC_NOSLIP || getFieldEntry(index) == vf::gpu::BC_SLIP ||
+ getFieldEntry(index) == vf::gpu::BC_STRESS);
+}
void GridImp::setPeriodicity(bool periodicityX, bool periodicityY, bool periodicityZ)
{
@@ -744,17 +749,17 @@ void GridImp::setPeriodicityZ(bool periodicity)
this->periodicityZ = periodicity;
}
-bool GridImp::getPeriodicityX()
+bool GridImp::getPeriodicityX() const
{
return this->periodicityX;
}
-bool GridImp::getPeriodicityY()
+bool GridImp::getPeriodicityY() const
{
return this->periodicityY;
}
-bool GridImp::getPeriodicityZ()
+bool GridImp::getPeriodicityZ() const
{
return this->periodicityZ;
}
@@ -770,7 +775,7 @@ uint GridImp::transCoordToIndex(const real &x, const real &y, const real &z) con
const uint yIndex = getYIndex(y);
const uint zIndex = getZIndex(z);
- if (xIndex >= nx || yIndex >= ny || zIndex >= nz)
+ if (xIndex >= nx || yIndex >= ny || zIndex >= nz)
return INVALID_INDEX;
return xIndex + nx * (yIndex + ny * zIndex);
@@ -819,20 +824,20 @@ TriangularMeshDiscretizationStrategy * GridImp::getTriangularMeshDiscretizationS
uint GridImp::getNumberOfSolidBoundaryNodes() const
{
- return this->numberOfSolidBoundaryNodes;
+ return this->numberOfSolidBoundaryNodes;
}
void GridImp::setNumberOfSolidBoundaryNodes(uint numberOfSolidBoundaryNodes)
{
- if (numberOfSolidBoundaryNodes < INVALID_INDEX)
- this->numberOfSolidBoundaryNodes = numberOfSolidBoundaryNodes;
+ if (numberOfSolidBoundaryNodes < INVALID_INDEX)
+ this->numberOfSolidBoundaryNodes = numberOfSolidBoundaryNodes;
}
real GridImp::getQValue(const uint index, const uint dir) const
{
- const int qIndex = dir * this->numberOfSolidBoundaryNodes + this->qIndices[index];
+ const int qIndex = dir * this->numberOfSolidBoundaryNodes + this->qIndices[index];
- return this->qValues[qIndex];
+ return this->qValues[qIndex];
}
uint GridImp::getQPatch(const uint index) const
@@ -858,7 +863,7 @@ void GridImp::findSparseIndices(SPtr<Grid> finerGrid)
{
*logging::out << logging::Logger::INFO_INTERMEDIATE << "Find sparse indices...";
auto fineGrid = std::static_pointer_cast<GridImp>(finerGrid);
-
+
this->updateSparseIndices();
#pragma omp parallel for
@@ -906,7 +911,7 @@ void GridImp::updateSparseIndices()
sparseSize = size - removedNodes;
}
-void GridImp::findFluidNodeIndices(bool splitDomain)
+void GridImp::findFluidNodeIndices(bool splitDomain)
{
// find sparse index of all fluid nodes
this->fluidNodeIndices.clear();
@@ -935,7 +940,7 @@ void GridImp::findFluidNodeIndicesBorder() {
// resize fluidNodeIndicesBorder (for better performance in copy operation)
size_t newSize = 0;
for (CommunicationIndices& ci : this->communicationIndices)
- newSize += ci.sendIndices.size();
+ newSize += ci.sendIndices.size();
this->fluidNodeIndicesBorder.reserve(newSize);
// copy all send indices to fluidNodeIndicesBorder
@@ -968,7 +973,7 @@ void GridImp::setNeighborIndices(uint index)
this->setStopperNeighborCoords(index);
return;
}
-
+
if (this->sparseIndices[index] == -1)
return;
@@ -1002,9 +1007,9 @@ void GridImp::setStopperNeighborCoords(uint index)
if (vf::Math::lessEqual(z + delta, endZ + (0.5 * delta)) && !this->field.isInvalidOutOfGrid(this->transCoordToIndex(x, y, z + delta)))
neighborIndexZ[index] = getSparseIndex(x, y, z + delta);
- if (vf::Math::greaterEqual(x - delta, endX) &&
- vf::Math::greaterEqual(y - delta, endY) &&
- vf::Math::greaterEqual(z - delta, endZ) &&
+ if (vf::Math::greaterEqual(x - delta, endX) &&
+ vf::Math::greaterEqual(y - delta, endY) &&
+ vf::Math::greaterEqual(z - delta, endZ) &&
!this->field.isInvalidOutOfGrid(this->transCoordToIndex(x - delta, y - delta, z - delta)))
{
neighborIndexNegative[index] = getSparseIndex(x - delta, y - delta, z - delta);
@@ -1035,7 +1040,7 @@ real GridImp::getNeighborCoord(bool periodicity, real startCoord, real coords[3]
return coords[direction] + delta;
}
-
+
return coords[direction] + delta;
}
@@ -1061,7 +1066,7 @@ real GridImp::getNegativeNeighborCoord(bool periodicity, real startCoord, real c
return getLastFluidNode(coords, direction, startCoord);
}
-
+
return coords[direction] - delta;
}
@@ -1155,7 +1160,7 @@ void GridImp::limitToSubDomain(SPtr<BoundingBox> subDomainBox, LbmOrGks lbmOrGks
if( lbmOrGks == LBM )
tmpSubDomainBox.extend(this->delta);
- if (!tmpSubDomainBox.isInside(x, y, z)
+ if (!tmpSubDomainBox.isInside(x, y, z)
&& ( this->getFieldEntry(index) == FLUID ||
this->getFieldEntry(index) == FLUID_CFC ||
this->getFieldEntry(index) == FLUID_CFF ||
@@ -1184,13 +1189,13 @@ void GridImp::limitToSubDomain(SPtr<BoundingBox> subDomainBox, LbmOrGks lbmOrGks
void GridImp::findGridInterfaceCF(uint index, GridImp& finerGrid, LbmOrGks lbmOrGks)
{
- if (lbmOrGks == LBM)
- {
- gridInterface->findInterfaceCF (index, this, &finerGrid);
- gridInterface->findBoundaryGridInterfaceCF(index, this, &finerGrid);
- }
- else if (lbmOrGks == GKS)
- gridInterface->findInterfaceCF_GKS(index, this, &finerGrid);
+ if (lbmOrGks == LBM)
+ {
+ gridInterface->findInterfaceCF (index, this, &finerGrid);
+ gridInterface->findBoundaryGridInterfaceCF(index, this, &finerGrid);
+ }
+ else if (lbmOrGks == GKS)
+ gridInterface->findInterfaceCF_GKS(index, this, &finerGrid);
}
void GridImp::findGridInterfaceFC(uint index, GridImp& finerGrid)
@@ -1217,16 +1222,16 @@ void GridImp::mesh(Object* object)
if (triangularMesh)
triangularMeshDiscretizationStrategy->discretize(triangularMesh, this, INVALID_SOLID, FLUID);
else
- //new method for geometric primitives (not cell based) to be implemented
+ //new method for geometric primitives (not cell based) to be implemented
this->discretize(object, INVALID_SOLID, FLUID);
this->closeNeedleCells();
- #pragma omp parallel for
+ #pragma omp parallel for
for (int index = 0; index < (int)this->size; index++)
this->findSolidStopperNode(index);
- //#pragma omp parallel for
+ //#pragma omp parallel for
for (int index = 0; index < (int)this->size; index++) {
this->findBoundarySolidNode(index);
}
@@ -1359,7 +1364,7 @@ void GridImp::findQs(Object* object) //TODO: enable qs for primitive objects
findQsPrimitive(object);
}
-void GridImp::allocateQs()
+void GridImp::allocateQs()
{
this->qPatches = new uint[this->getNumberOfSolidBoundaryNodes()];
@@ -1379,8 +1384,8 @@ void GridImp::findQs(TriangularMesh &triangularMesh)
if( this->qComputationStage == qComputationStageType::ComputeQs )
allocateQs();
-
-
+
+
#pragma omp parallel for
for (int i = 0; i < triangularMesh.size; i++)
this->findQs(triangularMesh.triangles[i]);
@@ -1406,15 +1411,15 @@ void GridImp::findQs(Triangle &triangle)
//if (!field.isFluid(index))
// continue;
- if( index == INVALID_INDEX ) continue;
+ if( index == INVALID_INDEX ) continue;
const Vertex point(x, y, z);
if( this->qComputationStage == qComputationStageType::ComputeQs ){
if(this->field.is(index, BC_SOLID))
{
- calculateQs(index, point, triangle);
- }
+ calculateQs(index, point, triangle);
+ }
}
else if( this->qComputationStage == qComputationStageType::FindSolidBoundaryNodes )
{
@@ -1449,14 +1454,14 @@ void GridImp::findQsPrimitive(Object * object)
real x,y,z;
this->transIndexToCoords(index,x,y,z);
-
+
const Vertex point(x, y, z);
if( this->qComputationStage == qComputationStageType::ComputeQs ){
if(this->field.is(index, BC_SOLID))
{
- calculateQs(index, point, object);
- }
+ calculateQs(index, point, object);
+ }
}
else if( this->qComputationStage == qComputationStageType::FindSolidBoundaryNodes )
{
@@ -1477,66 +1482,66 @@ void GridImp::calculateQs(const uint index, const Vertex &point, Object* object)
{
Vertex pointOnTriangle, direction;
- real subdistance;
- int error;
- for (int i = distribution.dir_start; i <= distribution.dir_end; i++)
- {
- direction = Vertex( real(distribution.dirs[i * DIMENSION + 0]),
+ real subdistance;
+ int error;
+ for (int i = distribution.dir_start; i <= distribution.dir_end; i++)
+ {
+ direction = Vertex( real(distribution.dirs[i * DIMENSION + 0]),
real(distribution.dirs[i * DIMENSION + 1]),
- real(distribution.dirs[i * DIMENSION + 2]) );
+ real(distribution.dirs[i * DIMENSION + 2]) );
- uint neighborIndex = this->transCoordToIndex(point.x + direction.x * this->delta,
- point.y + direction.y * this->delta,
- point.z + direction.z * this->delta);
+ uint neighborIndex = this->transCoordToIndex(point.x + direction.x * this->delta,
+ point.y + direction.y * this->delta,
+ point.z + direction.z * this->delta);
- if (neighborIndex == INVALID_INDEX) continue;
+ if (neighborIndex == INVALID_INDEX) continue;
- error = object->getIntersection(point, direction, pointOnTriangle, subdistance);
+ error = object->getIntersection(point, direction, pointOnTriangle, subdistance);
- subdistance /= this->delta;
+ subdistance /= this->delta;
- if (error == 0 && vf::Math::lessEqual(subdistance, 1.0) && vf::Math::greaterEqual(subdistance, 0.0))
- {
- if ( -0.5 > this->qValues[i*this->numberOfSolidBoundaryNodes + this->qIndices[index]] ||
+ if (error == 0 && vf::Math::lessEqual(subdistance, 1.0) && vf::Math::greaterEqual(subdistance, 0.0))
+ {
+ if ( -0.5 > this->qValues[i*this->numberOfSolidBoundaryNodes + this->qIndices[index]] ||
subdistance < this->qValues[i*this->numberOfSolidBoundaryNodes + this->qIndices[index]] )
- {
+ {
+
+ this->qValues[i*this->numberOfSolidBoundaryNodes + this->qIndices[index]] = subdistance;
- this->qValues[i*this->numberOfSolidBoundaryNodes + this->qIndices[index]] = subdistance;
-
this->qPatches[ this->qIndices[index] ] = 0;
- }
- }
- }
+ }
+ }
+ }
}
bool GridImp::checkIfAtLeastOneValidQ(const uint index, const Vertex &point, Object* object) const
{
Vertex pointOnTriangle, direction;
- real subdistance;
- int error;
- for (int i = distribution.dir_start; i <= distribution.dir_end; i++)
- {
- direction = Vertex( real(distribution.dirs[i * DIMENSION + 0]),
+ real subdistance;
+ int error;
+ for (int i = distribution.dir_start; i <= distribution.dir_end; i++)
+ {
+ direction = Vertex( real(distribution.dirs[i * DIMENSION + 0]),
real(distribution.dirs[i * DIMENSION + 1]),
- real(distribution.dirs[i * DIMENSION + 2]) );
+ real(distribution.dirs[i * DIMENSION + 2]) );
- uint neighborIndex = this->transCoordToIndex(point.x + direction.x * this->delta,
- point.y + direction.y * this->delta,
- point.z + direction.z * this->delta);
+ uint neighborIndex = this->transCoordToIndex(point.x + direction.x * this->delta,
+ point.y + direction.y * this->delta,
+ point.z + direction.z * this->delta);
- if (neighborIndex == INVALID_INDEX) continue;
+ if (neighborIndex == INVALID_INDEX) continue;
- error = object->getIntersection(point, direction, pointOnTriangle, subdistance);
+ error = object->getIntersection(point, direction, pointOnTriangle, subdistance);
- subdistance /= this->delta;
+ subdistance /= this->delta;
- if (error == 0 && vf::Math::lessEqual(subdistance, 1.0) && vf::Math::greaterEqual(subdistance, 0.0))
- {
- return true;
- }
- }
+ if (error == 0 && vf::Math::lessEqual(subdistance, 1.0) && vf::Math::greaterEqual(subdistance, 0.0))
+ {
+ return true;
+ }
+ }
return false;
}
@@ -1565,7 +1570,7 @@ void GridImp::calculateQs(const Vertex &point, const Triangle &triangle) const
error = triangle.getTriangleIntersection(point, direction, pointOnTriangle, subdistance);
- subdistance /= this->delta;
+ subdistance /= this->delta;
if (error == 0 && subdistance < 1.0 && subdistance > 0.0)
{
@@ -1577,81 +1582,80 @@ void GridImp::calculateQs(const Vertex &point, const Triangle &triangle) const
void GridImp::calculateQs(const uint index, const Vertex &point, const Triangle &triangle) const
{
- Vertex pointOnTriangle, direction;
- real subdistance;
- int error;
- for (int i = distribution.dir_start; i <= distribution.dir_end; i++)
- {
+ Vertex pointOnTriangle, direction;
+ real subdistance;
+ int error;
+ for (int i = distribution.dir_start; i <= distribution.dir_end; i++)
+ {
#if defined(__CUDA_ARCH__)
- direction = Vertex(DIRECTIONS[i][0], DIRECTIONS[i][1], DIRECTIONS[i][2]);
+ direction = Vertex(DIRECTIONS[i][0], DIRECTIONS[i][1], DIRECTIONS[i][2]);
#else
- direction = Vertex( real(distribution.dirs[i * DIMENSION + 0]),
+ direction = Vertex( real(distribution.dirs[i * DIMENSION + 0]),
real(distribution.dirs[i * DIMENSION + 1]),
- real(distribution.dirs[i * DIMENSION + 2]) );
+ real(distribution.dirs[i * DIMENSION + 2]) );
#endif
- uint neighborIndex = this->transCoordToIndex(point.x + direction.x * this->delta,
- point.y + direction.y * this->delta,
- point.z + direction.z * this->delta);
+ uint neighborIndex = this->transCoordToIndex(point.x + direction.x * this->delta,
+ point.y + direction.y * this->delta,
+ point.z + direction.z * this->delta);
- if (neighborIndex == INVALID_INDEX) continue;
+ if (neighborIndex == INVALID_INDEX) continue;
- error = triangle.getTriangleIntersection(point, direction, pointOnTriangle, subdistance);
+ error = triangle.getTriangleIntersection(point, direction, pointOnTriangle, subdistance);
- subdistance /= this->delta;
+ subdistance /= this->delta;
- if (error == 0 && vf::Math::lessEqual(subdistance, 1.0) && vf::Math::greaterEqual(subdistance, 0.0))
- {
- if ( -0.5 > this->qValues[i*this->numberOfSolidBoundaryNodes + this->qIndices[index]] ||
+ if (error == 0 && vf::Math::lessEqual(subdistance, 1.0) && vf::Math::greaterEqual(subdistance, 0.0))
+ {
+ if ( -0.5 > this->qValues[i*this->numberOfSolidBoundaryNodes + this->qIndices[index]] ||
subdistance < this->qValues[i*this->numberOfSolidBoundaryNodes + this->qIndices[index]] )
- {
- this->qValues[i*this->numberOfSolidBoundaryNodes + this->qIndices[index]] = subdistance;
+ {
+ this->qValues[i*this->numberOfSolidBoundaryNodes + this->qIndices[index]] = subdistance;
this->qPatches[ this->qIndices[index] ] = triangle.patchIndex;
- }
- }
- }
+ }
+ }
+ }
}
bool GridImp::checkIfAtLeastOneValidQ(const uint index, const Vertex & point, const Triangle & triangle) const
{
- Vertex pointOnTriangle, direction;
- real subdistance;
- int error;
- for (int i = distribution.dir_start; i <= distribution.dir_end; i++)
- {
+ Vertex pointOnTriangle, direction;
+ real subdistance;
+ int error;
+ for (int i = distribution.dir_start; i <= distribution.dir_end; i++)
+ {
#if defined(__CUDA_ARCH__)
- direction = Vertex(DIRECTIONS[i][0], DIRECTIONS[i][1], DIRECTIONS[i][2]);
+ direction = Vertex(DIRECTIONS[i][0], DIRECTIONS[i][1], DIRECTIONS[i][2]);
#else
- direction = Vertex(real(distribution.dirs[i * DIMENSION + 0]),
+ direction = Vertex(real(distribution.dirs[i * DIMENSION + 0]),
real(distribution.dirs[i * DIMENSION + 1]),
- real(distribution.dirs[i * DIMENSION + 2]));
+ real(distribution.dirs[i * DIMENSION + 2]));
#endif
- uint neighborIndex = this->transCoordToIndex(point.x + direction.x * this->delta,
- point.y + direction.y * this->delta,
- point.z + direction.z * this->delta);
- if (neighborIndex == INVALID_INDEX) continue;
+ uint neighborIndex = this->transCoordToIndex(point.x + direction.x * this->delta,
+ point.y + direction.y * this->delta,
+ point.z + direction.z * this->delta);
+ if (neighborIndex == INVALID_INDEX) continue;
- error = triangle.getTriangleIntersection(point, direction, pointOnTriangle, subdistance);
+ error = triangle.getTriangleIntersection(point, direction, pointOnTriangle, subdistance);
- subdistance /= this->delta;
+ subdistance /= this->delta;
- if (error == 0 && vf::Math::lessEqual(subdistance, 1.0) && vf::Math::greaterEqual(subdistance, 0.0))
- {
- return true;
- }
- }
+ if (error == 0 && vf::Math::lessEqual(subdistance, 1.0) && vf::Math::greaterEqual(subdistance, 0.0))
+ {
+ return true;
+ }
+ }
return false;
}
void GridImp::findCommunicationIndices(int direction, SPtr<BoundingBox> subDomainBox, LbmOrGks lbmOrGks)
{
for( uint index = 0; index < this->size; index++ ){
-
real x, y, z;
this->transIndexToCoords(index, x, y, z);
-
+
if( this->getFieldEntry(index) == INVALID_OUT_OF_GRID ||
this->getFieldEntry(index) == INVALID_SOLID ||
this->getFieldEntry(index) == INVALID_COARSE_UNDER_FINE ||
@@ -1660,7 +1664,6 @@ void GridImp::findCommunicationIndices(int direction, SPtr<BoundingBox> subDomai
if( lbmOrGks == LBM && this->getFieldEntry(index) == STOPPER_OUT_OF_GRID_BOUNDARY ) continue;
if( lbmOrGks == LBM && this->getFieldEntry(index) == STOPPER_SOLID ) continue;
-
if( direction == CommunicationDirections::MX ) findCommunicationIndex( index, x, subDomainBox->minX, direction);
if( direction == CommunicationDirections::PX ) findCommunicationIndex( index, x, subDomainBox->maxX, direction);
if( direction == CommunicationDirections::MY ) findCommunicationIndex( index, y, subDomainBox->minY, direction);
@@ -1672,16 +1675,13 @@ void GridImp::findCommunicationIndices(int direction, SPtr<BoundingBox> subDomai
void GridImp::findCommunicationIndex( uint index, real coordinate, real limit, int direction ){
// negative direction get a negative sign
- real s = ( direction % 2 == 0 ) ? ( -1.0 ) : ( 1.0 );
-
+ real s = ( direction % 2 == 0 ) ? ( -1.0 ) : ( 1.0 );
- if (std::abs(coordinate - (limit + s * 0.5 * this->delta)) < 0.1 * this->delta) {
- this->communicationIndices[direction].receiveIndices.push_back(index);
- }
+ if (std::abs(coordinate - (limit + s * 0.5 * this->delta)) < 0.1 * this->delta)
+ this->communicationIndices[direction].receiveIndices.push_back(index);
- if (std::abs(coordinate - (limit - s * 0.5 * this->delta)) < 0.1 * this->delta) {
- this->communicationIndices[direction].sendIndices.push_back(index);
- }
+ if (std::abs(coordinate - (limit - s * 0.5 * this->delta)) < 0.1 * this->delta)
+ this->communicationIndices[direction].sendIndices.push_back(index);
}
bool GridImp::isSendNode(int index) const
@@ -1727,14 +1727,14 @@ uint GridImp::getReceiveIndex(int direction, uint index)
void GridImp::repairCommunicationIndices(int direction)
{
- this->communicationIndices[direction].sendIndices.insert( this->communicationIndices[direction].sendIndices.end(),
- this->communicationIndices[direction+1].sendIndices.begin(),
+ this->communicationIndices[direction].sendIndices.insert( this->communicationIndices[direction].sendIndices.end(),
+ this->communicationIndices[direction+1].sendIndices.begin(),
this->communicationIndices[direction+1].sendIndices.end() );
- this->communicationIndices[direction+1].receiveIndices.insert( this->communicationIndices[direction+1].receiveIndices.end(),
- this->communicationIndices[direction].receiveIndices.begin(),
+ this->communicationIndices[direction+1].receiveIndices.insert( this->communicationIndices[direction+1].receiveIndices.end(),
+ this->communicationIndices[direction].receiveIndices.begin(),
this->communicationIndices[direction].receiveIndices.end() );
this->communicationIndices[direction].receiveIndices = this->communicationIndices[direction+1].receiveIndices;
@@ -1839,19 +1839,19 @@ real GridImp::getMaximumOnNodes(const real &maxExact, const real &decimalStart,
return maxNode;
}
-uint GridImp::getXIndex(real x) const
-{
- return std::lround((x - startX) / delta);
+uint GridImp::getXIndex(real x) const
+{
+ return std::lround((x - startX) / delta);
}
uint GridImp::getYIndex(real y) const
-{
- return std::lround((y - startY) / delta);
+{
+ return std::lround((y - startY) / delta);
}
uint GridImp::getZIndex(real z) const
-{
- return std::lround((z - startZ) / delta);
+{
+ return std::lround((z - startZ) / delta);
}
real GridImp::getDelta() const
@@ -1866,11 +1866,11 @@ uint GridImp::getSize() const
uint GridImp::getSparseSize() const
{
- return this->sparseSize;
+ return this->sparseSize;
}
-uint GridImp::getNumberOfFluidNodes() const {
- return (uint)this->fluidNodeIndices.size();
+uint GridImp::getNumberOfFluidNodes() const {
+ return (uint)this->fluidNodeIndices.size();
}
Field GridImp::getField() const
@@ -2063,23 +2063,147 @@ void GridImp::getNodeValues(real *xCoords, real *yCoords, real *zCoords, uint *n
}
}
-void GridImp::getFluidNodeIndices(uint *fluidNodeIndices) const
-{
+void GridImp::getFluidNodeIndices(uint *fluidNodeIndices) const
+{
for (uint nodeNumber = 0; nodeNumber < (uint)this->fluidNodeIndices.size(); nodeNumber++)
fluidNodeIndices[nodeNumber] = this->fluidNodeIndices[nodeNumber];
}
-uint GridImp::getNumberOfFluidNodesBorder() const
-{
- return (uint)this->fluidNodeIndicesBorder.size();
+uint GridImp::getNumberOfFluidNodesBorder() const
+{
+ return (uint)this->fluidNodeIndicesBorder.size();
}
-void GridImp::getFluidNodeIndicesBorder(uint *fluidNodeIndicesBorder) const
+void GridImp::getFluidNodeIndicesBorder(uint *fluidNodeIndicesBorder) const
{
for (uint nodeNumber = 0; nodeNumber < (uint)this->fluidNodeIndicesBorder.size(); nodeNumber++)
fluidNodeIndicesBorder[nodeNumber] = this->fluidNodeIndicesBorder[nodeNumber];
}
+void GridImp::addFluidNodeIndicesMacroVars(std::vector<uint> _fluidNodeIndicesMacroVars)
+{
+ size_t newSize = this->fluidNodeIndicesMacroVars.size()+_fluidNodeIndicesMacroVars.size();
+ this->fluidNodeIndicesMacroVars.reserve(newSize);
+ std::copy(_fluidNodeIndicesMacroVars.begin(), _fluidNodeIndicesMacroVars.end(), std::back_inserter(this->fluidNodeIndicesMacroVars));
+}
+
+void GridImp::addFluidNodeIndicesApplyBodyForce(std::vector<uint> _fluidNodeIndicesApplyBodyForce)
+{
+
+ size_t newSize = this->fluidNodeIndicesApplyBodyForce.size()+_fluidNodeIndicesApplyBodyForce.size();
+ this->fluidNodeIndicesApplyBodyForce.reserve(newSize);
+ std::copy(_fluidNodeIndicesApplyBodyForce.begin(), _fluidNodeIndicesApplyBodyForce.end(), std::back_inserter(this->fluidNodeIndicesApplyBodyForce));
+}
+
+void GridImp::addFluidNodeIndicesAllFeatures(std::vector<uint> _fluidNodeIndicesAllFeatures)
+{
+
+ size_t newSize = this->fluidNodeIndicesAllFeatures.size()+_fluidNodeIndicesAllFeatures.size();
+ this->fluidNodeIndicesAllFeatures.reserve(newSize);
+ std::copy(_fluidNodeIndicesAllFeatures.begin(), _fluidNodeIndicesAllFeatures.end(), std::back_inserter(this->fluidNodeIndicesAllFeatures));
+}
+
+void GridImp::sortFluidNodeIndicesMacroVars()
+{
+ if(this->fluidNodeIndicesMacroVars.size()>0)
+ {
+ sort(this->fluidNodeIndicesMacroVars.begin(), this->fluidNodeIndicesMacroVars.end());
+ // Remove duplicates
+ this->fluidNodeIndicesMacroVars.erase( unique( this->fluidNodeIndicesMacroVars.begin(), this->fluidNodeIndicesMacroVars.end() ), this->fluidNodeIndicesMacroVars.end() );
+
+ // Remove indices of fluidNodeIndicesAllFeatures from fluidNodeIndicesMacroVars
+ if(this->fluidNodeIndicesAllFeatures.size()>0)
+ {
+ this->fluidNodeIndicesMacroVars.erase( std::remove_if( this->fluidNodeIndicesMacroVars.begin(), this->fluidNodeIndicesMacroVars.end(),
+ [&](auto x){return binary_search(fluidNodeIndicesAllFeatures.begin(),fluidNodeIndicesAllFeatures.end(),x);} ),
+ this->fluidNodeIndicesMacroVars.end()
+ );
+ }
+
+ // Remove indices of fluidNodeIndicesMacroVars from fluidNodeIndices
+ this->fluidNodeIndices.erase( std::remove_if( this->fluidNodeIndices.begin(), this->fluidNodeIndices.end(),
+ [&](auto x){return binary_search(fluidNodeIndicesMacroVars.begin(),fluidNodeIndicesMacroVars.end(),x);} ),
+ this->fluidNodeIndices.end()
+ );
+ }
+}
+
+void GridImp::sortFluidNodeIndicesApplyBodyForce()
+{
+ if(this->fluidNodeIndicesApplyBodyForce.size()>0)
+ {
+ sort(this->fluidNodeIndicesApplyBodyForce.begin(), this->fluidNodeIndicesApplyBodyForce.end());
+ // Remove duplicates
+ this->fluidNodeIndicesApplyBodyForce.erase( unique( this->fluidNodeIndicesApplyBodyForce.begin(), this->fluidNodeIndicesApplyBodyForce.end() ), this->fluidNodeIndicesApplyBodyForce.end() );
+
+ // Remove indices of fluidNodeIndicesAllFeatures from fluidNodeIndicesMacroVars
+ if(this->fluidNodeIndicesAllFeatures.size()>0)
+ {
+ this->fluidNodeIndicesApplyBodyForce.erase( std::remove_if( this->fluidNodeIndicesApplyBodyForce.begin(), this->fluidNodeIndicesApplyBodyForce.end(),
+ [&](auto x){return binary_search(fluidNodeIndicesAllFeatures.begin(),fluidNodeIndicesAllFeatures.end(),x);} ),
+ this->fluidNodeIndicesApplyBodyForce.end()
+ );
+ }
+
+ // Remove indices of fluidNodeIndicesMacroVars from fluidNodeIndices
+ this->fluidNodeIndices.erase( std::remove_if( this->fluidNodeIndices.begin(), this->fluidNodeIndices.end(),
+ [&](auto x){return binary_search(fluidNodeIndicesApplyBodyForce.begin(),fluidNodeIndicesApplyBodyForce.end(),x);} ),
+ this->fluidNodeIndices.end()
+ );
+ }
+}
+
+void GridImp::sortFluidNodeIndicesAllFeatures()
+{
+ if(this->fluidNodeIndicesAllFeatures.size()>0)
+ {
+ sort(this->fluidNodeIndicesAllFeatures.begin(), this->fluidNodeIndicesAllFeatures.end());
+ // Remove duplicates
+ this->fluidNodeIndicesAllFeatures.erase( unique( this->fluidNodeIndicesAllFeatures.begin(), this->fluidNodeIndicesAllFeatures.end() ), this->fluidNodeIndicesAllFeatures.end() );
+ // Remove indices of fluidNodeIndicesMacroVars from fluidNodeIndices
+ this->fluidNodeIndices.erase( std::remove_if( this->fluidNodeIndices.begin(), this->fluidNodeIndices.end(),
+ [&](auto x){return binary_search(fluidNodeIndicesAllFeatures.begin(),fluidNodeIndicesAllFeatures.end(),x);} ),
+ this->fluidNodeIndices.end()
+ );
+ }
+}
+
+uint GridImp::getNumberOfFluidNodeIndicesMacroVars() const {
+ return (uint)this->fluidNodeIndicesMacroVars.size();
+}
+
+uint GridImp::getNumberOfFluidNodeIndicesApplyBodyForce() const {
+ return (uint)this->fluidNodeIndicesApplyBodyForce.size();
+}
+
+uint GridImp::getNumberOfFluidNodeIndicesAllFeatures() const {
+ return (uint)this->fluidNodeIndicesAllFeatures.size();
+}
+
+void GridImp::getFluidNodeIndicesMacroVars(uint *_fluidNodeIndicesMacroVars) const
+{
+ std::copy(fluidNodeIndicesMacroVars.begin(), fluidNodeIndicesMacroVars.end(), _fluidNodeIndicesMacroVars);
+}
+void GridImp::getFluidNodeIndicesApplyBodyForce(uint *_fluidNodeIndicesApplyBodyForce) const
+{
+ std::copy(fluidNodeIndicesApplyBodyForce.begin(), fluidNodeIndicesApplyBodyForce.end(), _fluidNodeIndicesApplyBodyForce);
+}
+void GridImp::getFluidNodeIndicesAllFeatures(uint *_fluidNodeIndicesAllFeatures) const
+{
+ std::copy(fluidNodeIndicesAllFeatures.begin(), fluidNodeIndicesAllFeatures.end(), _fluidNodeIndicesAllFeatures);
+}
+
+
+std::vector<SideType> GridImp::getBCAlreadySet() {
+ return this->bcAlreadySet;
+}
+
+void GridImp::addBCalreadySet(SideType side)
+{
+ this->bcAlreadySet.push_back(side);
+}
+
+
void GridImp::print() const
{
printf("min: (%2.4f, %2.4f, %2.4f), max: (%2.4f, %2.4f, %2.4f), size: %d, delta: %2.4f\n", startX, startY, startZ,
@@ -2087,3 +2211,10 @@ void GridImp::print() const
if(this->gridInterface)
this->gridInterface->print();
}
+
+bool GridImp::isStopperForBC(uint index) const
+{
+ return (this->getFieldEntry(index) == vf::gpu::STOPPER_OUT_OF_GRID_BOUNDARY ||
+ this->getFieldEntry(index) == vf::gpu::STOPPER_OUT_OF_GRID ||
+ this->getFieldEntry(index) == vf::gpu::STOPPER_SOLID);
+}
diff --git a/src/gpu/GridGenerator/grid/GridImp.h b/src/gpu/GridGenerator/grid/GridImp.h
index edb5ca916bf68dcf992ea214dcddb2dc43810352..2cd322ebed78daaf135ad97b881923ca5831bbcd 100644
--- a/src/gpu/GridGenerator/grid/GridImp.h
+++ b/src/gpu/GridGenerator/grid/GridImp.h
@@ -34,6 +34,7 @@
#define GRID_IMP_H
#include <array>
+#include <vector>
#include "Core/LbmOrGks.h"
@@ -52,6 +53,7 @@ class Object;
class BoundingBox;
class TriangularMeshDiscretizationStrategy;
+
#ifdef __GNUC__
#ifndef __clang__
#pragma push
@@ -76,7 +78,7 @@ protected:
public:
static SPtr<GridImp> makeShared(Object* object, real startX, real startY, real startZ, real endX, real endY, real endZ, real delta, std::string d3Qxx, uint level);
- virtual ~GridImp() = default;
+ ~GridImp() override = default;
private:
void initalNumberOfNodesAndSize();
@@ -92,6 +94,7 @@ private:
bool nodeInPreviousCellIs(int index, char type) const;
bool nodeInCellIs(Cell& cell, char type) const override;
+
uint getXIndex(real x) const;
uint getYIndex(real y) const;
uint getZIndex(real z) const;
@@ -115,8 +118,11 @@ private:
int *sparseIndices;
- std::vector<uint> fluidNodeIndices;
- std::vector<uint> fluidNodeIndicesBorder;
+ std::vector<uint> fluidNodeIndices; // run on CollisionTemplate::Default
+ std::vector<uint> fluidNodeIndicesBorder; // run on subdomain border nodes (CollisionTemplate::SubDomainBorder)
+ std::vector<uint> fluidNodeIndicesMacroVars; // run on CollisionTemplate::MacroVars
+ std::vector<uint> fluidNodeIndicesApplyBodyForce; // run on CollisionTemplate::ApplyBodyForce
+ std::vector<uint> fluidNodeIndicesAllFeatures; // run on CollisionTemplate::AllFeatures
uint *qIndices; //maps from matrix index to qIndex
real *qValues;
@@ -132,6 +138,8 @@ private:
bool enableFixRefinementIntoTheWall;
+ std::vector<SideType> bcAlreadySet;
+
protected:
Field field;
int *neighborIndexX, *neighborIndexY, *neighborIndexZ, *neighborIndexNegative;
@@ -146,9 +154,9 @@ public:
void setPeriodicityY(bool periodicity) override;
void setPeriodicityZ(bool periodicity) override;
- bool getPeriodicityX() override;
- bool getPeriodicityY() override;
- bool getPeriodicityZ() override;
+ bool getPeriodicityX() const override;
+ bool getPeriodicityY() const override;
+ bool getPeriodicityZ() const override;
void setEnableFixRefinementIntoTheWall(bool enableFixRefinementIntoTheWall) override;
@@ -182,6 +190,9 @@ public:
void setNumberOfLayers(uint numberOfLayers) override;
+ std::vector<SideType> getBCAlreadySet() override;
+ void addBCalreadySet(SideType side) override;
+
public:
Distribution distribution;
@@ -216,6 +227,7 @@ public:
bool nodeInNextCellIs(int index, char type) const;
bool hasAllNeighbors(uint index) const;
bool hasNeighborOfType(uint index, char type) const;
+ bool nodeHasBC(uint index) const override;
bool cellContainsOnly(Cell &cell, char type) const;
bool cellContainsOnly(Cell &cell, char typeA, char typeB) const;
@@ -256,6 +268,8 @@ public:
static void getGridInterface(uint *gridInterfaceList, const uint *oldGridInterfaceList, uint size);
bool isSparseIndexInFluidNodeIndicesBorder(uint &sparseIndex) const override;
+
+ bool isStopperForBC(uint index) const override;
int *getNeighborsX() const override;
int* getNeighborsY() const override;
@@ -273,7 +287,7 @@ public:
void print() const;
public:
- virtual void findSparseIndices(SPtr<Grid> fineGrid) override;
+ void findSparseIndices(SPtr<Grid> fineGrid) override;
void findForGridInterfaceNewIndices(SPtr<GridImp> fineGrid);
void updateSparseIndices();
@@ -364,6 +378,19 @@ public:
uint getNumberOfFluidNodesBorder() const override;
void getFluidNodeIndicesBorder(uint *fluidNodeIndicesBorder) const override;
+ void addFluidNodeIndicesMacroVars(std::vector<uint> _fluidNodeIndicesMacroVars) override;
+ void addFluidNodeIndicesApplyBodyForce(std::vector<uint> _fluidNodeIndicesApplyBodyForce) override;
+ void addFluidNodeIndicesAllFeatures(std::vector<uint> _fluidNodeIndicesAllFeatures) override;
+ void sortFluidNodeIndicesMacroVars() override;
+ void sortFluidNodeIndicesApplyBodyForce() override;
+ void sortFluidNodeIndicesAllFeatures() override;
+
+ uint getNumberOfFluidNodeIndicesMacroVars() const override;
+ uint getNumberOfFluidNodeIndicesApplyBodyForce() const override;
+ uint getNumberOfFluidNodeIndicesAllFeatures() const override;
+ void getFluidNodeIndicesMacroVars(uint *fluidNodeIndicesMacroVars) const override;
+ void getFluidNodeIndicesApplyBodyForce(uint *fluidNodeIndicesApplyBodyForce) const override;
+ void getFluidNodeIndicesAllFeatures(uint *fluidNodeIndicesAllFeatures) const override;
public:
struct CommunicationIndices {
diff --git a/src/gpu/GridGenerator/io/SimulationFileWriter/SimulationFileWriter.cpp b/src/gpu/GridGenerator/io/SimulationFileWriter/SimulationFileWriter.cpp
index 23fb0f4e7f3e16702e9cb2459606986af1032e49..0238434dc87b453dc21164577d8abd4ce1819793 100644
--- a/src/gpu/GridGenerator/io/SimulationFileWriter/SimulationFileWriter.cpp
+++ b/src/gpu/GridGenerator/io/SimulationFileWriter/SimulationFileWriter.cpp
@@ -417,10 +417,10 @@ void SimulationFileWriter::writeGridInterfaceToFile(SPtr<GridBuilder> builder, u
}
}
-void SimulationFileWriter::writeGridInterfaceToFile(const uint numberOfNodes, std::ofstream& coarseFile, uint* coarse, std::ofstream& fineFile, uint* fine)
+void SimulationFileWriter::writeGridInterfaceToFile(uint numberOfNodes, std::ofstream &coarseFile, uint *coarse,
+ std::ofstream &fineFile, uint *fine)
{
- for (uint index = 0; index < numberOfNodes; index++)
- {
+ for (uint index = 0; index < numberOfNodes; index++) {
coarseFile << coarse[index] << " \n";
fineFile << fine[index] << " \n";
}
@@ -428,17 +428,15 @@ void SimulationFileWriter::writeGridInterfaceToFile(const uint numberOfNodes, st
fineFile << "\n";
}
-void SimulationFileWriter::writeGridInterfaceOffsetToFile(uint numberOfNodes, std::ofstream & offsetFile, real* offset_X, real* offset_Y, real* offset_Z)
+void SimulationFileWriter::writeGridInterfaceOffsetToFile(uint numberOfNodes, std::ofstream &offsetFile, real *offset_X,
+ real *offset_Y, real *offset_Z)
{
- for (uint index = 0; index < numberOfNodes; index++)
- {
+ for (uint index = 0; index < numberOfNodes; index++) {
offsetFile << offset_X[index] << " " << offset_Y[index] << " " << offset_Z[index] << " \n";
}
offsetFile << "\n";
}
-
-
/*#################################################################################*/
/*---------------------------------private methods---------------------------------*/
/*---------------------------------------------------------------------------------*/
diff --git a/src/gpu/VirtualFluids_GPU/CMakeLists.txt b/src/gpu/VirtualFluids_GPU/CMakeLists.txt
index 759528e5346ba8d9899cb90eb64503b20a44c4fc..ed647cb406bca23ef90667b7d17171c7b3f46283 100644
--- a/src/gpu/VirtualFluids_GPU/CMakeLists.txt
+++ b/src/gpu/VirtualFluids_GPU/CMakeLists.txt
@@ -8,7 +8,7 @@ if(MSVC)
set(additional_libraries ws2_32 Traffic) # ws_32 throws an error on Phoenix
endif()
-vf_add_library(PUBLIC_LINK basics lbmCuda PRIVATE_LINK ${additional_libraries} GridGenerator MPI::MPI_CXX vf_cuda)
+vf_add_library(PUBLIC_LINK basics lbm PRIVATE_LINK ${additional_libraries} GridGenerator MPI::MPI_CXX vf_cuda)
#SET(TPN_WIN32 "/EHsc")
#https://stackoverflow.com/questions/6832666/lnk2019-when-including-asio-headers-solution-generated-with-cmake
diff --git a/src/gpu/VirtualFluids_GPU/Calculation/Calc2ndMoments.cpp b/src/gpu/VirtualFluids_GPU/Calculation/Calc2ndMoments.cpp
index f8f5c42b835a1a4ba55e378e624230bbb43dc05a..e3f344231dc9d5e19c09f7ce1fde7d31f1770232 100644
--- a/src/gpu/VirtualFluids_GPU/Calculation/Calc2ndMoments.cpp
+++ b/src/gpu/VirtualFluids_GPU/Calculation/Calc2ndMoments.cpp
@@ -17,11 +17,11 @@ void alloc2ndMoments(Parameter* para, CudaMemoryManager* cudaMemoryManager)
void init2ndMoments(Parameter* para)
{
- for (int lev=para->getCoarse(); lev <= para->getFine(); lev++)
+ for (int lev = para->getCoarse(); lev <= para->getFine(); lev++)
{
//////////////////////////////////////////////////////////////////////////
//init host arrays
- for (unsigned int pos=0;pos<para->getParH(lev)->numberOfNodes;pos++)
+ for (size_t pos = 0; pos < para->getParH(lev)->numberOfNodes; pos++)
{
para->getParH(lev)->kxyFromfcNEQ[pos] = 0.0;
para->getParH(lev)->kyzFromfcNEQ[pos] = 0.0;
@@ -116,7 +116,7 @@ void init3rdMoments(Parameter* para)
{
//////////////////////////////////////////////////////////////////////////
//init host arrays
- for (unsigned int pos=0;pos<para->getParH(lev)->numberOfNodes;pos++)
+ for (size_t pos = 0; pos < para->getParH(lev)->numberOfNodes; pos++)
{
para->getParH(lev)->CUMbbb[pos] = 0.0;
para->getParH(lev)->CUMabc[pos] = 0.0;
@@ -198,7 +198,7 @@ void calc3rdMoments(Parameter* para, CudaMemoryManager* cudaMemoryManager)
void allocHigherOrderMoments(Parameter* para, CudaMemoryManager* cudaMemoryManager)
{
- for (int lev=para->getCoarse(); lev <= para->getFine(); lev++)
+ for (int lev = para->getCoarse(); lev <= para->getFine(); lev++)
{
//////////////////////////////////////////////////////////////////////////
//allocation (device-memory + host-memory)
@@ -211,11 +211,11 @@ void allocHigherOrderMoments(Parameter* para, CudaMemoryManager* cudaMemoryManag
void initHigherOrderMoments(Parameter* para)
{
- for (int lev=para->getCoarse(); lev <= para->getFine(); lev++)
+ for (int lev = para->getCoarse(); lev <= para->getFine(); lev++)
{
//////////////////////////////////////////////////////////////////////////
//init host arrays
- for (unsigned int pos=0;pos<para->getParH(lev)->numberOfNodes;pos++)
+ for (size_t pos = 0; pos < para->getParH(lev)->numberOfNodes; pos++)
{
para->getParH(lev)->CUMcbb[pos] = 0.0;
para->getParH(lev)->CUMbcb[pos] = 0.0;
diff --git a/src/gpu/VirtualFluids_GPU/Calculation/CalcMedian.cpp b/src/gpu/VirtualFluids_GPU/Calculation/CalcMedian.cpp
index 77db571f7f10e0ea0bff827400270dd074d4e666..80a667f91976b745b619fed5d5763b5429a6559c 100644
--- a/src/gpu/VirtualFluids_GPU/Calculation/CalcMedian.cpp
+++ b/src/gpu/VirtualFluids_GPU/Calculation/CalcMedian.cpp
@@ -11,16 +11,16 @@
void allocMedian(Parameter* para, CudaMemoryManager* cudaMemoryManager)
{
- for (int lev=para->getCoarse(); lev <= para->getFine(); lev++)
+ for (int lev = para->getCoarse(); lev <= para->getFine(); lev++)
{
cudaMemoryManager->cudaAllocMedianOut(lev);
- for (unsigned int i = 0; i < para->getParH(lev)->numberOfNodes; i++)
+ for (size_t pos = 0; pos < para->getParH(lev)->numberOfNodes; pos++)
{
- para->getParH(lev)->vx_SP_Med_Out[i] = (real)0.0;
- para->getParH(lev)->vy_SP_Med_Out[i] = (real)0.0;
- para->getParH(lev)->vz_SP_Med_Out[i] = (real)0.0;
- para->getParH(lev)->rho_SP_Med_Out[i] = (real)0.0;
- para->getParH(lev)->press_SP_Med_Out[i] = (real)0.0;
+ para->getParH(lev)->vx_SP_Med_Out[pos] = (real)0.0;
+ para->getParH(lev)->vy_SP_Med_Out[pos] = (real)0.0;
+ para->getParH(lev)->vz_SP_Med_Out[pos] = (real)0.0;
+ para->getParH(lev)->rho_SP_Med_Out[pos] = (real)0.0;
+ para->getParH(lev)->press_SP_Med_Out[pos] = (real)0.0;
}
}
}
@@ -31,15 +31,15 @@ void allocMedian(Parameter* para, CudaMemoryManager* cudaMemoryManager)
void calcMedian(Parameter* para, uint tdiff)
{
- for (int lev=para->getCoarse(); lev <= para->getFine(); lev++)
+ for (int lev = para->getCoarse(); lev <= para->getFine(); lev++)
{
- for (uint i = 0; i < para->getParH(lev)->numberOfNodes; i++)
+ for (size_t pos = 0; pos < para->getParH(lev)->numberOfNodes; pos++)
{
- para->getParH(lev)->vx_SP_Med_Out[i] = para->getParH(lev)->vx_SP_Med[i] / (real)tdiff;
- para->getParH(lev)->vy_SP_Med_Out[i] = para->getParH(lev)->vy_SP_Med[i] / (real)tdiff;
- para->getParH(lev)->vz_SP_Med_Out[i] = para->getParH(lev)->vz_SP_Med[i] / (real)tdiff;
- para->getParH(lev)->rho_SP_Med_Out[i] = para->getParH(lev)->rho_SP_Med[i] / (real)tdiff;
- para->getParH(lev)->press_SP_Med_Out[i] = para->getParH(lev)->press_SP_Med[i]/ (real)tdiff;
+ para->getParH(lev)->vx_SP_Med_Out[pos] = para->getParH(lev)->vx_SP_Med[pos] / (real)tdiff;
+ para->getParH(lev)->vy_SP_Med_Out[pos] = para->getParH(lev)->vy_SP_Med[pos] / (real)tdiff;
+ para->getParH(lev)->vz_SP_Med_Out[pos] = para->getParH(lev)->vz_SP_Med[pos] / (real)tdiff;
+ para->getParH(lev)->rho_SP_Med_Out[pos] = para->getParH(lev)->rho_SP_Med[pos] / (real)tdiff;
+ para->getParH(lev)->press_SP_Med_Out[pos] = para->getParH(lev)->press_SP_Med[pos]/ (real)tdiff;
}
}
}
@@ -75,14 +75,14 @@ void allocMedianAD(Parameter* para, CudaMemoryManager* cudaMemoryManager)
for (int lev = para->getCoarse(); lev <= para->getFine(); lev++)
{
cudaMemoryManager->cudaAllocMedianOutAD(lev);
- for (unsigned int i = 0; i < para->getParH(lev)->numberOfNodes; i++)
+ for (size_t pos = 0; pos < para->getParH(lev)->numberOfNodes; pos++)
{
- para->getParH(lev)->vx_SP_Med_Out[i] = (real)0.0;
- para->getParH(lev)->vy_SP_Med_Out[i] = (real)0.0;
- para->getParH(lev)->vz_SP_Med_Out[i] = (real)0.0;
- para->getParH(lev)->rho_SP_Med_Out[i] = (real)0.0;
- para->getParH(lev)->press_SP_Med_Out[i] = (real)0.0;
- para->getParH(lev)->Conc_Med_Out[i] = (real)0.0;
+ para->getParH(lev)->vx_SP_Med_Out[pos] = (real)0.0;
+ para->getParH(lev)->vy_SP_Med_Out[pos] = (real)0.0;
+ para->getParH(lev)->vz_SP_Med_Out[pos] = (real)0.0;
+ para->getParH(lev)->rho_SP_Med_Out[pos] = (real)0.0;
+ para->getParH(lev)->press_SP_Med_Out[pos] = (real)0.0;
+ para->getParH(lev)->Conc_Med_Out[pos] = (real)0.0;
}
}
}
@@ -95,14 +95,14 @@ void calcMedianAD(Parameter* para, uint tdiff)
{
for (int lev = para->getCoarse(); lev <= para->getFine(); lev++)
{
- for (uint i = 0; i < para->getParH(lev)->numberOfNodes; i++)
+ for (size_t pos = 0; pos < para->getParH(lev)->numberOfNodes; pos++)
{
- para->getParH(lev)->vx_SP_Med_Out[i] = para->getParH(lev)->vx_SP_Med[i] / (real)tdiff;
- para->getParH(lev)->vy_SP_Med_Out[i] = para->getParH(lev)->vy_SP_Med[i] / (real)tdiff;
- para->getParH(lev)->vz_SP_Med_Out[i] = para->getParH(lev)->vz_SP_Med[i] / (real)tdiff;
- para->getParH(lev)->rho_SP_Med_Out[i] = para->getParH(lev)->rho_SP_Med[i] / (real)tdiff;
- para->getParH(lev)->press_SP_Med_Out[i] = para->getParH(lev)->press_SP_Med[i] / (real)tdiff;
- para->getParH(lev)->Conc_Med_Out[i] = para->getParH(lev)->Conc_Med[i] / (real)tdiff;
+ para->getParH(lev)->vx_SP_Med_Out[pos] = para->getParH(lev)->vx_SP_Med[pos] / (real)tdiff;
+ para->getParH(lev)->vy_SP_Med_Out[pos] = para->getParH(lev)->vy_SP_Med[pos] / (real)tdiff;
+ para->getParH(lev)->vz_SP_Med_Out[pos] = para->getParH(lev)->vz_SP_Med[pos] / (real)tdiff;
+ para->getParH(lev)->rho_SP_Med_Out[pos] = para->getParH(lev)->rho_SP_Med[pos] / (real)tdiff;
+ para->getParH(lev)->press_SP_Med_Out[pos] = para->getParH(lev)->press_SP_Med[pos] / (real)tdiff;
+ para->getParH(lev)->Conc_Med_Out[pos] = para->getParH(lev)->Conc_Med[pos] / (real)tdiff;
}
}
}
diff --git a/src/gpu/VirtualFluids_GPU/Calculation/CalcTurbulenceIntensity.cpp b/src/gpu/VirtualFluids_GPU/Calculation/CalcTurbulenceIntensity.cpp
index e91fb6f5c232bd98073a1c930149693f8af4b078..9572252965e1c619702370f8b9a3756bf035035e 100644
--- a/src/gpu/VirtualFluids_GPU/Calculation/CalcTurbulenceIntensity.cpp
+++ b/src/gpu/VirtualFluids_GPU/Calculation/CalcTurbulenceIntensity.cpp
@@ -25,32 +25,32 @@ void calcVelocityAndFluctuations(Parameter *para, CudaMemoryManager *cudaMemoryM
for (int lev = para->getCoarse(); lev <= para->getFine(); lev++) {
cudaMemoryManager->cudaCopyTurbulenceIntensityDH(lev, para->getParH(lev)->numberOfNodes);
- for (uint i = 0; i < para->getParH(lev)->numberOfNodes; i++) {
+ for (size_t pos = 0; pos < para->getParH(lev)->numberOfNodes; pos++) {
// mean velocity
- para->getParH(lev)->vx_mean[i] = para->getParH(lev)->vx_mean[i] / (real)tdiff;
- para->getParH(lev)->vy_mean[i] = para->getParH(lev)->vy_mean[i] / (real)tdiff;
- para->getParH(lev)->vz_mean[i] = para->getParH(lev)->vz_mean[i] / (real)tdiff;
+ para->getParH(lev)->vx_mean[pos] = para->getParH(lev)->vx_mean[pos] / (real)tdiff;
+ para->getParH(lev)->vy_mean[pos] = para->getParH(lev)->vy_mean[pos] / (real)tdiff;
+ para->getParH(lev)->vz_mean[pos] = para->getParH(lev)->vz_mean[pos] / (real)tdiff;
// fluctuations
- para->getParH(lev)->vxx[i] = para->getParH(lev)->vxx[i] / (real)tdiff;
- para->getParH(lev)->vyy[i] = para->getParH(lev)->vyy[i] / (real)tdiff;
- para->getParH(lev)->vzz[i] = para->getParH(lev)->vzz[i] / (real)tdiff;
- para->getParH(lev)->vxy[i] = para->getParH(lev)->vxy[i] / (real)tdiff;
- para->getParH(lev)->vxz[i] = para->getParH(lev)->vxz[i] / (real)tdiff;
- para->getParH(lev)->vyz[i] = para->getParH(lev)->vyz[i] / (real)tdiff;
-
- para->getParH(lev)->vxx[i] =
- para->getParH(lev)->vxx[i] - para->getParH(lev)->vx_mean[i] * para->getParH(lev)->vx_mean[i];
- para->getParH(lev)->vyy[i] =
- para->getParH(lev)->vyy[i] - para->getParH(lev)->vy_mean[i] * para->getParH(lev)->vy_mean[i];
- para->getParH(lev)->vzz[i] =
- para->getParH(lev)->vzz[i] - para->getParH(lev)->vz_mean[i] * para->getParH(lev)->vz_mean[i];
- para->getParH(lev)->vxy[i] =
- para->getParH(lev)->vxy[i] - para->getParH(lev)->vx_mean[i] * para->getParH(lev)->vy_mean[i];
- para->getParH(lev)->vxz[i] =
- para->getParH(lev)->vxz[i] - para->getParH(lev)->vx_mean[i] * para->getParH(lev)->vz_mean[i];
- para->getParH(lev)->vyz[i] =
- para->getParH(lev)->vyz[i] - para->getParH(lev)->vy_mean[i] * para->getParH(lev)->vz_mean[i];
+ para->getParH(lev)->vxx[pos] = para->getParH(lev)->vxx[pos] / (real)tdiff;
+ para->getParH(lev)->vyy[pos] = para->getParH(lev)->vyy[pos] / (real)tdiff;
+ para->getParH(lev)->vzz[pos] = para->getParH(lev)->vzz[pos] / (real)tdiff;
+ para->getParH(lev)->vxy[pos] = para->getParH(lev)->vxy[pos] / (real)tdiff;
+ para->getParH(lev)->vxz[pos] = para->getParH(lev)->vxz[pos] / (real)tdiff;
+ para->getParH(lev)->vyz[pos] = para->getParH(lev)->vyz[pos] / (real)tdiff;
+
+ para->getParH(lev)->vxx[pos] =
+ para->getParH(lev)->vxx[pos] - para->getParH(lev)->vx_mean[pos] * para->getParH(lev)->vx_mean[pos];
+ para->getParH(lev)->vyy[pos] =
+ para->getParH(lev)->vyy[pos] - para->getParH(lev)->vy_mean[pos] * para->getParH(lev)->vy_mean[pos];
+ para->getParH(lev)->vzz[pos] =
+ para->getParH(lev)->vzz[pos] - para->getParH(lev)->vz_mean[pos] * para->getParH(lev)->vz_mean[pos];
+ para->getParH(lev)->vxy[pos] =
+ para->getParH(lev)->vxy[pos] - para->getParH(lev)->vx_mean[pos] * para->getParH(lev)->vy_mean[pos];
+ para->getParH(lev)->vxz[pos] =
+ para->getParH(lev)->vxz[pos] - para->getParH(lev)->vx_mean[pos] * para->getParH(lev)->vz_mean[pos];
+ para->getParH(lev)->vyz[pos] =
+ para->getParH(lev)->vyz[pos] - para->getParH(lev)->vy_mean[pos] * para->getParH(lev)->vz_mean[pos];
}
}
}
@@ -146,7 +146,7 @@ void writeAllTiDatafToFile(Parameter *para, uint timestep)
}
}
-void writeTiStuffToFile(Parameter *para, uint timestep, int sizeOfTiArray, std::vector<real *> &data,
+void writeTiStuffToFile(Parameter *para, uint timestep, unsigned long long sizeOfTiArray, std::vector<real *> &data,
std::vector<std::string> &datanames)
{
////////////////////////////////////////////////////////////////////////
@@ -169,10 +169,10 @@ void writeTiStuffToFile(Parameter *para, uint timestep, int sizeOfTiArray, std::
ostr << std::endl;
////////////////////////////////////////////////////////////////////////
// fill file with data
- for (int i = 0; i < sizeOfTiArray; i++) {
- ostr << i;
+ for (size_t pos = 0; pos < sizeOfTiArray; pos++) {
+ ostr << pos;
for (auto dataset : data)
- ostr << "\t" << dataset[i];
+ ostr << "\t" << dataset[pos];
ostr << std::endl;
}
////////////////////////////////////////////////////////////////////////
diff --git a/src/gpu/VirtualFluids_GPU/Calculation/CalcTurbulenceIntensity.h b/src/gpu/VirtualFluids_GPU/Calculation/CalcTurbulenceIntensity.h
index f70973eb5921a17c3229a026623de2a0ef9f3ce4..a76c2d0dde99ad9fb3fd38137b6c72e5c3f5a6c3 100644
--- a/src/gpu/VirtualFluids_GPU/Calculation/CalcTurbulenceIntensity.h
+++ b/src/gpu/VirtualFluids_GPU/Calculation/CalcTurbulenceIntensity.h
@@ -18,7 +18,7 @@ void writeVeloFluctuationToFile(Parameter *para, uint timeste);
void writeVeloMeansToFile(Parameter *para, uint timestep);
void writeAllTiDatafToFile(Parameter *para, uint timestep);
-void writeTiStuffToFile(Parameter *para, uint timestep, int sizeOfTiArray, std::vector<real *> &data,
+void writeTiStuffToFile(Parameter *para, uint timestep, unsigned long long sizeOfTiArray, std::vector<real *> &data,
std::vector<std::string> &datanames);
#endif
diff --git a/src/gpu/VirtualFluids_GPU/Calculation/CollisisionStrategy.cpp b/src/gpu/VirtualFluids_GPU/Calculation/CollisisionStrategy.cpp
index 4a14d19c10936f84379f332ef24f081f0ebb0cb7..49543f37df7fb54290f4ab6c09edb8d10c0b67be 100644
--- a/src/gpu/VirtualFluids_GPU/Calculation/CollisisionStrategy.cpp
+++ b/src/gpu/VirtualFluids_GPU/Calculation/CollisisionStrategy.cpp
@@ -39,8 +39,14 @@ void CollisionAndExchange_noStreams_indexKernel::operator()(UpdateGrid27 *update
//!
//! 1. run collision
//!
- updateGrid->collisionUsingIndices(level, t, para->getParD(level)->fluidNodeIndices,
- para->getParD(level)->numberOfFluidNodes, -1);
+ for( CollisionTemplate tag: para->getParH(level)->allocatedBulkFluidNodeTags )
+ {
+ updateGrid->collisionUsingIndices( level, t,
+ para->getParD(level)->taggedFluidNodeIndices[tag],
+ para->getParD(level)->numberOfTaggedFluidNodes[tag],
+ tag,
+ CudaStreamIndex::Legacy);
+ }
//! 2. exchange information between GPUs
updateGrid->exchangeMultiGPU_noStreams_withPrepare(level, false);
@@ -61,28 +67,35 @@ void CollisionAndExchange_noStreams_oldKernel::operator()(UpdateGrid27 *updateGr
void CollisionAndExchange_streams::operator()(UpdateGrid27 *updateGrid, Parameter *para, int level, unsigned int t)
{
- int borderStreamIndex = para->getStreamManager()->getBorderStreamIndex();
- int bulkStreamIndex = para->getStreamManager()->getBulkStreamIndex();
-
//! \details steps:
//!
- //! 1. run collision for nodes which are at the border of the gpus/processes
- //!
- updateGrid->collisionUsingIndices(level, t, para->getParD(level)->fluidNodeIndicesBorder,
- para->getParD(level)->numberOfFluidNodesBorder, borderStreamIndex);
+ //! 1. run collision for nodes which are at the border of the gpus/processes, running with WriteMacroVars in case probes sample on these nodes
+ //!
+ updateGrid->collisionUsingIndices( level, t,
+ para->getParD(level)->taggedFluidNodeIndices[CollisionTemplate::SubDomainBorder],
+ para->getParD(level)->numberOfTaggedFluidNodes[CollisionTemplate::SubDomainBorder],
+ CollisionTemplate::WriteMacroVars,
+ CudaStreamIndex::SubDomainBorder);
//! 2. prepare the exchange between gpus (collect the send nodes for communication in a buffer on the gpu) and trigger bulk kernel execution when finished
//!
- updateGrid->prepareExchangeMultiGPU(level, borderStreamIndex);
+ updateGrid->prepareExchangeMultiGPU(level, CudaStreamIndex::SubDomainBorder);
if (para->getUseStreams())
- para->getStreamManager()->triggerStartBulkKernel(borderStreamIndex);
-
- //! 3. launch the collision kernel for bulk nodes
- //!
- para->getStreamManager()->waitOnStartBulkKernelEvent(bulkStreamIndex);
- updateGrid->collisionUsingIndices(level, t, para->getParD(level)->fluidNodeIndices,
- para->getParD(level)->numberOfFluidNodes, bulkStreamIndex);
-
+ para->getStreamManager()->triggerStartBulkKernel(CudaStreamIndex::SubDomainBorder);
+
+ //! 3. launch the collision kernel for bulk nodes. This includes nodes with \param tag Default, WriteMacroVars, ApplyBodyForce,
+ //! or AllFeatures. All assigned tags are listed in \param allocatedBulkFluidNodeTags during initialization in Simulation::init
+
+ para->getStreamManager()->waitOnStartBulkKernelEvent(CudaStreamIndex::Bulk);
+
+ for( CollisionTemplate tag: para->getParH(level)->allocatedBulkFluidNodeTags )
+ {
+ updateGrid->collisionUsingIndices( level, t,
+ para->getParD(level)->taggedFluidNodeIndices[tag],
+ para->getParD(level)->numberOfTaggedFluidNodes[tag],
+ tag,
+ CudaStreamIndex::Bulk);
+ }
//! 4. exchange information between GPUs
- updateGrid->exchangeMultiGPU(level, borderStreamIndex);
+ updateGrid->exchangeMultiGPU(level, CudaStreamIndex::SubDomainBorder);
}
diff --git a/src/gpu/VirtualFluids_GPU/Calculation/ForceCalculations.cpp b/src/gpu/VirtualFluids_GPU/Calculation/ForceCalculations.cpp
index d62e8fee24dad1cde7ccd2044a5a5f9573f7ff82..cc1d2eb748b01835b46f5fc69f47ed3ddc17a28d 100644
--- a/src/gpu/VirtualFluids_GPU/Calculation/ForceCalculations.cpp
+++ b/src/gpu/VirtualFluids_GPU/Calculation/ForceCalculations.cpp
@@ -53,7 +53,7 @@ void ForceCalculations::calcPIDControllerForForce(Parameter* para, CudaMemoryMan
{
//////////////////////////////////////////////////////////////////////
//measure the velocity
- int numberOfElements = para->getParH(lev)->numberOfNodes;
+ unsigned long long numberOfElements = para->getParH(lev)->numberOfNodes;
if (numberOfElements > 0)
{
CalcMacCompSP27(para->getParD(lev)->velocityX,
@@ -74,11 +74,11 @@ void ForceCalculations::calcPIDControllerForForce(Parameter* para, CudaMemoryMan
cudaMemoryManager->cudaCopyPrint(lev);
// para->cudaCopyForceVelo(i,numberOfElements);
//////////////////////////////////////////////////////////////////
- for (int j = 0; j < numberOfElements; j++)
+ for (size_t pos = 0; pos < numberOfElements; pos++)
{
- tempVeloX += (double)para->getParH(lev)->velocityX[j];
- tempVeloY += (double)para->getParH(lev)->velocityY[j];
- tempVeloZ += (double)para->getParH(lev)->velocityZ[j];
+ tempVeloX += (double)para->getParH(lev)->velocityX[pos];
+ tempVeloY += (double)para->getParH(lev)->velocityY[pos];
+ tempVeloZ += (double)para->getParH(lev)->velocityZ[pos];
}
tempVeloX /= (double)numberOfElements;
tempVeloY /= (double)numberOfElements;
diff --git a/src/gpu/VirtualFluids_GPU/Calculation/RefinementStrategy.cpp b/src/gpu/VirtualFluids_GPU/Calculation/RefinementStrategy.cpp
index cd74216e1fbe7b718c72046ace4b7d2e7cf451fe..b8ca4e9c2020e17cd0192267ac5d931b510afc3a 100644
--- a/src/gpu/VirtualFluids_GPU/Calculation/RefinementStrategy.cpp
+++ b/src/gpu/VirtualFluids_GPU/Calculation/RefinementStrategy.cpp
@@ -38,67 +38,62 @@ void NoRefinement::operator()(UpdateGrid27 *updateGrid, Parameter *para, int lev
void RefinementAndExchange_streams_exchangeInterface::operator()(UpdateGrid27 *updateGrid, Parameter *para, int level)
{
- int borderStreamIndex = para->getStreamManager()->getBorderStreamIndex();
- int bulkStreamIndex = para->getStreamManager()->getBulkStreamIndex();
-
//! \details steps:
//!
//! 1. Interpolation fine to coarse for nodes which are at the border of the gpus/processes
//!
- updateGrid->fineToCoarse(level, ¶->getParD(level)->intFCBorder, para->getParD(level)->offFC, borderStreamIndex);
+ updateGrid->fineToCoarse(level, ¶->getParD(level)->intFCBorder, para->getParD(level)->offFC, CudaStreamIndex::SubDomainBorder);
//! 2. prepare the exchange between gpus (collect the send nodes for communication in a buffer on the gpu) and trigger bulk kernel execution when finished
//!
- updateGrid->prepareExchangeMultiGPUAfterFtoC(level, borderStreamIndex);
+ updateGrid->prepareExchangeMultiGPUAfterFtoC(level, CudaStreamIndex::SubDomainBorder);
if (para->getUseStreams())
- para->getStreamManager()->triggerStartBulkKernel(borderStreamIndex);
+ para->getStreamManager()->triggerStartBulkKernel(CudaStreamIndex::SubDomainBorder);
//! 3. launch the bulk kernels for both interpolation processes (fine to coarse and coarse to fine)
//!
- para->getStreamManager()->waitOnStartBulkKernelEvent(bulkStreamIndex);
- updateGrid->fineToCoarse(level, ¶->getParD(level)->intFCBulk, para->getParD(level)->offFCBulk, bulkStreamIndex);
- updateGrid->coarseToFine(level, ¶->getParD(level)->intCFBulk, para->getParD(level)->offCFBulk, bulkStreamIndex);
+ para->getStreamManager()->waitOnStartBulkKernelEvent(CudaStreamIndex::Bulk);
+ updateGrid->fineToCoarse(level, ¶->getParD(level)->intFCBulk, para->getParD(level)->offFCBulk, CudaStreamIndex::SubDomainBorder);
+ updateGrid->coarseToFine(level, ¶->getParD(level)->intCFBulk, para->getParD(level)->offCFBulk, CudaStreamIndex::SubDomainBorder);
//! 4. exchange information between GPUs (only nodes which are part of the interpolation)
//!
- updateGrid->exchangeMultiGPUAfterFtoC(level, borderStreamIndex);
+ updateGrid->exchangeMultiGPUAfterFtoC(level, CudaStreamIndex::SubDomainBorder);
// 5. interpolation fine to coarse for nodes which are at the border of the gpus/processes
//!
- updateGrid->coarseToFine(level, ¶->getParD(level)->intCFBorder, para->getParD(level)->offCF, borderStreamIndex);
+ updateGrid->coarseToFine(level, ¶->getParD(level)->intCFBorder, para->getParD(level)->offCF, CudaStreamIndex::SubDomainBorder);
cudaDeviceSynchronize();
}
-void RefinementAndExchange_streams_exchangeAllNodes::operator()(UpdateGrid27 *updateGrid, Parameter *para, int level){
- int borderStreamIndex = para->getStreamManager()->getBorderStreamIndex();
- int bulkStreamIndex = para->getStreamManager()->getBulkStreamIndex();
-
+void RefinementAndExchange_streams_exchangeAllNodes::operator()(UpdateGrid27 *updateGrid, Parameter *para, int level)
+{
//! \details steps:
//!
//! 1. interpolation fine to coarse for nodes which are at the border of the gpus/processes
//!
- updateGrid->fineToCoarse(level, ¶->getParD(level)->intFCBorder, para->getParD(level)->offFC, borderStreamIndex);
+ updateGrid->fineToCoarse(level, ¶->getParD(level)->intFCBorder, para->getParD(level)->offFC, CudaStreamIndex::SubDomainBorder);
//! 2. prepare the exchange between gpus (collect the send nodes for communication in a buffer on the gpu) and trigger bulk kernel execution when finished
//!
- updateGrid->prepareExchangeMultiGPU(level, borderStreamIndex);
+ updateGrid->prepareExchangeMultiGPU(level, CudaStreamIndex::SubDomainBorder);
if (para->getUseStreams())
- para->getStreamManager()->triggerStartBulkKernel(borderStreamIndex);
+ para->getStreamManager()->triggerStartBulkKernel(CudaStreamIndex::SubDomainBorder);
//! 3. launch the bulk kernels for both interpolation processes (fine to coarse and coarse to fine)
//!
- para->getStreamManager()->waitOnStartBulkKernelEvent(bulkStreamIndex);
- updateGrid->fineToCoarse(level, ¶->getParD(level)->intFCBulk, para->getParD(level)->offFCBulk, bulkStreamIndex);
- updateGrid->coarseToFine(level, ¶->getParD(level)->intCFBulk, para->getParD(level)->offCFBulk, bulkStreamIndex);
+ para->getStreamManager()->waitOnStartBulkKernelEvent(CudaStreamIndex::Bulk);
+ updateGrid->fineToCoarse(level, ¶->getParD(level)->intFCBulk, para->getParD(level)->offFCBulk, CudaStreamIndex::SubDomainBorder);
+ updateGrid->coarseToFine(level, ¶->getParD(level)->intCFBulk, para->getParD(level)->offCFBulk, CudaStreamIndex::SubDomainBorder);
//! 4. exchange information between GPUs (all nodes)
//!
- updateGrid->exchangeMultiGPU(level, borderStreamIndex);
+ updateGrid->exchangeMultiGPU(level, CudaStreamIndex::SubDomainBorder);
// 5. interpolation fine to coarse for nodes which are at the border of the gpus/processes
//!
- updateGrid->coarseToFine(level, ¶->getParD(level)->intCFBorder, para->getParD(level)->offCF, borderStreamIndex);
+ updateGrid->coarseToFine(level, ¶->getParD(level)->intCFBorder, para->getParD(level)->offCF, CudaStreamIndex::SubDomainBorder);
cudaDeviceSynchronize();
}
@@ -109,14 +104,14 @@ void RefinementAndExchange_noStreams_exchangeInterface::operator()(UpdateGrid27
//!
//! 1. interpolation fine to coarse
//!
- updateGrid->fineToCoarse(level, ¶->getParD(level)->intFC, para->getParD(level)->offFC, -1);
+ updateGrid->fineToCoarse(level, ¶->getParD(level)->intFC, para->getParD(level)->offFC, CudaStreamIndex::Legacy);
//! 2. exchange information between GPUs (only nodes which are part of the interpolation)
//!
updateGrid->exchangeMultiGPU_noStreams_withPrepare(level, true);
//! 3. interpolation coarse to fine
- updateGrid->coarseToFine(level, ¶->getParD(level)->intCF, para->getParD(level)->offCF, -1);
+ updateGrid->coarseToFine(level, ¶->getParD(level)->intCF, para->getParD(level)->offCF, CudaStreamIndex::Legacy);
}
void RefinementAndExchange_noStreams_exchangeAllNodes::operator()(UpdateGrid27 *updateGrid, Parameter *para, int level)
@@ -125,14 +120,14 @@ void RefinementAndExchange_noStreams_exchangeAllNodes::operator()(UpdateGrid27 *
//!
//! 1. interpolation fine to coarse
//!
- updateGrid->fineToCoarse(level, ¶->getParD(level)->intFC, para->getParD(level)->offFC, -1);
+ updateGrid->fineToCoarse(level, ¶->getParD(level)->intFC, para->getParD(level)->offFC, CudaStreamIndex::Legacy);
//! 2. exchange information between GPUs (all nodes)
//!
updateGrid->exchangeMultiGPU_noStreams_withPrepare(level, false);
//! 3. interpolation coarse to fine
- updateGrid->coarseToFine(level, ¶->getParD(level)->intCF, para->getParD(level)->offCF, -1);
+ updateGrid->coarseToFine(level, ¶->getParD(level)->intCF, para->getParD(level)->offCF, CudaStreamIndex::Legacy);
}
void Refinement_noExchange::operator()(UpdateGrid27 *updateGrid, Parameter *para, int level)
@@ -141,7 +136,7 @@ void Refinement_noExchange::operator()(UpdateGrid27 *updateGrid, Parameter *para
//!
//! 1. interpolation fine to coarse
//!
- updateGrid->fineToCoarse(level, ¶->getParD(level)->intFC, para->getParD(level)->offFC, -1);
+ updateGrid->fineToCoarse(level, ¶->getParD(level)->intFC, para->getParD(level)->offFC, CudaStreamIndex::Legacy);
//! 2. interpolation coarse to fine
- updateGrid->coarseToFine(level, ¶->getParD(level)->intCF, para->getParD(level)->offCF, -1);
+ updateGrid->coarseToFine(level, ¶->getParD(level)->intCF, para->getParD(level)->offCF, CudaStreamIndex::Legacy);
}
diff --git a/src/gpu/VirtualFluids_GPU/Calculation/UpdateGrid27.cpp b/src/gpu/VirtualFluids_GPU/Calculation/UpdateGrid27.cpp
index 296ab819c5538a6b6d6a6827b5c28cbc475af838..4136614dfbfc9e0d2fc1bf7f4b01624f94eabb6f 100644
--- a/src/gpu/VirtualFluids_GPU/Calculation/UpdateGrid27.cpp
+++ b/src/gpu/VirtualFluids_GPU/Calculation/UpdateGrid27.cpp
@@ -22,13 +22,17 @@ void UpdateGrid27::updateGrid(int level, unsigned int t)
updateGrid(level + 1, t);
}
+ //////////////////////////////////////////////////////////////////////////
+
+ interactWithProbes(level, t);
+
//////////////////////////////////////////////////////////////////////////
collision(this, para.get(), level, t);
//////////////////////////////////////////////////////////////////////////
- postCollisionBC(level);
+ postCollisionBC(level, t);
//////////////////////////////////////////////////////////////////////////
@@ -47,13 +51,14 @@ void UpdateGrid27::updateGrid(int level, unsigned int t)
//////////////////////////////////////////////////////////////////////////
if( level != para->getFine() )
- {
+ {
refinement(this, para.get(), level);
}
+ //////////////////////////////////////////////////////////////////////////
+
interactWithActuators(level, t);
- interactWithProbes(level, t);
}
void UpdateGrid27::collisionAllNodes(int level, unsigned int t)
@@ -71,15 +76,16 @@ void UpdateGrid27::collisionAllNodes(int level, unsigned int t)
collisionAdvectionDiffusion(level);
}
-void UpdateGrid27::collisionUsingIndices(int level, unsigned int t, uint *fluidNodeIndices, uint numberOfFluidNodes, int stream)
+void UpdateGrid27::collisionUsingIndices(int level, unsigned int t, uint *taggedFluidNodeIndices, uint numberOfTaggedFluidNodes, CollisionTemplate collisionTemplate, CudaStreamIndex stream)
{
- if (fluidNodeIndices != nullptr && numberOfFluidNodes != 0)
- kernels.at(level)->runOnIndices(fluidNodeIndices, numberOfFluidNodes, stream);
+ if (taggedFluidNodeIndices != nullptr && numberOfTaggedFluidNodes != 0)
+ kernels.at(level)->runOnIndices(taggedFluidNodeIndices, numberOfTaggedFluidNodes, collisionTemplate, stream);
else
- std::cout << "In collision: fluidNodeIndices or numberOfFluidNodes not definded"
+ std::cout << "In collision: fluidNodeIndices or numberOfFluidNodes not defined"
<< std::endl;
//////////////////////////////////////////////////////////////////////////
+ //! \todo: AD collision and porousMedia should be called separately, not in collisionUsingIndices
if (para->getSimulatePorousMedia())
collisionPorousMedia(level);
@@ -118,21 +124,21 @@ void UpdateGrid27::collisionAdvectionDiffusion(int level)
this->adKernelManager->runADcollisionKernel(level);
}
-void UpdateGrid27::prepareExchangeMultiGPU(int level, int streamIndex)
+void UpdateGrid27::prepareExchangeMultiGPU(int level, CudaStreamIndex streamIndex)
{
prepareExchangeCollDataXGPU27AllNodes(para.get(), level, streamIndex);
prepareExchangeCollDataYGPU27AllNodes(para.get(), level, streamIndex);
prepareExchangeCollDataZGPU27AllNodes(para.get(), level, streamIndex);
}
-void UpdateGrid27::prepareExchangeMultiGPUAfterFtoC(int level, int streamIndex)
+void UpdateGrid27::prepareExchangeMultiGPUAfterFtoC(int level, CudaStreamIndex streamIndex)
{
prepareExchangeCollDataXGPU27AfterFtoC(para.get(), level, streamIndex);
prepareExchangeCollDataYGPU27AfterFtoC(para.get(), level, streamIndex);
prepareExchangeCollDataZGPU27AfterFtoC(para.get(), level, streamIndex);
}
-void UpdateGrid27::exchangeMultiGPU(int level, int streamIndex)
+void UpdateGrid27::exchangeMultiGPU(int level, CudaStreamIndex streamIndex)
{
//////////////////////////////////////////////////////////////////////////
// 3D domain decomposition
@@ -168,30 +174,30 @@ void UpdateGrid27::exchangeMultiGPU_noStreams_withPrepare(int level, bool useRed
// 3D domain decomposition
if (useReducedComm) {
// X
- prepareExchangeCollDataXGPU27AfterFtoC(para.get(), level, -1);
- exchangeCollDataXGPU27AfterFtoC(para.get(), comm, cudaMemoryManager.get(), level, -1);
- scatterNodesFromRecvBufferXGPU27AfterFtoC(para.get(), level, -1);
+ prepareExchangeCollDataXGPU27AfterFtoC(para.get(), level, CudaStreamIndex::Legacy);
+ exchangeCollDataXGPU27AfterFtoC(para.get(), comm, cudaMemoryManager.get(), level, CudaStreamIndex::Legacy);
+ scatterNodesFromRecvBufferXGPU27AfterFtoC(para.get(), level, CudaStreamIndex::Legacy);
// Y
- prepareExchangeCollDataYGPU27AfterFtoC(para.get(), level, -1);
- exchangeCollDataYGPU27AfterFtoC(para.get(), comm, cudaMemoryManager.get(), level, -1);
- scatterNodesFromRecvBufferYGPU27AfterFtoC(para.get(), level, -1);
+ prepareExchangeCollDataYGPU27AfterFtoC(para.get(), level, CudaStreamIndex::Legacy);
+ exchangeCollDataYGPU27AfterFtoC(para.get(), comm, cudaMemoryManager.get(), level, CudaStreamIndex::Legacy);
+ scatterNodesFromRecvBufferYGPU27AfterFtoC(para.get(), level, CudaStreamIndex::Legacy);
// Z
- prepareExchangeCollDataZGPU27AfterFtoC(para.get(), level, -1);
- exchangeCollDataZGPU27AfterFtoC(para.get(), comm, cudaMemoryManager.get(), level, -1);
- scatterNodesFromRecvBufferZGPU27AfterFtoC(para.get(), level, -1);
+ prepareExchangeCollDataZGPU27AfterFtoC(para.get(), level, CudaStreamIndex::Legacy);
+ exchangeCollDataZGPU27AfterFtoC(para.get(), comm, cudaMemoryManager.get(), level, CudaStreamIndex::Legacy);
+ scatterNodesFromRecvBufferZGPU27AfterFtoC(para.get(), level, CudaStreamIndex::Legacy);
} else {
// X
- prepareExchangeCollDataXGPU27AllNodes(para.get(), level, -1);
- exchangeCollDataXGPU27AllNodes(para.get(), comm, cudaMemoryManager.get(), level, -1);
- scatterNodesFromRecvBufferXGPU27AllNodes(para.get(), level, -1);
+ prepareExchangeCollDataXGPU27AllNodes(para.get(), level, CudaStreamIndex::Legacy);
+ exchangeCollDataXGPU27AllNodes(para.get(), comm, cudaMemoryManager.get(), level, CudaStreamIndex::Legacy);
+ scatterNodesFromRecvBufferXGPU27AllNodes(para.get(), level, CudaStreamIndex::Legacy);
// Y
- prepareExchangeCollDataYGPU27AllNodes(para.get(), level, -1);
- exchangeCollDataYGPU27AllNodes(para.get(), comm, cudaMemoryManager.get(), level, -1);
- scatterNodesFromRecvBufferYGPU27AllNodes(para.get(), level, -1);
+ prepareExchangeCollDataYGPU27AllNodes(para.get(), level, CudaStreamIndex::Legacy);
+ exchangeCollDataYGPU27AllNodes(para.get(), comm, cudaMemoryManager.get(), level, CudaStreamIndex::Legacy);
+ scatterNodesFromRecvBufferYGPU27AllNodes(para.get(), level, CudaStreamIndex::Legacy);
// Z
- prepareExchangeCollDataZGPU27AllNodes(para.get(), level, -1);
- exchangeCollDataZGPU27AllNodes(para.get(), comm, cudaMemoryManager.get(), level, -1);
- scatterNodesFromRecvBufferZGPU27AllNodes(para.get(), level, -1);
+ prepareExchangeCollDataZGPU27AllNodes(para.get(), level, CudaStreamIndex::Legacy);
+ exchangeCollDataZGPU27AllNodes(para.get(), comm, cudaMemoryManager.get(), level, CudaStreamIndex::Legacy);
+ scatterNodesFromRecvBufferZGPU27AllNodes(para.get(), level, CudaStreamIndex::Legacy);
}
//////////////////////////////////////////////////////////////////////////
@@ -204,7 +210,7 @@ void UpdateGrid27::exchangeMultiGPU_noStreams_withPrepare(int level, bool useRed
exchangePostCollDataADZGPU27(para.get(), comm, cudaMemoryManager.get(), level);
}
}
-void UpdateGrid27::exchangeMultiGPUAfterFtoC(int level, int streamIndex)
+void UpdateGrid27::exchangeMultiGPUAfterFtoC(int level, CudaStreamIndex streamIndex)
{
//////////////////////////////////////////////////////////////////////////
// 3D domain decomposition
@@ -227,9 +233,10 @@ void UpdateGrid27::exchangeMultiGPUAfterFtoC(int level, int streamIndex)
}
}
-void UpdateGrid27::postCollisionBC(int level)
+void UpdateGrid27::postCollisionBC(int level, uint t)
{
//////////////////////////////////////////////////////////////////////////
+ // G E O M E T R Y
// V E L O C I T Y (I N F L O W)
this->bcKernelManager->runVelocityBCKernelPost(level);
@@ -257,6 +264,10 @@ void UpdateGrid27::postCollisionBC(int level)
// P R E S S U R E
this->bcKernelManager->runPressureBCKernelPost(level);
+ //////////////////////////////////////////////////////////////////////////
+ // P R E C U R S O R
+ this->bcKernelManager->runPrecursorBCKernelPost(level, t, cudaMemoryManager.get());
+
//////////////////////////////////////////////////////////////////////////
// A D V E C T I O N D I F F U S I O N
if (para->getDiffOn())
@@ -317,13 +328,12 @@ void UpdateGrid27::preCollisionBC(int level, unsigned int t)
//////////////////////////////////////////////////////////////////////////////////
}
-void UpdateGrid27::fineToCoarse(int level, InterpolationCellFC* icellFC, OffFC &offFC,
- int streamIndex)
+void UpdateGrid27::fineToCoarse(int level, InterpolationCellFC* icellFC, OffFC &offFC, CudaStreamIndex streamIndex)
{
gridScalingKernelManager->runFineToCoarseKernelLB(level, icellFC, offFC, streamIndex);
if (para->getDiffOn()) {
- if (streamIndex != -1) {
+ if (para->getStreamManager()->streamIsRegistered(streamIndex)) {
printf("fineToCoarse Advection Diffusion not implemented"); // TODO
return;
}
@@ -331,14 +341,13 @@ void UpdateGrid27::fineToCoarse(int level, InterpolationCellFC* icellFC, OffFC &
}
}
-void UpdateGrid27::coarseToFine(int level, InterpolationCellCF* icellCF, OffCF &offCF,
- int streamIndex)
+void UpdateGrid27::coarseToFine(int level, InterpolationCellCF* icellCF, OffCF &offCF, CudaStreamIndex streamIndex)
{
this->gridScalingKernelManager->runCoarseToFineKernelLB(level, icellCF, offCF, streamIndex);
if (para->getDiffOn())
{
- if (streamIndex != -1){
+ if(para->getStreamManager()->streamIsRegistered(streamIndex)){
printf("CoarseToFineWithStream Advection Diffusion not implemented"); // TODO
return;
}
diff --git a/src/gpu/VirtualFluids_GPU/Calculation/UpdateGrid27.h b/src/gpu/VirtualFluids_GPU/Calculation/UpdateGrid27.h
index 8110923bf066412e2bb09ffa1f10efe3ddc983c7..8ce2cf5bfd72f9f53cdb35bc92502ee9ca0d3ad8 100644
--- a/src/gpu/VirtualFluids_GPU/Calculation/UpdateGrid27.h
+++ b/src/gpu/VirtualFluids_GPU/Calculation/UpdateGrid27.h
@@ -4,6 +4,7 @@
#include "LBM/LB.h"
#include "GPU/GPU_Interface.h"
#include "Parameter/Parameter.h"
+#include "Parameter/CudaStreamManager.h"
#include "GPU/CudaMemoryManager.h"
#include "Communication/Communicator.h"
#include "Calculation/PorousMedia.h"
@@ -15,7 +16,6 @@ class Kernel;
class BoundaryConditionFactory;
class GridScalingFactory;
class TurbulenceModelFactory;
-
class UpdateGrid27;
using CollisionStrategy = std::function<void (UpdateGrid27* updateGrid, Parameter* para, int level, unsigned int t)>;
using RefinementStrategy = std::function<void (UpdateGrid27* updateGrid, Parameter* para, int level)>;
@@ -31,21 +31,21 @@ public:
private:
void collisionAllNodes(int level, unsigned int t);
- void collisionUsingIndices(int level, unsigned int t, uint *fluidNodeIndices = nullptr, uint numberOfFluidNodes = 0, int stream = -1);
+ void collisionUsingIndices(int level, unsigned int t, uint *taggedFluidNodeIndices = nullptr, uint numberOfTaggedFluidNodes = 0, CollisionTemplate collisionTemplate = CollisionTemplate::Default, CudaStreamIndex streamIndex=CudaStreamIndex::Legacy);
void collisionAdvectionDiffusion(int level);
- void postCollisionBC(int level);
+ void postCollisionBC(int level, unsigned int t);
void preCollisionBC(int level, unsigned int t);
void collisionPorousMedia(int level);
- void fineToCoarse(int level, InterpolationCellFC* icellFC, OffFC &offFC, int streamIndex);
- void coarseToFine(int level, InterpolationCellCF* icellCF, OffCF &offCF, int streamIndex);
+ void fineToCoarse(int level, InterpolationCellFC* icellFC, OffFC &offFC, CudaStreamIndex streamIndex);
+ void coarseToFine(int level, InterpolationCellCF* icellCF, OffCF &offCF, CudaStreamIndex streamIndex);
- void prepareExchangeMultiGPU(int level, int streamIndex);
- void prepareExchangeMultiGPUAfterFtoC(int level, int streamIndex);
+ void prepareExchangeMultiGPU(int level, CudaStreamIndex streamIndex);
+ void prepareExchangeMultiGPUAfterFtoC(int level, CudaStreamIndex streamIndex);
- void exchangeMultiGPU(int level, int streamIndex);
- void exchangeMultiGPUAfterFtoC(int level, int streamIndex);
+ void exchangeMultiGPU(int level, CudaStreamIndex streamIndex);
+ void exchangeMultiGPUAfterFtoC(int level, CudaStreamIndex streamIndex);
void exchangeMultiGPU_noStreams_withPrepare(int level, bool useReducedComm);
void swapBetweenEvenAndOddTimestep(int level);
@@ -60,6 +60,7 @@ private:
friend class CollisionAndExchange_noStreams_indexKernel;
friend class CollisionAndExchange_noStreams_oldKernel;
friend class CollisionAndExchange_streams;
+ friend class CollisionAndExchange_noStreams_withReadWriteFlags;
RefinementStrategy refinement;
friend class RefinementAndExchange_streams_exchangeInterface;
diff --git a/src/gpu/VirtualFluids_GPU/Communication/ExchangeData27.cpp b/src/gpu/VirtualFluids_GPU/Communication/ExchangeData27.cpp
index 36c250401e0775b3abcc7d25c0f89fde0556631e..00a7b45668e2050467f3d1122455dc74d0ad4f1c 100644
--- a/src/gpu/VirtualFluids_GPU/Communication/ExchangeData27.cpp
+++ b/src/gpu/VirtualFluids_GPU/Communication/ExchangeData27.cpp
@@ -11,12 +11,12 @@ using namespace vf::lbm::dir;
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
// 3D domain decomposition: functions used by all directions
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
-void collectNodesInSendBufferGPU(Parameter *para, int level, int streamIndex,
+void collectNodesInSendBufferGPU(Parameter *para, int level, CudaStreamIndex streamIndex,
std::vector<ProcessNeighbor27> *sendProcessNeighbor,
unsigned int numberOfSendProcessNeighbors)
{
- cudaStream_t stream = (streamIndex == -1) ? CU_STREAM_LEGACY : para->getStreamManager()->getStream(streamIndex);
-
+ cudaStream_t stream = para->getStreamManager()->getStream(streamIndex);
+
for (unsigned int i = 0; i < numberOfSendProcessNeighbors; i++) {
GetSendFsPostDev27(para->getParD(level)->distributions.f[0],
(*sendProcessNeighbor)[i].f[0],
@@ -32,11 +32,11 @@ void collectNodesInSendBufferGPU(Parameter *para, int level, int streamIndex,
}
}
-void scatterNodesFromRecvBufferGPU(Parameter *para, int level, int streamIndex,
+void scatterNodesFromRecvBufferGPU(Parameter *para, int level, CudaStreamIndex streamIndex,
std::vector<ProcessNeighbor27> *recvProcessNeighborDev,
unsigned int numberOfRecvProcessNeighbors)
{
- cudaStream_t stream = (streamIndex == -1) ? CU_STREAM_LEGACY : para->getStreamManager()->getStream(streamIndex);
+ cudaStream_t stream = para->getStreamManager()->getStream(streamIndex);
for (unsigned int i = 0; i < numberOfRecvProcessNeighbors; i++) {
SetRecvFsPostDev27(para->getParD(level)->distributions.f[0],
(*recvProcessNeighborDev)[i].f[0],
@@ -105,22 +105,22 @@ void copyEdgeNodes(std::vector<LBMSimulationParameter::EdgeNodePositions> &edgeN
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
// X
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
-void prepareExchangeCollDataXGPU27AllNodes(Parameter *para, int level, int streamIndex)
+void prepareExchangeCollDataXGPU27AllNodes(Parameter *para, int level, CudaStreamIndex streamIndex)
{
collectNodesInSendBufferGPU(para, level, streamIndex, ¶->getParD(level)->sendProcessNeighborX,
(unsigned int)(para->getNumberOfProcessNeighborsX(level, "send")));
}
-void prepareExchangeCollDataXGPU27AfterFtoC(Parameter *para, int level, int streamIndex)
+void prepareExchangeCollDataXGPU27AfterFtoC(Parameter *para, int level, CudaStreamIndex streamIndex)
{
collectNodesInSendBufferGPU(para, level, streamIndex, ¶->getParD(level)->sendProcessNeighborsAfterFtoCX,
(unsigned int)(para->getNumberOfProcessNeighborsX(level, "send")));
}
void exchangeCollDataXGPU27AllNodes(Parameter *para, vf::gpu::Communicator &comm, CudaMemoryManager *cudaMemoryManager,
- int level, int streamIndex)
+ int level, CudaStreamIndex streamIndex)
{
- exchangeCollDataXGPU27(para, comm, cudaMemoryManager, level, streamIndex,
+ exchangeCollDataXGPU27(para, comm, cudaMemoryManager, level, streamIndex,
¶->getParD(level)->sendProcessNeighborX,
¶->getParD(level)->recvProcessNeighborX,
¶->getParH(level)->sendProcessNeighborX,
@@ -128,40 +128,40 @@ void exchangeCollDataXGPU27AllNodes(Parameter *para, vf::gpu::Communicator &comm
}
void exchangeCollDataXGPU27AfterFtoC(Parameter *para, vf::gpu::Communicator &comm, CudaMemoryManager *cudaMemoryManager,
- int level, int streamIndex)
+ int level, CudaStreamIndex streamIndex)
{
- exchangeCollDataXGPU27(para, comm, cudaMemoryManager, level, streamIndex,
+ exchangeCollDataXGPU27(para, comm, cudaMemoryManager, level, streamIndex,
¶->getParD(level)->sendProcessNeighborsAfterFtoCX,
¶->getParD(level)->recvProcessNeighborsAfterFtoCX,
¶->getParH(level)->sendProcessNeighborsAfterFtoCX,
¶->getParH(level)->recvProcessNeighborsAfterFtoCX);
}
-void scatterNodesFromRecvBufferXGPU27AllNodes(Parameter *para, int level, int streamIndex)
+void scatterNodesFromRecvBufferXGPU27AllNodes(Parameter *para, int level, CudaStreamIndex streamIndex)
{
- scatterNodesFromRecvBufferGPU(para, level, streamIndex, ¶->getParD(level)->recvProcessNeighborX,
+ scatterNodesFromRecvBufferGPU(para, level, streamIndex,¶->getParD(level)->recvProcessNeighborX,
(unsigned int)(para->getNumberOfProcessNeighborsX(level, "send")));
}
-void scatterNodesFromRecvBufferXGPU27AfterFtoC(Parameter *para, int level, int streamIndex)
+void scatterNodesFromRecvBufferXGPU27AfterFtoC(Parameter *para, int level, CudaStreamIndex streamIndex)
{
- scatterNodesFromRecvBufferGPU(para, level, streamIndex, ¶->getParD(level)->recvProcessNeighborsAfterFtoCX,
+ scatterNodesFromRecvBufferGPU(para, level, streamIndex,¶->getParD(level)->recvProcessNeighborsAfterFtoCX,
(unsigned int)(para->getNumberOfProcessNeighborsX(level, "send")));
}
-void exchangeCollDataXGPU27(Parameter *para, vf::gpu::Communicator &comm, CudaMemoryManager *cudaMemoryManager, int level,
- int streamIndex,
+void exchangeCollDataXGPU27(Parameter *para, vf::gpu::Communicator &comm, CudaMemoryManager *cudaMemoryManager,
+ int level, CudaStreamIndex streamIndex,
std::vector<ProcessNeighbor27> *sendProcessNeighborDev,
std::vector<ProcessNeighbor27> *recvProcessNeighborDev,
std::vector<ProcessNeighbor27> *sendProcessNeighborHost,
std::vector<ProcessNeighbor27> *recvProcessNeighborHost)
{
- cudaStream_t stream = (streamIndex == -1) ? CU_STREAM_LEGACY : para->getStreamManager()->getStream(streamIndex);
+ cudaStream_t stream = para->getStreamManager()->getStream(streamIndex);
///////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
//! \details steps:
//! 1. copy data from device to host
for (unsigned int i = 0; i < (unsigned int)(para->getNumberOfProcessNeighborsX(level, "send")); i++)
- cudaMemoryManager->cudaCopyProcessNeighborXFsDH(level, i, (*sendProcessNeighborDev)[i].memsizeFs, streamIndex);
+ cudaMemoryManager->cudaCopyProcessNeighborXFsDH(level, i, (*sendProcessNeighborDev)[i].memsizeFs);
///////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
//! 2. start non-blocking receive (MPI)
@@ -181,7 +181,7 @@ void exchangeCollDataXGPU27(Parameter *para, vf::gpu::Communicator &comm, CudaMe
///////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
//! 7. copy received data from host to device
for (unsigned int i = 0; i < (unsigned int)(para->getNumberOfProcessNeighborsX(level, "send")); i++)
- cudaMemoryManager->cudaCopyProcessNeighborXFsHD(level, i, (*recvProcessNeighborDev)[i].memsizeFs, streamIndex);
+ cudaMemoryManager->cudaCopyProcessNeighborXFsHD(level, i, (*recvProcessNeighborDev)[i].memsizeFs);
///////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
}
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
@@ -189,22 +189,22 @@ void exchangeCollDataXGPU27(Parameter *para, vf::gpu::Communicator &comm, CudaMe
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
// Y
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
-void prepareExchangeCollDataYGPU27AllNodes(Parameter *para, int level, int streamIndex)
+void prepareExchangeCollDataYGPU27AllNodes(Parameter *para, int level, CudaStreamIndex streamIndex)
{
collectNodesInSendBufferGPU(para, level, streamIndex, ¶->getParD(level)->sendProcessNeighborY,
(unsigned int)(para->getNumberOfProcessNeighborsY(level, "send")));
}
-void prepareExchangeCollDataYGPU27AfterFtoC(Parameter *para, int level, int streamIndex)
+void prepareExchangeCollDataYGPU27AfterFtoC(Parameter *para, int level, CudaStreamIndex streamIndex)
{
collectNodesInSendBufferGPU(para, level, streamIndex, ¶->getParD(level)->sendProcessNeighborsAfterFtoCY,
(unsigned int)(para->getNumberOfProcessNeighborsY(level, "send")));
}
void exchangeCollDataYGPU27AllNodes(Parameter *para, vf::gpu::Communicator &comm, CudaMemoryManager *cudaMemoryManager,
- int level, int streamIndex)
+ int level, CudaStreamIndex streamIndex)
{
- exchangeCollDataYGPU27(para, comm, cudaMemoryManager, level, streamIndex,
+ exchangeCollDataYGPU27(para, comm, cudaMemoryManager, level, streamIndex,
¶->getParD(level)->sendProcessNeighborY,
¶->getParD(level)->recvProcessNeighborY,
¶->getParH(level)->sendProcessNeighborY,
@@ -212,38 +212,39 @@ void exchangeCollDataYGPU27AllNodes(Parameter *para, vf::gpu::Communicator &comm
}
void exchangeCollDataYGPU27AfterFtoC(Parameter *para, vf::gpu::Communicator &comm, CudaMemoryManager *cudaMemoryManager,
- int level, int streamIndex)
+ int level, CudaStreamIndex streamIndex)
{
- exchangeCollDataYGPU27(para, comm, cudaMemoryManager, level, streamIndex,
+ exchangeCollDataYGPU27(para, comm, cudaMemoryManager, level, streamIndex,
¶->getParD(level)->sendProcessNeighborsAfterFtoCY,
¶->getParD(level)->recvProcessNeighborsAfterFtoCY,
¶->getParH(level)->sendProcessNeighborsAfterFtoCY,
¶->getParH(level)->recvProcessNeighborsAfterFtoCY);
}
-void scatterNodesFromRecvBufferYGPU27AllNodes(Parameter *para, int level, int streamIndex)
+void scatterNodesFromRecvBufferYGPU27AllNodes(Parameter *para, int level, CudaStreamIndex streamIndex)
{
scatterNodesFromRecvBufferGPU(para, level, streamIndex, ¶->getParD(level)->recvProcessNeighborY,
(unsigned int)(para->getNumberOfProcessNeighborsY(level, "send")));
}
-void scatterNodesFromRecvBufferYGPU27AfterFtoC(Parameter *para, int level, int streamIndex)
+void scatterNodesFromRecvBufferYGPU27AfterFtoC(Parameter *para, int level, CudaStreamIndex streamIndex)
{
scatterNodesFromRecvBufferGPU(para, level, streamIndex, ¶->getParD(level)->recvProcessNeighborsAfterFtoCY,
(unsigned int)(para->getNumberOfProcessNeighborsY(level, "send")));
}
void exchangeCollDataYGPU27(Parameter *para, vf::gpu::Communicator &comm, CudaMemoryManager *cudaMemoryManager, int level,
- int streamIndex, std::vector<ProcessNeighbor27> *sendProcessNeighborDev,
+ CudaStreamIndex streamIndex,
+ std::vector<ProcessNeighbor27> *sendProcessNeighborDev,
std::vector<ProcessNeighbor27> *recvProcessNeighborDev,
std::vector<ProcessNeighbor27> *sendProcessNeighborHost,
std::vector<ProcessNeighbor27> *recvProcessNeighborHost)
{
- cudaStream_t stream = (streamIndex == -1) ? CU_STREAM_LEGACY : para->getStreamManager()->getStream(streamIndex);
+ cudaStream_t stream = para->getStreamManager()->getStream(CudaStreamIndex::SubDomainBorder);
///////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
// copy Device to Host
for (unsigned int i = 0; i < (unsigned int)(para->getNumberOfProcessNeighborsY(level, "send")); i++)
- cudaMemoryManager->cudaCopyProcessNeighborYFsDH(level, i, (*sendProcessNeighborDev)[i].memsizeFs, streamIndex);
+ cudaMemoryManager->cudaCopyProcessNeighborYFsDH(level, i, (*sendProcessNeighborDev)[i].memsizeFs);
///////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
startNonBlockingMpiReceive((unsigned int)(*sendProcessNeighborHost).size(), comm, recvProcessNeighborHost);
@@ -276,7 +277,7 @@ void exchangeCollDataYGPU27(Parameter *para, vf::gpu::Communicator &comm, CudaMe
///////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
// copy Host to Device
for (unsigned int i = 0; i < (unsigned int)(para->getNumberOfProcessNeighborsY(level, "send")); i++) {
- cudaMemoryManager->cudaCopyProcessNeighborYFsHD(level, i, (*recvProcessNeighborDev)[i].memsizeFs, streamIndex);
+ cudaMemoryManager->cudaCopyProcessNeighborYFsHD(level, i, (*recvProcessNeighborDev)[i].memsizeFs);
}
///////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
}
@@ -285,61 +286,62 @@ void exchangeCollDataYGPU27(Parameter *para, vf::gpu::Communicator &comm, CudaMe
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
// Z
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
-void prepareExchangeCollDataZGPU27AllNodes(Parameter *para, int level, int streamIndex)
+void prepareExchangeCollDataZGPU27AllNodes(Parameter *para, int level, CudaStreamIndex streamIndex)
{
collectNodesInSendBufferGPU(para, level, streamIndex, ¶->getParD(level)->sendProcessNeighborZ,
(unsigned int)(para->getNumberOfProcessNeighborsZ(level, "send")));
}
-void prepareExchangeCollDataZGPU27AfterFtoC(Parameter *para, int level, int streamIndex)
+void prepareExchangeCollDataZGPU27AfterFtoC(Parameter *para, int level, CudaStreamIndex streamIndex)
{
collectNodesInSendBufferGPU(para, level, streamIndex, ¶->getParD(level)->sendProcessNeighborsAfterFtoCZ,
(unsigned int)(para->getNumberOfProcessNeighborsZ(level, "send")));
}
void exchangeCollDataZGPU27AllNodes(Parameter *para, vf::gpu::Communicator &comm, CudaMemoryManager *cudaMemoryManager,
- int level, int streamIndex)
+ int level, CudaStreamIndex streamIndex)
{
- exchangeCollDataZGPU27(para, comm, cudaMemoryManager, level, streamIndex,
+ exchangeCollDataZGPU27(para, comm, cudaMemoryManager, level, streamIndex,
¶->getParD(level)->sendProcessNeighborZ,
¶->getParD(level)->recvProcessNeighborZ,
¶->getParH(level)->sendProcessNeighborZ,
¶->getParH(level)->recvProcessNeighborZ);
}
void exchangeCollDataZGPU27AfterFtoC(Parameter *para, vf::gpu::Communicator &comm, CudaMemoryManager *cudaMemoryManager,
- int level, int streamIndex)
+ int level, CudaStreamIndex streamIndex)
{
- exchangeCollDataZGPU27(para, comm, cudaMemoryManager, level, streamIndex,
+ exchangeCollDataZGPU27(para, comm, cudaMemoryManager, level, streamIndex,
¶->getParD(level)->sendProcessNeighborsAfterFtoCZ,
¶->getParD(level)->recvProcessNeighborsAfterFtoCZ,
¶->getParH(level)->sendProcessNeighborsAfterFtoCZ,
¶->getParH(level)->recvProcessNeighborsAfterFtoCZ);
}
-void scatterNodesFromRecvBufferZGPU27AllNodes(Parameter *para, int level, int streamIndex)
+void scatterNodesFromRecvBufferZGPU27AllNodes(Parameter *para, int level, CudaStreamIndex streamIndex)
{
scatterNodesFromRecvBufferGPU(para, level, streamIndex, ¶->getParD(level)->recvProcessNeighborZ,
(unsigned int)(para->getNumberOfProcessNeighborsZ(level, "send")));
}
-void scatterNodesFromRecvBufferZGPU27AfterFtoC(Parameter *para, int level, int streamIndex)
+void scatterNodesFromRecvBufferZGPU27AfterFtoC(Parameter *para, int level, CudaStreamIndex streamIndex)
{
scatterNodesFromRecvBufferGPU(para, level, streamIndex, ¶->getParD(level)->recvProcessNeighborsAfterFtoCZ,
(unsigned int)(para->getNumberOfProcessNeighborsZ(level, "send")));
}
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
-void exchangeCollDataZGPU27(Parameter *para, vf::gpu::Communicator &comm, CudaMemoryManager *cudaMemoryManager, int level,
- int streamIndex, std::vector<ProcessNeighbor27> *sendProcessNeighborDev,
+void exchangeCollDataZGPU27(Parameter *para, vf::gpu::Communicator &comm, CudaMemoryManager *cudaMemoryManager, int level,
+ CudaStreamIndex streamIndex,
+ std::vector<ProcessNeighbor27> *sendProcessNeighborDev,
std::vector<ProcessNeighbor27> *recvProcessNeighborDev,
std::vector<ProcessNeighbor27> *sendProcessNeighborHost,
std::vector<ProcessNeighbor27> *recvProcessNeighborHost)
{
- cudaStream_t stream = (streamIndex == -1) ? CU_STREAM_LEGACY : para->getStreamManager()->getStream(streamIndex);
+ cudaStream_t stream = para->getStreamManager()->getStream(streamIndex);
///////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
// copy Device to Host
for (unsigned int i = 0; i < (unsigned int)(para->getNumberOfProcessNeighborsZ(level, "send")); i++)
- cudaMemoryManager->cudaCopyProcessNeighborZFsDH(level, i, (*sendProcessNeighborDev)[i].memsizeFs, streamIndex);
+ cudaMemoryManager->cudaCopyProcessNeighborZFsDH(level, i, (*sendProcessNeighborDev)[i].memsizeFs);
///////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
startNonBlockingMpiReceive((unsigned int)(*sendProcessNeighborHost).size(), comm, recvProcessNeighborHost);
/////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
@@ -386,7 +388,7 @@ void exchangeCollDataZGPU27(Parameter *para, vf::gpu::Communicator &comm, CudaMe
// copy Host to Device
for (unsigned int i = 0; i < (unsigned int)(para->getNumberOfProcessNeighborsZ(level, "send")); i++)
{
- cudaMemoryManager->cudaCopyProcessNeighborZFsHD(level, i, (*recvProcessNeighborDev)[i].memsizeFs, streamIndex);
+ cudaMemoryManager->cudaCopyProcessNeighborZFsHD(level, i, (*recvProcessNeighborDev)[i].memsizeFs);
}
///////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
}
diff --git a/src/gpu/VirtualFluids_GPU/Communication/ExchangeData27.h b/src/gpu/VirtualFluids_GPU/Communication/ExchangeData27.h
index ec930ebbc06554e948204b74e79e0e25b85f57b5..8302ffdc47bfa012c47df00f90c2491039f4eaee 100644
--- a/src/gpu/VirtualFluids_GPU/Communication/ExchangeData27.h
+++ b/src/gpu/VirtualFluids_GPU/Communication/ExchangeData27.h
@@ -6,6 +6,7 @@
#include "GPU/GPU_Interface.h"
#include "LBM/LB.h"
#include "Parameter/Parameter.h"
+#include "Parameter/CudaStreamManager.h"
//! \file ExchangeData27.h
//! \ingroup GPU
@@ -14,9 +15,9 @@
//////////////////////////////////////////////////////////////////////////
// 1D domain decomposition
-void exchangePreCollDataGPU27(Parameter *para, vf::gpu::Communicator &comm, CudaMemoryManager *cudaMemoryManager,
+void exchangePreCollDataGPU27(Parameter *para, vf::gpu::Communicator &comm, CudaMemoryManager *cudaMemoryManager,
int level);
-void exchangePostCollDataGPU27(Parameter *para, vf::gpu::Communicator &comm, CudaMemoryManager *cudaMemoryManager,
+void exchangePostCollDataGPU27(Parameter *para, vf::gpu::Communicator &comm, CudaMemoryManager *cudaMemoryManager,
int level);
//////////////////////////////////////////////////////////////////////////
// 3D domain decomposition
@@ -24,13 +25,13 @@ void exchangePostCollDataGPU27(Parameter *para, vf::gpu::Communicator &comm, Cud
// functions used for all directions
//! \brief Collect the send nodes in a buffer on the gpu
-void collectNodesInSendBufferGPU(Parameter *para, int level, int streamIndex,
- std::vector<ProcessNeighbor27> *sendProcessNeighbor,
- unsigned int numberOfSendProcessNeighbors);
+void collectNodesInSendBufferGPU(Parameter *para, int level, CudaStreamIndex streamIndex,
+ std::vector<ProcessNeighbor27> *sendProcessNeighbor,
+ unsigned int numberOfSendProcessNeighbors);
//! \brief Distribute the receive nodes from the buffer on the gpu
-void scatterNodesFromRecvBufferGPU(Parameter *para, int level, int streamIndex,
- std::vector<ProcessNeighbor27> *recvProcessNeighborDev,
- unsigned int numberOfRecvProcessNeighbors);
+void scatterNodesFromRecvBufferGPU(Parameter *para, int level, CudaStreamIndex streamIndex,
+ std::vector<ProcessNeighbor27> *recvProcessNeighborDev,
+ unsigned int numberOfRecvProcessNeighbors);
//! \brief Copy nodes which are part of the communication in multiple directions
//! \details The nodes are copied from the receive buffer in one direction to the send buffer in another direction. The
//! copy operation is conducted on the cpu.
@@ -49,21 +50,20 @@ void copyEdgeNodes(std::vector<LBMSimulationParameter::EdgeNodePositions> &edgeN
//! \brief Collect the send nodes for communication in the x direction in a buffer on the gpu
//! \details Needed to exchange all nodes, used in the communication after collision step
-void prepareExchangeCollDataXGPU27AllNodes(Parameter *para, int level, int streamIndex);
+void prepareExchangeCollDataXGPU27AllNodes(Parameter *para, int level, CudaStreamIndex streamIndex);
//! \brief Collect the send nodes for communication in the x direction in a buffer on the gpu
//! \details Only exchange nodes which are part of the interpolation process on refined grids. This function is used in
//! the exchange which takes place after the interpolation fine to coarse and before the interpolation coarse to fine.
//! See [master thesis of Anna Wellmann]
-void prepareExchangeCollDataXGPU27AfterFtoC(Parameter *para, int level, int streamIndex);
+void prepareExchangeCollDataXGPU27AfterFtoC(Parameter *para, int level, CudaStreamIndex streamIndex);
//! \brief Exchange routine in x direction for simulations on multiple gpus
//! \details Send and receive the nodes from the communication buffers on the gpus.
//! \param Communicator is needed for the communication between the processes with mpi
//! \param CudaMemoryManager is needed for moving the data between host and device
-//! \param streamIndex is the index of a CUDA Stream, which is needed for communication hiding
//! \param sendProcessNeighborDev, recvProcessNeighborDev, sendProcessNeighborHost, recvProcessNeighborHost are pointers
//! to the send and receive arrays, both on the device and the host
void exchangeCollDataXGPU27(Parameter *para, vf::gpu::Communicator &comm, CudaMemoryManager *cudaMemoryManager,
- int level, int streamIndex,
+ int level, CudaStreamIndex streamIndex,
std::vector<ProcessNeighbor27> *sendProcessNeighborDev,
std::vector<ProcessNeighbor27> *recvProcessNeighborDev,
std::vector<ProcessNeighbor27> *sendProcessNeighborHost,
@@ -71,59 +71,59 @@ void exchangeCollDataXGPU27(Parameter *para, vf::gpu::Communicator &comm, CudaMe
//! \brief Calls exchangeCollDataXGPU27() for exchanging all nodes
//! \details Used in the communication after collision step
void exchangeCollDataXGPU27AllNodes(Parameter *para, vf::gpu::Communicator &comm,
- CudaMemoryManager *cudaMemoryManager, int level, int streamIndex);
+ CudaMemoryManager *cudaMemoryManager, int level, CudaStreamIndex streamIndex);
//! \brief Calls exchangeCollDataXGPU27() for exchanging the nodes, which are part of the communication between the two
//! interpolation processes on refined grids
//! \details Only exchange nodes which are part of the interpolation process on
//! refined grids. This function is used in the exchange which takes place after the interpolation fine to coarse and
//! before the interpolation coarse to fine. See [master thesis of Anna Wellmann]
void exchangeCollDataXGPU27AfterFtoC(Parameter *para, vf::gpu::Communicator &comm,
- CudaMemoryManager *cudaMemoryManager, int level, int streamIndex);
+ CudaMemoryManager *cudaMemoryManager, int level, CudaStreamIndex streamIndex);
//! \brief Distribute the receive nodes (x direction) from the buffer on the gpu
//! \details Needed to exchange all nodes, used in the communication after collision step
-void scatterNodesFromRecvBufferXGPU27AllNodes(Parameter *para, int level, int streamIndex);
+void scatterNodesFromRecvBufferXGPU27AllNodes(Parameter *para, int level, CudaStreamIndex streamIndex);
//! \brief Distribute the receive nodes (x direction) from the buffer on the gpu
//! \details Only exchange nodes which are part of the interpolation process on refined grids. This function is used in
//! the exchange which takes place after the interpolation fine to coarse and before the interpolation coarse to fine.
//! See [master thesis of Anna Wellmann]
-void scatterNodesFromRecvBufferXGPU27AfterFtoC(Parameter *para, int level, int streamIndex);
+void scatterNodesFromRecvBufferXGPU27AfterFtoC(Parameter *para, int level, CudaStreamIndex streamIndex);
//////////////////////////////////////////////////////////////////////////
// y
-void prepareExchangeCollDataYGPU27AllNodes(Parameter *para, int level, int streamIndex);
-void prepareExchangeCollDataYGPU27AfterFtoC(Parameter *para, int level, int streamIndex);
+void prepareExchangeCollDataYGPU27AllNodes(Parameter *para, int level, CudaStreamIndex streamIndex);
+void prepareExchangeCollDataYGPU27AfterFtoC(Parameter *para, int level, CudaStreamIndex streamIndex);
void exchangeCollDataYGPU27(Parameter *para, vf::gpu::Communicator &comm, CudaMemoryManager *cudaMemoryManager,
- int level, int streamIndex,
+ int level,CudaStreamIndex streamIndex,
std::vector<ProcessNeighbor27> *sendProcessNeighborDev,
std::vector<ProcessNeighbor27> *recvProcessNeighborDev,
std::vector<ProcessNeighbor27> *sendProcessNeighborHost,
std::vector<ProcessNeighbor27> *recvProcessNeighborHos);
void exchangeCollDataYGPU27AllNodes(Parameter *para, vf::gpu::Communicator &comm,
- CudaMemoryManager *cudaMemoryManager, int level, int streamIndex);
+ CudaMemoryManager *cudaMemoryManager, int level, CudaStreamIndex streamIndex);
void exchangeCollDataYGPU27AfterFtoC(Parameter *para, vf::gpu::Communicator &comm,
- CudaMemoryManager *cudaMemoryManager, int level, int streamIndex);
-void scatterNodesFromRecvBufferYGPU27AllNodes(Parameter *para, int level, int streamIndex);
-void scatterNodesFromRecvBufferYGPU27AfterFtoC(Parameter *para, int level, int streamIndex);
+ CudaMemoryManager *cudaMemoryManager, int level, CudaStreamIndex streamIndex);
+void scatterNodesFromRecvBufferYGPU27AllNodes(Parameter *para, int level, CudaStreamIndex streamIndex);
+void scatterNodesFromRecvBufferYGPU27AfterFtoC(Parameter *para, int level, CudaStreamIndex streamIndex);
// z
-void prepareExchangeCollDataZGPU27AllNodes(Parameter *para, int level, int streamIndex);
-void prepareExchangeCollDataZGPU27AfterFtoC(Parameter *para, int level, int streamIndex);
+void prepareExchangeCollDataZGPU27AllNodes(Parameter *para, int level, CudaStreamIndex streamIndex);
+void prepareExchangeCollDataZGPU27AfterFtoC(Parameter *para, int level, CudaStreamIndex streamIndex);
void exchangeCollDataZGPU27(Parameter *para, vf::gpu::Communicator &comm, CudaMemoryManager *cudaMemoryManager,
- int level, int streamIndex,
+ int level, CudaStreamIndex streamIndex,
std::vector<ProcessNeighbor27> *sendProcessNeighborDev,
std::vector<ProcessNeighbor27> *recvProcessNeighborDev,
std::vector<ProcessNeighbor27> *sendProcessNeighborHost,
std::vector<ProcessNeighbor27> *recvProcessNeighborHost);
void exchangeCollDataZGPU27AllNodes(Parameter *para, vf::gpu::Communicator &comm,
- CudaMemoryManager *cudaMemoryManager, int level, int streamIndex);
+ CudaMemoryManager *cudaMemoryManager, int level, CudaStreamIndex streamIndex);
void exchangeCollDataZGPU27AfterFtoC(Parameter *para, vf::gpu::Communicator &comm,
- CudaMemoryManager *cudaMemoryManager, int level, int streamIndex);
+ CudaMemoryManager *cudaMemoryManager, int level, CudaStreamIndex streamIndex);
-void scatterNodesFromRecvBufferZGPU27AllNodes(Parameter *para, int level, int streamIndex);
-void scatterNodesFromRecvBufferZGPU27AfterFtoC(Parameter *para, int level, int streamIndex);
+void scatterNodesFromRecvBufferZGPU27AllNodes(Parameter *para, int level, CudaStreamIndex streamIndex);
+void scatterNodesFromRecvBufferZGPU27AfterFtoC(Parameter *para, int level, CudaStreamIndex streamIndex);
//////////////////////////////////////////////////////////////////////////
// 3D domain decomposition convection diffusion
diff --git a/src/gpu/VirtualFluids_GPU/DataStructureInitializer/GridProvider.cpp b/src/gpu/VirtualFluids_GPU/DataStructureInitializer/GridProvider.cpp
index e197fb5c28611e77406b30ab39aa6af2f54b9ef5..3b511264e9c7edc80bbe367cac4a9b6d8725674b 100644
--- a/src/gpu/VirtualFluids_GPU/DataStructureInitializer/GridProvider.cpp
+++ b/src/gpu/VirtualFluids_GPU/DataStructureInitializer/GridProvider.cpp
@@ -19,34 +19,29 @@ std::shared_ptr<GridProvider> GridProvider::makeGridReader(FILEFORMAT format, st
return std::shared_ptr<GridProvider>(new GridReader(format, para, cudaMemoryManager));
}
-void GridProvider::setNumberOfNodes(const int numberOfNodes, const int level) const
+void GridProvider::setNumberOfNodes(uint numberOfNodes, int level) const
{
- para->getParH(level)->numberOfNodes = numberOfNodes;
- para->getParD(level)->numberOfNodes = numberOfNodes;
- para->getParH(level)->mem_size_real_SP = sizeof(real) * para->getParH(level)->numberOfNodes;
- para->getParH(level)->mem_size_int_SP = sizeof(uint) * para->getParH(level)->numberOfNodes;
- para->getParD(level)->mem_size_real_SP = sizeof(real) * para->getParD(level)->numberOfNodes;
- para->getParD(level)->mem_size_int_SP = sizeof(uint) * para->getParD(level)->numberOfNodes;
+ para->getParH(level)->numberOfNodes = (unsigned long long)numberOfNodes;
+ para->getParD(level)->numberOfNodes = (unsigned long long)numberOfNodes;
+ para->getParH(level)->memSizeRealLBnodes = sizeof(real) * para->getParH(level)->numberOfNodes;
+ para->getParD(level)->memSizeRealLBnodes = sizeof(real) * para->getParD(level)->numberOfNodes;
+ para->getParH(level)->memSizeLonglongLBnodes = sizeof(unsigned long long) * para->getParH(level)->numberOfNodes;
+ para->getParD(level)->memSizeLonglongLBnodes = sizeof(unsigned long long) * para->getParD(level)->numberOfNodes;
}
-void GridProvider::setNumberOfFluidNodes(const int numberOfNodes, const int level) const
+void GridProvider::setNumberOfTaggedFluidNodes(uint numberOfNodes, CollisionTemplate tag, int level) const
{
- para->getParH(level)->numberOfFluidNodes = numberOfNodes;
- para->getParD(level)->numberOfFluidNodes = numberOfNodes;
+ para->getParH(level)->numberOfTaggedFluidNodes[tag] = numberOfNodes;
+ para->getParD(level)->numberOfTaggedFluidNodes[tag] = numberOfNodes;
}
-void GridProvider::setNumberOfFluidNodesBorder(const int numberOfNodes, const int level) const {
- para->getParH(level)->numberOfFluidNodesBorder = numberOfNodes;
- para->getParD(level)->numberOfFluidNodesBorder = numberOfNodes;
-}
-
-void GridProvider::setInitalNodeValues(const int numberOfNodes, const int level) const
+void GridProvider::setInitalNodeValues(uint numberOfNodes, int level) const
{
- for (int j = 1; j <= numberOfNodes; j++)
+ for (uint pos = 1; pos <= numberOfNodes; pos++)
{
- const real coordX = para->getParH(level)->coordinateX[j];
- const real coordY = para->getParH(level)->coordinateY[j];
- const real coordZ = para->getParH(level)->coordinateZ[j];
+ const real coordX = para->getParH(level)->coordinateX[pos];
+ const real coordY = para->getParH(level)->coordinateY[pos];
+ const real coordZ = para->getParH(level)->coordinateZ[pos];
real rho, vx, vy, vz;
@@ -63,40 +58,40 @@ void GridProvider::setInitalNodeValues(const int numberOfNodes, const int level)
vz = real(0.0);
}
- para->getParH(level)->rho[j] = rho;
- para->getParH(level)->velocityX[j] = vx;
- para->getParH(level)->velocityY[j] = vy;
- para->getParH(level)->velocityZ[j] = vz;
+ para->getParH(level)->rho[pos] = rho;
+ para->getParH(level)->velocityX[pos] = vx;
+ para->getParH(level)->velocityY[pos] = vy;
+ para->getParH(level)->velocityZ[pos] = vz;
//////////////////////////////////////////////////////////////////////////
if (para->getCalcMedian()) {
- para->getParH(level)->vx_SP_Med[j] = 0.0f;
- para->getParH(level)->vy_SP_Med[j] = 0.0f;
- para->getParH(level)->vz_SP_Med[j] = 0.0f;
- para->getParH(level)->rho_SP_Med[j] = 0.0f;
- para->getParH(level)->press_SP_Med[j] = 0.0f;
+ para->getParH(level)->vx_SP_Med[pos] = 0.0f;
+ para->getParH(level)->vy_SP_Med[pos] = 0.0f;
+ para->getParH(level)->vz_SP_Med[pos] = 0.0f;
+ para->getParH(level)->rho_SP_Med[pos] = 0.0f;
+ para->getParH(level)->press_SP_Med[pos] = 0.0f;
}
if (para->getUseWale()) {
- para->getParH(level)->turbViscosity[j] = 0.0f;
+ para->getParH(level)->turbViscosity[pos] = 0.0f;
//Debug
- para->getParH(level)->gSij[j] = 0.0f;
- para->getParH(level)->gSDij[j] = 0.0f;
- para->getParH(level)->gDxvx[j] = 0.0f;
- para->getParH(level)->gDyvx[j] = 0.0f;
- para->getParH(level)->gDzvx[j] = 0.0f;
- para->getParH(level)->gDxvy[j] = 0.0f;
- para->getParH(level)->gDyvy[j] = 0.0f;
- para->getParH(level)->gDzvy[j] = 0.0f;
- para->getParH(level)->gDxvz[j] = 0.0f;
- para->getParH(level)->gDyvz[j] = 0.0f;
- para->getParH(level)->gDzvz[j] = 0.0f;
+ para->getParH(level)->gSij[pos] = 0.0f;
+ para->getParH(level)->gSDij[pos] = 0.0f;
+ para->getParH(level)->gDxvx[pos] = 0.0f;
+ para->getParH(level)->gDyvx[pos] = 0.0f;
+ para->getParH(level)->gDzvx[pos] = 0.0f;
+ para->getParH(level)->gDxvy[pos] = 0.0f;
+ para->getParH(level)->gDyvy[pos] = 0.0f;
+ para->getParH(level)->gDzvy[pos] = 0.0f;
+ para->getParH(level)->gDxvz[pos] = 0.0f;
+ para->getParH(level)->gDyvz[pos] = 0.0f;
+ para->getParH(level)->gDzvz[pos] = 0.0f;
}
if (para->getIsBodyForce()) {
- para->getParH(level)->forceX_SP[j] = 0.0f;
- para->getParH(level)->forceY_SP[j] = 0.0f;
- para->getParH(level)->forceZ_SP[j] = 0.0f;
+ para->getParH(level)->forceX_SP[pos] = 0.0f;
+ para->getParH(level)->forceY_SP[pos] = 0.0f;
+ para->getParH(level)->forceZ_SP[pos] = 0.0f;
}
}
diff --git a/src/gpu/VirtualFluids_GPU/DataStructureInitializer/GridProvider.h b/src/gpu/VirtualFluids_GPU/DataStructureInitializer/GridProvider.h
index 5fc5826735643ec748da169160e782004d7e5fb7..007db1e0d8e27b3810aa38c089bae8069bbe5813 100644
--- a/src/gpu/VirtualFluids_GPU/DataStructureInitializer/GridProvider.h
+++ b/src/gpu/VirtualFluids_GPU/DataStructureInitializer/GridProvider.h
@@ -5,7 +5,7 @@
#include <vector>
#include <memory>
-
+#include "LBM/LB.h"
#include "PointerDefinitions.h"
#include "VirtualFluids_GPU_export.h"
#include "gpu/GridGenerator/io/SimulationFileWriter/SimulationFileWriter.h"
@@ -24,34 +24,35 @@ public:
static std::shared_ptr<GridProvider> makeGridGenerator(std::shared_ptr<GridBuilder> builder, std::shared_ptr<Parameter> para, std::shared_ptr<CudaMemoryManager> cudaMemoryManager, vf::gpu::Communicator& communicator);
static std::shared_ptr<GridProvider> makeGridReader(FILEFORMAT format, std::shared_ptr<Parameter> para, std::shared_ptr<CudaMemoryManager> cudaMemoryManager);
- virtual void allocArrays_CoordNeighborGeo() = 0;
- virtual void allocArrays_BoundaryValues() = 0;
- virtual void allocArrays_BoundaryQs() = 0;
+ virtual void allocArrays_CoordNeighborGeo() = 0;
+ virtual void allocArrays_BoundaryValues() = 0;
+ virtual void allocArrays_BoundaryQs() = 0;
virtual void allocArrays_OffsetScale() = 0;
- virtual void allocArrays_fluidNodeIndices() = 0;
- virtual void allocArrays_fluidNodeIndicesBorder() = 0;
+ virtual void allocArrays_taggedFluidNodes() = 0;
+
+ virtual void tagFluidNodeIndices(const std::vector<uint>& taggedFluidNodeIndices, CollisionTemplate tag, uint level) = 0;
+ virtual void sortFluidNodeTags() = 0;
- virtual void setDimensions() = 0;
- virtual void setBoundingBox() = 0;
- virtual void initPeriodicNeigh(std::vector<std::vector<std::vector<unsigned int> > > periodV, std::vector<std::vector<unsigned int> > periodIndex, std::string way) = 0;
+ virtual void setDimensions() = 0;
+ virtual void setBoundingBox() = 0;
+ virtual void initPeriodicNeigh(std::vector<std::vector<std::vector<unsigned int> > > periodV, std::vector<std::vector<unsigned int> > periodIndex, std::string way) = 0;
virtual void allocAndCopyForcing();
virtual void allocAndCopyQuadricLimiters();
virtual void freeMemoryOnHost();
virtual void cudaCopyDataToHost(int level);
- virtual ~GridProvider() = default;
+ virtual ~GridProvider() = default;
virtual void initalGridInformations() = 0;
protected:
- void setNumberOfNodes(const int numberOfNodes, const int level) const;
- void setNumberOfFluidNodes(const int numberOfNodes, const int level) const;
- void setNumberOfFluidNodesBorder(const int numberOfNodes, const int level) const;
- virtual void setInitalNodeValues(const int numberOfNodes, const int level) const;
-
- void setPressSizePerLevel(int level, int sizePerLevel) const;
- void setVelocitySizePerLevel(int level, int sizePerLevel) const;
- void setOutflowSizePerLevel(int level, int sizePerLevel) const;
+ void setNumberOfNodes(uint numberOfNodes, int level) const;
+ void setNumberOfTaggedFluidNodes(uint numberOfNodes, CollisionTemplate tag, int level) const;
+ virtual void setInitalNodeValues(uint numberOfNodes, int level) const;
+
+ void setPressSizePerLevel(int level, int sizePerLevel) const;
+ void setVelocitySizePerLevel(int level, int sizePerLevel) const;
+ void setOutflowSizePerLevel(int level, int sizePerLevel) const;
std::shared_ptr<Parameter> para;
std::shared_ptr<CudaMemoryManager> cudaMemoryManager;
diff --git a/src/gpu/VirtualFluids_GPU/DataStructureInitializer/GridReaderFiles/GridReader.cpp b/src/gpu/VirtualFluids_GPU/DataStructureInitializer/GridReaderFiles/GridReader.cpp
index fa432a1d9c3922b88e93588548db74083275ef1e..a1c8554cc4e262e9f1eca4204aed4ffcfd4c3a87 100644
--- a/src/gpu/VirtualFluids_GPU/DataStructureInitializer/GridReaderFiles/GridReader.cpp
+++ b/src/gpu/VirtualFluids_GPU/DataStructureInitializer/GridReaderFiles/GridReader.cpp
@@ -47,7 +47,7 @@ bool GridReader::getBinaer()
void rearrangeGeometry(Parameter* para, int lev)
{
- for (uint index = 0; index < para->getParH(lev)->numberOfNodes; index++)
+ for (size_t index = 0; index < para->getParH(lev)->numberOfNodes; index++)
{
if (para->getParH(lev)->typeOfGridNode[index] == GEO_FLUID_OLD)
{
@@ -74,11 +74,11 @@ void GridReader::allocArrays_CoordNeighborGeo()
uint numberOfNodesGlobal = 0;
std::cout << "Number of Nodes: " << std::endl;
- for (uint level = 0; level <= maxLevel; level++)
- {
- int numberOfNodesPerLevel = coordX.getSize(level) + 1;
- numberOfNodesGlobal += numberOfNodesPerLevel;
- std::cout << "Level " << level << " = " << numberOfNodesPerLevel << " Nodes" << std::endl;
+ for (uint level = 0; level <= maxLevel; level++)
+ {
+ const uint numberOfNodesPerLevel = coordX.getSize(level) + 1;
+ numberOfNodesGlobal += numberOfNodesPerLevel;
+ std::cout << "Level " << level << " = " << numberOfNodesPerLevel << " Nodes" << std::endl;
setNumberOfNodes(numberOfNodesPerLevel, level);
@@ -130,9 +130,9 @@ void GridReader::allocArrays_BoundaryValues()
for (uint i = 0; i < channelBoundaryConditions.size(); i++)
{
- if ( this->channelBoundaryConditions[i] == "velocity") { fillVelocityVectors(i); }
- else if (this->channelBoundaryConditions[i] == "pressure") { setPressureValues(i); }
- else if (this->channelBoundaryConditions[i] == "outflow") { setOutflowValues(i); }
+ if ( this->channelBoundaryConditions[i] == "velocity") { fillVelocityVectors(i); }
+ else if (this->channelBoundaryConditions[i] == "pressure") { setPressureValues(i); }
+ else if (this->channelBoundaryConditions[i] == "outflow") { setOutflowValues(i); }
}
setVelocityValues();
@@ -218,16 +218,20 @@ void GridReader::allocArrays_OffsetScale()
std::cout << "-----Ende OffsetScale------" << std::endl;
}
-void GridReader::allocArrays_fluidNodeIndices() {
+void GridReader::allocArrays_taggedFluidNodes() {
std::cout << "GridReader::allocArrays_fluidNodeIndices not implemented" << std::endl;
// TODO
}
-void GridReader::allocArrays_fluidNodeIndicesBorder() {
- std::cout << "GridReader::allocArrays_fluidNodeIndicesBorder not implemented" << std::endl;
+void GridReader::tagFluidNodeIndices(const std::vector<uint>& taggedFluidNodeIndices, CollisionTemplate tag, uint level){
+ std::cout << "GridReader::tagFluidNodeIndices not implemented" << std::endl;
// TODO
}
+void GridReader::sortFluidNodeTags(){
+ std::cout << "GridReader::sortFluidNodeTags not implemented" << std::endl;
+ // TODO
+}
void GridReader::setPressureValues(int channelSide) const
{
@@ -281,23 +285,23 @@ void GridReader::fillVelocityVectors(int channelSide)
delete[] veloX_ValuesPerSide;
delete[] veloY_ValuesPerSide;
delete[] veloZ_ValuesPerSide;
- }
- }
+ }
+ }
}
-void GridReader::setVelocityValues() {
+void GridReader::setVelocityValues() {
for (int level = 0; level < (int)(velocityX_BCvalues.size()); level++) {
-
- int sizePerLevel = (int) velocityX_BCvalues[level].size();
+
+ int sizePerLevel = (int) velocityX_BCvalues[level].size();
std::cout << "complete size velocity level " << level << " : " << sizePerLevel << std::endl;
setVelocitySizePerLevel(level, sizePerLevel);
-
- if (sizePerLevel > 1) {
+
+ if (sizePerLevel > 1) {
cudaMemoryManager->cudaAllocVeloBC(level);
setVelocity(level, sizePerLevel);
- cudaMemoryManager->cudaCopyVeloBC(level);
+ cudaMemoryManager->cudaCopyVeloBC(level);
}
}
}
@@ -668,8 +672,8 @@ void GridReader::modifyQElement(std::shared_ptr<BoundaryQs> boundaryQ, unsigned
/*------------------------------------------------------------------------------------------------*/
/*---------------------------------------private q methods----------------------------------------*/
/*------------------------------------------------------------------------------------------------*/
-void GridReader::initalVectorForQStruct(std::vector<std::vector<std::vector<real>>> &Qs, std::vector<std::vector<int>> &index,
- std::shared_ptr<BoundaryQs> boundaryQ, unsigned int level) const
+void GridReader::initalVectorForQStruct(std::vector<std::vector<std::vector<real>>> &Qs, std::vector<std::vector<int>> &index,
+ std::shared_ptr<BoundaryQs> boundaryQ, unsigned int level) const
{
boundaryQ->setValuesInVector(Qs, level);
boundaryQ->setIndexInVector(index, level);
@@ -685,7 +689,7 @@ void GridReader::copyVectorsToQStruct(std::vector<std::vector<real>> &Qs,
for (int direction = 0; direction < para->getD3Qxx(); direction++) {
for (size_t indexQ = 0; indexQ < sizeOfValues; indexQ++) {
- qTemp.q27[direction][indexQ] = Qs[direction][indexQ];
+ qTemp.q27[direction][indexQ] = Qs[direction][indexQ];
}
}
@@ -847,46 +851,46 @@ void GridReader::setBoundingBox()
void GridReader::initPeriodicNeigh(std::vector<std::vector<std::vector<unsigned int> > > periodV, std::vector<std::vector<unsigned int> > periodIndex, std::string boundaryCondition)
{
- std::vector<unsigned int>neighVec;
- std::vector<unsigned int>indexVec;
-
- int counter = 0;
-
- for(unsigned int i=0; i<neighX->getLevel();i++) {
- if(boundaryCondition =="periodic_y"){
- neighVec = neighY->getVec(i);
- }
- else if(boundaryCondition =="periodic_x"){
- neighVec = neighX->getVec(i);
- }
- else if(boundaryCondition =="periodic_z"){
- neighVec = neighZ->getVec(i);
- }
- else {
- std::cout << "wrong String in periodicValue" << std::endl;
- exit(1);
- }
+ std::vector<unsigned int>neighVec;
+ std::vector<unsigned int>indexVec;
- for (std::vector<unsigned int>::iterator it = periodIndex[i].begin(); it != periodIndex[i].end(); it++) {
- if(periodV[i][0][counter] != 0) {
- neighVec[*it]=periodV[i][0][counter];
- }
+ int counter = 0;
- counter++;
- }
+ for(unsigned int i=0; i<neighX->getLevel();i++) {
+ if(boundaryCondition =="periodic_y"){
+ neighVec = neighY->getVec(i);
+ }
+ else if(boundaryCondition =="periodic_x"){
+ neighVec = neighX->getVec(i);
+ }
+ else if(boundaryCondition =="periodic_z"){
+ neighVec = neighZ->getVec(i);
+ }
+ else {
+ std::cout << "wrong String in periodicValue" << std::endl;
+ exit(1);
+ }
+ for (std::vector<unsigned int>::iterator it = periodIndex[i].begin(); it != periodIndex[i].end(); it++) {
+ if(periodV[i][0][counter] != 0) {
+ neighVec[*it]=periodV[i][0][counter];
+ }
- if(boundaryCondition =="periodic_y"){
- neighY->setVec(i, neighVec);
- }
- else if(boundaryCondition =="periodic_x"){
- neighX->setVec(i, neighVec);
- }
- else if(boundaryCondition =="periodic_z"){
- neighZ->setVec(i, neighVec);
- }
+ counter++;
+ }
- }
+
+ if(boundaryCondition =="periodic_y"){
+ neighY->setVec(i, neighVec);
+ }
+ else if(boundaryCondition =="periodic_x"){
+ neighX->setVec(i, neighVec);
+ }
+ else if(boundaryCondition =="periodic_z"){
+ neighZ->setVec(i, neighVec);
+ }
+
+ }
}
void GridReader::makeReader(std::shared_ptr<Parameter> para)
@@ -917,9 +921,9 @@ void GridReader::makeReader(std::vector<std::shared_ptr<BoundaryQs> > &BC_Qs, st
void GridReader::setChannelBoundaryCondition()
{
- for (std::size_t i = 0; i < channelDirections.size(); i++)
- {
- this->channelBoundaryConditions[i] = BC_Values[i]->getBoundaryCondition();
- std::cout << this->channelDirections[i] << " Boundary: " << channelBoundaryConditions[i] << std::endl;
- }
-}
\ No newline at end of file
+ for (std::size_t i = 0; i < channelDirections.size(); i++)
+ {
+ this->channelBoundaryConditions[i] = BC_Values[i]->getBoundaryCondition();
+ std::cout << this->channelDirections[i] << " Boundary: " << channelBoundaryConditions[i] << std::endl;
+ }
+}
diff --git a/src/gpu/VirtualFluids_GPU/DataStructureInitializer/GridReaderFiles/GridReader.h b/src/gpu/VirtualFluids_GPU/DataStructureInitializer/GridReaderFiles/GridReader.h
index 18efb6a7885191312ea4e2fbb22eb45162ab1de1..041d2c3ce94592f792c5a850eebd14c07f4db1b4 100644
--- a/src/gpu/VirtualFluids_GPU/DataStructureInitializer/GridReaderFiles/GridReader.h
+++ b/src/gpu/VirtualFluids_GPU/DataStructureInitializer/GridReaderFiles/GridReader.h
@@ -3,9 +3,9 @@
#include "../GridProvider.h"
-#include <vector>
-#include <string>
#include <memory>
+#include <string>
+#include <vector>
#include "LBM/LB.h"
@@ -16,15 +16,14 @@ class BoundaryValues;
class BoundaryQs;
class CoordNeighborGeoV;
-class VIRTUALFLUIDS_GPU_EXPORT GridReader
- : public GridProvider
+class VIRTUALFLUIDS_GPU_EXPORT GridReader : public GridProvider
{
private:
- bool binaer;
- std::vector<std::string> channelDirections;
- std::vector<std::string> channelBoundaryConditions;
- std::shared_ptr<CoordNeighborGeoV> neighX, neighY, neighZ, neighWSB;
- std::vector<std::shared_ptr<BoundaryValues> > BC_Values;
+ bool binaer;
+ std::vector<std::string> channelDirections;
+ std::vector<std::string> channelBoundaryConditions;
+ std::shared_ptr<CoordNeighborGeoV> neighX, neighY, neighZ, neighWSB;
+ std::vector<std::shared_ptr<BoundaryValues>> BC_Values;
std::vector<std::vector<real>> velocityX_BCvalues, velocityY_BCvalues, velocityZ_BCvalues;
std::vector<std::vector<std::vector<real>>> velocityQs;
@@ -34,57 +33,62 @@ private:
std::vector<std::vector<real>> outflowBCvalues;
public:
- GridReader(FILEFORMAT format, std::shared_ptr<Parameter> para, std::shared_ptr<CudaMemoryManager> cudaMemoryManager);
- ~GridReader();
- void allocArrays_CoordNeighborGeo() override;
- void allocArrays_BoundaryValues() override;
+ GridReader(FILEFORMAT format, std::shared_ptr<Parameter> para,
+ std::shared_ptr<CudaMemoryManager> cudaMemoryManager);
+ ~GridReader() override;
+ void allocArrays_CoordNeighborGeo() override;
+ void allocArrays_BoundaryValues() override;
void allocArrays_OffsetScale() override;
- void allocArrays_fluidNodeIndices() override;
- void allocArrays_fluidNodeIndicesBorder() override;
+ void allocArrays_taggedFluidNodes() override;
- void initalValuesDomainDecompostion(int level);
+ void tagFluidNodeIndices(const std::vector<uint> &taggedFluidNodeIndices, CollisionTemplate tag, uint level) override;
- void setChannelBoundaryCondition();
+ void sortFluidNodeTags() override;
- void allocArrays_BoundaryQs() override;
- bool getBinaer();
- void setDimensions() override;
- void setBoundingBox() override;
- void initPeriodicNeigh(std::vector<std::vector<std::vector<unsigned int> > > periodV, std::vector<std::vector<unsigned int> > periodIndex, std::string way) override;
+ void initalValuesDomainDecompostion(int level);
+
+ void setChannelBoundaryCondition();
+
+ void allocArrays_BoundaryQs() override;
+ bool getBinaer();
+ void setDimensions() override;
+ void setBoundingBox() override;
+ void initPeriodicNeigh(std::vector<std::vector<std::vector<unsigned int>>> periodV,
+ std::vector<std::vector<unsigned int>> periodIndex, std::string way) override;
private:
- void makeReader(std::shared_ptr<Parameter> para);
- void makeReader(std::vector<std::shared_ptr<BoundaryQs> > &BC_Qs, std::shared_ptr<Parameter> para);
+ void makeReader(std::shared_ptr<Parameter> para);
+ void makeReader(std::vector<std::shared_ptr<BoundaryQs>> &BC_Qs, std::shared_ptr<Parameter> para);
- void setPressureValues(int channelSide) const;
- void setPressRhoBC(int sizePerLevel, int level, int channelSide) const;
+ void setPressureValues(int channelSide) const;
+ void setPressRhoBC(int sizePerLevel, int level, int channelSide) const;
- void fillVelocityVectors(int channelSide);
+ void fillVelocityVectors(int channelSide);
void setVelocityValues();
- void setVelocity(int level, int sizePerLevel) const;
+ void setVelocity(int level, int sizePerLevel) const;
- void setOutflowValues(int channelSide) const;
- void setOutflow(int level, int sizePerLevel, int channelSide) const;
+ void setOutflowValues(int channelSide) const;
+ void setOutflow(int level, int sizePerLevel, int channelSide) const;
-
- //void fillVelocityQVectors(int channelSide);
+ // void fillVelocityQVectors(int channelSide);
void setPressQs(std::shared_ptr<BoundaryQs> boundaryQ) const;
- void setVelocityQs(std::shared_ptr<BoundaryQs> boundaryQ);
- void setOutflowQs(std::shared_ptr<BoundaryQs> boundaryQ) const;
- void setNoSlipQs(std::shared_ptr<BoundaryQs> boundaryQ) const;
- void setGeoQs(std::shared_ptr<BoundaryQs> boundaryQ) const;
- void modifyQElement(std::shared_ptr<BoundaryQs> boundaryQ, unsigned int level) const;
+ void setVelocityQs(std::shared_ptr<BoundaryQs> boundaryQ);
+ void setOutflowQs(std::shared_ptr<BoundaryQs> boundaryQ) const;
+ void setNoSlipQs(std::shared_ptr<BoundaryQs> boundaryQ) const;
+ void setGeoQs(std::shared_ptr<BoundaryQs> boundaryQ) const;
+ void modifyQElement(std::shared_ptr<BoundaryQs> boundaryQ, unsigned int level) const;
- void initalVectorForQStruct(std::vector<std::vector<std::vector<real>>> &Qs, std::vector<std::vector<int>> &index,
+ void initalVectorForQStruct(std::vector<std::vector<std::vector<real>>> &Qs, std::vector<std::vector<int>> &index,
std::shared_ptr<BoundaryQs> boundaryQ, unsigned int level) const;
void copyVectorsToQStruct(std::vector<std::vector<real>> &Qs, std::vector<int> &index,
QforBoundaryConditions &Q) const;
void initalQStruct(QforBoundaryConditions &Q, std::shared_ptr<BoundaryQs> boundaryQ, unsigned int level) const;
- void printQSize(std::string bc, std::shared_ptr<BoundaryQs> boundaryQ, unsigned int level) const;
- void setSizeNoSlip(std::shared_ptr<BoundaryQs> boundaryQ, unsigned int level) const;
- void setSizeGeoQs(std::shared_ptr<BoundaryQs> boundaryQ, unsigned int level) const;
- void setQ27Size(QforBoundaryConditions &Q, real* QQ, unsigned int sizeQ) const;
- bool hasQs(std::shared_ptr<BoundaryQs> boundaryQ, unsigned int level) const;
+ void printQSize(std::string bc, std::shared_ptr<BoundaryQs> boundaryQ, unsigned int level) const;
+ void setSizeNoSlip(std::shared_ptr<BoundaryQs> boundaryQ, unsigned int level) const;
+ void setSizeGeoQs(std::shared_ptr<BoundaryQs> boundaryQ, unsigned int level) const;
+ void setQ27Size(QforBoundaryConditions &Q, real *QQ, unsigned int sizeQ) const;
+ bool hasQs(std::shared_ptr<BoundaryQs> boundaryQ, unsigned int level) const;
+
public:
void initalGridInformations() override;
};
diff --git a/src/gpu/VirtualFluids_GPU/DataStructureInitializer/GridReaderGenerator/GridGenerator.cpp b/src/gpu/VirtualFluids_GPU/DataStructureInitializer/GridReaderGenerator/GridGenerator.cpp
index 7f61b4357276f38d8fde71489dcf60348b402941..38a7eef7e356e2f2da4c1a819d8375035a37313a 100644
--- a/src/gpu/VirtualFluids_GPU/DataStructureInitializer/GridReaderGenerator/GridGenerator.cpp
+++ b/src/gpu/VirtualFluids_GPU/DataStructureInitializer/GridReaderGenerator/GridGenerator.cpp
@@ -1,5 +1,6 @@
#include "GridGenerator.h"
+#include "LBM/LB.h"
#include "Parameter/Parameter.h"
#include "GridGenerator/grid/GridBuilder/GridBuilder.h"
#include "GPU/CudaMemoryManager.h"
@@ -10,19 +11,24 @@
#include <algorithm>
#include "utilities/math/Math.h"
#include "Output/QDebugWriter.hpp"
+#include "GridGenerator/TransientBCSetter/TransientBCSetter.h"
#include "utilities/communication.h"
#include "Communication/Communicator.h"
+#include <logger/Logger.h>
+
using namespace vf::lbm::dir;
-GridGenerator::GridGenerator(std::shared_ptr<GridBuilder> builder, std::shared_ptr<Parameter> para, std::shared_ptr<CudaMemoryManager> cudaMemoryManager, vf::gpu::Communicator& communicator):
- mpiProcessID(communicator.getPID()), builder(builder)
+GridGenerator::GridGenerator(std::shared_ptr<GridBuilder> builder, std::shared_ptr<Parameter> para,
+ std::shared_ptr<CudaMemoryManager> cudaMemoryManager, vf::gpu::Communicator &communicator)
+ : mpiProcessID(communicator.getPID()), builder(builder)
{
this->para = para;
this->cudaMemoryManager = cudaMemoryManager;
this->indexRearrangement = std::make_unique<IndexRearrangementForStreams>(para, builder, communicator);
- this->interpolationGrouper = std::make_unique<InterpolationCellGrouper>(para->getParHallLevels(), para->getParDallLevels(), builder);
+ this->interpolationGrouper =
+ std::make_unique<InterpolationCellGrouper>(para->getParHallLevels(), para->getParDallLevels(), builder);
}
GridGenerator::~GridGenerator() = default;
@@ -55,15 +61,15 @@ void GridGenerator::allocArrays_CoordNeighborGeo()
std::cout << "Number of Level: " << numberOfLevels << std::endl;
int numberOfNodesGlobal = 0;
std::cout << "Number of Nodes: " << std::endl;
-
- for (uint level = 0; level < numberOfLevels; level++)
+
+ for (uint level = 0; level < numberOfLevels; level++)
{
- const int numberOfNodesPerLevel = builder->getNumberOfNodes(level) + 1;
+ const uint numberOfNodesPerLevel = builder->getNumberOfNodes(level) + 1;
numberOfNodesGlobal += numberOfNodesPerLevel;
std::cout << "Level " << level << " = " << numberOfNodesPerLevel << " Nodes" << std::endl;
-
+
setNumberOfNodes(numberOfNodesPerLevel, level);
-
+
cudaMemoryManager->cudaAllocCoord(level);
cudaMemoryManager->cudaAllocSP(level);
//cudaMemoryManager->cudaAllocF3SP(level);
@@ -71,7 +77,7 @@ void GridGenerator::allocArrays_CoordNeighborGeo()
if(para->getUseTurbulentViscosity())
cudaMemoryManager->cudaAllocTurbulentViscosity(level);
-
+
if(para->getIsBodyForce())
cudaMemoryManager->cudaAllocBodyForce(level);
@@ -100,28 +106,104 @@ void GridGenerator::allocArrays_CoordNeighborGeo()
std::cout << "-----finish Coord, Neighbor, Geo------" << std::endl;
}
-void GridGenerator::allocArrays_fluidNodeIndices() {
- for (uint level = 0; level < builder->getNumberOfGridLevels(); level++) {
- setNumberOfFluidNodes(builder->getNumberOfFluidNodes(level), level);
- cudaMemoryManager->cudaAllocFluidNodeIndices(level);
- builder->getFluidNodeIndices(para->getParH(level)->fluidNodeIndices, level);
- cudaMemoryManager->cudaCopyFluidNodeIndices(level);
- }
+void GridGenerator::allocArrays_taggedFluidNodes() {
+
+ for (uint level = 0; level < builder->getNumberOfGridLevels(); level++)
+ {
+ for ( CollisionTemplate tag: all_CollisionTemplate )
+ { //TODO: Need to add CollisionTemplate to GridBuilder to allow as argument and get rid of indivual get funtions for fluid node indices... and clean up this mess
+ switch(tag)
+ {
+ case CollisionTemplate::Default:
+ this->setNumberOfTaggedFluidNodes(builder->getNumberOfFluidNodes(level), CollisionTemplate::Default, level);
+ cudaMemoryManager->cudaAllocTaggedFluidNodeIndices(CollisionTemplate::Default, level);
+ builder->getFluidNodeIndices(para->getParH(level)->taggedFluidNodeIndices[CollisionTemplate::Default], level);
+ cudaMemoryManager->cudaCopyTaggedFluidNodeIndices(CollisionTemplate::Default, level);
+ if(para->getParH(level)->numberOfTaggedFluidNodes[tag]>0)
+ para->getParH(level)->allocatedBulkFluidNodeTags.push_back(tag);
+ break;
+ case CollisionTemplate::SubDomainBorder:
+ this->setNumberOfTaggedFluidNodes(builder->getNumberOfFluidNodesBorder(level), CollisionTemplate::SubDomainBorder, level);
+ cudaMemoryManager->cudaAllocTaggedFluidNodeIndices(CollisionTemplate::SubDomainBorder, level);
+ builder->getFluidNodeIndicesBorder(para->getParH(level)->taggedFluidNodeIndices[CollisionTemplate::SubDomainBorder], level);
+ cudaMemoryManager->cudaCopyTaggedFluidNodeIndices(CollisionTemplate::SubDomainBorder, level);
+ break;
+ case CollisionTemplate::WriteMacroVars:
+ this->setNumberOfTaggedFluidNodes(builder->getNumberOfFluidNodesMacroVars(level), CollisionTemplate::WriteMacroVars, level);
+ cudaMemoryManager->cudaAllocTaggedFluidNodeIndices(CollisionTemplate::WriteMacroVars, level);
+ builder->getFluidNodeIndicesMacroVars(para->getParH(level)->taggedFluidNodeIndices[CollisionTemplate::WriteMacroVars], level);
+ cudaMemoryManager->cudaCopyTaggedFluidNodeIndices(CollisionTemplate::WriteMacroVars, level);
+ if(para->getParH(level)->numberOfTaggedFluidNodes[tag]>0)
+ para->getParH(level)->allocatedBulkFluidNodeTags.push_back(tag);
+ break;
+ case CollisionTemplate::ApplyBodyForce:
+ this->setNumberOfTaggedFluidNodes(builder->getNumberOfFluidNodesApplyBodyForce(level), CollisionTemplate::ApplyBodyForce, level);
+ cudaMemoryManager->cudaAllocTaggedFluidNodeIndices(CollisionTemplate::ApplyBodyForce, level);
+ builder->getFluidNodeIndicesApplyBodyForce(para->getParH(level)->taggedFluidNodeIndices[CollisionTemplate::ApplyBodyForce], level);
+ cudaMemoryManager->cudaCopyTaggedFluidNodeIndices(CollisionTemplate::ApplyBodyForce, level);
+ if(para->getParH(level)->numberOfTaggedFluidNodes[tag]>0)
+ para->getParH(level)->allocatedBulkFluidNodeTags.push_back(tag);
+ break;
+ case CollisionTemplate::AllFeatures:
+ this->setNumberOfTaggedFluidNodes(builder->getNumberOfFluidNodesAllFeatures(level), CollisionTemplate::AllFeatures, level);
+ cudaMemoryManager->cudaAllocTaggedFluidNodeIndices(CollisionTemplate::AllFeatures, level);
+ builder->getFluidNodeIndicesAllFeatures(para->getParH(level)->taggedFluidNodeIndices[CollisionTemplate::AllFeatures], level);
+ cudaMemoryManager->cudaCopyTaggedFluidNodeIndices(CollisionTemplate::AllFeatures, level);
+ if(para->getParH(level)->numberOfTaggedFluidNodes[tag]>0)
+ para->getParH(level)->allocatedBulkFluidNodeTags.push_back(tag);
+ break;
+ default:
+ break;
+ }
+ }
+ VF_LOG_INFO("Number of tagged nodes on level {}:", level);
+ VF_LOG_INFO("Default: {}, Border: {}, WriteMacroVars: {}, ApplyBodyForce: {}, AllFeatures: {}",
+ para->getParH(level)->numberOfTaggedFluidNodes[CollisionTemplate::Default],
+ para->getParH(level)->numberOfTaggedFluidNodes[CollisionTemplate::SubDomainBorder],
+ para->getParH(level)->numberOfTaggedFluidNodes[CollisionTemplate::WriteMacroVars],
+ para->getParH(level)->numberOfTaggedFluidNodes[CollisionTemplate::ApplyBodyForce],
+ para->getParH(level)->numberOfTaggedFluidNodes[CollisionTemplate::AllFeatures] );
+ }
}
-void GridGenerator::allocArrays_fluidNodeIndicesBorder() {
- for (uint level = 0; level < builder->getNumberOfGridLevels(); level++) {
- setNumberOfFluidNodesBorder(builder->getNumberOfFluidNodesBorder(level), level);
- cudaMemoryManager->cudaAllocFluidNodeIndicesBorder(level);
- builder->getFluidNodeIndicesBorder(para->getParH(level)->fluidNodeIndicesBorder, level);
- cudaMemoryManager->cudaCopyFluidNodeIndicesBorder(level);
+void GridGenerator::tagFluidNodeIndices(const std::vector<uint>& taggedFluidNodeIndices, CollisionTemplate tag, uint level) {
+ switch(tag)
+ {
+ case CollisionTemplate::WriteMacroVars:
+ builder->addFluidNodeIndicesMacroVars( taggedFluidNodeIndices, level );
+ break;
+ case CollisionTemplate::ApplyBodyForce:
+ builder->addFluidNodeIndicesApplyBodyForce( taggedFluidNodeIndices, level );
+ break;
+ case CollisionTemplate::AllFeatures:
+ builder->addFluidNodeIndicesAllFeatures( taggedFluidNodeIndices, level );
+ break;
+ case CollisionTemplate::Default:
+ case CollisionTemplate::SubDomainBorder:
+ throw std::runtime_error("Cannot tag fluid nodes as Default or SubDomainBorder!");
+ default:
+ throw std::runtime_error("Tagging fluid nodes with invald tag!");
+ break;
+
}
+
+}
+
+void GridGenerator::sortFluidNodeTags() {
+ VF_LOG_INFO("Start sorting tagged fluid nodes...");
+ for (uint level = 0; level < builder->getNumberOfGridLevels(); level++)
+ {
+ builder->sortFluidNodeIndicesAllFeatures(level); //has to be called first!
+ builder->sortFluidNodeIndicesMacroVars(level);
+ builder->sortFluidNodeIndicesApplyBodyForce(level);
+ }
+ VF_LOG_INFO("done.");
}
void GridGenerator::allocArrays_BoundaryValues()
{
std::cout << "------read BoundaryValues------" << std::endl;
- int blocks = 0;
+ int blocks;
for (uint level = 0; level < builder->getNumberOfGridLevels(); level++) {
const auto numberOfPressureValues = int(builder->getPressureSize(level));
@@ -129,6 +211,7 @@ void GridGenerator::allocArrays_BoundaryValues()
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
para->getParH(level)->pressureBC.numberOfBCnodes = 0;
+ para->getParD(level)->outflowPressureCorrectionFactor = para->getOutflowPressureCorrectionFactor();
if (numberOfPressureValues > 1)
{
blocks = (numberOfPressureValues / para->getParH(level)->numberofthreads) + 1;
@@ -148,12 +231,12 @@ void GridGenerator::allocArrays_BoundaryValues()
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
para->getParH(level)->slipBC.numberOfBCnodes = 0;
- if (numberOfSlipValues > 1)
- {
+ if (numberOfSlipValues > 1) {
blocks = (numberOfSlipValues / para->getParH(level)->numberofthreads) + 1;
para->getParH(level)->slipBC.numberOfBCnodes = blocks * para->getParH(level)->numberofthreads;
cudaMemoryManager->cudaAllocSlipBC(level);
- builder->getSlipValues(para->getParH(level)->slipBC.normalX, para->getParH(level)->slipBC.normalY, para->getParH(level)->slipBC.normalZ, para->getParH(level)->slipBC.k, level);
+ builder->getSlipValues(para->getParH(level)->slipBC.normalX, para->getParH(level)->slipBC.normalY,
+ para->getParH(level)->slipBC.normalZ, para->getParH(level)->slipBC.k, level);
cudaMemoryManager->cudaCopySlipBC(level);
}
para->getParD(level)->slipBC.numberOfBCnodes = para->getParH(level)->slipBC.numberOfBCnodes;
@@ -173,11 +256,11 @@ void GridGenerator::allocArrays_BoundaryValues()
para->getParH(level)->stressBC.numberOfBCnodes = blocks * para->getParH(level)->numberofthreads;
cudaMemoryManager->cudaAllocStressBC(level);
cudaMemoryManager->cudaAllocWallModel(level, para->getHasWallModelMonitor());
- builder->getStressValues( para->getParH(level)->stressBC.normalX, para->getParH(level)->stressBC.normalY, para->getParH(level)->stressBC.normalZ,
+ builder->getStressValues( para->getParH(level)->stressBC.normalX, para->getParH(level)->stressBC.normalY, para->getParH(level)->stressBC.normalZ,
para->getParH(level)->stressBC.Vx, para->getParH(level)->stressBC.Vy, para->getParH(level)->stressBC.Vz,
para->getParH(level)->stressBC.Vx1, para->getParH(level)->stressBC.Vy1, para->getParH(level)->stressBC.Vz1,
- para->getParH(level)->stressBC.k, para->getParH(level)->stressBC.kN,
- para->getParH(level)->wallModel.samplingOffset, para->getParH(level)->wallModel.z0,
+ para->getParH(level)->stressBC.k, para->getParH(level)->stressBC.kN,
+ para->getParH(level)->wallModel.samplingOffset, para->getParH(level)->wallModel.z0,
level);
cudaMemoryManager->cudaCopyStressBC(level);
@@ -187,7 +270,7 @@ void GridGenerator::allocArrays_BoundaryValues()
para->getParH(level)->numberOfStressBCnodesRead = para->getParH(level)->stressBC.numberOfBCnodes * para->getD3Qxx();
para->getParD(level)->numberOfStressBCnodesRead = para->getParH(level)->stressBC.numberOfBCnodes * para->getD3Qxx();
}
-
+
for (uint level = 0; level < builder->getNumberOfGridLevels(); level++) {
const auto numberOfVelocityValues = int(builder->getVelocitySize(level));
@@ -204,7 +287,8 @@ void GridGenerator::allocArrays_BoundaryValues()
cudaMemoryManager->cudaAllocVeloBC(level);
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
- builder->getVelocityValues(para->getParH(level)->velocityBC.Vx, para->getParH(level)->velocityBC.Vy, para->getParH(level)->velocityBC.Vz, para->getParH(level)->velocityBC.k, level);
+ builder->getVelocityValues(para->getParH(level)->velocityBC.Vx, para->getParH(level)->velocityBC.Vy,
+ para->getParH(level)->velocityBC.Vz, para->getParH(level)->velocityBC.k, level);
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
@@ -242,6 +326,100 @@ void GridGenerator::allocArrays_BoundaryValues()
para->getParD(level)->numberOfVeloBCnodesRead = para->getParH(level)->velocityBC.numberOfBCnodes * para->getD3Qxx();
}
+ for (uint level = 0; level < builder->getNumberOfGridLevels(); level++) {
+ const auto numberOfPrecursorValues = int(builder->getPrecursorSize(level));
+ *logging::out << logging::Logger::INFO_INTERMEDIATE << "size precursor level " << level << " : " << numberOfPrecursorValues << "\n";
+ ////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
+ blocks = (numberOfPrecursorValues / para->getParH(level)->numberofthreads) + 1;
+ para->getParH(level)->precursorBC.sizeQ = blocks * para->getParH(level)->numberofthreads;
+ para->getParD(level)->precursorBC.sizeQ = para->getParH(level)->precursorBC.sizeQ;
+ ////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
+ para->getParH(level)->precursorBC.numberOfBCnodes = numberOfPrecursorValues;
+ para->getParD(level)->precursorBC.numberOfBCnodes = numberOfPrecursorValues;
+ para->getParH(level)->numberOfPrecursorBCnodesRead = numberOfPrecursorValues * para->getD3Qxx();
+ para->getParD(level)->numberOfPrecursorBCnodesRead = numberOfPrecursorValues * para->getD3Qxx();
+
+ if (numberOfPrecursorValues > 1)
+ {
+ ////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
+ cudaMemoryManager->cudaAllocPrecursorBC(level);
+ ////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
+ builder->getPrecursorValues(
+ para->getParH(level)->precursorBC.planeNeighbor0PP, para->getParH(level)->precursorBC.planeNeighbor0PM,
+ para->getParH(level)->precursorBC.planeNeighbor0MP, para->getParH(level)->precursorBC.planeNeighbor0MM,
+ para->getParH(level)->precursorBC.weights0PP, para->getParH(level)->precursorBC.weights0PM,
+ para->getParH(level)->precursorBC.weights0MP, para->getParH(level)->precursorBC.weights0MM,
+ para->getParH(level)->precursorBC.k, para->getParH(level)->transientBCInputFileReader, para->getParH(level)->precursorBC.numberOfPrecursorNodes,
+ para->getParH(level)->precursorBC.numberOfQuantities, para->getParH(level)->precursorBC.timeStepsBetweenReads,
+ para->getParH(level)->precursorBC.velocityX, para->getParH(level)->precursorBC.velocityY, para->getParH(level)->precursorBC.velocityZ,
+ level);
+ ////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
+ para->getParD(level)->precursorBC.numberOfPrecursorNodes = para->getParH(level)->precursorBC.numberOfPrecursorNodes;
+ para->getParD(level)->precursorBC.numberOfQuantities = para->getParH(level)->precursorBC.numberOfQuantities;
+ para->getParD(level)->precursorBC.timeStepsBetweenReads = para->getParH(level)->precursorBC.timeStepsBetweenReads;
+ para->getParD(level)->precursorBC.velocityX = para->getParH(level)->precursorBC.velocityX;
+ para->getParD(level)->precursorBC.velocityY = para->getParH(level)->precursorBC.velocityY;
+ para->getParD(level)->precursorBC.velocityZ = para->getParH(level)->precursorBC.velocityZ;
+
+ for(auto reader : para->getParH(level)->transientBCInputFileReader)
+ {
+ if(reader->getNumberOfQuantities() != para->getParD(level)->precursorBC.numberOfQuantities)
+ throw std::runtime_error(
+ "Number of quantities in reader and number of quantities needed for precursor don't match!");
+ }
+
+ cudaMemoryManager->cudaCopyPrecursorBC(level);
+ cudaMemoryManager->cudaAllocPrecursorData(level);
+
+ // read first timestep of precursor into next and copy to next on device
+ for(auto reader : para->getParH(level)->transientBCInputFileReader)
+ {
+ reader->getNextData(para->getParH(level)->precursorBC.next, para->getParH(level)->precursorBC.numberOfPrecursorNodes, 0);
+ }
+
+ cudaMemoryManager->cudaCopyPrecursorData(level);
+
+ //switch next with last pointers
+ real* tmp = para->getParD(level)->precursorBC.last;
+ para->getParD(level)->precursorBC.last = para->getParD(level)->precursorBC.next;
+ para->getParD(level)->precursorBC.next = tmp;
+
+ //read second timestep of precursor into next and copy next to device
+ real nextTime = para->getParD(level)->precursorBC.timeStepsBetweenReads*pow(2,-((real)level))*para->getTimeRatio();
+ for(auto reader : para->getParH(level)->transientBCInputFileReader)
+ {
+ reader->getNextData(para->getParH(level)->precursorBC.next, para->getParH(level)->precursorBC.numberOfPrecursorNodes, nextTime);
+ }
+
+ cudaMemoryManager->cudaCopyPrecursorData(level);
+
+ para->getParD(level)->precursorBC.nPrecursorReads = 1;
+
+
+ //switch next with current pointers
+ tmp = para->getParD(level)->precursorBC.current;
+ para->getParD(level)->precursorBC.current = para->getParD(level)->precursorBC.next;
+ para->getParD(level)->precursorBC.next = tmp;
+
+ //start usual cycle of loading, i.e. read velocities of timestep after current and copy asynchronously to device
+ for(auto reader : para->getParH(level)->transientBCInputFileReader)
+ {
+ reader->getNextData(para->getParH(level)->precursorBC.next, para->getParH(level)->precursorBC.numberOfPrecursorNodes, 2*nextTime);
+ }
+
+ cudaMemoryManager->cudaCopyPrecursorData(level);
+
+ para->getParD(level)->precursorBC.nPrecursorReads = 2;
+ }
+
+ ////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
+ // advection - diffusion stuff
+ if (para->getDiffOn()==true){
+ throw std::runtime_error(" Advection Diffusion not implemented for Precursor!");
+ }
+ //////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
+ }
+
if (builder->hasGeometryValues()) {
@@ -303,7 +481,7 @@ void GridGenerator::initalValuesDomainDecompostion()
if (para->getNumprocs() < 2)
return;
if ((para->getNumprocs() > 1) /*&& (procNeighborsSendX.size() == procNeighborsRecvX.size())*/) {
-
+
// direction has to be changed in case of periodic BCs and multiple sub domains
std::vector<int> fillOrder = { 0, 1, 2, 3, 4, 5 };
@@ -383,7 +561,7 @@ void GridGenerator::initalValuesDomainDecompostion()
builder->getReceiveIndices(para->getParH(level)->recvProcessNeighborX[indexProcessNeighbor].index, direction,
level);
if (level != builder->getNumberOfGridLevels() - 1 && para->useReducedCommunicationAfterFtoC)
- indexRearrangement->initCommunicationArraysForCommAfterFinetoCoarseX(level, indexProcessNeighbor, direction);
+ indexRearrangement->initCommunicationArraysForCommAfterFinetoCoarseX(level, indexProcessNeighbor, direction);
////////////////////////////////////////////////////////////////////////////////////////
cudaMemoryManager->cudaCopyProcessNeighborXIndex(level, indexProcessNeighbor);
////////////////////////////////////////////////////////////////////////////////////////
@@ -446,7 +624,7 @@ void GridGenerator::initalValuesDomainDecompostion()
////////////////////////////////////////////////////////////////////////////////////////
// malloc on host and device
cudaMemoryManager->cudaAllocProcessNeighborY(level, indexProcessNeighbor);
- ////////////////////////////////////////////////////////////////////////////////////////
+ ////////////////////////////////////////////////////////////////////////////////////////
// init index arrays
builder->getSendIndices(para->getParH(level)->sendProcessNeighborY[indexProcessNeighbor].index, direction, level);
builder->getReceiveIndices(para->getParH(level)->recvProcessNeighborY[indexProcessNeighbor].index, direction,
@@ -465,7 +643,7 @@ void GridGenerator::initalValuesDomainDecompostion()
if (tempSend > 0) {
int indexProcessNeighbor = (int)para->getParH(level)->sendProcessNeighborZ.size();
-
+
para->getParH(level)->sendProcessNeighborZ.emplace_back();
para->getParD(level)->sendProcessNeighborZ.emplace_back();
para->getParH(level)->recvProcessNeighborZ.emplace_back();
@@ -755,9 +933,9 @@ void GridGenerator::allocArrays_BoundaryQs()
//preprocessing
real* QQ = para->getParH(i)->pressureBC.q27[0];
unsigned int sizeQ = para->getParH(i)->pressureBC.numberOfBCnodes;
- QforBoundaryConditions Q;
+ QforBoundaryConditions &Q = para->getParH(i)->pressureBC;
getPointersToBoundaryConditions(Q, QQ, sizeQ);
-
+
builder->getPressureQs(Q.q27, i);
@@ -802,9 +980,9 @@ void GridGenerator::allocArrays_BoundaryQs()
//preprocessing
real* QQ = para->getParH(i)->slipBC.q27[0];
unsigned int sizeQ = para->getParH(i)->slipBC.numberOfBCnodes;
- QforBoundaryConditions Q;
+ QforBoundaryConditions &Q = para->getParH(i)->slipBC;
getPointersToBoundaryConditions(Q, QQ, sizeQ);
-
+
builder->getSlipQs(Q.q27, i);
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
cudaMemoryManager->cudaCopySlipBC(i);
@@ -822,9 +1000,9 @@ void GridGenerator::allocArrays_BoundaryQs()
//preprocessing
real* QQ = para->getParH(i)->stressBC.q27[0];
unsigned int sizeQ = para->getParH(i)->stressBC.numberOfBCnodes;
- QforBoundaryConditions Q;
+ QforBoundaryConditions &Q = para->getParH(i)->stressBC;
getPointersToBoundaryConditions(Q, QQ, sizeQ);
-
+
builder->getStressQs(Q.q27, i);
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
cudaMemoryManager->cudaCopyStressBC(i);
@@ -842,7 +1020,7 @@ void GridGenerator::allocArrays_BoundaryQs()
//preprocessing
real* QQ = para->getParH(i)->velocityBC.q27[0];
unsigned int sizeQ = para->getParH(i)->velocityBC.numberOfBCnodes;
- QforBoundaryConditions Q;
+ QforBoundaryConditions &Q = para->getParH(i)->velocityBC;
getPointersToBoundaryConditions(Q, QQ, sizeQ);
builder->getVelocityQs(Q.q27, i);
@@ -874,6 +1052,50 @@ void GridGenerator::allocArrays_BoundaryQs()
}
}
+ for (uint i = 0; i < builder->getNumberOfGridLevels(); i++) {
+ const auto numberOfPrecursorNodes = int(builder->getPrecursorSize(i));
+ if (numberOfPrecursorNodes > 0)
+ {
+ std::cout << "size velocity level " << i << " : " << numberOfPrecursorNodes << std::endl;
+ ////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
+ //preprocessing
+ real* QQ = para->getParH(i)->precursorBC.q27[0];
+ unsigned int sizeQ = para->getParH(i)->precursorBC.numberOfBCnodes;
+ QforBoundaryConditions Q;
+ getPointersToBoundaryConditions(Q, QQ, sizeQ);
+
+ builder->getPrecursorQs(Q.q27, i);
+
+ if (para->getDiffOn()) {
+ throw std::runtime_error("Advection diffusion not implemented for Precursor!");
+ //////////////////////////////////////////////////////////////////////////
+ // para->getParH(i)->TempVel.kTemp = numberOfVelocityNodes;
+ // para->getParD(i)->TempVel.kTemp = numberOfVelocityNodes;
+ // std::cout << "Groesse TempVel.kTemp = " << para->getParH(i)->TempPress.kTemp << std::endl;
+ // std::cout << "getTemperatureInit = " << para->getTemperatureInit() << std::endl;
+ // std::cout << "getTemperatureBC = " << para->getTemperatureBC() << std::endl;
+ // //////////////////////////////////////////////////////////////////////////
+ // cudaMemoryManager->cudaAllocTempVeloBC(i);
+ // //cout << "nach alloc " << std::endl;
+ // //////////////////////////////////////////////////////////////////////////
+ // for (int m = 0; m < numberOfVelocityNodes; m++)
+ // {
+ // para->getParH(i)->TempVel.temp[m] = para->getTemperatureInit();
+ // para->getParH(i)->TempVel.tempPulse[m] = para->getTemperatureBC();
+ // para->getParH(i)->TempVel.velo[m] = para->getVelocity();
+ // para->getParH(i)->TempVel.k[m] = para->getParH(i)->Qinflow.k[m];
+ // }
+ // //////////////////////////////////////////////////////////////////////////
+ // //cout << "vor copy " << std::endl;
+ // cudaMemoryManager->cudaCopyTempVeloBCHD(i);
+ // //cout << "nach copy " << std::endl;
+ //////////////////////////////////////////////////////////////////////////
+ }
+ cudaMemoryManager->cudaCopyPrecursorBC(i);
+ }
+ }
+
+
for (uint i = 0; i < builder->getNumberOfGridLevels(); i++) {
const int numberOfGeometryNodes = builder->getGeometrySize(i);
@@ -898,7 +1120,7 @@ void GridGenerator::allocArrays_BoundaryQs()
//preprocessing
real* QQ = para->getParH(i)->geometryBC.q27[0];
unsigned int sizeQ = para->getParH(i)->geometryBC.numberOfBCnodes;
- QforBoundaryConditions Q;
+ QforBoundaryConditions &Q = para->getParH(i)->geometryBC;
getPointersToBoundaryConditions(Q, QQ, sizeQ);
//////////////////////////////////////////////////////////////////
@@ -948,7 +1170,7 @@ void GridGenerator::allocArrays_BoundaryQs()
void GridGenerator::allocArrays_OffsetScale()
{
- for (uint level = 0; level < builder->getNumberOfGridLevels() - 1; level++)
+ for (uint level = 0; level < builder->getNumberOfGridLevels() - 1; level++)
{
const uint numberOfNodesPerLevelCF = builder->getNumberOfNodesCF(level);
const uint numberOfNodesPerLevelFC = builder->getNumberOfNodesFC(level);
@@ -987,7 +1209,7 @@ void GridGenerator::allocArrays_OffsetScale()
builder->getOffsetCF(para->getParH(level)->offCF.xOffCF, para->getParH(level)->offCF.yOffCF, para->getParH(level)->offCF.zOffCF, level);
builder->getOffsetFC(para->getParH(level)->offFC.xOffFC, para->getParH(level)->offFC.yOffFC, para->getParH(level)->offFC.zOffFC, level);
builder->getGridInterfaceIndices(para->getParH(level)->intCF.ICellCFC, para->getParH(level)->intCF.ICellCFF, para->getParH(level)->intFC.ICellFCC, para->getParH(level)->intFC.ICellFCF, level);
-
+
if (para->getUseStreams() || para->getNumprocs() > 1) {
// split fine-to-coarse indices into border and bulk
interpolationGrouper->splitFineToCoarseIntoBorderAndBulk(level);
@@ -1060,8 +1282,8 @@ std::string GridGenerator::verifyNeighborIndices(int level) const
int wrongNeighbors = 0;
int stopperNodes = 0;
- for (uint index = 0; index < para->getParH(level)->numberOfNodes; index++)
- oss << verifyNeighborIndex(level, index, invalidNodes, stopperNodes, wrongNeighbors);
+ for (size_t index = 0; index < para->getParH(level)->numberOfNodes; index++)
+ oss << verifyNeighborIndex(level, (int)index, invalidNodes, stopperNodes, wrongNeighbors);
oss << "invalid nodes found: " << invalidNodes << "\n";
@@ -1090,7 +1312,7 @@ std::string GridGenerator::verifyNeighborIndex(int level, int index , int &inval
//std::cout << para->getParH(level)->coordinateX[1] << ", " << para->getParH(level)->coordinateY[1] << ", " << para->getParH(level)->coordinateZ[1] << std::endl;
//std::cout << para->getParH(level)->coordinateX[para->getParH(level)->numberOfNodes - 1] << ", " << para->getParH(level)->coordinateY[para->getParH(level)->numberOfNodes - 1] << ", " << para->getParH(level)->coordinateZ[para->getParH(level)->numberOfNodes - 1] << std::endl;
-
+
real maxX = para->getParH(level)->coordinateX[para->getParH(level)->numberOfNodes - 1] - delta;
real maxY = para->getParH(level)->coordinateY[para->getParH(level)->numberOfNodes - 1] - delta;
real maxZ = para->getParH(level)->coordinateZ[para->getParH(level)->numberOfNodes - 1] - delta;
@@ -1131,8 +1353,8 @@ std::string GridGenerator::checkNeighbor(int level, real x, real y, real z, int
if (!neighborValid) {
oss << "NeighborX invalid from: (" << x << ", " << y << ", " << z << "), index: " << index << ", "
- << direction << " neighborIndex: " << neighborIndex <<
- ", actual neighborCoords : (" << neighborCoordX << ", " << neighborCoordY << ", " << neighborCoordZ <<
+ << direction << " neighborIndex: " << neighborIndex <<
+ ", actual neighborCoords : (" << neighborCoordX << ", " << neighborCoordY << ", " << neighborCoordZ <<
"), expected neighborCoords : (" << neighborX << ", " << neighborY << ", " << neighborZ << ")\n";
numberOfWrongNeihgbors++;
}
@@ -1140,31 +1362,31 @@ std::string GridGenerator::checkNeighbor(int level, real x, real y, real z, int
}
void GridGenerator::getPointersToBoundaryConditions(QforBoundaryConditions& boundaryConditionStruct, real* subgridDistances, const unsigned int numberOfBCnodes){
- boundaryConditionStruct.q27[DIR_P00] = &subgridDistances[DIR_P00 * numberOfBCnodes];
- boundaryConditionStruct.q27[DIR_M00] = &subgridDistances[DIR_M00 * numberOfBCnodes];
- boundaryConditionStruct.q27[DIR_0P0] = &subgridDistances[DIR_0P0 * numberOfBCnodes];
- boundaryConditionStruct.q27[DIR_0M0] = &subgridDistances[DIR_0M0 * numberOfBCnodes];
- boundaryConditionStruct.q27[DIR_00P] = &subgridDistances[DIR_00P * numberOfBCnodes];
- boundaryConditionStruct.q27[DIR_00M] = &subgridDistances[DIR_00M * numberOfBCnodes];
- boundaryConditionStruct.q27[DIR_PP0] = &subgridDistances[DIR_PP0 * numberOfBCnodes];
- boundaryConditionStruct.q27[DIR_MM0] = &subgridDistances[DIR_MM0 * numberOfBCnodes];
- boundaryConditionStruct.q27[DIR_PM0] = &subgridDistances[DIR_PM0 * numberOfBCnodes];
- boundaryConditionStruct.q27[DIR_MP0] = &subgridDistances[DIR_MP0 * numberOfBCnodes];
- boundaryConditionStruct.q27[DIR_P0P] = &subgridDistances[DIR_P0P * numberOfBCnodes];
- boundaryConditionStruct.q27[DIR_M0M] = &subgridDistances[DIR_M0M * numberOfBCnodes];
- boundaryConditionStruct.q27[DIR_P0M] = &subgridDistances[DIR_P0M * numberOfBCnodes];
- boundaryConditionStruct.q27[DIR_M0P] = &subgridDistances[DIR_M0P * numberOfBCnodes];
- boundaryConditionStruct.q27[DIR_0PP] = &subgridDistances[DIR_0PP * numberOfBCnodes];
- boundaryConditionStruct.q27[DIR_0MM] = &subgridDistances[DIR_0MM * numberOfBCnodes];
- boundaryConditionStruct.q27[DIR_0PM] = &subgridDistances[DIR_0PM * numberOfBCnodes];
- boundaryConditionStruct.q27[DIR_0MP] = &subgridDistances[DIR_0MP * numberOfBCnodes];
- boundaryConditionStruct.q27[DIR_000] = &subgridDistances[DIR_000* numberOfBCnodes];
- boundaryConditionStruct.q27[DIR_PPP] = &subgridDistances[DIR_PPP * numberOfBCnodes];
- boundaryConditionStruct.q27[DIR_MMP] = &subgridDistances[DIR_MMP * numberOfBCnodes];
- boundaryConditionStruct.q27[DIR_PMP] = &subgridDistances[DIR_PMP * numberOfBCnodes];
- boundaryConditionStruct.q27[DIR_MPP] = &subgridDistances[DIR_MPP * numberOfBCnodes];
- boundaryConditionStruct.q27[DIR_PPM] = &subgridDistances[DIR_PPM * numberOfBCnodes];
- boundaryConditionStruct.q27[DIR_MMM] = &subgridDistances[DIR_MMM * numberOfBCnodes];
- boundaryConditionStruct.q27[DIR_PMM] = &subgridDistances[DIR_PMM * numberOfBCnodes];
- boundaryConditionStruct.q27[DIR_MPM] = &subgridDistances[DIR_MPM * numberOfBCnodes];
-}
\ No newline at end of file
+ boundaryConditionStruct.q27[DIR_P00] = &subgridDistances[DIR_P00 * numberOfBCnodes];
+ boundaryConditionStruct.q27[DIR_M00] = &subgridDistances[DIR_M00 * numberOfBCnodes];
+ boundaryConditionStruct.q27[DIR_0P0] = &subgridDistances[DIR_0P0 * numberOfBCnodes];
+ boundaryConditionStruct.q27[DIR_0M0] = &subgridDistances[DIR_0M0 * numberOfBCnodes];
+ boundaryConditionStruct.q27[DIR_00P] = &subgridDistances[DIR_00P * numberOfBCnodes];
+ boundaryConditionStruct.q27[DIR_00M] = &subgridDistances[DIR_00M * numberOfBCnodes];
+ boundaryConditionStruct.q27[DIR_PP0] = &subgridDistances[DIR_PP0 * numberOfBCnodes];
+ boundaryConditionStruct.q27[DIR_MM0] = &subgridDistances[DIR_MM0 * numberOfBCnodes];
+ boundaryConditionStruct.q27[DIR_PM0] = &subgridDistances[DIR_PM0 * numberOfBCnodes];
+ boundaryConditionStruct.q27[DIR_MP0] = &subgridDistances[DIR_MP0 * numberOfBCnodes];
+ boundaryConditionStruct.q27[DIR_P0P] = &subgridDistances[DIR_P0P * numberOfBCnodes];
+ boundaryConditionStruct.q27[DIR_M0M] = &subgridDistances[DIR_M0M * numberOfBCnodes];
+ boundaryConditionStruct.q27[DIR_P0M] = &subgridDistances[DIR_P0M * numberOfBCnodes];
+ boundaryConditionStruct.q27[DIR_M0P] = &subgridDistances[DIR_M0P * numberOfBCnodes];
+ boundaryConditionStruct.q27[DIR_0PP] = &subgridDistances[DIR_0PP * numberOfBCnodes];
+ boundaryConditionStruct.q27[DIR_0MM] = &subgridDistances[DIR_0MM * numberOfBCnodes];
+ boundaryConditionStruct.q27[DIR_0PM] = &subgridDistances[DIR_0PM * numberOfBCnodes];
+ boundaryConditionStruct.q27[DIR_0MP] = &subgridDistances[DIR_0MP * numberOfBCnodes];
+ boundaryConditionStruct.q27[DIR_000] = &subgridDistances[DIR_000 * numberOfBCnodes];
+ boundaryConditionStruct.q27[DIR_PPP] = &subgridDistances[DIR_PPP * numberOfBCnodes];
+ boundaryConditionStruct.q27[DIR_MMP] = &subgridDistances[DIR_MMP * numberOfBCnodes];
+ boundaryConditionStruct.q27[DIR_PMP] = &subgridDistances[DIR_PMP * numberOfBCnodes];
+ boundaryConditionStruct.q27[DIR_MPP] = &subgridDistances[DIR_MPP * numberOfBCnodes];
+ boundaryConditionStruct.q27[DIR_PPM] = &subgridDistances[DIR_PPM * numberOfBCnodes];
+ boundaryConditionStruct.q27[DIR_MMM] = &subgridDistances[DIR_MMM * numberOfBCnodes];
+ boundaryConditionStruct.q27[DIR_PMM] = &subgridDistances[DIR_PMM * numberOfBCnodes];
+ boundaryConditionStruct.q27[DIR_MPM] = &subgridDistances[DIR_MPM * numberOfBCnodes];
+}
diff --git a/src/gpu/VirtualFluids_GPU/DataStructureInitializer/GridReaderGenerator/GridGenerator.h b/src/gpu/VirtualFluids_GPU/DataStructureInitializer/GridReaderGenerator/GridGenerator.h
index d2f56e1df4ee5658c61b8e8a3e94a820d1a4f2f1..c97ed02a64da1d5fafa18150c75d149f96484d44 100644
--- a/src/gpu/VirtualFluids_GPU/DataStructureInitializer/GridReaderGenerator/GridGenerator.h
+++ b/src/gpu/VirtualFluids_GPU/DataStructureInitializer/GridReaderGenerator/GridGenerator.h
@@ -40,6 +40,7 @@
#include "LBM/LB.h"
+
class Parameter;
class GridBuilder;
class IndexRearrangementForStreams;
@@ -75,8 +76,10 @@ public:
//! \brief allocates and initialized the sub-grid distances at the boundary conditions
void allocArrays_BoundaryQs() override;
void allocArrays_OffsetScale() override;
- void allocArrays_fluidNodeIndices() override;
- void allocArrays_fluidNodeIndicesBorder() override;
+ void allocArrays_taggedFluidNodes() override;
+
+ void tagFluidNodeIndices(const std::vector<uint>& taggedFluidNodeIndices, CollisionTemplate tag, uint level) override;
+ void sortFluidNodeTags() override;
virtual void setDimensions() override;
virtual void setBoundingBox() override;
diff --git a/src/gpu/VirtualFluids_GPU/Factories/BoundaryConditionFactory.cpp b/src/gpu/VirtualFluids_GPU/Factories/BoundaryConditionFactory.cpp
index bff054eb174a0f5fa34119deedde6f1c9733d83c..b1c398638cff1ec1b6d52f59f8e773183e270331 100644
--- a/src/gpu/VirtualFluids_GPU/Factories/BoundaryConditionFactory.cpp
+++ b/src/gpu/VirtualFluids_GPU/Factories/BoundaryConditionFactory.cpp
@@ -35,6 +35,11 @@ void BoundaryConditionFactory::setStressBoundaryCondition(const StressBC boundar
this->stressBoundaryCondition = boundaryConditionType;
}
+void BoundaryConditionFactory::setPrecursorBoundaryCondition(const PrecursorBC boundaryConditionType)
+{
+ this->precursorBoundaryCondition = boundaryConditionType;
+}
+
boundaryCondition BoundaryConditionFactory::getVelocityBoundaryConditionPost(bool isGeometryBC) const
{
const VelocityBC &boundaryCondition =
@@ -132,6 +137,22 @@ boundaryCondition BoundaryConditionFactory::getPressureBoundaryConditionPre() co
case PressureBC::OutflowNonReflective:
return QPressNoRhoDev27;
break;
+ case PressureBC::OutflowNonReflectivePressureCorrection:
+ return QPressZeroRhoOutflowDev27;
+ default:
+ return nullptr;
+ }
+}
+
+precursorBoundaryConditionFunc BoundaryConditionFactory::getPrecursorBoundaryConditionPost() const
+{
+ switch (this->precursorBoundaryCondition) {
+ case PrecursorBC::VelocityPrecursor:
+ return QPrecursorDevCompZeroPress;
+ break;
+ case PrecursorBC::DistributionsPrecursor:
+ return PrecursorDevDistributions;
+ break;
default:
return nullptr;
}
diff --git a/src/gpu/VirtualFluids_GPU/Factories/BoundaryConditionFactory.h b/src/gpu/VirtualFluids_GPU/Factories/BoundaryConditionFactory.h
index 9d6872c4847be72dff4be7137b774c8082e39e34..c6877cbfeffe5b32c0c2d336e46b02d68cd946a3 100644
--- a/src/gpu/VirtualFluids_GPU/Factories/BoundaryConditionFactory.h
+++ b/src/gpu/VirtualFluids_GPU/Factories/BoundaryConditionFactory.h
@@ -42,11 +42,13 @@
#include "Parameter/Parameter.h"
#include "gpu/GridGenerator/grid/BoundaryConditions/Side.h"
+
struct LBMSimulationParameter;
class Parameter;
using boundaryCondition = std::function<void(LBMSimulationParameter *, QforBoundaryConditions *)>;
using boundaryConditionWithParameter = std::function<void(Parameter *, QforBoundaryConditions *, const int level)>;
+using precursorBoundaryConditionFunc = std::function<void(LBMSimulationParameter *, QforPrecursorBoundaryConditions *, real timeRatio, real velocityRatio)>;
class BoundaryConditionFactory
{
@@ -109,6 +111,8 @@ public:
PressureNonEquilibriumCompressible,
//! - OutflowNonReflective = outflow boundary condition, should be combined with VelocityAndPressureCompressible
OutflowNonReflective,
+ //! - OutflowNonreflectivePressureCorrection = like OutflowNonReflective, but also reduces pressure overshoot
+ OutflowNonReflectivePressureCorrection,
//! - NotSpecified = the user did not set a boundary condition
NotSpecified
};
@@ -128,11 +132,21 @@ public:
// enum class OutflowBoundaryCondition {}; // TODO:
// https://git.rz.tu-bs.de/m.schoenherr/VirtualFluids_dev/-/issues/16
+ enum class PrecursorBC {
+ //! - VelocityPrecursor
+ VelocityPrecursor,
+ //! - DisitributionsPrecursor
+ DistributionsPrecursor,
+ //! - NotSpecified = the user did not set a boundary condition
+ NotSpecified
+ };
+
void setVelocityBoundaryCondition(const BoundaryConditionFactory::VelocityBC boundaryConditionType);
void setNoSlipBoundaryCondition(const BoundaryConditionFactory::NoSlipBC boundaryConditionType);
void setSlipBoundaryCondition(const BoundaryConditionFactory::SlipBC boundaryConditionType);
void setPressureBoundaryCondition(const BoundaryConditionFactory::PressureBC boundaryConditionType);
void setStressBoundaryCondition(const BoundaryConditionFactory::StressBC boundaryConditionType);
+ void setPrecursorBoundaryCondition(const BoundaryConditionFactory::PrecursorBC boundaryConditionType);
//! \brief set a boundary condition for the geometry
//! param boundaryConditionType: a velocity, no-slip or slip boundary condition
//! \details suggestions for boundaryConditionType:
@@ -152,6 +166,8 @@ public:
[[nodiscard]] boundaryCondition getSlipBoundaryConditionPost(bool isGeometryBC = false) const;
[[nodiscard]] boundaryCondition getPressureBoundaryConditionPre() const;
[[nodiscard]] boundaryCondition getGeometryBoundaryConditionPost() const;
+ [[nodiscard]] precursorBoundaryConditionFunc getPrecursorBoundaryConditionPost() const;
+
[[nodiscard]] boundaryConditionWithParameter getStressBoundaryConditionPost() const;
@@ -162,6 +178,7 @@ private:
PressureBC pressureBoundaryCondition = PressureBC::NotSpecified;
std::variant<VelocityBC, NoSlipBC, SlipBC> geometryBoundaryCondition = NoSlipBC::NoSlipImplicitBounceBack;
StressBC stressBoundaryCondition = StressBC::NotSpecified;
+ PrecursorBC precursorBoundaryCondition = PrecursorBC::NotSpecified;
// OutflowBoundaryConditon outflowBC // TODO: https://git.rz.tu-bs.de/m.schoenherr/VirtualFluids_dev/-/issues/16
};
diff --git a/src/gpu/VirtualFluids_GPU/GPU/AdvectionDiffusion27chim.cu b/src/gpu/VirtualFluids_GPU/GPU/AdvectionDiffusion27chim.cu
index 8f54358e04063c9063c873caf02a86e76bb7f936..04f6afe4cf9ebd99dc293ded16f55a56f0d77036 100644
--- a/src/gpu/VirtualFluids_GPU/GPU/AdvectionDiffusion27chim.cu
+++ b/src/gpu/VirtualFluids_GPU/GPU/AdvectionDiffusion27chim.cu
@@ -74,7 +74,7 @@ __global__ void Factorized_Central_Moments_Advection_Diffusion_Device_Kernel(
uint* neighborZ,
real* distributions,
real* distributionsAD,
- int size_Mat,
+ unsigned long long numberOfLBnodes,
real* forces,
bool isEvenTimestep)
{
@@ -100,7 +100,7 @@ __global__ void Factorized_Central_Moments_Advection_Diffusion_Device_Kernel(
//////////////////////////////////////////////////////////////////////////
// run for all indices in size_Mat and fluid nodes
- if ((k < size_Mat) && (typeOfGridNode[k] == GEO_FLUID))
+ if ((k < numberOfLBnodes) && (typeOfGridNode[k] == GEO_FLUID))
{
//////////////////////////////////////////////////////////////////////////
//! - Read distributions: style of reading and writing the distributions from/to stored arrays dependent on timestep is based on the esoteric twist algorithm \ref
@@ -109,125 +109,125 @@ __global__ void Factorized_Central_Moments_Advection_Diffusion_Device_Kernel(
Distributions27 dist;
if (isEvenTimestep)
{
- dist.f[DIR_P00 ] = &distributions[DIR_P00 *size_Mat];
- dist.f[DIR_M00 ] = &distributions[DIR_M00 *size_Mat];
- dist.f[DIR_0P0 ] = &distributions[DIR_0P0 *size_Mat];
- dist.f[DIR_0M0 ] = &distributions[DIR_0M0 *size_Mat];
- dist.f[DIR_00P ] = &distributions[DIR_00P *size_Mat];
- dist.f[DIR_00M ] = &distributions[DIR_00M *size_Mat];
- dist.f[DIR_PP0 ] = &distributions[DIR_PP0 *size_Mat];
- dist.f[DIR_MM0 ] = &distributions[DIR_MM0 *size_Mat];
- dist.f[DIR_PM0 ] = &distributions[DIR_PM0 *size_Mat];
- dist.f[DIR_MP0 ] = &distributions[DIR_MP0 *size_Mat];
- dist.f[DIR_P0P ] = &distributions[DIR_P0P *size_Mat];
- dist.f[DIR_M0M ] = &distributions[DIR_M0M *size_Mat];
- dist.f[DIR_P0M ] = &distributions[DIR_P0M *size_Mat];
- dist.f[DIR_M0P ] = &distributions[DIR_M0P *size_Mat];
- dist.f[DIR_0PP ] = &distributions[DIR_0PP *size_Mat];
- dist.f[DIR_0MM ] = &distributions[DIR_0MM *size_Mat];
- dist.f[DIR_0PM ] = &distributions[DIR_0PM *size_Mat];
- dist.f[DIR_0MP ] = &distributions[DIR_0MP *size_Mat];
- dist.f[DIR_000] = &distributions[DIR_000*size_Mat];
- dist.f[DIR_PPP ] = &distributions[DIR_PPP *size_Mat];
- dist.f[DIR_MMP ] = &distributions[DIR_MMP *size_Mat];
- dist.f[DIR_PMP ] = &distributions[DIR_PMP *size_Mat];
- dist.f[DIR_MPP ] = &distributions[DIR_MPP *size_Mat];
- dist.f[DIR_PPM ] = &distributions[DIR_PPM *size_Mat];
- dist.f[DIR_MMM ] = &distributions[DIR_MMM *size_Mat];
- dist.f[DIR_PMM ] = &distributions[DIR_PMM *size_Mat];
- dist.f[DIR_MPM ] = &distributions[DIR_MPM *size_Mat];
+ dist.f[DIR_P00] = &distributions[DIR_P00 * numberOfLBnodes];
+ dist.f[DIR_M00] = &distributions[DIR_M00 * numberOfLBnodes];
+ dist.f[DIR_0P0] = &distributions[DIR_0P0 * numberOfLBnodes];
+ dist.f[DIR_0M0] = &distributions[DIR_0M0 * numberOfLBnodes];
+ dist.f[DIR_00P] = &distributions[DIR_00P * numberOfLBnodes];
+ dist.f[DIR_00M] = &distributions[DIR_00M * numberOfLBnodes];
+ dist.f[DIR_PP0] = &distributions[DIR_PP0 * numberOfLBnodes];
+ dist.f[DIR_MM0] = &distributions[DIR_MM0 * numberOfLBnodes];
+ dist.f[DIR_PM0] = &distributions[DIR_PM0 * numberOfLBnodes];
+ dist.f[DIR_MP0] = &distributions[DIR_MP0 * numberOfLBnodes];
+ dist.f[DIR_P0P] = &distributions[DIR_P0P * numberOfLBnodes];
+ dist.f[DIR_M0M] = &distributions[DIR_M0M * numberOfLBnodes];
+ dist.f[DIR_P0M] = &distributions[DIR_P0M * numberOfLBnodes];
+ dist.f[DIR_M0P] = &distributions[DIR_M0P * numberOfLBnodes];
+ dist.f[DIR_0PP] = &distributions[DIR_0PP * numberOfLBnodes];
+ dist.f[DIR_0MM] = &distributions[DIR_0MM * numberOfLBnodes];
+ dist.f[DIR_0PM] = &distributions[DIR_0PM * numberOfLBnodes];
+ dist.f[DIR_0MP] = &distributions[DIR_0MP * numberOfLBnodes];
+ dist.f[DIR_000] = &distributions[DIR_000 * numberOfLBnodes];
+ dist.f[DIR_PPP] = &distributions[DIR_PPP * numberOfLBnodes];
+ dist.f[DIR_MMP] = &distributions[DIR_MMP * numberOfLBnodes];
+ dist.f[DIR_PMP] = &distributions[DIR_PMP * numberOfLBnodes];
+ dist.f[DIR_MPP] = &distributions[DIR_MPP * numberOfLBnodes];
+ dist.f[DIR_PPM] = &distributions[DIR_PPM * numberOfLBnodes];
+ dist.f[DIR_MMM] = &distributions[DIR_MMM * numberOfLBnodes];
+ dist.f[DIR_PMM] = &distributions[DIR_PMM * numberOfLBnodes];
+ dist.f[DIR_MPM] = &distributions[DIR_MPM * numberOfLBnodes];
}
else
{
- dist.f[DIR_M00 ] = &distributions[DIR_P00 *size_Mat];
- dist.f[DIR_P00 ] = &distributions[DIR_M00 *size_Mat];
- dist.f[DIR_0M0 ] = &distributions[DIR_0P0 *size_Mat];
- dist.f[DIR_0P0 ] = &distributions[DIR_0M0 *size_Mat];
- dist.f[DIR_00M ] = &distributions[DIR_00P *size_Mat];
- dist.f[DIR_00P ] = &distributions[DIR_00M *size_Mat];
- dist.f[DIR_MM0 ] = &distributions[DIR_PP0 *size_Mat];
- dist.f[DIR_PP0 ] = &distributions[DIR_MM0 *size_Mat];
- dist.f[DIR_MP0 ] = &distributions[DIR_PM0 *size_Mat];
- dist.f[DIR_PM0 ] = &distributions[DIR_MP0 *size_Mat];
- dist.f[DIR_M0M ] = &distributions[DIR_P0P *size_Mat];
- dist.f[DIR_P0P ] = &distributions[DIR_M0M *size_Mat];
- dist.f[DIR_M0P ] = &distributions[DIR_P0M *size_Mat];
- dist.f[DIR_P0M ] = &distributions[DIR_M0P *size_Mat];
- dist.f[DIR_0MM ] = &distributions[DIR_0PP *size_Mat];
- dist.f[DIR_0PP ] = &distributions[DIR_0MM *size_Mat];
- dist.f[DIR_0MP ] = &distributions[DIR_0PM *size_Mat];
- dist.f[DIR_0PM ] = &distributions[DIR_0MP *size_Mat];
- dist.f[DIR_000] = &distributions[DIR_000*size_Mat];
- dist.f[DIR_MMM ] = &distributions[DIR_PPP *size_Mat];
- dist.f[DIR_PPM ] = &distributions[DIR_MMP *size_Mat];
- dist.f[DIR_MPM ] = &distributions[DIR_PMP *size_Mat];
- dist.f[DIR_PMM ] = &distributions[DIR_MPP *size_Mat];
- dist.f[DIR_MMP ] = &distributions[DIR_PPM *size_Mat];
- dist.f[DIR_PPP ] = &distributions[DIR_MMM *size_Mat];
- dist.f[DIR_MPP ] = &distributions[DIR_PMM *size_Mat];
- dist.f[DIR_PMP ] = &distributions[DIR_MPM *size_Mat];
+ dist.f[DIR_M00] = &distributions[DIR_P00 * numberOfLBnodes];
+ dist.f[DIR_P00] = &distributions[DIR_M00 * numberOfLBnodes];
+ dist.f[DIR_0M0] = &distributions[DIR_0P0 * numberOfLBnodes];
+ dist.f[DIR_0P0] = &distributions[DIR_0M0 * numberOfLBnodes];
+ dist.f[DIR_00M] = &distributions[DIR_00P * numberOfLBnodes];
+ dist.f[DIR_00P] = &distributions[DIR_00M * numberOfLBnodes];
+ dist.f[DIR_MM0] = &distributions[DIR_PP0 * numberOfLBnodes];
+ dist.f[DIR_PP0] = &distributions[DIR_MM0 * numberOfLBnodes];
+ dist.f[DIR_MP0] = &distributions[DIR_PM0 * numberOfLBnodes];
+ dist.f[DIR_PM0] = &distributions[DIR_MP0 * numberOfLBnodes];
+ dist.f[DIR_M0M] = &distributions[DIR_P0P * numberOfLBnodes];
+ dist.f[DIR_P0P] = &distributions[DIR_M0M * numberOfLBnodes];
+ dist.f[DIR_M0P] = &distributions[DIR_P0M * numberOfLBnodes];
+ dist.f[DIR_P0M] = &distributions[DIR_M0P * numberOfLBnodes];
+ dist.f[DIR_0MM] = &distributions[DIR_0PP * numberOfLBnodes];
+ dist.f[DIR_0PP] = &distributions[DIR_0MM * numberOfLBnodes];
+ dist.f[DIR_0MP] = &distributions[DIR_0PM * numberOfLBnodes];
+ dist.f[DIR_0PM] = &distributions[DIR_0MP * numberOfLBnodes];
+ dist.f[DIR_000] = &distributions[DIR_000 * numberOfLBnodes];
+ dist.f[DIR_MMM] = &distributions[DIR_PPP * numberOfLBnodes];
+ dist.f[DIR_PPM] = &distributions[DIR_MMP * numberOfLBnodes];
+ dist.f[DIR_MPM] = &distributions[DIR_PMP * numberOfLBnodes];
+ dist.f[DIR_PMM] = &distributions[DIR_MPP * numberOfLBnodes];
+ dist.f[DIR_MMP] = &distributions[DIR_PPM * numberOfLBnodes];
+ dist.f[DIR_PPP] = &distributions[DIR_MMM * numberOfLBnodes];
+ dist.f[DIR_MPP] = &distributions[DIR_PMM * numberOfLBnodes];
+ dist.f[DIR_PMP] = &distributions[DIR_MPM * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////
Distributions27 distAD;
if (isEvenTimestep)
{
- distAD.f[DIR_P00 ] = &distributionsAD[DIR_P00 *size_Mat];
- distAD.f[DIR_M00 ] = &distributionsAD[DIR_M00 *size_Mat];
- distAD.f[DIR_0P0 ] = &distributionsAD[DIR_0P0 *size_Mat];
- distAD.f[DIR_0M0 ] = &distributionsAD[DIR_0M0 *size_Mat];
- distAD.f[DIR_00P ] = &distributionsAD[DIR_00P *size_Mat];
- distAD.f[DIR_00M ] = &distributionsAD[DIR_00M *size_Mat];
- distAD.f[DIR_PP0 ] = &distributionsAD[DIR_PP0 *size_Mat];
- distAD.f[DIR_MM0 ] = &distributionsAD[DIR_MM0 *size_Mat];
- distAD.f[DIR_PM0 ] = &distributionsAD[DIR_PM0 *size_Mat];
- distAD.f[DIR_MP0 ] = &distributionsAD[DIR_MP0 *size_Mat];
- distAD.f[DIR_P0P ] = &distributionsAD[DIR_P0P *size_Mat];
- distAD.f[DIR_M0M ] = &distributionsAD[DIR_M0M *size_Mat];
- distAD.f[DIR_P0M ] = &distributionsAD[DIR_P0M *size_Mat];
- distAD.f[DIR_M0P ] = &distributionsAD[DIR_M0P *size_Mat];
- distAD.f[DIR_0PP ] = &distributionsAD[DIR_0PP *size_Mat];
- distAD.f[DIR_0MM ] = &distributionsAD[DIR_0MM *size_Mat];
- distAD.f[DIR_0PM ] = &distributionsAD[DIR_0PM *size_Mat];
- distAD.f[DIR_0MP ] = &distributionsAD[DIR_0MP *size_Mat];
- distAD.f[DIR_000] = &distributionsAD[DIR_000*size_Mat];
- distAD.f[DIR_PPP ] = &distributionsAD[DIR_PPP *size_Mat];
- distAD.f[DIR_MMP ] = &distributionsAD[DIR_MMP *size_Mat];
- distAD.f[DIR_PMP ] = &distributionsAD[DIR_PMP *size_Mat];
- distAD.f[DIR_MPP ] = &distributionsAD[DIR_MPP *size_Mat];
- distAD.f[DIR_PPM ] = &distributionsAD[DIR_PPM *size_Mat];
- distAD.f[DIR_MMM ] = &distributionsAD[DIR_MMM *size_Mat];
- distAD.f[DIR_PMM ] = &distributionsAD[DIR_PMM *size_Mat];
- distAD.f[DIR_MPM ] = &distributionsAD[DIR_MPM *size_Mat];
+ distAD.f[DIR_P00] = &distributionsAD[DIR_P00 * numberOfLBnodes];
+ distAD.f[DIR_M00] = &distributionsAD[DIR_M00 * numberOfLBnodes];
+ distAD.f[DIR_0P0] = &distributionsAD[DIR_0P0 * numberOfLBnodes];
+ distAD.f[DIR_0M0] = &distributionsAD[DIR_0M0 * numberOfLBnodes];
+ distAD.f[DIR_00P] = &distributionsAD[DIR_00P * numberOfLBnodes];
+ distAD.f[DIR_00M] = &distributionsAD[DIR_00M * numberOfLBnodes];
+ distAD.f[DIR_PP0] = &distributionsAD[DIR_PP0 * numberOfLBnodes];
+ distAD.f[DIR_MM0] = &distributionsAD[DIR_MM0 * numberOfLBnodes];
+ distAD.f[DIR_PM0] = &distributionsAD[DIR_PM0 * numberOfLBnodes];
+ distAD.f[DIR_MP0] = &distributionsAD[DIR_MP0 * numberOfLBnodes];
+ distAD.f[DIR_P0P] = &distributionsAD[DIR_P0P * numberOfLBnodes];
+ distAD.f[DIR_M0M] = &distributionsAD[DIR_M0M * numberOfLBnodes];
+ distAD.f[DIR_P0M] = &distributionsAD[DIR_P0M * numberOfLBnodes];
+ distAD.f[DIR_M0P] = &distributionsAD[DIR_M0P * numberOfLBnodes];
+ distAD.f[DIR_0PP] = &distributionsAD[DIR_0PP * numberOfLBnodes];
+ distAD.f[DIR_0MM] = &distributionsAD[DIR_0MM * numberOfLBnodes];
+ distAD.f[DIR_0PM] = &distributionsAD[DIR_0PM * numberOfLBnodes];
+ distAD.f[DIR_0MP] = &distributionsAD[DIR_0MP * numberOfLBnodes];
+ distAD.f[DIR_000] = &distributionsAD[DIR_000 * numberOfLBnodes];
+ distAD.f[DIR_PPP] = &distributionsAD[DIR_PPP * numberOfLBnodes];
+ distAD.f[DIR_MMP] = &distributionsAD[DIR_MMP * numberOfLBnodes];
+ distAD.f[DIR_PMP] = &distributionsAD[DIR_PMP * numberOfLBnodes];
+ distAD.f[DIR_MPP] = &distributionsAD[DIR_MPP * numberOfLBnodes];
+ distAD.f[DIR_PPM] = &distributionsAD[DIR_PPM * numberOfLBnodes];
+ distAD.f[DIR_MMM] = &distributionsAD[DIR_MMM * numberOfLBnodes];
+ distAD.f[DIR_PMM] = &distributionsAD[DIR_PMM * numberOfLBnodes];
+ distAD.f[DIR_MPM] = &distributionsAD[DIR_MPM * numberOfLBnodes];
}
else
{
- distAD.f[DIR_M00 ] = &distributionsAD[DIR_P00 *size_Mat];
- distAD.f[DIR_P00 ] = &distributionsAD[DIR_M00 *size_Mat];
- distAD.f[DIR_0M0 ] = &distributionsAD[DIR_0P0 *size_Mat];
- distAD.f[DIR_0P0 ] = &distributionsAD[DIR_0M0 *size_Mat];
- distAD.f[DIR_00M ] = &distributionsAD[DIR_00P *size_Mat];
- distAD.f[DIR_00P ] = &distributionsAD[DIR_00M *size_Mat];
- distAD.f[DIR_MM0 ] = &distributionsAD[DIR_PP0 *size_Mat];
- distAD.f[DIR_PP0 ] = &distributionsAD[DIR_MM0 *size_Mat];
- distAD.f[DIR_MP0 ] = &distributionsAD[DIR_PM0 *size_Mat];
- distAD.f[DIR_PM0 ] = &distributionsAD[DIR_MP0 *size_Mat];
- distAD.f[DIR_M0M ] = &distributionsAD[DIR_P0P *size_Mat];
- distAD.f[DIR_P0P ] = &distributionsAD[DIR_M0M *size_Mat];
- distAD.f[DIR_M0P ] = &distributionsAD[DIR_P0M *size_Mat];
- distAD.f[DIR_P0M ] = &distributionsAD[DIR_M0P *size_Mat];
- distAD.f[DIR_0MM ] = &distributionsAD[DIR_0PP *size_Mat];
- distAD.f[DIR_0PP ] = &distributionsAD[DIR_0MM *size_Mat];
- distAD.f[DIR_0MP ] = &distributionsAD[DIR_0PM *size_Mat];
- distAD.f[DIR_0PM ] = &distributionsAD[DIR_0MP *size_Mat];
- distAD.f[DIR_000] = &distributionsAD[DIR_000*size_Mat];
- distAD.f[DIR_MMM ] = &distributionsAD[DIR_PPP *size_Mat];
- distAD.f[DIR_PPM ] = &distributionsAD[DIR_MMP *size_Mat];
- distAD.f[DIR_MPM ] = &distributionsAD[DIR_PMP *size_Mat];
- distAD.f[DIR_PMM ] = &distributionsAD[DIR_MPP *size_Mat];
- distAD.f[DIR_MMP ] = &distributionsAD[DIR_PPM *size_Mat];
- distAD.f[DIR_PPP ] = &distributionsAD[DIR_MMM *size_Mat];
- distAD.f[DIR_MPP ] = &distributionsAD[DIR_PMM *size_Mat];
- distAD.f[DIR_PMP ] = &distributionsAD[DIR_MPM *size_Mat];
+ distAD.f[DIR_M00] = &distributionsAD[DIR_P00 * numberOfLBnodes];
+ distAD.f[DIR_P00] = &distributionsAD[DIR_M00 * numberOfLBnodes];
+ distAD.f[DIR_0M0] = &distributionsAD[DIR_0P0 * numberOfLBnodes];
+ distAD.f[DIR_0P0] = &distributionsAD[DIR_0M0 * numberOfLBnodes];
+ distAD.f[DIR_00M] = &distributionsAD[DIR_00P * numberOfLBnodes];
+ distAD.f[DIR_00P] = &distributionsAD[DIR_00M * numberOfLBnodes];
+ distAD.f[DIR_MM0] = &distributionsAD[DIR_PP0 * numberOfLBnodes];
+ distAD.f[DIR_PP0] = &distributionsAD[DIR_MM0 * numberOfLBnodes];
+ distAD.f[DIR_MP0] = &distributionsAD[DIR_PM0 * numberOfLBnodes];
+ distAD.f[DIR_PM0] = &distributionsAD[DIR_MP0 * numberOfLBnodes];
+ distAD.f[DIR_M0M] = &distributionsAD[DIR_P0P * numberOfLBnodes];
+ distAD.f[DIR_P0P] = &distributionsAD[DIR_M0M * numberOfLBnodes];
+ distAD.f[DIR_M0P] = &distributionsAD[DIR_P0M * numberOfLBnodes];
+ distAD.f[DIR_P0M] = &distributionsAD[DIR_M0P * numberOfLBnodes];
+ distAD.f[DIR_0MM] = &distributionsAD[DIR_0PP * numberOfLBnodes];
+ distAD.f[DIR_0PP] = &distributionsAD[DIR_0MM * numberOfLBnodes];
+ distAD.f[DIR_0MP] = &distributionsAD[DIR_0PM * numberOfLBnodes];
+ distAD.f[DIR_0PM] = &distributionsAD[DIR_0MP * numberOfLBnodes];
+ distAD.f[DIR_000] = &distributionsAD[DIR_000 * numberOfLBnodes];
+ distAD.f[DIR_MMM] = &distributionsAD[DIR_PPP * numberOfLBnodes];
+ distAD.f[DIR_PPM] = &distributionsAD[DIR_MMP * numberOfLBnodes];
+ distAD.f[DIR_MPM] = &distributionsAD[DIR_PMP * numberOfLBnodes];
+ distAD.f[DIR_PMM] = &distributionsAD[DIR_MPP * numberOfLBnodes];
+ distAD.f[DIR_MMP] = &distributionsAD[DIR_PPM * numberOfLBnodes];
+ distAD.f[DIR_PPP] = &distributionsAD[DIR_MMM * numberOfLBnodes];
+ distAD.f[DIR_MPP] = &distributionsAD[DIR_PMM * numberOfLBnodes];
+ distAD.f[DIR_PMP] = &distributionsAD[DIR_MPM * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////
//! - Set neighbor indices (necessary for indirect addressing)
@@ -241,63 +241,63 @@ __global__ void Factorized_Central_Moments_Advection_Diffusion_Device_Kernel(
////////////////////////////////////////////////////////////////////////////////////
//! - Set local distributions Fluid
//!
- real fcbb = (dist.f[DIR_P00 ])[k];
- real fabb = (dist.f[DIR_M00 ])[kw];
- real fbcb = (dist.f[DIR_0P0 ])[k];
- real fbab = (dist.f[DIR_0M0 ])[ks];
- real fbbc = (dist.f[DIR_00P ])[k];
- real fbba = (dist.f[DIR_00M ])[kb];
- real fccb = (dist.f[DIR_PP0 ])[k];
- real faab = (dist.f[DIR_MM0 ])[ksw];
- real fcab = (dist.f[DIR_PM0 ])[ks];
- real facb = (dist.f[DIR_MP0 ])[kw];
- real fcbc = (dist.f[DIR_P0P ])[k];
- real faba = (dist.f[DIR_M0M ])[kbw];
- real fcba = (dist.f[DIR_P0M ])[kb];
- real fabc = (dist.f[DIR_M0P ])[kw];
- real fbcc = (dist.f[DIR_0PP ])[k];
- real fbaa = (dist.f[DIR_0MM ])[kbs];
- real fbca = (dist.f[DIR_0PM ])[kb];
- real fbac = (dist.f[DIR_0MP ])[ks];
+ real fcbb = (dist.f[DIR_P00])[k];
+ real fabb = (dist.f[DIR_M00])[kw];
+ real fbcb = (dist.f[DIR_0P0])[k];
+ real fbab = (dist.f[DIR_0M0])[ks];
+ real fbbc = (dist.f[DIR_00P])[k];
+ real fbba = (dist.f[DIR_00M])[kb];
+ real fccb = (dist.f[DIR_PP0])[k];
+ real faab = (dist.f[DIR_MM0])[ksw];
+ real fcab = (dist.f[DIR_PM0])[ks];
+ real facb = (dist.f[DIR_MP0])[kw];
+ real fcbc = (dist.f[DIR_P0P])[k];
+ real faba = (dist.f[DIR_M0M])[kbw];
+ real fcba = (dist.f[DIR_P0M])[kb];
+ real fabc = (dist.f[DIR_M0P])[kw];
+ real fbcc = (dist.f[DIR_0PP])[k];
+ real fbaa = (dist.f[DIR_0MM])[kbs];
+ real fbca = (dist.f[DIR_0PM])[kb];
+ real fbac = (dist.f[DIR_0MP])[ks];
real fbbb = (dist.f[DIR_000])[k];
- real fccc = (dist.f[DIR_PPP ])[k];
- real faac = (dist.f[DIR_MMP ])[ksw];
- real fcac = (dist.f[DIR_PMP ])[ks];
- real facc = (dist.f[DIR_MPP ])[kw];
- real fcca = (dist.f[DIR_PPM ])[kb];
- real faaa = (dist.f[DIR_MMM ])[kbsw];
- real fcaa = (dist.f[DIR_PMM ])[kbs];
- real faca = (dist.f[DIR_MPM ])[kbw];
+ real fccc = (dist.f[DIR_PPP])[k];
+ real faac = (dist.f[DIR_MMP])[ksw];
+ real fcac = (dist.f[DIR_PMP])[ks];
+ real facc = (dist.f[DIR_MPP])[kw];
+ real fcca = (dist.f[DIR_PPM])[kb];
+ real faaa = (dist.f[DIR_MMM])[kbsw];
+ real fcaa = (dist.f[DIR_PMM])[kbs];
+ real faca = (dist.f[DIR_MPM])[kbw];
////////////////////////////////////////////////////////////////////////////////////
//! - Set local distributions Advection Diffusion
//!
- real mfcbb = (distAD.f[DIR_P00 ])[k];
- real mfabb = (distAD.f[DIR_M00 ])[kw];
- real mfbcb = (distAD.f[DIR_0P0 ])[k];
- real mfbab = (distAD.f[DIR_0M0 ])[ks];
- real mfbbc = (distAD.f[DIR_00P ])[k];
- real mfbba = (distAD.f[DIR_00M ])[kb];
- real mfccb = (distAD.f[DIR_PP0 ])[k];
- real mfaab = (distAD.f[DIR_MM0 ])[ksw];
- real mfcab = (distAD.f[DIR_PM0 ])[ks];
- real mfacb = (distAD.f[DIR_MP0 ])[kw];
- real mfcbc = (distAD.f[DIR_P0P ])[k];
- real mfaba = (distAD.f[DIR_M0M ])[kbw];
- real mfcba = (distAD.f[DIR_P0M ])[kb];
- real mfabc = (distAD.f[DIR_M0P ])[kw];
- real mfbcc = (distAD.f[DIR_0PP ])[k];
- real mfbaa = (distAD.f[DIR_0MM ])[kbs];
- real mfbca = (distAD.f[DIR_0PM ])[kb];
- real mfbac = (distAD.f[DIR_0MP ])[ks];
+ real mfcbb = (distAD.f[DIR_P00])[k];
+ real mfabb = (distAD.f[DIR_M00])[kw];
+ real mfbcb = (distAD.f[DIR_0P0])[k];
+ real mfbab = (distAD.f[DIR_0M0])[ks];
+ real mfbbc = (distAD.f[DIR_00P])[k];
+ real mfbba = (distAD.f[DIR_00M])[kb];
+ real mfccb = (distAD.f[DIR_PP0])[k];
+ real mfaab = (distAD.f[DIR_MM0])[ksw];
+ real mfcab = (distAD.f[DIR_PM0])[ks];
+ real mfacb = (distAD.f[DIR_MP0])[kw];
+ real mfcbc = (distAD.f[DIR_P0P])[k];
+ real mfaba = (distAD.f[DIR_M0M])[kbw];
+ real mfcba = (distAD.f[DIR_P0M])[kb];
+ real mfabc = (distAD.f[DIR_M0P])[kw];
+ real mfbcc = (distAD.f[DIR_0PP])[k];
+ real mfbaa = (distAD.f[DIR_0MM])[kbs];
+ real mfbca = (distAD.f[DIR_0PM])[kb];
+ real mfbac = (distAD.f[DIR_0MP])[ks];
real mfbbb = (distAD.f[DIR_000])[k];
- real mfccc = (distAD.f[DIR_PPP ])[k];
- real mfaac = (distAD.f[DIR_MMP ])[ksw];
- real mfcac = (distAD.f[DIR_PMP ])[ks];
- real mfacc = (distAD.f[DIR_MPP ])[kw];
- real mfcca = (distAD.f[DIR_PPM ])[kb];
- real mfaaa = (distAD.f[DIR_MMM ])[kbsw];
- real mfcaa = (distAD.f[DIR_PMM ])[kbs];
- real mfaca = (distAD.f[DIR_MPM ])[kbw];
+ real mfccc = (distAD.f[DIR_PPP])[k];
+ real mfaac = (distAD.f[DIR_MMP])[ksw];
+ real mfcac = (distAD.f[DIR_PMP])[ks];
+ real mfacc = (distAD.f[DIR_MPP])[kw];
+ real mfcca = (distAD.f[DIR_PPM])[kb];
+ real mfaaa = (distAD.f[DIR_MMM])[kbsw];
+ real mfcaa = (distAD.f[DIR_PMM])[kbs];
+ real mfaca = (distAD.f[DIR_MPM])[kbw];
////////////////////////////////////////////////////////////////////////////////////
//! - Calculate density and velocity using pyramid summation for low round-off errors as in Eq. (J1)-(J3) \ref
//! <a href="https://doi.org/10.1016/j.camwa.2015.05.001"><b>[ M. Geier et al. (2015), DOI:10.1016/j.camwa.2015.05.001 ]</b></a>
@@ -503,33 +503,33 @@ __global__ void Factorized_Central_Moments_Advection_Diffusion_Device_Kernel(
//! stored arrays dependent on timestep is based on the esoteric twist algorithm
//! <a href="https://doi.org/10.3390/computation5020019"><b>[ M. Geier et al. (2017), DOI:10.3390/computation5020019 ]</b></a>
//!
- (distAD.f[DIR_P00 ])[k ] = mfabb;
- (distAD.f[DIR_M00 ])[kw ] = mfcbb;
- (distAD.f[DIR_0P0 ])[k ] = mfbab;
- (distAD.f[DIR_0M0 ])[ks ] = mfbcb;
- (distAD.f[DIR_00P ])[k ] = mfbba;
- (distAD.f[DIR_00M ])[kb ] = mfbbc;
- (distAD.f[DIR_PP0 ])[k ] = mfaab;
- (distAD.f[DIR_MM0 ])[ksw ] = mfccb;
- (distAD.f[DIR_PM0 ])[ks ] = mfacb;
- (distAD.f[DIR_MP0 ])[kw ] = mfcab;
- (distAD.f[DIR_P0P ])[k ] = mfaba;
- (distAD.f[DIR_M0M ])[kbw ] = mfcbc;
- (distAD.f[DIR_P0M ])[kb ] = mfabc;
- (distAD.f[DIR_M0P ])[kw ] = mfcba;
- (distAD.f[DIR_0PP ])[k ] = mfbaa;
- (distAD.f[DIR_0MM ])[kbs ] = mfbcc;
- (distAD.f[DIR_0PM ])[kb ] = mfbac;
- (distAD.f[DIR_0MP ])[ks ] = mfbca;
+ (distAD.f[DIR_P00])[k ] = mfabb;
+ (distAD.f[DIR_M00])[kw ] = mfcbb;
+ (distAD.f[DIR_0P0])[k ] = mfbab;
+ (distAD.f[DIR_0M0])[ks ] = mfbcb;
+ (distAD.f[DIR_00P])[k ] = mfbba;
+ (distAD.f[DIR_00M])[kb ] = mfbbc;
+ (distAD.f[DIR_PP0])[k ] = mfaab;
+ (distAD.f[DIR_MM0])[ksw ] = mfccb;
+ (distAD.f[DIR_PM0])[ks ] = mfacb;
+ (distAD.f[DIR_MP0])[kw ] = mfcab;
+ (distAD.f[DIR_P0P])[k ] = mfaba;
+ (distAD.f[DIR_M0M])[kbw ] = mfcbc;
+ (distAD.f[DIR_P0M])[kb ] = mfabc;
+ (distAD.f[DIR_M0P])[kw ] = mfcba;
+ (distAD.f[DIR_0PP])[k ] = mfbaa;
+ (distAD.f[DIR_0MM])[kbs ] = mfbcc;
+ (distAD.f[DIR_0PM])[kb ] = mfbac;
+ (distAD.f[DIR_0MP])[ks ] = mfbca;
(distAD.f[DIR_000])[k ] = mfbbb;
- (distAD.f[DIR_PPP ])[k ] = mfaaa;
- (distAD.f[DIR_PMP ])[ks ] = mfaca;
- (distAD.f[DIR_PPM ])[kb ] = mfaac;
- (distAD.f[DIR_PMM ])[kbs ] = mfacc;
- (distAD.f[DIR_MPP ])[kw ] = mfcaa;
- (distAD.f[DIR_MMP ])[ksw ] = mfcca;
- (distAD.f[DIR_MPM ])[kbw ] = mfcac;
- (distAD.f[DIR_MMM ])[kbsw] = mfccc;
+ (distAD.f[DIR_PPP])[k ] = mfaaa;
+ (distAD.f[DIR_PMP])[ks ] = mfaca;
+ (distAD.f[DIR_PPM])[kb ] = mfaac;
+ (distAD.f[DIR_PMM])[kbs ] = mfacc;
+ (distAD.f[DIR_MPP])[kw ] = mfcaa;
+ (distAD.f[DIR_MMP])[ksw ] = mfcca;
+ (distAD.f[DIR_MPM])[kbw ] = mfcac;
+ (distAD.f[DIR_MMM])[kbsw] = mfccc;
}
}
////////////////////////////////////////////////////////////////////////////////
diff --git a/src/gpu/VirtualFluids_GPU/GPU/AdvectionDiffusionBCs27.cu b/src/gpu/VirtualFluids_GPU/GPU/AdvectionDiffusionBCs27.cu
index ecf98a7494a0a5e1c81c1040917e941f066605e6..116ce20389985e0efa650598108224b2e3e25221 100644
--- a/src/gpu/VirtualFluids_GPU/GPU/AdvectionDiffusionBCs27.cu
+++ b/src/gpu/VirtualFluids_GPU/GPU/AdvectionDiffusionBCs27.cu
@@ -20,91 +20,91 @@ __global__ void QADPress7( real* DD,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep)
{
Distributions27 D;
if (isEvenTimestep==true)
{
- D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
+ D.f[DIR_M00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_MPM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_PMP * numberOfLBnodes];
}
Distributions7 D7;
if (isEvenTimestep==true)
{
- D7.f[0] = &DD7[0*size_Mat];
- D7.f[1] = &DD7[1*size_Mat];
- D7.f[2] = &DD7[2*size_Mat];
- D7.f[3] = &DD7[3*size_Mat];
- D7.f[4] = &DD7[4*size_Mat];
- D7.f[5] = &DD7[5*size_Mat];
- D7.f[6] = &DD7[6*size_Mat];
+ D7.f[0] = &DD7[0*numberOfLBnodes];
+ D7.f[1] = &DD7[1*numberOfLBnodes];
+ D7.f[2] = &DD7[2*numberOfLBnodes];
+ D7.f[3] = &DD7[3*numberOfLBnodes];
+ D7.f[4] = &DD7[4*numberOfLBnodes];
+ D7.f[5] = &DD7[5*numberOfLBnodes];
+ D7.f[6] = &DD7[6*numberOfLBnodes];
}
else
{
- D7.f[0] = &DD7[0*size_Mat];
- D7.f[2] = &DD7[1*size_Mat];
- D7.f[1] = &DD7[2*size_Mat];
- D7.f[4] = &DD7[3*size_Mat];
- D7.f[3] = &DD7[4*size_Mat];
- D7.f[6] = &DD7[5*size_Mat];
- D7.f[5] = &DD7[6*size_Mat];
+ D7.f[0] = &DD7[0*numberOfLBnodes];
+ D7.f[2] = &DD7[1*numberOfLBnodes];
+ D7.f[1] = &DD7[2*numberOfLBnodes];
+ D7.f[4] = &DD7[3*numberOfLBnodes];
+ D7.f[3] = &DD7[4*numberOfLBnodes];
+ D7.f[6] = &DD7[5*numberOfLBnodes];
+ D7.f[5] = &DD7[6*numberOfLBnodes];
}
@@ -128,24 +128,24 @@ __global__ void QADPress7( real* DD,
// *q_dirTNE, *q_dirTSW, *q_dirTSE, *q_dirTNW, *q_dirBNE, *q_dirBSW,
// *q_dirBSE, *q_dirBNW;
- q_dirE = &QQ[DIR_P00 * numberOfBCnodes];
- q_dirW = &QQ[DIR_M00 * numberOfBCnodes];
- q_dirN = &QQ[DIR_0P0 * numberOfBCnodes];
- q_dirS = &QQ[DIR_0M0 * numberOfBCnodes];
- q_dirT = &QQ[DIR_00P * numberOfBCnodes];
- q_dirB = &QQ[DIR_00M * numberOfBCnodes];
- //q_dirNE = &QQ[DIR_PP0 * numberOfBCnodes];
- //q_dirSW = &QQ[DIR_MM0 * numberOfBCnodes];
- //q_dirSE = &QQ[DIR_PM0 * numberOfBCnodes];
- //q_dirNW = &QQ[DIR_MP0 * numberOfBCnodes];
- //q_dirTE = &QQ[DIR_P0P * numberOfBCnodes];
- //q_dirBW = &QQ[DIR_M0M * numberOfBCnodes];
- //q_dirBE = &QQ[DIR_P0M * numberOfBCnodes];
- //q_dirTW = &QQ[DIR_M0P * numberOfBCnodes];
- //q_dirTN = &QQ[DIR_0PP * numberOfBCnodes];
- //q_dirBS = &QQ[DIR_0MM * numberOfBCnodes];
- //q_dirBN = &QQ[DIR_0PM * numberOfBCnodes];
- //q_dirTS = &QQ[DIR_0MP * numberOfBCnodes];
+ q_dirE = &QQ[DIR_P00 * numberOfBCnodes];
+ q_dirW = &QQ[DIR_M00 * numberOfBCnodes];
+ q_dirN = &QQ[DIR_0P0 * numberOfBCnodes];
+ q_dirS = &QQ[DIR_0M0 * numberOfBCnodes];
+ q_dirT = &QQ[DIR_00P * numberOfBCnodes];
+ q_dirB = &QQ[DIR_00M * numberOfBCnodes];
+ //q_dirNE = &QQ[DIR_PP0 * numberOfBCnodes];
+ //q_dirSW = &QQ[DIR_MM0 * numberOfBCnodes];
+ //q_dirSE = &QQ[DIR_PM0 * numberOfBCnodes];
+ //q_dirNW = &QQ[DIR_MP0 * numberOfBCnodes];
+ //q_dirTE = &QQ[DIR_P0P * numberOfBCnodes];
+ //q_dirBW = &QQ[DIR_M0M * numberOfBCnodes];
+ //q_dirBE = &QQ[DIR_P0M * numberOfBCnodes];
+ //q_dirTW = &QQ[DIR_M0P * numberOfBCnodes];
+ //q_dirTN = &QQ[DIR_0PP * numberOfBCnodes];
+ //q_dirBS = &QQ[DIR_0MM * numberOfBCnodes];
+ //q_dirBN = &QQ[DIR_0PM * numberOfBCnodes];
+ //q_dirTS = &QQ[DIR_0MP * numberOfBCnodes];
//q_dirTNE = &QQ[DIR_PPP * numberOfBCnodes];
//q_dirTSW = &QQ[DIR_MMP * numberOfBCnodes];
//q_dirTSE = &QQ[DIR_PMP * numberOfBCnodes];
@@ -188,32 +188,32 @@ __global__ void QADPress7( real* DD,
real f_E, f_W, f_N, f_S, f_T, f_B, f_NE, f_SW, f_SE, f_NW, f_TE, f_BW, f_BE,
f_TW, f_TN, f_BS, f_BN, f_TS, f_TNE, f_TSW, f_TSE, f_TNW, f_BNE, f_BSW, f_BSE, f_BNW;
- f_W = (D.f[DIR_P00 ])[ke ];
- f_E = (D.f[DIR_M00 ])[kw ];
- f_S = (D.f[DIR_0P0 ])[kn ];
- f_N = (D.f[DIR_0M0 ])[ks ];
- f_B = (D.f[DIR_00P ])[kt ];
- f_T = (D.f[DIR_00M ])[kb ];
- f_SW = (D.f[DIR_PP0 ])[kne ];
- f_NE = (D.f[DIR_MM0 ])[ksw ];
- f_NW = (D.f[DIR_PM0 ])[kse ];
- f_SE = (D.f[DIR_MP0 ])[knw ];
- f_BW = (D.f[DIR_P0P ])[kte ];
- f_TE = (D.f[DIR_M0M ])[kbw ];
- f_TW = (D.f[DIR_P0M ])[kbe ];
- f_BE = (D.f[DIR_M0P ])[ktw ];
- f_BS = (D.f[DIR_0PP ])[ktn ];
- f_TN = (D.f[DIR_0MM ])[kbs ];
- f_TS = (D.f[DIR_0PM ])[kbn ];
- f_BN = (D.f[DIR_0MP ])[kts ];
- f_BSW = (D.f[DIR_PPP ])[ktne ];
- f_BNE = (D.f[DIR_MMP ])[ktsw ];
- f_BNW = (D.f[DIR_PMP ])[ktse ];
- f_BSE = (D.f[DIR_MPP ])[ktnw ];
- f_TSW = (D.f[DIR_PPM ])[kbne ];
- f_TNE = (D.f[DIR_MMM ])[kbsw ];
- f_TNW = (D.f[DIR_PMM ])[kbse ];
- f_TSE = (D.f[DIR_MPM ])[kbnw ];
+ f_W = (D.f[DIR_P00])[ke ];
+ f_E = (D.f[DIR_M00])[kw ];
+ f_S = (D.f[DIR_0P0])[kn ];
+ f_N = (D.f[DIR_0M0])[ks ];
+ f_B = (D.f[DIR_00P])[kt ];
+ f_T = (D.f[DIR_00M])[kb ];
+ f_SW = (D.f[DIR_PP0])[kne ];
+ f_NE = (D.f[DIR_MM0])[ksw ];
+ f_NW = (D.f[DIR_PM0])[kse ];
+ f_SE = (D.f[DIR_MP0])[knw ];
+ f_BW = (D.f[DIR_P0P])[kte ];
+ f_TE = (D.f[DIR_M0M])[kbw ];
+ f_TW = (D.f[DIR_P0M])[kbe ];
+ f_BE = (D.f[DIR_M0P])[ktw ];
+ f_BS = (D.f[DIR_0PP])[ktn ];
+ f_TN = (D.f[DIR_0MM])[kbs ];
+ f_TS = (D.f[DIR_0PM])[kbn ];
+ f_BN = (D.f[DIR_0MP])[kts ];
+ f_BSW = (D.f[DIR_PPP])[ktne ];
+ f_BNE = (D.f[DIR_MMP])[ktsw ];
+ f_BNW = (D.f[DIR_PMP])[ktse ];
+ f_BSE = (D.f[DIR_MPP])[ktnw ];
+ f_TSW = (D.f[DIR_PPM])[kbne ];
+ f_TNE = (D.f[DIR_MMM])[kbsw ];
+ f_TNW = (D.f[DIR_PMM])[kbse ];
+ f_TSE = (D.f[DIR_MPM])[kbnw ];
////////////////////////////////////////////////////////////////////////////////
/*real drho*/;
//real vx1_Inflow = zero;
@@ -293,23 +293,23 @@ __global__ void QADPress7( real* DD,
//pointertausch
if (isEvenTimestep==false)
{
- D7.f[0] = &DD7[0*size_Mat];
- D7.f[1] = &DD7[1*size_Mat];
- D7.f[2] = &DD7[2*size_Mat];
- D7.f[3] = &DD7[3*size_Mat];
- D7.f[4] = &DD7[4*size_Mat];
- D7.f[5] = &DD7[5*size_Mat];
- D7.f[6] = &DD7[6*size_Mat];
+ D7.f[0] = &DD7[0*numberOfLBnodes];
+ D7.f[1] = &DD7[1*numberOfLBnodes];
+ D7.f[2] = &DD7[2*numberOfLBnodes];
+ D7.f[3] = &DD7[3*numberOfLBnodes];
+ D7.f[4] = &DD7[4*numberOfLBnodes];
+ D7.f[5] = &DD7[5*numberOfLBnodes];
+ D7.f[6] = &DD7[6*numberOfLBnodes];
}
else
{
- D7.f[0] = &DD7[0*size_Mat];
- D7.f[2] = &DD7[1*size_Mat];
- D7.f[1] = &DD7[2*size_Mat];
- D7.f[4] = &DD7[3*size_Mat];
- D7.f[3] = &DD7[4*size_Mat];
- D7.f[6] = &DD7[5*size_Mat];
- D7.f[5] = &DD7[6*size_Mat];
+ D7.f[0] = &DD7[0*numberOfLBnodes];
+ D7.f[2] = &DD7[1*numberOfLBnodes];
+ D7.f[1] = &DD7[2*numberOfLBnodes];
+ D7.f[4] = &DD7[3*numberOfLBnodes];
+ D7.f[3] = &DD7[4*numberOfLBnodes];
+ D7.f[6] = &DD7[5*numberOfLBnodes];
+ D7.f[5] = &DD7[6*numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
@@ -461,131 +461,131 @@ __global__ void QADPress27( real* DD,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep)
{
Distributions27 D;
if (isEvenTimestep==true)
{
- D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
+ D.f[DIR_M00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_MPM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_PMP * numberOfLBnodes];
}
Distributions27 D27;
if (isEvenTimestep==true)
{
- D27.f[DIR_P00 ] = &DD27[DIR_P00 *size_Mat];
- D27.f[DIR_M00 ] = &DD27[DIR_M00 *size_Mat];
- D27.f[DIR_0P0 ] = &DD27[DIR_0P0 *size_Mat];
- D27.f[DIR_0M0 ] = &DD27[DIR_0M0 *size_Mat];
- D27.f[DIR_00P ] = &DD27[DIR_00P *size_Mat];
- D27.f[DIR_00M ] = &DD27[DIR_00M *size_Mat];
- D27.f[DIR_PP0 ] = &DD27[DIR_PP0 *size_Mat];
- D27.f[DIR_MM0 ] = &DD27[DIR_MM0 *size_Mat];
- D27.f[DIR_PM0 ] = &DD27[DIR_PM0 *size_Mat];
- D27.f[DIR_MP0 ] = &DD27[DIR_MP0 *size_Mat];
- D27.f[DIR_P0P ] = &DD27[DIR_P0P *size_Mat];
- D27.f[DIR_M0M ] = &DD27[DIR_M0M *size_Mat];
- D27.f[DIR_P0M ] = &DD27[DIR_P0M *size_Mat];
- D27.f[DIR_M0P ] = &DD27[DIR_M0P *size_Mat];
- D27.f[DIR_0PP ] = &DD27[DIR_0PP *size_Mat];
- D27.f[DIR_0MM ] = &DD27[DIR_0MM *size_Mat];
- D27.f[DIR_0PM ] = &DD27[DIR_0PM *size_Mat];
- D27.f[DIR_0MP ] = &DD27[DIR_0MP *size_Mat];
- D27.f[DIR_000] = &DD27[DIR_000*size_Mat];
- D27.f[DIR_PPP ] = &DD27[DIR_PPP *size_Mat];
- D27.f[DIR_MMP ] = &DD27[DIR_MMP *size_Mat];
- D27.f[DIR_PMP ] = &DD27[DIR_PMP *size_Mat];
- D27.f[DIR_MPP ] = &DD27[DIR_MPP *size_Mat];
- D27.f[DIR_PPM ] = &DD27[DIR_PPM *size_Mat];
- D27.f[DIR_MMM ] = &DD27[DIR_MMM *size_Mat];
- D27.f[DIR_PMM ] = &DD27[DIR_PMM *size_Mat];
- D27.f[DIR_MPM ] = &DD27[DIR_MPM *size_Mat];
+ D27.f[DIR_P00] = &DD27[DIR_P00 * numberOfLBnodes];
+ D27.f[DIR_M00] = &DD27[DIR_M00 * numberOfLBnodes];
+ D27.f[DIR_0P0] = &DD27[DIR_0P0 * numberOfLBnodes];
+ D27.f[DIR_0M0] = &DD27[DIR_0M0 * numberOfLBnodes];
+ D27.f[DIR_00P] = &DD27[DIR_00P * numberOfLBnodes];
+ D27.f[DIR_00M] = &DD27[DIR_00M * numberOfLBnodes];
+ D27.f[DIR_PP0] = &DD27[DIR_PP0 * numberOfLBnodes];
+ D27.f[DIR_MM0] = &DD27[DIR_MM0 * numberOfLBnodes];
+ D27.f[DIR_PM0] = &DD27[DIR_PM0 * numberOfLBnodes];
+ D27.f[DIR_MP0] = &DD27[DIR_MP0 * numberOfLBnodes];
+ D27.f[DIR_P0P] = &DD27[DIR_P0P * numberOfLBnodes];
+ D27.f[DIR_M0M] = &DD27[DIR_M0M * numberOfLBnodes];
+ D27.f[DIR_P0M] = &DD27[DIR_P0M * numberOfLBnodes];
+ D27.f[DIR_M0P] = &DD27[DIR_M0P * numberOfLBnodes];
+ D27.f[DIR_0PP] = &DD27[DIR_0PP * numberOfLBnodes];
+ D27.f[DIR_0MM] = &DD27[DIR_0MM * numberOfLBnodes];
+ D27.f[DIR_0PM] = &DD27[DIR_0PM * numberOfLBnodes];
+ D27.f[DIR_0MP] = &DD27[DIR_0MP * numberOfLBnodes];
+ D27.f[DIR_000] = &DD27[DIR_000 * numberOfLBnodes];
+ D27.f[DIR_PPP] = &DD27[DIR_PPP * numberOfLBnodes];
+ D27.f[DIR_MMP] = &DD27[DIR_MMP * numberOfLBnodes];
+ D27.f[DIR_PMP] = &DD27[DIR_PMP * numberOfLBnodes];
+ D27.f[DIR_MPP] = &DD27[DIR_MPP * numberOfLBnodes];
+ D27.f[DIR_PPM] = &DD27[DIR_PPM * numberOfLBnodes];
+ D27.f[DIR_MMM] = &DD27[DIR_MMM * numberOfLBnodes];
+ D27.f[DIR_PMM] = &DD27[DIR_PMM * numberOfLBnodes];
+ D27.f[DIR_MPM] = &DD27[DIR_MPM * numberOfLBnodes];
}
else
{
- D27.f[DIR_M00 ] = &DD27[DIR_P00 *size_Mat];
- D27.f[DIR_P00 ] = &DD27[DIR_M00 *size_Mat];
- D27.f[DIR_0M0 ] = &DD27[DIR_0P0 *size_Mat];
- D27.f[DIR_0P0 ] = &DD27[DIR_0M0 *size_Mat];
- D27.f[DIR_00M ] = &DD27[DIR_00P *size_Mat];
- D27.f[DIR_00P ] = &DD27[DIR_00M *size_Mat];
- D27.f[DIR_MM0 ] = &DD27[DIR_PP0 *size_Mat];
- D27.f[DIR_PP0 ] = &DD27[DIR_MM0 *size_Mat];
- D27.f[DIR_MP0 ] = &DD27[DIR_PM0 *size_Mat];
- D27.f[DIR_PM0 ] = &DD27[DIR_MP0 *size_Mat];
- D27.f[DIR_M0M ] = &DD27[DIR_P0P *size_Mat];
- D27.f[DIR_P0P ] = &DD27[DIR_M0M *size_Mat];
- D27.f[DIR_M0P ] = &DD27[DIR_P0M *size_Mat];
- D27.f[DIR_P0M ] = &DD27[DIR_M0P *size_Mat];
- D27.f[DIR_0MM ] = &DD27[DIR_0PP *size_Mat];
- D27.f[DIR_0PP ] = &DD27[DIR_0MM *size_Mat];
- D27.f[DIR_0MP ] = &DD27[DIR_0PM *size_Mat];
- D27.f[DIR_0PM ] = &DD27[DIR_0MP *size_Mat];
- D27.f[DIR_000] = &DD27[DIR_000*size_Mat];
- D27.f[DIR_PPP ] = &DD27[DIR_MMM *size_Mat];
- D27.f[DIR_MMP ] = &DD27[DIR_PPM *size_Mat];
- D27.f[DIR_PMP ] = &DD27[DIR_MPM *size_Mat];
- D27.f[DIR_MPP ] = &DD27[DIR_PMM *size_Mat];
- D27.f[DIR_PPM ] = &DD27[DIR_MMP *size_Mat];
- D27.f[DIR_MMM ] = &DD27[DIR_PPP *size_Mat];
- D27.f[DIR_PMM ] = &DD27[DIR_MPP *size_Mat];
- D27.f[DIR_MPM ] = &DD27[DIR_PMP *size_Mat];
+ D27.f[DIR_M00] = &DD27[DIR_P00 * numberOfLBnodes];
+ D27.f[DIR_P00] = &DD27[DIR_M00 * numberOfLBnodes];
+ D27.f[DIR_0M0] = &DD27[DIR_0P0 * numberOfLBnodes];
+ D27.f[DIR_0P0] = &DD27[DIR_0M0 * numberOfLBnodes];
+ D27.f[DIR_00M] = &DD27[DIR_00P * numberOfLBnodes];
+ D27.f[DIR_00P] = &DD27[DIR_00M * numberOfLBnodes];
+ D27.f[DIR_MM0] = &DD27[DIR_PP0 * numberOfLBnodes];
+ D27.f[DIR_PP0] = &DD27[DIR_MM0 * numberOfLBnodes];
+ D27.f[DIR_MP0] = &DD27[DIR_PM0 * numberOfLBnodes];
+ D27.f[DIR_PM0] = &DD27[DIR_MP0 * numberOfLBnodes];
+ D27.f[DIR_M0M] = &DD27[DIR_P0P * numberOfLBnodes];
+ D27.f[DIR_P0P] = &DD27[DIR_M0M * numberOfLBnodes];
+ D27.f[DIR_M0P] = &DD27[DIR_P0M * numberOfLBnodes];
+ D27.f[DIR_P0M] = &DD27[DIR_M0P * numberOfLBnodes];
+ D27.f[DIR_0MM] = &DD27[DIR_0PP * numberOfLBnodes];
+ D27.f[DIR_0PP] = &DD27[DIR_0MM * numberOfLBnodes];
+ D27.f[DIR_0MP] = &DD27[DIR_0PM * numberOfLBnodes];
+ D27.f[DIR_0PM] = &DD27[DIR_0MP * numberOfLBnodes];
+ D27.f[DIR_000] = &DD27[DIR_000 * numberOfLBnodes];
+ D27.f[DIR_PPP] = &DD27[DIR_MMM * numberOfLBnodes];
+ D27.f[DIR_MMP] = &DD27[DIR_PPM * numberOfLBnodes];
+ D27.f[DIR_PMP] = &DD27[DIR_MPM * numberOfLBnodes];
+ D27.f[DIR_MPP] = &DD27[DIR_PMM * numberOfLBnodes];
+ D27.f[DIR_PPM] = &DD27[DIR_MMP * numberOfLBnodes];
+ D27.f[DIR_MMM] = &DD27[DIR_PPP * numberOfLBnodes];
+ D27.f[DIR_PMM] = &DD27[DIR_MPP * numberOfLBnodes];
+ D27.f[DIR_MPM] = &DD27[DIR_PMP * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////
const unsigned x = threadIdx.x; // Globaler x-Index
@@ -606,24 +606,24 @@ __global__ void QADPress27( real* DD,
*q_dirBE, *q_dirTW, *q_dirTN, *q_dirBS, *q_dirBN, *q_dirTS,
*q_dirTNE, *q_dirTSW, *q_dirTSE, *q_dirTNW, *q_dirBNE, *q_dirBSW,
*q_dirBSE, *q_dirBNW;
- q_dirE = &QQ[DIR_P00 * numberOfBCnodes];
- q_dirW = &QQ[DIR_M00 * numberOfBCnodes];
- q_dirN = &QQ[DIR_0P0 * numberOfBCnodes];
- q_dirS = &QQ[DIR_0M0 * numberOfBCnodes];
- q_dirT = &QQ[DIR_00P * numberOfBCnodes];
- q_dirB = &QQ[DIR_00M * numberOfBCnodes];
- q_dirNE = &QQ[DIR_PP0 * numberOfBCnodes];
- q_dirSW = &QQ[DIR_MM0 * numberOfBCnodes];
- q_dirSE = &QQ[DIR_PM0 * numberOfBCnodes];
- q_dirNW = &QQ[DIR_MP0 * numberOfBCnodes];
- q_dirTE = &QQ[DIR_P0P * numberOfBCnodes];
- q_dirBW = &QQ[DIR_M0M * numberOfBCnodes];
- q_dirBE = &QQ[DIR_P0M * numberOfBCnodes];
- q_dirTW = &QQ[DIR_M0P * numberOfBCnodes];
- q_dirTN = &QQ[DIR_0PP * numberOfBCnodes];
- q_dirBS = &QQ[DIR_0MM * numberOfBCnodes];
- q_dirBN = &QQ[DIR_0PM * numberOfBCnodes];
- q_dirTS = &QQ[DIR_0MP * numberOfBCnodes];
+ q_dirE = &QQ[DIR_P00 * numberOfBCnodes];
+ q_dirW = &QQ[DIR_M00 * numberOfBCnodes];
+ q_dirN = &QQ[DIR_0P0 * numberOfBCnodes];
+ q_dirS = &QQ[DIR_0M0 * numberOfBCnodes];
+ q_dirT = &QQ[DIR_00P * numberOfBCnodes];
+ q_dirB = &QQ[DIR_00M * numberOfBCnodes];
+ q_dirNE = &QQ[DIR_PP0 * numberOfBCnodes];
+ q_dirSW = &QQ[DIR_MM0 * numberOfBCnodes];
+ q_dirSE = &QQ[DIR_PM0 * numberOfBCnodes];
+ q_dirNW = &QQ[DIR_MP0 * numberOfBCnodes];
+ q_dirTE = &QQ[DIR_P0P * numberOfBCnodes];
+ q_dirBW = &QQ[DIR_M0M * numberOfBCnodes];
+ q_dirBE = &QQ[DIR_P0M * numberOfBCnodes];
+ q_dirTW = &QQ[DIR_M0P * numberOfBCnodes];
+ q_dirTN = &QQ[DIR_0PP * numberOfBCnodes];
+ q_dirBS = &QQ[DIR_0MM * numberOfBCnodes];
+ q_dirBN = &QQ[DIR_0PM * numberOfBCnodes];
+ q_dirTS = &QQ[DIR_0MP * numberOfBCnodes];
q_dirTNE = &QQ[DIR_PPP * numberOfBCnodes];
q_dirTSW = &QQ[DIR_MMP * numberOfBCnodes];
q_dirTSE = &QQ[DIR_PMP * numberOfBCnodes];
@@ -663,33 +663,33 @@ __global__ void QADPress27( real* DD,
unsigned int ktne = KQK;
unsigned int kbsw = neighborZ[ksw];
////////////////////////////////////////////////////////////////////////////////
- real f_W = (D.f[DIR_P00 ])[ke ];
- real f_E = (D.f[DIR_M00 ])[kw ];
- real f_S = (D.f[DIR_0P0 ])[kn ];
- real f_N = (D.f[DIR_0M0 ])[ks ];
- real f_B = (D.f[DIR_00P ])[kt ];
- real f_T = (D.f[DIR_00M ])[kb ];
- real f_SW = (D.f[DIR_PP0 ])[kne ];
- real f_NE = (D.f[DIR_MM0 ])[ksw ];
- real f_NW = (D.f[DIR_PM0 ])[kse ];
- real f_SE = (D.f[DIR_MP0 ])[knw ];
- real f_BW = (D.f[DIR_P0P ])[kte ];
- real f_TE = (D.f[DIR_M0M ])[kbw ];
- real f_TW = (D.f[DIR_P0M ])[kbe ];
- real f_BE = (D.f[DIR_M0P ])[ktw ];
- real f_BS = (D.f[DIR_0PP ])[ktn ];
- real f_TN = (D.f[DIR_0MM ])[kbs ];
- real f_TS = (D.f[DIR_0PM ])[kbn ];
- real f_BN = (D.f[DIR_0MP ])[kts ];
+ real f_W = (D.f[DIR_P00])[ke ];
+ real f_E = (D.f[DIR_M00])[kw ];
+ real f_S = (D.f[DIR_0P0])[kn ];
+ real f_N = (D.f[DIR_0M0])[ks ];
+ real f_B = (D.f[DIR_00P])[kt ];
+ real f_T = (D.f[DIR_00M])[kb ];
+ real f_SW = (D.f[DIR_PP0])[kne ];
+ real f_NE = (D.f[DIR_MM0])[ksw ];
+ real f_NW = (D.f[DIR_PM0])[kse ];
+ real f_SE = (D.f[DIR_MP0])[knw ];
+ real f_BW = (D.f[DIR_P0P])[kte ];
+ real f_TE = (D.f[DIR_M0M])[kbw ];
+ real f_TW = (D.f[DIR_P0M])[kbe ];
+ real f_BE = (D.f[DIR_M0P])[ktw ];
+ real f_BS = (D.f[DIR_0PP])[ktn ];
+ real f_TN = (D.f[DIR_0MM])[kbs ];
+ real f_TS = (D.f[DIR_0PM])[kbn ];
+ real f_BN = (D.f[DIR_0MP])[kts ];
real f_ZERO = (D.f[DIR_000])[kzero];
- real f_BSW = (D.f[DIR_PPP ])[ktne ];
- real f_BNE = (D.f[DIR_MMP ])[ktsw ];
- real f_BNW = (D.f[DIR_PMP ])[ktse ];
- real f_BSE = (D.f[DIR_MPP ])[ktnw ];
- real f_TSW = (D.f[DIR_PPM ])[kbne ];
- real f_TNE = (D.f[DIR_MMM ])[kbsw ];
- real f_TNW = (D.f[DIR_PMM ])[kbse ];
- real f_TSE = (D.f[DIR_MPM ])[kbnw ];
+ real f_BSW = (D.f[DIR_PPP])[ktne ];
+ real f_BNE = (D.f[DIR_MMP])[ktsw ];
+ real f_BNW = (D.f[DIR_PMP])[ktse ];
+ real f_BSE = (D.f[DIR_MPP])[ktnw ];
+ real f_TSW = (D.f[DIR_PPM])[kbne ];
+ real f_TNE = (D.f[DIR_MMM])[kbsw ];
+ real f_TNW = (D.f[DIR_PMM])[kbse ];
+ real f_TSE = (D.f[DIR_MPM])[kbnw ];
////////////////////////////////////////////////////////////////////////////////
real vx1, vx2, vx3, /*drho, feq,*/ q;
//drho = f_TSE + f_TNW + f_TNE + f_TSW + f_BSE + f_BNW + f_BNE + f_BSW +
@@ -715,33 +715,33 @@ __global__ void QADPress27( real* DD,
vx2 = OORho*((f_TNE-f_BSW)+(f_BNE-f_TSW)+(f_BNW-f_TSE)+(f_TNW-f_BSE) +(f_NE-f_SW)+(f_NW-f_SE)+(f_TN-f_BS)+(f_BN-f_TS)+(f_N-f_S));
vx3 = OORho*((f_TNE-f_BSW)+(f_TSW-f_BNE)+(f_TSE-f_BNW)+(f_TNW-f_BSE) +(f_TE-f_BW)+(f_TW-f_BE)+(f_TN-f_BS)+(f_TS-f_BN)+(f_T-f_B));
////////////////////////////////////////////////////////////////////////////////
- real f27_W = (D27.f[DIR_P00 ])[ke ];
- real f27_E = (D27.f[DIR_M00 ])[kw ];
- real f27_S = (D27.f[DIR_0P0 ])[kn ];
- real f27_N = (D27.f[DIR_0M0 ])[ks ];
- real f27_B = (D27.f[DIR_00P ])[kt ];
- real f27_T = (D27.f[DIR_00M ])[kb ];
- real f27_SW = (D27.f[DIR_PP0 ])[kne ];
- real f27_NE = (D27.f[DIR_MM0 ])[ksw ];
- real f27_NW = (D27.f[DIR_PM0 ])[kse ];
- real f27_SE = (D27.f[DIR_MP0 ])[knw ];
- real f27_BW = (D27.f[DIR_P0P ])[kte ];
- real f27_TE = (D27.f[DIR_M0M ])[kbw ];
- real f27_TW = (D27.f[DIR_P0M ])[kbe ];
- real f27_BE = (D27.f[DIR_M0P ])[ktw ];
- real f27_BS = (D27.f[DIR_0PP ])[ktn ];
- real f27_TN = (D27.f[DIR_0MM ])[kbs ];
- real f27_TS = (D27.f[DIR_0PM ])[kbn ];
- real f27_BN = (D27.f[DIR_0MP ])[kts ];
+ real f27_W = (D27.f[DIR_P00])[ke ];
+ real f27_E = (D27.f[DIR_M00])[kw ];
+ real f27_S = (D27.f[DIR_0P0])[kn ];
+ real f27_N = (D27.f[DIR_0M0])[ks ];
+ real f27_B = (D27.f[DIR_00P])[kt ];
+ real f27_T = (D27.f[DIR_00M])[kb ];
+ real f27_SW = (D27.f[DIR_PP0])[kne ];
+ real f27_NE = (D27.f[DIR_MM0])[ksw ];
+ real f27_NW = (D27.f[DIR_PM0])[kse ];
+ real f27_SE = (D27.f[DIR_MP0])[knw ];
+ real f27_BW = (D27.f[DIR_P0P])[kte ];
+ real f27_TE = (D27.f[DIR_M0M])[kbw ];
+ real f27_TW = (D27.f[DIR_P0M])[kbe ];
+ real f27_BE = (D27.f[DIR_M0P])[ktw ];
+ real f27_BS = (D27.f[DIR_0PP])[ktn ];
+ real f27_TN = (D27.f[DIR_0MM])[kbs ];
+ real f27_TS = (D27.f[DIR_0PM])[kbn ];
+ real f27_BN = (D27.f[DIR_0MP])[kts ];
real f27_ZERO = (D27.f[DIR_000])[kzero];
- real f27_BSW = (D27.f[DIR_PPP ])[ktne ];
- real f27_BNE = (D27.f[DIR_MMP ])[ktsw ];
- real f27_BNW = (D27.f[DIR_PMP ])[ktse ];
- real f27_BSE = (D27.f[DIR_MPP ])[ktnw ];
- real f27_TSW = (D27.f[DIR_PPM ])[kbne ];
- real f27_TNE = (D27.f[DIR_MMM ])[kbsw ];
- real f27_TNW = (D27.f[DIR_PMM ])[kbse ];
- real f27_TSE = (D27.f[DIR_MPM ])[kbnw ];
+ real f27_BSW = (D27.f[DIR_PPP])[ktne ];
+ real f27_BNE = (D27.f[DIR_MMP])[ktsw ];
+ real f27_BNW = (D27.f[DIR_PMP])[ktse ];
+ real f27_BSE = (D27.f[DIR_MPP])[ktnw ];
+ real f27_TSW = (D27.f[DIR_PPM])[kbne ];
+ real f27_TNE = (D27.f[DIR_MMM])[kbsw ];
+ real f27_TNW = (D27.f[DIR_PMM])[kbse ];
+ real f27_TSE = (D27.f[DIR_MPM])[kbnw ];
////////////////////////////////////////////////////////////////////////////////
real cu_sq=c3o2*(vx1*vx1+vx2*vx2+vx3*vx3);
////////////////////////////////////////////////////////////////////////////////
@@ -849,86 +849,86 @@ __global__ void QADPress27( real* DD,
//////////////////////////////////////////////////////////////////////////
if (isEvenTimestep==false)
{
- D27.f[DIR_P00 ] = &DD27[DIR_P00 *size_Mat];
- D27.f[DIR_M00 ] = &DD27[DIR_M00 *size_Mat];
- D27.f[DIR_0P0 ] = &DD27[DIR_0P0 *size_Mat];
- D27.f[DIR_0M0 ] = &DD27[DIR_0M0 *size_Mat];
- D27.f[DIR_00P ] = &DD27[DIR_00P *size_Mat];
- D27.f[DIR_00M ] = &DD27[DIR_00M *size_Mat];
- D27.f[DIR_PP0 ] = &DD27[DIR_PP0 *size_Mat];
- D27.f[DIR_MM0 ] = &DD27[DIR_MM0 *size_Mat];
- D27.f[DIR_PM0 ] = &DD27[DIR_PM0 *size_Mat];
- D27.f[DIR_MP0 ] = &DD27[DIR_MP0 *size_Mat];
- D27.f[DIR_P0P ] = &DD27[DIR_P0P *size_Mat];
- D27.f[DIR_M0M ] = &DD27[DIR_M0M *size_Mat];
- D27.f[DIR_P0M ] = &DD27[DIR_P0M *size_Mat];
- D27.f[DIR_M0P ] = &DD27[DIR_M0P *size_Mat];
- D27.f[DIR_0PP ] = &DD27[DIR_0PP *size_Mat];
- D27.f[DIR_0MM ] = &DD27[DIR_0MM *size_Mat];
- D27.f[DIR_0PM ] = &DD27[DIR_0PM *size_Mat];
- D27.f[DIR_0MP ] = &DD27[DIR_0MP *size_Mat];
- D27.f[DIR_000] = &DD27[DIR_000*size_Mat];
- D27.f[DIR_PPP ] = &DD27[DIR_PPP *size_Mat];
- D27.f[DIR_MMP ] = &DD27[DIR_MMP *size_Mat];
- D27.f[DIR_PMP ] = &DD27[DIR_PMP *size_Mat];
- D27.f[DIR_MPP ] = &DD27[DIR_MPP *size_Mat];
- D27.f[DIR_PPM ] = &DD27[DIR_PPM *size_Mat];
- D27.f[DIR_MMM ] = &DD27[DIR_MMM *size_Mat];
- D27.f[DIR_PMM ] = &DD27[DIR_PMM *size_Mat];
- D27.f[DIR_MPM ] = &DD27[DIR_MPM *size_Mat];
+ D27.f[DIR_P00] = &DD27[DIR_P00 * numberOfLBnodes];
+ D27.f[DIR_M00] = &DD27[DIR_M00 * numberOfLBnodes];
+ D27.f[DIR_0P0] = &DD27[DIR_0P0 * numberOfLBnodes];
+ D27.f[DIR_0M0] = &DD27[DIR_0M0 * numberOfLBnodes];
+ D27.f[DIR_00P] = &DD27[DIR_00P * numberOfLBnodes];
+ D27.f[DIR_00M] = &DD27[DIR_00M * numberOfLBnodes];
+ D27.f[DIR_PP0] = &DD27[DIR_PP0 * numberOfLBnodes];
+ D27.f[DIR_MM0] = &DD27[DIR_MM0 * numberOfLBnodes];
+ D27.f[DIR_PM0] = &DD27[DIR_PM0 * numberOfLBnodes];
+ D27.f[DIR_MP0] = &DD27[DIR_MP0 * numberOfLBnodes];
+ D27.f[DIR_P0P] = &DD27[DIR_P0P * numberOfLBnodes];
+ D27.f[DIR_M0M] = &DD27[DIR_M0M * numberOfLBnodes];
+ D27.f[DIR_P0M] = &DD27[DIR_P0M * numberOfLBnodes];
+ D27.f[DIR_M0P] = &DD27[DIR_M0P * numberOfLBnodes];
+ D27.f[DIR_0PP] = &DD27[DIR_0PP * numberOfLBnodes];
+ D27.f[DIR_0MM] = &DD27[DIR_0MM * numberOfLBnodes];
+ D27.f[DIR_0PM] = &DD27[DIR_0PM * numberOfLBnodes];
+ D27.f[DIR_0MP] = &DD27[DIR_0MP * numberOfLBnodes];
+ D27.f[DIR_000] = &DD27[DIR_000 * numberOfLBnodes];
+ D27.f[DIR_PPP] = &DD27[DIR_PPP * numberOfLBnodes];
+ D27.f[DIR_MMP] = &DD27[DIR_MMP * numberOfLBnodes];
+ D27.f[DIR_PMP] = &DD27[DIR_PMP * numberOfLBnodes];
+ D27.f[DIR_MPP] = &DD27[DIR_MPP * numberOfLBnodes];
+ D27.f[DIR_PPM] = &DD27[DIR_PPM * numberOfLBnodes];
+ D27.f[DIR_MMM] = &DD27[DIR_MMM * numberOfLBnodes];
+ D27.f[DIR_PMM] = &DD27[DIR_PMM * numberOfLBnodes];
+ D27.f[DIR_MPM] = &DD27[DIR_MPM * numberOfLBnodes];
}
else
{
- D27.f[DIR_M00 ] = &DD27[DIR_P00 *size_Mat];
- D27.f[DIR_P00 ] = &DD27[DIR_M00 *size_Mat];
- D27.f[DIR_0M0 ] = &DD27[DIR_0P0 *size_Mat];
- D27.f[DIR_0P0 ] = &DD27[DIR_0M0 *size_Mat];
- D27.f[DIR_00M ] = &DD27[DIR_00P *size_Mat];
- D27.f[DIR_00P ] = &DD27[DIR_00M *size_Mat];
- D27.f[DIR_MM0 ] = &DD27[DIR_PP0 *size_Mat];
- D27.f[DIR_PP0 ] = &DD27[DIR_MM0 *size_Mat];
- D27.f[DIR_MP0 ] = &DD27[DIR_PM0 *size_Mat];
- D27.f[DIR_PM0 ] = &DD27[DIR_MP0 *size_Mat];
- D27.f[DIR_M0M ] = &DD27[DIR_P0P *size_Mat];
- D27.f[DIR_P0P ] = &DD27[DIR_M0M *size_Mat];
- D27.f[DIR_M0P ] = &DD27[DIR_P0M *size_Mat];
- D27.f[DIR_P0M ] = &DD27[DIR_M0P *size_Mat];
- D27.f[DIR_0MM ] = &DD27[DIR_0PP *size_Mat];
- D27.f[DIR_0PP ] = &DD27[DIR_0MM *size_Mat];
- D27.f[DIR_0MP ] = &DD27[DIR_0PM *size_Mat];
- D27.f[DIR_0PM ] = &DD27[DIR_0MP *size_Mat];
- D27.f[DIR_000] = &DD27[DIR_000*size_Mat];
- D27.f[DIR_PPP ] = &DD27[DIR_MMM *size_Mat];
- D27.f[DIR_MMP ] = &DD27[DIR_PPM *size_Mat];
- D27.f[DIR_PMP ] = &DD27[DIR_MPM *size_Mat];
- D27.f[DIR_MPP ] = &DD27[DIR_PMM *size_Mat];
- D27.f[DIR_PPM ] = &DD27[DIR_MMP *size_Mat];
- D27.f[DIR_MMM ] = &DD27[DIR_PPP *size_Mat];
- D27.f[DIR_PMM ] = &DD27[DIR_MPP *size_Mat];
- D27.f[DIR_MPM ] = &DD27[DIR_PMP *size_Mat];
+ D27.f[DIR_M00] = &DD27[DIR_P00 * numberOfLBnodes];
+ D27.f[DIR_P00] = &DD27[DIR_M00 * numberOfLBnodes];
+ D27.f[DIR_0M0] = &DD27[DIR_0P0 * numberOfLBnodes];
+ D27.f[DIR_0P0] = &DD27[DIR_0M0 * numberOfLBnodes];
+ D27.f[DIR_00M] = &DD27[DIR_00P * numberOfLBnodes];
+ D27.f[DIR_00P] = &DD27[DIR_00M * numberOfLBnodes];
+ D27.f[DIR_MM0] = &DD27[DIR_PP0 * numberOfLBnodes];
+ D27.f[DIR_PP0] = &DD27[DIR_MM0 * numberOfLBnodes];
+ D27.f[DIR_MP0] = &DD27[DIR_PM0 * numberOfLBnodes];
+ D27.f[DIR_PM0] = &DD27[DIR_MP0 * numberOfLBnodes];
+ D27.f[DIR_M0M] = &DD27[DIR_P0P * numberOfLBnodes];
+ D27.f[DIR_P0P] = &DD27[DIR_M0M * numberOfLBnodes];
+ D27.f[DIR_M0P] = &DD27[DIR_P0M * numberOfLBnodes];
+ D27.f[DIR_P0M] = &DD27[DIR_M0P * numberOfLBnodes];
+ D27.f[DIR_0MM] = &DD27[DIR_0PP * numberOfLBnodes];
+ D27.f[DIR_0PP] = &DD27[DIR_0MM * numberOfLBnodes];
+ D27.f[DIR_0MP] = &DD27[DIR_0PM * numberOfLBnodes];
+ D27.f[DIR_0PM] = &DD27[DIR_0MP * numberOfLBnodes];
+ D27.f[DIR_000] = &DD27[DIR_000 * numberOfLBnodes];
+ D27.f[DIR_PPP] = &DD27[DIR_MMM * numberOfLBnodes];
+ D27.f[DIR_MMP] = &DD27[DIR_PPM * numberOfLBnodes];
+ D27.f[DIR_PMP] = &DD27[DIR_MPM * numberOfLBnodes];
+ D27.f[DIR_MPP] = &DD27[DIR_PMM * numberOfLBnodes];
+ D27.f[DIR_PPM] = &DD27[DIR_MMP * numberOfLBnodes];
+ D27.f[DIR_MMM] = &DD27[DIR_PPP * numberOfLBnodes];
+ D27.f[DIR_PMM] = &DD27[DIR_MPP * numberOfLBnodes];
+ D27.f[DIR_MPM] = &DD27[DIR_PMP * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
//Test
//(D.f[DIR_000])[k]=c1o10;
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
- q = q_dirE[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_M00 ])[kw ]=(c2o1*feqW27_W -(f27_E *(q*omegaD-c1o1)-omegaD*feq27_E *(q-c1o1))/(omegaD-c1o1)+f27_W *q)/(q+c1o1);
- q = q_dirW[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_P00 ])[ke ]=(c2o1*feqW27_E -(f27_W *(q*omegaD-c1o1)-omegaD*feq27_W *(q-c1o1))/(omegaD-c1o1)+f27_E *q)/(q+c1o1);
- q = q_dirN[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_0M0 ])[ks ]=(c2o1*feqW27_S -(f27_N *(q*omegaD-c1o1)-omegaD*feq27_N *(q-c1o1))/(omegaD-c1o1)+f27_S *q)/(q+c1o1);
- q = q_dirS[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_0P0 ])[kn ]=(c2o1*feqW27_N -(f27_S *(q*omegaD-c1o1)-omegaD*feq27_S *(q-c1o1))/(omegaD-c1o1)+f27_N *q)/(q+c1o1);
- q = q_dirT[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_00M ])[kb ]=(c2o1*feqW27_B -(f27_T *(q*omegaD-c1o1)-omegaD*feq27_T *(q-c1o1))/(omegaD-c1o1)+f27_B *q)/(q+c1o1);
- q = q_dirB[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_00P ])[kt ]=(c2o1*feqW27_T -(f27_B *(q*omegaD-c1o1)-omegaD*feq27_B *(q-c1o1))/(omegaD-c1o1)+f27_T *q)/(q+c1o1);
- q = q_dirNE[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_MM0 ])[ksw ]=(c2o1*feqW27_SW -(f27_NE *(q*omegaD-c1o1)-omegaD*feq27_NE *(q-c1o1))/(omegaD-c1o1)+f27_SW *q)/(q+c1o1);
- q = q_dirSW[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_PP0 ])[kne ]=(c2o1*feqW27_NE -(f27_SW *(q*omegaD-c1o1)-omegaD*feq27_SW *(q-c1o1))/(omegaD-c1o1)+f27_NE *q)/(q+c1o1);
- q = q_dirSE[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_MP0 ])[knw ]=(c2o1*feqW27_NW -(f27_SE *(q*omegaD-c1o1)-omegaD*feq27_SE *(q-c1o1))/(omegaD-c1o1)+f27_NW *q)/(q+c1o1);
- q = q_dirNW[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_PM0 ])[kse ]=(c2o1*feqW27_SE -(f27_NW *(q*omegaD-c1o1)-omegaD*feq27_NW *(q-c1o1))/(omegaD-c1o1)+f27_SE *q)/(q+c1o1);
- q = q_dirTE[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_M0M ])[kbw ]=(c2o1*feqW27_BW -(f27_TE *(q*omegaD-c1o1)-omegaD*feq27_TE *(q-c1o1))/(omegaD-c1o1)+f27_BW *q)/(q+c1o1);
- q = q_dirBW[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_P0P ])[kte ]=(c2o1*feqW27_TE -(f27_BW *(q*omegaD-c1o1)-omegaD*feq27_BW *(q-c1o1))/(omegaD-c1o1)+f27_TE *q)/(q+c1o1);
- q = q_dirBE[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_M0P ])[ktw ]=(c2o1*feqW27_TW -(f27_BE *(q*omegaD-c1o1)-omegaD*feq27_BE *(q-c1o1))/(omegaD-c1o1)+f27_TW *q)/(q+c1o1);
- q = q_dirTW[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_P0M ])[kbe ]=(c2o1*feqW27_BE -(f27_TW *(q*omegaD-c1o1)-omegaD*feq27_TW *(q-c1o1))/(omegaD-c1o1)+f27_BE *q)/(q+c1o1);
- q = q_dirTN[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_0MM ])[kbs ]=(c2o1*feqW27_BS -(f27_TN *(q*omegaD-c1o1)-omegaD*feq27_TN *(q-c1o1))/(omegaD-c1o1)+f27_BS *q)/(q+c1o1);
- q = q_dirBS[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_0PP ])[ktn ]=(c2o1*feqW27_TN -(f27_BS *(q*omegaD-c1o1)-omegaD*feq27_BS *(q-c1o1))/(omegaD-c1o1)+f27_TN *q)/(q+c1o1);
- q = q_dirBN[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_0MP ])[kts ]=(c2o1*feqW27_TS -(f27_BN *(q*omegaD-c1o1)-omegaD*feq27_BN *(q-c1o1))/(omegaD-c1o1)+f27_TS *q)/(q+c1o1);
- q = q_dirTS[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_0PM ])[kbn ]=(c2o1*feqW27_BN -(f27_TS *(q*omegaD-c1o1)-omegaD*feq27_TS *(q-c1o1))/(omegaD-c1o1)+f27_BN *q)/(q+c1o1);
+ q = q_dirE[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_M00])[kw ]=(c2o1*feqW27_W -(f27_E *(q*omegaD-c1o1)-omegaD*feq27_E *(q-c1o1))/(omegaD-c1o1)+f27_W *q)/(q+c1o1);
+ q = q_dirW[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_P00])[ke ]=(c2o1*feqW27_E -(f27_W *(q*omegaD-c1o1)-omegaD*feq27_W *(q-c1o1))/(omegaD-c1o1)+f27_E *q)/(q+c1o1);
+ q = q_dirN[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_0M0])[ks ]=(c2o1*feqW27_S -(f27_N *(q*omegaD-c1o1)-omegaD*feq27_N *(q-c1o1))/(omegaD-c1o1)+f27_S *q)/(q+c1o1);
+ q = q_dirS[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_0P0])[kn ]=(c2o1*feqW27_N -(f27_S *(q*omegaD-c1o1)-omegaD*feq27_S *(q-c1o1))/(omegaD-c1o1)+f27_N *q)/(q+c1o1);
+ q = q_dirT[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_00M])[kb ]=(c2o1*feqW27_B -(f27_T *(q*omegaD-c1o1)-omegaD*feq27_T *(q-c1o1))/(omegaD-c1o1)+f27_B *q)/(q+c1o1);
+ q = q_dirB[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_00P])[kt ]=(c2o1*feqW27_T -(f27_B *(q*omegaD-c1o1)-omegaD*feq27_B *(q-c1o1))/(omegaD-c1o1)+f27_T *q)/(q+c1o1);
+ q = q_dirNE[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_MM0])[ksw ]=(c2o1*feqW27_SW -(f27_NE *(q*omegaD-c1o1)-omegaD*feq27_NE *(q-c1o1))/(omegaD-c1o1)+f27_SW *q)/(q+c1o1);
+ q = q_dirSW[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_PP0])[kne ]=(c2o1*feqW27_NE -(f27_SW *(q*omegaD-c1o1)-omegaD*feq27_SW *(q-c1o1))/(omegaD-c1o1)+f27_NE *q)/(q+c1o1);
+ q = q_dirSE[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_MP0])[knw ]=(c2o1*feqW27_NW -(f27_SE *(q*omegaD-c1o1)-omegaD*feq27_SE *(q-c1o1))/(omegaD-c1o1)+f27_NW *q)/(q+c1o1);
+ q = q_dirNW[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_PM0])[kse ]=(c2o1*feqW27_SE -(f27_NW *(q*omegaD-c1o1)-omegaD*feq27_NW *(q-c1o1))/(omegaD-c1o1)+f27_SE *q)/(q+c1o1);
+ q = q_dirTE[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_M0M])[kbw ]=(c2o1*feqW27_BW -(f27_TE *(q*omegaD-c1o1)-omegaD*feq27_TE *(q-c1o1))/(omegaD-c1o1)+f27_BW *q)/(q+c1o1);
+ q = q_dirBW[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_P0P])[kte ]=(c2o1*feqW27_TE -(f27_BW *(q*omegaD-c1o1)-omegaD*feq27_BW *(q-c1o1))/(omegaD-c1o1)+f27_TE *q)/(q+c1o1);
+ q = q_dirBE[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_M0P])[ktw ]=(c2o1*feqW27_TW -(f27_BE *(q*omegaD-c1o1)-omegaD*feq27_BE *(q-c1o1))/(omegaD-c1o1)+f27_TW *q)/(q+c1o1);
+ q = q_dirTW[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_P0M])[kbe ]=(c2o1*feqW27_BE -(f27_TW *(q*omegaD-c1o1)-omegaD*feq27_TW *(q-c1o1))/(omegaD-c1o1)+f27_BE *q)/(q+c1o1);
+ q = q_dirTN[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_0MM])[kbs ]=(c2o1*feqW27_BS -(f27_TN *(q*omegaD-c1o1)-omegaD*feq27_TN *(q-c1o1))/(omegaD-c1o1)+f27_BS *q)/(q+c1o1);
+ q = q_dirBS[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_0PP])[ktn ]=(c2o1*feqW27_TN -(f27_BS *(q*omegaD-c1o1)-omegaD*feq27_BS *(q-c1o1))/(omegaD-c1o1)+f27_TN *q)/(q+c1o1);
+ q = q_dirBN[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_0MP])[kts ]=(c2o1*feqW27_TS -(f27_BN *(q*omegaD-c1o1)-omegaD*feq27_BN *(q-c1o1))/(omegaD-c1o1)+f27_TS *q)/(q+c1o1);
+ q = q_dirTS[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_0PM])[kbn ]=(c2o1*feqW27_BN -(f27_TS *(q*omegaD-c1o1)-omegaD*feq27_TS *(q-c1o1))/(omegaD-c1o1)+f27_BN *q)/(q+c1o1);
q = q_dirTNE[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_MMM])[kbsw]=(c2o1*feqW27_BSW-(f27_TNE*(q*omegaD-c1o1)-omegaD*feq27_TNE*(q-c1o1))/(omegaD-c1o1)+f27_BSW*q)/(q+c1o1);
q = q_dirBSW[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_PPP])[ktne]=(c2o1*feqW27_TNE-(f27_BSW*(q*omegaD-c1o1)-omegaD*feq27_BSW*(q-c1o1))/(omegaD-c1o1)+f27_TNE*q)/(q+c1o1);
q = q_dirBNE[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_MMP])[ktsw]=(c2o1*feqW27_TSW-(f27_BNE*(q*omegaD-c1o1)-omegaD*feq27_BNE*(q-c1o1))/(omegaD-c1o1)+f27_TSW*q)/(q+c1o1);
@@ -989,132 +989,132 @@ __global__ void QADPressNEQNeighbor27(
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep
)
{
Distributions27 D;
if (isEvenTimestep == true)
{
- D.f[DIR_P00] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_M00] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0P0] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0M0] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00P] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00M] = &DD[DIR_00M *size_Mat];
- D.f[DIR_PP0] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_MM0] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_PM0] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_MP0] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_P0P] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_M0M] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_P0M] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_M0P] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0PP] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0MM] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0PM] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0MP] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_MMP] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_PMP] = &DD[DIR_PMP *size_Mat];
- D.f[DIR_MPP] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_PPM] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_MMM] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_PMM] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_MPM] = &DD[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_P00] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0M0] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0P0] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00M] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00P] = &DD[DIR_00M *size_Mat];
- D.f[DIR_MM0] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_PP0] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_MP0] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_PM0] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_M0M] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_P0P] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_M0P] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_P0M] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0MM] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0PP] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0MP] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0PM] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_MMP] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_PMP] = &DD[DIR_MPM *size_Mat];
- D.f[DIR_MPP] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_PPM] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_MMM] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_PMM] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_MPM] = &DD[DIR_PMP *size_Mat];
+ D.f[DIR_M00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_MPM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_PMP * numberOfLBnodes];
}
Distributions27 D27;
if (isEvenTimestep == true)
{
- D27.f[DIR_P00] = &DD27[DIR_P00 *size_Mat];
- D27.f[DIR_M00] = &DD27[DIR_M00 *size_Mat];
- D27.f[DIR_0P0] = &DD27[DIR_0P0 *size_Mat];
- D27.f[DIR_0M0] = &DD27[DIR_0M0 *size_Mat];
- D27.f[DIR_00P] = &DD27[DIR_00P *size_Mat];
- D27.f[DIR_00M] = &DD27[DIR_00M *size_Mat];
- D27.f[DIR_PP0] = &DD27[DIR_PP0 *size_Mat];
- D27.f[DIR_MM0] = &DD27[DIR_MM0 *size_Mat];
- D27.f[DIR_PM0] = &DD27[DIR_PM0 *size_Mat];
- D27.f[DIR_MP0] = &DD27[DIR_MP0 *size_Mat];
- D27.f[DIR_P0P] = &DD27[DIR_P0P *size_Mat];
- D27.f[DIR_M0M] = &DD27[DIR_M0M *size_Mat];
- D27.f[DIR_P0M] = &DD27[DIR_P0M *size_Mat];
- D27.f[DIR_M0P] = &DD27[DIR_M0P *size_Mat];
- D27.f[DIR_0PP] = &DD27[DIR_0PP *size_Mat];
- D27.f[DIR_0MM] = &DD27[DIR_0MM *size_Mat];
- D27.f[DIR_0PM] = &DD27[DIR_0PM *size_Mat];
- D27.f[DIR_0MP] = &DD27[DIR_0MP *size_Mat];
- D27.f[DIR_000] = &DD27[DIR_000*size_Mat];
- D27.f[DIR_PPP] = &DD27[DIR_PPP *size_Mat];
- D27.f[DIR_MMP] = &DD27[DIR_MMP *size_Mat];
- D27.f[DIR_PMP] = &DD27[DIR_PMP *size_Mat];
- D27.f[DIR_MPP] = &DD27[DIR_MPP *size_Mat];
- D27.f[DIR_PPM] = &DD27[DIR_PPM *size_Mat];
- D27.f[DIR_MMM] = &DD27[DIR_MMM *size_Mat];
- D27.f[DIR_PMM] = &DD27[DIR_PMM *size_Mat];
- D27.f[DIR_MPM] = &DD27[DIR_MPM *size_Mat];
+ D27.f[DIR_P00] = &DD27[DIR_P00 * numberOfLBnodes];
+ D27.f[DIR_M00] = &DD27[DIR_M00 * numberOfLBnodes];
+ D27.f[DIR_0P0] = &DD27[DIR_0P0 * numberOfLBnodes];
+ D27.f[DIR_0M0] = &DD27[DIR_0M0 * numberOfLBnodes];
+ D27.f[DIR_00P] = &DD27[DIR_00P * numberOfLBnodes];
+ D27.f[DIR_00M] = &DD27[DIR_00M * numberOfLBnodes];
+ D27.f[DIR_PP0] = &DD27[DIR_PP0 * numberOfLBnodes];
+ D27.f[DIR_MM0] = &DD27[DIR_MM0 * numberOfLBnodes];
+ D27.f[DIR_PM0] = &DD27[DIR_PM0 * numberOfLBnodes];
+ D27.f[DIR_MP0] = &DD27[DIR_MP0 * numberOfLBnodes];
+ D27.f[DIR_P0P] = &DD27[DIR_P0P * numberOfLBnodes];
+ D27.f[DIR_M0M] = &DD27[DIR_M0M * numberOfLBnodes];
+ D27.f[DIR_P0M] = &DD27[DIR_P0M * numberOfLBnodes];
+ D27.f[DIR_M0P] = &DD27[DIR_M0P * numberOfLBnodes];
+ D27.f[DIR_0PP] = &DD27[DIR_0PP * numberOfLBnodes];
+ D27.f[DIR_0MM] = &DD27[DIR_0MM * numberOfLBnodes];
+ D27.f[DIR_0PM] = &DD27[DIR_0PM * numberOfLBnodes];
+ D27.f[DIR_0MP] = &DD27[DIR_0MP * numberOfLBnodes];
+ D27.f[DIR_000] = &DD27[DIR_000 * numberOfLBnodes];
+ D27.f[DIR_PPP] = &DD27[DIR_PPP * numberOfLBnodes];
+ D27.f[DIR_MMP] = &DD27[DIR_MMP * numberOfLBnodes];
+ D27.f[DIR_PMP] = &DD27[DIR_PMP * numberOfLBnodes];
+ D27.f[DIR_MPP] = &DD27[DIR_MPP * numberOfLBnodes];
+ D27.f[DIR_PPM] = &DD27[DIR_PPM * numberOfLBnodes];
+ D27.f[DIR_MMM] = &DD27[DIR_MMM * numberOfLBnodes];
+ D27.f[DIR_PMM] = &DD27[DIR_PMM * numberOfLBnodes];
+ D27.f[DIR_MPM] = &DD27[DIR_MPM * numberOfLBnodes];
}
else
{
- D27.f[DIR_M00] = &DD27[DIR_P00 *size_Mat];
- D27.f[DIR_P00] = &DD27[DIR_M00 *size_Mat];
- D27.f[DIR_0M0] = &DD27[DIR_0P0 *size_Mat];
- D27.f[DIR_0P0] = &DD27[DIR_0M0 *size_Mat];
- D27.f[DIR_00M] = &DD27[DIR_00P *size_Mat];
- D27.f[DIR_00P] = &DD27[DIR_00M *size_Mat];
- D27.f[DIR_MM0] = &DD27[DIR_PP0 *size_Mat];
- D27.f[DIR_PP0] = &DD27[DIR_MM0 *size_Mat];
- D27.f[DIR_MP0] = &DD27[DIR_PM0 *size_Mat];
- D27.f[DIR_PM0] = &DD27[DIR_MP0 *size_Mat];
- D27.f[DIR_M0M] = &DD27[DIR_P0P *size_Mat];
- D27.f[DIR_P0P] = &DD27[DIR_M0M *size_Mat];
- D27.f[DIR_M0P] = &DD27[DIR_P0M *size_Mat];
- D27.f[DIR_P0M] = &DD27[DIR_M0P *size_Mat];
- D27.f[DIR_0MM] = &DD27[DIR_0PP *size_Mat];
- D27.f[DIR_0PP] = &DD27[DIR_0MM *size_Mat];
- D27.f[DIR_0MP] = &DD27[DIR_0PM *size_Mat];
- D27.f[DIR_0PM] = &DD27[DIR_0MP *size_Mat];
- D27.f[DIR_000] = &DD27[DIR_000*size_Mat];
- D27.f[DIR_PPP] = &DD27[DIR_MMM *size_Mat];
- D27.f[DIR_MMP] = &DD27[DIR_PPM *size_Mat];
- D27.f[DIR_PMP] = &DD27[DIR_MPM *size_Mat];
- D27.f[DIR_MPP] = &DD27[DIR_PMM *size_Mat];
- D27.f[DIR_PPM] = &DD27[DIR_MMP *size_Mat];
- D27.f[DIR_MMM] = &DD27[DIR_PPP *size_Mat];
- D27.f[DIR_PMM] = &DD27[DIR_MPP *size_Mat];
- D27.f[DIR_MPM] = &DD27[DIR_PMP *size_Mat];
+ D27.f[DIR_M00] = &DD27[DIR_P00 * numberOfLBnodes];
+ D27.f[DIR_P00] = &DD27[DIR_M00 * numberOfLBnodes];
+ D27.f[DIR_0M0] = &DD27[DIR_0P0 * numberOfLBnodes];
+ D27.f[DIR_0P0] = &DD27[DIR_0M0 * numberOfLBnodes];
+ D27.f[DIR_00M] = &DD27[DIR_00P * numberOfLBnodes];
+ D27.f[DIR_00P] = &DD27[DIR_00M * numberOfLBnodes];
+ D27.f[DIR_MM0] = &DD27[DIR_PP0 * numberOfLBnodes];
+ D27.f[DIR_PP0] = &DD27[DIR_MM0 * numberOfLBnodes];
+ D27.f[DIR_MP0] = &DD27[DIR_PM0 * numberOfLBnodes];
+ D27.f[DIR_PM0] = &DD27[DIR_MP0 * numberOfLBnodes];
+ D27.f[DIR_M0M] = &DD27[DIR_P0P * numberOfLBnodes];
+ D27.f[DIR_P0P] = &DD27[DIR_M0M * numberOfLBnodes];
+ D27.f[DIR_M0P] = &DD27[DIR_P0M * numberOfLBnodes];
+ D27.f[DIR_P0M] = &DD27[DIR_M0P * numberOfLBnodes];
+ D27.f[DIR_0MM] = &DD27[DIR_0PP * numberOfLBnodes];
+ D27.f[DIR_0PP] = &DD27[DIR_0MM * numberOfLBnodes];
+ D27.f[DIR_0MP] = &DD27[DIR_0PM * numberOfLBnodes];
+ D27.f[DIR_0PM] = &DD27[DIR_0MP * numberOfLBnodes];
+ D27.f[DIR_000] = &DD27[DIR_000 * numberOfLBnodes];
+ D27.f[DIR_PPP] = &DD27[DIR_MMM * numberOfLBnodes];
+ D27.f[DIR_MMP] = &DD27[DIR_PPM * numberOfLBnodes];
+ D27.f[DIR_PMP] = &DD27[DIR_MPM * numberOfLBnodes];
+ D27.f[DIR_MPP] = &DD27[DIR_PMM * numberOfLBnodes];
+ D27.f[DIR_PPM] = &DD27[DIR_MMP * numberOfLBnodes];
+ D27.f[DIR_MMM] = &DD27[DIR_PPP * numberOfLBnodes];
+ D27.f[DIR_PMM] = &DD27[DIR_MPP * numberOfLBnodes];
+ D27.f[DIR_MPM] = &DD27[DIR_PMP * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////
const unsigned x = threadIdx.x; // Globaler x-Index
@@ -1345,33 +1345,33 @@ __global__ void QADPressNEQNeighbor27(
unsigned int kNbsw = neighborZ[kNsw];
////////////////////////////////////////////////////////////////////////////////
//update distributions at neighbor nodes
- (D27.f[DIR_P00 ])[kNe ] = f27_W ;
- (D27.f[DIR_M00 ])[kNw ] = f27_E ;
- (D27.f[DIR_0P0 ])[kNn ] = f27_S ;
- (D27.f[DIR_0M0 ])[kNs ] = f27_N ;
- (D27.f[DIR_00P ])[kNt ] = f27_B ;
- (D27.f[DIR_00M ])[kNb ] = f27_T ;
- (D27.f[DIR_PP0 ])[kNne ] = f27_SW ;
- (D27.f[DIR_MM0 ])[kNsw ] = f27_NE ;
- (D27.f[DIR_PM0 ])[kNse ] = f27_NW ;
- (D27.f[DIR_MP0 ])[kNnw ] = f27_SE ;
- (D27.f[DIR_P0P ])[kNte ] = f27_BW ;
- (D27.f[DIR_M0M ])[kNbw ] = f27_TE ;
- (D27.f[DIR_P0M ])[kNbe ] = f27_TW ;
- (D27.f[DIR_M0P ])[kNtw ] = f27_BE ;
- (D27.f[DIR_0PP ])[kNtn ] = f27_BS ;
- (D27.f[DIR_0MM ])[kNbs ] = f27_TN ;
- (D27.f[DIR_0PM ])[kNbn ] = f27_TS ;
- (D27.f[DIR_0MP ])[kNts ] = f27_BN ;
+ (D27.f[DIR_P00])[kNe ] = f27_W ;
+ (D27.f[DIR_M00])[kNw ] = f27_E ;
+ (D27.f[DIR_0P0])[kNn ] = f27_S ;
+ (D27.f[DIR_0M0])[kNs ] = f27_N ;
+ (D27.f[DIR_00P])[kNt ] = f27_B ;
+ (D27.f[DIR_00M])[kNb ] = f27_T ;
+ (D27.f[DIR_PP0])[kNne ] = f27_SW ;
+ (D27.f[DIR_MM0])[kNsw ] = f27_NE ;
+ (D27.f[DIR_PM0])[kNse ] = f27_NW ;
+ (D27.f[DIR_MP0])[kNnw ] = f27_SE ;
+ (D27.f[DIR_P0P])[kNte ] = f27_BW ;
+ (D27.f[DIR_M0M])[kNbw ] = f27_TE ;
+ (D27.f[DIR_P0M])[kNbe ] = f27_TW ;
+ (D27.f[DIR_M0P])[kNtw ] = f27_BE ;
+ (D27.f[DIR_0PP])[kNtn ] = f27_BS ;
+ (D27.f[DIR_0MM])[kNbs ] = f27_TN ;
+ (D27.f[DIR_0PM])[kNbn ] = f27_TS ;
+ (D27.f[DIR_0MP])[kNts ] = f27_BN ;
(D27.f[DIR_000])[kNzero] = f27_ZERO;
- (D27.f[DIR_PPP ])[kNtne ] = f27_BSW ;
- (D27.f[DIR_MMP ])[kNtsw ] = f27_BNE ;
- (D27.f[DIR_PMP ])[kNtse ] = f27_BNW ;
- (D27.f[DIR_MPP ])[kNtnw ] = f27_BSE ;
- (D27.f[DIR_PPM ])[kNbne ] = f27_TSW ;
- (D27.f[DIR_MMM ])[kNbsw ] = f27_TNE ;
- (D27.f[DIR_PMM ])[kNbse ] = f27_TNW ;
- (D27.f[DIR_MPM ])[kNbnw ] = f27_TSE ;
+ (D27.f[DIR_PPP])[kNtne ] = f27_BSW ;
+ (D27.f[DIR_MMP])[kNtsw ] = f27_BNE ;
+ (D27.f[DIR_PMP])[kNtse ] = f27_BNW ;
+ (D27.f[DIR_MPP])[kNtnw ] = f27_BSE ;
+ (D27.f[DIR_PPM])[kNbne ] = f27_TSW ;
+ (D27.f[DIR_MMM])[kNbsw ] = f27_TNE ;
+ (D27.f[DIR_PMM])[kNbse ] = f27_TNW ;
+ (D27.f[DIR_MPM])[kNbnw ] = f27_TSE ;
}
}
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
@@ -1427,91 +1427,91 @@ __global__ void QADVel7( real* DD,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep)
{
Distributions27 D;
if (isEvenTimestep==true)
{
- D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
+ D.f[DIR_M00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_MPM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_PMP * numberOfLBnodes];
}
Distributions7 D7;
if (isEvenTimestep==true)
{
- D7.f[0] = &DD7[0*size_Mat];
- D7.f[1] = &DD7[1*size_Mat];
- D7.f[2] = &DD7[2*size_Mat];
- D7.f[3] = &DD7[3*size_Mat];
- D7.f[4] = &DD7[4*size_Mat];
- D7.f[5] = &DD7[5*size_Mat];
- D7.f[6] = &DD7[6*size_Mat];
+ D7.f[0] = &DD7[0*numberOfLBnodes];
+ D7.f[1] = &DD7[1*numberOfLBnodes];
+ D7.f[2] = &DD7[2*numberOfLBnodes];
+ D7.f[3] = &DD7[3*numberOfLBnodes];
+ D7.f[4] = &DD7[4*numberOfLBnodes];
+ D7.f[5] = &DD7[5*numberOfLBnodes];
+ D7.f[6] = &DD7[6*numberOfLBnodes];
}
else
{
- D7.f[0] = &DD7[0*size_Mat];
- D7.f[2] = &DD7[1*size_Mat];
- D7.f[1] = &DD7[2*size_Mat];
- D7.f[4] = &DD7[3*size_Mat];
- D7.f[3] = &DD7[4*size_Mat];
- D7.f[6] = &DD7[5*size_Mat];
- D7.f[5] = &DD7[6*size_Mat];
+ D7.f[0] = &DD7[0*numberOfLBnodes];
+ D7.f[2] = &DD7[1*numberOfLBnodes];
+ D7.f[1] = &DD7[2*numberOfLBnodes];
+ D7.f[4] = &DD7[3*numberOfLBnodes];
+ D7.f[3] = &DD7[4*numberOfLBnodes];
+ D7.f[6] = &DD7[5*numberOfLBnodes];
+ D7.f[5] = &DD7[6*numberOfLBnodes];
}
@@ -1531,12 +1531,12 @@ __global__ void QADVel7( real* DD,
//////////////////////////////////////////////////////////////////////////////////
real *q_dirE, *q_dirW, *q_dirN, *q_dirS, *q_dirT, *q_dirB;//,
- q_dirE = &QQ[DIR_P00 * numberOfBCnodes];
- q_dirW = &QQ[DIR_M00 * numberOfBCnodes];
- q_dirN = &QQ[DIR_0P0 * numberOfBCnodes];
- q_dirS = &QQ[DIR_0M0 * numberOfBCnodes];
- q_dirT = &QQ[DIR_00P * numberOfBCnodes];
- q_dirB = &QQ[DIR_00M * numberOfBCnodes];
+ q_dirE = &QQ[DIR_P00 * numberOfBCnodes];
+ q_dirW = &QQ[DIR_M00 * numberOfBCnodes];
+ q_dirN = &QQ[DIR_0P0 * numberOfBCnodes];
+ q_dirS = &QQ[DIR_0M0 * numberOfBCnodes];
+ q_dirT = &QQ[DIR_00P * numberOfBCnodes];
+ q_dirB = &QQ[DIR_00M * numberOfBCnodes];
//////////////////////////////////////////////////////////////////////////////////
//index
unsigned int KQK = k_Q[k];
@@ -1571,32 +1571,32 @@ __global__ void QADVel7( real* DD,
real f_E, f_W, f_N, f_S, f_T, f_B, f_NE, f_SW, f_SE, f_NW, f_TE, f_BW, f_BE,
f_TW, f_TN, f_BS, f_BN, f_TS, f_TNE, f_TSW, f_TSE, f_TNW, f_BNE, f_BSW, f_BSE, f_BNW;
- f_W = (D.f[DIR_P00 ])[ke ];
- f_E = (D.f[DIR_M00 ])[kw ];
- f_S = (D.f[DIR_0P0 ])[kn ];
- f_N = (D.f[DIR_0M0 ])[ks ];
- f_B = (D.f[DIR_00P ])[kt ];
- f_T = (D.f[DIR_00M ])[kb ];
- f_SW = (D.f[DIR_PP0 ])[kne ];
- f_NE = (D.f[DIR_MM0 ])[ksw ];
- f_NW = (D.f[DIR_PM0 ])[kse ];
- f_SE = (D.f[DIR_MP0 ])[knw ];
- f_BW = (D.f[DIR_P0P ])[kte ];
- f_TE = (D.f[DIR_M0M ])[kbw ];
- f_TW = (D.f[DIR_P0M ])[kbe ];
- f_BE = (D.f[DIR_M0P ])[ktw ];
- f_BS = (D.f[DIR_0PP ])[ktn ];
- f_TN = (D.f[DIR_0MM ])[kbs ];
- f_TS = (D.f[DIR_0PM ])[kbn ];
- f_BN = (D.f[DIR_0MP ])[kts ];
- f_BSW = (D.f[DIR_PPP ])[ktne ];
- f_BNE = (D.f[DIR_MMP ])[ktsw ];
- f_BNW = (D.f[DIR_PMP ])[ktse ];
- f_BSE = (D.f[DIR_MPP ])[ktnw ];
- f_TSW = (D.f[DIR_PPM ])[kbne ];
- f_TNE = (D.f[DIR_MMM ])[kbsw ];
- f_TNW = (D.f[DIR_PMM ])[kbse ];
- f_TSE = (D.f[DIR_MPM ])[kbnw ];
+ f_W = (D.f[DIR_P00])[ke ];
+ f_E = (D.f[DIR_M00])[kw ];
+ f_S = (D.f[DIR_0P0])[kn ];
+ f_N = (D.f[DIR_0M0])[ks ];
+ f_B = (D.f[DIR_00P])[kt ];
+ f_T = (D.f[DIR_00M])[kb ];
+ f_SW = (D.f[DIR_PP0])[kne ];
+ f_NE = (D.f[DIR_MM0])[ksw ];
+ f_NW = (D.f[DIR_PM0])[kse ];
+ f_SE = (D.f[DIR_MP0])[knw ];
+ f_BW = (D.f[DIR_P0P])[kte ];
+ f_TE = (D.f[DIR_M0M])[kbw ];
+ f_TW = (D.f[DIR_P0M])[kbe ];
+ f_BE = (D.f[DIR_M0P])[ktw ];
+ f_BS = (D.f[DIR_0PP])[ktn ];
+ f_TN = (D.f[DIR_0MM])[kbs ];
+ f_TS = (D.f[DIR_0PM])[kbn ];
+ f_BN = (D.f[DIR_0MP])[kts ];
+ f_BSW = (D.f[DIR_PPP])[ktne ];
+ f_BNE = (D.f[DIR_MMP])[ktsw ];
+ f_BNW = (D.f[DIR_PMP])[ktse ];
+ f_BSE = (D.f[DIR_MPP])[ktnw ];
+ f_TSW = (D.f[DIR_PPM])[kbne ];
+ f_TNE = (D.f[DIR_MMM])[kbsw ];
+ f_TNW = (D.f[DIR_PMM])[kbse ];
+ f_TSE = (D.f[DIR_MPM])[kbnw ];
////////////////////////////////////////////////////////////////////////////////
/*real drho*/;
real vx1_Inflow = c0o1;
@@ -1676,23 +1676,23 @@ __global__ void QADVel7( real* DD,
//pointertausch
if (isEvenTimestep==false)
{
- D7.f[0] = &DD7[0*size_Mat];
- D7.f[1] = &DD7[1*size_Mat];
- D7.f[2] = &DD7[2*size_Mat];
- D7.f[3] = &DD7[3*size_Mat];
- D7.f[4] = &DD7[4*size_Mat];
- D7.f[5] = &DD7[5*size_Mat];
- D7.f[6] = &DD7[6*size_Mat];
+ D7.f[0] = &DD7[0*numberOfLBnodes];
+ D7.f[1] = &DD7[1*numberOfLBnodes];
+ D7.f[2] = &DD7[2*numberOfLBnodes];
+ D7.f[3] = &DD7[3*numberOfLBnodes];
+ D7.f[4] = &DD7[4*numberOfLBnodes];
+ D7.f[5] = &DD7[5*numberOfLBnodes];
+ D7.f[6] = &DD7[6*numberOfLBnodes];
}
else
{
- D7.f[0] = &DD7[0*size_Mat];
- D7.f[2] = &DD7[1*size_Mat];
- D7.f[1] = &DD7[2*size_Mat];
- D7.f[4] = &DD7[3*size_Mat];
- D7.f[3] = &DD7[4*size_Mat];
- D7.f[6] = &DD7[5*size_Mat];
- D7.f[5] = &DD7[6*size_Mat];
+ D7.f[0] = &DD7[0*numberOfLBnodes];
+ D7.f[2] = &DD7[1*numberOfLBnodes];
+ D7.f[1] = &DD7[2*numberOfLBnodes];
+ D7.f[4] = &DD7[3*numberOfLBnodes];
+ D7.f[3] = &DD7[4*numberOfLBnodes];
+ D7.f[6] = &DD7[5*numberOfLBnodes];
+ D7.f[5] = &DD7[6*numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
@@ -1844,131 +1844,131 @@ __global__ void QADVel27(real* DD,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep)
{
Distributions27 D;
if (isEvenTimestep==true)
{
- D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
+ D.f[DIR_M00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_MPM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_PMP * numberOfLBnodes];
}
Distributions27 D27;
if (isEvenTimestep==true)
{
- D27.f[DIR_P00 ] = &DD27[DIR_P00 *size_Mat];
- D27.f[DIR_M00 ] = &DD27[DIR_M00 *size_Mat];
- D27.f[DIR_0P0 ] = &DD27[DIR_0P0 *size_Mat];
- D27.f[DIR_0M0 ] = &DD27[DIR_0M0 *size_Mat];
- D27.f[DIR_00P ] = &DD27[DIR_00P *size_Mat];
- D27.f[DIR_00M ] = &DD27[DIR_00M *size_Mat];
- D27.f[DIR_PP0 ] = &DD27[DIR_PP0 *size_Mat];
- D27.f[DIR_MM0 ] = &DD27[DIR_MM0 *size_Mat];
- D27.f[DIR_PM0 ] = &DD27[DIR_PM0 *size_Mat];
- D27.f[DIR_MP0 ] = &DD27[DIR_MP0 *size_Mat];
- D27.f[DIR_P0P ] = &DD27[DIR_P0P *size_Mat];
- D27.f[DIR_M0M ] = &DD27[DIR_M0M *size_Mat];
- D27.f[DIR_P0M ] = &DD27[DIR_P0M *size_Mat];
- D27.f[DIR_M0P ] = &DD27[DIR_M0P *size_Mat];
- D27.f[DIR_0PP ] = &DD27[DIR_0PP *size_Mat];
- D27.f[DIR_0MM ] = &DD27[DIR_0MM *size_Mat];
- D27.f[DIR_0PM ] = &DD27[DIR_0PM *size_Mat];
- D27.f[DIR_0MP ] = &DD27[DIR_0MP *size_Mat];
- D27.f[DIR_000] = &DD27[DIR_000*size_Mat];
- D27.f[DIR_PPP ] = &DD27[DIR_PPP *size_Mat];
- D27.f[DIR_MMP ] = &DD27[DIR_MMP *size_Mat];
- D27.f[DIR_PMP ] = &DD27[DIR_PMP *size_Mat];
- D27.f[DIR_MPP ] = &DD27[DIR_MPP *size_Mat];
- D27.f[DIR_PPM ] = &DD27[DIR_PPM *size_Mat];
- D27.f[DIR_MMM ] = &DD27[DIR_MMM *size_Mat];
- D27.f[DIR_PMM ] = &DD27[DIR_PMM *size_Mat];
- D27.f[DIR_MPM ] = &DD27[DIR_MPM *size_Mat];
+ D27.f[DIR_P00] = &DD27[DIR_P00 * numberOfLBnodes];
+ D27.f[DIR_M00] = &DD27[DIR_M00 * numberOfLBnodes];
+ D27.f[DIR_0P0] = &DD27[DIR_0P0 * numberOfLBnodes];
+ D27.f[DIR_0M0] = &DD27[DIR_0M0 * numberOfLBnodes];
+ D27.f[DIR_00P] = &DD27[DIR_00P * numberOfLBnodes];
+ D27.f[DIR_00M] = &DD27[DIR_00M * numberOfLBnodes];
+ D27.f[DIR_PP0] = &DD27[DIR_PP0 * numberOfLBnodes];
+ D27.f[DIR_MM0] = &DD27[DIR_MM0 * numberOfLBnodes];
+ D27.f[DIR_PM0] = &DD27[DIR_PM0 * numberOfLBnodes];
+ D27.f[DIR_MP0] = &DD27[DIR_MP0 * numberOfLBnodes];
+ D27.f[DIR_P0P] = &DD27[DIR_P0P * numberOfLBnodes];
+ D27.f[DIR_M0M] = &DD27[DIR_M0M * numberOfLBnodes];
+ D27.f[DIR_P0M] = &DD27[DIR_P0M * numberOfLBnodes];
+ D27.f[DIR_M0P] = &DD27[DIR_M0P * numberOfLBnodes];
+ D27.f[DIR_0PP] = &DD27[DIR_0PP * numberOfLBnodes];
+ D27.f[DIR_0MM] = &DD27[DIR_0MM * numberOfLBnodes];
+ D27.f[DIR_0PM] = &DD27[DIR_0PM * numberOfLBnodes];
+ D27.f[DIR_0MP] = &DD27[DIR_0MP * numberOfLBnodes];
+ D27.f[DIR_000] = &DD27[DIR_000 * numberOfLBnodes];
+ D27.f[DIR_PPP] = &DD27[DIR_PPP * numberOfLBnodes];
+ D27.f[DIR_MMP] = &DD27[DIR_MMP * numberOfLBnodes];
+ D27.f[DIR_PMP] = &DD27[DIR_PMP * numberOfLBnodes];
+ D27.f[DIR_MPP] = &DD27[DIR_MPP * numberOfLBnodes];
+ D27.f[DIR_PPM] = &DD27[DIR_PPM * numberOfLBnodes];
+ D27.f[DIR_MMM] = &DD27[DIR_MMM * numberOfLBnodes];
+ D27.f[DIR_PMM] = &DD27[DIR_PMM * numberOfLBnodes];
+ D27.f[DIR_MPM] = &DD27[DIR_MPM * numberOfLBnodes];
}
else
{
- D27.f[DIR_M00 ] = &DD27[DIR_P00 *size_Mat];
- D27.f[DIR_P00 ] = &DD27[DIR_M00 *size_Mat];
- D27.f[DIR_0M0 ] = &DD27[DIR_0P0 *size_Mat];
- D27.f[DIR_0P0 ] = &DD27[DIR_0M0 *size_Mat];
- D27.f[DIR_00M ] = &DD27[DIR_00P *size_Mat];
- D27.f[DIR_00P ] = &DD27[DIR_00M *size_Mat];
- D27.f[DIR_MM0 ] = &DD27[DIR_PP0 *size_Mat];
- D27.f[DIR_PP0 ] = &DD27[DIR_MM0 *size_Mat];
- D27.f[DIR_MP0 ] = &DD27[DIR_PM0 *size_Mat];
- D27.f[DIR_PM0 ] = &DD27[DIR_MP0 *size_Mat];
- D27.f[DIR_M0M ] = &DD27[DIR_P0P *size_Mat];
- D27.f[DIR_P0P ] = &DD27[DIR_M0M *size_Mat];
- D27.f[DIR_M0P ] = &DD27[DIR_P0M *size_Mat];
- D27.f[DIR_P0M ] = &DD27[DIR_M0P *size_Mat];
- D27.f[DIR_0MM ] = &DD27[DIR_0PP *size_Mat];
- D27.f[DIR_0PP ] = &DD27[DIR_0MM *size_Mat];
- D27.f[DIR_0MP ] = &DD27[DIR_0PM *size_Mat];
- D27.f[DIR_0PM ] = &DD27[DIR_0MP *size_Mat];
- D27.f[DIR_000] = &DD27[DIR_000*size_Mat];
- D27.f[DIR_PPP ] = &DD27[DIR_MMM *size_Mat];
- D27.f[DIR_MMP ] = &DD27[DIR_PPM *size_Mat];
- D27.f[DIR_PMP ] = &DD27[DIR_MPM *size_Mat];
- D27.f[DIR_MPP ] = &DD27[DIR_PMM *size_Mat];
- D27.f[DIR_PPM ] = &DD27[DIR_MMP *size_Mat];
- D27.f[DIR_MMM ] = &DD27[DIR_PPP *size_Mat];
- D27.f[DIR_PMM ] = &DD27[DIR_MPP *size_Mat];
- D27.f[DIR_MPM ] = &DD27[DIR_PMP *size_Mat];
+ D27.f[DIR_M00] = &DD27[DIR_P00 * numberOfLBnodes];
+ D27.f[DIR_P00] = &DD27[DIR_M00 * numberOfLBnodes];
+ D27.f[DIR_0M0] = &DD27[DIR_0P0 * numberOfLBnodes];
+ D27.f[DIR_0P0] = &DD27[DIR_0M0 * numberOfLBnodes];
+ D27.f[DIR_00M] = &DD27[DIR_00P * numberOfLBnodes];
+ D27.f[DIR_00P] = &DD27[DIR_00M * numberOfLBnodes];
+ D27.f[DIR_MM0] = &DD27[DIR_PP0 * numberOfLBnodes];
+ D27.f[DIR_PP0] = &DD27[DIR_MM0 * numberOfLBnodes];
+ D27.f[DIR_MP0] = &DD27[DIR_PM0 * numberOfLBnodes];
+ D27.f[DIR_PM0] = &DD27[DIR_MP0 * numberOfLBnodes];
+ D27.f[DIR_M0M] = &DD27[DIR_P0P * numberOfLBnodes];
+ D27.f[DIR_P0P] = &DD27[DIR_M0M * numberOfLBnodes];
+ D27.f[DIR_M0P] = &DD27[DIR_P0M * numberOfLBnodes];
+ D27.f[DIR_P0M] = &DD27[DIR_M0P * numberOfLBnodes];
+ D27.f[DIR_0MM] = &DD27[DIR_0PP * numberOfLBnodes];
+ D27.f[DIR_0PP] = &DD27[DIR_0MM * numberOfLBnodes];
+ D27.f[DIR_0MP] = &DD27[DIR_0PM * numberOfLBnodes];
+ D27.f[DIR_0PM] = &DD27[DIR_0MP * numberOfLBnodes];
+ D27.f[DIR_000] = &DD27[DIR_000 * numberOfLBnodes];
+ D27.f[DIR_PPP] = &DD27[DIR_MMM * numberOfLBnodes];
+ D27.f[DIR_MMP] = &DD27[DIR_PPM * numberOfLBnodes];
+ D27.f[DIR_PMP] = &DD27[DIR_MPM * numberOfLBnodes];
+ D27.f[DIR_MPP] = &DD27[DIR_PMM * numberOfLBnodes];
+ D27.f[DIR_PPM] = &DD27[DIR_MMP * numberOfLBnodes];
+ D27.f[DIR_MMM] = &DD27[DIR_PPP * numberOfLBnodes];
+ D27.f[DIR_PMM] = &DD27[DIR_MPP * numberOfLBnodes];
+ D27.f[DIR_MPM] = &DD27[DIR_PMP * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////
const unsigned x = threadIdx.x; // Globaler x-Index
@@ -1989,24 +1989,24 @@ __global__ void QADVel27(real* DD,
*q_dirBE, *q_dirTW, *q_dirTN, *q_dirBS, *q_dirBN, *q_dirTS,
*q_dirTNE, *q_dirTSW, *q_dirTSE, *q_dirTNW, *q_dirBNE, *q_dirBSW,
*q_dirBSE, *q_dirBNW;
- q_dirE = &QQ[DIR_P00 * numberOfBCnodes];
- q_dirW = &QQ[DIR_M00 * numberOfBCnodes];
- q_dirN = &QQ[DIR_0P0 * numberOfBCnodes];
- q_dirS = &QQ[DIR_0M0 * numberOfBCnodes];
- q_dirT = &QQ[DIR_00P * numberOfBCnodes];
- q_dirB = &QQ[DIR_00M * numberOfBCnodes];
- q_dirNE = &QQ[DIR_PP0 * numberOfBCnodes];
- q_dirSW = &QQ[DIR_MM0 * numberOfBCnodes];
- q_dirSE = &QQ[DIR_PM0 * numberOfBCnodes];
- q_dirNW = &QQ[DIR_MP0 * numberOfBCnodes];
- q_dirTE = &QQ[DIR_P0P * numberOfBCnodes];
- q_dirBW = &QQ[DIR_M0M * numberOfBCnodes];
- q_dirBE = &QQ[DIR_P0M * numberOfBCnodes];
- q_dirTW = &QQ[DIR_M0P * numberOfBCnodes];
- q_dirTN = &QQ[DIR_0PP * numberOfBCnodes];
- q_dirBS = &QQ[DIR_0MM * numberOfBCnodes];
- q_dirBN = &QQ[DIR_0PM * numberOfBCnodes];
- q_dirTS = &QQ[DIR_0MP * numberOfBCnodes];
+ q_dirE = &QQ[DIR_P00 * numberOfBCnodes];
+ q_dirW = &QQ[DIR_M00 * numberOfBCnodes];
+ q_dirN = &QQ[DIR_0P0 * numberOfBCnodes];
+ q_dirS = &QQ[DIR_0M0 * numberOfBCnodes];
+ q_dirT = &QQ[DIR_00P * numberOfBCnodes];
+ q_dirB = &QQ[DIR_00M * numberOfBCnodes];
+ q_dirNE = &QQ[DIR_PP0 * numberOfBCnodes];
+ q_dirSW = &QQ[DIR_MM0 * numberOfBCnodes];
+ q_dirSE = &QQ[DIR_PM0 * numberOfBCnodes];
+ q_dirNW = &QQ[DIR_MP0 * numberOfBCnodes];
+ q_dirTE = &QQ[DIR_P0P * numberOfBCnodes];
+ q_dirBW = &QQ[DIR_M0M * numberOfBCnodes];
+ q_dirBE = &QQ[DIR_P0M * numberOfBCnodes];
+ q_dirTW = &QQ[DIR_M0P * numberOfBCnodes];
+ q_dirTN = &QQ[DIR_0PP * numberOfBCnodes];
+ q_dirBS = &QQ[DIR_0MM * numberOfBCnodes];
+ q_dirBN = &QQ[DIR_0PM * numberOfBCnodes];
+ q_dirTS = &QQ[DIR_0MP * numberOfBCnodes];
q_dirTNE = &QQ[DIR_PPP * numberOfBCnodes];
q_dirTSW = &QQ[DIR_MMP * numberOfBCnodes];
q_dirTSE = &QQ[DIR_PMP * numberOfBCnodes];
@@ -2046,33 +2046,33 @@ __global__ void QADVel27(real* DD,
unsigned int ktne = KQK;
unsigned int kbsw = neighborZ[ksw];
////////////////////////////////////////////////////////////////////////////////
- real f_W = (D.f[DIR_P00 ])[ke ];
- real f_E = (D.f[DIR_M00 ])[kw ];
- real f_S = (D.f[DIR_0P0 ])[kn ];
- real f_N = (D.f[DIR_0M0 ])[ks ];
- real f_B = (D.f[DIR_00P ])[kt ];
- real f_T = (D.f[DIR_00M ])[kb ];
- real f_SW = (D.f[DIR_PP0 ])[kne ];
- real f_NE = (D.f[DIR_MM0 ])[ksw ];
- real f_NW = (D.f[DIR_PM0 ])[kse ];
- real f_SE = (D.f[DIR_MP0 ])[knw ];
- real f_BW = (D.f[DIR_P0P ])[kte ];
- real f_TE = (D.f[DIR_M0M ])[kbw ];
- real f_TW = (D.f[DIR_P0M ])[kbe ];
- real f_BE = (D.f[DIR_M0P ])[ktw ];
- real f_BS = (D.f[DIR_0PP ])[ktn ];
- real f_TN = (D.f[DIR_0MM ])[kbs ];
- real f_TS = (D.f[DIR_0PM ])[kbn ];
- real f_BN = (D.f[DIR_0MP ])[kts ];
+ real f_W = (D.f[DIR_P00])[ke ];
+ real f_E = (D.f[DIR_M00])[kw ];
+ real f_S = (D.f[DIR_0P0])[kn ];
+ real f_N = (D.f[DIR_0M0])[ks ];
+ real f_B = (D.f[DIR_00P])[kt ];
+ real f_T = (D.f[DIR_00M])[kb ];
+ real f_SW = (D.f[DIR_PP0])[kne ];
+ real f_NE = (D.f[DIR_MM0])[ksw ];
+ real f_NW = (D.f[DIR_PM0])[kse ];
+ real f_SE = (D.f[DIR_MP0])[knw ];
+ real f_BW = (D.f[DIR_P0P])[kte ];
+ real f_TE = (D.f[DIR_M0M])[kbw ];
+ real f_TW = (D.f[DIR_P0M])[kbe ];
+ real f_BE = (D.f[DIR_M0P])[ktw ];
+ real f_BS = (D.f[DIR_0PP])[ktn ];
+ real f_TN = (D.f[DIR_0MM])[kbs ];
+ real f_TS = (D.f[DIR_0PM])[kbn ];
+ real f_BN = (D.f[DIR_0MP])[kts ];
real f_ZERO = (D.f[DIR_000])[kzero];
- real f_BSW = (D.f[DIR_PPP ])[ktne ];
- real f_BNE = (D.f[DIR_MMP ])[ktsw ];
- real f_BNW = (D.f[DIR_PMP ])[ktse ];
- real f_BSE = (D.f[DIR_MPP ])[ktnw ];
- real f_TSW = (D.f[DIR_PPM ])[kbne ];
- real f_TNE = (D.f[DIR_MMM ])[kbsw ];
- real f_TNW = (D.f[DIR_PMM ])[kbse ];
- real f_TSE = (D.f[DIR_MPM ])[kbnw ];
+ real f_BSW = (D.f[DIR_PPP])[ktne ];
+ real f_BNE = (D.f[DIR_MMP])[ktsw ];
+ real f_BNW = (D.f[DIR_PMP])[ktse ];
+ real f_BSE = (D.f[DIR_MPP])[ktnw ];
+ real f_TSW = (D.f[DIR_PPM])[kbne ];
+ real f_TNE = (D.f[DIR_MMM])[kbsw ];
+ real f_TNW = (D.f[DIR_PMM])[kbse ];
+ real f_TSE = (D.f[DIR_MPM])[kbnw ];
////////////////////////////////////////////////////////////////////////////////
real vx1, vx2, vx3, /*drho, feq,*/ q;
////drho = f_TSE + f_TNW + f_TNE + f_TSW + f_BSE + f_BNW + f_BNE + f_BSW +
@@ -2098,33 +2098,33 @@ __global__ void QADVel27(real* DD,
vx2 = OORho*((f_TNE-f_BSW)+(f_BNE-f_TSW)+(f_BNW-f_TSE)+(f_TNW-f_BSE) +(f_NE-f_SW)+(f_NW-f_SE)+(f_TN-f_BS)+(f_BN-f_TS)+(f_N-f_S));
vx3 = OORho*((f_TNE-f_BSW)+(f_TSW-f_BNE)+(f_TSE-f_BNW)+(f_TNW-f_BSE) +(f_TE-f_BW)+(f_TW-f_BE)+(f_TN-f_BS)+(f_TS-f_BN)+(f_T-f_B));
////////////////////////////////////////////////////////////////////////////////
- //real f27_W = (D27.f[DIR_P00 ])[ke ];
- //real f27_E = (D27.f[DIR_M00 ])[kw ];
- //real f27_S = (D27.f[DIR_0P0 ])[kn ];
- //real f27_N = (D27.f[DIR_0M0 ])[ks ];
- //real f27_B = (D27.f[DIR_00P ])[kt ];
- //real f27_T = (D27.f[DIR_00M ])[kb ];
- //real f27_SW = (D27.f[DIR_PP0 ])[kne ];
- //real f27_NE = (D27.f[DIR_MM0 ])[ksw ];
- //real f27_NW = (D27.f[DIR_PM0 ])[kse ];
- //real f27_SE = (D27.f[DIR_MP0 ])[knw ];
- //real f27_BW = (D27.f[DIR_P0P ])[kte ];
- //real f27_TE = (D27.f[DIR_M0M ])[kbw ];
- //real f27_TW = (D27.f[DIR_P0M ])[kbe ];
- //real f27_BE = (D27.f[DIR_M0P ])[ktw ];
- //real f27_BS = (D27.f[DIR_0PP ])[ktn ];
- //real f27_TN = (D27.f[DIR_0MM ])[kbs ];
- //real f27_TS = (D27.f[DIR_0PM ])[kbn ];
- //real f27_BN = (D27.f[DIR_0MP ])[kts ];
+ //real f27_W = (D27.f[DIR_P00])[ke ];
+ //real f27_E = (D27.f[DIR_M00])[kw ];
+ //real f27_S = (D27.f[DIR_0P0])[kn ];
+ //real f27_N = (D27.f[DIR_0M0])[ks ];
+ //real f27_B = (D27.f[DIR_00P])[kt ];
+ //real f27_T = (D27.f[DIR_00M])[kb ];
+ //real f27_SW = (D27.f[DIR_PP0])[kne ];
+ //real f27_NE = (D27.f[DIR_MM0])[ksw ];
+ //real f27_NW = (D27.f[DIR_PM0])[kse ];
+ //real f27_SE = (D27.f[DIR_MP0])[knw ];
+ //real f27_BW = (D27.f[DIR_P0P])[kte ];
+ //real f27_TE = (D27.f[DIR_M0M])[kbw ];
+ //real f27_TW = (D27.f[DIR_P0M])[kbe ];
+ //real f27_BE = (D27.f[DIR_M0P])[ktw ];
+ //real f27_BS = (D27.f[DIR_0PP])[ktn ];
+ //real f27_TN = (D27.f[DIR_0MM])[kbs ];
+ //real f27_TS = (D27.f[DIR_0PM])[kbn ];
+ //real f27_BN = (D27.f[DIR_0MP])[kts ];
//real f27_ZERO = (D27.f[DIR_000])[kzero];
- //real f27_BSW = (D27.f[DIR_PPP ])[ktne ];
- //real f27_BNE = (D27.f[DIR_MMP ])[ktsw ];
- //real f27_BNW = (D27.f[DIR_PMP ])[ktse ];
- //real f27_BSE = (D27.f[DIR_MPP ])[ktnw ];
- //real f27_TSW = (D27.f[DIR_PPM ])[kbne ];
- //real f27_TNE = (D27.f[DIR_MMM ])[kbsw ];
- //real f27_TNW = (D27.f[DIR_PMM ])[kbse ];
- //real f27_TSE = (D27.f[DIR_MPM ])[kbnw ];
+ //real f27_BSW = (D27.f[DIR_PPP])[ktne ];
+ //real f27_BNE = (D27.f[DIR_MMP])[ktsw ];
+ //real f27_BNW = (D27.f[DIR_PMP])[ktse ];
+ //real f27_BSE = (D27.f[DIR_MPP])[ktnw ];
+ //real f27_TSW = (D27.f[DIR_PPM])[kbne ];
+ //real f27_TNE = (D27.f[DIR_MMM])[kbsw ];
+ //real f27_TNW = (D27.f[DIR_PMM])[kbse ];
+ //real f27_TSE = (D27.f[DIR_MPM])[kbnw ];
////////////////////////////////////////////////////////////////////////////////
real cu_sq=c3o2*(vx1*vx1+vx2*vx2+vx3*vx3);
////////////////////////////////////////////////////////////////////////////////
@@ -2233,63 +2233,63 @@ __global__ void QADVel27(real* DD,
//////////////////////////////////////////////////////////////////////////
if (isEvenTimestep==false)
{
- D27.f[DIR_P00 ] = &DD27[DIR_P00 *size_Mat];
- D27.f[DIR_M00 ] = &DD27[DIR_M00 *size_Mat];
- D27.f[DIR_0P0 ] = &DD27[DIR_0P0 *size_Mat];
- D27.f[DIR_0M0 ] = &DD27[DIR_0M0 *size_Mat];
- D27.f[DIR_00P ] = &DD27[DIR_00P *size_Mat];
- D27.f[DIR_00M ] = &DD27[DIR_00M *size_Mat];
- D27.f[DIR_PP0 ] = &DD27[DIR_PP0 *size_Mat];
- D27.f[DIR_MM0 ] = &DD27[DIR_MM0 *size_Mat];
- D27.f[DIR_PM0 ] = &DD27[DIR_PM0 *size_Mat];
- D27.f[DIR_MP0 ] = &DD27[DIR_MP0 *size_Mat];
- D27.f[DIR_P0P ] = &DD27[DIR_P0P *size_Mat];
- D27.f[DIR_M0M ] = &DD27[DIR_M0M *size_Mat];
- D27.f[DIR_P0M ] = &DD27[DIR_P0M *size_Mat];
- D27.f[DIR_M0P ] = &DD27[DIR_M0P *size_Mat];
- D27.f[DIR_0PP ] = &DD27[DIR_0PP *size_Mat];
- D27.f[DIR_0MM ] = &DD27[DIR_0MM *size_Mat];
- D27.f[DIR_0PM ] = &DD27[DIR_0PM *size_Mat];
- D27.f[DIR_0MP ] = &DD27[DIR_0MP *size_Mat];
- D27.f[DIR_000] = &DD27[DIR_000*size_Mat];
- D27.f[DIR_PPP ] = &DD27[DIR_PPP *size_Mat];
- D27.f[DIR_MMP ] = &DD27[DIR_MMP *size_Mat];
- D27.f[DIR_PMP ] = &DD27[DIR_PMP *size_Mat];
- D27.f[DIR_MPP ] = &DD27[DIR_MPP *size_Mat];
- D27.f[DIR_PPM ] = &DD27[DIR_PPM *size_Mat];
- D27.f[DIR_MMM ] = &DD27[DIR_MMM *size_Mat];
- D27.f[DIR_PMM ] = &DD27[DIR_PMM *size_Mat];
- D27.f[DIR_MPM ] = &DD27[DIR_MPM *size_Mat];
+ D27.f[DIR_P00] = &DD27[DIR_P00 * numberOfLBnodes];
+ D27.f[DIR_M00] = &DD27[DIR_M00 * numberOfLBnodes];
+ D27.f[DIR_0P0] = &DD27[DIR_0P0 * numberOfLBnodes];
+ D27.f[DIR_0M0] = &DD27[DIR_0M0 * numberOfLBnodes];
+ D27.f[DIR_00P] = &DD27[DIR_00P * numberOfLBnodes];
+ D27.f[DIR_00M] = &DD27[DIR_00M * numberOfLBnodes];
+ D27.f[DIR_PP0] = &DD27[DIR_PP0 * numberOfLBnodes];
+ D27.f[DIR_MM0] = &DD27[DIR_MM0 * numberOfLBnodes];
+ D27.f[DIR_PM0] = &DD27[DIR_PM0 * numberOfLBnodes];
+ D27.f[DIR_MP0] = &DD27[DIR_MP0 * numberOfLBnodes];
+ D27.f[DIR_P0P] = &DD27[DIR_P0P * numberOfLBnodes];
+ D27.f[DIR_M0M] = &DD27[DIR_M0M * numberOfLBnodes];
+ D27.f[DIR_P0M] = &DD27[DIR_P0M * numberOfLBnodes];
+ D27.f[DIR_M0P] = &DD27[DIR_M0P * numberOfLBnodes];
+ D27.f[DIR_0PP] = &DD27[DIR_0PP * numberOfLBnodes];
+ D27.f[DIR_0MM] = &DD27[DIR_0MM * numberOfLBnodes];
+ D27.f[DIR_0PM] = &DD27[DIR_0PM * numberOfLBnodes];
+ D27.f[DIR_0MP] = &DD27[DIR_0MP * numberOfLBnodes];
+ D27.f[DIR_000] = &DD27[DIR_000 * numberOfLBnodes];
+ D27.f[DIR_PPP] = &DD27[DIR_PPP * numberOfLBnodes];
+ D27.f[DIR_MMP] = &DD27[DIR_MMP * numberOfLBnodes];
+ D27.f[DIR_PMP] = &DD27[DIR_PMP * numberOfLBnodes];
+ D27.f[DIR_MPP] = &DD27[DIR_MPP * numberOfLBnodes];
+ D27.f[DIR_PPM] = &DD27[DIR_PPM * numberOfLBnodes];
+ D27.f[DIR_MMM] = &DD27[DIR_MMM * numberOfLBnodes];
+ D27.f[DIR_PMM] = &DD27[DIR_PMM * numberOfLBnodes];
+ D27.f[DIR_MPM] = &DD27[DIR_MPM * numberOfLBnodes];
}
else
{
- D27.f[DIR_M00 ] = &DD27[DIR_P00 *size_Mat];
- D27.f[DIR_P00 ] = &DD27[DIR_M00 *size_Mat];
- D27.f[DIR_0M0 ] = &DD27[DIR_0P0 *size_Mat];
- D27.f[DIR_0P0 ] = &DD27[DIR_0M0 *size_Mat];
- D27.f[DIR_00M ] = &DD27[DIR_00P *size_Mat];
- D27.f[DIR_00P ] = &DD27[DIR_00M *size_Mat];
- D27.f[DIR_MM0 ] = &DD27[DIR_PP0 *size_Mat];
- D27.f[DIR_PP0 ] = &DD27[DIR_MM0 *size_Mat];
- D27.f[DIR_MP0 ] = &DD27[DIR_PM0 *size_Mat];
- D27.f[DIR_PM0 ] = &DD27[DIR_MP0 *size_Mat];
- D27.f[DIR_M0M ] = &DD27[DIR_P0P *size_Mat];
- D27.f[DIR_P0P ] = &DD27[DIR_M0M *size_Mat];
- D27.f[DIR_M0P ] = &DD27[DIR_P0M *size_Mat];
- D27.f[DIR_P0M ] = &DD27[DIR_M0P *size_Mat];
- D27.f[DIR_0MM ] = &DD27[DIR_0PP *size_Mat];
- D27.f[DIR_0PP ] = &DD27[DIR_0MM *size_Mat];
- D27.f[DIR_0MP ] = &DD27[DIR_0PM *size_Mat];
- D27.f[DIR_0PM ] = &DD27[DIR_0MP *size_Mat];
- D27.f[DIR_000] = &DD27[DIR_000*size_Mat];
- D27.f[DIR_PPP ] = &DD27[DIR_MMM *size_Mat];
- D27.f[DIR_MMP ] = &DD27[DIR_PPM *size_Mat];
- D27.f[DIR_PMP ] = &DD27[DIR_MPM *size_Mat];
- D27.f[DIR_MPP ] = &DD27[DIR_PMM *size_Mat];
- D27.f[DIR_PPM ] = &DD27[DIR_MMP *size_Mat];
- D27.f[DIR_MMM ] = &DD27[DIR_PPP *size_Mat];
- D27.f[DIR_PMM ] = &DD27[DIR_MPP *size_Mat];
- D27.f[DIR_MPM ] = &DD27[DIR_PMP *size_Mat];
+ D27.f[DIR_M00] = &DD27[DIR_P00 * numberOfLBnodes];
+ D27.f[DIR_P00] = &DD27[DIR_M00 * numberOfLBnodes];
+ D27.f[DIR_0M0] = &DD27[DIR_0P0 * numberOfLBnodes];
+ D27.f[DIR_0P0] = &DD27[DIR_0M0 * numberOfLBnodes];
+ D27.f[DIR_00M] = &DD27[DIR_00P * numberOfLBnodes];
+ D27.f[DIR_00P] = &DD27[DIR_00M * numberOfLBnodes];
+ D27.f[DIR_MM0] = &DD27[DIR_PP0 * numberOfLBnodes];
+ D27.f[DIR_PP0] = &DD27[DIR_MM0 * numberOfLBnodes];
+ D27.f[DIR_MP0] = &DD27[DIR_PM0 * numberOfLBnodes];
+ D27.f[DIR_PM0] = &DD27[DIR_MP0 * numberOfLBnodes];
+ D27.f[DIR_M0M] = &DD27[DIR_P0P * numberOfLBnodes];
+ D27.f[DIR_P0P] = &DD27[DIR_M0M * numberOfLBnodes];
+ D27.f[DIR_M0P] = &DD27[DIR_P0M * numberOfLBnodes];
+ D27.f[DIR_P0M] = &DD27[DIR_M0P * numberOfLBnodes];
+ D27.f[DIR_0MM] = &DD27[DIR_0PP * numberOfLBnodes];
+ D27.f[DIR_0PP] = &DD27[DIR_0MM * numberOfLBnodes];
+ D27.f[DIR_0MP] = &DD27[DIR_0PM * numberOfLBnodes];
+ D27.f[DIR_0PM] = &DD27[DIR_0MP * numberOfLBnodes];
+ D27.f[DIR_000] = &DD27[DIR_000 * numberOfLBnodes];
+ D27.f[DIR_PPP] = &DD27[DIR_MMM * numberOfLBnodes];
+ D27.f[DIR_MMP] = &DD27[DIR_PPM * numberOfLBnodes];
+ D27.f[DIR_PMP] = &DD27[DIR_MPM * numberOfLBnodes];
+ D27.f[DIR_MPP] = &DD27[DIR_PMM * numberOfLBnodes];
+ D27.f[DIR_PPM] = &DD27[DIR_MMP * numberOfLBnodes];
+ D27.f[DIR_MMM] = &DD27[DIR_PPP * numberOfLBnodes];
+ D27.f[DIR_PMM] = &DD27[DIR_MPP * numberOfLBnodes];
+ D27.f[DIR_MPM] = &DD27[DIR_PMP * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
//Test
@@ -2299,24 +2299,24 @@ __global__ void QADVel27(real* DD,
//Test
//(D.f[DIR_000])[k]=c1o10;
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
- //(D27.f[DIR_M00 ])[kw ]= four;
- //(D27.f[DIR_P00 ])[ke ]= four;
- //(D27.f[DIR_0M0 ])[ks ]= four;
- //(D27.f[DIR_0P0 ])[kn ]= four;
- //(D27.f[DIR_00M ])[kb ]= four;
- //(D27.f[DIR_00P ])[kt ]= four;
- //(D27.f[DIR_MM0 ])[ksw ]= four;
- //(D27.f[DIR_PP0 ])[kne ]= four;
- //(D27.f[DIR_MP0 ])[knw ]= four;
- //(D27.f[DIR_PM0 ])[kse ]= four;
- //(D27.f[DIR_M0M ])[kbw ]= four;
- //(D27.f[DIR_P0P ])[kte ]= four;
- //(D27.f[DIR_M0P ])[ktw ]= four;
- //(D27.f[DIR_P0M ])[kbe ]= four;
- //(D27.f[DIR_0MM ])[kbs ]= four;
- //(D27.f[DIR_0PP ])[ktn ]= four;
- //(D27.f[DIR_0MP ])[kts ]= four;
- //(D27.f[DIR_0PM ])[kbn ]= four;
+ //(D27.f[DIR_M00])[kw ]= four;
+ //(D27.f[DIR_P00])[ke ]= four;
+ //(D27.f[DIR_0M0])[ks ]= four;
+ //(D27.f[DIR_0P0])[kn ]= four;
+ //(D27.f[DIR_00M])[kb ]= four;
+ //(D27.f[DIR_00P])[kt ]= four;
+ //(D27.f[DIR_MM0])[ksw ]= four;
+ //(D27.f[DIR_PP0])[kne ]= four;
+ //(D27.f[DIR_MP0])[knw ]= four;
+ //(D27.f[DIR_PM0])[kse ]= four;
+ //(D27.f[DIR_M0M])[kbw ]= four;
+ //(D27.f[DIR_P0P])[kte ]= four;
+ //(D27.f[DIR_M0P])[ktw ]= four;
+ //(D27.f[DIR_P0M])[kbe ]= four;
+ //(D27.f[DIR_0MM])[kbs ]= four;
+ //(D27.f[DIR_0PP])[ktn ]= four;
+ //(D27.f[DIR_0MP])[kts ]= four;
+ //(D27.f[DIR_0PM])[kbn ]= four;
//(D27.f[DIR_MMM])[kbsw]= four;
//(D27.f[DIR_PPP])[ktne]= four;
//(D27.f[DIR_MMP])[ktsw]= four;
@@ -2325,24 +2325,24 @@ __global__ void QADVel27(real* DD,
//(D27.f[DIR_PMP])[ktse]= four;
//(D27.f[DIR_MPP])[ktnw]= four;
//(D27.f[DIR_PMM])[kbse]= four;
- q = q_dirE[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_M00 ])[kw ]= -feqW27_W + c2o1 * c2o27 * TempD;
- q = q_dirW[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_P00 ])[ke ]= -feqW27_E + c2o1 * c2o27 * TempD;
- q = q_dirN[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_0M0 ])[ks ]= -feqW27_S + c2o1 * c2o27 * TempD;
- q = q_dirS[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_0P0 ])[kn ]= -feqW27_N + c2o1 * c2o27 * TempD;
- q = q_dirT[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_00M ])[kb ]= -feqW27_B + c2o1 * c2o27 * TempD;
- q = q_dirB[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_00P ])[kt ]= -feqW27_T + c2o1 * c2o27 * TempD;
- q = q_dirNE[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_MM0 ])[ksw ]= -feqW27_SW + c2o1 * c1o54 * TempD;
- q = q_dirSW[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_PP0 ])[kne ]= -feqW27_NE + c2o1 * c1o54 * TempD;
- q = q_dirSE[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_MP0 ])[knw ]= -feqW27_NW + c2o1 * c1o54 * TempD;
- q = q_dirNW[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_PM0 ])[kse ]= -feqW27_SE + c2o1 * c1o54 * TempD;
- q = q_dirTE[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_M0M ])[kbw ]= -feqW27_BW + c2o1 * c1o54 * TempD;
- q = q_dirBW[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_P0P ])[kte ]= -feqW27_TE + c2o1 * c1o54 * TempD;
- q = q_dirBE[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_M0P ])[ktw ]= -feqW27_TW + c2o1 * c1o54 * TempD;
- q = q_dirTW[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_P0M ])[kbe ]= -feqW27_BE + c2o1 * c1o54 * TempD;
- q = q_dirTN[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_0MM ])[kbs ]= -feqW27_BS + c2o1 * c1o54 * TempD;
- q = q_dirBS[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_0PP ])[ktn ]= -feqW27_TN + c2o1 * c1o54 * TempD;
- q = q_dirBN[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_0MP ])[kts ]= -feqW27_TS + c2o1 * c1o54 * TempD;
- q = q_dirTS[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_0PM ])[kbn ]= -feqW27_BN + c2o1 * c1o54 * TempD;
+ q = q_dirE[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_M00])[kw ]= -feqW27_W + c2o1 * c2o27 * TempD;
+ q = q_dirW[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_P00])[ke ]= -feqW27_E + c2o1 * c2o27 * TempD;
+ q = q_dirN[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_0M0])[ks ]= -feqW27_S + c2o1 * c2o27 * TempD;
+ q = q_dirS[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_0P0])[kn ]= -feqW27_N + c2o1 * c2o27 * TempD;
+ q = q_dirT[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_00M])[kb ]= -feqW27_B + c2o1 * c2o27 * TempD;
+ q = q_dirB[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_00P])[kt ]= -feqW27_T + c2o1 * c2o27 * TempD;
+ q = q_dirNE[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_MM0])[ksw ]= -feqW27_SW + c2o1 * c1o54 * TempD;
+ q = q_dirSW[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_PP0])[kne ]= -feqW27_NE + c2o1 * c1o54 * TempD;
+ q = q_dirSE[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_MP0])[knw ]= -feqW27_NW + c2o1 * c1o54 * TempD;
+ q = q_dirNW[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_PM0])[kse ]= -feqW27_SE + c2o1 * c1o54 * TempD;
+ q = q_dirTE[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_M0M])[kbw ]= -feqW27_BW + c2o1 * c1o54 * TempD;
+ q = q_dirBW[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_P0P])[kte ]= -feqW27_TE + c2o1 * c1o54 * TempD;
+ q = q_dirBE[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_M0P])[ktw ]= -feqW27_TW + c2o1 * c1o54 * TempD;
+ q = q_dirTW[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_P0M])[kbe ]= -feqW27_BE + c2o1 * c1o54 * TempD;
+ q = q_dirTN[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_0MM])[kbs ]= -feqW27_BS + c2o1 * c1o54 * TempD;
+ q = q_dirBS[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_0PP])[ktn ]= -feqW27_TN + c2o1 * c1o54 * TempD;
+ q = q_dirBN[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_0MP])[kts ]= -feqW27_TS + c2o1 * c1o54 * TempD;
+ q = q_dirTS[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_0PM])[kbn ]= -feqW27_BN + c2o1 * c1o54 * TempD;
q = q_dirTNE[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_MMM])[kbsw]= -feqW27_BSW+ c2o1 * c1o216 * TempD;
q = q_dirBSW[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_PPP])[ktne]= -feqW27_TNE+ c2o1 * c1o216 * TempD;
q = q_dirBNE[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_MMP])[ktsw]= -feqW27_TSW+ c2o1 * c1o216 * TempD;
@@ -2351,24 +2351,24 @@ __global__ void QADVel27(real* DD,
q = q_dirBNW[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_PMP])[ktse]= -feqW27_TSE+ c2o1 * c1o216 * TempD;
q = q_dirBSE[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_MPP])[ktnw]= -feqW27_TNW+ c2o1 * c1o216 * TempD;
q = q_dirTNW[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_PMM])[kbse]= -feqW27_BSE+ c2o1 * c1o216 * TempD;
- //q = q_dirE[k]; if (q>=zero && q<=one) (D27.f[DIR_M00 ])[kw ]=(two*feqW27_W -(f27_E *(q*omegaD-one)-omegaD*feq27_E *(q-one))/(omegaD-one)+f27_W *q)/(q+one);
- //q = q_dirW[k]; if (q>=zero && q<=one) (D27.f[DIR_P00 ])[ke ]=(two*feqW27_E -(f27_W *(q*omegaD-one)-omegaD*feq27_W *(q-one))/(omegaD-one)+f27_E *q)/(q+one);
- //q = q_dirN[k]; if (q>=zero && q<=one) (D27.f[DIR_0M0 ])[ks ]=(two*feqW27_S -(f27_N *(q*omegaD-one)-omegaD*feq27_N *(q-one))/(omegaD-one)+f27_S *q)/(q+one);
- //q = q_dirS[k]; if (q>=zero && q<=one) (D27.f[DIR_0P0 ])[kn ]=(two*feqW27_N -(f27_S *(q*omegaD-one)-omegaD*feq27_S *(q-one))/(omegaD-one)+f27_N *q)/(q+one);
- //q = q_dirT[k]; if (q>=zero && q<=one) (D27.f[DIR_00M ])[kb ]=(two*feqW27_B -(f27_T *(q*omegaD-one)-omegaD*feq27_T *(q-one))/(omegaD-one)+f27_B *q)/(q+one);
- //q = q_dirB[k]; if (q>=zero && q<=one) (D27.f[DIR_00P ])[kt ]=(two*feqW27_T -(f27_B *(q*omegaD-one)-omegaD*feq27_B *(q-one))/(omegaD-one)+f27_T *q)/(q+one);
- //q = q_dirNE[k]; if (q>=zero && q<=one) (D27.f[DIR_MM0 ])[ksw ]=(two*feqW27_SW -(f27_NE *(q*omegaD-one)-omegaD*feq27_NE *(q-one))/(omegaD-one)+f27_SW *q)/(q+one);
- //q = q_dirSW[k]; if (q>=zero && q<=one) (D27.f[DIR_PP0 ])[kne ]=(two*feqW27_NE -(f27_SW *(q*omegaD-one)-omegaD*feq27_SW *(q-one))/(omegaD-one)+f27_NE *q)/(q+one);
- //q = q_dirSE[k]; if (q>=zero && q<=one) (D27.f[DIR_MP0 ])[knw ]=(two*feqW27_NW -(f27_SE *(q*omegaD-one)-omegaD*feq27_SE *(q-one))/(omegaD-one)+f27_NW *q)/(q+one);
- //q = q_dirNW[k]; if (q>=zero && q<=one) (D27.f[DIR_PM0 ])[kse ]=(two*feqW27_SE -(f27_NW *(q*omegaD-one)-omegaD*feq27_NW *(q-one))/(omegaD-one)+f27_SE *q)/(q+one);
- //q = q_dirTE[k]; if (q>=zero && q<=one) (D27.f[DIR_M0M ])[kbw ]=(two*feqW27_BW -(f27_TE *(q*omegaD-one)-omegaD*feq27_TE *(q-one))/(omegaD-one)+f27_BW *q)/(q+one);
- //q = q_dirBW[k]; if (q>=zero && q<=one) (D27.f[DIR_P0P ])[kte ]=(two*feqW27_TE -(f27_BW *(q*omegaD-one)-omegaD*feq27_BW *(q-one))/(omegaD-one)+f27_TE *q)/(q+one);
- //q = q_dirBE[k]; if (q>=zero && q<=one) (D27.f[DIR_M0P ])[ktw ]=(two*feqW27_TW -(f27_BE *(q*omegaD-one)-omegaD*feq27_BE *(q-one))/(omegaD-one)+f27_TW *q)/(q+one);
- //q = q_dirTW[k]; if (q>=zero && q<=one) (D27.f[DIR_P0M ])[kbe ]=(two*feqW27_BE -(f27_TW *(q*omegaD-one)-omegaD*feq27_TW *(q-one))/(omegaD-one)+f27_BE *q)/(q+one);
- //q = q_dirTN[k]; if (q>=zero && q<=one) (D27.f[DIR_0MM ])[kbs ]=(two*feqW27_BS -(f27_TN *(q*omegaD-one)-omegaD*feq27_TN *(q-one))/(omegaD-one)+f27_BS *q)/(q+one);
- //q = q_dirBS[k]; if (q>=zero && q<=one) (D27.f[DIR_0PP ])[ktn ]=(two*feqW27_TN -(f27_BS *(q*omegaD-one)-omegaD*feq27_BS *(q-one))/(omegaD-one)+f27_TN *q)/(q+one);
- //q = q_dirBN[k]; if (q>=zero && q<=one) (D27.f[DIR_0MP ])[kts ]=(two*feqW27_TS -(f27_BN *(q*omegaD-one)-omegaD*feq27_BN *(q-one))/(omegaD-one)+f27_TS *q)/(q+one);
- //q = q_dirTS[k]; if (q>=zero && q<=one) (D27.f[DIR_0PM ])[kbn ]=(two*feqW27_BN -(f27_TS *(q*omegaD-one)-omegaD*feq27_TS *(q-one))/(omegaD-one)+f27_BN *q)/(q+one);
+ //q = q_dirE[k]; if (q>=zero && q<=one) (D27.f[DIR_M00])[kw ]=(two*feqW27_W -(f27_E *(q*omegaD-one)-omegaD*feq27_E *(q-one))/(omegaD-one)+f27_W *q)/(q+one);
+ //q = q_dirW[k]; if (q>=zero && q<=one) (D27.f[DIR_P00])[ke ]=(two*feqW27_E -(f27_W *(q*omegaD-one)-omegaD*feq27_W *(q-one))/(omegaD-one)+f27_E *q)/(q+one);
+ //q = q_dirN[k]; if (q>=zero && q<=one) (D27.f[DIR_0M0])[ks ]=(two*feqW27_S -(f27_N *(q*omegaD-one)-omegaD*feq27_N *(q-one))/(omegaD-one)+f27_S *q)/(q+one);
+ //q = q_dirS[k]; if (q>=zero && q<=one) (D27.f[DIR_0P0])[kn ]=(two*feqW27_N -(f27_S *(q*omegaD-one)-omegaD*feq27_S *(q-one))/(omegaD-one)+f27_N *q)/(q+one);
+ //q = q_dirT[k]; if (q>=zero && q<=one) (D27.f[DIR_00M])[kb ]=(two*feqW27_B -(f27_T *(q*omegaD-one)-omegaD*feq27_T *(q-one))/(omegaD-one)+f27_B *q)/(q+one);
+ //q = q_dirB[k]; if (q>=zero && q<=one) (D27.f[DIR_00P])[kt ]=(two*feqW27_T -(f27_B *(q*omegaD-one)-omegaD*feq27_B *(q-one))/(omegaD-one)+f27_T *q)/(q+one);
+ //q = q_dirNE[k]; if (q>=zero && q<=one) (D27.f[DIR_MM0])[ksw ]=(two*feqW27_SW -(f27_NE *(q*omegaD-one)-omegaD*feq27_NE *(q-one))/(omegaD-one)+f27_SW *q)/(q+one);
+ //q = q_dirSW[k]; if (q>=zero && q<=one) (D27.f[DIR_PP0])[kne ]=(two*feqW27_NE -(f27_SW *(q*omegaD-one)-omegaD*feq27_SW *(q-one))/(omegaD-one)+f27_NE *q)/(q+one);
+ //q = q_dirSE[k]; if (q>=zero && q<=one) (D27.f[DIR_MP0])[knw ]=(two*feqW27_NW -(f27_SE *(q*omegaD-one)-omegaD*feq27_SE *(q-one))/(omegaD-one)+f27_NW *q)/(q+one);
+ //q = q_dirNW[k]; if (q>=zero && q<=one) (D27.f[DIR_PM0])[kse ]=(two*feqW27_SE -(f27_NW *(q*omegaD-one)-omegaD*feq27_NW *(q-one))/(omegaD-one)+f27_SE *q)/(q+one);
+ //q = q_dirTE[k]; if (q>=zero && q<=one) (D27.f[DIR_M0M])[kbw ]=(two*feqW27_BW -(f27_TE *(q*omegaD-one)-omegaD*feq27_TE *(q-one))/(omegaD-one)+f27_BW *q)/(q+one);
+ //q = q_dirBW[k]; if (q>=zero && q<=one) (D27.f[DIR_P0P])[kte ]=(two*feqW27_TE -(f27_BW *(q*omegaD-one)-omegaD*feq27_BW *(q-one))/(omegaD-one)+f27_TE *q)/(q+one);
+ //q = q_dirBE[k]; if (q>=zero && q<=one) (D27.f[DIR_M0P])[ktw ]=(two*feqW27_TW -(f27_BE *(q*omegaD-one)-omegaD*feq27_BE *(q-one))/(omegaD-one)+f27_TW *q)/(q+one);
+ //q = q_dirTW[k]; if (q>=zero && q<=one) (D27.f[DIR_P0M])[kbe ]=(two*feqW27_BE -(f27_TW *(q*omegaD-one)-omegaD*feq27_TW *(q-one))/(omegaD-one)+f27_BE *q)/(q+one);
+ //q = q_dirTN[k]; if (q>=zero && q<=one) (D27.f[DIR_0MM])[kbs ]=(two*feqW27_BS -(f27_TN *(q*omegaD-one)-omegaD*feq27_TN *(q-one))/(omegaD-one)+f27_BS *q)/(q+one);
+ //q = q_dirBS[k]; if (q>=zero && q<=one) (D27.f[DIR_0PP])[ktn ]=(two*feqW27_TN -(f27_BS *(q*omegaD-one)-omegaD*feq27_BS *(q-one))/(omegaD-one)+f27_TN *q)/(q+one);
+ //q = q_dirBN[k]; if (q>=zero && q<=one) (D27.f[DIR_0MP])[kts ]=(two*feqW27_TS -(f27_BN *(q*omegaD-one)-omegaD*feq27_BN *(q-one))/(omegaD-one)+f27_TS *q)/(q+one);
+ //q = q_dirTS[k]; if (q>=zero && q<=one) (D27.f[DIR_0PM])[kbn ]=(two*feqW27_BN -(f27_TS *(q*omegaD-one)-omegaD*feq27_TS *(q-one))/(omegaD-one)+f27_BN *q)/(q+one);
//q = q_dirTNE[k]; if (q>=zero && q<=one) (D27.f[DIR_MMM])[kbsw]=(two*feqW27_BSW-(f27_TNE*(q*omegaD-one)-omegaD*feq27_TNE*(q-one))/(omegaD-one)+f27_BSW*q)/(q+one);
//q = q_dirBSW[k]; if (q>=zero && q<=one) (D27.f[DIR_PPP])[ktne]=(two*feqW27_TNE-(f27_BSW*(q*omegaD-one)-omegaD*feq27_BSW*(q-one))/(omegaD-one)+f27_TNE*q)/(q+one);
//q = q_dirBNE[k]; if (q>=zero && q<=one) (D27.f[DIR_MMP])[ktsw]=(two*feqW27_TSW-(f27_BNE*(q*omegaD-one)-omegaD*feq27_BNE*(q-one))/(omegaD-one)+f27_TSW*q)/(q+one);
@@ -2431,91 +2431,91 @@ __global__ void QAD7( real* DD,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep)
{
Distributions27 D;
if (isEvenTimestep==true)
{
- D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
+ D.f[DIR_M00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_MPM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_PMP * numberOfLBnodes];
}
Distributions7 D7;
if (isEvenTimestep==true)
{
- D7.f[0] = &DD7[0*size_Mat];
- D7.f[1] = &DD7[1*size_Mat];
- D7.f[2] = &DD7[2*size_Mat];
- D7.f[3] = &DD7[3*size_Mat];
- D7.f[4] = &DD7[4*size_Mat];
- D7.f[5] = &DD7[5*size_Mat];
- D7.f[6] = &DD7[6*size_Mat];
+ D7.f[0] = &DD7[0*numberOfLBnodes];
+ D7.f[1] = &DD7[1*numberOfLBnodes];
+ D7.f[2] = &DD7[2*numberOfLBnodes];
+ D7.f[3] = &DD7[3*numberOfLBnodes];
+ D7.f[4] = &DD7[4*numberOfLBnodes];
+ D7.f[5] = &DD7[5*numberOfLBnodes];
+ D7.f[6] = &DD7[6*numberOfLBnodes];
}
else
{
- D7.f[0] = &DD7[0*size_Mat];
- D7.f[2] = &DD7[1*size_Mat];
- D7.f[1] = &DD7[2*size_Mat];
- D7.f[4] = &DD7[3*size_Mat];
- D7.f[3] = &DD7[4*size_Mat];
- D7.f[6] = &DD7[5*size_Mat];
- D7.f[5] = &DD7[6*size_Mat];
+ D7.f[0] = &DD7[0*numberOfLBnodes];
+ D7.f[2] = &DD7[1*numberOfLBnodes];
+ D7.f[1] = &DD7[2*numberOfLBnodes];
+ D7.f[4] = &DD7[3*numberOfLBnodes];
+ D7.f[3] = &DD7[4*numberOfLBnodes];
+ D7.f[6] = &DD7[5*numberOfLBnodes];
+ D7.f[5] = &DD7[6*numberOfLBnodes];
}
@@ -2539,24 +2539,24 @@ __global__ void QAD7( real* DD,
// *q_dirTNE, *q_dirTSW, *q_dirTSE, *q_dirTNW, *q_dirBNE, *q_dirBSW,
// *q_dirBSE, *q_dirBNW;
- q_dirE = &QQ[DIR_P00 * numberOfBCnodes];
- q_dirW = &QQ[DIR_M00 * numberOfBCnodes];
- q_dirN = &QQ[DIR_0P0 * numberOfBCnodes];
- q_dirS = &QQ[DIR_0M0 * numberOfBCnodes];
- q_dirT = &QQ[DIR_00P * numberOfBCnodes];
- q_dirB = &QQ[DIR_00M * numberOfBCnodes];
- //q_dirNE = &QQ[DIR_PP0 * numberOfBCnodes];
- //q_dirSW = &QQ[DIR_MM0 * numberOfBCnodes];
- //q_dirSE = &QQ[DIR_PM0 * numberOfBCnodes];
- //q_dirNW = &QQ[DIR_MP0 * numberOfBCnodes];
- //q_dirTE = &QQ[DIR_P0P * numberOfBCnodes];
- //q_dirBW = &QQ[DIR_M0M * numberOfBCnodes];
- //q_dirBE = &QQ[DIR_P0M * numberOfBCnodes];
- //q_dirTW = &QQ[DIR_M0P * numberOfBCnodes];
- //q_dirTN = &QQ[DIR_0PP * numberOfBCnodes];
- //q_dirBS = &QQ[DIR_0MM * numberOfBCnodes];
- //q_dirBN = &QQ[DIR_0PM * numberOfBCnodes];
- //q_dirTS = &QQ[DIR_0MP * numberOfBCnodes];
+ q_dirE = &QQ[DIR_P00 * numberOfBCnodes];
+ q_dirW = &QQ[DIR_M00 * numberOfBCnodes];
+ q_dirN = &QQ[DIR_0P0 * numberOfBCnodes];
+ q_dirS = &QQ[DIR_0M0 * numberOfBCnodes];
+ q_dirT = &QQ[DIR_00P * numberOfBCnodes];
+ q_dirB = &QQ[DIR_00M * numberOfBCnodes];
+ //q_dirNE = &QQ[DIR_PP0 * numberOfBCnodes];
+ //q_dirSW = &QQ[DIR_MM0 * numberOfBCnodes];
+ //q_dirSE = &QQ[DIR_PM0 * numberOfBCnodes];
+ //q_dirNW = &QQ[DIR_MP0 * numberOfBCnodes];
+ //q_dirTE = &QQ[DIR_P0P * numberOfBCnodes];
+ //q_dirBW = &QQ[DIR_M0M * numberOfBCnodes];
+ //q_dirBE = &QQ[DIR_P0M * numberOfBCnodes];
+ //q_dirTW = &QQ[DIR_M0P * numberOfBCnodes];
+ //q_dirTN = &QQ[DIR_0PP * numberOfBCnodes];
+ //q_dirBS = &QQ[DIR_0MM * numberOfBCnodes];
+ //q_dirBN = &QQ[DIR_0PM * numberOfBCnodes];
+ //q_dirTS = &QQ[DIR_0MP * numberOfBCnodes];
//q_dirTNE = &QQ[DIR_PPP * numberOfBCnodes];
//q_dirTSW = &QQ[DIR_MMP * numberOfBCnodes];
//q_dirTSE = &QQ[DIR_PMP * numberOfBCnodes];
@@ -2599,32 +2599,32 @@ __global__ void QAD7( real* DD,
real f_E, f_W, f_N, f_S, f_T, f_B, f_NE, f_SW, f_SE, f_NW, f_TE, f_BW, f_BE,
f_TW, f_TN, f_BS, f_BN, f_TS, f_TNE, f_TSW, f_TSE, f_TNW, f_BNE, f_BSW, f_BSE, f_BNW;
- f_W = (D.f[DIR_P00 ])[ke ];
- f_E = (D.f[DIR_M00 ])[kw ];
- f_S = (D.f[DIR_0P0 ])[kn ];
- f_N = (D.f[DIR_0M0 ])[ks ];
- f_B = (D.f[DIR_00P ])[kt ];
- f_T = (D.f[DIR_00M ])[kb ];
- f_SW = (D.f[DIR_PP0 ])[kne ];
- f_NE = (D.f[DIR_MM0 ])[ksw ];
- f_NW = (D.f[DIR_PM0 ])[kse ];
- f_SE = (D.f[DIR_MP0 ])[knw ];
- f_BW = (D.f[DIR_P0P ])[kte ];
- f_TE = (D.f[DIR_M0M ])[kbw ];
- f_TW = (D.f[DIR_P0M ])[kbe ];
- f_BE = (D.f[DIR_M0P ])[ktw ];
- f_BS = (D.f[DIR_0PP ])[ktn ];
- f_TN = (D.f[DIR_0MM ])[kbs ];
- f_TS = (D.f[DIR_0PM ])[kbn ];
- f_BN = (D.f[DIR_0MP ])[kts ];
- f_BSW = (D.f[DIR_PPP ])[ktne ];
- f_BNE = (D.f[DIR_MMP ])[ktsw ];
- f_BNW = (D.f[DIR_PMP ])[ktse ];
- f_BSE = (D.f[DIR_MPP ])[ktnw ];
- f_TSW = (D.f[DIR_PPM ])[kbne ];
- f_TNE = (D.f[DIR_MMM ])[kbsw ];
- f_TNW = (D.f[DIR_PMM ])[kbse ];
- f_TSE = (D.f[DIR_MPM ])[kbnw ];
+ f_W = (D.f[DIR_P00])[ke ];
+ f_E = (D.f[DIR_M00])[kw ];
+ f_S = (D.f[DIR_0P0])[kn ];
+ f_N = (D.f[DIR_0M0])[ks ];
+ f_B = (D.f[DIR_00P])[kt ];
+ f_T = (D.f[DIR_00M])[kb ];
+ f_SW = (D.f[DIR_PP0])[kne ];
+ f_NE = (D.f[DIR_MM0])[ksw ];
+ f_NW = (D.f[DIR_PM0])[kse ];
+ f_SE = (D.f[DIR_MP0])[knw ];
+ f_BW = (D.f[DIR_P0P])[kte ];
+ f_TE = (D.f[DIR_M0M])[kbw ];
+ f_TW = (D.f[DIR_P0M])[kbe ];
+ f_BE = (D.f[DIR_M0P])[ktw ];
+ f_BS = (D.f[DIR_0PP])[ktn ];
+ f_TN = (D.f[DIR_0MM])[kbs ];
+ f_TS = (D.f[DIR_0PM])[kbn ];
+ f_BN = (D.f[DIR_0MP])[kts ];
+ f_BSW = (D.f[DIR_PPP])[ktne ];
+ f_BNE = (D.f[DIR_MMP])[ktsw ];
+ f_BNW = (D.f[DIR_PMP])[ktse ];
+ f_BSE = (D.f[DIR_MPP])[ktnw ];
+ f_TSW = (D.f[DIR_PPM])[kbne ];
+ f_TNE = (D.f[DIR_MMM])[kbsw ];
+ f_TNW = (D.f[DIR_PMM])[kbse ];
+ f_TSE = (D.f[DIR_MPM])[kbnw ];
////////////////////////////////////////////////////////////////////////////////
real vx1, vx2, vx3/*, drho*/;
//drho = f_TSE + f_TNW + f_TNE + f_TSW + f_BSE + f_BNW + f_BNE + f_BSW +
@@ -2696,23 +2696,23 @@ __global__ void QAD7( real* DD,
//pointertausch
if (isEvenTimestep==false)
{
- D7.f[0] = &DD7[0*size_Mat];
- D7.f[1] = &DD7[1*size_Mat];
- D7.f[2] = &DD7[2*size_Mat];
- D7.f[3] = &DD7[3*size_Mat];
- D7.f[4] = &DD7[4*size_Mat];
- D7.f[5] = &DD7[5*size_Mat];
- D7.f[6] = &DD7[6*size_Mat];
+ D7.f[0] = &DD7[0*numberOfLBnodes];
+ D7.f[1] = &DD7[1*numberOfLBnodes];
+ D7.f[2] = &DD7[2*numberOfLBnodes];
+ D7.f[3] = &DD7[3*numberOfLBnodes];
+ D7.f[4] = &DD7[4*numberOfLBnodes];
+ D7.f[5] = &DD7[5*numberOfLBnodes];
+ D7.f[6] = &DD7[6*numberOfLBnodes];
}
else
{
- D7.f[0] = &DD7[0*size_Mat];
- D7.f[2] = &DD7[1*size_Mat];
- D7.f[1] = &DD7[2*size_Mat];
- D7.f[4] = &DD7[3*size_Mat];
- D7.f[3] = &DD7[4*size_Mat];
- D7.f[6] = &DD7[5*size_Mat];
- D7.f[5] = &DD7[6*size_Mat];
+ D7.f[0] = &DD7[0*numberOfLBnodes];
+ D7.f[2] = &DD7[1*numberOfLBnodes];
+ D7.f[1] = &DD7[2*numberOfLBnodes];
+ D7.f[4] = &DD7[3*numberOfLBnodes];
+ D7.f[3] = &DD7[4*numberOfLBnodes];
+ D7.f[6] = &DD7[5*numberOfLBnodes];
+ D7.f[5] = &DD7[6*numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
@@ -2864,131 +2864,131 @@ __global__ void QADDirichlet27(
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep)
{
Distributions27 D;
if (isEvenTimestep==true)
{
- D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
+ D.f[DIR_M00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_MPM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_PMP * numberOfLBnodes];
}
Distributions27 D27;
if (isEvenTimestep==true)
{
- D27.f[DIR_P00 ] = &DD27[DIR_P00 *size_Mat];
- D27.f[DIR_M00 ] = &DD27[DIR_M00 *size_Mat];
- D27.f[DIR_0P0 ] = &DD27[DIR_0P0 *size_Mat];
- D27.f[DIR_0M0 ] = &DD27[DIR_0M0 *size_Mat];
- D27.f[DIR_00P ] = &DD27[DIR_00P *size_Mat];
- D27.f[DIR_00M ] = &DD27[DIR_00M *size_Mat];
- D27.f[DIR_PP0 ] = &DD27[DIR_PP0 *size_Mat];
- D27.f[DIR_MM0 ] = &DD27[DIR_MM0 *size_Mat];
- D27.f[DIR_PM0 ] = &DD27[DIR_PM0 *size_Mat];
- D27.f[DIR_MP0 ] = &DD27[DIR_MP0 *size_Mat];
- D27.f[DIR_P0P ] = &DD27[DIR_P0P *size_Mat];
- D27.f[DIR_M0M ] = &DD27[DIR_M0M *size_Mat];
- D27.f[DIR_P0M ] = &DD27[DIR_P0M *size_Mat];
- D27.f[DIR_M0P ] = &DD27[DIR_M0P *size_Mat];
- D27.f[DIR_0PP ] = &DD27[DIR_0PP *size_Mat];
- D27.f[DIR_0MM ] = &DD27[DIR_0MM *size_Mat];
- D27.f[DIR_0PM ] = &DD27[DIR_0PM *size_Mat];
- D27.f[DIR_0MP ] = &DD27[DIR_0MP *size_Mat];
- D27.f[DIR_000] = &DD27[DIR_000*size_Mat];
- D27.f[DIR_PPP ] = &DD27[DIR_PPP *size_Mat];
- D27.f[DIR_MMP ] = &DD27[DIR_MMP *size_Mat];
- D27.f[DIR_PMP ] = &DD27[DIR_PMP *size_Mat];
- D27.f[DIR_MPP ] = &DD27[DIR_MPP *size_Mat];
- D27.f[DIR_PPM ] = &DD27[DIR_PPM *size_Mat];
- D27.f[DIR_MMM ] = &DD27[DIR_MMM *size_Mat];
- D27.f[DIR_PMM ] = &DD27[DIR_PMM *size_Mat];
- D27.f[DIR_MPM ] = &DD27[DIR_MPM *size_Mat];
+ D27.f[DIR_P00] = &DD27[DIR_P00 * numberOfLBnodes];
+ D27.f[DIR_M00] = &DD27[DIR_M00 * numberOfLBnodes];
+ D27.f[DIR_0P0] = &DD27[DIR_0P0 * numberOfLBnodes];
+ D27.f[DIR_0M0] = &DD27[DIR_0M0 * numberOfLBnodes];
+ D27.f[DIR_00P] = &DD27[DIR_00P * numberOfLBnodes];
+ D27.f[DIR_00M] = &DD27[DIR_00M * numberOfLBnodes];
+ D27.f[DIR_PP0] = &DD27[DIR_PP0 * numberOfLBnodes];
+ D27.f[DIR_MM0] = &DD27[DIR_MM0 * numberOfLBnodes];
+ D27.f[DIR_PM0] = &DD27[DIR_PM0 * numberOfLBnodes];
+ D27.f[DIR_MP0] = &DD27[DIR_MP0 * numberOfLBnodes];
+ D27.f[DIR_P0P] = &DD27[DIR_P0P * numberOfLBnodes];
+ D27.f[DIR_M0M] = &DD27[DIR_M0M * numberOfLBnodes];
+ D27.f[DIR_P0M] = &DD27[DIR_P0M * numberOfLBnodes];
+ D27.f[DIR_M0P] = &DD27[DIR_M0P * numberOfLBnodes];
+ D27.f[DIR_0PP] = &DD27[DIR_0PP * numberOfLBnodes];
+ D27.f[DIR_0MM] = &DD27[DIR_0MM * numberOfLBnodes];
+ D27.f[DIR_0PM] = &DD27[DIR_0PM * numberOfLBnodes];
+ D27.f[DIR_0MP] = &DD27[DIR_0MP * numberOfLBnodes];
+ D27.f[DIR_000] = &DD27[DIR_000 * numberOfLBnodes];
+ D27.f[DIR_PPP] = &DD27[DIR_PPP * numberOfLBnodes];
+ D27.f[DIR_MMP] = &DD27[DIR_MMP * numberOfLBnodes];
+ D27.f[DIR_PMP] = &DD27[DIR_PMP * numberOfLBnodes];
+ D27.f[DIR_MPP] = &DD27[DIR_MPP * numberOfLBnodes];
+ D27.f[DIR_PPM] = &DD27[DIR_PPM * numberOfLBnodes];
+ D27.f[DIR_MMM] = &DD27[DIR_MMM * numberOfLBnodes];
+ D27.f[DIR_PMM] = &DD27[DIR_PMM * numberOfLBnodes];
+ D27.f[DIR_MPM] = &DD27[DIR_MPM * numberOfLBnodes];
}
else
{
- D27.f[DIR_M00 ] = &DD27[DIR_P00 *size_Mat];
- D27.f[DIR_P00 ] = &DD27[DIR_M00 *size_Mat];
- D27.f[DIR_0M0 ] = &DD27[DIR_0P0 *size_Mat];
- D27.f[DIR_0P0 ] = &DD27[DIR_0M0 *size_Mat];
- D27.f[DIR_00M ] = &DD27[DIR_00P *size_Mat];
- D27.f[DIR_00P ] = &DD27[DIR_00M *size_Mat];
- D27.f[DIR_MM0 ] = &DD27[DIR_PP0 *size_Mat];
- D27.f[DIR_PP0 ] = &DD27[DIR_MM0 *size_Mat];
- D27.f[DIR_MP0 ] = &DD27[DIR_PM0 *size_Mat];
- D27.f[DIR_PM0 ] = &DD27[DIR_MP0 *size_Mat];
- D27.f[DIR_M0M ] = &DD27[DIR_P0P *size_Mat];
- D27.f[DIR_P0P ] = &DD27[DIR_M0M *size_Mat];
- D27.f[DIR_M0P ] = &DD27[DIR_P0M *size_Mat];
- D27.f[DIR_P0M ] = &DD27[DIR_M0P *size_Mat];
- D27.f[DIR_0MM ] = &DD27[DIR_0PP *size_Mat];
- D27.f[DIR_0PP ] = &DD27[DIR_0MM *size_Mat];
- D27.f[DIR_0MP ] = &DD27[DIR_0PM *size_Mat];
- D27.f[DIR_0PM ] = &DD27[DIR_0MP *size_Mat];
- D27.f[DIR_000] = &DD27[DIR_000*size_Mat];
- D27.f[DIR_PPP ] = &DD27[DIR_MMM *size_Mat];
- D27.f[DIR_MMP ] = &DD27[DIR_PPM *size_Mat];
- D27.f[DIR_PMP ] = &DD27[DIR_MPM *size_Mat];
- D27.f[DIR_MPP ] = &DD27[DIR_PMM *size_Mat];
- D27.f[DIR_PPM ] = &DD27[DIR_MMP *size_Mat];
- D27.f[DIR_MMM ] = &DD27[DIR_PPP *size_Mat];
- D27.f[DIR_PMM ] = &DD27[DIR_MPP *size_Mat];
- D27.f[DIR_MPM ] = &DD27[DIR_PMP *size_Mat];
+ D27.f[DIR_M00] = &DD27[DIR_P00 * numberOfLBnodes];
+ D27.f[DIR_P00] = &DD27[DIR_M00 * numberOfLBnodes];
+ D27.f[DIR_0M0] = &DD27[DIR_0P0 * numberOfLBnodes];
+ D27.f[DIR_0P0] = &DD27[DIR_0M0 * numberOfLBnodes];
+ D27.f[DIR_00M] = &DD27[DIR_00P * numberOfLBnodes];
+ D27.f[DIR_00P] = &DD27[DIR_00M * numberOfLBnodes];
+ D27.f[DIR_MM0] = &DD27[DIR_PP0 * numberOfLBnodes];
+ D27.f[DIR_PP0] = &DD27[DIR_MM0 * numberOfLBnodes];
+ D27.f[DIR_MP0] = &DD27[DIR_PM0 * numberOfLBnodes];
+ D27.f[DIR_PM0] = &DD27[DIR_MP0 * numberOfLBnodes];
+ D27.f[DIR_M0M] = &DD27[DIR_P0P * numberOfLBnodes];
+ D27.f[DIR_P0P] = &DD27[DIR_M0M * numberOfLBnodes];
+ D27.f[DIR_M0P] = &DD27[DIR_P0M * numberOfLBnodes];
+ D27.f[DIR_P0M] = &DD27[DIR_M0P * numberOfLBnodes];
+ D27.f[DIR_0MM] = &DD27[DIR_0PP * numberOfLBnodes];
+ D27.f[DIR_0PP] = &DD27[DIR_0MM * numberOfLBnodes];
+ D27.f[DIR_0MP] = &DD27[DIR_0PM * numberOfLBnodes];
+ D27.f[DIR_0PM] = &DD27[DIR_0MP * numberOfLBnodes];
+ D27.f[DIR_000] = &DD27[DIR_000 * numberOfLBnodes];
+ D27.f[DIR_PPP] = &DD27[DIR_MMM * numberOfLBnodes];
+ D27.f[DIR_MMP] = &DD27[DIR_PPM * numberOfLBnodes];
+ D27.f[DIR_PMP] = &DD27[DIR_MPM * numberOfLBnodes];
+ D27.f[DIR_MPP] = &DD27[DIR_PMM * numberOfLBnodes];
+ D27.f[DIR_PPM] = &DD27[DIR_MMP * numberOfLBnodes];
+ D27.f[DIR_MMM] = &DD27[DIR_PPP * numberOfLBnodes];
+ D27.f[DIR_PMM] = &DD27[DIR_MPP * numberOfLBnodes];
+ D27.f[DIR_MPM] = &DD27[DIR_PMP * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////
const unsigned x = threadIdx.x; // Globaler x-Index
@@ -3009,24 +3009,24 @@ __global__ void QADDirichlet27(
*q_dirBE, *q_dirTW, *q_dirTN, *q_dirBS, *q_dirBN, *q_dirTS,
*q_dirTNE, *q_dirTSW, *q_dirTSE, *q_dirTNW, *q_dirBNE, *q_dirBSW,
*q_dirBSE, *q_dirBNW;
- q_dirE = &QQ[DIR_P00 * numberOfBCnodes];
- q_dirW = &QQ[DIR_M00 * numberOfBCnodes];
- q_dirN = &QQ[DIR_0P0 * numberOfBCnodes];
- q_dirS = &QQ[DIR_0M0 * numberOfBCnodes];
- q_dirT = &QQ[DIR_00P * numberOfBCnodes];
- q_dirB = &QQ[DIR_00M * numberOfBCnodes];
- q_dirNE = &QQ[DIR_PP0 * numberOfBCnodes];
- q_dirSW = &QQ[DIR_MM0 * numberOfBCnodes];
- q_dirSE = &QQ[DIR_PM0 * numberOfBCnodes];
- q_dirNW = &QQ[DIR_MP0 * numberOfBCnodes];
- q_dirTE = &QQ[DIR_P0P * numberOfBCnodes];
- q_dirBW = &QQ[DIR_M0M * numberOfBCnodes];
- q_dirBE = &QQ[DIR_P0M * numberOfBCnodes];
- q_dirTW = &QQ[DIR_M0P * numberOfBCnodes];
- q_dirTN = &QQ[DIR_0PP * numberOfBCnodes];
- q_dirBS = &QQ[DIR_0MM * numberOfBCnodes];
- q_dirBN = &QQ[DIR_0PM * numberOfBCnodes];
- q_dirTS = &QQ[DIR_0MP * numberOfBCnodes];
+ q_dirE = &QQ[DIR_P00 * numberOfBCnodes];
+ q_dirW = &QQ[DIR_M00 * numberOfBCnodes];
+ q_dirN = &QQ[DIR_0P0 * numberOfBCnodes];
+ q_dirS = &QQ[DIR_0M0 * numberOfBCnodes];
+ q_dirT = &QQ[DIR_00P * numberOfBCnodes];
+ q_dirB = &QQ[DIR_00M * numberOfBCnodes];
+ q_dirNE = &QQ[DIR_PP0 * numberOfBCnodes];
+ q_dirSW = &QQ[DIR_MM0 * numberOfBCnodes];
+ q_dirSE = &QQ[DIR_PM0 * numberOfBCnodes];
+ q_dirNW = &QQ[DIR_MP0 * numberOfBCnodes];
+ q_dirTE = &QQ[DIR_P0P * numberOfBCnodes];
+ q_dirBW = &QQ[DIR_M0M * numberOfBCnodes];
+ q_dirBE = &QQ[DIR_P0M * numberOfBCnodes];
+ q_dirTW = &QQ[DIR_M0P * numberOfBCnodes];
+ q_dirTN = &QQ[DIR_0PP * numberOfBCnodes];
+ q_dirBS = &QQ[DIR_0MM * numberOfBCnodes];
+ q_dirBN = &QQ[DIR_0PM * numberOfBCnodes];
+ q_dirTS = &QQ[DIR_0MP * numberOfBCnodes];
q_dirTNE = &QQ[DIR_PPP * numberOfBCnodes];
q_dirTSW = &QQ[DIR_MMP * numberOfBCnodes];
q_dirTSE = &QQ[DIR_PMP * numberOfBCnodes];
@@ -3066,33 +3066,33 @@ __global__ void QADDirichlet27(
unsigned int ktne = KQK;
unsigned int kbsw = neighborZ[ksw];
////////////////////////////////////////////////////////////////////////////////
- real f_W = (D.f[DIR_P00 ])[ke ];
- real f_E = (D.f[DIR_M00 ])[kw ];
- real f_S = (D.f[DIR_0P0 ])[kn ];
- real f_N = (D.f[DIR_0M0 ])[ks ];
- real f_B = (D.f[DIR_00P ])[kt ];
- real f_T = (D.f[DIR_00M ])[kb ];
- real f_SW = (D.f[DIR_PP0 ])[kne ];
- real f_NE = (D.f[DIR_MM0 ])[ksw ];
- real f_NW = (D.f[DIR_PM0 ])[kse ];
- real f_SE = (D.f[DIR_MP0 ])[knw ];
- real f_BW = (D.f[DIR_P0P ])[kte ];
- real f_TE = (D.f[DIR_M0M ])[kbw ];
- real f_TW = (D.f[DIR_P0M ])[kbe ];
- real f_BE = (D.f[DIR_M0P ])[ktw ];
- real f_BS = (D.f[DIR_0PP ])[ktn ];
- real f_TN = (D.f[DIR_0MM ])[kbs ];
- real f_TS = (D.f[DIR_0PM ])[kbn ];
- real f_BN = (D.f[DIR_0MP ])[kts ];
+ real f_W = (D.f[DIR_P00])[ke ];
+ real f_E = (D.f[DIR_M00])[kw ];
+ real f_S = (D.f[DIR_0P0])[kn ];
+ real f_N = (D.f[DIR_0M0])[ks ];
+ real f_B = (D.f[DIR_00P])[kt ];
+ real f_T = (D.f[DIR_00M])[kb ];
+ real f_SW = (D.f[DIR_PP0])[kne ];
+ real f_NE = (D.f[DIR_MM0])[ksw ];
+ real f_NW = (D.f[DIR_PM0])[kse ];
+ real f_SE = (D.f[DIR_MP0])[knw ];
+ real f_BW = (D.f[DIR_P0P])[kte ];
+ real f_TE = (D.f[DIR_M0M])[kbw ];
+ real f_TW = (D.f[DIR_P0M])[kbe ];
+ real f_BE = (D.f[DIR_M0P])[ktw ];
+ real f_BS = (D.f[DIR_0PP])[ktn ];
+ real f_TN = (D.f[DIR_0MM])[kbs ];
+ real f_TS = (D.f[DIR_0PM])[kbn ];
+ real f_BN = (D.f[DIR_0MP])[kts ];
real f_ZERO = (D.f[DIR_000])[kzero];
- real f_BSW = (D.f[DIR_PPP ])[ktne ];
- real f_BNE = (D.f[DIR_MMP ])[ktsw ];
- real f_BNW = (D.f[DIR_PMP ])[ktse ];
- real f_BSE = (D.f[DIR_MPP ])[ktnw ];
- real f_TSW = (D.f[DIR_PPM ])[kbne ];
- real f_TNE = (D.f[DIR_MMM ])[kbsw ];
- real f_TNW = (D.f[DIR_PMM ])[kbse ];
- real f_TSE = (D.f[DIR_MPM ])[kbnw ];
+ real f_BSW = (D.f[DIR_PPP])[ktne ];
+ real f_BNE = (D.f[DIR_MMP])[ktsw ];
+ real f_BNW = (D.f[DIR_PMP])[ktse ];
+ real f_BSE = (D.f[DIR_MPP])[ktnw ];
+ real f_TSW = (D.f[DIR_PPM])[kbne ];
+ real f_TNE = (D.f[DIR_MMM])[kbsw ];
+ real f_TNW = (D.f[DIR_PMM])[kbse ];
+ real f_TSE = (D.f[DIR_MPM])[kbnw ];
////////////////////////////////////////////////////////////////////////////////
real vx1, vx2, vx3, /*drho, feq,*/ q;
////drho = f_TSE + f_TNW + f_TNE + f_TSW + f_BSE + f_BNW + f_BNE + f_BSW +
@@ -3118,33 +3118,33 @@ __global__ void QADDirichlet27(
vx2 = OORho*((f_TNE-f_BSW)+(f_BNE-f_TSW)+(f_BNW-f_TSE)+(f_TNW-f_BSE) +(f_NE-f_SW)+(f_NW-f_SE)+(f_TN-f_BS)+(f_BN-f_TS)+(f_N-f_S));
vx3 = OORho*((f_TNE-f_BSW)+(f_TSW-f_BNE)+(f_TSE-f_BNW)+(f_TNW-f_BSE) +(f_TE-f_BW)+(f_TW-f_BE)+(f_TN-f_BS)+(f_TS-f_BN)+(f_T-f_B));
////////////////////////////////////////////////////////////////////////////////
- real f27_W = (D27.f[DIR_P00 ])[ke ];
- real f27_E = (D27.f[DIR_M00 ])[kw ];
- real f27_S = (D27.f[DIR_0P0 ])[kn ];
- real f27_N = (D27.f[DIR_0M0 ])[ks ];
- real f27_B = (D27.f[DIR_00P ])[kt ];
- real f27_T = (D27.f[DIR_00M ])[kb ];
- real f27_SW = (D27.f[DIR_PP0 ])[kne ];
- real f27_NE = (D27.f[DIR_MM0 ])[ksw ];
- real f27_NW = (D27.f[DIR_PM0 ])[kse ];
- real f27_SE = (D27.f[DIR_MP0 ])[knw ];
- real f27_BW = (D27.f[DIR_P0P ])[kte ];
- real f27_TE = (D27.f[DIR_M0M ])[kbw ];
- real f27_TW = (D27.f[DIR_P0M ])[kbe ];
- real f27_BE = (D27.f[DIR_M0P ])[ktw ];
- real f27_BS = (D27.f[DIR_0PP ])[ktn ];
- real f27_TN = (D27.f[DIR_0MM ])[kbs ];
- real f27_TS = (D27.f[DIR_0PM ])[kbn ];
- real f27_BN = (D27.f[DIR_0MP ])[kts ];
+ real f27_W = (D27.f[DIR_P00])[ke ];
+ real f27_E = (D27.f[DIR_M00])[kw ];
+ real f27_S = (D27.f[DIR_0P0])[kn ];
+ real f27_N = (D27.f[DIR_0M0])[ks ];
+ real f27_B = (D27.f[DIR_00P])[kt ];
+ real f27_T = (D27.f[DIR_00M])[kb ];
+ real f27_SW = (D27.f[DIR_PP0])[kne ];
+ real f27_NE = (D27.f[DIR_MM0])[ksw ];
+ real f27_NW = (D27.f[DIR_PM0])[kse ];
+ real f27_SE = (D27.f[DIR_MP0])[knw ];
+ real f27_BW = (D27.f[DIR_P0P])[kte ];
+ real f27_TE = (D27.f[DIR_M0M])[kbw ];
+ real f27_TW = (D27.f[DIR_P0M])[kbe ];
+ real f27_BE = (D27.f[DIR_M0P])[ktw ];
+ real f27_BS = (D27.f[DIR_0PP])[ktn ];
+ real f27_TN = (D27.f[DIR_0MM])[kbs ];
+ real f27_TS = (D27.f[DIR_0PM])[kbn ];
+ real f27_BN = (D27.f[DIR_0MP])[kts ];
real f27_ZERO = (D27.f[DIR_000])[kzero];
- real f27_BSW = (D27.f[DIR_PPP ])[ktne ];
- real f27_BNE = (D27.f[DIR_MMP ])[ktsw ];
- real f27_BNW = (D27.f[DIR_PMP ])[ktse ];
- real f27_BSE = (D27.f[DIR_MPP ])[ktnw ];
- real f27_TSW = (D27.f[DIR_PPM ])[kbne ];
- real f27_TNE = (D27.f[DIR_MMM ])[kbsw ];
- real f27_TNW = (D27.f[DIR_PMM ])[kbse ];
- real f27_TSE = (D27.f[DIR_MPM ])[kbnw ];
+ real f27_BSW = (D27.f[DIR_PPP])[ktne ];
+ real f27_BNE = (D27.f[DIR_MMP])[ktsw ];
+ real f27_BNW = (D27.f[DIR_PMP])[ktse ];
+ real f27_BSE = (D27.f[DIR_MPP])[ktnw ];
+ real f27_TSW = (D27.f[DIR_PPM])[kbne ];
+ real f27_TNE = (D27.f[DIR_MMM])[kbsw ];
+ real f27_TNW = (D27.f[DIR_PMM])[kbse ];
+ real f27_TSE = (D27.f[DIR_MPM])[kbnw ];
////////////////////////////////////////////////////////////////////////////////
real cu_sq=c3o2*(vx1*vx1+vx2*vx2+vx3*vx3);
////////////////////////////////////////////////////////////////////////////////
@@ -3220,86 +3220,86 @@ __global__ void QADDirichlet27(
//////////////////////////////////////////////////////////////////////////
if (isEvenTimestep==false)
{
- D27.f[DIR_P00 ] = &DD27[DIR_P00 *size_Mat];
- D27.f[DIR_M00 ] = &DD27[DIR_M00 *size_Mat];
- D27.f[DIR_0P0 ] = &DD27[DIR_0P0 *size_Mat];
- D27.f[DIR_0M0 ] = &DD27[DIR_0M0 *size_Mat];
- D27.f[DIR_00P ] = &DD27[DIR_00P *size_Mat];
- D27.f[DIR_00M ] = &DD27[DIR_00M *size_Mat];
- D27.f[DIR_PP0 ] = &DD27[DIR_PP0 *size_Mat];
- D27.f[DIR_MM0 ] = &DD27[DIR_MM0 *size_Mat];
- D27.f[DIR_PM0 ] = &DD27[DIR_PM0 *size_Mat];
- D27.f[DIR_MP0 ] = &DD27[DIR_MP0 *size_Mat];
- D27.f[DIR_P0P ] = &DD27[DIR_P0P *size_Mat];
- D27.f[DIR_M0M ] = &DD27[DIR_M0M *size_Mat];
- D27.f[DIR_P0M ] = &DD27[DIR_P0M *size_Mat];
- D27.f[DIR_M0P ] = &DD27[DIR_M0P *size_Mat];
- D27.f[DIR_0PP ] = &DD27[DIR_0PP *size_Mat];
- D27.f[DIR_0MM ] = &DD27[DIR_0MM *size_Mat];
- D27.f[DIR_0PM ] = &DD27[DIR_0PM *size_Mat];
- D27.f[DIR_0MP ] = &DD27[DIR_0MP *size_Mat];
- D27.f[DIR_000] = &DD27[DIR_000*size_Mat];
- D27.f[DIR_PPP ] = &DD27[DIR_PPP *size_Mat];
- D27.f[DIR_MMP ] = &DD27[DIR_MMP *size_Mat];
- D27.f[DIR_PMP ] = &DD27[DIR_PMP *size_Mat];
- D27.f[DIR_MPP ] = &DD27[DIR_MPP *size_Mat];
- D27.f[DIR_PPM ] = &DD27[DIR_PPM *size_Mat];
- D27.f[DIR_MMM ] = &DD27[DIR_MMM *size_Mat];
- D27.f[DIR_PMM ] = &DD27[DIR_PMM *size_Mat];
- D27.f[DIR_MPM ] = &DD27[DIR_MPM *size_Mat];
+ D27.f[DIR_P00] = &DD27[DIR_P00 * numberOfLBnodes];
+ D27.f[DIR_M00] = &DD27[DIR_M00 * numberOfLBnodes];
+ D27.f[DIR_0P0] = &DD27[DIR_0P0 * numberOfLBnodes];
+ D27.f[DIR_0M0] = &DD27[DIR_0M0 * numberOfLBnodes];
+ D27.f[DIR_00P] = &DD27[DIR_00P * numberOfLBnodes];
+ D27.f[DIR_00M] = &DD27[DIR_00M * numberOfLBnodes];
+ D27.f[DIR_PP0] = &DD27[DIR_PP0 * numberOfLBnodes];
+ D27.f[DIR_MM0] = &DD27[DIR_MM0 * numberOfLBnodes];
+ D27.f[DIR_PM0] = &DD27[DIR_PM0 * numberOfLBnodes];
+ D27.f[DIR_MP0] = &DD27[DIR_MP0 * numberOfLBnodes];
+ D27.f[DIR_P0P] = &DD27[DIR_P0P * numberOfLBnodes];
+ D27.f[DIR_M0M] = &DD27[DIR_M0M * numberOfLBnodes];
+ D27.f[DIR_P0M] = &DD27[DIR_P0M * numberOfLBnodes];
+ D27.f[DIR_M0P] = &DD27[DIR_M0P * numberOfLBnodes];
+ D27.f[DIR_0PP] = &DD27[DIR_0PP * numberOfLBnodes];
+ D27.f[DIR_0MM] = &DD27[DIR_0MM * numberOfLBnodes];
+ D27.f[DIR_0PM] = &DD27[DIR_0PM * numberOfLBnodes];
+ D27.f[DIR_0MP] = &DD27[DIR_0MP * numberOfLBnodes];
+ D27.f[DIR_000] = &DD27[DIR_000 * numberOfLBnodes];
+ D27.f[DIR_PPP] = &DD27[DIR_PPP * numberOfLBnodes];
+ D27.f[DIR_MMP] = &DD27[DIR_MMP * numberOfLBnodes];
+ D27.f[DIR_PMP] = &DD27[DIR_PMP * numberOfLBnodes];
+ D27.f[DIR_MPP] = &DD27[DIR_MPP * numberOfLBnodes];
+ D27.f[DIR_PPM] = &DD27[DIR_PPM * numberOfLBnodes];
+ D27.f[DIR_MMM] = &DD27[DIR_MMM * numberOfLBnodes];
+ D27.f[DIR_PMM] = &DD27[DIR_PMM * numberOfLBnodes];
+ D27.f[DIR_MPM] = &DD27[DIR_MPM * numberOfLBnodes];
}
else
{
- D27.f[DIR_M00 ] = &DD27[DIR_P00 *size_Mat];
- D27.f[DIR_P00 ] = &DD27[DIR_M00 *size_Mat];
- D27.f[DIR_0M0 ] = &DD27[DIR_0P0 *size_Mat];
- D27.f[DIR_0P0 ] = &DD27[DIR_0M0 *size_Mat];
- D27.f[DIR_00M ] = &DD27[DIR_00P *size_Mat];
- D27.f[DIR_00P ] = &DD27[DIR_00M *size_Mat];
- D27.f[DIR_MM0 ] = &DD27[DIR_PP0 *size_Mat];
- D27.f[DIR_PP0 ] = &DD27[DIR_MM0 *size_Mat];
- D27.f[DIR_MP0 ] = &DD27[DIR_PM0 *size_Mat];
- D27.f[DIR_PM0 ] = &DD27[DIR_MP0 *size_Mat];
- D27.f[DIR_M0M ] = &DD27[DIR_P0P *size_Mat];
- D27.f[DIR_P0P ] = &DD27[DIR_M0M *size_Mat];
- D27.f[DIR_M0P ] = &DD27[DIR_P0M *size_Mat];
- D27.f[DIR_P0M ] = &DD27[DIR_M0P *size_Mat];
- D27.f[DIR_0MM ] = &DD27[DIR_0PP *size_Mat];
- D27.f[DIR_0PP ] = &DD27[DIR_0MM *size_Mat];
- D27.f[DIR_0MP ] = &DD27[DIR_0PM *size_Mat];
- D27.f[DIR_0PM ] = &DD27[DIR_0MP *size_Mat];
- D27.f[DIR_000] = &DD27[DIR_000*size_Mat];
- D27.f[DIR_PPP ] = &DD27[DIR_MMM *size_Mat];
- D27.f[DIR_MMP ] = &DD27[DIR_PPM *size_Mat];
- D27.f[DIR_PMP ] = &DD27[DIR_MPM *size_Mat];
- D27.f[DIR_MPP ] = &DD27[DIR_PMM *size_Mat];
- D27.f[DIR_PPM ] = &DD27[DIR_MMP *size_Mat];
- D27.f[DIR_MMM ] = &DD27[DIR_PPP *size_Mat];
- D27.f[DIR_PMM ] = &DD27[DIR_MPP *size_Mat];
- D27.f[DIR_MPM ] = &DD27[DIR_PMP *size_Mat];
+ D27.f[DIR_M00] = &DD27[DIR_P00 * numberOfLBnodes];
+ D27.f[DIR_P00] = &DD27[DIR_M00 * numberOfLBnodes];
+ D27.f[DIR_0M0] = &DD27[DIR_0P0 * numberOfLBnodes];
+ D27.f[DIR_0P0] = &DD27[DIR_0M0 * numberOfLBnodes];
+ D27.f[DIR_00M] = &DD27[DIR_00P * numberOfLBnodes];
+ D27.f[DIR_00P] = &DD27[DIR_00M * numberOfLBnodes];
+ D27.f[DIR_MM0] = &DD27[DIR_PP0 * numberOfLBnodes];
+ D27.f[DIR_PP0] = &DD27[DIR_MM0 * numberOfLBnodes];
+ D27.f[DIR_MP0] = &DD27[DIR_PM0 * numberOfLBnodes];
+ D27.f[DIR_PM0] = &DD27[DIR_MP0 * numberOfLBnodes];
+ D27.f[DIR_M0M] = &DD27[DIR_P0P * numberOfLBnodes];
+ D27.f[DIR_P0P] = &DD27[DIR_M0M * numberOfLBnodes];
+ D27.f[DIR_M0P] = &DD27[DIR_P0M * numberOfLBnodes];
+ D27.f[DIR_P0M] = &DD27[DIR_M0P * numberOfLBnodes];
+ D27.f[DIR_0MM] = &DD27[DIR_0PP * numberOfLBnodes];
+ D27.f[DIR_0PP] = &DD27[DIR_0MM * numberOfLBnodes];
+ D27.f[DIR_0MP] = &DD27[DIR_0PM * numberOfLBnodes];
+ D27.f[DIR_0PM] = &DD27[DIR_0MP * numberOfLBnodes];
+ D27.f[DIR_000] = &DD27[DIR_000 * numberOfLBnodes];
+ D27.f[DIR_PPP] = &DD27[DIR_MMM * numberOfLBnodes];
+ D27.f[DIR_MMP] = &DD27[DIR_PPM * numberOfLBnodes];
+ D27.f[DIR_PMP] = &DD27[DIR_MPM * numberOfLBnodes];
+ D27.f[DIR_MPP] = &DD27[DIR_PMM * numberOfLBnodes];
+ D27.f[DIR_PPM] = &DD27[DIR_MMP * numberOfLBnodes];
+ D27.f[DIR_MMM] = &DD27[DIR_PPP * numberOfLBnodes];
+ D27.f[DIR_PMM] = &DD27[DIR_MPP * numberOfLBnodes];
+ D27.f[DIR_MPM] = &DD27[DIR_PMP * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
//Test
//(D.f[DIR_000])[k]=0.1f;
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
- q = q_dirE[ ke ]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_M00 ])[kw ]=(c2o1*feqW27_W -(f27_E *(q*omegaD-c1o1)-omegaD*feq27_E *(q-c1o1))/(omegaD-c1o1)+f27_W *q)/(q+c1o1);
- q = q_dirW[ kw ]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_P00 ])[ke ]=(c2o1*feqW27_E -(f27_W *(q*omegaD-c1o1)-omegaD*feq27_W *(q-c1o1))/(omegaD-c1o1)+f27_E *q)/(q+c1o1);
- q = q_dirN[ kn ]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_0M0 ])[ks ]=(c2o1*feqW27_S -(f27_N *(q*omegaD-c1o1)-omegaD*feq27_N *(q-c1o1))/(omegaD-c1o1)+f27_S *q)/(q+c1o1);
- q = q_dirS[ ks ]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_0P0 ])[kn ]=(c2o1*feqW27_N -(f27_S *(q*omegaD-c1o1)-omegaD*feq27_S *(q-c1o1))/(omegaD-c1o1)+f27_N *q)/(q+c1o1);
- q = q_dirT[ kt ]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_00M ])[kb ]=(c2o1*feqW27_B -(f27_T *(q*omegaD-c1o1)-omegaD*feq27_T *(q-c1o1))/(omegaD-c1o1)+f27_B *q)/(q+c1o1);
- q = q_dirB[ kb ]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_00P ])[kt ]=(c2o1*feqW27_T -(f27_B *(q*omegaD-c1o1)-omegaD*feq27_B *(q-c1o1))/(omegaD-c1o1)+f27_T *q)/(q+c1o1);
- q = q_dirNE[ kne ]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_MM0 ])[ksw ]=(c2o1*feqW27_SW -(f27_NE *(q*omegaD-c1o1)-omegaD*feq27_NE *(q-c1o1))/(omegaD-c1o1)+f27_SW *q)/(q+c1o1);
- q = q_dirSW[ ksw ]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_PP0 ])[kne ]=(c2o1*feqW27_NE -(f27_SW *(q*omegaD-c1o1)-omegaD*feq27_SW *(q-c1o1))/(omegaD-c1o1)+f27_NE *q)/(q+c1o1);
- q = q_dirSE[ kse ]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_MP0 ])[knw ]=(c2o1*feqW27_NW -(f27_SE *(q*omegaD-c1o1)-omegaD*feq27_SE *(q-c1o1))/(omegaD-c1o1)+f27_NW *q)/(q+c1o1);
- q = q_dirNW[ knw ]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_PM0 ])[kse ]=(c2o1*feqW27_SE -(f27_NW *(q*omegaD-c1o1)-omegaD*feq27_NW *(q-c1o1))/(omegaD-c1o1)+f27_SE *q)/(q+c1o1);
- q = q_dirTE[ kte ]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_M0M ])[kbw ]=(c2o1*feqW27_BW -(f27_TE *(q*omegaD-c1o1)-omegaD*feq27_TE *(q-c1o1))/(omegaD-c1o1)+f27_BW *q)/(q+c1o1);
- q = q_dirBW[ kbw ]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_P0P ])[kte ]=(c2o1*feqW27_TE -(f27_BW *(q*omegaD-c1o1)-omegaD*feq27_BW *(q-c1o1))/(omegaD-c1o1)+f27_TE *q)/(q+c1o1);
- q = q_dirBE[ kbe ]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_M0P ])[ktw ]=(c2o1*feqW27_TW -(f27_BE *(q*omegaD-c1o1)-omegaD*feq27_BE *(q-c1o1))/(omegaD-c1o1)+f27_TW *q)/(q+c1o1);
- q = q_dirTW[ ktw ]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_P0M ])[kbe ]=(c2o1*feqW27_BE -(f27_TW *(q*omegaD-c1o1)-omegaD*feq27_TW *(q-c1o1))/(omegaD-c1o1)+f27_BE *q)/(q+c1o1);
- q = q_dirTN[ ktn ]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_0MM ])[kbs ]=(c2o1*feqW27_BS -(f27_TN *(q*omegaD-c1o1)-omegaD*feq27_TN *(q-c1o1))/(omegaD-c1o1)+f27_BS *q)/(q+c1o1);
- q = q_dirBS[ kbs ]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_0PP ])[ktn ]=(c2o1*feqW27_TN -(f27_BS *(q*omegaD-c1o1)-omegaD*feq27_BS *(q-c1o1))/(omegaD-c1o1)+f27_TN *q)/(q+c1o1);
- q = q_dirBN[ kbn ]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_0MP ])[kts ]=(c2o1*feqW27_TS -(f27_BN *(q*omegaD-c1o1)-omegaD*feq27_BN *(q-c1o1))/(omegaD-c1o1)+f27_TS *q)/(q+c1o1);
- q = q_dirTS[ kts ]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_0PM ])[kbn ]=(c2o1*feqW27_BN -(f27_TS *(q*omegaD-c1o1)-omegaD*feq27_TS *(q-c1o1))/(omegaD-c1o1)+f27_BN *q)/(q+c1o1);
+ q = q_dirE[ ke ]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_M00])[kw ]=(c2o1*feqW27_W -(f27_E *(q*omegaD-c1o1)-omegaD*feq27_E *(q-c1o1))/(omegaD-c1o1)+f27_W *q)/(q+c1o1);
+ q = q_dirW[ kw ]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_P00])[ke ]=(c2o1*feqW27_E -(f27_W *(q*omegaD-c1o1)-omegaD*feq27_W *(q-c1o1))/(omegaD-c1o1)+f27_E *q)/(q+c1o1);
+ q = q_dirN[ kn ]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_0M0])[ks ]=(c2o1*feqW27_S -(f27_N *(q*omegaD-c1o1)-omegaD*feq27_N *(q-c1o1))/(omegaD-c1o1)+f27_S *q)/(q+c1o1);
+ q = q_dirS[ ks ]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_0P0])[kn ]=(c2o1*feqW27_N -(f27_S *(q*omegaD-c1o1)-omegaD*feq27_S *(q-c1o1))/(omegaD-c1o1)+f27_N *q)/(q+c1o1);
+ q = q_dirT[ kt ]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_00M])[kb ]=(c2o1*feqW27_B -(f27_T *(q*omegaD-c1o1)-omegaD*feq27_T *(q-c1o1))/(omegaD-c1o1)+f27_B *q)/(q+c1o1);
+ q = q_dirB[ kb ]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_00P])[kt ]=(c2o1*feqW27_T -(f27_B *(q*omegaD-c1o1)-omegaD*feq27_B *(q-c1o1))/(omegaD-c1o1)+f27_T *q)/(q+c1o1);
+ q = q_dirNE[ kne ]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_MM0])[ksw ]=(c2o1*feqW27_SW -(f27_NE *(q*omegaD-c1o1)-omegaD*feq27_NE *(q-c1o1))/(omegaD-c1o1)+f27_SW *q)/(q+c1o1);
+ q = q_dirSW[ ksw ]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_PP0])[kne ]=(c2o1*feqW27_NE -(f27_SW *(q*omegaD-c1o1)-omegaD*feq27_SW *(q-c1o1))/(omegaD-c1o1)+f27_NE *q)/(q+c1o1);
+ q = q_dirSE[ kse ]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_MP0])[knw ]=(c2o1*feqW27_NW -(f27_SE *(q*omegaD-c1o1)-omegaD*feq27_SE *(q-c1o1))/(omegaD-c1o1)+f27_NW *q)/(q+c1o1);
+ q = q_dirNW[ knw ]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_PM0])[kse ]=(c2o1*feqW27_SE -(f27_NW *(q*omegaD-c1o1)-omegaD*feq27_NW *(q-c1o1))/(omegaD-c1o1)+f27_SE *q)/(q+c1o1);
+ q = q_dirTE[ kte ]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_M0M])[kbw ]=(c2o1*feqW27_BW -(f27_TE *(q*omegaD-c1o1)-omegaD*feq27_TE *(q-c1o1))/(omegaD-c1o1)+f27_BW *q)/(q+c1o1);
+ q = q_dirBW[ kbw ]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_P0P])[kte ]=(c2o1*feqW27_TE -(f27_BW *(q*omegaD-c1o1)-omegaD*feq27_BW *(q-c1o1))/(omegaD-c1o1)+f27_TE *q)/(q+c1o1);
+ q = q_dirBE[ kbe ]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_M0P])[ktw ]=(c2o1*feqW27_TW -(f27_BE *(q*omegaD-c1o1)-omegaD*feq27_BE *(q-c1o1))/(omegaD-c1o1)+f27_TW *q)/(q+c1o1);
+ q = q_dirTW[ ktw ]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_P0M])[kbe ]=(c2o1*feqW27_BE -(f27_TW *(q*omegaD-c1o1)-omegaD*feq27_TW *(q-c1o1))/(omegaD-c1o1)+f27_BE *q)/(q+c1o1);
+ q = q_dirTN[ ktn ]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_0MM])[kbs ]=(c2o1*feqW27_BS -(f27_TN *(q*omegaD-c1o1)-omegaD*feq27_TN *(q-c1o1))/(omegaD-c1o1)+f27_BS *q)/(q+c1o1);
+ q = q_dirBS[ kbs ]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_0PP])[ktn ]=(c2o1*feqW27_TN -(f27_BS *(q*omegaD-c1o1)-omegaD*feq27_BS *(q-c1o1))/(omegaD-c1o1)+f27_TN *q)/(q+c1o1);
+ q = q_dirBN[ kbn ]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_0MP])[kts ]=(c2o1*feqW27_TS -(f27_BN *(q*omegaD-c1o1)-omegaD*feq27_BN *(q-c1o1))/(omegaD-c1o1)+f27_TS *q)/(q+c1o1);
+ q = q_dirTS[ kts ]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_0PM])[kbn ]=(c2o1*feqW27_BN -(f27_TS *(q*omegaD-c1o1)-omegaD*feq27_TS *(q-c1o1))/(omegaD-c1o1)+f27_BN *q)/(q+c1o1);
q = q_dirTNE[ktne ]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_MMM])[kbsw]=(c2o1*feqW27_BSW-(f27_TNE*(q*omegaD-c1o1)-omegaD*feq27_TNE*(q-c1o1))/(omegaD-c1o1)+f27_BSW*q)/(q+c1o1);
q = q_dirBSW[kbsw ]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_PPP])[ktne]=(c2o1*feqW27_TNE-(f27_BSW*(q*omegaD-c1o1)-omegaD*feq27_BSW*(q-c1o1))/(omegaD-c1o1)+f27_TNE*q)/(q+c1o1);
q = q_dirBNE[kbne ]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_MMP])[ktsw]=(c2o1*feqW27_TSW-(f27_BNE*(q*omegaD-c1o1)-omegaD*feq27_BNE*(q-c1o1))/(omegaD-c1o1)+f27_TSW*q)/(q+c1o1);
@@ -3308,24 +3308,24 @@ __global__ void QADDirichlet27(
q = q_dirBNW[kbnw ]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_PMP])[ktse]=(c2o1*feqW27_TSE-(f27_BNW*(q*omegaD-c1o1)-omegaD*feq27_BNW*(q-c1o1))/(omegaD-c1o1)+f27_TSE*q)/(q+c1o1);
q = q_dirBSE[kbse ]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_MPP])[ktnw]=(c2o1*feqW27_TNW-(f27_BSE*(q*omegaD-c1o1)-omegaD*feq27_BSE*(q-c1o1))/(omegaD-c1o1)+f27_TNW*q)/(q+c1o1);
q = q_dirTNW[ktnw ]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_PMM])[kbse]=(c2o1*feqW27_BSE-(f27_TNW*(q*omegaD-c1o1)-omegaD*feq27_TNW*(q-c1o1))/(omegaD-c1o1)+f27_BSE*q)/(q+c1o1);
- //q = q_dirE[k]; if (q>=zero && q<=one) (D27.f[DIR_M00 ])[kw ]=(two*feqW27_W -(f27_E *(q*omegaD-one)-omegaD*feq27_E *(q-one))/(omegaD-one)+f27_W *q)/(q+one);
- //q = q_dirW[k]; if (q>=zero && q<=one) (D27.f[DIR_P00 ])[ke ]=(two*feqW27_E -(f27_W *(q*omegaD-one)-omegaD*feq27_W *(q-one))/(omegaD-one)+f27_E *q)/(q+one);
- //q = q_dirN[k]; if (q>=zero && q<=one) (D27.f[DIR_0M0 ])[ks ]=(two*feqW27_S -(f27_N *(q*omegaD-one)-omegaD*feq27_N *(q-one))/(omegaD-one)+f27_S *q)/(q+one);
- //q = q_dirS[k]; if (q>=zero && q<=one) (D27.f[DIR_0P0 ])[kn ]=(two*feqW27_N -(f27_S *(q*omegaD-one)-omegaD*feq27_S *(q-one))/(omegaD-one)+f27_N *q)/(q+one);
- //q = q_dirT[k]; if (q>=zero && q<=one) (D27.f[DIR_00M ])[kb ]=(two*feqW27_B -(f27_T *(q*omegaD-one)-omegaD*feq27_T *(q-one))/(omegaD-one)+f27_B *q)/(q+one);
- //q = q_dirB[k]; if (q>=zero && q<=one) (D27.f[DIR_00P ])[kt ]=(two*feqW27_T -(f27_B *(q*omegaD-one)-omegaD*feq27_B *(q-one))/(omegaD-one)+f27_T *q)/(q+one);
- //q = q_dirNE[k]; if (q>=zero && q<=one) (D27.f[DIR_MM0 ])[ksw ]=(two*feqW27_SW -(f27_NE *(q*omegaD-one)-omegaD*feq27_NE *(q-one))/(omegaD-one)+f27_SW *q)/(q+one);
- //q = q_dirSW[k]; if (q>=zero && q<=one) (D27.f[DIR_PP0 ])[kne ]=(two*feqW27_NE -(f27_SW *(q*omegaD-one)-omegaD*feq27_SW *(q-one))/(omegaD-one)+f27_NE *q)/(q+one);
- //q = q_dirSE[k]; if (q>=zero && q<=one) (D27.f[DIR_MP0 ])[knw ]=(two*feqW27_NW -(f27_SE *(q*omegaD-one)-omegaD*feq27_SE *(q-one))/(omegaD-one)+f27_NW *q)/(q+one);
- //q = q_dirNW[k]; if (q>=zero && q<=one) (D27.f[DIR_PM0 ])[kse ]=(two*feqW27_SE -(f27_NW *(q*omegaD-one)-omegaD*feq27_NW *(q-one))/(omegaD-one)+f27_SE *q)/(q+one);
- //q = q_dirTE[k]; if (q>=zero && q<=one) (D27.f[DIR_M0M ])[kbw ]=(two*feqW27_BW -(f27_TE *(q*omegaD-one)-omegaD*feq27_TE *(q-one))/(omegaD-one)+f27_BW *q)/(q+one);
- //q = q_dirBW[k]; if (q>=zero && q<=one) (D27.f[DIR_P0P ])[kte ]=(two*feqW27_TE -(f27_BW *(q*omegaD-one)-omegaD*feq27_BW *(q-one))/(omegaD-one)+f27_TE *q)/(q+one);
- //q = q_dirBE[k]; if (q>=zero && q<=one) (D27.f[DIR_M0P ])[ktw ]=(two*feqW27_TW -(f27_BE *(q*omegaD-one)-omegaD*feq27_BE *(q-one))/(omegaD-one)+f27_TW *q)/(q+one);
- //q = q_dirTW[k]; if (q>=zero && q<=one) (D27.f[DIR_P0M ])[kbe ]=(two*feqW27_BE -(f27_TW *(q*omegaD-one)-omegaD*feq27_TW *(q-one))/(omegaD-one)+f27_BE *q)/(q+one);
- //q = q_dirTN[k]; if (q>=zero && q<=one) (D27.f[DIR_0MM ])[kbs ]=(two*feqW27_BS -(f27_TN *(q*omegaD-one)-omegaD*feq27_TN *(q-one))/(omegaD-one)+f27_BS *q)/(q+one);
- //q = q_dirBS[k]; if (q>=zero && q<=one) (D27.f[DIR_0PP ])[ktn ]=(two*feqW27_TN -(f27_BS *(q*omegaD-one)-omegaD*feq27_BS *(q-one))/(omegaD-one)+f27_TN *q)/(q+one);
- //q = q_dirBN[k]; if (q>=zero && q<=one) (D27.f[DIR_0MP ])[kts ]=(two*feqW27_TS -(f27_BN *(q*omegaD-one)-omegaD*feq27_BN *(q-one))/(omegaD-one)+f27_TS *q)/(q+one);
- //q = q_dirTS[k]; if (q>=zero && q<=one) (D27.f[DIR_0PM ])[kbn ]=(two*feqW27_BN -(f27_TS *(q*omegaD-one)-omegaD*feq27_TS *(q-one))/(omegaD-one)+f27_BN *q)/(q+one);
+ //q = q_dirE[k]; if (q>=zero && q<=one) (D27.f[DIR_M00])[kw ]=(two*feqW27_W -(f27_E *(q*omegaD-one)-omegaD*feq27_E *(q-one))/(omegaD-one)+f27_W *q)/(q+one);
+ //q = q_dirW[k]; if (q>=zero && q<=one) (D27.f[DIR_P00])[ke ]=(two*feqW27_E -(f27_W *(q*omegaD-one)-omegaD*feq27_W *(q-one))/(omegaD-one)+f27_E *q)/(q+one);
+ //q = q_dirN[k]; if (q>=zero && q<=one) (D27.f[DIR_0M0])[ks ]=(two*feqW27_S -(f27_N *(q*omegaD-one)-omegaD*feq27_N *(q-one))/(omegaD-one)+f27_S *q)/(q+one);
+ //q = q_dirS[k]; if (q>=zero && q<=one) (D27.f[DIR_0P0])[kn ]=(two*feqW27_N -(f27_S *(q*omegaD-one)-omegaD*feq27_S *(q-one))/(omegaD-one)+f27_N *q)/(q+one);
+ //q = q_dirT[k]; if (q>=zero && q<=one) (D27.f[DIR_00M])[kb ]=(two*feqW27_B -(f27_T *(q*omegaD-one)-omegaD*feq27_T *(q-one))/(omegaD-one)+f27_B *q)/(q+one);
+ //q = q_dirB[k]; if (q>=zero && q<=one) (D27.f[DIR_00P])[kt ]=(two*feqW27_T -(f27_B *(q*omegaD-one)-omegaD*feq27_B *(q-one))/(omegaD-one)+f27_T *q)/(q+one);
+ //q = q_dirNE[k]; if (q>=zero && q<=one) (D27.f[DIR_MM0])[ksw ]=(two*feqW27_SW -(f27_NE *(q*omegaD-one)-omegaD*feq27_NE *(q-one))/(omegaD-one)+f27_SW *q)/(q+one);
+ //q = q_dirSW[k]; if (q>=zero && q<=one) (D27.f[DIR_PP0])[kne ]=(two*feqW27_NE -(f27_SW *(q*omegaD-one)-omegaD*feq27_SW *(q-one))/(omegaD-one)+f27_NE *q)/(q+one);
+ //q = q_dirSE[k]; if (q>=zero && q<=one) (D27.f[DIR_MP0])[knw ]=(two*feqW27_NW -(f27_SE *(q*omegaD-one)-omegaD*feq27_SE *(q-one))/(omegaD-one)+f27_NW *q)/(q+one);
+ //q = q_dirNW[k]; if (q>=zero && q<=one) (D27.f[DIR_PM0])[kse ]=(two*feqW27_SE -(f27_NW *(q*omegaD-one)-omegaD*feq27_NW *(q-one))/(omegaD-one)+f27_SE *q)/(q+one);
+ //q = q_dirTE[k]; if (q>=zero && q<=one) (D27.f[DIR_M0M])[kbw ]=(two*feqW27_BW -(f27_TE *(q*omegaD-one)-omegaD*feq27_TE *(q-one))/(omegaD-one)+f27_BW *q)/(q+one);
+ //q = q_dirBW[k]; if (q>=zero && q<=one) (D27.f[DIR_P0P])[kte ]=(two*feqW27_TE -(f27_BW *(q*omegaD-one)-omegaD*feq27_BW *(q-one))/(omegaD-one)+f27_TE *q)/(q+one);
+ //q = q_dirBE[k]; if (q>=zero && q<=one) (D27.f[DIR_M0P])[ktw ]=(two*feqW27_TW -(f27_BE *(q*omegaD-one)-omegaD*feq27_BE *(q-one))/(omegaD-one)+f27_TW *q)/(q+one);
+ //q = q_dirTW[k]; if (q>=zero && q<=one) (D27.f[DIR_P0M])[kbe ]=(two*feqW27_BE -(f27_TW *(q*omegaD-one)-omegaD*feq27_TW *(q-one))/(omegaD-one)+f27_BE *q)/(q+one);
+ //q = q_dirTN[k]; if (q>=zero && q<=one) (D27.f[DIR_0MM])[kbs ]=(two*feqW27_BS -(f27_TN *(q*omegaD-one)-omegaD*feq27_TN *(q-one))/(omegaD-one)+f27_BS *q)/(q+one);
+ //q = q_dirBS[k]; if (q>=zero && q<=one) (D27.f[DIR_0PP])[ktn ]=(two*feqW27_TN -(f27_BS *(q*omegaD-one)-omegaD*feq27_BS *(q-one))/(omegaD-one)+f27_TN *q)/(q+one);
+ //q = q_dirBN[k]; if (q>=zero && q<=one) (D27.f[DIR_0MP])[kts ]=(two*feqW27_TS -(f27_BN *(q*omegaD-one)-omegaD*feq27_BN *(q-one))/(omegaD-one)+f27_TS *q)/(q+one);
+ //q = q_dirTS[k]; if (q>=zero && q<=one) (D27.f[DIR_0PM])[kbn ]=(two*feqW27_BN -(f27_TS *(q*omegaD-one)-omegaD*feq27_TS *(q-one))/(omegaD-one)+f27_BN *q)/(q+one);
//q = q_dirTNE[k]; if (q>=zero && q<=one) (D27.f[DIR_MMM])[kbsw]=(two*feqW27_BSW-(f27_TNE*(q*omegaD-one)-omegaD*feq27_TNE*(q-one))/(omegaD-one)+f27_BSW*q)/(q+one);
//q = q_dirBSW[k]; if (q>=zero && q<=one) (D27.f[DIR_PPP])[ktne]=(two*feqW27_TNE-(f27_BSW*(q*omegaD-one)-omegaD*feq27_BSW*(q-one))/(omegaD-one)+f27_TNE*q)/(q+one);
//q = q_dirBNE[k]; if (q>=zero && q<=one) (D27.f[DIR_MMP])[ktsw]=(two*feqW27_TSW-(f27_BNE*(q*omegaD-one)-omegaD*feq27_BNE*(q-one))/(omegaD-one)+f27_TSW*q)/(q+one);
@@ -3389,131 +3389,131 @@ __global__ void QADBB27( real* DD,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep)
{
//Distributions27 D;
//if (isEvenTimestep==true)
//{
- // D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- // D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- // D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- // D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- // D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- // D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- // D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- // D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- // D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- // D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- // D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- // D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- // D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- // D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- // D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- // D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- // D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- // D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- // D.f[DIR_000] = &DD[DIR_000*size_Mat];
- // D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- // D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- // D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- // D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- // D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- // D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- // D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- // D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
+ // D.f[DIR_P00] = &DD[DIR_P00 * size_Mat];
+ // D.f[DIR_M00] = &DD[DIR_M00 * size_Mat];
+ // D.f[DIR_0P0] = &DD[DIR_0P0 * size_Mat];
+ // D.f[DIR_0M0] = &DD[DIR_0M0 * size_Mat];
+ // D.f[DIR_00P] = &DD[DIR_00P * size_Mat];
+ // D.f[DIR_00M] = &DD[DIR_00M * size_Mat];
+ // D.f[DIR_PP0] = &DD[DIR_PP0 * size_Mat];
+ // D.f[DIR_MM0] = &DD[DIR_MM0 * size_Mat];
+ // D.f[DIR_PM0] = &DD[DIR_PM0 * size_Mat];
+ // D.f[DIR_MP0] = &DD[DIR_MP0 * size_Mat];
+ // D.f[DIR_P0P] = &DD[DIR_P0P * size_Mat];
+ // D.f[DIR_M0M] = &DD[DIR_M0M * size_Mat];
+ // D.f[DIR_P0M] = &DD[DIR_P0M * size_Mat];
+ // D.f[DIR_M0P] = &DD[DIR_M0P * size_Mat];
+ // D.f[DIR_0PP] = &DD[DIR_0PP * size_Mat];
+ // D.f[DIR_0MM] = &DD[DIR_0MM * size_Mat];
+ // D.f[DIR_0PM] = &DD[DIR_0PM * size_Mat];
+ // D.f[DIR_0MP] = &DD[DIR_0MP * size_Mat];
+ // D.f[DIR_000] = &DD[DIR_000 * size_Mat];
+ // D.f[DIR_PPP] = &DD[DIR_PPP * size_Mat];
+ // D.f[DIR_MMP] = &DD[DIR_MMP * size_Mat];
+ // D.f[DIR_PMP] = &DD[DIR_PMP * size_Mat];
+ // D.f[DIR_MPP] = &DD[DIR_MPP * size_Mat];
+ // D.f[DIR_PPM] = &DD[DIR_PPM * size_Mat];
+ // D.f[DIR_MMM] = &DD[DIR_MMM * size_Mat];
+ // D.f[DIR_PMM] = &DD[DIR_PMM * size_Mat];
+ // D.f[DIR_MPM] = &DD[DIR_MPM * size_Mat];
//}
//else
//{
- // D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
- // D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
- // D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
- // D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
- // D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
- // D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
- // D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
- // D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
- // D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
- // D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
- // D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
- // D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
- // D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
- // D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
- // D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
- // D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
- // D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
- // D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
- // D.f[DIR_000] = &DD[DIR_000*size_Mat];
- // D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
- // D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
- // D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
- // D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
- // D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
- // D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
- // D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
- // D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
+ // D.f[DIR_M00] = &DD[DIR_P00 * size_Mat];
+ // D.f[DIR_P00] = &DD[DIR_M00 * size_Mat];
+ // D.f[DIR_0M0] = &DD[DIR_0P0 * size_Mat];
+ // D.f[DIR_0P0] = &DD[DIR_0M0 * size_Mat];
+ // D.f[DIR_00M] = &DD[DIR_00P * size_Mat];
+ // D.f[DIR_00P] = &DD[DIR_00M * size_Mat];
+ // D.f[DIR_MM0] = &DD[DIR_PP0 * size_Mat];
+ // D.f[DIR_PP0] = &DD[DIR_MM0 * size_Mat];
+ // D.f[DIR_MP0] = &DD[DIR_PM0 * size_Mat];
+ // D.f[DIR_PM0] = &DD[DIR_MP0 * size_Mat];
+ // D.f[DIR_M0M] = &DD[DIR_P0P * size_Mat];
+ // D.f[DIR_P0P] = &DD[DIR_M0M * size_Mat];
+ // D.f[DIR_M0P] = &DD[DIR_P0M * size_Mat];
+ // D.f[DIR_P0M] = &DD[DIR_M0P * size_Mat];
+ // D.f[DIR_0MM] = &DD[DIR_0PP * size_Mat];
+ // D.f[DIR_0PP] = &DD[DIR_0MM * size_Mat];
+ // D.f[DIR_0MP] = &DD[DIR_0PM * size_Mat];
+ // D.f[DIR_0PM] = &DD[DIR_0MP * size_Mat];
+ // D.f[DIR_000] = &DD[DIR_000 * size_Mat];
+ // D.f[DIR_PPP] = &DD[DIR_MMM * size_Mat];
+ // D.f[DIR_MMP] = &DD[DIR_PPM * size_Mat];
+ // D.f[DIR_PMP] = &DD[DIR_MPM * size_Mat];
+ // D.f[DIR_MPP] = &DD[DIR_PMM * size_Mat];
+ // D.f[DIR_PPM] = &DD[DIR_MMP * size_Mat];
+ // D.f[DIR_MMM] = &DD[DIR_PPP * size_Mat];
+ // D.f[DIR_PMM] = &DD[DIR_MPP * size_Mat];
+ // D.f[DIR_MPM] = &DD[DIR_PMP * size_Mat];
//}
Distributions27 D27;
if (isEvenTimestep==true)
{
- D27.f[DIR_P00 ] = &DD27[DIR_P00 *size_Mat];
- D27.f[DIR_M00 ] = &DD27[DIR_M00 *size_Mat];
- D27.f[DIR_0P0 ] = &DD27[DIR_0P0 *size_Mat];
- D27.f[DIR_0M0 ] = &DD27[DIR_0M0 *size_Mat];
- D27.f[DIR_00P ] = &DD27[DIR_00P *size_Mat];
- D27.f[DIR_00M ] = &DD27[DIR_00M *size_Mat];
- D27.f[DIR_PP0 ] = &DD27[DIR_PP0 *size_Mat];
- D27.f[DIR_MM0 ] = &DD27[DIR_MM0 *size_Mat];
- D27.f[DIR_PM0 ] = &DD27[DIR_PM0 *size_Mat];
- D27.f[DIR_MP0 ] = &DD27[DIR_MP0 *size_Mat];
- D27.f[DIR_P0P ] = &DD27[DIR_P0P *size_Mat];
- D27.f[DIR_M0M ] = &DD27[DIR_M0M *size_Mat];
- D27.f[DIR_P0M ] = &DD27[DIR_P0M *size_Mat];
- D27.f[DIR_M0P ] = &DD27[DIR_M0P *size_Mat];
- D27.f[DIR_0PP ] = &DD27[DIR_0PP *size_Mat];
- D27.f[DIR_0MM ] = &DD27[DIR_0MM *size_Mat];
- D27.f[DIR_0PM ] = &DD27[DIR_0PM *size_Mat];
- D27.f[DIR_0MP ] = &DD27[DIR_0MP *size_Mat];
- D27.f[DIR_000] = &DD27[DIR_000*size_Mat];
- D27.f[DIR_PPP ] = &DD27[DIR_PPP *size_Mat];
- D27.f[DIR_MMP ] = &DD27[DIR_MMP *size_Mat];
- D27.f[DIR_PMP ] = &DD27[DIR_PMP *size_Mat];
- D27.f[DIR_MPP ] = &DD27[DIR_MPP *size_Mat];
- D27.f[DIR_PPM ] = &DD27[DIR_PPM *size_Mat];
- D27.f[DIR_MMM ] = &DD27[DIR_MMM *size_Mat];
- D27.f[DIR_PMM ] = &DD27[DIR_PMM *size_Mat];
- D27.f[DIR_MPM ] = &DD27[DIR_MPM *size_Mat];
+ D27.f[DIR_P00] = &DD27[DIR_P00 * numberOfLBnodes];
+ D27.f[DIR_M00] = &DD27[DIR_M00 * numberOfLBnodes];
+ D27.f[DIR_0P0] = &DD27[DIR_0P0 * numberOfLBnodes];
+ D27.f[DIR_0M0] = &DD27[DIR_0M0 * numberOfLBnodes];
+ D27.f[DIR_00P] = &DD27[DIR_00P * numberOfLBnodes];
+ D27.f[DIR_00M] = &DD27[DIR_00M * numberOfLBnodes];
+ D27.f[DIR_PP0] = &DD27[DIR_PP0 * numberOfLBnodes];
+ D27.f[DIR_MM0] = &DD27[DIR_MM0 * numberOfLBnodes];
+ D27.f[DIR_PM0] = &DD27[DIR_PM0 * numberOfLBnodes];
+ D27.f[DIR_MP0] = &DD27[DIR_MP0 * numberOfLBnodes];
+ D27.f[DIR_P0P] = &DD27[DIR_P0P * numberOfLBnodes];
+ D27.f[DIR_M0M] = &DD27[DIR_M0M * numberOfLBnodes];
+ D27.f[DIR_P0M] = &DD27[DIR_P0M * numberOfLBnodes];
+ D27.f[DIR_M0P] = &DD27[DIR_M0P * numberOfLBnodes];
+ D27.f[DIR_0PP] = &DD27[DIR_0PP * numberOfLBnodes];
+ D27.f[DIR_0MM] = &DD27[DIR_0MM * numberOfLBnodes];
+ D27.f[DIR_0PM] = &DD27[DIR_0PM * numberOfLBnodes];
+ D27.f[DIR_0MP] = &DD27[DIR_0MP * numberOfLBnodes];
+ D27.f[DIR_000] = &DD27[DIR_000 * numberOfLBnodes];
+ D27.f[DIR_PPP] = &DD27[DIR_PPP * numberOfLBnodes];
+ D27.f[DIR_MMP] = &DD27[DIR_MMP * numberOfLBnodes];
+ D27.f[DIR_PMP] = &DD27[DIR_PMP * numberOfLBnodes];
+ D27.f[DIR_MPP] = &DD27[DIR_MPP * numberOfLBnodes];
+ D27.f[DIR_PPM] = &DD27[DIR_PPM * numberOfLBnodes];
+ D27.f[DIR_MMM] = &DD27[DIR_MMM * numberOfLBnodes];
+ D27.f[DIR_PMM] = &DD27[DIR_PMM * numberOfLBnodes];
+ D27.f[DIR_MPM] = &DD27[DIR_MPM * numberOfLBnodes];
}
else
{
- D27.f[DIR_M00 ] = &DD27[DIR_P00 *size_Mat];
- D27.f[DIR_P00 ] = &DD27[DIR_M00 *size_Mat];
- D27.f[DIR_0M0 ] = &DD27[DIR_0P0 *size_Mat];
- D27.f[DIR_0P0 ] = &DD27[DIR_0M0 *size_Mat];
- D27.f[DIR_00M ] = &DD27[DIR_00P *size_Mat];
- D27.f[DIR_00P ] = &DD27[DIR_00M *size_Mat];
- D27.f[DIR_MM0 ] = &DD27[DIR_PP0 *size_Mat];
- D27.f[DIR_PP0 ] = &DD27[DIR_MM0 *size_Mat];
- D27.f[DIR_MP0 ] = &DD27[DIR_PM0 *size_Mat];
- D27.f[DIR_PM0 ] = &DD27[DIR_MP0 *size_Mat];
- D27.f[DIR_M0M ] = &DD27[DIR_P0P *size_Mat];
- D27.f[DIR_P0P ] = &DD27[DIR_M0M *size_Mat];
- D27.f[DIR_M0P ] = &DD27[DIR_P0M *size_Mat];
- D27.f[DIR_P0M ] = &DD27[DIR_M0P *size_Mat];
- D27.f[DIR_0MM ] = &DD27[DIR_0PP *size_Mat];
- D27.f[DIR_0PP ] = &DD27[DIR_0MM *size_Mat];
- D27.f[DIR_0MP ] = &DD27[DIR_0PM *size_Mat];
- D27.f[DIR_0PM ] = &DD27[DIR_0MP *size_Mat];
- D27.f[DIR_000] = &DD27[DIR_000*size_Mat];
- D27.f[DIR_PPP ] = &DD27[DIR_MMM *size_Mat];
- D27.f[DIR_MMP ] = &DD27[DIR_PPM *size_Mat];
- D27.f[DIR_PMP ] = &DD27[DIR_MPM *size_Mat];
- D27.f[DIR_MPP ] = &DD27[DIR_PMM *size_Mat];
- D27.f[DIR_PPM ] = &DD27[DIR_MMP *size_Mat];
- D27.f[DIR_MMM ] = &DD27[DIR_PPP *size_Mat];
- D27.f[DIR_PMM ] = &DD27[DIR_MPP *size_Mat];
- D27.f[DIR_MPM ] = &DD27[DIR_PMP *size_Mat];
+ D27.f[DIR_M00] = &DD27[DIR_P00 * numberOfLBnodes];
+ D27.f[DIR_P00] = &DD27[DIR_M00 * numberOfLBnodes];
+ D27.f[DIR_0M0] = &DD27[DIR_0P0 * numberOfLBnodes];
+ D27.f[DIR_0P0] = &DD27[DIR_0M0 * numberOfLBnodes];
+ D27.f[DIR_00M] = &DD27[DIR_00P * numberOfLBnodes];
+ D27.f[DIR_00P] = &DD27[DIR_00M * numberOfLBnodes];
+ D27.f[DIR_MM0] = &DD27[DIR_PP0 * numberOfLBnodes];
+ D27.f[DIR_PP0] = &DD27[DIR_MM0 * numberOfLBnodes];
+ D27.f[DIR_MP0] = &DD27[DIR_PM0 * numberOfLBnodes];
+ D27.f[DIR_PM0] = &DD27[DIR_MP0 * numberOfLBnodes];
+ D27.f[DIR_M0M] = &DD27[DIR_P0P * numberOfLBnodes];
+ D27.f[DIR_P0P] = &DD27[DIR_M0M * numberOfLBnodes];
+ D27.f[DIR_M0P] = &DD27[DIR_P0M * numberOfLBnodes];
+ D27.f[DIR_P0M] = &DD27[DIR_M0P * numberOfLBnodes];
+ D27.f[DIR_0MM] = &DD27[DIR_0PP * numberOfLBnodes];
+ D27.f[DIR_0PP] = &DD27[DIR_0MM * numberOfLBnodes];
+ D27.f[DIR_0MP] = &DD27[DIR_0PM * numberOfLBnodes];
+ D27.f[DIR_0PM] = &DD27[DIR_0MP * numberOfLBnodes];
+ D27.f[DIR_000] = &DD27[DIR_000 * numberOfLBnodes];
+ D27.f[DIR_PPP] = &DD27[DIR_MMM * numberOfLBnodes];
+ D27.f[DIR_MMP] = &DD27[DIR_PPM * numberOfLBnodes];
+ D27.f[DIR_PMP] = &DD27[DIR_MPM * numberOfLBnodes];
+ D27.f[DIR_MPP] = &DD27[DIR_PMM * numberOfLBnodes];
+ D27.f[DIR_PPM] = &DD27[DIR_MMP * numberOfLBnodes];
+ D27.f[DIR_MMM] = &DD27[DIR_PPP * numberOfLBnodes];
+ D27.f[DIR_PMM] = &DD27[DIR_MPP * numberOfLBnodes];
+ D27.f[DIR_MPM] = &DD27[DIR_PMP * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////
const unsigned x = threadIdx.x; // Globaler x-Index
@@ -3534,24 +3534,24 @@ __global__ void QADBB27( real* DD,
*q_dirBE, *q_dirTW, *q_dirTN, *q_dirBS, *q_dirBN, *q_dirTS,
*q_dirTNE, *q_dirTSW, *q_dirTSE, *q_dirTNW, *q_dirBNE, *q_dirBSW,
*q_dirBSE, *q_dirBNW;
- q_dirE = &QQ[DIR_P00 * numberOfBCnodes];
- q_dirW = &QQ[DIR_M00 * numberOfBCnodes];
- q_dirN = &QQ[DIR_0P0 * numberOfBCnodes];
- q_dirS = &QQ[DIR_0M0 * numberOfBCnodes];
- q_dirT = &QQ[DIR_00P * numberOfBCnodes];
- q_dirB = &QQ[DIR_00M * numberOfBCnodes];
- q_dirNE = &QQ[DIR_PP0 * numberOfBCnodes];
- q_dirSW = &QQ[DIR_MM0 * numberOfBCnodes];
- q_dirSE = &QQ[DIR_PM0 * numberOfBCnodes];
- q_dirNW = &QQ[DIR_MP0 * numberOfBCnodes];
- q_dirTE = &QQ[DIR_P0P * numberOfBCnodes];
- q_dirBW = &QQ[DIR_M0M * numberOfBCnodes];
- q_dirBE = &QQ[DIR_P0M * numberOfBCnodes];
- q_dirTW = &QQ[DIR_M0P * numberOfBCnodes];
- q_dirTN = &QQ[DIR_0PP * numberOfBCnodes];
- q_dirBS = &QQ[DIR_0MM * numberOfBCnodes];
- q_dirBN = &QQ[DIR_0PM * numberOfBCnodes];
- q_dirTS = &QQ[DIR_0MP * numberOfBCnodes];
+ q_dirE = &QQ[DIR_P00 * numberOfBCnodes];
+ q_dirW = &QQ[DIR_M00 * numberOfBCnodes];
+ q_dirN = &QQ[DIR_0P0 * numberOfBCnodes];
+ q_dirS = &QQ[DIR_0M0 * numberOfBCnodes];
+ q_dirT = &QQ[DIR_00P * numberOfBCnodes];
+ q_dirB = &QQ[DIR_00M * numberOfBCnodes];
+ q_dirNE = &QQ[DIR_PP0 * numberOfBCnodes];
+ q_dirSW = &QQ[DIR_MM0 * numberOfBCnodes];
+ q_dirSE = &QQ[DIR_PM0 * numberOfBCnodes];
+ q_dirNW = &QQ[DIR_MP0 * numberOfBCnodes];
+ q_dirTE = &QQ[DIR_P0P * numberOfBCnodes];
+ q_dirBW = &QQ[DIR_M0M * numberOfBCnodes];
+ q_dirBE = &QQ[DIR_P0M * numberOfBCnodes];
+ q_dirTW = &QQ[DIR_M0P * numberOfBCnodes];
+ q_dirTN = &QQ[DIR_0PP * numberOfBCnodes];
+ q_dirBS = &QQ[DIR_0MM * numberOfBCnodes];
+ q_dirBN = &QQ[DIR_0PM * numberOfBCnodes];
+ q_dirTS = &QQ[DIR_0MP * numberOfBCnodes];
q_dirTNE = &QQ[DIR_PPP * numberOfBCnodes];
q_dirTSW = &QQ[DIR_MMP * numberOfBCnodes];
q_dirTSE = &QQ[DIR_PMP * numberOfBCnodes];
@@ -3591,33 +3591,33 @@ __global__ void QADBB27( real* DD,
unsigned int ktne = KQK;
unsigned int kbsw = neighborZ[ksw];
////////////////////////////////////////////////////////////////////////////////
- //real f_W = (D.f[DIR_P00 ])[ke ];
- //real f_E = (D.f[DIR_M00 ])[kw ];
- //real f_S = (D.f[DIR_0P0 ])[kn ];
- //real f_N = (D.f[DIR_0M0 ])[ks ];
- //real f_B = (D.f[DIR_00P ])[kt ];
- //real f_T = (D.f[DIR_00M ])[kb ];
- //real f_SW = (D.f[DIR_PP0 ])[kne ];
- //real f_NE = (D.f[DIR_MM0 ])[ksw ];
- //real f_NW = (D.f[DIR_PM0 ])[kse ];
- //real f_SE = (D.f[DIR_MP0 ])[knw ];
- //real f_BW = (D.f[DIR_P0P ])[kte ];
- //real f_TE = (D.f[DIR_M0M ])[kbw ];
- //real f_TW = (D.f[DIR_P0M ])[kbe ];
- //real f_BE = (D.f[DIR_M0P ])[ktw ];
- //real f_BS = (D.f[DIR_0PP ])[ktn ];
- //real f_TN = (D.f[DIR_0MM ])[kbs ];
- //real f_TS = (D.f[DIR_0PM ])[kbn ];
- //real f_BN = (D.f[DIR_0MP ])[kts ];
+ //real f_W = (D.f[DIR_P00])[ke ];
+ //real f_E = (D.f[DIR_M00])[kw ];
+ //real f_S = (D.f[DIR_0P0])[kn ];
+ //real f_N = (D.f[DIR_0M0])[ks ];
+ //real f_B = (D.f[DIR_00P])[kt ];
+ //real f_T = (D.f[DIR_00M])[kb ];
+ //real f_SW = (D.f[DIR_PP0])[kne ];
+ //real f_NE = (D.f[DIR_MM0])[ksw ];
+ //real f_NW = (D.f[DIR_PM0])[kse ];
+ //real f_SE = (D.f[DIR_MP0])[knw ];
+ //real f_BW = (D.f[DIR_P0P])[kte ];
+ //real f_TE = (D.f[DIR_M0M])[kbw ];
+ //real f_TW = (D.f[DIR_P0M])[kbe ];
+ //real f_BE = (D.f[DIR_M0P])[ktw ];
+ //real f_BS = (D.f[DIR_0PP])[ktn ];
+ //real f_TN = (D.f[DIR_0MM])[kbs ];
+ //real f_TS = (D.f[DIR_0PM])[kbn ];
+ //real f_BN = (D.f[DIR_0MP])[kts ];
//real f_ZERO = (D.f[DIR_000])[kzero];
- //real f_BSW = (D.f[DIR_PPP ])[ktne ];
- //real f_BNE = (D.f[DIR_MMP ])[ktsw ];
- //real f_BNW = (D.f[DIR_PMP ])[ktse ];
- //real f_BSE = (D.f[DIR_MPP ])[ktnw ];
- //real f_TSW = (D.f[DIR_PPM ])[kbne ];
- //real f_TNE = (D.f[DIR_MMM ])[kbsw ];
- //real f_TNW = (D.f[DIR_PMM ])[kbse ];
- //real f_TSE = (D.f[DIR_MPM ])[kbnw ];
+ //real f_BSW = (D.f[DIR_PPP])[ktne ];
+ //real f_BNE = (D.f[DIR_MMP])[ktsw ];
+ //real f_BNW = (D.f[DIR_PMP])[ktse ];
+ //real f_BSE = (D.f[DIR_MPP])[ktnw ];
+ //real f_TSW = (D.f[DIR_PPM])[kbne ];
+ //real f_TNE = (D.f[DIR_MMM])[kbsw ];
+ //real f_TNW = (D.f[DIR_PMM])[kbse ];
+ //real f_TSE = (D.f[DIR_MPM])[kbnw ];
////////////////////////////////////////////////////////////////////////////////
//real vx1, vx2, vx3, /*drho, feq,*/ q;
real q;
@@ -3644,33 +3644,33 @@ __global__ void QADBB27( real* DD,
//vx2 = OORho*((f_TNE-f_BSW)+(f_BNE-f_TSW)+(f_BNW-f_TSE)+(f_TNW-f_BSE) +(f_NE-f_SW)+(f_NW-f_SE)+(f_TN-f_BS)+(f_BN-f_TS)+(f_N-f_S));
//vx3 = OORho*((f_TNE-f_BSW)+(f_TSW-f_BNE)+(f_TSE-f_BNW)+(f_TNW-f_BSE) +(f_TE-f_BW)+(f_TW-f_BE)+(f_TN-f_BS)+(f_TS-f_BN)+(f_T-f_B));
////////////////////////////////////////////////////////////////////////////////
- real f27_W = (D27.f[DIR_P00 ])[ke ];
- real f27_E = (D27.f[DIR_M00 ])[kw ];
- real f27_S = (D27.f[DIR_0P0 ])[kn ];
- real f27_N = (D27.f[DIR_0M0 ])[ks ];
- real f27_B = (D27.f[DIR_00P ])[kt ];
- real f27_T = (D27.f[DIR_00M ])[kb ];
- real f27_SW = (D27.f[DIR_PP0 ])[kne ];
- real f27_NE = (D27.f[DIR_MM0 ])[ksw ];
- real f27_NW = (D27.f[DIR_PM0 ])[kse ];
- real f27_SE = (D27.f[DIR_MP0 ])[knw ];
- real f27_BW = (D27.f[DIR_P0P ])[kte ];
- real f27_TE = (D27.f[DIR_M0M ])[kbw ];
- real f27_TW = (D27.f[DIR_P0M ])[kbe ];
- real f27_BE = (D27.f[DIR_M0P ])[ktw ];
- real f27_BS = (D27.f[DIR_0PP ])[ktn ];
- real f27_TN = (D27.f[DIR_0MM ])[kbs ];
- real f27_TS = (D27.f[DIR_0PM ])[kbn ];
- real f27_BN = (D27.f[DIR_0MP ])[kts ];
+ real f27_W = (D27.f[DIR_P00])[ke ];
+ real f27_E = (D27.f[DIR_M00])[kw ];
+ real f27_S = (D27.f[DIR_0P0])[kn ];
+ real f27_N = (D27.f[DIR_0M0])[ks ];
+ real f27_B = (D27.f[DIR_00P])[kt ];
+ real f27_T = (D27.f[DIR_00M])[kb ];
+ real f27_SW = (D27.f[DIR_PP0])[kne ];
+ real f27_NE = (D27.f[DIR_MM0])[ksw ];
+ real f27_NW = (D27.f[DIR_PM0])[kse ];
+ real f27_SE = (D27.f[DIR_MP0])[knw ];
+ real f27_BW = (D27.f[DIR_P0P])[kte ];
+ real f27_TE = (D27.f[DIR_M0M])[kbw ];
+ real f27_TW = (D27.f[DIR_P0M])[kbe ];
+ real f27_BE = (D27.f[DIR_M0P])[ktw ];
+ real f27_BS = (D27.f[DIR_0PP])[ktn ];
+ real f27_TN = (D27.f[DIR_0MM])[kbs ];
+ real f27_TS = (D27.f[DIR_0PM])[kbn ];
+ real f27_BN = (D27.f[DIR_0MP])[kts ];
//real f27_ZERO = (D27.f[DIR_000])[kzero];
- real f27_BSW = (D27.f[DIR_PPP ])[ktne ];
- real f27_BNE = (D27.f[DIR_MMP ])[ktsw ];
- real f27_BNW = (D27.f[DIR_PMP ])[ktse ];
- real f27_BSE = (D27.f[DIR_MPP ])[ktnw ];
- real f27_TSW = (D27.f[DIR_PPM ])[kbne ];
- real f27_TNE = (D27.f[DIR_MMM ])[kbsw ];
- real f27_TNW = (D27.f[DIR_PMM ])[kbse ];
- real f27_TSE = (D27.f[DIR_MPM ])[kbnw ];
+ real f27_BSW = (D27.f[DIR_PPP])[ktne ];
+ real f27_BNE = (D27.f[DIR_MMP])[ktsw ];
+ real f27_BNW = (D27.f[DIR_PMP])[ktse ];
+ real f27_BSE = (D27.f[DIR_MPP])[ktnw ];
+ real f27_TSW = (D27.f[DIR_PPM])[kbne ];
+ real f27_TNE = (D27.f[DIR_MMM])[kbsw ];
+ real f27_TNW = (D27.f[DIR_PMM])[kbse ];
+ real f27_TSE = (D27.f[DIR_MPM])[kbnw ];
////////////////////////////////////////////////////////////////////////////////
//real cu_sq=c3o2*(vx1*vx1+vx2*vx2+vx3*vx3);
////////////////////////////////////////////////////////////////////////////////
@@ -3746,86 +3746,86 @@ __global__ void QADBB27( real* DD,
//////////////////////////////////////////////////////////////////////////
if (isEvenTimestep==false)
{
- D27.f[DIR_P00 ] = &DD27[DIR_P00 *size_Mat];
- D27.f[DIR_M00 ] = &DD27[DIR_M00 *size_Mat];
- D27.f[DIR_0P0 ] = &DD27[DIR_0P0 *size_Mat];
- D27.f[DIR_0M0 ] = &DD27[DIR_0M0 *size_Mat];
- D27.f[DIR_00P ] = &DD27[DIR_00P *size_Mat];
- D27.f[DIR_00M ] = &DD27[DIR_00M *size_Mat];
- D27.f[DIR_PP0 ] = &DD27[DIR_PP0 *size_Mat];
- D27.f[DIR_MM0 ] = &DD27[DIR_MM0 *size_Mat];
- D27.f[DIR_PM0 ] = &DD27[DIR_PM0 *size_Mat];
- D27.f[DIR_MP0 ] = &DD27[DIR_MP0 *size_Mat];
- D27.f[DIR_P0P ] = &DD27[DIR_P0P *size_Mat];
- D27.f[DIR_M0M ] = &DD27[DIR_M0M *size_Mat];
- D27.f[DIR_P0M ] = &DD27[DIR_P0M *size_Mat];
- D27.f[DIR_M0P ] = &DD27[DIR_M0P *size_Mat];
- D27.f[DIR_0PP ] = &DD27[DIR_0PP *size_Mat];
- D27.f[DIR_0MM ] = &DD27[DIR_0MM *size_Mat];
- D27.f[DIR_0PM ] = &DD27[DIR_0PM *size_Mat];
- D27.f[DIR_0MP ] = &DD27[DIR_0MP *size_Mat];
- D27.f[DIR_000] = &DD27[DIR_000*size_Mat];
- D27.f[DIR_PPP ] = &DD27[DIR_PPP *size_Mat];
- D27.f[DIR_MMP ] = &DD27[DIR_MMP *size_Mat];
- D27.f[DIR_PMP ] = &DD27[DIR_PMP *size_Mat];
- D27.f[DIR_MPP ] = &DD27[DIR_MPP *size_Mat];
- D27.f[DIR_PPM ] = &DD27[DIR_PPM *size_Mat];
- D27.f[DIR_MMM ] = &DD27[DIR_MMM *size_Mat];
- D27.f[DIR_PMM ] = &DD27[DIR_PMM *size_Mat];
- D27.f[DIR_MPM ] = &DD27[DIR_MPM *size_Mat];
+ D27.f[DIR_P00] = &DD27[DIR_P00 * numberOfLBnodes];
+ D27.f[DIR_M00] = &DD27[DIR_M00 * numberOfLBnodes];
+ D27.f[DIR_0P0] = &DD27[DIR_0P0 * numberOfLBnodes];
+ D27.f[DIR_0M0] = &DD27[DIR_0M0 * numberOfLBnodes];
+ D27.f[DIR_00P] = &DD27[DIR_00P * numberOfLBnodes];
+ D27.f[DIR_00M] = &DD27[DIR_00M * numberOfLBnodes];
+ D27.f[DIR_PP0] = &DD27[DIR_PP0 * numberOfLBnodes];
+ D27.f[DIR_MM0] = &DD27[DIR_MM0 * numberOfLBnodes];
+ D27.f[DIR_PM0] = &DD27[DIR_PM0 * numberOfLBnodes];
+ D27.f[DIR_MP0] = &DD27[DIR_MP0 * numberOfLBnodes];
+ D27.f[DIR_P0P] = &DD27[DIR_P0P * numberOfLBnodes];
+ D27.f[DIR_M0M] = &DD27[DIR_M0M * numberOfLBnodes];
+ D27.f[DIR_P0M] = &DD27[DIR_P0M * numberOfLBnodes];
+ D27.f[DIR_M0P] = &DD27[DIR_M0P * numberOfLBnodes];
+ D27.f[DIR_0PP] = &DD27[DIR_0PP * numberOfLBnodes];
+ D27.f[DIR_0MM] = &DD27[DIR_0MM * numberOfLBnodes];
+ D27.f[DIR_0PM] = &DD27[DIR_0PM * numberOfLBnodes];
+ D27.f[DIR_0MP] = &DD27[DIR_0MP * numberOfLBnodes];
+ D27.f[DIR_000] = &DD27[DIR_000 * numberOfLBnodes];
+ D27.f[DIR_PPP] = &DD27[DIR_PPP * numberOfLBnodes];
+ D27.f[DIR_MMP] = &DD27[DIR_MMP * numberOfLBnodes];
+ D27.f[DIR_PMP] = &DD27[DIR_PMP * numberOfLBnodes];
+ D27.f[DIR_MPP] = &DD27[DIR_MPP * numberOfLBnodes];
+ D27.f[DIR_PPM] = &DD27[DIR_PPM * numberOfLBnodes];
+ D27.f[DIR_MMM] = &DD27[DIR_MMM * numberOfLBnodes];
+ D27.f[DIR_PMM] = &DD27[DIR_PMM * numberOfLBnodes];
+ D27.f[DIR_MPM] = &DD27[DIR_MPM * numberOfLBnodes];
}
else
{
- D27.f[DIR_M00 ] = &DD27[DIR_P00 *size_Mat];
- D27.f[DIR_P00 ] = &DD27[DIR_M00 *size_Mat];
- D27.f[DIR_0M0 ] = &DD27[DIR_0P0 *size_Mat];
- D27.f[DIR_0P0 ] = &DD27[DIR_0M0 *size_Mat];
- D27.f[DIR_00M ] = &DD27[DIR_00P *size_Mat];
- D27.f[DIR_00P ] = &DD27[DIR_00M *size_Mat];
- D27.f[DIR_MM0 ] = &DD27[DIR_PP0 *size_Mat];
- D27.f[DIR_PP0 ] = &DD27[DIR_MM0 *size_Mat];
- D27.f[DIR_MP0 ] = &DD27[DIR_PM0 *size_Mat];
- D27.f[DIR_PM0 ] = &DD27[DIR_MP0 *size_Mat];
- D27.f[DIR_M0M ] = &DD27[DIR_P0P *size_Mat];
- D27.f[DIR_P0P ] = &DD27[DIR_M0M *size_Mat];
- D27.f[DIR_M0P ] = &DD27[DIR_P0M *size_Mat];
- D27.f[DIR_P0M ] = &DD27[DIR_M0P *size_Mat];
- D27.f[DIR_0MM ] = &DD27[DIR_0PP *size_Mat];
- D27.f[DIR_0PP ] = &DD27[DIR_0MM *size_Mat];
- D27.f[DIR_0MP ] = &DD27[DIR_0PM *size_Mat];
- D27.f[DIR_0PM ] = &DD27[DIR_0MP *size_Mat];
- D27.f[DIR_000] = &DD27[DIR_000*size_Mat];
- D27.f[DIR_PPP ] = &DD27[DIR_MMM *size_Mat];
- D27.f[DIR_MMP ] = &DD27[DIR_PPM *size_Mat];
- D27.f[DIR_PMP ] = &DD27[DIR_MPM *size_Mat];
- D27.f[DIR_MPP ] = &DD27[DIR_PMM *size_Mat];
- D27.f[DIR_PPM ] = &DD27[DIR_MMP *size_Mat];
- D27.f[DIR_MMM ] = &DD27[DIR_PPP *size_Mat];
- D27.f[DIR_PMM ] = &DD27[DIR_MPP *size_Mat];
- D27.f[DIR_MPM ] = &DD27[DIR_PMP *size_Mat];
+ D27.f[DIR_M00] = &DD27[DIR_P00 * numberOfLBnodes];
+ D27.f[DIR_P00] = &DD27[DIR_M00 * numberOfLBnodes];
+ D27.f[DIR_0M0] = &DD27[DIR_0P0 * numberOfLBnodes];
+ D27.f[DIR_0P0] = &DD27[DIR_0M0 * numberOfLBnodes];
+ D27.f[DIR_00M] = &DD27[DIR_00P * numberOfLBnodes];
+ D27.f[DIR_00P] = &DD27[DIR_00M * numberOfLBnodes];
+ D27.f[DIR_MM0] = &DD27[DIR_PP0 * numberOfLBnodes];
+ D27.f[DIR_PP0] = &DD27[DIR_MM0 * numberOfLBnodes];
+ D27.f[DIR_MP0] = &DD27[DIR_PM0 * numberOfLBnodes];
+ D27.f[DIR_PM0] = &DD27[DIR_MP0 * numberOfLBnodes];
+ D27.f[DIR_M0M] = &DD27[DIR_P0P * numberOfLBnodes];
+ D27.f[DIR_P0P] = &DD27[DIR_M0M * numberOfLBnodes];
+ D27.f[DIR_M0P] = &DD27[DIR_P0M * numberOfLBnodes];
+ D27.f[DIR_P0M] = &DD27[DIR_M0P * numberOfLBnodes];
+ D27.f[DIR_0MM] = &DD27[DIR_0PP * numberOfLBnodes];
+ D27.f[DIR_0PP] = &DD27[DIR_0MM * numberOfLBnodes];
+ D27.f[DIR_0MP] = &DD27[DIR_0PM * numberOfLBnodes];
+ D27.f[DIR_0PM] = &DD27[DIR_0MP * numberOfLBnodes];
+ D27.f[DIR_000] = &DD27[DIR_000 * numberOfLBnodes];
+ D27.f[DIR_PPP] = &DD27[DIR_MMM * numberOfLBnodes];
+ D27.f[DIR_MMP] = &DD27[DIR_PPM * numberOfLBnodes];
+ D27.f[DIR_PMP] = &DD27[DIR_MPM * numberOfLBnodes];
+ D27.f[DIR_MPP] = &DD27[DIR_PMM * numberOfLBnodes];
+ D27.f[DIR_PPM] = &DD27[DIR_MMP * numberOfLBnodes];
+ D27.f[DIR_MMM] = &DD27[DIR_PPP * numberOfLBnodes];
+ D27.f[DIR_PMM] = &DD27[DIR_MPP * numberOfLBnodes];
+ D27.f[DIR_MPM] = &DD27[DIR_PMP * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
//Test
//(D.f[DIR_000])[k]=0.1f;
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
- q = q_dirE[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_M00 ])[kw ]=f27_E ;
- q = q_dirW[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_P00 ])[ke ]=f27_W ;
- q = q_dirN[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_0M0 ])[ks ]=f27_N ;
- q = q_dirS[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_0P0 ])[kn ]=f27_S ;
- q = q_dirT[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_00M ])[kb ]=f27_T ;
- q = q_dirB[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_00P ])[kt ]=f27_B ;
- q = q_dirNE[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_MM0 ])[ksw ]=f27_NE ;
- q = q_dirSW[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_PP0 ])[kne ]=f27_SW ;
- q = q_dirSE[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_MP0 ])[knw ]=f27_SE ;
- q = q_dirNW[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_PM0 ])[kse ]=f27_NW ;
- q = q_dirTE[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_M0M ])[kbw ]=f27_TE ;
- q = q_dirBW[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_P0P ])[kte ]=f27_BW ;
- q = q_dirBE[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_M0P ])[ktw ]=f27_BE ;
- q = q_dirTW[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_P0M ])[kbe ]=f27_TW ;
- q = q_dirTN[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_0MM ])[kbs ]=f27_TN ;
- q = q_dirBS[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_0PP ])[ktn ]=f27_BS ;
- q = q_dirBN[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_0MP ])[kts ]=f27_BN ;
- q = q_dirTS[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_0PM ])[kbn ]=f27_TS ;
+ q = q_dirE[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_M00])[kw ]=f27_E ;
+ q = q_dirW[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_P00])[ke ]=f27_W ;
+ q = q_dirN[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_0M0])[ks ]=f27_N ;
+ q = q_dirS[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_0P0])[kn ]=f27_S ;
+ q = q_dirT[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_00M])[kb ]=f27_T ;
+ q = q_dirB[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_00P])[kt ]=f27_B ;
+ q = q_dirNE[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_MM0])[ksw ]=f27_NE ;
+ q = q_dirSW[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_PP0])[kne ]=f27_SW ;
+ q = q_dirSE[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_MP0])[knw ]=f27_SE ;
+ q = q_dirNW[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_PM0])[kse ]=f27_NW ;
+ q = q_dirTE[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_M0M])[kbw ]=f27_TE ;
+ q = q_dirBW[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_P0P])[kte ]=f27_BW ;
+ q = q_dirBE[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_M0P])[ktw ]=f27_BE ;
+ q = q_dirTW[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_P0M])[kbe ]=f27_TW ;
+ q = q_dirTN[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_0MM])[kbs ]=f27_TN ;
+ q = q_dirBS[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_0PP])[ktn ]=f27_BS ;
+ q = q_dirBN[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_0MP])[kts ]=f27_BN ;
+ q = q_dirTS[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_0PM])[kbn ]=f27_TS ;
q = q_dirTNE[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_MMM])[kbsw]=f27_TNE;
q = q_dirBSW[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_PPP])[ktne]=f27_BSW;
q = q_dirBNE[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_MMP])[ktsw]=f27_BNE;
@@ -3905,91 +3905,91 @@ __global__ void QNoSlipADincomp7(
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep)
{
//Distributions27 D;
//if (isEvenTimestep==true)
//{
- // D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- // D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- // D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- // D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- // D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- // D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- // D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- // D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- // D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- // D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- // D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- // D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- // D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- // D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- // D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- // D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- // D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- // D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- // D.f[DIR_000] = &DD[DIR_000*size_Mat];
- // D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- // D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- // D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- // D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- // D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- // D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- // D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- // D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
+ // D.f[DIR_P00] = &DD[DIR_P00 * size_Mat];
+ // D.f[DIR_M00] = &DD[DIR_M00 * size_Mat];
+ // D.f[DIR_0P0] = &DD[DIR_0P0 * size_Mat];
+ // D.f[DIR_0M0] = &DD[DIR_0M0 * size_Mat];
+ // D.f[DIR_00P] = &DD[DIR_00P * size_Mat];
+ // D.f[DIR_00M] = &DD[DIR_00M * size_Mat];
+ // D.f[DIR_PP0] = &DD[DIR_PP0 * size_Mat];
+ // D.f[DIR_MM0] = &DD[DIR_MM0 * size_Mat];
+ // D.f[DIR_PM0] = &DD[DIR_PM0 * size_Mat];
+ // D.f[DIR_MP0] = &DD[DIR_MP0 * size_Mat];
+ // D.f[DIR_P0P] = &DD[DIR_P0P * size_Mat];
+ // D.f[DIR_M0M] = &DD[DIR_M0M * size_Mat];
+ // D.f[DIR_P0M] = &DD[DIR_P0M * size_Mat];
+ // D.f[DIR_M0P] = &DD[DIR_M0P * size_Mat];
+ // D.f[DIR_0PP] = &DD[DIR_0PP * size_Mat];
+ // D.f[DIR_0MM] = &DD[DIR_0MM * size_Mat];
+ // D.f[DIR_0PM] = &DD[DIR_0PM * size_Mat];
+ // D.f[DIR_0MP] = &DD[DIR_0MP * size_Mat];
+ // D.f[DIR_000] = &DD[DIR_000 * size_Mat];
+ // D.f[DIR_PPP] = &DD[DIR_PPP * size_Mat];
+ // D.f[DIR_MMP] = &DD[DIR_MMP * size_Mat];
+ // D.f[DIR_PMP] = &DD[DIR_PMP * size_Mat];
+ // D.f[DIR_MPP] = &DD[DIR_MPP * size_Mat];
+ // D.f[DIR_PPM] = &DD[DIR_PPM * size_Mat];
+ // D.f[DIR_MMM] = &DD[DIR_MMM * size_Mat];
+ // D.f[DIR_PMM] = &DD[DIR_PMM * size_Mat];
+ // D.f[DIR_MPM] = &DD[DIR_MPM * size_Mat];
//}
//else
//{
- // D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
- // D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
- // D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
- // D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
- // D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
- // D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
- // D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
- // D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
- // D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
- // D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
- // D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
- // D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
- // D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
- // D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
- // D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
- // D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
- // D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
- // D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
- // D.f[DIR_000] = &DD[DIR_000*size_Mat];
- // D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
- // D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
- // D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
- // D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
- // D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
- // D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
- // D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
- // D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
+ // D.f[DIR_M00] = &DD[DIR_P00 * size_Mat];
+ // D.f[DIR_P00] = &DD[DIR_M00 * size_Mat];
+ // D.f[DIR_0M0] = &DD[DIR_0P0 * size_Mat];
+ // D.f[DIR_0P0] = &DD[DIR_0M0 * size_Mat];
+ // D.f[DIR_00M] = &DD[DIR_00P * size_Mat];
+ // D.f[DIR_00P] = &DD[DIR_00M * size_Mat];
+ // D.f[DIR_MM0] = &DD[DIR_PP0 * size_Mat];
+ // D.f[DIR_PP0] = &DD[DIR_MM0 * size_Mat];
+ // D.f[DIR_MP0] = &DD[DIR_PM0 * size_Mat];
+ // D.f[DIR_PM0] = &DD[DIR_MP0 * size_Mat];
+ // D.f[DIR_M0M] = &DD[DIR_P0P * size_Mat];
+ // D.f[DIR_P0P] = &DD[DIR_M0M * size_Mat];
+ // D.f[DIR_M0P] = &DD[DIR_P0M * size_Mat];
+ // D.f[DIR_P0M] = &DD[DIR_M0P * size_Mat];
+ // D.f[DIR_0MM] = &DD[DIR_0PP * size_Mat];
+ // D.f[DIR_0PP] = &DD[DIR_0MM * size_Mat];
+ // D.f[DIR_0MP] = &DD[DIR_0PM * size_Mat];
+ // D.f[DIR_0PM] = &DD[DIR_0MP * size_Mat];
+ // D.f[DIR_000] = &DD[DIR_000 * size_Mat];
+ // D.f[DIR_PPP] = &DD[DIR_MMM * size_Mat];
+ // D.f[DIR_MMP] = &DD[DIR_PPM * size_Mat];
+ // D.f[DIR_PMP] = &DD[DIR_MPM * size_Mat];
+ // D.f[DIR_MPP] = &DD[DIR_PMM * size_Mat];
+ // D.f[DIR_PPM] = &DD[DIR_MMP * size_Mat];
+ // D.f[DIR_MMM] = &DD[DIR_PPP * size_Mat];
+ // D.f[DIR_PMM] = &DD[DIR_MPP * size_Mat];
+ // D.f[DIR_MPM] = &DD[DIR_PMP * size_Mat];
//}
Distributions7 D7;
if (isEvenTimestep==true)
{
- D7.f[0] = &DD7[0*size_Mat];
- D7.f[1] = &DD7[1*size_Mat];
- D7.f[2] = &DD7[2*size_Mat];
- D7.f[3] = &DD7[3*size_Mat];
- D7.f[4] = &DD7[4*size_Mat];
- D7.f[5] = &DD7[5*size_Mat];
- D7.f[6] = &DD7[6*size_Mat];
+ D7.f[0] = &DD7[0*numberOfLBnodes];
+ D7.f[1] = &DD7[1*numberOfLBnodes];
+ D7.f[2] = &DD7[2*numberOfLBnodes];
+ D7.f[3] = &DD7[3*numberOfLBnodes];
+ D7.f[4] = &DD7[4*numberOfLBnodes];
+ D7.f[5] = &DD7[5*numberOfLBnodes];
+ D7.f[6] = &DD7[6*numberOfLBnodes];
}
else
{
- D7.f[0] = &DD7[0*size_Mat];
- D7.f[2] = &DD7[1*size_Mat];
- D7.f[1] = &DD7[2*size_Mat];
- D7.f[4] = &DD7[3*size_Mat];
- D7.f[3] = &DD7[4*size_Mat];
- D7.f[6] = &DD7[5*size_Mat];
- D7.f[5] = &DD7[6*size_Mat];
+ D7.f[0] = &DD7[0*numberOfLBnodes];
+ D7.f[2] = &DD7[1*numberOfLBnodes];
+ D7.f[1] = &DD7[2*numberOfLBnodes];
+ D7.f[4] = &DD7[3*numberOfLBnodes];
+ D7.f[3] = &DD7[4*numberOfLBnodes];
+ D7.f[6] = &DD7[5*numberOfLBnodes];
+ D7.f[5] = &DD7[6*numberOfLBnodes];
}
@@ -4009,12 +4009,12 @@ __global__ void QNoSlipADincomp7(
//////////////////////////////////////////////////////////////////////////////////
real *q_dirE, *q_dirW, *q_dirN, *q_dirS, *q_dirT, *q_dirB;
- q_dirE = &QQ[DIR_P00 * numberOfBCnodes];
- q_dirW = &QQ[DIR_M00 * numberOfBCnodes];
- q_dirN = &QQ[DIR_0P0 * numberOfBCnodes];
- q_dirS = &QQ[DIR_0M0 * numberOfBCnodes];
- q_dirT = &QQ[DIR_00P * numberOfBCnodes];
- q_dirB = &QQ[DIR_00M * numberOfBCnodes];
+ q_dirE = &QQ[DIR_P00 * numberOfBCnodes];
+ q_dirW = &QQ[DIR_M00 * numberOfBCnodes];
+ q_dirN = &QQ[DIR_0P0 * numberOfBCnodes];
+ q_dirS = &QQ[DIR_0M0 * numberOfBCnodes];
+ q_dirT = &QQ[DIR_00P * numberOfBCnodes];
+ q_dirB = &QQ[DIR_00M * numberOfBCnodes];
//////////////////////////////////////////////////////////////////////////////////
//index
unsigned int KQK = k_Q[k];
@@ -4046,32 +4046,32 @@ __global__ void QNoSlipADincomp7(
//unsigned int ktne = KQK;
//unsigned int kbsw = neighborZ[ksw];
////////////////////////////////////////////////////////////////////////////////
- //real f_W = (D.f[DIR_P00 ])[ke ];
- //real f_E = (D.f[DIR_M00 ])[kw ];
- //real f_S = (D.f[DIR_0P0 ])[kn ];
- //real f_N = (D.f[DIR_0M0 ])[ks ];
- //real f_B = (D.f[DIR_00P ])[kt ];
- //real f_T = (D.f[DIR_00M ])[kb ];
- //real f_SW = (D.f[DIR_PP0 ])[kne ];
- //real f_NE = (D.f[DIR_MM0 ])[ksw ];
- //real f_NW = (D.f[DIR_PM0 ])[kse ];
- //real f_SE = (D.f[DIR_MP0 ])[knw ];
- //real f_BW = (D.f[DIR_P0P ])[kte ];
- //real f_TE = (D.f[DIR_M0M ])[kbw ];
- //real f_TW = (D.f[DIR_P0M ])[kbe ];
- //real f_BE = (D.f[DIR_M0P ])[ktw ];
- //real f_BS = (D.f[DIR_0PP ])[ktn ];
- //real f_TN = (D.f[DIR_0MM ])[kbs ];
- //real f_TS = (D.f[DIR_0PM ])[kbn ];
- //real f_BN = (D.f[DIR_0MP ])[kts ];
- //real f_BSW = (D.f[DIR_PPP ])[ktne ];
- //real f_BNE = (D.f[DIR_MMP ])[ktsw ];
- //real f_BNW = (D.f[DIR_PMP ])[ktse ];
- //real f_BSE = (D.f[DIR_MPP ])[ktnw ];
- //real f_TSW = (D.f[DIR_PPM ])[kbne ];
- //real f_TNE = (D.f[DIR_MMM ])[kbsw ];
- //real f_TNW = (D.f[DIR_PMM ])[kbse ];
- //real f_TSE = (D.f[DIR_MPM ])[kbnw ];
+ //real f_W = (D.f[DIR_P00])[ke ];
+ //real f_E = (D.f[DIR_M00])[kw ];
+ //real f_S = (D.f[DIR_0P0])[kn ];
+ //real f_N = (D.f[DIR_0M0])[ks ];
+ //real f_B = (D.f[DIR_00P])[kt ];
+ //real f_T = (D.f[DIR_00M])[kb ];
+ //real f_SW = (D.f[DIR_PP0])[kne ];
+ //real f_NE = (D.f[DIR_MM0])[ksw ];
+ //real f_NW = (D.f[DIR_PM0])[kse ];
+ //real f_SE = (D.f[DIR_MP0])[knw ];
+ //real f_BW = (D.f[DIR_P0P])[kte ];
+ //real f_TE = (D.f[DIR_M0M])[kbw ];
+ //real f_TW = (D.f[DIR_P0M])[kbe ];
+ //real f_BE = (D.f[DIR_M0P])[ktw ];
+ //real f_BS = (D.f[DIR_0PP])[ktn ];
+ //real f_TN = (D.f[DIR_0MM])[kbs ];
+ //real f_TS = (D.f[DIR_0PM])[kbn ];
+ //real f_BN = (D.f[DIR_0MP])[kts ];
+ //real f_BSW = (D.f[DIR_PPP])[ktne ];
+ //real f_BNE = (D.f[DIR_MMP])[ktsw ];
+ //real f_BNW = (D.f[DIR_PMP])[ktse ];
+ //real f_BSE = (D.f[DIR_MPP])[ktnw ];
+ //real f_TSW = (D.f[DIR_PPM])[kbne ];
+ //real f_TNE = (D.f[DIR_MMM])[kbsw ];
+ //real f_TNW = (D.f[DIR_PMM])[kbse ];
+ //real f_TSE = (D.f[DIR_MPM])[kbnw ];
////////////////////////////////////////////////////////////////////////////////
//real vx1 = ((f_TNE-f_BSW)+(f_BNE-f_TSW)+(f_TSE-f_BNW)+(f_BSE-f_TNW) +(f_NE-f_SW)+(f_SE-f_NW)+(f_TE-f_BW)+(f_BE-f_TW)+(f_E-f_W));
//real vx2 = ((f_TNE-f_BSW)+(f_BNE-f_TSW)+(f_BNW-f_TSE)+(f_TNW-f_BSE) +(f_NE-f_SW)+(f_NW-f_SE)+(f_TN-f_BS)+(f_BN-f_TS)+(f_N-f_S));
@@ -4131,23 +4131,23 @@ __global__ void QNoSlipADincomp7(
//pointertausch
if (isEvenTimestep==false)
{
- D7.f[0] = &DD7[0*size_Mat];
- D7.f[1] = &DD7[1*size_Mat];
- D7.f[2] = &DD7[2*size_Mat];
- D7.f[3] = &DD7[3*size_Mat];
- D7.f[4] = &DD7[4*size_Mat];
- D7.f[5] = &DD7[5*size_Mat];
- D7.f[6] = &DD7[6*size_Mat];
+ D7.f[0] = &DD7[0*numberOfLBnodes];
+ D7.f[1] = &DD7[1*numberOfLBnodes];
+ D7.f[2] = &DD7[2*numberOfLBnodes];
+ D7.f[3] = &DD7[3*numberOfLBnodes];
+ D7.f[4] = &DD7[4*numberOfLBnodes];
+ D7.f[5] = &DD7[5*numberOfLBnodes];
+ D7.f[6] = &DD7[6*numberOfLBnodes];
}
else
{
- D7.f[0] = &DD7[0*size_Mat];
- D7.f[2] = &DD7[1*size_Mat];
- D7.f[1] = &DD7[2*size_Mat];
- D7.f[4] = &DD7[3*size_Mat];
- D7.f[3] = &DD7[4*size_Mat];
- D7.f[6] = &DD7[5*size_Mat];
- D7.f[5] = &DD7[6*size_Mat];
+ D7.f[0] = &DD7[0*numberOfLBnodes];
+ D7.f[2] = &DD7[1*numberOfLBnodes];
+ D7.f[1] = &DD7[2*numberOfLBnodes];
+ D7.f[4] = &DD7[3*numberOfLBnodes];
+ D7.f[3] = &DD7[4*numberOfLBnodes];
+ D7.f[6] = &DD7[5*numberOfLBnodes];
+ D7.f[5] = &DD7[6*numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////
@@ -4329,131 +4329,131 @@ __global__ void QNoSlipADincomp27(
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep)
{
Distributions27 D;
if (isEvenTimestep==true)
{
- D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
+ D.f[DIR_M00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_MPM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_PMP * numberOfLBnodes];
}
Distributions27 D27;
if (isEvenTimestep==true)
{
- D27.f[DIR_P00 ] = &DD27[DIR_P00 *size_Mat];
- D27.f[DIR_M00 ] = &DD27[DIR_M00 *size_Mat];
- D27.f[DIR_0P0 ] = &DD27[DIR_0P0 *size_Mat];
- D27.f[DIR_0M0 ] = &DD27[DIR_0M0 *size_Mat];
- D27.f[DIR_00P ] = &DD27[DIR_00P *size_Mat];
- D27.f[DIR_00M ] = &DD27[DIR_00M *size_Mat];
- D27.f[DIR_PP0 ] = &DD27[DIR_PP0 *size_Mat];
- D27.f[DIR_MM0 ] = &DD27[DIR_MM0 *size_Mat];
- D27.f[DIR_PM0 ] = &DD27[DIR_PM0 *size_Mat];
- D27.f[DIR_MP0 ] = &DD27[DIR_MP0 *size_Mat];
- D27.f[DIR_P0P ] = &DD27[DIR_P0P *size_Mat];
- D27.f[DIR_M0M ] = &DD27[DIR_M0M *size_Mat];
- D27.f[DIR_P0M ] = &DD27[DIR_P0M *size_Mat];
- D27.f[DIR_M0P ] = &DD27[DIR_M0P *size_Mat];
- D27.f[DIR_0PP ] = &DD27[DIR_0PP *size_Mat];
- D27.f[DIR_0MM ] = &DD27[DIR_0MM *size_Mat];
- D27.f[DIR_0PM ] = &DD27[DIR_0PM *size_Mat];
- D27.f[DIR_0MP ] = &DD27[DIR_0MP *size_Mat];
- D27.f[DIR_000] = &DD27[DIR_000*size_Mat];
- D27.f[DIR_PPP ] = &DD27[DIR_PPP *size_Mat];
- D27.f[DIR_MMP ] = &DD27[DIR_MMP *size_Mat];
- D27.f[DIR_PMP ] = &DD27[DIR_PMP *size_Mat];
- D27.f[DIR_MPP ] = &DD27[DIR_MPP *size_Mat];
- D27.f[DIR_PPM ] = &DD27[DIR_PPM *size_Mat];
- D27.f[DIR_MMM ] = &DD27[DIR_MMM *size_Mat];
- D27.f[DIR_PMM ] = &DD27[DIR_PMM *size_Mat];
- D27.f[DIR_MPM ] = &DD27[DIR_MPM *size_Mat];
+ D27.f[DIR_P00] = &DD27[DIR_P00 * numberOfLBnodes];
+ D27.f[DIR_M00] = &DD27[DIR_M00 * numberOfLBnodes];
+ D27.f[DIR_0P0] = &DD27[DIR_0P0 * numberOfLBnodes];
+ D27.f[DIR_0M0] = &DD27[DIR_0M0 * numberOfLBnodes];
+ D27.f[DIR_00P] = &DD27[DIR_00P * numberOfLBnodes];
+ D27.f[DIR_00M] = &DD27[DIR_00M * numberOfLBnodes];
+ D27.f[DIR_PP0] = &DD27[DIR_PP0 * numberOfLBnodes];
+ D27.f[DIR_MM0] = &DD27[DIR_MM0 * numberOfLBnodes];
+ D27.f[DIR_PM0] = &DD27[DIR_PM0 * numberOfLBnodes];
+ D27.f[DIR_MP0] = &DD27[DIR_MP0 * numberOfLBnodes];
+ D27.f[DIR_P0P] = &DD27[DIR_P0P * numberOfLBnodes];
+ D27.f[DIR_M0M] = &DD27[DIR_M0M * numberOfLBnodes];
+ D27.f[DIR_P0M] = &DD27[DIR_P0M * numberOfLBnodes];
+ D27.f[DIR_M0P] = &DD27[DIR_M0P * numberOfLBnodes];
+ D27.f[DIR_0PP] = &DD27[DIR_0PP * numberOfLBnodes];
+ D27.f[DIR_0MM] = &DD27[DIR_0MM * numberOfLBnodes];
+ D27.f[DIR_0PM] = &DD27[DIR_0PM * numberOfLBnodes];
+ D27.f[DIR_0MP] = &DD27[DIR_0MP * numberOfLBnodes];
+ D27.f[DIR_000] = &DD27[DIR_000 * numberOfLBnodes];
+ D27.f[DIR_PPP] = &DD27[DIR_PPP * numberOfLBnodes];
+ D27.f[DIR_MMP] = &DD27[DIR_MMP * numberOfLBnodes];
+ D27.f[DIR_PMP] = &DD27[DIR_PMP * numberOfLBnodes];
+ D27.f[DIR_MPP] = &DD27[DIR_MPP * numberOfLBnodes];
+ D27.f[DIR_PPM] = &DD27[DIR_PPM * numberOfLBnodes];
+ D27.f[DIR_MMM] = &DD27[DIR_MMM * numberOfLBnodes];
+ D27.f[DIR_PMM] = &DD27[DIR_PMM * numberOfLBnodes];
+ D27.f[DIR_MPM] = &DD27[DIR_MPM * numberOfLBnodes];
}
else
{
- D27.f[DIR_M00 ] = &DD27[DIR_P00 *size_Mat];
- D27.f[DIR_P00 ] = &DD27[DIR_M00 *size_Mat];
- D27.f[DIR_0M0 ] = &DD27[DIR_0P0 *size_Mat];
- D27.f[DIR_0P0 ] = &DD27[DIR_0M0 *size_Mat];
- D27.f[DIR_00M ] = &DD27[DIR_00P *size_Mat];
- D27.f[DIR_00P ] = &DD27[DIR_00M *size_Mat];
- D27.f[DIR_MM0 ] = &DD27[DIR_PP0 *size_Mat];
- D27.f[DIR_PP0 ] = &DD27[DIR_MM0 *size_Mat];
- D27.f[DIR_MP0 ] = &DD27[DIR_PM0 *size_Mat];
- D27.f[DIR_PM0 ] = &DD27[DIR_MP0 *size_Mat];
- D27.f[DIR_M0M ] = &DD27[DIR_P0P *size_Mat];
- D27.f[DIR_P0P ] = &DD27[DIR_M0M *size_Mat];
- D27.f[DIR_M0P ] = &DD27[DIR_P0M *size_Mat];
- D27.f[DIR_P0M ] = &DD27[DIR_M0P *size_Mat];
- D27.f[DIR_0MM ] = &DD27[DIR_0PP *size_Mat];
- D27.f[DIR_0PP ] = &DD27[DIR_0MM *size_Mat];
- D27.f[DIR_0MP ] = &DD27[DIR_0PM *size_Mat];
- D27.f[DIR_0PM ] = &DD27[DIR_0MP *size_Mat];
- D27.f[DIR_000] = &DD27[DIR_000*size_Mat];
- D27.f[DIR_PPP ] = &DD27[DIR_MMM *size_Mat];
- D27.f[DIR_MMP ] = &DD27[DIR_PPM *size_Mat];
- D27.f[DIR_PMP ] = &DD27[DIR_MPM *size_Mat];
- D27.f[DIR_MPP ] = &DD27[DIR_PMM *size_Mat];
- D27.f[DIR_PPM ] = &DD27[DIR_MMP *size_Mat];
- D27.f[DIR_MMM ] = &DD27[DIR_PPP *size_Mat];
- D27.f[DIR_PMM ] = &DD27[DIR_MPP *size_Mat];
- D27.f[DIR_MPM ] = &DD27[DIR_PMP *size_Mat];
+ D27.f[DIR_M00] = &DD27[DIR_P00 * numberOfLBnodes];
+ D27.f[DIR_P00] = &DD27[DIR_M00 * numberOfLBnodes];
+ D27.f[DIR_0M0] = &DD27[DIR_0P0 * numberOfLBnodes];
+ D27.f[DIR_0P0] = &DD27[DIR_0M0 * numberOfLBnodes];
+ D27.f[DIR_00M] = &DD27[DIR_00P * numberOfLBnodes];
+ D27.f[DIR_00P] = &DD27[DIR_00M * numberOfLBnodes];
+ D27.f[DIR_MM0] = &DD27[DIR_PP0 * numberOfLBnodes];
+ D27.f[DIR_PP0] = &DD27[DIR_MM0 * numberOfLBnodes];
+ D27.f[DIR_MP0] = &DD27[DIR_PM0 * numberOfLBnodes];
+ D27.f[DIR_PM0] = &DD27[DIR_MP0 * numberOfLBnodes];
+ D27.f[DIR_M0M] = &DD27[DIR_P0P * numberOfLBnodes];
+ D27.f[DIR_P0P] = &DD27[DIR_M0M * numberOfLBnodes];
+ D27.f[DIR_M0P] = &DD27[DIR_P0M * numberOfLBnodes];
+ D27.f[DIR_P0M] = &DD27[DIR_M0P * numberOfLBnodes];
+ D27.f[DIR_0MM] = &DD27[DIR_0PP * numberOfLBnodes];
+ D27.f[DIR_0PP] = &DD27[DIR_0MM * numberOfLBnodes];
+ D27.f[DIR_0MP] = &DD27[DIR_0PM * numberOfLBnodes];
+ D27.f[DIR_0PM] = &DD27[DIR_0MP * numberOfLBnodes];
+ D27.f[DIR_000] = &DD27[DIR_000 * numberOfLBnodes];
+ D27.f[DIR_PPP] = &DD27[DIR_MMM * numberOfLBnodes];
+ D27.f[DIR_MMP] = &DD27[DIR_PPM * numberOfLBnodes];
+ D27.f[DIR_PMP] = &DD27[DIR_MPM * numberOfLBnodes];
+ D27.f[DIR_MPP] = &DD27[DIR_PMM * numberOfLBnodes];
+ D27.f[DIR_PPM] = &DD27[DIR_MMP * numberOfLBnodes];
+ D27.f[DIR_MMM] = &DD27[DIR_PPP * numberOfLBnodes];
+ D27.f[DIR_PMM] = &DD27[DIR_MPP * numberOfLBnodes];
+ D27.f[DIR_MPM] = &DD27[DIR_PMP * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////
const unsigned x = threadIdx.x; // Globaler x-Index
@@ -4474,24 +4474,24 @@ __global__ void QNoSlipADincomp27(
*q_dirBE, *q_dirTW, *q_dirTN, *q_dirBS, *q_dirBN, *q_dirTS,
*q_dirTNE, *q_dirTSW, *q_dirTSE, *q_dirTNW, *q_dirBNE, *q_dirBSW,
*q_dirBSE, *q_dirBNW;
- q_dirE = &QQ[DIR_P00 * numberOfBCnodes];
- q_dirW = &QQ[DIR_M00 * numberOfBCnodes];
- q_dirN = &QQ[DIR_0P0 * numberOfBCnodes];
- q_dirS = &QQ[DIR_0M0 * numberOfBCnodes];
- q_dirT = &QQ[DIR_00P * numberOfBCnodes];
- q_dirB = &QQ[DIR_00M * numberOfBCnodes];
- q_dirNE = &QQ[DIR_PP0 * numberOfBCnodes];
- q_dirSW = &QQ[DIR_MM0 * numberOfBCnodes];
- q_dirSE = &QQ[DIR_PM0 * numberOfBCnodes];
- q_dirNW = &QQ[DIR_MP0 * numberOfBCnodes];
- q_dirTE = &QQ[DIR_P0P * numberOfBCnodes];
- q_dirBW = &QQ[DIR_M0M * numberOfBCnodes];
- q_dirBE = &QQ[DIR_P0M * numberOfBCnodes];
- q_dirTW = &QQ[DIR_M0P * numberOfBCnodes];
- q_dirTN = &QQ[DIR_0PP * numberOfBCnodes];
- q_dirBS = &QQ[DIR_0MM * numberOfBCnodes];
- q_dirBN = &QQ[DIR_0PM * numberOfBCnodes];
- q_dirTS = &QQ[DIR_0MP * numberOfBCnodes];
+ q_dirE = &QQ[DIR_P00 * numberOfBCnodes];
+ q_dirW = &QQ[DIR_M00 * numberOfBCnodes];
+ q_dirN = &QQ[DIR_0P0 * numberOfBCnodes];
+ q_dirS = &QQ[DIR_0M0 * numberOfBCnodes];
+ q_dirT = &QQ[DIR_00P * numberOfBCnodes];
+ q_dirB = &QQ[DIR_00M * numberOfBCnodes];
+ q_dirNE = &QQ[DIR_PP0 * numberOfBCnodes];
+ q_dirSW = &QQ[DIR_MM0 * numberOfBCnodes];
+ q_dirSE = &QQ[DIR_PM0 * numberOfBCnodes];
+ q_dirNW = &QQ[DIR_MP0 * numberOfBCnodes];
+ q_dirTE = &QQ[DIR_P0P * numberOfBCnodes];
+ q_dirBW = &QQ[DIR_M0M * numberOfBCnodes];
+ q_dirBE = &QQ[DIR_P0M * numberOfBCnodes];
+ q_dirTW = &QQ[DIR_M0P * numberOfBCnodes];
+ q_dirTN = &QQ[DIR_0PP * numberOfBCnodes];
+ q_dirBS = &QQ[DIR_0MM * numberOfBCnodes];
+ q_dirBN = &QQ[DIR_0PM * numberOfBCnodes];
+ q_dirTS = &QQ[DIR_0MP * numberOfBCnodes];
q_dirTNE = &QQ[DIR_PPP * numberOfBCnodes];
q_dirTSW = &QQ[DIR_MMP * numberOfBCnodes];
q_dirTSE = &QQ[DIR_PMP * numberOfBCnodes];
@@ -4531,65 +4531,65 @@ __global__ void QNoSlipADincomp27(
unsigned int ktne = KQK;
unsigned int kbsw = neighborZ[ksw];
////////////////////////////////////////////////////////////////////////////////
- real f_W = (D.f[DIR_P00 ])[ke ];
- real f_E = (D.f[DIR_M00 ])[kw ];
- real f_S = (D.f[DIR_0P0 ])[kn ];
- real f_N = (D.f[DIR_0M0 ])[ks ];
- real f_B = (D.f[DIR_00P ])[kt ];
- real f_T = (D.f[DIR_00M ])[kb ];
- real f_SW = (D.f[DIR_PP0 ])[kne ];
- real f_NE = (D.f[DIR_MM0 ])[ksw ];
- real f_NW = (D.f[DIR_PM0 ])[kse ];
- real f_SE = (D.f[DIR_MP0 ])[knw ];
- real f_BW = (D.f[DIR_P0P ])[kte ];
- real f_TE = (D.f[DIR_M0M ])[kbw ];
- real f_TW = (D.f[DIR_P0M ])[kbe ];
- real f_BE = (D.f[DIR_M0P ])[ktw ];
- real f_BS = (D.f[DIR_0PP ])[ktn ];
- real f_TN = (D.f[DIR_0MM ])[kbs ];
- real f_TS = (D.f[DIR_0PM ])[kbn ];
- real f_BN = (D.f[DIR_0MP ])[kts ];
+ real f_W = (D.f[DIR_P00])[ke ];
+ real f_E = (D.f[DIR_M00])[kw ];
+ real f_S = (D.f[DIR_0P0])[kn ];
+ real f_N = (D.f[DIR_0M0])[ks ];
+ real f_B = (D.f[DIR_00P])[kt ];
+ real f_T = (D.f[DIR_00M])[kb ];
+ real f_SW = (D.f[DIR_PP0])[kne ];
+ real f_NE = (D.f[DIR_MM0])[ksw ];
+ real f_NW = (D.f[DIR_PM0])[kse ];
+ real f_SE = (D.f[DIR_MP0])[knw ];
+ real f_BW = (D.f[DIR_P0P])[kte ];
+ real f_TE = (D.f[DIR_M0M])[kbw ];
+ real f_TW = (D.f[DIR_P0M])[kbe ];
+ real f_BE = (D.f[DIR_M0P])[ktw ];
+ real f_BS = (D.f[DIR_0PP])[ktn ];
+ real f_TN = (D.f[DIR_0MM])[kbs ];
+ real f_TS = (D.f[DIR_0PM])[kbn ];
+ real f_BN = (D.f[DIR_0MP])[kts ];
//real f_ZERO = (D.f[DIR_000])[kzero];
- real f_BSW = (D.f[DIR_PPP ])[ktne ];
- real f_BNE = (D.f[DIR_MMP ])[ktsw ];
- real f_BNW = (D.f[DIR_PMP ])[ktse ];
- real f_BSE = (D.f[DIR_MPP ])[ktnw ];
- real f_TSW = (D.f[DIR_PPM ])[kbne ];
- real f_TNE = (D.f[DIR_MMM ])[kbsw ];
- real f_TNW = (D.f[DIR_PMM ])[kbse ];
- real f_TSE = (D.f[DIR_MPM ])[kbnw ];
+ real f_BSW = (D.f[DIR_PPP])[ktne ];
+ real f_BNE = (D.f[DIR_MMP])[ktsw ];
+ real f_BNW = (D.f[DIR_PMP])[ktse ];
+ real f_BSE = (D.f[DIR_MPP])[ktnw ];
+ real f_TSW = (D.f[DIR_PPM])[kbne ];
+ real f_TNE = (D.f[DIR_MMM])[kbsw ];
+ real f_TNW = (D.f[DIR_PMM])[kbse ];
+ real f_TSE = (D.f[DIR_MPM])[kbnw ];
////////////////////////////////////////////////////////////////////////////////
real vx1 = ((f_TNE-f_BSW)+(f_BNE-f_TSW)+(f_TSE-f_BNW)+(f_BSE-f_TNW) +(f_NE-f_SW)+(f_SE-f_NW)+(f_TE-f_BW)+(f_BE-f_TW)+(f_E-f_W));
real vx2 = ((f_TNE-f_BSW)+(f_BNE-f_TSW)+(f_BNW-f_TSE)+(f_TNW-f_BSE) +(f_NE-f_SW)+(f_NW-f_SE)+(f_TN-f_BS)+(f_BN-f_TS)+(f_N-f_S));
real vx3 = ((f_TNE-f_BSW)+(f_TSW-f_BNE)+(f_TSE-f_BNW)+(f_TNW-f_BSE) +(f_TE-f_BW)+(f_TW-f_BE)+(f_TN-f_BS)+(f_TS-f_BN)+(f_T-f_B));
////////////////////////////////////////////////////////////////////////////////
- real f27_W = (D27.f[DIR_P00 ])[ke ];
- real f27_E = (D27.f[DIR_M00 ])[kw ];
- real f27_S = (D27.f[DIR_0P0 ])[kn ];
- real f27_N = (D27.f[DIR_0M0 ])[ks ];
- real f27_B = (D27.f[DIR_00P ])[kt ];
- real f27_T = (D27.f[DIR_00M ])[kb ];
- real f27_SW = (D27.f[DIR_PP0 ])[kne ];
- real f27_NE = (D27.f[DIR_MM0 ])[ksw ];
- real f27_NW = (D27.f[DIR_PM0 ])[kse ];
- real f27_SE = (D27.f[DIR_MP0 ])[knw ];
- real f27_BW = (D27.f[DIR_P0P ])[kte ];
- real f27_TE = (D27.f[DIR_M0M ])[kbw ];
- real f27_TW = (D27.f[DIR_P0M ])[kbe ];
- real f27_BE = (D27.f[DIR_M0P ])[ktw ];
- real f27_BS = (D27.f[DIR_0PP ])[ktn ];
- real f27_TN = (D27.f[DIR_0MM ])[kbs ];
- real f27_TS = (D27.f[DIR_0PM ])[kbn ];
- real f27_BN = (D27.f[DIR_0MP ])[kts ];
+ real f27_W = (D27.f[DIR_P00])[ke ];
+ real f27_E = (D27.f[DIR_M00])[kw ];
+ real f27_S = (D27.f[DIR_0P0])[kn ];
+ real f27_N = (D27.f[DIR_0M0])[ks ];
+ real f27_B = (D27.f[DIR_00P])[kt ];
+ real f27_T = (D27.f[DIR_00M])[kb ];
+ real f27_SW = (D27.f[DIR_PP0])[kne ];
+ real f27_NE = (D27.f[DIR_MM0])[ksw ];
+ real f27_NW = (D27.f[DIR_PM0])[kse ];
+ real f27_SE = (D27.f[DIR_MP0])[knw ];
+ real f27_BW = (D27.f[DIR_P0P])[kte ];
+ real f27_TE = (D27.f[DIR_M0M])[kbw ];
+ real f27_TW = (D27.f[DIR_P0M])[kbe ];
+ real f27_BE = (D27.f[DIR_M0P])[ktw ];
+ real f27_BS = (D27.f[DIR_0PP])[ktn ];
+ real f27_TN = (D27.f[DIR_0MM])[kbs ];
+ real f27_TS = (D27.f[DIR_0PM])[kbn ];
+ real f27_BN = (D27.f[DIR_0MP])[kts ];
real f27_ZERO = (D27.f[DIR_000])[kzero];
- real f27_BSW = (D27.f[DIR_PPP ])[ktne ];
- real f27_BNE = (D27.f[DIR_MMP ])[ktsw ];
- real f27_BNW = (D27.f[DIR_PMP ])[ktse ];
- real f27_BSE = (D27.f[DIR_MPP ])[ktnw ];
- real f27_TSW = (D27.f[DIR_PPM ])[kbne ];
- real f27_TNE = (D27.f[DIR_MMM ])[kbsw ];
- real f27_TNW = (D27.f[DIR_PMM ])[kbse ];
- real f27_TSE = (D27.f[DIR_MPM ])[kbnw ];
+ real f27_BSW = (D27.f[DIR_PPP])[ktne ];
+ real f27_BNE = (D27.f[DIR_MMP])[ktsw ];
+ real f27_BNW = (D27.f[DIR_PMP])[ktse ];
+ real f27_BSE = (D27.f[DIR_MPP])[ktnw ];
+ real f27_TSW = (D27.f[DIR_PPM])[kbne ];
+ real f27_TNE = (D27.f[DIR_MMM])[kbsw ];
+ real f27_TNW = (D27.f[DIR_PMM])[kbse ];
+ real f27_TSE = (D27.f[DIR_MPM])[kbnw ];
////////////////////////////////////////////////////////////////////////////////
real cu_sq=c3o2*(vx1*vx1+vx2*vx2+vx3*vx3);
////////////////////////////////////////////////////////////////////////////////
@@ -4665,63 +4665,63 @@ __global__ void QNoSlipADincomp27(
//////////////////////////////////////////////////////////////////////////
if (isEvenTimestep==false)
{
- D27.f[DIR_P00 ] = &DD27[DIR_P00 *size_Mat];
- D27.f[DIR_M00 ] = &DD27[DIR_M00 *size_Mat];
- D27.f[DIR_0P0 ] = &DD27[DIR_0P0 *size_Mat];
- D27.f[DIR_0M0 ] = &DD27[DIR_0M0 *size_Mat];
- D27.f[DIR_00P ] = &DD27[DIR_00P *size_Mat];
- D27.f[DIR_00M ] = &DD27[DIR_00M *size_Mat];
- D27.f[DIR_PP0 ] = &DD27[DIR_PP0 *size_Mat];
- D27.f[DIR_MM0 ] = &DD27[DIR_MM0 *size_Mat];
- D27.f[DIR_PM0 ] = &DD27[DIR_PM0 *size_Mat];
- D27.f[DIR_MP0 ] = &DD27[DIR_MP0 *size_Mat];
- D27.f[DIR_P0P ] = &DD27[DIR_P0P *size_Mat];
- D27.f[DIR_M0M ] = &DD27[DIR_M0M *size_Mat];
- D27.f[DIR_P0M ] = &DD27[DIR_P0M *size_Mat];
- D27.f[DIR_M0P ] = &DD27[DIR_M0P *size_Mat];
- D27.f[DIR_0PP ] = &DD27[DIR_0PP *size_Mat];
- D27.f[DIR_0MM ] = &DD27[DIR_0MM *size_Mat];
- D27.f[DIR_0PM ] = &DD27[DIR_0PM *size_Mat];
- D27.f[DIR_0MP ] = &DD27[DIR_0MP *size_Mat];
- D27.f[DIR_000] = &DD27[DIR_000*size_Mat];
- D27.f[DIR_PPP ] = &DD27[DIR_PPP *size_Mat];
- D27.f[DIR_MMP ] = &DD27[DIR_MMP *size_Mat];
- D27.f[DIR_PMP ] = &DD27[DIR_PMP *size_Mat];
- D27.f[DIR_MPP ] = &DD27[DIR_MPP *size_Mat];
- D27.f[DIR_PPM ] = &DD27[DIR_PPM *size_Mat];
- D27.f[DIR_MMM ] = &DD27[DIR_MMM *size_Mat];
- D27.f[DIR_PMM ] = &DD27[DIR_PMM *size_Mat];
- D27.f[DIR_MPM ] = &DD27[DIR_MPM *size_Mat];
+ D27.f[DIR_P00] = &DD27[DIR_P00 * numberOfLBnodes];
+ D27.f[DIR_M00] = &DD27[DIR_M00 * numberOfLBnodes];
+ D27.f[DIR_0P0] = &DD27[DIR_0P0 * numberOfLBnodes];
+ D27.f[DIR_0M0] = &DD27[DIR_0M0 * numberOfLBnodes];
+ D27.f[DIR_00P] = &DD27[DIR_00P * numberOfLBnodes];
+ D27.f[DIR_00M] = &DD27[DIR_00M * numberOfLBnodes];
+ D27.f[DIR_PP0] = &DD27[DIR_PP0 * numberOfLBnodes];
+ D27.f[DIR_MM0] = &DD27[DIR_MM0 * numberOfLBnodes];
+ D27.f[DIR_PM0] = &DD27[DIR_PM0 * numberOfLBnodes];
+ D27.f[DIR_MP0] = &DD27[DIR_MP0 * numberOfLBnodes];
+ D27.f[DIR_P0P] = &DD27[DIR_P0P * numberOfLBnodes];
+ D27.f[DIR_M0M] = &DD27[DIR_M0M * numberOfLBnodes];
+ D27.f[DIR_P0M] = &DD27[DIR_P0M * numberOfLBnodes];
+ D27.f[DIR_M0P] = &DD27[DIR_M0P * numberOfLBnodes];
+ D27.f[DIR_0PP] = &DD27[DIR_0PP * numberOfLBnodes];
+ D27.f[DIR_0MM] = &DD27[DIR_0MM * numberOfLBnodes];
+ D27.f[DIR_0PM] = &DD27[DIR_0PM * numberOfLBnodes];
+ D27.f[DIR_0MP] = &DD27[DIR_0MP * numberOfLBnodes];
+ D27.f[DIR_000] = &DD27[DIR_000 * numberOfLBnodes];
+ D27.f[DIR_PPP] = &DD27[DIR_PPP * numberOfLBnodes];
+ D27.f[DIR_MMP] = &DD27[DIR_MMP * numberOfLBnodes];
+ D27.f[DIR_PMP] = &DD27[DIR_PMP * numberOfLBnodes];
+ D27.f[DIR_MPP] = &DD27[DIR_MPP * numberOfLBnodes];
+ D27.f[DIR_PPM] = &DD27[DIR_PPM * numberOfLBnodes];
+ D27.f[DIR_MMM] = &DD27[DIR_MMM * numberOfLBnodes];
+ D27.f[DIR_PMM] = &DD27[DIR_PMM * numberOfLBnodes];
+ D27.f[DIR_MPM] = &DD27[DIR_MPM * numberOfLBnodes];
}
else
{
- D27.f[DIR_M00 ] = &DD27[DIR_P00 *size_Mat];
- D27.f[DIR_P00 ] = &DD27[DIR_M00 *size_Mat];
- D27.f[DIR_0M0 ] = &DD27[DIR_0P0 *size_Mat];
- D27.f[DIR_0P0 ] = &DD27[DIR_0M0 *size_Mat];
- D27.f[DIR_00M ] = &DD27[DIR_00P *size_Mat];
- D27.f[DIR_00P ] = &DD27[DIR_00M *size_Mat];
- D27.f[DIR_MM0 ] = &DD27[DIR_PP0 *size_Mat];
- D27.f[DIR_PP0 ] = &DD27[DIR_MM0 *size_Mat];
- D27.f[DIR_MP0 ] = &DD27[DIR_PM0 *size_Mat];
- D27.f[DIR_PM0 ] = &DD27[DIR_MP0 *size_Mat];
- D27.f[DIR_M0M ] = &DD27[DIR_P0P *size_Mat];
- D27.f[DIR_P0P ] = &DD27[DIR_M0M *size_Mat];
- D27.f[DIR_M0P ] = &DD27[DIR_P0M *size_Mat];
- D27.f[DIR_P0M ] = &DD27[DIR_M0P *size_Mat];
- D27.f[DIR_0MM ] = &DD27[DIR_0PP *size_Mat];
- D27.f[DIR_0PP ] = &DD27[DIR_0MM *size_Mat];
- D27.f[DIR_0MP ] = &DD27[DIR_0PM *size_Mat];
- D27.f[DIR_0PM ] = &DD27[DIR_0MP *size_Mat];
- D27.f[DIR_000] = &DD27[DIR_000*size_Mat];
- D27.f[DIR_PPP ] = &DD27[DIR_MMM *size_Mat];
- D27.f[DIR_MMP ] = &DD27[DIR_PPM *size_Mat];
- D27.f[DIR_PMP ] = &DD27[DIR_MPM *size_Mat];
- D27.f[DIR_MPP ] = &DD27[DIR_PMM *size_Mat];
- D27.f[DIR_PPM ] = &DD27[DIR_MMP *size_Mat];
- D27.f[DIR_MMM ] = &DD27[DIR_PPP *size_Mat];
- D27.f[DIR_PMM ] = &DD27[DIR_MPP *size_Mat];
- D27.f[DIR_MPM ] = &DD27[DIR_PMP *size_Mat];
+ D27.f[DIR_M00] = &DD27[DIR_P00 * numberOfLBnodes];
+ D27.f[DIR_P00] = &DD27[DIR_M00 * numberOfLBnodes];
+ D27.f[DIR_0M0] = &DD27[DIR_0P0 * numberOfLBnodes];
+ D27.f[DIR_0P0] = &DD27[DIR_0M0 * numberOfLBnodes];
+ D27.f[DIR_00M] = &DD27[DIR_00P * numberOfLBnodes];
+ D27.f[DIR_00P] = &DD27[DIR_00M * numberOfLBnodes];
+ D27.f[DIR_MM0] = &DD27[DIR_PP0 * numberOfLBnodes];
+ D27.f[DIR_PP0] = &DD27[DIR_MM0 * numberOfLBnodes];
+ D27.f[DIR_MP0] = &DD27[DIR_PM0 * numberOfLBnodes];
+ D27.f[DIR_PM0] = &DD27[DIR_MP0 * numberOfLBnodes];
+ D27.f[DIR_M0M] = &DD27[DIR_P0P * numberOfLBnodes];
+ D27.f[DIR_P0P] = &DD27[DIR_M0M * numberOfLBnodes];
+ D27.f[DIR_M0P] = &DD27[DIR_P0M * numberOfLBnodes];
+ D27.f[DIR_P0M] = &DD27[DIR_M0P * numberOfLBnodes];
+ D27.f[DIR_0MM] = &DD27[DIR_0PP * numberOfLBnodes];
+ D27.f[DIR_0PP] = &DD27[DIR_0MM * numberOfLBnodes];
+ D27.f[DIR_0MP] = &DD27[DIR_0PM * numberOfLBnodes];
+ D27.f[DIR_0PM] = &DD27[DIR_0MP * numberOfLBnodes];
+ D27.f[DIR_000] = &DD27[DIR_000 * numberOfLBnodes];
+ D27.f[DIR_PPP] = &DD27[DIR_MMM * numberOfLBnodes];
+ D27.f[DIR_MMP] = &DD27[DIR_PPM * numberOfLBnodes];
+ D27.f[DIR_PMP] = &DD27[DIR_MPM * numberOfLBnodes];
+ D27.f[DIR_MPP] = &DD27[DIR_PMM * numberOfLBnodes];
+ D27.f[DIR_PPM] = &DD27[DIR_MMP * numberOfLBnodes];
+ D27.f[DIR_MMM] = &DD27[DIR_PPP * numberOfLBnodes];
+ D27.f[DIR_PMM] = &DD27[DIR_MPP * numberOfLBnodes];
+ D27.f[DIR_MPM] = &DD27[DIR_PMP * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
//Test
@@ -4729,24 +4729,24 @@ __global__ void QNoSlipADincomp27(
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
real q;
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
- q = q_dirE[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_M00 ])[kw ]=(c2o1*feqW27_W -(f27_E *(q*omegaD-c1o1)-omegaD*feq27_E *(q-c1o1))/(omegaD-c1o1)+f27_W *q)/(q+c1o1);
- q = q_dirW[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_P00 ])[ke ]=(c2o1*feqW27_E -(f27_W *(q*omegaD-c1o1)-omegaD*feq27_W *(q-c1o1))/(omegaD-c1o1)+f27_E *q)/(q+c1o1);
- q = q_dirN[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_0M0 ])[ks ]=(c2o1*feqW27_S -(f27_N *(q*omegaD-c1o1)-omegaD*feq27_N *(q-c1o1))/(omegaD-c1o1)+f27_S *q)/(q+c1o1);
- q = q_dirS[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_0P0 ])[kn ]=(c2o1*feqW27_N -(f27_S *(q*omegaD-c1o1)-omegaD*feq27_S *(q-c1o1))/(omegaD-c1o1)+f27_N *q)/(q+c1o1);
- q = q_dirT[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_00M ])[kb ]=(c2o1*feqW27_B -(f27_T *(q*omegaD-c1o1)-omegaD*feq27_T *(q-c1o1))/(omegaD-c1o1)+f27_B *q)/(q+c1o1);
- q = q_dirB[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_00P ])[kt ]=(c2o1*feqW27_T -(f27_B *(q*omegaD-c1o1)-omegaD*feq27_B *(q-c1o1))/(omegaD-c1o1)+f27_T *q)/(q+c1o1);
- q = q_dirNE[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_MM0 ])[ksw ]=(c2o1*feqW27_SW -(f27_NE *(q*omegaD-c1o1)-omegaD*feq27_NE *(q-c1o1))/(omegaD-c1o1)+f27_SW *q)/(q+c1o1);
- q = q_dirSW[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_PP0 ])[kne ]=(c2o1*feqW27_NE -(f27_SW *(q*omegaD-c1o1)-omegaD*feq27_SW *(q-c1o1))/(omegaD-c1o1)+f27_NE *q)/(q+c1o1);
- q = q_dirSE[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_MP0 ])[knw ]=(c2o1*feqW27_NW -(f27_SE *(q*omegaD-c1o1)-omegaD*feq27_SE *(q-c1o1))/(omegaD-c1o1)+f27_NW *q)/(q+c1o1);
- q = q_dirNW[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_PM0 ])[kse ]=(c2o1*feqW27_SE -(f27_NW *(q*omegaD-c1o1)-omegaD*feq27_NW *(q-c1o1))/(omegaD-c1o1)+f27_SE *q)/(q+c1o1);
- q = q_dirTE[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_M0M ])[kbw ]=(c2o1*feqW27_BW -(f27_TE *(q*omegaD-c1o1)-omegaD*feq27_TE *(q-c1o1))/(omegaD-c1o1)+f27_BW *q)/(q+c1o1);
- q = q_dirBW[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_P0P ])[kte ]=(c2o1*feqW27_TE -(f27_BW *(q*omegaD-c1o1)-omegaD*feq27_BW *(q-c1o1))/(omegaD-c1o1)+f27_TE *q)/(q+c1o1);
- q = q_dirBE[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_M0P ])[ktw ]=(c2o1*feqW27_TW -(f27_BE *(q*omegaD-c1o1)-omegaD*feq27_BE *(q-c1o1))/(omegaD-c1o1)+f27_TW *q)/(q+c1o1);
- q = q_dirTW[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_P0M ])[kbe ]=(c2o1*feqW27_BE -(f27_TW *(q*omegaD-c1o1)-omegaD*feq27_TW *(q-c1o1))/(omegaD-c1o1)+f27_BE *q)/(q+c1o1);
- q = q_dirTN[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_0MM ])[kbs ]=(c2o1*feqW27_BS -(f27_TN *(q*omegaD-c1o1)-omegaD*feq27_TN *(q-c1o1))/(omegaD-c1o1)+f27_BS *q)/(q+c1o1);
- q = q_dirBS[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_0PP ])[ktn ]=(c2o1*feqW27_TN -(f27_BS *(q*omegaD-c1o1)-omegaD*feq27_BS *(q-c1o1))/(omegaD-c1o1)+f27_TN *q)/(q+c1o1);
- q = q_dirBN[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_0MP ])[kts ]=(c2o1*feqW27_TS -(f27_BN *(q*omegaD-c1o1)-omegaD*feq27_BN *(q-c1o1))/(omegaD-c1o1)+f27_TS *q)/(q+c1o1);
- q = q_dirTS[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_0PM ])[kbn ]=(c2o1*feqW27_BN -(f27_TS *(q*omegaD-c1o1)-omegaD*feq27_TS *(q-c1o1))/(omegaD-c1o1)+f27_BN *q)/(q+c1o1);
+ q = q_dirE[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_M00])[kw ]=(c2o1*feqW27_W -(f27_E *(q*omegaD-c1o1)-omegaD*feq27_E *(q-c1o1))/(omegaD-c1o1)+f27_W *q)/(q+c1o1);
+ q = q_dirW[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_P00])[ke ]=(c2o1*feqW27_E -(f27_W *(q*omegaD-c1o1)-omegaD*feq27_W *(q-c1o1))/(omegaD-c1o1)+f27_E *q)/(q+c1o1);
+ q = q_dirN[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_0M0])[ks ]=(c2o1*feqW27_S -(f27_N *(q*omegaD-c1o1)-omegaD*feq27_N *(q-c1o1))/(omegaD-c1o1)+f27_S *q)/(q+c1o1);
+ q = q_dirS[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_0P0])[kn ]=(c2o1*feqW27_N -(f27_S *(q*omegaD-c1o1)-omegaD*feq27_S *(q-c1o1))/(omegaD-c1o1)+f27_N *q)/(q+c1o1);
+ q = q_dirT[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_00M])[kb ]=(c2o1*feqW27_B -(f27_T *(q*omegaD-c1o1)-omegaD*feq27_T *(q-c1o1))/(omegaD-c1o1)+f27_B *q)/(q+c1o1);
+ q = q_dirB[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_00P])[kt ]=(c2o1*feqW27_T -(f27_B *(q*omegaD-c1o1)-omegaD*feq27_B *(q-c1o1))/(omegaD-c1o1)+f27_T *q)/(q+c1o1);
+ q = q_dirNE[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_MM0])[ksw ]=(c2o1*feqW27_SW -(f27_NE *(q*omegaD-c1o1)-omegaD*feq27_NE *(q-c1o1))/(omegaD-c1o1)+f27_SW *q)/(q+c1o1);
+ q = q_dirSW[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_PP0])[kne ]=(c2o1*feqW27_NE -(f27_SW *(q*omegaD-c1o1)-omegaD*feq27_SW *(q-c1o1))/(omegaD-c1o1)+f27_NE *q)/(q+c1o1);
+ q = q_dirSE[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_MP0])[knw ]=(c2o1*feqW27_NW -(f27_SE *(q*omegaD-c1o1)-omegaD*feq27_SE *(q-c1o1))/(omegaD-c1o1)+f27_NW *q)/(q+c1o1);
+ q = q_dirNW[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_PM0])[kse ]=(c2o1*feqW27_SE -(f27_NW *(q*omegaD-c1o1)-omegaD*feq27_NW *(q-c1o1))/(omegaD-c1o1)+f27_SE *q)/(q+c1o1);
+ q = q_dirTE[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_M0M])[kbw ]=(c2o1*feqW27_BW -(f27_TE *(q*omegaD-c1o1)-omegaD*feq27_TE *(q-c1o1))/(omegaD-c1o1)+f27_BW *q)/(q+c1o1);
+ q = q_dirBW[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_P0P])[kte ]=(c2o1*feqW27_TE -(f27_BW *(q*omegaD-c1o1)-omegaD*feq27_BW *(q-c1o1))/(omegaD-c1o1)+f27_TE *q)/(q+c1o1);
+ q = q_dirBE[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_M0P])[ktw ]=(c2o1*feqW27_TW -(f27_BE *(q*omegaD-c1o1)-omegaD*feq27_BE *(q-c1o1))/(omegaD-c1o1)+f27_TW *q)/(q+c1o1);
+ q = q_dirTW[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_P0M])[kbe ]=(c2o1*feqW27_BE -(f27_TW *(q*omegaD-c1o1)-omegaD*feq27_TW *(q-c1o1))/(omegaD-c1o1)+f27_BE *q)/(q+c1o1);
+ q = q_dirTN[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_0MM])[kbs ]=(c2o1*feqW27_BS -(f27_TN *(q*omegaD-c1o1)-omegaD*feq27_TN *(q-c1o1))/(omegaD-c1o1)+f27_BS *q)/(q+c1o1);
+ q = q_dirBS[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_0PP])[ktn ]=(c2o1*feqW27_TN -(f27_BS *(q*omegaD-c1o1)-omegaD*feq27_BS *(q-c1o1))/(omegaD-c1o1)+f27_TN *q)/(q+c1o1);
+ q = q_dirBN[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_0MP])[kts ]=(c2o1*feqW27_TS -(f27_BN *(q*omegaD-c1o1)-omegaD*feq27_BN *(q-c1o1))/(omegaD-c1o1)+f27_TS *q)/(q+c1o1);
+ q = q_dirTS[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_0PM])[kbn ]=(c2o1*feqW27_BN -(f27_TS *(q*omegaD-c1o1)-omegaD*feq27_TS *(q-c1o1))/(omegaD-c1o1)+f27_BN *q)/(q+c1o1);
q = q_dirTNE[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_MMM])[kbsw]=(c2o1*feqW27_BSW-(f27_TNE*(q*omegaD-c1o1)-omegaD*feq27_TNE*(q-c1o1))/(omegaD-c1o1)+f27_BSW*q)/(q+c1o1);
q = q_dirBSW[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_PPP])[ktne]=(c2o1*feqW27_TNE-(f27_BSW*(q*omegaD-c1o1)-omegaD*feq27_BSW*(q-c1o1))/(omegaD-c1o1)+f27_TNE*q)/(q+c1o1);
q = q_dirBNE[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_MMP])[ktsw]=(c2o1*feqW27_TSW-(f27_BNE*(q*omegaD-c1o1)-omegaD*feq27_BNE*(q-c1o1))/(omegaD-c1o1)+f27_TSW*q)/(q+c1o1);
@@ -4811,91 +4811,91 @@ __global__ void QADVeloIncomp7(
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep)
{
//Distributions27 D;
//if (isEvenTimestep==true)
//{
- // D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- // D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- // D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- // D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- // D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- // D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- // D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- // D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- // D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- // D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- // D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- // D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- // D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- // D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- // D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- // D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- // D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- // D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- // D.f[DIR_000] = &DD[DIR_000*size_Mat];
- // D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- // D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- // D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- // D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- // D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- // D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- // D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- // D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
+ // D.f[DIR_P00] = &DD[DIR_P00 * size_Mat];
+ // D.f[DIR_M00] = &DD[DIR_M00 * size_Mat];
+ // D.f[DIR_0P0] = &DD[DIR_0P0 * size_Mat];
+ // D.f[DIR_0M0] = &DD[DIR_0M0 * size_Mat];
+ // D.f[DIR_00P] = &DD[DIR_00P * size_Mat];
+ // D.f[DIR_00M] = &DD[DIR_00M * size_Mat];
+ // D.f[DIR_PP0] = &DD[DIR_PP0 * size_Mat];
+ // D.f[DIR_MM0] = &DD[DIR_MM0 * size_Mat];
+ // D.f[DIR_PM0] = &DD[DIR_PM0 * size_Mat];
+ // D.f[DIR_MP0] = &DD[DIR_MP0 * size_Mat];
+ // D.f[DIR_P0P] = &DD[DIR_P0P * size_Mat];
+ // D.f[DIR_M0M] = &DD[DIR_M0M * size_Mat];
+ // D.f[DIR_P0M] = &DD[DIR_P0M * size_Mat];
+ // D.f[DIR_M0P] = &DD[DIR_M0P * size_Mat];
+ // D.f[DIR_0PP] = &DD[DIR_0PP * size_Mat];
+ // D.f[DIR_0MM] = &DD[DIR_0MM * size_Mat];
+ // D.f[DIR_0PM] = &DD[DIR_0PM * size_Mat];
+ // D.f[DIR_0MP] = &DD[DIR_0MP * size_Mat];
+ // D.f[DIR_000] = &DD[DIR_000 * size_Mat];
+ // D.f[DIR_PPP] = &DD[DIR_PPP * size_Mat];
+ // D.f[DIR_MMP] = &DD[DIR_MMP * size_Mat];
+ // D.f[DIR_PMP] = &DD[DIR_PMP * size_Mat];
+ // D.f[DIR_MPP] = &DD[DIR_MPP * size_Mat];
+ // D.f[DIR_PPM] = &DD[DIR_PPM * size_Mat];
+ // D.f[DIR_MMM] = &DD[DIR_MMM * size_Mat];
+ // D.f[DIR_PMM] = &DD[DIR_PMM * size_Mat];
+ // D.f[DIR_MPM] = &DD[DIR_MPM * size_Mat];
//}
//else
//{
- // D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
- // D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
- // D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
- // D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
- // D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
- // D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
- // D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
- // D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
- // D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
- // D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
- // D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
- // D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
- // D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
- // D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
- // D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
- // D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
- // D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
- // D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
- // D.f[DIR_000] = &DD[DIR_000*size_Mat];
- // D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
- // D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
- // D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
- // D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
- // D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
- // D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
- // D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
- // D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
+ // D.f[DIR_M00] = &DD[DIR_P00 * size_Mat];
+ // D.f[DIR_P00] = &DD[DIR_M00 * size_Mat];
+ // D.f[DIR_0M0] = &DD[DIR_0P0 * size_Mat];
+ // D.f[DIR_0P0] = &DD[DIR_0M0 * size_Mat];
+ // D.f[DIR_00M] = &DD[DIR_00P * size_Mat];
+ // D.f[DIR_00P] = &DD[DIR_00M * size_Mat];
+ // D.f[DIR_MM0] = &DD[DIR_PP0 * size_Mat];
+ // D.f[DIR_PP0] = &DD[DIR_MM0 * size_Mat];
+ // D.f[DIR_MP0] = &DD[DIR_PM0 * size_Mat];
+ // D.f[DIR_PM0] = &DD[DIR_MP0 * size_Mat];
+ // D.f[DIR_M0M] = &DD[DIR_P0P * size_Mat];
+ // D.f[DIR_P0P] = &DD[DIR_M0M * size_Mat];
+ // D.f[DIR_M0P] = &DD[DIR_P0M * size_Mat];
+ // D.f[DIR_P0M] = &DD[DIR_M0P * size_Mat];
+ // D.f[DIR_0MM] = &DD[DIR_0PP * size_Mat];
+ // D.f[DIR_0PP] = &DD[DIR_0MM * size_Mat];
+ // D.f[DIR_0MP] = &DD[DIR_0PM * size_Mat];
+ // D.f[DIR_0PM] = &DD[DIR_0MP * size_Mat];
+ // D.f[DIR_000] = &DD[DIR_000 * size_Mat];
+ // D.f[DIR_PPP] = &DD[DIR_MMM * size_Mat];
+ // D.f[DIR_MMP] = &DD[DIR_PPM * size_Mat];
+ // D.f[DIR_PMP] = &DD[DIR_MPM * size_Mat];
+ // D.f[DIR_MPP] = &DD[DIR_PMM * size_Mat];
+ // D.f[DIR_PPM] = &DD[DIR_MMP * size_Mat];
+ // D.f[DIR_MMM] = &DD[DIR_PPP * size_Mat];
+ // D.f[DIR_PMM] = &DD[DIR_MPP * size_Mat];
+ // D.f[DIR_MPM] = &DD[DIR_PMP * size_Mat];
//}
Distributions7 D7;
if (isEvenTimestep==true)
{
- D7.f[0] = &DD7[0*size_Mat];
- D7.f[1] = &DD7[1*size_Mat];
- D7.f[2] = &DD7[2*size_Mat];
- D7.f[3] = &DD7[3*size_Mat];
- D7.f[4] = &DD7[4*size_Mat];
- D7.f[5] = &DD7[5*size_Mat];
- D7.f[6] = &DD7[6*size_Mat];
+ D7.f[0] = &DD7[0*numberOfLBnodes];
+ D7.f[1] = &DD7[1*numberOfLBnodes];
+ D7.f[2] = &DD7[2*numberOfLBnodes];
+ D7.f[3] = &DD7[3*numberOfLBnodes];
+ D7.f[4] = &DD7[4*numberOfLBnodes];
+ D7.f[5] = &DD7[5*numberOfLBnodes];
+ D7.f[6] = &DD7[6*numberOfLBnodes];
}
else
{
- D7.f[0] = &DD7[0*size_Mat];
- D7.f[2] = &DD7[1*size_Mat];
- D7.f[1] = &DD7[2*size_Mat];
- D7.f[4] = &DD7[3*size_Mat];
- D7.f[3] = &DD7[4*size_Mat];
- D7.f[6] = &DD7[5*size_Mat];
- D7.f[5] = &DD7[6*size_Mat];
+ D7.f[0] = &DD7[0*numberOfLBnodes];
+ D7.f[2] = &DD7[1*numberOfLBnodes];
+ D7.f[1] = &DD7[2*numberOfLBnodes];
+ D7.f[4] = &DD7[3*numberOfLBnodes];
+ D7.f[3] = &DD7[4*numberOfLBnodes];
+ D7.f[6] = &DD7[5*numberOfLBnodes];
+ D7.f[5] = &DD7[6*numberOfLBnodes];
}
@@ -4915,12 +4915,12 @@ __global__ void QADVeloIncomp7(
//////////////////////////////////////////////////////////////////////////////////
real *q_dirE, *q_dirW, *q_dirN, *q_dirS, *q_dirT, *q_dirB;
- q_dirE = &QQ[DIR_P00 * numberOfBCnodes];
- q_dirW = &QQ[DIR_M00 * numberOfBCnodes];
- q_dirN = &QQ[DIR_0P0 * numberOfBCnodes];
- q_dirS = &QQ[DIR_0M0 * numberOfBCnodes];
- q_dirT = &QQ[DIR_00P * numberOfBCnodes];
- q_dirB = &QQ[DIR_00M * numberOfBCnodes];
+ q_dirE = &QQ[DIR_P00 * numberOfBCnodes];
+ q_dirW = &QQ[DIR_M00 * numberOfBCnodes];
+ q_dirN = &QQ[DIR_0P0 * numberOfBCnodes];
+ q_dirS = &QQ[DIR_0M0 * numberOfBCnodes];
+ q_dirT = &QQ[DIR_00P * numberOfBCnodes];
+ q_dirB = &QQ[DIR_00M * numberOfBCnodes];
//////////////////////////////////////////////////////////////////////////////////
//index
unsigned int KQK = k_Q[k];
@@ -4952,32 +4952,32 @@ __global__ void QADVeloIncomp7(
//unsigned int ktne = KQK;
//unsigned int kbsw = neighborZ[ksw];
////////////////////////////////////////////////////////////////////////////////
- //real f_W = (D.f[DIR_P00 ])[ke ];
- //real f_E = (D.f[DIR_M00 ])[kw ];
- //real f_S = (D.f[DIR_0P0 ])[kn ];
- //real f_N = (D.f[DIR_0M0 ])[ks ];
- //real f_B = (D.f[DIR_00P ])[kt ];
- //real f_T = (D.f[DIR_00M ])[kb ];
- //real f_SW = (D.f[DIR_PP0 ])[kne ];
- //real f_NE = (D.f[DIR_MM0 ])[ksw ];
- //real f_NW = (D.f[DIR_PM0 ])[kse ];
- //real f_SE = (D.f[DIR_MP0 ])[knw ];
- //real f_BW = (D.f[DIR_P0P ])[kte ];
- //real f_TE = (D.f[DIR_M0M ])[kbw ];
- //real f_TW = (D.f[DIR_P0M ])[kbe ];
- //real f_BE = (D.f[DIR_M0P ])[ktw ];
- //real f_BS = (D.f[DIR_0PP ])[ktn ];
- //real f_TN = (D.f[DIR_0MM ])[kbs ];
- //real f_TS = (D.f[DIR_0PM ])[kbn ];
- //real f_BN = (D.f[DIR_0MP ])[kts ];
- //real f_BSW = (D.f[DIR_PPP ])[ktne ];
- //real f_BNE = (D.f[DIR_MMP ])[ktsw ];
- //real f_BNW = (D.f[DIR_PMP ])[ktse ];
- //real f_BSE = (D.f[DIR_MPP ])[ktnw ];
- //real f_TSW = (D.f[DIR_PPM ])[kbne ];
- //real f_TNE = (D.f[DIR_MMM ])[kbsw ];
- //real f_TNW = (D.f[DIR_PMM ])[kbse ];
- //real f_TSE = (D.f[DIR_MPM ])[kbnw ];
+ //real f_W = (D.f[DIR_P00])[ke ];
+ //real f_E = (D.f[DIR_M00])[kw ];
+ //real f_S = (D.f[DIR_0P0])[kn ];
+ //real f_N = (D.f[DIR_0M0])[ks ];
+ //real f_B = (D.f[DIR_00P])[kt ];
+ //real f_T = (D.f[DIR_00M])[kb ];
+ //real f_SW = (D.f[DIR_PP0])[kne ];
+ //real f_NE = (D.f[DIR_MM0])[ksw ];
+ //real f_NW = (D.f[DIR_PM0])[kse ];
+ //real f_SE = (D.f[DIR_MP0])[knw ];
+ //real f_BW = (D.f[DIR_P0P])[kte ];
+ //real f_TE = (D.f[DIR_M0M])[kbw ];
+ //real f_TW = (D.f[DIR_P0M])[kbe ];
+ //real f_BE = (D.f[DIR_M0P])[ktw ];
+ //real f_BS = (D.f[DIR_0PP])[ktn ];
+ //real f_TN = (D.f[DIR_0MM])[kbs ];
+ //real f_TS = (D.f[DIR_0PM])[kbn ];
+ //real f_BN = (D.f[DIR_0MP])[kts ];
+ //real f_BSW = (D.f[DIR_PPP])[ktne ];
+ //real f_BNE = (D.f[DIR_MMP])[ktsw ];
+ //real f_BNW = (D.f[DIR_PMP])[ktse ];
+ //real f_BSE = (D.f[DIR_MPP])[ktnw ];
+ //real f_TSW = (D.f[DIR_PPM])[kbne ];
+ //real f_TNE = (D.f[DIR_MMM])[kbsw ];
+ //real f_TNW = (D.f[DIR_PMM])[kbse ];
+ //real f_TSE = (D.f[DIR_MPM])[kbnw ];
////////////////////////////////////////////////////////////////////////////////
//real vx1_Inflow = c0o1;
//real vx2_Inflow = velo[k];
@@ -5091,23 +5091,23 @@ __global__ void QADVeloIncomp7(
//pointertausch
if (isEvenTimestep==false)
{
- D7.f[0] = &DD7[0*size_Mat];
- D7.f[1] = &DD7[1*size_Mat];
- D7.f[2] = &DD7[2*size_Mat];
- D7.f[3] = &DD7[3*size_Mat];
- D7.f[4] = &DD7[4*size_Mat];
- D7.f[5] = &DD7[5*size_Mat];
- D7.f[6] = &DD7[6*size_Mat];
+ D7.f[0] = &DD7[0*numberOfLBnodes];
+ D7.f[1] = &DD7[1*numberOfLBnodes];
+ D7.f[2] = &DD7[2*numberOfLBnodes];
+ D7.f[3] = &DD7[3*numberOfLBnodes];
+ D7.f[4] = &DD7[4*numberOfLBnodes];
+ D7.f[5] = &DD7[5*numberOfLBnodes];
+ D7.f[6] = &DD7[6*numberOfLBnodes];
}
else
{
- D7.f[0] = &DD7[0*size_Mat];
- D7.f[2] = &DD7[1*size_Mat];
- D7.f[1] = &DD7[2*size_Mat];
- D7.f[4] = &DD7[3*size_Mat];
- D7.f[3] = &DD7[4*size_Mat];
- D7.f[6] = &DD7[5*size_Mat];
- D7.f[5] = &DD7[6*size_Mat];
+ D7.f[0] = &DD7[0*numberOfLBnodes];
+ D7.f[2] = &DD7[1*numberOfLBnodes];
+ D7.f[1] = &DD7[2*numberOfLBnodes];
+ D7.f[4] = &DD7[3*numberOfLBnodes];
+ D7.f[3] = &DD7[4*numberOfLBnodes];
+ D7.f[6] = &DD7[5*numberOfLBnodes];
+ D7.f[5] = &DD7[6*numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////
@@ -5289,131 +5289,131 @@ __global__ void QADVeloIncomp27(
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep)
{
Distributions27 D;
if (isEvenTimestep==true)
{
- D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
+ D.f[DIR_M00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_MPM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_PMP * numberOfLBnodes];
}
Distributions27 D27;
if (isEvenTimestep==true)
{
- D27.f[DIR_P00 ] = &DD27[DIR_P00 *size_Mat];
- D27.f[DIR_M00 ] = &DD27[DIR_M00 *size_Mat];
- D27.f[DIR_0P0 ] = &DD27[DIR_0P0 *size_Mat];
- D27.f[DIR_0M0 ] = &DD27[DIR_0M0 *size_Mat];
- D27.f[DIR_00P ] = &DD27[DIR_00P *size_Mat];
- D27.f[DIR_00M ] = &DD27[DIR_00M *size_Mat];
- D27.f[DIR_PP0 ] = &DD27[DIR_PP0 *size_Mat];
- D27.f[DIR_MM0 ] = &DD27[DIR_MM0 *size_Mat];
- D27.f[DIR_PM0 ] = &DD27[DIR_PM0 *size_Mat];
- D27.f[DIR_MP0 ] = &DD27[DIR_MP0 *size_Mat];
- D27.f[DIR_P0P ] = &DD27[DIR_P0P *size_Mat];
- D27.f[DIR_M0M ] = &DD27[DIR_M0M *size_Mat];
- D27.f[DIR_P0M ] = &DD27[DIR_P0M *size_Mat];
- D27.f[DIR_M0P ] = &DD27[DIR_M0P *size_Mat];
- D27.f[DIR_0PP ] = &DD27[DIR_0PP *size_Mat];
- D27.f[DIR_0MM ] = &DD27[DIR_0MM *size_Mat];
- D27.f[DIR_0PM ] = &DD27[DIR_0PM *size_Mat];
- D27.f[DIR_0MP ] = &DD27[DIR_0MP *size_Mat];
- D27.f[DIR_000] = &DD27[DIR_000*size_Mat];
- D27.f[DIR_PPP ] = &DD27[DIR_PPP *size_Mat];
- D27.f[DIR_MMP ] = &DD27[DIR_MMP *size_Mat];
- D27.f[DIR_PMP ] = &DD27[DIR_PMP *size_Mat];
- D27.f[DIR_MPP ] = &DD27[DIR_MPP *size_Mat];
- D27.f[DIR_PPM ] = &DD27[DIR_PPM *size_Mat];
- D27.f[DIR_MMM ] = &DD27[DIR_MMM *size_Mat];
- D27.f[DIR_PMM ] = &DD27[DIR_PMM *size_Mat];
- D27.f[DIR_MPM ] = &DD27[DIR_MPM *size_Mat];
+ D27.f[DIR_P00] = &DD27[DIR_P00 * numberOfLBnodes];
+ D27.f[DIR_M00] = &DD27[DIR_M00 * numberOfLBnodes];
+ D27.f[DIR_0P0] = &DD27[DIR_0P0 * numberOfLBnodes];
+ D27.f[DIR_0M0] = &DD27[DIR_0M0 * numberOfLBnodes];
+ D27.f[DIR_00P] = &DD27[DIR_00P * numberOfLBnodes];
+ D27.f[DIR_00M] = &DD27[DIR_00M * numberOfLBnodes];
+ D27.f[DIR_PP0] = &DD27[DIR_PP0 * numberOfLBnodes];
+ D27.f[DIR_MM0] = &DD27[DIR_MM0 * numberOfLBnodes];
+ D27.f[DIR_PM0] = &DD27[DIR_PM0 * numberOfLBnodes];
+ D27.f[DIR_MP0] = &DD27[DIR_MP0 * numberOfLBnodes];
+ D27.f[DIR_P0P] = &DD27[DIR_P0P * numberOfLBnodes];
+ D27.f[DIR_M0M] = &DD27[DIR_M0M * numberOfLBnodes];
+ D27.f[DIR_P0M] = &DD27[DIR_P0M * numberOfLBnodes];
+ D27.f[DIR_M0P] = &DD27[DIR_M0P * numberOfLBnodes];
+ D27.f[DIR_0PP] = &DD27[DIR_0PP * numberOfLBnodes];
+ D27.f[DIR_0MM] = &DD27[DIR_0MM * numberOfLBnodes];
+ D27.f[DIR_0PM] = &DD27[DIR_0PM * numberOfLBnodes];
+ D27.f[DIR_0MP] = &DD27[DIR_0MP * numberOfLBnodes];
+ D27.f[DIR_000] = &DD27[DIR_000 * numberOfLBnodes];
+ D27.f[DIR_PPP] = &DD27[DIR_PPP * numberOfLBnodes];
+ D27.f[DIR_MMP] = &DD27[DIR_MMP * numberOfLBnodes];
+ D27.f[DIR_PMP] = &DD27[DIR_PMP * numberOfLBnodes];
+ D27.f[DIR_MPP] = &DD27[DIR_MPP * numberOfLBnodes];
+ D27.f[DIR_PPM] = &DD27[DIR_PPM * numberOfLBnodes];
+ D27.f[DIR_MMM] = &DD27[DIR_MMM * numberOfLBnodes];
+ D27.f[DIR_PMM] = &DD27[DIR_PMM * numberOfLBnodes];
+ D27.f[DIR_MPM] = &DD27[DIR_MPM * numberOfLBnodes];
}
else
{
- D27.f[DIR_M00 ] = &DD27[DIR_P00 *size_Mat];
- D27.f[DIR_P00 ] = &DD27[DIR_M00 *size_Mat];
- D27.f[DIR_0M0 ] = &DD27[DIR_0P0 *size_Mat];
- D27.f[DIR_0P0 ] = &DD27[DIR_0M0 *size_Mat];
- D27.f[DIR_00M ] = &DD27[DIR_00P *size_Mat];
- D27.f[DIR_00P ] = &DD27[DIR_00M *size_Mat];
- D27.f[DIR_MM0 ] = &DD27[DIR_PP0 *size_Mat];
- D27.f[DIR_PP0 ] = &DD27[DIR_MM0 *size_Mat];
- D27.f[DIR_MP0 ] = &DD27[DIR_PM0 *size_Mat];
- D27.f[DIR_PM0 ] = &DD27[DIR_MP0 *size_Mat];
- D27.f[DIR_M0M ] = &DD27[DIR_P0P *size_Mat];
- D27.f[DIR_P0P ] = &DD27[DIR_M0M *size_Mat];
- D27.f[DIR_M0P ] = &DD27[DIR_P0M *size_Mat];
- D27.f[DIR_P0M ] = &DD27[DIR_M0P *size_Mat];
- D27.f[DIR_0MM ] = &DD27[DIR_0PP *size_Mat];
- D27.f[DIR_0PP ] = &DD27[DIR_0MM *size_Mat];
- D27.f[DIR_0MP ] = &DD27[DIR_0PM *size_Mat];
- D27.f[DIR_0PM ] = &DD27[DIR_0MP *size_Mat];
- D27.f[DIR_000] = &DD27[DIR_000*size_Mat];
- D27.f[DIR_PPP ] = &DD27[DIR_MMM *size_Mat];
- D27.f[DIR_MMP ] = &DD27[DIR_PPM *size_Mat];
- D27.f[DIR_PMP ] = &DD27[DIR_MPM *size_Mat];
- D27.f[DIR_MPP ] = &DD27[DIR_PMM *size_Mat];
- D27.f[DIR_PPM ] = &DD27[DIR_MMP *size_Mat];
- D27.f[DIR_MMM ] = &DD27[DIR_PPP *size_Mat];
- D27.f[DIR_PMM ] = &DD27[DIR_MPP *size_Mat];
- D27.f[DIR_MPM ] = &DD27[DIR_PMP *size_Mat];
+ D27.f[DIR_M00] = &DD27[DIR_P00 * numberOfLBnodes];
+ D27.f[DIR_P00] = &DD27[DIR_M00 * numberOfLBnodes];
+ D27.f[DIR_0M0] = &DD27[DIR_0P0 * numberOfLBnodes];
+ D27.f[DIR_0P0] = &DD27[DIR_0M0 * numberOfLBnodes];
+ D27.f[DIR_00M] = &DD27[DIR_00P * numberOfLBnodes];
+ D27.f[DIR_00P] = &DD27[DIR_00M * numberOfLBnodes];
+ D27.f[DIR_MM0] = &DD27[DIR_PP0 * numberOfLBnodes];
+ D27.f[DIR_PP0] = &DD27[DIR_MM0 * numberOfLBnodes];
+ D27.f[DIR_MP0] = &DD27[DIR_PM0 * numberOfLBnodes];
+ D27.f[DIR_PM0] = &DD27[DIR_MP0 * numberOfLBnodes];
+ D27.f[DIR_M0M] = &DD27[DIR_P0P * numberOfLBnodes];
+ D27.f[DIR_P0P] = &DD27[DIR_M0M * numberOfLBnodes];
+ D27.f[DIR_M0P] = &DD27[DIR_P0M * numberOfLBnodes];
+ D27.f[DIR_P0M] = &DD27[DIR_M0P * numberOfLBnodes];
+ D27.f[DIR_0MM] = &DD27[DIR_0PP * numberOfLBnodes];
+ D27.f[DIR_0PP] = &DD27[DIR_0MM * numberOfLBnodes];
+ D27.f[DIR_0MP] = &DD27[DIR_0PM * numberOfLBnodes];
+ D27.f[DIR_0PM] = &DD27[DIR_0MP * numberOfLBnodes];
+ D27.f[DIR_000] = &DD27[DIR_000 * numberOfLBnodes];
+ D27.f[DIR_PPP] = &DD27[DIR_MMM * numberOfLBnodes];
+ D27.f[DIR_MMP] = &DD27[DIR_PPM * numberOfLBnodes];
+ D27.f[DIR_PMP] = &DD27[DIR_MPM * numberOfLBnodes];
+ D27.f[DIR_MPP] = &DD27[DIR_PMM * numberOfLBnodes];
+ D27.f[DIR_PPM] = &DD27[DIR_MMP * numberOfLBnodes];
+ D27.f[DIR_MMM] = &DD27[DIR_PPP * numberOfLBnodes];
+ D27.f[DIR_PMM] = &DD27[DIR_MPP * numberOfLBnodes];
+ D27.f[DIR_MPM] = &DD27[DIR_PMP * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////
const unsigned x = threadIdx.x; // Globaler x-Index
@@ -5434,24 +5434,24 @@ __global__ void QADVeloIncomp27(
*q_dirBE, *q_dirTW, *q_dirTN, *q_dirBS, *q_dirBN, *q_dirTS,
*q_dirTNE, *q_dirTSW, *q_dirTSE, *q_dirTNW, *q_dirBNE, *q_dirBSW,
*q_dirBSE, *q_dirBNW;
- q_dirE = &QQ[DIR_P00 * numberOfBCnodes];
- q_dirW = &QQ[DIR_M00 * numberOfBCnodes];
- q_dirN = &QQ[DIR_0P0 * numberOfBCnodes];
- q_dirS = &QQ[DIR_0M0 * numberOfBCnodes];
- q_dirT = &QQ[DIR_00P * numberOfBCnodes];
- q_dirB = &QQ[DIR_00M * numberOfBCnodes];
- q_dirNE = &QQ[DIR_PP0 * numberOfBCnodes];
- q_dirSW = &QQ[DIR_MM0 * numberOfBCnodes];
- q_dirSE = &QQ[DIR_PM0 * numberOfBCnodes];
- q_dirNW = &QQ[DIR_MP0 * numberOfBCnodes];
- q_dirTE = &QQ[DIR_P0P * numberOfBCnodes];
- q_dirBW = &QQ[DIR_M0M * numberOfBCnodes];
- q_dirBE = &QQ[DIR_P0M * numberOfBCnodes];
- q_dirTW = &QQ[DIR_M0P * numberOfBCnodes];
- q_dirTN = &QQ[DIR_0PP * numberOfBCnodes];
- q_dirBS = &QQ[DIR_0MM * numberOfBCnodes];
- q_dirBN = &QQ[DIR_0PM * numberOfBCnodes];
- q_dirTS = &QQ[DIR_0MP * numberOfBCnodes];
+ q_dirE = &QQ[DIR_P00 * numberOfBCnodes];
+ q_dirW = &QQ[DIR_M00 * numberOfBCnodes];
+ q_dirN = &QQ[DIR_0P0 * numberOfBCnodes];
+ q_dirS = &QQ[DIR_0M0 * numberOfBCnodes];
+ q_dirT = &QQ[DIR_00P * numberOfBCnodes];
+ q_dirB = &QQ[DIR_00M * numberOfBCnodes];
+ q_dirNE = &QQ[DIR_PP0 * numberOfBCnodes];
+ q_dirSW = &QQ[DIR_MM0 * numberOfBCnodes];
+ q_dirSE = &QQ[DIR_PM0 * numberOfBCnodes];
+ q_dirNW = &QQ[DIR_MP0 * numberOfBCnodes];
+ q_dirTE = &QQ[DIR_P0P * numberOfBCnodes];
+ q_dirBW = &QQ[DIR_M0M * numberOfBCnodes];
+ q_dirBE = &QQ[DIR_P0M * numberOfBCnodes];
+ q_dirTW = &QQ[DIR_M0P * numberOfBCnodes];
+ q_dirTN = &QQ[DIR_0PP * numberOfBCnodes];
+ q_dirBS = &QQ[DIR_0MM * numberOfBCnodes];
+ q_dirBN = &QQ[DIR_0PM * numberOfBCnodes];
+ q_dirTS = &QQ[DIR_0MP * numberOfBCnodes];
q_dirTNE = &QQ[DIR_PPP * numberOfBCnodes];
q_dirTSW = &QQ[DIR_MMP * numberOfBCnodes];
q_dirTSE = &QQ[DIR_PMP * numberOfBCnodes];
@@ -5491,65 +5491,65 @@ __global__ void QADVeloIncomp27(
unsigned int ktne = KQK;
unsigned int kbsw = neighborZ[ksw];
////////////////////////////////////////////////////////////////////////////////
- real f_W = (D.f[DIR_P00 ])[ke ];
- real f_E = (D.f[DIR_M00 ])[kw ];
- real f_S = (D.f[DIR_0P0 ])[kn ];
- real f_N = (D.f[DIR_0M0 ])[ks ];
- real f_B = (D.f[DIR_00P ])[kt ];
- real f_T = (D.f[DIR_00M ])[kb ];
- real f_SW = (D.f[DIR_PP0 ])[kne ];
- real f_NE = (D.f[DIR_MM0 ])[ksw ];
- real f_NW = (D.f[DIR_PM0 ])[kse ];
- real f_SE = (D.f[DIR_MP0 ])[knw ];
- real f_BW = (D.f[DIR_P0P ])[kte ];
- real f_TE = (D.f[DIR_M0M ])[kbw ];
- real f_TW = (D.f[DIR_P0M ])[kbe ];
- real f_BE = (D.f[DIR_M0P ])[ktw ];
- real f_BS = (D.f[DIR_0PP ])[ktn ];
- real f_TN = (D.f[DIR_0MM ])[kbs ];
- real f_TS = (D.f[DIR_0PM ])[kbn ];
- real f_BN = (D.f[DIR_0MP ])[kts ];
+ real f_W = (D.f[DIR_P00])[ke ];
+ real f_E = (D.f[DIR_M00])[kw ];
+ real f_S = (D.f[DIR_0P0])[kn ];
+ real f_N = (D.f[DIR_0M0])[ks ];
+ real f_B = (D.f[DIR_00P])[kt ];
+ real f_T = (D.f[DIR_00M])[kb ];
+ real f_SW = (D.f[DIR_PP0])[kne ];
+ real f_NE = (D.f[DIR_MM0])[ksw ];
+ real f_NW = (D.f[DIR_PM0])[kse ];
+ real f_SE = (D.f[DIR_MP0])[knw ];
+ real f_BW = (D.f[DIR_P0P])[kte ];
+ real f_TE = (D.f[DIR_M0M])[kbw ];
+ real f_TW = (D.f[DIR_P0M])[kbe ];
+ real f_BE = (D.f[DIR_M0P])[ktw ];
+ real f_BS = (D.f[DIR_0PP])[ktn ];
+ real f_TN = (D.f[DIR_0MM])[kbs ];
+ real f_TS = (D.f[DIR_0PM])[kbn ];
+ real f_BN = (D.f[DIR_0MP])[kts ];
//real f_ZERO = (D.f[DIR_000])[kzero];
- real f_BSW = (D.f[DIR_PPP ])[ktne ];
- real f_BNE = (D.f[DIR_MMP ])[ktsw ];
- real f_BNW = (D.f[DIR_PMP ])[ktse ];
- real f_BSE = (D.f[DIR_MPP ])[ktnw ];
- real f_TSW = (D.f[DIR_PPM ])[kbne ];
- real f_TNE = (D.f[DIR_MMM ])[kbsw ];
- real f_TNW = (D.f[DIR_PMM ])[kbse ];
- real f_TSE = (D.f[DIR_MPM ])[kbnw ];
+ real f_BSW = (D.f[DIR_PPP])[ktne ];
+ real f_BNE = (D.f[DIR_MMP])[ktsw ];
+ real f_BNW = (D.f[DIR_PMP])[ktse ];
+ real f_BSE = (D.f[DIR_MPP])[ktnw ];
+ real f_TSW = (D.f[DIR_PPM])[kbne ];
+ real f_TNE = (D.f[DIR_MMM])[kbsw ];
+ real f_TNW = (D.f[DIR_PMM])[kbse ];
+ real f_TSE = (D.f[DIR_MPM])[kbnw ];
////////////////////////////////////////////////////////////////////////////////
real vx1 = ((f_TNE-f_BSW)+(f_BNE-f_TSW)+(f_TSE-f_BNW)+(f_BSE-f_TNW) +(f_NE-f_SW)+(f_SE-f_NW)+(f_TE-f_BW)+(f_BE-f_TW)+(f_E-f_W));
real vx2 = ((f_TNE-f_BSW)+(f_BNE-f_TSW)+(f_BNW-f_TSE)+(f_TNW-f_BSE) +(f_NE-f_SW)+(f_NW-f_SE)+(f_TN-f_BS)+(f_BN-f_TS)+(f_N-f_S));
real vx3 = ((f_TNE-f_BSW)+(f_TSW-f_BNE)+(f_TSE-f_BNW)+(f_TNW-f_BSE) +(f_TE-f_BW)+(f_TW-f_BE)+(f_TN-f_BS)+(f_TS-f_BN)+(f_T-f_B));
////////////////////////////////////////////////////////////////////////////////
- //real f27_W = (D27.f[DIR_P00 ])[ke ];
- //real f27_E = (D27.f[DIR_M00 ])[kw ];
- //real f27_S = (D27.f[DIR_0P0 ])[kn ];
- //real f27_N = (D27.f[DIR_0M0 ])[ks ];
- //real f27_B = (D27.f[DIR_00P ])[kt ];
- //real f27_T = (D27.f[DIR_00M ])[kb ];
- //real f27_SW = (D27.f[DIR_PP0 ])[kne ];
- //real f27_NE = (D27.f[DIR_MM0 ])[ksw ];
- //real f27_NW = (D27.f[DIR_PM0 ])[kse ];
- //real f27_SE = (D27.f[DIR_MP0 ])[knw ];
- //real f27_BW = (D27.f[DIR_P0P ])[kte ];
- //real f27_TE = (D27.f[DIR_M0M ])[kbw ];
- //real f27_TW = (D27.f[DIR_P0M ])[kbe ];
- //real f27_BE = (D27.f[DIR_M0P ])[ktw ];
- //real f27_BS = (D27.f[DIR_0PP ])[ktn ];
- //real f27_TN = (D27.f[DIR_0MM ])[kbs ];
- //real f27_TS = (D27.f[DIR_0PM ])[kbn ];
- //real f27_BN = (D27.f[DIR_0MP ])[kts ];
+ //real f27_W = (D27.f[DIR_P00])[ke ];
+ //real f27_E = (D27.f[DIR_M00])[kw ];
+ //real f27_S = (D27.f[DIR_0P0])[kn ];
+ //real f27_N = (D27.f[DIR_0M0])[ks ];
+ //real f27_B = (D27.f[DIR_00P])[kt ];
+ //real f27_T = (D27.f[DIR_00M])[kb ];
+ //real f27_SW = (D27.f[DIR_PP0])[kne ];
+ //real f27_NE = (D27.f[DIR_MM0])[ksw ];
+ //real f27_NW = (D27.f[DIR_PM0])[kse ];
+ //real f27_SE = (D27.f[DIR_MP0])[knw ];
+ //real f27_BW = (D27.f[DIR_P0P])[kte ];
+ //real f27_TE = (D27.f[DIR_M0M])[kbw ];
+ //real f27_TW = (D27.f[DIR_P0M])[kbe ];
+ //real f27_BE = (D27.f[DIR_M0P])[ktw ];
+ //real f27_BS = (D27.f[DIR_0PP])[ktn ];
+ //real f27_TN = (D27.f[DIR_0MM])[kbs ];
+ //real f27_TS = (D27.f[DIR_0PM])[kbn ];
+ //real f27_BN = (D27.f[DIR_0MP])[kts ];
//real f27_ZERO = (D27.f[DIR_000])[kzero];
- //real f27_BSW = (D27.f[DIR_PPP ])[ktne ];
- //real f27_BNE = (D27.f[DIR_MMP ])[ktsw ];
- //real f27_BNW = (D27.f[DIR_PMP ])[ktse ];
- //real f27_BSE = (D27.f[DIR_MPP ])[ktnw ];
- //real f27_TSW = (D27.f[DIR_PPM ])[kbne ];
- //real f27_TNE = (D27.f[DIR_MMM ])[kbsw ];
- //real f27_TNW = (D27.f[DIR_PMM ])[kbse ];
- //real f27_TSE = (D27.f[DIR_MPM ])[kbnw ];
+ //real f27_BSW = (D27.f[DIR_PPP])[ktne ];
+ //real f27_BNE = (D27.f[DIR_MMP])[ktsw ];
+ //real f27_BNW = (D27.f[DIR_PMP])[ktse ];
+ //real f27_BSE = (D27.f[DIR_MPP])[ktnw ];
+ //real f27_TSW = (D27.f[DIR_PPM])[kbne ];
+ //real f27_TNE = (D27.f[DIR_MMM])[kbsw ];
+ //real f27_TNW = (D27.f[DIR_PMM])[kbse ];
+ //real f27_TSE = (D27.f[DIR_MPM])[kbnw ];
////////////////////////////////////////////////////////////////////////////////
real cu_sq=c3o2*(vx1*vx1+vx2*vx2+vx3*vx3);
////////////////////////////////////////////////////////////////////////////////
@@ -5630,63 +5630,63 @@ __global__ void QADVeloIncomp27(
//////////////////////////////////////////////////////////////////////////
if (isEvenTimestep==false)
{
- D27.f[DIR_P00 ] = &DD27[DIR_P00 *size_Mat];
- D27.f[DIR_M00 ] = &DD27[DIR_M00 *size_Mat];
- D27.f[DIR_0P0 ] = &DD27[DIR_0P0 *size_Mat];
- D27.f[DIR_0M0 ] = &DD27[DIR_0M0 *size_Mat];
- D27.f[DIR_00P ] = &DD27[DIR_00P *size_Mat];
- D27.f[DIR_00M ] = &DD27[DIR_00M *size_Mat];
- D27.f[DIR_PP0 ] = &DD27[DIR_PP0 *size_Mat];
- D27.f[DIR_MM0 ] = &DD27[DIR_MM0 *size_Mat];
- D27.f[DIR_PM0 ] = &DD27[DIR_PM0 *size_Mat];
- D27.f[DIR_MP0 ] = &DD27[DIR_MP0 *size_Mat];
- D27.f[DIR_P0P ] = &DD27[DIR_P0P *size_Mat];
- D27.f[DIR_M0M ] = &DD27[DIR_M0M *size_Mat];
- D27.f[DIR_P0M ] = &DD27[DIR_P0M *size_Mat];
- D27.f[DIR_M0P ] = &DD27[DIR_M0P *size_Mat];
- D27.f[DIR_0PP ] = &DD27[DIR_0PP *size_Mat];
- D27.f[DIR_0MM ] = &DD27[DIR_0MM *size_Mat];
- D27.f[DIR_0PM ] = &DD27[DIR_0PM *size_Mat];
- D27.f[DIR_0MP ] = &DD27[DIR_0MP *size_Mat];
- D27.f[DIR_000] = &DD27[DIR_000*size_Mat];
- D27.f[DIR_PPP ] = &DD27[DIR_PPP *size_Mat];
- D27.f[DIR_MMP ] = &DD27[DIR_MMP *size_Mat];
- D27.f[DIR_PMP ] = &DD27[DIR_PMP *size_Mat];
- D27.f[DIR_MPP ] = &DD27[DIR_MPP *size_Mat];
- D27.f[DIR_PPM ] = &DD27[DIR_PPM *size_Mat];
- D27.f[DIR_MMM ] = &DD27[DIR_MMM *size_Mat];
- D27.f[DIR_PMM ] = &DD27[DIR_PMM *size_Mat];
- D27.f[DIR_MPM ] = &DD27[DIR_MPM *size_Mat];
+ D27.f[DIR_P00] = &DD27[DIR_P00 * numberOfLBnodes];
+ D27.f[DIR_M00] = &DD27[DIR_M00 * numberOfLBnodes];
+ D27.f[DIR_0P0] = &DD27[DIR_0P0 * numberOfLBnodes];
+ D27.f[DIR_0M0] = &DD27[DIR_0M0 * numberOfLBnodes];
+ D27.f[DIR_00P] = &DD27[DIR_00P * numberOfLBnodes];
+ D27.f[DIR_00M] = &DD27[DIR_00M * numberOfLBnodes];
+ D27.f[DIR_PP0] = &DD27[DIR_PP0 * numberOfLBnodes];
+ D27.f[DIR_MM0] = &DD27[DIR_MM0 * numberOfLBnodes];
+ D27.f[DIR_PM0] = &DD27[DIR_PM0 * numberOfLBnodes];
+ D27.f[DIR_MP0] = &DD27[DIR_MP0 * numberOfLBnodes];
+ D27.f[DIR_P0P] = &DD27[DIR_P0P * numberOfLBnodes];
+ D27.f[DIR_M0M] = &DD27[DIR_M0M * numberOfLBnodes];
+ D27.f[DIR_P0M] = &DD27[DIR_P0M * numberOfLBnodes];
+ D27.f[DIR_M0P] = &DD27[DIR_M0P * numberOfLBnodes];
+ D27.f[DIR_0PP] = &DD27[DIR_0PP * numberOfLBnodes];
+ D27.f[DIR_0MM] = &DD27[DIR_0MM * numberOfLBnodes];
+ D27.f[DIR_0PM] = &DD27[DIR_0PM * numberOfLBnodes];
+ D27.f[DIR_0MP] = &DD27[DIR_0MP * numberOfLBnodes];
+ D27.f[DIR_000] = &DD27[DIR_000 * numberOfLBnodes];
+ D27.f[DIR_PPP] = &DD27[DIR_PPP * numberOfLBnodes];
+ D27.f[DIR_MMP] = &DD27[DIR_MMP * numberOfLBnodes];
+ D27.f[DIR_PMP] = &DD27[DIR_PMP * numberOfLBnodes];
+ D27.f[DIR_MPP] = &DD27[DIR_MPP * numberOfLBnodes];
+ D27.f[DIR_PPM] = &DD27[DIR_PPM * numberOfLBnodes];
+ D27.f[DIR_MMM] = &DD27[DIR_MMM * numberOfLBnodes];
+ D27.f[DIR_PMM] = &DD27[DIR_PMM * numberOfLBnodes];
+ D27.f[DIR_MPM] = &DD27[DIR_MPM * numberOfLBnodes];
}
else
{
- D27.f[DIR_M00 ] = &DD27[DIR_P00 *size_Mat];
- D27.f[DIR_P00 ] = &DD27[DIR_M00 *size_Mat];
- D27.f[DIR_0M0 ] = &DD27[DIR_0P0 *size_Mat];
- D27.f[DIR_0P0 ] = &DD27[DIR_0M0 *size_Mat];
- D27.f[DIR_00M ] = &DD27[DIR_00P *size_Mat];
- D27.f[DIR_00P ] = &DD27[DIR_00M *size_Mat];
- D27.f[DIR_MM0 ] = &DD27[DIR_PP0 *size_Mat];
- D27.f[DIR_PP0 ] = &DD27[DIR_MM0 *size_Mat];
- D27.f[DIR_MP0 ] = &DD27[DIR_PM0 *size_Mat];
- D27.f[DIR_PM0 ] = &DD27[DIR_MP0 *size_Mat];
- D27.f[DIR_M0M ] = &DD27[DIR_P0P *size_Mat];
- D27.f[DIR_P0P ] = &DD27[DIR_M0M *size_Mat];
- D27.f[DIR_M0P ] = &DD27[DIR_P0M *size_Mat];
- D27.f[DIR_P0M ] = &DD27[DIR_M0P *size_Mat];
- D27.f[DIR_0MM ] = &DD27[DIR_0PP *size_Mat];
- D27.f[DIR_0PP ] = &DD27[DIR_0MM *size_Mat];
- D27.f[DIR_0MP ] = &DD27[DIR_0PM *size_Mat];
- D27.f[DIR_0PM ] = &DD27[DIR_0MP *size_Mat];
- D27.f[DIR_000] = &DD27[DIR_000*size_Mat];
- D27.f[DIR_PPP ] = &DD27[DIR_MMM *size_Mat];
- D27.f[DIR_MMP ] = &DD27[DIR_PPM *size_Mat];
- D27.f[DIR_PMP ] = &DD27[DIR_MPM *size_Mat];
- D27.f[DIR_MPP ] = &DD27[DIR_PMM *size_Mat];
- D27.f[DIR_PPM ] = &DD27[DIR_MMP *size_Mat];
- D27.f[DIR_MMM ] = &DD27[DIR_PPP *size_Mat];
- D27.f[DIR_PMM ] = &DD27[DIR_MPP *size_Mat];
- D27.f[DIR_MPM ] = &DD27[DIR_PMP *size_Mat];
+ D27.f[DIR_M00] = &DD27[DIR_P00 * numberOfLBnodes];
+ D27.f[DIR_P00] = &DD27[DIR_M00 * numberOfLBnodes];
+ D27.f[DIR_0M0] = &DD27[DIR_0P0 * numberOfLBnodes];
+ D27.f[DIR_0P0] = &DD27[DIR_0M0 * numberOfLBnodes];
+ D27.f[DIR_00M] = &DD27[DIR_00P * numberOfLBnodes];
+ D27.f[DIR_00P] = &DD27[DIR_00M * numberOfLBnodes];
+ D27.f[DIR_MM0] = &DD27[DIR_PP0 * numberOfLBnodes];
+ D27.f[DIR_PP0] = &DD27[DIR_MM0 * numberOfLBnodes];
+ D27.f[DIR_MP0] = &DD27[DIR_PM0 * numberOfLBnodes];
+ D27.f[DIR_PM0] = &DD27[DIR_MP0 * numberOfLBnodes];
+ D27.f[DIR_M0M] = &DD27[DIR_P0P * numberOfLBnodes];
+ D27.f[DIR_P0P] = &DD27[DIR_M0M * numberOfLBnodes];
+ D27.f[DIR_M0P] = &DD27[DIR_P0M * numberOfLBnodes];
+ D27.f[DIR_P0M] = &DD27[DIR_M0P * numberOfLBnodes];
+ D27.f[DIR_0MM] = &DD27[DIR_0PP * numberOfLBnodes];
+ D27.f[DIR_0PP] = &DD27[DIR_0MM * numberOfLBnodes];
+ D27.f[DIR_0MP] = &DD27[DIR_0PM * numberOfLBnodes];
+ D27.f[DIR_0PM] = &DD27[DIR_0MP * numberOfLBnodes];
+ D27.f[DIR_000] = &DD27[DIR_000 * numberOfLBnodes];
+ D27.f[DIR_PPP] = &DD27[DIR_MMM * numberOfLBnodes];
+ D27.f[DIR_MMP] = &DD27[DIR_PPM * numberOfLBnodes];
+ D27.f[DIR_PMP] = &DD27[DIR_MPM * numberOfLBnodes];
+ D27.f[DIR_MPP] = &DD27[DIR_PMM * numberOfLBnodes];
+ D27.f[DIR_PPM] = &DD27[DIR_MMP * numberOfLBnodes];
+ D27.f[DIR_MMM] = &DD27[DIR_PPP * numberOfLBnodes];
+ D27.f[DIR_PMM] = &DD27[DIR_MPP * numberOfLBnodes];
+ D27.f[DIR_MPM] = &DD27[DIR_PMP * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
//Test
@@ -5694,24 +5694,24 @@ __global__ void QADVeloIncomp27(
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
real q;
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
- q = q_dirE[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_M00 ])[kw ]= -feqW27_W + c2o1 * c2o27 * TempD;
- q = q_dirW[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_P00 ])[ke ]= -feqW27_E + c2o1 * c2o27 * TempD;
- q = q_dirN[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_0M0 ])[ks ]= -feqW27_S + c2o1 * c2o27 * TempD;
- q = q_dirS[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_0P0 ])[kn ]= -feqW27_N + c2o1 * c2o27 * TempD;
- q = q_dirT[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_00M ])[kb ]= -feqW27_B + c2o1 * c2o27 * TempD;
- q = q_dirB[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_00P ])[kt ]= -feqW27_T + c2o1 * c2o27 * TempD;
- q = q_dirNE[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_MM0 ])[ksw ]= -feqW27_SW + c2o1 * c1o54 * TempD;
- q = q_dirSW[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_PP0 ])[kne ]= -feqW27_NE + c2o1 * c1o54 * TempD;
- q = q_dirSE[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_MP0 ])[knw ]= -feqW27_NW + c2o1 * c1o54 * TempD;
- q = q_dirNW[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_PM0 ])[kse ]= -feqW27_SE + c2o1 * c1o54 * TempD;
- q = q_dirTE[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_M0M ])[kbw ]= -feqW27_BW + c2o1 * c1o54 * TempD;
- q = q_dirBW[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_P0P ])[kte ]= -feqW27_TE + c2o1 * c1o54 * TempD;
- q = q_dirBE[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_M0P ])[ktw ]= -feqW27_TW + c2o1 * c1o54 * TempD;
- q = q_dirTW[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_P0M ])[kbe ]= -feqW27_BE + c2o1 * c1o54 * TempD;
- q = q_dirTN[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_0MM ])[kbs ]= -feqW27_BS + c2o1 * c1o54 * TempD;
- q = q_dirBS[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_0PP ])[ktn ]= -feqW27_TN + c2o1 * c1o54 * TempD;
- q = q_dirBN[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_0MP ])[kts ]= -feqW27_TS + c2o1 * c1o54 * TempD;
- q = q_dirTS[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_0PM ])[kbn ]= -feqW27_BN + c2o1 * c1o54 * TempD;
+ q = q_dirE[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_M00])[kw ]= -feqW27_W + c2o1 * c2o27 * TempD;
+ q = q_dirW[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_P00])[ke ]= -feqW27_E + c2o1 * c2o27 * TempD;
+ q = q_dirN[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_0M0])[ks ]= -feqW27_S + c2o1 * c2o27 * TempD;
+ q = q_dirS[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_0P0])[kn ]= -feqW27_N + c2o1 * c2o27 * TempD;
+ q = q_dirT[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_00M])[kb ]= -feqW27_B + c2o1 * c2o27 * TempD;
+ q = q_dirB[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_00P])[kt ]= -feqW27_T + c2o1 * c2o27 * TempD;
+ q = q_dirNE[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_MM0])[ksw ]= -feqW27_SW + c2o1 * c1o54 * TempD;
+ q = q_dirSW[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_PP0])[kne ]= -feqW27_NE + c2o1 * c1o54 * TempD;
+ q = q_dirSE[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_MP0])[knw ]= -feqW27_NW + c2o1 * c1o54 * TempD;
+ q = q_dirNW[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_PM0])[kse ]= -feqW27_SE + c2o1 * c1o54 * TempD;
+ q = q_dirTE[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_M0M])[kbw ]= -feqW27_BW + c2o1 * c1o54 * TempD;
+ q = q_dirBW[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_P0P])[kte ]= -feqW27_TE + c2o1 * c1o54 * TempD;
+ q = q_dirBE[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_M0P])[ktw ]= -feqW27_TW + c2o1 * c1o54 * TempD;
+ q = q_dirTW[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_P0M])[kbe ]= -feqW27_BE + c2o1 * c1o54 * TempD;
+ q = q_dirTN[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_0MM])[kbs ]= -feqW27_BS + c2o1 * c1o54 * TempD;
+ q = q_dirBS[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_0PP])[ktn ]= -feqW27_TN + c2o1 * c1o54 * TempD;
+ q = q_dirBN[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_0MP])[kts ]= -feqW27_TS + c2o1 * c1o54 * TempD;
+ q = q_dirTS[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_0PM])[kbn ]= -feqW27_BN + c2o1 * c1o54 * TempD;
q = q_dirTNE[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_MMM])[kbsw]= -feqW27_BSW+ c2o1 * c1o216 * TempD;
q = q_dirBSW[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_PPP])[ktne]= -feqW27_TNE+ c2o1 * c1o216 * TempD;
q = q_dirBNE[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_MMP])[ktsw]= -feqW27_TSW+ c2o1 * c1o216 * TempD;
@@ -5720,24 +5720,24 @@ __global__ void QADVeloIncomp27(
q = q_dirBNW[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_PMP])[ktse]= -feqW27_TSE+ c2o1 * c1o216 * TempD;
q = q_dirBSE[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_MPP])[ktnw]= -feqW27_TNW+ c2o1 * c1o216 * TempD;
q = q_dirTNW[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_PMM])[kbse]= -feqW27_BSE+ c2o1 * c1o216 * TempD;
- //q = q_dirE[k]; if (q>=zero && q<=one) (D27.f[DIR_M00 ])[kw ]=(two*feqW27_W -(f27_E *(q*omegaD-one)-omegaD*feq27_E *(q-one))/(omegaD-one)+f27_W *q)/(q+one);
- //q = q_dirW[k]; if (q>=zero && q<=one) (D27.f[DIR_P00 ])[ke ]=(two*feqW27_E -(f27_W *(q*omegaD-one)-omegaD*feq27_W *(q-one))/(omegaD-one)+f27_E *q)/(q+one);
- //q = q_dirN[k]; if (q>=zero && q<=one) (D27.f[DIR_0M0 ])[ks ]=(two*feqW27_S -(f27_N *(q*omegaD-one)-omegaD*feq27_N *(q-one))/(omegaD-one)+f27_S *q)/(q+one);
- //q = q_dirS[k]; if (q>=zero && q<=one) (D27.f[DIR_0P0 ])[kn ]=(two*feqW27_N -(f27_S *(q*omegaD-one)-omegaD*feq27_S *(q-one))/(omegaD-one)+f27_N *q)/(q+one);
- //q = q_dirT[k]; if (q>=zero && q<=one) (D27.f[DIR_00M ])[kb ]=(two*feqW27_B -(f27_T *(q*omegaD-one)-omegaD*feq27_T *(q-one))/(omegaD-one)+f27_B *q)/(q+one);
- //q = q_dirB[k]; if (q>=zero && q<=one) (D27.f[DIR_00P ])[kt ]=(two*feqW27_T -(f27_B *(q*omegaD-one)-omegaD*feq27_B *(q-one))/(omegaD-one)+f27_T *q)/(q+one);
- //q = q_dirNE[k]; if (q>=zero && q<=one) (D27.f[DIR_MM0 ])[ksw ]=(two*feqW27_SW -(f27_NE *(q*omegaD-one)-omegaD*feq27_NE *(q-one))/(omegaD-one)+f27_SW *q)/(q+one);
- //q = q_dirSW[k]; if (q>=zero && q<=one) (D27.f[DIR_PP0 ])[kne ]=(two*feqW27_NE -(f27_SW *(q*omegaD-one)-omegaD*feq27_SW *(q-one))/(omegaD-one)+f27_NE *q)/(q+one);
- //q = q_dirSE[k]; if (q>=zero && q<=one) (D27.f[DIR_MP0 ])[knw ]=(two*feqW27_NW -(f27_SE *(q*omegaD-one)-omegaD*feq27_SE *(q-one))/(omegaD-one)+f27_NW *q)/(q+one);
- //q = q_dirNW[k]; if (q>=zero && q<=one) (D27.f[DIR_PM0 ])[kse ]=(two*feqW27_SE -(f27_NW *(q*omegaD-one)-omegaD*feq27_NW *(q-one))/(omegaD-one)+f27_SE *q)/(q+one);
- //q = q_dirTE[k]; if (q>=zero && q<=one) (D27.f[DIR_M0M ])[kbw ]=(two*feqW27_BW -(f27_TE *(q*omegaD-one)-omegaD*feq27_TE *(q-one))/(omegaD-one)+f27_BW *q)/(q+one);
- //q = q_dirBW[k]; if (q>=zero && q<=one) (D27.f[DIR_P0P ])[kte ]=(two*feqW27_TE -(f27_BW *(q*omegaD-one)-omegaD*feq27_BW *(q-one))/(omegaD-one)+f27_TE *q)/(q+one);
- //q = q_dirBE[k]; if (q>=zero && q<=one) (D27.f[DIR_M0P ])[ktw ]=(two*feqW27_TW -(f27_BE *(q*omegaD-one)-omegaD*feq27_BE *(q-one))/(omegaD-one)+f27_TW *q)/(q+one);
- //q = q_dirTW[k]; if (q>=zero && q<=one) (D27.f[DIR_P0M ])[kbe ]=(two*feqW27_BE -(f27_TW *(q*omegaD-one)-omegaD*feq27_TW *(q-one))/(omegaD-one)+f27_BE *q)/(q+one);
- //q = q_dirTN[k]; if (q>=zero && q<=one) (D27.f[DIR_0MM ])[kbs ]=(two*feqW27_BS -(f27_TN *(q*omegaD-one)-omegaD*feq27_TN *(q-one))/(omegaD-one)+f27_BS *q)/(q+one);
- //q = q_dirBS[k]; if (q>=zero && q<=one) (D27.f[DIR_0PP ])[ktn ]=(two*feqW27_TN -(f27_BS *(q*omegaD-one)-omegaD*feq27_BS *(q-one))/(omegaD-one)+f27_TN *q)/(q+one);
- //q = q_dirBN[k]; if (q>=zero && q<=one) (D27.f[DIR_0MP ])[kts ]=(two*feqW27_TS -(f27_BN *(q*omegaD-one)-omegaD*feq27_BN *(q-one))/(omegaD-one)+f27_TS *q)/(q+one);
- //q = q_dirTS[k]; if (q>=zero && q<=one) (D27.f[DIR_0PM ])[kbn ]=(two*feqW27_BN -(f27_TS *(q*omegaD-one)-omegaD*feq27_TS *(q-one))/(omegaD-one)+f27_BN *q)/(q+one);
+ //q = q_dirE[k]; if (q>=zero && q<=one) (D27.f[DIR_M00])[kw ]=(two*feqW27_W -(f27_E *(q*omegaD-one)-omegaD*feq27_E *(q-one))/(omegaD-one)+f27_W *q)/(q+one);
+ //q = q_dirW[k]; if (q>=zero && q<=one) (D27.f[DIR_P00])[ke ]=(two*feqW27_E -(f27_W *(q*omegaD-one)-omegaD*feq27_W *(q-one))/(omegaD-one)+f27_E *q)/(q+one);
+ //q = q_dirN[k]; if (q>=zero && q<=one) (D27.f[DIR_0M0])[ks ]=(two*feqW27_S -(f27_N *(q*omegaD-one)-omegaD*feq27_N *(q-one))/(omegaD-one)+f27_S *q)/(q+one);
+ //q = q_dirS[k]; if (q>=zero && q<=one) (D27.f[DIR_0P0])[kn ]=(two*feqW27_N -(f27_S *(q*omegaD-one)-omegaD*feq27_S *(q-one))/(omegaD-one)+f27_N *q)/(q+one);
+ //q = q_dirT[k]; if (q>=zero && q<=one) (D27.f[DIR_00M])[kb ]=(two*feqW27_B -(f27_T *(q*omegaD-one)-omegaD*feq27_T *(q-one))/(omegaD-one)+f27_B *q)/(q+one);
+ //q = q_dirB[k]; if (q>=zero && q<=one) (D27.f[DIR_00P])[kt ]=(two*feqW27_T -(f27_B *(q*omegaD-one)-omegaD*feq27_B *(q-one))/(omegaD-one)+f27_T *q)/(q+one);
+ //q = q_dirNE[k]; if (q>=zero && q<=one) (D27.f[DIR_MM0])[ksw ]=(two*feqW27_SW -(f27_NE *(q*omegaD-one)-omegaD*feq27_NE *(q-one))/(omegaD-one)+f27_SW *q)/(q+one);
+ //q = q_dirSW[k]; if (q>=zero && q<=one) (D27.f[DIR_PP0])[kne ]=(two*feqW27_NE -(f27_SW *(q*omegaD-one)-omegaD*feq27_SW *(q-one))/(omegaD-one)+f27_NE *q)/(q+one);
+ //q = q_dirSE[k]; if (q>=zero && q<=one) (D27.f[DIR_MP0])[knw ]=(two*feqW27_NW -(f27_SE *(q*omegaD-one)-omegaD*feq27_SE *(q-one))/(omegaD-one)+f27_NW *q)/(q+one);
+ //q = q_dirNW[k]; if (q>=zero && q<=one) (D27.f[DIR_PM0])[kse ]=(two*feqW27_SE -(f27_NW *(q*omegaD-one)-omegaD*feq27_NW *(q-one))/(omegaD-one)+f27_SE *q)/(q+one);
+ //q = q_dirTE[k]; if (q>=zero && q<=one) (D27.f[DIR_M0M])[kbw ]=(two*feqW27_BW -(f27_TE *(q*omegaD-one)-omegaD*feq27_TE *(q-one))/(omegaD-one)+f27_BW *q)/(q+one);
+ //q = q_dirBW[k]; if (q>=zero && q<=one) (D27.f[DIR_P0P])[kte ]=(two*feqW27_TE -(f27_BW *(q*omegaD-one)-omegaD*feq27_BW *(q-one))/(omegaD-one)+f27_TE *q)/(q+one);
+ //q = q_dirBE[k]; if (q>=zero && q<=one) (D27.f[DIR_M0P])[ktw ]=(two*feqW27_TW -(f27_BE *(q*omegaD-one)-omegaD*feq27_BE *(q-one))/(omegaD-one)+f27_TW *q)/(q+one);
+ //q = q_dirTW[k]; if (q>=zero && q<=one) (D27.f[DIR_P0M])[kbe ]=(two*feqW27_BE -(f27_TW *(q*omegaD-one)-omegaD*feq27_TW *(q-one))/(omegaD-one)+f27_BE *q)/(q+one);
+ //q = q_dirTN[k]; if (q>=zero && q<=one) (D27.f[DIR_0MM])[kbs ]=(two*feqW27_BS -(f27_TN *(q*omegaD-one)-omegaD*feq27_TN *(q-one))/(omegaD-one)+f27_BS *q)/(q+one);
+ //q = q_dirBS[k]; if (q>=zero && q<=one) (D27.f[DIR_0PP])[ktn ]=(two*feqW27_TN -(f27_BS *(q*omegaD-one)-omegaD*feq27_BS *(q-one))/(omegaD-one)+f27_TN *q)/(q+one);
+ //q = q_dirBN[k]; if (q>=zero && q<=one) (D27.f[DIR_0MP])[kts ]=(two*feqW27_TS -(f27_BN *(q*omegaD-one)-omegaD*feq27_BN *(q-one))/(omegaD-one)+f27_TS *q)/(q+one);
+ //q = q_dirTS[k]; if (q>=zero && q<=one) (D27.f[DIR_0PM])[kbn ]=(two*feqW27_BN -(f27_TS *(q*omegaD-one)-omegaD*feq27_TS *(q-one))/(omegaD-one)+f27_BN *q)/(q+one);
//q = q_dirTNE[k]; if (q>=zero && q<=one) (D27.f[DIR_MMM])[kbsw]=(two*feqW27_BSW-(f27_TNE*(q*omegaD-one)-omegaD*feq27_TNE*(q-one))/(omegaD-one)+f27_BSW*q)/(q+one);
//q = q_dirBSW[k]; if (q>=zero && q<=one) (D27.f[DIR_PPP])[ktne]=(two*feqW27_TNE-(f27_BSW*(q*omegaD-one)-omegaD*feq27_BSW*(q-one))/(omegaD-one)+f27_TNE*q)/(q+one);
//q = q_dirBNE[k]; if (q>=zero && q<=one) (D27.f[DIR_MMP])[ktsw]=(two*feqW27_TSW-(f27_BNE*(q*omegaD-one)-omegaD*feq27_BNE*(q-one))/(omegaD-one)+f27_TSW*q)/(q+one);
@@ -5801,91 +5801,91 @@ __global__ void QADPressIncomp7( real* DD,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep)
{
/* Distributions27 D;
if (isEvenTimestep==true)
{
- D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DD[DIR_P00 * size_Mat];
+ D.f[DIR_M00] = &DD[DIR_M00 * size_Mat];
+ D.f[DIR_0P0] = &DD[DIR_0P0 * size_Mat];
+ D.f[DIR_0M0] = &DD[DIR_0M0 * size_Mat];
+ D.f[DIR_00P] = &DD[DIR_00P * size_Mat];
+ D.f[DIR_00M] = &DD[DIR_00M * size_Mat];
+ D.f[DIR_PP0] = &DD[DIR_PP0 * size_Mat];
+ D.f[DIR_MM0] = &DD[DIR_MM0 * size_Mat];
+ D.f[DIR_PM0] = &DD[DIR_PM0 * size_Mat];
+ D.f[DIR_MP0] = &DD[DIR_MP0 * size_Mat];
+ D.f[DIR_P0P] = &DD[DIR_P0P * size_Mat];
+ D.f[DIR_M0M] = &DD[DIR_M0M * size_Mat];
+ D.f[DIR_P0M] = &DD[DIR_P0M * size_Mat];
+ D.f[DIR_M0P] = &DD[DIR_M0P * size_Mat];
+ D.f[DIR_0PP] = &DD[DIR_0PP * size_Mat];
+ D.f[DIR_0MM] = &DD[DIR_0MM * size_Mat];
+ D.f[DIR_0PM] = &DD[DIR_0PM * size_Mat];
+ D.f[DIR_0MP] = &DD[DIR_0MP * size_Mat];
+ D.f[DIR_000] = &DD[DIR_000 * size_Mat];
+ D.f[DIR_PPP] = &DD[DIR_PPP * size_Mat];
+ D.f[DIR_MMP] = &DD[DIR_MMP * size_Mat];
+ D.f[DIR_PMP] = &DD[DIR_PMP * size_Mat];
+ D.f[DIR_MPP] = &DD[DIR_MPP * size_Mat];
+ D.f[DIR_PPM] = &DD[DIR_PPM * size_Mat];
+ D.f[DIR_MMM] = &DD[DIR_MMM * size_Mat];
+ D.f[DIR_PMM] = &DD[DIR_PMM * size_Mat];
+ D.f[DIR_MPM] = &DD[DIR_MPM * size_Mat];
}
else
{
- D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
+ D.f[DIR_M00] = &DD[DIR_P00 * size_Mat];
+ D.f[DIR_P00] = &DD[DIR_M00 * size_Mat];
+ D.f[DIR_0M0] = &DD[DIR_0P0 * size_Mat];
+ D.f[DIR_0P0] = &DD[DIR_0M0 * size_Mat];
+ D.f[DIR_00M] = &DD[DIR_00P * size_Mat];
+ D.f[DIR_00P] = &DD[DIR_00M * size_Mat];
+ D.f[DIR_MM0] = &DD[DIR_PP0 * size_Mat];
+ D.f[DIR_PP0] = &DD[DIR_MM0 * size_Mat];
+ D.f[DIR_MP0] = &DD[DIR_PM0 * size_Mat];
+ D.f[DIR_PM0] = &DD[DIR_MP0 * size_Mat];
+ D.f[DIR_M0M] = &DD[DIR_P0P * size_Mat];
+ D.f[DIR_P0P] = &DD[DIR_M0M * size_Mat];
+ D.f[DIR_M0P] = &DD[DIR_P0M * size_Mat];
+ D.f[DIR_P0M] = &DD[DIR_M0P * size_Mat];
+ D.f[DIR_0MM] = &DD[DIR_0PP * size_Mat];
+ D.f[DIR_0PP] = &DD[DIR_0MM * size_Mat];
+ D.f[DIR_0MP] = &DD[DIR_0PM * size_Mat];
+ D.f[DIR_0PM] = &DD[DIR_0MP * size_Mat];
+ D.f[DIR_000] = &DD[DIR_000 * size_Mat];
+ D.f[DIR_PPP] = &DD[DIR_MMM * size_Mat];
+ D.f[DIR_MMP] = &DD[DIR_PPM * size_Mat];
+ D.f[DIR_PMP] = &DD[DIR_MPM * size_Mat];
+ D.f[DIR_MPP] = &DD[DIR_PMM * size_Mat];
+ D.f[DIR_PPM] = &DD[DIR_MMP * size_Mat];
+ D.f[DIR_MMM] = &DD[DIR_PPP * size_Mat];
+ D.f[DIR_PMM] = &DD[DIR_MPP * size_Mat];
+ D.f[DIR_MPM] = &DD[DIR_PMP * size_Mat];
}*/
Distributions7 D7;
if (isEvenTimestep==true)
{
- D7.f[0] = &DD7[0*size_Mat];
- D7.f[1] = &DD7[1*size_Mat];
- D7.f[2] = &DD7[2*size_Mat];
- D7.f[3] = &DD7[3*size_Mat];
- D7.f[4] = &DD7[4*size_Mat];
- D7.f[5] = &DD7[5*size_Mat];
- D7.f[6] = &DD7[6*size_Mat];
+ D7.f[0] = &DD7[0*numberOfLBnodes];
+ D7.f[1] = &DD7[1*numberOfLBnodes];
+ D7.f[2] = &DD7[2*numberOfLBnodes];
+ D7.f[3] = &DD7[3*numberOfLBnodes];
+ D7.f[4] = &DD7[4*numberOfLBnodes];
+ D7.f[5] = &DD7[5*numberOfLBnodes];
+ D7.f[6] = &DD7[6*numberOfLBnodes];
}
else
{
- D7.f[0] = &DD7[0*size_Mat];
- D7.f[2] = &DD7[1*size_Mat];
- D7.f[1] = &DD7[2*size_Mat];
- D7.f[4] = &DD7[3*size_Mat];
- D7.f[3] = &DD7[4*size_Mat];
- D7.f[6] = &DD7[5*size_Mat];
- D7.f[5] = &DD7[6*size_Mat];
+ D7.f[0] = &DD7[0*numberOfLBnodes];
+ D7.f[2] = &DD7[1*numberOfLBnodes];
+ D7.f[1] = &DD7[2*numberOfLBnodes];
+ D7.f[4] = &DD7[3*numberOfLBnodes];
+ D7.f[3] = &DD7[4*numberOfLBnodes];
+ D7.f[6] = &DD7[5*numberOfLBnodes];
+ D7.f[5] = &DD7[6*numberOfLBnodes];
}
@@ -5905,12 +5905,12 @@ __global__ void QADPressIncomp7( real* DD,
//////////////////////////////////////////////////////////////////////////////////
real *q_dirE, *q_dirW, *q_dirN, *q_dirS, *q_dirT, *q_dirB;
- q_dirE = &QQ[DIR_P00 * numberOfBCnodes];
- q_dirW = &QQ[DIR_M00 * numberOfBCnodes];
- q_dirN = &QQ[DIR_0P0 * numberOfBCnodes];
- q_dirS = &QQ[DIR_0M0 * numberOfBCnodes];
- q_dirT = &QQ[DIR_00P * numberOfBCnodes];
- q_dirB = &QQ[DIR_00M * numberOfBCnodes];
+ q_dirE = &QQ[DIR_P00 * numberOfBCnodes];
+ q_dirW = &QQ[DIR_M00 * numberOfBCnodes];
+ q_dirN = &QQ[DIR_0P0 * numberOfBCnodes];
+ q_dirS = &QQ[DIR_0M0 * numberOfBCnodes];
+ q_dirT = &QQ[DIR_00P * numberOfBCnodes];
+ q_dirB = &QQ[DIR_00M * numberOfBCnodes];
//////////////////////////////////////////////////////////////////////////////////
//index
unsigned int KQK = k_Q[k];
@@ -5945,32 +5945,32 @@ __global__ void QADPressIncomp7( real* DD,
/* real f_E, f_W, f_N, f_S, f_T, f_B, f_NE, f_SW, f_SE, f_NW, f_TE, f_BW, f_BE,
f_TW, f_TN, f_BS, f_BN, f_TS, f_TNE, f_TSW, f_TSE, f_TNW, f_BNE, f_BSW, f_BSE, f_BNW;
- f_W = (D.f[DIR_P00 ])[ke ];
- f_E = (D.f[DIR_M00 ])[kw ];
- f_S = (D.f[DIR_0P0 ])[kn ];
- f_N = (D.f[DIR_0M0 ])[ks ];
- f_B = (D.f[DIR_00P ])[kt ];
- f_T = (D.f[DIR_00M ])[kb ];
- f_SW = (D.f[DIR_PP0 ])[kne ];
- f_NE = (D.f[DIR_MM0 ])[ksw ];
- f_NW = (D.f[DIR_PM0 ])[kse ];
- f_SE = (D.f[DIR_MP0 ])[knw ];
- f_BW = (D.f[DIR_P0P ])[kte ];
- f_TE = (D.f[DIR_M0M ])[kbw ];
- f_TW = (D.f[DIR_P0M ])[kbe ];
- f_BE = (D.f[DIR_M0P ])[ktw ];
- f_BS = (D.f[DIR_0PP ])[ktn ];
- f_TN = (D.f[DIR_0MM ])[kbs ];
- f_TS = (D.f[DIR_0PM ])[kbn ];
- f_BN = (D.f[DIR_0MP ])[kts ];
- f_BSW = (D.f[DIR_PPP ])[ktne ];
- f_BNE = (D.f[DIR_MMP ])[ktsw ];
- f_BNW = (D.f[DIR_PMP ])[ktse ];
- f_BSE = (D.f[DIR_MPP ])[ktnw ];
- f_TSW = (D.f[DIR_PPM ])[kbne ];
- f_TNE = (D.f[DIR_MMM ])[kbsw ];
- f_TNW = (D.f[DIR_PMM ])[kbse ];
- f_TSE = (D.f[DIR_MPM ])[kbnw ];*/
+ f_W = (D.f[DIR_P00])[ke ];
+ f_E = (D.f[DIR_M00])[kw ];
+ f_S = (D.f[DIR_0P0])[kn ];
+ f_N = (D.f[DIR_0M0])[ks ];
+ f_B = (D.f[DIR_00P])[kt ];
+ f_T = (D.f[DIR_00M])[kb ];
+ f_SW = (D.f[DIR_PP0])[kne ];
+ f_NE = (D.f[DIR_MM0])[ksw ];
+ f_NW = (D.f[DIR_PM0])[kse ];
+ f_SE = (D.f[DIR_MP0])[knw ];
+ f_BW = (D.f[DIR_P0P])[kte ];
+ f_TE = (D.f[DIR_M0M])[kbw ];
+ f_TW = (D.f[DIR_P0M])[kbe ];
+ f_BE = (D.f[DIR_M0P])[ktw ];
+ f_BS = (D.f[DIR_0PP])[ktn ];
+ f_TN = (D.f[DIR_0MM])[kbs ];
+ f_TS = (D.f[DIR_0PM])[kbn ];
+ f_BN = (D.f[DIR_0MP])[kts ];
+ f_BSW = (D.f[DIR_PPP])[ktne ];
+ f_BNE = (D.f[DIR_MMP])[ktsw ];
+ f_BNW = (D.f[DIR_PMP])[ktse ];
+ f_BSE = (D.f[DIR_MPP])[ktnw ];
+ f_TSW = (D.f[DIR_PPM])[kbne ];
+ f_TNE = (D.f[DIR_MMM])[kbsw ];
+ f_TNW = (D.f[DIR_PMM])[kbse ];
+ f_TSE = (D.f[DIR_MPM])[kbnw ];*/
////////////////////////////////////////////////////////////////////////////////
//real vx1 = ((f_TNE-f_BSW)+(f_BNE-f_TSW)+(f_TSE-f_BNW)+(f_BSE-f_TNW) +(f_NE-f_SW)+(f_SE-f_NW)+(f_TE-f_BW)+(f_BE-f_TW)+(f_E-f_W));
//real vx2 = ((f_TNE-f_BSW)+(f_BNE-f_TSW)+(f_BNW-f_TSE)+(f_TNW-f_BSE) +(f_NE-f_SW)+(f_NW-f_SE)+(f_TN-f_BS)+(f_BN-f_TS)+(f_N-f_S));
@@ -6035,23 +6035,23 @@ __global__ void QADPressIncomp7( real* DD,
//pointertausch
if (isEvenTimestep==false)
{
- D7.f[0] = &DD7[0*size_Mat];
- D7.f[1] = &DD7[1*size_Mat];
- D7.f[2] = &DD7[2*size_Mat];
- D7.f[3] = &DD7[3*size_Mat];
- D7.f[4] = &DD7[4*size_Mat];
- D7.f[5] = &DD7[5*size_Mat];
- D7.f[6] = &DD7[6*size_Mat];
+ D7.f[0] = &DD7[0*numberOfLBnodes];
+ D7.f[1] = &DD7[1*numberOfLBnodes];
+ D7.f[2] = &DD7[2*numberOfLBnodes];
+ D7.f[3] = &DD7[3*numberOfLBnodes];
+ D7.f[4] = &DD7[4*numberOfLBnodes];
+ D7.f[5] = &DD7[5*numberOfLBnodes];
+ D7.f[6] = &DD7[6*numberOfLBnodes];
}
else
{
- D7.f[0] = &DD7[0*size_Mat];
- D7.f[2] = &DD7[1*size_Mat];
- D7.f[1] = &DD7[2*size_Mat];
- D7.f[4] = &DD7[3*size_Mat];
- D7.f[3] = &DD7[4*size_Mat];
- D7.f[6] = &DD7[5*size_Mat];
- D7.f[5] = &DD7[6*size_Mat];
+ D7.f[0] = &DD7[0*numberOfLBnodes];
+ D7.f[2] = &DD7[1*numberOfLBnodes];
+ D7.f[1] = &DD7[2*numberOfLBnodes];
+ D7.f[4] = &DD7[3*numberOfLBnodes];
+ D7.f[3] = &DD7[4*numberOfLBnodes];
+ D7.f[6] = &DD7[5*numberOfLBnodes];
+ D7.f[5] = &DD7[6*numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////
@@ -6240,131 +6240,131 @@ __global__ void QADPressIncomp27(
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep)
{
Distributions27 D;
if (isEvenTimestep==true)
{
- D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
+ D.f[DIR_M00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_MPM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_PMP * numberOfLBnodes];
}
Distributions27 D27;
if (isEvenTimestep==true)
{
- D27.f[DIR_P00 ] = &DD27[DIR_P00 *size_Mat];
- D27.f[DIR_M00 ] = &DD27[DIR_M00 *size_Mat];
- D27.f[DIR_0P0 ] = &DD27[DIR_0P0 *size_Mat];
- D27.f[DIR_0M0 ] = &DD27[DIR_0M0 *size_Mat];
- D27.f[DIR_00P ] = &DD27[DIR_00P *size_Mat];
- D27.f[DIR_00M ] = &DD27[DIR_00M *size_Mat];
- D27.f[DIR_PP0 ] = &DD27[DIR_PP0 *size_Mat];
- D27.f[DIR_MM0 ] = &DD27[DIR_MM0 *size_Mat];
- D27.f[DIR_PM0 ] = &DD27[DIR_PM0 *size_Mat];
- D27.f[DIR_MP0 ] = &DD27[DIR_MP0 *size_Mat];
- D27.f[DIR_P0P ] = &DD27[DIR_P0P *size_Mat];
- D27.f[DIR_M0M ] = &DD27[DIR_M0M *size_Mat];
- D27.f[DIR_P0M ] = &DD27[DIR_P0M *size_Mat];
- D27.f[DIR_M0P ] = &DD27[DIR_M0P *size_Mat];
- D27.f[DIR_0PP ] = &DD27[DIR_0PP *size_Mat];
- D27.f[DIR_0MM ] = &DD27[DIR_0MM *size_Mat];
- D27.f[DIR_0PM ] = &DD27[DIR_0PM *size_Mat];
- D27.f[DIR_0MP ] = &DD27[DIR_0MP *size_Mat];
- D27.f[DIR_000] = &DD27[DIR_000*size_Mat];
- D27.f[DIR_PPP ] = &DD27[DIR_PPP *size_Mat];
- D27.f[DIR_MMP ] = &DD27[DIR_MMP *size_Mat];
- D27.f[DIR_PMP ] = &DD27[DIR_PMP *size_Mat];
- D27.f[DIR_MPP ] = &DD27[DIR_MPP *size_Mat];
- D27.f[DIR_PPM ] = &DD27[DIR_PPM *size_Mat];
- D27.f[DIR_MMM ] = &DD27[DIR_MMM *size_Mat];
- D27.f[DIR_PMM ] = &DD27[DIR_PMM *size_Mat];
- D27.f[DIR_MPM ] = &DD27[DIR_MPM *size_Mat];
+ D27.f[DIR_P00] = &DD27[DIR_P00 * numberOfLBnodes];
+ D27.f[DIR_M00] = &DD27[DIR_M00 * numberOfLBnodes];
+ D27.f[DIR_0P0] = &DD27[DIR_0P0 * numberOfLBnodes];
+ D27.f[DIR_0M0] = &DD27[DIR_0M0 * numberOfLBnodes];
+ D27.f[DIR_00P] = &DD27[DIR_00P * numberOfLBnodes];
+ D27.f[DIR_00M] = &DD27[DIR_00M * numberOfLBnodes];
+ D27.f[DIR_PP0] = &DD27[DIR_PP0 * numberOfLBnodes];
+ D27.f[DIR_MM0] = &DD27[DIR_MM0 * numberOfLBnodes];
+ D27.f[DIR_PM0] = &DD27[DIR_PM0 * numberOfLBnodes];
+ D27.f[DIR_MP0] = &DD27[DIR_MP0 * numberOfLBnodes];
+ D27.f[DIR_P0P] = &DD27[DIR_P0P * numberOfLBnodes];
+ D27.f[DIR_M0M] = &DD27[DIR_M0M * numberOfLBnodes];
+ D27.f[DIR_P0M] = &DD27[DIR_P0M * numberOfLBnodes];
+ D27.f[DIR_M0P] = &DD27[DIR_M0P * numberOfLBnodes];
+ D27.f[DIR_0PP] = &DD27[DIR_0PP * numberOfLBnodes];
+ D27.f[DIR_0MM] = &DD27[DIR_0MM * numberOfLBnodes];
+ D27.f[DIR_0PM] = &DD27[DIR_0PM * numberOfLBnodes];
+ D27.f[DIR_0MP] = &DD27[DIR_0MP * numberOfLBnodes];
+ D27.f[DIR_000] = &DD27[DIR_000 * numberOfLBnodes];
+ D27.f[DIR_PPP] = &DD27[DIR_PPP * numberOfLBnodes];
+ D27.f[DIR_MMP] = &DD27[DIR_MMP * numberOfLBnodes];
+ D27.f[DIR_PMP] = &DD27[DIR_PMP * numberOfLBnodes];
+ D27.f[DIR_MPP] = &DD27[DIR_MPP * numberOfLBnodes];
+ D27.f[DIR_PPM] = &DD27[DIR_PPM * numberOfLBnodes];
+ D27.f[DIR_MMM] = &DD27[DIR_MMM * numberOfLBnodes];
+ D27.f[DIR_PMM] = &DD27[DIR_PMM * numberOfLBnodes];
+ D27.f[DIR_MPM] = &DD27[DIR_MPM * numberOfLBnodes];
}
else
{
- D27.f[DIR_M00 ] = &DD27[DIR_P00 *size_Mat];
- D27.f[DIR_P00 ] = &DD27[DIR_M00 *size_Mat];
- D27.f[DIR_0M0 ] = &DD27[DIR_0P0 *size_Mat];
- D27.f[DIR_0P0 ] = &DD27[DIR_0M0 *size_Mat];
- D27.f[DIR_00M ] = &DD27[DIR_00P *size_Mat];
- D27.f[DIR_00P ] = &DD27[DIR_00M *size_Mat];
- D27.f[DIR_MM0 ] = &DD27[DIR_PP0 *size_Mat];
- D27.f[DIR_PP0 ] = &DD27[DIR_MM0 *size_Mat];
- D27.f[DIR_MP0 ] = &DD27[DIR_PM0 *size_Mat];
- D27.f[DIR_PM0 ] = &DD27[DIR_MP0 *size_Mat];
- D27.f[DIR_M0M ] = &DD27[DIR_P0P *size_Mat];
- D27.f[DIR_P0P ] = &DD27[DIR_M0M *size_Mat];
- D27.f[DIR_M0P ] = &DD27[DIR_P0M *size_Mat];
- D27.f[DIR_P0M ] = &DD27[DIR_M0P *size_Mat];
- D27.f[DIR_0MM ] = &DD27[DIR_0PP *size_Mat];
- D27.f[DIR_0PP ] = &DD27[DIR_0MM *size_Mat];
- D27.f[DIR_0MP ] = &DD27[DIR_0PM *size_Mat];
- D27.f[DIR_0PM ] = &DD27[DIR_0MP *size_Mat];
- D27.f[DIR_000] = &DD27[DIR_000*size_Mat];
- D27.f[DIR_PPP ] = &DD27[DIR_MMM *size_Mat];
- D27.f[DIR_MMP ] = &DD27[DIR_PPM *size_Mat];
- D27.f[DIR_PMP ] = &DD27[DIR_MPM *size_Mat];
- D27.f[DIR_MPP ] = &DD27[DIR_PMM *size_Mat];
- D27.f[DIR_PPM ] = &DD27[DIR_MMP *size_Mat];
- D27.f[DIR_MMM ] = &DD27[DIR_PPP *size_Mat];
- D27.f[DIR_PMM ] = &DD27[DIR_MPP *size_Mat];
- D27.f[DIR_MPM ] = &DD27[DIR_PMP *size_Mat];
+ D27.f[DIR_M00] = &DD27[DIR_P00 * numberOfLBnodes];
+ D27.f[DIR_P00] = &DD27[DIR_M00 * numberOfLBnodes];
+ D27.f[DIR_0M0] = &DD27[DIR_0P0 * numberOfLBnodes];
+ D27.f[DIR_0P0] = &DD27[DIR_0M0 * numberOfLBnodes];
+ D27.f[DIR_00M] = &DD27[DIR_00P * numberOfLBnodes];
+ D27.f[DIR_00P] = &DD27[DIR_00M * numberOfLBnodes];
+ D27.f[DIR_MM0] = &DD27[DIR_PP0 * numberOfLBnodes];
+ D27.f[DIR_PP0] = &DD27[DIR_MM0 * numberOfLBnodes];
+ D27.f[DIR_MP0] = &DD27[DIR_PM0 * numberOfLBnodes];
+ D27.f[DIR_PM0] = &DD27[DIR_MP0 * numberOfLBnodes];
+ D27.f[DIR_M0M] = &DD27[DIR_P0P * numberOfLBnodes];
+ D27.f[DIR_P0P] = &DD27[DIR_M0M * numberOfLBnodes];
+ D27.f[DIR_M0P] = &DD27[DIR_P0M * numberOfLBnodes];
+ D27.f[DIR_P0M] = &DD27[DIR_M0P * numberOfLBnodes];
+ D27.f[DIR_0MM] = &DD27[DIR_0PP * numberOfLBnodes];
+ D27.f[DIR_0PP] = &DD27[DIR_0MM * numberOfLBnodes];
+ D27.f[DIR_0MP] = &DD27[DIR_0PM * numberOfLBnodes];
+ D27.f[DIR_0PM] = &DD27[DIR_0MP * numberOfLBnodes];
+ D27.f[DIR_000] = &DD27[DIR_000 * numberOfLBnodes];
+ D27.f[DIR_PPP] = &DD27[DIR_MMM * numberOfLBnodes];
+ D27.f[DIR_MMP] = &DD27[DIR_PPM * numberOfLBnodes];
+ D27.f[DIR_PMP] = &DD27[DIR_MPM * numberOfLBnodes];
+ D27.f[DIR_MPP] = &DD27[DIR_PMM * numberOfLBnodes];
+ D27.f[DIR_PPM] = &DD27[DIR_MMP * numberOfLBnodes];
+ D27.f[DIR_MMM] = &DD27[DIR_PPP * numberOfLBnodes];
+ D27.f[DIR_PMM] = &DD27[DIR_MPP * numberOfLBnodes];
+ D27.f[DIR_MPM] = &DD27[DIR_PMP * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////
const unsigned x = threadIdx.x; // Globaler x-Index
@@ -6385,24 +6385,24 @@ __global__ void QADPressIncomp27(
*q_dirBE, *q_dirTW, *q_dirTN, *q_dirBS, *q_dirBN, *q_dirTS,
*q_dirTNE, *q_dirTSW, *q_dirTSE, *q_dirTNW, *q_dirBNE, *q_dirBSW,
*q_dirBSE, *q_dirBNW;
- q_dirE = &QQ[DIR_P00 * numberOfBCnodes];
- q_dirW = &QQ[DIR_M00 * numberOfBCnodes];
- q_dirN = &QQ[DIR_0P0 * numberOfBCnodes];
- q_dirS = &QQ[DIR_0M0 * numberOfBCnodes];
- q_dirT = &QQ[DIR_00P * numberOfBCnodes];
- q_dirB = &QQ[DIR_00M * numberOfBCnodes];
- q_dirNE = &QQ[DIR_PP0 * numberOfBCnodes];
- q_dirSW = &QQ[DIR_MM0 * numberOfBCnodes];
- q_dirSE = &QQ[DIR_PM0 * numberOfBCnodes];
- q_dirNW = &QQ[DIR_MP0 * numberOfBCnodes];
- q_dirTE = &QQ[DIR_P0P * numberOfBCnodes];
- q_dirBW = &QQ[DIR_M0M * numberOfBCnodes];
- q_dirBE = &QQ[DIR_P0M * numberOfBCnodes];
- q_dirTW = &QQ[DIR_M0P * numberOfBCnodes];
- q_dirTN = &QQ[DIR_0PP * numberOfBCnodes];
- q_dirBS = &QQ[DIR_0MM * numberOfBCnodes];
- q_dirBN = &QQ[DIR_0PM * numberOfBCnodes];
- q_dirTS = &QQ[DIR_0MP * numberOfBCnodes];
+ q_dirE = &QQ[DIR_P00 * numberOfBCnodes];
+ q_dirW = &QQ[DIR_M00 * numberOfBCnodes];
+ q_dirN = &QQ[DIR_0P0 * numberOfBCnodes];
+ q_dirS = &QQ[DIR_0M0 * numberOfBCnodes];
+ q_dirT = &QQ[DIR_00P * numberOfBCnodes];
+ q_dirB = &QQ[DIR_00M * numberOfBCnodes];
+ q_dirNE = &QQ[DIR_PP0 * numberOfBCnodes];
+ q_dirSW = &QQ[DIR_MM0 * numberOfBCnodes];
+ q_dirSE = &QQ[DIR_PM0 * numberOfBCnodes];
+ q_dirNW = &QQ[DIR_MP0 * numberOfBCnodes];
+ q_dirTE = &QQ[DIR_P0P * numberOfBCnodes];
+ q_dirBW = &QQ[DIR_M0M * numberOfBCnodes];
+ q_dirBE = &QQ[DIR_P0M * numberOfBCnodes];
+ q_dirTW = &QQ[DIR_M0P * numberOfBCnodes];
+ q_dirTN = &QQ[DIR_0PP * numberOfBCnodes];
+ q_dirBS = &QQ[DIR_0MM * numberOfBCnodes];
+ q_dirBN = &QQ[DIR_0PM * numberOfBCnodes];
+ q_dirTS = &QQ[DIR_0MP * numberOfBCnodes];
q_dirTNE = &QQ[DIR_PPP * numberOfBCnodes];
q_dirTSW = &QQ[DIR_MMP * numberOfBCnodes];
q_dirTSE = &QQ[DIR_PMP * numberOfBCnodes];
@@ -6442,65 +6442,65 @@ __global__ void QADPressIncomp27(
unsigned int ktne = KQK;
unsigned int kbsw = neighborZ[ksw];
////////////////////////////////////////////////////////////////////////////////
- real f_W = (D.f[DIR_P00 ])[ke ];
- real f_E = (D.f[DIR_M00 ])[kw ];
- real f_S = (D.f[DIR_0P0 ])[kn ];
- real f_N = (D.f[DIR_0M0 ])[ks ];
- real f_B = (D.f[DIR_00P ])[kt ];
- real f_T = (D.f[DIR_00M ])[kb ];
- real f_SW = (D.f[DIR_PP0 ])[kne ];
- real f_NE = (D.f[DIR_MM0 ])[ksw ];
- real f_NW = (D.f[DIR_PM0 ])[kse ];
- real f_SE = (D.f[DIR_MP0 ])[knw ];
- real f_BW = (D.f[DIR_P0P ])[kte ];
- real f_TE = (D.f[DIR_M0M ])[kbw ];
- real f_TW = (D.f[DIR_P0M ])[kbe ];
- real f_BE = (D.f[DIR_M0P ])[ktw ];
- real f_BS = (D.f[DIR_0PP ])[ktn ];
- real f_TN = (D.f[DIR_0MM ])[kbs ];
- real f_TS = (D.f[DIR_0PM ])[kbn ];
- real f_BN = (D.f[DIR_0MP ])[kts ];
+ real f_W = (D.f[DIR_P00])[ke ];
+ real f_E = (D.f[DIR_M00])[kw ];
+ real f_S = (D.f[DIR_0P0])[kn ];
+ real f_N = (D.f[DIR_0M0])[ks ];
+ real f_B = (D.f[DIR_00P])[kt ];
+ real f_T = (D.f[DIR_00M])[kb ];
+ real f_SW = (D.f[DIR_PP0])[kne ];
+ real f_NE = (D.f[DIR_MM0])[ksw ];
+ real f_NW = (D.f[DIR_PM0])[kse ];
+ real f_SE = (D.f[DIR_MP0])[knw ];
+ real f_BW = (D.f[DIR_P0P])[kte ];
+ real f_TE = (D.f[DIR_M0M])[kbw ];
+ real f_TW = (D.f[DIR_P0M])[kbe ];
+ real f_BE = (D.f[DIR_M0P])[ktw ];
+ real f_BS = (D.f[DIR_0PP])[ktn ];
+ real f_TN = (D.f[DIR_0MM])[kbs ];
+ real f_TS = (D.f[DIR_0PM])[kbn ];
+ real f_BN = (D.f[DIR_0MP])[kts ];
//real f_ZERO = (D.f[DIR_000])[kzero];
- real f_BSW = (D.f[DIR_PPP ])[ktne ];
- real f_BNE = (D.f[DIR_MMP ])[ktsw ];
- real f_BNW = (D.f[DIR_PMP ])[ktse ];
- real f_BSE = (D.f[DIR_MPP ])[ktnw ];
- real f_TSW = (D.f[DIR_PPM ])[kbne ];
- real f_TNE = (D.f[DIR_MMM ])[kbsw ];
- real f_TNW = (D.f[DIR_PMM ])[kbse ];
- real f_TSE = (D.f[DIR_MPM ])[kbnw ];
+ real f_BSW = (D.f[DIR_PPP])[ktne ];
+ real f_BNE = (D.f[DIR_MMP])[ktsw ];
+ real f_BNW = (D.f[DIR_PMP])[ktse ];
+ real f_BSE = (D.f[DIR_MPP])[ktnw ];
+ real f_TSW = (D.f[DIR_PPM])[kbne ];
+ real f_TNE = (D.f[DIR_MMM])[kbsw ];
+ real f_TNW = (D.f[DIR_PMM])[kbse ];
+ real f_TSE = (D.f[DIR_MPM])[kbnw ];
////////////////////////////////////////////////////////////////////////////////
real vx1 = ((f_TNE-f_BSW)+(f_BNE-f_TSW)+(f_TSE-f_BNW)+(f_BSE-f_TNW) +(f_NE-f_SW)+(f_SE-f_NW)+(f_TE-f_BW)+(f_BE-f_TW)+(f_E-f_W));
real vx2 = ((f_TNE-f_BSW)+(f_BNE-f_TSW)+(f_BNW-f_TSE)+(f_TNW-f_BSE) +(f_NE-f_SW)+(f_NW-f_SE)+(f_TN-f_BS)+(f_BN-f_TS)+(f_N-f_S));
real vx3 = ((f_TNE-f_BSW)+(f_TSW-f_BNE)+(f_TSE-f_BNW)+(f_TNW-f_BSE) +(f_TE-f_BW)+(f_TW-f_BE)+(f_TN-f_BS)+(f_TS-f_BN)+(f_T-f_B));
////////////////////////////////////////////////////////////////////////////////
- //real f27_W = (D27.f[DIR_P00 ])[ke ];
- //real f27_E = (D27.f[DIR_M00 ])[kw ];
- //real f27_S = (D27.f[DIR_0P0 ])[kn ];
- //real f27_N = (D27.f[DIR_0M0 ])[ks ];
- //real f27_B = (D27.f[DIR_00P ])[kt ];
- //real f27_T = (D27.f[DIR_00M ])[kb ];
- //real f27_SW = (D27.f[DIR_PP0 ])[kne ];
- //real f27_NE = (D27.f[DIR_MM0 ])[ksw ];
- //real f27_NW = (D27.f[DIR_PM0 ])[kse ];
- //real f27_SE = (D27.f[DIR_MP0 ])[knw ];
- //real f27_BW = (D27.f[DIR_P0P ])[kte ];
- //real f27_TE = (D27.f[DIR_M0M ])[kbw ];
- //real f27_TW = (D27.f[DIR_P0M ])[kbe ];
- //real f27_BE = (D27.f[DIR_M0P ])[ktw ];
- //real f27_BS = (D27.f[DIR_0PP ])[ktn ];
- //real f27_TN = (D27.f[DIR_0MM ])[kbs ];
- //real f27_TS = (D27.f[DIR_0PM ])[kbn ];
- //real f27_BN = (D27.f[DIR_0MP ])[kts ];
+ //real f27_W = (D27.f[DIR_P00])[ke ];
+ //real f27_E = (D27.f[DIR_M00])[kw ];
+ //real f27_S = (D27.f[DIR_0P0])[kn ];
+ //real f27_N = (D27.f[DIR_0M0])[ks ];
+ //real f27_B = (D27.f[DIR_00P])[kt ];
+ //real f27_T = (D27.f[DIR_00M])[kb ];
+ //real f27_SW = (D27.f[DIR_PP0])[kne ];
+ //real f27_NE = (D27.f[DIR_MM0])[ksw ];
+ //real f27_NW = (D27.f[DIR_PM0])[kse ];
+ //real f27_SE = (D27.f[DIR_MP0])[knw ];
+ //real f27_BW = (D27.f[DIR_P0P])[kte ];
+ //real f27_TE = (D27.f[DIR_M0M])[kbw ];
+ //real f27_TW = (D27.f[DIR_P0M])[kbe ];
+ //real f27_BE = (D27.f[DIR_M0P])[ktw ];
+ //real f27_BS = (D27.f[DIR_0PP])[ktn ];
+ //real f27_TN = (D27.f[DIR_0MM])[kbs ];
+ //real f27_TS = (D27.f[DIR_0PM])[kbn ];
+ //real f27_BN = (D27.f[DIR_0MP])[kts ];
//real f27_ZERO = (D27.f[DIR_000])[kzero];
- //real f27_BSW = (D27.f[DIR_PPP ])[ktne ];
- //real f27_BNE = (D27.f[DIR_MMP ])[ktsw ];
- //real f27_BNW = (D27.f[DIR_PMP ])[ktse ];
- //real f27_BSE = (D27.f[DIR_MPP ])[ktnw ];
- //real f27_TSW = (D27.f[DIR_PPM ])[kbne ];
- //real f27_TNE = (D27.f[DIR_MMM ])[kbsw ];
- //real f27_TNW = (D27.f[DIR_PMM ])[kbse ];
- //real f27_TSE = (D27.f[DIR_MPM ])[kbnw ];
+ //real f27_BSW = (D27.f[DIR_PPP])[ktne ];
+ //real f27_BNE = (D27.f[DIR_MMP])[ktsw ];
+ //real f27_BNW = (D27.f[DIR_PMP])[ktse ];
+ //real f27_BSE = (D27.f[DIR_MPP])[ktnw ];
+ //real f27_TSW = (D27.f[DIR_PPM])[kbne ];
+ //real f27_TNE = (D27.f[DIR_MMM])[kbsw ];
+ //real f27_TNW = (D27.f[DIR_PMM])[kbse ];
+ //real f27_TSE = (D27.f[DIR_MPM])[kbnw ];
////////////////////////////////////////////////////////////////////////////////
real cu_sq=c3o2*(vx1*vx1+vx2*vx2+vx3*vx3);
////////////////////////////////////////////////////////////////////////////////
@@ -6581,63 +6581,63 @@ __global__ void QADPressIncomp27(
//////////////////////////////////////////////////////////////////////////
if (isEvenTimestep==false)
{
- D27.f[DIR_P00 ] = &DD27[DIR_P00 *size_Mat];
- D27.f[DIR_M00 ] = &DD27[DIR_M00 *size_Mat];
- D27.f[DIR_0P0 ] = &DD27[DIR_0P0 *size_Mat];
- D27.f[DIR_0M0 ] = &DD27[DIR_0M0 *size_Mat];
- D27.f[DIR_00P ] = &DD27[DIR_00P *size_Mat];
- D27.f[DIR_00M ] = &DD27[DIR_00M *size_Mat];
- D27.f[DIR_PP0 ] = &DD27[DIR_PP0 *size_Mat];
- D27.f[DIR_MM0 ] = &DD27[DIR_MM0 *size_Mat];
- D27.f[DIR_PM0 ] = &DD27[DIR_PM0 *size_Mat];
- D27.f[DIR_MP0 ] = &DD27[DIR_MP0 *size_Mat];
- D27.f[DIR_P0P ] = &DD27[DIR_P0P *size_Mat];
- D27.f[DIR_M0M ] = &DD27[DIR_M0M *size_Mat];
- D27.f[DIR_P0M ] = &DD27[DIR_P0M *size_Mat];
- D27.f[DIR_M0P ] = &DD27[DIR_M0P *size_Mat];
- D27.f[DIR_0PP ] = &DD27[DIR_0PP *size_Mat];
- D27.f[DIR_0MM ] = &DD27[DIR_0MM *size_Mat];
- D27.f[DIR_0PM ] = &DD27[DIR_0PM *size_Mat];
- D27.f[DIR_0MP ] = &DD27[DIR_0MP *size_Mat];
- D27.f[DIR_000] = &DD27[DIR_000*size_Mat];
- D27.f[DIR_PPP ] = &DD27[DIR_PPP *size_Mat];
- D27.f[DIR_MMP ] = &DD27[DIR_MMP *size_Mat];
- D27.f[DIR_PMP ] = &DD27[DIR_PMP *size_Mat];
- D27.f[DIR_MPP ] = &DD27[DIR_MPP *size_Mat];
- D27.f[DIR_PPM ] = &DD27[DIR_PPM *size_Mat];
- D27.f[DIR_MMM ] = &DD27[DIR_MMM *size_Mat];
- D27.f[DIR_PMM ] = &DD27[DIR_PMM *size_Mat];
- D27.f[DIR_MPM ] = &DD27[DIR_MPM *size_Mat];
+ D27.f[DIR_P00] = &DD27[DIR_P00 * numberOfLBnodes];
+ D27.f[DIR_M00] = &DD27[DIR_M00 * numberOfLBnodes];
+ D27.f[DIR_0P0] = &DD27[DIR_0P0 * numberOfLBnodes];
+ D27.f[DIR_0M0] = &DD27[DIR_0M0 * numberOfLBnodes];
+ D27.f[DIR_00P] = &DD27[DIR_00P * numberOfLBnodes];
+ D27.f[DIR_00M] = &DD27[DIR_00M * numberOfLBnodes];
+ D27.f[DIR_PP0] = &DD27[DIR_PP0 * numberOfLBnodes];
+ D27.f[DIR_MM0] = &DD27[DIR_MM0 * numberOfLBnodes];
+ D27.f[DIR_PM0] = &DD27[DIR_PM0 * numberOfLBnodes];
+ D27.f[DIR_MP0] = &DD27[DIR_MP0 * numberOfLBnodes];
+ D27.f[DIR_P0P] = &DD27[DIR_P0P * numberOfLBnodes];
+ D27.f[DIR_M0M] = &DD27[DIR_M0M * numberOfLBnodes];
+ D27.f[DIR_P0M] = &DD27[DIR_P0M * numberOfLBnodes];
+ D27.f[DIR_M0P] = &DD27[DIR_M0P * numberOfLBnodes];
+ D27.f[DIR_0PP] = &DD27[DIR_0PP * numberOfLBnodes];
+ D27.f[DIR_0MM] = &DD27[DIR_0MM * numberOfLBnodes];
+ D27.f[DIR_0PM] = &DD27[DIR_0PM * numberOfLBnodes];
+ D27.f[DIR_0MP] = &DD27[DIR_0MP * numberOfLBnodes];
+ D27.f[DIR_000] = &DD27[DIR_000 * numberOfLBnodes];
+ D27.f[DIR_PPP] = &DD27[DIR_PPP * numberOfLBnodes];
+ D27.f[DIR_MMP] = &DD27[DIR_MMP * numberOfLBnodes];
+ D27.f[DIR_PMP] = &DD27[DIR_PMP * numberOfLBnodes];
+ D27.f[DIR_MPP] = &DD27[DIR_MPP * numberOfLBnodes];
+ D27.f[DIR_PPM] = &DD27[DIR_PPM * numberOfLBnodes];
+ D27.f[DIR_MMM] = &DD27[DIR_MMM * numberOfLBnodes];
+ D27.f[DIR_PMM] = &DD27[DIR_PMM * numberOfLBnodes];
+ D27.f[DIR_MPM] = &DD27[DIR_MPM * numberOfLBnodes];
}
else
{
- D27.f[DIR_M00 ] = &DD27[DIR_P00 *size_Mat];
- D27.f[DIR_P00 ] = &DD27[DIR_M00 *size_Mat];
- D27.f[DIR_0M0 ] = &DD27[DIR_0P0 *size_Mat];
- D27.f[DIR_0P0 ] = &DD27[DIR_0M0 *size_Mat];
- D27.f[DIR_00M ] = &DD27[DIR_00P *size_Mat];
- D27.f[DIR_00P ] = &DD27[DIR_00M *size_Mat];
- D27.f[DIR_MM0 ] = &DD27[DIR_PP0 *size_Mat];
- D27.f[DIR_PP0 ] = &DD27[DIR_MM0 *size_Mat];
- D27.f[DIR_MP0 ] = &DD27[DIR_PM0 *size_Mat];
- D27.f[DIR_PM0 ] = &DD27[DIR_MP0 *size_Mat];
- D27.f[DIR_M0M ] = &DD27[DIR_P0P *size_Mat];
- D27.f[DIR_P0P ] = &DD27[DIR_M0M *size_Mat];
- D27.f[DIR_M0P ] = &DD27[DIR_P0M *size_Mat];
- D27.f[DIR_P0M ] = &DD27[DIR_M0P *size_Mat];
- D27.f[DIR_0MM ] = &DD27[DIR_0PP *size_Mat];
- D27.f[DIR_0PP ] = &DD27[DIR_0MM *size_Mat];
- D27.f[DIR_0MP ] = &DD27[DIR_0PM *size_Mat];
- D27.f[DIR_0PM ] = &DD27[DIR_0MP *size_Mat];
- D27.f[DIR_000] = &DD27[DIR_000*size_Mat];
- D27.f[DIR_PPP ] = &DD27[DIR_MMM *size_Mat];
- D27.f[DIR_MMP ] = &DD27[DIR_PPM *size_Mat];
- D27.f[DIR_PMP ] = &DD27[DIR_MPM *size_Mat];
- D27.f[DIR_MPP ] = &DD27[DIR_PMM *size_Mat];
- D27.f[DIR_PPM ] = &DD27[DIR_MMP *size_Mat];
- D27.f[DIR_MMM ] = &DD27[DIR_PPP *size_Mat];
- D27.f[DIR_PMM ] = &DD27[DIR_MPP *size_Mat];
- D27.f[DIR_MPM ] = &DD27[DIR_PMP *size_Mat];
+ D27.f[DIR_M00] = &DD27[DIR_P00 * numberOfLBnodes];
+ D27.f[DIR_P00] = &DD27[DIR_M00 * numberOfLBnodes];
+ D27.f[DIR_0M0] = &DD27[DIR_0P0 * numberOfLBnodes];
+ D27.f[DIR_0P0] = &DD27[DIR_0M0 * numberOfLBnodes];
+ D27.f[DIR_00M] = &DD27[DIR_00P * numberOfLBnodes];
+ D27.f[DIR_00P] = &DD27[DIR_00M * numberOfLBnodes];
+ D27.f[DIR_MM0] = &DD27[DIR_PP0 * numberOfLBnodes];
+ D27.f[DIR_PP0] = &DD27[DIR_MM0 * numberOfLBnodes];
+ D27.f[DIR_MP0] = &DD27[DIR_PM0 * numberOfLBnodes];
+ D27.f[DIR_PM0] = &DD27[DIR_MP0 * numberOfLBnodes];
+ D27.f[DIR_M0M] = &DD27[DIR_P0P * numberOfLBnodes];
+ D27.f[DIR_P0P] = &DD27[DIR_M0M * numberOfLBnodes];
+ D27.f[DIR_M0P] = &DD27[DIR_P0M * numberOfLBnodes];
+ D27.f[DIR_P0M] = &DD27[DIR_M0P * numberOfLBnodes];
+ D27.f[DIR_0MM] = &DD27[DIR_0PP * numberOfLBnodes];
+ D27.f[DIR_0PP] = &DD27[DIR_0MM * numberOfLBnodes];
+ D27.f[DIR_0MP] = &DD27[DIR_0PM * numberOfLBnodes];
+ D27.f[DIR_0PM] = &DD27[DIR_0MP * numberOfLBnodes];
+ D27.f[DIR_000] = &DD27[DIR_000 * numberOfLBnodes];
+ D27.f[DIR_PPP] = &DD27[DIR_MMM * numberOfLBnodes];
+ D27.f[DIR_MMP] = &DD27[DIR_PPM * numberOfLBnodes];
+ D27.f[DIR_PMP] = &DD27[DIR_MPM * numberOfLBnodes];
+ D27.f[DIR_MPP] = &DD27[DIR_PMM * numberOfLBnodes];
+ D27.f[DIR_PPM] = &DD27[DIR_MMP * numberOfLBnodes];
+ D27.f[DIR_MMM] = &DD27[DIR_PPP * numberOfLBnodes];
+ D27.f[DIR_PMM] = &DD27[DIR_MPP * numberOfLBnodes];
+ D27.f[DIR_MPM] = &DD27[DIR_PMP * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
//Test
@@ -6645,24 +6645,24 @@ __global__ void QADPressIncomp27(
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
real q;
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
- q = q_dirE[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_M00 ])[kw ]= -feqW27_W + c2o1 * c2o27 * TempD;
- q = q_dirW[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_P00 ])[ke ]= -feqW27_E + c2o1 * c2o27 * TempD;
- q = q_dirN[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_0M0 ])[ks ]= -feqW27_S + c2o1 * c2o27 * TempD;
- q = q_dirS[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_0P0 ])[kn ]= -feqW27_N + c2o1 * c2o27 * TempD;
- q = q_dirT[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_00M ])[kb ]= -feqW27_B + c2o1 * c2o27 * TempD;
- q = q_dirB[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_00P ])[kt ]= -feqW27_T + c2o1 * c2o27 * TempD;
- q = q_dirNE[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_MM0 ])[ksw ]= -feqW27_SW + c2o1 * c1o54 * TempD;
- q = q_dirSW[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_PP0 ])[kne ]= -feqW27_NE + c2o1 * c1o54 * TempD;
- q = q_dirSE[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_MP0 ])[knw ]= -feqW27_NW + c2o1 * c1o54 * TempD;
- q = q_dirNW[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_PM0 ])[kse ]= -feqW27_SE + c2o1 * c1o54 * TempD;
- q = q_dirTE[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_M0M ])[kbw ]= -feqW27_BW + c2o1 * c1o54 * TempD;
- q = q_dirBW[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_P0P ])[kte ]= -feqW27_TE + c2o1 * c1o54 * TempD;
- q = q_dirBE[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_M0P ])[ktw ]= -feqW27_TW + c2o1 * c1o54 * TempD;
- q = q_dirTW[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_P0M ])[kbe ]= -feqW27_BE + c2o1 * c1o54 * TempD;
- q = q_dirTN[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_0MM ])[kbs ]= -feqW27_BS + c2o1 * c1o54 * TempD;
- q = q_dirBS[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_0PP ])[ktn ]= -feqW27_TN + c2o1 * c1o54 * TempD;
- q = q_dirBN[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_0MP ])[kts ]= -feqW27_TS + c2o1 * c1o54 * TempD;
- q = q_dirTS[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_0PM ])[kbn ]= -feqW27_BN + c2o1 * c1o54 * TempD;
+ q = q_dirE[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_M00])[kw ]= -feqW27_W + c2o1 * c2o27 * TempD;
+ q = q_dirW[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_P00])[ke ]= -feqW27_E + c2o1 * c2o27 * TempD;
+ q = q_dirN[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_0M0])[ks ]= -feqW27_S + c2o1 * c2o27 * TempD;
+ q = q_dirS[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_0P0])[kn ]= -feqW27_N + c2o1 * c2o27 * TempD;
+ q = q_dirT[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_00M])[kb ]= -feqW27_B + c2o1 * c2o27 * TempD;
+ q = q_dirB[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_00P])[kt ]= -feqW27_T + c2o1 * c2o27 * TempD;
+ q = q_dirNE[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_MM0])[ksw ]= -feqW27_SW + c2o1 * c1o54 * TempD;
+ q = q_dirSW[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_PP0])[kne ]= -feqW27_NE + c2o1 * c1o54 * TempD;
+ q = q_dirSE[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_MP0])[knw ]= -feqW27_NW + c2o1 * c1o54 * TempD;
+ q = q_dirNW[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_PM0])[kse ]= -feqW27_SE + c2o1 * c1o54 * TempD;
+ q = q_dirTE[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_M0M])[kbw ]= -feqW27_BW + c2o1 * c1o54 * TempD;
+ q = q_dirBW[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_P0P])[kte ]= -feqW27_TE + c2o1 * c1o54 * TempD;
+ q = q_dirBE[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_M0P])[ktw ]= -feqW27_TW + c2o1 * c1o54 * TempD;
+ q = q_dirTW[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_P0M])[kbe ]= -feqW27_BE + c2o1 * c1o54 * TempD;
+ q = q_dirTN[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_0MM])[kbs ]= -feqW27_BS + c2o1 * c1o54 * TempD;
+ q = q_dirBS[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_0PP])[ktn ]= -feqW27_TN + c2o1 * c1o54 * TempD;
+ q = q_dirBN[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_0MP])[kts ]= -feqW27_TS + c2o1 * c1o54 * TempD;
+ q = q_dirTS[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_0PM])[kbn ]= -feqW27_BN + c2o1 * c1o54 * TempD;
q = q_dirTNE[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_MMM])[kbsw]= -feqW27_BSW+ c2o1 * c1o216 * TempD;
q = q_dirBSW[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_PPP])[ktne]= -feqW27_TNE+ c2o1 * c1o216 * TempD;
q = q_dirBNE[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_MMP])[ktsw]= -feqW27_TSW+ c2o1 * c1o216 * TempD;
@@ -6671,24 +6671,24 @@ __global__ void QADPressIncomp27(
q = q_dirBNW[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_PMP])[ktse]= -feqW27_TSE+ c2o1 * c1o216 * TempD;
q = q_dirBSE[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_MPP])[ktnw]= -feqW27_TNW+ c2o1 * c1o216 * TempD;
q = q_dirTNW[k]; if (q>=c0o1 && q<=c1o1) (D27.f[DIR_PMM])[kbse]= -feqW27_BSE+ c2o1 * c1o216 * TempD;
- //q = q_dirE[k]; if (q>=zero && q<=one) (D27.f[DIR_M00 ])[kw ]=(two*feqW27_W -(f27_E *(q*omegaD-one)-omegaD*feq27_E *(q-one))/(omegaD-one)+f27_W *q)/(q+one);
- //q = q_dirW[k]; if (q>=zero && q<=one) (D27.f[DIR_P00 ])[ke ]=(two*feqW27_E -(f27_W *(q*omegaD-one)-omegaD*feq27_W *(q-one))/(omegaD-one)+f27_E *q)/(q+one);
- //q = q_dirN[k]; if (q>=zero && q<=one) (D27.f[DIR_0M0 ])[ks ]=(two*feqW27_S -(f27_N *(q*omegaD-one)-omegaD*feq27_N *(q-one))/(omegaD-one)+f27_S *q)/(q+one);
- //q = q_dirS[k]; if (q>=zero && q<=one) (D27.f[DIR_0P0 ])[kn ]=(two*feqW27_N -(f27_S *(q*omegaD-one)-omegaD*feq27_S *(q-one))/(omegaD-one)+f27_N *q)/(q+one);
- //q = q_dirT[k]; if (q>=zero && q<=one) (D27.f[DIR_00M ])[kb ]=(two*feqW27_B -(f27_T *(q*omegaD-one)-omegaD*feq27_T *(q-one))/(omegaD-one)+f27_B *q)/(q+one);
- //q = q_dirB[k]; if (q>=zero && q<=one) (D27.f[DIR_00P ])[kt ]=(two*feqW27_T -(f27_B *(q*omegaD-one)-omegaD*feq27_B *(q-one))/(omegaD-one)+f27_T *q)/(q+one);
- //q = q_dirNE[k]; if (q>=zero && q<=one) (D27.f[DIR_MM0 ])[ksw ]=(two*feqW27_SW -(f27_NE *(q*omegaD-one)-omegaD*feq27_NE *(q-one))/(omegaD-one)+f27_SW *q)/(q+one);
- //q = q_dirSW[k]; if (q>=zero && q<=one) (D27.f[DIR_PP0 ])[kne ]=(two*feqW27_NE -(f27_SW *(q*omegaD-one)-omegaD*feq27_SW *(q-one))/(omegaD-one)+f27_NE *q)/(q+one);
- //q = q_dirSE[k]; if (q>=zero && q<=one) (D27.f[DIR_MP0 ])[knw ]=(two*feqW27_NW -(f27_SE *(q*omegaD-one)-omegaD*feq27_SE *(q-one))/(omegaD-one)+f27_NW *q)/(q+one);
- //q = q_dirNW[k]; if (q>=zero && q<=one) (D27.f[DIR_PM0 ])[kse ]=(two*feqW27_SE -(f27_NW *(q*omegaD-one)-omegaD*feq27_NW *(q-one))/(omegaD-one)+f27_SE *q)/(q+one);
- //q = q_dirTE[k]; if (q>=zero && q<=one) (D27.f[DIR_M0M ])[kbw ]=(two*feqW27_BW -(f27_TE *(q*omegaD-one)-omegaD*feq27_TE *(q-one))/(omegaD-one)+f27_BW *q)/(q+one);
- //q = q_dirBW[k]; if (q>=zero && q<=one) (D27.f[DIR_P0P ])[kte ]=(two*feqW27_TE -(f27_BW *(q*omegaD-one)-omegaD*feq27_BW *(q-one))/(omegaD-one)+f27_TE *q)/(q+one);
- //q = q_dirBE[k]; if (q>=zero && q<=one) (D27.f[DIR_M0P ])[ktw ]=(two*feqW27_TW -(f27_BE *(q*omegaD-one)-omegaD*feq27_BE *(q-one))/(omegaD-one)+f27_TW *q)/(q+one);
- //q = q_dirTW[k]; if (q>=zero && q<=one) (D27.f[DIR_P0M ])[kbe ]=(two*feqW27_BE -(f27_TW *(q*omegaD-one)-omegaD*feq27_TW *(q-one))/(omegaD-one)+f27_BE *q)/(q+one);
- //q = q_dirTN[k]; if (q>=zero && q<=one) (D27.f[DIR_0MM ])[kbs ]=(two*feqW27_BS -(f27_TN *(q*omegaD-one)-omegaD*feq27_TN *(q-one))/(omegaD-one)+f27_BS *q)/(q+one);
- //q = q_dirBS[k]; if (q>=zero && q<=one) (D27.f[DIR_0PP ])[ktn ]=(two*feqW27_TN -(f27_BS *(q*omegaD-one)-omegaD*feq27_BS *(q-one))/(omegaD-one)+f27_TN *q)/(q+one);
- //q = q_dirBN[k]; if (q>=zero && q<=one) (D27.f[DIR_0MP ])[kts ]=(two*feqW27_TS -(f27_BN *(q*omegaD-one)-omegaD*feq27_BN *(q-one))/(omegaD-one)+f27_TS *q)/(q+one);
- //q = q_dirTS[k]; if (q>=zero && q<=one) (D27.f[DIR_0PM ])[kbn ]=(two*feqW27_BN -(f27_TS *(q*omegaD-one)-omegaD*feq27_TS *(q-one))/(omegaD-one)+f27_BN *q)/(q+one);
+ //q = q_dirE[k]; if (q>=zero && q<=one) (D27.f[DIR_M00])[kw ]=(two*feqW27_W -(f27_E *(q*omegaD-one)-omegaD*feq27_E *(q-one))/(omegaD-one)+f27_W *q)/(q+one);
+ //q = q_dirW[k]; if (q>=zero && q<=one) (D27.f[DIR_P00])[ke ]=(two*feqW27_E -(f27_W *(q*omegaD-one)-omegaD*feq27_W *(q-one))/(omegaD-one)+f27_E *q)/(q+one);
+ //q = q_dirN[k]; if (q>=zero && q<=one) (D27.f[DIR_0M0])[ks ]=(two*feqW27_S -(f27_N *(q*omegaD-one)-omegaD*feq27_N *(q-one))/(omegaD-one)+f27_S *q)/(q+one);
+ //q = q_dirS[k]; if (q>=zero && q<=one) (D27.f[DIR_0P0])[kn ]=(two*feqW27_N -(f27_S *(q*omegaD-one)-omegaD*feq27_S *(q-one))/(omegaD-one)+f27_N *q)/(q+one);
+ //q = q_dirT[k]; if (q>=zero && q<=one) (D27.f[DIR_00M])[kb ]=(two*feqW27_B -(f27_T *(q*omegaD-one)-omegaD*feq27_T *(q-one))/(omegaD-one)+f27_B *q)/(q+one);
+ //q = q_dirB[k]; if (q>=zero && q<=one) (D27.f[DIR_00P])[kt ]=(two*feqW27_T -(f27_B *(q*omegaD-one)-omegaD*feq27_B *(q-one))/(omegaD-one)+f27_T *q)/(q+one);
+ //q = q_dirNE[k]; if (q>=zero && q<=one) (D27.f[DIR_MM0])[ksw ]=(two*feqW27_SW -(f27_NE *(q*omegaD-one)-omegaD*feq27_NE *(q-one))/(omegaD-one)+f27_SW *q)/(q+one);
+ //q = q_dirSW[k]; if (q>=zero && q<=one) (D27.f[DIR_PP0])[kne ]=(two*feqW27_NE -(f27_SW *(q*omegaD-one)-omegaD*feq27_SW *(q-one))/(omegaD-one)+f27_NE *q)/(q+one);
+ //q = q_dirSE[k]; if (q>=zero && q<=one) (D27.f[DIR_MP0])[knw ]=(two*feqW27_NW -(f27_SE *(q*omegaD-one)-omegaD*feq27_SE *(q-one))/(omegaD-one)+f27_NW *q)/(q+one);
+ //q = q_dirNW[k]; if (q>=zero && q<=one) (D27.f[DIR_PM0])[kse ]=(two*feqW27_SE -(f27_NW *(q*omegaD-one)-omegaD*feq27_NW *(q-one))/(omegaD-one)+f27_SE *q)/(q+one);
+ //q = q_dirTE[k]; if (q>=zero && q<=one) (D27.f[DIR_M0M])[kbw ]=(two*feqW27_BW -(f27_TE *(q*omegaD-one)-omegaD*feq27_TE *(q-one))/(omegaD-one)+f27_BW *q)/(q+one);
+ //q = q_dirBW[k]; if (q>=zero && q<=one) (D27.f[DIR_P0P])[kte ]=(two*feqW27_TE -(f27_BW *(q*omegaD-one)-omegaD*feq27_BW *(q-one))/(omegaD-one)+f27_TE *q)/(q+one);
+ //q = q_dirBE[k]; if (q>=zero && q<=one) (D27.f[DIR_M0P])[ktw ]=(two*feqW27_TW -(f27_BE *(q*omegaD-one)-omegaD*feq27_BE *(q-one))/(omegaD-one)+f27_TW *q)/(q+one);
+ //q = q_dirTW[k]; if (q>=zero && q<=one) (D27.f[DIR_P0M])[kbe ]=(two*feqW27_BE -(f27_TW *(q*omegaD-one)-omegaD*feq27_TW *(q-one))/(omegaD-one)+f27_BE *q)/(q+one);
+ //q = q_dirTN[k]; if (q>=zero && q<=one) (D27.f[DIR_0MM])[kbs ]=(two*feqW27_BS -(f27_TN *(q*omegaD-one)-omegaD*feq27_TN *(q-one))/(omegaD-one)+f27_BS *q)/(q+one);
+ //q = q_dirBS[k]; if (q>=zero && q<=one) (D27.f[DIR_0PP])[ktn ]=(two*feqW27_TN -(f27_BS *(q*omegaD-one)-omegaD*feq27_BS *(q-one))/(omegaD-one)+f27_TN *q)/(q+one);
+ //q = q_dirBN[k]; if (q>=zero && q<=one) (D27.f[DIR_0MP])[kts ]=(two*feqW27_TS -(f27_BN *(q*omegaD-one)-omegaD*feq27_BN *(q-one))/(omegaD-one)+f27_TS *q)/(q+one);
+ //q = q_dirTS[k]; if (q>=zero && q<=one) (D27.f[DIR_0PM])[kbn ]=(two*feqW27_BN -(f27_TS *(q*omegaD-one)-omegaD*feq27_TS *(q-one))/(omegaD-one)+f27_BN *q)/(q+one);
//q = q_dirTNE[k]; if (q>=zero && q<=one) (D27.f[DIR_MMM])[kbsw]=(two*feqW27_BSW-(f27_TNE*(q*omegaD-one)-omegaD*feq27_TNE*(q-one))/(omegaD-one)+f27_BSW*q)/(q+one);
//q = q_dirBSW[k]; if (q>=zero && q<=one) (D27.f[DIR_PPP])[ktne]=(two*feqW27_TNE-(f27_BSW*(q*omegaD-one)-omegaD*feq27_BSW*(q-one))/(omegaD-one)+f27_TNE*q)/(q+one);
//q = q_dirBNE[k]; if (q>=zero && q<=one) (D27.f[DIR_MMP])[ktsw]=(two*feqW27_TSW-(f27_BNE*(q*omegaD-one)-omegaD*feq27_BNE*(q-one))/(omegaD-one)+f27_TSW*q)/(q+one);
@@ -6739,131 +6739,131 @@ __global__ void AD_SlipVelDeviceComp(
uint* neighborX,
uint* neighborY,
uint* neighborZ,
- uint size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep)
{
Distributions27 D;
if (isEvenTimestep)
{
- D.f[DIR_P00 ] = &distributions[DIR_P00 * size_Mat];
- D.f[DIR_M00 ] = &distributions[DIR_M00 * size_Mat];
- D.f[DIR_0P0 ] = &distributions[DIR_0P0 * size_Mat];
- D.f[DIR_0M0 ] = &distributions[DIR_0M0 * size_Mat];
- D.f[DIR_00P ] = &distributions[DIR_00P * size_Mat];
- D.f[DIR_00M ] = &distributions[DIR_00M * size_Mat];
- D.f[DIR_PP0 ] = &distributions[DIR_PP0 * size_Mat];
- D.f[DIR_MM0 ] = &distributions[DIR_MM0 * size_Mat];
- D.f[DIR_PM0 ] = &distributions[DIR_PM0 * size_Mat];
- D.f[DIR_MP0 ] = &distributions[DIR_MP0 * size_Mat];
- D.f[DIR_P0P ] = &distributions[DIR_P0P * size_Mat];
- D.f[DIR_M0M ] = &distributions[DIR_M0M * size_Mat];
- D.f[DIR_P0M ] = &distributions[DIR_P0M * size_Mat];
- D.f[DIR_M0P ] = &distributions[DIR_M0P * size_Mat];
- D.f[DIR_0PP ] = &distributions[DIR_0PP * size_Mat];
- D.f[DIR_0MM ] = &distributions[DIR_0MM * size_Mat];
- D.f[DIR_0PM ] = &distributions[DIR_0PM * size_Mat];
- D.f[DIR_0MP ] = &distributions[DIR_0MP * size_Mat];
- D.f[DIR_000] = &distributions[DIR_000 * size_Mat];
- D.f[DIR_PPP ] = &distributions[DIR_PPP * size_Mat];
- D.f[DIR_MMP ] = &distributions[DIR_MMP * size_Mat];
- D.f[DIR_PMP ] = &distributions[DIR_PMP * size_Mat];
- D.f[DIR_MPP ] = &distributions[DIR_MPP * size_Mat];
- D.f[DIR_PPM ] = &distributions[DIR_PPM * size_Mat];
- D.f[DIR_MMM ] = &distributions[DIR_MMM * size_Mat];
- D.f[DIR_PMM ] = &distributions[DIR_PMM * size_Mat];
- D.f[DIR_MPM ] = &distributions[DIR_MPM * size_Mat];
+ D.f[DIR_P00] = &distributions[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &distributions[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &distributions[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &distributions[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &distributions[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &distributions[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &distributions[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &distributions[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &distributions[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &distributions[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &distributions[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &distributions[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &distributions[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &distributions[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &distributions[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &distributions[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &distributions[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &distributions[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &distributions[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &distributions[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &distributions[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &distributions[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &distributions[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &distributions[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &distributions[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &distributions[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &distributions[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00 ] = &distributions[DIR_P00 * size_Mat];
- D.f[DIR_P00 ] = &distributions[DIR_M00 * size_Mat];
- D.f[DIR_0M0 ] = &distributions[DIR_0P0 * size_Mat];
- D.f[DIR_0P0 ] = &distributions[DIR_0M0 * size_Mat];
- D.f[DIR_00M ] = &distributions[DIR_00P * size_Mat];
- D.f[DIR_00P ] = &distributions[DIR_00M * size_Mat];
- D.f[DIR_MM0 ] = &distributions[DIR_PP0 * size_Mat];
- D.f[DIR_PP0 ] = &distributions[DIR_MM0 * size_Mat];
- D.f[DIR_MP0 ] = &distributions[DIR_PM0 * size_Mat];
- D.f[DIR_PM0 ] = &distributions[DIR_MP0 * size_Mat];
- D.f[DIR_M0M ] = &distributions[DIR_P0P * size_Mat];
- D.f[DIR_P0P ] = &distributions[DIR_M0M * size_Mat];
- D.f[DIR_M0P ] = &distributions[DIR_P0M * size_Mat];
- D.f[DIR_P0M ] = &distributions[DIR_M0P * size_Mat];
- D.f[DIR_0MM ] = &distributions[DIR_0PP * size_Mat];
- D.f[DIR_0PP ] = &distributions[DIR_0MM * size_Mat];
- D.f[DIR_0MP ] = &distributions[DIR_0PM * size_Mat];
- D.f[DIR_0PM ] = &distributions[DIR_0MP * size_Mat];
- D.f[DIR_000] = &distributions[DIR_000 * size_Mat];
- D.f[DIR_PPP ] = &distributions[DIR_MMM * size_Mat];
- D.f[DIR_MMP ] = &distributions[DIR_PPM * size_Mat];
- D.f[DIR_PMP ] = &distributions[DIR_MPM * size_Mat];
- D.f[DIR_MPP ] = &distributions[DIR_PMM * size_Mat];
- D.f[DIR_PPM ] = &distributions[DIR_MMP * size_Mat];
- D.f[DIR_MMM ] = &distributions[DIR_PPP * size_Mat];
- D.f[DIR_PMM ] = &distributions[DIR_MPP * size_Mat];
- D.f[DIR_MPM ] = &distributions[DIR_PMP * size_Mat];
+ D.f[DIR_M00] = &distributions[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &distributions[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &distributions[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &distributions[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &distributions[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &distributions[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &distributions[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &distributions[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &distributions[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &distributions[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &distributions[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &distributions[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &distributions[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &distributions[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &distributions[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &distributions[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &distributions[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &distributions[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &distributions[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &distributions[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MMP] = &distributions[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PMP] = &distributions[DIR_MPM * numberOfLBnodes];
+ D.f[DIR_MPP] = &distributions[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PPM] = &distributions[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MMM] = &distributions[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PMM] = &distributions[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MPM] = &distributions[DIR_PMP * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////
Distributions27 DAD;
if (isEvenTimestep)
{
- DAD.f[DIR_P00 ] = &distributionsAD[DIR_P00 * size_Mat];
- DAD.f[DIR_M00 ] = &distributionsAD[DIR_M00 * size_Mat];
- DAD.f[DIR_0P0 ] = &distributionsAD[DIR_0P0 * size_Mat];
- DAD.f[DIR_0M0 ] = &distributionsAD[DIR_0M0 * size_Mat];
- DAD.f[DIR_00P ] = &distributionsAD[DIR_00P * size_Mat];
- DAD.f[DIR_00M ] = &distributionsAD[DIR_00M * size_Mat];
- DAD.f[DIR_PP0 ] = &distributionsAD[DIR_PP0 * size_Mat];
- DAD.f[DIR_MM0 ] = &distributionsAD[DIR_MM0 * size_Mat];
- DAD.f[DIR_PM0 ] = &distributionsAD[DIR_PM0 * size_Mat];
- DAD.f[DIR_MP0 ] = &distributionsAD[DIR_MP0 * size_Mat];
- DAD.f[DIR_P0P ] = &distributionsAD[DIR_P0P * size_Mat];
- DAD.f[DIR_M0M ] = &distributionsAD[DIR_M0M * size_Mat];
- DAD.f[DIR_P0M ] = &distributionsAD[DIR_P0M * size_Mat];
- DAD.f[DIR_M0P ] = &distributionsAD[DIR_M0P * size_Mat];
- DAD.f[DIR_0PP ] = &distributionsAD[DIR_0PP * size_Mat];
- DAD.f[DIR_0MM ] = &distributionsAD[DIR_0MM * size_Mat];
- DAD.f[DIR_0PM ] = &distributionsAD[DIR_0PM * size_Mat];
- DAD.f[DIR_0MP ] = &distributionsAD[DIR_0MP * size_Mat];
- DAD.f[DIR_000] = &distributionsAD[DIR_000 * size_Mat];
- DAD.f[DIR_PPP ] = &distributionsAD[DIR_PPP * size_Mat];
- DAD.f[DIR_MMP ] = &distributionsAD[DIR_MMP * size_Mat];
- DAD.f[DIR_PMP ] = &distributionsAD[DIR_PMP * size_Mat];
- DAD.f[DIR_MPP ] = &distributionsAD[DIR_MPP * size_Mat];
- DAD.f[DIR_PPM ] = &distributionsAD[DIR_PPM * size_Mat];
- DAD.f[DIR_MMM ] = &distributionsAD[DIR_MMM * size_Mat];
- DAD.f[DIR_PMM ] = &distributionsAD[DIR_PMM * size_Mat];
- DAD.f[DIR_MPM ] = &distributionsAD[DIR_MPM * size_Mat];
+ DAD.f[DIR_P00] = &distributionsAD[DIR_P00 * numberOfLBnodes];
+ DAD.f[DIR_M00] = &distributionsAD[DIR_M00 * numberOfLBnodes];
+ DAD.f[DIR_0P0] = &distributionsAD[DIR_0P0 * numberOfLBnodes];
+ DAD.f[DIR_0M0] = &distributionsAD[DIR_0M0 * numberOfLBnodes];
+ DAD.f[DIR_00P] = &distributionsAD[DIR_00P * numberOfLBnodes];
+ DAD.f[DIR_00M] = &distributionsAD[DIR_00M * numberOfLBnodes];
+ DAD.f[DIR_PP0] = &distributionsAD[DIR_PP0 * numberOfLBnodes];
+ DAD.f[DIR_MM0] = &distributionsAD[DIR_MM0 * numberOfLBnodes];
+ DAD.f[DIR_PM0] = &distributionsAD[DIR_PM0 * numberOfLBnodes];
+ DAD.f[DIR_MP0] = &distributionsAD[DIR_MP0 * numberOfLBnodes];
+ DAD.f[DIR_P0P] = &distributionsAD[DIR_P0P * numberOfLBnodes];
+ DAD.f[DIR_M0M] = &distributionsAD[DIR_M0M * numberOfLBnodes];
+ DAD.f[DIR_P0M] = &distributionsAD[DIR_P0M * numberOfLBnodes];
+ DAD.f[DIR_M0P] = &distributionsAD[DIR_M0P * numberOfLBnodes];
+ DAD.f[DIR_0PP] = &distributionsAD[DIR_0PP * numberOfLBnodes];
+ DAD.f[DIR_0MM] = &distributionsAD[DIR_0MM * numberOfLBnodes];
+ DAD.f[DIR_0PM] = &distributionsAD[DIR_0PM * numberOfLBnodes];
+ DAD.f[DIR_0MP] = &distributionsAD[DIR_0MP * numberOfLBnodes];
+ DAD.f[DIR_000] = &distributionsAD[DIR_000 * numberOfLBnodes];
+ DAD.f[DIR_PPP] = &distributionsAD[DIR_PPP * numberOfLBnodes];
+ DAD.f[DIR_MMP] = &distributionsAD[DIR_MMP * numberOfLBnodes];
+ DAD.f[DIR_PMP] = &distributionsAD[DIR_PMP * numberOfLBnodes];
+ DAD.f[DIR_MPP] = &distributionsAD[DIR_MPP * numberOfLBnodes];
+ DAD.f[DIR_PPM] = &distributionsAD[DIR_PPM * numberOfLBnodes];
+ DAD.f[DIR_MMM] = &distributionsAD[DIR_MMM * numberOfLBnodes];
+ DAD.f[DIR_PMM] = &distributionsAD[DIR_PMM * numberOfLBnodes];
+ DAD.f[DIR_MPM] = &distributionsAD[DIR_MPM * numberOfLBnodes];
}
else
{
- DAD.f[DIR_M00 ] = &distributionsAD[DIR_P00 * size_Mat];
- DAD.f[DIR_P00 ] = &distributionsAD[DIR_M00 * size_Mat];
- DAD.f[DIR_0M0 ] = &distributionsAD[DIR_0P0 * size_Mat];
- DAD.f[DIR_0P0 ] = &distributionsAD[DIR_0M0 * size_Mat];
- DAD.f[DIR_00M ] = &distributionsAD[DIR_00P * size_Mat];
- DAD.f[DIR_00P ] = &distributionsAD[DIR_00M * size_Mat];
- DAD.f[DIR_MM0 ] = &distributionsAD[DIR_PP0 * size_Mat];
- DAD.f[DIR_PP0 ] = &distributionsAD[DIR_MM0 * size_Mat];
- DAD.f[DIR_MP0 ] = &distributionsAD[DIR_PM0 * size_Mat];
- DAD.f[DIR_PM0 ] = &distributionsAD[DIR_MP0 * size_Mat];
- DAD.f[DIR_M0M ] = &distributionsAD[DIR_P0P * size_Mat];
- DAD.f[DIR_P0P ] = &distributionsAD[DIR_M0M * size_Mat];
- DAD.f[DIR_M0P ] = &distributionsAD[DIR_P0M * size_Mat];
- DAD.f[DIR_P0M ] = &distributionsAD[DIR_M0P * size_Mat];
- DAD.f[DIR_0MM ] = &distributionsAD[DIR_0PP * size_Mat];
- DAD.f[DIR_0PP ] = &distributionsAD[DIR_0MM * size_Mat];
- DAD.f[DIR_0MP ] = &distributionsAD[DIR_0PM * size_Mat];
- DAD.f[DIR_0PM ] = &distributionsAD[DIR_0MP * size_Mat];
- DAD.f[DIR_000] = &distributionsAD[DIR_000 * size_Mat];
- DAD.f[DIR_PPP ] = &distributionsAD[DIR_MMM * size_Mat];
- DAD.f[DIR_MMP ] = &distributionsAD[DIR_PPM * size_Mat];
- DAD.f[DIR_PMP ] = &distributionsAD[DIR_MPM * size_Mat];
- DAD.f[DIR_MPP ] = &distributionsAD[DIR_PMM * size_Mat];
- DAD.f[DIR_PPM ] = &distributionsAD[DIR_MMP * size_Mat];
- DAD.f[DIR_MMM ] = &distributionsAD[DIR_PPP * size_Mat];
- DAD.f[DIR_PMM ] = &distributionsAD[DIR_MPP * size_Mat];
- DAD.f[DIR_MPM ] = &distributionsAD[DIR_PMP * size_Mat];
+ DAD.f[DIR_M00] = &distributionsAD[DIR_P00 * numberOfLBnodes];
+ DAD.f[DIR_P00] = &distributionsAD[DIR_M00 * numberOfLBnodes];
+ DAD.f[DIR_0M0] = &distributionsAD[DIR_0P0 * numberOfLBnodes];
+ DAD.f[DIR_0P0] = &distributionsAD[DIR_0M0 * numberOfLBnodes];
+ DAD.f[DIR_00M] = &distributionsAD[DIR_00P * numberOfLBnodes];
+ DAD.f[DIR_00P] = &distributionsAD[DIR_00M * numberOfLBnodes];
+ DAD.f[DIR_MM0] = &distributionsAD[DIR_PP0 * numberOfLBnodes];
+ DAD.f[DIR_PP0] = &distributionsAD[DIR_MM0 * numberOfLBnodes];
+ DAD.f[DIR_MP0] = &distributionsAD[DIR_PM0 * numberOfLBnodes];
+ DAD.f[DIR_PM0] = &distributionsAD[DIR_MP0 * numberOfLBnodes];
+ DAD.f[DIR_M0M] = &distributionsAD[DIR_P0P * numberOfLBnodes];
+ DAD.f[DIR_P0P] = &distributionsAD[DIR_M0M * numberOfLBnodes];
+ DAD.f[DIR_M0P] = &distributionsAD[DIR_P0M * numberOfLBnodes];
+ DAD.f[DIR_P0M] = &distributionsAD[DIR_M0P * numberOfLBnodes];
+ DAD.f[DIR_0MM] = &distributionsAD[DIR_0PP * numberOfLBnodes];
+ DAD.f[DIR_0PP] = &distributionsAD[DIR_0MM * numberOfLBnodes];
+ DAD.f[DIR_0MP] = &distributionsAD[DIR_0PM * numberOfLBnodes];
+ DAD.f[DIR_0PM] = &distributionsAD[DIR_0MP * numberOfLBnodes];
+ DAD.f[DIR_000] = &distributionsAD[DIR_000 * numberOfLBnodes];
+ DAD.f[DIR_PPP] = &distributionsAD[DIR_MMM * numberOfLBnodes];
+ DAD.f[DIR_MMP] = &distributionsAD[DIR_PPM * numberOfLBnodes];
+ DAD.f[DIR_PMP] = &distributionsAD[DIR_MPM * numberOfLBnodes];
+ DAD.f[DIR_MPP] = &distributionsAD[DIR_PMM * numberOfLBnodes];
+ DAD.f[DIR_PPM] = &distributionsAD[DIR_MMP * numberOfLBnodes];
+ DAD.f[DIR_MMM] = &distributionsAD[DIR_PPP * numberOfLBnodes];
+ DAD.f[DIR_PMM] = &distributionsAD[DIR_MPP * numberOfLBnodes];
+ DAD.f[DIR_MPM] = &distributionsAD[DIR_PMP * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////
const unsigned x = threadIdx.x; // Globaler x-Index
@@ -6888,24 +6888,24 @@ __global__ void AD_SlipVelDeviceComp(
* q_dirBE, * q_dirTW, * q_dirTN, * q_dirBS, * q_dirBN, * q_dirTS,
* q_dirTNE, * q_dirTSW, * q_dirTSE, * q_dirTNW, * q_dirBNE, * q_dirBSW,
* q_dirBSE, * q_dirBNW;
- q_dirE = &Qarrays[DIR_P00 * numberOfBCnodes];
- q_dirW = &Qarrays[DIR_M00 * numberOfBCnodes];
- q_dirN = &Qarrays[DIR_0P0 * numberOfBCnodes];
- q_dirS = &Qarrays[DIR_0M0 * numberOfBCnodes];
- q_dirT = &Qarrays[DIR_00P * numberOfBCnodes];
- q_dirB = &Qarrays[DIR_00M * numberOfBCnodes];
- q_dirNE = &Qarrays[DIR_PP0 * numberOfBCnodes];
- q_dirSW = &Qarrays[DIR_MM0 * numberOfBCnodes];
- q_dirSE = &Qarrays[DIR_PM0 * numberOfBCnodes];
- q_dirNW = &Qarrays[DIR_MP0 * numberOfBCnodes];
- q_dirTE = &Qarrays[DIR_P0P * numberOfBCnodes];
- q_dirBW = &Qarrays[DIR_M0M * numberOfBCnodes];
- q_dirBE = &Qarrays[DIR_P0M * numberOfBCnodes];
- q_dirTW = &Qarrays[DIR_M0P * numberOfBCnodes];
- q_dirTN = &Qarrays[DIR_0PP * numberOfBCnodes];
- q_dirBS = &Qarrays[DIR_0MM * numberOfBCnodes];
- q_dirBN = &Qarrays[DIR_0PM * numberOfBCnodes];
- q_dirTS = &Qarrays[DIR_0MP * numberOfBCnodes];
+ q_dirE = &Qarrays[DIR_P00 * numberOfBCnodes];
+ q_dirW = &Qarrays[DIR_M00 * numberOfBCnodes];
+ q_dirN = &Qarrays[DIR_0P0 * numberOfBCnodes];
+ q_dirS = &Qarrays[DIR_0M0 * numberOfBCnodes];
+ q_dirT = &Qarrays[DIR_00P * numberOfBCnodes];
+ q_dirB = &Qarrays[DIR_00M * numberOfBCnodes];
+ q_dirNE = &Qarrays[DIR_PP0 * numberOfBCnodes];
+ q_dirSW = &Qarrays[DIR_MM0 * numberOfBCnodes];
+ q_dirSE = &Qarrays[DIR_PM0 * numberOfBCnodes];
+ q_dirNW = &Qarrays[DIR_MP0 * numberOfBCnodes];
+ q_dirTE = &Qarrays[DIR_P0P * numberOfBCnodes];
+ q_dirBW = &Qarrays[DIR_M0M * numberOfBCnodes];
+ q_dirBE = &Qarrays[DIR_P0M * numberOfBCnodes];
+ q_dirTW = &Qarrays[DIR_M0P * numberOfBCnodes];
+ q_dirTN = &Qarrays[DIR_0PP * numberOfBCnodes];
+ q_dirBS = &Qarrays[DIR_0MM * numberOfBCnodes];
+ q_dirBN = &Qarrays[DIR_0PM * numberOfBCnodes];
+ q_dirTS = &Qarrays[DIR_0MP * numberOfBCnodes];
q_dirTNE = &Qarrays[DIR_PPP * numberOfBCnodes];
q_dirTSW = &Qarrays[DIR_MMP * numberOfBCnodes];
q_dirTSE = &Qarrays[DIR_PMP * numberOfBCnodes];
@@ -7025,63 +7025,63 @@ __global__ void AD_SlipVelDeviceComp(
//////////////////////////////////////////////////////////////////////////
if (!isEvenTimestep)
{
- DAD.f[DIR_P00 ] = &distributionsAD[DIR_P00 * size_Mat];
- DAD.f[DIR_M00 ] = &distributionsAD[DIR_M00 * size_Mat];
- DAD.f[DIR_0P0 ] = &distributionsAD[DIR_0P0 * size_Mat];
- DAD.f[DIR_0M0 ] = &distributionsAD[DIR_0M0 * size_Mat];
- DAD.f[DIR_00P ] = &distributionsAD[DIR_00P * size_Mat];
- DAD.f[DIR_00M ] = &distributionsAD[DIR_00M * size_Mat];
- DAD.f[DIR_PP0 ] = &distributionsAD[DIR_PP0 * size_Mat];
- DAD.f[DIR_MM0 ] = &distributionsAD[DIR_MM0 * size_Mat];
- DAD.f[DIR_PM0 ] = &distributionsAD[DIR_PM0 * size_Mat];
- DAD.f[DIR_MP0 ] = &distributionsAD[DIR_MP0 * size_Mat];
- DAD.f[DIR_P0P ] = &distributionsAD[DIR_P0P * size_Mat];
- DAD.f[DIR_M0M ] = &distributionsAD[DIR_M0M * size_Mat];
- DAD.f[DIR_P0M ] = &distributionsAD[DIR_P0M * size_Mat];
- DAD.f[DIR_M0P ] = &distributionsAD[DIR_M0P * size_Mat];
- DAD.f[DIR_0PP ] = &distributionsAD[DIR_0PP * size_Mat];
- DAD.f[DIR_0MM ] = &distributionsAD[DIR_0MM * size_Mat];
- DAD.f[DIR_0PM ] = &distributionsAD[DIR_0PM * size_Mat];
- DAD.f[DIR_0MP ] = &distributionsAD[DIR_0MP * size_Mat];
- DAD.f[DIR_000] = &distributionsAD[DIR_000 * size_Mat];
- DAD.f[DIR_PPP ] = &distributionsAD[DIR_PPP * size_Mat];
- DAD.f[DIR_MMP ] = &distributionsAD[DIR_MMP * size_Mat];
- DAD.f[DIR_PMP ] = &distributionsAD[DIR_PMP * size_Mat];
- DAD.f[DIR_MPP ] = &distributionsAD[DIR_MPP * size_Mat];
- DAD.f[DIR_PPM ] = &distributionsAD[DIR_PPM * size_Mat];
- DAD.f[DIR_MMM ] = &distributionsAD[DIR_MMM * size_Mat];
- DAD.f[DIR_PMM ] = &distributionsAD[DIR_PMM * size_Mat];
- DAD.f[DIR_MPM ] = &distributionsAD[DIR_MPM * size_Mat];
+ DAD.f[DIR_P00] = &distributionsAD[DIR_P00 * numberOfLBnodes];
+ DAD.f[DIR_M00] = &distributionsAD[DIR_M00 * numberOfLBnodes];
+ DAD.f[DIR_0P0] = &distributionsAD[DIR_0P0 * numberOfLBnodes];
+ DAD.f[DIR_0M0] = &distributionsAD[DIR_0M0 * numberOfLBnodes];
+ DAD.f[DIR_00P] = &distributionsAD[DIR_00P * numberOfLBnodes];
+ DAD.f[DIR_00M] = &distributionsAD[DIR_00M * numberOfLBnodes];
+ DAD.f[DIR_PP0] = &distributionsAD[DIR_PP0 * numberOfLBnodes];
+ DAD.f[DIR_MM0] = &distributionsAD[DIR_MM0 * numberOfLBnodes];
+ DAD.f[DIR_PM0] = &distributionsAD[DIR_PM0 * numberOfLBnodes];
+ DAD.f[DIR_MP0] = &distributionsAD[DIR_MP0 * numberOfLBnodes];
+ DAD.f[DIR_P0P] = &distributionsAD[DIR_P0P * numberOfLBnodes];
+ DAD.f[DIR_M0M] = &distributionsAD[DIR_M0M * numberOfLBnodes];
+ DAD.f[DIR_P0M] = &distributionsAD[DIR_P0M * numberOfLBnodes];
+ DAD.f[DIR_M0P] = &distributionsAD[DIR_M0P * numberOfLBnodes];
+ DAD.f[DIR_0PP] = &distributionsAD[DIR_0PP * numberOfLBnodes];
+ DAD.f[DIR_0MM] = &distributionsAD[DIR_0MM * numberOfLBnodes];
+ DAD.f[DIR_0PM] = &distributionsAD[DIR_0PM * numberOfLBnodes];
+ DAD.f[DIR_0MP] = &distributionsAD[DIR_0MP * numberOfLBnodes];
+ DAD.f[DIR_000] = &distributionsAD[DIR_000 * numberOfLBnodes];
+ DAD.f[DIR_PPP] = &distributionsAD[DIR_PPP * numberOfLBnodes];
+ DAD.f[DIR_MMP] = &distributionsAD[DIR_MMP * numberOfLBnodes];
+ DAD.f[DIR_PMP] = &distributionsAD[DIR_PMP * numberOfLBnodes];
+ DAD.f[DIR_MPP] = &distributionsAD[DIR_MPP * numberOfLBnodes];
+ DAD.f[DIR_PPM] = &distributionsAD[DIR_PPM * numberOfLBnodes];
+ DAD.f[DIR_MMM] = &distributionsAD[DIR_MMM * numberOfLBnodes];
+ DAD.f[DIR_PMM] = &distributionsAD[DIR_PMM * numberOfLBnodes];
+ DAD.f[DIR_MPM] = &distributionsAD[DIR_MPM * numberOfLBnodes];
}
else
{
- DAD.f[DIR_M00 ] = &distributionsAD[DIR_P00 * size_Mat];
- DAD.f[DIR_P00 ] = &distributionsAD[DIR_M00 * size_Mat];
- DAD.f[DIR_0M0 ] = &distributionsAD[DIR_0P0 * size_Mat];
- DAD.f[DIR_0P0 ] = &distributionsAD[DIR_0M0 * size_Mat];
- DAD.f[DIR_00M ] = &distributionsAD[DIR_00P * size_Mat];
- DAD.f[DIR_00P ] = &distributionsAD[DIR_00M * size_Mat];
- DAD.f[DIR_MM0 ] = &distributionsAD[DIR_PP0 * size_Mat];
- DAD.f[DIR_PP0 ] = &distributionsAD[DIR_MM0 * size_Mat];
- DAD.f[DIR_MP0 ] = &distributionsAD[DIR_PM0 * size_Mat];
- DAD.f[DIR_PM0 ] = &distributionsAD[DIR_MP0 * size_Mat];
- DAD.f[DIR_M0M ] = &distributionsAD[DIR_P0P * size_Mat];
- DAD.f[DIR_P0P ] = &distributionsAD[DIR_M0M * size_Mat];
- DAD.f[DIR_M0P ] = &distributionsAD[DIR_P0M * size_Mat];
- DAD.f[DIR_P0M ] = &distributionsAD[DIR_M0P * size_Mat];
- DAD.f[DIR_0MM ] = &distributionsAD[DIR_0PP * size_Mat];
- DAD.f[DIR_0PP ] = &distributionsAD[DIR_0MM * size_Mat];
- DAD.f[DIR_0MP ] = &distributionsAD[DIR_0PM * size_Mat];
- DAD.f[DIR_0PM ] = &distributionsAD[DIR_0MP * size_Mat];
- DAD.f[DIR_000] = &distributionsAD[DIR_000 * size_Mat];
- DAD.f[DIR_PPP ] = &distributionsAD[DIR_MMM * size_Mat];
- DAD.f[DIR_MMP ] = &distributionsAD[DIR_PPM * size_Mat];
- DAD.f[DIR_PMP ] = &distributionsAD[DIR_MPM * size_Mat];
- DAD.f[DIR_MPP ] = &distributionsAD[DIR_PMM * size_Mat];
- DAD.f[DIR_PPM ] = &distributionsAD[DIR_MMP * size_Mat];
- DAD.f[DIR_MMM ] = &distributionsAD[DIR_PPP * size_Mat];
- DAD.f[DIR_PMM ] = &distributionsAD[DIR_MPP * size_Mat];
- DAD.f[DIR_MPM ] = &distributionsAD[DIR_PMP * size_Mat];
+ DAD.f[DIR_M00] = &distributionsAD[DIR_P00 * numberOfLBnodes];
+ DAD.f[DIR_P00] = &distributionsAD[DIR_M00 * numberOfLBnodes];
+ DAD.f[DIR_0M0] = &distributionsAD[DIR_0P0 * numberOfLBnodes];
+ DAD.f[DIR_0P0] = &distributionsAD[DIR_0M0 * numberOfLBnodes];
+ DAD.f[DIR_00M] = &distributionsAD[DIR_00P * numberOfLBnodes];
+ DAD.f[DIR_00P] = &distributionsAD[DIR_00M * numberOfLBnodes];
+ DAD.f[DIR_MM0] = &distributionsAD[DIR_PP0 * numberOfLBnodes];
+ DAD.f[DIR_PP0] = &distributionsAD[DIR_MM0 * numberOfLBnodes];
+ DAD.f[DIR_MP0] = &distributionsAD[DIR_PM0 * numberOfLBnodes];
+ DAD.f[DIR_PM0] = &distributionsAD[DIR_MP0 * numberOfLBnodes];
+ DAD.f[DIR_M0M] = &distributionsAD[DIR_P0P * numberOfLBnodes];
+ DAD.f[DIR_P0P] = &distributionsAD[DIR_M0M * numberOfLBnodes];
+ DAD.f[DIR_M0P] = &distributionsAD[DIR_P0M * numberOfLBnodes];
+ DAD.f[DIR_P0M] = &distributionsAD[DIR_M0P * numberOfLBnodes];
+ DAD.f[DIR_0MM] = &distributionsAD[DIR_0PP * numberOfLBnodes];
+ DAD.f[DIR_0PP] = &distributionsAD[DIR_0MM * numberOfLBnodes];
+ DAD.f[DIR_0MP] = &distributionsAD[DIR_0PM * numberOfLBnodes];
+ DAD.f[DIR_0PM] = &distributionsAD[DIR_0MP * numberOfLBnodes];
+ DAD.f[DIR_000] = &distributionsAD[DIR_000 * numberOfLBnodes];
+ DAD.f[DIR_PPP] = &distributionsAD[DIR_MMM * numberOfLBnodes];
+ DAD.f[DIR_MMP] = &distributionsAD[DIR_PPM * numberOfLBnodes];
+ DAD.f[DIR_PMP] = &distributionsAD[DIR_MPM * numberOfLBnodes];
+ DAD.f[DIR_MPP] = &distributionsAD[DIR_PMM * numberOfLBnodes];
+ DAD.f[DIR_PPM] = &distributionsAD[DIR_MMP * numberOfLBnodes];
+ DAD.f[DIR_MMM] = &distributionsAD[DIR_PPP * numberOfLBnodes];
+ DAD.f[DIR_PMM] = &distributionsAD[DIR_MPP * numberOfLBnodes];
+ DAD.f[DIR_MPM] = &distributionsAD[DIR_PMP * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
real concentration =
@@ -7115,24 +7115,24 @@ __global__ void AD_SlipVelDeviceComp(
real jTan3 = jx3 - NormJ * NormZ;
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
- q = q_dirE[k]; if (q >= c0o1 && q <= c1o1) { (DAD.f[DIR_M00 ])[kw ] = calcDistributionBC_AD(q, c2o27, vx1, cu_sq, f_E, f_W, omegaDiffusivity, jTan1, concentration); }
- q = q_dirW[k]; if (q >= c0o1 && q <= c1o1) { (DAD.f[DIR_P00 ])[ke ] = calcDistributionBC_AD(q, c2o27, -vx1, cu_sq, f_W, f_E, omegaDiffusivity, -jTan1, concentration); }
- q = q_dirN[k]; if (q >= c0o1 && q <= c1o1) { (DAD.f[DIR_0M0 ])[ks ] = calcDistributionBC_AD(q, c2o27, vx2, cu_sq, f_N, f_S, omegaDiffusivity, jTan2, concentration); }
- q = q_dirS[k]; if (q >= c0o1 && q <= c1o1) { (DAD.f[DIR_0P0 ])[kn ] = calcDistributionBC_AD(q, c2o27, -vx2, cu_sq, f_S, f_N, omegaDiffusivity, -jTan2, concentration); }
- q = q_dirT[k]; if (q >= c0o1 && q <= c1o1) { (DAD.f[DIR_00M ])[kb ] = calcDistributionBC_AD(q, c2o27, vx3, cu_sq, f_T, f_B, omegaDiffusivity, jTan3, concentration); }
- q = q_dirB[k]; if (q >= c0o1 && q <= c1o1) { (DAD.f[DIR_00P ])[kt ] = calcDistributionBC_AD(q, c2o27, -vx3, cu_sq, f_B, f_T, omegaDiffusivity, -jTan3, concentration); }
- q = q_dirNE[k]; if (q >= c0o1 && q <= c1o1) { (DAD.f[DIR_MM0 ])[ksw ] = calcDistributionBC_AD(q, c1o54, vx1+vx2, cu_sq, f_NE, f_SW, omegaDiffusivity, jTan1+jTan2, concentration); }
- q = q_dirSW[k]; if (q >= c0o1 && q <= c1o1) { (DAD.f[DIR_PP0 ])[kne ] = calcDistributionBC_AD(q, c1o54, -vx1-vx2, cu_sq, f_SW, f_NE, omegaDiffusivity, -jTan1-jTan2, concentration); }
- q = q_dirSE[k]; if (q >= c0o1 && q <= c1o1) { (DAD.f[DIR_MP0 ])[knw ] = calcDistributionBC_AD(q, c1o54, vx1-vx2, cu_sq, f_SE, f_NW, omegaDiffusivity, jTan1-jTan2, concentration); }
- q = q_dirNW[k]; if (q >= c0o1 && q <= c1o1) { (DAD.f[DIR_PM0 ])[kse ] = calcDistributionBC_AD(q, c1o54, -vx1+vx2, cu_sq, f_NW, f_SE, omegaDiffusivity, -jTan1+jTan2, concentration); }
- q = q_dirTE[k]; if (q >= c0o1 && q <= c1o1) { (DAD.f[DIR_M0M ])[kbw ] = calcDistributionBC_AD(q, c1o54, vx1 +vx3, cu_sq, f_TE, f_BW, omegaDiffusivity, jTan1 +jTan3, concentration); }
- q = q_dirBW[k]; if (q >= c0o1 && q <= c1o1) { (DAD.f[DIR_P0P ])[kte ] = calcDistributionBC_AD(q, c1o54, -vx1 -vx3, cu_sq, f_BW, f_TE, omegaDiffusivity, -jTan1 -jTan3, concentration); }
- q = q_dirBE[k]; if (q >= c0o1 && q <= c1o1) { (DAD.f[DIR_M0P ])[ktw ] = calcDistributionBC_AD(q, c1o54, vx1 -vx3, cu_sq, f_BE, f_TW, omegaDiffusivity, jTan1 -jTan3, concentration); }
- q = q_dirTW[k]; if (q >= c0o1 && q <= c1o1) { (DAD.f[DIR_P0M ])[kbe ] = calcDistributionBC_AD(q, c1o54, -vx1 +vx3, cu_sq, f_TW, f_BE, omegaDiffusivity, -jTan1 +jTan3, concentration); }
- q = q_dirTN[k]; if (q >= c0o1 && q <= c1o1) { (DAD.f[DIR_0MM ])[kbs ] = calcDistributionBC_AD(q, c1o54, vx2+vx3, cu_sq, f_TN, f_BS, omegaDiffusivity, jTan2+jTan3, concentration); }
- q = q_dirBS[k]; if (q >= c0o1 && q <= c1o1) { (DAD.f[DIR_0PP ])[ktn ] = calcDistributionBC_AD(q, c1o54, -vx2-vx3, cu_sq, f_BS, f_TN, omegaDiffusivity, -jTan2-jTan3, concentration); }
- q = q_dirBN[k]; if (q >= c0o1 && q <= c1o1) { (DAD.f[DIR_0MP ])[kts ] = calcDistributionBC_AD(q, c1o54, vx2-vx3, cu_sq, f_BN, f_TS, omegaDiffusivity, jTan2-jTan3, concentration); }
- q = q_dirTS[k]; if (q >= c0o1 && q <= c1o1) { (DAD.f[DIR_0PM ])[kbn ] = calcDistributionBC_AD(q, c1o54, -vx2+vx3, cu_sq, f_TS, f_BN, omegaDiffusivity, -jTan2+jTan3, concentration); }
+ q = q_dirE[k]; if (q >= c0o1 && q <= c1o1) { (DAD.f[DIR_M00])[kw ] = calcDistributionBC_AD(q, c2o27, vx1, cu_sq, f_E, f_W, omegaDiffusivity, jTan1, concentration); }
+ q = q_dirW[k]; if (q >= c0o1 && q <= c1o1) { (DAD.f[DIR_P00])[ke ] = calcDistributionBC_AD(q, c2o27, -vx1, cu_sq, f_W, f_E, omegaDiffusivity, -jTan1, concentration); }
+ q = q_dirN[k]; if (q >= c0o1 && q <= c1o1) { (DAD.f[DIR_0M0])[ks ] = calcDistributionBC_AD(q, c2o27, vx2, cu_sq, f_N, f_S, omegaDiffusivity, jTan2, concentration); }
+ q = q_dirS[k]; if (q >= c0o1 && q <= c1o1) { (DAD.f[DIR_0P0])[kn ] = calcDistributionBC_AD(q, c2o27, -vx2, cu_sq, f_S, f_N, omegaDiffusivity, -jTan2, concentration); }
+ q = q_dirT[k]; if (q >= c0o1 && q <= c1o1) { (DAD.f[DIR_00M])[kb ] = calcDistributionBC_AD(q, c2o27, vx3, cu_sq, f_T, f_B, omegaDiffusivity, jTan3, concentration); }
+ q = q_dirB[k]; if (q >= c0o1 && q <= c1o1) { (DAD.f[DIR_00P])[kt ] = calcDistributionBC_AD(q, c2o27, -vx3, cu_sq, f_B, f_T, omegaDiffusivity, -jTan3, concentration); }
+ q = q_dirNE[k]; if (q >= c0o1 && q <= c1o1) { (DAD.f[DIR_MM0])[ksw ] = calcDistributionBC_AD(q, c1o54, vx1+vx2, cu_sq, f_NE, f_SW, omegaDiffusivity, jTan1+jTan2, concentration); }
+ q = q_dirSW[k]; if (q >= c0o1 && q <= c1o1) { (DAD.f[DIR_PP0])[kne ] = calcDistributionBC_AD(q, c1o54, -vx1-vx2, cu_sq, f_SW, f_NE, omegaDiffusivity, -jTan1-jTan2, concentration); }
+ q = q_dirSE[k]; if (q >= c0o1 && q <= c1o1) { (DAD.f[DIR_MP0])[knw ] = calcDistributionBC_AD(q, c1o54, vx1-vx2, cu_sq, f_SE, f_NW, omegaDiffusivity, jTan1-jTan2, concentration); }
+ q = q_dirNW[k]; if (q >= c0o1 && q <= c1o1) { (DAD.f[DIR_PM0])[kse ] = calcDistributionBC_AD(q, c1o54, -vx1+vx2, cu_sq, f_NW, f_SE, omegaDiffusivity, -jTan1+jTan2, concentration); }
+ q = q_dirTE[k]; if (q >= c0o1 && q <= c1o1) { (DAD.f[DIR_M0M])[kbw ] = calcDistributionBC_AD(q, c1o54, vx1 +vx3, cu_sq, f_TE, f_BW, omegaDiffusivity, jTan1 +jTan3, concentration); }
+ q = q_dirBW[k]; if (q >= c0o1 && q <= c1o1) { (DAD.f[DIR_P0P])[kte ] = calcDistributionBC_AD(q, c1o54, -vx1 -vx3, cu_sq, f_BW, f_TE, omegaDiffusivity, -jTan1 -jTan3, concentration); }
+ q = q_dirBE[k]; if (q >= c0o1 && q <= c1o1) { (DAD.f[DIR_M0P])[ktw ] = calcDistributionBC_AD(q, c1o54, vx1 -vx3, cu_sq, f_BE, f_TW, omegaDiffusivity, jTan1 -jTan3, concentration); }
+ q = q_dirTW[k]; if (q >= c0o1 && q <= c1o1) { (DAD.f[DIR_P0M])[kbe ] = calcDistributionBC_AD(q, c1o54, -vx1 +vx3, cu_sq, f_TW, f_BE, omegaDiffusivity, -jTan1 +jTan3, concentration); }
+ q = q_dirTN[k]; if (q >= c0o1 && q <= c1o1) { (DAD.f[DIR_0MM])[kbs ] = calcDistributionBC_AD(q, c1o54, vx2+vx3, cu_sq, f_TN, f_BS, omegaDiffusivity, jTan2+jTan3, concentration); }
+ q = q_dirBS[k]; if (q >= c0o1 && q <= c1o1) { (DAD.f[DIR_0PP])[ktn ] = calcDistributionBC_AD(q, c1o54, -vx2-vx3, cu_sq, f_BS, f_TN, omegaDiffusivity, -jTan2-jTan3, concentration); }
+ q = q_dirBN[k]; if (q >= c0o1 && q <= c1o1) { (DAD.f[DIR_0MP])[kts ] = calcDistributionBC_AD(q, c1o54, vx2-vx3, cu_sq, f_BN, f_TS, omegaDiffusivity, jTan2-jTan3, concentration); }
+ q = q_dirTS[k]; if (q >= c0o1 && q <= c1o1) { (DAD.f[DIR_0PM])[kbn ] = calcDistributionBC_AD(q, c1o54, -vx2+vx3, cu_sq, f_TS, f_BN, omegaDiffusivity, -jTan2+jTan3, concentration); }
q = q_dirTNE[k]; if (q >= c0o1 && q <= c1o1) { (DAD.f[DIR_MMM])[kbsw] = calcDistributionBC_AD(q, c1o216, vx1+vx2+vx3, cu_sq, f_TNE, f_BSW, omegaDiffusivity, jTan1+jTan2+jTan3, concentration); }
q = q_dirBSW[k]; if (q >= c0o1 && q <= c1o1) { (DAD.f[DIR_PPP])[ktne] = calcDistributionBC_AD(q, c1o216, -vx1-vx2-vx3, cu_sq, f_BSW, f_TNE, omegaDiffusivity, -jTan1-jTan2-jTan3, concentration); }
q = q_dirBNE[k]; if (q >= c0o1 && q <= c1o1) { (DAD.f[DIR_MMP])[ktsw] = calcDistributionBC_AD(q, c1o216, vx1+vx2-vx3, cu_sq, f_BNE, f_TSW, omegaDiffusivity, jTan1+jTan2-jTan3, concentration); }
diff --git a/src/gpu/VirtualFluids_GPU/GPU/CP27.cu b/src/gpu/VirtualFluids_GPU/GPU/CP27.cu
index 1ef111330c0d4293c14d66893847689ad8fac77f..8d02f4e1c110fc82b65adda4db67976f29796d07 100644
--- a/src/gpu/VirtualFluids_GPU/GPU/CP27.cu
+++ b/src/gpu/VirtualFluids_GPU/GPU/CP27.cu
@@ -14,69 +14,69 @@ __global__ void CalcCP27(real* DD,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep)
{
Distributions27 D;
if (isEvenTimestep==true)
{
- D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
+ D.f[DIR_M00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_MPM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_PMP * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////
const unsigned x = threadIdx.x; // Globaler x-Index
@@ -124,20 +124,20 @@ __global__ void CalcCP27(real* DD,
////////////////////////////////////////////////////////////////////////////////
double PressCP;
- PressCP = (D.f[DIR_P00 ])[ke ]+ (D.f[DIR_M00 ])[kw ]+
- (D.f[DIR_0P0 ])[kn ]+ (D.f[DIR_0M0 ])[ks ]+
- (D.f[DIR_00P ])[kt ]+ (D.f[DIR_00M ])[kb ]+
- (D.f[DIR_PP0 ])[kne ]+ (D.f[DIR_MM0 ])[ksw ]+
- (D.f[DIR_PM0 ])[kse ]+ (D.f[DIR_MP0 ])[knw ]+
- (D.f[DIR_P0P ])[kte ]+ (D.f[DIR_M0M ])[kbw ]+
- (D.f[DIR_P0M ])[kbe ]+ (D.f[DIR_M0P ])[ktw ]+
- (D.f[DIR_0PP ])[ktn ]+ (D.f[DIR_0MM ])[kbs ]+
- (D.f[DIR_0PM ])[kbn ]+ (D.f[DIR_0MP ])[kts ]+
+ PressCP = (D.f[DIR_P00])[ke ]+ (D.f[DIR_M00])[kw ]+
+ (D.f[DIR_0P0])[kn ]+ (D.f[DIR_0M0])[ks ]+
+ (D.f[DIR_00P])[kt ]+ (D.f[DIR_00M])[kb ]+
+ (D.f[DIR_PP0])[kne ]+ (D.f[DIR_MM0])[ksw ]+
+ (D.f[DIR_PM0])[kse ]+ (D.f[DIR_MP0])[knw ]+
+ (D.f[DIR_P0P])[kte ]+ (D.f[DIR_M0M])[kbw ]+
+ (D.f[DIR_P0M])[kbe ]+ (D.f[DIR_M0P])[ktw ]+
+ (D.f[DIR_0PP])[ktn ]+ (D.f[DIR_0MM])[kbs ]+
+ (D.f[DIR_0PM])[kbn ]+ (D.f[DIR_0MP])[kts ]+
(D.f[DIR_000])[kzero]+
- (D.f[DIR_PPP ])[ktne]+ (D.f[DIR_MMP ])[ktsw]+
- (D.f[DIR_PMP ])[ktse]+ (D.f[DIR_MPP ])[ktnw]+
- (D.f[DIR_PPM ])[kbne]+ (D.f[DIR_MMM ])[kbsw]+
- (D.f[DIR_PMM ])[kbse]+ (D.f[DIR_MPM ])[kbnw];
+ (D.f[DIR_PPP])[ktne]+ (D.f[DIR_MMP])[ktsw]+
+ (D.f[DIR_PMP])[ktse]+ (D.f[DIR_MPP])[ktnw]+
+ (D.f[DIR_PPM])[kbne]+ (D.f[DIR_MMM])[kbsw]+
+ (D.f[DIR_PMM])[kbse]+ (D.f[DIR_MPM])[kbnw];
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
cpPress[k] = PressCP;
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
diff --git a/src/gpu/VirtualFluids_GPU/GPU/Calc2ndMoments27.cu b/src/gpu/VirtualFluids_GPU/GPU/Calc2ndMoments27.cu
index ce8fe68cd6a2e8f09f150cb0ccdec502a6278b50..c41751dc1b5cea53983d94d9cc7c3c75c8a84101 100644
--- a/src/gpu/VirtualFluids_GPU/GPU/Calc2ndMoments27.cu
+++ b/src/gpu/VirtualFluids_GPU/GPU/Calc2ndMoments27.cu
@@ -16,70 +16,70 @@ __global__ void LBCalc2ndMomentsIncompSP27( real* kxyFromfcNEQ,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
real* DD,
bool isEvenTimestep)
{
Distributions27 D;
if (isEvenTimestep==true)
{
- D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
+ D.f[DIR_M00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_MPM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_PMP * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////
const unsigned x = threadIdx.x; // Globaler x-Index
@@ -92,7 +92,7 @@ __global__ void LBCalc2ndMomentsIncompSP27( real* kxyFromfcNEQ,
const unsigned k = nx*(ny*z + y) + x;
//////////////////////////////////////////////////////////////////////////
- if(k < size_Mat)
+ if(k < numberOfLBnodes)
{
//////////////////////////////////////////////////////////////////////////
//index
@@ -125,33 +125,33 @@ __global__ void LBCalc2ndMomentsIncompSP27( real* kxyFromfcNEQ,
unsigned int kbsw = neighborZ[ksw];
//////////////////////////////////////////////////////////////////////////
real f_E,f_W,f_N,f_S,f_T,f_B,f_NE,f_SW,f_SE,f_NW,f_TE,f_BW,f_BE,f_TW,f_TN,f_BS,f_BN,f_TS,/*f_ZERO,*/f_TNE, f_TSW, f_TSE, f_TNW, f_BNE, f_BSW, f_BSE, f_BNW;
- f_E = (D.f[DIR_P00 ])[ke ];
- f_W = (D.f[DIR_M00 ])[kw ];
- f_N = (D.f[DIR_0P0 ])[kn ];
- f_S = (D.f[DIR_0M0 ])[ks ];
- f_T = (D.f[DIR_00P ])[kt ];
- f_B = (D.f[DIR_00M ])[kb ];
- f_NE = (D.f[DIR_PP0 ])[kne ];
- f_SW = (D.f[DIR_MM0 ])[ksw ];
- f_SE = (D.f[DIR_PM0 ])[kse ];
- f_NW = (D.f[DIR_MP0 ])[knw ];
- f_TE = (D.f[DIR_P0P ])[kte ];
- f_BW = (D.f[DIR_M0M ])[kbw ];
- f_BE = (D.f[DIR_P0M ])[kbe ];
- f_TW = (D.f[DIR_M0P ])[ktw ];
- f_TN = (D.f[DIR_0PP ])[ktn ];
- f_BS = (D.f[DIR_0MM ])[kbs ];
- f_BN = (D.f[DIR_0PM ])[kbn ];
- f_TS = (D.f[DIR_0MP ])[kts ];
+ f_E = (D.f[DIR_P00])[ke ];
+ f_W = (D.f[DIR_M00])[kw ];
+ f_N = (D.f[DIR_0P0])[kn ];
+ f_S = (D.f[DIR_0M0])[ks ];
+ f_T = (D.f[DIR_00P])[kt ];
+ f_B = (D.f[DIR_00M])[kb ];
+ f_NE = (D.f[DIR_PP0])[kne ];
+ f_SW = (D.f[DIR_MM0])[ksw ];
+ f_SE = (D.f[DIR_PM0])[kse ];
+ f_NW = (D.f[DIR_MP0])[knw ];
+ f_TE = (D.f[DIR_P0P])[kte ];
+ f_BW = (D.f[DIR_M0M])[kbw ];
+ f_BE = (D.f[DIR_P0M])[kbe ];
+ f_TW = (D.f[DIR_M0P])[ktw ];
+ f_TN = (D.f[DIR_0PP])[ktn ];
+ f_BS = (D.f[DIR_0MM])[kbs ];
+ f_BN = (D.f[DIR_0PM])[kbn ];
+ f_TS = (D.f[DIR_0MP])[kts ];
//f_ZERO = (D.f[DIR_000])[kzero];
- f_TNE = (D.f[DIR_PPP ])[ktne ];
- f_TSW = (D.f[DIR_MMP ])[ktsw ];
- f_TSE = (D.f[DIR_PMP ])[ktse ];
- f_TNW = (D.f[DIR_MPP ])[ktnw ];
- f_BNE = (D.f[DIR_PPM ])[kbne ];
- f_BSW = (D.f[DIR_MMM ])[kbsw ];
- f_BSE = (D.f[DIR_PMM ])[kbse ];
- f_BNW = (D.f[DIR_MPM ])[kbnw ];
+ f_TNE = (D.f[DIR_PPP])[ktne ];
+ f_TSW = (D.f[DIR_MMP])[ktsw ];
+ f_TSE = (D.f[DIR_PMP])[ktse ];
+ f_TNW = (D.f[DIR_MPP])[ktnw ];
+ f_BNE = (D.f[DIR_PPM])[kbne ];
+ f_BSW = (D.f[DIR_MMM])[kbsw ];
+ f_BSE = (D.f[DIR_PMM])[kbse ];
+ f_BNW = (D.f[DIR_MPM])[kbnw ];
//////////////////////////////////////////////////////////////////////////
real vx1, vx2, vx3;
kxyFromfcNEQ[k] = c0o1;
@@ -215,70 +215,70 @@ __global__ void LBCalc2ndMomentsCompSP27(real* kxyFromfcNEQ,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
real* DD,
bool isEvenTimestep)
{
Distributions27 D;
if (isEvenTimestep==true)
{
- D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
+ D.f[DIR_M00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_MPM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_PMP * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////
const unsigned x = threadIdx.x; // Globaler x-Index
@@ -291,7 +291,7 @@ __global__ void LBCalc2ndMomentsCompSP27(real* kxyFromfcNEQ,
const unsigned k = nx*(ny*z + y) + x;
//////////////////////////////////////////////////////////////////////////
- if(k < size_Mat)
+ if(k < numberOfLBnodes)
{
//////////////////////////////////////////////////////////////////////////
//index
@@ -325,33 +325,33 @@ __global__ void LBCalc2ndMomentsCompSP27(real* kxyFromfcNEQ,
//////////////////////////////////////////////////////////////////////////
real f_ZERO;
real f_E,f_W,f_N,f_S,f_T,f_B,f_NE,f_SW,f_SE,f_NW,f_TE,f_BW,f_BE,f_TW,f_TN,f_BS,f_BN,f_TS,f_TNE, f_TSW, f_TSE, f_TNW, f_BNE, f_BSW, f_BSE, f_BNW;
- f_E = (D.f[DIR_P00 ])[ke ];
- f_W = (D.f[DIR_M00 ])[kw ];
- f_N = (D.f[DIR_0P0 ])[kn ];
- f_S = (D.f[DIR_0M0 ])[ks ];
- f_T = (D.f[DIR_00P ])[kt ];
- f_B = (D.f[DIR_00M ])[kb ];
- f_NE = (D.f[DIR_PP0 ])[kne ];
- f_SW = (D.f[DIR_MM0 ])[ksw ];
- f_SE = (D.f[DIR_PM0 ])[kse ];
- f_NW = (D.f[DIR_MP0 ])[knw ];
- f_TE = (D.f[DIR_P0P ])[kte ];
- f_BW = (D.f[DIR_M0M ])[kbw ];
- f_BE = (D.f[DIR_P0M ])[kbe ];
- f_TW = (D.f[DIR_M0P ])[ktw ];
- f_TN = (D.f[DIR_0PP ])[ktn ];
- f_BS = (D.f[DIR_0MM ])[kbs ];
- f_BN = (D.f[DIR_0PM ])[kbn ];
- f_TS = (D.f[DIR_0MP ])[kts ];
+ f_E = (D.f[DIR_P00])[ke ];
+ f_W = (D.f[DIR_M00])[kw ];
+ f_N = (D.f[DIR_0P0])[kn ];
+ f_S = (D.f[DIR_0M0])[ks ];
+ f_T = (D.f[DIR_00P])[kt ];
+ f_B = (D.f[DIR_00M])[kb ];
+ f_NE = (D.f[DIR_PP0])[kne ];
+ f_SW = (D.f[DIR_MM0])[ksw ];
+ f_SE = (D.f[DIR_PM0])[kse ];
+ f_NW = (D.f[DIR_MP0])[knw ];
+ f_TE = (D.f[DIR_P0P])[kte ];
+ f_BW = (D.f[DIR_M0M])[kbw ];
+ f_BE = (D.f[DIR_P0M])[kbe ];
+ f_TW = (D.f[DIR_M0P])[ktw ];
+ f_TN = (D.f[DIR_0PP])[ktn ];
+ f_BS = (D.f[DIR_0MM])[kbs ];
+ f_BN = (D.f[DIR_0PM])[kbn ];
+ f_TS = (D.f[DIR_0MP])[kts ];
f_ZERO = (D.f[DIR_000])[kzero];
- f_TNE = (D.f[DIR_PPP ])[ktne ];
- f_TSW = (D.f[DIR_MMP ])[ktsw ];
- f_TSE = (D.f[DIR_PMP ])[ktse ];
- f_TNW = (D.f[DIR_MPP ])[ktnw ];
- f_BNE = (D.f[DIR_PPM ])[kbne ];
- f_BSW = (D.f[DIR_MMM ])[kbsw ];
- f_BSE = (D.f[DIR_PMM ])[kbse ];
- f_BNW = (D.f[DIR_MPM ])[kbnw ];
+ f_TNE = (D.f[DIR_PPP])[ktne ];
+ f_TSW = (D.f[DIR_MMP])[ktsw ];
+ f_TSE = (D.f[DIR_PMP])[ktse ];
+ f_TNW = (D.f[DIR_MPP])[ktnw ];
+ f_BNE = (D.f[DIR_PPM])[kbne ];
+ f_BSW = (D.f[DIR_MMM])[kbsw ];
+ f_BSE = (D.f[DIR_PMM])[kbse ];
+ f_BNW = (D.f[DIR_MPM])[kbnw ];
//////////////////////////////////////////////////////////////////////////
real drho;
real vx1, vx2, vx3, rho;
@@ -423,7 +423,7 @@ __global__ void LBCalc3rdMomentsIncompSP27( real* CUMbbb,
unsigned int* neighborY,
unsigned int* neighborZ,
real* DDStart,
- int size_Mat,
+ unsigned long long numberOfLBnodes,
bool EvenOrOdd)
{
////////////////////////////////////////////////////////////////////////////////
@@ -437,7 +437,7 @@ __global__ void LBCalc3rdMomentsIncompSP27( real* CUMbbb,
const unsigned k = nx*(ny*z + y) + x;
//////////////////////////////////////////////////////////////////////////
- if(k<size_Mat)
+ if(k<numberOfLBnodes)
{
////////////////////////////////////////////////////////////////////////////////
unsigned int BC;
@@ -448,63 +448,63 @@ __global__ void LBCalc3rdMomentsIncompSP27( real* CUMbbb,
Distributions27 D;
if (EvenOrOdd==true)
{
- D.f[DIR_P00 ] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DDStart[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DDStart[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DDStart[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DDStart[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DDStart[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DDStart[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DDStart[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DDStart[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DDStart[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DDStart[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DDStart[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DDStart[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DDStart[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DDStart[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DDStart[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DDStart[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DDStart[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DDStart[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DDStart[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DDStart[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DDStart[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DDStart[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DDStart[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DDStart[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DDStart[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DDStart[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DDStart[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00 ] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_MMM ] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_PPM ] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_MPM ] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_PMM ] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_MMP ] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_PPP ] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_MPP ] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_PMP ] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_M00] = &DDStart[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DDStart[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DDStart[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DDStart[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DDStart[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DDStart[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DDStart[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DDStart[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DDStart[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DDStart[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DDStart[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DDStart[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DDStart[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DDStart[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DDStart[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DDStart[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DDStart[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DDStart[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DDStart[DIR_000 * numberOfLBnodes];
+ D.f[DIR_MMM] = &DDStart[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DDStart[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DDStart[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DDStart[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DDStart[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PPP] = &DDStart[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DDStart[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DDStart[DIR_MPM * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////
@@ -517,33 +517,33 @@ __global__ void LBCalc3rdMomentsIncompSP27( real* CUMbbb,
unsigned int kbs = neighborZ[ks];
unsigned int kbsw = neighborZ[ksw];
//////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
- real mfcbb = (D.f[DIR_P00 ])[k ];
- real mfabb = (D.f[DIR_M00 ])[kw ];
- real mfbcb = (D.f[DIR_0P0 ])[k ];
- real mfbab = (D.f[DIR_0M0 ])[ks ];
- real mfbbc = (D.f[DIR_00P ])[k ];
- real mfbba = (D.f[DIR_00M ])[kb ];
- real mfccb = (D.f[DIR_PP0 ])[k ];
- real mfaab = (D.f[DIR_MM0 ])[ksw];
- real mfcab = (D.f[DIR_PM0 ])[ks ];
- real mfacb = (D.f[DIR_MP0 ])[kw ];
- real mfcbc = (D.f[DIR_P0P ])[k ];
- real mfaba = (D.f[DIR_M0M ])[kbw];
- real mfcba = (D.f[DIR_P0M ])[kb ];
- real mfabc = (D.f[DIR_M0P ])[kw ];
- real mfbcc = (D.f[DIR_0PP ])[k ];
- real mfbaa = (D.f[DIR_0MM ])[kbs];
- real mfbca = (D.f[DIR_0PM ])[kb ];
- real mfbac = (D.f[DIR_0MP ])[ks ];
+ real mfcbb = (D.f[DIR_P00])[k ];
+ real mfabb = (D.f[DIR_M00])[kw ];
+ real mfbcb = (D.f[DIR_0P0])[k ];
+ real mfbab = (D.f[DIR_0M0])[ks ];
+ real mfbbc = (D.f[DIR_00P])[k ];
+ real mfbba = (D.f[DIR_00M])[kb ];
+ real mfccb = (D.f[DIR_PP0])[k ];
+ real mfaab = (D.f[DIR_MM0])[ksw];
+ real mfcab = (D.f[DIR_PM0])[ks ];
+ real mfacb = (D.f[DIR_MP0])[kw ];
+ real mfcbc = (D.f[DIR_P0P])[k ];
+ real mfaba = (D.f[DIR_M0M])[kbw];
+ real mfcba = (D.f[DIR_P0M])[kb ];
+ real mfabc = (D.f[DIR_M0P])[kw ];
+ real mfbcc = (D.f[DIR_0PP])[k ];
+ real mfbaa = (D.f[DIR_0MM])[kbs];
+ real mfbca = (D.f[DIR_0PM])[kb ];
+ real mfbac = (D.f[DIR_0MP])[ks ];
real mfbbb = (D.f[DIR_000])[k ];
- real mfccc = (D.f[DIR_PPP ])[k ];
- real mfaac = (D.f[DIR_MMP ])[ksw];
- real mfcac = (D.f[DIR_PMP ])[ks ];
- real mfacc = (D.f[DIR_MPP ])[kw ];
- real mfcca = (D.f[DIR_PPM ])[kb ];
- real mfaaa = (D.f[DIR_MMM ])[kbsw];
- real mfcaa = (D.f[DIR_PMM ])[kbs];
- real mfaca = (D.f[DIR_MPM ])[kbw];
+ real mfccc = (D.f[DIR_PPP])[k ];
+ real mfaac = (D.f[DIR_MMP])[ksw];
+ real mfcac = (D.f[DIR_PMP])[ks ];
+ real mfacc = (D.f[DIR_MPP])[kw ];
+ real mfcca = (D.f[DIR_PPM])[kb ];
+ real mfaaa = (D.f[DIR_MMM])[kbsw];
+ real mfcaa = (D.f[DIR_PMM])[kbs];
+ real mfaca = (D.f[DIR_MPM])[kbw];
////////////////////////////////////////////////////////////////////////////////////
real vvx =((((mfccc-mfaaa) + (mfcac-mfaca)) + ((mfcaa-mfacc) + (mfcca-mfaac))) +
(((mfcba-mfabc) + (mfcbc-mfaba)) + ((mfcab-mfacb) + (mfccb-mfaab))) +
@@ -857,7 +857,7 @@ __global__ void LBCalc3rdMomentsCompSP27(real* CUMbbb,
unsigned int* neighborY,
unsigned int* neighborZ,
real* DDStart,
- int size_Mat,
+ unsigned long long numberOfLBnodes,
bool EvenOrOdd)
{
////////////////////////////////////////////////////////////////////////////////
@@ -871,7 +871,7 @@ __global__ void LBCalc3rdMomentsCompSP27(real* CUMbbb,
const unsigned k = nx*(ny*z + y) + x;
//////////////////////////////////////////////////////////////////////////
- if(k<size_Mat)
+ if(k<numberOfLBnodes)
{
////////////////////////////////////////////////////////////////////////////////
unsigned int BC;
@@ -882,63 +882,63 @@ __global__ void LBCalc3rdMomentsCompSP27(real* CUMbbb,
Distributions27 D;
if (EvenOrOdd==true)
{
- D.f[DIR_P00 ] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DDStart[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DDStart[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DDStart[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DDStart[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DDStart[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DDStart[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DDStart[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DDStart[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DDStart[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DDStart[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DDStart[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DDStart[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DDStart[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DDStart[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DDStart[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DDStart[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DDStart[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DDStart[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DDStart[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DDStart[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DDStart[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DDStart[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DDStart[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DDStart[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DDStart[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DDStart[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DDStart[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00 ] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_MMM ] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_PPM ] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_MPM ] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_PMM ] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_MMP ] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_PPP ] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_MPP ] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_PMP ] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_M00] = &DDStart[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DDStart[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DDStart[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DDStart[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DDStart[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DDStart[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DDStart[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DDStart[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DDStart[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DDStart[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DDStart[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DDStart[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DDStart[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DDStart[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DDStart[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DDStart[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DDStart[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DDStart[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DDStart[DIR_000 * numberOfLBnodes];
+ D.f[DIR_MMM] = &DDStart[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DDStart[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DDStart[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DDStart[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DDStart[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PPP] = &DDStart[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DDStart[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DDStart[DIR_MPM * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////
@@ -951,33 +951,33 @@ __global__ void LBCalc3rdMomentsCompSP27(real* CUMbbb,
unsigned int kbs = neighborZ[ks];
unsigned int kbsw = neighborZ[ksw];
//////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
- real mfcbb = (D.f[DIR_P00 ])[k ];
- real mfabb = (D.f[DIR_M00 ])[kw ];
- real mfbcb = (D.f[DIR_0P0 ])[k ];
- real mfbab = (D.f[DIR_0M0 ])[ks ];
- real mfbbc = (D.f[DIR_00P ])[k ];
- real mfbba = (D.f[DIR_00M ])[kb ];
- real mfccb = (D.f[DIR_PP0 ])[k ];
- real mfaab = (D.f[DIR_MM0 ])[ksw];
- real mfcab = (D.f[DIR_PM0 ])[ks ];
- real mfacb = (D.f[DIR_MP0 ])[kw ];
- real mfcbc = (D.f[DIR_P0P ])[k ];
- real mfaba = (D.f[DIR_M0M ])[kbw];
- real mfcba = (D.f[DIR_P0M ])[kb ];
- real mfabc = (D.f[DIR_M0P ])[kw ];
- real mfbcc = (D.f[DIR_0PP ])[k ];
- real mfbaa = (D.f[DIR_0MM ])[kbs];
- real mfbca = (D.f[DIR_0PM ])[kb ];
- real mfbac = (D.f[DIR_0MP ])[ks ];
+ real mfcbb = (D.f[DIR_P00])[k ];
+ real mfabb = (D.f[DIR_M00])[kw ];
+ real mfbcb = (D.f[DIR_0P0])[k ];
+ real mfbab = (D.f[DIR_0M0])[ks ];
+ real mfbbc = (D.f[DIR_00P])[k ];
+ real mfbba = (D.f[DIR_00M])[kb ];
+ real mfccb = (D.f[DIR_PP0])[k ];
+ real mfaab = (D.f[DIR_MM0])[ksw];
+ real mfcab = (D.f[DIR_PM0])[ks ];
+ real mfacb = (D.f[DIR_MP0])[kw ];
+ real mfcbc = (D.f[DIR_P0P])[k ];
+ real mfaba = (D.f[DIR_M0M])[kbw];
+ real mfcba = (D.f[DIR_P0M])[kb ];
+ real mfabc = (D.f[DIR_M0P])[kw ];
+ real mfbcc = (D.f[DIR_0PP])[k ];
+ real mfbaa = (D.f[DIR_0MM])[kbs];
+ real mfbca = (D.f[DIR_0PM])[kb ];
+ real mfbac = (D.f[DIR_0MP])[ks ];
real mfbbb = (D.f[DIR_000])[k ];
- real mfccc = (D.f[DIR_PPP ])[k ];
- real mfaac = (D.f[DIR_MMP ])[ksw];
- real mfcac = (D.f[DIR_PMP ])[ks ];
- real mfacc = (D.f[DIR_MPP ])[kw ];
- real mfcca = (D.f[DIR_PPM ])[kb ];
- real mfaaa = (D.f[DIR_MMM ])[kbsw];
- real mfcaa = (D.f[DIR_PMM ])[kbs];
- real mfaca = (D.f[DIR_MPM ])[kbw];
+ real mfccc = (D.f[DIR_PPP])[k ];
+ real mfaac = (D.f[DIR_MMP])[ksw];
+ real mfcac = (D.f[DIR_PMP])[ks ];
+ real mfacc = (D.f[DIR_MPP])[kw ];
+ real mfcca = (D.f[DIR_PPM])[kb ];
+ real mfaaa = (D.f[DIR_MMM])[kbsw];
+ real mfcaa = (D.f[DIR_PMM])[kbs];
+ real mfaca = (D.f[DIR_MPM])[kbw];
////////////////////////////////////////////////////////////////////////////////////
real drho = ((((mfccc+mfaaa) + (mfaca+mfcac)) + ((mfacc+mfcaa) + (mfaac+mfcca))) +
(((mfbac+mfbca) + (mfbaa+mfbcc)) + ((mfabc+mfcba) + (mfaba+mfcbc)) + ((mfacb+mfcab) + (mfaab+mfccb))) +
@@ -1298,7 +1298,7 @@ __global__ void LBCalcHigherMomentsIncompSP27( real* CUMcbb,
unsigned int* neighborY,
unsigned int* neighborZ,
real* DDStart,
- int size_Mat,
+ unsigned long long numberOfLBnodes,
bool EvenOrOdd)
{
////////////////////////////////////////////////////////////////////////////////
@@ -1312,7 +1312,7 @@ __global__ void LBCalcHigherMomentsIncompSP27( real* CUMcbb,
const unsigned k = nx*(ny*z + y) + x;
//////////////////////////////////////////////////////////////////////////
- if(k<size_Mat)
+ if(k<numberOfLBnodes)
{
////////////////////////////////////////////////////////////////////////////////
unsigned int BC;
@@ -1323,63 +1323,63 @@ __global__ void LBCalcHigherMomentsIncompSP27( real* CUMcbb,
Distributions27 D;
if (EvenOrOdd==true)
{
- D.f[DIR_P00 ] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DDStart[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DDStart[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DDStart[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DDStart[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DDStart[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DDStart[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DDStart[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DDStart[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DDStart[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DDStart[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DDStart[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DDStart[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DDStart[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DDStart[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DDStart[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DDStart[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DDStart[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DDStart[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DDStart[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DDStart[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DDStart[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DDStart[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DDStart[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DDStart[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DDStart[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DDStart[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DDStart[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00 ] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_MMM ] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_PPM ] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_MPM ] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_PMM ] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_MMP ] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_PPP ] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_MPP ] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_PMP ] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_M00] = &DDStart[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DDStart[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DDStart[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DDStart[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DDStart[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DDStart[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DDStart[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DDStart[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DDStart[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DDStart[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DDStart[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DDStart[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DDStart[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DDStart[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DDStart[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DDStart[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DDStart[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DDStart[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DDStart[DIR_000 * numberOfLBnodes];
+ D.f[DIR_MMM] = &DDStart[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DDStart[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DDStart[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DDStart[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DDStart[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PPP] = &DDStart[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DDStart[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DDStart[DIR_MPM * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////
@@ -1392,33 +1392,33 @@ __global__ void LBCalcHigherMomentsIncompSP27( real* CUMcbb,
unsigned int kbs = neighborZ[ks];
unsigned int kbsw = neighborZ[ksw];
//////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
- real mfcbb = (D.f[DIR_P00 ])[k ];
- real mfabb = (D.f[DIR_M00 ])[kw ];
- real mfbcb = (D.f[DIR_0P0 ])[k ];
- real mfbab = (D.f[DIR_0M0 ])[ks ];
- real mfbbc = (D.f[DIR_00P ])[k ];
- real mfbba = (D.f[DIR_00M ])[kb ];
- real mfccb = (D.f[DIR_PP0 ])[k ];
- real mfaab = (D.f[DIR_MM0 ])[ksw];
- real mfcab = (D.f[DIR_PM0 ])[ks ];
- real mfacb = (D.f[DIR_MP0 ])[kw ];
- real mfcbc = (D.f[DIR_P0P ])[k ];
- real mfaba = (D.f[DIR_M0M ])[kbw];
- real mfcba = (D.f[DIR_P0M ])[kb ];
- real mfabc = (D.f[DIR_M0P ])[kw ];
- real mfbcc = (D.f[DIR_0PP ])[k ];
- real mfbaa = (D.f[DIR_0MM ])[kbs];
- real mfbca = (D.f[DIR_0PM ])[kb ];
- real mfbac = (D.f[DIR_0MP ])[ks ];
+ real mfcbb = (D.f[DIR_P00])[k ];
+ real mfabb = (D.f[DIR_M00])[kw ];
+ real mfbcb = (D.f[DIR_0P0])[k ];
+ real mfbab = (D.f[DIR_0M0])[ks ];
+ real mfbbc = (D.f[DIR_00P])[k ];
+ real mfbba = (D.f[DIR_00M])[kb ];
+ real mfccb = (D.f[DIR_PP0])[k ];
+ real mfaab = (D.f[DIR_MM0])[ksw];
+ real mfcab = (D.f[DIR_PM0])[ks ];
+ real mfacb = (D.f[DIR_MP0])[kw ];
+ real mfcbc = (D.f[DIR_P0P])[k ];
+ real mfaba = (D.f[DIR_M0M])[kbw];
+ real mfcba = (D.f[DIR_P0M])[kb ];
+ real mfabc = (D.f[DIR_M0P])[kw ];
+ real mfbcc = (D.f[DIR_0PP])[k ];
+ real mfbaa = (D.f[DIR_0MM])[kbs];
+ real mfbca = (D.f[DIR_0PM])[kb ];
+ real mfbac = (D.f[DIR_0MP])[ks ];
real mfbbb = (D.f[DIR_000])[k ];
- real mfccc = (D.f[DIR_PPP ])[k ];
- real mfaac = (D.f[DIR_MMP ])[ksw];
- real mfcac = (D.f[DIR_PMP ])[ks ];
- real mfacc = (D.f[DIR_MPP ])[kw ];
- real mfcca = (D.f[DIR_PPM ])[kb ];
- real mfaaa = (D.f[DIR_MMM ])[kbsw];
- real mfcaa = (D.f[DIR_PMM ])[kbs];
- real mfaca = (D.f[DIR_MPM ])[kbw];
+ real mfccc = (D.f[DIR_PPP])[k ];
+ real mfaac = (D.f[DIR_MMP])[ksw];
+ real mfcac = (D.f[DIR_PMP])[ks ];
+ real mfacc = (D.f[DIR_MPP])[kw ];
+ real mfcca = (D.f[DIR_PPM])[kb ];
+ real mfaaa = (D.f[DIR_MMM])[kbsw];
+ real mfcaa = (D.f[DIR_PMM])[kbs];
+ real mfaca = (D.f[DIR_MPM])[kbw];
////////////////////////////////////////////////////////////////////////////////////
real vvx =((((mfccc-mfaaa) + (mfcac-mfaca)) + ((mfcaa-mfacc) + (mfcca-mfaac))) +
(((mfcba-mfabc) + (mfcbc-mfaba)) + ((mfcab-mfacb) + (mfccb-mfaab))) +
@@ -1752,7 +1752,7 @@ __global__ void LBCalcHigherMomentsCompSP27( real* CUMcbb,
unsigned int* neighborY,
unsigned int* neighborZ,
real* DDStart,
- int size_Mat,
+ unsigned long long numberOfLBnodes,
bool EvenOrOdd)
{
////////////////////////////////////////////////////////////////////////////////
@@ -1766,7 +1766,7 @@ __global__ void LBCalcHigherMomentsCompSP27( real* CUMcbb,
const unsigned k = nx*(ny*z + y) + x;
//////////////////////////////////////////////////////////////////////////
- if(k<size_Mat)
+ if(k<numberOfLBnodes)
{
////////////////////////////////////////////////////////////////////////////////
unsigned int BC;
@@ -1777,63 +1777,63 @@ __global__ void LBCalcHigherMomentsCompSP27( real* CUMcbb,
Distributions27 D;
if (EvenOrOdd==true)
{
- D.f[DIR_P00 ] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DDStart[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DDStart[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DDStart[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DDStart[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DDStart[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DDStart[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DDStart[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DDStart[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DDStart[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DDStart[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DDStart[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DDStart[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DDStart[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DDStart[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DDStart[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DDStart[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DDStart[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DDStart[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DDStart[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DDStart[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DDStart[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DDStart[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DDStart[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DDStart[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DDStart[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DDStart[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DDStart[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00 ] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_MMM ] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_PPM ] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_MPM ] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_PMM ] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_MMP ] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_PPP ] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_MPP ] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_PMP ] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_M00] = &DDStart[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DDStart[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DDStart[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DDStart[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DDStart[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DDStart[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DDStart[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DDStart[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DDStart[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DDStart[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DDStart[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DDStart[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DDStart[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DDStart[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DDStart[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DDStart[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DDStart[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DDStart[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DDStart[DIR_000 * numberOfLBnodes];
+ D.f[DIR_MMM] = &DDStart[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DDStart[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DDStart[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DDStart[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DDStart[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PPP] = &DDStart[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DDStart[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DDStart[DIR_MPM * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////
@@ -1846,33 +1846,33 @@ __global__ void LBCalcHigherMomentsCompSP27( real* CUMcbb,
unsigned int kbs = neighborZ[ks];
unsigned int kbsw = neighborZ[ksw];
//////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
- real mfcbb = (D.f[DIR_P00 ])[k ];
- real mfabb = (D.f[DIR_M00 ])[kw ];
- real mfbcb = (D.f[DIR_0P0 ])[k ];
- real mfbab = (D.f[DIR_0M0 ])[ks ];
- real mfbbc = (D.f[DIR_00P ])[k ];
- real mfbba = (D.f[DIR_00M ])[kb ];
- real mfccb = (D.f[DIR_PP0 ])[k ];
- real mfaab = (D.f[DIR_MM0 ])[ksw];
- real mfcab = (D.f[DIR_PM0 ])[ks ];
- real mfacb = (D.f[DIR_MP0 ])[kw ];
- real mfcbc = (D.f[DIR_P0P ])[k ];
- real mfaba = (D.f[DIR_M0M ])[kbw];
- real mfcba = (D.f[DIR_P0M ])[kb ];
- real mfabc = (D.f[DIR_M0P ])[kw ];
- real mfbcc = (D.f[DIR_0PP ])[k ];
- real mfbaa = (D.f[DIR_0MM ])[kbs];
- real mfbca = (D.f[DIR_0PM ])[kb ];
- real mfbac = (D.f[DIR_0MP ])[ks ];
+ real mfcbb = (D.f[DIR_P00])[k ];
+ real mfabb = (D.f[DIR_M00])[kw ];
+ real mfbcb = (D.f[DIR_0P0])[k ];
+ real mfbab = (D.f[DIR_0M0])[ks ];
+ real mfbbc = (D.f[DIR_00P])[k ];
+ real mfbba = (D.f[DIR_00M])[kb ];
+ real mfccb = (D.f[DIR_PP0])[k ];
+ real mfaab = (D.f[DIR_MM0])[ksw];
+ real mfcab = (D.f[DIR_PM0])[ks ];
+ real mfacb = (D.f[DIR_MP0])[kw ];
+ real mfcbc = (D.f[DIR_P0P])[k ];
+ real mfaba = (D.f[DIR_M0M])[kbw];
+ real mfcba = (D.f[DIR_P0M])[kb ];
+ real mfabc = (D.f[DIR_M0P])[kw ];
+ real mfbcc = (D.f[DIR_0PP])[k ];
+ real mfbaa = (D.f[DIR_0MM])[kbs];
+ real mfbca = (D.f[DIR_0PM])[kb ];
+ real mfbac = (D.f[DIR_0MP])[ks ];
real mfbbb = (D.f[DIR_000])[k ];
- real mfccc = (D.f[DIR_PPP ])[k ];
- real mfaac = (D.f[DIR_MMP ])[ksw];
- real mfcac = (D.f[DIR_PMP ])[ks ];
- real mfacc = (D.f[DIR_MPP ])[kw ];
- real mfcca = (D.f[DIR_PPM ])[kb ];
- real mfaaa = (D.f[DIR_MMM ])[kbsw];
- real mfcaa = (D.f[DIR_PMM ])[kbs];
- real mfaca = (D.f[DIR_MPM ])[kbw];
+ real mfccc = (D.f[DIR_PPP])[k ];
+ real mfaac = (D.f[DIR_MMP])[ksw];
+ real mfcac = (D.f[DIR_PMP])[ks ];
+ real mfacc = (D.f[DIR_MPP])[kw ];
+ real mfcca = (D.f[DIR_PPM])[kb ];
+ real mfaaa = (D.f[DIR_MMM])[kbsw];
+ real mfcaa = (D.f[DIR_PMM])[kbs];
+ real mfaca = (D.f[DIR_MPM])[kbw];
////////////////////////////////////////////////////////////////////////////////////
real drho = ((((mfccc+mfaaa) + (mfaca+mfcac)) + ((mfacc+mfcaa) + (mfaac+mfcca))) +
(((mfbac+mfbca) + (mfbaa+mfbcc)) + ((mfabc+mfcba) + (mfaba+mfcbc)) + ((mfacb+mfcab) + (mfaab+mfccb))) +
diff --git a/src/gpu/VirtualFluids_GPU/GPU/CalcConc27.cu b/src/gpu/VirtualFluids_GPU/GPU/CalcConc27.cu
index d246f39a030b6df0b249aee17f37b7d5258ff00d..ad5a05b12a1b3ae2541e36ccffae4635fccfe62a 100644
--- a/src/gpu/VirtualFluids_GPU/GPU/CalcConc27.cu
+++ b/src/gpu/VirtualFluids_GPU/GPU/CalcConc27.cu
@@ -45,7 +45,7 @@ __global__ void CalcConc27(
uint* neighborX,
uint* neighborY,
uint* neighborZ,
- uint size_Mat,
+ unsigned long long numberOfLBnodes,
real* distributionsAD,
bool isEvenTimestep)
{
@@ -67,7 +67,7 @@ __global__ void CalcConc27(
//////////////////////////////////////////////////////////////////////////
// run for all indices in size_Mat and fluid nodes
- if ((k < size_Mat) && (typeOfGridNode[k] == GEO_FLUID))
+ if ((k < numberOfLBnodes) && (typeOfGridNode[k] == GEO_FLUID))
{
//////////////////////////////////////////////////////////////////////////
//! - Read distributions: style of reading and writing the distributions from/to stored arrays dependent on timestep is based on the esoteric twist algorithm \ref
@@ -76,63 +76,63 @@ __global__ void CalcConc27(
Distributions27 distAD;
if (isEvenTimestep)
{
- distAD.f[DIR_P00 ] = &distributionsAD[DIR_P00 *size_Mat];
- distAD.f[DIR_M00 ] = &distributionsAD[DIR_M00 *size_Mat];
- distAD.f[DIR_0P0 ] = &distributionsAD[DIR_0P0 *size_Mat];
- distAD.f[DIR_0M0 ] = &distributionsAD[DIR_0M0 *size_Mat];
- distAD.f[DIR_00P ] = &distributionsAD[DIR_00P *size_Mat];
- distAD.f[DIR_00M ] = &distributionsAD[DIR_00M *size_Mat];
- distAD.f[DIR_PP0 ] = &distributionsAD[DIR_PP0 *size_Mat];
- distAD.f[DIR_MM0 ] = &distributionsAD[DIR_MM0 *size_Mat];
- distAD.f[DIR_PM0 ] = &distributionsAD[DIR_PM0 *size_Mat];
- distAD.f[DIR_MP0 ] = &distributionsAD[DIR_MP0 *size_Mat];
- distAD.f[DIR_P0P ] = &distributionsAD[DIR_P0P *size_Mat];
- distAD.f[DIR_M0M ] = &distributionsAD[DIR_M0M *size_Mat];
- distAD.f[DIR_P0M ] = &distributionsAD[DIR_P0M *size_Mat];
- distAD.f[DIR_M0P ] = &distributionsAD[DIR_M0P *size_Mat];
- distAD.f[DIR_0PP ] = &distributionsAD[DIR_0PP *size_Mat];
- distAD.f[DIR_0MM ] = &distributionsAD[DIR_0MM *size_Mat];
- distAD.f[DIR_0PM ] = &distributionsAD[DIR_0PM *size_Mat];
- distAD.f[DIR_0MP ] = &distributionsAD[DIR_0MP *size_Mat];
- distAD.f[DIR_000] = &distributionsAD[DIR_000*size_Mat];
- distAD.f[DIR_PPP ] = &distributionsAD[DIR_PPP *size_Mat];
- distAD.f[DIR_MMP ] = &distributionsAD[DIR_MMP *size_Mat];
- distAD.f[DIR_PMP ] = &distributionsAD[DIR_PMP *size_Mat];
- distAD.f[DIR_MPP ] = &distributionsAD[DIR_MPP *size_Mat];
- distAD.f[DIR_PPM ] = &distributionsAD[DIR_PPM *size_Mat];
- distAD.f[DIR_MMM ] = &distributionsAD[DIR_MMM *size_Mat];
- distAD.f[DIR_PMM ] = &distributionsAD[DIR_PMM *size_Mat];
- distAD.f[DIR_MPM ] = &distributionsAD[DIR_MPM *size_Mat];
+ distAD.f[DIR_P00] = &distributionsAD[DIR_P00 * numberOfLBnodes];
+ distAD.f[DIR_M00] = &distributionsAD[DIR_M00 * numberOfLBnodes];
+ distAD.f[DIR_0P0] = &distributionsAD[DIR_0P0 * numberOfLBnodes];
+ distAD.f[DIR_0M0] = &distributionsAD[DIR_0M0 * numberOfLBnodes];
+ distAD.f[DIR_00P] = &distributionsAD[DIR_00P * numberOfLBnodes];
+ distAD.f[DIR_00M] = &distributionsAD[DIR_00M * numberOfLBnodes];
+ distAD.f[DIR_PP0] = &distributionsAD[DIR_PP0 * numberOfLBnodes];
+ distAD.f[DIR_MM0] = &distributionsAD[DIR_MM0 * numberOfLBnodes];
+ distAD.f[DIR_PM0] = &distributionsAD[DIR_PM0 * numberOfLBnodes];
+ distAD.f[DIR_MP0] = &distributionsAD[DIR_MP0 * numberOfLBnodes];
+ distAD.f[DIR_P0P] = &distributionsAD[DIR_P0P * numberOfLBnodes];
+ distAD.f[DIR_M0M] = &distributionsAD[DIR_M0M * numberOfLBnodes];
+ distAD.f[DIR_P0M] = &distributionsAD[DIR_P0M * numberOfLBnodes];
+ distAD.f[DIR_M0P] = &distributionsAD[DIR_M0P * numberOfLBnodes];
+ distAD.f[DIR_0PP] = &distributionsAD[DIR_0PP * numberOfLBnodes];
+ distAD.f[DIR_0MM] = &distributionsAD[DIR_0MM * numberOfLBnodes];
+ distAD.f[DIR_0PM] = &distributionsAD[DIR_0PM * numberOfLBnodes];
+ distAD.f[DIR_0MP] = &distributionsAD[DIR_0MP * numberOfLBnodes];
+ distAD.f[DIR_000] = &distributionsAD[DIR_000 * numberOfLBnodes];
+ distAD.f[DIR_PPP] = &distributionsAD[DIR_PPP * numberOfLBnodes];
+ distAD.f[DIR_MMP] = &distributionsAD[DIR_MMP * numberOfLBnodes];
+ distAD.f[DIR_PMP] = &distributionsAD[DIR_PMP * numberOfLBnodes];
+ distAD.f[DIR_MPP] = &distributionsAD[DIR_MPP * numberOfLBnodes];
+ distAD.f[DIR_PPM] = &distributionsAD[DIR_PPM * numberOfLBnodes];
+ distAD.f[DIR_MMM] = &distributionsAD[DIR_MMM * numberOfLBnodes];
+ distAD.f[DIR_PMM] = &distributionsAD[DIR_PMM * numberOfLBnodes];
+ distAD.f[DIR_MPM] = &distributionsAD[DIR_MPM * numberOfLBnodes];
}
else
{
- distAD.f[DIR_M00 ] = &distributionsAD[DIR_P00 *size_Mat];
- distAD.f[DIR_P00 ] = &distributionsAD[DIR_M00 *size_Mat];
- distAD.f[DIR_0M0 ] = &distributionsAD[DIR_0P0 *size_Mat];
- distAD.f[DIR_0P0 ] = &distributionsAD[DIR_0M0 *size_Mat];
- distAD.f[DIR_00M ] = &distributionsAD[DIR_00P *size_Mat];
- distAD.f[DIR_00P ] = &distributionsAD[DIR_00M *size_Mat];
- distAD.f[DIR_MM0 ] = &distributionsAD[DIR_PP0 *size_Mat];
- distAD.f[DIR_PP0 ] = &distributionsAD[DIR_MM0 *size_Mat];
- distAD.f[DIR_MP0 ] = &distributionsAD[DIR_PM0 *size_Mat];
- distAD.f[DIR_PM0 ] = &distributionsAD[DIR_MP0 *size_Mat];
- distAD.f[DIR_M0M ] = &distributionsAD[DIR_P0P *size_Mat];
- distAD.f[DIR_P0P ] = &distributionsAD[DIR_M0M *size_Mat];
- distAD.f[DIR_M0P ] = &distributionsAD[DIR_P0M *size_Mat];
- distAD.f[DIR_P0M ] = &distributionsAD[DIR_M0P *size_Mat];
- distAD.f[DIR_0MM ] = &distributionsAD[DIR_0PP *size_Mat];
- distAD.f[DIR_0PP ] = &distributionsAD[DIR_0MM *size_Mat];
- distAD.f[DIR_0MP ] = &distributionsAD[DIR_0PM *size_Mat];
- distAD.f[DIR_0PM ] = &distributionsAD[DIR_0MP *size_Mat];
- distAD.f[DIR_000] = &distributionsAD[DIR_000*size_Mat];
- distAD.f[DIR_PPP ] = &distributionsAD[DIR_MMM *size_Mat];
- distAD.f[DIR_MMP ] = &distributionsAD[DIR_PPM *size_Mat];
- distAD.f[DIR_PMP ] = &distributionsAD[DIR_MPM *size_Mat];
- distAD.f[DIR_MPP ] = &distributionsAD[DIR_PMM *size_Mat];
- distAD.f[DIR_PPM ] = &distributionsAD[DIR_MMP *size_Mat];
- distAD.f[DIR_MMM ] = &distributionsAD[DIR_PPP *size_Mat];
- distAD.f[DIR_PMM ] = &distributionsAD[DIR_MPP *size_Mat];
- distAD.f[DIR_MPM ] = &distributionsAD[DIR_PMP *size_Mat];
+ distAD.f[DIR_M00] = &distributionsAD[DIR_P00 * numberOfLBnodes];
+ distAD.f[DIR_P00] = &distributionsAD[DIR_M00 * numberOfLBnodes];
+ distAD.f[DIR_0M0] = &distributionsAD[DIR_0P0 * numberOfLBnodes];
+ distAD.f[DIR_0P0] = &distributionsAD[DIR_0M0 * numberOfLBnodes];
+ distAD.f[DIR_00M] = &distributionsAD[DIR_00P * numberOfLBnodes];
+ distAD.f[DIR_00P] = &distributionsAD[DIR_00M * numberOfLBnodes];
+ distAD.f[DIR_MM0] = &distributionsAD[DIR_PP0 * numberOfLBnodes];
+ distAD.f[DIR_PP0] = &distributionsAD[DIR_MM0 * numberOfLBnodes];
+ distAD.f[DIR_MP0] = &distributionsAD[DIR_PM0 * numberOfLBnodes];
+ distAD.f[DIR_PM0] = &distributionsAD[DIR_MP0 * numberOfLBnodes];
+ distAD.f[DIR_M0M] = &distributionsAD[DIR_P0P * numberOfLBnodes];
+ distAD.f[DIR_P0P] = &distributionsAD[DIR_M0M * numberOfLBnodes];
+ distAD.f[DIR_M0P] = &distributionsAD[DIR_P0M * numberOfLBnodes];
+ distAD.f[DIR_P0M] = &distributionsAD[DIR_M0P * numberOfLBnodes];
+ distAD.f[DIR_0MM] = &distributionsAD[DIR_0PP * numberOfLBnodes];
+ distAD.f[DIR_0PP] = &distributionsAD[DIR_0MM * numberOfLBnodes];
+ distAD.f[DIR_0MP] = &distributionsAD[DIR_0PM * numberOfLBnodes];
+ distAD.f[DIR_0PM] = &distributionsAD[DIR_0MP * numberOfLBnodes];
+ distAD.f[DIR_000] = &distributionsAD[DIR_000 * numberOfLBnodes];
+ distAD.f[DIR_PPP] = &distributionsAD[DIR_MMM * numberOfLBnodes];
+ distAD.f[DIR_MMP] = &distributionsAD[DIR_PPM * numberOfLBnodes];
+ distAD.f[DIR_PMP] = &distributionsAD[DIR_MPM * numberOfLBnodes];
+ distAD.f[DIR_MPP] = &distributionsAD[DIR_PMM * numberOfLBnodes];
+ distAD.f[DIR_PPM] = &distributionsAD[DIR_MMP * numberOfLBnodes];
+ distAD.f[DIR_MMM] = &distributionsAD[DIR_PPP * numberOfLBnodes];
+ distAD.f[DIR_PMM] = &distributionsAD[DIR_MPP * numberOfLBnodes];
+ distAD.f[DIR_MPM] = &distributionsAD[DIR_PMP * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////
//! - Set neighbor indices (necessary for indirect addressing)
@@ -166,33 +166,33 @@ __global__ void CalcConc27(
////////////////////////////////////////////////////////////////////////////////
//! - Set local distributions
//!
- real mfcbb = (distAD.f[DIR_P00 ])[ke ];
- real mfabb = (distAD.f[DIR_M00 ])[kw ];
- real mfbcb = (distAD.f[DIR_0P0 ])[kn ];
- real mfbab = (distAD.f[DIR_0M0 ])[ks ];
- real mfbbc = (distAD.f[DIR_00P ])[kt ];
- real mfbba = (distAD.f[DIR_00M ])[kb ];
- real mfccb = (distAD.f[DIR_PP0 ])[kne ];
- real mfaab = (distAD.f[DIR_MM0 ])[ksw ];
- real mfcab = (distAD.f[DIR_PM0 ])[kse ];
- real mfacb = (distAD.f[DIR_MP0 ])[knw ];
- real mfcbc = (distAD.f[DIR_P0P ])[kte ];
- real mfaba = (distAD.f[DIR_M0M ])[kbw ];
- real mfcba = (distAD.f[DIR_P0M ])[kbe ];
- real mfabc = (distAD.f[DIR_M0P ])[ktw ];
- real mfbcc = (distAD.f[DIR_0PP ])[ktn ];
- real mfbaa = (distAD.f[DIR_0MM ])[kbs ];
- real mfbca = (distAD.f[DIR_0PM ])[kbn ];
- real mfbac = (distAD.f[DIR_0MP ])[kts ];
+ real mfcbb = (distAD.f[DIR_P00])[ke ];
+ real mfabb = (distAD.f[DIR_M00])[kw ];
+ real mfbcb = (distAD.f[DIR_0P0])[kn ];
+ real mfbab = (distAD.f[DIR_0M0])[ks ];
+ real mfbbc = (distAD.f[DIR_00P])[kt ];
+ real mfbba = (distAD.f[DIR_00M])[kb ];
+ real mfccb = (distAD.f[DIR_PP0])[kne ];
+ real mfaab = (distAD.f[DIR_MM0])[ksw ];
+ real mfcab = (distAD.f[DIR_PM0])[kse ];
+ real mfacb = (distAD.f[DIR_MP0])[knw ];
+ real mfcbc = (distAD.f[DIR_P0P])[kte ];
+ real mfaba = (distAD.f[DIR_M0M])[kbw ];
+ real mfcba = (distAD.f[DIR_P0M])[kbe ];
+ real mfabc = (distAD.f[DIR_M0P])[ktw ];
+ real mfbcc = (distAD.f[DIR_0PP])[ktn ];
+ real mfbaa = (distAD.f[DIR_0MM])[kbs ];
+ real mfbca = (distAD.f[DIR_0PM])[kbn ];
+ real mfbac = (distAD.f[DIR_0MP])[kts ];
real mfbbb = (distAD.f[DIR_000])[k ];
- real mfccc = (distAD.f[DIR_PPP ])[ktne];
- real mfaac = (distAD.f[DIR_MMP ])[ktsw];
- real mfcac = (distAD.f[DIR_PMP ])[ktse];
- real mfacc = (distAD.f[DIR_MPP ])[ktnw];
- real mfcca = (distAD.f[DIR_PPM ])[kbne];
- real mfaaa = (distAD.f[DIR_MMM ])[kbsw];
- real mfcaa = (distAD.f[DIR_PMM ])[kbse];
- real mfaca = (distAD.f[DIR_MPM ])[kbnw];
+ real mfccc = (distAD.f[DIR_PPP])[ktne];
+ real mfaac = (distAD.f[DIR_MMP])[ktsw];
+ real mfcac = (distAD.f[DIR_PMP])[ktse];
+ real mfacc = (distAD.f[DIR_MPP])[ktnw];
+ real mfcca = (distAD.f[DIR_PPM])[kbne];
+ real mfaaa = (distAD.f[DIR_MMM])[kbsw];
+ real mfcaa = (distAD.f[DIR_PMM])[kbse];
+ real mfaca = (distAD.f[DIR_MPM])[kbnw];
//////////////////////////////////////////////////////////////////////////
//! - Calculate concentration using pyramid summation for low round-off errors as in Eq. (J1)-(J3) \ref
//! <a href="https://doi.org/10.1016/j.camwa.2015.05.001"><b>[ M. Geier et al. (2015), DOI:10.1016/j.camwa.2015.05.001 ]</b></a>
@@ -229,30 +229,30 @@ __global__ void CalcConc7( real* Conc,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
real* DD7,
bool isEvenTimestep)
{
Distributions7 D7;
if (isEvenTimestep==true)
{
- D7.f[0] = &DD7[0*size_Mat];
- D7.f[1] = &DD7[1*size_Mat];
- D7.f[2] = &DD7[2*size_Mat];
- D7.f[3] = &DD7[3*size_Mat];
- D7.f[4] = &DD7[4*size_Mat];
- D7.f[5] = &DD7[5*size_Mat];
- D7.f[6] = &DD7[6*size_Mat];
+ D7.f[0] = &DD7[0*numberOfLBnodes];
+ D7.f[1] = &DD7[1*numberOfLBnodes];
+ D7.f[2] = &DD7[2*numberOfLBnodes];
+ D7.f[3] = &DD7[3*numberOfLBnodes];
+ D7.f[4] = &DD7[4*numberOfLBnodes];
+ D7.f[5] = &DD7[5*numberOfLBnodes];
+ D7.f[6] = &DD7[6*numberOfLBnodes];
}
else
{
- D7.f[0] = &DD7[0*size_Mat];
- D7.f[2] = &DD7[1*size_Mat];
- D7.f[1] = &DD7[2*size_Mat];
- D7.f[4] = &DD7[3*size_Mat];
- D7.f[3] = &DD7[4*size_Mat];
- D7.f[6] = &DD7[5*size_Mat];
- D7.f[5] = &DD7[6*size_Mat];
+ D7.f[0] = &DD7[0*numberOfLBnodes];
+ D7.f[2] = &DD7[1*numberOfLBnodes];
+ D7.f[1] = &DD7[2*numberOfLBnodes];
+ D7.f[4] = &DD7[3*numberOfLBnodes];
+ D7.f[3] = &DD7[4*numberOfLBnodes];
+ D7.f[6] = &DD7[5*numberOfLBnodes];
+ D7.f[5] = &DD7[6*numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////
const unsigned x = threadIdx.x; // Globaler x-Index
@@ -265,7 +265,7 @@ __global__ void CalcConc7( real* Conc,
const unsigned k = nx*(ny*z + y) + x;
//////////////////////////////////////////////////////////////////////////
- if(k<size_Mat)
+ if(k<numberOfLBnodes)
{
//////////////////////////////////////////////////////////////////////////
//index
@@ -327,63 +327,63 @@ __global__ void CalcConc7( real* Conc,
// Distributions27 D27;
// if (isEvenTimestep==true)
// {
-// D27.f[DIR_P00 ] = &DD27[DIR_P00 *size_Mat];
-// D27.f[DIR_M00 ] = &DD27[DIR_M00 *size_Mat];
-// D27.f[DIR_0P0 ] = &DD27[DIR_0P0 *size_Mat];
-// D27.f[DIR_0M0 ] = &DD27[DIR_0M0 *size_Mat];
-// D27.f[DIR_00P ] = &DD27[DIR_00P *size_Mat];
-// D27.f[DIR_00M ] = &DD27[DIR_00M *size_Mat];
-// D27.f[DIR_PP0 ] = &DD27[DIR_PP0 *size_Mat];
-// D27.f[DIR_MM0 ] = &DD27[DIR_MM0 *size_Mat];
-// D27.f[DIR_PM0 ] = &DD27[DIR_PM0 *size_Mat];
-// D27.f[DIR_MP0 ] = &DD27[DIR_MP0 *size_Mat];
-// D27.f[DIR_P0P ] = &DD27[DIR_P0P *size_Mat];
-// D27.f[DIR_M0M ] = &DD27[DIR_M0M *size_Mat];
-// D27.f[DIR_P0M ] = &DD27[DIR_P0M *size_Mat];
-// D27.f[DIR_M0P ] = &DD27[DIR_M0P *size_Mat];
-// D27.f[DIR_0PP ] = &DD27[DIR_0PP *size_Mat];
-// D27.f[DIR_0MM ] = &DD27[DIR_0MM *size_Mat];
-// D27.f[DIR_0PM ] = &DD27[DIR_0PM *size_Mat];
-// D27.f[DIR_0MP ] = &DD27[DIR_0MP *size_Mat];
-// D27.f[DIR_000] = &DD27[DIR_000*size_Mat];
-// D27.f[DIR_PPP ] = &DD27[DIR_PPP *size_Mat];
-// D27.f[DIR_MMP ] = &DD27[DIR_MMP *size_Mat];
-// D27.f[DIR_PMP ] = &DD27[DIR_PMP *size_Mat];
-// D27.f[DIR_MPP ] = &DD27[DIR_MPP *size_Mat];
-// D27.f[DIR_PPM ] = &DD27[DIR_PPM *size_Mat];
-// D27.f[DIR_MMM ] = &DD27[DIR_MMM *size_Mat];
-// D27.f[DIR_PMM ] = &DD27[DIR_PMM *size_Mat];
-// D27.f[DIR_MPM ] = &DD27[DIR_MPM *size_Mat];
+// D27.f[DIR_P00] = &DD27[DIR_P00 * size_Mat];
+// D27.f[DIR_M00] = &DD27[DIR_M00 * size_Mat];
+// D27.f[DIR_0P0] = &DD27[DIR_0P0 * size_Mat];
+// D27.f[DIR_0M0] = &DD27[DIR_0M0 * size_Mat];
+// D27.f[DIR_00P] = &DD27[DIR_00P * size_Mat];
+// D27.f[DIR_00M] = &DD27[DIR_00M * size_Mat];
+// D27.f[DIR_PP0] = &DD27[DIR_PP0 * size_Mat];
+// D27.f[DIR_MM0] = &DD27[DIR_MM0 * size_Mat];
+// D27.f[DIR_PM0] = &DD27[DIR_PM0 * size_Mat];
+// D27.f[DIR_MP0] = &DD27[DIR_MP0 * size_Mat];
+// D27.f[DIR_P0P] = &DD27[DIR_P0P * size_Mat];
+// D27.f[DIR_M0M] = &DD27[DIR_M0M * size_Mat];
+// D27.f[DIR_P0M] = &DD27[DIR_P0M * size_Mat];
+// D27.f[DIR_M0P] = &DD27[DIR_M0P * size_Mat];
+// D27.f[DIR_0PP] = &DD27[DIR_0PP * size_Mat];
+// D27.f[DIR_0MM] = &DD27[DIR_0MM * size_Mat];
+// D27.f[DIR_0PM] = &DD27[DIR_0PM * size_Mat];
+// D27.f[DIR_0MP] = &DD27[DIR_0MP * size_Mat];
+// D27.f[DIR_000] = &DD27[DIR_000 * size_Mat];
+// D27.f[DIR_PPP] = &DD27[DIR_PPP * size_Mat];
+// D27.f[DIR_MMP] = &DD27[DIR_MMP * size_Mat];
+// D27.f[DIR_PMP] = &DD27[DIR_PMP * size_Mat];
+// D27.f[DIR_MPP] = &DD27[DIR_MPP * size_Mat];
+// D27.f[DIR_PPM] = &DD27[DIR_PPM * size_Mat];
+// D27.f[DIR_MMM] = &DD27[DIR_MMM * size_Mat];
+// D27.f[DIR_PMM] = &DD27[DIR_PMM * size_Mat];
+// D27.f[DIR_MPM] = &DD27[DIR_MPM * size_Mat];
// }
// else
// {
-// D27.f[DIR_M00 ] = &DD27[DIR_P00 *size_Mat];
-// D27.f[DIR_P00 ] = &DD27[DIR_M00 *size_Mat];
-// D27.f[DIR_0M0 ] = &DD27[DIR_0P0 *size_Mat];
-// D27.f[DIR_0P0 ] = &DD27[DIR_0M0 *size_Mat];
-// D27.f[DIR_00M ] = &DD27[DIR_00P *size_Mat];
-// D27.f[DIR_00P ] = &DD27[DIR_00M *size_Mat];
-// D27.f[DIR_MM0 ] = &DD27[DIR_PP0 *size_Mat];
-// D27.f[DIR_PP0 ] = &DD27[DIR_MM0 *size_Mat];
-// D27.f[DIR_MP0 ] = &DD27[DIR_PM0 *size_Mat];
-// D27.f[DIR_PM0 ] = &DD27[DIR_MP0 *size_Mat];
-// D27.f[DIR_M0M ] = &DD27[DIR_P0P *size_Mat];
-// D27.f[DIR_P0P ] = &DD27[DIR_M0M *size_Mat];
-// D27.f[DIR_M0P ] = &DD27[DIR_P0M *size_Mat];
-// D27.f[DIR_P0M ] = &DD27[DIR_M0P *size_Mat];
-// D27.f[DIR_0MM ] = &DD27[DIR_0PP *size_Mat];
-// D27.f[DIR_0PP ] = &DD27[DIR_0MM *size_Mat];
-// D27.f[DIR_0MP ] = &DD27[DIR_0PM *size_Mat];
-// D27.f[DIR_0PM ] = &DD27[DIR_0MP *size_Mat];
-// D27.f[DIR_000] = &DD27[DIR_000*size_Mat];
-// D27.f[DIR_MMM ] = &DD27[DIR_PPP *size_Mat];
-// D27.f[DIR_PPM ] = &DD27[DIR_MMP *size_Mat];
-// D27.f[DIR_MPM ] = &DD27[DIR_PMP *size_Mat];
-// D27.f[DIR_PMM ] = &DD27[DIR_MPP *size_Mat];
-// D27.f[DIR_MMP ] = &DD27[DIR_PPM *size_Mat];
-// D27.f[DIR_PPP ] = &DD27[DIR_MMM *size_Mat];
-// D27.f[DIR_MPP ] = &DD27[DIR_PMM *size_Mat];
-// D27.f[DIR_PMP ] = &DD27[DIR_MPM *size_Mat];
+// D27.f[DIR_M00] = &DD27[DIR_P00 * size_Mat];
+// D27.f[DIR_P00] = &DD27[DIR_M00 * size_Mat];
+// D27.f[DIR_0M0] = &DD27[DIR_0P0 * size_Mat];
+// D27.f[DIR_0P0] = &DD27[DIR_0M0 * size_Mat];
+// D27.f[DIR_00M] = &DD27[DIR_00P * size_Mat];
+// D27.f[DIR_00P] = &DD27[DIR_00M * size_Mat];
+// D27.f[DIR_MM0] = &DD27[DIR_PP0 * size_Mat];
+// D27.f[DIR_PP0] = &DD27[DIR_MM0 * size_Mat];
+// D27.f[DIR_MP0] = &DD27[DIR_PM0 * size_Mat];
+// D27.f[DIR_PM0] = &DD27[DIR_MP0 * size_Mat];
+// D27.f[DIR_M0M] = &DD27[DIR_P0P * size_Mat];
+// D27.f[DIR_P0P] = &DD27[DIR_M0M * size_Mat];
+// D27.f[DIR_M0P] = &DD27[DIR_P0M * size_Mat];
+// D27.f[DIR_P0M] = &DD27[DIR_M0P * size_Mat];
+// D27.f[DIR_0MM] = &DD27[DIR_0PP * size_Mat];
+// D27.f[DIR_0PP] = &DD27[DIR_0MM * size_Mat];
+// D27.f[DIR_0MP] = &DD27[DIR_0PM * size_Mat];
+// D27.f[DIR_0PM] = &DD27[DIR_0MP * size_Mat];
+// D27.f[DIR_000] = &DD27[DIR_000 * size_Mat];
+// D27.f[DIR_MMM] = &DD27[DIR_PPP * size_Mat];
+// D27.f[DIR_PPM] = &DD27[DIR_MMP * size_Mat];
+// D27.f[DIR_MPM] = &DD27[DIR_PMP * size_Mat];
+// D27.f[DIR_PMM] = &DD27[DIR_MPP * size_Mat];
+// D27.f[DIR_MMP] = &DD27[DIR_PPM * size_Mat];
+// D27.f[DIR_PPP] = &DD27[DIR_MMM * size_Mat];
+// D27.f[DIR_MPP] = &DD27[DIR_PMM * size_Mat];
+// D27.f[DIR_PMP] = &DD27[DIR_MPM * size_Mat];
// }
// ////////////////////////////////////////////////////////////////////////////////
// const unsigned x = threadIdx.x; // Globaler x-Index
@@ -432,20 +432,20 @@ __global__ void CalcConc7( real* Conc,
// if(geoD[k] == GEO_FLUID)
// {
-// Conc[k] = (D27.f[DIR_P00 ])[ke ]+ (D27.f[DIR_M00 ])[kw ]+
-// (D27.f[DIR_0P0 ])[kn ]+ (D27.f[DIR_0M0 ])[ks ]+
-// (D27.f[DIR_00P ])[kt ]+ (D27.f[DIR_00M ])[kb ]+
-// (D27.f[DIR_PP0 ])[kne ]+ (D27.f[DIR_MM0 ])[ksw ]+
-// (D27.f[DIR_PM0 ])[kse ]+ (D27.f[DIR_MP0 ])[knw ]+
-// (D27.f[DIR_P0P ])[kte ]+ (D27.f[DIR_M0M ])[kbw ]+
-// (D27.f[DIR_P0M ])[kbe ]+ (D27.f[DIR_M0P ])[ktw ]+
-// (D27.f[DIR_0PP ])[ktn ]+ (D27.f[DIR_0MM ])[kbs ]+
-// (D27.f[DIR_0PM ])[kbn ]+ (D27.f[DIR_0MP ])[kts ]+
+// Conc[k] = (D27.f[DIR_P00])[ke ]+ (D27.f[DIR_M00])[kw ]+
+// (D27.f[DIR_0P0])[kn ]+ (D27.f[DIR_0M0])[ks ]+
+// (D27.f[DIR_00P])[kt ]+ (D27.f[DIR_00M])[kb ]+
+// (D27.f[DIR_PP0])[kne ]+ (D27.f[DIR_MM0])[ksw ]+
+// (D27.f[DIR_PM0])[kse ]+ (D27.f[DIR_MP0])[knw ]+
+// (D27.f[DIR_P0P])[kte ]+ (D27.f[DIR_M0M])[kbw ]+
+// (D27.f[DIR_P0M])[kbe ]+ (D27.f[DIR_M0P])[ktw ]+
+// (D27.f[DIR_0PP])[ktn ]+ (D27.f[DIR_0MM])[kbs ]+
+// (D27.f[DIR_0PM])[kbn ]+ (D27.f[DIR_0MP])[kts ]+
// (D27.f[DIR_000])[kzero]+
-// (D27.f[DIR_PPP ])[ktne]+ (D27.f[DIR_MMP ])[ktsw]+
-// (D27.f[DIR_PMP ])[ktse]+ (D27.f[DIR_MPP ])[ktnw]+
-// (D27.f[DIR_PPM ])[kbne]+ (D27.f[DIR_MMM ])[kbsw]+
-// (D27.f[DIR_PMM ])[kbse]+ (D27.f[DIR_MPM ])[kbnw];
+// (D27.f[DIR_PPP])[ktne]+ (D27.f[DIR_MMP])[ktsw]+
+// (D27.f[DIR_PMP])[ktse]+ (D27.f[DIR_MPP])[ktnw]+
+// (D27.f[DIR_PPM])[kbne]+ (D27.f[DIR_MMM])[kbsw]+
+// (D27.f[DIR_PMM])[kbse]+ (D27.f[DIR_MPM])[kbnw];
// }
// }
// }
@@ -476,30 +476,30 @@ __global__ void GetPlaneConc7(real* Conc,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
real* DD7,
bool isEvenTimestep)
{
Distributions7 D7;
if (isEvenTimestep==true)
{
- D7.f[0] = &DD7[0*size_Mat];
- D7.f[1] = &DD7[1*size_Mat];
- D7.f[2] = &DD7[2*size_Mat];
- D7.f[3] = &DD7[3*size_Mat];
- D7.f[4] = &DD7[4*size_Mat];
- D7.f[5] = &DD7[5*size_Mat];
- D7.f[6] = &DD7[6*size_Mat];
+ D7.f[0] = &DD7[0*numberOfLBnodes];
+ D7.f[1] = &DD7[1*numberOfLBnodes];
+ D7.f[2] = &DD7[2*numberOfLBnodes];
+ D7.f[3] = &DD7[3*numberOfLBnodes];
+ D7.f[4] = &DD7[4*numberOfLBnodes];
+ D7.f[5] = &DD7[5*numberOfLBnodes];
+ D7.f[6] = &DD7[6*numberOfLBnodes];
}
else
{
- D7.f[0] = &DD7[0*size_Mat];
- D7.f[2] = &DD7[1*size_Mat];
- D7.f[1] = &DD7[2*size_Mat];
- D7.f[4] = &DD7[3*size_Mat];
- D7.f[3] = &DD7[4*size_Mat];
- D7.f[6] = &DD7[5*size_Mat];
- D7.f[5] = &DD7[6*size_Mat];
+ D7.f[0] = &DD7[0*numberOfLBnodes];
+ D7.f[2] = &DD7[1*numberOfLBnodes];
+ D7.f[1] = &DD7[2*numberOfLBnodes];
+ D7.f[4] = &DD7[3*numberOfLBnodes];
+ D7.f[3] = &DD7[4*numberOfLBnodes];
+ D7.f[6] = &DD7[5*numberOfLBnodes];
+ D7.f[5] = &DD7[6*numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////
const unsigned x = threadIdx.x; // Globaler x-Index
@@ -581,70 +581,70 @@ __global__ void GetPlaneConc27(real* Conc,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
real* DD27,
bool isEvenTimestep)
{
Distributions27 D27;
if (isEvenTimestep==true)
{
- D27.f[DIR_P00 ] = &DD27[DIR_P00 *size_Mat];
- D27.f[DIR_M00 ] = &DD27[DIR_M00 *size_Mat];
- D27.f[DIR_0P0 ] = &DD27[DIR_0P0 *size_Mat];
- D27.f[DIR_0M0 ] = &DD27[DIR_0M0 *size_Mat];
- D27.f[DIR_00P ] = &DD27[DIR_00P *size_Mat];
- D27.f[DIR_00M ] = &DD27[DIR_00M *size_Mat];
- D27.f[DIR_PP0 ] = &DD27[DIR_PP0 *size_Mat];
- D27.f[DIR_MM0 ] = &DD27[DIR_MM0 *size_Mat];
- D27.f[DIR_PM0 ] = &DD27[DIR_PM0 *size_Mat];
- D27.f[DIR_MP0 ] = &DD27[DIR_MP0 *size_Mat];
- D27.f[DIR_P0P ] = &DD27[DIR_P0P *size_Mat];
- D27.f[DIR_M0M ] = &DD27[DIR_M0M *size_Mat];
- D27.f[DIR_P0M ] = &DD27[DIR_P0M *size_Mat];
- D27.f[DIR_M0P ] = &DD27[DIR_M0P *size_Mat];
- D27.f[DIR_0PP ] = &DD27[DIR_0PP *size_Mat];
- D27.f[DIR_0MM ] = &DD27[DIR_0MM *size_Mat];
- D27.f[DIR_0PM ] = &DD27[DIR_0PM *size_Mat];
- D27.f[DIR_0MP ] = &DD27[DIR_0MP *size_Mat];
- D27.f[DIR_000] = &DD27[DIR_000*size_Mat];
- D27.f[DIR_PPP ] = &DD27[DIR_PPP *size_Mat];
- D27.f[DIR_MMP ] = &DD27[DIR_MMP *size_Mat];
- D27.f[DIR_PMP ] = &DD27[DIR_PMP *size_Mat];
- D27.f[DIR_MPP ] = &DD27[DIR_MPP *size_Mat];
- D27.f[DIR_PPM ] = &DD27[DIR_PPM *size_Mat];
- D27.f[DIR_MMM ] = &DD27[DIR_MMM *size_Mat];
- D27.f[DIR_PMM ] = &DD27[DIR_PMM *size_Mat];
- D27.f[DIR_MPM ] = &DD27[DIR_MPM *size_Mat];
+ D27.f[DIR_P00] = &DD27[DIR_P00 * numberOfLBnodes];
+ D27.f[DIR_M00] = &DD27[DIR_M00 * numberOfLBnodes];
+ D27.f[DIR_0P0] = &DD27[DIR_0P0 * numberOfLBnodes];
+ D27.f[DIR_0M0] = &DD27[DIR_0M0 * numberOfLBnodes];
+ D27.f[DIR_00P] = &DD27[DIR_00P * numberOfLBnodes];
+ D27.f[DIR_00M] = &DD27[DIR_00M * numberOfLBnodes];
+ D27.f[DIR_PP0] = &DD27[DIR_PP0 * numberOfLBnodes];
+ D27.f[DIR_MM0] = &DD27[DIR_MM0 * numberOfLBnodes];
+ D27.f[DIR_PM0] = &DD27[DIR_PM0 * numberOfLBnodes];
+ D27.f[DIR_MP0] = &DD27[DIR_MP0 * numberOfLBnodes];
+ D27.f[DIR_P0P] = &DD27[DIR_P0P * numberOfLBnodes];
+ D27.f[DIR_M0M] = &DD27[DIR_M0M * numberOfLBnodes];
+ D27.f[DIR_P0M] = &DD27[DIR_P0M * numberOfLBnodes];
+ D27.f[DIR_M0P] = &DD27[DIR_M0P * numberOfLBnodes];
+ D27.f[DIR_0PP] = &DD27[DIR_0PP * numberOfLBnodes];
+ D27.f[DIR_0MM] = &DD27[DIR_0MM * numberOfLBnodes];
+ D27.f[DIR_0PM] = &DD27[DIR_0PM * numberOfLBnodes];
+ D27.f[DIR_0MP] = &DD27[DIR_0MP * numberOfLBnodes];
+ D27.f[DIR_000] = &DD27[DIR_000 * numberOfLBnodes];
+ D27.f[DIR_PPP] = &DD27[DIR_PPP * numberOfLBnodes];
+ D27.f[DIR_MMP] = &DD27[DIR_MMP * numberOfLBnodes];
+ D27.f[DIR_PMP] = &DD27[DIR_PMP * numberOfLBnodes];
+ D27.f[DIR_MPP] = &DD27[DIR_MPP * numberOfLBnodes];
+ D27.f[DIR_PPM] = &DD27[DIR_PPM * numberOfLBnodes];
+ D27.f[DIR_MMM] = &DD27[DIR_MMM * numberOfLBnodes];
+ D27.f[DIR_PMM] = &DD27[DIR_PMM * numberOfLBnodes];
+ D27.f[DIR_MPM] = &DD27[DIR_MPM * numberOfLBnodes];
}
else
{
- D27.f[DIR_M00 ] = &DD27[DIR_P00 *size_Mat];
- D27.f[DIR_P00 ] = &DD27[DIR_M00 *size_Mat];
- D27.f[DIR_0M0 ] = &DD27[DIR_0P0 *size_Mat];
- D27.f[DIR_0P0 ] = &DD27[DIR_0M0 *size_Mat];
- D27.f[DIR_00M ] = &DD27[DIR_00P *size_Mat];
- D27.f[DIR_00P ] = &DD27[DIR_00M *size_Mat];
- D27.f[DIR_MM0 ] = &DD27[DIR_PP0 *size_Mat];
- D27.f[DIR_PP0 ] = &DD27[DIR_MM0 *size_Mat];
- D27.f[DIR_MP0 ] = &DD27[DIR_PM0 *size_Mat];
- D27.f[DIR_PM0 ] = &DD27[DIR_MP0 *size_Mat];
- D27.f[DIR_M0M ] = &DD27[DIR_P0P *size_Mat];
- D27.f[DIR_P0P ] = &DD27[DIR_M0M *size_Mat];
- D27.f[DIR_M0P ] = &DD27[DIR_P0M *size_Mat];
- D27.f[DIR_P0M ] = &DD27[DIR_M0P *size_Mat];
- D27.f[DIR_0MM ] = &DD27[DIR_0PP *size_Mat];
- D27.f[DIR_0PP ] = &DD27[DIR_0MM *size_Mat];
- D27.f[DIR_0MP ] = &DD27[DIR_0PM *size_Mat];
- D27.f[DIR_0PM ] = &DD27[DIR_0MP *size_Mat];
- D27.f[DIR_000] = &DD27[DIR_000*size_Mat];
- D27.f[DIR_MMM ] = &DD27[DIR_PPP *size_Mat];
- D27.f[DIR_PPM ] = &DD27[DIR_MMP *size_Mat];
- D27.f[DIR_MPM ] = &DD27[DIR_PMP *size_Mat];
- D27.f[DIR_PMM ] = &DD27[DIR_MPP *size_Mat];
- D27.f[DIR_MMP ] = &DD27[DIR_PPM *size_Mat];
- D27.f[DIR_PPP ] = &DD27[DIR_MMM *size_Mat];
- D27.f[DIR_MPP ] = &DD27[DIR_PMM *size_Mat];
- D27.f[DIR_PMP ] = &DD27[DIR_MPM *size_Mat];
+ D27.f[DIR_M00] = &DD27[DIR_P00 * numberOfLBnodes];
+ D27.f[DIR_P00] = &DD27[DIR_M00 * numberOfLBnodes];
+ D27.f[DIR_0M0] = &DD27[DIR_0P0 * numberOfLBnodes];
+ D27.f[DIR_0P0] = &DD27[DIR_0M0 * numberOfLBnodes];
+ D27.f[DIR_00M] = &DD27[DIR_00P * numberOfLBnodes];
+ D27.f[DIR_00P] = &DD27[DIR_00M * numberOfLBnodes];
+ D27.f[DIR_MM0] = &DD27[DIR_PP0 * numberOfLBnodes];
+ D27.f[DIR_PP0] = &DD27[DIR_MM0 * numberOfLBnodes];
+ D27.f[DIR_MP0] = &DD27[DIR_PM0 * numberOfLBnodes];
+ D27.f[DIR_PM0] = &DD27[DIR_MP0 * numberOfLBnodes];
+ D27.f[DIR_M0M] = &DD27[DIR_P0P * numberOfLBnodes];
+ D27.f[DIR_P0P] = &DD27[DIR_M0M * numberOfLBnodes];
+ D27.f[DIR_M0P] = &DD27[DIR_P0M * numberOfLBnodes];
+ D27.f[DIR_P0M] = &DD27[DIR_M0P * numberOfLBnodes];
+ D27.f[DIR_0MM] = &DD27[DIR_0PP * numberOfLBnodes];
+ D27.f[DIR_0PP] = &DD27[DIR_0MM * numberOfLBnodes];
+ D27.f[DIR_0MP] = &DD27[DIR_0PM * numberOfLBnodes];
+ D27.f[DIR_0PM] = &DD27[DIR_0MP * numberOfLBnodes];
+ D27.f[DIR_000] = &DD27[DIR_000 * numberOfLBnodes];
+ D27.f[DIR_MMM] = &DD27[DIR_PPP * numberOfLBnodes];
+ D27.f[DIR_PPM] = &DD27[DIR_MMP * numberOfLBnodes];
+ D27.f[DIR_MPM] = &DD27[DIR_PMP * numberOfLBnodes];
+ D27.f[DIR_PMM] = &DD27[DIR_MPP * numberOfLBnodes];
+ D27.f[DIR_MMP] = &DD27[DIR_PPM * numberOfLBnodes];
+ D27.f[DIR_PPP] = &DD27[DIR_MMM * numberOfLBnodes];
+ D27.f[DIR_MPP] = &DD27[DIR_PMM * numberOfLBnodes];
+ D27.f[DIR_PMP] = &DD27[DIR_MPM * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////
const unsigned x = threadIdx.x; // Globaler x-Index
@@ -693,20 +693,20 @@ __global__ void GetPlaneConc27(real* Conc,
if(geoD[k] == GEO_FLUID)
{
- Conc[k] = (D27.f[DIR_P00 ])[ke ]+ (D27.f[DIR_M00 ])[kw ]+
- (D27.f[DIR_0P0 ])[kn ]+ (D27.f[DIR_0M0 ])[ks ]+
- (D27.f[DIR_00P ])[kt ]+ (D27.f[DIR_00M ])[kb ]+
- (D27.f[DIR_PP0 ])[kne ]+ (D27.f[DIR_MM0 ])[ksw ]+
- (D27.f[DIR_PM0 ])[kse ]+ (D27.f[DIR_MP0 ])[knw ]+
- (D27.f[DIR_P0P ])[kte ]+ (D27.f[DIR_M0M ])[kbw ]+
- (D27.f[DIR_P0M ])[kbe ]+ (D27.f[DIR_M0P ])[ktw ]+
- (D27.f[DIR_0PP ])[ktn ]+ (D27.f[DIR_0MM ])[kbs ]+
- (D27.f[DIR_0PM ])[kbn ]+ (D27.f[DIR_0MP ])[kts ]+
+ Conc[k] = (D27.f[DIR_P00])[ke ]+ (D27.f[DIR_M00])[kw ]+
+ (D27.f[DIR_0P0])[kn ]+ (D27.f[DIR_0M0])[ks ]+
+ (D27.f[DIR_00P])[kt ]+ (D27.f[DIR_00M])[kb ]+
+ (D27.f[DIR_PP0])[kne ]+ (D27.f[DIR_MM0])[ksw ]+
+ (D27.f[DIR_PM0])[kse ]+ (D27.f[DIR_MP0])[knw ]+
+ (D27.f[DIR_P0P])[kte ]+ (D27.f[DIR_M0M])[kbw ]+
+ (D27.f[DIR_P0M])[kbe ]+ (D27.f[DIR_M0P])[ktw ]+
+ (D27.f[DIR_0PP])[ktn ]+ (D27.f[DIR_0MM])[kbs ]+
+ (D27.f[DIR_0PM])[kbn ]+ (D27.f[DIR_0MP])[kts ]+
(D27.f[DIR_000])[kzero]+
- (D27.f[DIR_PPP ])[ktne]+ (D27.f[DIR_MMP ])[ktsw]+
- (D27.f[DIR_PMP ])[ktse]+ (D27.f[DIR_MPP ])[ktnw]+
- (D27.f[DIR_PPM ])[kbne]+ (D27.f[DIR_MMM ])[kbsw]+
- (D27.f[DIR_PMM ])[kbse]+ (D27.f[DIR_MPM ])[kbnw];
+ (D27.f[DIR_PPP])[ktne]+ (D27.f[DIR_MMP])[ktsw]+
+ (D27.f[DIR_PMP])[ktse]+ (D27.f[DIR_MPP])[ktnw]+
+ (D27.f[DIR_PPM])[kbne]+ (D27.f[DIR_MMM])[kbsw]+
+ (D27.f[DIR_PMM])[kbse]+ (D27.f[DIR_MPM])[kbnw];
}
}
}
\ No newline at end of file
diff --git a/src/gpu/VirtualFluids_GPU/GPU/CalcMac27.cu b/src/gpu/VirtualFluids_GPU/GPU/CalcMac27.cu
index 4792b8846b2612383c07a97419e0473b21ebd187..f7bb09f816f45973fd4e2319a1bfa35cf9172caa 100644
--- a/src/gpu/VirtualFluids_GPU/GPU/CalcMac27.cu
+++ b/src/gpu/VirtualFluids_GPU/GPU/CalcMac27.cu
@@ -1,306 +1,310 @@
-// _ ___ __ __________ _ __ ______________ __
-// | | / (_)____/ /___ ______ _/ / ____/ /_ __(_)___/ /____ / ___/ __ / / / /
-// | | / / / ___/ __/ / / / __ `/ / /_ / / / / / / __ / ___/ / /___/ /_/ / / / /
-// | |/ / / / / /_/ /_/ / /_/ / / __/ / / /_/ / / /_/ (__ ) / /_) / ____/ /__/ /
-// |___/_/_/ \__/\__,_/\__,_/_/_/ /_/\__,_/_/\__,_/____/ \____/_/ \_____/
+//=======================================================================================
+// ____ ____ __ ______ __________ __ __ __ __
+// \ \ | | | | | _ \ |___ ___| | | | | / \ | |
+// \ \ | | | | | |_) | | | | | | | / \ | |
+// \ \ | | | | | _ / | | | | | | / /\ \ | |
+// \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
+// \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
+// \ \ | | ________________________________________________________________
+// \ \ | | | ______________________________________________________________|
+// \ \| | | | __ __ __ __ ______ _______
+// \ | | |_____ | | | | | | | | | _ \ / _____)
+// \ | | _____| | | | | | | | | | | \ \ \_______
+// \ | | | | |_____ | \_/ | | | | |_/ / _____ |
+// \ _____| |__| |________| \_______/ |__| |______/ (_______/
//
-//////////////////////////////////////////////////////////////////////////
-/* Device code */
+// This file is part of VirtualFluids. VirtualFluids is free software: you can
+// redistribute it and/or modify it under the terms of the GNU General Public
+// License as published by the Free Software Foundation, either version 3 of
+// the License, or (at your option) any later version.
+//
+// VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
+// ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+// FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
+// for more details.
+//
+// You should have received a copy of the GNU General Public License along
+// with VirtualFluids (see COPYING.txt). If not, see <http://www.gnu.org/licenses/>.
+//
+//! \file CalcMac27.cu
+//! \ingroup GPU
+//! \author Martin Schoenherr, Soeren Peters
+//======================================================================================
#include "LBM/LB.h"
#include "lbm/constants/D3Q27.h"
#include "lbm/constants/NumericConstants.h"
+#include "lbm/MacroscopicQuantities.h"
+
+#include "Kernel/Utilities/DistributionHelper.cuh"
using namespace vf::lbm::constant;
using namespace vf::lbm::dir;
+using namespace vf::gpu;
+
+////////////////////////////////////////////////////////////////////////////////
+__global__ void LBCalcMac27(
+ real* vxD,
+ real* vyD,
+ real* vzD,
+ real* rhoD,
+ unsigned int* geoD,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ real* distributions,
+ bool isEvenTimestep)
+{
+ const unsigned int tx = threadIdx.x; // Thread index = lokaler i index
+ const unsigned int by = blockIdx.x; // Block index x
+ const unsigned int bz = blockIdx.y; // Block index y
+ const unsigned int x = tx + STARTOFFX; // Globaler x-Index
+ const unsigned int y = by + STARTOFFY; // Globaler y-Index
+ const unsigned int z = bz + STARTOFFZ; // Globaler z-Index
+
+ const unsigned nx = blockDim.x + 2 * STARTOFFX;
+ const unsigned ny = gridDim.x + 2 * STARTOFFY;
+
+ const unsigned int k = nx*(ny*z + y) + x; // Zugriff auf arrays im device
+
+
+ if(k >= numberOfLBnodes)
+ return;
+
+ if(!isValidFluidNode(geoD[k]))
+ return;
+
+ rhoD[k] = c0o1;
+ vxD[k] = c0o1;
+ vyD[k] = c0o1;
+ vzD[k] = c0o1;
+
+ DistributionWrapper distr_wrapper(distributions, numberOfLBnodes, isEvenTimestep, k, neighborX, neighborY, neighborZ);
+ const auto& distribution = distr_wrapper.distribution;
+
+ rhoD[k] = vf::lbm::getDensity(distribution.f);
+ vxD[k] = vf::lbm::getIncompressibleVelocityX1(distribution.f);
+ vyD[k] = vf::lbm::getIncompressibleVelocityX2(distribution.f);
+ vzD[k] = vf::lbm::getIncompressibleVelocityX3(distribution.f);
+}
+
-#include "lbm/MacroscopicQuantities.h"
-#include "../Kernel/Utilities/DistributionHelper.cuh"
////////////////////////////////////////////////////////////////////////////////
-__global__ void LBCalcMac27( real* vxD,
- real* vyD,
- real* vzD,
- real* rhoD,
- unsigned int* geoD,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- real* distributions,
- bool isEvenTimestep)
+__global__ void LBCalcMacSP27(
+ real* vxD,
+ real* vyD,
+ real* vzD,
+ real* rhoD,
+ real* pressD,
+ unsigned int* geoD,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ real* distributions,
+ bool isEvenTimestep)
{
- const unsigned int tx = threadIdx.x; // Thread index = lokaler i index
- const unsigned int by = blockIdx.x; // Block index x
- const unsigned int bz = blockIdx.y; // Block index y
- const unsigned int x = tx + STARTOFFX; // Globaler x-Index
- const unsigned int y = by + STARTOFFY; // Globaler y-Index
- const unsigned int z = bz + STARTOFFZ; // Globaler z-Index
+ ////////////////////////////////////////////////////////////////////////////////
+ //! - Get node index coordinates from threadIdx, blockIdx, blockDim and gridDim.
+ //!
+ const unsigned nodeIndex = getNodeIndex();
+
+ //////////////////////////////////////////////////////////////////////////
+ if(nodeIndex<numberOfLBnodes)
+ {
+ //////////////////////////////////////////////////////////////////////////
+ //! - Read distributions: style of reading and writing the distributions from/to stored arrays dependent on
+ //! timestep is based on the esoteric twist algorithm \ref <a
+ //! href="https://doi.org/10.3390/computation5020019"><b>[ M. Geier et al. (2017),
+ //! DOI:10.3390/computation5020019 ]</b></a>
+ //!
+ Distributions27 dist;
+ getPointersToDistributions(dist, distributions, numberOfLBnodes, isEvenTimestep);
+
+ //////////////////////////////////////////////////////////////////////////
+ //index
+ unsigned int kzero= nodeIndex;
+ unsigned int ke = nodeIndex;
+ unsigned int kw = neighborX[nodeIndex];
+ unsigned int kn = nodeIndex;
+ unsigned int ks = neighborY[nodeIndex];
+ unsigned int kt = nodeIndex;
+ unsigned int kb = neighborZ[nodeIndex];
+ unsigned int ksw = neighborY[kw];
+ unsigned int kne = nodeIndex;
+ unsigned int kse = ks;
+ unsigned int knw = kw;
+ unsigned int kbw = neighborZ[kw];
+ unsigned int kte = nodeIndex;
+ unsigned int kbe = kb;
+ unsigned int ktw = kw;
+ unsigned int kbs = neighborZ[ks];
+ unsigned int ktn = nodeIndex;
+ unsigned int kbn = kb;
+ unsigned int kts = ks;
+ unsigned int ktse = ks;
+ unsigned int kbnw = kbw;
+ unsigned int ktnw = kw;
+ unsigned int kbse = kbs;
+ unsigned int ktsw = ksw;
+ unsigned int kbne = kb;
+ unsigned int ktne = nodeIndex;
+ unsigned int kbsw = neighborZ[ksw];
+ //////////////////////////////////////////////////////////////////////////
+ pressD[nodeIndex] = c0o1;
+ rhoD[nodeIndex] = c0o1;
+ vxD[nodeIndex] = c0o1;
+ vyD[nodeIndex] = c0o1;
+ vzD[nodeIndex] = c0o1;
+
+ if(geoD[nodeIndex] == GEO_FLUID)
+ {
+ rhoD[nodeIndex] =
+ (dist.f[DIR_P00])[ke ]+ (dist.f[DIR_M00])[kw ]+
+ (dist.f[DIR_0P0])[kn ]+ (dist.f[DIR_0M0])[ks ]+
+ (dist.f[DIR_00P])[kt ]+ (dist.f[DIR_00M])[kb ]+
+ (dist.f[DIR_PP0])[kne ]+ (dist.f[DIR_MM0])[ksw ]+
+ (dist.f[DIR_PM0])[kse ]+ (dist.f[DIR_MP0])[knw ]+
+ (dist.f[DIR_P0P])[kte ]+ (dist.f[DIR_M0M])[kbw ]+
+ (dist.f[DIR_P0M])[kbe ]+ (dist.f[DIR_M0P])[ktw ]+
+ (dist.f[DIR_0PP])[ktn ]+ (dist.f[DIR_0MM])[kbs ]+
+ (dist.f[DIR_0PM])[kbn ]+ (dist.f[DIR_0MP])[kts ]+
+ (dist.f[DIR_000])[kzero]+
+ (dist.f[DIR_PPP])[ktne]+ (dist.f[DIR_MMP])[ktsw]+
+ (dist.f[DIR_PMP])[ktse]+ (dist.f[DIR_MPP])[ktnw]+
+ (dist.f[DIR_PPM])[kbne]+ (dist.f[DIR_MMM])[kbsw]+
+ (dist.f[DIR_PMM])[kbse]+ (dist.f[DIR_MPM])[kbnw];
+
+ vxD[nodeIndex] =
+ (dist.f[DIR_P00])[ke ]- (dist.f[DIR_M00])[kw ]+
+ (dist.f[DIR_PP0])[kne ]- (dist.f[DIR_MM0])[ksw ]+
+ (dist.f[DIR_PM0])[kse ]- (dist.f[DIR_MP0])[knw ]+
+ (dist.f[DIR_P0P])[kte ]- (dist.f[DIR_M0M])[kbw ]+
+ (dist.f[DIR_P0M])[kbe ]- (dist.f[DIR_M0P])[ktw ]+
+ (dist.f[DIR_PPP])[ktne]- (dist.f[DIR_MMP])[ktsw]+
+ (dist.f[DIR_PMP])[ktse]- (dist.f[DIR_MPP])[ktnw]+
+ (dist.f[DIR_PPM])[kbne]- (dist.f[DIR_MMM])[kbsw]+
+ (dist.f[DIR_PMM])[kbse]- (dist.f[DIR_MPM])[kbnw];
+
+ vyD[nodeIndex] =
+ (dist.f[DIR_0P0])[kn ]- (dist.f[DIR_0M0])[ks ]+
+ (dist.f[DIR_PP0])[kne ]- (dist.f[DIR_MM0])[ksw ]-
+ (dist.f[DIR_PM0])[kse ]+ (dist.f[DIR_MP0])[knw ]+
+ (dist.f[DIR_0PP])[ktn ]- (dist.f[DIR_0MM])[kbs ]+
+ (dist.f[DIR_0PM])[kbn ]- (dist.f[DIR_0MP])[kts ]+
+ (dist.f[DIR_PPP])[ktne]- (dist.f[DIR_MMP])[ktsw]-
+ (dist.f[DIR_PMP])[ktse]+ (dist.f[DIR_MPP])[ktnw]+
+ (dist.f[DIR_PPM])[kbne]- (dist.f[DIR_MMM])[kbsw]-
+ (dist.f[DIR_PMM])[kbse]+ (dist.f[DIR_MPM])[kbnw];
+
+ vzD[nodeIndex] =
+ (dist.f[DIR_00P])[kt ]- (dist.f[DIR_00M])[kb ]+
+ (dist.f[DIR_P0P])[kte ]- (dist.f[DIR_M0M])[kbw ]-
+ (dist.f[DIR_P0M])[kbe ]+ (dist.f[DIR_M0P])[ktw ]+
+ (dist.f[DIR_0PP])[ktn ]- (dist.f[DIR_0MM])[kbs ]-
+ (dist.f[DIR_0PM])[kbn ]+ (dist.f[DIR_0MP])[kts ]+
+ (dist.f[DIR_PPP])[ktne]+ (dist.f[DIR_MMP])[ktsw]+
+ (dist.f[DIR_PMP])[ktse]+ (dist.f[DIR_MPP])[ktnw]-
+ (dist.f[DIR_PPM])[kbne]- (dist.f[DIR_MMM])[kbsw]-
+ (dist.f[DIR_PMM])[kbse]- (dist.f[DIR_MPM])[kbnw];
+
+ pressD[nodeIndex] =
+ ((dist.f[DIR_P00])[ke ]+ (dist.f[DIR_M00])[kw ]+
+ (dist.f[DIR_0P0])[kn ]+ (dist.f[DIR_0M0])[ks ]+
+ (dist.f[DIR_00P])[kt ]+ (dist.f[DIR_00M])[kb ]+
+ 2.f*(
+ (dist.f[DIR_PP0])[kne ]+ (dist.f[DIR_MM0])[ksw ]+
+ (dist.f[DIR_PM0])[kse ]+ (dist.f[DIR_MP0])[knw ]+
+ (dist.f[DIR_P0P])[kte ]+ (dist.f[DIR_M0M])[kbw ]+
+ (dist.f[DIR_P0M])[kbe ]+ (dist.f[DIR_M0P])[ktw ]+
+ (dist.f[DIR_0PP])[ktn ]+ (dist.f[DIR_0MM])[kbs ]+
+ (dist.f[DIR_0PM])[kbn ]+ (dist.f[DIR_0MP])[kts ])+
+ 3.f*(
+ (dist.f[DIR_PPP])[ktne]+ (dist.f[DIR_MMP])[ktsw]+
+ (dist.f[DIR_PMP])[ktse]+ (dist.f[DIR_MPP])[ktnw]+
+ (dist.f[DIR_PPM])[kbne]+ (dist.f[DIR_MMM])[kbsw]+
+ (dist.f[DIR_PMM])[kbse]+ (dist.f[DIR_MPM])[kbnw])-
+ rhoD[nodeIndex]-(vxD[nodeIndex] * vxD[nodeIndex] + vyD[nodeIndex] * vyD[nodeIndex] + vzD[nodeIndex] * vzD[nodeIndex]) * (c1o1+c0o1*rhoD[nodeIndex])) * c1o2+rhoD[nodeIndex]; // times zero for incompressible case
+ //achtung op hart gesetzt Annahme op = 1 ; ^^^^(1.0/op-0.5)=0.5
+ }
+ }
+}
+////////////////////////////////////////////////////////////////////////////////
+
+
+
+
+
+
+
+
+
+
+
+
+
- const unsigned nx = blockDim.x + 2 * STARTOFFX;
- const unsigned ny = gridDim.x + 2 * STARTOFFY;
- const unsigned int k = nx*(ny*z + y) + x; // Zugriff auf arrays im device
- if(k >= size_Mat)
- return;
- if(!vf::gpu::isValidFluidNode(geoD[k]))
- return;
- rhoD[k] = c0o1;
- vxD[k] = c0o1;
- vyD[k] = c0o1;
- vzD[k] = c0o1;
- vf::gpu::DistributionWrapper distr_wrapper(distributions, size_Mat, isEvenTimestep, k, neighborX, neighborY, neighborZ);
- const auto& distribution = distr_wrapper.distribution;
- rhoD[k] = vf::lbm::getDensity(distribution.f);
- vxD[k] = vf::lbm::getIncompressibleVelocityX1(distribution.f);
- vyD[k] = vf::lbm::getIncompressibleVelocityX2(distribution.f);
- vzD[k] = vf::lbm::getIncompressibleVelocityX3(distribution.f);
-}
-////////////////////////////////////////////////////////////////////////////////
-__global__ void LBCalcMacSP27( real* vxD,
- real* vyD,
- real* vzD,
- real* rhoD,
- real* pressD,
- unsigned int* geoD,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- real* DD,
- bool isEvenTimestep)
-{
- Distributions27 D;
- if (isEvenTimestep==true)
- {
- D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
- }
- else
- {
- D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
- }
- ////////////////////////////////////////////////////////////////////////////////
- const unsigned x = threadIdx.x; // Globaler x-Index
- const unsigned y = blockIdx.x; // Globaler y-Index
- const unsigned z = blockIdx.y; // Globaler z-Index
- const unsigned nx = blockDim.x;
- const unsigned ny = gridDim.x;
- const unsigned k = nx*(ny*z + y) + x;
- //////////////////////////////////////////////////////////////////////////
- if(k<size_Mat)
- {
- //////////////////////////////////////////////////////////////////////////
- //index
- unsigned int kzero= k;
- unsigned int ke = k;
- unsigned int kw = neighborX[k];
- unsigned int kn = k;
- unsigned int ks = neighborY[k];
- unsigned int kt = k;
- unsigned int kb = neighborZ[k];
- unsigned int ksw = neighborY[kw];
- unsigned int kne = k;
- unsigned int kse = ks;
- unsigned int knw = kw;
- unsigned int kbw = neighborZ[kw];
- unsigned int kte = k;
- unsigned int kbe = kb;
- unsigned int ktw = kw;
- unsigned int kbs = neighborZ[ks];
- unsigned int ktn = k;
- unsigned int kbn = kb;
- unsigned int kts = ks;
- unsigned int ktse = ks;
- unsigned int kbnw = kbw;
- unsigned int ktnw = kw;
- unsigned int kbse = kbs;
- unsigned int ktsw = ksw;
- unsigned int kbne = kb;
- unsigned int ktne = k;
- unsigned int kbsw = neighborZ[ksw];
- //////////////////////////////////////////////////////////////////////////
- pressD[k] = c0o1;
- rhoD[k] = c0o1;
- vxD[k] = c0o1;
- vyD[k] = c0o1;
- vzD[k] = c0o1;
-
- if(geoD[k] == GEO_FLUID)
- {
- rhoD[k] = (D.f[DIR_P00 ])[ke ]+ (D.f[DIR_M00 ])[kw ]+
- (D.f[DIR_0P0 ])[kn ]+ (D.f[DIR_0M0 ])[ks ]+
- (D.f[DIR_00P ])[kt ]+ (D.f[DIR_00M ])[kb ]+
- (D.f[DIR_PP0 ])[kne ]+ (D.f[DIR_MM0 ])[ksw ]+
- (D.f[DIR_PM0 ])[kse ]+ (D.f[DIR_MP0 ])[knw ]+
- (D.f[DIR_P0P ])[kte ]+ (D.f[DIR_M0M ])[kbw ]+
- (D.f[DIR_P0M ])[kbe ]+ (D.f[DIR_M0P ])[ktw ]+
- (D.f[DIR_0PP ])[ktn ]+ (D.f[DIR_0MM ])[kbs ]+
- (D.f[DIR_0PM ])[kbn ]+ (D.f[DIR_0MP ])[kts ]+
- (D.f[DIR_000])[kzero]+
- (D.f[DIR_PPP ])[ktne]+ (D.f[DIR_MMP ])[ktsw]+
- (D.f[DIR_PMP ])[ktse]+ (D.f[DIR_MPP ])[ktnw]+
- (D.f[DIR_PPM ])[kbne]+ (D.f[DIR_MMM ])[kbsw]+
- (D.f[DIR_PMM ])[kbse]+ (D.f[DIR_MPM ])[kbnw];
-
- vxD[k] = (D.f[DIR_P00 ])[ke ]- (D.f[DIR_M00 ])[kw ]+
- (D.f[DIR_PP0 ])[kne ]- (D.f[DIR_MM0 ])[ksw ]+
- (D.f[DIR_PM0 ])[kse ]- (D.f[DIR_MP0 ])[knw ]+
- (D.f[DIR_P0P ])[kte ]- (D.f[DIR_M0M ])[kbw ]+
- (D.f[DIR_P0M ])[kbe ]- (D.f[DIR_M0P ])[ktw ]+
- (D.f[DIR_PPP ])[ktne]- (D.f[DIR_MMP ])[ktsw]+
- (D.f[DIR_PMP ])[ktse]- (D.f[DIR_MPP ])[ktnw]+
- (D.f[DIR_PPM ])[kbne]- (D.f[DIR_MMM ])[kbsw]+
- (D.f[DIR_PMM ])[kbse]- (D.f[DIR_MPM ])[kbnw];
-
- vyD[k] = (D.f[DIR_0P0 ])[kn ]- (D.f[DIR_0M0 ])[ks ]+
- (D.f[DIR_PP0 ])[kne ]- (D.f[DIR_MM0 ])[ksw ]-
- (D.f[DIR_PM0 ])[kse ]+ (D.f[DIR_MP0 ])[knw ]+
- (D.f[DIR_0PP ])[ktn ]- (D.f[DIR_0MM ])[kbs ]+
- (D.f[DIR_0PM ])[kbn ]- (D.f[DIR_0MP ])[kts ]+
- (D.f[DIR_PPP ])[ktne]- (D.f[DIR_MMP ])[ktsw]-
- (D.f[DIR_PMP ])[ktse]+ (D.f[DIR_MPP ])[ktnw]+
- (D.f[DIR_PPM ])[kbne]- (D.f[DIR_MMM ])[kbsw]-
- (D.f[DIR_PMM ])[kbse]+ (D.f[DIR_MPM ])[kbnw];
-
- vzD[k] = (D.f[DIR_00P ])[kt ]- (D.f[DIR_00M ])[kb ]+
- (D.f[DIR_P0P ])[kte ]- (D.f[DIR_M0M ])[kbw ]-
- (D.f[DIR_P0M ])[kbe ]+ (D.f[DIR_M0P ])[ktw ]+
- (D.f[DIR_0PP ])[ktn ]- (D.f[DIR_0MM ])[kbs ]-
- (D.f[DIR_0PM ])[kbn ]+ (D.f[DIR_0MP ])[kts ]+
- (D.f[DIR_PPP ])[ktne]+ (D.f[DIR_MMP ])[ktsw]+
- (D.f[DIR_PMP ])[ktse]+ (D.f[DIR_MPP ])[ktnw]-
- (D.f[DIR_PPM ])[kbne]- (D.f[DIR_MMM ])[kbsw]-
- (D.f[DIR_PMM ])[kbse]- (D.f[DIR_MPM ])[kbnw];
-
- pressD[k] = ((D.f[DIR_P00 ])[ke ]+ (D.f[DIR_M00 ])[kw ]+
- (D.f[DIR_0P0 ])[kn ]+ (D.f[DIR_0M0 ])[ks ]+
- (D.f[DIR_00P ])[kt ]+ (D.f[DIR_00M ])[kb ]+
- 2.f*(
- (D.f[DIR_PP0 ])[kne ]+ (D.f[DIR_MM0 ])[ksw ]+
- (D.f[DIR_PM0 ])[kse ]+ (D.f[DIR_MP0 ])[knw ]+
- (D.f[DIR_P0P ])[kte ]+ (D.f[DIR_M0M ])[kbw ]+
- (D.f[DIR_P0M ])[kbe ]+ (D.f[DIR_M0P ])[ktw ]+
- (D.f[DIR_0PP ])[ktn ]+ (D.f[DIR_0MM ])[kbs ]+
- (D.f[DIR_0PM ])[kbn ]+ (D.f[DIR_0MP ])[kts ])+
- 3.f*(
- (D.f[DIR_PPP ])[ktne]+ (D.f[DIR_MMP ])[ktsw]+
- (D.f[DIR_PMP ])[ktse]+ (D.f[DIR_MPP ])[ktnw]+
- (D.f[DIR_PPM ])[kbne]+ (D.f[DIR_MMM ])[kbsw]+
- (D.f[DIR_PMM ])[kbse]+ (D.f[DIR_MPM ])[kbnw])-
- rhoD[k]-(vxD[k] * vxD[k] + vyD[k] * vyD[k] + vzD[k] * vzD[k]) * (c1o1+c0o1*rhoD[k])) * c1o2+rhoD[k]; // times zero for incompressible case
- //achtung op hart gesetzt Annahme op = 1 ; ^^^^(1.0/op-0.5)=0.5
-
- }
- }
-}
////////////////////////////////////////////////////////////////////////////////
__global__ void LBCalcMacCompSP27(
- real *vxD,
- real *vyD,
- real *vzD,
- real *rhoD,
- real *pressD,
- unsigned int *geoD,
- unsigned int *neighborX,
- unsigned int *neighborY,
- unsigned int *neighborZ,
- unsigned int size_Mat,
- real *distributions,
- bool isEvenTimestep)
+ real *vxD,
+ real *vyD,
+ real *vzD,
+ real *rhoD,
+ real *pressD,
+ unsigned int *geoD,
+ unsigned int *neighborX,
+ unsigned int *neighborY,
+ unsigned int *neighborZ,
+ unsigned long long numberOfLBnodes,
+ real *distributions,
+ bool isEvenTimestep)
{
- const unsigned k = vf::gpu::getNodeIndex();
+ ////////////////////////////////////////////////////////////////////////////////
+ //! - Get node index coordinates from threadIdx, blockIdx, blockDim and gridDim.
+ //!
+ const unsigned nodeIndex = getNodeIndex();
- if(k >= size_Mat)
+ if(nodeIndex >= numberOfLBnodes)
return;
- pressD[k] = c0o1;
- rhoD[k] = c0o1;
- vxD[k] = c0o1;
- vyD[k] = c0o1;
- vzD[k] = c0o1;
+ pressD[nodeIndex] = c0o1;
+ rhoD[nodeIndex] = c0o1;
+ vxD[nodeIndex] = c0o1;
+ vyD[nodeIndex] = c0o1;
+ vzD[nodeIndex] = c0o1;
- if (!vf::gpu::isValidFluidNode(geoD[k]))
+ if (!isValidFluidNode(geoD[nodeIndex]))
return;
- vf::gpu::DistributionWrapper distr_wrapper(distributions, size_Mat, isEvenTimestep, k, neighborX, neighborY,
- neighborZ);
+ DistributionWrapper distr_wrapper(distributions, numberOfLBnodes, isEvenTimestep, nodeIndex, neighborX, neighborY, neighborZ);
const auto &distribution = distr_wrapper.distribution;
- rhoD[k] = vf::lbm::getDensity(distribution.f);
- vxD[k] = vf::lbm::getCompressibleVelocityX1(distribution.f, rhoD[k]);
- vyD[k] = vf::lbm::getCompressibleVelocityX2(distribution.f, rhoD[k]);
- vzD[k] = vf::lbm::getCompressibleVelocityX3(distribution.f, rhoD[k]);
- pressD[k] = vf::lbm::getPressure(distribution.f, rhoD[k], vxD[k], vyD[k], vzD[k]);
+ rhoD[nodeIndex] = vf::lbm::getDensity(distribution.f);
+ vxD[nodeIndex] = vf::lbm::getCompressibleVelocityX1(distribution.f, rhoD[nodeIndex]);
+ vyD[nodeIndex] = vf::lbm::getCompressibleVelocityX2(distribution.f, rhoD[nodeIndex]);
+ vzD[nodeIndex] = vf::lbm::getCompressibleVelocityX3(distribution.f, rhoD[nodeIndex]);
+ pressD[nodeIndex] = vf::lbm::getPressure(distribution.f, rhoD[nodeIndex], vxD[nodeIndex], vyD[nodeIndex], vzD[nodeIndex]);
}
@@ -339,206 +343,155 @@ __global__ void LBCalcMacCompSP27(
////////////////////////////////////////////////////////////////////////////////
-__global__ void LBCalcMedSP27( real* vxD,
- real* vyD,
- real* vzD,
- real* rhoD,
- real* pressD,
- unsigned int* geoD,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- real* DD,
- bool isEvenTimestep)
+__global__ void LBCalcMedSP27(
+ real* vxD,
+ real* vyD,
+ real* vzD,
+ real* rhoD,
+ real* pressD,
+ unsigned int* geoD,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ real* distributions,
+ bool isEvenTimestep)
{
- Distributions27 D;
- if (isEvenTimestep==true)
- {
- D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
- }
- else
- {
- D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
- }
- ////////////////////////////////////////////////////////////////////////////////
- const unsigned x = threadIdx.x; // Globaler x-Index
- const unsigned y = blockIdx.x; // Globaler y-Index
- const unsigned z = blockIdx.y; // Globaler z-Index
-
- const unsigned nx = blockDim.x;
- const unsigned ny = gridDim.x;
-
- const unsigned k = nx*(ny*z + y) + x;
- //////////////////////////////////////////////////////////////////////////
-
- if(k<size_Mat)
- {
- //////////////////////////////////////////////////////////////////////////
- //index
- unsigned int kzero= k;
- unsigned int ke = k;
- unsigned int kw = neighborX[k];
- unsigned int kn = k;
- unsigned int ks = neighborY[k];
- unsigned int kt = k;
- unsigned int kb = neighborZ[k];
- unsigned int ksw = neighborY[kw];
- unsigned int kne = k;
- unsigned int kse = ks;
- unsigned int knw = kw;
- unsigned int kbw = neighborZ[kw];
- unsigned int kte = k;
- unsigned int kbe = kb;
- unsigned int ktw = kw;
- unsigned int kbs = neighborZ[ks];
- unsigned int ktn = k;
- unsigned int kbn = kb;
- unsigned int kts = ks;
- unsigned int ktse = ks;
- unsigned int kbnw = kbw;
- unsigned int ktnw = kw;
- unsigned int kbse = kbs;
- unsigned int ktsw = ksw;
- unsigned int kbne = kb;
- unsigned int ktne = k;
- unsigned int kbsw = neighborZ[ksw];
- //////////////////////////////////////////////////////////////////////////
- real PRESS = pressD[k];
- real RHO = rhoD[k];
- real VX = vxD[k];
- real VY = vyD[k];
- real VZ = vzD[k];
- //////////////////////////////////////////////////////////////////////////
- pressD[k] = c0o1;
- rhoD[k] = c0o1;
- vxD[k] = c0o1;
- vyD[k] = c0o1;
- vzD[k] = c0o1;
-
- if(geoD[k] == GEO_FLUID)
- {
- rhoD[k] = (D.f[DIR_P00 ])[ke ]+ (D.f[DIR_M00 ])[kw ]+
- (D.f[DIR_0P0 ])[kn ]+ (D.f[DIR_0M0 ])[ks ]+
- (D.f[DIR_00P ])[kt ]+ (D.f[DIR_00M ])[kb ]+
- (D.f[DIR_PP0 ])[kne ]+ (D.f[DIR_MM0 ])[ksw ]+
- (D.f[DIR_PM0 ])[kse ]+ (D.f[DIR_MP0 ])[knw ]+
- (D.f[DIR_P0P ])[kte ]+ (D.f[DIR_M0M ])[kbw ]+
- (D.f[DIR_P0M ])[kbe ]+ (D.f[DIR_M0P ])[ktw ]+
- (D.f[DIR_0PP ])[ktn ]+ (D.f[DIR_0MM ])[kbs ]+
- (D.f[DIR_0PM ])[kbn ]+ (D.f[DIR_0MP ])[kts ]+
- (D.f[DIR_000])[kzero]+
- (D.f[DIR_PPP ])[ktne]+ (D.f[DIR_MMP ])[ktsw]+
- (D.f[DIR_PMP ])[ktse]+ (D.f[DIR_MPP ])[ktnw]+
- (D.f[DIR_PPM ])[kbne]+ (D.f[DIR_MMM ])[kbsw]+
- (D.f[DIR_PMM ])[kbse]+ (D.f[DIR_MPM ])[kbnw]+
- RHO;
-
- vxD[k] = (D.f[DIR_P00 ])[ke ]- (D.f[DIR_M00 ])[kw ]+
- (D.f[DIR_PP0 ])[kne ]- (D.f[DIR_MM0 ])[ksw ]+
- (D.f[DIR_PM0 ])[kse ]- (D.f[DIR_MP0 ])[knw ]+
- (D.f[DIR_P0P ])[kte ]- (D.f[DIR_M0M ])[kbw ]+
- (D.f[DIR_P0M ])[kbe ]- (D.f[DIR_M0P ])[ktw ]+
- (D.f[DIR_PPP ])[ktne]- (D.f[DIR_MMP ])[ktsw]+
- (D.f[DIR_PMP ])[ktse]- (D.f[DIR_MPP ])[ktnw]+
- (D.f[DIR_PPM ])[kbne]- (D.f[DIR_MMM ])[kbsw]+
- (D.f[DIR_PMM ])[kbse]- (D.f[DIR_MPM ])[kbnw]+
- VX;
-
- vyD[k] = (D.f[DIR_0P0 ])[kn ]- (D.f[DIR_0M0 ])[ks ]+
- (D.f[DIR_PP0 ])[kne ]- (D.f[DIR_MM0 ])[ksw ]-
- (D.f[DIR_PM0 ])[kse ]+ (D.f[DIR_MP0 ])[knw ]+
- (D.f[DIR_0PP ])[ktn ]- (D.f[DIR_0MM ])[kbs ]+
- (D.f[DIR_0PM ])[kbn ]- (D.f[DIR_0MP ])[kts ]+
- (D.f[DIR_PPP ])[ktne]- (D.f[DIR_MMP ])[ktsw]-
- (D.f[DIR_PMP ])[ktse]+ (D.f[DIR_MPP ])[ktnw]+
- (D.f[DIR_PPM ])[kbne]- (D.f[DIR_MMM ])[kbsw]-
- (D.f[DIR_PMM ])[kbse]+ (D.f[DIR_MPM ])[kbnw]+
- VY;
-
- vzD[k] = (D.f[DIR_00P ])[kt ]- (D.f[DIR_00M ])[kb ]+
- (D.f[DIR_P0P ])[kte ]- (D.f[DIR_M0M ])[kbw ]-
- (D.f[DIR_P0M ])[kbe ]+ (D.f[DIR_M0P ])[ktw ]+
- (D.f[DIR_0PP ])[ktn ]- (D.f[DIR_0MM ])[kbs ]-
- (D.f[DIR_0PM ])[kbn ]+ (D.f[DIR_0MP ])[kts ]+
- (D.f[DIR_PPP ])[ktne]+ (D.f[DIR_MMP ])[ktsw]+
- (D.f[DIR_PMP ])[ktse]+ (D.f[DIR_MPP ])[ktnw]-
- (D.f[DIR_PPM ])[kbne]- (D.f[DIR_MMM ])[kbsw]-
- (D.f[DIR_PMM ])[kbse]- (D.f[DIR_MPM ])[kbnw]+
- VZ;
-
- pressD[k] = ((D.f[DIR_P00 ])[ke ]+ (D.f[DIR_M00 ])[kw ]+
- (D.f[DIR_0P0 ])[kn ]+ (D.f[DIR_0M0 ])[ks ]+
- (D.f[DIR_00P ])[kt ]+ (D.f[DIR_00M ])[kb ]+
- c2o1*(
- (D.f[DIR_PP0 ])[kne ]+ (D.f[DIR_MM0 ])[ksw ]+
- (D.f[DIR_PM0 ])[kse ]+ (D.f[DIR_MP0 ])[knw ]+
- (D.f[DIR_P0P ])[kte ]+ (D.f[DIR_M0M ])[kbw ]+
- (D.f[DIR_P0M ])[kbe ]+ (D.f[DIR_M0P ])[ktw ]+
- (D.f[DIR_0PP ])[ktn ]+ (D.f[DIR_0MM ])[kbs ]+
- (D.f[DIR_0PM ])[kbn ]+ (D.f[DIR_0MP ])[kts ])+
- c3o1*(
- (D.f[DIR_PPP ])[ktne]+ (D.f[DIR_MMP ])[ktsw]+
- (D.f[DIR_PMP ])[ktse]+ (D.f[DIR_MPP ])[ktnw]+
- (D.f[DIR_PPM ])[kbne]+ (D.f[DIR_MMM ])[kbsw]+
- (D.f[DIR_PMM ])[kbse]+ (D.f[DIR_MPM ])[kbnw])-
- rhoD[k]-(vxD[k] * vxD[k] + vyD[k] * vyD[k] + vzD[k] * vzD[k]) * (c1o1+rhoD[k])) * c1o2+rhoD[k]+
- PRESS;
- //achtung op hart gesetzt Annahme op = 1 ; ^^^^(1.0/op-0.5)=0.5
- }
- }
+ ////////////////////////////////////////////////////////////////////////////////
+ //! - Get node index coordinates from threadIdx, blockIdx, blockDim and gridDim.
+ //!
+ const unsigned nodeIndex = getNodeIndex();
+
+ //////////////////////////////////////////////////////////////////////////
+ if( nodeIndex < numberOfLBnodes )
+ {
+ //////////////////////////////////////////////////////////////////////////
+ //! - Read distributions: style of reading and writing the distributions from/to stored arrays dependent on
+ //! timestep is based on the esoteric twist algorithm \ref <a
+ //! href="https://doi.org/10.3390/computation5020019"><b>[ M. Geier et al. (2017),
+ //! DOI:10.3390/computation5020019 ]</b></a>
+ //!
+ Distributions27 dist;
+ getPointersToDistributions(dist, distributions, numberOfLBnodes, isEvenTimestep);
+
+ //////////////////////////////////////////////////////////////////////////
+ //index
+ unsigned int kzero= nodeIndex;
+ unsigned int ke = nodeIndex;
+ unsigned int kw = neighborX[nodeIndex];
+ unsigned int kn = nodeIndex;
+ unsigned int ks = neighborY[nodeIndex];
+ unsigned int kt = nodeIndex;
+ unsigned int kb = neighborZ[nodeIndex];
+ unsigned int ksw = neighborY[kw];
+ unsigned int kne = nodeIndex;
+ unsigned int kse = ks;
+ unsigned int knw = kw;
+ unsigned int kbw = neighborZ[kw];
+ unsigned int kte = nodeIndex;
+ unsigned int kbe = kb;
+ unsigned int ktw = kw;
+ unsigned int kbs = neighborZ[ks];
+ unsigned int ktn = nodeIndex;
+ unsigned int kbn = kb;
+ unsigned int kts = ks;
+ unsigned int ktse = ks;
+ unsigned int kbnw = kbw;
+ unsigned int ktnw = kw;
+ unsigned int kbse = kbs;
+ unsigned int ktsw = ksw;
+ unsigned int kbne = kb;
+ unsigned int ktne = nodeIndex;
+ unsigned int kbsw = neighborZ[ksw];
+ //////////////////////////////////////////////////////////////////////////
+ real PRESS = pressD[nodeIndex];
+ real RHO = rhoD[nodeIndex];
+ real VX = vxD[nodeIndex];
+ real VY = vyD[nodeIndex];
+ real VZ = vzD[nodeIndex];
+ //////////////////////////////////////////////////////////////////////////
+ pressD[nodeIndex] = c0o1;
+ rhoD[nodeIndex] = c0o1;
+ vxD[nodeIndex] = c0o1;
+ vyD[nodeIndex] = c0o1;
+ vzD[nodeIndex] = c0o1;
+
+ if(geoD[nodeIndex] == GEO_FLUID)
+ {
+ rhoD[nodeIndex] =
+ (dist.f[DIR_P00])[ke ]+ (dist.f[DIR_M00])[kw ]+
+ (dist.f[DIR_0P0])[kn ]+ (dist.f[DIR_0M0])[ks ]+
+ (dist.f[DIR_00P])[kt ]+ (dist.f[DIR_00M])[kb ]+
+ (dist.f[DIR_PP0])[kne ]+ (dist.f[DIR_MM0])[ksw ]+
+ (dist.f[DIR_PM0])[kse ]+ (dist.f[DIR_MP0])[knw ]+
+ (dist.f[DIR_P0P])[kte ]+ (dist.f[DIR_M0M])[kbw ]+
+ (dist.f[DIR_P0M])[kbe ]+ (dist.f[DIR_M0P])[ktw ]+
+ (dist.f[DIR_0PP])[ktn ]+ (dist.f[DIR_0MM])[kbs ]+
+ (dist.f[DIR_0PM])[kbn ]+ (dist.f[DIR_0MP])[kts ]+
+ (dist.f[DIR_000])[kzero]+
+ (dist.f[DIR_PPP])[ktne]+ (dist.f[DIR_MMP])[ktsw]+
+ (dist.f[DIR_PMP])[ktse]+ (dist.f[DIR_MPP])[ktnw]+
+ (dist.f[DIR_PPM])[kbne]+ (dist.f[DIR_MMM])[kbsw]+
+ (dist.f[DIR_PMM])[kbse]+ (dist.f[DIR_MPM])[kbnw]+
+ RHO;
+
+ vxD[nodeIndex] =
+ (dist.f[DIR_P00])[ke ]- (dist.f[DIR_M00])[kw ]+
+ (dist.f[DIR_PP0])[kne ]- (dist.f[DIR_MM0])[ksw ]+
+ (dist.f[DIR_PM0])[kse ]- (dist.f[DIR_MP0])[knw ]+
+ (dist.f[DIR_P0P])[kte ]- (dist.f[DIR_M0M])[kbw ]+
+ (dist.f[DIR_P0M])[kbe ]- (dist.f[DIR_M0P])[ktw ]+
+ (dist.f[DIR_PPP])[ktne]- (dist.f[DIR_MMP])[ktsw]+
+ (dist.f[DIR_PMP])[ktse]- (dist.f[DIR_MPP])[ktnw]+
+ (dist.f[DIR_PPM])[kbne]- (dist.f[DIR_MMM])[kbsw]+
+ (dist.f[DIR_PMM])[kbse]- (dist.f[DIR_MPM])[kbnw]+
+ VX;
+
+ vyD[nodeIndex] =
+ (dist.f[DIR_0P0])[kn ]- (dist.f[DIR_0M0])[ks ]+
+ (dist.f[DIR_PP0])[kne ]- (dist.f[DIR_MM0])[ksw ]-
+ (dist.f[DIR_PM0])[kse ]+ (dist.f[DIR_MP0])[knw ]+
+ (dist.f[DIR_0PP])[ktn ]- (dist.f[DIR_0MM])[kbs ]+
+ (dist.f[DIR_0PM])[kbn ]- (dist.f[DIR_0MP])[kts ]+
+ (dist.f[DIR_PPP])[ktne]- (dist.f[DIR_MMP])[ktsw]-
+ (dist.f[DIR_PMP])[ktse]+ (dist.f[DIR_MPP])[ktnw]+
+ (dist.f[DIR_PPM])[kbne]- (dist.f[DIR_MMM])[kbsw]-
+ (dist.f[DIR_PMM])[kbse]+ (dist.f[DIR_MPM])[kbnw]+
+ VY;
+
+ vzD[nodeIndex] =
+ (dist.f[DIR_00P])[kt ]- (dist.f[DIR_00M])[kb ]+
+ (dist.f[DIR_P0P])[kte ]- (dist.f[DIR_M0M])[kbw ]-
+ (dist.f[DIR_P0M])[kbe ]+ (dist.f[DIR_M0P])[ktw ]+
+ (dist.f[DIR_0PP])[ktn ]- (dist.f[DIR_0MM])[kbs ]-
+ (dist.f[DIR_0PM])[kbn ]+ (dist.f[DIR_0MP])[kts ]+
+ (dist.f[DIR_PPP])[ktne]+ (dist.f[DIR_MMP])[ktsw]+
+ (dist.f[DIR_PMP])[ktse]+ (dist.f[DIR_MPP])[ktnw]-
+ (dist.f[DIR_PPM])[kbne]- (dist.f[DIR_MMM])[kbsw]-
+ (dist.f[DIR_PMM])[kbse]- (dist.f[DIR_MPM])[kbnw]+
+ VZ;
+
+ pressD[nodeIndex] =
+ ((dist.f[DIR_P00])[ke ]+ (dist.f[DIR_M00])[kw ]+
+ (dist.f[DIR_0P0])[kn ]+ (dist.f[DIR_0M0])[ks ]+
+ (dist.f[DIR_00P])[kt ]+ (dist.f[DIR_00M])[kb ]+
+ c2o1*(
+ (dist.f[DIR_PP0])[kne ]+ (dist.f[DIR_MM0])[ksw ]+
+ (dist.f[DIR_PM0])[kse ]+ (dist.f[DIR_MP0])[knw ]+
+ (dist.f[DIR_P0P])[kte ]+ (dist.f[DIR_M0M])[kbw ]+
+ (dist.f[DIR_P0M])[kbe ]+ (dist.f[DIR_M0P])[ktw ]+
+ (dist.f[DIR_0PP])[ktn ]+ (dist.f[DIR_0MM])[kbs ]+
+ (dist.f[DIR_0PM])[kbn ]+ (dist.f[DIR_0MP])[kts ])+
+ c3o1*(
+ (dist.f[DIR_PPP])[ktne]+ (dist.f[DIR_MMP])[ktsw]+
+ (dist.f[DIR_PMP])[ktse]+ (dist.f[DIR_MPP])[ktnw]+
+ (dist.f[DIR_PPM])[kbne]+ (dist.f[DIR_MMM])[kbsw]+
+ (dist.f[DIR_PMM])[kbse]+ (dist.f[DIR_MPM])[kbnw])-
+ rhoD[nodeIndex]-(vxD[nodeIndex] * vxD[nodeIndex] + vyD[nodeIndex] * vyD[nodeIndex] + vzD[nodeIndex] * vzD[nodeIndex]) * (c1o1+rhoD[nodeIndex])) * c1o2+rhoD[nodeIndex]+
+ PRESS;
+ //achtung op hart gesetzt Annahme op = 1 ; ^^^^(1.0/op-0.5)=0.5
+ }
+ }
}
////////////////////////////////////////////////////////////////////////////////
@@ -563,259 +516,152 @@ __global__ void LBCalcMedSP27( real* vxD,
////////////////////////////////////////////////////////////////////////////////
-__global__ void LBCalcMedCompSP27( real* vxD,
- real* vyD,
- real* vzD,
- real* rhoD,
- real* pressD,
- unsigned int* geoD,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- real* DD,
- bool isEvenTimestep)
+__global__ void LBCalcMedCompSP27(
+ real* vxD,
+ real* vyD,
+ real* vzD,
+ real* rhoD,
+ real* pressD,
+ unsigned int* geoD,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ real* distributions,
+ bool isEvenTimestep)
{
- Distributions27 D;
- if (isEvenTimestep==true)
- {
- D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
- }
- else
- {
- D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
- }
- ////////////////////////////////////////////////////////////////////////////////
- const unsigned x = threadIdx.x; // Globaler x-Index
- const unsigned y = blockIdx.x; // Globaler y-Index
- const unsigned z = blockIdx.y; // Globaler z-Index
-
- const unsigned nx = blockDim.x;
- const unsigned ny = gridDim.x;
-
- const unsigned k = nx*(ny*z + y) + x;
- //////////////////////////////////////////////////////////////////////////
-
- if(k<size_Mat)
- {
- //////////////////////////////////////////////////////////////////////////
- //index
- //unsigned int kzero= k;
- unsigned int ke = k;
- unsigned int kw = neighborX[k];
- unsigned int kn = k;
- unsigned int ks = neighborY[k];
- unsigned int kt = k;
- unsigned int kb = neighborZ[k];
- unsigned int ksw = neighborY[kw];
- unsigned int kne = k;
- unsigned int kse = ks;
- unsigned int knw = kw;
- unsigned int kbw = neighborZ[kw];
- unsigned int kte = k;
- unsigned int kbe = kb;
- unsigned int ktw = kw;
- unsigned int kbs = neighborZ[ks];
- unsigned int ktn = k;
- unsigned int kbn = kb;
- unsigned int kts = ks;
- unsigned int ktse = ks;
- unsigned int kbnw = kbw;
- unsigned int ktnw = kw;
- unsigned int kbse = kbs;
- unsigned int ktsw = ksw;
- unsigned int kbne = kb;
- unsigned int ktne = k;
- unsigned int kbsw = neighborZ[ksw];
- //////////////////////////////////////////////////////////////////////////
- real PRESS = pressD[k];
- real RHO = rhoD[k];
- real VX = vxD[k];
- real VY = vyD[k];
- real VZ = vzD[k];
- //////////////////////////////////////////////////////////////////////////
- pressD[k] = c0o1;
- rhoD[k] = c0o1;
- vxD[k] = c0o1;
- vyD[k] = c0o1;
- vzD[k] = c0o1;
-
- if(geoD[k] == GEO_FLUID)
- {
- real mfcbb = (D.f[DIR_P00])[k];//[ke ];
- real mfabb = (D.f[DIR_M00])[kw];//[kw ];
- real mfbcb = (D.f[DIR_0P0])[k];//[kn ];
- real mfbab = (D.f[DIR_0M0])[ks];//[ks ];
- real mfbbc = (D.f[DIR_00P])[k];//[kt ];
- real mfbba = (D.f[DIR_00M])[kb];//[kb ];
- real mfccb = (D.f[DIR_PP0])[k];//[kne ];
- real mfaab = (D.f[DIR_MM0])[ksw];//[ksw ];
- real mfcab = (D.f[DIR_PM0])[ks];//[kse ];
- real mfacb = (D.f[DIR_MP0])[kw];//[knw ];
- real mfcbc = (D.f[DIR_P0P])[k];//[kte ];
- real mfaba = (D.f[DIR_M0M])[kbw];//[kbw ];
- real mfcba = (D.f[DIR_P0M])[kb];//[kbe ];
- real mfabc = (D.f[DIR_M0P])[kw];//[ktw ];
- real mfbcc = (D.f[DIR_0PP])[k];//[ktn ];
- real mfbaa = (D.f[DIR_0MM])[kbs];//[kbs ];
- real mfbca = (D.f[DIR_0PM])[kb];//[kbn ];
- real mfbac = (D.f[DIR_0MP])[ks];//[kts ];
- real mfbbb = (D.f[DIR_000])[k];//[kzero];
- real mfccc = (D.f[DIR_PPP])[k];//[ktne ];
- real mfaac = (D.f[DIR_MMP])[ksw];//[ktsw ];
- real mfcac = (D.f[DIR_PMP])[ks];//[ktse ];
- real mfacc = (D.f[DIR_MPP])[kw];//[ktnw ];
- real mfcca = (D.f[DIR_PPM])[kb];//[kbne ];
- real mfaaa = (D.f[DIR_MMM])[kbsw];//[kbsw ];
- real mfcaa = (D.f[DIR_PMM])[kbs];//[kbse ];
- real mfaca = (D.f[DIR_MPM])[kbw];//[kbnw ];
- ////////////////////////////////////////////////////////////////////////////////////
- real drho =
- ((((mfccc + mfaaa) + (mfaca + mfcac)) + ((mfacc + mfcaa) + (mfaac + mfcca))) +
- (((mfbac + mfbca) + (mfbaa + mfbcc)) + ((mfabc + mfcba) + (mfaba + mfcbc)) + ((mfacb + mfcab) + (mfaab + mfccb))) +
- ((mfabb + mfcbb) + (mfbab + mfbcb) + (mfbba + mfbbc))) + mfbbb;
-
- real rho = c1o1 + drho;
-
- rhoD[k] = drho + RHO;
-
- vxD[k] =
- (((((mfccc - mfaaa) + (mfcac - mfaca)) + ((mfcaa - mfacc) + (mfcca - mfaac))) +
- (((mfcba - mfabc) + (mfcbc - mfaba)) + ((mfcab - mfacb) + (mfccb - mfaab))) +
- (mfcbb - mfabb)) / rho) + VX;
- vyD[k] =
- (((((mfccc - mfaaa) + (mfaca - mfcac)) + ((mfacc - mfcaa) + (mfcca - mfaac))) +
- (((mfbca - mfbac) + (mfbcc - mfbaa)) + ((mfacb - mfcab) + (mfccb - mfaab))) +
- (mfbcb - mfbab)) / rho) + VY;
- vzD[k] =
- (((((mfccc - mfaaa) + (mfcac - mfaca)) + ((mfacc - mfcaa) + (mfaac - mfcca))) +
- (((mfbac - mfbca) + (mfbcc - mfbaa)) + ((mfabc - mfcba) + (mfcbc - mfaba))) +
- (mfbbc - mfbba)) / rho) + VZ;
-
- //rhoD[k] =
- // (D.f[DIR_P00])[ke] + (D.f[DIR_M00])[kw] +
- // (D.f[DIR_0P0])[kn] + (D.f[DIR_0M0])[ks] +
- // (D.f[DIR_00P])[kt] + (D.f[DIR_00M])[kb] +
- // (D.f[DIR_PP0])[kne] + (D.f[DIR_MM0])[ksw] +
- // (D.f[DIR_PM0])[kse] + (D.f[DIR_MP0])[knw] +
- // (D.f[DIR_P0P])[kte] + (D.f[DIR_M0M])[kbw] +
- // (D.f[DIR_P0M])[kbe] + (D.f[DIR_M0P])[ktw] +
- // (D.f[DIR_0PP])[ktn] + (D.f[DIR_0MM])[kbs] +
- // (D.f[DIR_0PM])[kbn] + (D.f[DIR_0MP])[kts] +
- // (D.f[DIR_000])[kzero] +
- // (D.f[DIR_PPP])[ktne] + (D.f[DIR_MMP])[ktsw] +
- // (D.f[DIR_PMP])[ktse] + (D.f[DIR_MPP])[ktnw] +
- // (D.f[DIR_PPM])[kbne] + (D.f[DIR_MMM])[kbsw] +
- // (D.f[DIR_PMM])[kbse] + (D.f[DIR_MPM])[kbnw];// +RHO;
-
- // vxD[k] =
- //((D.f[DIR_P00 ])[ke ]- (D.f[DIR_M00 ])[kw ]+
- // (D.f[DIR_PP0 ])[kne ]- (D.f[DIR_MM0 ])[ksw ]+
- // (D.f[DIR_PM0 ])[kse ]- (D.f[DIR_MP0 ])[knw ]+
- // (D.f[DIR_P0P ])[kte ]- (D.f[DIR_M0M ])[kbw ]+
- // (D.f[DIR_P0M ])[kbe ]- (D.f[DIR_M0P ])[ktw ]+
- // (D.f[DIR_PPP ])[ktne]- (D.f[DIR_MMP ])[ktsw]+
- // (D.f[DIR_PMP ])[ktse]- (D.f[DIR_MPP ])[ktnw]+
- // (D.f[DIR_PPM ])[kbne]- (D.f[DIR_MMM ])[kbsw]+
- // (D.f[DIR_PMM ])[kbse]- (D.f[DIR_MPM ])[kbnw]) / (one + rhoD[k])+
- // VX;
-
- // vyD[k] =
- //((D.f[DIR_0P0 ])[kn ]- (D.f[DIR_0M0 ])[ks ]+
- // (D.f[DIR_PP0 ])[kne ]- (D.f[DIR_MM0 ])[ksw ]-
- // (D.f[DIR_PM0 ])[kse ]+ (D.f[DIR_MP0 ])[knw ]+
- // (D.f[DIR_0PP ])[ktn ]- (D.f[DIR_0MM ])[kbs ]+
- // (D.f[DIR_0PM ])[kbn ]- (D.f[DIR_0MP ])[kts ]+
- // (D.f[DIR_PPP ])[ktne]- (D.f[DIR_MMP ])[ktsw]-
- // (D.f[DIR_PMP ])[ktse]+ (D.f[DIR_MPP ])[ktnw]+
- // (D.f[DIR_PPM ])[kbne]- (D.f[DIR_MMM ])[kbsw]-
- // (D.f[DIR_PMM ])[kbse]+ (D.f[DIR_MPM ])[kbnw]) / (one + rhoD[k])+
- // VY;
-
- // vzD[k] =
- //((D.f[DIR_00P ])[kt ]- (D.f[DIR_00M ])[kb ]+
- // (D.f[DIR_P0P ])[kte ]- (D.f[DIR_M0M ])[kbw ]-
- // (D.f[DIR_P0M ])[kbe ]+ (D.f[DIR_M0P ])[ktw ]+
- // (D.f[DIR_0PP ])[ktn ]- (D.f[DIR_0MM ])[kbs ]-
- // (D.f[DIR_0PM ])[kbn ]+ (D.f[DIR_0MP ])[kts ]+
- // (D.f[DIR_PPP ])[ktne]+ (D.f[DIR_MMP ])[ktsw]+
- // (D.f[DIR_PMP ])[ktse]+ (D.f[DIR_MPP ])[ktnw]-
- // (D.f[DIR_PPM ])[kbne]- (D.f[DIR_MMM ])[kbsw]-
- // (D.f[DIR_PMM ])[kbse]- (D.f[DIR_MPM ])[kbnw]) / (one + rhoD[k])+
- // VZ;
-
- pressD[k] = ((D.f[DIR_P00 ])[ke ]+ (D.f[DIR_M00 ])[kw ]+
- (D.f[DIR_0P0 ])[kn ]+ (D.f[DIR_0M0 ])[ks ]+
- (D.f[DIR_00P ])[kt ]+ (D.f[DIR_00M ])[kb ]+
- c2o1*(
- (D.f[DIR_PP0 ])[kne ]+ (D.f[DIR_MM0 ])[ksw ]+
- (D.f[DIR_PM0 ])[kse ]+ (D.f[DIR_MP0 ])[knw ]+
- (D.f[DIR_P0P ])[kte ]+ (D.f[DIR_M0M ])[kbw ]+
- (D.f[DIR_P0M ])[kbe ]+ (D.f[DIR_M0P ])[ktw ]+
- (D.f[DIR_0PP ])[ktn ]+ (D.f[DIR_0MM ])[kbs ]+
- (D.f[DIR_0PM ])[kbn ]+ (D.f[DIR_0MP ])[kts ])+
- c3o1*(
- (D.f[DIR_PPP ])[ktne]+ (D.f[DIR_MMP ])[ktsw]+
- (D.f[DIR_PMP ])[ktse]+ (D.f[DIR_MPP ])[ktnw]+
- (D.f[DIR_PPM ])[kbne]+ (D.f[DIR_MMM ])[kbsw]+
- (D.f[DIR_PMM ])[kbse]+ (D.f[DIR_MPM ])[kbnw])-
- rhoD[k]-(vxD[k] * vxD[k] + vyD[k] * vyD[k] + vzD[k] * vzD[k]) * (c1o1+rhoD[k])) * c1o2+rhoD[k]+
- PRESS;
- //achtung op hart gesetzt Annahme op = 1 ; ^^^^(1.0/op-0.5)=0.5
- }
- }
+ ////////////////////////////////////////////////////////////////////////////////
+ //! - Get node index coordinates from threadIdx, blockIdx, blockDim and gridDim.
+ //!
+ const unsigned nodeIndex = getNodeIndex();
+
+ //////////////////////////////////////////////////////////////////////////
+ if( nodeIndex < numberOfLBnodes )
+ {
+ //////////////////////////////////////////////////////////////////////////
+ //! - Read distributions: style of reading and writing the distributions from/to stored arrays dependent on
+ //! timestep is based on the esoteric twist algorithm \ref <a
+ //! href="https://doi.org/10.3390/computation5020019"><b>[ M. Geier et al. (2017),
+ //! DOI:10.3390/computation5020019 ]</b></a>
+ //!
+ Distributions27 dist;
+ getPointersToDistributions(dist, distributions, numberOfLBnodes, isEvenTimestep);
+
+ //////////////////////////////////////////////////////////////////////////
+ //index
+ //unsigned int kzero= k;
+ unsigned int ke = nodeIndex;
+ unsigned int kw = neighborX[nodeIndex];
+ unsigned int kn = nodeIndex;
+ unsigned int ks = neighborY[nodeIndex];
+ unsigned int kt = nodeIndex;
+ unsigned int kb = neighborZ[nodeIndex];
+ unsigned int ksw = neighborY[kw];
+ unsigned int kne = nodeIndex;
+ unsigned int kse = ks;
+ unsigned int knw = kw;
+ unsigned int kbw = neighborZ[kw];
+ unsigned int kte = nodeIndex;
+ unsigned int kbe = kb;
+ unsigned int ktw = kw;
+ unsigned int kbs = neighborZ[ks];
+ unsigned int ktn = nodeIndex;
+ unsigned int kbn = kb;
+ unsigned int kts = ks;
+ unsigned int ktse = ks;
+ unsigned int kbnw = kbw;
+ unsigned int ktnw = kw;
+ unsigned int kbse = kbs;
+ unsigned int ktsw = ksw;
+ unsigned int kbne = kb;
+ unsigned int ktne = nodeIndex;
+ unsigned int kbsw = neighborZ[ksw];
+ //////////////////////////////////////////////////////////////////////////
+ real PRESS = pressD[nodeIndex];
+ real RHO = rhoD[nodeIndex];
+ real VX = vxD[nodeIndex];
+ real VY = vyD[nodeIndex];
+ real VZ = vzD[nodeIndex];
+ //////////////////////////////////////////////////////////////////////////
+ pressD[nodeIndex] = c0o1;
+ rhoD[nodeIndex] = c0o1;
+ vxD[nodeIndex] = c0o1;
+ vyD[nodeIndex] = c0o1;
+ vzD[nodeIndex] = c0o1;
+
+ if(geoD[nodeIndex] == GEO_FLUID)
+ {
+ real mfcbb = (dist.f[DIR_P00])[nodeIndex];//[ke ];
+ real mfabb = (dist.f[DIR_M00])[kw];//[kw ];
+ real mfbcb = (dist.f[DIR_0P0])[nodeIndex];//[kn ];
+ real mfbab = (dist.f[DIR_0M0])[ks];//[ks ];
+ real mfbbc = (dist.f[DIR_00P])[nodeIndex];//[kt ];
+ real mfbba = (dist.f[DIR_00M])[kb];//[kb ];
+ real mfccb = (dist.f[DIR_PP0])[nodeIndex];//[kne ];
+ real mfaab = (dist.f[DIR_MM0])[ksw];//[ksw ];
+ real mfcab = (dist.f[DIR_PM0])[ks];//[kse ];
+ real mfacb = (dist.f[DIR_MP0])[kw];//[knw ];
+ real mfcbc = (dist.f[DIR_P0P])[nodeIndex];//[kte ];
+ real mfaba = (dist.f[DIR_M0M])[kbw];//[kbw ];
+ real mfcba = (dist.f[DIR_P0M])[kb];//[kbe ];
+ real mfabc = (dist.f[DIR_M0P])[kw];//[ktw ];
+ real mfbcc = (dist.f[DIR_0PP])[nodeIndex];//[ktn ];
+ real mfbaa = (dist.f[DIR_0MM])[kbs];//[kbs ];
+ real mfbca = (dist.f[DIR_0PM])[kb];//[kbn ];
+ real mfbac = (dist.f[DIR_0MP])[ks];//[kts ];
+ real mfbbb = (dist.f[DIR_000])[nodeIndex];//[kzero];
+ real mfccc = (dist.f[DIR_PPP])[nodeIndex];//[ktne ];
+ real mfaac = (dist.f[DIR_MMP])[ksw];//[ktsw ];
+ real mfcac = (dist.f[DIR_PMP])[ks];//[ktse ];
+ real mfacc = (dist.f[DIR_MPP])[kw];//[ktnw ];
+ real mfcca = (dist.f[DIR_PPM])[kb];//[kbne ];
+ real mfaaa = (dist.f[DIR_MMM])[kbsw];//[kbsw ];
+ real mfcaa = (dist.f[DIR_PMM])[kbs];//[kbse ];
+ real mfaca = (dist.f[DIR_MPM])[kbw];//[kbnw ];
+ ////////////////////////////////////////////////////////////////////////////////////
+ real drho =
+ ((((mfccc + mfaaa) + (mfaca + mfcac)) + ((mfacc + mfcaa) + (mfaac + mfcca))) +
+ (((mfbac + mfbca) + (mfbaa + mfbcc)) + ((mfabc + mfcba) + (mfaba + mfcbc)) + ((mfacb + mfcab) + (mfaab + mfccb))) +
+ ((mfabb + mfcbb) + (mfbab + mfbcb) + (mfbba + mfbbc))) + mfbbb;
+
+ real rho = c1o1 + drho;
+
+ rhoD[nodeIndex] = drho + RHO;
+
+ vxD[nodeIndex] =
+ (((((mfccc - mfaaa) + (mfcac - mfaca)) + ((mfcaa - mfacc) + (mfcca - mfaac))) +
+ (((mfcba - mfabc) + (mfcbc - mfaba)) + ((mfcab - mfacb) + (mfccb - mfaab))) +
+ (mfcbb - mfabb)) / rho) + VX;
+ vyD[nodeIndex] =
+ (((((mfccc - mfaaa) + (mfaca - mfcac)) + ((mfacc - mfcaa) + (mfcca - mfaac))) +
+ (((mfbca - mfbac) + (mfbcc - mfbaa)) + ((mfacb - mfcab) + (mfccb - mfaab))) +
+ (mfbcb - mfbab)) / rho) + VY;
+ vzD[nodeIndex] =
+ (((((mfccc - mfaaa) + (mfcac - mfaca)) + ((mfacc - mfcaa) + (mfaac - mfcca))) +
+ (((mfbac - mfbca) + (mfbcc - mfbaa)) + ((mfabc - mfcba) + (mfcbc - mfaba))) +
+ (mfbbc - mfbba)) / rho) + VZ;
+
+ pressD[nodeIndex] =
+ ((dist.f[DIR_P00])[ke ]+ (dist.f[DIR_M00])[kw ]+
+ (dist.f[DIR_0P0])[kn ]+ (dist.f[DIR_0M0])[ks ]+
+ (dist.f[DIR_00P])[kt ]+ (dist.f[DIR_00M])[kb ]+
+ c2o1*(
+ (dist.f[DIR_PP0])[kne ]+ (dist.f[DIR_MM0])[ksw ]+
+ (dist.f[DIR_PM0])[kse ]+ (dist.f[DIR_MP0])[knw ]+
+ (dist.f[DIR_P0P])[kte ]+ (dist.f[DIR_M0M])[kbw ]+
+ (dist.f[DIR_P0M])[kbe ]+ (dist.f[DIR_M0P])[ktw ]+
+ (dist.f[DIR_0PP])[ktn ]+ (dist.f[DIR_0MM])[kbs ]+
+ (dist.f[DIR_0PM])[kbn ]+ (dist.f[DIR_0MP])[kts ])+
+ c3o1*(
+ (dist.f[DIR_PPP])[ktne]+ (dist.f[DIR_MMP])[ktsw]+
+ (dist.f[DIR_PMP])[ktse]+ (dist.f[DIR_MPP])[ktnw]+
+ (dist.f[DIR_PPM])[kbne]+ (dist.f[DIR_MMM])[kbsw]+
+ (dist.f[DIR_PMM])[kbse]+ (dist.f[DIR_MPM])[kbnw])-
+ rhoD[nodeIndex]-(vxD[nodeIndex] * vxD[nodeIndex] + vyD[nodeIndex] * vyD[nodeIndex] + vzD[nodeIndex] * vzD[nodeIndex]) * (c1o1+rhoD[nodeIndex])) * c1o2+rhoD[nodeIndex]+
+ PRESS;
+ //achtung op hart gesetzt Annahme op = 1 ; ^^^^(1.0/op-0.5)=0.5
+ }
+ }
}
////////////////////////////////////////////////////////////////////////////////
@@ -841,309 +687,191 @@ __global__ void LBCalcMedCompSP27( real* vxD,
////////////////////////////////////////////////////////////////////////////////
__global__ void LBCalcMedCompAD27(
- real* vxD,
- real* vyD,
- real* vzD,
- real* rhoD,
- real* pressD,
- real* concD,
- unsigned int* geoD,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- real* DD,
- real* DD_AD,
- bool isEvenTimestep)
+ real* vxD,
+ real* vyD,
+ real* vzD,
+ real* rhoD,
+ real* pressD,
+ real* concD,
+ unsigned int* geoD,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ real* distributions,
+ real* distributionsAD,
+ bool isEvenTimestep)
{
- Distributions27 D;
- if (isEvenTimestep == true)
- {
- D.f[DIR_P00] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_M00] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0P0] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0M0] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00P] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00M] = &DD[DIR_00M *size_Mat];
- D.f[DIR_PP0] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_MM0] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_PM0] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_MP0] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_P0P] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_M0M] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_P0M] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_M0P] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0PP] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0MM] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0PM] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0MP] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_MMP] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_PMP] = &DD[DIR_PMP *size_Mat];
- D.f[DIR_MPP] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_PPM] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_MMM] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_PMM] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_MPM] = &DD[DIR_MPM *size_Mat];
- }
- else
- {
- D.f[DIR_M00] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_P00] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0M0] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0P0] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00M] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00P] = &DD[DIR_00M *size_Mat];
- D.f[DIR_MM0] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_PP0] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_MP0] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_PM0] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_M0M] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_P0P] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_M0P] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_P0M] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0MM] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0PP] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0MP] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0PM] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_MMP] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_PMP] = &DD[DIR_MPM *size_Mat];
- D.f[DIR_MPP] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_PPM] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_MMM] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_PMM] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_MPM] = &DD[DIR_PMP *size_Mat];
- }
- ////////////////////////////////////////////////////////////////////////////////
- Distributions27 Dad;
- if (isEvenTimestep == true)
- {
- Dad.f[DIR_P00] = &DD_AD[DIR_P00 *size_Mat];
- Dad.f[DIR_M00] = &DD_AD[DIR_M00 *size_Mat];
- Dad.f[DIR_0P0] = &DD_AD[DIR_0P0 *size_Mat];
- Dad.f[DIR_0M0] = &DD_AD[DIR_0M0 *size_Mat];
- Dad.f[DIR_00P] = &DD_AD[DIR_00P *size_Mat];
- Dad.f[DIR_00M] = &DD_AD[DIR_00M *size_Mat];
- Dad.f[DIR_PP0] = &DD_AD[DIR_PP0 *size_Mat];
- Dad.f[DIR_MM0] = &DD_AD[DIR_MM0 *size_Mat];
- Dad.f[DIR_PM0] = &DD_AD[DIR_PM0 *size_Mat];
- Dad.f[DIR_MP0] = &DD_AD[DIR_MP0 *size_Mat];
- Dad.f[DIR_P0P] = &DD_AD[DIR_P0P *size_Mat];
- Dad.f[DIR_M0M] = &DD_AD[DIR_M0M *size_Mat];
- Dad.f[DIR_P0M] = &DD_AD[DIR_P0M *size_Mat];
- Dad.f[DIR_M0P] = &DD_AD[DIR_M0P *size_Mat];
- Dad.f[DIR_0PP] = &DD_AD[DIR_0PP *size_Mat];
- Dad.f[DIR_0MM] = &DD_AD[DIR_0MM *size_Mat];
- Dad.f[DIR_0PM] = &DD_AD[DIR_0PM *size_Mat];
- Dad.f[DIR_0MP] = &DD_AD[DIR_0MP *size_Mat];
- Dad.f[DIR_000] = &DD_AD[DIR_000*size_Mat];
- Dad.f[DIR_PPP] = &DD_AD[DIR_PPP *size_Mat];
- Dad.f[DIR_MMP] = &DD_AD[DIR_MMP *size_Mat];
- Dad.f[DIR_PMP] = &DD_AD[DIR_PMP *size_Mat];
- Dad.f[DIR_MPP] = &DD_AD[DIR_MPP *size_Mat];
- Dad.f[DIR_PPM] = &DD_AD[DIR_PPM *size_Mat];
- Dad.f[DIR_MMM] = &DD_AD[DIR_MMM *size_Mat];
- Dad.f[DIR_PMM] = &DD_AD[DIR_PMM *size_Mat];
- Dad.f[DIR_MPM] = &DD_AD[DIR_MPM *size_Mat];
- }
- else
- {
- Dad.f[DIR_M00] = &DD_AD[DIR_P00 *size_Mat];
- Dad.f[DIR_P00] = &DD_AD[DIR_M00 *size_Mat];
- Dad.f[DIR_0M0] = &DD_AD[DIR_0P0 *size_Mat];
- Dad.f[DIR_0P0] = &DD_AD[DIR_0M0 *size_Mat];
- Dad.f[DIR_00M] = &DD_AD[DIR_00P *size_Mat];
- Dad.f[DIR_00P] = &DD_AD[DIR_00M *size_Mat];
- Dad.f[DIR_MM0] = &DD_AD[DIR_PP0 *size_Mat];
- Dad.f[DIR_PP0] = &DD_AD[DIR_MM0 *size_Mat];
- Dad.f[DIR_MP0] = &DD_AD[DIR_PM0 *size_Mat];
- Dad.f[DIR_PM0] = &DD_AD[DIR_MP0 *size_Mat];
- Dad.f[DIR_M0M] = &DD_AD[DIR_P0P *size_Mat];
- Dad.f[DIR_P0P] = &DD_AD[DIR_M0M *size_Mat];
- Dad.f[DIR_M0P] = &DD_AD[DIR_P0M *size_Mat];
- Dad.f[DIR_P0M] = &DD_AD[DIR_M0P *size_Mat];
- Dad.f[DIR_0MM] = &DD_AD[DIR_0PP *size_Mat];
- Dad.f[DIR_0PP] = &DD_AD[DIR_0MM *size_Mat];
- Dad.f[DIR_0MP] = &DD_AD[DIR_0PM *size_Mat];
- Dad.f[DIR_0PM] = &DD_AD[DIR_0MP *size_Mat];
- Dad.f[DIR_000] = &DD_AD[DIR_000*size_Mat];
- Dad.f[DIR_PPP] = &DD_AD[DIR_MMM *size_Mat];
- Dad.f[DIR_MMP] = &DD_AD[DIR_PPM *size_Mat];
- Dad.f[DIR_PMP] = &DD_AD[DIR_MPM *size_Mat];
- Dad.f[DIR_MPP] = &DD_AD[DIR_PMM *size_Mat];
- Dad.f[DIR_PPM] = &DD_AD[DIR_MMP *size_Mat];
- Dad.f[DIR_MMM] = &DD_AD[DIR_PPP *size_Mat];
- Dad.f[DIR_PMM] = &DD_AD[DIR_MPP *size_Mat];
- Dad.f[DIR_MPM] = &DD_AD[DIR_PMP *size_Mat];
- }
- ////////////////////////////////////////////////////////////////////////////////
- const unsigned x = threadIdx.x; // Globaler x-Index
- const unsigned y = blockIdx.x; // Globaler y-Index
- const unsigned z = blockIdx.y; // Globaler z-Index
-
- const unsigned nx = blockDim.x;
- const unsigned ny = gridDim.x;
-
- const unsigned k = nx*(ny*z + y) + x;
- //////////////////////////////////////////////////////////////////////////
-
- if (k < size_Mat)
- {
- //////////////////////////////////////////////////////////////////////////
- //index
- //unsigned int kzero = k;
- unsigned int ke = k;
- unsigned int kw = neighborX[k];
- unsigned int kn = k;
- unsigned int ks = neighborY[k];
- unsigned int kt = k;
- unsigned int kb = neighborZ[k];
- unsigned int ksw = neighborY[kw];
- unsigned int kne = k;
- unsigned int kse = ks;
- unsigned int knw = kw;
- unsigned int kbw = neighborZ[kw];
- unsigned int kte = k;
- unsigned int kbe = kb;
- unsigned int ktw = kw;
- unsigned int kbs = neighborZ[ks];
- unsigned int ktn = k;
- unsigned int kbn = kb;
- unsigned int kts = ks;
- unsigned int ktse = ks;
- unsigned int kbnw = kbw;
- unsigned int ktnw = kw;
- unsigned int kbse = kbs;
- unsigned int ktsw = ksw;
- unsigned int kbne = kb;
- unsigned int ktne = k;
- unsigned int kbsw = neighborZ[ksw];
- //////////////////////////////////////////////////////////////////////////
- real CONC = concD[k];
- real PRESS = pressD[k];
- real RHO = rhoD[k];
- real VX = vxD[k];
- real VY = vyD[k];
- real VZ = vzD[k];
- //////////////////////////////////////////////////////////////////////////
- concD[k] = c0o1;
- pressD[k] = c0o1;
- rhoD[k] = c0o1;
- vxD[k] = c0o1;
- vyD[k] = c0o1;
- vzD[k] = c0o1;
-
- if (geoD[k] == GEO_FLUID)
- {
- real mfcbb = (D.f[DIR_P00])[k];//[ke ];
- real mfabb = (D.f[DIR_M00])[kw];//[kw ];
- real mfbcb = (D.f[DIR_0P0])[k];//[kn ];
- real mfbab = (D.f[DIR_0M0])[ks];//[ks ];
- real mfbbc = (D.f[DIR_00P])[k];//[kt ];
- real mfbba = (D.f[DIR_00M])[kb];//[kb ];
- real mfccb = (D.f[DIR_PP0])[k];//[kne ];
- real mfaab = (D.f[DIR_MM0])[ksw];//[ksw ];
- real mfcab = (D.f[DIR_PM0])[ks];//[kse ];
- real mfacb = (D.f[DIR_MP0])[kw];//[knw ];
- real mfcbc = (D.f[DIR_P0P])[k];//[kte ];
- real mfaba = (D.f[DIR_M0M])[kbw];//[kbw ];
- real mfcba = (D.f[DIR_P0M])[kb];//[kbe ];
- real mfabc = (D.f[DIR_M0P])[kw];//[ktw ];
- real mfbcc = (D.f[DIR_0PP])[k];//[ktn ];
- real mfbaa = (D.f[DIR_0MM])[kbs];//[kbs ];
- real mfbca = (D.f[DIR_0PM])[kb];//[kbn ];
- real mfbac = (D.f[DIR_0MP])[ks];//[kts ];
- real mfbbb = (D.f[DIR_000])[k];//[kzero];
- real mfccc = (D.f[DIR_PPP])[k];//[ktne ];
- real mfaac = (D.f[DIR_MMP])[ksw];//[ktsw ];
- real mfcac = (D.f[DIR_PMP])[ks];//[ktse ];
- real mfacc = (D.f[DIR_MPP])[kw];//[ktnw ];
- real mfcca = (D.f[DIR_PPM])[kb];//[kbne ];
- real mfaaa = (D.f[DIR_MMM])[kbsw];//[kbsw ];
- real mfcaa = (D.f[DIR_PMM])[kbs];//[kbse ];
- real mfaca = (D.f[DIR_MPM])[kbw];//[kbnw ];
- ////////////////////////////////////////////////////////////////////////////////////
- real drho =
- ((((mfccc + mfaaa) + (mfaca + mfcac)) + ((mfacc + mfcaa) + (mfaac + mfcca))) +
- (((mfbac + mfbca) + (mfbaa + mfbcc)) + ((mfabc + mfcba) + (mfaba + mfcbc)) + ((mfacb + mfcab) + (mfaab + mfccb))) +
- ((mfabb + mfcbb) + (mfbab + mfbcb) + (mfbba + mfbbc))) + mfbbb;
- real rho = c1o1 + drho;
- ////////////////////////////////////////////////////////////////////////////////////
-
- rhoD[k] = drho + RHO;
-
- vxD[k] =
- (((((mfccc - mfaaa) + (mfcac - mfaca)) + ((mfcaa - mfacc) + (mfcca - mfaac))) +
- (((mfcba - mfabc) + (mfcbc - mfaba)) + ((mfcab - mfacb) + (mfccb - mfaab))) +
- (mfcbb - mfabb)) / rho) + VX;
-
- vyD[k] =
- (((((mfccc - mfaaa) + (mfaca - mfcac)) + ((mfacc - mfcaa) + (mfcca - mfaac))) +
- (((mfbca - mfbac) + (mfbcc - mfbaa)) + ((mfacb - mfcab) + (mfccb - mfaab))) +
- (mfbcb - mfbab)) / rho) + VY;
-
- vzD[k] =
- (((((mfccc - mfaaa) + (mfcac - mfaca)) + ((mfacc - mfcaa) + (mfaac - mfcca))) +
- (((mfbac - mfbca) + (mfbcc - mfbaa)) + ((mfabc - mfcba) + (mfcbc - mfaba))) +
- (mfbbc - mfbba)) / rho) + VZ;
-
- pressD[k] =
- ((D.f[DIR_P00])[ke] + (D.f[DIR_M00])[kw] +
- (D.f[DIR_0P0])[kn] + (D.f[DIR_0M0])[ks] +
- (D.f[DIR_00P])[kt] + (D.f[DIR_00M])[kb] +
- c2o1*(
- (D.f[DIR_PP0])[kne] + (D.f[DIR_MM0])[ksw] +
- (D.f[DIR_PM0])[kse] + (D.f[DIR_MP0])[knw] +
- (D.f[DIR_P0P])[kte] + (D.f[DIR_M0M])[kbw] +
- (D.f[DIR_P0M])[kbe] + (D.f[DIR_M0P])[ktw] +
- (D.f[DIR_0PP])[ktn] + (D.f[DIR_0MM])[kbs] +
- (D.f[DIR_0PM])[kbn] + (D.f[DIR_0MP])[kts]) +
- c3o1*(
- (D.f[DIR_PPP])[ktne] + (D.f[DIR_MMP])[ktsw] +
- (D.f[DIR_PMP])[ktse] + (D.f[DIR_MPP])[ktnw] +
- (D.f[DIR_PPM])[kbne] + (D.f[DIR_MMM])[kbsw] +
- (D.f[DIR_PMM])[kbse] + (D.f[DIR_MPM])[kbnw]) -
- rhoD[k] - (vxD[k] * vxD[k] + vyD[k] * vyD[k] + vzD[k] * vzD[k]) * (c1o1 + rhoD[k])) * c1o2 + rhoD[k] +
- PRESS;
- //achtung op hart gesetzt Annahme op = 1 ; ^^^^(1.0/op-0.5)=0.5
- //////////////////////////////////////////////////////////////////////////
- mfcbb = (Dad.f[DIR_P00 ])[k ];
- mfabb = (Dad.f[DIR_M00 ])[kw ];
- mfbcb = (Dad.f[DIR_0P0 ])[k ];
- mfbab = (Dad.f[DIR_0M0 ])[ks ];
- mfbbc = (Dad.f[DIR_00P ])[k ];
- mfbba = (Dad.f[DIR_00M ])[kb ];
- mfccb = (Dad.f[DIR_PP0 ])[k ];
- mfaab = (Dad.f[DIR_MM0 ])[ksw ];
- mfcab = (Dad.f[DIR_PM0 ])[ks ];
- mfacb = (Dad.f[DIR_MP0 ])[kw ];
- mfcbc = (Dad.f[DIR_P0P ])[k ];
- mfaba = (Dad.f[DIR_M0M ])[kbw ];
- mfcba = (Dad.f[DIR_P0M ])[kb ];
- mfabc = (Dad.f[DIR_M0P ])[kw ];
- mfbcc = (Dad.f[DIR_0PP ])[k ];
- mfbaa = (Dad.f[DIR_0MM ])[kbs ];
- mfbca = (Dad.f[DIR_0PM ])[kb ];
- mfbac = (Dad.f[DIR_0MP ])[ks ];
- mfbbb = (Dad.f[DIR_000])[k ];
- mfccc = (Dad.f[DIR_PPP ])[k ];
- mfaac = (Dad.f[DIR_MMP ])[ksw ];
- mfcac = (Dad.f[DIR_PMP ])[ks ];
- mfacc = (Dad.f[DIR_MPP ])[kw ];
- mfcca = (Dad.f[DIR_PPM ])[kb ];
- mfaaa = (Dad.f[DIR_MMM ])[kbsw];
- mfcaa = (Dad.f[DIR_PMM ])[kbs ];
- mfaca = (Dad.f[DIR_MPM ])[kbw ];
- //////////////////////////////////////////////////////////////////////////
- concD[k] =
- ((((mfccc + mfaaa) + (mfaca + mfcac)) + ((mfacc + mfcaa) + (mfaac + mfcca))) +
- (((mfbac + mfbca) + (mfbaa + mfbcc)) + ((mfabc + mfcba) + (mfaba + mfcbc)) + ((mfacb + mfcab) + (mfaab + mfccb))) +
- ((mfabb + mfcbb) + (mfbab + mfbcb) + (mfbba + mfbbc))) + mfbbb + CONC;
- }
- }
+ ////////////////////////////////////////////////////////////////////////////////
+ //! - Get node index coordinates from threadIdx, blockIdx, blockDim and gridDim.
+ //!
+ const unsigned nodeIndex = getNodeIndex();
+
+ //////////////////////////////////////////////////////////////////////////
+ if ( nodeIndex < numberOfLBnodes )
+ {
+ //////////////////////////////////////////////////////////////////////////
+ //! - Read distributions: style of reading and writing the distributions from/to stored arrays dependent on
+ //! timestep is based on the esoteric twist algorithm \ref <a
+ //! href="https://doi.org/10.3390/computation5020019"><b>[ M. Geier et al. (2017),
+ //! DOI:10.3390/computation5020019 ]</b></a>
+ //!
+ Distributions27 dist, distAD;
+ getPointersToDistributions(dist, distributions, numberOfLBnodes, isEvenTimestep);
+ getPointersToDistributions(distAD, distributionsAD, numberOfLBnodes, isEvenTimestep);
+
+ //////////////////////////////////////////////////////////////////////////
+ //index
+ //unsigned int kzero = k;
+ unsigned int ke = nodeIndex;
+ unsigned int kw = neighborX[nodeIndex];
+ unsigned int kn = nodeIndex;
+ unsigned int ks = neighborY[nodeIndex];
+ unsigned int kt = nodeIndex;
+ unsigned int kb = neighborZ[nodeIndex];
+ unsigned int ksw = neighborY[kw];
+ unsigned int kne = nodeIndex;
+ unsigned int kse = ks;
+ unsigned int knw = kw;
+ unsigned int kbw = neighborZ[kw];
+ unsigned int kte = nodeIndex;
+ unsigned int kbe = kb;
+ unsigned int ktw = kw;
+ unsigned int kbs = neighborZ[ks];
+ unsigned int ktn = nodeIndex;
+ unsigned int kbn = kb;
+ unsigned int kts = ks;
+ unsigned int ktse = ks;
+ unsigned int kbnw = kbw;
+ unsigned int ktnw = kw;
+ unsigned int kbse = kbs;
+ unsigned int ktsw = ksw;
+ unsigned int kbne = kb;
+ unsigned int ktne = nodeIndex;
+ unsigned int kbsw = neighborZ[ksw];
+ //////////////////////////////////////////////////////////////////////////
+ real CONC = concD[nodeIndex];
+ real PRESS = pressD[nodeIndex];
+ real RHO = rhoD[nodeIndex];
+ real VX = vxD[nodeIndex];
+ real VY = vyD[nodeIndex];
+ real VZ = vzD[nodeIndex];
+ //////////////////////////////////////////////////////////////////////////
+ concD[nodeIndex] = c0o1;
+ pressD[nodeIndex] = c0o1;
+ rhoD[nodeIndex] = c0o1;
+ vxD[nodeIndex] = c0o1;
+ vyD[nodeIndex] = c0o1;
+ vzD[nodeIndex] = c0o1;
+
+ if (geoD[nodeIndex] == GEO_FLUID)
+ {
+ real mfcbb = (dist.f[DIR_P00])[nodeIndex];//[ke ];
+ real mfabb = (dist.f[DIR_M00])[kw];//[kw ];
+ real mfbcb = (dist.f[DIR_0P0])[nodeIndex];//[kn ];
+ real mfbab = (dist.f[DIR_0M0])[ks];//[ks ];
+ real mfbbc = (dist.f[DIR_00P])[nodeIndex];//[kt ];
+ real mfbba = (dist.f[DIR_00M])[kb];//[kb ];
+ real mfccb = (dist.f[DIR_PP0])[nodeIndex];//[kne ];
+ real mfaab = (dist.f[DIR_MM0])[ksw];//[ksw ];
+ real mfcab = (dist.f[DIR_PM0])[ks];//[kse ];
+ real mfacb = (dist.f[DIR_MP0])[kw];//[knw ];
+ real mfcbc = (dist.f[DIR_P0P])[nodeIndex];//[kte ];
+ real mfaba = (dist.f[DIR_M0M])[kbw];//[kbw ];
+ real mfcba = (dist.f[DIR_P0M])[kb];//[kbe ];
+ real mfabc = (dist.f[DIR_M0P])[kw];//[ktw ];
+ real mfbcc = (dist.f[DIR_0PP])[nodeIndex];//[ktn ];
+ real mfbaa = (dist.f[DIR_0MM])[kbs];//[kbs ];
+ real mfbca = (dist.f[DIR_0PM])[kb];//[kbn ];
+ real mfbac = (dist.f[DIR_0MP])[ks];//[kts ];
+ real mfbbb = (dist.f[DIR_000])[nodeIndex];//[kzero];
+ real mfccc = (dist.f[DIR_PPP])[nodeIndex];//[ktne ];
+ real mfaac = (dist.f[DIR_MMP])[ksw];//[ktsw ];
+ real mfcac = (dist.f[DIR_PMP])[ks];//[ktse ];
+ real mfacc = (dist.f[DIR_MPP])[kw];//[ktnw ];
+ real mfcca = (dist.f[DIR_PPM])[kb];//[kbne ];
+ real mfaaa = (dist.f[DIR_MMM])[kbsw];//[kbsw ];
+ real mfcaa = (dist.f[DIR_PMM])[kbs];//[kbse ];
+ real mfaca = (dist.f[DIR_MPM])[kbw];//[kbnw ];
+ ////////////////////////////////////////////////////////////////////////////////////
+ real drho =
+ ((((mfccc + mfaaa) + (mfaca + mfcac)) + ((mfacc + mfcaa) + (mfaac + mfcca))) +
+ (((mfbac + mfbca) + (mfbaa + mfbcc)) + ((mfabc + mfcba) + (mfaba + mfcbc)) + ((mfacb + mfcab) + (mfaab + mfccb))) +
+ ((mfabb + mfcbb) + (mfbab + mfbcb) + (mfbba + mfbbc))) + mfbbb;
+ real rho = c1o1 + drho;
+ ////////////////////////////////////////////////////////////////////////////////////
+
+ rhoD[nodeIndex] = drho + RHO;
+
+ vxD[nodeIndex] =
+ (((((mfccc - mfaaa) + (mfcac - mfaca)) + ((mfcaa - mfacc) + (mfcca - mfaac))) +
+ (((mfcba - mfabc) + (mfcbc - mfaba)) + ((mfcab - mfacb) + (mfccb - mfaab))) +
+ (mfcbb - mfabb)) / rho) + VX;
+
+ vyD[nodeIndex] =
+ (((((mfccc - mfaaa) + (mfaca - mfcac)) + ((mfacc - mfcaa) + (mfcca - mfaac))) +
+ (((mfbca - mfbac) + (mfbcc - mfbaa)) + ((mfacb - mfcab) + (mfccb - mfaab))) +
+ (mfbcb - mfbab)) / rho) + VY;
+
+ vzD[nodeIndex] =
+ (((((mfccc - mfaaa) + (mfcac - mfaca)) + ((mfacc - mfcaa) + (mfaac - mfcca))) +
+ (((mfbac - mfbca) + (mfbcc - mfbaa)) + ((mfabc - mfcba) + (mfcbc - mfaba))) +
+ (mfbbc - mfbba)) / rho) + VZ;
+
+ pressD[nodeIndex] =
+ ((dist.f[DIR_P00])[ke] + (dist.f[DIR_M00])[kw] +
+ (dist.f[DIR_0P0])[kn] + (dist.f[DIR_0M0])[ks] +
+ (dist.f[DIR_00P])[kt] + (dist.f[DIR_00M])[kb] +
+ c2o1*(
+ (dist.f[DIR_PP0])[kne] + (dist.f[DIR_MM0])[ksw] +
+ (dist.f[DIR_PM0])[kse] + (dist.f[DIR_MP0])[knw] +
+ (dist.f[DIR_P0P])[kte] + (dist.f[DIR_M0M])[kbw] +
+ (dist.f[DIR_P0M])[kbe] + (dist.f[DIR_M0P])[ktw] +
+ (dist.f[DIR_0PP])[ktn] + (dist.f[DIR_0MM])[kbs] +
+ (dist.f[DIR_0PM])[kbn] + (dist.f[DIR_0MP])[kts]) +
+ c3o1*(
+ (dist.f[DIR_PPP])[ktne] + (dist.f[DIR_MMP])[ktsw] +
+ (dist.f[DIR_PMP])[ktse] + (dist.f[DIR_MPP])[ktnw] +
+ (dist.f[DIR_PPM])[kbne] + (dist.f[DIR_MMM])[kbsw] +
+ (dist.f[DIR_PMM])[kbse] + (dist.f[DIR_MPM])[kbnw]) -
+ rhoD[nodeIndex] - (vxD[nodeIndex] * vxD[nodeIndex] + vyD[nodeIndex] * vyD[nodeIndex] + vzD[nodeIndex] * vzD[nodeIndex]) * (c1o1 + rhoD[nodeIndex])) * c1o2 + rhoD[nodeIndex] +
+ PRESS;
+ //achtung op hart gesetzt Annahme op = 1 ; ^^^^(1.0/op-0.5)=0.5
+ //////////////////////////////////////////////////////////////////////////
+ mfcbb = (distAD.f[DIR_P00])[nodeIndex ];
+ mfabb = (distAD.f[DIR_M00])[kw ];
+ mfbcb = (distAD.f[DIR_0P0])[nodeIndex ];
+ mfbab = (distAD.f[DIR_0M0])[ks ];
+ mfbbc = (distAD.f[DIR_00P])[nodeIndex ];
+ mfbba = (distAD.f[DIR_00M])[kb ];
+ mfccb = (distAD.f[DIR_PP0])[nodeIndex ];
+ mfaab = (distAD.f[DIR_MM0])[ksw ];
+ mfcab = (distAD.f[DIR_PM0])[ks ];
+ mfacb = (distAD.f[DIR_MP0])[kw ];
+ mfcbc = (distAD.f[DIR_P0P])[nodeIndex ];
+ mfaba = (distAD.f[DIR_M0M])[kbw ];
+ mfcba = (distAD.f[DIR_P0M])[kb ];
+ mfabc = (distAD.f[DIR_M0P])[kw ];
+ mfbcc = (distAD.f[DIR_0PP])[nodeIndex ];
+ mfbaa = (distAD.f[DIR_0MM])[kbs ];
+ mfbca = (distAD.f[DIR_0PM])[kb ];
+ mfbac = (distAD.f[DIR_0MP])[ks ];
+ mfbbb = (distAD.f[DIR_000])[nodeIndex ];
+ mfccc = (distAD.f[DIR_PPP])[nodeIndex ];
+ mfaac = (distAD.f[DIR_MMP])[ksw ];
+ mfcac = (distAD.f[DIR_PMP])[ks ];
+ mfacc = (distAD.f[DIR_MPP])[kw ];
+ mfcca = (distAD.f[DIR_PPM])[kb ];
+ mfaaa = (distAD.f[DIR_MMM])[kbsw];
+ mfcaa = (distAD.f[DIR_PMM])[kbs ];
+ mfaca = (distAD.f[DIR_MPM])[kbw ];
+ //////////////////////////////////////////////////////////////////////////
+ concD[nodeIndex] =
+ ((((mfccc + mfaaa) + (mfaca + mfcac)) + ((mfacc + mfcaa) + (mfaac + mfcca))) +
+ (((mfbac + mfbca) + (mfbaa + mfbcc)) + ((mfabc + mfcba) + (mfaba + mfcbc)) + ((mfacb + mfcab) + (mfaab + mfccb))) +
+ ((mfabb + mfcbb) + (mfbab + mfbcb) + (mfbba + mfbbc))) + mfbbb + CONC;
+ }
+ }
}
////////////////////////////////////////////////////////////////////////////////
@@ -1168,54 +896,50 @@ __global__ void LBCalcMedCompAD27(
////////////////////////////////////////////////////////////////////////////////
-__global__ void LBCalcMacMedSP27( real* vxD,
- real* vyD,
- real* vzD,
- real* rhoD,
- real* pressD,
- unsigned int* geoD,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int tdiff,
- unsigned int size_Mat,
- bool isEvenTimestep)
+__global__ void LBCalcMacMedSP27(
+ real* vxD,
+ real* vyD,
+ real* vzD,
+ real* rhoD,
+ real* pressD,
+ unsigned int* geoD,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned int tdiff,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep)
{
- ////////////////////////////////////////////////////////////////////////////////
- const unsigned x = threadIdx.x; // Globaler x-Index
- const unsigned y = blockIdx.x; // Globaler y-Index
- const unsigned z = blockIdx.y; // Globaler z-Index
-
- const unsigned nx = blockDim.x;
- const unsigned ny = gridDim.x;
-
- const unsigned k = nx*(ny*z + y) + x;
- //////////////////////////////////////////////////////////////////////////
-
- if(k<size_Mat)
- {
- //////////////////////////////////////////////////////////////////////////
- real PRESS = pressD[k];
- real RHO = rhoD[k];
- real VX = vxD[k];
- real VY = vyD[k];
- real VZ = vzD[k];
- //////////////////////////////////////////////////////////////////////////
- pressD[k] = c0o1;
- rhoD[k] = c0o1;
- vxD[k] = c0o1;
- vyD[k] = c0o1;
- vzD[k] = c0o1;
-
- if(geoD[k] == GEO_FLUID)
- {
- rhoD[k] = RHO / tdiff;
- vxD[k] = VX / tdiff;
- vyD[k] = VY / tdiff;
- vzD[k] = VZ / tdiff;
- pressD[k] = PRESS / tdiff;
- }
- }
+ ////////////////////////////////////////////////////////////////////////////////
+ //! - Get node index coordinates from threadIdx, blockIdx, blockDim and gridDim.
+ //!
+ const unsigned nodeIndex = getNodeIndex();
+
+ //////////////////////////////////////////////////////////////////////////
+ if(nodeIndex<numberOfLBnodes)
+ {
+ //////////////////////////////////////////////////////////////////////////
+ real PRESS = pressD[nodeIndex];
+ real RHO = rhoD[nodeIndex];
+ real VX = vxD[nodeIndex];
+ real VY = vyD[nodeIndex];
+ real VZ = vzD[nodeIndex];
+ //////////////////////////////////////////////////////////////////////////
+ pressD[nodeIndex] = c0o1;
+ rhoD[nodeIndex] = c0o1;
+ vxD[nodeIndex] = c0o1;
+ vyD[nodeIndex] = c0o1;
+ vzD[nodeIndex] = c0o1;
+
+ if(geoD[nodeIndex] == GEO_FLUID)
+ {
+ rhoD[nodeIndex] = RHO / tdiff;
+ vxD[nodeIndex] = VX / tdiff;
+ vyD[nodeIndex] = VY / tdiff;
+ vzD[nodeIndex] = VZ / tdiff;
+ pressD[nodeIndex] = PRESS / tdiff;
+ }
+ }
}
////////////////////////////////////////////////////////////////////////////////
@@ -1241,34 +965,29 @@ __global__ void LBCalcMacMedSP27( real* vxD,
////////////////////////////////////////////////////////////////////////////////
__global__ void LBResetMedianValuesSP27(
- real* vxD,
- real* vyD,
- real* vzD,
- real* rhoD,
- real* pressD,
- unsigned int size_Mat,
- bool isEvenTimestep)
+ real* vxD,
+ real* vyD,
+ real* vzD,
+ real* rhoD,
+ real* pressD,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep)
{
- ////////////////////////////////////////////////////////////////////////////////
- const unsigned x = threadIdx.x; // Globaler x-Index
- const unsigned y = blockIdx.x; // Globaler y-Index
- const unsigned z = blockIdx.y; // Globaler z-Index
-
- const unsigned nx = blockDim.x;
- const unsigned ny = gridDim.x;
-
- const unsigned k = nx*(ny*z + y) + x;
- //////////////////////////////////////////////////////////////////////////
-
- if (k<size_Mat)
- {
- //////////////////////////////////////////////////////////////////////////
- pressD[k] = c0o1;
- rhoD[k] = c0o1;
- vxD[k] = c0o1;
- vyD[k] = c0o1;
- vzD[k] = c0o1;
- }
+ ////////////////////////////////////////////////////////////////////////////////
+ //! - Get node index coordinates from threadIdx, blockIdx, blockDim and gridDim.
+ //!
+ const unsigned nodeIndex = getNodeIndex();
+
+ //////////////////////////////////////////////////////////////////////////
+ if ( nodeIndex < numberOfLBnodes )
+ {
+ //////////////////////////////////////////////////////////////////////////
+ pressD[nodeIndex] = c0o1;
+ rhoD[nodeIndex] = c0o1;
+ vxD[nodeIndex] = c0o1;
+ vyD[nodeIndex] = c0o1;
+ vzD[nodeIndex] = c0o1;
+ }
}
////////////////////////////////////////////////////////////////////////////////
@@ -1294,36 +1013,30 @@ __global__ void LBResetMedianValuesSP27(
////////////////////////////////////////////////////////////////////////////////
__global__ void LBResetMedianValuesAD27(
- real* vxD,
- real* vyD,
- real* vzD,
- real* rhoD,
- real* pressD,
- real* concD,
- unsigned int size_Mat,
- bool isEvenTimestep)
+ real* vxD,
+ real* vyD,
+ real* vzD,
+ real* rhoD,
+ real* pressD,
+ real* concD,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep)
{
- ////////////////////////////////////////////////////////////////////////////////
- const unsigned x = threadIdx.x; // Globaler x-Index
- const unsigned y = blockIdx.x; // Globaler y-Index
- const unsigned z = blockIdx.y; // Globaler z-Index
-
- const unsigned nx = blockDim.x;
- const unsigned ny = gridDim.x;
-
- const unsigned k = nx*(ny*z + y) + x;
- //////////////////////////////////////////////////////////////////////////
-
- if (k < size_Mat)
- {
- //////////////////////////////////////////////////////////////////////////
- concD[k] = c0o1;
- pressD[k] = c0o1;
- rhoD[k] = c0o1;
- vxD[k] = c0o1;
- vyD[k] = c0o1;
- vzD[k] = c0o1;
- }
+ ////////////////////////////////////////////////////////////////////////////////
+ //! - Get node index coordinates from threadIdx, blockIdx, blockDim and gridDim.
+ //!
+ const unsigned nodeIndex = getNodeIndex();
+
+ //////////////////////////////////////////////////////////////////////////
+ if (nodeIndex < numberOfLBnodes)
+ {
+ concD[nodeIndex] = c0o1;
+ pressD[nodeIndex] = c0o1;
+ rhoD[nodeIndex] = c0o1;
+ vxD[nodeIndex] = c0o1;
+ vyD[nodeIndex] = c0o1;
+ vzD[nodeIndex] = c0o1;
+ }
}
////////////////////////////////////////////////////////////////////////////////
@@ -1348,177 +1061,121 @@ __global__ void LBResetMedianValuesAD27(
////////////////////////////////////////////////////////////////////////////////
-__global__ void LBCalcMeasurePoints( real* vxMP,
- real* vyMP,
- real* vzMP,
- real* rhoMP,
- unsigned int* kMP,
- unsigned int numberOfPointskMP,
- unsigned int MPClockCycle,
- unsigned int t,
- unsigned int* geoD,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- real* DD,
- bool isEvenTimestep)
+__global__ void LBCalcMeasurePoints(
+ real* vxMP,
+ real* vyMP,
+ real* vzMP,
+ real* rhoMP,
+ unsigned int* kMP,
+ unsigned int numberOfPointskMP,
+ unsigned int MPClockCycle,
+ unsigned int t,
+ unsigned int* geoD,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ real* distributions,
+ bool isEvenTimestep)
{
- Distributions27 D;
- if (isEvenTimestep==true)
- {
- D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
- }
- else
- {
- D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
- }
- ////////////////////////////////////////////////////////////////////////////////
- const unsigned x = threadIdx.x; // Globaler x-Index
- const unsigned y = blockIdx.x; // Globaler y-Index
- const unsigned z = blockIdx.y; // Globaler z-Index
-
- const unsigned nx = blockDim.x;
- const unsigned ny = gridDim.x;
-
- const unsigned k = nx*(ny*z + y) + x;
- //////////////////////////////////////////////////////////////////////////
-
- if(k<numberOfPointskMP)
- {
- //////////////////////////////////////////////////////////////////////////
- //index
- unsigned int kzero= kMP[k];//k;
- unsigned int ke = kzero;
- unsigned int kw = neighborX[kzero];
- unsigned int kn = kzero;
- unsigned int ks = neighborY[kzero];
- unsigned int kt = kzero;
- unsigned int kb = neighborZ[kzero];
- unsigned int ksw = neighborY[kw];
- unsigned int kne = kzero;
- unsigned int kse = ks;
- unsigned int knw = kw;
- unsigned int kbw = neighborZ[kw];
- unsigned int kte = kzero;
- unsigned int kbe = kb;
- unsigned int ktw = kw;
- unsigned int kbs = neighborZ[ks];
- unsigned int ktn = kzero;
- unsigned int kbn = kb;
- unsigned int kts = ks;
- unsigned int ktse = ks;
- unsigned int kbnw = kbw;
- unsigned int ktnw = kw;
- unsigned int kbse = kbs;
- unsigned int ktsw = ksw;
- unsigned int kbne = kb;
- unsigned int ktne = kzero;
- unsigned int kbsw = neighborZ[ksw];
- //////////////////////////////////////////////////////////////////////////
- unsigned int kMac = k*MPClockCycle + t;
- //////////////////////////////////////////////////////////////////////////
-
- if(geoD[kzero] == GEO_FLUID)
- {
- rhoMP[kMac]= (D.f[DIR_P00 ])[ke ]+ (D.f[DIR_M00 ])[kw ]+
- (D.f[DIR_0P0 ])[kn ]+ (D.f[DIR_0M0 ])[ks ]+
- (D.f[DIR_00P ])[kt ]+ (D.f[DIR_00M ])[kb ]+
- (D.f[DIR_PP0 ])[kne ]+ (D.f[DIR_MM0 ])[ksw ]+
- (D.f[DIR_PM0 ])[kse ]+ (D.f[DIR_MP0 ])[knw ]+
- (D.f[DIR_P0P ])[kte ]+ (D.f[DIR_M0M ])[kbw ]+
- (D.f[DIR_P0M ])[kbe ]+ (D.f[DIR_M0P ])[ktw ]+
- (D.f[DIR_0PP ])[ktn ]+ (D.f[DIR_0MM ])[kbs ]+
- (D.f[DIR_0PM ])[kbn ]+ (D.f[DIR_0MP ])[kts ]+
- (D.f[DIR_000])[kzero]+
- (D.f[DIR_PPP ])[ktne]+ (D.f[DIR_MMP ])[ktsw]+
- (D.f[DIR_PMP ])[ktse]+ (D.f[DIR_MPP ])[ktnw]+
- (D.f[DIR_PPM ])[kbne]+ (D.f[DIR_MMM ])[kbsw]+
- (D.f[DIR_PMM ])[kbse]+ (D.f[DIR_MPM ])[kbnw];
-
- vxMP[kMac] = (D.f[DIR_P00 ])[ke ]- (D.f[DIR_M00 ])[kw ]+
- (D.f[DIR_PP0 ])[kne ]- (D.f[DIR_MM0 ])[ksw ]+
- (D.f[DIR_PM0 ])[kse ]- (D.f[DIR_MP0 ])[knw ]+
- (D.f[DIR_P0P ])[kte ]- (D.f[DIR_M0M ])[kbw ]+
- (D.f[DIR_P0M ])[kbe ]- (D.f[DIR_M0P ])[ktw ]+
- (D.f[DIR_PPP ])[ktne]- (D.f[DIR_MMP ])[ktsw]+
- (D.f[DIR_PMP ])[ktse]- (D.f[DIR_MPP ])[ktnw]+
- (D.f[DIR_PPM ])[kbne]- (D.f[DIR_MMM ])[kbsw]+
- (D.f[DIR_PMM ])[kbse]- (D.f[DIR_MPM ])[kbnw];
-
- vyMP[kMac] = (D.f[DIR_0P0 ])[kn ]- (D.f[DIR_0M0 ])[ks ]+
- (D.f[DIR_PP0 ])[kne ]- (D.f[DIR_MM0 ])[ksw ]-
- (D.f[DIR_PM0 ])[kse ]+ (D.f[DIR_MP0 ])[knw ]+
- (D.f[DIR_0PP ])[ktn ]- (D.f[DIR_0MM ])[kbs ]+
- (D.f[DIR_0PM ])[kbn ]- (D.f[DIR_0MP ])[kts ]+
- (D.f[DIR_PPP ])[ktne]- (D.f[DIR_MMP ])[ktsw]-
- (D.f[DIR_PMP ])[ktse]+ (D.f[DIR_MPP ])[ktnw]+
- (D.f[DIR_PPM ])[kbne]- (D.f[DIR_MMM ])[kbsw]-
- (D.f[DIR_PMM ])[kbse]+ (D.f[DIR_MPM ])[kbnw];
-
- vzMP[kMac] = (D.f[DIR_00P ])[kt ]- (D.f[DIR_00M ])[kb ]+
- (D.f[DIR_P0P ])[kte ]- (D.f[DIR_M0M ])[kbw ]-
- (D.f[DIR_P0M ])[kbe ]+ (D.f[DIR_M0P ])[ktw ]+
- (D.f[DIR_0PP ])[ktn ]- (D.f[DIR_0MM ])[kbs ]-
- (D.f[DIR_0PM ])[kbn ]+ (D.f[DIR_0MP ])[kts ]+
- (D.f[DIR_PPP ])[ktne]+ (D.f[DIR_MMP ])[ktsw]+
- (D.f[DIR_PMP ])[ktse]+ (D.f[DIR_MPP ])[ktnw]-
- (D.f[DIR_PPM ])[kbne]- (D.f[DIR_MMM ])[kbsw]-
- (D.f[DIR_PMM ])[kbse]- (D.f[DIR_MPM ])[kbnw];
- }
- }
+ ////////////////////////////////////////////////////////////////////////////////
+ //! - Get node index coordinates from threadIdx, blockIdx, blockDim and gridDim.
+ //!
+ const unsigned nodeIndex = getNodeIndex();
+
+ //////////////////////////////////////////////////////////////////////////
+ if( nodeIndex < numberOfPointskMP )
+ {
+ //////////////////////////////////////////////////////////////////////////
+ //! - Read distributions: style of reading and writing the distributions from/to stored arrays dependent on
+ //! timestep is based on the esoteric twist algorithm \ref <a
+ //! href="https://doi.org/10.3390/computation5020019"><b>[ M. Geier et al. (2017),
+ //! DOI:10.3390/computation5020019 ]</b></a>
+ //!
+ Distributions27 dist;
+ getPointersToDistributions(dist, distributions, numberOfLBnodes, isEvenTimestep);
+
+ //////////////////////////////////////////////////////////////////////////
+ //index
+ unsigned int kzero= kMP[nodeIndex];//k;
+ unsigned int ke = kzero;
+ unsigned int kw = neighborX[kzero];
+ unsigned int kn = kzero;
+ unsigned int ks = neighborY[kzero];
+ unsigned int kt = kzero;
+ unsigned int kb = neighborZ[kzero];
+ unsigned int ksw = neighborY[kw];
+ unsigned int kne = kzero;
+ unsigned int kse = ks;
+ unsigned int knw = kw;
+ unsigned int kbw = neighborZ[kw];
+ unsigned int kte = kzero;
+ unsigned int kbe = kb;
+ unsigned int ktw = kw;
+ unsigned int kbs = neighborZ[ks];
+ unsigned int ktn = kzero;
+ unsigned int kbn = kb;
+ unsigned int kts = ks;
+ unsigned int ktse = ks;
+ unsigned int kbnw = kbw;
+ unsigned int ktnw = kw;
+ unsigned int kbse = kbs;
+ unsigned int ktsw = ksw;
+ unsigned int kbne = kb;
+ unsigned int ktne = kzero;
+ unsigned int kbsw = neighborZ[ksw];
+ //////////////////////////////////////////////////////////////////////////
+ unsigned int kMac = nodeIndex*MPClockCycle + t;
+ //////////////////////////////////////////////////////////////////////////
+
+ if(geoD[kzero] == GEO_FLUID)
+ {
+ rhoMP[kMac]= (dist.f[DIR_P00])[ke ]+ (dist.f[DIR_M00])[kw ]+
+ (dist.f[DIR_0P0])[kn ]+ (dist.f[DIR_0M0])[ks ]+
+ (dist.f[DIR_00P])[kt ]+ (dist.f[DIR_00M])[kb ]+
+ (dist.f[DIR_PP0])[kne ]+ (dist.f[DIR_MM0])[ksw ]+
+ (dist.f[DIR_PM0])[kse ]+ (dist.f[DIR_MP0])[knw ]+
+ (dist.f[DIR_P0P])[kte ]+ (dist.f[DIR_M0M])[kbw ]+
+ (dist.f[DIR_P0M])[kbe ]+ (dist.f[DIR_M0P])[ktw ]+
+ (dist.f[DIR_0PP])[ktn ]+ (dist.f[DIR_0MM])[kbs ]+
+ (dist.f[DIR_0PM])[kbn ]+ (dist.f[DIR_0MP])[kts ]+
+ (dist.f[DIR_000])[kzero]+
+ (dist.f[DIR_PPP])[ktne]+ (dist.f[DIR_MMP])[ktsw]+
+ (dist.f[DIR_PMP])[ktse]+ (dist.f[DIR_MPP])[ktnw]+
+ (dist.f[DIR_PPM])[kbne]+ (dist.f[DIR_MMM])[kbsw]+
+ (dist.f[DIR_PMM])[kbse]+ (dist.f[DIR_MPM])[kbnw];
+
+ vxMP[kMac] = (dist.f[DIR_P00])[ke ]- (dist.f[DIR_M00])[kw ]+
+ (dist.f[DIR_PP0])[kne ]- (dist.f[DIR_MM0])[ksw ]+
+ (dist.f[DIR_PM0])[kse ]- (dist.f[DIR_MP0])[knw ]+
+ (dist.f[DIR_P0P])[kte ]- (dist.f[DIR_M0M])[kbw ]+
+ (dist.f[DIR_P0M])[kbe ]- (dist.f[DIR_M0P])[ktw ]+
+ (dist.f[DIR_PPP])[ktne]- (dist.f[DIR_MMP])[ktsw]+
+ (dist.f[DIR_PMP])[ktse]- (dist.f[DIR_MPP])[ktnw]+
+ (dist.f[DIR_PPM])[kbne]- (dist.f[DIR_MMM])[kbsw]+
+ (dist.f[DIR_PMM])[kbse]- (dist.f[DIR_MPM])[kbnw];
+
+ vyMP[kMac] = (dist.f[DIR_0P0])[kn ]- (dist.f[DIR_0M0])[ks ]+
+ (dist.f[DIR_PP0])[kne ]- (dist.f[DIR_MM0])[ksw ]-
+ (dist.f[DIR_PM0])[kse ]+ (dist.f[DIR_MP0])[knw ]+
+ (dist.f[DIR_0PP])[ktn ]- (dist.f[DIR_0MM])[kbs ]+
+ (dist.f[DIR_0PM])[kbn ]- (dist.f[DIR_0MP])[kts ]+
+ (dist.f[DIR_PPP])[ktne]- (dist.f[DIR_MMP])[ktsw]-
+ (dist.f[DIR_PMP])[ktse]+ (dist.f[DIR_MPP])[ktnw]+
+ (dist.f[DIR_PPM])[kbne]- (dist.f[DIR_MMM])[kbsw]-
+ (dist.f[DIR_PMM])[kbse]+ (dist.f[DIR_MPM])[kbnw];
+
+ vzMP[kMac] = (dist.f[DIR_00P])[kt ]- (dist.f[DIR_00M])[kb ]+
+ (dist.f[DIR_P0P])[kte ]- (dist.f[DIR_M0M])[kbw ]-
+ (dist.f[DIR_P0M])[kbe ]+ (dist.f[DIR_M0P])[ktw ]+
+ (dist.f[DIR_0PP])[ktn ]- (dist.f[DIR_0MM])[kbs ]-
+ (dist.f[DIR_0PM])[kbn ]+ (dist.f[DIR_0MP])[kts ]+
+ (dist.f[DIR_PPP])[ktne]+ (dist.f[DIR_MMP])[ktsw]+
+ (dist.f[DIR_PMP])[ktse]+ (dist.f[DIR_MPP])[ktnw]-
+ (dist.f[DIR_PPM])[kbne]- (dist.f[DIR_MMM])[kbsw]-
+ (dist.f[DIR_PMM])[kbse]- (dist.f[DIR_MPM])[kbnw];
+ }
+ }
}
////////////////////////////////////////////////////////////////////////////////
@@ -1559,40 +1216,36 @@ __global__ void LBCalcMeasurePoints( real* vxMP,
////////////////////////////////////////////////////////////////////////////////
-__global__ void LBSetOutputWallVelocitySP27( real* vxD,
- real* vyD,
- real* vzD,
- real* vxWall,
- real* vyWall,
- real* vzWall,
- int numberOfWallNodes,
- int* kWallNodes,
- real* rhoD,
- real* pressD,
- unsigned int* geoD,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- real* DD,
- bool isEvenTimestep)
+__global__ void LBSetOutputWallVelocitySP27(
+ real* vxD,
+ real* vyD,
+ real* vzD,
+ real* vxWall,
+ real* vyWall,
+ real* vzWall,
+ int numberOfWallNodes,
+ int* kWallNodes,
+ real* rhoD,
+ real* pressD,
+ unsigned int* geoD,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ real* DD,
+ bool isEvenTimestep)
{
////////////////////////////////////////////////////////////////////////////////
- const unsigned x = threadIdx.x; // Globaler x-Index
- const unsigned y = blockIdx.x; // Globaler y-Index
- const unsigned z = blockIdx.y; // Globaler z-Index
+ //! - Get node index coordinates from threadIdx, blockIdx, blockDim and gridDim.
+ //!
+ const unsigned nodeIndex = getNodeIndex();
- const unsigned nx = blockDim.x;
- const unsigned ny = gridDim.x;
-
- const unsigned k = nx*(ny*z + y) + x;
//////////////////////////////////////////////////////////////////////////
-
- if(k<numberOfWallNodes)
+ if(nodeIndex<numberOfWallNodes)
{
//////////////////////////////////////////////////////////////////////////
//index
- unsigned int KWN = kWallNodes[k];
+ unsigned int KWN = kWallNodes[nodeIndex];
//////////////////////////////////////////////////////////////////////////
vxD[KWN] = 0.0;//vxWall[k];
vyD[KWN] = 0.0;//vyWall[k];
diff --git a/src/gpu/VirtualFluids_GPU/GPU/Cascade27.cu b/src/gpu/VirtualFluids_GPU/GPU/Cascade27.cu
index a79588421a624cae62ec32127739efb47bb7b2ef..457623d4ee62b624248306b6b900fcff3f026286 100644
--- a/src/gpu/VirtualFluids_GPU/GPU/Cascade27.cu
+++ b/src/gpu/VirtualFluids_GPU/GPU/Cascade27.cu
@@ -15,7 +15,7 @@ __global__ void LB_Kernel_Cascade_SP_27( real omega,
unsigned int* neighborY,
unsigned int* neighborZ,
real* DDStart,
- int size_Mat,
+ unsigned long long numberOfLBnodes,
bool EvenOrOdd)
{
////////////////////////////////////////////////////////////////////////////////
@@ -29,7 +29,7 @@ __global__ void LB_Kernel_Cascade_SP_27( real omega,
const unsigned k = nx*(ny*z + y) + x;
//////////////////////////////////////////////////////////////////////////
- if(k<size_Mat)
+ if(k<numberOfLBnodes)
{
////////////////////////////////////////////////////////////////////////////////
unsigned int BC;
@@ -40,63 +40,63 @@ __global__ void LB_Kernel_Cascade_SP_27( real omega,
Distributions27 D;
if (EvenOrOdd==true)
{
- D.f[DIR_P00 ] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DDStart[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DDStart[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DDStart[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DDStart[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DDStart[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DDStart[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DDStart[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DDStart[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DDStart[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DDStart[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DDStart[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DDStart[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DDStart[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DDStart[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DDStart[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DDStart[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DDStart[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DDStart[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DDStart[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DDStart[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DDStart[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DDStart[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DDStart[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DDStart[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DDStart[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DDStart[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DDStart[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00 ] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_MMM ] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_PPM ] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_MPM ] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_PMM ] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_MMP ] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_PPP ] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_MPP ] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_PMP ] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_M00] = &DDStart[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DDStart[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DDStart[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DDStart[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DDStart[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DDStart[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DDStart[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DDStart[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DDStart[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DDStart[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DDStart[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DDStart[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DDStart[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DDStart[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DDStart[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DDStart[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DDStart[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DDStart[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DDStart[DIR_000 * numberOfLBnodes];
+ D.f[DIR_MMM] = &DDStart[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DDStart[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DDStart[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DDStart[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DDStart[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PPP] = &DDStart[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DDStart[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DDStart[DIR_MPM * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////
@@ -129,33 +129,33 @@ __global__ void LB_Kernel_Cascade_SP_27( real omega,
//unsigned int ktne = k;
unsigned int kbsw = neighborZ[ksw];
//////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
- real mfcbb = (D.f[DIR_P00 ])[k ];//[ke ];// + c2over27 ;(D.f[DIR_P00 ])[k ];//ke
- real mfabb = (D.f[DIR_M00 ])[kw ];//[kw ];// + c2over27 ;(D.f[DIR_M00 ])[kw ];
- real mfbcb = (D.f[DIR_0P0 ])[k ];//[kn ];// + c2over27 ;(D.f[DIR_0P0 ])[k ];//kn
- real mfbab = (D.f[DIR_0M0 ])[ks ];//[ks ];// + c2over27 ;(D.f[DIR_0M0 ])[ks ];
- real mfbbc = (D.f[DIR_00P ])[k ];//[kt ];// + c2over27 ;(D.f[DIR_00P ])[k ];//kt
- real mfbba = (D.f[DIR_00M ])[kb ];//[kb ];// + c2over27 ;(D.f[DIR_00M ])[kb ];
- real mfccb = (D.f[DIR_PP0 ])[k ];//[kne ];// + c1over54 ;(D.f[DIR_PP0 ])[k ];//kne
- real mfaab = (D.f[DIR_MM0 ])[ksw];//[ksw ];// + c1over54 ;(D.f[DIR_MM0 ])[ksw];
- real mfcab = (D.f[DIR_PM0 ])[ks ];//[kse ];// + c1over54 ;(D.f[DIR_PM0 ])[ks ];//kse
- real mfacb = (D.f[DIR_MP0 ])[kw ];//[knw ];// + c1over54 ;(D.f[DIR_MP0 ])[kw ];//knw
- real mfcbc = (D.f[DIR_P0P ])[k ];//[kte ];// + c1over54 ;(D.f[DIR_P0P ])[k ];//kte
- real mfaba = (D.f[DIR_M0M ])[kbw];//[kbw ];// + c1over54 ;(D.f[DIR_M0M ])[kbw];
- real mfcba = (D.f[DIR_P0M ])[kb ];//[kbe ];// + c1over54 ;(D.f[DIR_P0M ])[kb ];//kbe
- real mfabc = (D.f[DIR_M0P ])[kw ];//[ktw ];// + c1over54 ;(D.f[DIR_M0P ])[kw ];//ktw
- real mfbcc = (D.f[DIR_0PP ])[k ];//[ktn ];// + c1over54 ;(D.f[DIR_0PP ])[k ];//ktn
- real mfbaa = (D.f[DIR_0MM ])[kbs];//[kbs ];// + c1over54 ;(D.f[DIR_0MM ])[kbs];
- real mfbca = (D.f[DIR_0PM ])[kb ];//[kbn ];// + c1over54 ;(D.f[DIR_0PM ])[kb ];//kbn
- real mfbac = (D.f[DIR_0MP ])[ks ];//[kts ];// + c1over54 ;(D.f[DIR_0MP ])[ks ];//kts
+ real mfcbb = (D.f[DIR_P00])[k ];//[ke ];// + c2over27 ;(D.f[DIR_P00])[k ];//ke
+ real mfabb = (D.f[DIR_M00])[kw ];//[kw ];// + c2over27 ;(D.f[DIR_M00])[kw ];
+ real mfbcb = (D.f[DIR_0P0])[k ];//[kn ];// + c2over27 ;(D.f[DIR_0P0])[k ];//kn
+ real mfbab = (D.f[DIR_0M0])[ks ];//[ks ];// + c2over27 ;(D.f[DIR_0M0])[ks ];
+ real mfbbc = (D.f[DIR_00P])[k ];//[kt ];// + c2over27 ;(D.f[DIR_00P])[k ];//kt
+ real mfbba = (D.f[DIR_00M])[kb ];//[kb ];// + c2over27 ;(D.f[DIR_00M])[kb ];
+ real mfccb = (D.f[DIR_PP0])[k ];//[kne ];// + c1over54 ;(D.f[DIR_PP0])[k ];//kne
+ real mfaab = (D.f[DIR_MM0])[ksw];//[ksw ];// + c1over54 ;(D.f[DIR_MM0])[ksw];
+ real mfcab = (D.f[DIR_PM0])[ks ];//[kse ];// + c1over54 ;(D.f[DIR_PM0])[ks ];//kse
+ real mfacb = (D.f[DIR_MP0])[kw ];//[knw ];// + c1over54 ;(D.f[DIR_MP0])[kw ];//knw
+ real mfcbc = (D.f[DIR_P0P])[k ];//[kte ];// + c1over54 ;(D.f[DIR_P0P])[k ];//kte
+ real mfaba = (D.f[DIR_M0M])[kbw];//[kbw ];// + c1over54 ;(D.f[DIR_M0M])[kbw];
+ real mfcba = (D.f[DIR_P0M])[kb ];//[kbe ];// + c1over54 ;(D.f[DIR_P0M])[kb ];//kbe
+ real mfabc = (D.f[DIR_M0P])[kw ];//[ktw ];// + c1over54 ;(D.f[DIR_M0P])[kw ];//ktw
+ real mfbcc = (D.f[DIR_0PP])[k ];//[ktn ];// + c1over54 ;(D.f[DIR_0PP])[k ];//ktn
+ real mfbaa = (D.f[DIR_0MM])[kbs];//[kbs ];// + c1over54 ;(D.f[DIR_0MM])[kbs];
+ real mfbca = (D.f[DIR_0PM])[kb ];//[kbn ];// + c1over54 ;(D.f[DIR_0PM])[kb ];//kbn
+ real mfbac = (D.f[DIR_0MP])[ks ];//[kts ];// + c1over54 ;(D.f[DIR_0MP])[ks ];//kts
real mfbbb = (D.f[DIR_000])[k ];//[kzero];// + c8over27 ;(D.f[DIR_000])[k ];//kzero
- real mfccc = (D.f[DIR_PPP ])[k ];//[ktne ];// + c1over216;(D.f[DIR_PPP ])[k ];//ktne
- real mfaac = (D.f[DIR_MMP ])[ksw];//[ktsw ];// + c1over216;(D.f[DIR_MMP ])[ksw];//ktsw
- real mfcac = (D.f[DIR_PMP ])[ks ];//[ktse ];// + c1over216;(D.f[DIR_PMP ])[ks ];//ktse
- real mfacc = (D.f[DIR_MPP ])[kw ];//[ktnw ];// + c1over216;(D.f[DIR_MPP ])[kw ];//ktnw
- real mfcca = (D.f[DIR_PPM ])[kb ];//[kbne ];// + c1over216;(D.f[DIR_PPM ])[kb ];//kbne
- real mfaaa = (D.f[DIR_MMM ])[kbsw];//[kbsw ];// + c1over216;(D.f[DIR_MMM ])[kbsw];
- real mfcaa = (D.f[DIR_PMM ])[kbs];//[kbse ];// + c1over216;(D.f[DIR_PMM ])[kbs];//kbse
- real mfaca = (D.f[DIR_MPM ])[kbw];//[kbnw ];// + c1over216;(D.f[DIR_MPM ])[kbw];//kbnw
+ real mfccc = (D.f[DIR_PPP])[k ];//[ktne ];// + c1over216;(D.f[DIR_PPP])[k ];//ktne
+ real mfaac = (D.f[DIR_MMP])[ksw];//[ktsw ];// + c1over216;(D.f[DIR_MMP])[ksw];//ktsw
+ real mfcac = (D.f[DIR_PMP])[ks ];//[ktse ];// + c1over216;(D.f[DIR_PMP])[ks ];//ktse
+ real mfacc = (D.f[DIR_MPP])[kw ];//[ktnw ];// + c1over216;(D.f[DIR_MPP])[kw ];//ktnw
+ real mfcca = (D.f[DIR_PPM])[kb ];//[kbne ];// + c1over216;(D.f[DIR_PPM])[kb ];//kbne
+ real mfaaa = (D.f[DIR_MMM])[kbsw];//[kbsw ];// + c1over216;(D.f[DIR_MMM])[kbsw];
+ real mfcaa = (D.f[DIR_PMM])[kbs];//[kbse ];// + c1over216;(D.f[DIR_PMM])[kbs];//kbse
+ real mfaca = (D.f[DIR_MPM])[kbw];//[kbnw ];// + c1over216;(D.f[DIR_MPM])[kbw];//kbnw
////////////////////////////////////////////////////////////////////////////////////
//slow
//real oMdrho = one - ((((mfccc+mfaaa) + (mfaca+mfcac)) + ((mfacc+mfcaa) + (mfaac+mfcca))) +
@@ -842,7 +842,7 @@ __global__ void LB_Kernel_Casc_Comp_SP_27( real omega,
unsigned int* neighborY,
unsigned int* neighborZ,
real* DDStart,
- int size_Mat,
+ unsigned long long numberOfLBnodes,
bool EvenOrOdd)
{
////////////////////////////////////////////////////////////////////////////////
@@ -856,7 +856,7 @@ __global__ void LB_Kernel_Casc_Comp_SP_27( real omega,
const unsigned k = nx*(ny*z + y) + x;
//////////////////////////////////////////////////////////////////////////
- if(k<size_Mat)
+ if(k<numberOfLBnodes)
{
////////////////////////////////////////////////////////////////////////////////
unsigned int BC;
@@ -867,63 +867,63 @@ __global__ void LB_Kernel_Casc_Comp_SP_27( real omega,
Distributions27 D;
if (EvenOrOdd==true)
{
- D.f[DIR_P00 ] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DDStart[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DDStart[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DDStart[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DDStart[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DDStart[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DDStart[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DDStart[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DDStart[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DDStart[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DDStart[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DDStart[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DDStart[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DDStart[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DDStart[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DDStart[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DDStart[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DDStart[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DDStart[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DDStart[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DDStart[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DDStart[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DDStart[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DDStart[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DDStart[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DDStart[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DDStart[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DDStart[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00 ] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_MMM ] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_PPM ] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_MPM ] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_PMM ] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_MMP ] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_PPP ] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_MPP ] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_PMP ] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_M00] = &DDStart[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DDStart[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DDStart[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DDStart[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DDStart[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DDStart[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DDStart[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DDStart[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DDStart[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DDStart[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DDStart[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DDStart[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DDStart[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DDStart[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DDStart[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DDStart[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DDStart[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DDStart[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DDStart[DIR_000 * numberOfLBnodes];
+ D.f[DIR_MMM] = &DDStart[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DDStart[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DDStart[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DDStart[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DDStart[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PPP] = &DDStart[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DDStart[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DDStart[DIR_MPM * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////
@@ -956,33 +956,33 @@ __global__ void LB_Kernel_Casc_Comp_SP_27( real omega,
unsigned int ktne = k;
unsigned int kbsw = neighborZ[ksw];
//////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
- real f_E = (D.f[DIR_P00 ])[ke ];// + c2over27 ;
- real f_W = (D.f[DIR_M00 ])[kw ];// + c2over27 ;
- real f_N = (D.f[DIR_0P0 ])[kn ];// + c2over27 ;
- real f_S = (D.f[DIR_0M0 ])[ks ];// + c2over27 ;
- real f_F = (D.f[DIR_00P ])[kt ];// + c2over27 ;
- real f_B = (D.f[DIR_00M ])[kb ];// + c2over27 ;
- real f_NE = (D.f[DIR_PP0 ])[kne ];// + c1over54 ;
- real f_SW = (D.f[DIR_MM0 ])[ksw ];// + c1over54 ;
- real f_SE = (D.f[DIR_PM0 ])[kse ];// + c1over54 ;
- real f_NW = (D.f[DIR_MP0 ])[knw ];// + c1over54 ;
- real f_Ef = (D.f[DIR_P0P ])[kte ];// + c1over54 ;
- real f_Wb = (D.f[DIR_M0M ])[kbw ];// + c1over54 ;
- real f_Eb = (D.f[DIR_P0M ])[kbe ];// + c1over54 ;
- real f_Wf = (D.f[DIR_M0P ])[ktw ];// + c1over54 ;
- real f_Nf = (D.f[DIR_0PP ])[ktn ];// + c1over54 ;
- real f_Sb = (D.f[DIR_0MM ])[kbs ];// + c1over54 ;
- real f_Nb = (D.f[DIR_0PM ])[kbn ];// + c1over54 ;
- real f_Sf = (D.f[DIR_0MP ])[kts ];// + c1over54 ;
+ real f_E = (D.f[DIR_P00])[ke ];// + c2over27 ;
+ real f_W = (D.f[DIR_M00])[kw ];// + c2over27 ;
+ real f_N = (D.f[DIR_0P0])[kn ];// + c2over27 ;
+ real f_S = (D.f[DIR_0M0])[ks ];// + c2over27 ;
+ real f_F = (D.f[DIR_00P])[kt ];// + c2over27 ;
+ real f_B = (D.f[DIR_00M])[kb ];// + c2over27 ;
+ real f_NE = (D.f[DIR_PP0])[kne ];// + c1over54 ;
+ real f_SW = (D.f[DIR_MM0])[ksw ];// + c1over54 ;
+ real f_SE = (D.f[DIR_PM0])[kse ];// + c1over54 ;
+ real f_NW = (D.f[DIR_MP0])[knw ];// + c1over54 ;
+ real f_Ef = (D.f[DIR_P0P])[kte ];// + c1over54 ;
+ real f_Wb = (D.f[DIR_M0M])[kbw ];// + c1over54 ;
+ real f_Eb = (D.f[DIR_P0M])[kbe ];// + c1over54 ;
+ real f_Wf = (D.f[DIR_M0P])[ktw ];// + c1over54 ;
+ real f_Nf = (D.f[DIR_0PP])[ktn ];// + c1over54 ;
+ real f_Sb = (D.f[DIR_0MM])[kbs ];// + c1over54 ;
+ real f_Nb = (D.f[DIR_0PM])[kbn ];// + c1over54 ;
+ real f_Sf = (D.f[DIR_0MP])[kts ];// + c1over54 ;
real f_R = (D.f[DIR_000])[kzero];// + c8over27 ;
- real f_Nef = (D.f[DIR_PPP ])[ktne ];// + c1over216;
- real f_Swf = (D.f[DIR_MMP ])[ktsw ];// + c1over216;
- real f_Sef = (D.f[DIR_PMP ])[ktse ];// + c1over216;
- real f_Nwf = (D.f[DIR_MPP ])[ktnw ];// + c1over216;
- real f_Neb = (D.f[DIR_PPM ])[kbne ];// + c1over216;
- real f_Swb = (D.f[DIR_MMM ])[kbsw ];// + c1over216;
- real f_Seb = (D.f[DIR_PMM ])[kbse ];// + c1over216;
- real f_Nwb = (D.f[DIR_MPM ])[kbnw ];// + c1over216;
+ real f_Nef = (D.f[DIR_PPP])[ktne ];// + c1over216;
+ real f_Swf = (D.f[DIR_MMP])[ktsw ];// + c1over216;
+ real f_Sef = (D.f[DIR_PMP])[ktse ];// + c1over216;
+ real f_Nwf = (D.f[DIR_MPP])[ktnw ];// + c1over216;
+ real f_Neb = (D.f[DIR_PPM])[kbne ];// + c1over216;
+ real f_Swb = (D.f[DIR_MMM])[kbsw ];// + c1over216;
+ real f_Seb = (D.f[DIR_PMM])[kbse ];// + c1over216;
+ real f_Nwb = (D.f[DIR_MPM])[kbnw ];// + c1over216;
////////////////////////////////////////////////////////////////////////////////////
real rho=f_NW+f_W+f_SW+f_S+f_SE+f_E+f_NE+f_N+f_R+f_Nf+f_Nb+f_Sf+f_Sb+f_Ef+f_Eb+f_Wf+f_Wb+f_Nwf+f_Nwb+f_Nef+f_Neb+f_Swf+f_Swb+f_Sef+f_Seb+f_F+f_B+c1o1;// ACHTUNG ne EINS !!!!!!!!
real pix=(f_NE+f_E+f_SE+f_Ef+f_Eb-f_NW-f_W-f_SW-f_Wf-f_Wb+f_Nef+f_Neb+f_Sef+f_Seb-f_Nwf-f_Nwb-f_Swf-f_Swb);
@@ -1689,7 +1689,7 @@ __global__ void LB_Kernel_Casc_SP_MS_OHM_27( real omega,
unsigned int* neighborY,
unsigned int* neighborZ,
real* DDStart,
- int size_Mat,
+ unsigned long long numberOfLBnodes,
bool EvenOrOdd)
{
////////////////////////////////////////////////////////////////////////////////
@@ -1703,7 +1703,7 @@ __global__ void LB_Kernel_Casc_SP_MS_OHM_27( real omega,
const unsigned k = nx*(ny*z + y) + x;
//////////////////////////////////////////////////////////////////////////
- if(k<size_Mat)
+ if(k<numberOfLBnodes)
{
////////////////////////////////////////////////////////////////////////////////
unsigned int BC;
@@ -1714,63 +1714,63 @@ __global__ void LB_Kernel_Casc_SP_MS_OHM_27( real omega,
Distributions27 D;
if (EvenOrOdd==true)
{
- D.f[DIR_P00 ] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DDStart[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DDStart[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DDStart[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DDStart[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DDStart[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DDStart[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DDStart[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DDStart[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DDStart[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DDStart[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DDStart[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DDStart[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DDStart[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DDStart[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DDStart[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DDStart[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DDStart[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DDStart[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DDStart[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DDStart[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DDStart[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DDStart[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DDStart[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DDStart[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DDStart[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DDStart[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DDStart[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00 ] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_MMM ] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_PPM ] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_MPM ] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_PMM ] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_MMP ] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_PPP ] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_MPP ] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_PMP ] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_M00] = &DDStart[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DDStart[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DDStart[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DDStart[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DDStart[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DDStart[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DDStart[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DDStart[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DDStart[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DDStart[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DDStart[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DDStart[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DDStart[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DDStart[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DDStart[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DDStart[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DDStart[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DDStart[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DDStart[DIR_000 * numberOfLBnodes];
+ D.f[DIR_MMM] = &DDStart[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DDStart[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DDStart[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DDStart[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DDStart[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PPP] = &DDStart[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DDStart[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DDStart[DIR_MPM * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////
@@ -1803,33 +1803,33 @@ __global__ void LB_Kernel_Casc_SP_MS_OHM_27( real omega,
//unsigned int ktne = k;
unsigned int kbsw = neighborZ[ksw];
//////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
- real fE = (D.f[DIR_P00 ])[k ];//ke
- real fW = (D.f[DIR_M00 ])[kw ];
- real fN = (D.f[DIR_0P0 ])[k ];//kn
- real fS = (D.f[DIR_0M0 ])[ks ];
- real fT = (D.f[DIR_00P ])[k ];//kt
- real fB = (D.f[DIR_00M ])[kb ];
- real fNE = (D.f[DIR_PP0 ])[k ];//kne
- real fSW = (D.f[DIR_MM0 ])[ksw];
- real fSE = (D.f[DIR_PM0 ])[ks ];//kse
- real fNW = (D.f[DIR_MP0 ])[kw ];//knw
- real fTE = (D.f[DIR_P0P ])[k ];//kte
- real fBW = (D.f[DIR_M0M ])[kbw];
- real fBE = (D.f[DIR_P0M ])[kb ];//kbe
- real fTW = (D.f[DIR_M0P ])[kw ];//ktw
- real fTN = (D.f[DIR_0PP ])[k ];//ktn
- real fBS = (D.f[DIR_0MM ])[kbs];
- real fBN = (D.f[DIR_0PM ])[kb ];//kbn
- real fTS = (D.f[DIR_0MP ])[ks ];//kts
+ real fE = (D.f[DIR_P00])[k ];//ke
+ real fW = (D.f[DIR_M00])[kw ];
+ real fN = (D.f[DIR_0P0])[k ];//kn
+ real fS = (D.f[DIR_0M0])[ks ];
+ real fT = (D.f[DIR_00P])[k ];//kt
+ real fB = (D.f[DIR_00M])[kb ];
+ real fNE = (D.f[DIR_PP0])[k ];//kne
+ real fSW = (D.f[DIR_MM0])[ksw];
+ real fSE = (D.f[DIR_PM0])[ks ];//kse
+ real fNW = (D.f[DIR_MP0])[kw ];//knw
+ real fTE = (D.f[DIR_P0P])[k ];//kte
+ real fBW = (D.f[DIR_M0M])[kbw];
+ real fBE = (D.f[DIR_P0M])[kb ];//kbe
+ real fTW = (D.f[DIR_M0P])[kw ];//ktw
+ real fTN = (D.f[DIR_0PP])[k ];//ktn
+ real fBS = (D.f[DIR_0MM])[kbs];
+ real fBN = (D.f[DIR_0PM])[kb ];//kbn
+ real fTS = (D.f[DIR_0MP])[ks ];//kts
real fZERO = (D.f[DIR_000])[k ];//kzero
- real fTNE = (D.f[DIR_PPP ])[k ];//ktne
- real fTSW = (D.f[DIR_MMP ])[ksw];//ktsw
- real fTSE = (D.f[DIR_PMP ])[ks ];//ktse
- real fTNW = (D.f[DIR_MPP ])[kw ];//ktnw
- real fBNE = (D.f[DIR_PPM ])[kb ];//kbne
- real fBSW = (D.f[DIR_MMM ])[kbsw];
- real fBSE = (D.f[DIR_PMM ])[kbs];//kbse
- real fBNW = (D.f[DIR_MPM ])[kbw];//kbnw
+ real fTNE = (D.f[DIR_PPP])[k ];//ktne
+ real fTSW = (D.f[DIR_MMP])[ksw];//ktsw
+ real fTSE = (D.f[DIR_PMP])[ks ];//ktse
+ real fTNW = (D.f[DIR_MPP])[kw ];//ktnw
+ real fBNE = (D.f[DIR_PPM])[kb ];//kbne
+ real fBSW = (D.f[DIR_MMM])[kbsw];
+ real fBSE = (D.f[DIR_PMM])[kbs];//kbse
+ real fBNW = (D.f[DIR_MPM])[kbw];//kbnw
////////////////////////////////////////////////////////////////////////////////
real rho0 = (fTNE+fBSW)+(fTSW+fBNE)+(fTSE+fBNW)+(fTNW+fBSE)+(fNE+fSW)+(fNW+fSE)+(fTE+fBW)+(fBE+fTW)+(fTN+fBS)+(fBN+fTS)+(fE+fW)+(fN+fS)+(fT+fB)+fZERO;
real rho = rho0 + c1o1;
@@ -2321,7 +2321,7 @@ __global__ void LB_Kernel_Casc_SP_MS_27( real omega,
unsigned int* neighborY,
unsigned int* neighborZ,
real* DDStart,
- int size_Mat,
+ unsigned long long numberOfLBnodes,
bool EvenOrOdd)
{
////////////////////////////////////////////////////////////////////////////////
@@ -2335,7 +2335,7 @@ __global__ void LB_Kernel_Casc_SP_MS_27( real omega,
const unsigned k = nx*(ny*z + y) + x;
//////////////////////////////////////////////////////////////////////////
- if(k<size_Mat)
+ if(k<numberOfLBnodes)
{
////////////////////////////////////////////////////////////////////////////////
unsigned int BC;
@@ -2346,63 +2346,63 @@ __global__ void LB_Kernel_Casc_SP_MS_27( real omega,
Distributions27 D;
if (EvenOrOdd==true)
{
- D.f[DIR_P00 ] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DDStart[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DDStart[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DDStart[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DDStart[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DDStart[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DDStart[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DDStart[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DDStart[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DDStart[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DDStart[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DDStart[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DDStart[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DDStart[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DDStart[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DDStart[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DDStart[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DDStart[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DDStart[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DDStart[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DDStart[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DDStart[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DDStart[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DDStart[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DDStart[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DDStart[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DDStart[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DDStart[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00 ] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_MMM ] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_PPM ] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_MPM ] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_PMM ] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_MMP ] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_PPP ] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_MPP ] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_PMP ] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_M00] = &DDStart[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DDStart[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DDStart[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DDStart[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DDStart[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DDStart[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DDStart[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DDStart[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DDStart[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DDStart[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DDStart[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DDStart[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DDStart[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DDStart[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DDStart[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DDStart[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DDStart[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DDStart[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DDStart[DIR_000 * numberOfLBnodes];
+ D.f[DIR_MMM] = &DDStart[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DDStart[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DDStart[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DDStart[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DDStart[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PPP] = &DDStart[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DDStart[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DDStart[DIR_MPM * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////
@@ -2435,33 +2435,33 @@ __global__ void LB_Kernel_Casc_SP_MS_27( real omega,
//unsigned int ktne = k;
unsigned int kbsw = neighborZ[ksw];
//////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
- real fE = (D.f[DIR_P00 ])[k ];//ke
- real fW = (D.f[DIR_M00 ])[kw ];
- real fN = (D.f[DIR_0P0 ])[k ];//kn
- real fS = (D.f[DIR_0M0 ])[ks ];
- real fT = (D.f[DIR_00P ])[k ];//kt
- real fB = (D.f[DIR_00M ])[kb ];
- real fNE = (D.f[DIR_PP0 ])[k ];//kne
- real fSW = (D.f[DIR_MM0 ])[ksw];
- real fSE = (D.f[DIR_PM0 ])[ks ];//kse
- real fNW = (D.f[DIR_MP0 ])[kw ];//knw
- real fTE = (D.f[DIR_P0P ])[k ];//kte
- real fBW = (D.f[DIR_M0M ])[kbw];
- real fBE = (D.f[DIR_P0M ])[kb ];//kbe
- real fTW = (D.f[DIR_M0P ])[kw ];//ktw
- real fTN = (D.f[DIR_0PP ])[k ];//ktn
- real fBS = (D.f[DIR_0MM ])[kbs];
- real fBN = (D.f[DIR_0PM ])[kb ];//kbn
- real fTS = (D.f[DIR_0MP ])[ks ];//kts
+ real fE = (D.f[DIR_P00])[k ];//ke
+ real fW = (D.f[DIR_M00])[kw ];
+ real fN = (D.f[DIR_0P0])[k ];//kn
+ real fS = (D.f[DIR_0M0])[ks ];
+ real fT = (D.f[DIR_00P])[k ];//kt
+ real fB = (D.f[DIR_00M])[kb ];
+ real fNE = (D.f[DIR_PP0])[k ];//kne
+ real fSW = (D.f[DIR_MM0])[ksw];
+ real fSE = (D.f[DIR_PM0])[ks ];//kse
+ real fNW = (D.f[DIR_MP0])[kw ];//knw
+ real fTE = (D.f[DIR_P0P])[k ];//kte
+ real fBW = (D.f[DIR_M0M])[kbw];
+ real fBE = (D.f[DIR_P0M])[kb ];//kbe
+ real fTW = (D.f[DIR_M0P])[kw ];//ktw
+ real fTN = (D.f[DIR_0PP])[k ];//ktn
+ real fBS = (D.f[DIR_0MM])[kbs];
+ real fBN = (D.f[DIR_0PM])[kb ];//kbn
+ real fTS = (D.f[DIR_0MP])[ks ];//kts
real fZERO = (D.f[DIR_000])[k ];//kzero
- real fTNE = (D.f[DIR_PPP ])[k ];//ktne
- real fTSW = (D.f[DIR_MMP ])[ksw];//ktsw
- real fTSE = (D.f[DIR_PMP ])[ks ];//ktse
- real fTNW = (D.f[DIR_MPP ])[kw ];//ktnw
- real fBNE = (D.f[DIR_PPM ])[kb ];//kbne
- real fBSW = (D.f[DIR_MMM ])[kbsw];
- real fBSE = (D.f[DIR_PMM ])[kbs];//kbse
- real fBNW = (D.f[DIR_MPM ])[kbw];//kbnw
+ real fTNE = (D.f[DIR_PPP])[k ];//ktne
+ real fTSW = (D.f[DIR_MMP])[ksw];//ktsw
+ real fTSE = (D.f[DIR_PMP])[ks ];//ktse
+ real fTNW = (D.f[DIR_MPP])[kw ];//ktnw
+ real fBNE = (D.f[DIR_PPM])[kb ];//kbne
+ real fBSW = (D.f[DIR_MMM])[kbsw];
+ real fBSE = (D.f[DIR_PMM])[kbs];//kbse
+ real fBNW = (D.f[DIR_MPM])[kbw];//kbnw
////////////////////////////////////////////////////////////////////////////////
real rho0 = fZERO+fE+fW+fN+fS+fT+fB+fNE+fSW+fSE+fNW+fTE+fBW+fBE+fTW+fTN+fBS+fBN+fTS+fTNE+fTSW+fTSE+fTNW+fBNE+fBSW+fBSE+fBNW;
real rho = rho0 + c1o1;
@@ -2846,7 +2846,7 @@ __global__ void LB_Kernel_Casc_SP_MS_Diff_27(real omega,
unsigned int* neighborY,
unsigned int* neighborZ,
real* DDStart,
- int size_Mat,
+ unsigned long long numberOfLBnodes,
bool EvenOrOdd)
{
////////////////////////////////////////////////////////////////////////////////
@@ -2860,7 +2860,7 @@ __global__ void LB_Kernel_Casc_SP_MS_Diff_27(real omega,
const unsigned k = nx*(ny*z + y) + x;
//////////////////////////////////////////////////////////////////////////
- if(k<size_Mat)
+ if(k<numberOfLBnodes)
{
////////////////////////////////////////////////////////////////////////////////
unsigned int BC;
@@ -2871,63 +2871,63 @@ __global__ void LB_Kernel_Casc_SP_MS_Diff_27(real omega,
Distributions27 D;
if (EvenOrOdd==true)
{
- D.f[DIR_P00 ] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DDStart[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DDStart[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DDStart[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DDStart[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DDStart[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DDStart[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DDStart[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DDStart[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DDStart[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DDStart[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DDStart[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DDStart[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DDStart[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DDStart[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DDStart[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DDStart[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DDStart[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DDStart[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DDStart[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DDStart[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DDStart[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DDStart[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DDStart[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DDStart[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DDStart[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DDStart[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DDStart[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00 ] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_MMM ] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_PPM ] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_MPM ] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_PMM ] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_MMP ] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_PPP ] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_MPP ] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_PMP ] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_M00] = &DDStart[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DDStart[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DDStart[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DDStart[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DDStart[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DDStart[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DDStart[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DDStart[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DDStart[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DDStart[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DDStart[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DDStart[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DDStart[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DDStart[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DDStart[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DDStart[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DDStart[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DDStart[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DDStart[DIR_000 * numberOfLBnodes];
+ D.f[DIR_MMM] = &DDStart[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DDStart[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DDStart[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DDStart[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DDStart[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PPP] = &DDStart[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DDStart[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DDStart[DIR_MPM * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////
@@ -2960,33 +2960,33 @@ __global__ void LB_Kernel_Casc_SP_MS_Diff_27(real omega,
//unsigned int ktne = k;
unsigned int kbsw = neighborZ[ksw];
//////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
- real fE = (D.f[DIR_P00 ])[k ];//ke
- real fW = (D.f[DIR_M00 ])[kw ];
- real fN = (D.f[DIR_0P0 ])[k ];//kn
- real fS = (D.f[DIR_0M0 ])[ks ];
- real fT = (D.f[DIR_00P ])[k ];//kt
- real fB = (D.f[DIR_00M ])[kb ];
- real fNE = (D.f[DIR_PP0 ])[k ];//kne
- real fSW = (D.f[DIR_MM0 ])[ksw];
- real fSE = (D.f[DIR_PM0 ])[ks ];//kse
- real fNW = (D.f[DIR_MP0 ])[kw ];//knw
- real fTE = (D.f[DIR_P0P ])[k ];//kte
- real fBW = (D.f[DIR_M0M ])[kbw];
- real fBE = (D.f[DIR_P0M ])[kb ];//kbe
- real fTW = (D.f[DIR_M0P ])[kw ];//ktw
- real fTN = (D.f[DIR_0PP ])[k ];//ktn
- real fBS = (D.f[DIR_0MM ])[kbs];
- real fBN = (D.f[DIR_0PM ])[kb ];//kbn
- real fTS = (D.f[DIR_0MP ])[ks ];//kts
+ real fE = (D.f[DIR_P00])[k ];//ke
+ real fW = (D.f[DIR_M00])[kw ];
+ real fN = (D.f[DIR_0P0])[k ];//kn
+ real fS = (D.f[DIR_0M0])[ks ];
+ real fT = (D.f[DIR_00P])[k ];//kt
+ real fB = (D.f[DIR_00M])[kb ];
+ real fNE = (D.f[DIR_PP0])[k ];//kne
+ real fSW = (D.f[DIR_MM0])[ksw];
+ real fSE = (D.f[DIR_PM0])[ks ];//kse
+ real fNW = (D.f[DIR_MP0])[kw ];//knw
+ real fTE = (D.f[DIR_P0P])[k ];//kte
+ real fBW = (D.f[DIR_M0M])[kbw];
+ real fBE = (D.f[DIR_P0M])[kb ];//kbe
+ real fTW = (D.f[DIR_M0P])[kw ];//ktw
+ real fTN = (D.f[DIR_0PP])[k ];//ktn
+ real fBS = (D.f[DIR_0MM])[kbs];
+ real fBN = (D.f[DIR_0PM])[kb ];//kbn
+ real fTS = (D.f[DIR_0MP])[ks ];//kts
real fZERO = (D.f[DIR_000])[k ];//kzero
- real fTNE = (D.f[DIR_PPP ])[k ];//ktne
- real fTSW = (D.f[DIR_MMP ])[ksw];//ktsw
- real fTSE = (D.f[DIR_PMP ])[ks ];//ktse
- real fTNW = (D.f[DIR_MPP ])[kw ];//ktnw
- real fBNE = (D.f[DIR_PPM ])[kb ];//kbne
- real fBSW = (D.f[DIR_MMM ])[kbsw];
- real fBSE = (D.f[DIR_PMM ])[kbs];//kbse
- real fBNW = (D.f[DIR_MPM ])[kbw];//kbnw
+ real fTNE = (D.f[DIR_PPP])[k ];//ktne
+ real fTSW = (D.f[DIR_MMP])[ksw];//ktsw
+ real fTSE = (D.f[DIR_PMP])[ks ];//ktse
+ real fTNW = (D.f[DIR_MPP])[kw ];//ktnw
+ real fBNE = (D.f[DIR_PPM])[kb ];//kbne
+ real fBSW = (D.f[DIR_MMM])[kbsw];
+ real fBSE = (D.f[DIR_PMM])[kbs];//kbse
+ real fBNW = (D.f[DIR_MPM])[kbw];//kbnw
////////////////////////////////////////////////////////////////////////////////
real rho0 = fZERO+fE+fW+fN+fS+fT+fB+fNE+fSW+fSE+fNW+fTE+fBW+fBE+fTW+fTN+fBS+fBN+fTS+fTNE+fTSW+fTSE+fTNW+fBNE+fBSW+fBSE+fBNW;
real rho = rho0 + c1o1;
@@ -3368,7 +3368,7 @@ __global__ void LB_Kernel_Casc_SP_27( real omega,
unsigned int* neighborY,
unsigned int* neighborZ,
real* DDStart,
- int size_Mat,
+ unsigned long long numberOfLBnodes,
bool EvenOrOdd)
{
////////////////////////////////////////////////////////////////////////////////
@@ -3382,7 +3382,7 @@ __global__ void LB_Kernel_Casc_SP_27( real omega,
const unsigned k = nx*(ny*z + y) + x;
//////////////////////////////////////////////////////////////////////////
- if(k<size_Mat)
+ if(k<numberOfLBnodes)
{
////////////////////////////////////////////////////////////////////////////////
unsigned int BC;
@@ -3393,63 +3393,63 @@ __global__ void LB_Kernel_Casc_SP_27( real omega,
Distributions27 D;
if (EvenOrOdd==true)
{
- D.f[DIR_P00 ] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DDStart[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DDStart[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DDStart[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DDStart[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DDStart[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DDStart[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DDStart[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DDStart[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DDStart[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DDStart[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DDStart[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DDStart[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DDStart[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DDStart[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DDStart[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DDStart[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DDStart[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DDStart[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DDStart[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DDStart[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DDStart[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DDStart[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DDStart[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DDStart[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DDStart[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DDStart[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DDStart[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00 ] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_MMM ] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_PPM ] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_MPM ] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_PMM ] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_MMP ] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_PPP ] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_MPP ] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_PMP ] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_M00] = &DDStart[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DDStart[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DDStart[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DDStart[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DDStart[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DDStart[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DDStart[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DDStart[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DDStart[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DDStart[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DDStart[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DDStart[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DDStart[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DDStart[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DDStart[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DDStart[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DDStart[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DDStart[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DDStart[DIR_000 * numberOfLBnodes];
+ D.f[DIR_MMM] = &DDStart[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DDStart[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DDStart[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DDStart[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DDStart[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PPP] = &DDStart[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DDStart[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DDStart[DIR_MPM * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////
@@ -3512,33 +3512,33 @@ __global__ void LB_Kernel_Casc_SP_27( real omega,
//////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
real f_E,f_W,f_N,f_S,f_T,f_B,f_NE,f_SW,f_SE,f_NW,f_TE,f_BW,f_BE,f_TW,f_TN,f_BS,f_BN,f_TS,f_ZERO, f_TNE,f_TNW,f_TSE,f_TSW, f_BNE,f_BNW,f_BSE,f_BSW;
//////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
- f_E = (D.f[DIR_P00 ])[ke]+c2o27;
- f_W = (D.f[DIR_M00 ])[kw]+c2o27;
- f_N = (D.f[DIR_0P0 ])[kn]+c2o27;
- f_S = (D.f[DIR_0M0 ])[ks]+c2o27;
- f_T = (D.f[DIR_00P ])[kt]+c2o27;
- f_B = (D.f[DIR_00M ])[kb]+c2o27;
- f_NE = (D.f[DIR_PP0 ])[kne]+c1o54;
- f_SW = (D.f[DIR_MM0 ])[ksw]+c1o54;
- f_SE = (D.f[DIR_PM0 ])[kse]+c1o54;
- f_NW = (D.f[DIR_MP0 ])[knw]+c1o54;
- f_TE = (D.f[DIR_P0P ])[kte]+c1o54;
- f_BW = (D.f[DIR_M0M ])[kbw]+c1o54;
- f_BE = (D.f[DIR_P0M ])[kbe]+c1o54;
- f_TW = (D.f[DIR_M0P ])[ktw]+c1o54;
- f_TN = (D.f[DIR_0PP ])[ktn]+c1o54;
- f_BS = (D.f[DIR_0MM ])[kbs]+c1o54;
- f_BN = (D.f[DIR_0PM ])[kbn]+c1o54;
- f_TS = (D.f[DIR_0MP ])[kts]+c1o54;
+ f_E = (D.f[DIR_P00])[ke]+c2o27;
+ f_W = (D.f[DIR_M00])[kw]+c2o27;
+ f_N = (D.f[DIR_0P0])[kn]+c2o27;
+ f_S = (D.f[DIR_0M0])[ks]+c2o27;
+ f_T = (D.f[DIR_00P])[kt]+c2o27;
+ f_B = (D.f[DIR_00M])[kb]+c2o27;
+ f_NE = (D.f[DIR_PP0])[kne]+c1o54;
+ f_SW = (D.f[DIR_MM0])[ksw]+c1o54;
+ f_SE = (D.f[DIR_PM0])[kse]+c1o54;
+ f_NW = (D.f[DIR_MP0])[knw]+c1o54;
+ f_TE = (D.f[DIR_P0P])[kte]+c1o54;
+ f_BW = (D.f[DIR_M0M])[kbw]+c1o54;
+ f_BE = (D.f[DIR_P0M])[kbe]+c1o54;
+ f_TW = (D.f[DIR_M0P])[ktw]+c1o54;
+ f_TN = (D.f[DIR_0PP])[ktn]+c1o54;
+ f_BS = (D.f[DIR_0MM])[kbs]+c1o54;
+ f_BN = (D.f[DIR_0PM])[kbn]+c1o54;
+ f_TS = (D.f[DIR_0MP])[kts]+c1o54;
f_ZERO = (D.f[DIR_000])[kzero]+c8o27;
- f_TNE = (D.f[DIR_PPP ])[ktne]+c1o216;
- f_TSW = (D.f[DIR_MMP ])[ktsw]+c1o216;
- f_TSE = (D.f[DIR_PMP ])[ktse]+c1o216;
- f_TNW = (D.f[DIR_MPP ])[ktnw]+c1o216;
- f_BNE = (D.f[DIR_PPM ])[kbne]+c1o216;
- f_BSW = (D.f[DIR_MMM ])[kbsw]+c1o216;
- f_BSE = (D.f[DIR_PMM ])[kbse]+c1o216;
- f_BNW = (D.f[DIR_MPM ])[kbnw]+c1o216;
+ f_TNE = (D.f[DIR_PPP])[ktne]+c1o216;
+ f_TSW = (D.f[DIR_MMP])[ktsw]+c1o216;
+ f_TSE = (D.f[DIR_PMP])[ktse]+c1o216;
+ f_TNW = (D.f[DIR_MPP])[ktnw]+c1o216;
+ f_BNE = (D.f[DIR_PPM])[kbne]+c1o216;
+ f_BSW = (D.f[DIR_MMM])[kbsw]+c1o216;
+ f_BSE = (D.f[DIR_PMM])[kbse]+c1o216;
+ f_BNW = (D.f[DIR_MPM])[kbnw]+c1o216;
////////////////////////////////////////////////////////////////////////////////
if( BC == GEO_FLUID || BC == GEO_VELO)
@@ -4060,7 +4060,7 @@ __global__ void LB_Kernel_Casc27(real omega,
unsigned int* neighborY,
unsigned int* neighborZ,
real* DDStart,
- int size_Mat,
+ unsigned long long numberOfLBnodes,
bool EvenOrOdd)
{
////////////////////////////////////////////////////////////////////////////////
@@ -4089,63 +4089,63 @@ __global__ void LB_Kernel_Casc27(real omega,
Distributions27 D;
if (EvenOrOdd==true)
{
- D.f[DIR_P00 ] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DDStart[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DDStart[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DDStart[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DDStart[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DDStart[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DDStart[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DDStart[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DDStart[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DDStart[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DDStart[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DDStart[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DDStart[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DDStart[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DDStart[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DDStart[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DDStart[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DDStart[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DDStart[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DDStart[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DDStart[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DDStart[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DDStart[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DDStart[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DDStart[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DDStart[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DDStart[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DDStart[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00 ] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_MMM ] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_PPM ] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_MPM ] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_PMM ] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_MMP ] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_PPP ] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_MPP ] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_PMP ] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_M00] = &DDStart[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DDStart[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DDStart[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DDStart[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DDStart[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DDStart[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DDStart[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DDStart[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DDStart[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DDStart[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DDStart[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DDStart[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DDStart[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DDStart[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DDStart[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DDStart[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DDStart[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DDStart[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DDStart[DIR_000 * numberOfLBnodes];
+ D.f[DIR_MMM] = &DDStart[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DDStart[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DDStart[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DDStart[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DDStart[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PPP] = &DDStart[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DDStart[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DDStart[DIR_MPM * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////
@@ -4208,33 +4208,33 @@ __global__ void LB_Kernel_Casc27(real omega,
//////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
real f_E,f_W,f_N,f_S,f_T,f_B,f_NE,f_SW,f_SE,f_NW,f_TE,f_BW,f_BE,f_TW,f_TN,f_BS,f_BN,f_TS,f_ZERO, f_TNE,f_TNW,f_TSE,f_TSW, f_BNE,f_BNW,f_BSE,f_BSW;
//////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
- f_E = (D.f[DIR_P00 ])[ke]+c2o27;
- f_W = (D.f[DIR_M00 ])[kw]+c2o27;
- f_N = (D.f[DIR_0P0 ])[kn]+c2o27;
- f_S = (D.f[DIR_0M0 ])[ks]+c2o27;
- f_T = (D.f[DIR_00P ])[kt]+c2o27;
- f_B = (D.f[DIR_00M ])[kb]+c2o27;
- f_NE = (D.f[DIR_PP0 ])[kne]+c1o54;
- f_SW = (D.f[DIR_MM0 ])[ksw]+c1o54;
- f_SE = (D.f[DIR_PM0 ])[kse]+c1o54;
- f_NW = (D.f[DIR_MP0 ])[knw]+c1o54;
- f_TE = (D.f[DIR_P0P ])[kte]+c1o54;
- f_BW = (D.f[DIR_M0M ])[kbw]+c1o54;
- f_BE = (D.f[DIR_P0M ])[kbe]+c1o54;
- f_TW = (D.f[DIR_M0P ])[ktw]+c1o54;
- f_TN = (D.f[DIR_0PP ])[ktn]+c1o54;
- f_BS = (D.f[DIR_0MM ])[kbs]+c1o54;
- f_BN = (D.f[DIR_0PM ])[kbn]+c1o54;
- f_TS = (D.f[DIR_0MP ])[kts]+c1o54;
+ f_E = (D.f[DIR_P00])[ke]+c2o27;
+ f_W = (D.f[DIR_M00])[kw]+c2o27;
+ f_N = (D.f[DIR_0P0])[kn]+c2o27;
+ f_S = (D.f[DIR_0M0])[ks]+c2o27;
+ f_T = (D.f[DIR_00P])[kt]+c2o27;
+ f_B = (D.f[DIR_00M])[kb]+c2o27;
+ f_NE = (D.f[DIR_PP0])[kne]+c1o54;
+ f_SW = (D.f[DIR_MM0])[ksw]+c1o54;
+ f_SE = (D.f[DIR_PM0])[kse]+c1o54;
+ f_NW = (D.f[DIR_MP0])[knw]+c1o54;
+ f_TE = (D.f[DIR_P0P])[kte]+c1o54;
+ f_BW = (D.f[DIR_M0M])[kbw]+c1o54;
+ f_BE = (D.f[DIR_P0M])[kbe]+c1o54;
+ f_TW = (D.f[DIR_M0P])[ktw]+c1o54;
+ f_TN = (D.f[DIR_0PP])[ktn]+c1o54;
+ f_BS = (D.f[DIR_0MM])[kbs]+c1o54;
+ f_BN = (D.f[DIR_0PM])[kbn]+c1o54;
+ f_TS = (D.f[DIR_0MP])[kts]+c1o54;
f_ZERO = (D.f[DIR_000])[kzero]+c8o27;
- f_TNE = (D.f[DIR_PPP ])[ktne]+c1o216;
- f_TSW = (D.f[DIR_MMP ])[ktsw]+c1o216;
- f_TSE = (D.f[DIR_PMP ])[ktse]+c1o216;
- f_TNW = (D.f[DIR_MPP ])[ktnw]+c1o216;
- f_BNE = (D.f[DIR_PPM ])[kbne]+c1o216;
- f_BSW = (D.f[DIR_MMM ])[kbsw]+c1o216;
- f_BSE = (D.f[DIR_PMM ])[kbse]+c1o216;
- f_BNW = (D.f[DIR_MPM ])[kbnw]+c1o216;
+ f_TNE = (D.f[DIR_PPP])[ktne]+c1o216;
+ f_TSW = (D.f[DIR_MMP])[ktsw]+c1o216;
+ f_TSE = (D.f[DIR_PMP])[ktse]+c1o216;
+ f_TNW = (D.f[DIR_MPP])[ktnw]+c1o216;
+ f_BNE = (D.f[DIR_PPM])[kbne]+c1o216;
+ f_BSW = (D.f[DIR_MMM])[kbsw]+c1o216;
+ f_BSE = (D.f[DIR_PMM])[kbse]+c1o216;
+ f_BNW = (D.f[DIR_MPM])[kbnw]+c1o216;
////////////////////////////////////////////////////////////////////////////////
if( BC == GEO_FLUID || BC == GEO_VELO)
diff --git a/src/gpu/VirtualFluids_GPU/GPU/CudaMemoryManager.cpp b/src/gpu/VirtualFluids_GPU/GPU/CudaMemoryManager.cpp
index 22192216927f91c33fafc23c54c3fae334abdd34..9fd2a6b2f5c5c10a36856852db47f989ace714ce 100644
--- a/src/gpu/VirtualFluids_GPU/GPU/CudaMemoryManager.cpp
+++ b/src/gpu/VirtualFluids_GPU/GPU/CudaMemoryManager.cpp
@@ -5,72 +5,65 @@
#include <math.h>
#include <Parameter/Parameter.h>
+
#include "Parameter/CudaStreamManager.h"
-#include "PreCollisionInteractor/ActuatorLine.h"
+#include "PreCollisionInteractor/ActuatorFarm.h"
#include "PreCollisionInteractor/Probes/Probe.h"
+#include <PreCollisionInteractor/PrecursorWriter.h>
#include "Calculation/PorousMedia.h"
#include "lbm/constants/NumericConstants.h"
-void CudaMemoryManager::cudaAllocFull(int lev)
-{
- checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->geo ), parameter->getParH(lev)->mem_size_int ));
- checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->k ), parameter->getParH(lev)->mem_size_int ));
-}
-void CudaMemoryManager::cudaFreeFull(int lev)
-{
- checkCudaErrors( cudaFreeHost(parameter->getParH(lev)->geo ));
- checkCudaErrors( cudaFreeHost(parameter->getParH(lev)->k ));
-}
+
void CudaMemoryManager::cudaCopyPrint(int lev)
{
- checkCudaErrors( cudaMemcpy(parameter->getParH(lev)->velocityX , parameter->getParD(lev)->velocityX , parameter->getParH(lev)->mem_size_real_SP , cudaMemcpyDeviceToHost));
- checkCudaErrors( cudaMemcpy(parameter->getParH(lev)->velocityY , parameter->getParD(lev)->velocityY , parameter->getParH(lev)->mem_size_real_SP , cudaMemcpyDeviceToHost));
- checkCudaErrors( cudaMemcpy(parameter->getParH(lev)->velocityZ , parameter->getParD(lev)->velocityZ , parameter->getParH(lev)->mem_size_real_SP , cudaMemcpyDeviceToHost));
- checkCudaErrors( cudaMemcpy(parameter->getParH(lev)->rho , parameter->getParD(lev)->rho , parameter->getParH(lev)->mem_size_real_SP , cudaMemcpyDeviceToHost));
- checkCudaErrors( cudaMemcpy(parameter->getParH(lev)->pressure, parameter->getParD(lev)->pressure, parameter->getParH(lev)->mem_size_real_SP , cudaMemcpyDeviceToHost));
+ checkCudaErrors( cudaMemcpy(parameter->getParH(lev)->velocityX , parameter->getParD(lev)->velocityX , parameter->getParH(lev)->memSizeRealLBnodes , cudaMemcpyDeviceToHost));
+ checkCudaErrors( cudaMemcpy(parameter->getParH(lev)->velocityY , parameter->getParD(lev)->velocityY , parameter->getParH(lev)->memSizeRealLBnodes , cudaMemcpyDeviceToHost));
+ checkCudaErrors( cudaMemcpy(parameter->getParH(lev)->velocityZ , parameter->getParD(lev)->velocityZ , parameter->getParH(lev)->memSizeRealLBnodes , cudaMemcpyDeviceToHost));
+ checkCudaErrors( cudaMemcpy(parameter->getParH(lev)->rho , parameter->getParD(lev)->rho , parameter->getParH(lev)->memSizeRealLBnodes , cudaMemcpyDeviceToHost));
+ checkCudaErrors( cudaMemcpy(parameter->getParH(lev)->pressure , parameter->getParD(lev)->pressure , parameter->getParH(lev)->memSizeRealLBnodes , cudaMemcpyDeviceToHost));
if(parameter->getIsBodyForce())
{
- checkCudaErrors( cudaMemcpy(parameter->getParH(lev)->forceX_SP , parameter->getParD(lev)->forceX_SP , parameter->getParH(lev)->mem_size_real_SP , cudaMemcpyDeviceToHost));
- checkCudaErrors( cudaMemcpy(parameter->getParH(lev)->forceY_SP , parameter->getParD(lev)->forceY_SP , parameter->getParH(lev)->mem_size_real_SP , cudaMemcpyDeviceToHost));
- checkCudaErrors( cudaMemcpy(parameter->getParH(lev)->forceZ_SP , parameter->getParD(lev)->forceZ_SP , parameter->getParH(lev)->mem_size_real_SP , cudaMemcpyDeviceToHost));
+ checkCudaErrors( cudaMemcpy(parameter->getParH(lev)->forceX_SP , parameter->getParD(lev)->forceX_SP , parameter->getParH(lev)->memSizeRealLBnodes , cudaMemcpyDeviceToHost));
+ checkCudaErrors( cudaMemcpy(parameter->getParH(lev)->forceY_SP , parameter->getParD(lev)->forceY_SP , parameter->getParH(lev)->memSizeRealLBnodes , cudaMemcpyDeviceToHost));
+ checkCudaErrors( cudaMemcpy(parameter->getParH(lev)->forceZ_SP , parameter->getParD(lev)->forceZ_SP , parameter->getParH(lev)->memSizeRealLBnodes , cudaMemcpyDeviceToHost));
}
if(parameter->getUseTurbulentViscosity())
{
- checkCudaErrors( cudaMemcpy(parameter->getParH(lev)->turbViscosity , parameter->getParD(lev)->turbViscosity , parameter->getParH(lev)->mem_size_real_SP , cudaMemcpyDeviceToHost));
+ checkCudaErrors( cudaMemcpy(parameter->getParH(lev)->turbViscosity , parameter->getParD(lev)->turbViscosity , parameter->getParH(lev)->memSizeRealLBnodes , cudaMemcpyDeviceToHost));
}
}
void CudaMemoryManager::cudaCopyMedianPrint(int lev)
{
- checkCudaErrors( cudaMemcpy(parameter->getParH(lev)->vx_SP_Med , parameter->getParD(lev)->vx_SP_Med , parameter->getParH(lev)->mem_size_real_SP , cudaMemcpyDeviceToHost));
- checkCudaErrors( cudaMemcpy(parameter->getParH(lev)->vy_SP_Med , parameter->getParD(lev)->vy_SP_Med , parameter->getParH(lev)->mem_size_real_SP , cudaMemcpyDeviceToHost));
- checkCudaErrors( cudaMemcpy(parameter->getParH(lev)->vz_SP_Med , parameter->getParD(lev)->vz_SP_Med , parameter->getParH(lev)->mem_size_real_SP , cudaMemcpyDeviceToHost));
- checkCudaErrors( cudaMemcpy(parameter->getParH(lev)->rho_SP_Med , parameter->getParD(lev)->rho_SP_Med , parameter->getParH(lev)->mem_size_real_SP , cudaMemcpyDeviceToHost));
- checkCudaErrors( cudaMemcpy(parameter->getParH(lev)->press_SP_Med, parameter->getParD(lev)->press_SP_Med, parameter->getParH(lev)->mem_size_real_SP , cudaMemcpyDeviceToHost));
+ checkCudaErrors( cudaMemcpy(parameter->getParH(lev)->vx_SP_Med , parameter->getParD(lev)->vx_SP_Med , parameter->getParH(lev)->memSizeRealLBnodes , cudaMemcpyDeviceToHost));
+ checkCudaErrors( cudaMemcpy(parameter->getParH(lev)->vy_SP_Med , parameter->getParD(lev)->vy_SP_Med , parameter->getParH(lev)->memSizeRealLBnodes , cudaMemcpyDeviceToHost));
+ checkCudaErrors( cudaMemcpy(parameter->getParH(lev)->vz_SP_Med , parameter->getParD(lev)->vz_SP_Med , parameter->getParH(lev)->memSizeRealLBnodes , cudaMemcpyDeviceToHost));
+ checkCudaErrors( cudaMemcpy(parameter->getParH(lev)->rho_SP_Med , parameter->getParD(lev)->rho_SP_Med , parameter->getParH(lev)->memSizeRealLBnodes , cudaMemcpyDeviceToHost));
+ checkCudaErrors( cudaMemcpy(parameter->getParH(lev)->press_SP_Med, parameter->getParD(lev)->press_SP_Med, parameter->getParH(lev)->memSizeRealLBnodes , cudaMemcpyDeviceToHost));
}
void CudaMemoryManager::cudaAllocCoord(int lev)
{
//Host
- checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->coordinateX ), parameter->getParH(lev)->mem_size_real_SP ));
- checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->coordinateY ), parameter->getParH(lev)->mem_size_real_SP ));
- checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->coordinateZ ), parameter->getParH(lev)->mem_size_real_SP ));
+ checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->coordinateX ), parameter->getParH(lev)->memSizeRealLBnodes ));
+ checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->coordinateY ), parameter->getParH(lev)->memSizeRealLBnodes ));
+ checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->coordinateZ ), parameter->getParH(lev)->memSizeRealLBnodes ));
//Device (spinning ship + uppsala)
- checkCudaErrors( cudaMalloc((void**) &(parameter->getParD(lev)->coordinateX ), parameter->getParH(lev)->mem_size_real_SP ));
- checkCudaErrors( cudaMalloc((void**) &(parameter->getParD(lev)->coordinateY ), parameter->getParH(lev)->mem_size_real_SP ));
- checkCudaErrors( cudaMalloc((void**) &(parameter->getParD(lev)->coordinateZ ), parameter->getParH(lev)->mem_size_real_SP ));
+ checkCudaErrors( cudaMalloc((void**) &(parameter->getParD(lev)->coordinateX ), parameter->getParH(lev)->memSizeRealLBnodes ));
+ checkCudaErrors( cudaMalloc((void**) &(parameter->getParD(lev)->coordinateY ), parameter->getParH(lev)->memSizeRealLBnodes ));
+ checkCudaErrors( cudaMalloc((void**) &(parameter->getParD(lev)->coordinateZ ), parameter->getParH(lev)->memSizeRealLBnodes ));
//////////////////////////////////////////////////////////////////////////
- double tmp = 3. * (double)parameter->getParH(lev)->mem_size_real_SP;
+ double tmp = 3. * (double)parameter->getParH(lev)->memSizeRealLBnodes;
setMemsizeGPU(tmp, false);
}
void CudaMemoryManager::cudaCopyCoord(int lev)
{
//copy host to device
- checkCudaErrors( cudaMemcpy(parameter->getParD(lev)->coordinateX, parameter->getParH(lev)->coordinateX, parameter->getParH(lev)->mem_size_real_SP , cudaMemcpyHostToDevice));
- checkCudaErrors( cudaMemcpy(parameter->getParD(lev)->coordinateY, parameter->getParH(lev)->coordinateY, parameter->getParH(lev)->mem_size_real_SP , cudaMemcpyHostToDevice));
- checkCudaErrors( cudaMemcpy(parameter->getParD(lev)->coordinateZ, parameter->getParH(lev)->coordinateZ, parameter->getParH(lev)->mem_size_real_SP , cudaMemcpyHostToDevice));
+ checkCudaErrors( cudaMemcpy(parameter->getParD(lev)->coordinateX, parameter->getParH(lev)->coordinateX, parameter->getParH(lev)->memSizeRealLBnodes , cudaMemcpyHostToDevice));
+ checkCudaErrors( cudaMemcpy(parameter->getParD(lev)->coordinateY, parameter->getParH(lev)->coordinateY, parameter->getParH(lev)->memSizeRealLBnodes , cudaMemcpyHostToDevice));
+ checkCudaErrors( cudaMemcpy(parameter->getParD(lev)->coordinateZ, parameter->getParH(lev)->coordinateZ, parameter->getParH(lev)->memSizeRealLBnodes , cudaMemcpyHostToDevice));
}
void CudaMemoryManager::cudaFreeCoord(int lev)
{
@@ -81,24 +74,24 @@ void CudaMemoryManager::cudaFreeCoord(int lev)
void CudaMemoryManager::cudaAllocBodyForce(int lev)
{
//Host
- checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->forceX_SP ), parameter->getParH(lev)->mem_size_real_SP ));
- checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->forceY_SP ), parameter->getParH(lev)->mem_size_real_SP ));
- checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->forceZ_SP ), parameter->getParH(lev)->mem_size_real_SP ));
+ checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->forceX_SP ), parameter->getParH(lev)->memSizeRealLBnodes ));
+ checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->forceY_SP ), parameter->getParH(lev)->memSizeRealLBnodes ));
+ checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->forceZ_SP ), parameter->getParH(lev)->memSizeRealLBnodes ));
//Device
- checkCudaErrors( cudaMalloc((void**) &(parameter->getParD(lev)->forceX_SP ), parameter->getParH(lev)->mem_size_real_SP ));
- checkCudaErrors( cudaMalloc((void**) &(parameter->getParD(lev)->forceY_SP ), parameter->getParH(lev)->mem_size_real_SP ));
- checkCudaErrors( cudaMalloc((void**) &(parameter->getParD(lev)->forceZ_SP ), parameter->getParH(lev)->mem_size_real_SP ));
+ checkCudaErrors( cudaMalloc((void**) &(parameter->getParD(lev)->forceX_SP ), parameter->getParH(lev)->memSizeRealLBnodes ));
+ checkCudaErrors( cudaMalloc((void**) &(parameter->getParD(lev)->forceY_SP ), parameter->getParH(lev)->memSizeRealLBnodes ));
+ checkCudaErrors( cudaMalloc((void**) &(parameter->getParD(lev)->forceZ_SP ), parameter->getParH(lev)->memSizeRealLBnodes ));
//////////////////////////////////////////////////////////////////////////
- double tmp = 3. * (double)parameter->getParH(lev)->mem_size_real_SP;
+ double tmp = 3. * (double)parameter->getParH(lev)->memSizeRealLBnodes;
setMemsizeGPU(tmp, false);
}
void CudaMemoryManager::cudaCopyBodyForce(int lev)
{
//copy host to device
- checkCudaErrors( cudaMemcpy(parameter->getParD(lev)->forceX_SP, parameter->getParH(lev)->forceX_SP, parameter->getParH(lev)->mem_size_real_SP , cudaMemcpyHostToDevice));
- checkCudaErrors( cudaMemcpy(parameter->getParD(lev)->forceY_SP, parameter->getParH(lev)->forceY_SP, parameter->getParH(lev)->mem_size_real_SP , cudaMemcpyHostToDevice));
- checkCudaErrors( cudaMemcpy(parameter->getParD(lev)->forceZ_SP, parameter->getParH(lev)->forceZ_SP, parameter->getParH(lev)->mem_size_real_SP , cudaMemcpyHostToDevice));
+ checkCudaErrors( cudaMemcpy(parameter->getParD(lev)->forceX_SP, parameter->getParH(lev)->forceX_SP, parameter->getParH(lev)->memSizeRealLBnodes , cudaMemcpyHostToDevice));
+ checkCudaErrors( cudaMemcpy(parameter->getParD(lev)->forceY_SP, parameter->getParH(lev)->forceY_SP, parameter->getParH(lev)->memSizeRealLBnodes , cudaMemcpyHostToDevice));
+ checkCudaErrors( cudaMemcpy(parameter->getParD(lev)->forceZ_SP, parameter->getParH(lev)->forceZ_SP, parameter->getParH(lev)->memSizeRealLBnodes , cudaMemcpyHostToDevice));
}
void CudaMemoryManager::cudaFreeBodyForce(int lev)
@@ -111,71 +104,71 @@ void CudaMemoryManager::cudaFreeBodyForce(int lev)
//print
void CudaMemoryManager::cudaCopyDataToHost(int lev)
{
- checkCudaErrors( cudaMemcpy(parameter->getParH(lev)->velocityX , parameter->getParD(lev)->velocityX , parameter->getParH(lev)->mem_size_real_SP , cudaMemcpyDeviceToHost));
- checkCudaErrors( cudaMemcpy(parameter->getParH(lev)->velocityY , parameter->getParD(lev)->velocityY , parameter->getParH(lev)->mem_size_real_SP , cudaMemcpyDeviceToHost));
- checkCudaErrors( cudaMemcpy(parameter->getParH(lev)->velocityZ , parameter->getParD(lev)->velocityZ , parameter->getParH(lev)->mem_size_real_SP , cudaMemcpyDeviceToHost));
- checkCudaErrors( cudaMemcpy(parameter->getParH(lev)->rho , parameter->getParD(lev)->rho , parameter->getParH(lev)->mem_size_real_SP , cudaMemcpyDeviceToHost));
- checkCudaErrors( cudaMemcpy(parameter->getParH(lev)->pressure, parameter->getParD(lev)->pressure, parameter->getParH(lev)->mem_size_real_SP , cudaMemcpyDeviceToHost));
+ checkCudaErrors( cudaMemcpy(parameter->getParH(lev)->velocityX , parameter->getParD(lev)->velocityX , parameter->getParH(lev)->memSizeRealLBnodes , cudaMemcpyDeviceToHost));
+ checkCudaErrors( cudaMemcpy(parameter->getParH(lev)->velocityY , parameter->getParD(lev)->velocityY , parameter->getParH(lev)->memSizeRealLBnodes , cudaMemcpyDeviceToHost));
+ checkCudaErrors( cudaMemcpy(parameter->getParH(lev)->velocityZ , parameter->getParD(lev)->velocityZ , parameter->getParH(lev)->memSizeRealLBnodes , cudaMemcpyDeviceToHost));
+ checkCudaErrors( cudaMemcpy(parameter->getParH(lev)->rho , parameter->getParD(lev)->rho , parameter->getParH(lev)->memSizeRealLBnodes , cudaMemcpyDeviceToHost));
+ checkCudaErrors( cudaMemcpy(parameter->getParH(lev)->pressure , parameter->getParD(lev)->pressure , parameter->getParH(lev)->memSizeRealLBnodes , cudaMemcpyDeviceToHost));
}
//sparse
void CudaMemoryManager::cudaAllocSP(int lev)
{
//Host
- checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->typeOfGridNode ), parameter->getParH(lev)->mem_size_int_SP ));
- checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->neighborX ), parameter->getParH(lev)->mem_size_int_SP ));
- checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->neighborY ), parameter->getParH(lev)->mem_size_int_SP ));
- checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->neighborZ ), parameter->getParH(lev)->mem_size_int_SP ));
- checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->rho ), parameter->getParH(lev)->mem_size_real_SP));
- checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->velocityX ), parameter->getParH(lev)->mem_size_real_SP));
- checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->velocityY ), parameter->getParH(lev)->mem_size_real_SP));
- checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->velocityZ ), parameter->getParH(lev)->mem_size_real_SP));
- checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->pressure ), parameter->getParH(lev)->mem_size_real_SP));
+ checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->typeOfGridNode), parameter->getParH(lev)->memSizeLonglongLBnodes));
+ checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->neighborX ), parameter->getParH(lev)->memSizeLonglongLBnodes));
+ checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->neighborY ), parameter->getParH(lev)->memSizeLonglongLBnodes));
+ checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->neighborZ ), parameter->getParH(lev)->memSizeLonglongLBnodes));
+ checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->rho ), parameter->getParH(lev)->memSizeRealLBnodes ));
+ checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->velocityX ), parameter->getParH(lev)->memSizeRealLBnodes ));
+ checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->velocityY ), parameter->getParH(lev)->memSizeRealLBnodes ));
+ checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->velocityZ ), parameter->getParH(lev)->memSizeRealLBnodes ));
+ checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->pressure ), parameter->getParH(lev)->memSizeRealLBnodes ));
//Device
- checkCudaErrors( cudaMalloc((void**) &(parameter->getParD(lev)->typeOfGridNode ), parameter->getParD(lev)->mem_size_int_SP ));
- checkCudaErrors( cudaMalloc((void**) &(parameter->getParD(lev)->neighborX ), parameter->getParD(lev)->mem_size_int_SP ));
- checkCudaErrors( cudaMalloc((void**) &(parameter->getParD(lev)->neighborY ), parameter->getParD(lev)->mem_size_int_SP ));
- checkCudaErrors( cudaMalloc((void**) &(parameter->getParD(lev)->neighborZ ), parameter->getParD(lev)->mem_size_int_SP ));
- checkCudaErrors( cudaMalloc((void**) &(parameter->getParD(lev)->rho ), parameter->getParD(lev)->mem_size_real_SP));
- checkCudaErrors( cudaMalloc((void**) &(parameter->getParD(lev)->velocityX ), parameter->getParD(lev)->mem_size_real_SP));
- checkCudaErrors( cudaMalloc((void**) &(parameter->getParD(lev)->velocityY ), parameter->getParD(lev)->mem_size_real_SP));
- checkCudaErrors( cudaMalloc((void**) &(parameter->getParD(lev)->velocityZ ), parameter->getParD(lev)->mem_size_real_SP));
- checkCudaErrors( cudaMalloc((void**) &(parameter->getParD(lev)->pressure ), parameter->getParD(lev)->mem_size_real_SP));
- checkCudaErrors( cudaMalloc((void**) &(parameter->getParD(lev)->distributions.f[0] ), (unsigned long long)parameter->getD3Qxx()*(unsigned long long)parameter->getParD(lev)->mem_size_real_SP));
+ checkCudaErrors( cudaMalloc((void**) &(parameter->getParD(lev)->typeOfGridNode ), parameter->getParD(lev)->memSizeLonglongLBnodes));
+ checkCudaErrors( cudaMalloc((void**) &(parameter->getParD(lev)->neighborX ), parameter->getParD(lev)->memSizeLonglongLBnodes));
+ checkCudaErrors( cudaMalloc((void**) &(parameter->getParD(lev)->neighborY ), parameter->getParD(lev)->memSizeLonglongLBnodes));
+ checkCudaErrors( cudaMalloc((void**) &(parameter->getParD(lev)->neighborZ ), parameter->getParD(lev)->memSizeLonglongLBnodes));
+ checkCudaErrors( cudaMalloc((void**) &(parameter->getParD(lev)->rho ), parameter->getParD(lev)->memSizeRealLBnodes ));
+ checkCudaErrors( cudaMalloc((void**) &(parameter->getParD(lev)->velocityX ), parameter->getParD(lev)->memSizeRealLBnodes ));
+ checkCudaErrors( cudaMalloc((void**) &(parameter->getParD(lev)->velocityY ), parameter->getParD(lev)->memSizeRealLBnodes ));
+ checkCudaErrors( cudaMalloc((void**) &(parameter->getParD(lev)->velocityZ ), parameter->getParD(lev)->memSizeRealLBnodes ));
+ checkCudaErrors( cudaMalloc((void**) &(parameter->getParD(lev)->pressure ), parameter->getParD(lev)->memSizeRealLBnodes ));
+ checkCudaErrors( cudaMalloc((void**) &(parameter->getParD(lev)->distributions.f[0]), (unsigned long long)parameter->getD3Qxx()*(unsigned long long)parameter->getParD(lev)->memSizeRealLBnodes));
//////////////////////////////////////////////////////////////////////////
- double tmp = 4. * (double)parameter->getParH(lev)->mem_size_int_SP + 5. * (double)parameter->getParH(lev)->mem_size_real_SP + (double)parameter->getD3Qxx() * (double)parameter->getParH(lev)->mem_size_real_SP;
+ double tmp = 4. * (double)parameter->getParH(lev)->memSizeLonglongLBnodes + 5. * (double)parameter->getParH(lev)->memSizeRealLBnodes + (double)parameter->getD3Qxx() * (double)parameter->getParH(lev)->memSizeRealLBnodes;
setMemsizeGPU(tmp, false);
}
void CudaMemoryManager::cudaCopySP(int lev)
{
//copy host to device
- checkCudaErrors( cudaMemcpy(parameter->getParD(lev)->typeOfGridNode , parameter->getParH(lev)->typeOfGridNode , parameter->getParH(lev)->mem_size_int_SP , cudaMemcpyHostToDevice));
- checkCudaErrors( cudaMemcpy(parameter->getParD(lev)->neighborX, parameter->getParH(lev)->neighborX, parameter->getParH(lev)->mem_size_int_SP , cudaMemcpyHostToDevice));
- checkCudaErrors( cudaMemcpy(parameter->getParD(lev)->neighborY, parameter->getParH(lev)->neighborY, parameter->getParH(lev)->mem_size_int_SP , cudaMemcpyHostToDevice));
- checkCudaErrors( cudaMemcpy(parameter->getParD(lev)->neighborZ, parameter->getParH(lev)->neighborZ, parameter->getParH(lev)->mem_size_int_SP , cudaMemcpyHostToDevice));
- checkCudaErrors( cudaMemcpy(parameter->getParD(lev)->rho , parameter->getParH(lev)->rho , parameter->getParH(lev)->mem_size_real_SP , cudaMemcpyHostToDevice));
- checkCudaErrors( cudaMemcpy(parameter->getParD(lev)->velocityX , parameter->getParH(lev)->velocityX , parameter->getParH(lev)->mem_size_real_SP , cudaMemcpyHostToDevice));
- checkCudaErrors( cudaMemcpy(parameter->getParD(lev)->velocityY , parameter->getParH(lev)->velocityY , parameter->getParH(lev)->mem_size_real_SP , cudaMemcpyHostToDevice));
- checkCudaErrors( cudaMemcpy(parameter->getParD(lev)->velocityZ , parameter->getParH(lev)->velocityZ , parameter->getParH(lev)->mem_size_real_SP , cudaMemcpyHostToDevice));
- checkCudaErrors( cudaMemcpy(parameter->getParD(lev)->pressure , parameter->getParH(lev)->pressure , parameter->getParH(lev)->mem_size_real_SP , cudaMemcpyHostToDevice));
+ checkCudaErrors( cudaMemcpy(parameter->getParD(lev)->typeOfGridNode, parameter->getParH(lev)->typeOfGridNode, parameter->getParH(lev)->memSizeLonglongLBnodes , cudaMemcpyHostToDevice));
+ checkCudaErrors( cudaMemcpy(parameter->getParD(lev)->neighborX , parameter->getParH(lev)->neighborX , parameter->getParH(lev)->memSizeLonglongLBnodes , cudaMemcpyHostToDevice));
+ checkCudaErrors( cudaMemcpy(parameter->getParD(lev)->neighborY , parameter->getParH(lev)->neighborY , parameter->getParH(lev)->memSizeLonglongLBnodes , cudaMemcpyHostToDevice));
+ checkCudaErrors( cudaMemcpy(parameter->getParD(lev)->neighborZ , parameter->getParH(lev)->neighborZ , parameter->getParH(lev)->memSizeLonglongLBnodes , cudaMemcpyHostToDevice));
+ checkCudaErrors( cudaMemcpy(parameter->getParD(lev)->rho , parameter->getParH(lev)->rho , parameter->getParH(lev)->memSizeRealLBnodes , cudaMemcpyHostToDevice));
+ checkCudaErrors( cudaMemcpy(parameter->getParD(lev)->velocityX , parameter->getParH(lev)->velocityX , parameter->getParH(lev)->memSizeRealLBnodes , cudaMemcpyHostToDevice));
+ checkCudaErrors( cudaMemcpy(parameter->getParD(lev)->velocityY , parameter->getParH(lev)->velocityY , parameter->getParH(lev)->memSizeRealLBnodes , cudaMemcpyHostToDevice));
+ checkCudaErrors( cudaMemcpy(parameter->getParD(lev)->velocityZ , parameter->getParH(lev)->velocityZ , parameter->getParH(lev)->memSizeRealLBnodes , cudaMemcpyHostToDevice));
+ checkCudaErrors( cudaMemcpy(parameter->getParD(lev)->pressure , parameter->getParH(lev)->pressure , parameter->getParH(lev)->memSizeRealLBnodes , cudaMemcpyHostToDevice));
}
void CudaMemoryManager::cudaFreeSP(int lev)
{
- checkCudaErrors( cudaFreeHost(parameter->getParH(lev)->typeOfGridNode ));
- checkCudaErrors( cudaFreeHost(parameter->getParH(lev)->velocityX ));
- checkCudaErrors( cudaFreeHost(parameter->getParH(lev)->velocityY ));
- checkCudaErrors( cudaFreeHost(parameter->getParH(lev)->velocityZ ));
- checkCudaErrors( cudaFreeHost(parameter->getParH(lev)->rho ));
- checkCudaErrors( cudaFreeHost(parameter->getParH(lev)->pressure ));
- checkCudaErrors( cudaFreeHost(parameter->getParH(lev)->neighborX));
- checkCudaErrors( cudaFreeHost(parameter->getParH(lev)->neighborY));
- checkCudaErrors( cudaFreeHost(parameter->getParH(lev)->neighborZ));
+ checkCudaErrors( cudaFreeHost(parameter->getParH(lev)->typeOfGridNode ));
+ checkCudaErrors( cudaFreeHost(parameter->getParH(lev)->velocityX ));
+ checkCudaErrors( cudaFreeHost(parameter->getParH(lev)->velocityY ));
+ checkCudaErrors( cudaFreeHost(parameter->getParH(lev)->velocityZ ));
+ checkCudaErrors( cudaFreeHost(parameter->getParH(lev)->rho ));
+ checkCudaErrors( cudaFreeHost(parameter->getParH(lev)->pressure ));
+ checkCudaErrors( cudaFreeHost(parameter->getParH(lev)->neighborX ));
+ checkCudaErrors( cudaFreeHost(parameter->getParH(lev)->neighborY ));
+ checkCudaErrors( cudaFreeHost(parameter->getParH(lev)->neighborZ ));
}
void CudaMemoryManager::cudaAllocF3SP(int lev)
{
//Device
- checkCudaErrors(cudaMalloc((void**) &(parameter->getParD(lev)->g6.g[0]), (unsigned long long)6*(unsigned long long)parameter->getParD(lev)->mem_size_real_SP));
+ checkCudaErrors(cudaMalloc((void**) &(parameter->getParD(lev)->g6.g[0]), (unsigned long long)6*(unsigned long long)parameter->getParD(lev)->memSizeRealLBnodes));
//////////////////////////////////////////////////////////////////////////
- double tmp = (double)6 * (double)parameter->getParH(lev)->mem_size_real_SP;
+ double tmp = (double)6 * (double)parameter->getParH(lev)->memSizeRealLBnodes;
setMemsizeGPU(tmp, false);
}
@@ -207,20 +200,20 @@ void CudaMemoryManager::cudaAllocVeloBC(int lev)
//Host
checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->velocityBC.q27[0]), parameter->getD3Qxx()*mem_size_inflow_Q_q ));
- checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->velocityBC.k), mem_size_inflow_Q_k ));
- checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->velocityBC.Vx), mem_size_inflow_Q_q ));
- checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->velocityBC.Vy), mem_size_inflow_Q_q ));
- checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->velocityBC.Vz), mem_size_inflow_Q_q ));
- checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->velocityBC.deltaVz), mem_size_inflow_Q_q ));
- checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->velocityBC.RhoBC), mem_size_inflow_Q_q ));
+ checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->velocityBC.k), mem_size_inflow_Q_k ));
+ checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->velocityBC.Vx), mem_size_inflow_Q_q ));
+ checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->velocityBC.Vy), mem_size_inflow_Q_q ));
+ checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->velocityBC.Vz), mem_size_inflow_Q_q ));
+ checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->velocityBC.deltaVz), mem_size_inflow_Q_q ));
+ checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->velocityBC.RhoBC), mem_size_inflow_Q_q ));
//Device
checkCudaErrors( cudaMalloc((void**) &(parameter->getParD(lev)->velocityBC.q27[0]), parameter->getD3Qxx()*mem_size_inflow_Q_q ));
- checkCudaErrors( cudaMalloc((void**) &(parameter->getParD(lev)->velocityBC.k), mem_size_inflow_Q_k ));
- checkCudaErrors( cudaMalloc((void**) &(parameter->getParD(lev)->velocityBC.Vx), mem_size_inflow_Q_q ));
- checkCudaErrors( cudaMalloc((void**) &(parameter->getParD(lev)->velocityBC.Vy), mem_size_inflow_Q_q ));
- checkCudaErrors( cudaMalloc((void**) &(parameter->getParD(lev)->velocityBC.Vz), mem_size_inflow_Q_q ));
- checkCudaErrors( cudaMalloc((void**) &(parameter->getParD(lev)->velocityBC.deltaVz), mem_size_inflow_Q_q ));
+ checkCudaErrors( cudaMalloc((void**) &(parameter->getParD(lev)->velocityBC.k), mem_size_inflow_Q_k ));
+ checkCudaErrors( cudaMalloc((void**) &(parameter->getParD(lev)->velocityBC.Vx), mem_size_inflow_Q_q ));
+ checkCudaErrors( cudaMalloc((void**) &(parameter->getParD(lev)->velocityBC.Vy), mem_size_inflow_Q_q ));
+ checkCudaErrors( cudaMalloc((void**) &(parameter->getParD(lev)->velocityBC.Vz), mem_size_inflow_Q_q ));
+ checkCudaErrors( cudaMalloc((void**) &(parameter->getParD(lev)->velocityBC.deltaVz), mem_size_inflow_Q_q ));
//////////////////////////////////////////////////////////////////////////
double tmp = (double)mem_size_inflow_Q_k + 4. * (double)mem_size_inflow_Q_q + (double)parameter->getD3Qxx() * (double)mem_size_inflow_Q_q;
@@ -232,13 +225,14 @@ void CudaMemoryManager::cudaCopyVeloBC(int lev)
unsigned int mem_size_inflow_Q_q = sizeof(real)*parameter->getParH(lev)->velocityBC.numberOfBCnodes;
checkCudaErrors( cudaMemcpy(parameter->getParD(lev)->velocityBC.q27[0], parameter->getParH(lev)->velocityBC.q27[0], parameter->getD3Qxx()* mem_size_inflow_Q_q, cudaMemcpyHostToDevice));
- checkCudaErrors( cudaMemcpy(parameter->getParD(lev)->velocityBC.k, parameter->getParH(lev)->velocityBC.k, mem_size_inflow_Q_k, cudaMemcpyHostToDevice));
- checkCudaErrors( cudaMemcpy(parameter->getParD(lev)->velocityBC.Vx, parameter->getParH(lev)->velocityBC.Vx, mem_size_inflow_Q_q, cudaMemcpyHostToDevice));
- checkCudaErrors( cudaMemcpy(parameter->getParD(lev)->velocityBC.Vy, parameter->getParH(lev)->velocityBC.Vy, mem_size_inflow_Q_q, cudaMemcpyHostToDevice));
- checkCudaErrors( cudaMemcpy(parameter->getParD(lev)->velocityBC.Vz, parameter->getParH(lev)->velocityBC.Vz, mem_size_inflow_Q_q, cudaMemcpyHostToDevice));
- checkCudaErrors( cudaMemcpy(parameter->getParD(lev)->velocityBC.deltaVz, parameter->getParH(lev)->velocityBC.deltaVz, mem_size_inflow_Q_q, cudaMemcpyHostToDevice));
+ checkCudaErrors( cudaMemcpy(parameter->getParD(lev)->velocityBC.k, parameter->getParH(lev)->velocityBC.k, mem_size_inflow_Q_k, cudaMemcpyHostToDevice));
+ checkCudaErrors( cudaMemcpy(parameter->getParD(lev)->velocityBC.Vx, parameter->getParH(lev)->velocityBC.Vx, mem_size_inflow_Q_q, cudaMemcpyHostToDevice));
+ checkCudaErrors( cudaMemcpy(parameter->getParD(lev)->velocityBC.Vy, parameter->getParH(lev)->velocityBC.Vy, mem_size_inflow_Q_q, cudaMemcpyHostToDevice));
+ checkCudaErrors( cudaMemcpy(parameter->getParD(lev)->velocityBC.Vz, parameter->getParH(lev)->velocityBC.Vz, mem_size_inflow_Q_q, cudaMemcpyHostToDevice));
+ checkCudaErrors( cudaMemcpy(parameter->getParD(lev)->velocityBC.deltaVz, parameter->getParH(lev)->velocityBC.deltaVz, mem_size_inflow_Q_q, cudaMemcpyHostToDevice));
}
+
void CudaMemoryManager::cudaFreeVeloBC(int lev)
{
checkCudaErrors( cudaFreeHost(parameter->getParH(lev)->velocityBC.q27[0] ));
@@ -256,15 +250,15 @@ void CudaMemoryManager::cudaAllocOutflowBC(int lev)
//Host
checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->outflowBC.q27[0]), parameter->getD3Qxx()*mem_size_outflow_Q_q ));
- checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->outflowBC.k), mem_size_outflow_Q_k ));
- checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->outflowBC.kN), mem_size_outflow_Q_k ));
- checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->outflowBC.RhoBC), mem_size_outflow_Q_q ));
+ checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->outflowBC.k), mem_size_outflow_Q_k ));
+ checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->outflowBC.kN), mem_size_outflow_Q_k ));
+ checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->outflowBC.RhoBC), mem_size_outflow_Q_q ));
//Device
checkCudaErrors( cudaMalloc((void**) &(parameter->getParD(lev)->outflowBC.q27[0]), parameter->getD3Qxx()* mem_size_outflow_Q_q ));
- checkCudaErrors( cudaMalloc((void**) &(parameter->getParD(lev)->outflowBC.k), mem_size_outflow_Q_k ));
- checkCudaErrors( cudaMalloc((void**) &(parameter->getParD(lev)->outflowBC.kN), mem_size_outflow_Q_k ));
- checkCudaErrors( cudaMalloc((void**) &(parameter->getParD(lev)->outflowBC.RhoBC), mem_size_outflow_Q_q ));
+ checkCudaErrors( cudaMalloc((void**) &(parameter->getParD(lev)->outflowBC.k), mem_size_outflow_Q_k ));
+ checkCudaErrors( cudaMalloc((void**) &(parameter->getParD(lev)->outflowBC.kN), mem_size_outflow_Q_k ));
+ checkCudaErrors( cudaMalloc((void**) &(parameter->getParD(lev)->outflowBC.RhoBC), mem_size_outflow_Q_q ));
//////////////////////////////////////////////////////////////////////////
double tmp = (double)mem_size_outflow_Q_q + 2. * (double)mem_size_outflow_Q_k + (double)parameter->getD3Qxx()*(double)mem_size_outflow_Q_q;
@@ -276,9 +270,9 @@ void CudaMemoryManager::cudaCopyOutflowBC(int lev)
unsigned int mem_size_outflow_Q_q = sizeof(real)*parameter->getParH(lev)->outflowBC.numberOfBCnodes;
checkCudaErrors( cudaMemcpy(parameter->getParD(lev)->outflowBC.q27[0], parameter->getParH(lev)->outflowBC.q27[0], parameter->getD3Qxx()* mem_size_outflow_Q_q, cudaMemcpyHostToDevice));
- checkCudaErrors( cudaMemcpy(parameter->getParD(lev)->outflowBC.k, parameter->getParH(lev)->outflowBC.k, mem_size_outflow_Q_k, cudaMemcpyHostToDevice));
- checkCudaErrors( cudaMemcpy(parameter->getParD(lev)->outflowBC.kN, parameter->getParH(lev)->outflowBC.kN, mem_size_outflow_Q_k, cudaMemcpyHostToDevice));
- checkCudaErrors( cudaMemcpy(parameter->getParD(lev)->outflowBC.RhoBC, parameter->getParH(lev)->outflowBC.RhoBC, mem_size_outflow_Q_q, cudaMemcpyHostToDevice));
+ checkCudaErrors( cudaMemcpy(parameter->getParD(lev)->outflowBC.k, parameter->getParH(lev)->outflowBC.k, mem_size_outflow_Q_k, cudaMemcpyHostToDevice));
+ checkCudaErrors( cudaMemcpy(parameter->getParD(lev)->outflowBC.kN, parameter->getParH(lev)->outflowBC.kN, mem_size_outflow_Q_k, cudaMemcpyHostToDevice));
+ checkCudaErrors( cudaMemcpy(parameter->getParD(lev)->outflowBC.RhoBC, parameter->getParH(lev)->outflowBC.RhoBC, mem_size_outflow_Q_q, cudaMemcpyHostToDevice));
}
void CudaMemoryManager::cudaFreeOutflowBC(int lev)
{
@@ -297,13 +291,13 @@ void CudaMemoryManager::cudaAllocNoSlipBC(int lev)
//Host
checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->noSlipBC.q27[0]), parameter->getD3Qxx()*mem_size_Q_q ));
- checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->noSlipBC.k), mem_size_Q_k ));
- checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->noSlipBC.qread), mem_size_Q_q_read ));//Geller
- checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->noSlipBC.valueQ), mem_size_Q_value ));//Geller
+ checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->noSlipBC.k), mem_size_Q_k ));
+ checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->noSlipBC.qread), mem_size_Q_q_read ));//Geller
+ checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->noSlipBC.valueQ), mem_size_Q_value ));//Geller
//Device
checkCudaErrors( cudaMalloc((void**) &(parameter->getParD(lev)->noSlipBC.q27[0]), parameter->getD3Qxx()* mem_size_Q_q ));
- checkCudaErrors( cudaMalloc((void**) &(parameter->getParD(lev)->noSlipBC.k), mem_size_Q_k ));
+ checkCudaErrors( cudaMalloc((void**) &(parameter->getParD(lev)->noSlipBC.k), mem_size_Q_k ));
//////////////////////////////////////////////////////////////////////////
double tmp = (double)mem_size_Q_k + (double)parameter->getD3Qxx()*(double)mem_size_Q_q;
@@ -315,7 +309,7 @@ void CudaMemoryManager::cudaCopyNoSlipBC(int lev)
unsigned int mem_size_Q_q = sizeof(real)*parameter->getParH(lev)->noSlipBC.numberOfBCnodes;
checkCudaErrors( cudaMemcpy(parameter->getParD(lev)->noSlipBC.q27[0], parameter->getParH(lev)->noSlipBC.q27[0], parameter->getD3Qxx()* mem_size_Q_q, cudaMemcpyHostToDevice));
- checkCudaErrors( cudaMemcpy(parameter->getParD(lev)->noSlipBC.k, parameter->getParH(lev)->noSlipBC.k, mem_size_Q_k, cudaMemcpyHostToDevice));
+ checkCudaErrors( cudaMemcpy(parameter->getParD(lev)->noSlipBC.k, parameter->getParH(lev)->noSlipBC.k, mem_size_Q_k, cudaMemcpyHostToDevice));
}
void CudaMemoryManager::cudaFreeNoSlipBC(int lev)
{
@@ -332,11 +326,11 @@ void CudaMemoryManager::cudaAllocGeomBC(int lev)
//Host
checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->geometryBC.q27[0]), parameter->getD3Qxx()*mem_size_Q_q ));
- checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->geometryBC.k), mem_size_Q_k ));
+ checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->geometryBC.k), mem_size_Q_k ));
//Device
checkCudaErrors( cudaMalloc((void**) &(parameter->getParD(lev)->geometryBC.q27[0]), parameter->getD3Qxx()* mem_size_Q_q ));
- checkCudaErrors( cudaMalloc((void**) &(parameter->getParD(lev)->geometryBC.k), mem_size_Q_k ));
+ checkCudaErrors( cudaMalloc((void**) &(parameter->getParD(lev)->geometryBC.k), mem_size_Q_k ));
//////////////////////////////////////////////////////////////////////////
double tmp = (double)mem_size_Q_k + (double)parameter->getD3Qxx()*(double)mem_size_Q_q;
@@ -348,7 +342,7 @@ void CudaMemoryManager::cudaCopyGeomBC(int lev)
unsigned int mem_size_Q_q = sizeof(real)*parameter->getParH(lev)->geometryBC.numberOfBCnodes;
checkCudaErrors( cudaMemcpy(parameter->getParD(lev)->geometryBC.q27[0], parameter->getParH(lev)->geometryBC.q27[0], parameter->getD3Qxx()* mem_size_Q_q, cudaMemcpyHostToDevice));
- checkCudaErrors( cudaMemcpy(parameter->getParD(lev)->geometryBC.k, parameter->getParH(lev)->geometryBC.k, mem_size_Q_k, cudaMemcpyHostToDevice));
+ checkCudaErrors( cudaMemcpy(parameter->getParD(lev)->geometryBC.k, parameter->getParH(lev)->geometryBC.k, mem_size_Q_k, cudaMemcpyHostToDevice));
}
void CudaMemoryManager::cudaFreeGeomBC(int lev)
{
@@ -363,15 +357,15 @@ void CudaMemoryManager::cudaAllocPress(int lev)
//Host
checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->pressureBC.q27[0]), parameter->getD3Qxx()*mem_size_Q_q ));
- checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->pressureBC.k), mem_size_Q_k ));
- checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->pressureBC.kN), mem_size_Q_k ));
- checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->pressureBC.RhoBC), mem_size_Q_q ));
+ checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->pressureBC.k), mem_size_Q_k ));
+ checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->pressureBC.kN), mem_size_Q_k ));
+ checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->pressureBC.RhoBC), mem_size_Q_q ));
//Device
checkCudaErrors( cudaMalloc((void**) &(parameter->getParD(lev)->pressureBC.q27[0]), parameter->getD3Qxx()* mem_size_Q_q ));
- checkCudaErrors( cudaMalloc((void**) &(parameter->getParD(lev)->pressureBC.k), mem_size_Q_k ));
- checkCudaErrors( cudaMalloc((void**) &(parameter->getParD(lev)->pressureBC.kN), mem_size_Q_k ));
- checkCudaErrors( cudaMalloc((void**) &(parameter->getParD(lev)->pressureBC.RhoBC), mem_size_Q_q ));
+ checkCudaErrors( cudaMalloc((void**) &(parameter->getParD(lev)->pressureBC.k), mem_size_Q_k ));
+ checkCudaErrors( cudaMalloc((void**) &(parameter->getParD(lev)->pressureBC.kN), mem_size_Q_k ));
+ checkCudaErrors( cudaMalloc((void**) &(parameter->getParD(lev)->pressureBC.RhoBC), mem_size_Q_q ));
//////////////////////////////////////////////////////////////////////////
double tmp = 2. * (double)mem_size_Q_k + (double)mem_size_Q_q + (double)parameter->getD3Qxx()*(double)mem_size_Q_q;
@@ -524,24 +518,24 @@ void CudaMemoryManager::cudaCopyProcessNeighborXIndex(int lev, unsigned int proc
cudaMemcpyHostToDevice));
}
void CudaMemoryManager::cudaCopyProcessNeighborXFsHD(int lev, unsigned int processNeighbor,
- const unsigned int &memsizeFsRecv, int streamIndex)
+ const unsigned int &memsizeFsRecv)
{
- if (streamIndex == -1)
+ if (!parameter->getStreamManager()->streamIsRegistered(CudaStreamIndex::SubDomainBorder))
checkCudaErrors( cudaMemcpy(parameter->getParD(lev)->recvProcessNeighborX[processNeighbor].f[0],
parameter->getParH(lev)->recvProcessNeighborX[processNeighbor].f[0],
parameter->getD3Qxx() * memsizeFsRecv,
cudaMemcpyHostToDevice));
else
checkCudaErrors( cudaMemcpyAsync(parameter->getParD(lev)->recvProcessNeighborX[processNeighbor].f[0],
- parameter->getParH(lev)->recvProcessNeighborX[processNeighbor].f[0],
- parameter->getD3Qxx() * memsizeFsRecv,
- cudaMemcpyHostToDevice,
- parameter->getStreamManager()->getStream(streamIndex)));
+ parameter->getParH(lev)->recvProcessNeighborX[processNeighbor].f[0],
+ parameter->getD3Qxx() * memsizeFsRecv,
+ cudaMemcpyHostToDevice,
+ parameter->getStreamManager()->getStream(CudaStreamIndex::SubDomainBorder)));
}
void CudaMemoryManager::cudaCopyProcessNeighborXFsDH(int lev, unsigned int processNeighbor,
- const unsigned int &memsizeFsSend, int streamIndex)
-{
- if (streamIndex == -1)
+ const unsigned int &memsizeFsSend)
+{
+ if (!parameter->getStreamManager()->streamIsRegistered(CudaStreamIndex::SubDomainBorder))
checkCudaErrors( cudaMemcpy(parameter->getParH(lev)->sendProcessNeighborX[processNeighbor].f[0],
parameter->getParD(lev)->sendProcessNeighborX[processNeighbor].f[0],
parameter->getD3Qxx() * memsizeFsSend,
@@ -551,7 +545,7 @@ void CudaMemoryManager::cudaCopyProcessNeighborXFsDH(int lev, unsigned int proce
parameter->getParD(lev)->sendProcessNeighborX[processNeighbor].f[0],
parameter->getD3Qxx() * memsizeFsSend,
cudaMemcpyDeviceToHost,
- parameter->getStreamManager()->getStream(streamIndex)));
+ parameter->getStreamManager()->getStream(CudaStreamIndex::SubDomainBorder)));
}
void CudaMemoryManager::cudaFreeProcessNeighborX(int lev, unsigned int processNeighbor)
{
@@ -594,35 +588,33 @@ void CudaMemoryManager::cudaCopyProcessNeighborYIndex(int lev, unsigned int proc
parameter->getParH(lev)->recvProcessNeighborY[processNeighbor].memsizeIndex,
cudaMemcpyHostToDevice));
}
-void CudaMemoryManager::cudaCopyProcessNeighborYFsHD(int lev, unsigned int processNeighbor, const unsigned int &memsizeFsRecv,
- int streamIndex)
+void CudaMemoryManager::cudaCopyProcessNeighborYFsHD(int lev, unsigned int processNeighbor, const unsigned int &memsizeFsRecv)
{
- if (streamIndex == -1)
+ if (!parameter->getStreamManager()->streamIsRegistered(CudaStreamIndex::SubDomainBorder))
checkCudaErrors( cudaMemcpy(parameter->getParD(lev)->recvProcessNeighborY[processNeighbor].f[0],
parameter->getParH(lev)->recvProcessNeighborY[processNeighbor].f[0],
parameter->getD3Qxx() * memsizeFsRecv,
cudaMemcpyHostToDevice));
else
- checkCudaErrors(cudaMemcpyAsync(parameter->getParD(lev)->recvProcessNeighborY[processNeighbor].f[0],
- parameter->getParH(lev)->recvProcessNeighborY[processNeighbor].f[0],
- parameter->getD3Qxx() * memsizeFsRecv,
- cudaMemcpyHostToDevice,
- parameter->getStreamManager()->getStream(streamIndex)));
+ checkCudaErrors( cudaMemcpyAsync(parameter->getParD(lev)->recvProcessNeighborY[processNeighbor].f[0],
+ parameter->getParH(lev)->recvProcessNeighborY[processNeighbor].f[0],
+ parameter->getD3Qxx() * memsizeFsRecv,
+ cudaMemcpyHostToDevice,
+ parameter->getStreamManager()->getStream(CudaStreamIndex::SubDomainBorder)));
}
-void CudaMemoryManager::cudaCopyProcessNeighborYFsDH(int lev, unsigned int processNeighbor, const unsigned int &memsizeFsSend,
- int streamIndex)
+void CudaMemoryManager::cudaCopyProcessNeighborYFsDH(int lev, unsigned int processNeighbor, const unsigned int &memsizeFsSend)
{
- if (streamIndex == -1)
+ if (!parameter->getStreamManager()->streamIsRegistered(CudaStreamIndex::SubDomainBorder))
checkCudaErrors( cudaMemcpy(parameter->getParH(lev)->sendProcessNeighborY[processNeighbor].f[0],
parameter->getParD(lev)->sendProcessNeighborY[processNeighbor].f[0],
parameter->getD3Qxx() * memsizeFsSend,
cudaMemcpyDeviceToHost));
else
- checkCudaErrors(
- cudaMemcpyAsync(parameter->getParH(lev)->sendProcessNeighborY[processNeighbor].f[0],
- parameter->getParD(lev)->sendProcessNeighborY[processNeighbor].f[0],
- parameter->getD3Qxx() * memsizeFsSend,
- cudaMemcpyDeviceToHost, parameter->getStreamManager()->getStream(streamIndex)));
+ checkCudaErrors( cudaMemcpyAsync(parameter->getParH(lev)->sendProcessNeighborY[processNeighbor].f[0],
+ parameter->getParD(lev)->sendProcessNeighborY[processNeighbor].f[0],
+ parameter->getD3Qxx() * memsizeFsSend,
+ cudaMemcpyDeviceToHost,
+ parameter->getStreamManager()->getStream(CudaStreamIndex::SubDomainBorder)));
}
void CudaMemoryManager::cudaFreeProcessNeighborY(int lev, unsigned int processNeighbor)
{
@@ -666,9 +658,9 @@ void CudaMemoryManager::cudaCopyProcessNeighborZIndex(int lev, unsigned int proc
cudaMemcpyHostToDevice));
}
void CudaMemoryManager::cudaCopyProcessNeighborZFsHD(int lev, unsigned int processNeighbor,
- const unsigned int &memsizeFsRecv, int streamIndex)
+ const unsigned int &memsizeFsRecv)
{
- if (streamIndex == -1)
+ if (!parameter->getStreamManager()->streamIsRegistered(CudaStreamIndex::SubDomainBorder))
checkCudaErrors( cudaMemcpy(parameter->getParD(lev)->recvProcessNeighborZ[processNeighbor].f[0],
parameter->getParH(lev)->recvProcessNeighborZ[processNeighbor].f[0],
parameter->getD3Qxx() * memsizeFsRecv,
@@ -678,12 +670,12 @@ void CudaMemoryManager::cudaCopyProcessNeighborZFsHD(int lev, unsigned int proce
parameter->getParH(lev)->recvProcessNeighborZ[processNeighbor].f[0],
parameter->getD3Qxx() * memsizeFsRecv,
cudaMemcpyHostToDevice,
- parameter->getStreamManager()->getStream(streamIndex)));
+ parameter->getStreamManager()->getStream(CudaStreamIndex::SubDomainBorder)));
}
void CudaMemoryManager::cudaCopyProcessNeighborZFsDH(int lev, unsigned int processNeighbor,
- const unsigned int &memsizeFsSend, int streamIndex)
+ const unsigned int &memsizeFsSend)
{
- if (streamIndex == -1)
+ if (!parameter->getStreamManager()->streamIsRegistered(CudaStreamIndex::SubDomainBorder))
checkCudaErrors( cudaMemcpy(parameter->getParH(lev)->sendProcessNeighborZ[processNeighbor].f[0],
parameter->getParD(lev)->sendProcessNeighborZ[processNeighbor].f[0],
parameter->getD3Qxx() * memsizeFsSend,
@@ -693,7 +685,7 @@ void CudaMemoryManager::cudaCopyProcessNeighborZFsDH(int lev, unsigned int proce
parameter->getParD(lev)->sendProcessNeighborZ[processNeighbor].f[0],
parameter->getD3Qxx() * memsizeFsSend,
cudaMemcpyDeviceToHost,
- parameter->getStreamManager()->getStream(streamIndex)));
+ parameter->getStreamManager()->getStream(CudaStreamIndex::SubDomainBorder)));
}
void CudaMemoryManager::cudaFreeProcessNeighborZ(int lev, unsigned int processNeighbor)
{
@@ -887,17 +879,17 @@ void CudaMemoryManager::cudaFreeProcessNeighborF3Z(int lev, unsigned int process
void CudaMemoryManager::cudaAllocNeighborWSB(int lev)
{
//Host
- checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->neighborInverse ), parameter->getParH(lev)->mem_size_int_SP ));
+ checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->neighborInverse ), parameter->getParH(lev)->memSizeLonglongLBnodes ));
//Device
- checkCudaErrors( cudaMalloc((void**) &(parameter->getParD(lev)->neighborInverse ), parameter->getParD(lev)->mem_size_int_SP ));
+ checkCudaErrors( cudaMalloc((void**) &(parameter->getParD(lev)->neighborInverse ), parameter->getParD(lev)->memSizeLonglongLBnodes ));
//////////////////////////////////////////////////////////////////////////
- double tmp = (double)parameter->getParH(lev)->mem_size_int_SP;
+ double tmp = (double)parameter->getParH(lev)->memSizeLonglongLBnodes;
setMemsizeGPU(tmp, false);
}
void CudaMemoryManager::cudaCopyNeighborWSB(int lev)
{
//copy host to device
- checkCudaErrors( cudaMemcpy(parameter->getParD(lev)->neighborInverse, parameter->getParH(lev)->neighborInverse, parameter->getParH(lev)->mem_size_int_SP , cudaMemcpyHostToDevice));
+ checkCudaErrors( cudaMemcpy(parameter->getParD(lev)->neighborInverse, parameter->getParH(lev)->neighborInverse, parameter->getParH(lev)->memSizeLonglongLBnodes , cudaMemcpyHostToDevice));
}
void CudaMemoryManager::cudaFreeNeighborWSB(int lev)
{
@@ -907,7 +899,7 @@ void CudaMemoryManager::cudaFreeNeighborWSB(int lev)
void CudaMemoryManager::cudaAllocTurbulentViscosity(int lev)
{
//Host
- checkCudaErrors(cudaMallocHost((void**) &(parameter->getParH(lev)->turbViscosity), parameter->getParH(lev)->mem_size_real_SP));
+ checkCudaErrors(cudaMallocHost((void**) &(parameter->getParH(lev)->turbViscosity), parameter->getParH(lev)->memSizeRealLBnodes));
//Debug
// checkCudaErrors(cudaMallocHost((void**) &(parameter->getParH(lev)->gSij ), parameter->getParH(lev)->mem_size_real_SP));
// checkCudaErrors(cudaMallocHost((void**) &(parameter->getParH(lev)->gSDij), parameter->getParH(lev)->mem_size_real_SP));
@@ -922,7 +914,7 @@ void CudaMemoryManager::cudaAllocTurbulentViscosity(int lev)
// checkCudaErrors(cudaMallocHost((void**) &(parameter->getParH(lev)->gDzvz), parameter->getParH(lev)->mem_size_real_SP));
//Device
- checkCudaErrors(cudaMalloc((void**) &(parameter->getParD(lev)->turbViscosity), parameter->getParD(lev)->mem_size_real_SP));
+ checkCudaErrors(cudaMalloc((void**) &(parameter->getParD(lev)->turbViscosity), parameter->getParD(lev)->memSizeRealLBnodes));
//Debug
// checkCudaErrors(cudaMalloc((void**) &(parameter->getParD(lev)->gSij ), parameter->getParD(lev)->mem_size_real_SP));
// checkCudaErrors(cudaMalloc((void**) &(parameter->getParD(lev)->gSDij), parameter->getParD(lev)->mem_size_real_SP));
@@ -937,13 +929,13 @@ void CudaMemoryManager::cudaAllocTurbulentViscosity(int lev)
// checkCudaErrors(cudaMalloc((void**) &(parameter->getParD(lev)->gDzvz), parameter->getParD(lev)->mem_size_real_SP));
// //////////////////////////////////////////////////////////////////////////
// double tmp = (double)parameter->getParH(lev)->mem_size_real_SP * 12.0;
- double tmp = (double)parameter->getParH(lev)->mem_size_real_SP;
+ double tmp = (double)parameter->getParH(lev)->memSizeRealLBnodes;
setMemsizeGPU(tmp, false);
}
void CudaMemoryManager::cudaCopyTurbulentViscosityHD(int lev)
{
//copy host to device
- checkCudaErrors(cudaMemcpy(parameter->getParD(lev)->turbViscosity, parameter->getParH(lev)->turbViscosity, parameter->getParH(lev)->mem_size_real_SP, cudaMemcpyHostToDevice));
+ checkCudaErrors(cudaMemcpy(parameter->getParD(lev)->turbViscosity, parameter->getParH(lev)->turbViscosity, parameter->getParH(lev)->memSizeRealLBnodes, cudaMemcpyHostToDevice));
//Debug
// checkCudaErrors(cudaMemcpy(parameter->getParD(lev)->gSij , parameter->getParH(lev)->gSij , parameter->getParH(lev)->mem_size_real_SP, cudaMemcpyHostToDevice));
// checkCudaErrors(cudaMemcpy(parameter->getParD(lev)->gSDij, parameter->getParH(lev)->gSDij, parameter->getParH(lev)->mem_size_real_SP, cudaMemcpyHostToDevice));
@@ -960,7 +952,7 @@ void CudaMemoryManager::cudaCopyTurbulentViscosityHD(int lev)
void CudaMemoryManager::cudaCopyTurbulentViscosityDH(int lev)
{
//copy device to host
- checkCudaErrors(cudaMemcpy(parameter->getParH(lev)->turbViscosity, parameter->getParD(lev)->turbViscosity, parameter->getParH(lev)->mem_size_real_SP, cudaMemcpyDeviceToHost));
+ checkCudaErrors(cudaMemcpy(parameter->getParH(lev)->turbViscosity, parameter->getParD(lev)->turbViscosity, parameter->getParH(lev)->memSizeRealLBnodes, cudaMemcpyDeviceToHost));
//Debug
// checkCudaErrors(cudaMemcpy(parameter->getParH(lev)->gSij , parameter->getParD(lev)->gSij , parameter->getParH(lev)->mem_size_real_SP, cudaMemcpyDeviceToHost));
// checkCudaErrors(cudaMemcpy(parameter->getParH(lev)->gSDij, parameter->getParD(lev)->gSDij, parameter->getParH(lev)->mem_size_real_SP, cudaMemcpyDeviceToHost));
@@ -1062,29 +1054,29 @@ void CudaMemoryManager::cudaFreeTurbulenceIntensity(int lev)
void CudaMemoryManager::cudaAllocMedianSP(int lev)
{
//Host
- checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->rho_SP_Med ), parameter->getParH(lev)->mem_size_real_SP));
- checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->vx_SP_Med ), parameter->getParH(lev)->mem_size_real_SP));
- checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->vy_SP_Med ), parameter->getParH(lev)->mem_size_real_SP));
- checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->vz_SP_Med ), parameter->getParH(lev)->mem_size_real_SP));
- checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->press_SP_Med ), parameter->getParH(lev)->mem_size_real_SP));
+ checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->rho_SP_Med ), parameter->getParH(lev)->memSizeRealLBnodes));
+ checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->vx_SP_Med ), parameter->getParH(lev)->memSizeRealLBnodes));
+ checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->vy_SP_Med ), parameter->getParH(lev)->memSizeRealLBnodes));
+ checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->vz_SP_Med ), parameter->getParH(lev)->memSizeRealLBnodes));
+ checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->press_SP_Med ), parameter->getParH(lev)->memSizeRealLBnodes));
//Device
- checkCudaErrors( cudaMalloc((void**) &(parameter->getParD(lev)->rho_SP_Med ), parameter->getParD(lev)->mem_size_real_SP));
- checkCudaErrors( cudaMalloc((void**) &(parameter->getParD(lev)->vx_SP_Med ), parameter->getParD(lev)->mem_size_real_SP));
- checkCudaErrors( cudaMalloc((void**) &(parameter->getParD(lev)->vy_SP_Med ), parameter->getParD(lev)->mem_size_real_SP));
- checkCudaErrors( cudaMalloc((void**) &(parameter->getParD(lev)->vz_SP_Med ), parameter->getParD(lev)->mem_size_real_SP));
- checkCudaErrors( cudaMalloc((void**) &(parameter->getParD(lev)->press_SP_Med ), parameter->getParD(lev)->mem_size_real_SP));
+ checkCudaErrors( cudaMalloc((void**) &(parameter->getParD(lev)->rho_SP_Med ), parameter->getParD(lev)->memSizeRealLBnodes));
+ checkCudaErrors( cudaMalloc((void**) &(parameter->getParD(lev)->vx_SP_Med ), parameter->getParD(lev)->memSizeRealLBnodes));
+ checkCudaErrors( cudaMalloc((void**) &(parameter->getParD(lev)->vy_SP_Med ), parameter->getParD(lev)->memSizeRealLBnodes));
+ checkCudaErrors( cudaMalloc((void**) &(parameter->getParD(lev)->vz_SP_Med ), parameter->getParD(lev)->memSizeRealLBnodes));
+ checkCudaErrors( cudaMalloc((void**) &(parameter->getParD(lev)->press_SP_Med ), parameter->getParD(lev)->memSizeRealLBnodes));
//////////////////////////////////////////////////////////////////////////
- double tmp = 5. * (double)parameter->getParH(lev)->mem_size_real_SP;
+ double tmp = 5. * (double)parameter->getParH(lev)->memSizeRealLBnodes;
setMemsizeGPU(tmp, false);
}
void CudaMemoryManager::cudaCopyMedianSP(int lev)
{
//copy host to device
- checkCudaErrors( cudaMemcpy(parameter->getParD(lev)->rho_SP_Med , parameter->getParH(lev)->rho_SP_Med , parameter->getParH(lev)->mem_size_real_SP , cudaMemcpyHostToDevice));
- checkCudaErrors( cudaMemcpy(parameter->getParD(lev)->vx_SP_Med , parameter->getParH(lev)->vx_SP_Med , parameter->getParH(lev)->mem_size_real_SP , cudaMemcpyHostToDevice));
- checkCudaErrors( cudaMemcpy(parameter->getParD(lev)->vy_SP_Med , parameter->getParH(lev)->vy_SP_Med , parameter->getParH(lev)->mem_size_real_SP , cudaMemcpyHostToDevice));
- checkCudaErrors( cudaMemcpy(parameter->getParD(lev)->vz_SP_Med , parameter->getParH(lev)->vz_SP_Med , parameter->getParH(lev)->mem_size_real_SP , cudaMemcpyHostToDevice));
- checkCudaErrors( cudaMemcpy(parameter->getParD(lev)->press_SP_Med, parameter->getParH(lev)->press_SP_Med, parameter->getParH(lev)->mem_size_real_SP , cudaMemcpyHostToDevice));
+ checkCudaErrors( cudaMemcpy(parameter->getParD(lev)->rho_SP_Med , parameter->getParH(lev)->rho_SP_Med , parameter->getParH(lev)->memSizeRealLBnodes , cudaMemcpyHostToDevice));
+ checkCudaErrors( cudaMemcpy(parameter->getParD(lev)->vx_SP_Med , parameter->getParH(lev)->vx_SP_Med , parameter->getParH(lev)->memSizeRealLBnodes , cudaMemcpyHostToDevice));
+ checkCudaErrors( cudaMemcpy(parameter->getParD(lev)->vy_SP_Med , parameter->getParH(lev)->vy_SP_Med , parameter->getParH(lev)->memSizeRealLBnodes , cudaMemcpyHostToDevice));
+ checkCudaErrors( cudaMemcpy(parameter->getParD(lev)->vz_SP_Med , parameter->getParH(lev)->vz_SP_Med , parameter->getParH(lev)->memSizeRealLBnodes , cudaMemcpyHostToDevice));
+ checkCudaErrors( cudaMemcpy(parameter->getParD(lev)->press_SP_Med, parameter->getParH(lev)->press_SP_Med, parameter->getParH(lev)->memSizeRealLBnodes , cudaMemcpyHostToDevice));
}
void CudaMemoryManager::cudaFreeMedianSP(int lev)
{
@@ -1097,11 +1089,11 @@ void CudaMemoryManager::cudaFreeMedianSP(int lev)
void CudaMemoryManager::cudaAllocMedianOut(int lev)
{
//Host
- checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->rho_SP_Med_Out ), parameter->getParH(lev)->mem_size_real_SP));
- checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->vx_SP_Med_Out ), parameter->getParH(lev)->mem_size_real_SP));
- checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->vy_SP_Med_Out ), parameter->getParH(lev)->mem_size_real_SP));
- checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->vz_SP_Med_Out ), parameter->getParH(lev)->mem_size_real_SP));
- checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->press_SP_Med_Out ), parameter->getParH(lev)->mem_size_real_SP));
+ checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->rho_SP_Med_Out ), parameter->getParH(lev)->memSizeRealLBnodes));
+ checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->vx_SP_Med_Out ), parameter->getParH(lev)->memSizeRealLBnodes));
+ checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->vy_SP_Med_Out ), parameter->getParH(lev)->memSizeRealLBnodes));
+ checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->vz_SP_Med_Out ), parameter->getParH(lev)->memSizeRealLBnodes));
+ checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->press_SP_Med_Out ), parameter->getParH(lev)->memSizeRealLBnodes));
}
void CudaMemoryManager::cudaFreeMedianOut(int lev)
{
@@ -1655,6 +1647,133 @@ void CudaMemoryManager::cudaFreeWallModel(int lev, bool hasWallModelMonitor)
}
}
+
+//Precursor BC
+void CudaMemoryManager::cudaAllocPrecursorBC(int lev)
+{
+ uint memSizeQInt = parameter->getParH(lev)->precursorBC.numberOfBCnodes*sizeof(int);
+ uint memSizeQUint = parameter->getParH(lev)->precursorBC.numberOfBCnodes*sizeof(uint);
+ uint memSizeQReal = parameter->getParH(lev)->precursorBC.numberOfBCnodes*sizeof(real);
+
+ checkCudaErrors( cudaMallocHost((void**) ¶meter->getParH(lev)->precursorBC.k, memSizeQInt));
+ checkCudaErrors( cudaMallocHost((void**) ¶meter->getParH(lev)->precursorBC.q27[0], parameter->getD3Qxx()*memSizeQReal));
+
+
+ checkCudaErrors( cudaMallocHost((void**) ¶meter->getParH(lev)->precursorBC.planeNeighbor0PP, memSizeQUint));
+ checkCudaErrors( cudaMallocHost((void**) ¶meter->getParH(lev)->precursorBC.planeNeighbor0PM, memSizeQUint));
+ checkCudaErrors( cudaMallocHost((void**) ¶meter->getParH(lev)->precursorBC.planeNeighbor0MP, memSizeQUint));
+ checkCudaErrors( cudaMallocHost((void**) ¶meter->getParH(lev)->precursorBC.planeNeighbor0MM, memSizeQUint));
+
+ checkCudaErrors( cudaMallocHost((void**) ¶meter->getParH(lev)->precursorBC.weights0PP, memSizeQReal));
+ checkCudaErrors( cudaMallocHost((void**) ¶meter->getParH(lev)->precursorBC.weights0PM, memSizeQReal));
+ checkCudaErrors( cudaMallocHost((void**) ¶meter->getParH(lev)->precursorBC.weights0MP, memSizeQReal));
+ checkCudaErrors( cudaMallocHost((void**) ¶meter->getParH(lev)->precursorBC.weights0MM, memSizeQReal));
+
+ checkCudaErrors( cudaMalloc((void**) ¶meter->getParD(lev)->precursorBC.k, memSizeQInt));
+ checkCudaErrors( cudaMalloc((void**) ¶meter->getParD(lev)->precursorBC.q27[0], parameter->getD3Qxx()*memSizeQReal));
+
+ checkCudaErrors( cudaMalloc((void**) ¶meter->getParD(lev)->precursorBC.planeNeighbor0PP, memSizeQUint));
+ checkCudaErrors( cudaMalloc((void**) ¶meter->getParD(lev)->precursorBC.planeNeighbor0PM, memSizeQUint));
+ checkCudaErrors( cudaMalloc((void**) ¶meter->getParD(lev)->precursorBC.planeNeighbor0MP, memSizeQUint));
+ checkCudaErrors( cudaMalloc((void**) ¶meter->getParD(lev)->precursorBC.planeNeighbor0MM, memSizeQUint));
+
+ checkCudaErrors( cudaMalloc((void**) ¶meter->getParD(lev)->precursorBC.weights0PP, memSizeQReal));
+ checkCudaErrors( cudaMalloc((void**) ¶meter->getParD(lev)->precursorBC.weights0PM, memSizeQReal));
+ checkCudaErrors( cudaMalloc((void**) ¶meter->getParD(lev)->precursorBC.weights0MP, memSizeQReal));
+ checkCudaErrors( cudaMalloc((void**) ¶meter->getParD(lev)->precursorBC.weights0MM, memSizeQReal));
+
+ real memSize = memSizeQInt+4*memSizeQUint+(4+parameter->getD3Qxx())*memSizeQReal;
+ setMemsizeGPU(memSize, false);
+
+}
+
+
+void CudaMemoryManager::cudaAllocPrecursorData(int lev)
+{
+ size_t size = parameter->getParH(lev)->precursorBC.numberOfPrecursorNodes*sizeof(real)*parameter->getParH(lev)->precursorBC.numberOfQuantities;
+
+ checkCudaErrors( cudaMallocHost((void**) ¶meter->getParH(lev)->precursorBC.last, size));
+ checkCudaErrors( cudaMallocHost((void**) ¶meter->getParH(lev)->precursorBC.current, size));
+ checkCudaErrors( cudaMallocHost((void**) ¶meter->getParH(lev)->precursorBC.next, size));
+
+ checkCudaErrors( cudaMalloc((void**) ¶meter->getParD(lev)->precursorBC.last, size));
+ checkCudaErrors( cudaMalloc((void**) ¶meter->getParD(lev)->precursorBC.current, size));
+ checkCudaErrors( cudaMalloc((void**) ¶meter->getParD(lev)->precursorBC.next, size));
+ setMemsizeGPU(3*size, false);
+}
+
+
+void CudaMemoryManager::cudaCopyPrecursorBC(int lev)
+{
+ uint memSizeQInt = parameter->getParH(lev)->precursorBC.numberOfBCnodes*sizeof(int);
+ uint memSizeQUint = parameter->getParH(lev)->precursorBC.numberOfBCnodes*sizeof(uint);
+ uint memSizeQReal = parameter->getParH(lev)->precursorBC.numberOfBCnodes*sizeof(real);
+
+ checkCudaErrors( cudaMemcpy(parameter->getParD(lev)->precursorBC.k, parameter->getParH(lev)->precursorBC.k, memSizeQInt, cudaMemcpyHostToDevice));
+
+ checkCudaErrors( cudaMemcpy(parameter->getParD(lev)->precursorBC.q27[0], parameter->getParH(lev)->precursorBC.q27[0], memSizeQReal*parameter->getD3Qxx(), cudaMemcpyHostToDevice));
+
+ checkCudaErrors( cudaMemcpy(parameter->getParD(lev)->precursorBC.planeNeighbor0PP, parameter->getParH(lev)->precursorBC.planeNeighbor0PP, memSizeQUint, cudaMemcpyHostToDevice));
+ checkCudaErrors( cudaMemcpy(parameter->getParD(lev)->precursorBC.planeNeighbor0PM, parameter->getParH(lev)->precursorBC.planeNeighbor0PM, memSizeQUint, cudaMemcpyHostToDevice));
+ checkCudaErrors( cudaMemcpy(parameter->getParD(lev)->precursorBC.planeNeighbor0MP, parameter->getParH(lev)->precursorBC.planeNeighbor0MP, memSizeQUint, cudaMemcpyHostToDevice));
+ checkCudaErrors( cudaMemcpy(parameter->getParD(lev)->precursorBC.planeNeighbor0MM, parameter->getParH(lev)->precursorBC.planeNeighbor0MM, memSizeQUint, cudaMemcpyHostToDevice));
+
+ checkCudaErrors( cudaMemcpy(parameter->getParD(lev)->precursorBC.weights0PP, parameter->getParH(lev)->precursorBC.weights0PP, memSizeQReal, cudaMemcpyHostToDevice));
+ checkCudaErrors( cudaMemcpy(parameter->getParD(lev)->precursorBC.weights0PM, parameter->getParH(lev)->precursorBC.weights0PM, memSizeQReal, cudaMemcpyHostToDevice));
+ checkCudaErrors( cudaMemcpy(parameter->getParD(lev)->precursorBC.weights0MP, parameter->getParH(lev)->precursorBC.weights0MP, memSizeQReal, cudaMemcpyHostToDevice));
+ checkCudaErrors( cudaMemcpy(parameter->getParD(lev)->precursorBC.weights0MM, parameter->getParH(lev)->precursorBC.weights0MM, memSizeQReal, cudaMemcpyHostToDevice));
+}
+void CudaMemoryManager::cudaCopyPrecursorData(int lev)
+{
+ auto prec = ¶meter->getParH(lev)->precursorBC;
+ auto precStream = parameter->getStreamManager()->getStream(CudaStreamIndex::Precursor);
+ size_t memSize = prec->numberOfPrecursorNodes*sizeof(real)*prec->numberOfQuantities;
+ checkCudaErrors( cudaStreamSynchronize(precStream) );
+ checkCudaErrors( cudaMemcpyAsync(parameter->getParD(lev)->precursorBC.next, prec->next, memSize, cudaMemcpyHostToDevice, precStream) );
+}
+
+
+void CudaMemoryManager::cudaFreePrecursorBC(int lev)
+{
+ checkCudaErrors( cudaFreeHost( parameter->getParH(lev)->precursorBC.k));
+
+ checkCudaErrors( cudaFreeHost( parameter->getParH(lev)->precursorBC.q27[0]));
+
+ checkCudaErrors( cudaFreeHost( parameter->getParH(lev)->precursorBC.planeNeighbor0PP));
+ checkCudaErrors( cudaFreeHost( parameter->getParH(lev)->precursorBC.planeNeighbor0PM));
+ checkCudaErrors( cudaFreeHost( parameter->getParH(lev)->precursorBC.planeNeighbor0MP));
+ checkCudaErrors( cudaFreeHost( parameter->getParH(lev)->precursorBC.planeNeighbor0MM));
+
+ checkCudaErrors( cudaFreeHost( parameter->getParH(lev)->precursorBC.weights0PP));
+ checkCudaErrors( cudaFreeHost( parameter->getParH(lev)->precursorBC.weights0PM));
+ checkCudaErrors( cudaFreeHost( parameter->getParH(lev)->precursorBC.weights0MP));
+ checkCudaErrors( cudaFreeHost( parameter->getParH(lev)->precursorBC.weights0MM));
+
+ checkCudaErrors( cudaFree( parameter->getParD(lev)->precursorBC.k));
+
+ checkCudaErrors( cudaFree( parameter->getParD(lev)->precursorBC.q27[0]));
+
+ checkCudaErrors( cudaFree( parameter->getParD(lev)->precursorBC.planeNeighbor0PP));
+ checkCudaErrors( cudaFree( parameter->getParD(lev)->precursorBC.planeNeighbor0PM));
+ checkCudaErrors( cudaFree( parameter->getParD(lev)->precursorBC.planeNeighbor0MP));
+ checkCudaErrors( cudaFree( parameter->getParD(lev)->precursorBC.planeNeighbor0MM));
+
+ checkCudaErrors( cudaFree( parameter->getParD(lev)->precursorBC.weights0PP));
+ checkCudaErrors( cudaFree( parameter->getParD(lev)->precursorBC.weights0PM));
+ checkCudaErrors( cudaFree( parameter->getParD(lev)->precursorBC.weights0MP));
+ checkCudaErrors( cudaFree( parameter->getParD(lev)->precursorBC.weights0MM));
+}
+
+void CudaMemoryManager::cudaFreePrecursorData(int lev)
+{
+ checkCudaErrors( cudaFreeHost( parameter->getParH(lev)->precursorBC.last));
+ checkCudaErrors( cudaFreeHost( parameter->getParH(lev)->precursorBC.current));
+ checkCudaErrors( cudaFreeHost( parameter->getParH(lev)->precursorBC.next));
+
+ checkCudaErrors( cudaFree( parameter->getParD(lev)->precursorBC.last));
+ checkCudaErrors( cudaFree( parameter->getParD(lev)->precursorBC.current));
+ checkCudaErrors( cudaFree( parameter->getParD(lev)->precursorBC.next));
+}
//Test roundoff error
void CudaMemoryManager::cudaAllocTestRE(int lev, unsigned int size)
{
@@ -1913,15 +2032,15 @@ void CudaMemoryManager::cudaFreeMeasurePointsIndex(int lev)
}
void CudaMemoryManager::cudaAllocFsForCheckPointAndRestart(int lev)
{
- checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->distributions.f[0] ), (unsigned long long)parameter->getD3Qxx()*(unsigned long long)parameter->getParH(lev)->mem_size_real_SP));
+ checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->distributions.f[0] ), (unsigned long long)parameter->getD3Qxx()*(unsigned long long)parameter->getParH(lev)->memSizeRealLBnodes));
}
void CudaMemoryManager::cudaCopyFsForRestart(int lev)
{
- checkCudaErrors( cudaMemcpy(parameter->getParD(lev)->distributions.f[0], parameter->getParH(lev)->distributions.f[0], (unsigned long long)parameter->getD3Qxx()*(unsigned long long)parameter->getParH(lev)->mem_size_real_SP , cudaMemcpyHostToDevice));
+ checkCudaErrors( cudaMemcpy(parameter->getParD(lev)->distributions.f[0], parameter->getParH(lev)->distributions.f[0], (unsigned long long)parameter->getD3Qxx()*(unsigned long long)parameter->getParH(lev)->memSizeRealLBnodes , cudaMemcpyHostToDevice));
}
void CudaMemoryManager::cudaCopyFsForCheckPoint(int lev)
{
- checkCudaErrors( cudaMemcpy(parameter->getParH(lev)->distributions.f[0], parameter->getParD(lev)->distributions.f[0], (unsigned long long)parameter->getD3Qxx()*(unsigned long long)parameter->getParH(lev)->mem_size_real_SP , cudaMemcpyDeviceToHost));
+ checkCudaErrors( cudaMemcpy(parameter->getParH(lev)->distributions.f[0], parameter->getParD(lev)->distributions.f[0], (unsigned long long)parameter->getD3Qxx()*(unsigned long long)parameter->getParH(lev)->memSizeRealLBnodes , cudaMemcpyDeviceToHost));
}
void CudaMemoryManager::cudaFreeFsForCheckPointAndRestart(int lev)
{
@@ -2400,20 +2519,20 @@ void CudaMemoryManager::cudaFreePorousMedia(PorousMedia* pm, int lev)
void CudaMemoryManager::cudaAllocConcentration(int lev)
{
//Host
- checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->Conc), parameter->getParH(lev)->mem_size_real_SP));
+ checkCudaErrors( cudaMallocHost((void**) &(parameter->getParH(lev)->Conc), parameter->getParH(lev)->memSizeRealLBnodes));
//Device
- checkCudaErrors( cudaMalloc((void**) &(parameter->getParD(lev)->Conc), parameter->getParD(lev)->mem_size_real_SP));
+ checkCudaErrors( cudaMalloc((void**) &(parameter->getParD(lev)->Conc), parameter->getParD(lev)->memSizeRealLBnodes));
//////////////////////////////////////////////////////////////////////////
- double tmp = (double)parameter->getParH(lev)->mem_size_real_SP;
+ double tmp = (double)parameter->getParH(lev)->memSizeRealLBnodes;
setMemsizeGPU(tmp, false);
}
void CudaMemoryManager::cudaCopyConcentrationDeviceToHost(int lev)
{
- checkCudaErrors( cudaMemcpy(parameter->getParH(lev)->Conc, parameter->getParD(lev)->Conc, parameter->getParH(lev)->mem_size_real_SP , cudaMemcpyDeviceToHost));
+ checkCudaErrors( cudaMemcpy(parameter->getParH(lev)->Conc, parameter->getParD(lev)->Conc, parameter->getParH(lev)->memSizeRealLBnodes , cudaMemcpyDeviceToHost));
}
void CudaMemoryManager::cudaCopyConcentrationHostToDevice(int lev)
{
- checkCudaErrors( cudaMemcpy(parameter->getParD(lev)->Conc, parameter->getParH(lev)->Conc, parameter->getParH(lev)->mem_size_real_SP, cudaMemcpyHostToDevice));
+ checkCudaErrors( cudaMemcpy(parameter->getParD(lev)->Conc, parameter->getParH(lev)->Conc, parameter->getParH(lev)->memSizeRealLBnodes, cudaMemcpyHostToDevice));
}
void CudaMemoryManager::cudaFreeConcentration(int lev)
{
@@ -2425,14 +2544,14 @@ void CudaMemoryManager::cudaAllocTempFs(int lev)
//Device
if (parameter->getDiffMod() == 7)
{
- checkCudaErrors( cudaMalloc((void**) &(parameter->getParD(lev)->distributionsAD7.f[0]), parameter->getDiffMod()*parameter->getParH(lev)->mem_size_real_SP));
+ checkCudaErrors( cudaMalloc((void**) &(parameter->getParD(lev)->distributionsAD7.f[0]), parameter->getDiffMod()*parameter->getParH(lev)->memSizeRealLBnodes));
}
else if (parameter->getDiffMod() == 27)
{
- checkCudaErrors( cudaMalloc((void**) &(parameter->getParD(lev)->distributionsAD27.f[0]), parameter->getDiffMod()*parameter->getParH(lev)->mem_size_real_SP));
+ checkCudaErrors( cudaMalloc((void**) &(parameter->getParD(lev)->distributionsAD27.f[0]), parameter->getDiffMod()*parameter->getParH(lev)->memSizeRealLBnodes));
}
//////////////////////////////////////////////////////////////////////////
- double tmp = (double)(parameter->getDiffMod() * parameter->getParH(lev)->mem_size_real_SP);
+ double tmp = (double)(parameter->getDiffMod() * parameter->getParH(lev)->memSizeRealLBnodes);
setMemsizeGPU(tmp, false);
}
//////////////////////////////////////////////////////////////////////////
@@ -2627,12 +2746,12 @@ void CudaMemoryManager::cudaFreeConcFile(int lev)
void CudaMemoryManager::cudaAllocMedianOutAD(int lev)
{
//Host
- checkCudaErrors(cudaMallocHost((void**) &(parameter->getParH(lev)->rho_SP_Med_Out), parameter->getParH(lev)->mem_size_real_SP));
- checkCudaErrors(cudaMallocHost((void**) &(parameter->getParH(lev)->vx_SP_Med_Out), parameter->getParH(lev)->mem_size_real_SP));
- checkCudaErrors(cudaMallocHost((void**) &(parameter->getParH(lev)->vy_SP_Med_Out), parameter->getParH(lev)->mem_size_real_SP));
- checkCudaErrors(cudaMallocHost((void**) &(parameter->getParH(lev)->vz_SP_Med_Out), parameter->getParH(lev)->mem_size_real_SP));
- checkCudaErrors(cudaMallocHost((void**) &(parameter->getParH(lev)->press_SP_Med_Out), parameter->getParH(lev)->mem_size_real_SP));
- checkCudaErrors(cudaMallocHost((void**) &(parameter->getParH(lev)->Conc_Med_Out), parameter->getParH(lev)->mem_size_real_SP));
+ checkCudaErrors(cudaMallocHost((void**) &(parameter->getParH(lev)->rho_SP_Med_Out), parameter->getParH(lev)->memSizeRealLBnodes));
+ checkCudaErrors(cudaMallocHost((void**) &(parameter->getParH(lev)->vx_SP_Med_Out), parameter->getParH(lev)->memSizeRealLBnodes));
+ checkCudaErrors(cudaMallocHost((void**) &(parameter->getParH(lev)->vy_SP_Med_Out), parameter->getParH(lev)->memSizeRealLBnodes));
+ checkCudaErrors(cudaMallocHost((void**) &(parameter->getParH(lev)->vz_SP_Med_Out), parameter->getParH(lev)->memSizeRealLBnodes));
+ checkCudaErrors(cudaMallocHost((void**) &(parameter->getParH(lev)->press_SP_Med_Out), parameter->getParH(lev)->memSizeRealLBnodes));
+ checkCudaErrors(cudaMallocHost((void**) &(parameter->getParH(lev)->Conc_Med_Out), parameter->getParH(lev)->memSizeRealLBnodes));
}
void CudaMemoryManager::cudaFreeMedianOutAD(int lev)
{
@@ -2886,31 +3005,31 @@ void CudaMemoryManager::cudaFreeProcessNeighborADZ(int lev, unsigned int process
void CudaMemoryManager::cudaAlloc2ndOrderDerivitivesIsoTest(int lev)
{
//Host
- checkCudaErrors(cudaMallocHost((void**) &(parameter->getParH(lev)->dxxUx), parameter->getParH(lev)->mem_size_real_SP));
- checkCudaErrors(cudaMallocHost((void**) &(parameter->getParH(lev)->dyyUy), parameter->getParH(lev)->mem_size_real_SP));
- checkCudaErrors(cudaMallocHost((void**) &(parameter->getParH(lev)->dzzUz), parameter->getParH(lev)->mem_size_real_SP));
+ checkCudaErrors(cudaMallocHost((void**) &(parameter->getParH(lev)->dxxUx), parameter->getParH(lev)->memSizeRealLBnodes));
+ checkCudaErrors(cudaMallocHost((void**) &(parameter->getParH(lev)->dyyUy), parameter->getParH(lev)->memSizeRealLBnodes));
+ checkCudaErrors(cudaMallocHost((void**) &(parameter->getParH(lev)->dzzUz), parameter->getParH(lev)->memSizeRealLBnodes));
//Device (spinning ship)
- checkCudaErrors(cudaMalloc((void**) &(parameter->getParD(lev)->dxxUx), parameter->getParH(lev)->mem_size_real_SP));
- checkCudaErrors(cudaMalloc((void**) &(parameter->getParD(lev)->dyyUy), parameter->getParH(lev)->mem_size_real_SP));
- checkCudaErrors(cudaMalloc((void**) &(parameter->getParD(lev)->dzzUz), parameter->getParH(lev)->mem_size_real_SP));
+ checkCudaErrors(cudaMalloc((void**) &(parameter->getParD(lev)->dxxUx), parameter->getParH(lev)->memSizeRealLBnodes));
+ checkCudaErrors(cudaMalloc((void**) &(parameter->getParD(lev)->dyyUy), parameter->getParH(lev)->memSizeRealLBnodes));
+ checkCudaErrors(cudaMalloc((void**) &(parameter->getParD(lev)->dzzUz), parameter->getParH(lev)->memSizeRealLBnodes));
//////////////////////////////////////////////////////////////////////////
- double tmp = 3. * (double)parameter->getParH(lev)->mem_size_real_SP;
+ double tmp = 3. * (double)parameter->getParH(lev)->memSizeRealLBnodes;
setMemsizeGPU(tmp, false);
//printf("Coord = %f MB",tmp/1000000.);
}
void CudaMemoryManager::cudaCopy2ndOrderDerivitivesIsoTestDH(int lev)
{
//copy device to host
- checkCudaErrors(cudaMemcpy(parameter->getParH(lev)->dxxUx, parameter->getParD(lev)->dxxUx, parameter->getParH(lev)->mem_size_real_SP, cudaMemcpyDeviceToHost));
- checkCudaErrors(cudaMemcpy(parameter->getParH(lev)->dyyUy, parameter->getParD(lev)->dyyUy, parameter->getParH(lev)->mem_size_real_SP, cudaMemcpyDeviceToHost));
- checkCudaErrors(cudaMemcpy(parameter->getParH(lev)->dzzUz, parameter->getParD(lev)->dzzUz, parameter->getParH(lev)->mem_size_real_SP, cudaMemcpyDeviceToHost));
+ checkCudaErrors(cudaMemcpy(parameter->getParH(lev)->dxxUx, parameter->getParD(lev)->dxxUx, parameter->getParH(lev)->memSizeRealLBnodes, cudaMemcpyDeviceToHost));
+ checkCudaErrors(cudaMemcpy(parameter->getParH(lev)->dyyUy, parameter->getParD(lev)->dyyUy, parameter->getParH(lev)->memSizeRealLBnodes, cudaMemcpyDeviceToHost));
+ checkCudaErrors(cudaMemcpy(parameter->getParH(lev)->dzzUz, parameter->getParD(lev)->dzzUz, parameter->getParH(lev)->memSizeRealLBnodes, cudaMemcpyDeviceToHost));
}
void CudaMemoryManager::cudaCopy2ndOrderDerivitivesIsoTestHD(int lev)
{
//copy host to device
- checkCudaErrors(cudaMemcpy(parameter->getParD(lev)->dxxUx, parameter->getParH(lev)->dxxUx, parameter->getParH(lev)->mem_size_real_SP, cudaMemcpyHostToDevice));
- checkCudaErrors(cudaMemcpy(parameter->getParD(lev)->dyyUy, parameter->getParH(lev)->dyyUy, parameter->getParH(lev)->mem_size_real_SP, cudaMemcpyHostToDevice));
- checkCudaErrors(cudaMemcpy(parameter->getParD(lev)->dzzUz, parameter->getParH(lev)->dzzUz, parameter->getParH(lev)->mem_size_real_SP, cudaMemcpyHostToDevice));
+ checkCudaErrors(cudaMemcpy(parameter->getParD(lev)->dxxUx, parameter->getParH(lev)->dxxUx, parameter->getParH(lev)->memSizeRealLBnodes, cudaMemcpyHostToDevice));
+ checkCudaErrors(cudaMemcpy(parameter->getParD(lev)->dyyUy, parameter->getParH(lev)->dyyUy, parameter->getParH(lev)->memSizeRealLBnodes, cudaMemcpyHostToDevice));
+ checkCudaErrors(cudaMemcpy(parameter->getParD(lev)->dzzUz, parameter->getParH(lev)->dzzUz, parameter->getParH(lev)->memSizeRealLBnodes, cudaMemcpyHostToDevice));
}
void CudaMemoryManager::cudaFree2ndOrderDerivitivesIsoTest(int lev)
@@ -2921,231 +3040,297 @@ void CudaMemoryManager::cudaFree2ndOrderDerivitivesIsoTest(int lev)
}
-void CudaMemoryManager::cudaAllocFluidNodeIndices(int lev) {
- uint mem_size_geo_fluid_nodes = sizeof(uint) * parameter->getParH(lev)->numberOfFluidNodes;
+void CudaMemoryManager::cudaAllocTaggedFluidNodeIndices(CollisionTemplate tag, int lev) {
+ uint mem_size_tagged_fluid_nodes = sizeof(uint) * parameter->getParH(lev)->numberOfTaggedFluidNodes[tag];
// Host
- checkCudaErrors(cudaMallocHost((void **)&(parameter->getParH(lev)->fluidNodeIndices), mem_size_geo_fluid_nodes));
+ checkCudaErrors(cudaMallocHost((void **)&(parameter->getParH(lev)->taggedFluidNodeIndices[tag]), mem_size_tagged_fluid_nodes));
// Device
- checkCudaErrors(cudaMalloc((void **)&(parameter->getParD(lev)->fluidNodeIndices), mem_size_geo_fluid_nodes));
+ checkCudaErrors(cudaMalloc((void **)&(parameter->getParD(lev)->taggedFluidNodeIndices[tag]), mem_size_tagged_fluid_nodes));
//////////////////////////////////////////////////////////////////////////
- setMemsizeGPU((double)mem_size_geo_fluid_nodes, false);
+ setMemsizeGPU((double)mem_size_tagged_fluid_nodes, false);
}
-void CudaMemoryManager::cudaCopyFluidNodeIndices(int lev) {
- uint mem_size_geo_fluid_nodes = sizeof(uint) * parameter->getParH(lev)->numberOfFluidNodes;
- checkCudaErrors(cudaMemcpy(parameter->getParD(lev)->fluidNodeIndices,
- parameter->getParH(lev)->fluidNodeIndices,
- mem_size_geo_fluid_nodes, cudaMemcpyHostToDevice));
+void CudaMemoryManager::cudaCopyTaggedFluidNodeIndices(CollisionTemplate tag, int lev) {
+ uint mem_size_tagged_fluid_nodes = sizeof(uint) * parameter->getParH(lev)->numberOfTaggedFluidNodes[tag];
+ checkCudaErrors(cudaMemcpy(parameter->getParD(lev)->taggedFluidNodeIndices[tag],
+ parameter->getParH(lev)->taggedFluidNodeIndices[tag],
+ mem_size_tagged_fluid_nodes, cudaMemcpyHostToDevice));
}
-void CudaMemoryManager::cudaFreeFluidNodeIndices(int lev) {
- checkCudaErrors(cudaFreeHost(parameter->getParH(lev)->fluidNodeIndices));
+void CudaMemoryManager::cudaFreeTaggedFluidNodeIndices(CollisionTemplate tag, int lev) {
+ checkCudaErrors(cudaFreeHost(parameter->getParH(lev)->taggedFluidNodeIndices[tag]));
}
-void CudaMemoryManager::cudaAllocFluidNodeIndicesBorder(int lev) {
- uint mem_size_fluid_nodes_border = sizeof(uint) * parameter->getParH(lev)->numberOfFluidNodesBorder;
- // Host
- checkCudaErrors(
- cudaMallocHost((void **)&(parameter->getParH(lev)->fluidNodeIndicesBorder), mem_size_fluid_nodes_border));
- // Device
- checkCudaErrors(
- cudaMalloc((void **)&(parameter->getParD(lev)->fluidNodeIndicesBorder), mem_size_fluid_nodes_border));
- //////////////////////////////////////////////////////////////////////////
- setMemsizeGPU((double)mem_size_fluid_nodes_border, false);
-}
+////////////////////////////////////////////////////////////////////////////////////
+// ActuatorFarm
+///////////////////////////////////////////////////////////////////////////////
+void CudaMemoryManager::cudaAllocBladeGeometries(ActuatorFarm* actuatorFarm)
+{
+ uint sizeRealTurbine = sizeof(real)*actuatorFarm->getNumberOfTurbines();
+ checkCudaErrors( cudaMallocHost((void**) &actuatorFarm->bladeRadiiH, sizeRealTurbine*actuatorFarm->getNumberOfNodesPerBlade()) );
+ checkCudaErrors( cudaMallocHost((void**) &actuatorFarm->diametersH, sizeRealTurbine) );
+ checkCudaErrors( cudaMallocHost((void**) &actuatorFarm->turbinePosXH, sizeRealTurbine) );
+ checkCudaErrors( cudaMallocHost((void**) &actuatorFarm->turbinePosYH, sizeRealTurbine) );
+ checkCudaErrors( cudaMallocHost((void**) &actuatorFarm->turbinePosZH, sizeRealTurbine) );
+
+ checkCudaErrors( cudaMalloc((void**) &actuatorFarm->bladeRadiiD, sizeRealTurbine*actuatorFarm->getNumberOfNodesPerBlade()) );
+ checkCudaErrors( cudaMalloc((void**) &actuatorFarm->diametersD, sizeRealTurbine) );
+ checkCudaErrors( cudaMalloc((void**) &actuatorFarm->turbinePosXD, sizeRealTurbine) );
+ checkCudaErrors( cudaMalloc((void**) &actuatorFarm->turbinePosYD, sizeRealTurbine) );
+ checkCudaErrors( cudaMalloc((void**) &actuatorFarm->turbinePosZD, sizeRealTurbine) );
+ setMemsizeGPU(sizeof(real)*(actuatorFarm->getNumberOfNodesPerBlade()+4)*actuatorFarm->getNumberOfTurbines(), false);
-void CudaMemoryManager::cudaCopyFluidNodeIndicesBorder(int lev) {
- uint mem_size_fluid_nodes_border = sizeof(uint) * parameter->getParH(lev)->numberOfFluidNodesBorder;
- checkCudaErrors(cudaMemcpy(parameter->getParD(lev)->fluidNodeIndicesBorder,
- parameter->getParH(lev)->fluidNodeIndicesBorder,
- mem_size_fluid_nodes_border, cudaMemcpyHostToDevice));
}
+void CudaMemoryManager::cudaCopyBladeGeometriesHtoD(ActuatorFarm* actuatorFarm)
+{
+ uint sizeRealTurbine = sizeof(real)*actuatorFarm->getNumberOfTurbines();
+ checkCudaErrors( cudaMemcpy(actuatorFarm->bladeRadiiD, actuatorFarm->bladeRadiiH, sizeRealTurbine*actuatorFarm->getNumberOfNodesPerBlade(), cudaMemcpyHostToDevice) );
+ checkCudaErrors( cudaMemcpy(actuatorFarm->diametersD, actuatorFarm->diametersH, sizeRealTurbine, cudaMemcpyHostToDevice) );
+ checkCudaErrors( cudaMemcpy(actuatorFarm->turbinePosXD, actuatorFarm->turbinePosXH, sizeRealTurbine, cudaMemcpyHostToDevice) );
+ checkCudaErrors( cudaMemcpy(actuatorFarm->turbinePosYD, actuatorFarm->turbinePosYH, sizeRealTurbine, cudaMemcpyHostToDevice) );
+ checkCudaErrors( cudaMemcpy(actuatorFarm->turbinePosZD, actuatorFarm->turbinePosZH, sizeRealTurbine, cudaMemcpyHostToDevice) );
-void CudaMemoryManager::cudaFreeFluidNodeIndicesBorder(int lev) {
- checkCudaErrors(cudaFreeHost(parameter->getParH(lev)->fluidNodeIndicesBorder));
}
+void CudaMemoryManager::cudaCopyBladeGeometriesDtoH(ActuatorFarm* actuatorFarm)
+{
+ uint sizeRealTurbine = sizeof(real)*actuatorFarm->getNumberOfTurbines();
+ checkCudaErrors( cudaMemcpy(actuatorFarm->bladeRadiiH, actuatorFarm->bladeRadiiD, sizeRealTurbine*actuatorFarm->getNumberOfNodesPerBlade(), cudaMemcpyDeviceToHost) );
+ checkCudaErrors( cudaMemcpy(actuatorFarm->diametersH, actuatorFarm->diametersD, sizeRealTurbine, cudaMemcpyDeviceToHost) );
+ checkCudaErrors( cudaMemcpy(actuatorFarm->turbinePosXH, actuatorFarm->turbinePosXD, sizeRealTurbine, cudaMemcpyDeviceToHost) );
+ checkCudaErrors( cudaMemcpy(actuatorFarm->turbinePosYH, actuatorFarm->turbinePosYD, sizeRealTurbine, cudaMemcpyDeviceToHost) );
+ checkCudaErrors( cudaMemcpy(actuatorFarm->turbinePosZH, actuatorFarm->turbinePosZD, sizeRealTurbine, cudaMemcpyDeviceToHost) );
-////////////////////////////////////////////////////////////////////////////////////
-// ActuatorLine
-///////////////////////////////////////////////////////////////////////////////
+}
+void CudaMemoryManager::cudaFreeBladeGeometries(ActuatorFarm* actuatorFarm)
+{
+ checkCudaErrors( cudaFree(actuatorFarm->bladeRadiiD) );
+ checkCudaErrors( cudaFree(actuatorFarm->diametersD) );
+ checkCudaErrors( cudaFree(actuatorFarm->turbinePosXD) );
+ checkCudaErrors( cudaFree(actuatorFarm->turbinePosYD) );
+ checkCudaErrors( cudaFree(actuatorFarm->turbinePosZD) );
+
+ checkCudaErrors( cudaFreeHost(actuatorFarm->bladeRadiiH) );
+ checkCudaErrors( cudaFreeHost(actuatorFarm->diametersH) );
+ checkCudaErrors( cudaFreeHost(actuatorFarm->turbinePosXH) );
+ checkCudaErrors( cudaFreeHost(actuatorFarm->turbinePosYH) );
+ checkCudaErrors( cudaFreeHost(actuatorFarm->turbinePosZH) );
+}
-void CudaMemoryManager::cudaAllocBladeRadii(ActuatorLine* actuatorLine)
+void CudaMemoryManager::cudaAllocBladeOrientations(ActuatorFarm* actuatorFarm)
{
- checkCudaErrors( cudaMallocHost((void**) &actuatorLine->bladeRadiiH, sizeof(real)*actuatorLine->getNBladeNodes()) );
+ uint sizeRealTurbine = sizeof(real)*actuatorFarm->getNumberOfTurbines();
+ checkCudaErrors( cudaMallocHost((void**) &actuatorFarm->omegasH, sizeRealTurbine) );
+ checkCudaErrors( cudaMallocHost((void**) &actuatorFarm->azimuthsH, sizeRealTurbine) );
+ checkCudaErrors( cudaMallocHost((void**) &actuatorFarm->yawsH, sizeRealTurbine) );
- checkCudaErrors( cudaMalloc((void**) &actuatorLine->bladeRadiiD, sizeof(real)*actuatorLine->getNBladeNodes()) );
+ checkCudaErrors( cudaMalloc((void**) &actuatorFarm->omegasD, sizeRealTurbine) );
+ checkCudaErrors( cudaMalloc((void**) &actuatorFarm->azimuthsD, sizeRealTurbine) );
+ checkCudaErrors( cudaMalloc((void**) &actuatorFarm->yawsD, sizeRealTurbine) );
- setMemsizeGPU(sizeof(real)*actuatorLine->getNBladeNodes(), false);
-}
+ setMemsizeGPU(3*sizeRealTurbine, false);
-void CudaMemoryManager::cudaCopyBladeRadiiHtoD(ActuatorLine* actuatorLine)
-{
- checkCudaErrors( cudaMemcpy(actuatorLine->bladeRadiiD, actuatorLine->bladeRadiiH, sizeof(real)*actuatorLine->getNBladeNodes(), cudaMemcpyHostToDevice) );
}
+void CudaMemoryManager::cudaCopyBladeOrientationsHtoD(ActuatorFarm* actuatorFarm)
+{
+ uint sizeRealTurbine = sizeof(real)*actuatorFarm->getNumberOfTurbines();
+ checkCudaErrors( cudaMemcpy(actuatorFarm->omegasD, actuatorFarm->omegasH, sizeRealTurbine, cudaMemcpyHostToDevice) );
+ checkCudaErrors( cudaMemcpy(actuatorFarm->azimuthsD, actuatorFarm->azimuthsH, sizeRealTurbine, cudaMemcpyHostToDevice) );
+ checkCudaErrors( cudaMemcpy(actuatorFarm->yawsD, actuatorFarm->yawsH, sizeRealTurbine, cudaMemcpyHostToDevice) );
-void CudaMemoryManager::cudaCopyBladeRadiiDtoH(ActuatorLine* actuatorLine)
+}
+void CudaMemoryManager::cudaCopyBladeOrientationsDtoH(ActuatorFarm* actuatorFarm)
{
- checkCudaErrors( cudaMemcpy(actuatorLine->bladeRadiiH, actuatorLine->bladeRadiiD, sizeof(real)*actuatorLine->getNBladeNodes(), cudaMemcpyDeviceToHost) );
+ uint sizeRealTurbine = sizeof(real)*actuatorFarm->getNumberOfTurbines();
+ checkCudaErrors( cudaMemcpy(actuatorFarm->omegasH, actuatorFarm->omegasD, sizeRealTurbine, cudaMemcpyDeviceToHost) );
+ checkCudaErrors( cudaMemcpy(actuatorFarm->azimuthsH, actuatorFarm->azimuthsD, sizeRealTurbine, cudaMemcpyDeviceToHost) );
+ checkCudaErrors( cudaMemcpy(actuatorFarm->yawsH, actuatorFarm->yawsD, sizeRealTurbine, cudaMemcpyDeviceToHost) );
}
-
-void CudaMemoryManager::cudaFreeBladeRadii(ActuatorLine* actuatorLine)
+void CudaMemoryManager::cudaFreeBladeOrientations(ActuatorFarm* actuatorFarm)
{
- checkCudaErrors( cudaFree(actuatorLine->bladeRadiiD) );
+ checkCudaErrors( cudaFree((void**) &actuatorFarm->omegasD) );
+ checkCudaErrors( cudaFree((void**) &actuatorFarm->azimuthsD) );
+ checkCudaErrors( cudaFree((void**) &actuatorFarm->yawsD) );
- checkCudaErrors( cudaFreeHost(actuatorLine->bladeRadiiH) );
+ checkCudaErrors( cudaFreeHost((void**) &actuatorFarm->omegasH) );
+ checkCudaErrors( cudaFreeHost((void**) &actuatorFarm->azimuthsH) );
+ checkCudaErrors( cudaFreeHost((void**) &actuatorFarm->yawsH) );
}
-void CudaMemoryManager::cudaAllocBladeCoords(ActuatorLine* actuatorLine)
+void CudaMemoryManager::cudaAllocBladeCoords(ActuatorFarm* actuatorFarm)
{
- checkCudaErrors( cudaMallocHost((void**) &actuatorLine->bladeCoordsXH, sizeof(real)*actuatorLine->getNNodes()) );
- checkCudaErrors( cudaMallocHost((void**) &actuatorLine->bladeCoordsYH, sizeof(real)*actuatorLine->getNNodes()) );
- checkCudaErrors( cudaMallocHost((void**) &actuatorLine->bladeCoordsZH, sizeof(real)*actuatorLine->getNNodes()) );
+ checkCudaErrors( cudaMallocHost((void**) &actuatorFarm->bladeCoordsXH, sizeof(real)*actuatorFarm->getNumberOfNodes()) );
+ checkCudaErrors( cudaMallocHost((void**) &actuatorFarm->bladeCoordsYH, sizeof(real)*actuatorFarm->getNumberOfNodes()) );
+ checkCudaErrors( cudaMallocHost((void**) &actuatorFarm->bladeCoordsZH, sizeof(real)*actuatorFarm->getNumberOfNodes()) );
- checkCudaErrors( cudaMalloc((void**) &actuatorLine->bladeCoordsXD, sizeof(real)*actuatorLine->getNNodes()) );
- checkCudaErrors( cudaMalloc((void**) &actuatorLine->bladeCoordsYD, sizeof(real)*actuatorLine->getNNodes()) );
- checkCudaErrors( cudaMalloc((void**) &actuatorLine->bladeCoordsZD, sizeof(real)*actuatorLine->getNNodes()) );
+ checkCudaErrors( cudaMalloc((void**) &actuatorFarm->bladeCoordsXDCurrentTimestep, sizeof(real)*actuatorFarm->getNumberOfNodes()) );
+ checkCudaErrors( cudaMalloc((void**) &actuatorFarm->bladeCoordsYDCurrentTimestep, sizeof(real)*actuatorFarm->getNumberOfNodes()) );
+ checkCudaErrors( cudaMalloc((void**) &actuatorFarm->bladeCoordsZDCurrentTimestep, sizeof(real)*actuatorFarm->getNumberOfNodes()) );
+
+ checkCudaErrors( cudaMalloc((void**) &actuatorFarm->bladeCoordsXDPreviousTimestep, sizeof(real)*actuatorFarm->getNumberOfNodes()) );
+ checkCudaErrors( cudaMalloc((void**) &actuatorFarm->bladeCoordsYDPreviousTimestep, sizeof(real)*actuatorFarm->getNumberOfNodes()) );
+ checkCudaErrors( cudaMalloc((void**) &actuatorFarm->bladeCoordsZDPreviousTimestep, sizeof(real)*actuatorFarm->getNumberOfNodes()) );
- setMemsizeGPU(3.f*actuatorLine->getNNodes(), false);
+ setMemsizeGPU(6.f*actuatorFarm->getNumberOfNodes(), false);
}
-void CudaMemoryManager::cudaCopyBladeCoordsHtoD(ActuatorLine* actuatorLine)
+void CudaMemoryManager::cudaCopyBladeCoordsHtoD(ActuatorFarm* actuatorFarm)
{
- checkCudaErrors( cudaMemcpy(actuatorLine->bladeCoordsXD, actuatorLine->bladeCoordsXH, sizeof(real)*actuatorLine->getNNodes(), cudaMemcpyHostToDevice) );
- checkCudaErrors( cudaMemcpy(actuatorLine->bladeCoordsYD, actuatorLine->bladeCoordsYH, sizeof(real)*actuatorLine->getNNodes(), cudaMemcpyHostToDevice) );
- checkCudaErrors( cudaMemcpy(actuatorLine->bladeCoordsZD, actuatorLine->bladeCoordsZH, sizeof(real)*actuatorLine->getNNodes(), cudaMemcpyHostToDevice) );
+ checkCudaErrors( cudaMemcpy(actuatorFarm->bladeCoordsXDCurrentTimestep, actuatorFarm->bladeCoordsXH, sizeof(real)*actuatorFarm->getNumberOfNodes(), cudaMemcpyHostToDevice) );
+ checkCudaErrors( cudaMemcpy(actuatorFarm->bladeCoordsYDCurrentTimestep, actuatorFarm->bladeCoordsYH, sizeof(real)*actuatorFarm->getNumberOfNodes(), cudaMemcpyHostToDevice) );
+ checkCudaErrors( cudaMemcpy(actuatorFarm->bladeCoordsZDCurrentTimestep, actuatorFarm->bladeCoordsZH, sizeof(real)*actuatorFarm->getNumberOfNodes(), cudaMemcpyHostToDevice) );
}
-void CudaMemoryManager::cudaCopyBladeCoordsDtoH(ActuatorLine* actuatorLine)
+void CudaMemoryManager::cudaCopyBladeCoordsDtoH(ActuatorFarm* actuatorFarm)
{
- checkCudaErrors( cudaMemcpy(actuatorLine->bladeCoordsXH, actuatorLine->bladeCoordsXD, sizeof(real)*actuatorLine->getNNodes(), cudaMemcpyDeviceToHost) );
- checkCudaErrors( cudaMemcpy(actuatorLine->bladeCoordsYH, actuatorLine->bladeCoordsYD, sizeof(real)*actuatorLine->getNNodes(), cudaMemcpyDeviceToHost) );
- checkCudaErrors( cudaMemcpy(actuatorLine->bladeCoordsZH, actuatorLine->bladeCoordsZD, sizeof(real)*actuatorLine->getNNodes(), cudaMemcpyDeviceToHost) );
+ checkCudaErrors( cudaMemcpy(actuatorFarm->bladeCoordsXH, actuatorFarm->bladeCoordsXDCurrentTimestep, sizeof(real)*actuatorFarm->getNumberOfNodes(), cudaMemcpyDeviceToHost) );
+ checkCudaErrors( cudaMemcpy(actuatorFarm->bladeCoordsYH, actuatorFarm->bladeCoordsYDCurrentTimestep, sizeof(real)*actuatorFarm->getNumberOfNodes(), cudaMemcpyDeviceToHost) );
+ checkCudaErrors( cudaMemcpy(actuatorFarm->bladeCoordsZH, actuatorFarm->bladeCoordsZDCurrentTimestep, sizeof(real)*actuatorFarm->getNumberOfNodes(), cudaMemcpyDeviceToHost) );
}
-void CudaMemoryManager::cudaFreeBladeCoords(ActuatorLine* actuatorLine)
+void CudaMemoryManager::cudaFreeBladeCoords(ActuatorFarm* actuatorFarm)
{
- checkCudaErrors( cudaFree(actuatorLine->bladeCoordsXD) );
- checkCudaErrors( cudaFree(actuatorLine->bladeCoordsYD) );
- checkCudaErrors( cudaFree(actuatorLine->bladeCoordsZD) );
+ checkCudaErrors( cudaFree(actuatorFarm->bladeCoordsXDCurrentTimestep) );
+ checkCudaErrors( cudaFree(actuatorFarm->bladeCoordsYDCurrentTimestep) );
+ checkCudaErrors( cudaFree(actuatorFarm->bladeCoordsZDCurrentTimestep) );
- checkCudaErrors( cudaFreeHost(actuatorLine->bladeCoordsXH) );
- checkCudaErrors( cudaFreeHost(actuatorLine->bladeCoordsYH) );
- checkCudaErrors( cudaFreeHost(actuatorLine->bladeCoordsZH) );
+ checkCudaErrors( cudaFree(actuatorFarm->bladeCoordsXDPreviousTimestep) );
+ checkCudaErrors( cudaFree(actuatorFarm->bladeCoordsYDPreviousTimestep) );
+ checkCudaErrors( cudaFree(actuatorFarm->bladeCoordsZDPreviousTimestep) );
+
+ checkCudaErrors( cudaFreeHost(actuatorFarm->bladeCoordsXH) );
+ checkCudaErrors( cudaFreeHost(actuatorFarm->bladeCoordsYH) );
+ checkCudaErrors( cudaFreeHost(actuatorFarm->bladeCoordsZH) );
}
-void CudaMemoryManager::cudaAllocBladeIndices(ActuatorLine* actuatorLine)
+void CudaMemoryManager::cudaAllocBladeIndices(ActuatorFarm* actuatorFarm)
{
- checkCudaErrors( cudaMallocHost((void**) &actuatorLine->bladeIndicesH, sizeof(uint)*actuatorLine->getNNodes()) );
+ checkCudaErrors( cudaMallocHost((void**) &actuatorFarm->bladeIndicesH, sizeof(uint)*actuatorFarm->getNumberOfNodes()) );
- checkCudaErrors( cudaMalloc((void**) &actuatorLine->bladeIndicesD, sizeof(uint)*actuatorLine->getNNodes()) );
+ checkCudaErrors( cudaMalloc((void**) &actuatorFarm->bladeIndicesD, sizeof(uint)*actuatorFarm->getNumberOfNodes()) );
- setMemsizeGPU(sizeof(uint)*actuatorLine->getNNodes(), false);
+ setMemsizeGPU(sizeof(uint)*actuatorFarm->getNumberOfNodes(), false);
}
-void CudaMemoryManager::cudaCopyBladeIndicesHtoD(ActuatorLine* actuatorLine)
+void CudaMemoryManager::cudaCopyBladeIndicesHtoD(ActuatorFarm* actuatorFarm)
{
- checkCudaErrors( cudaMemcpy(actuatorLine->bladeIndicesD, actuatorLine->bladeIndicesH, sizeof(uint)*actuatorLine->getNNodes(), cudaMemcpyHostToDevice) );
+ checkCudaErrors( cudaMemcpy(actuatorFarm->bladeIndicesD, actuatorFarm->bladeIndicesH, sizeof(uint)*actuatorFarm->getNumberOfNodes(), cudaMemcpyHostToDevice) );
}
-void CudaMemoryManager::cudaFreeBladeIndices(ActuatorLine* actuatorLine)
+void CudaMemoryManager::cudaFreeBladeIndices(ActuatorFarm* actuatorFarm)
{
- checkCudaErrors( cudaFree(actuatorLine->bladeIndicesD) );
+ checkCudaErrors( cudaFree(actuatorFarm->bladeIndicesD) );
- checkCudaErrors( cudaFreeHost(actuatorLine->bladeIndicesH) );
+ checkCudaErrors( cudaFreeHost(actuatorFarm->bladeIndicesH) );
}
-void CudaMemoryManager::cudaAllocBladeVelocities(ActuatorLine* actuatorLine)
+void CudaMemoryManager::cudaAllocBladeVelocities(ActuatorFarm* actuatorFarm)
{
- checkCudaErrors( cudaMallocHost((void**) &actuatorLine->bladeVelocitiesXH, sizeof(real)*actuatorLine->getNNodes()) );
- checkCudaErrors( cudaMallocHost((void**) &actuatorLine->bladeVelocitiesYH, sizeof(real)*actuatorLine->getNNodes()) );
- checkCudaErrors( cudaMallocHost((void**) &actuatorLine->bladeVelocitiesZH, sizeof(real)*actuatorLine->getNNodes()) );
+ checkCudaErrors( cudaMallocHost((void**) &actuatorFarm->bladeVelocitiesXH, sizeof(real)*actuatorFarm->getNumberOfNodes()) );
+ checkCudaErrors( cudaMallocHost((void**) &actuatorFarm->bladeVelocitiesYH, sizeof(real)*actuatorFarm->getNumberOfNodes()) );
+ checkCudaErrors( cudaMallocHost((void**) &actuatorFarm->bladeVelocitiesZH, sizeof(real)*actuatorFarm->getNumberOfNodes()) );
+
+ checkCudaErrors( cudaMalloc((void**) &actuatorFarm->bladeVelocitiesXDCurrentTimestep, sizeof(real)*actuatorFarm->getNumberOfNodes()) );
+ checkCudaErrors( cudaMalloc((void**) &actuatorFarm->bladeVelocitiesYDCurrentTimestep, sizeof(real)*actuatorFarm->getNumberOfNodes()) );
+ checkCudaErrors( cudaMalloc((void**) &actuatorFarm->bladeVelocitiesZDCurrentTimestep, sizeof(real)*actuatorFarm->getNumberOfNodes()) );
- checkCudaErrors( cudaMalloc((void**) &actuatorLine->bladeVelocitiesXD, sizeof(real)*actuatorLine->getNNodes()) );
- checkCudaErrors( cudaMalloc((void**) &actuatorLine->bladeVelocitiesYD, sizeof(real)*actuatorLine->getNNodes()) );
- checkCudaErrors( cudaMalloc((void**) &actuatorLine->bladeVelocitiesZD, sizeof(real)*actuatorLine->getNNodes()) );
+ checkCudaErrors( cudaMalloc((void**) &actuatorFarm->bladeVelocitiesXDPreviousTimestep, sizeof(real)*actuatorFarm->getNumberOfNodes()) );
+ checkCudaErrors( cudaMalloc((void**) &actuatorFarm->bladeVelocitiesYDPreviousTimestep, sizeof(real)*actuatorFarm->getNumberOfNodes()) );
+ checkCudaErrors( cudaMalloc((void**) &actuatorFarm->bladeVelocitiesZDPreviousTimestep, sizeof(real)*actuatorFarm->getNumberOfNodes()) );
- setMemsizeGPU(3.*sizeof(real)*actuatorLine->getNNodes(), false);
+ setMemsizeGPU(3.*sizeof(real)*actuatorFarm->getNumberOfNodes(), false);
}
-void CudaMemoryManager::cudaCopyBladeVelocitiesHtoD(ActuatorLine* actuatorLine)
+void CudaMemoryManager::cudaCopyBladeVelocitiesHtoD(ActuatorFarm* actuatorFarm)
{
- checkCudaErrors( cudaMemcpy(actuatorLine->bladeVelocitiesXD, actuatorLine->bladeVelocitiesXH, sizeof(real)*actuatorLine->getNNodes(), cudaMemcpyHostToDevice) );
- checkCudaErrors( cudaMemcpy(actuatorLine->bladeVelocitiesYD, actuatorLine->bladeVelocitiesYH, sizeof(real)*actuatorLine->getNNodes(), cudaMemcpyHostToDevice) );
- checkCudaErrors( cudaMemcpy(actuatorLine->bladeVelocitiesZD, actuatorLine->bladeVelocitiesZH, sizeof(real)*actuatorLine->getNNodes(), cudaMemcpyHostToDevice) );
+ checkCudaErrors( cudaMemcpy(actuatorFarm->bladeVelocitiesXDCurrentTimestep, actuatorFarm->bladeVelocitiesXH, sizeof(real)*actuatorFarm->getNumberOfNodes(), cudaMemcpyHostToDevice) );
+ checkCudaErrors( cudaMemcpy(actuatorFarm->bladeVelocitiesYDCurrentTimestep, actuatorFarm->bladeVelocitiesYH, sizeof(real)*actuatorFarm->getNumberOfNodes(), cudaMemcpyHostToDevice) );
+ checkCudaErrors( cudaMemcpy(actuatorFarm->bladeVelocitiesZDCurrentTimestep, actuatorFarm->bladeVelocitiesZH, sizeof(real)*actuatorFarm->getNumberOfNodes(), cudaMemcpyHostToDevice) );
}
-void CudaMemoryManager::cudaCopyBladeVelocitiesDtoH(ActuatorLine* actuatorLine)
+void CudaMemoryManager::cudaCopyBladeVelocitiesDtoH(ActuatorFarm* actuatorFarm)
{
- checkCudaErrors( cudaMemcpy(actuatorLine->bladeVelocitiesXH, actuatorLine->bladeVelocitiesXD, sizeof(real)*actuatorLine->getNNodes(), cudaMemcpyDeviceToHost) );
- checkCudaErrors( cudaMemcpy(actuatorLine->bladeVelocitiesYH, actuatorLine->bladeVelocitiesYD, sizeof(real)*actuatorLine->getNNodes(), cudaMemcpyDeviceToHost) );
- checkCudaErrors( cudaMemcpy(actuatorLine->bladeVelocitiesZH, actuatorLine->bladeVelocitiesZD, sizeof(real)*actuatorLine->getNNodes(), cudaMemcpyDeviceToHost) );
+ checkCudaErrors( cudaMemcpy(actuatorFarm->bladeVelocitiesXH, actuatorFarm->bladeVelocitiesXDCurrentTimestep, sizeof(real)*actuatorFarm->getNumberOfNodes(), cudaMemcpyDeviceToHost) );
+ checkCudaErrors( cudaMemcpy(actuatorFarm->bladeVelocitiesYH, actuatorFarm->bladeVelocitiesYDCurrentTimestep, sizeof(real)*actuatorFarm->getNumberOfNodes(), cudaMemcpyDeviceToHost) );
+ checkCudaErrors( cudaMemcpy(actuatorFarm->bladeVelocitiesZH, actuatorFarm->bladeVelocitiesZDCurrentTimestep, sizeof(real)*actuatorFarm->getNumberOfNodes(), cudaMemcpyDeviceToHost) );
}
-void CudaMemoryManager::cudaFreeBladeVelocities(ActuatorLine* actuatorLine)
+void CudaMemoryManager::cudaFreeBladeVelocities(ActuatorFarm* actuatorFarm)
{
- checkCudaErrors( cudaFree(actuatorLine->bladeVelocitiesXD) );
- checkCudaErrors( cudaFree(actuatorLine->bladeVelocitiesYD) );
- checkCudaErrors( cudaFree(actuatorLine->bladeVelocitiesZD) );
+ checkCudaErrors( cudaFree(actuatorFarm->bladeVelocitiesXDCurrentTimestep) );
+ checkCudaErrors( cudaFree(actuatorFarm->bladeVelocitiesYDCurrentTimestep) );
+ checkCudaErrors( cudaFree(actuatorFarm->bladeVelocitiesZDCurrentTimestep) );
+
+ checkCudaErrors( cudaFree(actuatorFarm->bladeVelocitiesXDPreviousTimestep) );
+ checkCudaErrors( cudaFree(actuatorFarm->bladeVelocitiesYDPreviousTimestep) );
+ checkCudaErrors( cudaFree(actuatorFarm->bladeVelocitiesZDPreviousTimestep) );
- checkCudaErrors( cudaFreeHost(actuatorLine->bladeVelocitiesXH) );
- checkCudaErrors( cudaFreeHost(actuatorLine->bladeVelocitiesYH) );
- checkCudaErrors( cudaFreeHost(actuatorLine->bladeVelocitiesZH) );
+ checkCudaErrors( cudaFreeHost(actuatorFarm->bladeVelocitiesXH) );
+ checkCudaErrors( cudaFreeHost(actuatorFarm->bladeVelocitiesYH) );
+ checkCudaErrors( cudaFreeHost(actuatorFarm->bladeVelocitiesZH) );
}
-void CudaMemoryManager::cudaAllocBladeForces(ActuatorLine* actuatorLine)
+void CudaMemoryManager::cudaAllocBladeForces(ActuatorFarm* actuatorFarm)
{
- checkCudaErrors( cudaMallocHost((void**) &actuatorLine->bladeForcesXH, sizeof(real)*actuatorLine->getNNodes()) );
- checkCudaErrors( cudaMallocHost((void**) &actuatorLine->bladeForcesYH, sizeof(real)*actuatorLine->getNNodes()) );
- checkCudaErrors( cudaMallocHost((void**) &actuatorLine->bladeForcesZH, sizeof(real)*actuatorLine->getNNodes()) );
+ checkCudaErrors( cudaMallocHost((void**) &actuatorFarm->bladeForcesXH, sizeof(real)*actuatorFarm->getNumberOfNodes()) );
+ checkCudaErrors( cudaMallocHost((void**) &actuatorFarm->bladeForcesYH, sizeof(real)*actuatorFarm->getNumberOfNodes()) );
+ checkCudaErrors( cudaMallocHost((void**) &actuatorFarm->bladeForcesZH, sizeof(real)*actuatorFarm->getNumberOfNodes()) );
+
+ checkCudaErrors( cudaMalloc((void**) &actuatorFarm->bladeForcesXDCurrentTimestep, sizeof(real)*actuatorFarm->getNumberOfNodes()) );
+ checkCudaErrors( cudaMalloc((void**) &actuatorFarm->bladeForcesYDCurrentTimestep, sizeof(real)*actuatorFarm->getNumberOfNodes()) );
+ checkCudaErrors( cudaMalloc((void**) &actuatorFarm->bladeForcesZDCurrentTimestep, sizeof(real)*actuatorFarm->getNumberOfNodes()) );
- checkCudaErrors( cudaMalloc((void**) &actuatorLine->bladeForcesXD, sizeof(real)*actuatorLine->getNNodes()) );
- checkCudaErrors( cudaMalloc((void**) &actuatorLine->bladeForcesYD, sizeof(real)*actuatorLine->getNNodes()) );
- checkCudaErrors( cudaMalloc((void**) &actuatorLine->bladeForcesZD, sizeof(real)*actuatorLine->getNNodes()) );
+ checkCudaErrors( cudaMalloc((void**) &actuatorFarm->bladeForcesXDPreviousTimestep, sizeof(real)*actuatorFarm->getNumberOfNodes()) );
+ checkCudaErrors( cudaMalloc((void**) &actuatorFarm->bladeForcesYDPreviousTimestep, sizeof(real)*actuatorFarm->getNumberOfNodes()) );
+ checkCudaErrors( cudaMalloc((void**) &actuatorFarm->bladeForcesZDPreviousTimestep, sizeof(real)*actuatorFarm->getNumberOfNodes()) );
- setMemsizeGPU(3.*sizeof(real)*actuatorLine->getNNodes(), false);
+ setMemsizeGPU(3.*sizeof(real)*actuatorFarm->getNumberOfNodes(), false);
}
-void CudaMemoryManager::cudaCopyBladeForcesHtoD(ActuatorLine* actuatorLine)
+void CudaMemoryManager::cudaCopyBladeForcesHtoD(ActuatorFarm* actuatorFarm)
{
- checkCudaErrors( cudaMemcpy(actuatorLine->bladeForcesXD, actuatorLine->bladeForcesXH, sizeof(real)*actuatorLine->getNNodes(), cudaMemcpyHostToDevice) );
- checkCudaErrors( cudaMemcpy(actuatorLine->bladeForcesYD, actuatorLine->bladeForcesYH, sizeof(real)*actuatorLine->getNNodes(), cudaMemcpyHostToDevice) );
- checkCudaErrors( cudaMemcpy(actuatorLine->bladeForcesZD, actuatorLine->bladeForcesZH, sizeof(real)*actuatorLine->getNNodes(), cudaMemcpyHostToDevice) );
+ checkCudaErrors( cudaMemcpy(actuatorFarm->bladeForcesXDCurrentTimestep, actuatorFarm->bladeForcesXH, sizeof(real)*actuatorFarm->getNumberOfNodes(), cudaMemcpyHostToDevice) );
+ checkCudaErrors( cudaMemcpy(actuatorFarm->bladeForcesYDCurrentTimestep, actuatorFarm->bladeForcesYH, sizeof(real)*actuatorFarm->getNumberOfNodes(), cudaMemcpyHostToDevice) );
+ checkCudaErrors( cudaMemcpy(actuatorFarm->bladeForcesZDCurrentTimestep, actuatorFarm->bladeForcesZH, sizeof(real)*actuatorFarm->getNumberOfNodes(), cudaMemcpyHostToDevice) );
}
-void CudaMemoryManager::cudaCopyBladeForcesDtoH(ActuatorLine* actuatorLine)
+void CudaMemoryManager::cudaCopyBladeForcesDtoH(ActuatorFarm* actuatorFarm)
{
- checkCudaErrors( cudaMemcpy(actuatorLine->bladeForcesXH, actuatorLine->bladeForcesXD, sizeof(real)*actuatorLine->getNNodes(), cudaMemcpyDeviceToHost) );
- checkCudaErrors( cudaMemcpy(actuatorLine->bladeForcesYH, actuatorLine->bladeForcesYD, sizeof(real)*actuatorLine->getNNodes(), cudaMemcpyDeviceToHost) );
- checkCudaErrors( cudaMemcpy(actuatorLine->bladeForcesZH, actuatorLine->bladeForcesZD, sizeof(real)*actuatorLine->getNNodes(), cudaMemcpyDeviceToHost) );
+ checkCudaErrors( cudaMemcpy(actuatorFarm->bladeForcesXH, actuatorFarm->bladeForcesXDCurrentTimestep, sizeof(real)*actuatorFarm->getNumberOfNodes(), cudaMemcpyDeviceToHost) );
+ checkCudaErrors( cudaMemcpy(actuatorFarm->bladeForcesYH, actuatorFarm->bladeForcesYDCurrentTimestep, sizeof(real)*actuatorFarm->getNumberOfNodes(), cudaMemcpyDeviceToHost) );
+ checkCudaErrors( cudaMemcpy(actuatorFarm->bladeForcesZH, actuatorFarm->bladeForcesZDCurrentTimestep, sizeof(real)*actuatorFarm->getNumberOfNodes(), cudaMemcpyDeviceToHost) );
}
-void CudaMemoryManager::cudaFreeBladeForces(ActuatorLine* actuatorLine)
+void CudaMemoryManager::cudaFreeBladeForces(ActuatorFarm* actuatorFarm)
{
- checkCudaErrors( cudaFree(actuatorLine->bladeForcesXD) );
- checkCudaErrors( cudaFree(actuatorLine->bladeForcesYD) );
- checkCudaErrors( cudaFree(actuatorLine->bladeForcesZD) );
+ checkCudaErrors( cudaFree(actuatorFarm->bladeForcesXDCurrentTimestep) );
+ checkCudaErrors( cudaFree(actuatorFarm->bladeForcesYDCurrentTimestep) );
+ checkCudaErrors( cudaFree(actuatorFarm->bladeForcesZDCurrentTimestep) );
- checkCudaErrors( cudaFreeHost(actuatorLine->bladeForcesXH) );
- checkCudaErrors( cudaFreeHost(actuatorLine->bladeForcesYH) );
- checkCudaErrors( cudaFreeHost(actuatorLine->bladeForcesZH) );
+ checkCudaErrors( cudaFree(actuatorFarm->bladeForcesXDPreviousTimestep) );
+ checkCudaErrors( cudaFree(actuatorFarm->bladeForcesYDPreviousTimestep) );
+ checkCudaErrors( cudaFree(actuatorFarm->bladeForcesZDPreviousTimestep) );
+
+ checkCudaErrors( cudaFreeHost(actuatorFarm->bladeForcesXH) );
+ checkCudaErrors( cudaFreeHost(actuatorFarm->bladeForcesYH) );
+ checkCudaErrors( cudaFreeHost(actuatorFarm->bladeForcesZH) );
}
-void CudaMemoryManager::cudaAllocSphereIndices(ActuatorLine* actuatorLine)
+void CudaMemoryManager::cudaAllocSphereIndices(ActuatorFarm* actuatorFarm)
{
- checkCudaErrors( cudaMallocHost((void**) &(actuatorLine->boundingSphereIndicesH), sizeof(int)*actuatorLine->getNIndices()));
- checkCudaErrors( cudaMalloc((void**) &(actuatorLine->boundingSphereIndicesD), sizeof(int)*actuatorLine->getNIndices()));
- setMemsizeGPU(sizeof(int)*actuatorLine->getNIndices(), false);
+ checkCudaErrors( cudaMallocHost((void**) &(actuatorFarm->boundingSphereIndicesH), sizeof(int)*actuatorFarm->getNumberOfIndices()));
+ checkCudaErrors( cudaMalloc((void**) &(actuatorFarm->boundingSphereIndicesD), sizeof(int)*actuatorFarm->getNumberOfIndices()));
+ setMemsizeGPU(sizeof(int)*actuatorFarm->getNumberOfIndices(), false);
}
-void CudaMemoryManager::cudaCopySphereIndicesHtoD(ActuatorLine* actuatorLine)
+void CudaMemoryManager::cudaCopySphereIndicesHtoD(ActuatorFarm* actuatorFarm)
{
- checkCudaErrors( cudaMemcpy(actuatorLine->boundingSphereIndicesD, actuatorLine->boundingSphereIndicesH, sizeof(int)*actuatorLine->getNIndices(), cudaMemcpyHostToDevice) );
+ checkCudaErrors( cudaMemcpy(actuatorFarm->boundingSphereIndicesD, actuatorFarm->boundingSphereIndicesH, sizeof(int)*actuatorFarm->getNumberOfIndices(), cudaMemcpyHostToDevice) );
}
-void CudaMemoryManager::cudaFreeSphereIndices(ActuatorLine* actuatorLine)
+void CudaMemoryManager::cudaFreeSphereIndices(ActuatorFarm* actuatorFarm)
{
- checkCudaErrors( cudaFreeHost(actuatorLine->boundingSphereIndicesH) );
- checkCudaErrors( cudaFree(actuatorLine->boundingSphereIndicesD) );
+ checkCudaErrors( cudaFreeHost(actuatorFarm->boundingSphereIndicesH) );
+ checkCudaErrors( cudaFree(actuatorFarm->boundingSphereIndicesD) );
}
////////////////////////////////////////////////////////////////////////////////////
@@ -3231,8 +3416,11 @@ void CudaMemoryManager::cudaCopyProbeQuantityArrayHtoD(Probe* probe, int level)
}
void CudaMemoryManager::cudaCopyProbeQuantityArrayDtoH(Probe* probe, int level)
{
- checkCudaErrors( cudaMemcpy(probe->getProbeStruct(level)->quantitiesArrayH, probe->getProbeStruct(level)->quantitiesArrayD, probe->getProbeStruct(level)->nArrays*sizeof(real)*probe->getProbeStruct(level)->nPoints, cudaMemcpyDeviceToHost) );
+ auto probeStruct = probe->getProbeStruct(level);
+
+ checkCudaErrors( cudaMemcpy(probeStruct->quantitiesArrayH, probeStruct->quantitiesArrayD, probeStruct->nArrays*sizeof(real)*probeStruct->nPoints, cudaMemcpyDeviceToHost) );
}
+
void CudaMemoryManager::cudaFreeProbeQuantityArray(Probe* probe, int level)
{
checkCudaErrors( cudaFreeHost(probe->getProbeStruct(level)->quantitiesArrayH) );
@@ -3262,6 +3450,7 @@ void CudaMemoryManager::cudaCopyProbeQuantitiesAndOffsetsDtoH(Probe* probe, int
checkCudaErrors( cudaMemcpy(probe->getProbeStruct(level)->quantitiesH, probe->getProbeStruct(level)->quantitiesD, int(Statistic::LAST)*sizeof(bool), cudaMemcpyDeviceToHost) );
checkCudaErrors( cudaMemcpy(probe->getProbeStruct(level)->arrayOffsetsH, probe->getProbeStruct(level)->arrayOffsetsD, int(Statistic::LAST)*sizeof(int), cudaMemcpyDeviceToHost) );
}
+
void CudaMemoryManager::cudaFreeProbeQuantitiesAndOffsets(Probe* probe, int level)
{
checkCudaErrors( cudaFreeHost(probe->getProbeStruct(level)->quantitiesH) );
@@ -3270,23 +3459,51 @@ void CudaMemoryManager::cudaFreeProbeQuantitiesAndOffsets(Probe* probe, int leve
checkCudaErrors( cudaFree (probe->getProbeStruct(level)->arrayOffsetsD) );
}
+void CudaMemoryManager::cudaAllocPrecursorWriter(PrecursorWriter* writer, int level)
+{
+ auto prec = writer->getPrecursorStruct(level);
+ size_t indSize = prec->numberOfPointsInBC*sizeof(uint);
+ checkCudaErrors( cudaStreamCreate(&prec->stream) );
+ checkCudaErrors( cudaMallocHost((void**) &prec->indicesH, indSize));
+ checkCudaErrors( cudaMalloc((void**) &prec->indicesD, indSize));
+ size_t dataSize = prec->numberOfPointsInBC*sizeof(real)*prec->numberOfQuantities;
+ size_t dataSizeH = dataSize * prec->numberOfTimestepsPerFile;
+
+ checkCudaErrors( cudaMallocHost((void**) &prec->dataH, dataSizeH));
+ checkCudaErrors( cudaMallocHost((void**) &prec->bufferH, dataSizeH));
+ checkCudaErrors( cudaMalloc((void**) &prec->dataD, dataSize));
+ checkCudaErrors( cudaMalloc((void**) &prec->bufferD, dataSize));
+ setMemsizeGPU(indSize+2*dataSize, false);
+}
+void CudaMemoryManager::cudaCopyPrecursorWriterIndicesHtoD(PrecursorWriter* writer, int level)
+{
+ checkCudaErrors( cudaMemcpy(writer->getPrecursorStruct(level)->indicesD, writer->getPrecursorStruct(level)->indicesH, writer->getPrecursorStruct(level)->numberOfPointsInBC*sizeof(uint), cudaMemcpyHostToDevice) );
+}
+void CudaMemoryManager::cudaCopyPrecursorWriterOutputVariablesDtoH(PrecursorWriter* writer, int level)
+{
+ auto prec = writer->getPrecursorStruct(level);
+ int sizeTimestep = prec->numberOfPointsInBC*prec->numberOfQuantities;
+ checkCudaErrors( cudaStreamSynchronize(prec->stream) );
+ checkCudaErrors( cudaMemcpyAsync( &prec->bufferH[prec->numberOfTimestepsBuffered*sizeTimestep], prec->bufferD, sizeof(real)*sizeTimestep, cudaMemcpyDeviceToHost, prec->stream));
+}
+void CudaMemoryManager::cudaFreePrecursorWriter(PrecursorWriter* writer, int level)
+{
+ checkCudaErrors( cudaFreeHost(writer->getPrecursorStruct(level)->indicesH));
+ checkCudaErrors( cudaFree(writer->getPrecursorStruct(level)->indicesD));
-
-
-
-
-
-
-
-
+ checkCudaErrors( cudaFreeHost(writer->getPrecursorStruct(level)->dataH));
+ checkCudaErrors( cudaFreeHost(writer->getPrecursorStruct(level)->bufferH));
+ checkCudaErrors( cudaFree(writer->getPrecursorStruct(level)->dataD));
+ checkCudaErrors( cudaFree(writer->getPrecursorStruct(level)->bufferD));
+}
CudaMemoryManager::CudaMemoryManager(std::shared_ptr<Parameter> parameter) : parameter(parameter)
diff --git a/src/gpu/VirtualFluids_GPU/GPU/CudaMemoryManager.h b/src/gpu/VirtualFluids_GPU/GPU/CudaMemoryManager.h
index d410340d2de7797cf23a781a64d11f592d62a6fb..e2f2e8658b6ef7a9453546454dd8e1f643574e17 100644
--- a/src/gpu/VirtualFluids_GPU/GPU/CudaMemoryManager.h
+++ b/src/gpu/VirtualFluids_GPU/GPU/CudaMemoryManager.h
@@ -18,8 +18,10 @@
class Parameter;
class PorousMedia;
-class ActuatorLine;
+class ActuatorFarm;
class Probe;
+class VelocitySetter;
+class PrecursorWriter;
class VIRTUALFLUIDS_GPU_EXPORT CudaMemoryManager
{
@@ -30,8 +32,8 @@ public:
void setMemsizeGPU(double admem, bool reset);
double getMemsizeGPU();
- void cudaAllocFull(int lev);
- void cudaFreeFull(int lev);
+ //void cudaAllocFull(int lev); //DEPRECATED: related to full matrix
+ //void cudaFreeFull(int lev); //DEPRECATED: related to full matrix
void cudaCopyPrint(int lev);
void cudaCopyMedianPrint(int lev);
@@ -92,26 +94,20 @@ public:
//////////////////////////////////////////////////////////////////////////
//3D domain decomposition
virtual void cudaAllocProcessNeighborX(int lev, unsigned int processNeighbor);
- void cudaCopyProcessNeighborXFsHD(int lev, unsigned int processNeighbor, const unsigned int &memsizeFsRecv,
- int streamIndex);
- void cudaCopyProcessNeighborXFsDH(int lev, unsigned int processNeighbor, const unsigned int &memsizeFsSend,
- int streamIndex);
+ void cudaCopyProcessNeighborXFsHD(int lev, unsigned int processNeighbor, const unsigned int &memsizeFsRecv);
+ void cudaCopyProcessNeighborXFsDH(int lev, unsigned int processNeighbor, const unsigned int &memsizeFsSend);
virtual void cudaCopyProcessNeighborXIndex(int lev, unsigned int processNeighbor);
void cudaFreeProcessNeighborX(int lev, unsigned int processNeighbor);
//
virtual void cudaAllocProcessNeighborY(int lev, unsigned int processNeighbor);
- void cudaCopyProcessNeighborYFsHD(int lev, unsigned int processNeighbor, const unsigned int &memsizeFsRecv,
- int streamIndex);
- void cudaCopyProcessNeighborYFsDH(int lev, unsigned int processNeighbor, const unsigned int &memsizeFsSend,
- int streamIndex);
+ void cudaCopyProcessNeighborYFsHD(int lev, unsigned int processNeighbor, const unsigned int &memsizeFsRecv);
+ void cudaCopyProcessNeighborYFsDH(int lev, unsigned int processNeighbor, const unsigned int &memsizeFsSend);
virtual void cudaCopyProcessNeighborYIndex(int lev, unsigned int processNeighbor);
void cudaFreeProcessNeighborY(int lev, unsigned int processNeighbor);
//
virtual void cudaAllocProcessNeighborZ(int lev, unsigned int processNeighbor);
- void cudaCopyProcessNeighborZFsHD(int lev, unsigned int processNeighbor, const unsigned int &memsizeFsRecv,
- int streamIndex);
- void cudaCopyProcessNeighborZFsDH(int lev, unsigned int processNeighbor, const unsigned int &memsizeFsSend,
- int streamIndex);
+ void cudaCopyProcessNeighborZFsHD(int lev, unsigned int processNeighbor, const unsigned int &memsizeFsRecv);
+ void cudaCopyProcessNeighborZFsDH(int lev, unsigned int processNeighbor, const unsigned int &memsizeFsSend);
virtual void cudaCopyProcessNeighborZIndex(int lev, unsigned int processNeighbor);
void cudaFreeProcessNeighborZ(int lev, unsigned int processNeighbor);
@@ -183,6 +179,13 @@ public:
void cudaCopyStressBC(int lev);
void cudaFreeStressBC(int lev);
+ void cudaAllocPrecursorBC(int lev);
+ void cudaAllocPrecursorData(int lev);
+ void cudaCopyPrecursorBC(int lev);
+ void cudaCopyPrecursorData(int lev);
+ void cudaFreePrecursorBC(int lev);
+ void cudaFreePrecursorData(int lev);
+
void cudaAllocWallModel(int lev, bool hasWallModelMonitor);
void cudaCopyWallModel(int lev, bool hasWallModelMonitor);
void cudaFreeWallModel(int lev, bool hasWallModelMonitor);
@@ -346,42 +349,44 @@ public:
void cudaCopyProcessNeighborADZIndex(int lev, unsigned int processNeighbor);
void cudaFreeProcessNeighborADZ(int lev, unsigned int processNeighbor);
- void cudaAllocFluidNodeIndices(int lev);
- void cudaCopyFluidNodeIndices(int lev);
- void cudaFreeFluidNodeIndices(int lev);
- void cudaAllocFluidNodeIndicesBorder(int lev);
- void cudaCopyFluidNodeIndicesBorder(int lev);
- void cudaFreeFluidNodeIndicesBorder(int lev);
-
- // Actuator Line
- void cudaAllocBladeRadii(ActuatorLine* actuatorLine);
- void cudaCopyBladeRadiiHtoD(ActuatorLine* actuatorLine);
- void cudaCopyBladeRadiiDtoH(ActuatorLine* actuatorLine);
- void cudaFreeBladeRadii(ActuatorLine* actuatorLine);
-
- void cudaAllocBladeCoords(ActuatorLine* actuatorLine);
- void cudaCopyBladeCoordsHtoD(ActuatorLine* actuatorLine);
- void cudaCopyBladeCoordsDtoH(ActuatorLine* actuatorLine);
- void cudaFreeBladeCoords(ActuatorLine* actuatorLine);
-
- void cudaAllocBladeIndices(ActuatorLine* actuatorLine);
- void cudaCopyBladeIndicesHtoD(ActuatorLine* actuatorLine);
- void cudaFreeBladeIndices(ActuatorLine* actuatorLine);
-
- void cudaAllocBladeVelocities(ActuatorLine* actuatorLine);
- void cudaCopyBladeVelocitiesHtoD(ActuatorLine* actuatorLine);
- void cudaCopyBladeVelocitiesDtoH(ActuatorLine* actuatorLine);
- void cudaFreeBladeVelocities(ActuatorLine* actuatorLine);
-
- void cudaAllocBladeForces(ActuatorLine* actuatorLine);
- void cudaCopyBladeForcesHtoD(ActuatorLine* actuatorLine);
- void cudaCopyBladeForcesDtoH(ActuatorLine* actuatorLine);
- void cudaFreeBladeForces(ActuatorLine* actuatorLine);
-
- void cudaAllocSphereIndices(ActuatorLine* actuatorLine);
- void cudaCopySphereIndicesHtoD(ActuatorLine* actuatorLine);
- void cudaFreeSphereIndices(ActuatorLine* actuatorLine);
-
+ void cudaAllocTaggedFluidNodeIndices(CollisionTemplate tag, int lev);
+ void cudaCopyTaggedFluidNodeIndices(CollisionTemplate tag, int lev);
+ void cudaFreeTaggedFluidNodeIndices(CollisionTemplate tag, int lev);
+
+ // ActuatorFarm
+ void cudaAllocBladeGeometries(ActuatorFarm* actuatorFarm);
+ void cudaCopyBladeGeometriesHtoD(ActuatorFarm* actuatorFarm);
+ void cudaCopyBladeGeometriesDtoH(ActuatorFarm* actuatorFarm);
+ void cudaFreeBladeGeometries(ActuatorFarm* actuatorFarm);
+
+ void cudaAllocBladeOrientations(ActuatorFarm* actuatorFarm);
+ void cudaCopyBladeOrientationsHtoD(ActuatorFarm* actuatorFarm);
+ void cudaCopyBladeOrientationsDtoH(ActuatorFarm* actuatorFarm);
+ void cudaFreeBladeOrientations(ActuatorFarm* actuatorFarm);
+
+ void cudaAllocBladeCoords(ActuatorFarm* actuatorFarm);
+ void cudaCopyBladeCoordsHtoD(ActuatorFarm* actuatorFarm);
+ void cudaCopyBladeCoordsDtoH(ActuatorFarm* actuatorFarm);
+ void cudaFreeBladeCoords(ActuatorFarm* actuatorFarm);
+
+ void cudaAllocBladeIndices(ActuatorFarm* actuatorFarm);
+ void cudaCopyBladeIndicesHtoD(ActuatorFarm* actuatorFarm);
+ void cudaFreeBladeIndices(ActuatorFarm* actuatorFarm);
+
+ void cudaAllocBladeVelocities(ActuatorFarm* actuatorFarm);
+ void cudaCopyBladeVelocitiesHtoD(ActuatorFarm* actuatorFarm);
+ void cudaCopyBladeVelocitiesDtoH(ActuatorFarm* actuatorFarm);
+ void cudaFreeBladeVelocities(ActuatorFarm* actuatorFarm);
+
+ void cudaAllocBladeForces(ActuatorFarm* actuatorFarm);
+ void cudaCopyBladeForcesHtoD(ActuatorFarm* actuatorFarm);
+ void cudaCopyBladeForcesDtoH(ActuatorFarm* actuatorFarm);
+ void cudaFreeBladeForces(ActuatorFarm* actuatorFarm);
+
+ void cudaAllocSphereIndices(ActuatorFarm* actuatorFarm);
+ void cudaCopySphereIndicesHtoD(ActuatorFarm* actuatorFarm);
+ void cudaFreeSphereIndices(ActuatorFarm* actuatorFarm);
+ // Probes
void cudaAllocProbeDistances(Probe* probe, int level);
void cudaCopyProbeDistancesHtoD(Probe* probe, int level);
void cudaCopyProbeDistancesDtoH(Probe* probe, int level);
@@ -402,6 +407,12 @@ public:
void cudaCopyProbeQuantitiesAndOffsetsDtoH(Probe* probe, int level);
void cudaFreeProbeQuantitiesAndOffsets(Probe* probe, int level);
+ //Precursor Writer
+ void cudaAllocPrecursorWriter(PrecursorWriter* writer, int level);
+ void cudaCopyPrecursorWriterIndicesHtoD(PrecursorWriter* writer, int level);
+ void cudaCopyPrecursorWriterOutputVariablesDtoH(PrecursorWriter* writer, int level);
+ void cudaFreePrecursorWriter(PrecursorWriter* writer, int level);
+
private:
std::shared_ptr<Parameter> parameter;
double memsizeGPU = 0.0;
diff --git a/src/gpu/VirtualFluids_GPU/GPU/Cumulant27.cu b/src/gpu/VirtualFluids_GPU/GPU/Cumulant27.cu
index bbce8181d814fc8b9dbb086764becb73a86c0eda..553e1f34f7993a42682605b66d53407ede9292fd 100644
--- a/src/gpu/VirtualFluids_GPU/GPU/Cumulant27.cu
+++ b/src/gpu/VirtualFluids_GPU/GPU/Cumulant27.cu
@@ -21,7 +21,7 @@ __global__ void LB_Kernel_Kum_AA2016_Comp_Bulk_SP_27(real omega,
unsigned int* neighborY,
unsigned int* neighborZ,
real* DDStart,
- int size_Mat,
+ unsigned long long numberOfLBnodes,
int level,
real* forces,
bool EvenOrOdd)
@@ -37,7 +37,7 @@ __global__ void LB_Kernel_Kum_AA2016_Comp_Bulk_SP_27(real omega,
const unsigned k = nx*(ny*z + y) + x;
//////////////////////////////////////////////////////////////////////////
- if(k<size_Mat)
+ if(k<numberOfLBnodes)
{
////////////////////////////////////////////////////////////////////////////////
unsigned int BC;
@@ -48,63 +48,63 @@ __global__ void LB_Kernel_Kum_AA2016_Comp_Bulk_SP_27(real omega,
Distributions27 D;
if (EvenOrOdd==true)
{
- D.f[DIR_P00 ] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DDStart[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DDStart[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DDStart[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DDStart[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DDStart[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DDStart[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DDStart[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DDStart[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DDStart[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DDStart[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DDStart[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DDStart[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DDStart[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DDStart[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DDStart[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DDStart[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DDStart[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DDStart[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DDStart[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DDStart[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DDStart[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DDStart[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DDStart[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DDStart[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DDStart[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DDStart[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DDStart[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00 ] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_MMM ] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_PPM ] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_MPM ] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_PMM ] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_MMP ] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_PPP ] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_MPP ] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_PMP ] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_M00] = &DDStart[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DDStart[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DDStart[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DDStart[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DDStart[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DDStart[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DDStart[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DDStart[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DDStart[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DDStart[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DDStart[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DDStart[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DDStart[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DDStart[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DDStart[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DDStart[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DDStart[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DDStart[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DDStart[DIR_000 * numberOfLBnodes];
+ D.f[DIR_MMM] = &DDStart[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DDStart[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DDStart[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DDStart[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DDStart[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PPP] = &DDStart[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DDStart[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DDStart[DIR_MPM * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////
@@ -137,33 +137,33 @@ __global__ void LB_Kernel_Kum_AA2016_Comp_Bulk_SP_27(real omega,
//unsigned int ktne = k;
unsigned int kbsw = neighborZ[ksw];
//////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
- real mfcbb = (D.f[DIR_P00 ])[k ];//[ke ];// + c2over27 ;(D.f[DIR_P00 ])[k ];//ke
- real mfabb = (D.f[DIR_M00 ])[kw ];//[kw ];// + c2over27 ;(D.f[DIR_M00 ])[kw ];
- real mfbcb = (D.f[DIR_0P0 ])[k ];//[kn ];// + c2over27 ;(D.f[DIR_0P0 ])[k ];//kn
- real mfbab = (D.f[DIR_0M0 ])[ks ];//[ks ];// + c2over27 ;(D.f[DIR_0M0 ])[ks ];
- real mfbbc = (D.f[DIR_00P ])[k ];//[kt ];// + c2over27 ;(D.f[DIR_00P ])[k ];//kt
- real mfbba = (D.f[DIR_00M ])[kb ];//[kb ];// + c2over27 ;(D.f[DIR_00M ])[kb ];
- real mfccb = (D.f[DIR_PP0 ])[k ];//[kne ];// + c1over54 ;(D.f[DIR_PP0 ])[k ];//kne
- real mfaab = (D.f[DIR_MM0 ])[ksw];//[ksw ];// + c1over54 ;(D.f[DIR_MM0 ])[ksw];
- real mfcab = (D.f[DIR_PM0 ])[ks ];//[kse ];// + c1over54 ;(D.f[DIR_PM0 ])[ks ];//kse
- real mfacb = (D.f[DIR_MP0 ])[kw ];//[knw ];// + c1over54 ;(D.f[DIR_MP0 ])[kw ];//knw
- real mfcbc = (D.f[DIR_P0P ])[k ];//[kte ];// + c1over54 ;(D.f[DIR_P0P ])[k ];//kte
- real mfaba = (D.f[DIR_M0M ])[kbw];//[kbw ];// + c1over54 ;(D.f[DIR_M0M ])[kbw];
- real mfcba = (D.f[DIR_P0M ])[kb ];//[kbe ];// + c1over54 ;(D.f[DIR_P0M ])[kb ];//kbe
- real mfabc = (D.f[DIR_M0P ])[kw ];//[ktw ];// + c1over54 ;(D.f[DIR_M0P ])[kw ];//ktw
- real mfbcc = (D.f[DIR_0PP ])[k ];//[ktn ];// + c1over54 ;(D.f[DIR_0PP ])[k ];//ktn
- real mfbaa = (D.f[DIR_0MM ])[kbs];//[kbs ];// + c1over54 ;(D.f[DIR_0MM ])[kbs];
- real mfbca = (D.f[DIR_0PM ])[kb ];//[kbn ];// + c1over54 ;(D.f[DIR_0PM ])[kb ];//kbn
- real mfbac = (D.f[DIR_0MP ])[ks ];//[kts ];// + c1over54 ;(D.f[DIR_0MP ])[ks ];//kts
+ real mfcbb = (D.f[DIR_P00])[k ];//[ke ];// + c2over27 ;(D.f[DIR_P00])[k ];//ke
+ real mfabb = (D.f[DIR_M00])[kw ];//[kw ];// + c2over27 ;(D.f[DIR_M00])[kw ];
+ real mfbcb = (D.f[DIR_0P0])[k ];//[kn ];// + c2over27 ;(D.f[DIR_0P0])[k ];//kn
+ real mfbab = (D.f[DIR_0M0])[ks ];//[ks ];// + c2over27 ;(D.f[DIR_0M0])[ks ];
+ real mfbbc = (D.f[DIR_00P])[k ];//[kt ];// + c2over27 ;(D.f[DIR_00P])[k ];//kt
+ real mfbba = (D.f[DIR_00M])[kb ];//[kb ];// + c2over27 ;(D.f[DIR_00M])[kb ];
+ real mfccb = (D.f[DIR_PP0])[k ];//[kne ];// + c1over54 ;(D.f[DIR_PP0])[k ];//kne
+ real mfaab = (D.f[DIR_MM0])[ksw];//[ksw ];// + c1over54 ;(D.f[DIR_MM0])[ksw];
+ real mfcab = (D.f[DIR_PM0])[ks ];//[kse ];// + c1over54 ;(D.f[DIR_PM0])[ks ];//kse
+ real mfacb = (D.f[DIR_MP0])[kw ];//[knw ];// + c1over54 ;(D.f[DIR_MP0])[kw ];//knw
+ real mfcbc = (D.f[DIR_P0P])[k ];//[kte ];// + c1over54 ;(D.f[DIR_P0P])[k ];//kte
+ real mfaba = (D.f[DIR_M0M])[kbw];//[kbw ];// + c1over54 ;(D.f[DIR_M0M])[kbw];
+ real mfcba = (D.f[DIR_P0M])[kb ];//[kbe ];// + c1over54 ;(D.f[DIR_P0M])[kb ];//kbe
+ real mfabc = (D.f[DIR_M0P])[kw ];//[ktw ];// + c1over54 ;(D.f[DIR_M0P])[kw ];//ktw
+ real mfbcc = (D.f[DIR_0PP])[k ];//[ktn ];// + c1over54 ;(D.f[DIR_0PP])[k ];//ktn
+ real mfbaa = (D.f[DIR_0MM])[kbs];//[kbs ];// + c1over54 ;(D.f[DIR_0MM])[kbs];
+ real mfbca = (D.f[DIR_0PM])[kb ];//[kbn ];// + c1over54 ;(D.f[DIR_0PM])[kb ];//kbn
+ real mfbac = (D.f[DIR_0MP])[ks ];//[kts ];// + c1over54 ;(D.f[DIR_0MP])[ks ];//kts
real mfbbb = (D.f[DIR_000])[k ];//[kzero];// + c8over27 ;(D.f[DIR_000])[k ];//kzero
- real mfccc = (D.f[DIR_PPP ])[k ];//[ktne ];// + c1over216;(D.f[DIR_PPP ])[k ];//ktne
- real mfaac = (D.f[DIR_MMP ])[ksw];//[ktsw ];// + c1over216;(D.f[DIR_MMP ])[ksw];//ktsw
- real mfcac = (D.f[DIR_PMP ])[ks ];//[ktse ];// + c1over216;(D.f[DIR_PMP ])[ks ];//ktse
- real mfacc = (D.f[DIR_MPP ])[kw ];//[ktnw ];// + c1over216;(D.f[DIR_MPP ])[kw ];//ktnw
- real mfcca = (D.f[DIR_PPM ])[kb ];//[kbne ];// + c1over216;(D.f[DIR_PPM ])[kb ];//kbne
- real mfaaa = (D.f[DIR_MMM ])[kbsw];//[kbsw ];// + c1over216;(D.f[DIR_MMM ])[kbsw];
- real mfcaa = (D.f[DIR_PMM ])[kbs];//[kbse ];// + c1over216;(D.f[DIR_PMM ])[kbs];//kbse
- real mfaca = (D.f[DIR_MPM ])[kbw];//[kbnw ];// + c1over216;(D.f[DIR_MPM ])[kbw];//kbnw
+ real mfccc = (D.f[DIR_PPP])[k ];//[ktne ];// + c1over216;(D.f[DIR_PPP])[k ];//ktne
+ real mfaac = (D.f[DIR_MMP])[ksw];//[ktsw ];// + c1over216;(D.f[DIR_MMP])[ksw];//ktsw
+ real mfcac = (D.f[DIR_PMP])[ks ];//[ktse ];// + c1over216;(D.f[DIR_PMP])[ks ];//ktse
+ real mfacc = (D.f[DIR_MPP])[kw ];//[ktnw ];// + c1over216;(D.f[DIR_MPP])[kw ];//ktnw
+ real mfcca = (D.f[DIR_PPM])[kb ];//[kbne ];// + c1over216;(D.f[DIR_PPM])[kb ];//kbne
+ real mfaaa = (D.f[DIR_MMM])[kbsw];//[kbsw ];// + c1over216;(D.f[DIR_MMM])[kbsw];
+ real mfcaa = (D.f[DIR_PMM])[kbs];//[kbse ];// + c1over216;(D.f[DIR_PMM])[kbs];//kbse
+ real mfaca = (D.f[DIR_MPM])[kbw];//[kbnw ];// + c1over216;(D.f[DIR_MPM])[kbw];//kbnw
////////////////////////////////////////////////////////////////////////////////////
real drho = ((((mfccc+mfaaa) + (mfaca+mfcac)) + ((mfacc+mfcaa) + (mfaac+mfcca))) +
(((mfbac+mfbca) + (mfbaa+mfbcc)) + ((mfabc+mfcba) + (mfaba+mfcbc)) + ((mfacb+mfcab) + (mfaab+mfccb))) +
@@ -975,7 +975,7 @@ __global__ void LB_Kernel_Kum_IsoTest_SP_27( real omega,
real* dxxUx,
real* dyyUy,
real* dzzUz,
- int size_Mat,
+ unsigned long long numberOfLBnodes,
bool EvenOrOdd)
{
////////////////////////////////////////////////////////////////////////////////
@@ -989,7 +989,7 @@ __global__ void LB_Kernel_Kum_IsoTest_SP_27( real omega,
const unsigned k = nx*(ny*z + y) + x;
//////////////////////////////////////////////////////////////////////////
- if(k<size_Mat)
+ if(k<numberOfLBnodes)
{
////////////////////////////////////////////////////////////////////////////////
unsigned int BC;
@@ -1000,63 +1000,63 @@ __global__ void LB_Kernel_Kum_IsoTest_SP_27( real omega,
Distributions27 D;
if (EvenOrOdd==true)
{
- D.f[DIR_P00 ] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DDStart[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DDStart[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DDStart[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DDStart[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DDStart[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DDStart[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DDStart[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DDStart[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DDStart[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DDStart[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DDStart[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DDStart[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DDStart[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DDStart[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DDStart[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DDStart[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DDStart[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DDStart[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DDStart[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DDStart[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DDStart[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DDStart[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DDStart[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DDStart[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DDStart[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DDStart[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DDStart[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00 ] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_MMM ] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_PPM ] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_MPM ] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_PMM ] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_MMP ] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_PPP ] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_MPP ] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_PMP ] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_M00] = &DDStart[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DDStart[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DDStart[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DDStart[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DDStart[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DDStart[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DDStart[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DDStart[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DDStart[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DDStart[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DDStart[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DDStart[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DDStart[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DDStart[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DDStart[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DDStart[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DDStart[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DDStart[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DDStart[DIR_000 * numberOfLBnodes];
+ D.f[DIR_MMM] = &DDStart[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DDStart[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DDStart[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DDStart[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DDStart[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PPP] = &DDStart[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DDStart[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DDStart[DIR_MPM * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////
@@ -1089,33 +1089,33 @@ __global__ void LB_Kernel_Kum_IsoTest_SP_27( real omega,
//unsigned int ktne = k;
unsigned int kbsw = neighborZ[ksw];
//////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
- real mfcbb = (D.f[DIR_P00 ])[k ];//[ke ];// + c2over27 ;(D.f[DIR_P00 ])[k ];//ke
- real mfabb = (D.f[DIR_M00 ])[kw ];//[kw ];// + c2over27 ;(D.f[DIR_M00 ])[kw ];
- real mfbcb = (D.f[DIR_0P0 ])[k ];//[kn ];// + c2over27 ;(D.f[DIR_0P0 ])[k ];//kn
- real mfbab = (D.f[DIR_0M0 ])[ks ];//[ks ];// + c2over27 ;(D.f[DIR_0M0 ])[ks ];
- real mfbbc = (D.f[DIR_00P ])[k ];//[kt ];// + c2over27 ;(D.f[DIR_00P ])[k ];//kt
- real mfbba = (D.f[DIR_00M ])[kb ];//[kb ];// + c2over27 ;(D.f[DIR_00M ])[kb ];
- real mfccb = (D.f[DIR_PP0 ])[k ];//[kne ];// + c1over54 ;(D.f[DIR_PP0 ])[k ];//kne
- real mfaab = (D.f[DIR_MM0 ])[ksw];//[ksw ];// + c1over54 ;(D.f[DIR_MM0 ])[ksw];
- real mfcab = (D.f[DIR_PM0 ])[ks ];//[kse ];// + c1over54 ;(D.f[DIR_PM0 ])[ks ];//kse
- real mfacb = (D.f[DIR_MP0 ])[kw ];//[knw ];// + c1over54 ;(D.f[DIR_MP0 ])[kw ];//knw
- real mfcbc = (D.f[DIR_P0P ])[k ];//[kte ];// + c1over54 ;(D.f[DIR_P0P ])[k ];//kte
- real mfaba = (D.f[DIR_M0M ])[kbw];//[kbw ];// + c1over54 ;(D.f[DIR_M0M ])[kbw];
- real mfcba = (D.f[DIR_P0M ])[kb ];//[kbe ];// + c1over54 ;(D.f[DIR_P0M ])[kb ];//kbe
- real mfabc = (D.f[DIR_M0P ])[kw ];//[ktw ];// + c1over54 ;(D.f[DIR_M0P ])[kw ];//ktw
- real mfbcc = (D.f[DIR_0PP ])[k ];//[ktn ];// + c1over54 ;(D.f[DIR_0PP ])[k ];//ktn
- real mfbaa = (D.f[DIR_0MM ])[kbs];//[kbs ];// + c1over54 ;(D.f[DIR_0MM ])[kbs];
- real mfbca = (D.f[DIR_0PM ])[kb ];//[kbn ];// + c1over54 ;(D.f[DIR_0PM ])[kb ];//kbn
- real mfbac = (D.f[DIR_0MP ])[ks ];//[kts ];// + c1over54 ;(D.f[DIR_0MP ])[ks ];//kts
+ real mfcbb = (D.f[DIR_P00])[k ];//[ke ];// + c2over27 ;(D.f[DIR_P00])[k ];//ke
+ real mfabb = (D.f[DIR_M00])[kw ];//[kw ];// + c2over27 ;(D.f[DIR_M00])[kw ];
+ real mfbcb = (D.f[DIR_0P0])[k ];//[kn ];// + c2over27 ;(D.f[DIR_0P0])[k ];//kn
+ real mfbab = (D.f[DIR_0M0])[ks ];//[ks ];// + c2over27 ;(D.f[DIR_0M0])[ks ];
+ real mfbbc = (D.f[DIR_00P])[k ];//[kt ];// + c2over27 ;(D.f[DIR_00P])[k ];//kt
+ real mfbba = (D.f[DIR_00M])[kb ];//[kb ];// + c2over27 ;(D.f[DIR_00M])[kb ];
+ real mfccb = (D.f[DIR_PP0])[k ];//[kne ];// + c1over54 ;(D.f[DIR_PP0])[k ];//kne
+ real mfaab = (D.f[DIR_MM0])[ksw];//[ksw ];// + c1over54 ;(D.f[DIR_MM0])[ksw];
+ real mfcab = (D.f[DIR_PM0])[ks ];//[kse ];// + c1over54 ;(D.f[DIR_PM0])[ks ];//kse
+ real mfacb = (D.f[DIR_MP0])[kw ];//[knw ];// + c1over54 ;(D.f[DIR_MP0])[kw ];//knw
+ real mfcbc = (D.f[DIR_P0P])[k ];//[kte ];// + c1over54 ;(D.f[DIR_P0P])[k ];//kte
+ real mfaba = (D.f[DIR_M0M])[kbw];//[kbw ];// + c1over54 ;(D.f[DIR_M0M])[kbw];
+ real mfcba = (D.f[DIR_P0M])[kb ];//[kbe ];// + c1over54 ;(D.f[DIR_P0M])[kb ];//kbe
+ real mfabc = (D.f[DIR_M0P])[kw ];//[ktw ];// + c1over54 ;(D.f[DIR_M0P])[kw ];//ktw
+ real mfbcc = (D.f[DIR_0PP])[k ];//[ktn ];// + c1over54 ;(D.f[DIR_0PP])[k ];//ktn
+ real mfbaa = (D.f[DIR_0MM])[kbs];//[kbs ];// + c1over54 ;(D.f[DIR_0MM])[kbs];
+ real mfbca = (D.f[DIR_0PM])[kb ];//[kbn ];// + c1over54 ;(D.f[DIR_0PM])[kb ];//kbn
+ real mfbac = (D.f[DIR_0MP])[ks ];//[kts ];// + c1over54 ;(D.f[DIR_0MP])[ks ];//kts
real mfbbb = (D.f[DIR_000])[k ];//[kzero];// + c8over27 ;(D.f[DIR_000])[k ];//kzero
- real mfccc = (D.f[DIR_PPP ])[k ];//[ktne ];// + c1over216;(D.f[DIR_PPP ])[k ];//ktne
- real mfaac = (D.f[DIR_MMP ])[ksw];//[ktsw ];// + c1over216;(D.f[DIR_MMP ])[ksw];//ktsw
- real mfcac = (D.f[DIR_PMP ])[ks ];//[ktse ];// + c1over216;(D.f[DIR_PMP ])[ks ];//ktse
- real mfacc = (D.f[DIR_MPP ])[kw ];//[ktnw ];// + c1over216;(D.f[DIR_MPP ])[kw ];//ktnw
- real mfcca = (D.f[DIR_PPM ])[kb ];//[kbne ];// + c1over216;(D.f[DIR_PPM ])[kb ];//kbne
- real mfaaa = (D.f[DIR_MMM ])[kbsw];//[kbsw ];// + c1over216;(D.f[DIR_MMM ])[kbsw];
- real mfcaa = (D.f[DIR_PMM ])[kbs];//[kbse ];// + c1over216;(D.f[DIR_PMM ])[kbs];//kbse
- real mfaca = (D.f[DIR_MPM ])[kbw];//[kbnw ];// + c1over216;(D.f[DIR_MPM ])[kbw];//kbnw
+ real mfccc = (D.f[DIR_PPP])[k ];//[ktne ];// + c1over216;(D.f[DIR_PPP])[k ];//ktne
+ real mfaac = (D.f[DIR_MMP])[ksw];//[ktsw ];// + c1over216;(D.f[DIR_MMP])[ksw];//ktsw
+ real mfcac = (D.f[DIR_PMP])[ks ];//[ktse ];// + c1over216;(D.f[DIR_PMP])[ks ];//ktse
+ real mfacc = (D.f[DIR_MPP])[kw ];//[ktnw ];// + c1over216;(D.f[DIR_MPP])[kw ];//ktnw
+ real mfcca = (D.f[DIR_PPM])[kb ];//[kbne ];// + c1over216;(D.f[DIR_PPM])[kb ];//kbne
+ real mfaaa = (D.f[DIR_MMM])[kbsw];//[kbsw ];// + c1over216;(D.f[DIR_MMM])[kbsw];
+ real mfcaa = (D.f[DIR_PMM])[kbs];//[kbse ];// + c1over216;(D.f[DIR_PMM])[kbs];//kbse
+ real mfaca = (D.f[DIR_MPM])[kbw];//[kbnw ];// + c1over216;(D.f[DIR_MPM])[kbw];//kbnw
////////////////////////////////////////////////////////////////////////////////////
//slow
//real oMdrho = one - ((((mfccc+mfaaa) + (mfaca+mfcac)) + ((mfacc+mfcaa) + (mfaac+mfcca))) +
@@ -2016,7 +2016,7 @@ __global__ void LB_Kernel_Kum_1h_SP_27( real omega,
real* coordY,
real* coordZ,
real* DDStart,
- int size_Mat,
+ unsigned long long numberOfLBnodes,
bool EvenOrOdd)
{
////////////////////////////////////////////////////////////////////////////////
@@ -2030,7 +2030,7 @@ __global__ void LB_Kernel_Kum_1h_SP_27( real omega,
const unsigned k = nx*(ny*z + y) + x;
//////////////////////////////////////////////////////////////////////////
- if(k<size_Mat)
+ if(k<numberOfLBnodes)
{
////////////////////////////////////////////////////////////////////////////////
unsigned int BC;
@@ -2041,63 +2041,63 @@ __global__ void LB_Kernel_Kum_1h_SP_27( real omega,
Distributions27 D;
if (EvenOrOdd==true)
{
- D.f[DIR_P00 ] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DDStart[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DDStart[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DDStart[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DDStart[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DDStart[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DDStart[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DDStart[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DDStart[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DDStart[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DDStart[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DDStart[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DDStart[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DDStart[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DDStart[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DDStart[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DDStart[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DDStart[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DDStart[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DDStart[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DDStart[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DDStart[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DDStart[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DDStart[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DDStart[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DDStart[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DDStart[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DDStart[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00 ] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_MMM ] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_PPM ] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_MPM ] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_PMM ] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_MMP ] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_PPP ] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_MPP ] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_PMP ] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_M00] = &DDStart[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DDStart[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DDStart[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DDStart[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DDStart[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DDStart[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DDStart[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DDStart[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DDStart[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DDStart[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DDStart[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DDStart[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DDStart[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DDStart[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DDStart[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DDStart[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DDStart[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DDStart[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DDStart[DIR_000 * numberOfLBnodes];
+ D.f[DIR_MMM] = &DDStart[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DDStart[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DDStart[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DDStart[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DDStart[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PPP] = &DDStart[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DDStart[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DDStart[DIR_MPM * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////
@@ -2158,33 +2158,33 @@ __global__ void LB_Kernel_Kum_1h_SP_27( real omega,
//unsigned int ktne = k;
//unsigned int kbsw = neighborZ[ksw];
//////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
- real mfcbb = (D.f[DIR_P00 ])[k ];//[ke ];// + c2over27 ;(D.f[DIR_P00 ])[k ];//ke
- real mfabb = (D.f[DIR_M00 ])[kw ];//[kw ];// + c2over27 ;(D.f[DIR_M00 ])[kw ];
- real mfbcb = (D.f[DIR_0P0 ])[k ];//[kn ];// + c2over27 ;(D.f[DIR_0P0 ])[k ];//kn
- real mfbab = (D.f[DIR_0M0 ])[ks ];//[ks ];// + c2over27 ;(D.f[DIR_0M0 ])[ks ];
- real mfbbc = (D.f[DIR_00P ])[k ];//[kt ];// + c2over27 ;(D.f[DIR_00P ])[k ];//kt
- real mfbba = (D.f[DIR_00M ])[kb ];//[kb ];// + c2over27 ;(D.f[DIR_00M ])[kb ];
- real mfccb = (D.f[DIR_PP0 ])[k ];//[kne ];// + c1over54 ;(D.f[DIR_PP0 ])[k ];//kne
- real mfaab = (D.f[DIR_MM0 ])[ksw];//[ksw ];// + c1over54 ;(D.f[DIR_MM0 ])[ksw];
- real mfcab = (D.f[DIR_PM0 ])[ks ];//[kse ];// + c1over54 ;(D.f[DIR_PM0 ])[ks ];//kse
- real mfacb = (D.f[DIR_MP0 ])[kw ];//[knw ];// + c1over54 ;(D.f[DIR_MP0 ])[kw ];//knw
- real mfcbc = (D.f[DIR_P0P ])[k ];//[kte ];// + c1over54 ;(D.f[DIR_P0P ])[k ];//kte
- real mfaba = (D.f[DIR_M0M ])[kbw];//[kbw ];// + c1over54 ;(D.f[DIR_M0M ])[kbw];
- real mfcba = (D.f[DIR_P0M ])[kb ];//[kbe ];// + c1over54 ;(D.f[DIR_P0M ])[kb ];//kbe
- real mfabc = (D.f[DIR_M0P ])[kw ];//[ktw ];// + c1over54 ;(D.f[DIR_M0P ])[kw ];//ktw
- real mfbcc = (D.f[DIR_0PP ])[k ];//[ktn ];// + c1over54 ;(D.f[DIR_0PP ])[k ];//ktn
- real mfbaa = (D.f[DIR_0MM ])[kbs];//[kbs ];// + c1over54 ;(D.f[DIR_0MM ])[kbs];
- real mfbca = (D.f[DIR_0PM ])[kb ];//[kbn ];// + c1over54 ;(D.f[DIR_0PM ])[kb ];//kbn
- real mfbac = (D.f[DIR_0MP ])[ks ];//[kts ];// + c1over54 ;(D.f[DIR_0MP ])[ks ];//kts
+ real mfcbb = (D.f[DIR_P00])[k ];//[ke ];// + c2over27 ;(D.f[DIR_P00])[k ];//ke
+ real mfabb = (D.f[DIR_M00])[kw ];//[kw ];// + c2over27 ;(D.f[DIR_M00])[kw ];
+ real mfbcb = (D.f[DIR_0P0])[k ];//[kn ];// + c2over27 ;(D.f[DIR_0P0])[k ];//kn
+ real mfbab = (D.f[DIR_0M0])[ks ];//[ks ];// + c2over27 ;(D.f[DIR_0M0])[ks ];
+ real mfbbc = (D.f[DIR_00P])[k ];//[kt ];// + c2over27 ;(D.f[DIR_00P])[k ];//kt
+ real mfbba = (D.f[DIR_00M])[kb ];//[kb ];// + c2over27 ;(D.f[DIR_00M])[kb ];
+ real mfccb = (D.f[DIR_PP0])[k ];//[kne ];// + c1over54 ;(D.f[DIR_PP0])[k ];//kne
+ real mfaab = (D.f[DIR_MM0])[ksw];//[ksw ];// + c1over54 ;(D.f[DIR_MM0])[ksw];
+ real mfcab = (D.f[DIR_PM0])[ks ];//[kse ];// + c1over54 ;(D.f[DIR_PM0])[ks ];//kse
+ real mfacb = (D.f[DIR_MP0])[kw ];//[knw ];// + c1over54 ;(D.f[DIR_MP0])[kw ];//knw
+ real mfcbc = (D.f[DIR_P0P])[k ];//[kte ];// + c1over54 ;(D.f[DIR_P0P])[k ];//kte
+ real mfaba = (D.f[DIR_M0M])[kbw];//[kbw ];// + c1over54 ;(D.f[DIR_M0M])[kbw];
+ real mfcba = (D.f[DIR_P0M])[kb ];//[kbe ];// + c1over54 ;(D.f[DIR_P0M])[kb ];//kbe
+ real mfabc = (D.f[DIR_M0P])[kw ];//[ktw ];// + c1over54 ;(D.f[DIR_M0P])[kw ];//ktw
+ real mfbcc = (D.f[DIR_0PP])[k ];//[ktn ];// + c1over54 ;(D.f[DIR_0PP])[k ];//ktn
+ real mfbaa = (D.f[DIR_0MM])[kbs];//[kbs ];// + c1over54 ;(D.f[DIR_0MM])[kbs];
+ real mfbca = (D.f[DIR_0PM])[kb ];//[kbn ];// + c1over54 ;(D.f[DIR_0PM])[kb ];//kbn
+ real mfbac = (D.f[DIR_0MP])[ks ];//[kts ];// + c1over54 ;(D.f[DIR_0MP])[ks ];//kts
real mfbbb = (D.f[DIR_000])[k ];//[kzero];// + c8over27 ;(D.f[DIR_000])[k ];//kzero
- real mfccc = (D.f[DIR_PPP ])[k ];//[ktne ];// + c1over216;(D.f[DIR_PPP ])[k ];//ktne
- real mfaac = (D.f[DIR_MMP ])[ksw];//[ktsw ];// + c1over216;(D.f[DIR_MMP ])[ksw];//ktsw
- real mfcac = (D.f[DIR_PMP ])[ks ];//[ktse ];// + c1over216;(D.f[DIR_PMP ])[ks ];//ktse
- real mfacc = (D.f[DIR_MPP ])[kw ];//[ktnw ];// + c1over216;(D.f[DIR_MPP ])[kw ];//ktnw
- real mfcca = (D.f[DIR_PPM ])[kb ];//[kbne ];// + c1over216;(D.f[DIR_PPM ])[kb ];//kbne
- real mfaaa = (D.f[DIR_MMM ])[kbsw];//[kbsw ];// + c1over216;(D.f[DIR_MMM ])[kbsw];
- real mfcaa = (D.f[DIR_PMM ])[kbs];//[kbse ];// + c1over216;(D.f[DIR_PMM ])[kbs];//kbse
- real mfaca = (D.f[DIR_MPM ])[kbw];//[kbnw ];// + c1over216;(D.f[DIR_MPM ])[kbw];//kbnw
+ real mfccc = (D.f[DIR_PPP])[k ];//[ktne ];// + c1over216;(D.f[DIR_PPP])[k ];//ktne
+ real mfaac = (D.f[DIR_MMP])[ksw];//[ktsw ];// + c1over216;(D.f[DIR_MMP])[ksw];//ktsw
+ real mfcac = (D.f[DIR_PMP])[ks ];//[ktse ];// + c1over216;(D.f[DIR_PMP])[ks ];//ktse
+ real mfacc = (D.f[DIR_MPP])[kw ];//[ktnw ];// + c1over216;(D.f[DIR_MPP])[kw ];//ktnw
+ real mfcca = (D.f[DIR_PPM])[kb ];//[kbne ];// + c1over216;(D.f[DIR_PPM])[kb ];//kbne
+ real mfaaa = (D.f[DIR_MMM])[kbsw];//[kbsw ];// + c1over216;(D.f[DIR_MMM])[kbsw];
+ real mfcaa = (D.f[DIR_PMM])[kbs];//[kbse ];// + c1over216;(D.f[DIR_PMM])[kbs];//kbse
+ real mfaca = (D.f[DIR_MPM])[kbw];//[kbnw ];// + c1over216;(D.f[DIR_MPM])[kbw];//kbnw
////////////////////////////////////////////////////////////////////////////////////
//Ship
real coord0X = 281.125f;//7.5f;
@@ -3238,7 +3238,7 @@ __global__ void LB_Kernel_Kum_New_SP_27( real omega,
unsigned int* neighborY,
unsigned int* neighborZ,
real* DDStart,
- int size_Mat,
+ unsigned long long numberOfLBnodes,
bool EvenOrOdd)
{
////////////////////////////////////////////////////////////////////////////////
@@ -3252,7 +3252,7 @@ __global__ void LB_Kernel_Kum_New_SP_27( real omega,
const unsigned k = nx*(ny*z + y) + x;
//////////////////////////////////////////////////////////////////////////
- if(k<size_Mat)
+ if(k<numberOfLBnodes)
{
////////////////////////////////////////////////////////////////////////////////
unsigned int BC;
@@ -3263,63 +3263,63 @@ __global__ void LB_Kernel_Kum_New_SP_27( real omega,
Distributions27 D;
if (EvenOrOdd==true)
{
- D.f[DIR_P00 ] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DDStart[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DDStart[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DDStart[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DDStart[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DDStart[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DDStart[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DDStart[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DDStart[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DDStart[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DDStart[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DDStart[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DDStart[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DDStart[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DDStart[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DDStart[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DDStart[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DDStart[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DDStart[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DDStart[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DDStart[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DDStart[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DDStart[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DDStart[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DDStart[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DDStart[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DDStart[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DDStart[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00 ] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_MMM ] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_PPM ] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_MPM ] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_PMM ] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_MMP ] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_PPP ] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_MPP ] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_PMP ] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_M00] = &DDStart[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DDStart[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DDStart[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DDStart[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DDStart[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DDStart[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DDStart[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DDStart[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DDStart[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DDStart[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DDStart[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DDStart[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DDStart[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DDStart[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DDStart[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DDStart[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DDStart[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DDStart[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DDStart[DIR_000 * numberOfLBnodes];
+ D.f[DIR_MMM] = &DDStart[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DDStart[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DDStart[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DDStart[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DDStart[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PPP] = &DDStart[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DDStart[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DDStart[DIR_MPM * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////
@@ -3380,33 +3380,33 @@ __global__ void LB_Kernel_Kum_New_SP_27( real omega,
//unsigned int ktne = k;
//unsigned int kbsw = neighborZ[ksw];
//////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
- real mfcbb = (D.f[DIR_P00 ])[k ];//[ke ];// + c2over27 ;(D.f[DIR_P00 ])[k ];//ke
- real mfabb = (D.f[DIR_M00 ])[kw ];//[kw ];// + c2over27 ;(D.f[DIR_M00 ])[kw ];
- real mfbcb = (D.f[DIR_0P0 ])[k ];//[kn ];// + c2over27 ;(D.f[DIR_0P0 ])[k ];//kn
- real mfbab = (D.f[DIR_0M0 ])[ks ];//[ks ];// + c2over27 ;(D.f[DIR_0M0 ])[ks ];
- real mfbbc = (D.f[DIR_00P ])[k ];//[kt ];// + c2over27 ;(D.f[DIR_00P ])[k ];//kt
- real mfbba = (D.f[DIR_00M ])[kb ];//[kb ];// + c2over27 ;(D.f[DIR_00M ])[kb ];
- real mfccb = (D.f[DIR_PP0 ])[k ];//[kne ];// + c1over54 ;(D.f[DIR_PP0 ])[k ];//kne
- real mfaab = (D.f[DIR_MM0 ])[ksw];//[ksw ];// + c1over54 ;(D.f[DIR_MM0 ])[ksw];
- real mfcab = (D.f[DIR_PM0 ])[ks ];//[kse ];// + c1over54 ;(D.f[DIR_PM0 ])[ks ];//kse
- real mfacb = (D.f[DIR_MP0 ])[kw ];//[knw ];// + c1over54 ;(D.f[DIR_MP0 ])[kw ];//knw
- real mfcbc = (D.f[DIR_P0P ])[k ];//[kte ];// + c1over54 ;(D.f[DIR_P0P ])[k ];//kte
- real mfaba = (D.f[DIR_M0M ])[kbw];//[kbw ];// + c1over54 ;(D.f[DIR_M0M ])[kbw];
- real mfcba = (D.f[DIR_P0M ])[kb ];//[kbe ];// + c1over54 ;(D.f[DIR_P0M ])[kb ];//kbe
- real mfabc = (D.f[DIR_M0P ])[kw ];//[ktw ];// + c1over54 ;(D.f[DIR_M0P ])[kw ];//ktw
- real mfbcc = (D.f[DIR_0PP ])[k ];//[ktn ];// + c1over54 ;(D.f[DIR_0PP ])[k ];//ktn
- real mfbaa = (D.f[DIR_0MM ])[kbs];//[kbs ];// + c1over54 ;(D.f[DIR_0MM ])[kbs];
- real mfbca = (D.f[DIR_0PM ])[kb ];//[kbn ];// + c1over54 ;(D.f[DIR_0PM ])[kb ];//kbn
- real mfbac = (D.f[DIR_0MP ])[ks ];//[kts ];// + c1over54 ;(D.f[DIR_0MP ])[ks ];//kts
+ real mfcbb = (D.f[DIR_P00])[k ];//[ke ];// + c2over27 ;(D.f[DIR_P00])[k ];//ke
+ real mfabb = (D.f[DIR_M00])[kw ];//[kw ];// + c2over27 ;(D.f[DIR_M00])[kw ];
+ real mfbcb = (D.f[DIR_0P0])[k ];//[kn ];// + c2over27 ;(D.f[DIR_0P0])[k ];//kn
+ real mfbab = (D.f[DIR_0M0])[ks ];//[ks ];// + c2over27 ;(D.f[DIR_0M0])[ks ];
+ real mfbbc = (D.f[DIR_00P])[k ];//[kt ];// + c2over27 ;(D.f[DIR_00P])[k ];//kt
+ real mfbba = (D.f[DIR_00M])[kb ];//[kb ];// + c2over27 ;(D.f[DIR_00M])[kb ];
+ real mfccb = (D.f[DIR_PP0])[k ];//[kne ];// + c1over54 ;(D.f[DIR_PP0])[k ];//kne
+ real mfaab = (D.f[DIR_MM0])[ksw];//[ksw ];// + c1over54 ;(D.f[DIR_MM0])[ksw];
+ real mfcab = (D.f[DIR_PM0])[ks ];//[kse ];// + c1over54 ;(D.f[DIR_PM0])[ks ];//kse
+ real mfacb = (D.f[DIR_MP0])[kw ];//[knw ];// + c1over54 ;(D.f[DIR_MP0])[kw ];//knw
+ real mfcbc = (D.f[DIR_P0P])[k ];//[kte ];// + c1over54 ;(D.f[DIR_P0P])[k ];//kte
+ real mfaba = (D.f[DIR_M0M])[kbw];//[kbw ];// + c1over54 ;(D.f[DIR_M0M])[kbw];
+ real mfcba = (D.f[DIR_P0M])[kb ];//[kbe ];// + c1over54 ;(D.f[DIR_P0M])[kb ];//kbe
+ real mfabc = (D.f[DIR_M0P])[kw ];//[ktw ];// + c1over54 ;(D.f[DIR_M0P])[kw ];//ktw
+ real mfbcc = (D.f[DIR_0PP])[k ];//[ktn ];// + c1over54 ;(D.f[DIR_0PP])[k ];//ktn
+ real mfbaa = (D.f[DIR_0MM])[kbs];//[kbs ];// + c1over54 ;(D.f[DIR_0MM])[kbs];
+ real mfbca = (D.f[DIR_0PM])[kb ];//[kbn ];// + c1over54 ;(D.f[DIR_0PM])[kb ];//kbn
+ real mfbac = (D.f[DIR_0MP])[ks ];//[kts ];// + c1over54 ;(D.f[DIR_0MP])[ks ];//kts
real mfbbb = (D.f[DIR_000])[k ];//[kzero];// + c8over27 ;(D.f[DIR_000])[k ];//kzero
- real mfccc = (D.f[DIR_PPP ])[k ];//[ktne ];// + c1over216;(D.f[DIR_PPP ])[k ];//ktne
- real mfaac = (D.f[DIR_MMP ])[ksw];//[ktsw ];// + c1over216;(D.f[DIR_MMP ])[ksw];//ktsw
- real mfcac = (D.f[DIR_PMP ])[ks ];//[ktse ];// + c1over216;(D.f[DIR_PMP ])[ks ];//ktse
- real mfacc = (D.f[DIR_MPP ])[kw ];//[ktnw ];// + c1over216;(D.f[DIR_MPP ])[kw ];//ktnw
- real mfcca = (D.f[DIR_PPM ])[kb ];//[kbne ];// + c1over216;(D.f[DIR_PPM ])[kb ];//kbne
- real mfaaa = (D.f[DIR_MMM ])[kbsw];//[kbsw ];// + c1over216;(D.f[DIR_MMM ])[kbsw];
- real mfcaa = (D.f[DIR_PMM ])[kbs];//[kbse ];// + c1over216;(D.f[DIR_PMM ])[kbs];//kbse
- real mfaca = (D.f[DIR_MPM ])[kbw];//[kbnw ];// + c1over216;(D.f[DIR_MPM ])[kbw];//kbnw
+ real mfccc = (D.f[DIR_PPP])[k ];//[ktne ];// + c1over216;(D.f[DIR_PPP])[k ];//ktne
+ real mfaac = (D.f[DIR_MMP])[ksw];//[ktsw ];// + c1over216;(D.f[DIR_MMP])[ksw];//ktsw
+ real mfcac = (D.f[DIR_PMP])[ks ];//[ktse ];// + c1over216;(D.f[DIR_PMP])[ks ];//ktse
+ real mfacc = (D.f[DIR_MPP])[kw ];//[ktnw ];// + c1over216;(D.f[DIR_MPP])[kw ];//ktnw
+ real mfcca = (D.f[DIR_PPM])[kb ];//[kbne ];// + c1over216;(D.f[DIR_PPM])[kb ];//kbne
+ real mfaaa = (D.f[DIR_MMM])[kbsw];//[kbsw ];// + c1over216;(D.f[DIR_MMM])[kbsw];
+ real mfcaa = (D.f[DIR_PMM])[kbs];//[kbse ];// + c1over216;(D.f[DIR_PMM])[kbs];//kbse
+ real mfaca = (D.f[DIR_MPM])[kbw];//[kbnw ];// + c1over216;(D.f[DIR_MPM])[kbw];//kbnw
////////////////////////////////////////////////////////////////////////////////////
//slow
//real oMdrho = one - ((((mfccc+mfaaa) + (mfaca+mfcac)) + ((mfacc+mfcaa) + (mfaac+mfcca))) +
@@ -4510,7 +4510,7 @@ __global__ void LB_Kernel_Kum_Comp_SP_27( real omega,
unsigned int* neighborY,
unsigned int* neighborZ,
real* DDStart,
- int size_Mat,
+ unsigned long long numberOfLBnodes,
bool EvenOrOdd)
{
////////////////////////////////////////////////////////////////////////////////
@@ -4524,7 +4524,7 @@ __global__ void LB_Kernel_Kum_Comp_SP_27( real omega,
const unsigned k = nx*(ny*z + y) + x;
//////////////////////////////////////////////////////////////////////////
- if(k<size_Mat)
+ if(k<numberOfLBnodes)
{
////////////////////////////////////////////////////////////////////////////////
unsigned int BC;
@@ -4535,63 +4535,63 @@ __global__ void LB_Kernel_Kum_Comp_SP_27( real omega,
Distributions27 D;
if (EvenOrOdd==true)
{
- D.f[DIR_P00 ] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DDStart[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DDStart[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DDStart[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DDStart[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DDStart[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DDStart[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DDStart[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DDStart[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DDStart[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DDStart[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DDStart[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DDStart[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DDStart[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DDStart[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DDStart[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DDStart[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DDStart[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DDStart[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DDStart[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DDStart[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DDStart[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DDStart[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DDStart[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DDStart[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DDStart[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DDStart[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DDStart[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00 ] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_MMM ] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_PPM ] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_MPM ] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_PMM ] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_MMP ] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_PPP ] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_MPP ] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_PMP ] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_M00] = &DDStart[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DDStart[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DDStart[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DDStart[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DDStart[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DDStart[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DDStart[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DDStart[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DDStart[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DDStart[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DDStart[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DDStart[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DDStart[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DDStart[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DDStart[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DDStart[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DDStart[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DDStart[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DDStart[DIR_000 * numberOfLBnodes];
+ D.f[DIR_MMM] = &DDStart[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DDStart[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DDStart[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DDStart[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DDStart[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PPP] = &DDStart[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DDStart[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DDStart[DIR_MPM * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////
@@ -4624,33 +4624,33 @@ __global__ void LB_Kernel_Kum_Comp_SP_27( real omega,
unsigned int ktne = k;
unsigned int kbsw = neighborZ[ksw];
//////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
- real f_E = (D.f[DIR_P00 ])[ke ];// + c2over27 ;
- real f_W = (D.f[DIR_M00 ])[kw ];// + c2over27 ;
- real f_N = (D.f[DIR_0P0 ])[kn ];// + c2over27 ;
- real f_S = (D.f[DIR_0M0 ])[ks ];// + c2over27 ;
- real f_T = (D.f[DIR_00P ])[kt ];// + c2over27 ;
- real f_B = (D.f[DIR_00M ])[kb ];// + c2over27 ;
- real f_NE = (D.f[DIR_PP0 ])[kne ];// + c1over54 ;
- real f_SW = (D.f[DIR_MM0 ])[ksw ];// + c1over54 ;
- real f_SE = (D.f[DIR_PM0 ])[kse ];// + c1over54 ;
- real f_NW = (D.f[DIR_MP0 ])[knw ];// + c1over54 ;
- real f_TE = (D.f[DIR_P0P ])[kte ];// + c1over54 ;
- real f_BW = (D.f[DIR_M0M ])[kbw ];// + c1over54 ;
- real f_BE = (D.f[DIR_P0M ])[kbe ];// + c1over54 ;
- real f_TW = (D.f[DIR_M0P ])[ktw ];// + c1over54 ;
- real f_TN = (D.f[DIR_0PP ])[ktn ];// + c1over54 ;
- real f_BS = (D.f[DIR_0MM ])[kbs ];// + c1over54 ;
- real f_BN = (D.f[DIR_0PM ])[kbn ];// + c1over54 ;
- real f_TS = (D.f[DIR_0MP ])[kts ];// + c1over54 ;
+ real f_E = (D.f[DIR_P00])[ke ];// + c2over27 ;
+ real f_W = (D.f[DIR_M00])[kw ];// + c2over27 ;
+ real f_N = (D.f[DIR_0P0])[kn ];// + c2over27 ;
+ real f_S = (D.f[DIR_0M0])[ks ];// + c2over27 ;
+ real f_T = (D.f[DIR_00P])[kt ];// + c2over27 ;
+ real f_B = (D.f[DIR_00M])[kb ];// + c2over27 ;
+ real f_NE = (D.f[DIR_PP0])[kne ];// + c1over54 ;
+ real f_SW = (D.f[DIR_MM0])[ksw ];// + c1over54 ;
+ real f_SE = (D.f[DIR_PM0])[kse ];// + c1over54 ;
+ real f_NW = (D.f[DIR_MP0])[knw ];// + c1over54 ;
+ real f_TE = (D.f[DIR_P0P])[kte ];// + c1over54 ;
+ real f_BW = (D.f[DIR_M0M])[kbw ];// + c1over54 ;
+ real f_BE = (D.f[DIR_P0M])[kbe ];// + c1over54 ;
+ real f_TW = (D.f[DIR_M0P])[ktw ];// + c1over54 ;
+ real f_TN = (D.f[DIR_0PP])[ktn ];// + c1over54 ;
+ real f_BS = (D.f[DIR_0MM])[kbs ];// + c1over54 ;
+ real f_BN = (D.f[DIR_0PM])[kbn ];// + c1over54 ;
+ real f_TS = (D.f[DIR_0MP])[kts ];// + c1over54 ;
real f_R = (D.f[DIR_000])[kzero];// + c8over27 ;
- real f_TNE = (D.f[DIR_PPP ])[ktne ];// + c1over216;
- real f_TSW = (D.f[DIR_MMP ])[ktsw ];// + c1over216;
- real f_TSE = (D.f[DIR_PMP ])[ktse ];// + c1over216;
- real f_TNW = (D.f[DIR_MPP ])[ktnw ];// + c1over216;
- real f_BNE = (D.f[DIR_PPM ])[kbne ];// + c1over216;
- real f_BSW = (D.f[DIR_MMM ])[kbsw ];// + c1over216;
- real f_BSE = (D.f[DIR_PMM ])[kbse ];// + c1over216;
- real f_BNW = (D.f[DIR_MPM ])[kbnw ];// + c1over216;
+ real f_TNE = (D.f[DIR_PPP])[ktne ];// + c1over216;
+ real f_TSW = (D.f[DIR_MMP])[ktsw ];// + c1over216;
+ real f_TSE = (D.f[DIR_PMP])[ktse ];// + c1over216;
+ real f_TNW = (D.f[DIR_MPP])[ktnw ];// + c1over216;
+ real f_BNE = (D.f[DIR_PPM])[kbne ];// + c1over216;
+ real f_BSW = (D.f[DIR_MMM])[kbsw ];// + c1over216;
+ real f_BSE = (D.f[DIR_PMM])[kbse ];// + c1over216;
+ real f_BNW = (D.f[DIR_MPM])[kbnw ];// + c1over216;
////////////////////////////////////////////////////////////////////////////////////
real fx = c0o1;
real fy = c0o1;
@@ -5451,7 +5451,7 @@ __global__ void LB_Kernel_Kum_New_Comp_SRT_SP_27(
unsigned int* neighborY,
unsigned int* neighborZ,
real* DDStart,
- int size_Mat,
+ unsigned long long numberOfLBnodes,
int level,
real* forces,
bool EvenOrOdd)
@@ -5467,7 +5467,7 @@ __global__ void LB_Kernel_Kum_New_Comp_SRT_SP_27(
const unsigned k = nx*(ny*z + y) + x;
//////////////////////////////////////////////////////////////////////////
- if (k<size_Mat)
+ if (k<numberOfLBnodes)
{
////////////////////////////////////////////////////////////////////////////////
unsigned int BC;
@@ -5478,63 +5478,63 @@ __global__ void LB_Kernel_Kum_New_Comp_SRT_SP_27(
Distributions27 D;
if (EvenOrOdd == true)
{
- D.f[DIR_P00] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_M00] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0P0] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0M0] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00P] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00M] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_PP0] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_MM0] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_PM0] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_MP0] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_P0P] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_M0M] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_P0M] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_M0P] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0PP] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0MM] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0PM] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0MP] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_PPP] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_MMP] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_PMP] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_MPP] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_PPM] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_MMM] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_PMM] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_MPM] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DDStart[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DDStart[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DDStart[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DDStart[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DDStart[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DDStart[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DDStart[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DDStart[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DDStart[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DDStart[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DDStart[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DDStart[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DDStart[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DDStart[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DDStart[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DDStart[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DDStart[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DDStart[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DDStart[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DDStart[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DDStart[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DDStart[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DDStart[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DDStart[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DDStart[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DDStart[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DDStart[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_P00] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0M0] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0P0] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00M] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00P] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_MM0] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_PP0] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_MP0] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_PM0] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_M0M] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_P0P] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_M0P] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_P0M] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0MM] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0PP] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0MP] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0PM] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_MMM] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_PPM] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_MPM] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_PMM] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_MMP] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_PPP] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_MPP] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_PMP] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_M00] = &DDStart[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DDStart[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DDStart[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DDStart[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DDStart[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DDStart[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DDStart[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DDStart[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DDStart[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DDStart[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DDStart[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DDStart[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DDStart[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DDStart[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DDStart[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DDStart[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DDStart[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DDStart[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DDStart[DIR_000 * numberOfLBnodes];
+ D.f[DIR_MMM] = &DDStart[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DDStart[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DDStart[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DDStart[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DDStart[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PPP] = &DDStart[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DDStart[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DDStart[DIR_MPM * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////
@@ -5568,33 +5568,33 @@ __global__ void LB_Kernel_Kum_New_Comp_SRT_SP_27(
unsigned int kbsw = neighborZ[ksw];
//////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
- real mfcbb = (D.f[DIR_P00 ])[k ];
- real mfabb = (D.f[DIR_M00 ])[kw ];
- real mfbcb = (D.f[DIR_0P0 ])[k ];
- real mfbab = (D.f[DIR_0M0 ])[ks ];
- real mfbbc = (D.f[DIR_00P ])[k ];
- real mfbba = (D.f[DIR_00M ])[kb ];
- real mfccb = (D.f[DIR_PP0 ])[k ];
- real mfaab = (D.f[DIR_MM0 ])[ksw ];
- real mfcab = (D.f[DIR_PM0 ])[ks ];
- real mfacb = (D.f[DIR_MP0 ])[kw ];
- real mfcbc = (D.f[DIR_P0P ])[k ];
- real mfaba = (D.f[DIR_M0M ])[kbw ];
- real mfcba = (D.f[DIR_P0M ])[kb ];
- real mfabc = (D.f[DIR_M0P ])[kw ];
- real mfbcc = (D.f[DIR_0PP ])[k ];
- real mfbaa = (D.f[DIR_0MM ])[kbs ];
- real mfbca = (D.f[DIR_0PM ])[kb ];
- real mfbac = (D.f[DIR_0MP ])[ks ];
+ real mfcbb = (D.f[DIR_P00])[k ];
+ real mfabb = (D.f[DIR_M00])[kw ];
+ real mfbcb = (D.f[DIR_0P0])[k ];
+ real mfbab = (D.f[DIR_0M0])[ks ];
+ real mfbbc = (D.f[DIR_00P])[k ];
+ real mfbba = (D.f[DIR_00M])[kb ];
+ real mfccb = (D.f[DIR_PP0])[k ];
+ real mfaab = (D.f[DIR_MM0])[ksw ];
+ real mfcab = (D.f[DIR_PM0])[ks ];
+ real mfacb = (D.f[DIR_MP0])[kw ];
+ real mfcbc = (D.f[DIR_P0P])[k ];
+ real mfaba = (D.f[DIR_M0M])[kbw ];
+ real mfcba = (D.f[DIR_P0M])[kb ];
+ real mfabc = (D.f[DIR_M0P])[kw ];
+ real mfbcc = (D.f[DIR_0PP])[k ];
+ real mfbaa = (D.f[DIR_0MM])[kbs ];
+ real mfbca = (D.f[DIR_0PM])[kb ];
+ real mfbac = (D.f[DIR_0MP])[ks ];
real mfbbb = (D.f[DIR_000])[k ];
- real mfccc = (D.f[DIR_PPP ])[k ];
- real mfaac = (D.f[DIR_MMP ])[ksw ];
- real mfcac = (D.f[DIR_PMP ])[ks ];
- real mfacc = (D.f[DIR_MPP ])[kw ];
- real mfcca = (D.f[DIR_PPM ])[kb ];
- real mfaaa = (D.f[DIR_MMM ])[kbsw];
- real mfcaa = (D.f[DIR_PMM ])[kbs ];
- real mfaca = (D.f[DIR_MPM ])[kbw ];
+ real mfccc = (D.f[DIR_PPP])[k ];
+ real mfaac = (D.f[DIR_MMP])[ksw ];
+ real mfcac = (D.f[DIR_PMP])[ks ];
+ real mfacc = (D.f[DIR_MPP])[kw ];
+ real mfcca = (D.f[DIR_PPM])[kb ];
+ real mfaaa = (D.f[DIR_MMM])[kbsw];
+ real mfcaa = (D.f[DIR_PMM])[kbs ];
+ real mfaca = (D.f[DIR_MPM])[kbw ];
////////////////////////////////////////////////////////////////////////////////////
real drho = ((((mfccc + mfaaa) + (mfaca + mfcac)) + ((mfacc + mfcaa) + (mfaac + mfcca))) +
(((mfbac + mfbca) + (mfbaa + mfbcc)) + ((mfabc + mfcba) + (mfaba + mfcbc)) + ((mfacb + mfcab) + (mfaab + mfccb))) +
@@ -6349,33 +6349,33 @@ __global__ void LB_Kernel_Kum_New_Comp_SRT_SP_27(
((mfabb + mfcbb) + (mfbab + mfbcb) + (mfbba + mfbbc))) + mfbbb;
mfbbb += drho - drhoPost;
////////////////////////////////////////////////////////////////////////////////////
- (D.f[DIR_P00 ])[k ] = mfabb;
- (D.f[DIR_M00 ])[kw ] = mfcbb;
- (D.f[DIR_0P0 ])[k ] = mfbab;
- (D.f[DIR_0M0 ])[ks ] = mfbcb;
- (D.f[DIR_00P ])[k ] = mfbba;
- (D.f[DIR_00M ])[kb ] = mfbbc;
- (D.f[DIR_PP0 ])[k ] = mfaab;
- (D.f[DIR_MM0 ])[ksw ] = mfccb;
- (D.f[DIR_PM0 ])[ks ] = mfacb;
- (D.f[DIR_MP0 ])[kw ] = mfcab;
- (D.f[DIR_P0P ])[k ] = mfaba;
- (D.f[DIR_M0M ])[kbw ] = mfcbc;
- (D.f[DIR_P0M ])[kb ] = mfabc;
- (D.f[DIR_M0P ])[kw ] = mfcba;
- (D.f[DIR_0PP ])[k ] = mfbaa;
- (D.f[DIR_0MM ])[kbs ] = mfbcc;
- (D.f[DIR_0PM ])[kb ] = mfbac;
- (D.f[DIR_0MP ])[ks ] = mfbca;
+ (D.f[DIR_P00])[k ] = mfabb;
+ (D.f[DIR_M00])[kw ] = mfcbb;
+ (D.f[DIR_0P0])[k ] = mfbab;
+ (D.f[DIR_0M0])[ks ] = mfbcb;
+ (D.f[DIR_00P])[k ] = mfbba;
+ (D.f[DIR_00M])[kb ] = mfbbc;
+ (D.f[DIR_PP0])[k ] = mfaab;
+ (D.f[DIR_MM0])[ksw ] = mfccb;
+ (D.f[DIR_PM0])[ks ] = mfacb;
+ (D.f[DIR_MP0])[kw ] = mfcab;
+ (D.f[DIR_P0P])[k ] = mfaba;
+ (D.f[DIR_M0M])[kbw ] = mfcbc;
+ (D.f[DIR_P0M])[kb ] = mfabc;
+ (D.f[DIR_M0P])[kw ] = mfcba;
+ (D.f[DIR_0PP])[k ] = mfbaa;
+ (D.f[DIR_0MM])[kbs ] = mfbcc;
+ (D.f[DIR_0PM])[kb ] = mfbac;
+ (D.f[DIR_0MP])[ks ] = mfbca;
(D.f[DIR_000])[k ] = mfbbb;
- (D.f[DIR_PPP ])[k ] = mfaaa;
- (D.f[DIR_PMP ])[ks ] = mfaca;
- (D.f[DIR_PPM ])[kb ] = mfaac;
- (D.f[DIR_PMM ])[kbs ] = mfacc;
- (D.f[DIR_MPP ])[kw ] = mfcaa;
- (D.f[DIR_MMP ])[ksw ] = mfcca;
- (D.f[DIR_MPM ])[kbw ] = mfcac;
- (D.f[DIR_MMM ])[kbsw] = mfccc;
+ (D.f[DIR_PPP])[k ] = mfaaa;
+ (D.f[DIR_PMP])[ks ] = mfaca;
+ (D.f[DIR_PPM])[kb ] = mfaac;
+ (D.f[DIR_PMM])[kbs ] = mfacc;
+ (D.f[DIR_MPP])[kw ] = mfcaa;
+ (D.f[DIR_MMP])[ksw ] = mfcca;
+ (D.f[DIR_MPM])[kbw ] = mfcac;
+ (D.f[DIR_MMM])[kbsw] = mfccc;
}
}
}
diff --git a/src/gpu/VirtualFluids_GPU/GPU/Cumulant27chim.cu b/src/gpu/VirtualFluids_GPU/GPU/Cumulant27chim.cu
index 97c1aff4d26cb85deaf1dd0d145245f28affc2e3..3706e5f929b50a2a72c107a982525ec3172eb144 100644
--- a/src/gpu/VirtualFluids_GPU/GPU/Cumulant27chim.cu
+++ b/src/gpu/VirtualFluids_GPU/GPU/Cumulant27chim.cu
@@ -51,7 +51,7 @@ __global__ void Cumulant_One_preconditioned_errorDiffusion_chim_Comp_SP_27(
unsigned int* neighborY,
unsigned int* neighborZ,
real* DDStart,
- int size_Mat,
+ unsigned long long numberOfLBnodes,
int level,
real* forces,
bool EvenOrOdd)
@@ -67,7 +67,7 @@ __global__ void Cumulant_One_preconditioned_errorDiffusion_chim_Comp_SP_27(
const unsigned k = nx*(ny*z + y) + x;
//////////////////////////////////////////////////////////////////////////
- if (k<size_Mat)
+ if (k<numberOfLBnodes)
{
////////////////////////////////////////////////////////////////////////////////
unsigned int BC;
@@ -78,63 +78,63 @@ __global__ void Cumulant_One_preconditioned_errorDiffusion_chim_Comp_SP_27(
Distributions27 D;
if (EvenOrOdd == true)
{
- D.f[DIR_P00] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_M00] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0P0] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0M0] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00P] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00M] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_PP0] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_MM0] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_PM0] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_MP0] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_P0P] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_M0M] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_P0M] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_M0P] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0PP] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0MM] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0PM] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0MP] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_PPP] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_MMP] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_PMP] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_MPP] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_PPM] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_MMM] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_PMM] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_MPM] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DDStart[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DDStart[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DDStart[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DDStart[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DDStart[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DDStart[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DDStart[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DDStart[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DDStart[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DDStart[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DDStart[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DDStart[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DDStart[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DDStart[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DDStart[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DDStart[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DDStart[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DDStart[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DDStart[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DDStart[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DDStart[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DDStart[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DDStart[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DDStart[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DDStart[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DDStart[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DDStart[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_P00] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0M0] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0P0] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00M] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00P] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_MM0] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_PP0] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_MP0] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_PM0] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_M0M] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_P0P] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_M0P] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_P0M] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0MM] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0PP] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0MP] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0PM] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_MMM] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_PPM] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_MPM] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_PMM] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_MMP] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_PPP] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_MPP] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_PMP] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_M00] = &DDStart[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DDStart[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DDStart[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DDStart[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DDStart[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DDStart[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DDStart[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DDStart[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DDStart[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DDStart[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DDStart[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DDStart[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DDStart[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DDStart[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DDStart[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DDStart[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DDStart[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DDStart[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DDStart[DIR_000 * numberOfLBnodes];
+ D.f[DIR_MMM] = &DDStart[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DDStart[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DDStart[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DDStart[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DDStart[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PPP] = &DDStart[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DDStart[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DDStart[DIR_MPM * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////
@@ -170,33 +170,33 @@ __global__ void Cumulant_One_preconditioned_errorDiffusion_chim_Comp_SP_27(
//////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
- real mfcbb = (D.f[DIR_P00])[k];//[ke ];// + c2over27 ;(D.f[DIR_P00 ])[k ];//ke
- real mfabb = (D.f[DIR_M00])[kw];//[kw ];// + c2over27 ;(D.f[DIR_M00 ])[kw ];
- real mfbcb = (D.f[DIR_0P0])[k];//[kn ];// + c2over27 ;(D.f[DIR_0P0 ])[k ];//kn
- real mfbab = (D.f[DIR_0M0])[ks];//[ks ];// + c2over27 ;(D.f[DIR_0M0 ])[ks ];
- real mfbbc = (D.f[DIR_00P])[k];//[kt ];// + c2over27 ;(D.f[DIR_00P ])[k ];//kt
- real mfbba = (D.f[DIR_00M])[kb];//[kb ];// + c2over27 ;(D.f[DIR_00M ])[kb ];
- real mfccb = (D.f[DIR_PP0])[k];//[kne ];// + c1over54 ;(D.f[DIR_PP0 ])[k ];//kne
- real mfaab = (D.f[DIR_MM0])[ksw];//[ksw ];// + c1over54 ;(D.f[DIR_MM0 ])[ksw];
- real mfcab = (D.f[DIR_PM0])[ks];//[kse ];// + c1over54 ;(D.f[DIR_PM0 ])[ks ];//kse
- real mfacb = (D.f[DIR_MP0])[kw];//[knw ];// + c1over54 ;(D.f[DIR_MP0 ])[kw ];//knw
- real mfcbc = (D.f[DIR_P0P])[k];//[kte ];// + c1over54 ;(D.f[DIR_P0P ])[k ];//kte
- real mfaba = (D.f[DIR_M0M])[kbw];//[kbw ];// + c1over54 ;(D.f[DIR_M0M ])[kbw];
- real mfcba = (D.f[DIR_P0M])[kb];//[kbe ];// + c1over54 ;(D.f[DIR_P0M ])[kb ];//kbe
- real mfabc = (D.f[DIR_M0P])[kw];//[ktw ];// + c1over54 ;(D.f[DIR_M0P ])[kw ];//ktw
- real mfbcc = (D.f[DIR_0PP])[k];//[ktn ];// + c1over54 ;(D.f[DIR_0PP ])[k ];//ktn
- real mfbaa = (D.f[DIR_0MM])[kbs];//[kbs ];// + c1over54 ;(D.f[DIR_0MM ])[kbs];
- real mfbca = (D.f[DIR_0PM])[kb];//[kbn ];// + c1over54 ;(D.f[DIR_0PM ])[kb ];//kbn
- real mfbac = (D.f[DIR_0MP])[ks];//[kts ];// + c1over54 ;(D.f[DIR_0MP ])[ks ];//kts
+ real mfcbb = (D.f[DIR_P00])[k];//[ke ];// + c2over27 ;(D.f[DIR_P00])[k ];//ke
+ real mfabb = (D.f[DIR_M00])[kw];//[kw ];// + c2over27 ;(D.f[DIR_M00])[kw ];
+ real mfbcb = (D.f[DIR_0P0])[k];//[kn ];// + c2over27 ;(D.f[DIR_0P0])[k ];//kn
+ real mfbab = (D.f[DIR_0M0])[ks];//[ks ];// + c2over27 ;(D.f[DIR_0M0])[ks ];
+ real mfbbc = (D.f[DIR_00P])[k];//[kt ];// + c2over27 ;(D.f[DIR_00P])[k ];//kt
+ real mfbba = (D.f[DIR_00M])[kb];//[kb ];// + c2over27 ;(D.f[DIR_00M])[kb ];
+ real mfccb = (D.f[DIR_PP0])[k];//[kne ];// + c1over54 ;(D.f[DIR_PP0])[k ];//kne
+ real mfaab = (D.f[DIR_MM0])[ksw];//[ksw ];// + c1over54 ;(D.f[DIR_MM0])[ksw];
+ real mfcab = (D.f[DIR_PM0])[ks];//[kse ];// + c1over54 ;(D.f[DIR_PM0])[ks ];//kse
+ real mfacb = (D.f[DIR_MP0])[kw];//[knw ];// + c1over54 ;(D.f[DIR_MP0])[kw ];//knw
+ real mfcbc = (D.f[DIR_P0P])[k];//[kte ];// + c1over54 ;(D.f[DIR_P0P])[k ];//kte
+ real mfaba = (D.f[DIR_M0M])[kbw];//[kbw ];// + c1over54 ;(D.f[DIR_M0M])[kbw];
+ real mfcba = (D.f[DIR_P0M])[kb];//[kbe ];// + c1over54 ;(D.f[DIR_P0M])[kb ];//kbe
+ real mfabc = (D.f[DIR_M0P])[kw];//[ktw ];// + c1over54 ;(D.f[DIR_M0P])[kw ];//ktw
+ real mfbcc = (D.f[DIR_0PP])[k];//[ktn ];// + c1over54 ;(D.f[DIR_0PP])[k ];//ktn
+ real mfbaa = (D.f[DIR_0MM])[kbs];//[kbs ];// + c1over54 ;(D.f[DIR_0MM])[kbs];
+ real mfbca = (D.f[DIR_0PM])[kb];//[kbn ];// + c1over54 ;(D.f[DIR_0PM])[kb ];//kbn
+ real mfbac = (D.f[DIR_0MP])[ks];//[kts ];// + c1over54 ;(D.f[DIR_0MP])[ks ];//kts
real mfbbb = (D.f[DIR_000])[k];//[kzero];// + c8over27 ;(D.f[DIR_000])[k ];//kzero
- real mfccc = (D.f[DIR_PPP])[k];//[ktne ];// + c1over216;(D.f[DIR_PPP ])[k ];//ktne
- real mfaac = (D.f[DIR_MMP])[ksw];//[ktsw ];// + c1over216;(D.f[DIR_MMP ])[ksw];//ktsw
- real mfcac = (D.f[DIR_PMP])[ks];//[ktse ];// + c1over216;(D.f[DIR_PMP ])[ks ];//ktse
- real mfacc = (D.f[DIR_MPP])[kw];//[ktnw ];// + c1over216;(D.f[DIR_MPP ])[kw ];//ktnw
- real mfcca = (D.f[DIR_PPM])[kb];//[kbne ];// + c1over216;(D.f[DIR_PPM ])[kb ];//kbne
- real mfaaa = (D.f[DIR_MMM])[kbsw];//[kbsw ];// + c1over216;(D.f[DIR_MMM ])[kbsw];
- real mfcaa = (D.f[DIR_PMM])[kbs];//[kbse ];// + c1over216;(D.f[DIR_PMM ])[kbs];//kbse
- real mfaca = (D.f[DIR_MPM])[kbw];//[kbnw ];// + c1over216;(D.f[DIR_MPM ])[kbw];//kbnw
+ real mfccc = (D.f[DIR_PPP])[k];//[ktne ];// + c1over216;(D.f[DIR_PPP])[k ];//ktne
+ real mfaac = (D.f[DIR_MMP])[ksw];//[ktsw ];// + c1over216;(D.f[DIR_MMP])[ksw];//ktsw
+ real mfcac = (D.f[DIR_PMP])[ks];//[ktse ];// + c1over216;(D.f[DIR_PMP])[ks ];//ktse
+ real mfacc = (D.f[DIR_MPP])[kw];//[ktnw ];// + c1over216;(D.f[DIR_MPP])[kw ];//ktnw
+ real mfcca = (D.f[DIR_PPM])[kb];//[kbne ];// + c1over216;(D.f[DIR_PPM])[kb ];//kbne
+ real mfaaa = (D.f[DIR_MMM])[kbsw];//[kbsw ];// + c1over216;(D.f[DIR_MMM])[kbsw];
+ real mfcaa = (D.f[DIR_PMM])[kbs];//[kbse ];// + c1over216;(D.f[DIR_PMM])[kbs];//kbse
+ real mfaca = (D.f[DIR_MPM])[kbw];//[kbnw ];// + c1over216;(D.f[DIR_MPM])[kbw];//kbnw
////////////////////////////////////////////////////////////////////////////////////
real drho = ((((mfccc + mfaaa) + (mfaca + mfcac)) + ((mfacc + mfcaa) + (mfaac + mfcca))) +
(((mfbac + mfbca) + (mfbaa + mfbcc)) + ((mfabc + mfcba) + (mfaba + mfcbc)) + ((mfacb + mfcab) + (mfaab + mfccb))) +
@@ -966,7 +966,7 @@ __global__ void Cumulant_One_preconditioned_chim_Comp_SP_27(
unsigned int* neighborY,
unsigned int* neighborZ,
real* DDStart,
- int size_Mat,
+ unsigned long long numberOfLBnodes,
int level,
real* forces,
bool EvenOrOdd)
@@ -982,7 +982,7 @@ __global__ void Cumulant_One_preconditioned_chim_Comp_SP_27(
const unsigned k = nx*(ny*z + y) + x;
//////////////////////////////////////////////////////////////////////////
- if (k<size_Mat)
+ if (k<numberOfLBnodes)
{
////////////////////////////////////////////////////////////////////////////////
unsigned int BC;
@@ -993,63 +993,63 @@ __global__ void Cumulant_One_preconditioned_chim_Comp_SP_27(
Distributions27 D;
if (EvenOrOdd == true)
{
- D.f[DIR_P00] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_M00] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0P0] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0M0] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00P] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00M] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_PP0] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_MM0] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_PM0] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_MP0] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_P0P] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_M0M] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_P0M] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_M0P] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0PP] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0MM] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0PM] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0MP] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_PPP] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_MMP] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_PMP] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_MPP] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_PPM] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_MMM] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_PMM] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_MPM] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DDStart[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DDStart[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DDStart[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DDStart[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DDStart[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DDStart[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DDStart[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DDStart[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DDStart[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DDStart[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DDStart[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DDStart[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DDStart[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DDStart[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DDStart[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DDStart[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DDStart[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DDStart[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DDStart[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DDStart[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DDStart[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DDStart[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DDStart[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DDStart[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DDStart[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DDStart[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DDStart[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_P00] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0M0] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0P0] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00M] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00P] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_MM0] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_PP0] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_MP0] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_PM0] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_M0M] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_P0P] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_M0P] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_P0M] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0MM] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0PP] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0MP] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0PM] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_MMM] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_PPM] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_MPM] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_PMM] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_MMP] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_PPP] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_MPP] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_PMP] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_M00] = &DDStart[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DDStart[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DDStart[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DDStart[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DDStart[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DDStart[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DDStart[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DDStart[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DDStart[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DDStart[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DDStart[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DDStart[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DDStart[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DDStart[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DDStart[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DDStart[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DDStart[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DDStart[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DDStart[DIR_000 * numberOfLBnodes];
+ D.f[DIR_MMM] = &DDStart[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DDStart[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DDStart[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DDStart[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DDStart[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PPP] = &DDStart[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DDStart[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DDStart[DIR_MPM * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////
@@ -1085,33 +1085,33 @@ __global__ void Cumulant_One_preconditioned_chim_Comp_SP_27(
//////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
- real mfcbb = (D.f[DIR_P00])[k];//[ke ];// + c2over27 ;(D.f[DIR_P00 ])[k ];//ke
- real mfabb = (D.f[DIR_M00])[kw];//[kw ];// + c2over27 ;(D.f[DIR_M00 ])[kw ];
- real mfbcb = (D.f[DIR_0P0])[k];//[kn ];// + c2over27 ;(D.f[DIR_0P0 ])[k ];//kn
- real mfbab = (D.f[DIR_0M0])[ks];//[ks ];// + c2over27 ;(D.f[DIR_0M0 ])[ks ];
- real mfbbc = (D.f[DIR_00P])[k];//[kt ];// + c2over27 ;(D.f[DIR_00P ])[k ];//kt
- real mfbba = (D.f[DIR_00M])[kb];//[kb ];// + c2over27 ;(D.f[DIR_00M ])[kb ];
- real mfccb = (D.f[DIR_PP0])[k];//[kne ];// + c1over54 ;(D.f[DIR_PP0 ])[k ];//kne
- real mfaab = (D.f[DIR_MM0])[ksw];//[ksw ];// + c1over54 ;(D.f[DIR_MM0 ])[ksw];
- real mfcab = (D.f[DIR_PM0])[ks];//[kse ];// + c1over54 ;(D.f[DIR_PM0 ])[ks ];//kse
- real mfacb = (D.f[DIR_MP0])[kw];//[knw ];// + c1over54 ;(D.f[DIR_MP0 ])[kw ];//knw
- real mfcbc = (D.f[DIR_P0P])[k];//[kte ];// + c1over54 ;(D.f[DIR_P0P ])[k ];//kte
- real mfaba = (D.f[DIR_M0M])[kbw];//[kbw ];// + c1over54 ;(D.f[DIR_M0M ])[kbw];
- real mfcba = (D.f[DIR_P0M])[kb];//[kbe ];// + c1over54 ;(D.f[DIR_P0M ])[kb ];//kbe
- real mfabc = (D.f[DIR_M0P])[kw];//[ktw ];// + c1over54 ;(D.f[DIR_M0P ])[kw ];//ktw
- real mfbcc = (D.f[DIR_0PP])[k];//[ktn ];// + c1over54 ;(D.f[DIR_0PP ])[k ];//ktn
- real mfbaa = (D.f[DIR_0MM])[kbs];//[kbs ];// + c1over54 ;(D.f[DIR_0MM ])[kbs];
- real mfbca = (D.f[DIR_0PM])[kb];//[kbn ];// + c1over54 ;(D.f[DIR_0PM ])[kb ];//kbn
- real mfbac = (D.f[DIR_0MP])[ks];//[kts ];// + c1over54 ;(D.f[DIR_0MP ])[ks ];//kts
+ real mfcbb = (D.f[DIR_P00])[k];//[ke ];// + c2over27 ;(D.f[DIR_P00])[k ];//ke
+ real mfabb = (D.f[DIR_M00])[kw];//[kw ];// + c2over27 ;(D.f[DIR_M00])[kw ];
+ real mfbcb = (D.f[DIR_0P0])[k];//[kn ];// + c2over27 ;(D.f[DIR_0P0])[k ];//kn
+ real mfbab = (D.f[DIR_0M0])[ks];//[ks ];// + c2over27 ;(D.f[DIR_0M0])[ks ];
+ real mfbbc = (D.f[DIR_00P])[k];//[kt ];// + c2over27 ;(D.f[DIR_00P])[k ];//kt
+ real mfbba = (D.f[DIR_00M])[kb];//[kb ];// + c2over27 ;(D.f[DIR_00M])[kb ];
+ real mfccb = (D.f[DIR_PP0])[k];//[kne ];// + c1over54 ;(D.f[DIR_PP0])[k ];//kne
+ real mfaab = (D.f[DIR_MM0])[ksw];//[ksw ];// + c1over54 ;(D.f[DIR_MM0])[ksw];
+ real mfcab = (D.f[DIR_PM0])[ks];//[kse ];// + c1over54 ;(D.f[DIR_PM0])[ks ];//kse
+ real mfacb = (D.f[DIR_MP0])[kw];//[knw ];// + c1over54 ;(D.f[DIR_MP0])[kw ];//knw
+ real mfcbc = (D.f[DIR_P0P])[k];//[kte ];// + c1over54 ;(D.f[DIR_P0P])[k ];//kte
+ real mfaba = (D.f[DIR_M0M])[kbw];//[kbw ];// + c1over54 ;(D.f[DIR_M0M])[kbw];
+ real mfcba = (D.f[DIR_P0M])[kb];//[kbe ];// + c1over54 ;(D.f[DIR_P0M])[kb ];//kbe
+ real mfabc = (D.f[DIR_M0P])[kw];//[ktw ];// + c1over54 ;(D.f[DIR_M0P])[kw ];//ktw
+ real mfbcc = (D.f[DIR_0PP])[k];//[ktn ];// + c1over54 ;(D.f[DIR_0PP])[k ];//ktn
+ real mfbaa = (D.f[DIR_0MM])[kbs];//[kbs ];// + c1over54 ;(D.f[DIR_0MM])[kbs];
+ real mfbca = (D.f[DIR_0PM])[kb];//[kbn ];// + c1over54 ;(D.f[DIR_0PM])[kb ];//kbn
+ real mfbac = (D.f[DIR_0MP])[ks];//[kts ];// + c1over54 ;(D.f[DIR_0MP])[ks ];//kts
real mfbbb = (D.f[DIR_000])[k];//[kzero];// + c8over27 ;(D.f[DIR_000])[k ];//kzero
- real mfccc = (D.f[DIR_PPP])[k];//[ktne ];// + c1over216;(D.f[DIR_PPP ])[k ];//ktne
- real mfaac = (D.f[DIR_MMP])[ksw];//[ktsw ];// + c1over216;(D.f[DIR_MMP ])[ksw];//ktsw
- real mfcac = (D.f[DIR_PMP])[ks];//[ktse ];// + c1over216;(D.f[DIR_PMP ])[ks ];//ktse
- real mfacc = (D.f[DIR_MPP])[kw];//[ktnw ];// + c1over216;(D.f[DIR_MPP ])[kw ];//ktnw
- real mfcca = (D.f[DIR_PPM])[kb];//[kbne ];// + c1over216;(D.f[DIR_PPM ])[kb ];//kbne
- real mfaaa = (D.f[DIR_MMM])[kbsw];//[kbsw ];// + c1over216;(D.f[DIR_MMM ])[kbsw];
- real mfcaa = (D.f[DIR_PMM])[kbs];//[kbse ];// + c1over216;(D.f[DIR_PMM ])[kbs];//kbse
- real mfaca = (D.f[DIR_MPM])[kbw];//[kbnw ];// + c1over216;(D.f[DIR_MPM ])[kbw];//kbnw
+ real mfccc = (D.f[DIR_PPP])[k];//[ktne ];// + c1over216;(D.f[DIR_PPP])[k ];//ktne
+ real mfaac = (D.f[DIR_MMP])[ksw];//[ktsw ];// + c1over216;(D.f[DIR_MMP])[ksw];//ktsw
+ real mfcac = (D.f[DIR_PMP])[ks];//[ktse ];// + c1over216;(D.f[DIR_PMP])[ks ];//ktse
+ real mfacc = (D.f[DIR_MPP])[kw];//[ktnw ];// + c1over216;(D.f[DIR_MPP])[kw ];//ktnw
+ real mfcca = (D.f[DIR_PPM])[kb];//[kbne ];// + c1over216;(D.f[DIR_PPM])[kb ];//kbne
+ real mfaaa = (D.f[DIR_MMM])[kbsw];//[kbsw ];// + c1over216;(D.f[DIR_MMM])[kbsw];
+ real mfcaa = (D.f[DIR_PMM])[kbs];//[kbse ];// + c1over216;(D.f[DIR_PMM])[kbs];//kbse
+ real mfaca = (D.f[DIR_MPM])[kbw];//[kbnw ];// + c1over216;(D.f[DIR_MPM])[kbw];//kbnw
////////////////////////////////////////////////////////////////////////////////////
real drho = ((((mfccc + mfaaa) + (mfaca + mfcac)) + ((mfacc + mfcaa) + (mfaac + mfcca))) +
(((mfbac + mfbca) + (mfbaa + mfbcc)) + ((mfabc + mfcba) + (mfaba + mfcbc)) + ((mfacb + mfcab) + (mfaab + mfccb))) +
@@ -1762,7 +1762,7 @@ __global__ void Cumulant_One_chim_Comp_SP_27(
unsigned int* neighborY,
unsigned int* neighborZ,
real* DDStart,
- int size_Mat,
+ unsigned long long numberOfLBnodes,
int level,
real* forces,
bool EvenOrOdd)
@@ -1778,7 +1778,7 @@ __global__ void Cumulant_One_chim_Comp_SP_27(
const unsigned k = nx*(ny*z + y) + x;
//////////////////////////////////////////////////////////////////////////
- if (k<size_Mat)
+ if (k<numberOfLBnodes)
{
////////////////////////////////////////////////////////////////////////////////
unsigned int BC;
@@ -1789,63 +1789,63 @@ __global__ void Cumulant_One_chim_Comp_SP_27(
Distributions27 D;
if (EvenOrOdd == true)
{
- D.f[DIR_P00 ] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DDStart[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DDStart[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DDStart[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DDStart[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DDStart[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DDStart[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DDStart[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DDStart[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DDStart[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DDStart[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DDStart[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DDStart[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DDStart[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DDStart[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DDStart[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DDStart[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DDStart[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DDStart[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DDStart[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DDStart[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DDStart[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DDStart[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DDStart[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DDStart[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DDStart[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DDStart[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DDStart[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00 ] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_MMM ] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_PPM ] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_MPM ] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_PMM ] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_MMP ] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_PPP ] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_MPP ] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_PMP ] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_M00] = &DDStart[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DDStart[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DDStart[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DDStart[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DDStart[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DDStart[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DDStart[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DDStart[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DDStart[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DDStart[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DDStart[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DDStart[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DDStart[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DDStart[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DDStart[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DDStart[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DDStart[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DDStart[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DDStart[DIR_000 * numberOfLBnodes];
+ D.f[DIR_MMM] = &DDStart[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DDStart[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DDStart[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DDStart[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DDStart[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PPP] = &DDStart[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DDStart[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DDStart[DIR_MPM * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////
//index
@@ -1857,33 +1857,33 @@ __global__ void Cumulant_One_chim_Comp_SP_27(
unsigned int kbs = neighborZ[ks];
unsigned int kbsw = neighborZ[ksw];
////////////////////////////////////////////////////////////////////////////////////
- real mfcbb = (D.f[DIR_P00 ])[k ];
- real mfabb = (D.f[DIR_M00 ])[kw ];
- real mfbcb = (D.f[DIR_0P0 ])[k ];
- real mfbab = (D.f[DIR_0M0 ])[ks ];
- real mfbbc = (D.f[DIR_00P ])[k ];
- real mfbba = (D.f[DIR_00M ])[kb ];
- real mfccb = (D.f[DIR_PP0 ])[k ];
- real mfaab = (D.f[DIR_MM0 ])[ksw ];
- real mfcab = (D.f[DIR_PM0 ])[ks ];
- real mfacb = (D.f[DIR_MP0 ])[kw ];
- real mfcbc = (D.f[DIR_P0P ])[k ];
- real mfaba = (D.f[DIR_M0M ])[kbw ];
- real mfcba = (D.f[DIR_P0M ])[kb ];
- real mfabc = (D.f[DIR_M0P ])[kw ];
- real mfbcc = (D.f[DIR_0PP ])[k ];
- real mfbaa = (D.f[DIR_0MM ])[kbs ];
- real mfbca = (D.f[DIR_0PM ])[kb ];
- real mfbac = (D.f[DIR_0MP ])[ks ];
+ real mfcbb = (D.f[DIR_P00])[k ];
+ real mfabb = (D.f[DIR_M00])[kw ];
+ real mfbcb = (D.f[DIR_0P0])[k ];
+ real mfbab = (D.f[DIR_0M0])[ks ];
+ real mfbbc = (D.f[DIR_00P])[k ];
+ real mfbba = (D.f[DIR_00M])[kb ];
+ real mfccb = (D.f[DIR_PP0])[k ];
+ real mfaab = (D.f[DIR_MM0])[ksw ];
+ real mfcab = (D.f[DIR_PM0])[ks ];
+ real mfacb = (D.f[DIR_MP0])[kw ];
+ real mfcbc = (D.f[DIR_P0P])[k ];
+ real mfaba = (D.f[DIR_M0M])[kbw ];
+ real mfcba = (D.f[DIR_P0M])[kb ];
+ real mfabc = (D.f[DIR_M0P])[kw ];
+ real mfbcc = (D.f[DIR_0PP])[k ];
+ real mfbaa = (D.f[DIR_0MM])[kbs ];
+ real mfbca = (D.f[DIR_0PM])[kb ];
+ real mfbac = (D.f[DIR_0MP])[ks ];
real mfbbb = (D.f[DIR_000])[k ];
- real mfccc = (D.f[DIR_PPP ])[k ];
- real mfaac = (D.f[DIR_MMP ])[ksw ];
- real mfcac = (D.f[DIR_PMP ])[ks ];
- real mfacc = (D.f[DIR_MPP ])[kw ];
- real mfcca = (D.f[DIR_PPM ])[kb ];
- real mfaaa = (D.f[DIR_MMM ])[kbsw];
- real mfcaa = (D.f[DIR_PMM ])[kbs ];
- real mfaca = (D.f[DIR_MPM ])[kbw ];
+ real mfccc = (D.f[DIR_PPP])[k ];
+ real mfaac = (D.f[DIR_MMP])[ksw ];
+ real mfcac = (D.f[DIR_PMP])[ks ];
+ real mfacc = (D.f[DIR_MPP])[kw ];
+ real mfcca = (D.f[DIR_PPM])[kb ];
+ real mfaaa = (D.f[DIR_MMM])[kbsw];
+ real mfcaa = (D.f[DIR_PMM])[kbs ];
+ real mfaca = (D.f[DIR_MPM])[kbw ];
////////////////////////////////////////////////////////////////////////////////////
real drho = ((((mfccc + mfaaa) + (mfaca + mfcac)) + ((mfacc + mfcaa) + (mfaac + mfcca))) +
(((mfbac + mfbca) + (mfbaa + mfbcc)) + ((mfabc + mfcba) + (mfaba + mfcbc)) + ((mfacb + mfcab) + (mfaab + mfccb))) +
@@ -2204,33 +2204,33 @@ __global__ void Cumulant_One_chim_Comp_SP_27(
((mfabb + mfcbb) + (mfbab + mfbcb) + (mfbba + mfbbc))) + mfbbb;
mfbbb += drho - drhoPost;
////////////////////////////////////////////////////////////////////////////////////
- (D.f[DIR_P00 ])[k ] = mfabb;
- (D.f[DIR_M00 ])[kw ] = mfcbb;
- (D.f[DIR_0P0 ])[k ] = mfbab;
- (D.f[DIR_0M0 ])[ks ] = mfbcb;
- (D.f[DIR_00P ])[k ] = mfbba;
- (D.f[DIR_00M ])[kb ] = mfbbc;
- (D.f[DIR_PP0 ])[k ] = mfaab;
- (D.f[DIR_MM0 ])[ksw ] = mfccb;
- (D.f[DIR_PM0 ])[ks ] = mfacb;
- (D.f[DIR_MP0 ])[kw ] = mfcab;
- (D.f[DIR_P0P ])[k ] = mfaba;
- (D.f[DIR_M0M ])[kbw ] = mfcbc;
- (D.f[DIR_P0M ])[kb ] = mfabc;
- (D.f[DIR_M0P ])[kw ] = mfcba;
- (D.f[DIR_0PP ])[k ] = mfbaa;
- (D.f[DIR_0MM ])[kbs ] = mfbcc;
- (D.f[DIR_0PM ])[kb ] = mfbac;
- (D.f[DIR_0MP ])[ks ] = mfbca;
+ (D.f[DIR_P00])[k ] = mfabb;
+ (D.f[DIR_M00])[kw ] = mfcbb;
+ (D.f[DIR_0P0])[k ] = mfbab;
+ (D.f[DIR_0M0])[ks ] = mfbcb;
+ (D.f[DIR_00P])[k ] = mfbba;
+ (D.f[DIR_00M])[kb ] = mfbbc;
+ (D.f[DIR_PP0])[k ] = mfaab;
+ (D.f[DIR_MM0])[ksw ] = mfccb;
+ (D.f[DIR_PM0])[ks ] = mfacb;
+ (D.f[DIR_MP0])[kw ] = mfcab;
+ (D.f[DIR_P0P])[k ] = mfaba;
+ (D.f[DIR_M0M])[kbw ] = mfcbc;
+ (D.f[DIR_P0M])[kb ] = mfabc;
+ (D.f[DIR_M0P])[kw ] = mfcba;
+ (D.f[DIR_0PP])[k ] = mfbaa;
+ (D.f[DIR_0MM])[kbs ] = mfbcc;
+ (D.f[DIR_0PM])[kb ] = mfbac;
+ (D.f[DIR_0MP])[ks ] = mfbca;
(D.f[DIR_000])[k ] = mfbbb;
- (D.f[DIR_PPP ])[k ] = mfaaa;
- (D.f[DIR_PMP ])[ks ] = mfaca;
- (D.f[DIR_PPM ])[kb ] = mfaac;
- (D.f[DIR_PMM ])[kbs ] = mfacc;
- (D.f[DIR_MPP ])[kw ] = mfcaa;
- (D.f[DIR_MMP ])[ksw ] = mfcca;
- (D.f[DIR_MPM ])[kbw ] = mfcac;
- (D.f[DIR_MMM ])[kbsw] = mfccc;
+ (D.f[DIR_PPP])[k ] = mfaaa;
+ (D.f[DIR_PMP])[ks ] = mfaca;
+ (D.f[DIR_PPM])[kb ] = mfaac;
+ (D.f[DIR_PMM])[kbs ] = mfacc;
+ (D.f[DIR_MPP])[kw ] = mfcaa;
+ (D.f[DIR_MMP])[ksw ] = mfcca;
+ (D.f[DIR_MPM])[kbw ] = mfcac;
+ (D.f[DIR_MMM])[kbsw] = mfccc;
}
}
}
diff --git a/src/gpu/VirtualFluids_GPU/GPU/Cumulant_F3_27.cu b/src/gpu/VirtualFluids_GPU/GPU/Cumulant_F3_27.cu
index 7adfd40da157d825d83c63b084bf1f855ea6dca2..c89c3cfe87560c808d47163b45d512fa0d7e494f 100644
--- a/src/gpu/VirtualFluids_GPU/GPU/Cumulant_F3_27.cu
+++ b/src/gpu/VirtualFluids_GPU/GPU/Cumulant_F3_27.cu
@@ -27,7 +27,7 @@ __global__ void LB_PostProcessor_F3_2018_Fehlberg(real omega,
real* vzOut,
real* DDStart,
real* G6,
- int size_Mat,
+ unsigned long long numberOfLBnodes,
int level,
real* forces,
bool EvenOrOdd)
@@ -43,7 +43,7 @@ __global__ void LB_PostProcessor_F3_2018_Fehlberg(real omega,
const unsigned k = nx*(ny*z + y) + x;
//////////////////////////////////////////////////////////////////////////
- if (k < size_Mat)
+ if (k < numberOfLBnodes)
{
////////////////////////////////////////////////////////////////////////////////
unsigned int BC;
@@ -54,83 +54,83 @@ __global__ void LB_PostProcessor_F3_2018_Fehlberg(real omega,
Distributions27 D;
if (EvenOrOdd == true)
{
- D.f[DIR_P00] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_M00] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0P0] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0M0] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00P] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00M] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_PP0] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_MM0] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_PM0] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_MP0] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_P0P] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_M0M] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_P0M] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_M0P] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0PP] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0MM] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0PM] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0MP] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_PPP] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_MMP] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_PMP] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_MPP] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_PPM] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_MMM] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_PMM] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_MPM] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DDStart[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DDStart[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DDStart[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DDStart[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DDStart[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DDStart[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DDStart[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DDStart[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DDStart[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DDStart[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DDStart[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DDStart[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DDStart[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DDStart[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DDStart[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DDStart[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DDStart[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DDStart[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DDStart[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DDStart[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DDStart[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DDStart[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DDStart[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DDStart[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DDStart[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DDStart[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DDStart[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_P00] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0M0] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0P0] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00M] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00P] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_MM0] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_PP0] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_MP0] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_PM0] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_M0M] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_P0P] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_M0P] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_P0M] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0MM] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0PP] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0MP] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0PM] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_MMM] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_PPM] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_MPM] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_PMM] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_MMP] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_PPP] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_MPP] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_PMP] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_M00] = &DDStart[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DDStart[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DDStart[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DDStart[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DDStart[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DDStart[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DDStart[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DDStart[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DDStart[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DDStart[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DDStart[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DDStart[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DDStart[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DDStart[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DDStart[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DDStart[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DDStart[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DDStart[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DDStart[DIR_000 * numberOfLBnodes];
+ D.f[DIR_MMM] = &DDStart[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DDStart[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DDStart[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DDStart[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DDStart[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PPP] = &DDStart[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DDStart[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DDStart[DIR_MPM * numberOfLBnodes];
}
Distributions6 G;
if (EvenOrOdd == true)
{
- G.g[DIR_P00] = &G6[DIR_P00 *size_Mat];
- G.g[DIR_M00] = &G6[DIR_M00 *size_Mat];
- G.g[DIR_0P0] = &G6[DIR_0P0 *size_Mat];
- G.g[DIR_0M0] = &G6[DIR_0M0 *size_Mat];
- G.g[DIR_00P] = &G6[DIR_00P *size_Mat];
- G.g[DIR_00M] = &G6[DIR_00M *size_Mat];
+ G.g[DIR_P00] = &G6[DIR_P00 * numberOfLBnodes];
+ G.g[DIR_M00] = &G6[DIR_M00 * numberOfLBnodes];
+ G.g[DIR_0P0] = &G6[DIR_0P0 * numberOfLBnodes];
+ G.g[DIR_0M0] = &G6[DIR_0M0 * numberOfLBnodes];
+ G.g[DIR_00P] = &G6[DIR_00P * numberOfLBnodes];
+ G.g[DIR_00M] = &G6[DIR_00M * numberOfLBnodes];
}
else
{
- G.g[DIR_M00] = &G6[DIR_P00 *size_Mat];
- G.g[DIR_P00] = &G6[DIR_M00 *size_Mat];
- G.g[DIR_0M0] = &G6[DIR_0P0 *size_Mat];
- G.g[DIR_0P0] = &G6[DIR_0M0 *size_Mat];
- G.g[DIR_00M] = &G6[DIR_00P *size_Mat];
- G.g[DIR_00P] = &G6[DIR_00M *size_Mat];
+ G.g[DIR_M00] = &G6[DIR_P00 * numberOfLBnodes];
+ G.g[DIR_P00] = &G6[DIR_M00 * numberOfLBnodes];
+ G.g[DIR_0M0] = &G6[DIR_0P0 * numberOfLBnodes];
+ G.g[DIR_0P0] = &G6[DIR_0M0 * numberOfLBnodes];
+ G.g[DIR_00M] = &G6[DIR_00P * numberOfLBnodes];
+ G.g[DIR_00P] = &G6[DIR_00M * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////
@@ -1026,83 +1026,83 @@ __global__ void LB_PostProcessor_F3_2018_Fehlberg(real omega,
// Distributions27 D;
// if (EvenOrOdd == true)
// {
-// D.f[DIR_P00] = &DDStart[DIR_P00 *size_Mat];
-// D.f[DIR_M00] = &DDStart[DIR_M00 *size_Mat];
-// D.f[DIR_0P0] = &DDStart[DIR_0P0 *size_Mat];
-// D.f[DIR_0M0] = &DDStart[DIR_0M0 *size_Mat];
-// D.f[DIR_00P] = &DDStart[DIR_00P *size_Mat];
-// D.f[DIR_00M] = &DDStart[DIR_00M *size_Mat];
-// D.f[DIR_PP0] = &DDStart[DIR_PP0 *size_Mat];
-// D.f[DIR_MM0] = &DDStart[DIR_MM0 *size_Mat];
-// D.f[DIR_PM0] = &DDStart[DIR_PM0 *size_Mat];
-// D.f[DIR_MP0] = &DDStart[DIR_MP0 *size_Mat];
-// D.f[DIR_P0P] = &DDStart[DIR_P0P *size_Mat];
-// D.f[DIR_M0M] = &DDStart[DIR_M0M *size_Mat];
-// D.f[DIR_P0M] = &DDStart[DIR_P0M *size_Mat];
-// D.f[DIR_M0P] = &DDStart[DIR_M0P *size_Mat];
-// D.f[DIR_0PP] = &DDStart[DIR_0PP *size_Mat];
-// D.f[DIR_0MM] = &DDStart[DIR_0MM *size_Mat];
-// D.f[DIR_0PM] = &DDStart[DIR_0PM *size_Mat];
-// D.f[DIR_0MP] = &DDStart[DIR_0MP *size_Mat];
-// D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
-// D.f[DIR_PPP] = &DDStart[DIR_PPP *size_Mat];
-// D.f[DIR_MMP] = &DDStart[DIR_MMP *size_Mat];
-// D.f[DIR_PMP] = &DDStart[DIR_PMP *size_Mat];
-// D.f[DIR_MPP] = &DDStart[DIR_MPP *size_Mat];
-// D.f[DIR_PPM] = &DDStart[DIR_PPM *size_Mat];
-// D.f[DIR_MMM] = &DDStart[DIR_MMM *size_Mat];
-// D.f[DIR_PMM] = &DDStart[DIR_PMM *size_Mat];
-// D.f[DIR_MPM] = &DDStart[DIR_MPM *size_Mat];
+// D.f[DIR_P00] = &DDStart[DIR_P00 * size_Mat];
+// D.f[DIR_M00] = &DDStart[DIR_M00 * size_Mat];
+// D.f[DIR_0P0] = &DDStart[DIR_0P0 * size_Mat];
+// D.f[DIR_0M0] = &DDStart[DIR_0M0 * size_Mat];
+// D.f[DIR_00P] = &DDStart[DIR_00P * size_Mat];
+// D.f[DIR_00M] = &DDStart[DIR_00M * size_Mat];
+// D.f[DIR_PP0] = &DDStart[DIR_PP0 * size_Mat];
+// D.f[DIR_MM0] = &DDStart[DIR_MM0 * size_Mat];
+// D.f[DIR_PM0] = &DDStart[DIR_PM0 * size_Mat];
+// D.f[DIR_MP0] = &DDStart[DIR_MP0 * size_Mat];
+// D.f[DIR_P0P] = &DDStart[DIR_P0P * size_Mat];
+// D.f[DIR_M0M] = &DDStart[DIR_M0M * size_Mat];
+// D.f[DIR_P0M] = &DDStart[DIR_P0M * size_Mat];
+// D.f[DIR_M0P] = &DDStart[DIR_M0P * size_Mat];
+// D.f[DIR_0PP] = &DDStart[DIR_0PP * size_Mat];
+// D.f[DIR_0MM] = &DDStart[DIR_0MM * size_Mat];
+// D.f[DIR_0PM] = &DDStart[DIR_0PM * size_Mat];
+// D.f[DIR_0MP] = &DDStart[DIR_0MP * size_Mat];
+// D.f[DIR_000] = &DDStart[DIR_000 * size_Mat];
+// D.f[DIR_PPP] = &DDStart[DIR_PPP * size_Mat];
+// D.f[DIR_MMP] = &DDStart[DIR_MMP * size_Mat];
+// D.f[DIR_PMP] = &DDStart[DIR_PMP * size_Mat];
+// D.f[DIR_MPP] = &DDStart[DIR_MPP * size_Mat];
+// D.f[DIR_PPM] = &DDStart[DIR_PPM * size_Mat];
+// D.f[DIR_MMM] = &DDStart[DIR_MMM * size_Mat];
+// D.f[DIR_PMM] = &DDStart[DIR_PMM * size_Mat];
+// D.f[DIR_MPM] = &DDStart[DIR_MPM * size_Mat];
// }
// else
// {
-// D.f[DIR_M00] = &DDStart[DIR_P00 *size_Mat];
-// D.f[DIR_P00] = &DDStart[DIR_M00 *size_Mat];
-// D.f[DIR_0M0] = &DDStart[DIR_0P0 *size_Mat];
-// D.f[DIR_0P0] = &DDStart[DIR_0M0 *size_Mat];
-// D.f[DIR_00M] = &DDStart[DIR_00P *size_Mat];
-// D.f[DIR_00P] = &DDStart[DIR_00M *size_Mat];
-// D.f[DIR_MM0] = &DDStart[DIR_PP0 *size_Mat];
-// D.f[DIR_PP0] = &DDStart[DIR_MM0 *size_Mat];
-// D.f[DIR_MP0] = &DDStart[DIR_PM0 *size_Mat];
-// D.f[DIR_PM0] = &DDStart[DIR_MP0 *size_Mat];
-// D.f[DIR_M0M] = &DDStart[DIR_P0P *size_Mat];
-// D.f[DIR_P0P] = &DDStart[DIR_M0M *size_Mat];
-// D.f[DIR_M0P] = &DDStart[DIR_P0M *size_Mat];
-// D.f[DIR_P0M] = &DDStart[DIR_M0P *size_Mat];
-// D.f[DIR_0MM] = &DDStart[DIR_0PP *size_Mat];
-// D.f[DIR_0PP] = &DDStart[DIR_0MM *size_Mat];
-// D.f[DIR_0MP] = &DDStart[DIR_0PM *size_Mat];
-// D.f[DIR_0PM] = &DDStart[DIR_0MP *size_Mat];
-// D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
-// D.f[DIR_MMM] = &DDStart[DIR_PPP *size_Mat];
-// D.f[DIR_PPM] = &DDStart[DIR_MMP *size_Mat];
-// D.f[DIR_MPM] = &DDStart[DIR_PMP *size_Mat];
-// D.f[DIR_PMM] = &DDStart[DIR_MPP *size_Mat];
-// D.f[DIR_MMP] = &DDStart[DIR_PPM *size_Mat];
-// D.f[DIR_PPP] = &DDStart[DIR_MMM *size_Mat];
-// D.f[DIR_MPP] = &DDStart[DIR_PMM *size_Mat];
-// D.f[DIR_PMP] = &DDStart[DIR_MPM *size_Mat];
+// D.f[DIR_M00] = &DDStart[DIR_P00 * size_Mat];
+// D.f[DIR_P00] = &DDStart[DIR_M00 * size_Mat];
+// D.f[DIR_0M0] = &DDStart[DIR_0P0 * size_Mat];
+// D.f[DIR_0P0] = &DDStart[DIR_0M0 * size_Mat];
+// D.f[DIR_00M] = &DDStart[DIR_00P * size_Mat];
+// D.f[DIR_00P] = &DDStart[DIR_00M * size_Mat];
+// D.f[DIR_MM0] = &DDStart[DIR_PP0 * size_Mat];
+// D.f[DIR_PP0] = &DDStart[DIR_MM0 * size_Mat];
+// D.f[DIR_MP0] = &DDStart[DIR_PM0 * size_Mat];
+// D.f[DIR_PM0] = &DDStart[DIR_MP0 * size_Mat];
+// D.f[DIR_M0M] = &DDStart[DIR_P0P * size_Mat];
+// D.f[DIR_P0P] = &DDStart[DIR_M0M * size_Mat];
+// D.f[DIR_M0P] = &DDStart[DIR_P0M * size_Mat];
+// D.f[DIR_P0M] = &DDStart[DIR_M0P * size_Mat];
+// D.f[DIR_0MM] = &DDStart[DIR_0PP * size_Mat];
+// D.f[DIR_0PP] = &DDStart[DIR_0MM * size_Mat];
+// D.f[DIR_0MP] = &DDStart[DIR_0PM * size_Mat];
+// D.f[DIR_0PM] = &DDStart[DIR_0MP * size_Mat];
+// D.f[DIR_000] = &DDStart[DIR_000 * size_Mat];
+// D.f[DIR_MMM] = &DDStart[DIR_PPP * size_Mat];
+// D.f[DIR_PPM] = &DDStart[DIR_MMP * size_Mat];
+// D.f[DIR_MPM] = &DDStart[DIR_PMP * size_Mat];
+// D.f[DIR_PMM] = &DDStart[DIR_MPP * size_Mat];
+// D.f[DIR_MMP] = &DDStart[DIR_PPM * size_Mat];
+// D.f[DIR_PPP] = &DDStart[DIR_MMM * size_Mat];
+// D.f[DIR_MPP] = &DDStart[DIR_PMM * size_Mat];
+// D.f[DIR_PMP] = &DDStart[DIR_MPM * size_Mat];
// }
//
// Distributions6 G;
// if (EvenOrOdd == true)
// {
-// G.g[DIR_P00] = &G6[DIR_P00 *size_Mat];
-// G.g[DIR_M00] = &G6[DIR_M00 *size_Mat];
-// G.g[DIR_0P0] = &G6[DIR_0P0 *size_Mat];
-// G.g[DIR_0M0] = &G6[DIR_0M0 *size_Mat];
-// G.g[DIR_00P] = &G6[DIR_00P *size_Mat];
-// G.g[DIR_00M] = &G6[DIR_00M *size_Mat];
+// G.g[DIR_P00] = &G6[DIR_P00 * size_Mat];
+// G.g[DIR_M00] = &G6[DIR_M00 * size_Mat];
+// G.g[DIR_0P0] = &G6[DIR_0P0 * size_Mat];
+// G.g[DIR_0M0] = &G6[DIR_0M0 * size_Mat];
+// G.g[DIR_00P] = &G6[DIR_00P * size_Mat];
+// G.g[DIR_00M] = &G6[DIR_00M * size_Mat];
// }
// else
// {
-// G.g[DIR_M00] = &G6[DIR_P00 *size_Mat];
-// G.g[DIR_P00] = &G6[DIR_M00 *size_Mat];
-// G.g[DIR_0M0] = &G6[DIR_0P0 *size_Mat];
-// G.g[DIR_0P0] = &G6[DIR_0M0 *size_Mat];
-// G.g[DIR_00M] = &G6[DIR_00P *size_Mat];
-// G.g[DIR_00P] = &G6[DIR_00M *size_Mat];
+// G.g[DIR_M00] = &G6[DIR_P00 * size_Mat];
+// G.g[DIR_P00] = &G6[DIR_M00 * size_Mat];
+// G.g[DIR_0M0] = &G6[DIR_0P0 * size_Mat];
+// G.g[DIR_0P0] = &G6[DIR_0M0 * size_Mat];
+// G.g[DIR_00M] = &G6[DIR_00P * size_Mat];
+// G.g[DIR_00P] = &G6[DIR_00M * size_Mat];
// }
//
// ////////////////////////////////////////////////////////////////////////////////
@@ -2006,83 +2006,83 @@ __global__ void LB_PostProcessor_F3_2018_Fehlberg(real omega,
// Distributions27 D;
// if (EvenOrOdd == true)
// {
-// D.f[DIR_P00] = &DDStart[DIR_P00 *size_Mat];
-// D.f[DIR_M00] = &DDStart[DIR_M00 *size_Mat];
-// D.f[DIR_0P0] = &DDStart[DIR_0P0 *size_Mat];
-// D.f[DIR_0M0] = &DDStart[DIR_0M0 *size_Mat];
-// D.f[DIR_00P] = &DDStart[DIR_00P *size_Mat];
-// D.f[DIR_00M] = &DDStart[DIR_00M *size_Mat];
-// D.f[DIR_PP0] = &DDStart[DIR_PP0 *size_Mat];
-// D.f[DIR_MM0] = &DDStart[DIR_MM0 *size_Mat];
-// D.f[DIR_PM0] = &DDStart[DIR_PM0 *size_Mat];
-// D.f[DIR_MP0] = &DDStart[DIR_MP0 *size_Mat];
-// D.f[DIR_P0P] = &DDStart[DIR_P0P *size_Mat];
-// D.f[DIR_M0M] = &DDStart[DIR_M0M *size_Mat];
-// D.f[DIR_P0M] = &DDStart[DIR_P0M *size_Mat];
-// D.f[DIR_M0P] = &DDStart[DIR_M0P *size_Mat];
-// D.f[DIR_0PP] = &DDStart[DIR_0PP *size_Mat];
-// D.f[DIR_0MM] = &DDStart[DIR_0MM *size_Mat];
-// D.f[DIR_0PM] = &DDStart[DIR_0PM *size_Mat];
-// D.f[DIR_0MP] = &DDStart[DIR_0MP *size_Mat];
-// D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
-// D.f[DIR_PPP] = &DDStart[DIR_PPP *size_Mat];
-// D.f[DIR_MMP] = &DDStart[DIR_MMP *size_Mat];
-// D.f[DIR_PMP] = &DDStart[DIR_PMP *size_Mat];
-// D.f[DIR_MPP] = &DDStart[DIR_MPP *size_Mat];
-// D.f[DIR_PPM] = &DDStart[DIR_PPM *size_Mat];
-// D.f[DIR_MMM] = &DDStart[DIR_MMM *size_Mat];
-// D.f[DIR_PMM] = &DDStart[DIR_PMM *size_Mat];
-// D.f[DIR_MPM] = &DDStart[DIR_MPM *size_Mat];
+// D.f[DIR_P00] = &DDStart[DIR_P00 * size_Mat];
+// D.f[DIR_M00] = &DDStart[DIR_M00 * size_Mat];
+// D.f[DIR_0P0] = &DDStart[DIR_0P0 * size_Mat];
+// D.f[DIR_0M0] = &DDStart[DIR_0M0 * size_Mat];
+// D.f[DIR_00P] = &DDStart[DIR_00P * size_Mat];
+// D.f[DIR_00M] = &DDStart[DIR_00M * size_Mat];
+// D.f[DIR_PP0] = &DDStart[DIR_PP0 * size_Mat];
+// D.f[DIR_MM0] = &DDStart[DIR_MM0 * size_Mat];
+// D.f[DIR_PM0] = &DDStart[DIR_PM0 * size_Mat];
+// D.f[DIR_MP0] = &DDStart[DIR_MP0 * size_Mat];
+// D.f[DIR_P0P] = &DDStart[DIR_P0P * size_Mat];
+// D.f[DIR_M0M] = &DDStart[DIR_M0M * size_Mat];
+// D.f[DIR_P0M] = &DDStart[DIR_P0M * size_Mat];
+// D.f[DIR_M0P] = &DDStart[DIR_M0P * size_Mat];
+// D.f[DIR_0PP] = &DDStart[DIR_0PP * size_Mat];
+// D.f[DIR_0MM] = &DDStart[DIR_0MM * size_Mat];
+// D.f[DIR_0PM] = &DDStart[DIR_0PM * size_Mat];
+// D.f[DIR_0MP] = &DDStart[DIR_0MP * size_Mat];
+// D.f[DIR_000] = &DDStart[DIR_000 * size_Mat];
+// D.f[DIR_PPP] = &DDStart[DIR_PPP * size_Mat];
+// D.f[DIR_MMP] = &DDStart[DIR_MMP * size_Mat];
+// D.f[DIR_PMP] = &DDStart[DIR_PMP * size_Mat];
+// D.f[DIR_MPP] = &DDStart[DIR_MPP * size_Mat];
+// D.f[DIR_PPM] = &DDStart[DIR_PPM * size_Mat];
+// D.f[DIR_MMM] = &DDStart[DIR_MMM * size_Mat];
+// D.f[DIR_PMM] = &DDStart[DIR_PMM * size_Mat];
+// D.f[DIR_MPM] = &DDStart[DIR_MPM * size_Mat];
// }
// else
// {
-// D.f[DIR_M00] = &DDStart[DIR_P00 *size_Mat];
-// D.f[DIR_P00] = &DDStart[DIR_M00 *size_Mat];
-// D.f[DIR_0M0] = &DDStart[DIR_0P0 *size_Mat];
-// D.f[DIR_0P0] = &DDStart[DIR_0M0 *size_Mat];
-// D.f[DIR_00M] = &DDStart[DIR_00P *size_Mat];
-// D.f[DIR_00P] = &DDStart[DIR_00M *size_Mat];
-// D.f[DIR_MM0] = &DDStart[DIR_PP0 *size_Mat];
-// D.f[DIR_PP0] = &DDStart[DIR_MM0 *size_Mat];
-// D.f[DIR_MP0] = &DDStart[DIR_PM0 *size_Mat];
-// D.f[DIR_PM0] = &DDStart[DIR_MP0 *size_Mat];
-// D.f[DIR_M0M] = &DDStart[DIR_P0P *size_Mat];
-// D.f[DIR_P0P] = &DDStart[DIR_M0M *size_Mat];
-// D.f[DIR_M0P] = &DDStart[DIR_P0M *size_Mat];
-// D.f[DIR_P0M] = &DDStart[DIR_M0P *size_Mat];
-// D.f[DIR_0MM] = &DDStart[DIR_0PP *size_Mat];
-// D.f[DIR_0PP] = &DDStart[DIR_0MM *size_Mat];
-// D.f[DIR_0MP] = &DDStart[DIR_0PM *size_Mat];
-// D.f[DIR_0PM] = &DDStart[DIR_0MP *size_Mat];
-// D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
-// D.f[DIR_MMM] = &DDStart[DIR_PPP *size_Mat];
-// D.f[DIR_PPM] = &DDStart[DIR_MMP *size_Mat];
-// D.f[DIR_MPM] = &DDStart[DIR_PMP *size_Mat];
-// D.f[DIR_PMM] = &DDStart[DIR_MPP *size_Mat];
-// D.f[DIR_MMP] = &DDStart[DIR_PPM *size_Mat];
-// D.f[DIR_PPP] = &DDStart[DIR_MMM *size_Mat];
-// D.f[DIR_MPP] = &DDStart[DIR_PMM *size_Mat];
-// D.f[DIR_PMP] = &DDStart[DIR_MPM *size_Mat];
+// D.f[DIR_M00] = &DDStart[DIR_P00 * size_Mat];
+// D.f[DIR_P00] = &DDStart[DIR_M00 * size_Mat];
+// D.f[DIR_0M0] = &DDStart[DIR_0P0 * size_Mat];
+// D.f[DIR_0P0] = &DDStart[DIR_0M0 * size_Mat];
+// D.f[DIR_00M] = &DDStart[DIR_00P * size_Mat];
+// D.f[DIR_00P] = &DDStart[DIR_00M * size_Mat];
+// D.f[DIR_MM0] = &DDStart[DIR_PP0 * size_Mat];
+// D.f[DIR_PP0] = &DDStart[DIR_MM0 * size_Mat];
+// D.f[DIR_MP0] = &DDStart[DIR_PM0 * size_Mat];
+// D.f[DIR_PM0] = &DDStart[DIR_MP0 * size_Mat];
+// D.f[DIR_M0M] = &DDStart[DIR_P0P * size_Mat];
+// D.f[DIR_P0P] = &DDStart[DIR_M0M * size_Mat];
+// D.f[DIR_M0P] = &DDStart[DIR_P0M * size_Mat];
+// D.f[DIR_P0M] = &DDStart[DIR_M0P * size_Mat];
+// D.f[DIR_0MM] = &DDStart[DIR_0PP * size_Mat];
+// D.f[DIR_0PP] = &DDStart[DIR_0MM * size_Mat];
+// D.f[DIR_0MP] = &DDStart[DIR_0PM * size_Mat];
+// D.f[DIR_0PM] = &DDStart[DIR_0MP * size_Mat];
+// D.f[DIR_000] = &DDStart[DIR_000 * size_Mat];
+// D.f[DIR_MMM] = &DDStart[DIR_PPP * size_Mat];
+// D.f[DIR_PPM] = &DDStart[DIR_MMP * size_Mat];
+// D.f[DIR_MPM] = &DDStart[DIR_PMP * size_Mat];
+// D.f[DIR_PMM] = &DDStart[DIR_MPP * size_Mat];
+// D.f[DIR_MMP] = &DDStart[DIR_PPM * size_Mat];
+// D.f[DIR_PPP] = &DDStart[DIR_MMM * size_Mat];
+// D.f[DIR_MPP] = &DDStart[DIR_PMM * size_Mat];
+// D.f[DIR_PMP] = &DDStart[DIR_MPM * size_Mat];
// }
//
// Distributions6 G;
// if (EvenOrOdd == true)
// {
-// G.g[DIR_P00] = &G6[DIR_P00 *size_Mat];
-// G.g[DIR_M00] = &G6[DIR_M00 *size_Mat];
-// G.g[DIR_0P0] = &G6[DIR_0P0 *size_Mat];
-// G.g[DIR_0M0] = &G6[DIR_0M0 *size_Mat];
-// G.g[DIR_00P] = &G6[DIR_00P *size_Mat];
-// G.g[DIR_00M] = &G6[DIR_00M *size_Mat];
+// G.g[DIR_P00] = &G6[DIR_P00 * size_Mat];
+// G.g[DIR_M00] = &G6[DIR_M00 * size_Mat];
+// G.g[DIR_0P0] = &G6[DIR_0P0 * size_Mat];
+// G.g[DIR_0M0] = &G6[DIR_0M0 * size_Mat];
+// G.g[DIR_00P] = &G6[DIR_00P * size_Mat];
+// G.g[DIR_00M] = &G6[DIR_00M * size_Mat];
// }
// else
// {
-// G.g[DIR_M00] = &G6[DIR_P00 *size_Mat];
-// G.g[DIR_P00] = &G6[DIR_M00 *size_Mat];
-// G.g[DIR_0M0] = &G6[DIR_0P0 *size_Mat];
-// G.g[DIR_0P0] = &G6[DIR_0M0 *size_Mat];
-// G.g[DIR_00M] = &G6[DIR_00P *size_Mat];
-// G.g[DIR_00P] = &G6[DIR_00M *size_Mat];
+// G.g[DIR_M00] = &G6[DIR_P00 * size_Mat];
+// G.g[DIR_P00] = &G6[DIR_M00 * size_Mat];
+// G.g[DIR_0M0] = &G6[DIR_0P0 * size_Mat];
+// G.g[DIR_0P0] = &G6[DIR_0M0 * size_Mat];
+// G.g[DIR_00M] = &G6[DIR_00P * size_Mat];
+// G.g[DIR_00P] = &G6[DIR_00M * size_Mat];
// }
//
// ////////////////////////////////////////////////////////////////////////////////
@@ -2153,33 +2153,33 @@ __global__ void LB_PostProcessor_F3_2018_Fehlberg(real omega,
// real dyyuy = c1o2 * (-mgbcb + mgbab);
// real dzzuz = c1o2 * (-mgbbc + mgbba);
// //////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
-// real mfcbb = (D.f[DIR_P00])[k];//[ke ];// + c2over27 ;(D.f[DIR_P00 ])[k ];//ke
-// real mfabb = (D.f[DIR_M00])[kw];//[kw ];// + c2over27 ;(D.f[DIR_M00 ])[kw ];
-// real mfbcb = (D.f[DIR_0P0])[k];//[kn ];// + c2over27 ;(D.f[DIR_0P0 ])[k ];//kn
-// real mfbab = (D.f[DIR_0M0])[ks];//[ks ];// + c2over27 ;(D.f[DIR_0M0 ])[ks ];
-// real mfbbc = (D.f[DIR_00P])[k];//[kt ];// + c2over27 ;(D.f[DIR_00P ])[k ];//kt
-// real mfbba = (D.f[DIR_00M])[kb];//[kb ];// + c2over27 ;(D.f[DIR_00M ])[kb ];
-// real mfccb = (D.f[DIR_PP0])[k];//[kne ];// + c1over54 ;(D.f[DIR_PP0 ])[k ];//kne
-// real mfaab = (D.f[DIR_MM0])[ksw];//[ksw ];// + c1over54 ;(D.f[DIR_MM0 ])[ksw];
-// real mfcab = (D.f[DIR_PM0])[ks];//[kse ];// + c1over54 ;(D.f[DIR_PM0 ])[ks ];//kse
-// real mfacb = (D.f[DIR_MP0])[kw];//[knw ];// + c1over54 ;(D.f[DIR_MP0 ])[kw ];//knw
-// real mfcbc = (D.f[DIR_P0P])[k];//[kte ];// + c1over54 ;(D.f[DIR_P0P ])[k ];//kte
-// real mfaba = (D.f[DIR_M0M])[kbw];//[kbw ];// + c1over54 ;(D.f[DIR_M0M ])[kbw];
-// real mfcba = (D.f[DIR_P0M])[kb];//[kbe ];// + c1over54 ;(D.f[DIR_P0M ])[kb ];//kbe
-// real mfabc = (D.f[DIR_M0P])[kw];//[ktw ];// + c1over54 ;(D.f[DIR_M0P ])[kw ];//ktw
-// real mfbcc = (D.f[DIR_0PP])[k];//[ktn ];// + c1over54 ;(D.f[DIR_0PP ])[k ];//ktn
-// real mfbaa = (D.f[DIR_0MM])[kbs];//[kbs ];// + c1over54 ;(D.f[DIR_0MM ])[kbs];
-// real mfbca = (D.f[DIR_0PM])[kb];//[kbn ];// + c1over54 ;(D.f[DIR_0PM ])[kb ];//kbn
-// real mfbac = (D.f[DIR_0MP])[ks];//[kts ];// + c1over54 ;(D.f[DIR_0MP ])[ks ];//kts
+// real mfcbb = (D.f[DIR_P00])[k];//[ke ];// + c2over27 ;(D.f[DIR_P00])[k ];//ke
+// real mfabb = (D.f[DIR_M00])[kw];//[kw ];// + c2over27 ;(D.f[DIR_M00])[kw ];
+// real mfbcb = (D.f[DIR_0P0])[k];//[kn ];// + c2over27 ;(D.f[DIR_0P0])[k ];//kn
+// real mfbab = (D.f[DIR_0M0])[ks];//[ks ];// + c2over27 ;(D.f[DIR_0M0])[ks ];
+// real mfbbc = (D.f[DIR_00P])[k];//[kt ];// + c2over27 ;(D.f[DIR_00P])[k ];//kt
+// real mfbba = (D.f[DIR_00M])[kb];//[kb ];// + c2over27 ;(D.f[DIR_00M])[kb ];
+// real mfccb = (D.f[DIR_PP0])[k];//[kne ];// + c1over54 ;(D.f[DIR_PP0])[k ];//kne
+// real mfaab = (D.f[DIR_MM0])[ksw];//[ksw ];// + c1over54 ;(D.f[DIR_MM0])[ksw];
+// real mfcab = (D.f[DIR_PM0])[ks];//[kse ];// + c1over54 ;(D.f[DIR_PM0])[ks ];//kse
+// real mfacb = (D.f[DIR_MP0])[kw];//[knw ];// + c1over54 ;(D.f[DIR_MP0])[kw ];//knw
+// real mfcbc = (D.f[DIR_P0P])[k];//[kte ];// + c1over54 ;(D.f[DIR_P0P])[k ];//kte
+// real mfaba = (D.f[DIR_M0M])[kbw];//[kbw ];// + c1over54 ;(D.f[DIR_M0M])[kbw];
+// real mfcba = (D.f[DIR_P0M])[kb];//[kbe ];// + c1over54 ;(D.f[DIR_P0M])[kb ];//kbe
+// real mfabc = (D.f[DIR_M0P])[kw];//[ktw ];// + c1over54 ;(D.f[DIR_M0P])[kw ];//ktw
+// real mfbcc = (D.f[DIR_0PP])[k];//[ktn ];// + c1over54 ;(D.f[DIR_0PP])[k ];//ktn
+// real mfbaa = (D.f[DIR_0MM])[kbs];//[kbs ];// + c1over54 ;(D.f[DIR_0MM])[kbs];
+// real mfbca = (D.f[DIR_0PM])[kb];//[kbn ];// + c1over54 ;(D.f[DIR_0PM])[kb ];//kbn
+// real mfbac = (D.f[DIR_0MP])[ks];//[kts ];// + c1over54 ;(D.f[DIR_0MP])[ks ];//kts
// real mfbbb = (D.f[DIR_000])[k];//[kzero];// + c8over27 ;(D.f[DIR_000])[k ];//kzero
-// real mfccc = (D.f[DIR_PPP])[k];//[ktne ];// + c1over216;(D.f[DIR_PPP ])[k ];//ktne
-// real mfaac = (D.f[DIR_MMP])[ksw];//[ktsw ];// + c1over216;(D.f[DIR_MMP ])[ksw];//ktsw
-// real mfcac = (D.f[DIR_PMP])[ks];//[ktse ];// + c1over216;(D.f[DIR_PMP ])[ks ];//ktse
-// real mfacc = (D.f[DIR_MPP])[kw];//[ktnw ];// + c1over216;(D.f[DIR_MPP ])[kw ];//ktnw
-// real mfcca = (D.f[DIR_PPM])[kb];//[kbne ];// + c1over216;(D.f[DIR_PPM ])[kb ];//kbne
-// real mfaaa = (D.f[DIR_MMM])[kbsw];//[kbsw ];// + c1over216;(D.f[DIR_MMM ])[kbsw];
-// real mfcaa = (D.f[DIR_PMM])[kbs];//[kbse ];// + c1over216;(D.f[DIR_PMM ])[kbs];//kbse
-// real mfaca = (D.f[DIR_MPM])[kbw];//[kbnw ];// + c1over216;(D.f[DIR_MPM ])[kbw];//kbnw
+// real mfccc = (D.f[DIR_PPP])[k];//[ktne ];// + c1over216;(D.f[DIR_PPP])[k ];//ktne
+// real mfaac = (D.f[DIR_MMP])[ksw];//[ktsw ];// + c1over216;(D.f[DIR_MMP])[ksw];//ktsw
+// real mfcac = (D.f[DIR_PMP])[ks];//[ktse ];// + c1over216;(D.f[DIR_PMP])[ks ];//ktse
+// real mfacc = (D.f[DIR_MPP])[kw];//[ktnw ];// + c1over216;(D.f[DIR_MPP])[kw ];//ktnw
+// real mfcca = (D.f[DIR_PPM])[kb];//[kbne ];// + c1over216;(D.f[DIR_PPM])[kb ];//kbne
+// real mfaaa = (D.f[DIR_MMM])[kbsw];//[kbsw ];// + c1over216;(D.f[DIR_MMM])[kbsw];
+// real mfcaa = (D.f[DIR_PMM])[kbs];//[kbse ];// + c1over216;(D.f[DIR_PMM])[kbs];//kbse
+// real mfaca = (D.f[DIR_MPM])[kbw];//[kbnw ];// + c1over216;(D.f[DIR_MPM])[kbw];//kbnw
// ////////////////////////////////////////////////////////////////////////////////////
// real drho = ((((mfccc + mfaaa) + (mfaca + mfcac)) + ((mfacc + mfcaa) + (mfaac + mfcca))) +
// (((mfbac + mfbca) + (mfbaa + mfbcc)) + ((mfabc + mfcba) + (mfaba + mfcbc)) + ((mfacb + mfcab) + (mfaab + mfccb))) +
diff --git a/src/gpu/VirtualFluids_GPU/GPU/DragLift27.cu b/src/gpu/VirtualFluids_GPU/GPU/DragLift27.cu
index 5146242fed374a919b6dcc02774db1d8ce4f864a..0e3945829725c0614ed4da01d0bae3b99ba2720a 100644
--- a/src/gpu/VirtualFluids_GPU/GPU/DragLift27.cu
+++ b/src/gpu/VirtualFluids_GPU/GPU/DragLift27.cu
@@ -17,69 +17,69 @@ __global__ void DragLiftPost27( real* DD,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep)
{
Distributions27 D;
if (isEvenTimestep==true)
{
- D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
+ D.f[DIR_M00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_MPM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_PMP * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////
const unsigned x = threadIdx.x; // Globaler x-Index
@@ -100,24 +100,24 @@ __global__ void DragLiftPost27( real* DD,
*q_dirBE, *q_dirTW, *q_dirTN, *q_dirBS, *q_dirBN, *q_dirTS,
*q_dirTNE, *q_dirTSW, *q_dirTSE, *q_dirTNW, *q_dirBNE, *q_dirBSW,
*q_dirBSE, *q_dirBNW;
- q_dirE = &QQ[DIR_P00 * numberOfBCnodes];
- q_dirW = &QQ[DIR_M00 * numberOfBCnodes];
- q_dirN = &QQ[DIR_0P0 * numberOfBCnodes];
- q_dirS = &QQ[DIR_0M0 * numberOfBCnodes];
- q_dirT = &QQ[DIR_00P * numberOfBCnodes];
- q_dirB = &QQ[DIR_00M * numberOfBCnodes];
- q_dirNE = &QQ[DIR_PP0 * numberOfBCnodes];
- q_dirSW = &QQ[DIR_MM0 * numberOfBCnodes];
- q_dirSE = &QQ[DIR_PM0 * numberOfBCnodes];
- q_dirNW = &QQ[DIR_MP0 * numberOfBCnodes];
- q_dirTE = &QQ[DIR_P0P * numberOfBCnodes];
- q_dirBW = &QQ[DIR_M0M * numberOfBCnodes];
- q_dirBE = &QQ[DIR_P0M * numberOfBCnodes];
- q_dirTW = &QQ[DIR_M0P * numberOfBCnodes];
- q_dirTN = &QQ[DIR_0PP * numberOfBCnodes];
- q_dirBS = &QQ[DIR_0MM * numberOfBCnodes];
- q_dirBN = &QQ[DIR_0PM * numberOfBCnodes];
- q_dirTS = &QQ[DIR_0MP * numberOfBCnodes];
+ q_dirE = &QQ[DIR_P00 * numberOfBCnodes];
+ q_dirW = &QQ[DIR_M00 * numberOfBCnodes];
+ q_dirN = &QQ[DIR_0P0 * numberOfBCnodes];
+ q_dirS = &QQ[DIR_0M0 * numberOfBCnodes];
+ q_dirT = &QQ[DIR_00P * numberOfBCnodes];
+ q_dirB = &QQ[DIR_00M * numberOfBCnodes];
+ q_dirNE = &QQ[DIR_PP0 * numberOfBCnodes];
+ q_dirSW = &QQ[DIR_MM0 * numberOfBCnodes];
+ q_dirSE = &QQ[DIR_PM0 * numberOfBCnodes];
+ q_dirNW = &QQ[DIR_MP0 * numberOfBCnodes];
+ q_dirTE = &QQ[DIR_P0P * numberOfBCnodes];
+ q_dirBW = &QQ[DIR_M0M * numberOfBCnodes];
+ q_dirBE = &QQ[DIR_P0M * numberOfBCnodes];
+ q_dirTW = &QQ[DIR_M0P * numberOfBCnodes];
+ q_dirTN = &QQ[DIR_0PP * numberOfBCnodes];
+ q_dirBS = &QQ[DIR_0MM * numberOfBCnodes];
+ q_dirBN = &QQ[DIR_0PM * numberOfBCnodes];
+ q_dirTS = &QQ[DIR_0MP * numberOfBCnodes];
q_dirTNE = &QQ[DIR_PPP * numberOfBCnodes];
q_dirTSW = &QQ[DIR_MMP * numberOfBCnodes];
q_dirTSE = &QQ[DIR_PMP * numberOfBCnodes];
@@ -160,32 +160,32 @@ __global__ void DragLiftPost27( real* DD,
real f_E, f_W, f_N, f_S, f_T, f_B, f_NE, f_SW, f_SE, f_NW, f_TE, f_BW, f_BE,
f_TW, f_TN, f_BS, f_BN, f_TS, f_TNE, f_TSW, f_TSE, f_TNW, f_BNE, f_BSW, f_BSE, f_BNW;
- f_W = (D.f[DIR_P00 ])[ke ];
- f_E = (D.f[DIR_M00 ])[kw ];
- f_S = (D.f[DIR_0P0 ])[kn ];
- f_N = (D.f[DIR_0M0 ])[ks ];
- f_B = (D.f[DIR_00P ])[kt ];
- f_T = (D.f[DIR_00M ])[kb ];
- f_SW = (D.f[DIR_PP0 ])[kne ];
- f_NE = (D.f[DIR_MM0 ])[ksw ];
- f_NW = (D.f[DIR_PM0 ])[kse ];
- f_SE = (D.f[DIR_MP0 ])[knw ];
- f_BW = (D.f[DIR_P0P ])[kte ];
- f_TE = (D.f[DIR_M0M ])[kbw ];
- f_TW = (D.f[DIR_P0M ])[kbe ];
- f_BE = (D.f[DIR_M0P ])[ktw ];
- f_BS = (D.f[DIR_0PP ])[ktn ];
- f_TN = (D.f[DIR_0MM ])[kbs ];
- f_TS = (D.f[DIR_0PM ])[kbn ];
- f_BN = (D.f[DIR_0MP ])[kts ];
- f_BSW = (D.f[DIR_PPP ])[ktne ];
- f_BNE = (D.f[DIR_MMP ])[ktsw ];
- f_BNW = (D.f[DIR_PMP ])[ktse ];
- f_BSE = (D.f[DIR_MPP ])[ktnw ];
- f_TSW = (D.f[DIR_PPM ])[kbne ];
- f_TNE = (D.f[DIR_MMM ])[kbsw ];
- f_TNW = (D.f[DIR_PMM ])[kbse ];
- f_TSE = (D.f[DIR_MPM ])[kbnw ];
+ f_W = (D.f[DIR_P00])[ke ];
+ f_E = (D.f[DIR_M00])[kw ];
+ f_S = (D.f[DIR_0P0])[kn ];
+ f_N = (D.f[DIR_0M0])[ks ];
+ f_B = (D.f[DIR_00P])[kt ];
+ f_T = (D.f[DIR_00M])[kb ];
+ f_SW = (D.f[DIR_PP0])[kne ];
+ f_NE = (D.f[DIR_MM0])[ksw ];
+ f_NW = (D.f[DIR_PM0])[kse ];
+ f_SE = (D.f[DIR_MP0])[knw ];
+ f_BW = (D.f[DIR_P0P])[kte ];
+ f_TE = (D.f[DIR_M0M])[kbw ];
+ f_TW = (D.f[DIR_P0M])[kbe ];
+ f_BE = (D.f[DIR_M0P])[ktw ];
+ f_BS = (D.f[DIR_0PP])[ktn ];
+ f_TN = (D.f[DIR_0MM])[kbs ];
+ f_TS = (D.f[DIR_0PM])[kbn ];
+ f_BN = (D.f[DIR_0MP])[kts ];
+ f_BSW = (D.f[DIR_PPP])[ktne ];
+ f_BNE = (D.f[DIR_MMP])[ktsw ];
+ f_BNW = (D.f[DIR_PMP])[ktse ];
+ f_BSE = (D.f[DIR_MPP])[ktnw ];
+ f_TSW = (D.f[DIR_PPM])[kbne ];
+ f_TNE = (D.f[DIR_MMM])[kbsw ];
+ f_TNW = (D.f[DIR_PMM])[kbse ];
+ f_TSE = (D.f[DIR_MPM])[kbnw ];
////////////////////////////////////////////////////////////////////////////////
double OnE = c0o1, OnW = c0o1, OnN = c0o1, OnS = c0o1, OnT = c0o1, OnB = c0o1,
OnNE = c0o1, OnSW = c0o1, OnSE = c0o1, OnNW = c0o1,
@@ -282,69 +282,69 @@ __global__ void DragLiftPre27( real* DD,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep)
{
Distributions27 D;
if (isEvenTimestep==true)
{
- D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
+ D.f[DIR_M00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_MPM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_PMP * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////
const unsigned x = threadIdx.x; // Globaler x-Index
@@ -365,24 +365,24 @@ __global__ void DragLiftPre27( real* DD,
*q_dirBE, *q_dirTW, *q_dirTN, *q_dirBS, *q_dirBN, *q_dirTS,
*q_dirTNE, *q_dirTSW, *q_dirTSE, *q_dirTNW, *q_dirBNE, *q_dirBSW,
*q_dirBSE, *q_dirBNW;
- q_dirE = &QQ[DIR_P00 * numberOfBCnodes];
- q_dirW = &QQ[DIR_M00 * numberOfBCnodes];
- q_dirN = &QQ[DIR_0P0 * numberOfBCnodes];
- q_dirS = &QQ[DIR_0M0 * numberOfBCnodes];
- q_dirT = &QQ[DIR_00P * numberOfBCnodes];
- q_dirB = &QQ[DIR_00M * numberOfBCnodes];
- q_dirNE = &QQ[DIR_PP0 * numberOfBCnodes];
- q_dirSW = &QQ[DIR_MM0 * numberOfBCnodes];
- q_dirSE = &QQ[DIR_PM0 * numberOfBCnodes];
- q_dirNW = &QQ[DIR_MP0 * numberOfBCnodes];
- q_dirTE = &QQ[DIR_P0P * numberOfBCnodes];
- q_dirBW = &QQ[DIR_M0M * numberOfBCnodes];
- q_dirBE = &QQ[DIR_P0M * numberOfBCnodes];
- q_dirTW = &QQ[DIR_M0P * numberOfBCnodes];
- q_dirTN = &QQ[DIR_0PP * numberOfBCnodes];
- q_dirBS = &QQ[DIR_0MM * numberOfBCnodes];
- q_dirBN = &QQ[DIR_0PM * numberOfBCnodes];
- q_dirTS = &QQ[DIR_0MP * numberOfBCnodes];
+ q_dirE = &QQ[DIR_P00 * numberOfBCnodes];
+ q_dirW = &QQ[DIR_M00 * numberOfBCnodes];
+ q_dirN = &QQ[DIR_0P0 * numberOfBCnodes];
+ q_dirS = &QQ[DIR_0M0 * numberOfBCnodes];
+ q_dirT = &QQ[DIR_00P * numberOfBCnodes];
+ q_dirB = &QQ[DIR_00M * numberOfBCnodes];
+ q_dirNE = &QQ[DIR_PP0 * numberOfBCnodes];
+ q_dirSW = &QQ[DIR_MM0 * numberOfBCnodes];
+ q_dirSE = &QQ[DIR_PM0 * numberOfBCnodes];
+ q_dirNW = &QQ[DIR_MP0 * numberOfBCnodes];
+ q_dirTE = &QQ[DIR_P0P * numberOfBCnodes];
+ q_dirBW = &QQ[DIR_M0M * numberOfBCnodes];
+ q_dirBE = &QQ[DIR_P0M * numberOfBCnodes];
+ q_dirTW = &QQ[DIR_M0P * numberOfBCnodes];
+ q_dirTN = &QQ[DIR_0PP * numberOfBCnodes];
+ q_dirBS = &QQ[DIR_0MM * numberOfBCnodes];
+ q_dirBN = &QQ[DIR_0PM * numberOfBCnodes];
+ q_dirTS = &QQ[DIR_0MP * numberOfBCnodes];
q_dirTNE = &QQ[DIR_PPP * numberOfBCnodes];
q_dirTSW = &QQ[DIR_MMP * numberOfBCnodes];
q_dirTSE = &QQ[DIR_PMP * numberOfBCnodes];
@@ -425,32 +425,32 @@ __global__ void DragLiftPre27( real* DD,
real f_E, f_W, f_N, f_S, f_T, f_B, f_NE, f_SW, f_SE, f_NW, f_TE, f_BW, f_BE,
f_TW, f_TN, f_BS, f_BN, f_TS, f_TNE, f_TSW, f_TSE, f_TNW, f_BNE, f_BSW, f_BSE, f_BNW;
- f_E = (D.f[DIR_P00 ])[ke ];
- f_W = (D.f[DIR_M00 ])[kw ];
- f_N = (D.f[DIR_0P0 ])[kn ];
- f_S = (D.f[DIR_0M0 ])[ks ];
- f_T = (D.f[DIR_00P ])[kt ];
- f_B = (D.f[DIR_00M ])[kb ];
- f_NE = (D.f[DIR_PP0 ])[kne ];
- f_SW = (D.f[DIR_MM0 ])[ksw ];
- f_SE = (D.f[DIR_PM0 ])[kse ];
- f_NW = (D.f[DIR_MP0 ])[knw ];
- f_TE = (D.f[DIR_P0P ])[kte ];
- f_BW = (D.f[DIR_M0M ])[kbw ];
- f_BE = (D.f[DIR_P0M ])[kbe ];
- f_TW = (D.f[DIR_M0P ])[ktw ];
- f_TN = (D.f[DIR_0PP ])[ktn ];
- f_BS = (D.f[DIR_0MM ])[kbs ];
- f_BN = (D.f[DIR_0PM ])[kbn ];
- f_TS = (D.f[DIR_0MP ])[kts ];
- f_TNE = (D.f[DIR_PPP ])[ktne ];
- f_TSW = (D.f[DIR_MMP ])[ktsw ];
- f_TSE = (D.f[DIR_PMP ])[ktse ];
- f_TNW = (D.f[DIR_MPP ])[ktnw ];
- f_BNE = (D.f[DIR_PPM ])[kbne ];
- f_BSW = (D.f[DIR_MMM ])[kbsw ];
- f_BSE = (D.f[DIR_PMM ])[kbse ];
- f_BNW = (D.f[DIR_MPM ])[kbnw ];
+ f_E = (D.f[DIR_P00])[ke ];
+ f_W = (D.f[DIR_M00])[kw ];
+ f_N = (D.f[DIR_0P0])[kn ];
+ f_S = (D.f[DIR_0M0])[ks ];
+ f_T = (D.f[DIR_00P])[kt ];
+ f_B = (D.f[DIR_00M])[kb ];
+ f_NE = (D.f[DIR_PP0])[kne ];
+ f_SW = (D.f[DIR_MM0])[ksw ];
+ f_SE = (D.f[DIR_PM0])[kse ];
+ f_NW = (D.f[DIR_MP0])[knw ];
+ f_TE = (D.f[DIR_P0P])[kte ];
+ f_BW = (D.f[DIR_M0M])[kbw ];
+ f_BE = (D.f[DIR_P0M])[kbe ];
+ f_TW = (D.f[DIR_M0P])[ktw ];
+ f_TN = (D.f[DIR_0PP])[ktn ];
+ f_BS = (D.f[DIR_0MM])[kbs ];
+ f_BN = (D.f[DIR_0PM])[kbn ];
+ f_TS = (D.f[DIR_0MP])[kts ];
+ f_TNE = (D.f[DIR_PPP])[ktne ];
+ f_TSW = (D.f[DIR_MMP])[ktsw ];
+ f_TSE = (D.f[DIR_PMP])[ktse ];
+ f_TNW = (D.f[DIR_MPP])[ktnw ];
+ f_BNE = (D.f[DIR_PPM])[kbne ];
+ f_BSW = (D.f[DIR_MMM])[kbsw ];
+ f_BSE = (D.f[DIR_PMM])[kbse ];
+ f_BNW = (D.f[DIR_MPM])[kbnw ];
////////////////////////////////////////////////////////////////////////////////
double OnE = c0o1, OnW = c0o1, OnN = c0o1, OnS = c0o1, OnT = c0o1, OnB = c0o1,
OnNE = c0o1, OnSW = c0o1, OnSE = c0o1, OnNW = c0o1,
diff --git a/src/gpu/VirtualFluids_GPU/GPU/EnstrophyAnalyzer.cu b/src/gpu/VirtualFluids_GPU/GPU/EnstrophyAnalyzer.cu
index acd62b46c5666fc5f621c3772438e42b7ebef5c6..93879d73a32458d5403fd3fd16e68e0fcea7753d 100644
--- a/src/gpu/VirtualFluids_GPU/GPU/EnstrophyAnalyzer.cu
+++ b/src/gpu/VirtualFluids_GPU/GPU/EnstrophyAnalyzer.cu
@@ -13,7 +13,7 @@
#include <iomanip>
-//#include "Core/Logger/Logger.h"
+#include "cuda/CudaGrid.h"
#include "Parameter/Parameter.h"
// includes, kernels
@@ -22,7 +22,7 @@
using namespace vf::lbm::constant;
-__global__ void enstrophyKernel ( real* veloX, real* veloY, real* veloZ, real* rho, uint* neighborX, uint* neighborY, uint* neighborZ, uint* neighborWSB, uint* geo, real* enstrophy, uint* isFluid, uint size_Mat );
+__global__ void enstrophyKernel ( real* veloX, real* veloY, real* veloZ, real* rho, uint* neighborX, uint* neighborY, uint* neighborZ, uint* neighborWSB, uint* geo, real* enstrophy, uint* isFluid, unsigned long long numberOfLBnodes );
__host__ __device__ inline void enstrophyFunction( real* veloX, real* veloY, real* veloZ, real* rho, uint* neighborX, uint* neighborY, uint* neighborZ, uint* neighborWSB, uint* geo, real* enstrophy, uint* isFluid, uint index );
@@ -32,55 +32,40 @@ bool EnstrophyAnalyzer::run(uint iter)
{
if( iter % this->analyzeIter != 0 ) return false;
- int lev = 0;
- int size_Mat = this->para->getParD(lev)->numberOfNodes;
-
- thrust::device_vector<real> enstrophy( size_Mat, c0o1 );
- thrust::device_vector<uint> isFluid ( size_Mat, 0);
-
- unsigned int numberOfThreads = 128;
- int Grid = (size_Mat / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 grid(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- LBCalcMacCompSP27<<< grid, threads >>> (para->getParD(lev)->velocityX,
- para->getParD(lev)->velocityY,
- para->getParD(lev)->velocityZ,
- para->getParD(lev)->rho,
- para->getParD(lev)->pressure,
- para->getParD(lev)->typeOfGridNode,
- para->getParD(lev)->neighborX,
- para->getParD(lev)->neighborY,
- para->getParD(lev)->neighborZ,
- para->getParD(lev)->numberOfNodes,
- para->getParD(lev)->distributions.f[0],
- para->getParD(lev)->isEvenTimestep);
- //cudaDeviceSynchronize();
- getLastCudaError("LBCalcMacSP27 execution failed");
-
- enstrophyKernel <<< grid, threads >>> ( para->getParD(lev)->velocityX,
- para->getParD(lev)->velocityY,
- para->getParD(lev)->velocityZ,
- para->getParD(lev)->rho,
- para->getParD(lev)->neighborX,
- para->getParD(lev)->neighborY,
- para->getParD(lev)->neighborZ,
- para->getParD(lev)->neighborInverse,
- para->getParD(lev)->typeOfGridNode,
- enstrophy.data().get(),
- isFluid.data().get(),
- size_Mat);
+ int lev = 0;
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(para->getParD(lev)->numberofthreads, para->getParD(lev)->numberOfNodes);
+
+ thrust::device_vector<real> enstrophy( this->para->getParD(lev)->numberOfNodes, c0o1);
+ thrust::device_vector<uint> isFluid ( this->para->getParD(lev)->numberOfNodes, 0);
+
+ LBCalcMacCompSP27<<< grid.grid, grid.threads >>>(
+ para->getParD(lev)->velocityX,
+ para->getParD(lev)->velocityY,
+ para->getParD(lev)->velocityZ,
+ para->getParD(lev)->rho,
+ para->getParD(lev)->pressure,
+ para->getParD(lev)->typeOfGridNode,
+ para->getParD(lev)->neighborX,
+ para->getParD(lev)->neighborY,
+ para->getParD(lev)->neighborZ,
+ para->getParD(lev)->numberOfNodes,
+ para->getParD(lev)->distributions.f[0],
+ para->getParD(lev)->isEvenTimestep);
+ getLastCudaError("LBCalcMacCompSP27 execution failed");
+
+ enstrophyKernel<<< grid.grid, grid.threads >>>(
+ para->getParD(lev)->velocityX,
+ para->getParD(lev)->velocityY,
+ para->getParD(lev)->velocityZ,
+ para->getParD(lev)->rho,
+ para->getParD(lev)->neighborX,
+ para->getParD(lev)->neighborY,
+ para->getParD(lev)->neighborZ,
+ para->getParD(lev)->neighborInverse,
+ para->getParD(lev)->typeOfGridNode,
+ enstrophy.data().get(),
+ isFluid.data().get(),
+ para->getParD(lev)->numberOfNodes);
cudaDeviceSynchronize();
getLastCudaError("enstrophyKernel execution failed");
@@ -97,7 +82,7 @@ bool EnstrophyAnalyzer::run(uint iter)
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
-__global__ void enstrophyKernel(real* veloX, real* veloY, real* veloZ, real* rho, uint* neighborX, uint* neighborY, uint* neighborZ, uint* neighborWSB, uint* geo, real* enstrophy, uint* isFluid, uint size_Mat)
+__global__ void enstrophyKernel(real* veloX, real* veloY, real* veloZ, real* rho, uint* neighborX, uint* neighborY, uint* neighborZ, uint* neighborWSB, uint* geo, real* enstrophy, uint* isFluid, unsigned long long numberOfLBnodes)
{
//////////////////////////////////////////////////////////////////////////
const uint x = threadIdx.x; // Globaler x-Index
@@ -113,7 +98,7 @@ __global__ void enstrophyKernel(real* veloX, real* veloY, real* veloZ, real* rho
//if( index % 34 == 0 || index % 34 == 33 ) return;
- if( index >= size_Mat) return;
+ if( index >= (uint)numberOfLBnodes) return;
unsigned int BC;
BC = geo[index];
diff --git a/src/gpu/VirtualFluids_GPU/GPU/ExchangeData27.cu b/src/gpu/VirtualFluids_GPU/GPU/ExchangeData27.cu
index 5470da46342c85e57370227313c8c82674a17e6e..4ced64c0152bdbbd9752f736e2edca2c51fbc2ff 100644
--- a/src/gpu/VirtualFluids_GPU/GPU/ExchangeData27.cu
+++ b/src/gpu/VirtualFluids_GPU/GPU/ExchangeData27.cu
@@ -14,7 +14,7 @@ __global__ void getSendFsPost27(real* DD,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep)
{
////////////////////////////////////////////////////////////////////////////////
@@ -65,150 +65,150 @@ __global__ void getSendFsPost27(real* DD,
Distributions27 D;
if (isEvenTimestep==true)
{
- D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
+ D.f[DIR_M00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_MPM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_PMP * numberOfLBnodes];
}
//////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
//set Pointer for Buffer Fs
Distributions27 Dbuff;
- Dbuff.f[DIR_P00 ] = &bufferFs[DIR_P00 *buffmax];
- Dbuff.f[DIR_M00 ] = &bufferFs[DIR_M00 *buffmax];
- Dbuff.f[DIR_0P0 ] = &bufferFs[DIR_0P0 *buffmax];
- Dbuff.f[DIR_0M0 ] = &bufferFs[DIR_0M0 *buffmax];
- Dbuff.f[DIR_00P ] = &bufferFs[DIR_00P *buffmax];
- Dbuff.f[DIR_00M ] = &bufferFs[DIR_00M *buffmax];
- Dbuff.f[DIR_PP0 ] = &bufferFs[DIR_PP0 *buffmax];
- Dbuff.f[DIR_MM0 ] = &bufferFs[DIR_MM0 *buffmax];
- Dbuff.f[DIR_PM0 ] = &bufferFs[DIR_PM0 *buffmax];
- Dbuff.f[DIR_MP0 ] = &bufferFs[DIR_MP0 *buffmax];
- Dbuff.f[DIR_P0P ] = &bufferFs[DIR_P0P *buffmax];
- Dbuff.f[DIR_M0M ] = &bufferFs[DIR_M0M *buffmax];
- Dbuff.f[DIR_P0M ] = &bufferFs[DIR_P0M *buffmax];
- Dbuff.f[DIR_M0P ] = &bufferFs[DIR_M0P *buffmax];
- Dbuff.f[DIR_0PP ] = &bufferFs[DIR_0PP *buffmax];
- Dbuff.f[DIR_0MM ] = &bufferFs[DIR_0MM *buffmax];
- Dbuff.f[DIR_0PM ] = &bufferFs[DIR_0PM *buffmax];
- Dbuff.f[DIR_0MP ] = &bufferFs[DIR_0MP *buffmax];
- Dbuff.f[DIR_000] = &bufferFs[DIR_000*buffmax];
- Dbuff.f[DIR_PPP ] = &bufferFs[DIR_PPP *buffmax];
- Dbuff.f[DIR_MMP ] = &bufferFs[DIR_MMP *buffmax];
- Dbuff.f[DIR_PMP ] = &bufferFs[DIR_PMP *buffmax];
- Dbuff.f[DIR_MPP ] = &bufferFs[DIR_MPP *buffmax];
- Dbuff.f[DIR_PPM ] = &bufferFs[DIR_PPM *buffmax];
- Dbuff.f[DIR_MMM ] = &bufferFs[DIR_MMM *buffmax];
- Dbuff.f[DIR_PMM ] = &bufferFs[DIR_PMM *buffmax];
- Dbuff.f[DIR_MPM ] = &bufferFs[DIR_MPM *buffmax];
+ Dbuff.f[DIR_P00] = &bufferFs[DIR_P00 * buffmax];
+ Dbuff.f[DIR_M00] = &bufferFs[DIR_M00 * buffmax];
+ Dbuff.f[DIR_0P0] = &bufferFs[DIR_0P0 * buffmax];
+ Dbuff.f[DIR_0M0] = &bufferFs[DIR_0M0 * buffmax];
+ Dbuff.f[DIR_00P] = &bufferFs[DIR_00P * buffmax];
+ Dbuff.f[DIR_00M] = &bufferFs[DIR_00M * buffmax];
+ Dbuff.f[DIR_PP0] = &bufferFs[DIR_PP0 * buffmax];
+ Dbuff.f[DIR_MM0] = &bufferFs[DIR_MM0 * buffmax];
+ Dbuff.f[DIR_PM0] = &bufferFs[DIR_PM0 * buffmax];
+ Dbuff.f[DIR_MP0] = &bufferFs[DIR_MP0 * buffmax];
+ Dbuff.f[DIR_P0P] = &bufferFs[DIR_P0P * buffmax];
+ Dbuff.f[DIR_M0M] = &bufferFs[DIR_M0M * buffmax];
+ Dbuff.f[DIR_P0M] = &bufferFs[DIR_P0M * buffmax];
+ Dbuff.f[DIR_M0P] = &bufferFs[DIR_M0P * buffmax];
+ Dbuff.f[DIR_0PP] = &bufferFs[DIR_0PP * buffmax];
+ Dbuff.f[DIR_0MM] = &bufferFs[DIR_0MM * buffmax];
+ Dbuff.f[DIR_0PM] = &bufferFs[DIR_0PM * buffmax];
+ Dbuff.f[DIR_0MP] = &bufferFs[DIR_0MP * buffmax];
+ Dbuff.f[DIR_000] = &bufferFs[DIR_000 * buffmax];
+ Dbuff.f[DIR_PPP] = &bufferFs[DIR_PPP * buffmax];
+ Dbuff.f[DIR_MMP] = &bufferFs[DIR_MMP * buffmax];
+ Dbuff.f[DIR_PMP] = &bufferFs[DIR_PMP * buffmax];
+ Dbuff.f[DIR_MPP] = &bufferFs[DIR_MPP * buffmax];
+ Dbuff.f[DIR_PPM] = &bufferFs[DIR_PPM * buffmax];
+ Dbuff.f[DIR_MMM] = &bufferFs[DIR_MMM * buffmax];
+ Dbuff.f[DIR_PMM] = &bufferFs[DIR_PMM * buffmax];
+ Dbuff.f[DIR_MPM] = &bufferFs[DIR_MPM * buffmax];
//////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
//copy to buffer
- //(Dbuff.f[DIR_P00 ])[k] = (D.f[DIR_P00 ])[ke ];
- //(Dbuff.f[DIR_M00 ])[k] = (D.f[DIR_M00 ])[kw ];
- //(Dbuff.f[DIR_0P0 ])[k] = (D.f[DIR_0P0 ])[kn ];
- //(Dbuff.f[DIR_0M0 ])[k] = (D.f[DIR_0M0 ])[ks ];
- //(Dbuff.f[DIR_00P ])[k] = (D.f[DIR_00P ])[kt ];
- //(Dbuff.f[DIR_00M ])[k] = (D.f[DIR_00M ])[kb ];
- //(Dbuff.f[DIR_PP0 ])[k] = (D.f[DIR_PP0 ])[kne ];
- //(Dbuff.f[DIR_MM0 ])[k] = (D.f[DIR_MM0 ])[ksw ];
- //(Dbuff.f[DIR_PM0 ])[k] = (D.f[DIR_PM0 ])[kse ];
- //(Dbuff.f[DIR_MP0 ])[k] = (D.f[DIR_MP0 ])[knw ];
- //(Dbuff.f[DIR_P0P ])[k] = (D.f[DIR_P0P ])[kte ];
- //(Dbuff.f[DIR_M0M ])[k] = (D.f[DIR_M0M ])[kbw ];
- //(Dbuff.f[DIR_P0M ])[k] = (D.f[DIR_P0M ])[kbe ];
- //(Dbuff.f[DIR_M0P ])[k] = (D.f[DIR_M0P ])[ktw ];
- //(Dbuff.f[DIR_0PP ])[k] = (D.f[DIR_0PP ])[ktn ];
- //(Dbuff.f[DIR_0MM ])[k] = (D.f[DIR_0MM ])[kbs ];
- //(Dbuff.f[DIR_0PM ])[k] = (D.f[DIR_0PM ])[kbn ];
- //(Dbuff.f[DIR_0MP ])[k] = (D.f[DIR_0MP ])[kts ];
+ //(Dbuff.f[DIR_P00])[k] = (D.f[DIR_P00])[ke ];
+ //(Dbuff.f[DIR_M00])[k] = (D.f[DIR_M00])[kw ];
+ //(Dbuff.f[DIR_0P0])[k] = (D.f[DIR_0P0])[kn ];
+ //(Dbuff.f[DIR_0M0])[k] = (D.f[DIR_0M0])[ks ];
+ //(Dbuff.f[DIR_00P])[k] = (D.f[DIR_00P])[kt ];
+ //(Dbuff.f[DIR_00M])[k] = (D.f[DIR_00M])[kb ];
+ //(Dbuff.f[DIR_PP0])[k] = (D.f[DIR_PP0])[kne ];
+ //(Dbuff.f[DIR_MM0])[k] = (D.f[DIR_MM0])[ksw ];
+ //(Dbuff.f[DIR_PM0])[k] = (D.f[DIR_PM0])[kse ];
+ //(Dbuff.f[DIR_MP0])[k] = (D.f[DIR_MP0])[knw ];
+ //(Dbuff.f[DIR_P0P])[k] = (D.f[DIR_P0P])[kte ];
+ //(Dbuff.f[DIR_M0M])[k] = (D.f[DIR_M0M])[kbw ];
+ //(Dbuff.f[DIR_P0M])[k] = (D.f[DIR_P0M])[kbe ];
+ //(Dbuff.f[DIR_M0P])[k] = (D.f[DIR_M0P])[ktw ];
+ //(Dbuff.f[DIR_0PP])[k] = (D.f[DIR_0PP])[ktn ];
+ //(Dbuff.f[DIR_0MM])[k] = (D.f[DIR_0MM])[kbs ];
+ //(Dbuff.f[DIR_0PM])[k] = (D.f[DIR_0PM])[kbn ];
+ //(Dbuff.f[DIR_0MP])[k] = (D.f[DIR_0MP])[kts ];
//(Dbuff.f[DIR_000])[k] = (D.f[DIR_000])[kzero];
- //(Dbuff.f[DIR_PPP ])[k] = (D.f[DIR_PPP ])[ktne ];
- //(Dbuff.f[DIR_MMP ])[k] = (D.f[DIR_MMP ])[ktsw ];
- //(Dbuff.f[DIR_PMP ])[k] = (D.f[DIR_PMP ])[ktse ];
- //(Dbuff.f[DIR_MPP ])[k] = (D.f[DIR_MPP ])[ktnw ];
- //(Dbuff.f[DIR_PPM ])[k] = (D.f[DIR_PPM ])[kbne ];
- //(Dbuff.f[DIR_MMM ])[k] = (D.f[DIR_MMM ])[kbsw ];
- //(Dbuff.f[DIR_PMM ])[k] = (D.f[DIR_PMM ])[kbse ];
- //(Dbuff.f[DIR_MPM ])[k] = (D.f[DIR_MPM ])[kbnw ];
- (Dbuff.f[DIR_P00 ])[k] = (D.f[DIR_M00 ])[kw ];
- (Dbuff.f[DIR_M00 ])[k] = (D.f[DIR_P00 ])[ke ];
- (Dbuff.f[DIR_0P0 ])[k] = (D.f[DIR_0M0 ])[ks ];
- (Dbuff.f[DIR_0M0 ])[k] = (D.f[DIR_0P0 ])[kn ];
- (Dbuff.f[DIR_00P ])[k] = (D.f[DIR_00M ])[kb ];
- (Dbuff.f[DIR_00M ])[k] = (D.f[DIR_00P ])[kt ];
- (Dbuff.f[DIR_PP0 ])[k] = (D.f[DIR_MM0 ])[ksw ];
- (Dbuff.f[DIR_MM0 ])[k] = (D.f[DIR_PP0 ])[kne ];
- (Dbuff.f[DIR_PM0 ])[k] = (D.f[DIR_MP0 ])[knw ];
- (Dbuff.f[DIR_MP0 ])[k] = (D.f[DIR_PM0 ])[kse ];
- (Dbuff.f[DIR_P0P ])[k] = (D.f[DIR_M0M ])[kbw ];
- (Dbuff.f[DIR_M0M ])[k] = (D.f[DIR_P0P ])[kte ];
- (Dbuff.f[DIR_P0M ])[k] = (D.f[DIR_M0P ])[ktw ];
- (Dbuff.f[DIR_M0P ])[k] = (D.f[DIR_P0M ])[kbe ];
- (Dbuff.f[DIR_0PP ])[k] = (D.f[DIR_0MM ])[kbs ];
- (Dbuff.f[DIR_0MM ])[k] = (D.f[DIR_0PP ])[ktn ];
- (Dbuff.f[DIR_0PM ])[k] = (D.f[DIR_0MP ])[kts ];
- (Dbuff.f[DIR_0MP ])[k] = (D.f[DIR_0PM ])[kbn ];
+ //(Dbuff.f[DIR_PPP])[k] = (D.f[DIR_PPP])[ktne ];
+ //(Dbuff.f[DIR_MMP])[k] = (D.f[DIR_MMP])[ktsw ];
+ //(Dbuff.f[DIR_PMP])[k] = (D.f[DIR_PMP])[ktse ];
+ //(Dbuff.f[DIR_MPP])[k] = (D.f[DIR_MPP])[ktnw ];
+ //(Dbuff.f[DIR_PPM])[k] = (D.f[DIR_PPM])[kbne ];
+ //(Dbuff.f[DIR_MMM])[k] = (D.f[DIR_MMM])[kbsw ];
+ //(Dbuff.f[DIR_PMM])[k] = (D.f[DIR_PMM])[kbse ];
+ //(Dbuff.f[DIR_MPM])[k] = (D.f[DIR_MPM])[kbnw ];
+ (Dbuff.f[DIR_P00])[k] = (D.f[DIR_M00])[kw ];
+ (Dbuff.f[DIR_M00])[k] = (D.f[DIR_P00])[ke ];
+ (Dbuff.f[DIR_0P0])[k] = (D.f[DIR_0M0])[ks ];
+ (Dbuff.f[DIR_0M0])[k] = (D.f[DIR_0P0])[kn ];
+ (Dbuff.f[DIR_00P])[k] = (D.f[DIR_00M])[kb ];
+ (Dbuff.f[DIR_00M])[k] = (D.f[DIR_00P])[kt ];
+ (Dbuff.f[DIR_PP0])[k] = (D.f[DIR_MM0])[ksw ];
+ (Dbuff.f[DIR_MM0])[k] = (D.f[DIR_PP0])[kne ];
+ (Dbuff.f[DIR_PM0])[k] = (D.f[DIR_MP0])[knw ];
+ (Dbuff.f[DIR_MP0])[k] = (D.f[DIR_PM0])[kse ];
+ (Dbuff.f[DIR_P0P])[k] = (D.f[DIR_M0M])[kbw ];
+ (Dbuff.f[DIR_M0M])[k] = (D.f[DIR_P0P])[kte ];
+ (Dbuff.f[DIR_P0M])[k] = (D.f[DIR_M0P])[ktw ];
+ (Dbuff.f[DIR_M0P])[k] = (D.f[DIR_P0M])[kbe ];
+ (Dbuff.f[DIR_0PP])[k] = (D.f[DIR_0MM])[kbs ];
+ (Dbuff.f[DIR_0MM])[k] = (D.f[DIR_0PP])[ktn ];
+ (Dbuff.f[DIR_0PM])[k] = (D.f[DIR_0MP])[kts ];
+ (Dbuff.f[DIR_0MP])[k] = (D.f[DIR_0PM])[kbn ];
(Dbuff.f[DIR_000])[k] = (D.f[DIR_000])[kzero];
- (Dbuff.f[DIR_PPP ])[k] = (D.f[DIR_MMM ])[kbsw ];
- (Dbuff.f[DIR_MMP ])[k] = (D.f[DIR_PPM ])[kbne ];
- (Dbuff.f[DIR_PMP ])[k] = (D.f[DIR_MPM ])[kbnw ];
- (Dbuff.f[DIR_MPP ])[k] = (D.f[DIR_PMM ])[kbse ];
- (Dbuff.f[DIR_PPM ])[k] = (D.f[DIR_MMP ])[ktsw ];
- (Dbuff.f[DIR_MMM ])[k] = (D.f[DIR_PPP ])[ktne ];
- (Dbuff.f[DIR_PMM ])[k] = (D.f[DIR_MPP ])[ktnw ];
- (Dbuff.f[DIR_MPM ])[k] = (D.f[DIR_PMP ])[ktse ];
+ (Dbuff.f[DIR_PPP])[k] = (D.f[DIR_MMM])[kbsw ];
+ (Dbuff.f[DIR_MMP])[k] = (D.f[DIR_PPM])[kbne ];
+ (Dbuff.f[DIR_PMP])[k] = (D.f[DIR_MPM])[kbnw ];
+ (Dbuff.f[DIR_MPP])[k] = (D.f[DIR_PMM])[kbse ];
+ (Dbuff.f[DIR_PPM])[k] = (D.f[DIR_MMP])[ktsw ];
+ (Dbuff.f[DIR_MMM])[k] = (D.f[DIR_PPP])[ktne ];
+ (Dbuff.f[DIR_PMM])[k] = (D.f[DIR_MPP])[ktnw ];
+ (Dbuff.f[DIR_MPM])[k] = (D.f[DIR_PMP])[ktse ];
}
}
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
@@ -250,7 +250,7 @@ __global__ void setRecvFsPost27(real* DD,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep)
{
////////////////////////////////////////////////////////////////////////////////
@@ -301,150 +301,150 @@ __global__ void setRecvFsPost27(real* DD,
Distributions27 D;
if (isEvenTimestep==true)
{
- D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
+ D.f[DIR_M00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_MPM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_PMP * numberOfLBnodes];
}
//////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
//set Pointer for Buffer Fs
Distributions27 Dbuff;
- Dbuff.f[DIR_P00 ] = &bufferFs[DIR_P00 *buffmax];
- Dbuff.f[DIR_M00 ] = &bufferFs[DIR_M00 *buffmax];
- Dbuff.f[DIR_0P0 ] = &bufferFs[DIR_0P0 *buffmax];
- Dbuff.f[DIR_0M0 ] = &bufferFs[DIR_0M0 *buffmax];
- Dbuff.f[DIR_00P ] = &bufferFs[DIR_00P *buffmax];
- Dbuff.f[DIR_00M ] = &bufferFs[DIR_00M *buffmax];
- Dbuff.f[DIR_PP0 ] = &bufferFs[DIR_PP0 *buffmax];
- Dbuff.f[DIR_MM0 ] = &bufferFs[DIR_MM0 *buffmax];
- Dbuff.f[DIR_PM0 ] = &bufferFs[DIR_PM0 *buffmax];
- Dbuff.f[DIR_MP0 ] = &bufferFs[DIR_MP0 *buffmax];
- Dbuff.f[DIR_P0P ] = &bufferFs[DIR_P0P *buffmax];
- Dbuff.f[DIR_M0M ] = &bufferFs[DIR_M0M *buffmax];
- Dbuff.f[DIR_P0M ] = &bufferFs[DIR_P0M *buffmax];
- Dbuff.f[DIR_M0P ] = &bufferFs[DIR_M0P *buffmax];
- Dbuff.f[DIR_0PP ] = &bufferFs[DIR_0PP *buffmax];
- Dbuff.f[DIR_0MM ] = &bufferFs[DIR_0MM *buffmax];
- Dbuff.f[DIR_0PM ] = &bufferFs[DIR_0PM *buffmax];
- Dbuff.f[DIR_0MP ] = &bufferFs[DIR_0MP *buffmax];
- Dbuff.f[DIR_000] = &bufferFs[DIR_000*buffmax];
- Dbuff.f[DIR_PPP ] = &bufferFs[DIR_PPP *buffmax];
- Dbuff.f[DIR_MMP ] = &bufferFs[DIR_MMP *buffmax];
- Dbuff.f[DIR_PMP ] = &bufferFs[DIR_PMP *buffmax];
- Dbuff.f[DIR_MPP ] = &bufferFs[DIR_MPP *buffmax];
- Dbuff.f[DIR_PPM ] = &bufferFs[DIR_PPM *buffmax];
- Dbuff.f[DIR_MMM ] = &bufferFs[DIR_MMM *buffmax];
- Dbuff.f[DIR_PMM ] = &bufferFs[DIR_PMM *buffmax];
- Dbuff.f[DIR_MPM ] = &bufferFs[DIR_MPM *buffmax];
+ Dbuff.f[DIR_P00] = &bufferFs[DIR_P00 * buffmax];
+ Dbuff.f[DIR_M00] = &bufferFs[DIR_M00 * buffmax];
+ Dbuff.f[DIR_0P0] = &bufferFs[DIR_0P0 * buffmax];
+ Dbuff.f[DIR_0M0] = &bufferFs[DIR_0M0 * buffmax];
+ Dbuff.f[DIR_00P] = &bufferFs[DIR_00P * buffmax];
+ Dbuff.f[DIR_00M] = &bufferFs[DIR_00M * buffmax];
+ Dbuff.f[DIR_PP0] = &bufferFs[DIR_PP0 * buffmax];
+ Dbuff.f[DIR_MM0] = &bufferFs[DIR_MM0 * buffmax];
+ Dbuff.f[DIR_PM0] = &bufferFs[DIR_PM0 * buffmax];
+ Dbuff.f[DIR_MP0] = &bufferFs[DIR_MP0 * buffmax];
+ Dbuff.f[DIR_P0P] = &bufferFs[DIR_P0P * buffmax];
+ Dbuff.f[DIR_M0M] = &bufferFs[DIR_M0M * buffmax];
+ Dbuff.f[DIR_P0M] = &bufferFs[DIR_P0M * buffmax];
+ Dbuff.f[DIR_M0P] = &bufferFs[DIR_M0P * buffmax];
+ Dbuff.f[DIR_0PP] = &bufferFs[DIR_0PP * buffmax];
+ Dbuff.f[DIR_0MM] = &bufferFs[DIR_0MM * buffmax];
+ Dbuff.f[DIR_0PM] = &bufferFs[DIR_0PM * buffmax];
+ Dbuff.f[DIR_0MP] = &bufferFs[DIR_0MP * buffmax];
+ Dbuff.f[DIR_000] = &bufferFs[DIR_000 * buffmax];
+ Dbuff.f[DIR_PPP] = &bufferFs[DIR_PPP * buffmax];
+ Dbuff.f[DIR_MMP] = &bufferFs[DIR_MMP * buffmax];
+ Dbuff.f[DIR_PMP] = &bufferFs[DIR_PMP * buffmax];
+ Dbuff.f[DIR_MPP] = &bufferFs[DIR_MPP * buffmax];
+ Dbuff.f[DIR_PPM] = &bufferFs[DIR_PPM * buffmax];
+ Dbuff.f[DIR_MMM] = &bufferFs[DIR_MMM * buffmax];
+ Dbuff.f[DIR_PMM] = &bufferFs[DIR_PMM * buffmax];
+ Dbuff.f[DIR_MPM] = &bufferFs[DIR_MPM * buffmax];
//////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
//copy from buffer
- //(D.f[DIR_P00 ])[ke ] = (Dbuff.f[DIR_P00 ])[k];
- //(D.f[DIR_M00 ])[kw ] = (Dbuff.f[DIR_M00 ])[k];
- //(D.f[DIR_0P0 ])[kn ] = (Dbuff.f[DIR_0P0 ])[k];
- //(D.f[DIR_0M0 ])[ks ] = (Dbuff.f[DIR_0M0 ])[k];
- //(D.f[DIR_00P ])[kt ] = (Dbuff.f[DIR_00P ])[k];
- //(D.f[DIR_00M ])[kb ] = (Dbuff.f[DIR_00M ])[k];
- //(D.f[DIR_PP0 ])[kne ] = (Dbuff.f[DIR_PP0 ])[k];
- //(D.f[DIR_MM0 ])[ksw ] = (Dbuff.f[DIR_MM0 ])[k];
- //(D.f[DIR_PM0 ])[kse ] = (Dbuff.f[DIR_PM0 ])[k];
- //(D.f[DIR_MP0 ])[knw ] = (Dbuff.f[DIR_MP0 ])[k];
- //(D.f[DIR_P0P ])[kte ] = (Dbuff.f[DIR_P0P ])[k];
- //(D.f[DIR_M0M ])[kbw ] = (Dbuff.f[DIR_M0M ])[k];
- //(D.f[DIR_P0M ])[kbe ] = (Dbuff.f[DIR_P0M ])[k];
- //(D.f[DIR_M0P ])[ktw ] = (Dbuff.f[DIR_M0P ])[k];
- //(D.f[DIR_0PP ])[ktn ] = (Dbuff.f[DIR_0PP ])[k];
- //(D.f[DIR_0MM ])[kbs ] = (Dbuff.f[DIR_0MM ])[k];
- //(D.f[DIR_0PM ])[kbn ] = (Dbuff.f[DIR_0PM ])[k];
- //(D.f[DIR_0MP ])[kts ] = (Dbuff.f[DIR_0MP ])[k];
+ //(D.f[DIR_P00])[ke ] = (Dbuff.f[DIR_P00])[k];
+ //(D.f[DIR_M00])[kw ] = (Dbuff.f[DIR_M00])[k];
+ //(D.f[DIR_0P0])[kn ] = (Dbuff.f[DIR_0P0])[k];
+ //(D.f[DIR_0M0])[ks ] = (Dbuff.f[DIR_0M0])[k];
+ //(D.f[DIR_00P])[kt ] = (Dbuff.f[DIR_00P])[k];
+ //(D.f[DIR_00M])[kb ] = (Dbuff.f[DIR_00M])[k];
+ //(D.f[DIR_PP0])[kne ] = (Dbuff.f[DIR_PP0])[k];
+ //(D.f[DIR_MM0])[ksw ] = (Dbuff.f[DIR_MM0])[k];
+ //(D.f[DIR_PM0])[kse ] = (Dbuff.f[DIR_PM0])[k];
+ //(D.f[DIR_MP0])[knw ] = (Dbuff.f[DIR_MP0])[k];
+ //(D.f[DIR_P0P])[kte ] = (Dbuff.f[DIR_P0P])[k];
+ //(D.f[DIR_M0M])[kbw ] = (Dbuff.f[DIR_M0M])[k];
+ //(D.f[DIR_P0M])[kbe ] = (Dbuff.f[DIR_P0M])[k];
+ //(D.f[DIR_M0P])[ktw ] = (Dbuff.f[DIR_M0P])[k];
+ //(D.f[DIR_0PP])[ktn ] = (Dbuff.f[DIR_0PP])[k];
+ //(D.f[DIR_0MM])[kbs ] = (Dbuff.f[DIR_0MM])[k];
+ //(D.f[DIR_0PM])[kbn ] = (Dbuff.f[DIR_0PM])[k];
+ //(D.f[DIR_0MP])[kts ] = (Dbuff.f[DIR_0MP])[k];
//(D.f[DIR_000])[kzero] = (Dbuff.f[DIR_000])[k];
- //(D.f[DIR_PPP ])[ktne ] = (Dbuff.f[DIR_PPP ])[k];
- //(D.f[DIR_MMP ])[ktsw ] = (Dbuff.f[DIR_MMP ])[k];
- //(D.f[DIR_PMP ])[ktse ] = (Dbuff.f[DIR_PMP ])[k];
- //(D.f[DIR_MPP ])[ktnw ] = (Dbuff.f[DIR_MPP ])[k];
- //(D.f[DIR_PPM ])[kbne ] = (Dbuff.f[DIR_PPM ])[k];
- //(D.f[DIR_MMM ])[kbsw ] = (Dbuff.f[DIR_MMM ])[k];
- //(D.f[DIR_PMM ])[kbse ] = (Dbuff.f[DIR_PMM ])[k];
- //(D.f[DIR_MPM ])[kbnw ] = (Dbuff.f[DIR_MPM ])[k];
- (D.f[DIR_M00 ])[kw ] = (Dbuff.f[DIR_P00 ])[k];
- (D.f[DIR_P00 ])[ke ] = (Dbuff.f[DIR_M00 ])[k];
- (D.f[DIR_0M0 ])[ks ] = (Dbuff.f[DIR_0P0 ])[k];
- (D.f[DIR_0P0 ])[kn ] = (Dbuff.f[DIR_0M0 ])[k];
- (D.f[DIR_00M ])[kb ] = (Dbuff.f[DIR_00P ])[k];
- (D.f[DIR_00P ])[kt ] = (Dbuff.f[DIR_00M ])[k];
- (D.f[DIR_MM0 ])[ksw ] = (Dbuff.f[DIR_PP0 ])[k];
- (D.f[DIR_PP0 ])[kne ] = (Dbuff.f[DIR_MM0 ])[k];
- (D.f[DIR_MP0 ])[knw ] = (Dbuff.f[DIR_PM0 ])[k];
- (D.f[DIR_PM0 ])[kse ] = (Dbuff.f[DIR_MP0 ])[k];
- (D.f[DIR_M0M ])[kbw ] = (Dbuff.f[DIR_P0P ])[k];
- (D.f[DIR_P0P ])[kte ] = (Dbuff.f[DIR_M0M ])[k];
- (D.f[DIR_M0P ])[ktw ] = (Dbuff.f[DIR_P0M ])[k];
- (D.f[DIR_P0M ])[kbe ] = (Dbuff.f[DIR_M0P ])[k];
- (D.f[DIR_0MM ])[kbs ] = (Dbuff.f[DIR_0PP ])[k];
- (D.f[DIR_0PP ])[ktn ] = (Dbuff.f[DIR_0MM ])[k];
- (D.f[DIR_0MP ])[kts ] = (Dbuff.f[DIR_0PM ])[k];
- (D.f[DIR_0PM ])[kbn ] = (Dbuff.f[DIR_0MP ])[k];
+ //(D.f[DIR_PPP])[ktne ] = (Dbuff.f[DIR_PPP])[k];
+ //(D.f[DIR_MMP])[ktsw ] = (Dbuff.f[DIR_MMP])[k];
+ //(D.f[DIR_PMP])[ktse ] = (Dbuff.f[DIR_PMP])[k];
+ //(D.f[DIR_MPP])[ktnw ] = (Dbuff.f[DIR_MPP])[k];
+ //(D.f[DIR_PPM])[kbne ] = (Dbuff.f[DIR_PPM])[k];
+ //(D.f[DIR_MMM])[kbsw ] = (Dbuff.f[DIR_MMM])[k];
+ //(D.f[DIR_PMM])[kbse ] = (Dbuff.f[DIR_PMM])[k];
+ //(D.f[DIR_MPM])[kbnw ] = (Dbuff.f[DIR_MPM])[k];
+ (D.f[DIR_M00])[kw ] = (Dbuff.f[DIR_P00])[k];
+ (D.f[DIR_P00])[ke ] = (Dbuff.f[DIR_M00])[k];
+ (D.f[DIR_0M0])[ks ] = (Dbuff.f[DIR_0P0])[k];
+ (D.f[DIR_0P0])[kn ] = (Dbuff.f[DIR_0M0])[k];
+ (D.f[DIR_00M])[kb ] = (Dbuff.f[DIR_00P])[k];
+ (D.f[DIR_00P])[kt ] = (Dbuff.f[DIR_00M])[k];
+ (D.f[DIR_MM0])[ksw ] = (Dbuff.f[DIR_PP0])[k];
+ (D.f[DIR_PP0])[kne ] = (Dbuff.f[DIR_MM0])[k];
+ (D.f[DIR_MP0])[knw ] = (Dbuff.f[DIR_PM0])[k];
+ (D.f[DIR_PM0])[kse ] = (Dbuff.f[DIR_MP0])[k];
+ (D.f[DIR_M0M])[kbw ] = (Dbuff.f[DIR_P0P])[k];
+ (D.f[DIR_P0P])[kte ] = (Dbuff.f[DIR_M0M])[k];
+ (D.f[DIR_M0P])[ktw ] = (Dbuff.f[DIR_P0M])[k];
+ (D.f[DIR_P0M])[kbe ] = (Dbuff.f[DIR_M0P])[k];
+ (D.f[DIR_0MM])[kbs ] = (Dbuff.f[DIR_0PP])[k];
+ (D.f[DIR_0PP])[ktn ] = (Dbuff.f[DIR_0MM])[k];
+ (D.f[DIR_0MP])[kts ] = (Dbuff.f[DIR_0PM])[k];
+ (D.f[DIR_0PM])[kbn ] = (Dbuff.f[DIR_0MP])[k];
(D.f[DIR_000])[kzero] = (Dbuff.f[DIR_000])[k];
- (D.f[DIR_MMM ])[kbsw ] = (Dbuff.f[DIR_PPP ])[k];
- (D.f[DIR_PPM ])[kbne ] = (Dbuff.f[DIR_MMP ])[k];
- (D.f[DIR_MPM ])[kbnw ] = (Dbuff.f[DIR_PMP ])[k];
- (D.f[DIR_PMM ])[kbse ] = (Dbuff.f[DIR_MPP ])[k];
- (D.f[DIR_MMP ])[ktsw ] = (Dbuff.f[DIR_PPM ])[k];
- (D.f[DIR_PPP ])[ktne ] = (Dbuff.f[DIR_MMM ])[k];
- (D.f[DIR_MPP ])[ktnw ] = (Dbuff.f[DIR_PMM ])[k];
- (D.f[DIR_PMP ])[ktse ] = (Dbuff.f[DIR_MPM ])[k];
+ (D.f[DIR_MMM])[kbsw ] = (Dbuff.f[DIR_PPP])[k];
+ (D.f[DIR_PPM])[kbne ] = (Dbuff.f[DIR_MMP])[k];
+ (D.f[DIR_MPM])[kbnw ] = (Dbuff.f[DIR_PMP])[k];
+ (D.f[DIR_PMM])[kbse ] = (Dbuff.f[DIR_MPP])[k];
+ (D.f[DIR_MMP])[ktsw ] = (Dbuff.f[DIR_PPM])[k];
+ (D.f[DIR_PPP])[ktne ] = (Dbuff.f[DIR_MMM])[k];
+ (D.f[DIR_MPP])[ktnw ] = (Dbuff.f[DIR_PMM])[k];
+ (D.f[DIR_PMP])[ktse ] = (Dbuff.f[DIR_MPM])[k];
}
}
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
@@ -485,7 +485,7 @@ __global__ void getSendFsPre27(real* DD,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep)
{
////////////////////////////////////////////////////////////////////////////////
@@ -536,123 +536,123 @@ __global__ void getSendFsPre27(real* DD,
Distributions27 D;
if (isEvenTimestep==true)
{
- D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
+ D.f[DIR_M00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_MPM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_PMP * numberOfLBnodes];
}
//////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
//set Pointer for Buffer Fs
Distributions27 Dbuff;
- Dbuff.f[DIR_P00 ] = &bufferFs[DIR_P00 *buffmax];
- Dbuff.f[DIR_M00 ] = &bufferFs[DIR_M00 *buffmax];
- Dbuff.f[DIR_0P0 ] = &bufferFs[DIR_0P0 *buffmax];
- Dbuff.f[DIR_0M0 ] = &bufferFs[DIR_0M0 *buffmax];
- Dbuff.f[DIR_00P ] = &bufferFs[DIR_00P *buffmax];
- Dbuff.f[DIR_00M ] = &bufferFs[DIR_00M *buffmax];
- Dbuff.f[DIR_PP0 ] = &bufferFs[DIR_PP0 *buffmax];
- Dbuff.f[DIR_MM0 ] = &bufferFs[DIR_MM0 *buffmax];
- Dbuff.f[DIR_PM0 ] = &bufferFs[DIR_PM0 *buffmax];
- Dbuff.f[DIR_MP0 ] = &bufferFs[DIR_MP0 *buffmax];
- Dbuff.f[DIR_P0P ] = &bufferFs[DIR_P0P *buffmax];
- Dbuff.f[DIR_M0M ] = &bufferFs[DIR_M0M *buffmax];
- Dbuff.f[DIR_P0M ] = &bufferFs[DIR_P0M *buffmax];
- Dbuff.f[DIR_M0P ] = &bufferFs[DIR_M0P *buffmax];
- Dbuff.f[DIR_0PP ] = &bufferFs[DIR_0PP *buffmax];
- Dbuff.f[DIR_0MM ] = &bufferFs[DIR_0MM *buffmax];
- Dbuff.f[DIR_0PM ] = &bufferFs[DIR_0PM *buffmax];
- Dbuff.f[DIR_0MP ] = &bufferFs[DIR_0MP *buffmax];
- Dbuff.f[DIR_000] = &bufferFs[DIR_000*buffmax];
- Dbuff.f[DIR_PPP ] = &bufferFs[DIR_PPP *buffmax];
- Dbuff.f[DIR_MMP ] = &bufferFs[DIR_MMP *buffmax];
- Dbuff.f[DIR_PMP ] = &bufferFs[DIR_PMP *buffmax];
- Dbuff.f[DIR_MPP ] = &bufferFs[DIR_MPP *buffmax];
- Dbuff.f[DIR_PPM ] = &bufferFs[DIR_PPM *buffmax];
- Dbuff.f[DIR_MMM ] = &bufferFs[DIR_MMM *buffmax];
- Dbuff.f[DIR_PMM ] = &bufferFs[DIR_PMM *buffmax];
- Dbuff.f[DIR_MPM ] = &bufferFs[DIR_MPM *buffmax];
+ Dbuff.f[DIR_P00] = &bufferFs[DIR_P00 * buffmax];
+ Dbuff.f[DIR_M00] = &bufferFs[DIR_M00 * buffmax];
+ Dbuff.f[DIR_0P0] = &bufferFs[DIR_0P0 * buffmax];
+ Dbuff.f[DIR_0M0] = &bufferFs[DIR_0M0 * buffmax];
+ Dbuff.f[DIR_00P] = &bufferFs[DIR_00P * buffmax];
+ Dbuff.f[DIR_00M] = &bufferFs[DIR_00M * buffmax];
+ Dbuff.f[DIR_PP0] = &bufferFs[DIR_PP0 * buffmax];
+ Dbuff.f[DIR_MM0] = &bufferFs[DIR_MM0 * buffmax];
+ Dbuff.f[DIR_PM0] = &bufferFs[DIR_PM0 * buffmax];
+ Dbuff.f[DIR_MP0] = &bufferFs[DIR_MP0 * buffmax];
+ Dbuff.f[DIR_P0P] = &bufferFs[DIR_P0P * buffmax];
+ Dbuff.f[DIR_M0M] = &bufferFs[DIR_M0M * buffmax];
+ Dbuff.f[DIR_P0M] = &bufferFs[DIR_P0M * buffmax];
+ Dbuff.f[DIR_M0P] = &bufferFs[DIR_M0P * buffmax];
+ Dbuff.f[DIR_0PP] = &bufferFs[DIR_0PP * buffmax];
+ Dbuff.f[DIR_0MM] = &bufferFs[DIR_0MM * buffmax];
+ Dbuff.f[DIR_0PM] = &bufferFs[DIR_0PM * buffmax];
+ Dbuff.f[DIR_0MP] = &bufferFs[DIR_0MP * buffmax];
+ Dbuff.f[DIR_000] = &bufferFs[DIR_000 * buffmax];
+ Dbuff.f[DIR_PPP] = &bufferFs[DIR_PPP * buffmax];
+ Dbuff.f[DIR_MMP] = &bufferFs[DIR_MMP * buffmax];
+ Dbuff.f[DIR_PMP] = &bufferFs[DIR_PMP * buffmax];
+ Dbuff.f[DIR_MPP] = &bufferFs[DIR_MPP * buffmax];
+ Dbuff.f[DIR_PPM] = &bufferFs[DIR_PPM * buffmax];
+ Dbuff.f[DIR_MMM] = &bufferFs[DIR_MMM * buffmax];
+ Dbuff.f[DIR_PMM] = &bufferFs[DIR_PMM * buffmax];
+ Dbuff.f[DIR_MPM] = &bufferFs[DIR_MPM * buffmax];
//////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
//copy to buffer
- (Dbuff.f[DIR_P00 ])[k] = (D.f[DIR_P00 ])[ke ];
- (Dbuff.f[DIR_M00 ])[k] = (D.f[DIR_M00 ])[kw ];
- (Dbuff.f[DIR_0P0 ])[k] = (D.f[DIR_0P0 ])[kn ];
- (Dbuff.f[DIR_0M0 ])[k] = (D.f[DIR_0M0 ])[ks ];
- (Dbuff.f[DIR_00P ])[k] = (D.f[DIR_00P ])[kt ];
- (Dbuff.f[DIR_00M ])[k] = (D.f[DIR_00M ])[kb ];
- (Dbuff.f[DIR_PP0 ])[k] = (D.f[DIR_PP0 ])[kne ];
- (Dbuff.f[DIR_MM0 ])[k] = (D.f[DIR_MM0 ])[ksw ];
- (Dbuff.f[DIR_PM0 ])[k] = (D.f[DIR_PM0 ])[kse ];
- (Dbuff.f[DIR_MP0 ])[k] = (D.f[DIR_MP0 ])[knw ];
- (Dbuff.f[DIR_P0P ])[k] = (D.f[DIR_P0P ])[kte ];
- (Dbuff.f[DIR_M0M ])[k] = (D.f[DIR_M0M ])[kbw ];
- (Dbuff.f[DIR_P0M ])[k] = (D.f[DIR_P0M ])[kbe ];
- (Dbuff.f[DIR_M0P ])[k] = (D.f[DIR_M0P ])[ktw ];
- (Dbuff.f[DIR_0PP ])[k] = (D.f[DIR_0PP ])[ktn ];
- (Dbuff.f[DIR_0MM ])[k] = (D.f[DIR_0MM ])[kbs ];
- (Dbuff.f[DIR_0PM ])[k] = (D.f[DIR_0PM ])[kbn ];
- (Dbuff.f[DIR_0MP ])[k] = (D.f[DIR_0MP ])[kts ];
+ (Dbuff.f[DIR_P00])[k] = (D.f[DIR_P00])[ke ];
+ (Dbuff.f[DIR_M00])[k] = (D.f[DIR_M00])[kw ];
+ (Dbuff.f[DIR_0P0])[k] = (D.f[DIR_0P0])[kn ];
+ (Dbuff.f[DIR_0M0])[k] = (D.f[DIR_0M0])[ks ];
+ (Dbuff.f[DIR_00P])[k] = (D.f[DIR_00P])[kt ];
+ (Dbuff.f[DIR_00M])[k] = (D.f[DIR_00M])[kb ];
+ (Dbuff.f[DIR_PP0])[k] = (D.f[DIR_PP0])[kne ];
+ (Dbuff.f[DIR_MM0])[k] = (D.f[DIR_MM0])[ksw ];
+ (Dbuff.f[DIR_PM0])[k] = (D.f[DIR_PM0])[kse ];
+ (Dbuff.f[DIR_MP0])[k] = (D.f[DIR_MP0])[knw ];
+ (Dbuff.f[DIR_P0P])[k] = (D.f[DIR_P0P])[kte ];
+ (Dbuff.f[DIR_M0M])[k] = (D.f[DIR_M0M])[kbw ];
+ (Dbuff.f[DIR_P0M])[k] = (D.f[DIR_P0M])[kbe ];
+ (Dbuff.f[DIR_M0P])[k] = (D.f[DIR_M0P])[ktw ];
+ (Dbuff.f[DIR_0PP])[k] = (D.f[DIR_0PP])[ktn ];
+ (Dbuff.f[DIR_0MM])[k] = (D.f[DIR_0MM])[kbs ];
+ (Dbuff.f[DIR_0PM])[k] = (D.f[DIR_0PM])[kbn ];
+ (Dbuff.f[DIR_0MP])[k] = (D.f[DIR_0MP])[kts ];
(Dbuff.f[DIR_000])[k] = (D.f[DIR_000])[kzero];
- (Dbuff.f[DIR_PPP ])[k] = (D.f[DIR_PPP ])[ktne ];
- (Dbuff.f[DIR_MMP ])[k] = (D.f[DIR_MMP ])[ktsw ];
- (Dbuff.f[DIR_PMP ])[k] = (D.f[DIR_PMP ])[ktse ];
- (Dbuff.f[DIR_MPP ])[k] = (D.f[DIR_MPP ])[ktnw ];
- (Dbuff.f[DIR_PPM ])[k] = (D.f[DIR_PPM ])[kbne ];
- (Dbuff.f[DIR_MMM ])[k] = (D.f[DIR_MMM ])[kbsw ];
- (Dbuff.f[DIR_PMM ])[k] = (D.f[DIR_PMM ])[kbse ];
- (Dbuff.f[DIR_MPM ])[k] = (D.f[DIR_MPM ])[kbnw ];
+ (Dbuff.f[DIR_PPP])[k] = (D.f[DIR_PPP])[ktne ];
+ (Dbuff.f[DIR_MMP])[k] = (D.f[DIR_MMP])[ktsw ];
+ (Dbuff.f[DIR_PMP])[k] = (D.f[DIR_PMP])[ktse ];
+ (Dbuff.f[DIR_MPP])[k] = (D.f[DIR_MPP])[ktnw ];
+ (Dbuff.f[DIR_PPM])[k] = (D.f[DIR_PPM])[kbne ];
+ (Dbuff.f[DIR_MMM])[k] = (D.f[DIR_MMM])[kbsw ];
+ (Dbuff.f[DIR_PMM])[k] = (D.f[DIR_PMM])[kbse ];
+ (Dbuff.f[DIR_MPM])[k] = (D.f[DIR_MPM])[kbnw ];
}
}
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
@@ -694,7 +694,7 @@ __global__ void setRecvFsPre27(real* DD,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep)
{
////////////////////////////////////////////////////////////////////////////////
@@ -745,123 +745,123 @@ __global__ void setRecvFsPre27(real* DD,
Distributions27 D;
if (isEvenTimestep==true)
{
- D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
+ D.f[DIR_M00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_MPM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_PMP * numberOfLBnodes];
}
//////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
//set Pointer for Buffer Fs
Distributions27 Dbuff;
- Dbuff.f[DIR_P00 ] = &bufferFs[DIR_P00 *buffmax];
- Dbuff.f[DIR_M00 ] = &bufferFs[DIR_M00 *buffmax];
- Dbuff.f[DIR_0P0 ] = &bufferFs[DIR_0P0 *buffmax];
- Dbuff.f[DIR_0M0 ] = &bufferFs[DIR_0M0 *buffmax];
- Dbuff.f[DIR_00P ] = &bufferFs[DIR_00P *buffmax];
- Dbuff.f[DIR_00M ] = &bufferFs[DIR_00M *buffmax];
- Dbuff.f[DIR_PP0 ] = &bufferFs[DIR_PP0 *buffmax];
- Dbuff.f[DIR_MM0 ] = &bufferFs[DIR_MM0 *buffmax];
- Dbuff.f[DIR_PM0 ] = &bufferFs[DIR_PM0 *buffmax];
- Dbuff.f[DIR_MP0 ] = &bufferFs[DIR_MP0 *buffmax];
- Dbuff.f[DIR_P0P ] = &bufferFs[DIR_P0P *buffmax];
- Dbuff.f[DIR_M0M ] = &bufferFs[DIR_M0M *buffmax];
- Dbuff.f[DIR_P0M ] = &bufferFs[DIR_P0M *buffmax];
- Dbuff.f[DIR_M0P ] = &bufferFs[DIR_M0P *buffmax];
- Dbuff.f[DIR_0PP ] = &bufferFs[DIR_0PP *buffmax];
- Dbuff.f[DIR_0MM ] = &bufferFs[DIR_0MM *buffmax];
- Dbuff.f[DIR_0PM ] = &bufferFs[DIR_0PM *buffmax];
- Dbuff.f[DIR_0MP ] = &bufferFs[DIR_0MP *buffmax];
- Dbuff.f[DIR_000] = &bufferFs[DIR_000*buffmax];
- Dbuff.f[DIR_PPP ] = &bufferFs[DIR_PPP *buffmax];
- Dbuff.f[DIR_MMP ] = &bufferFs[DIR_MMP *buffmax];
- Dbuff.f[DIR_PMP ] = &bufferFs[DIR_PMP *buffmax];
- Dbuff.f[DIR_MPP ] = &bufferFs[DIR_MPP *buffmax];
- Dbuff.f[DIR_PPM ] = &bufferFs[DIR_PPM *buffmax];
- Dbuff.f[DIR_MMM ] = &bufferFs[DIR_MMM *buffmax];
- Dbuff.f[DIR_PMM ] = &bufferFs[DIR_PMM *buffmax];
- Dbuff.f[DIR_MPM ] = &bufferFs[DIR_MPM *buffmax];
+ Dbuff.f[DIR_P00] = &bufferFs[DIR_P00 * buffmax];
+ Dbuff.f[DIR_M00] = &bufferFs[DIR_M00 * buffmax];
+ Dbuff.f[DIR_0P0] = &bufferFs[DIR_0P0 * buffmax];
+ Dbuff.f[DIR_0M0] = &bufferFs[DIR_0M0 * buffmax];
+ Dbuff.f[DIR_00P] = &bufferFs[DIR_00P * buffmax];
+ Dbuff.f[DIR_00M] = &bufferFs[DIR_00M * buffmax];
+ Dbuff.f[DIR_PP0] = &bufferFs[DIR_PP0 * buffmax];
+ Dbuff.f[DIR_MM0] = &bufferFs[DIR_MM0 * buffmax];
+ Dbuff.f[DIR_PM0] = &bufferFs[DIR_PM0 * buffmax];
+ Dbuff.f[DIR_MP0] = &bufferFs[DIR_MP0 * buffmax];
+ Dbuff.f[DIR_P0P] = &bufferFs[DIR_P0P * buffmax];
+ Dbuff.f[DIR_M0M] = &bufferFs[DIR_M0M * buffmax];
+ Dbuff.f[DIR_P0M] = &bufferFs[DIR_P0M * buffmax];
+ Dbuff.f[DIR_M0P] = &bufferFs[DIR_M0P * buffmax];
+ Dbuff.f[DIR_0PP] = &bufferFs[DIR_0PP * buffmax];
+ Dbuff.f[DIR_0MM] = &bufferFs[DIR_0MM * buffmax];
+ Dbuff.f[DIR_0PM] = &bufferFs[DIR_0PM * buffmax];
+ Dbuff.f[DIR_0MP] = &bufferFs[DIR_0MP * buffmax];
+ Dbuff.f[DIR_000] = &bufferFs[DIR_000 * buffmax];
+ Dbuff.f[DIR_PPP] = &bufferFs[DIR_PPP * buffmax];
+ Dbuff.f[DIR_MMP] = &bufferFs[DIR_MMP * buffmax];
+ Dbuff.f[DIR_PMP] = &bufferFs[DIR_PMP * buffmax];
+ Dbuff.f[DIR_MPP] = &bufferFs[DIR_MPP * buffmax];
+ Dbuff.f[DIR_PPM] = &bufferFs[DIR_PPM * buffmax];
+ Dbuff.f[DIR_MMM] = &bufferFs[DIR_MMM * buffmax];
+ Dbuff.f[DIR_PMM] = &bufferFs[DIR_PMM * buffmax];
+ Dbuff.f[DIR_MPM] = &bufferFs[DIR_MPM * buffmax];
//////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
//copy from buffer
- (D.f[DIR_P00 ])[ke ] = (Dbuff.f[DIR_P00 ])[k];
- (D.f[DIR_M00 ])[kw ] = (Dbuff.f[DIR_M00 ])[k];
- (D.f[DIR_0P0 ])[kn ] = (Dbuff.f[DIR_0P0 ])[k];
- (D.f[DIR_0M0 ])[ks ] = (Dbuff.f[DIR_0M0 ])[k];
- (D.f[DIR_00P ])[kt ] = (Dbuff.f[DIR_00P ])[k];
- (D.f[DIR_00M ])[kb ] = (Dbuff.f[DIR_00M ])[k];
- (D.f[DIR_PP0 ])[kne ] = (Dbuff.f[DIR_PP0 ])[k];
- (D.f[DIR_MM0 ])[ksw ] = (Dbuff.f[DIR_MM0 ])[k];
- (D.f[DIR_PM0 ])[kse ] = (Dbuff.f[DIR_PM0 ])[k];
- (D.f[DIR_MP0 ])[knw ] = (Dbuff.f[DIR_MP0 ])[k];
- (D.f[DIR_P0P ])[kte ] = (Dbuff.f[DIR_P0P ])[k];
- (D.f[DIR_M0M ])[kbw ] = (Dbuff.f[DIR_M0M ])[k];
- (D.f[DIR_P0M ])[kbe ] = (Dbuff.f[DIR_P0M ])[k];
- (D.f[DIR_M0P ])[ktw ] = (Dbuff.f[DIR_M0P ])[k];
- (D.f[DIR_0PP ])[ktn ] = (Dbuff.f[DIR_0PP ])[k];
- (D.f[DIR_0MM ])[kbs ] = (Dbuff.f[DIR_0MM ])[k];
- (D.f[DIR_0PM ])[kbn ] = (Dbuff.f[DIR_0PM ])[k];
- (D.f[DIR_0MP ])[kts ] = (Dbuff.f[DIR_0MP ])[k];
+ (D.f[DIR_P00])[ke ] = (Dbuff.f[DIR_P00])[k];
+ (D.f[DIR_M00])[kw ] = (Dbuff.f[DIR_M00])[k];
+ (D.f[DIR_0P0])[kn ] = (Dbuff.f[DIR_0P0])[k];
+ (D.f[DIR_0M0])[ks ] = (Dbuff.f[DIR_0M0])[k];
+ (D.f[DIR_00P])[kt ] = (Dbuff.f[DIR_00P])[k];
+ (D.f[DIR_00M])[kb ] = (Dbuff.f[DIR_00M])[k];
+ (D.f[DIR_PP0])[kne ] = (Dbuff.f[DIR_PP0])[k];
+ (D.f[DIR_MM0])[ksw ] = (Dbuff.f[DIR_MM0])[k];
+ (D.f[DIR_PM0])[kse ] = (Dbuff.f[DIR_PM0])[k];
+ (D.f[DIR_MP0])[knw ] = (Dbuff.f[DIR_MP0])[k];
+ (D.f[DIR_P0P])[kte ] = (Dbuff.f[DIR_P0P])[k];
+ (D.f[DIR_M0M])[kbw ] = (Dbuff.f[DIR_M0M])[k];
+ (D.f[DIR_P0M])[kbe ] = (Dbuff.f[DIR_P0M])[k];
+ (D.f[DIR_M0P])[ktw ] = (Dbuff.f[DIR_M0P])[k];
+ (D.f[DIR_0PP])[ktn ] = (Dbuff.f[DIR_0PP])[k];
+ (D.f[DIR_0MM])[kbs ] = (Dbuff.f[DIR_0MM])[k];
+ (D.f[DIR_0PM])[kbn ] = (Dbuff.f[DIR_0PM])[k];
+ (D.f[DIR_0MP])[kts ] = (Dbuff.f[DIR_0MP])[k];
(D.f[DIR_000])[kzero] = (Dbuff.f[DIR_000])[k];
- (D.f[DIR_PPP ])[ktne ] = (Dbuff.f[DIR_PPP ])[k];
- (D.f[DIR_MMP ])[ktsw ] = (Dbuff.f[DIR_MMP ])[k];
- (D.f[DIR_PMP ])[ktse ] = (Dbuff.f[DIR_PMP ])[k];
- (D.f[DIR_MPP ])[ktnw ] = (Dbuff.f[DIR_MPP ])[k];
- (D.f[DIR_PPM ])[kbne ] = (Dbuff.f[DIR_PPM ])[k];
- (D.f[DIR_MMM ])[kbsw ] = (Dbuff.f[DIR_MMM ])[k];
- (D.f[DIR_PMM ])[kbse ] = (Dbuff.f[DIR_PMM ])[k];
- (D.f[DIR_MPM ])[kbnw ] = (Dbuff.f[DIR_MPM ])[k];
+ (D.f[DIR_PPP])[ktne ] = (Dbuff.f[DIR_PPP])[k];
+ (D.f[DIR_MMP])[ktsw ] = (Dbuff.f[DIR_MMP])[k];
+ (D.f[DIR_PMP])[ktse ] = (Dbuff.f[DIR_PMP])[k];
+ (D.f[DIR_MPP])[ktnw ] = (Dbuff.f[DIR_MPP])[k];
+ (D.f[DIR_PPM])[kbne ] = (Dbuff.f[DIR_PPM])[k];
+ (D.f[DIR_MMM])[kbsw ] = (Dbuff.f[DIR_MMM])[k];
+ (D.f[DIR_PMM])[kbse ] = (Dbuff.f[DIR_PMM])[k];
+ (D.f[DIR_MPM])[kbnw ] = (Dbuff.f[DIR_MPM])[k];
}
}
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
@@ -903,7 +903,7 @@ __global__ void getSendGsF3(
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep)
{
////////////////////////////////////////////////////////////////////////////////
@@ -931,31 +931,31 @@ __global__ void getSendGsF3(
Distributions6 G;
if (isEvenTimestep)
{
- G.g[DIR_P00] = &G6[DIR_P00 *size_Mat];
- G.g[DIR_M00] = &G6[DIR_M00 *size_Mat];
- G.g[DIR_0P0] = &G6[DIR_0P0 *size_Mat];
- G.g[DIR_0M0] = &G6[DIR_0M0 *size_Mat];
- G.g[DIR_00P] = &G6[DIR_00P *size_Mat];
- G.g[DIR_00M] = &G6[DIR_00M *size_Mat];
+ G.g[DIR_P00] = &G6[DIR_P00 * numberOfLBnodes];
+ G.g[DIR_M00] = &G6[DIR_M00 * numberOfLBnodes];
+ G.g[DIR_0P0] = &G6[DIR_0P0 * numberOfLBnodes];
+ G.g[DIR_0M0] = &G6[DIR_0M0 * numberOfLBnodes];
+ G.g[DIR_00P] = &G6[DIR_00P * numberOfLBnodes];
+ G.g[DIR_00M] = &G6[DIR_00M * numberOfLBnodes];
}
else
{
- G.g[DIR_M00] = &G6[DIR_P00 *size_Mat];
- G.g[DIR_P00] = &G6[DIR_M00 *size_Mat];
- G.g[DIR_0M0] = &G6[DIR_0P0 *size_Mat];
- G.g[DIR_0P0] = &G6[DIR_0M0 *size_Mat];
- G.g[DIR_00M] = &G6[DIR_00P *size_Mat];
- G.g[DIR_00P] = &G6[DIR_00M *size_Mat];
+ G.g[DIR_M00] = &G6[DIR_P00 * numberOfLBnodes];
+ G.g[DIR_P00] = &G6[DIR_M00 * numberOfLBnodes];
+ G.g[DIR_0M0] = &G6[DIR_0P0 * numberOfLBnodes];
+ G.g[DIR_0P0] = &G6[DIR_0M0 * numberOfLBnodes];
+ G.g[DIR_00M] = &G6[DIR_00P * numberOfLBnodes];
+ G.g[DIR_00P] = &G6[DIR_00M * numberOfLBnodes];
}
//////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
//set Pointer for Buffer Gs
Distributions6 Dbuff;
- Dbuff.g[DIR_P00] = &bufferGs[DIR_P00 *buffmax];
- Dbuff.g[DIR_M00] = &bufferGs[DIR_M00 *buffmax];
- Dbuff.g[DIR_0P0] = &bufferGs[DIR_0P0 *buffmax];
- Dbuff.g[DIR_0M0] = &bufferGs[DIR_0M0 *buffmax];
- Dbuff.g[DIR_00P] = &bufferGs[DIR_00P *buffmax];
- Dbuff.g[DIR_00M] = &bufferGs[DIR_00M *buffmax];
+ Dbuff.g[DIR_P00] = &bufferGs[DIR_P00 * buffmax];
+ Dbuff.g[DIR_M00] = &bufferGs[DIR_M00 * buffmax];
+ Dbuff.g[DIR_0P0] = &bufferGs[DIR_0P0 * buffmax];
+ Dbuff.g[DIR_0M0] = &bufferGs[DIR_0M0 * buffmax];
+ Dbuff.g[DIR_00P] = &bufferGs[DIR_00P * buffmax];
+ Dbuff.g[DIR_00M] = &bufferGs[DIR_00M * buffmax];
//////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
//write Gs to buffer
(Dbuff.g[DIR_P00])[k] = (G.g[DIR_M00])[kw];
@@ -1006,7 +1006,7 @@ __global__ void setRecvGsF3(
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep)
{
////////////////////////////////////////////////////////////////////////////////
@@ -1034,31 +1034,31 @@ __global__ void setRecvGsF3(
Distributions6 G;
if (isEvenTimestep)
{
- G.g[DIR_P00] = &G6[DIR_P00 *size_Mat];
- G.g[DIR_M00] = &G6[DIR_M00 *size_Mat];
- G.g[DIR_0P0] = &G6[DIR_0P0 *size_Mat];
- G.g[DIR_0M0] = &G6[DIR_0M0 *size_Mat];
- G.g[DIR_00P] = &G6[DIR_00P *size_Mat];
- G.g[DIR_00M] = &G6[DIR_00M *size_Mat];
+ G.g[DIR_P00] = &G6[DIR_P00 * numberOfLBnodes];
+ G.g[DIR_M00] = &G6[DIR_M00 * numberOfLBnodes];
+ G.g[DIR_0P0] = &G6[DIR_0P0 * numberOfLBnodes];
+ G.g[DIR_0M0] = &G6[DIR_0M0 * numberOfLBnodes];
+ G.g[DIR_00P] = &G6[DIR_00P * numberOfLBnodes];
+ G.g[DIR_00M] = &G6[DIR_00M * numberOfLBnodes];
}
else
{
- G.g[DIR_M00] = &G6[DIR_P00 *size_Mat];
- G.g[DIR_P00] = &G6[DIR_M00 *size_Mat];
- G.g[DIR_0M0] = &G6[DIR_0P0 *size_Mat];
- G.g[DIR_0P0] = &G6[DIR_0M0 *size_Mat];
- G.g[DIR_00M] = &G6[DIR_00P *size_Mat];
- G.g[DIR_00P] = &G6[DIR_00M *size_Mat];
+ G.g[DIR_M00] = &G6[DIR_P00 * numberOfLBnodes];
+ G.g[DIR_P00] = &G6[DIR_M00 * numberOfLBnodes];
+ G.g[DIR_0M0] = &G6[DIR_0P0 * numberOfLBnodes];
+ G.g[DIR_0P0] = &G6[DIR_0M0 * numberOfLBnodes];
+ G.g[DIR_00M] = &G6[DIR_00P * numberOfLBnodes];
+ G.g[DIR_00P] = &G6[DIR_00M * numberOfLBnodes];
}
//////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
//set Pointer for Buffer Gs
Distributions6 Dbuff;
- Dbuff.g[DIR_P00] = &bufferGs[DIR_P00 *buffmax];
- Dbuff.g[DIR_M00] = &bufferGs[DIR_M00 *buffmax];
- Dbuff.g[DIR_0P0] = &bufferGs[DIR_0P0 *buffmax];
- Dbuff.g[DIR_0M0] = &bufferGs[DIR_0M0 *buffmax];
- Dbuff.g[DIR_00P] = &bufferGs[DIR_00P *buffmax];
- Dbuff.g[DIR_00M] = &bufferGs[DIR_00M *buffmax];
+ Dbuff.g[DIR_P00] = &bufferGs[DIR_P00 * buffmax];
+ Dbuff.g[DIR_M00] = &bufferGs[DIR_M00 * buffmax];
+ Dbuff.g[DIR_0P0] = &bufferGs[DIR_0P0 * buffmax];
+ Dbuff.g[DIR_0M0] = &bufferGs[DIR_0M0 * buffmax];
+ Dbuff.g[DIR_00P] = &bufferGs[DIR_00P * buffmax];
+ Dbuff.g[DIR_00M] = &bufferGs[DIR_00M * buffmax];
//////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
//write buffer to Gs
(G.g[DIR_M00])[kw] = (Dbuff.g[DIR_P00])[k];
diff --git a/src/gpu/VirtualFluids_GPU/GPU/GPU_Interface.h b/src/gpu/VirtualFluids_GPU/GPU/GPU_Interface.h
index ee987ae23402ef304220349db77084cc341ccd5a..ae8cbb77ec2493126d64b90a7119cbfa3efee666 100644
--- a/src/gpu/VirtualFluids_GPU/GPU/GPU_Interface.h
+++ b/src/gpu/VirtualFluids_GPU/GPU/GPU_Interface.h
@@ -29,7 +29,7 @@ void KernelCas27(unsigned int grid_nx,
unsigned int* neighborY,
unsigned int* neighborZ,
real* DD,
- int size_Mat,
+ unsigned long long numberOfLBnodes,
bool EvenOrOdd);
void KernelCasSP27(unsigned int numberOfThreads,
@@ -39,7 +39,7 @@ void KernelCasSP27(unsigned int numberOfThreads,
unsigned int* neighborY,
unsigned int* neighborZ,
real* DD,
- int size_Mat,
+ unsigned long long numberOfLBnodes,
bool EvenOrOdd);
void KernelCasSPMS27(unsigned int numberOfThreads,
@@ -49,7 +49,7 @@ void KernelCasSPMS27(unsigned int numberOfThreads,
unsigned int* neighborY,
unsigned int* neighborZ,
real* DD,
- int size_Mat,
+ unsigned long long numberOfLBnodes,
bool EvenOrOdd);
void KernelCasSPMSOHM27( unsigned int numberOfThreads,
@@ -59,7 +59,7 @@ void KernelCasSPMSOHM27( unsigned int numberOfThreads,
unsigned int* neighborY,
unsigned int* neighborZ,
real* DD,
- int size_Mat,
+ unsigned long long numberOfLBnodes,
bool EvenOrOdd);
void KernelKumCompSRTSP27(
@@ -70,7 +70,7 @@ void KernelKumCompSRTSP27(
unsigned int* neighborY,
unsigned int* neighborZ,
real* DDStart,
- int size_Mat,
+ unsigned long long numberOfLBnodes,
int level,
real* forces,
bool EvenOrOdd);
@@ -82,7 +82,7 @@ void KernelCumulantD3Q27All4(unsigned int numberOfThreads,
unsigned int* neighborY,
unsigned int* neighborZ,
real* DD,
- int size_Mat,
+ unsigned long long numberOfLBnodes,
int level,
real* forces,
bool EvenOrOdd);
@@ -94,7 +94,7 @@ void KernelKumAA2016CompBulkSP27(unsigned int numberOfThreads,
unsigned int* neighborY,
unsigned int* neighborZ,
real* DD,
- int size_Mat,
+ unsigned long long numberOfLBnodes,
int size_Array,
int level,
real* forces,
@@ -112,7 +112,7 @@ void KernelKum1hSP27( unsigned int numberOfThreads,
real* coordY,
real* coordZ,
real* DDStart,
- int size_Mat,
+ unsigned long long numberOfLBnodes,
bool EvenOrOdd);
void KernelCascadeSP27(unsigned int numberOfThreads,
@@ -122,7 +122,7 @@ void KernelCascadeSP27(unsigned int numberOfThreads,
unsigned int* neighborY,
unsigned int* neighborZ,
real* DD,
- int size_Mat,
+ unsigned long long numberOfLBnodes,
bool EvenOrOdd);
void KernelKumNewSP27( unsigned int numberOfThreads,
@@ -132,7 +132,7 @@ void KernelKumNewSP27( unsigned int numberOfThreads,
unsigned int* neighborY,
unsigned int* neighborZ,
real* DD,
- int size_Mat,
+ unsigned long long numberOfLBnodes,
bool EvenOrOdd);
@@ -144,7 +144,7 @@ void CumulantOnePreconditionedErrorDiffusionChimCompSP27(
unsigned int* neighborY,
unsigned int* neighborZ,
real* DD,
- int size_Mat,
+ unsigned long long numberOfLBnodes,
int size_Array,
int level,
real* forces,
@@ -158,7 +158,7 @@ void CumulantOnePreconditionedChimCompSP27(
unsigned int* neighborY,
unsigned int* neighborZ,
real* DD,
- int size_Mat,
+ unsigned long long numberOfLBnodes,
int size_Array,
int level,
real* forces,
@@ -172,7 +172,7 @@ void CumulantOneChimCompSP27(
unsigned int* neighborY,
unsigned int* neighborZ,
real* DD,
- int size_Mat,
+ unsigned long long numberOfLBnodes,
int size_Array,
int level,
real* forces,
@@ -189,7 +189,7 @@ void KernelKumIsoTestSP27(unsigned int numberOfThreads,
real* dxxUx,
real* dyyUy,
real* dzzUz,
- int size_Mat,
+ unsigned long long numberOfLBnodes,
bool EvenOrOdd);
void KernelKumCompSP27( unsigned int numberOfThreads,
@@ -199,7 +199,7 @@ void KernelKumCompSP27( unsigned int numberOfThreads,
unsigned int* neighborY,
unsigned int* neighborZ,
real* DD,
- int size_Mat,
+ unsigned long long numberOfLBnodes,
bool EvenOrOdd);
void KernelWaleBySoniMalavCumAA2016CompSP27(
@@ -215,7 +215,7 @@ void KernelWaleBySoniMalavCumAA2016CompSP27(
real* veloZ,
real* DD,
real* turbulentViscosity,
- int size_Mat,
+ unsigned long long numberOfLBnodes,
int size_Array,
int level,
real* forces,
@@ -227,7 +227,7 @@ void KernelPMCumOneCompSP27(unsigned int numberOfThreads,
unsigned int* neighborY,
unsigned int* neighborZ,
real* DD,
- int size_Mat,
+ unsigned long long numberOfLBnodes,
int level,
real* forces,
real porosity,
@@ -245,7 +245,7 @@ void KernelADincomp7( unsigned int numberOfThreads,
unsigned int* neighborZ,
real* DD,
real* DD7,
- int size_Mat,
+ unsigned long long numberOfLBnodes,
bool EvenOrOdd);
void KernelADincomp27( unsigned int numberOfThreads,
@@ -256,7 +256,7 @@ void KernelADincomp27( unsigned int numberOfThreads,
unsigned int* neighborZ,
real* DD,
real* DD7,
- int size_Mat,
+ unsigned long long numberOfLBnodes,
bool EvenOrOdd);
void Init27(int myid,
@@ -267,7 +267,7 @@ void Init27(int myid,
unsigned int* neighborY,
unsigned int* neighborZ,
real* vParab,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
unsigned int grid_nx,
unsigned int grid_ny,
unsigned int grid_nz,
@@ -285,7 +285,7 @@ void InitNonEqPartSP27(unsigned int numberOfThreads,
real* ux,
real* uy,
real* uz,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
real* DD,
real omega,
bool EvenOrOdd);
@@ -300,7 +300,7 @@ void InitThS7( unsigned int numberOfThreads,
real* ux,
real* uy,
real* uz,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
real* DD7,
bool EvenOrOdd);
@@ -313,7 +313,7 @@ void InitADDev27( unsigned int numberOfThreads,
real* ux,
real* uy,
real* uz,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
real* DD27,
bool EvenOrOdd);
@@ -330,7 +330,7 @@ void PostProcessorF3_2018Fehlberg(
real* vzOut,
real* DDStart,
real* G6,
- int size_Mat,
+ unsigned long long numberOfLBnodes,
int level,
real* forces,
bool EvenOrOdd);
@@ -343,7 +343,7 @@ void CalcMac27( real* vxD,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
unsigned int grid_nx,
unsigned int grid_ny,
unsigned int grid_nz,
@@ -359,7 +359,7 @@ void CalcMacSP27(real* vxD,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
unsigned int numberOfThreads,
real* DD,
bool isEvenTimestep);
@@ -373,7 +373,7 @@ void CalcMacCompSP27(real* vxD,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
unsigned int numberOfThreads,
real* DD,
bool isEvenTimestep);
@@ -383,7 +383,7 @@ void CalcMacThS7( real* Conc,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
unsigned int numberOfThreads,
real* DD7,
bool isEvenTimestep);
@@ -395,7 +395,7 @@ void PlaneConcThS7(real* Conc,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
unsigned int numberOfThreads,
real* DD7,
bool isEvenTimestep);
@@ -407,7 +407,7 @@ void PlaneConcThS27(real* Conc,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
unsigned int numberOfThreads,
real* DD27,
bool isEvenTimestep);
@@ -418,7 +418,7 @@ void CalcConcentration27( unsigned int numberOfThreads,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
real* DD27,
bool isEvenTimestep);
@@ -431,7 +431,7 @@ void CalcMedSP27( real* vxD,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
unsigned int numberOfThreads,
real* DD,
bool isEvenTimestep);
@@ -445,7 +445,7 @@ void CalcMedCompSP27(real* vxD,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
unsigned int numberOfThreads,
real* DD,
bool isEvenTimestep);
@@ -461,7 +461,7 @@ void CalcMedCompAD27(
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
unsigned int numberOfThreads,
real* DD,
real* DD_AD,
@@ -477,7 +477,7 @@ void CalcMacMedSP27( real* vxD,
unsigned int* neighborY,
unsigned int* neighborZ,
unsigned int tdiff,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
unsigned int numberOfThreads,
bool isEvenTimestep);
@@ -487,7 +487,7 @@ void ResetMedianValuesSP27(
real* vzD,
real* rhoD,
real* pressD,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
unsigned int numberOfThreads,
bool isEvenTimestep);
@@ -498,7 +498,7 @@ void ResetMedianValuesAD27(
real* rhoD,
real* pressD,
real* concD,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
unsigned int numberOfThreads,
bool isEvenTimestep);
@@ -511,7 +511,7 @@ void Calc2ndMomentsIncompSP27(real* kxyFromfcNEQ,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
unsigned int numberOfThreads,
real* DD,
bool isEvenTimestep);
@@ -525,7 +525,7 @@ void Calc2ndMomentsCompSP27(real* kxyFromfcNEQ,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
unsigned int numberOfThreads,
real* DD,
bool isEvenTimestep);
@@ -541,7 +541,7 @@ void Calc3rdMomentsIncompSP27(real* CUMbbb,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
unsigned int numberOfThreads,
real* DD,
bool isEvenTimestep);
@@ -557,7 +557,7 @@ void Calc3rdMomentsCompSP27(real* CUMbbb,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
unsigned int numberOfThreads,
real* DD,
bool isEvenTimestep);
@@ -576,7 +576,7 @@ void CalcHigherMomentsIncompSP27(real* CUMcbb,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
unsigned int numberOfThreads,
real* DD,
bool isEvenTimestep);
@@ -595,7 +595,7 @@ void CalcHigherMomentsCompSP27(real* CUMcbb,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
unsigned int numberOfThreads,
real* DD,
bool isEvenTimestep);
@@ -612,7 +612,7 @@ void LBCalcMeasurePoints27(real* vxMP,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
real* DD,
unsigned int numberOfThreads,
bool isEvenTimestep);
@@ -627,7 +627,7 @@ void BcPress27(int nx,
unsigned int* neighborY,
unsigned int* neighborZ,
real* DD,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep);
void BcVel27(int nx,
@@ -641,7 +641,7 @@ void BcVel27(int nx,
unsigned int* neighborY,
unsigned int* neighborZ,
real* DD,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep,
real u0x,
real om);
@@ -661,7 +661,7 @@ void QDevCompThinWalls27(unsigned int numberOfThreads,
unsigned int* neighborY,
unsigned int* neighborZ,
unsigned int* neighborWSB,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep);
void QDev3rdMomentsComp27(LBMSimulationParameter* parameterDevice, QforBoundaryConditions* boundaryCondition);
@@ -675,7 +675,7 @@ void QDevIncompHighNu27( unsigned int numberOfThreads,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep);
void QDevCompHighNu27(unsigned int numberOfThreads,
@@ -687,7 +687,7 @@ void QDevCompHighNu27(unsigned int numberOfThreads,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep);
void QVelDevicePlainBB27(LBMSimulationParameter* parameterDevice, QforBoundaryConditions* boundaryCondition);
@@ -704,7 +704,7 @@ void QVelDeviceCouette27(unsigned int numberOfThreads,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep);
void QVelDevice1h27( unsigned int numberOfThreads,
@@ -726,7 +726,7 @@ void QVelDevice1h27( unsigned int numberOfThreads,
real* coordX,
real* coordY,
real* coordZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep);
void QVelDev27(LBMSimulationParameter* parameterDevice, QforBoundaryConditions* boundaryCondition);
@@ -743,7 +743,7 @@ void QVelDevCompPlusSlip27(unsigned int numberOfThreads,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep);
void QVelDevComp27(LBMSimulationParameter* parameterDevice, QforBoundaryConditions* boundaryCondition);
@@ -762,7 +762,7 @@ void QVelDevCompThinWalls27(unsigned int numberOfThreads,
unsigned int* neighborY,
unsigned int* neighborZ,
unsigned int* neighborWSB,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep);
void QVelDevCompZeroPress27(LBMSimulationParameter* parameterDevice, QforBoundaryConditions* boundaryCondition);
@@ -779,7 +779,7 @@ void QVelDevIncompHighNu27( unsigned int numberOfThreads,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep);
void QVelDevCompHighNu27(unsigned int numberOfThreads,
@@ -794,7 +794,7 @@ void QVelDevCompHighNu27(unsigned int numberOfThreads,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep);
void QVeloDevEQ27(unsigned int numberOfThreads,
@@ -808,7 +808,7 @@ void QVeloDevEQ27(unsigned int numberOfThreads,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep);
void QVeloStreetDevEQ27(
@@ -848,7 +848,7 @@ void QSlipGeomDevComp27( unsigned int numberOfThreads,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep);
void QSlipNormDevComp27(unsigned int numberOfThreads,
@@ -863,7 +863,7 @@ void QSlipNormDevComp27(unsigned int numberOfThreads,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep);
void QStressDevComp27(Parameter *para, QforBoundaryConditions* boundaryCondition, const int level);
@@ -883,7 +883,7 @@ void QPressDevFixBackflow27(unsigned int numberOfThreads,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep);
void QPressDevDirDepBot27(unsigned int numberOfThreads,
@@ -895,11 +895,13 @@ void QPressDevDirDepBot27(unsigned int numberOfThreads,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep);
void QPressNoRhoDev27(LBMSimulationParameter* parameterDevice, QforBoundaryConditions* boundaryCondition);
+void QPressZeroRhoOutflowDev27(LBMSimulationParameter* parameterDevice, QforBoundaryConditions* boundaryCondition);
+
void QInflowScaleByPressDev27(LBMSimulationParameter* parameterDevice, QforBoundaryConditions* boundaryCondition);
void QPressDevOld27(unsigned int numberOfThreads,
@@ -912,7 +914,7 @@ void QPressDevOld27(unsigned int numberOfThreads,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep);
void QPressDevIncompNEQ27(LBMSimulationParameter* parameterDevice, QforBoundaryConditions* boundaryCondition);
@@ -928,7 +930,7 @@ void QPressDevZero27(unsigned int numberOfThreads,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep);
void QPressDevFake27( unsigned int numberOfThreads,
@@ -941,7 +943,7 @@ void QPressDevFake27( unsigned int numberOfThreads,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep);
void BBDev27(LBMSimulationParameter* parameterDevice, QforBoundaryConditions* boundaryCondition);
@@ -956,7 +958,7 @@ void QPressDev27_IntBB( unsigned int numberOfThreads,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep);
void QPressDevAntiBB27( unsigned int numberOfThreads,
@@ -972,7 +974,7 @@ void QPressDevAntiBB27( unsigned int numberOfThreads,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep);
void PressSchlaffer27(unsigned int numberOfThreads,
@@ -989,7 +991,7 @@ void PressSchlaffer27(unsigned int numberOfThreads,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep);
void VelSchlaffer27( unsigned int numberOfThreads,
@@ -1004,9 +1006,17 @@ void VelSchlaffer27( unsigned int numberOfThreads,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep);
+void QPrecursorDevCompZeroPress(LBMSimulationParameter* parameterDevice, QforPrecursorBoundaryConditions* boundaryCondition, real tRatio, real velocityRatio);
+
+void PrecursorDevEQ27(LBMSimulationParameter* parameterDevice, QforPrecursorBoundaryConditions* boundaryCondition, real tRatio, real velocityRatio);
+
+void PrecursorDevDistributions(LBMSimulationParameter* parameterDevice, QforPrecursorBoundaryConditions* boundaryCondition, real tRatio, real velocityRatio);
+
+void QPrecursorDevDistributions(LBMSimulationParameter* parameterDevice, QforPrecursorBoundaryConditions* boundaryCondition, real tRatio, real velocityRatio);
+
void QADDev7(unsigned int numberOfThreads,
real* DD,
real* DD7,
@@ -1019,7 +1029,7 @@ void QADDev7(unsigned int numberOfThreads,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep);
//////////////////////////////////////////////////////////////////////////
@@ -1033,7 +1043,7 @@ void FactorizedCentralMomentsAdvectionDiffusionDeviceKernel(
uint* neighborZ,
real* distributions,
real* distributionsAD,
- int size_Mat,
+ unsigned long long numberOfLBnodes,
real* forces,
bool isEvenTimestep);
@@ -1053,7 +1063,7 @@ void ADSlipVelDevComp(
uint * neighborX,
uint * neighborY,
uint * neighborZ,
- uint size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep);
void QADDirichletDev27( unsigned int numberOfThreads,
@@ -1068,7 +1078,7 @@ void QADDirichletDev27( unsigned int numberOfThreads,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep);
void QADBBDev27( unsigned int numberOfThreads,
@@ -1083,7 +1093,7 @@ void QADBBDev27( unsigned int numberOfThreads,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep);
void QADVelDev7(unsigned int numberOfThreads,
@@ -1099,7 +1109,7 @@ void QADVelDev7(unsigned int numberOfThreads,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep);
@@ -1116,7 +1126,7 @@ void QADVelDev27( unsigned int numberOfThreads,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep);
void QADPressDev7( unsigned int numberOfThreads,
@@ -1132,7 +1142,7 @@ void QADPressDev7( unsigned int numberOfThreads,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep);
void QADPressDev27(unsigned int numberOfThreads,
@@ -1148,7 +1158,7 @@ void QADPressDev27(unsigned int numberOfThreads,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep);
void QADPressNEQNeighborDev27(
@@ -1161,7 +1171,7 @@ void QADPressNEQNeighborDev27(
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep
);
@@ -1177,7 +1187,7 @@ void QNoSlipADincompDev7(unsigned int numberOfThreads,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep);
void QNoSlipADincompDev27(unsigned int numberOfThreads,
@@ -1192,7 +1202,7 @@ void QNoSlipADincompDev27(unsigned int numberOfThreads,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep);
void QADVeloIncompDev7( unsigned int numberOfThreads,
@@ -1208,7 +1218,7 @@ void QADVeloIncompDev7( unsigned int numberOfThreads,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep);
@@ -1225,7 +1235,7 @@ void QADVeloIncompDev27( unsigned int numberOfThreads,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep);
void QADPressIncompDev7( unsigned int numberOfThreads,
@@ -1241,7 +1251,7 @@ void QADPressIncompDev7( unsigned int numberOfThreads,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep);
void QADPressIncompDev27( unsigned int numberOfThreads,
@@ -1257,7 +1267,7 @@ void QADPressIncompDev27( unsigned int numberOfThreads,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep);
void PropVelo( unsigned int numberOfThreads,
@@ -1270,7 +1280,7 @@ void PropVelo( unsigned int numberOfThreads,
real* uz,
int* k_Q,
unsigned int size_Prop,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
unsigned int* bcMatD,
real* DD,
bool EvenOrOdd);
@@ -1283,8 +1293,8 @@ void ScaleCF27( real* DC,
unsigned int* neighborFX,
unsigned int* neighborFY,
unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
bool isEvenTimestep,
unsigned int* posCSWB,
unsigned int* posFSWB,
@@ -1306,8 +1316,8 @@ void ScaleFC27( real* DC,
unsigned int* neighborFX,
unsigned int* neighborFY,
unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
bool isEvenTimestep,
unsigned int* posC,
unsigned int* posFSWB,
@@ -1329,8 +1339,8 @@ void ScaleCFEff27(real* DC,
unsigned int* neighborFX,
unsigned int* neighborFY,
unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
bool isEvenTimestep,
unsigned int* posCSWB,
unsigned int* posFSWB,
@@ -1353,8 +1363,8 @@ void ScaleFCEff27(real* DC,
unsigned int* neighborFX,
unsigned int* neighborFY,
unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
bool isEvenTimestep,
unsigned int* posC,
unsigned int* posFSWB,
@@ -1377,8 +1387,8 @@ void ScaleCFLast27(real* DC,
unsigned int* neighborFX,
unsigned int* neighborFY,
unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
bool isEvenTimestep,
unsigned int* posCSWB,
unsigned int* posFSWB,
@@ -1401,8 +1411,8 @@ void ScaleFCLast27(real* DC,
unsigned int* neighborFX,
unsigned int* neighborFY,
unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
bool isEvenTimestep,
unsigned int* posC,
unsigned int* posFSWB,
@@ -1425,8 +1435,8 @@ void ScaleCFpress27(real* DC,
unsigned int* neighborFX,
unsigned int* neighborFY,
unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
bool isEvenTimestep,
unsigned int* posCSWB,
unsigned int* posFSWB,
@@ -1449,8 +1459,8 @@ void ScaleFCpress27( real* DC,
unsigned int* neighborFX,
unsigned int* neighborFY,
unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
bool isEvenTimestep,
unsigned int* posC,
unsigned int* posFSWB,
@@ -1473,8 +1483,8 @@ void ScaleCF_Fix_27(real* DC,
unsigned int* neighborFX,
unsigned int* neighborFY,
unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
bool isEvenTimestep,
unsigned int* posCSWB,
unsigned int* posFSWB,
@@ -1497,8 +1507,8 @@ void ScaleCF_Fix_comp_27( real* DC,
unsigned int* neighborFX,
unsigned int* neighborFY,
unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
bool isEvenTimestep,
unsigned int* posCSWB,
unsigned int* posFSWB,
@@ -1521,8 +1531,8 @@ void ScaleCF_0817_comp_27( real* DC,
unsigned int* neighborFX,
unsigned int* neighborFY,
unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
bool isEvenTimestep,
unsigned int* posCSWB,
unsigned int* posFSWB,
@@ -1547,8 +1557,8 @@ void ScaleCF_comp_D3Q27F3_2018( real* DC,
unsigned int* neighborFX,
unsigned int* neighborFY,
unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
bool isEvenTimestep,
unsigned int* posCSWB,
unsigned int* posFSWB,
@@ -1572,8 +1582,8 @@ void ScaleCF_comp_D3Q27F3(real* DC,
unsigned int* neighborFX,
unsigned int* neighborFY,
unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
bool isEvenTimestep,
unsigned int* posCSWB,
unsigned int* posFSWB,
@@ -1597,8 +1607,8 @@ void ScaleCF_staggered_time_comp_27( real* DC,
unsigned int* neighborFX,
unsigned int* neighborFY,
unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
bool isEvenTimestep,
unsigned int* posCSWB,
unsigned int* posFSWB,
@@ -1624,8 +1634,8 @@ void ScaleCF_RhoSq_3rdMom_comp_27( real* DC,
unsigned int* neighborFX,
unsigned int* neighborFY,
unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
bool isEvenTimestep,
unsigned int* posCSWB,
unsigned int* posFSWB,
@@ -1649,8 +1659,8 @@ void ScaleCF_AA2016_comp_27( real* DC,
unsigned int* neighborFX,
unsigned int* neighborFY,
unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
bool isEvenTimestep,
unsigned int* posCSWB,
unsigned int* posFSWB,
@@ -1674,8 +1684,8 @@ void ScaleCF_NSPress_27(real* DC,
unsigned int* neighborFX,
unsigned int* neighborFY,
unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
bool isEvenTimestep,
unsigned int* posCSWB,
unsigned int* posFSWB,
@@ -1698,8 +1708,8 @@ void ScaleFC_Fix_27( real* DC,
unsigned int* neighborFX,
unsigned int* neighborFY,
unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
bool isEvenTimestep,
unsigned int* posC,
unsigned int* posFSWB,
@@ -1722,8 +1732,8 @@ void ScaleFC_Fix_comp_27( real* DC,
unsigned int* neighborFX,
unsigned int* neighborFY,
unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
bool isEvenTimestep,
unsigned int* posC,
unsigned int* posFSWB,
@@ -1746,8 +1756,8 @@ void ScaleFC_0817_comp_27( real* DC,
unsigned int* neighborFX,
unsigned int* neighborFY,
unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
bool isEvenTimestep,
unsigned int* posC,
unsigned int* posFSWB,
@@ -1772,8 +1782,8 @@ void ScaleFC_comp_D3Q27F3_2018(real* DC,
unsigned int* neighborFX,
unsigned int* neighborFY,
unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
bool isEvenTimestep,
unsigned int* posC,
unsigned int* posFSWB,
@@ -1797,8 +1807,8 @@ void ScaleFC_comp_D3Q27F3( real* DC,
unsigned int* neighborFX,
unsigned int* neighborFY,
unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
bool isEvenTimestep,
unsigned int* posC,
unsigned int* posFSWB,
@@ -1822,8 +1832,8 @@ void ScaleFC_staggered_time_comp_27( real* DC,
unsigned int* neighborFX,
unsigned int* neighborFY,
unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
bool isEvenTimestep,
unsigned int* posC,
unsigned int* posFSWB,
@@ -1849,8 +1859,8 @@ void ScaleFC_RhoSq_3rdMom_comp_27( real* DC,
unsigned int* neighborFX,
unsigned int* neighborFY,
unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
bool isEvenTimestep,
unsigned int* posC,
unsigned int* posFSWB,
@@ -1874,8 +1884,8 @@ void ScaleFC_AA2016_comp_27( real* DC,
unsigned int* neighborFX,
unsigned int* neighborFY,
unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
bool isEvenTimestep,
unsigned int* posC,
unsigned int* posFSWB,
@@ -1899,8 +1909,8 @@ void ScaleFC_NSPress_27( real* DC,
unsigned int* neighborFX,
unsigned int* neighborFY,
unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
bool isEvenTimestep,
unsigned int* posC,
unsigned int* posFSWB,
@@ -1925,8 +1935,8 @@ void ScaleCFThS7( real* DC,
unsigned int* neighborFX,
unsigned int* neighborFY,
unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
bool isEvenTimestep,
unsigned int* posCSWB,
unsigned int* posFSWB,
@@ -1945,8 +1955,8 @@ void ScaleFCThS7( real* DC,
unsigned int* neighborFX,
unsigned int* neighborFY,
unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
bool isEvenTimestep,
unsigned int* posC,
unsigned int* posFSWB,
@@ -1965,8 +1975,8 @@ void ScaleCFThSMG7( real* DC,
unsigned int* neighborFX,
unsigned int* neighborFY,
unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
bool isEvenTimestep,
unsigned int* posCSWB,
unsigned int* posFSWB,
@@ -1986,8 +1996,8 @@ void ScaleFCThSMG7(real* DC,
unsigned int* neighborFX,
unsigned int* neighborFY,
unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
bool isEvenTimestep,
unsigned int* posC,
unsigned int* posFSWB,
@@ -2007,8 +2017,8 @@ void ScaleCFThS27( real* DC,
unsigned int* neighborFX,
unsigned int* neighborFY,
unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
bool isEvenTimestep,
unsigned int* posCSWB,
unsigned int* posFSWB,
@@ -2028,8 +2038,8 @@ void ScaleFCThS27( real* DC,
unsigned int* neighborFX,
unsigned int* neighborFY,
unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
bool isEvenTimestep,
unsigned int* posC,
unsigned int* posFSWB,
@@ -2049,7 +2059,7 @@ void DragLiftPostD27(real* DD,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep,
unsigned int numberOfThreads);
@@ -2063,7 +2073,7 @@ void DragLiftPreD27( real* DD,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep,
unsigned int numberOfThreads);
@@ -2074,7 +2084,7 @@ void CalcCPtop27(real* DD,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep,
unsigned int numberOfThreads);
@@ -2085,7 +2095,7 @@ void CalcCPbottom27(real* DD,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep,
unsigned int numberOfThreads);
@@ -2096,7 +2106,7 @@ void GetSendFsPreDev27(real* DD,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep,
unsigned int numberOfThreads,
cudaStream_t stream = CU_STREAM_LEGACY);
@@ -2108,7 +2118,7 @@ void GetSendFsPostDev27(real* DD,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep,
unsigned int numberOfThreads,
cudaStream_t stream = CU_STREAM_LEGACY);
@@ -2120,7 +2130,7 @@ void SetRecvFsPreDev27(real* DD,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep, unsigned int numberOfThreads,
cudaStream_t stream = CU_STREAM_LEGACY);
@@ -2131,7 +2141,7 @@ void SetRecvFsPostDev27(real* DD,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep,
unsigned int numberOfThreads,
cudaStream_t stream = CU_STREAM_LEGACY);
@@ -2144,7 +2154,7 @@ void getSendGsDevF3(
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep,
unsigned int numberOfThreads);
@@ -2156,7 +2166,7 @@ void setRecvGsDevF3(
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep,
unsigned int numberOfThreads);
@@ -2172,7 +2182,7 @@ void WallFuncDev27(unsigned int numberOfThreads,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep);
void SetOutputWallVelocitySP27(unsigned int numberOfThreads,
@@ -2190,7 +2200,7 @@ void SetOutputWallVelocitySP27(unsigned int numberOfThreads,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
real* DD,
bool isEvenTimestep);
@@ -2204,7 +2214,7 @@ void GetVelotoForce27(unsigned int numberOfThreads,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep);
void InitParticlesDevice(real* coordX,
@@ -2229,7 +2239,7 @@ void InitParticlesDevice(real* coordX,
unsigned int* neighborWSB,
int level,
unsigned int numberOfParticles,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
unsigned int numberOfThreads);
void MoveParticlesDevice(real* coordX,
@@ -2257,16 +2267,16 @@ void MoveParticlesDevice(real* coordX,
unsigned int timestep,
unsigned int numberOfTimesteps,
unsigned int numberOfParticles,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
unsigned int numberOfThreads,
bool isEvenTimestep);
void initRandomDevice(curandState* state,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
unsigned int numberOfThreads);
void generateRandomValuesDevice(curandState* state,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
real* randArray,
unsigned int numberOfThreads);
@@ -2285,7 +2295,7 @@ void CalcTurbulenceIntensityDevice(
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep,
uint numberOfThreads);
diff --git a/src/gpu/VirtualFluids_GPU/GPU/GPU_Kernels.cuh b/src/gpu/VirtualFluids_GPU/GPU/GPU_Kernels.cuh
index 94b9704b7ca57df4cd985f5aff9521b8a087b97f..3134db44346ee7f465a5c8f04505ee5749482fbf 100644
--- a/src/gpu/VirtualFluids_GPU/GPU/GPU_Kernels.cuh
+++ b/src/gpu/VirtualFluids_GPU/GPU/GPU_Kernels.cuh
@@ -22,7 +22,7 @@ __global__ void LB_Kernel_Casc27(real s9,
unsigned int* neighborY,
unsigned int* neighborZ,
real* DDStart,
- int size_Mat,
+ unsigned long long numberOfLBnodes,
bool EvenOrOdd);
__global__ void LB_Kernel_Casc_SP_27( real s9,
@@ -31,7 +31,7 @@ __global__ void LB_Kernel_Casc_SP_27( real s9,
unsigned int* neighborY,
unsigned int* neighborZ,
real* DDStart,
- int size_Mat,
+ unsigned long long numberOfLBnodes,
bool EvenOrOdd);
__global__ void LB_Kernel_Casc_SP_MS_27( real s9,
@@ -40,7 +40,7 @@ __global__ void LB_Kernel_Casc_SP_MS_27( real s9,
unsigned int* neighborY,
unsigned int* neighborZ,
real* DDStart,
- int size_Mat,
+ unsigned long long numberOfLBnodes,
bool EvenOrOdd);
__global__ void LB_Kernel_Casc_SP_MS_OHM_27( real s9,
@@ -49,134 +49,134 @@ __global__ void LB_Kernel_Casc_SP_MS_OHM_27( real s9,
unsigned int* neighborY,
unsigned int* neighborZ,
real* DDStart,
- int size_Mat,
+ unsigned long long numberOfLBnodes,
bool EvenOrOdd);
__global__ void LB_Kernel_Kum_New_Comp_SRT_SP_27(
- real omega,
- unsigned int* bcMatD,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- real* DDStart,
- int size_Mat,
- int level,
- real* forces,
- bool EvenOrOdd);
+ real omega,
+ unsigned int* bcMatD,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ real* DDStart,
+ unsigned long long numberOfLBnodes,
+ int level,
+ real* forces,
+ bool EvenOrOdd);
__global__ void LB_Kernel_Cumulant_D3Q27All4(real omega,
- unsigned int* bcMatD,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- real* DDStart,
- int size_Mat,
- int level,
- real* forces,
- bool EvenOrOdd);
+ unsigned int* bcMatD,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ real* DDStart,
+ unsigned long long numberOfLBnodes,
+ int level,
+ real* forces,
+ bool EvenOrOdd);
__global__ void LB_Kernel_Kum_AA2016_Comp_Bulk_SP_27(real omega,
- unsigned int* bcMatD,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- real* DDStart,
- int size_Mat,
- int level,
- real* forces,
- bool EvenOrOdd);
+ unsigned int* bcMatD,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ real* DDStart,
+ unsigned long long numberOfLBnodes,
+ int level,
+ real* forces,
+ bool EvenOrOdd);
__global__ void LB_Kernel_Kum_1h_SP_27( real omega,
- real deltaPhi,
- real angularVelocity,
- unsigned int* bcMatD,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- real* coordX,
- real* coordY,
- real* coordZ,
- real* DDStart,
- int size_Mat,
- bool EvenOrOdd);
+ real deltaPhi,
+ real angularVelocity,
+ unsigned int* bcMatD,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ real* coordX,
+ real* coordY,
+ real* coordZ,
+ real* DDStart,
+ unsigned long long numberOfLBnodes,
+ bool EvenOrOdd);
__global__ void LB_Kernel_Cascade_SP_27( real s9,
- unsigned int* bcMatD,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- real* DDStart,
- int size_Mat,
- bool EvenOrOdd);
+ unsigned int* bcMatD,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ real* DDStart,
+ unsigned long long numberOfLBnodes,
+ bool EvenOrOdd);
__global__ void LB_Kernel_Kum_New_SP_27( real s9,
- unsigned int* bcMatD,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- real* DDStart,
- int size_Mat,
- bool EvenOrOdd);
+ unsigned int* bcMatD,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ real* DDStart,
+ unsigned long long numberOfLBnodes,
+ bool EvenOrOdd);
__global__ void LB_Kernel_Kum_IsoTest_SP_27( real omega,
- unsigned int* bcMatD,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- real* DDStart,
- real* dxxUx,
- real* dyyUy,
- real* dzzUz,
- int size_Mat,
- bool EvenOrOdd);
+ unsigned int* bcMatD,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ real* DDStart,
+ real* dxxUx,
+ real* dyyUy,
+ real* dzzUz,
+ unsigned long long numberOfLBnodes,
+ bool EvenOrOdd);
__global__ void LB_Kernel_Kum_Comp_SP_27(real s9,
- unsigned int* bcMatD,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- real* DDStart,
- int size_Mat,
- bool EvenOrOdd);
+ unsigned int* bcMatD,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ real* DDStart,
+ unsigned long long numberOfLBnodes,
+ bool EvenOrOdd);
__global__ void Cumulant_One_preconditioned_errorDiffusion_chim_Comp_SP_27(
- real omega,
- unsigned int* bcMatD,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- real* DDStart,
- int size_Mat,
- int level,
- real* forces,
- bool EvenOrOdd);
+ real omega,
+ unsigned int* bcMatD,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ real* DDStart,
+ unsigned long long numberOfLBnodes,
+ int level,
+ real* forces,
+ bool EvenOrOdd);
__global__ void Cumulant_One_preconditioned_chim_Comp_SP_27(
- real omega,
- unsigned int* bcMatD,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- real* DDStart,
- int size_Mat,
- int level,
- real* forces,
- bool EvenOrOdd);
+ real omega,
+ unsigned int* bcMatD,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ real* DDStart,
+ unsigned long long numberOfLBnodes,
+ int level,
+ real* forces,
+ bool EvenOrOdd);
__global__ void Cumulant_One_chim_Comp_SP_27(
- real omega,
- unsigned int* bcMatD,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- real* DDStart,
- int size_Mat,
- int level,
- real* forces,
- bool EvenOrOdd);
+ real omega,
+ unsigned int* bcMatD,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ real* DDStart,
+ unsigned long long numberOfLBnodes,
+ int level,
+ real* forces,
+ bool EvenOrOdd);
inline __device__ void forwardChimeraWithK(real &mfa, real &mfb, real &mfc, real vv, real v2, real K);
@@ -189,57 +189,57 @@ inline __device__ void backwardChimeraWithK(real &mfa, real &mfb, real &mfc, rea
__global__ void LB_Kernel_WaleBySoniMalav_Cum_AA2016_Comp_SP_27(
- real omega_in,
- unsigned int* bcMatD,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int* neighborWSB,
- real* veloX,
- real* veloY,
- real* veloZ,
- real* DDStart,
- real* turbulentViscosity,
- int size_Mat,
- int level,
- real* forces,
- bool EvenOrOdd);
+ real omega_in,
+ unsigned int* bcMatD,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned int* neighborWSB,
+ real* veloX,
+ real* veloY,
+ real* veloZ,
+ real* DDStart,
+ real* turbulentViscosity,
+ unsigned long long numberOfLBnodes,
+ int level,
+ real* forces,
+ bool EvenOrOdd);
__global__ void LB_Kernel_PM_Cum_One_Comp_SP_27( real omega,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- real* DDStart,
- int size_Mat,
- int level,
- real* forces,
- real porosity,
- real darcy,
- real forchheimer,
- unsigned int sizeOfPorousMedia,
- unsigned int* nodeIdsPorousMedia,
- bool EvenOrOdd);
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ real* DDStart,
+ unsigned long long numberOfLBnodes,
+ int level,
+ real* forces,
+ real porosity,
+ real darcy,
+ real forchheimer,
+ unsigned int sizeOfPorousMedia,
+ unsigned int* nodeIdsPorousMedia,
+ bool EvenOrOdd);
__global__ void LB_Kernel_AD_Incomp_7( real diffusivity,
- unsigned int* bcMatD,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- real* DDStart,
- real* DD7,
- int size_Mat,
- bool EvenOrOdd);
+ unsigned int* bcMatD,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ real* DDStart,
+ real* DD7,
+ unsigned long long numberOfLBnodes,
+ bool EvenOrOdd);
__global__ void LB_Kernel_AD_Incomp_27( real diffusivity,
- unsigned int* bcMatD,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- real* DDStart,
- real* DD27,
- int size_Mat,
- bool EvenOrOdd);
+ unsigned int* bcMatD,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ real* DDStart,
+ real* DD27,
+ unsigned long long numberOfLBnodes,
+ bool EvenOrOdd);
__global__ void LBInit27( int myid,
int numprocs,
@@ -249,7 +249,7 @@ __global__ void LBInit27( int myid,
unsigned int* neighborY,
unsigned int* neighborZ,
real* vParabel,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
unsigned int grid_nx,
unsigned int grid_ny,
unsigned int grid_nz,
@@ -266,7 +266,7 @@ __global__ void LBInitNonEqPartSP27(unsigned int* neighborX,
real* ux,
real* uy,
real* uz,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
real* DD,
real omega,
bool EvenOrOdd);
@@ -279,7 +279,7 @@ __global__ void InitAD7( unsigned int* neighborX,
real* ux,
real* uy,
real* uz,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
real* DD7,
bool EvenOrOdd);
@@ -291,26 +291,26 @@ __global__ void InitAD27(unsigned int* neighborX,
real* ux,
real* uy,
real* uz,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
real* DD27,
bool EvenOrOdd);
__global__ void LB_PostProcessor_F3_2018_Fehlberg(
- real omega,
- unsigned int* bcMatD,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- real* rhoOut,
- real* vxOut,
- real* vyOut,
- real* vzOut,
- real* DDStart,
- real* G6,
- int size_Mat,
- int level,
- real* forces,
- bool EvenOrOdd);
+ real omega,
+ unsigned int* bcMatD,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ real* rhoOut,
+ real* vxOut,
+ real* vyOut,
+ real* vzOut,
+ real* DDStart,
+ real* G6,
+ unsigned long long numberOfLBnodes,
+ int level,
+ real* forces,
+ bool EvenOrOdd);
__global__ void LBCalcMac27( real* vxD,
real* vyD,
@@ -320,7 +320,7 @@ __global__ void LBCalcMac27( real* vxD,
unsigned int* neighborY,
unsigned int* neighborZ,
unsigned int* geoD,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
real* DD,
bool isEvenTimestep);
@@ -333,60 +333,60 @@ __global__ void LBCalcMacSP27( real* vxD,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
real* DD,
bool isEvenTimestep);
__global__ void LBCalcMacCompSP27( real* vxD,
- real* vyD,
- real* vzD,
- real* rhoD,
- real* pressD,
- unsigned int* geoD,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- real* DD,
- bool isEvenTimestep);
+ real* vyD,
+ real* vzD,
+ real* rhoD,
+ real* pressD,
+ unsigned int* geoD,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ real* DD,
+ bool isEvenTimestep);
__global__ void CalcConc7( real* Conc,
unsigned int* geoD,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
real* DD7,
bool isEvenTimestep);
__global__ void GetPlaneConc7(real* Conc,
- int* kPC,
- unsigned int numberOfPointskPC,
- unsigned int* geoD,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- real* DD7,
- bool isEvenTimestep);
+ int* kPC,
+ unsigned int numberOfPointskPC,
+ unsigned int* geoD,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ real* DD7,
+ bool isEvenTimestep);
__global__ void GetPlaneConc27(real* Conc,
- int* kPC,
- unsigned int numberOfPointskPC,
- unsigned int* geoD,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- real* DD27,
- bool isEvenTimestep);
+ int* kPC,
+ unsigned int numberOfPointskPC,
+ unsigned int* geoD,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ real* DD27,
+ bool isEvenTimestep);
__global__ void CalcConc27(real* Conc,
unsigned int* geoD,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
real* DD27,
bool isEvenTimestep);
@@ -399,38 +399,38 @@ __global__ void LBCalcMedSP27( real* vxD,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
real* DD,
bool isEvenTimestep);
__global__ void LBCalcMedCompSP27( real* vxD,
- real* vyD,
- real* vzD,
- real* rhoD,
- real* pressD,
- unsigned int* geoD,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- real* DD,
- bool isEvenTimestep);
+ real* vyD,
+ real* vzD,
+ real* rhoD,
+ real* pressD,
+ unsigned int* geoD,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ real* DD,
+ bool isEvenTimestep);
__global__ void LBCalcMedCompAD27(
- real* vxD,
- real* vyD,
- real* vzD,
- real* rhoD,
- real* pressD,
- real* concD,
- unsigned int* geoD,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- real* DD,
- real* DD_AD,
- bool isEvenTimestep);
+ real* vxD,
+ real* vyD,
+ real* vzD,
+ real* rhoD,
+ real* pressD,
+ real* concD,
+ unsigned int* geoD,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ real* DD,
+ real* DD_AD,
+ bool isEvenTimestep);
__global__ void LBCalcMacMedSP27( real* vxD,
real* vyD,
@@ -442,119 +442,119 @@ __global__ void LBCalcMacMedSP27( real* vxD,
unsigned int* neighborY,
unsigned int* neighborZ,
unsigned int tdiff,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep);
__global__ void LBResetMedianValuesSP27(
- real* vxD,
- real* vyD,
- real* vzD,
- real* rhoD,
- real* pressD,
- unsigned int size_Mat,
- bool isEvenTimestep);
+ real* vxD,
+ real* vyD,
+ real* vzD,
+ real* rhoD,
+ real* pressD,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep);
__global__ void LBResetMedianValuesAD27(
- real* vxD,
- real* vyD,
- real* vzD,
- real* rhoD,
- real* pressD,
- real* concD,
- unsigned int size_Mat,
- bool isEvenTimestep);
+ real* vxD,
+ real* vyD,
+ real* vzD,
+ real* rhoD,
+ real* pressD,
+ real* concD,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep);
__global__ void LBCalc2ndMomentsIncompSP27( real* kxyFromfcNEQ,
- real* kyzFromfcNEQ,
- real* kxzFromfcNEQ,
- real* kxxMyyFromfcNEQ,
- real* kxxMzzFromfcNEQ,
- unsigned int* geoD,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- real* DD,
- bool isEvenTimestep);
+ real* kyzFromfcNEQ,
+ real* kxzFromfcNEQ,
+ real* kxxMyyFromfcNEQ,
+ real* kxxMzzFromfcNEQ,
+ unsigned int* geoD,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ real* DD,
+ bool isEvenTimestep);
__global__ void LBCalc2ndMomentsCompSP27(real* kxyFromfcNEQ,
- real* kyzFromfcNEQ,
- real* kxzFromfcNEQ,
- real* kxxMyyFromfcNEQ,
- real* kxxMzzFromfcNEQ,
- unsigned int* geoD,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- real* DD,
- bool isEvenTimestep);
+ real* kyzFromfcNEQ,
+ real* kxzFromfcNEQ,
+ real* kxxMyyFromfcNEQ,
+ real* kxxMzzFromfcNEQ,
+ unsigned int* geoD,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ real* DD,
+ bool isEvenTimestep);
__global__ void LBCalc3rdMomentsIncompSP27( real* CUMbbb,
- real* CUMabc,
- real* CUMbac,
- real* CUMbca,
- real* CUMcba,
- real* CUMacb,
- real* CUMcab,
- unsigned int* bcMatD,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- real* DDStart,
- int size_Mat,
- bool EvenOrOdd);
+ real* CUMabc,
+ real* CUMbac,
+ real* CUMbca,
+ real* CUMcba,
+ real* CUMacb,
+ real* CUMcab,
+ unsigned int* bcMatD,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ real* DDStart,
+ unsigned long long numberOfLBnodes,
+ bool EvenOrOdd);
__global__ void LBCalc3rdMomentsCompSP27(real* CUMbbb,
- real* CUMabc,
- real* CUMbac,
- real* CUMbca,
- real* CUMcba,
- real* CUMacb,
- real* CUMcab,
- unsigned int* bcMatD,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- real* DDStart,
- int size_Mat,
- bool EvenOrOdd);
+ real* CUMabc,
+ real* CUMbac,
+ real* CUMbca,
+ real* CUMcba,
+ real* CUMacb,
+ real* CUMcab,
+ unsigned int* bcMatD,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ real* DDStart,
+ unsigned long long numberOfLBnodes,
+ bool EvenOrOdd);
__global__ void LBCalcHigherMomentsIncompSP27( real* CUMcbb,
- real* CUMbcb,
- real* CUMbbc,
- real* CUMcca,
- real* CUMcac,
- real* CUMacc,
- real* CUMbcc,
- real* CUMcbc,
- real* CUMccb,
- real* CUMccc,
- unsigned int* bcMatD,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- real* DDStart,
- int size_Mat,
- bool EvenOrOdd);
+ real* CUMbcb,
+ real* CUMbbc,
+ real* CUMcca,
+ real* CUMcac,
+ real* CUMacc,
+ real* CUMbcc,
+ real* CUMcbc,
+ real* CUMccb,
+ real* CUMccc,
+ unsigned int* bcMatD,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ real* DDStart,
+ unsigned long long numberOfLBnodes,
+ bool EvenOrOdd);
__global__ void LBCalcHigherMomentsCompSP27( real* CUMcbb,
- real* CUMbcb,
- real* CUMbbc,
- real* CUMcca,
- real* CUMcac,
- real* CUMacc,
- real* CUMbcc,
- real* CUMcbc,
- real* CUMccb,
- real* CUMccc,
- unsigned int* bcMatD,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- real* DDStart,
- int size_Mat,
- bool EvenOrOdd);
+ real* CUMbcb,
+ real* CUMbbc,
+ real* CUMcca,
+ real* CUMcac,
+ real* CUMacc,
+ real* CUMbcc,
+ real* CUMcbc,
+ real* CUMccb,
+ real* CUMccc,
+ unsigned int* bcMatD,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ real* DDStart,
+ unsigned long long numberOfLBnodes,
+ bool EvenOrOdd);
__global__ void LBCalcMeasurePoints(real* vxMP,
real* vyMP,
@@ -568,7 +568,7 @@ __global__ void LBCalcMeasurePoints(real* vxMP,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
real* DD,
bool isEvenTimestep);
@@ -580,7 +580,7 @@ __global__ void LB_BC_Press_East27( int nx,
unsigned int* neighborY,
unsigned int* neighborZ,
real* DD,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep) ;
__global__ void LB_BC_Vel_West_27( int nx,
@@ -592,7 +592,7 @@ __global__ void LB_BC_Vel_West_27( int nx,
unsigned int* neighborY,
unsigned int* neighborZ,
real* DD,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep,
real u0x,
unsigned int grid_nx,
@@ -608,64 +608,64 @@ __global__ void QDevice27(real* distributions,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int numberOfLBnodes,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep);
__global__ void QDeviceComp27(
- real* distributions,
- int* subgridDistanceIndices,
- real* subgridDistances,
- unsigned int numberOfBCnodes,
- real omega,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int numberOfLBnodes,
- bool isEvenTimestep);
+ real* distributions,
+ int* subgridDistanceIndices,
+ real* subgridDistances,
+ unsigned int numberOfBCnodes,
+ real omega,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep);
__global__ void QDeviceCompThinWallsPartOne27(real* DD,
- int* k_Q,
- real* QQ,
- unsigned int numberOfBCnodes,
- real om1,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep);
-
-__global__ void QDevice3rdMomentsComp27( real* distributions,
- int* subgridDistanceIndices,
- real* subgridDistances,
- unsigned int numberOfBCnodes,
- real omega,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int numberOfLBnodes,
- bool isEvenTimestep);
+ int* k_Q,
+ real* QQ,
+ unsigned int numberOfBCnodes,
+ real om1,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep);
+
+__global__ void QDevice3rdMomentsComp27( real* distributions,
+ int* subgridDistanceIndices,
+ real* subgridDistances,
+ unsigned int numberOfBCnodes,
+ real omega,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep);
__global__ void QDeviceIncompHighNu27(real* DD,
- int* k_Q,
- real* QQ,
- unsigned int numberOfBCnodes,
- real om1,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int numberOfLBnodes,
- bool isEvenTimestep);
+ int* k_Q,
+ real* QQ,
+ unsigned int numberOfBCnodes,
+ real om1,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep);
__global__ void QDeviceCompHighNu27( real* DD,
- int* k_Q,
- real* QQ,
- unsigned int numberOfBCnodes,
- real om1,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep);
+ int* k_Q,
+ real* QQ,
+ unsigned int numberOfBCnodes,
+ real om1,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep);
//Velocity BCs
__global__ void QVelDevPlainBB27(
@@ -679,43 +679,43 @@ __global__ void QVelDevPlainBB27(
uint* neighborX,
uint* neighborY,
uint* neighborZ,
- uint numberOfLBnodes,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep);
__global__ void QVelDevCouette27(real* vx,
- real* vy,
- real* vz,
- real* DD,
- int* k_Q,
- real* QQ,
- unsigned int numberOfBCnodes,
- real om1,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep);
+ real* vy,
+ real* vz,
+ real* DD,
+ int* k_Q,
+ real* QQ,
+ unsigned int numberOfBCnodes,
+ real om1,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep);
__global__ void QVelDev1h27( int inx,
- int iny,
- real* vx,
- real* vy,
- real* vz,
- real* DD,
- int* k_Q,
- real* QQ,
- unsigned int numberOfBCnodes,
- real om1,
- real Phi,
- real angularVelocity,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- real* coordX,
- real* coordY,
- real* coordZ,
- unsigned int size_Mat,
- bool isEvenTimestep);
+ int iny,
+ real* vx,
+ real* vy,
+ real* vz,
+ real* DD,
+ int* k_Q,
+ real* QQ,
+ unsigned int numberOfBCnodes,
+ real om1,
+ real Phi,
+ real angularVelocity,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ real* coordX,
+ real* coordY,
+ real* coordZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep);
__global__ void QVelDevice27(int inx,
int iny,
@@ -730,111 +730,111 @@ __global__ void QVelDevice27(int inx,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep);
__global__ void QVelDeviceCompPlusSlip27(real* vx,
- real* vy,
- real* vz,
- real* DD,
- int* k_Q,
- real* QQ,
- unsigned int numberOfBCnodes,
- real om1,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep);
+ real* vy,
+ real* vz,
+ real* DD,
+ int* k_Q,
+ real* QQ,
+ unsigned int numberOfBCnodes,
+ real om1,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep);
__global__ void QVelDeviceComp27(real* velocityX,
- real* velocityY,
- real* velocityZ,
- real* distribution,
- int* subgridDistanceIndices,
- real* subgridDistances,
- unsigned int numberOfBCnodes,
- real omega,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int numberOfLBnodes,
- bool isEvenTimestep);
+ real* velocityY,
+ real* velocityZ,
+ real* distribution,
+ int* subgridDistanceIndices,
+ real* subgridDistances,
+ unsigned int numberOfBCnodes,
+ real omega,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep);
__global__ void QVelDeviceCompThinWallsPartOne27(
- real* vx,
- real* vy,
- real* vz,
- real* DD,
- int* k_Q,
- real* QQ,
- uint numberOfBCnodes,
- real om1,
- uint* neighborX,
- uint* neighborY,
- uint* neighborZ,
- uint size_Mat,
- bool isEvenTimestep);
+ real* vx,
+ real* vy,
+ real* vz,
+ real* DD,
+ int* k_Q,
+ real* QQ,
+ uint numberOfBCnodes,
+ real om1,
+ uint* neighborX,
+ uint* neighborY,
+ uint* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep);
__global__ void QThinWallsPartTwo27(
- real* DD,
- int* k_Q,
- real* QQ,
- uint numberOfBCnodes,
- uint* geom,
- uint* neighborX,
- uint* neighborY,
- uint* neighborZ,
- uint* neighborWSB,
- uint size_Mat,
- bool isEvenTimestep);
+ real* DD,
+ int* k_Q,
+ real* QQ,
+ uint numberOfBCnodes,
+ uint* geom,
+ uint* neighborX,
+ uint* neighborY,
+ uint* neighborZ,
+ uint* neighborWSB,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep);
__global__ void QVelDeviceCompZeroPress27(
- real* velocityX,
- real* velocityY,
- real* velocityZ,
- real* distribution,
- int* subgridDistanceIndices,
- real* subgridDistances,
- unsigned int numberOfBCnodes,
- real omega,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int numberOfLBnodes,
- bool isEvenTimestep);
+ real* velocityX,
+ real* velocityY,
+ real* velocityZ,
+ real* distribution,
+ int* subgridDistanceIndices,
+ real* subgridDistances,
+ unsigned int numberOfBCnodes,
+ real omega,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep);
__global__ void QVelDeviceIncompHighNu27(real* vx,
- real* vy,
- real* vz,
- real* DD,
- int* k_Q,
- real* QQ,
- unsigned int numberOfBCnodes,
- real om1,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep);
+ real* vy,
+ real* vz,
+ real* DD,
+ int* k_Q,
+ real* QQ,
+ unsigned int numberOfBCnodes,
+ real om1,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep);
__global__ void QVelDeviceCompHighNu27( real* vx,
- real* vy,
- real* vz,
- real* DD,
- int* k_Q,
- real* QQ,
- unsigned int numberOfBCnodes,
- real om1,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep);
+ real* vy,
+ real* vz,
+ real* DD,
+ int* k_Q,
+ real* QQ,
+ unsigned int numberOfBCnodes,
+ real om1,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep);
__global__ void QVeloDeviceEQ27(real* VeloX,
- real* VeloY,
- real* VeloZ,
+ real* VeloY,
+ real* VeloZ,
real* DD,
int* k_Q,
int numberOfBCnodes,
@@ -842,22 +842,22 @@ __global__ void QVeloDeviceEQ27(real* VeloX,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep);
__global__ void QVeloStreetDeviceEQ27(
- real* veloXfraction,
- real* veloYfraction,
- int* naschVelo,
- real* DD,
- int* naschIndex,
- int numberOfStreetNodes,
- real velocityRatio,
- uint* neighborX,
- uint* neighborY,
- uint* neighborZ,
- uint size_Mat,
- bool isEvenTimestep);
+ real* veloXfraction,
+ real* veloYfraction,
+ int* naschVelo,
+ real* DD,
+ int* naschIndex,
+ int numberOfStreetNodes,
+ real velocityRatio,
+ uint* neighborX,
+ uint* neighborY,
+ uint* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep);
//Slip BCs
__global__ void QSlipDevice27(real* DD,
@@ -868,139 +868,150 @@ __global__ void QSlipDevice27(real* DD,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep);
__global__ void QSlipDeviceComp27(real* DD,
- int* k_Q,
- real* QQ,
- unsigned int numberOfBCnodes,
- real om1,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep);
+ int* k_Q,
+ real* QQ,
+ unsigned int numberOfBCnodes,
+ real om1,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep);
__global__ void QSlipDeviceComp27TurbViscosity(
- real* distributions,
- int* subgridDistanceIndices,
+ real* distributions,
+ int* subgridDistanceIndices,
real* subgridDistances,
unsigned int numberOfBCnodes,
- real omega,
+ real omega,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
real* turbViscosity,
- unsigned int numberOfLBnodes,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep);
__global__ void QSlipPressureDeviceComp27TurbViscosity(
- real* distributions,
- int* subgridDistanceIndices,
+ real* distributions,
+ int* subgridDistanceIndices,
real* subgridDistances,
unsigned int numberOfBCnodes,
- real omega,
+ real omega,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
real* turbViscosity,
- unsigned int numberOfLBnodes,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep);
__global__ void QSlipGeomDeviceComp27(real* DD,
- int* k_Q,
- real* QQ,
- unsigned int numberOfBCnodes,
- real om1,
- real* NormalX,
- real* NormalY,
- real* NormalZ,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep);
+ int* k_Q,
+ real* QQ,
+ unsigned int numberOfBCnodes,
+ real om1,
+ real* NormalX,
+ real* NormalY,
+ real* NormalZ,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep);
__global__ void QSlipNormDeviceComp27(real* DD,
- int* k_Q,
- real* QQ,
- unsigned int numberOfBCnodes,
- real om1,
- real* NormalX,
- real* NormalY,
- real* NormalZ,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep);
+ int* k_Q,
+ real* QQ,
+ unsigned int numberOfBCnodes,
+ real om1,
+ real* NormalX,
+ real* NormalY,
+ real* NormalZ,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep);
+
+__global__ void BBSlipDeviceComp27(
+ real* distributions,
+ int* subgridDistanceIndices,
+ real* subgridDistances,
+ unsigned int numberOfBCnodes,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep);
// Stress BCs (wall model)
__global__ void QStressDeviceComp27(real* DD,
- int* k_Q,
- int* k_N,
- real* QQ,
- unsigned int numberOfBCnodes,
- real om1,
- real* turbViscosity,
- real* vx,
- real* vy,
- real* vz,
- real* normalX,
- real* normalY,
- real* normalZ,
- real* vx_bc,
- real* vy_bc,
- real* vz_bc,
- real* vx1,
- real* vy1,
- real* vz1,
- int* samplingOffset,
- real* z0,
- bool hasWallModelMonitor,
- real* u_star_monitor,
- real* Fx_monitor,
- real* Fy_monitor,
- real* Fz_monitor,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep);
+ int* k_Q,
+ int* k_N,
+ real* QQ,
+ unsigned int numberOfBCnodes,
+ real om1,
+ real* turbViscosity,
+ real* vx,
+ real* vy,
+ real* vz,
+ real* normalX,
+ real* normalY,
+ real* normalZ,
+ real* vx_bc,
+ real* vy_bc,
+ real* vz_bc,
+ real* vx1,
+ real* vy1,
+ real* vz1,
+ int* samplingOffset,
+ real* z0,
+ bool hasWallModelMonitor,
+ real* u_star_monitor,
+ real* Fx_monitor,
+ real* Fy_monitor,
+ real* Fz_monitor,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep);
__global__ void BBStressDevice27( real* DD,
- int* k_Q,
- int* k_N,
- real* QQ,
- unsigned int numberOfBCnodes,
- real* vx,
- real* vy,
- real* vz,
- real* normalX,
- real* normalY,
- real* normalZ,
- real* vx_bc,
- real* vy_bc,
- real* vz_bc,
- real* vx1,
- real* vy1,
- real* vz1,
- int* samplingOffset,
- real* z0,
- bool hasWallModelMonitor,
- real* u_star_monitor,
- real* Fx_monitor,
- real* Fy_monitor,
- real* Fz_monitor,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep);
+ int* k_Q,
+ int* k_N,
+ real* QQ,
+ unsigned int numberOfBCnodes,
+ real* vx,
+ real* vy,
+ real* vz,
+ real* normalX,
+ real* normalY,
+ real* normalZ,
+ real* vx_bc,
+ real* vy_bc,
+ real* vz_bc,
+ real* vx1,
+ real* vy1,
+ real* vz1,
+ int* samplingOffset,
+ real* z0,
+ bool hasWallModelMonitor,
+ real* u_star_monitor,
+ real* Fx_monitor,
+ real* Fy_monitor,
+ real* Fz_monitor,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep);
__global__ void BBStressPressureDevice27( real* DD,
- int* k_Q,
+ int* k_Q,
int* k_N,
real* QQ,
unsigned int numberOfBCnodes,
@@ -1026,7 +1037,7 @@ __global__ void BBStressPressureDevice27( real* DD,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep);
//Pressure BCs
@@ -1039,23 +1050,23 @@ __global__ void QPressDevice27( real* rhoBC,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep);
__global__ void QPressDeviceAntiBB27( real* rhoBC,
- real* vx,
- real* vy,
- real* vz,
- real* DD,
- int* k_Q,
- real* QQ,
- int numberOfBCnodes,
- real om1,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep);
+ real* vx,
+ real* vy,
+ real* vz,
+ real* DD,
+ int* k_Q,
+ real* QQ,
+ int numberOfBCnodes,
+ real om1,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep);
__global__ void QPressDeviceFixBackflow27( real* rhoBC,
real* DD,
@@ -1065,7 +1076,7 @@ __global__ void QPressDeviceFixBackflow27( real* rhoBC,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep);
__global__ void QPressDeviceDirDepBot27( real* rhoBC,
@@ -1076,32 +1087,47 @@ __global__ void QPressDeviceDirDepBot27( real* rhoBC,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep);
__global__ void QPressNoRhoDevice27( real* rhoBC,
- real* DD,
- int* k_Q,
- int* k_N,
- int numberOfBCnodes,
- real om1,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep);
+ real* distributions,
+ int* k_Q,
+ int* k_N,
+ int numberOfBCnodes,
+ real om1,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep,
+ int direction);
+
+__global__ void QPressZeroRhoOutflowDevice27( real* rhoBC,
+ real* distributions,
+ int* k_Q,
+ int* k_N,
+ int numberOfBCnodes,
+ real om1,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep,
+ int direction,
+ real densityCorrectionFactor);
__global__ void QInflowScaleByPressDevice27( real* rhoBC,
- real* DD,
- int* k_Q,
- int* k_N,
- int numberOfBCnodes,
- real om1,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep);
+ real* DD,
+ int* k_Q,
+ int* k_N,
+ int numberOfBCnodes,
+ real om1,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep);
__global__ void QPressDeviceOld27(real* rhoBC,
real* DD,
@@ -1112,20 +1138,20 @@ __global__ void QPressDeviceOld27(real* rhoBC,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep);
__global__ void QPressDeviceIncompNEQ27( real* rhoBC,
- real* DD,
- int* k_Q,
- int* k_N,
- int numberOfBCnodes,
- real om1,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep);
+ real* DD,
+ int* k_Q,
+ int* k_N,
+ int numberOfBCnodes,
+ real om1,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep);
__global__ void QPressDeviceNEQ27(real* rhoBC,
real* distribution,
@@ -1136,7 +1162,7 @@ __global__ void QPressDeviceNEQ27(real* rhoBC,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep);
__global__ void QPressDeviceEQZ27(real* rhoBC,
@@ -1149,17 +1175,17 @@ __global__ void QPressDeviceEQZ27(real* rhoBC,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep);
__global__ void QPressDeviceZero27( real* DD,
- int* k_Q,
- unsigned int numberOfBCnodes,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep);
+ int* k_Q,
+ unsigned int numberOfBCnodes,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep);
__global__ void QPressDeviceFake27(real* rhoBC,
real* DD,
@@ -1170,7 +1196,7 @@ __global__ void QPressDeviceFake27(real* rhoBC,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep);
__global__ void BBDevice27(real* distributions,
@@ -1180,20 +1206,20 @@ __global__ void BBDevice27(real* distributions,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int numberOfLBnodes,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep);
__global__ void QPressDevice27_IntBB(real* rho,
- real* DD,
- int* k_Q,
- real* QQ,
- unsigned int numberOfBCnodes,
- real om1,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep);
+ real* DD,
+ int* k_Q,
+ real* QQ,
+ unsigned int numberOfBCnodes,
+ real om1,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep);
// TODO: https://git.rz.tu-bs.de/irmb/VirtualFluids_dev/-/issues/29
//Schlaffer BCs
@@ -1210,7 +1236,7 @@ __global__ void PressSchlaff27(real* rhoBC,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep);
// TODO: https://git.rz.tu-bs.de/irmb/VirtualFluids_dev/-/issues/29
@@ -1225,9 +1251,106 @@ __global__ void VelSchlaff27( int t,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep);
+__global__ void QPrecursorDeviceCompZeroPress( int* subgridDistanceIndices,
+ int numberOfBCnodes,
+ int numberOfPrecursorNodes,
+ int sizeQ,
+ real omega,
+ real* distributions,
+ real* subgridDistances,
+ uint* neighborX,
+ uint* neighborY,
+ uint* neighborZ,
+ uint* neighborsNT,
+ uint* neighborsNB,
+ uint* neighborsST,
+ uint* neighborsSB,
+ real* weights0PP,
+ real* weights0PM,
+ real* weights0MP,
+ real* weights0MM,
+ real* vLast,
+ real* vCurrent,
+ real velocityX,
+ real velocityY,
+ real velocityZ,
+ real timeRatio,
+ real velocityRatio,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep);
+
+__global__ void PrecursorDeviceEQ27( int* subgridDistanceIndices,
+ int numberOfBCnodes,
+ int numberOfPrecursorNodes,
+ real omega,
+ real* distributions,
+ uint* neighborX,
+ uint* neighborY,
+ uint* neighborZ,
+ uint* neighborsNT,
+ uint* neighborsNB,
+ uint* neighborsST,
+ uint* neighborsSB,
+ real* weights0PP,
+ real* weights0PM,
+ real* weights0MP,
+ real* weights0MM,
+ real* vLast,
+ real* vCurrent,
+ real velocityX,
+ real velocityY,
+ real velocityZ,
+ real timeRatio,
+ real velocityRatio,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep);
+
+__global__ void PrecursorDeviceDistributions( int* subgridDistanceIndices,
+ int numberOfBCNodes,
+ int numberOfPrecursorNodes,
+ real* distributions,
+ uint* neighborX,
+ uint* neighborY,
+ uint* neighborZ,
+ uint* neighborsNT,
+ uint* neighborsNB,
+ uint* neighborsST,
+ uint* neighborsSB,
+ real* weights0PP,
+ real* weights0PM,
+ real* weights0MP,
+ real* weights0MM,
+ real* fsLast,
+ real* fsNext,
+ real timeRatio,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep);
+__global__ void QPrecursorDeviceDistributions( int* subgridDistanceIndices,
+ real* subgridDistances,
+ int sizeQ,
+ int numberOfBCNodes,
+ int numberOfPrecursorNodes,
+ real* distributions,
+ uint* neighborX,
+ uint* neighborY,
+ uint* neighborZ,
+ uint* neighborsNT,
+ uint* neighborsNB,
+ uint* neighborsST,
+ uint* neighborsSB,
+ real* weights0PP,
+ real* weights0PM,
+ real* weights0MP,
+ real* weights0MM,
+ real* fsLast,
+ real* fsNext,
+ real timeRatio,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep);
+
//Advection / Diffusion BCs
__global__ void QAD7( real* DD,
real* DD7,
@@ -1240,68 +1363,68 @@ __global__ void QAD7( real* DD,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep);
//////////////////////////////////////////////////////////////////////////
//! \brief \ref Advection_Diffusion_Device_Kernel : Factorized central moments for Advection Diffusion Equation
__global__ void Factorized_Central_Moments_Advection_Diffusion_Device_Kernel(
- real omegaDiffusivity,
- uint* typeOfGridNode,
- uint* neighborX,
- uint* neighborY,
- uint* neighborZ,
- real* distributions,
- real* distributionsAD,
- int size_Mat,
- real* forces,
- bool isEvenTimestep);
+ real omegaDiffusivity,
+ uint* typeOfGridNode,
+ uint* neighborX,
+ uint* neighborY,
+ uint* neighborZ,
+ real* distributions,
+ real* distributionsAD,
+ unsigned long long numberOfLBnodes,
+ real* forces,
+ bool isEvenTimestep);
//////////////////////////////////////////////////////////////////////////
//! \brief \ref AD_SlipVelDeviceComp : device function for the slip-AD boundary condition
__global__ void AD_SlipVelDeviceComp(
- real * normalX,
- real * normalY,
- real * normalZ,
- real * distributions,
- real * distributionsAD,
- int* QindexArray,
- real * Qarrays,
- uint numberOfBCnodes,
- real omegaDiffusivity,
- uint * neighborX,
- uint * neighborY,
- uint * neighborZ,
- uint size_Mat,
- bool isEvenTimestep);
+ real * normalX,
+ real * normalY,
+ real * normalZ,
+ real * distributions,
+ real * distributionsAD,
+ int* QindexArray,
+ real * Qarrays,
+ uint numberOfBCnodes,
+ real omegaDiffusivity,
+ uint * neighborX,
+ uint * neighborY,
+ uint * neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep);
__global__ void QADDirichlet27( real* DD,
- real* DD27,
- real* temp,
- real diffusivity,
- int* k_Q,
- real* QQ,
- unsigned int numberOfBCnodes,
- real om1,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep);
+ real* DD27,
+ real* temp,
+ real diffusivity,
+ int* k_Q,
+ real* QQ,
+ unsigned int numberOfBCnodes,
+ real om1,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep);
__global__ void QADBB27( real* DD,
- real* DD27,
- real* temp,
- real diffusivity,
- int* k_Q,
- real* QQ,
- unsigned int numberOfBCnodes,
- real om1,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep);
+ real* DD27,
+ real* temp,
+ real diffusivity,
+ int* k_Q,
+ real* QQ,
+ unsigned int numberOfBCnodes,
+ real om1,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep);
__global__ void QADVel7( real* DD,
real* DD7,
@@ -1315,7 +1438,7 @@ __global__ void QADVel7( real* DD,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep);
__global__ void QADVel27(real* DD,
@@ -1330,7 +1453,7 @@ __global__ void QADVel27(real* DD,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep);
__global__ void QADPress7( real* DD,
@@ -1345,7 +1468,7 @@ __global__ void QADPress7( real* DD,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep);
__global__ void QADPress27( real* DD,
@@ -1360,109 +1483,109 @@ __global__ void QADPress27( real* DD,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep);
__global__ void QADPressNEQNeighbor27(
- real* DD,
- real* DD27,
- int* k_Q,
- int* k_N,
- int numberOfBCnodes,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep
- );
+ real* DD,
+ real* DD27,
+ int* k_Q,
+ int* k_N,
+ int numberOfBCnodes,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep
+ );
__global__ void QNoSlipADincomp7( real* DD,
- real* DD7,
- real* temp,
- real diffusivity,
- int* k_Q,
- real* QQ,
- unsigned int numberOfBCnodes,
- real om1,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep);
+ real* DD7,
+ real* temp,
+ real diffusivity,
+ int* k_Q,
+ real* QQ,
+ unsigned int numberOfBCnodes,
+ real om1,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep);
__global__ void QNoSlipADincomp27( real* DD,
- real* DD27,
- real* temp,
- real diffusivity,
- int* k_Q,
- real* QQ,
- unsigned int numberOfBCnodes,
- real om1,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep);
+ real* DD27,
+ real* temp,
+ real diffusivity,
+ int* k_Q,
+ real* QQ,
+ unsigned int numberOfBCnodes,
+ real om1,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep);
__global__ void QADVeloIncomp7( real* DD,
- real* DD7,
- real* temp,
- real* velo,
- real diffusivity,
- int* k_Q,
- real* QQ,
- unsigned int numberOfBCnodes,
- real om1,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep);
+ real* DD7,
+ real* temp,
+ real* velo,
+ real diffusivity,
+ int* k_Q,
+ real* QQ,
+ unsigned int numberOfBCnodes,
+ real om1,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep);
__global__ void QADVeloIncomp27( real* DD,
- real* DD27,
- real* temp,
- real* velo,
- real diffusivity,
- int* k_Q,
- real* QQ,
- unsigned int numberOfBCnodes,
- real om1,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep);
+ real* DD27,
+ real* temp,
+ real* velo,
+ real diffusivity,
+ int* k_Q,
+ real* QQ,
+ unsigned int numberOfBCnodes,
+ real om1,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep);
__global__ void QADPressIncomp7(real* DD,
- real* DD7,
- real* temp,
- real* velo,
- real diffusivity,
- int* k_Q,
- real* QQ,
- unsigned int numberOfBCnodes,
- real om1,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep);
+ real* DD7,
+ real* temp,
+ real* velo,
+ real diffusivity,
+ int* k_Q,
+ real* QQ,
+ unsigned int numberOfBCnodes,
+ real om1,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep);
__global__ void QADPressIncomp27( real* DD,
- real* DD27,
- real* temp,
- real* velo,
- real diffusivity,
- int* k_Q,
- real* QQ,
- unsigned int numberOfBCnodes,
- real om1,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep);
+ real* DD27,
+ real* temp,
+ real* velo,
+ real diffusivity,
+ int* k_Q,
+ real* QQ,
+ unsigned int numberOfBCnodes,
+ real om1,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep);
//Propeller BC
__global__ void PropellerBC(unsigned int* neighborX,
@@ -1473,8 +1596,8 @@ __global__ void PropellerBC(unsigned int* neighborX,
real* uy,
real* uz,
int* k_Q,
- unsigned int size_Prop,
- unsigned int size_Mat,
+ unsigned int size_Prop,
+ unsigned long long numberOfLBnodes,
unsigned int* bcMatD,
real* DD,
bool EvenOrOdd);
@@ -1490,19 +1613,19 @@ __global__ void scaleCF27(real* DC,
unsigned int* neighborFX,
unsigned int* neighborFY,
unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
- bool isEvenTimestep,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
+ bool isEvenTimestep,
unsigned int* posCSWB,
unsigned int* posFSWB,
unsigned int kCF,
- real omCoarse,
- real omFine,
- real nu,
- unsigned int nxC,
- unsigned int nyC,
- unsigned int nxF,
- unsigned int nyF);
+ real omCoarse,
+ real omFine,
+ real nu,
+ unsigned int nxC,
+ unsigned int nyC,
+ unsigned int nxF,
+ unsigned int nyF);
__global__ void scaleCFEff27(real* DC,
real* DF,
@@ -1512,18 +1635,18 @@ __global__ void scaleCFEff27(real* DC,
unsigned int* neighborFX,
unsigned int* neighborFY,
unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
- bool isEvenTimestep,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
+ bool isEvenTimestep,
unsigned int* posCSWB,
unsigned int* posFSWB,
unsigned int kCF,
- real omCoarse,
- real omFine,
- real nu,
- unsigned int nxC,
- unsigned int nyC,
- unsigned int nxF,
+ real omCoarse,
+ real omFine,
+ real nu,
+ unsigned int nxC,
+ unsigned int nyC,
+ unsigned int nxF,
unsigned int nyF,
OffCF offCF);
@@ -1535,8 +1658,8 @@ __global__ void scaleCFLast27( real* DC,
unsigned int* neighborFX,
unsigned int* neighborFY,
unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
bool isEvenTimestep,
unsigned int* posCSWB,
unsigned int* posFSWB,
@@ -1558,8 +1681,8 @@ __global__ void scaleCFpress27(real* DC,
unsigned int* neighborFX,
unsigned int* neighborFY,
unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
bool isEvenTimestep,
unsigned int* posCSWB,
unsigned int* posFSWB,
@@ -1581,8 +1704,8 @@ __global__ void scaleCF_Fix_27(real* DC,
unsigned int* neighborFX,
unsigned int* neighborFY,
unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
bool isEvenTimestep,
unsigned int* posCSWB,
unsigned int* posFSWB,
@@ -1597,233 +1720,233 @@ __global__ void scaleCF_Fix_27(real* DC,
OffCF offCF);
__global__ void scaleCF_Fix_comp_27( real* DC,
- real* DF,
- unsigned int* neighborCX,
- unsigned int* neighborCY,
- unsigned int* neighborCZ,
- unsigned int* neighborFX,
- unsigned int* neighborFY,
- unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
- bool isEvenTimestep,
- unsigned int* posCSWB,
- unsigned int* posFSWB,
- unsigned int kCF,
- real omCoarse,
- real omFine,
- real nu,
- unsigned int nxC,
- unsigned int nyC,
- unsigned int nxF,
- unsigned int nyF,
- OffCF offCF);
+ real* DF,
+ unsigned int* neighborCX,
+ unsigned int* neighborCY,
+ unsigned int* neighborCZ,
+ unsigned int* neighborFX,
+ unsigned int* neighborFY,
+ unsigned int* neighborFZ,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
+ bool isEvenTimestep,
+ unsigned int* posCSWB,
+ unsigned int* posFSWB,
+ unsigned int kCF,
+ real omCoarse,
+ real omFine,
+ real nu,
+ unsigned int nxC,
+ unsigned int nyC,
+ unsigned int nxF,
+ unsigned int nyF,
+ OffCF offCF);
__global__ void scaleCF_0817_comp_27( real* DC,
- real* DF,
- unsigned int* neighborCX,
- unsigned int* neighborCY,
- unsigned int* neighborCZ,
- unsigned int* neighborFX,
- unsigned int* neighborFY,
- unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
- bool isEvenTimestep,
- unsigned int* posCSWB,
- unsigned int* posFSWB,
- unsigned int kCF,
- real omCoarse,
- real omFine,
- real nu,
- unsigned int nxC,
- unsigned int nyC,
- unsigned int nxF,
- unsigned int nyF,
- OffCF offCF);
+ real* DF,
+ unsigned int* neighborCX,
+ unsigned int* neighborCY,
+ unsigned int* neighborCZ,
+ unsigned int* neighborFX,
+ unsigned int* neighborFY,
+ unsigned int* neighborFZ,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
+ bool isEvenTimestep,
+ unsigned int* posCSWB,
+ unsigned int* posFSWB,
+ unsigned int kCF,
+ real omCoarse,
+ real omFine,
+ real nu,
+ unsigned int nxC,
+ unsigned int nyC,
+ unsigned int nxF,
+ unsigned int nyF,
+ OffCF offCF);
__global__ void scaleCF_comp_D3Q27F3_2018( real* DC,
- real* DF,
- real* G6,
- unsigned int* neighborCX,
- unsigned int* neighborCY,
- unsigned int* neighborCZ,
- unsigned int* neighborFX,
- unsigned int* neighborFY,
- unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
- bool isEvenTimestep,
- unsigned int* posCSWB,
- unsigned int* posFSWB,
- unsigned int kCF,
- real omCoarse,
- real omFine,
- real nu,
- unsigned int nxC,
- unsigned int nyC,
- unsigned int nxF,
- unsigned int nyF,
- OffCF offCF);
+ real* DF,
+ real* G6,
+ unsigned int* neighborCX,
+ unsigned int* neighborCY,
+ unsigned int* neighborCZ,
+ unsigned int* neighborFX,
+ unsigned int* neighborFY,
+ unsigned int* neighborFZ,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
+ bool isEvenTimestep,
+ unsigned int* posCSWB,
+ unsigned int* posFSWB,
+ unsigned int kCF,
+ real omCoarse,
+ real omFine,
+ real nu,
+ unsigned int nxC,
+ unsigned int nyC,
+ unsigned int nxF,
+ unsigned int nyF,
+ OffCF offCF);
__global__ void scaleCF_comp_D3Q27F3( real* DC,
- real* DF,
- real* G6,
- unsigned int* neighborCX,
- unsigned int* neighborCY,
- unsigned int* neighborCZ,
- unsigned int* neighborFX,
- unsigned int* neighborFY,
- unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
- bool isEvenTimestep,
- unsigned int* posCSWB,
- unsigned int* posFSWB,
- unsigned int kCF,
- real omCoarse,
- real omFine,
- real nu,
- unsigned int nxC,
- unsigned int nyC,
- unsigned int nxF,
- unsigned int nyF,
- OffCF offCF);
+ real* DF,
+ real* G6,
+ unsigned int* neighborCX,
+ unsigned int* neighborCY,
+ unsigned int* neighborCZ,
+ unsigned int* neighborFX,
+ unsigned int* neighborFY,
+ unsigned int* neighborFZ,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
+ bool isEvenTimestep,
+ unsigned int* posCSWB,
+ unsigned int* posFSWB,
+ unsigned int kCF,
+ real omCoarse,
+ real omFine,
+ real nu,
+ unsigned int nxC,
+ unsigned int nyC,
+ unsigned int nxF,
+ unsigned int nyF,
+ OffCF offCF);
__global__ void scaleCF_staggered_time_comp_27(real* DC,
- real* DF,
- unsigned int* neighborCX,
- unsigned int* neighborCY,
- unsigned int* neighborCZ,
- unsigned int* neighborFX,
- unsigned int* neighborFY,
- unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
- bool isEvenTimestep,
- unsigned int* posCSWB,
- unsigned int* posFSWB,
- unsigned int kCF,
- real omCoarse,
- real omFine,
- real nu,
- unsigned int nxC,
- unsigned int nyC,
- unsigned int nxF,
- unsigned int nyF,
- OffCF offCF);
+ real* DF,
+ unsigned int* neighborCX,
+ unsigned int* neighborCY,
+ unsigned int* neighborCZ,
+ unsigned int* neighborFX,
+ unsigned int* neighborFY,
+ unsigned int* neighborFZ,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
+ bool isEvenTimestep,
+ unsigned int* posCSWB,
+ unsigned int* posFSWB,
+ unsigned int kCF,
+ real omCoarse,
+ real omFine,
+ real nu,
+ unsigned int nxC,
+ unsigned int nyC,
+ unsigned int nxF,
+ unsigned int nyF,
+ OffCF offCF);
__global__ void scaleCF_RhoSq_comp_27( real* DC,
- real* DF,
- unsigned int* neighborCX,
- unsigned int* neighborCY,
- unsigned int* neighborCZ,
- unsigned int* neighborFX,
- unsigned int* neighborFY,
- unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
- bool isEvenTimestep,
- unsigned int* posCSWB,
- unsigned int* posFSWB,
- unsigned int kCF,
- real omCoarse,
- real omFine,
- real nu,
- unsigned int nxC,
- unsigned int nyC,
- unsigned int nxF,
- unsigned int nyF,
- OffCF offCF);
+ real* DF,
+ unsigned int* neighborCX,
+ unsigned int* neighborCY,
+ unsigned int* neighborCZ,
+ unsigned int* neighborFX,
+ unsigned int* neighborFY,
+ unsigned int* neighborFZ,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
+ bool isEvenTimestep,
+ unsigned int* posCSWB,
+ unsigned int* posFSWB,
+ unsigned int kCF,
+ real omCoarse,
+ real omFine,
+ real nu,
+ unsigned int nxC,
+ unsigned int nyC,
+ unsigned int nxF,
+ unsigned int nyF,
+ OffCF offCF);
__global__ void scaleCF_compressible(
- real* distributionsCoarse,
- real* distributionsFine,
+ real* distributionsCoarse,
+ real* distributionsFine,
unsigned int* neighborXcoarse,
unsigned int* neighborYcoarse,
unsigned int* neighborZcoarse,
unsigned int* neighborXfine,
unsigned int* neighborYfine,
unsigned int* neighborZfine,
- unsigned int numberOfLBnodesCoarse,
- unsigned int numberOfLBnodesFine,
+ unsigned long long numberOfLBnodesCoarse,
+ unsigned long long numberOfLBnodesFine,
bool isEvenTimestep,
- unsigned int* indicesCoarseMMM,
- unsigned int* indicesFineMMM,
- unsigned int numberOfInterfaceNodes,
- real omegaCoarse,
- real omegaFine,
+ unsigned int* indicesCoarseMMM,
+ unsigned int* indicesFineMMM,
+ unsigned int numberOfInterfaceNodes,
+ real omegaCoarse,
+ real omegaFine,
OffCF offsetCF);
__global__ void scaleCF_RhoSq_3rdMom_comp_27(real* DC,
- real* DF,
- unsigned int* neighborCX,
- unsigned int* neighborCY,
- unsigned int* neighborCZ,
- unsigned int* neighborFX,
- unsigned int* neighborFY,
- unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
- bool isEvenTimestep,
- unsigned int* posCSWB,
- unsigned int* posFSWB,
- unsigned int kCF,
- real omCoarse,
- real omFine,
- real nu,
- unsigned int nxC,
- unsigned int nyC,
- unsigned int nxF,
- unsigned int nyF,
- OffCF offCF);
+ real* DF,
+ unsigned int* neighborCX,
+ unsigned int* neighborCY,
+ unsigned int* neighborCZ,
+ unsigned int* neighborFX,
+ unsigned int* neighborFY,
+ unsigned int* neighborFZ,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
+ bool isEvenTimestep,
+ unsigned int* posCSWB,
+ unsigned int* posFSWB,
+ unsigned int kCF,
+ real omCoarse,
+ real omFine,
+ real nu,
+ unsigned int nxC,
+ unsigned int nyC,
+ unsigned int nxF,
+ unsigned int nyF,
+ OffCF offCF);
__global__ void scaleCF_AA2016_comp_27(real* DC,
- real* DF,
- unsigned int* neighborCX,
- unsigned int* neighborCY,
- unsigned int* neighborCZ,
- unsigned int* neighborFX,
- unsigned int* neighborFY,
- unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
- bool isEvenTimestep,
- unsigned int* posCSWB,
- unsigned int* posFSWB,
- unsigned int kCF,
- real omCoarse,
- real omFine,
- real nu,
- unsigned int nxC,
- unsigned int nyC,
- unsigned int nxF,
- unsigned int nyF,
- OffCF offCF);
+ real* DF,
+ unsigned int* neighborCX,
+ unsigned int* neighborCY,
+ unsigned int* neighborCZ,
+ unsigned int* neighborFX,
+ unsigned int* neighborFY,
+ unsigned int* neighborFZ,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
+ bool isEvenTimestep,
+ unsigned int* posCSWB,
+ unsigned int* posFSWB,
+ unsigned int kCF,
+ real omCoarse,
+ real omFine,
+ real nu,
+ unsigned int nxC,
+ unsigned int nyC,
+ unsigned int nxF,
+ unsigned int nyF,
+ OffCF offCF);
__global__ void scaleCF_NSPress_27(real* DC,
- real* DF,
- unsigned int* neighborCX,
- unsigned int* neighborCY,
- unsigned int* neighborCZ,
- unsigned int* neighborFX,
- unsigned int* neighborFY,
- unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
- bool isEvenTimestep,
- unsigned int* posCSWB,
- unsigned int* posFSWB,
- unsigned int kCF,
- real omCoarse,
- real omFine,
- real nu,
- unsigned int nxC,
- unsigned int nyC,
- unsigned int nxF,
- unsigned int nyF,
- OffCF offCF);
+ real* DF,
+ unsigned int* neighborCX,
+ unsigned int* neighborCY,
+ unsigned int* neighborCZ,
+ unsigned int* neighborFX,
+ unsigned int* neighborFY,
+ unsigned int* neighborFZ,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
+ bool isEvenTimestep,
+ unsigned int* posCSWB,
+ unsigned int* posFSWB,
+ unsigned int kCF,
+ real omCoarse,
+ real omFine,
+ real nu,
+ unsigned int nxC,
+ unsigned int nyC,
+ unsigned int nxF,
+ unsigned int nyF,
+ OffCF offCF);
__global__ void scaleCFThSMG7( real* DC,
real* DF,
@@ -1835,8 +1958,8 @@ __global__ void scaleCFThSMG7( real* DC,
unsigned int* neighborFX,
unsigned int* neighborFY,
unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
bool isEvenTimestep,
unsigned int* posCSWB,
unsigned int* posFSWB,
@@ -1855,8 +1978,8 @@ __global__ void scaleCFThS7(real* DC,
unsigned int* neighborFX,
unsigned int* neighborFY,
unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
bool isEvenTimestep,
unsigned int* posCSWB,
unsigned int* posFSWB,
@@ -1874,15 +1997,15 @@ __global__ void scaleCFThS27(real* DC,
unsigned int* neighborFX,
unsigned int* neighborFY,
unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
bool isEvenTimestep,
unsigned int* posCSWB,
unsigned int* posFSWB,
unsigned int kCF,
real nu,
real diffusivity_fine,
- OffCF offCF);
+ OffCF offCF);
//fine to coarse
__global__ void scaleFC27(real* DC,
@@ -1893,18 +2016,18 @@ __global__ void scaleFC27(real* DC,
unsigned int* neighborFX,
unsigned int* neighborFY,
unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
- bool isEvenTimestep,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
+ bool isEvenTimestep,
unsigned int* posC,
unsigned int* posFSWB,
unsigned int kFC,
- real omCoarse,
- real omFine,
- real nu,
- unsigned int nxC,
- unsigned int nyC,
- unsigned int nxF,
+ real omCoarse,
+ real omFine,
+ real nu,
+ unsigned int nxC,
+ unsigned int nyC,
+ unsigned int nxF,
unsigned int nyF);
__global__ void scaleFCEff27(real* DC,
@@ -1915,8 +2038,8 @@ __global__ void scaleFCEff27(real* DC,
unsigned int* neighborFX,
unsigned int* neighborFY,
unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
bool isEvenTimestep,
unsigned int* posC,
unsigned int* posFSWB,
@@ -1938,8 +2061,8 @@ __global__ void scaleFCLast27( real* DC,
unsigned int* neighborFX,
unsigned int* neighborFY,
unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
bool isEvenTimestep,
unsigned int* posC,
unsigned int* posFSWB,
@@ -1961,8 +2084,8 @@ __global__ void scaleFCpress27( real* DC,
unsigned int* neighborFX,
unsigned int* neighborFY,
unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
bool isEvenTimestep,
unsigned int* posC,
unsigned int* posFSWB,
@@ -1984,8 +2107,8 @@ __global__ void scaleFC_Fix_27( real* DC,
unsigned int* neighborFX,
unsigned int* neighborFY,
unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
bool isEvenTimestep,
unsigned int* posC,
unsigned int* posFSWB,
@@ -2000,145 +2123,145 @@ __global__ void scaleFC_Fix_27( real* DC,
OffFC offFC);
__global__ void scaleFC_Fix_comp_27( real* DC,
- real* DF,
- unsigned int* neighborCX,
- unsigned int* neighborCY,
- unsigned int* neighborCZ,
- unsigned int* neighborFX,
- unsigned int* neighborFY,
- unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
- bool isEvenTimestep,
- unsigned int* posC,
- unsigned int* posFSWB,
- unsigned int kFC,
- real omCoarse,
- real omFine,
- real nu,
- unsigned int nxC,
- unsigned int nyC,
- unsigned int nxF,
- unsigned int nyF,
- OffFC offFC);
+ real* DF,
+ unsigned int* neighborCX,
+ unsigned int* neighborCY,
+ unsigned int* neighborCZ,
+ unsigned int* neighborFX,
+ unsigned int* neighborFY,
+ unsigned int* neighborFZ,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
+ bool isEvenTimestep,
+ unsigned int* posC,
+ unsigned int* posFSWB,
+ unsigned int kFC,
+ real omCoarse,
+ real omFine,
+ real nu,
+ unsigned int nxC,
+ unsigned int nyC,
+ unsigned int nxF,
+ unsigned int nyF,
+ OffFC offFC);
__global__ void scaleFC_0817_comp_27( real* DC,
- real* DF,
- unsigned int* neighborCX,
- unsigned int* neighborCY,
- unsigned int* neighborCZ,
- unsigned int* neighborFX,
- unsigned int* neighborFY,
- unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
- bool isEvenTimestep,
- unsigned int* posC,
- unsigned int* posFSWB,
- unsigned int kFC,
- real omCoarse,
- real omFine,
- real nu,
- unsigned int nxC,
- unsigned int nyC,
- unsigned int nxF,
- unsigned int nyF,
- OffFC offFC);
+ real* DF,
+ unsigned int* neighborCX,
+ unsigned int* neighborCY,
+ unsigned int* neighborCZ,
+ unsigned int* neighborFX,
+ unsigned int* neighborFY,
+ unsigned int* neighborFZ,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
+ bool isEvenTimestep,
+ unsigned int* posC,
+ unsigned int* posFSWB,
+ unsigned int kFC,
+ real omCoarse,
+ real omFine,
+ real nu,
+ unsigned int nxC,
+ unsigned int nyC,
+ unsigned int nxF,
+ unsigned int nyF,
+ OffFC offFC);
__global__ void scaleFC_comp_D3Q27F3_2018( real* DC,
- real* DF,
- real* G6,
- unsigned int* neighborCX,
- unsigned int* neighborCY,
- unsigned int* neighborCZ,
- unsigned int* neighborFX,
- unsigned int* neighborFY,
- unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
- bool isEvenTimestep,
- unsigned int* posC,
- unsigned int* posFSWB,
- unsigned int kFC,
- real omCoarse,
- real omFine,
- real nu,
- unsigned int nxC,
- unsigned int nyC,
- unsigned int nxF,
- unsigned int nyF,
- OffFC offFC);
+ real* DF,
+ real* G6,
+ unsigned int* neighborCX,
+ unsigned int* neighborCY,
+ unsigned int* neighborCZ,
+ unsigned int* neighborFX,
+ unsigned int* neighborFY,
+ unsigned int* neighborFZ,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
+ bool isEvenTimestep,
+ unsigned int* posC,
+ unsigned int* posFSWB,
+ unsigned int kFC,
+ real omCoarse,
+ real omFine,
+ real nu,
+ unsigned int nxC,
+ unsigned int nyC,
+ unsigned int nxF,
+ unsigned int nyF,
+ OffFC offFC);
__global__ void scaleFC_comp_D3Q27F3( real* DC,
- real* DF,
- real* G6,
- unsigned int* neighborCX,
- unsigned int* neighborCY,
- unsigned int* neighborCZ,
- unsigned int* neighborFX,
- unsigned int* neighborFY,
- unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
- bool isEvenTimestep,
- unsigned int* posC,
- unsigned int* posFSWB,
- unsigned int kFC,
- real omCoarse,
- real omFine,
- real nu,
- unsigned int nxC,
- unsigned int nyC,
- unsigned int nxF,
- unsigned int nyF,
- OffFC offFC);
+ real* DF,
+ real* G6,
+ unsigned int* neighborCX,
+ unsigned int* neighborCY,
+ unsigned int* neighborCZ,
+ unsigned int* neighborFX,
+ unsigned int* neighborFY,
+ unsigned int* neighborFZ,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
+ bool isEvenTimestep,
+ unsigned int* posC,
+ unsigned int* posFSWB,
+ unsigned int kFC,
+ real omCoarse,
+ real omFine,
+ real nu,
+ unsigned int nxC,
+ unsigned int nyC,
+ unsigned int nxF,
+ unsigned int nyF,
+ OffFC offFC);
__global__ void scaleFC_staggered_time_comp_27(real* DC,
- real* DF,
- unsigned int* neighborCX,
- unsigned int* neighborCY,
- unsigned int* neighborCZ,
- unsigned int* neighborFX,
- unsigned int* neighborFY,
- unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
- bool isEvenTimestep,
- unsigned int* posC,
- unsigned int* posFSWB,
- unsigned int kFC,
- real omCoarse,
- real omFine,
- real nu,
- unsigned int nxC,
- unsigned int nyC,
- unsigned int nxF,
- unsigned int nyF,
- OffFC offFC);
+ real* DF,
+ unsigned int* neighborCX,
+ unsigned int* neighborCY,
+ unsigned int* neighborCZ,
+ unsigned int* neighborFX,
+ unsigned int* neighborFY,
+ unsigned int* neighborFZ,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
+ bool isEvenTimestep,
+ unsigned int* posC,
+ unsigned int* posFSWB,
+ unsigned int kFC,
+ real omCoarse,
+ real omFine,
+ real nu,
+ unsigned int nxC,
+ unsigned int nyC,
+ unsigned int nxF,
+ unsigned int nyF,
+ OffFC offFC);
__global__ void scaleFC_RhoSq_comp_27( real* DC,
- real* DF,
- unsigned int* neighborCX,
- unsigned int* neighborCY,
- unsigned int* neighborCZ,
- unsigned int* neighborFX,
- unsigned int* neighborFY,
- unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
- bool isEvenTimestep,
- unsigned int* posC,
- unsigned int* posFSWB,
- unsigned int kFC,
- real omCoarse,
- real omFine,
- real nu,
- unsigned int nxC,
- unsigned int nyC,
- unsigned int nxF,
- unsigned int nyF,
- OffFC offFC);
+ real* DF,
+ unsigned int* neighborCX,
+ unsigned int* neighborCY,
+ unsigned int* neighborCZ,
+ unsigned int* neighborFX,
+ unsigned int* neighborFY,
+ unsigned int* neighborFZ,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
+ bool isEvenTimestep,
+ unsigned int* posC,
+ unsigned int* posFSWB,
+ unsigned int kFC,
+ real omCoarse,
+ real omFine,
+ real nu,
+ unsigned int nxC,
+ unsigned int nyC,
+ unsigned int nxF,
+ unsigned int nyF,
+ OffFC offFC);
__global__ void scaleFC_compressible(
real *distributionsCoarse,
@@ -2149,8 +2272,8 @@ __global__ void scaleFC_compressible(
unsigned int *neighborXfine,
unsigned int *neighborYfine,
unsigned int *neighborZfine,
- unsigned int numberOfLBnodesCoarse,
- unsigned int numberOfLBnodesFine,
+ unsigned long long numberOfLBnodesCoarse,
+ unsigned long long numberOfLBnodesFine,
bool isEvenTimestep,
unsigned int *indicesCoarse000,
unsigned int *indicesFineMMM,
@@ -2160,73 +2283,73 @@ __global__ void scaleFC_compressible(
OffFC offsetFC);
__global__ void scaleFC_RhoSq_3rdMom_comp_27(real* DC,
- real* DF,
- unsigned int* neighborCX,
- unsigned int* neighborCY,
- unsigned int* neighborCZ,
- unsigned int* neighborFX,
- unsigned int* neighborFY,
- unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
- bool isEvenTimestep,
- unsigned int* posC,
- unsigned int* posFSWB,
- unsigned int kFC,
- real omCoarse,
- real omFine,
- real nu,
- unsigned int nxC,
- unsigned int nyC,
- unsigned int nxF,
- unsigned int nyF,
- OffFC offFC);
+ real* DF,
+ unsigned int* neighborCX,
+ unsigned int* neighborCY,
+ unsigned int* neighborCZ,
+ unsigned int* neighborFX,
+ unsigned int* neighborFY,
+ unsigned int* neighborFZ,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
+ bool isEvenTimestep,
+ unsigned int* posC,
+ unsigned int* posFSWB,
+ unsigned int kFC,
+ real omCoarse,
+ real omFine,
+ real nu,
+ unsigned int nxC,
+ unsigned int nyC,
+ unsigned int nxF,
+ unsigned int nyF,
+ OffFC offFC);
__global__ void scaleFC_AA2016_comp_27(real* DC,
- real* DF,
- unsigned int* neighborCX,
- unsigned int* neighborCY,
- unsigned int* neighborCZ,
- unsigned int* neighborFX,
- unsigned int* neighborFY,
- unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
- bool isEvenTimestep,
- unsigned int* posC,
- unsigned int* posFSWB,
- unsigned int kFC,
- real omCoarse,
- real omFine,
- real nu,
- unsigned int nxC,
- unsigned int nyC,
- unsigned int nxF,
- unsigned int nyF,
- OffFC offFC);
+ real* DF,
+ unsigned int* neighborCX,
+ unsigned int* neighborCY,
+ unsigned int* neighborCZ,
+ unsigned int* neighborFX,
+ unsigned int* neighborFY,
+ unsigned int* neighborFZ,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
+ bool isEvenTimestep,
+ unsigned int* posC,
+ unsigned int* posFSWB,
+ unsigned int kFC,
+ real omCoarse,
+ real omFine,
+ real nu,
+ unsigned int nxC,
+ unsigned int nyC,
+ unsigned int nxF,
+ unsigned int nyF,
+ OffFC offFC);
__global__ void scaleFC_NSPress_27(real* DC,
- real* DF,
- unsigned int* neighborCX,
- unsigned int* neighborCY,
- unsigned int* neighborCZ,
- unsigned int* neighborFX,
- unsigned int* neighborFY,
- unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
- bool isEvenTimestep,
- unsigned int* posC,
- unsigned int* posFSWB,
- unsigned int kFC,
- real omCoarse,
- real omFine,
- real nu,
- unsigned int nxC,
- unsigned int nyC,
- unsigned int nxF,
- unsigned int nyF,
- OffFC offFC);
+ real* DF,
+ unsigned int* neighborCX,
+ unsigned int* neighborCY,
+ unsigned int* neighborCZ,
+ unsigned int* neighborFX,
+ unsigned int* neighborFY,
+ unsigned int* neighborFZ,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
+ bool isEvenTimestep,
+ unsigned int* posC,
+ unsigned int* posFSWB,
+ unsigned int kFC,
+ real omCoarse,
+ real omFine,
+ real nu,
+ unsigned int nxC,
+ unsigned int nyC,
+ unsigned int nxF,
+ unsigned int nyF,
+ OffFC offFC);
__global__ void scaleFCThSMG7( real* DC,
real* DF,
@@ -2238,8 +2361,8 @@ __global__ void scaleFCThSMG7( real* DC,
unsigned int* neighborFX,
unsigned int* neighborFY,
unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
bool isEvenTimestep,
unsigned int* posC,
unsigned int* posFSWB,
@@ -2258,8 +2381,8 @@ __global__ void scaleFCThS7(real* DC,
unsigned int* neighborFX,
unsigned int* neighborFY,
unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
bool isEvenTimestep,
unsigned int* posC,
unsigned int* posFSWB,
@@ -2277,242 +2400,242 @@ __global__ void scaleFCThS27( real* DC,
unsigned int* neighborFX,
unsigned int* neighborFY,
unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
bool isEvenTimestep,
unsigned int* posC,
unsigned int* posFSWB,
unsigned int kFC,
real nu,
real diffusivity_coarse,
- OffFC offFC);
+ OffFC offFC);
__global__ void DragLiftPost27( real* DD,
- int* k_Q,
- real* QQ,
- int numberOfBCnodes,
- double *DragX,
- double *DragY,
- double *DragZ,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep);
+ int* k_Q,
+ real* QQ,
+ int numberOfBCnodes,
+ double *DragX,
+ double *DragY,
+ double *DragZ,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep);
__global__ void DragLiftPre27( real* DD,
- int* k_Q,
- real* QQ,
- int numberOfBCnodes,
- double *DragX,
- double *DragY,
- double *DragZ,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep);
+ int* k_Q,
+ real* QQ,
+ int numberOfBCnodes,
+ double *DragX,
+ double *DragY,
+ double *DragZ,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep);
__global__ void CalcCP27(real* DD,
- int* cpIndex,
- int nonCp,
- double *cpPress,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep);
+ int* cpIndex,
+ int nonCp,
+ double *cpPress,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep);
__global__ void getSendFsPre27(real* DD,
- real* bufferFs,
- int* sendIndex,
+ real* bufferFs,
+ int* sendIndex,
int buffmax,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep);
__global__ void getSendFsPost27(real* DD,
- real* bufferFs,
- int* sendIndex,
+ real* bufferFs,
+ int* sendIndex,
int buffmax,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep);
__global__ void setRecvFsPre27(real* DD,
- real* bufferFs,
- int* recvIndex,
+ real* bufferFs,
+ int* recvIndex,
int buffmax,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep);
__global__ void setRecvFsPost27(real* DD,
- real* bufferFs,
- int* recvIndex,
+ real* bufferFs,
+ int* recvIndex,
int buffmax,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep);
__global__ void getSendGsF3(
- real* G6,
- real* bufferGs,
- int* sendIndex,
- int buffmax,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep);
+ real* G6,
+ real* bufferGs,
+ int* sendIndex,
+ int buffmax,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep);
__global__ void setRecvGsF3(
- real* G6,
- real* bufferGs,
- int* recvIndex,
- int buffmax,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep);
+ real* G6,
+ real* bufferGs,
+ int* recvIndex,
+ int buffmax,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep);
__global__ void WallFunction27( real* vx,
- real* vy,
- real* vz,
- real* DD,
- int* k_Q,
- real* QQ,
- unsigned int numberOfBCnodes,
- real om1,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep);
+ real* vy,
+ real* vz,
+ real* DD,
+ int* k_Q,
+ real* QQ,
+ unsigned int numberOfBCnodes,
+ real om1,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep);
__global__ void LBSetOutputWallVelocitySP27( real* vxD,
- real* vyD,
- real* vzD,
- real* vxWall,
- real* vyWall,
- real* vzWall,
- int numberOfWallNodes,
- int* kWallNodes,
- real* rhoD,
- real* pressD,
- unsigned int* geoD,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- real* DD,
- bool isEvenTimestep);
+ real* vyD,
+ real* vzD,
+ real* vxWall,
+ real* vyWall,
+ real* vzWall,
+ int numberOfWallNodes,
+ int* kWallNodes,
+ real* rhoD,
+ real* pressD,
+ unsigned int* geoD,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ real* DD,
+ bool isEvenTimestep);
__global__ void GetVeloforForcing27( real* DD,
- int* bcIndex,
- int nonAtBC,
- real* Vx,
- real* Vy,
- real* Vz,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep);
+ int* bcIndex,
+ int nonAtBC,
+ real* Vx,
+ real* Vy,
+ real* Vz,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep);
__global__ void InitParticles( real* coordX,
- real* coordY,
- real* coordZ,
- real* coordParticleXlocal,
- real* coordParticleYlocal,
- real* coordParticleZlocal,
- real* coordParticleXglobal,
- real* coordParticleYglobal,
- real* coordParticleZglobal,
- real* veloParticleX,
- real* veloParticleY,
- real* veloParticleZ,
- real* randArray,
- unsigned int* particleID,
- unsigned int* cellBaseID,
- unsigned int* bcMatD,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int* neighborWSB,
- int level,
- unsigned int numberOfParticles,
- unsigned int size_Mat);
+ real* coordY,
+ real* coordZ,
+ real* coordParticleXlocal,
+ real* coordParticleYlocal,
+ real* coordParticleZlocal,
+ real* coordParticleXglobal,
+ real* coordParticleYglobal,
+ real* coordParticleZglobal,
+ real* veloParticleX,
+ real* veloParticleY,
+ real* veloParticleZ,
+ real* randArray,
+ unsigned int* particleID,
+ unsigned int* cellBaseID,
+ unsigned int* bcMatD,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned int* neighborWSB,
+ int level,
+ unsigned int numberOfParticles,
+ unsigned long long numberOfLBnodes);
__global__ void MoveParticles( real* coordX,
- real* coordY,
- real* coordZ,
- real* coordParticleXlocal,
- real* coordParticleYlocal,
- real* coordParticleZlocal,
- real* coordParticleXglobal,
- real* coordParticleYglobal,
- real* coordParticleZglobal,
- real* veloParticleX,
- real* veloParticleY,
- real* veloParticleZ,
- real* DD,
- real omega,
- unsigned int* particleID,
- unsigned int* cellBaseID,
- unsigned int* bcMatD,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int* neighborWSB,
- int level,
- unsigned int timestep,
- unsigned int numberOfTimesteps,
- unsigned int numberOfParticles,
- unsigned int size_Mat,
- bool isEvenTimestep);
+ real* coordY,
+ real* coordZ,
+ real* coordParticleXlocal,
+ real* coordParticleYlocal,
+ real* coordParticleZlocal,
+ real* coordParticleXglobal,
+ real* coordParticleYglobal,
+ real* coordParticleZglobal,
+ real* veloParticleX,
+ real* veloParticleY,
+ real* veloParticleZ,
+ real* DD,
+ real omega,
+ unsigned int* particleID,
+ unsigned int* cellBaseID,
+ unsigned int* bcMatD,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned int* neighborWSB,
+ int level,
+ unsigned int timestep,
+ unsigned int numberOfTimesteps,
+ unsigned int numberOfParticles,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep);
__global__ void MoveParticlesWithoutBCs( real* coordX,
- real* coordY,
- real* coordZ,
- real* coordParticleXlocal,
- real* coordParticleYlocal,
- real* coordParticleZlocal,
- real* coordParticleXglobal,
- real* coordParticleYglobal,
- real* coordParticleZglobal,
- real* veloParticleX,
- real* veloParticleY,
- real* veloParticleZ,
- real* DD,
- real omega,
- unsigned int* particleID,
- unsigned int* cellBaseID,
- unsigned int* bcMatD,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int* neighborWSB,
- int level,
- unsigned int timestep,
- unsigned int numberOfTimesteps,
- unsigned int numberOfParticles,
- unsigned int size_Mat,
- bool isEvenTimestep);
+ real* coordY,
+ real* coordZ,
+ real* coordParticleXlocal,
+ real* coordParticleYlocal,
+ real* coordParticleZlocal,
+ real* coordParticleXglobal,
+ real* coordParticleYglobal,
+ real* coordParticleZglobal,
+ real* veloParticleX,
+ real* veloParticleY,
+ real* veloParticleZ,
+ real* DD,
+ real omega,
+ unsigned int* particleID,
+ unsigned int* cellBaseID,
+ unsigned int* bcMatD,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned int* neighborWSB,
+ int level,
+ unsigned int timestep,
+ unsigned int numberOfTimesteps,
+ unsigned int numberOfParticles,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep);
__global__ void initRandom(curandState* state);
__global__ void generateRandomValues(curandState* state,
- real* randArray);
+ real* randArray);
__global__ void CalcTurbulenceIntensity(
real* vxx,
@@ -2529,7 +2652,7 @@ __global__ void CalcTurbulenceIntensity(
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep);
#endif
diff --git a/src/gpu/VirtualFluids_GPU/GPU/GridScaling/ScaleCF27.cu b/src/gpu/VirtualFluids_GPU/GPU/GridScaling/ScaleCF27.cu
index 619d68c87d7a707e70be4c56d434191994144148..641d6519669b1522430fe88990c00d0630d00e9b 100644
--- a/src/gpu/VirtualFluids_GPU/GPU/GridScaling/ScaleCF27.cu
+++ b/src/gpu/VirtualFluids_GPU/GPU/GridScaling/ScaleCF27.cu
@@ -22,8 +22,8 @@ __global__ void scaleCF_0817_comp_27( real* DC,
unsigned int* neighborFX,
unsigned int* neighborFY,
unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
+ unsigned long long numberOfLBnodesCoarse,
+ unsigned long long numberOfLBnodesFine,
bool isEvenTimestep,
unsigned int* posCSWB,
unsigned int* posFSWB,
@@ -43,33 +43,33 @@ __global__ void scaleCF_0817_comp_27( real* DC,
*f000dest, *fMMMdest, *fMMPdest, *fMPPdest, *fMPMdest, *fPPMdest, *fPPPdest, *fPMPdest, *fPMMdest;
- fP00dest = &DF[DIR_P00 *size_MatF];
- fM00dest = &DF[DIR_M00 *size_MatF];
- f0P0dest = &DF[DIR_0P0 *size_MatF];
- f0M0dest = &DF[DIR_0M0 *size_MatF];
- f00Pdest = &DF[DIR_00P *size_MatF];
- f00Mdest = &DF[DIR_00M *size_MatF];
- fPP0dest = &DF[DIR_PP0 *size_MatF];
- fMM0dest = &DF[DIR_MM0 *size_MatF];
- fPM0dest = &DF[DIR_PM0 *size_MatF];
- fMP0dest = &DF[DIR_MP0 *size_MatF];
- fP0Pdest = &DF[DIR_P0P *size_MatF];
- fM0Mdest = &DF[DIR_M0M *size_MatF];
- fP0Mdest = &DF[DIR_P0M *size_MatF];
- fM0Pdest = &DF[DIR_M0P *size_MatF];
- f0PPdest = &DF[DIR_0PP *size_MatF];
- f0MMdest = &DF[DIR_0MM *size_MatF];
- f0PMdest = &DF[DIR_0PM *size_MatF];
- f0MPdest = &DF[DIR_0MP *size_MatF];
- f000dest = &DF[DIR_000*size_MatF];
- fMMMdest = &DF[DIR_MMM *size_MatF];
- fMMPdest = &DF[DIR_MMP *size_MatF];
- fMPPdest = &DF[DIR_MPP *size_MatF];
- fMPMdest = &DF[DIR_MPM *size_MatF];
- fPPMdest = &DF[DIR_PPM *size_MatF];
- fPPPdest = &DF[DIR_PPP *size_MatF];
- fPMPdest = &DF[DIR_PMP *size_MatF];
- fPMMdest = &DF[DIR_PMM *size_MatF];
+ fP00dest = &DF[DIR_P00 * numberOfLBnodesFine];
+ fM00dest = &DF[DIR_M00 * numberOfLBnodesFine];
+ f0P0dest = &DF[DIR_0P0 * numberOfLBnodesFine];
+ f0M0dest = &DF[DIR_0M0 * numberOfLBnodesFine];
+ f00Pdest = &DF[DIR_00P * numberOfLBnodesFine];
+ f00Mdest = &DF[DIR_00M * numberOfLBnodesFine];
+ fPP0dest = &DF[DIR_PP0 * numberOfLBnodesFine];
+ fMM0dest = &DF[DIR_MM0 * numberOfLBnodesFine];
+ fPM0dest = &DF[DIR_PM0 * numberOfLBnodesFine];
+ fMP0dest = &DF[DIR_MP0 * numberOfLBnodesFine];
+ fP0Pdest = &DF[DIR_P0P * numberOfLBnodesFine];
+ fM0Mdest = &DF[DIR_M0M * numberOfLBnodesFine];
+ fP0Mdest = &DF[DIR_P0M * numberOfLBnodesFine];
+ fM0Pdest = &DF[DIR_M0P * numberOfLBnodesFine];
+ f0PPdest = &DF[DIR_0PP * numberOfLBnodesFine];
+ f0MMdest = &DF[DIR_0MM * numberOfLBnodesFine];
+ f0PMdest = &DF[DIR_0PM * numberOfLBnodesFine];
+ f0MPdest = &DF[DIR_0MP * numberOfLBnodesFine];
+ f000dest = &DF[DIR_000 * numberOfLBnodesFine];
+ fMMMdest = &DF[DIR_MMM * numberOfLBnodesFine];
+ fMMPdest = &DF[DIR_MMP * numberOfLBnodesFine];
+ fMPPdest = &DF[DIR_MPP * numberOfLBnodesFine];
+ fMPMdest = &DF[DIR_MPM * numberOfLBnodesFine];
+ fPPMdest = &DF[DIR_PPM * numberOfLBnodesFine];
+ fPPPdest = &DF[DIR_PPP * numberOfLBnodesFine];
+ fPMPdest = &DF[DIR_PMP * numberOfLBnodesFine];
+ fPMMdest = &DF[DIR_PMM * numberOfLBnodesFine];
real
*fP00source, *fM00source, *f0P0source, *f0M0source, *f00Psource, *f00Msource, *fPP0source, *fMM0source, *fPM0source,
@@ -78,63 +78,63 @@ __global__ void scaleCF_0817_comp_27( real* DC,
if (isEvenTimestep == true)
{
- fP00source = &DC[DIR_P00 *size_MatC];
- fM00source = &DC[DIR_M00 *size_MatC];
- f0P0source = &DC[DIR_0P0 *size_MatC];
- f0M0source = &DC[DIR_0M0 *size_MatC];
- f00Psource = &DC[DIR_00P *size_MatC];
- f00Msource = &DC[DIR_00M *size_MatC];
- fPP0source = &DC[DIR_PP0 *size_MatC];
- fMM0source = &DC[DIR_MM0 *size_MatC];
- fPM0source = &DC[DIR_PM0 *size_MatC];
- fMP0source = &DC[DIR_MP0 *size_MatC];
- fP0Psource = &DC[DIR_P0P *size_MatC];
- fM0Msource = &DC[DIR_M0M *size_MatC];
- fP0Msource = &DC[DIR_P0M *size_MatC];
- fM0Psource = &DC[DIR_M0P *size_MatC];
- f0PPsource = &DC[DIR_0PP *size_MatC];
- f0MMsource = &DC[DIR_0MM *size_MatC];
- f0PMsource = &DC[DIR_0PM *size_MatC];
- f0MPsource = &DC[DIR_0MP *size_MatC];
- f000source = &DC[DIR_000*size_MatC];
- fMMMsource = &DC[DIR_MMM *size_MatC];
- fMMPsource = &DC[DIR_MMP *size_MatC];
- fMPPsource = &DC[DIR_MPP *size_MatC];
- fMPMsource = &DC[DIR_MPM *size_MatC];
- fPPMsource = &DC[DIR_PPM *size_MatC];
- fPPPsource = &DC[DIR_PPP *size_MatC];
- fPMPsource = &DC[DIR_PMP *size_MatC];
- fPMMsource = &DC[DIR_PMM *size_MatC];
+ fP00source = &DC[DIR_P00 * numberOfLBnodesCoarse];
+ fM00source = &DC[DIR_M00 * numberOfLBnodesCoarse];
+ f0P0source = &DC[DIR_0P0 * numberOfLBnodesCoarse];
+ f0M0source = &DC[DIR_0M0 * numberOfLBnodesCoarse];
+ f00Psource = &DC[DIR_00P * numberOfLBnodesCoarse];
+ f00Msource = &DC[DIR_00M * numberOfLBnodesCoarse];
+ fPP0source = &DC[DIR_PP0 * numberOfLBnodesCoarse];
+ fMM0source = &DC[DIR_MM0 * numberOfLBnodesCoarse];
+ fPM0source = &DC[DIR_PM0 * numberOfLBnodesCoarse];
+ fMP0source = &DC[DIR_MP0 * numberOfLBnodesCoarse];
+ fP0Psource = &DC[DIR_P0P * numberOfLBnodesCoarse];
+ fM0Msource = &DC[DIR_M0M * numberOfLBnodesCoarse];
+ fP0Msource = &DC[DIR_P0M * numberOfLBnodesCoarse];
+ fM0Psource = &DC[DIR_M0P * numberOfLBnodesCoarse];
+ f0PPsource = &DC[DIR_0PP * numberOfLBnodesCoarse];
+ f0MMsource = &DC[DIR_0MM * numberOfLBnodesCoarse];
+ f0PMsource = &DC[DIR_0PM * numberOfLBnodesCoarse];
+ f0MPsource = &DC[DIR_0MP * numberOfLBnodesCoarse];
+ f000source = &DC[DIR_000 * numberOfLBnodesCoarse];
+ fMMMsource = &DC[DIR_MMM * numberOfLBnodesCoarse];
+ fMMPsource = &DC[DIR_MMP * numberOfLBnodesCoarse];
+ fMPPsource = &DC[DIR_MPP * numberOfLBnodesCoarse];
+ fMPMsource = &DC[DIR_MPM * numberOfLBnodesCoarse];
+ fPPMsource = &DC[DIR_PPM * numberOfLBnodesCoarse];
+ fPPPsource = &DC[DIR_PPP * numberOfLBnodesCoarse];
+ fPMPsource = &DC[DIR_PMP * numberOfLBnodesCoarse];
+ fPMMsource = &DC[DIR_PMM * numberOfLBnodesCoarse];
}
else
{
- fP00source = &DC[DIR_M00 *size_MatC];
- fM00source = &DC[DIR_P00 *size_MatC];
- f0P0source = &DC[DIR_0M0 *size_MatC];
- f0M0source = &DC[DIR_0P0 *size_MatC];
- f00Psource = &DC[DIR_00M *size_MatC];
- f00Msource = &DC[DIR_00P *size_MatC];
- fPP0source = &DC[DIR_MM0 *size_MatC];
- fMM0source = &DC[DIR_PP0 *size_MatC];
- fPM0source = &DC[DIR_MP0 *size_MatC];
- fMP0source = &DC[DIR_PM0 *size_MatC];
- fP0Psource = &DC[DIR_M0M *size_MatC];
- fM0Msource = &DC[DIR_P0P *size_MatC];
- fP0Msource = &DC[DIR_M0P *size_MatC];
- fM0Psource = &DC[DIR_P0M *size_MatC];
- f0PPsource = &DC[DIR_0MM *size_MatC];
- f0MMsource = &DC[DIR_0PP *size_MatC];
- f0PMsource = &DC[DIR_0MP *size_MatC];
- f0MPsource = &DC[DIR_0PM *size_MatC];
- f000source = &DC[DIR_000*size_MatC];
- fMMMsource = &DC[DIR_PPP *size_MatC];
- fMMPsource = &DC[DIR_PPM *size_MatC];
- fMPPsource = &DC[DIR_PMM *size_MatC];
- fMPMsource = &DC[DIR_PMP *size_MatC];
- fPPMsource = &DC[DIR_MMP *size_MatC];
- fPPPsource = &DC[DIR_MMM *size_MatC];
- fPMPsource = &DC[DIR_MPM *size_MatC];
- fPMMsource = &DC[DIR_MPP *size_MatC];
+ fP00source = &DC[DIR_M00 * numberOfLBnodesCoarse];
+ fM00source = &DC[DIR_P00 * numberOfLBnodesCoarse];
+ f0P0source = &DC[DIR_0M0 * numberOfLBnodesCoarse];
+ f0M0source = &DC[DIR_0P0 * numberOfLBnodesCoarse];
+ f00Psource = &DC[DIR_00M * numberOfLBnodesCoarse];
+ f00Msource = &DC[DIR_00P * numberOfLBnodesCoarse];
+ fPP0source = &DC[DIR_MM0 * numberOfLBnodesCoarse];
+ fMM0source = &DC[DIR_PP0 * numberOfLBnodesCoarse];
+ fPM0source = &DC[DIR_MP0 * numberOfLBnodesCoarse];
+ fMP0source = &DC[DIR_PM0 * numberOfLBnodesCoarse];
+ fP0Psource = &DC[DIR_M0M * numberOfLBnodesCoarse];
+ fM0Msource = &DC[DIR_P0P * numberOfLBnodesCoarse];
+ fP0Msource = &DC[DIR_M0P * numberOfLBnodesCoarse];
+ fM0Psource = &DC[DIR_P0M * numberOfLBnodesCoarse];
+ f0PPsource = &DC[DIR_0MM * numberOfLBnodesCoarse];
+ f0MMsource = &DC[DIR_0PP * numberOfLBnodesCoarse];
+ f0PMsource = &DC[DIR_0MP * numberOfLBnodesCoarse];
+ f0MPsource = &DC[DIR_0PM * numberOfLBnodesCoarse];
+ f000source = &DC[DIR_000 * numberOfLBnodesCoarse];
+ fMMMsource = &DC[DIR_PPP * numberOfLBnodesCoarse];
+ fMMPsource = &DC[DIR_PPM * numberOfLBnodesCoarse];
+ fMPPsource = &DC[DIR_PMM * numberOfLBnodesCoarse];
+ fMPMsource = &DC[DIR_PMP * numberOfLBnodesCoarse];
+ fPPMsource = &DC[DIR_MMP * numberOfLBnodesCoarse];
+ fPPPsource = &DC[DIR_MMM * numberOfLBnodesCoarse];
+ fPMPsource = &DC[DIR_MPM * numberOfLBnodesCoarse];
+ fPMMsource = &DC[DIR_MPP * numberOfLBnodesCoarse];
}
////////////////////////////////////////////////////////////////////////////////
const unsigned ix = threadIdx.x; // Globaler x-Index
@@ -4091,8 +4091,8 @@ __global__ void scaleCF_AA2016_comp_27(real* DC,
unsigned int* neighborFX,
unsigned int* neighborFY,
unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
+ unsigned long long numberOfLBnodesCoarse,
+ unsigned long long numberOfLBnodesFine,
bool isEvenTimestep,
unsigned int* posCSWB,
unsigned int* posFSWB,
@@ -4109,96 +4109,96 @@ __global__ void scaleCF_AA2016_comp_27(real* DC,
real *feF, *fwF, *fnF, *fsF, *ftF, *fbF, *fneF, *fswF, *fseF, *fnwF, *fteF, *fbwF, *fbeF, *ftwF, *ftnF, *fbsF, *fbnF, *ftsF, *fzeroF,
*ftneF, *ftswF, *ftseF, *ftnwF, *fbneF, *fbswF, *fbseF, *fbnwF;
- feF = &DF[DIR_P00 *size_MatF];
- fwF = &DF[DIR_M00 *size_MatF];
- fnF = &DF[DIR_0P0 *size_MatF];
- fsF = &DF[DIR_0M0 *size_MatF];
- ftF = &DF[DIR_00P *size_MatF];
- fbF = &DF[DIR_00M *size_MatF];
- fneF = &DF[DIR_PP0 *size_MatF];
- fswF = &DF[DIR_MM0 *size_MatF];
- fseF = &DF[DIR_PM0 *size_MatF];
- fnwF = &DF[DIR_MP0 *size_MatF];
- fteF = &DF[DIR_P0P *size_MatF];
- fbwF = &DF[DIR_M0M *size_MatF];
- fbeF = &DF[DIR_P0M *size_MatF];
- ftwF = &DF[DIR_M0P *size_MatF];
- ftnF = &DF[DIR_0PP *size_MatF];
- fbsF = &DF[DIR_0MM *size_MatF];
- fbnF = &DF[DIR_0PM *size_MatF];
- ftsF = &DF[DIR_0MP *size_MatF];
- fzeroF = &DF[DIR_000*size_MatF];
- ftneF = &DF[DIR_PPP *size_MatF];
- ftswF = &DF[DIR_MMP *size_MatF];
- ftseF = &DF[DIR_PMP *size_MatF];
- ftnwF = &DF[DIR_MPP *size_MatF];
- fbneF = &DF[DIR_PPM *size_MatF];
- fbswF = &DF[DIR_MMM *size_MatF];
- fbseF = &DF[DIR_PMM *size_MatF];
- fbnwF = &DF[DIR_MPM *size_MatF];
+ feF = &DF[DIR_P00 * numberOfLBnodesFine];
+ fwF = &DF[DIR_M00 * numberOfLBnodesFine];
+ fnF = &DF[DIR_0P0 * numberOfLBnodesFine];
+ fsF = &DF[DIR_0M0 * numberOfLBnodesFine];
+ ftF = &DF[DIR_00P * numberOfLBnodesFine];
+ fbF = &DF[DIR_00M * numberOfLBnodesFine];
+ fneF = &DF[DIR_PP0 * numberOfLBnodesFine];
+ fswF = &DF[DIR_MM0 * numberOfLBnodesFine];
+ fseF = &DF[DIR_PM0 * numberOfLBnodesFine];
+ fnwF = &DF[DIR_MP0 * numberOfLBnodesFine];
+ fteF = &DF[DIR_P0P * numberOfLBnodesFine];
+ fbwF = &DF[DIR_M0M * numberOfLBnodesFine];
+ fbeF = &DF[DIR_P0M * numberOfLBnodesFine];
+ ftwF = &DF[DIR_M0P * numberOfLBnodesFine];
+ ftnF = &DF[DIR_0PP * numberOfLBnodesFine];
+ fbsF = &DF[DIR_0MM * numberOfLBnodesFine];
+ fbnF = &DF[DIR_0PM * numberOfLBnodesFine];
+ ftsF = &DF[DIR_0MP * numberOfLBnodesFine];
+ fzeroF = &DF[DIR_000 * numberOfLBnodesFine];
+ ftneF = &DF[DIR_PPP * numberOfLBnodesFine];
+ ftswF = &DF[DIR_MMP * numberOfLBnodesFine];
+ ftseF = &DF[DIR_PMP * numberOfLBnodesFine];
+ ftnwF = &DF[DIR_MPP * numberOfLBnodesFine];
+ fbneF = &DF[DIR_PPM * numberOfLBnodesFine];
+ fbswF = &DF[DIR_MMM * numberOfLBnodesFine];
+ fbseF = &DF[DIR_PMM * numberOfLBnodesFine];
+ fbnwF = &DF[DIR_MPM * numberOfLBnodesFine];
real *feC, *fwC, *fnC, *fsC, *ftC, *fbC, *fneC, *fswC, *fseC, *fnwC, *fteC, *fbwC, *fbeC, *ftwC, *ftnC, *fbsC, *fbnC, *ftsC, *fzeroC,
*ftneC, *ftswC, *ftseC, *ftnwC, *fbneC, *fbswC, *fbseC, *fbnwC;
if (isEvenTimestep==true)
{
- feC = &DC[DIR_P00 *size_MatC];
- fwC = &DC[DIR_M00 *size_MatC];
- fnC = &DC[DIR_0P0 *size_MatC];
- fsC = &DC[DIR_0M0 *size_MatC];
- ftC = &DC[DIR_00P *size_MatC];
- fbC = &DC[DIR_00M *size_MatC];
- fneC = &DC[DIR_PP0 *size_MatC];
- fswC = &DC[DIR_MM0 *size_MatC];
- fseC = &DC[DIR_PM0 *size_MatC];
- fnwC = &DC[DIR_MP0 *size_MatC];
- fteC = &DC[DIR_P0P *size_MatC];
- fbwC = &DC[DIR_M0M *size_MatC];
- fbeC = &DC[DIR_P0M *size_MatC];
- ftwC = &DC[DIR_M0P *size_MatC];
- ftnC = &DC[DIR_0PP *size_MatC];
- fbsC = &DC[DIR_0MM *size_MatC];
- fbnC = &DC[DIR_0PM *size_MatC];
- ftsC = &DC[DIR_0MP *size_MatC];
- fzeroC = &DC[DIR_000*size_MatC];
- ftneC = &DC[DIR_PPP *size_MatC];
- ftswC = &DC[DIR_MMP *size_MatC];
- ftseC = &DC[DIR_PMP *size_MatC];
- ftnwC = &DC[DIR_MPP *size_MatC];
- fbneC = &DC[DIR_PPM *size_MatC];
- fbswC = &DC[DIR_MMM *size_MatC];
- fbseC = &DC[DIR_PMM *size_MatC];
- fbnwC = &DC[DIR_MPM *size_MatC];
+ feC = &DC[DIR_P00 * numberOfLBnodesCoarse];
+ fwC = &DC[DIR_M00 * numberOfLBnodesCoarse];
+ fnC = &DC[DIR_0P0 * numberOfLBnodesCoarse];
+ fsC = &DC[DIR_0M0 * numberOfLBnodesCoarse];
+ ftC = &DC[DIR_00P * numberOfLBnodesCoarse];
+ fbC = &DC[DIR_00M * numberOfLBnodesCoarse];
+ fneC = &DC[DIR_PP0 * numberOfLBnodesCoarse];
+ fswC = &DC[DIR_MM0 * numberOfLBnodesCoarse];
+ fseC = &DC[DIR_PM0 * numberOfLBnodesCoarse];
+ fnwC = &DC[DIR_MP0 * numberOfLBnodesCoarse];
+ fteC = &DC[DIR_P0P * numberOfLBnodesCoarse];
+ fbwC = &DC[DIR_M0M * numberOfLBnodesCoarse];
+ fbeC = &DC[DIR_P0M * numberOfLBnodesCoarse];
+ ftwC = &DC[DIR_M0P * numberOfLBnodesCoarse];
+ ftnC = &DC[DIR_0PP * numberOfLBnodesCoarse];
+ fbsC = &DC[DIR_0MM * numberOfLBnodesCoarse];
+ fbnC = &DC[DIR_0PM * numberOfLBnodesCoarse];
+ ftsC = &DC[DIR_0MP * numberOfLBnodesCoarse];
+ fzeroC = &DC[DIR_000 * numberOfLBnodesCoarse];
+ ftneC = &DC[DIR_PPP * numberOfLBnodesCoarse];
+ ftswC = &DC[DIR_MMP * numberOfLBnodesCoarse];
+ ftseC = &DC[DIR_PMP * numberOfLBnodesCoarse];
+ ftnwC = &DC[DIR_MPP * numberOfLBnodesCoarse];
+ fbneC = &DC[DIR_PPM * numberOfLBnodesCoarse];
+ fbswC = &DC[DIR_MMM * numberOfLBnodesCoarse];
+ fbseC = &DC[DIR_PMM * numberOfLBnodesCoarse];
+ fbnwC = &DC[DIR_MPM * numberOfLBnodesCoarse];
}
else
{
- fwC = &DC[DIR_P00 *size_MatC];
- feC = &DC[DIR_M00 *size_MatC];
- fsC = &DC[DIR_0P0 *size_MatC];
- fnC = &DC[DIR_0M0 *size_MatC];
- fbC = &DC[DIR_00P *size_MatC];
- ftC = &DC[DIR_00M *size_MatC];
- fswC = &DC[DIR_PP0 *size_MatC];
- fneC = &DC[DIR_MM0 *size_MatC];
- fnwC = &DC[DIR_PM0 *size_MatC];
- fseC = &DC[DIR_MP0 *size_MatC];
- fbwC = &DC[DIR_P0P *size_MatC];
- fteC = &DC[DIR_M0M *size_MatC];
- ftwC = &DC[DIR_P0M *size_MatC];
- fbeC = &DC[DIR_M0P *size_MatC];
- fbsC = &DC[DIR_0PP *size_MatC];
- ftnC = &DC[DIR_0MM *size_MatC];
- ftsC = &DC[DIR_0PM *size_MatC];
- fbnC = &DC[DIR_0MP *size_MatC];
- fzeroC = &DC[DIR_000*size_MatC];
- fbswC = &DC[DIR_PPP *size_MatC];
- fbneC = &DC[DIR_MMP *size_MatC];
- fbnwC = &DC[DIR_PMP *size_MatC];
- fbseC = &DC[DIR_MPP *size_MatC];
- ftswC = &DC[DIR_PPM *size_MatC];
- ftneC = &DC[DIR_MMM *size_MatC];
- ftnwC = &DC[DIR_PMM *size_MatC];
- ftseC = &DC[DIR_MPM *size_MatC];
+ fwC = &DC[DIR_P00 * numberOfLBnodesCoarse];
+ feC = &DC[DIR_M00 * numberOfLBnodesCoarse];
+ fsC = &DC[DIR_0P0 * numberOfLBnodesCoarse];
+ fnC = &DC[DIR_0M0 * numberOfLBnodesCoarse];
+ fbC = &DC[DIR_00P * numberOfLBnodesCoarse];
+ ftC = &DC[DIR_00M * numberOfLBnodesCoarse];
+ fswC = &DC[DIR_PP0 * numberOfLBnodesCoarse];
+ fneC = &DC[DIR_MM0 * numberOfLBnodesCoarse];
+ fnwC = &DC[DIR_PM0 * numberOfLBnodesCoarse];
+ fseC = &DC[DIR_MP0 * numberOfLBnodesCoarse];
+ fbwC = &DC[DIR_P0P * numberOfLBnodesCoarse];
+ fteC = &DC[DIR_M0M * numberOfLBnodesCoarse];
+ ftwC = &DC[DIR_P0M * numberOfLBnodesCoarse];
+ fbeC = &DC[DIR_M0P * numberOfLBnodesCoarse];
+ fbsC = &DC[DIR_0PP * numberOfLBnodesCoarse];
+ ftnC = &DC[DIR_0MM * numberOfLBnodesCoarse];
+ ftsC = &DC[DIR_0PM * numberOfLBnodesCoarse];
+ fbnC = &DC[DIR_0MP * numberOfLBnodesCoarse];
+ fzeroC = &DC[DIR_000 * numberOfLBnodesCoarse];
+ fbswC = &DC[DIR_PPP * numberOfLBnodesCoarse];
+ fbneC = &DC[DIR_MMP * numberOfLBnodesCoarse];
+ fbnwC = &DC[DIR_PMP * numberOfLBnodesCoarse];
+ fbseC = &DC[DIR_MPP * numberOfLBnodesCoarse];
+ ftswC = &DC[DIR_PPM * numberOfLBnodesCoarse];
+ ftneC = &DC[DIR_MMM * numberOfLBnodesCoarse];
+ ftnwC = &DC[DIR_PMM * numberOfLBnodesCoarse];
+ ftseC = &DC[DIR_MPM * numberOfLBnodesCoarse];
}
////////////////////////////////////////////////////////////////////////////////
@@ -10974,8 +10974,8 @@ __global__ void scaleCF_RhoSq_3rdMom_comp_27(real* DC,
unsigned int* neighborFX,
unsigned int* neighborFY,
unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
+ unsigned long long numberOfLBnodesCoarse,
+ unsigned long long numberOfLBnodesFine,
bool isEvenTimestep,
unsigned int* posCSWB,
unsigned int* posFSWB,
@@ -10992,96 +10992,96 @@ __global__ void scaleCF_RhoSq_3rdMom_comp_27(real* DC,
real *feF, *fwF, *fnF, *fsF, *ftF, *fbF, *fneF, *fswF, *fseF, *fnwF, *fteF, *fbwF, *fbeF, *ftwF, *ftnF, *fbsF, *fbnF, *ftsF, *fzeroF,
*ftneF, *ftswF, *ftseF, *ftnwF, *fbneF, *fbswF, *fbseF, *fbnwF;
- feF = &DF[DIR_P00 *size_MatF];
- fwF = &DF[DIR_M00 *size_MatF];
- fnF = &DF[DIR_0P0 *size_MatF];
- fsF = &DF[DIR_0M0 *size_MatF];
- ftF = &DF[DIR_00P *size_MatF];
- fbF = &DF[DIR_00M *size_MatF];
- fneF = &DF[DIR_PP0 *size_MatF];
- fswF = &DF[DIR_MM0 *size_MatF];
- fseF = &DF[DIR_PM0 *size_MatF];
- fnwF = &DF[DIR_MP0 *size_MatF];
- fteF = &DF[DIR_P0P *size_MatF];
- fbwF = &DF[DIR_M0M *size_MatF];
- fbeF = &DF[DIR_P0M *size_MatF];
- ftwF = &DF[DIR_M0P *size_MatF];
- ftnF = &DF[DIR_0PP *size_MatF];
- fbsF = &DF[DIR_0MM *size_MatF];
- fbnF = &DF[DIR_0PM *size_MatF];
- ftsF = &DF[DIR_0MP *size_MatF];
- fzeroF = &DF[DIR_000*size_MatF];
- ftneF = &DF[DIR_PPP *size_MatF];
- ftswF = &DF[DIR_MMP *size_MatF];
- ftseF = &DF[DIR_PMP *size_MatF];
- ftnwF = &DF[DIR_MPP *size_MatF];
- fbneF = &DF[DIR_PPM *size_MatF];
- fbswF = &DF[DIR_MMM *size_MatF];
- fbseF = &DF[DIR_PMM *size_MatF];
- fbnwF = &DF[DIR_MPM *size_MatF];
+ feF = &DF[DIR_P00 * numberOfLBnodesFine];
+ fwF = &DF[DIR_M00 * numberOfLBnodesFine];
+ fnF = &DF[DIR_0P0 * numberOfLBnodesFine];
+ fsF = &DF[DIR_0M0 * numberOfLBnodesFine];
+ ftF = &DF[DIR_00P * numberOfLBnodesFine];
+ fbF = &DF[DIR_00M * numberOfLBnodesFine];
+ fneF = &DF[DIR_PP0 * numberOfLBnodesFine];
+ fswF = &DF[DIR_MM0 * numberOfLBnodesFine];
+ fseF = &DF[DIR_PM0 * numberOfLBnodesFine];
+ fnwF = &DF[DIR_MP0 * numberOfLBnodesFine];
+ fteF = &DF[DIR_P0P * numberOfLBnodesFine];
+ fbwF = &DF[DIR_M0M * numberOfLBnodesFine];
+ fbeF = &DF[DIR_P0M * numberOfLBnodesFine];
+ ftwF = &DF[DIR_M0P * numberOfLBnodesFine];
+ ftnF = &DF[DIR_0PP * numberOfLBnodesFine];
+ fbsF = &DF[DIR_0MM * numberOfLBnodesFine];
+ fbnF = &DF[DIR_0PM * numberOfLBnodesFine];
+ ftsF = &DF[DIR_0MP * numberOfLBnodesFine];
+ fzeroF = &DF[DIR_000 * numberOfLBnodesFine];
+ ftneF = &DF[DIR_PPP * numberOfLBnodesFine];
+ ftswF = &DF[DIR_MMP * numberOfLBnodesFine];
+ ftseF = &DF[DIR_PMP * numberOfLBnodesFine];
+ ftnwF = &DF[DIR_MPP * numberOfLBnodesFine];
+ fbneF = &DF[DIR_PPM * numberOfLBnodesFine];
+ fbswF = &DF[DIR_MMM * numberOfLBnodesFine];
+ fbseF = &DF[DIR_PMM * numberOfLBnodesFine];
+ fbnwF = &DF[DIR_MPM * numberOfLBnodesFine];
real *feC, *fwC, *fnC, *fsC, *ftC, *fbC, *fneC, *fswC, *fseC, *fnwC, *fteC, *fbwC, *fbeC, *ftwC, *ftnC, *fbsC, *fbnC, *ftsC, *fzeroC,
*ftneC, *ftswC, *ftseC, *ftnwC, *fbneC, *fbswC, *fbseC, *fbnwC;
if (isEvenTimestep==true)
{
- feC = &DC[DIR_P00 *size_MatC];
- fwC = &DC[DIR_M00 *size_MatC];
- fnC = &DC[DIR_0P0 *size_MatC];
- fsC = &DC[DIR_0M0 *size_MatC];
- ftC = &DC[DIR_00P *size_MatC];
- fbC = &DC[DIR_00M *size_MatC];
- fneC = &DC[DIR_PP0 *size_MatC];
- fswC = &DC[DIR_MM0 *size_MatC];
- fseC = &DC[DIR_PM0 *size_MatC];
- fnwC = &DC[DIR_MP0 *size_MatC];
- fteC = &DC[DIR_P0P *size_MatC];
- fbwC = &DC[DIR_M0M *size_MatC];
- fbeC = &DC[DIR_P0M *size_MatC];
- ftwC = &DC[DIR_M0P *size_MatC];
- ftnC = &DC[DIR_0PP *size_MatC];
- fbsC = &DC[DIR_0MM *size_MatC];
- fbnC = &DC[DIR_0PM *size_MatC];
- ftsC = &DC[DIR_0MP *size_MatC];
- fzeroC = &DC[DIR_000*size_MatC];
- ftneC = &DC[DIR_PPP *size_MatC];
- ftswC = &DC[DIR_MMP *size_MatC];
- ftseC = &DC[DIR_PMP *size_MatC];
- ftnwC = &DC[DIR_MPP *size_MatC];
- fbneC = &DC[DIR_PPM *size_MatC];
- fbswC = &DC[DIR_MMM *size_MatC];
- fbseC = &DC[DIR_PMM *size_MatC];
- fbnwC = &DC[DIR_MPM *size_MatC];
+ feC = &DC[DIR_P00 * numberOfLBnodesCoarse];
+ fwC = &DC[DIR_M00 * numberOfLBnodesCoarse];
+ fnC = &DC[DIR_0P0 * numberOfLBnodesCoarse];
+ fsC = &DC[DIR_0M0 * numberOfLBnodesCoarse];
+ ftC = &DC[DIR_00P * numberOfLBnodesCoarse];
+ fbC = &DC[DIR_00M * numberOfLBnodesCoarse];
+ fneC = &DC[DIR_PP0 * numberOfLBnodesCoarse];
+ fswC = &DC[DIR_MM0 * numberOfLBnodesCoarse];
+ fseC = &DC[DIR_PM0 * numberOfLBnodesCoarse];
+ fnwC = &DC[DIR_MP0 * numberOfLBnodesCoarse];
+ fteC = &DC[DIR_P0P * numberOfLBnodesCoarse];
+ fbwC = &DC[DIR_M0M * numberOfLBnodesCoarse];
+ fbeC = &DC[DIR_P0M * numberOfLBnodesCoarse];
+ ftwC = &DC[DIR_M0P * numberOfLBnodesCoarse];
+ ftnC = &DC[DIR_0PP * numberOfLBnodesCoarse];
+ fbsC = &DC[DIR_0MM * numberOfLBnodesCoarse];
+ fbnC = &DC[DIR_0PM * numberOfLBnodesCoarse];
+ ftsC = &DC[DIR_0MP * numberOfLBnodesCoarse];
+ fzeroC = &DC[DIR_000 * numberOfLBnodesCoarse];
+ ftneC = &DC[DIR_PPP * numberOfLBnodesCoarse];
+ ftswC = &DC[DIR_MMP * numberOfLBnodesCoarse];
+ ftseC = &DC[DIR_PMP * numberOfLBnodesCoarse];
+ ftnwC = &DC[DIR_MPP * numberOfLBnodesCoarse];
+ fbneC = &DC[DIR_PPM * numberOfLBnodesCoarse];
+ fbswC = &DC[DIR_MMM * numberOfLBnodesCoarse];
+ fbseC = &DC[DIR_PMM * numberOfLBnodesCoarse];
+ fbnwC = &DC[DIR_MPM * numberOfLBnodesCoarse];
}
else
{
- fwC = &DC[DIR_P00 *size_MatC];
- feC = &DC[DIR_M00 *size_MatC];
- fsC = &DC[DIR_0P0 *size_MatC];
- fnC = &DC[DIR_0M0 *size_MatC];
- fbC = &DC[DIR_00P *size_MatC];
- ftC = &DC[DIR_00M *size_MatC];
- fswC = &DC[DIR_PP0 *size_MatC];
- fneC = &DC[DIR_MM0 *size_MatC];
- fnwC = &DC[DIR_PM0 *size_MatC];
- fseC = &DC[DIR_MP0 *size_MatC];
- fbwC = &DC[DIR_P0P *size_MatC];
- fteC = &DC[DIR_M0M *size_MatC];
- ftwC = &DC[DIR_P0M *size_MatC];
- fbeC = &DC[DIR_M0P *size_MatC];
- fbsC = &DC[DIR_0PP *size_MatC];
- ftnC = &DC[DIR_0MM *size_MatC];
- ftsC = &DC[DIR_0PM *size_MatC];
- fbnC = &DC[DIR_0MP *size_MatC];
- fzeroC = &DC[DIR_000*size_MatC];
- fbswC = &DC[DIR_PPP *size_MatC];
- fbneC = &DC[DIR_MMP *size_MatC];
- fbnwC = &DC[DIR_PMP *size_MatC];
- fbseC = &DC[DIR_MPP *size_MatC];
- ftswC = &DC[DIR_PPM *size_MatC];
- ftneC = &DC[DIR_MMM *size_MatC];
- ftnwC = &DC[DIR_PMM *size_MatC];
- ftseC = &DC[DIR_MPM *size_MatC];
+ fwC = &DC[DIR_P00 * numberOfLBnodesCoarse];
+ feC = &DC[DIR_M00 * numberOfLBnodesCoarse];
+ fsC = &DC[DIR_0P0 * numberOfLBnodesCoarse];
+ fnC = &DC[DIR_0M0 * numberOfLBnodesCoarse];
+ fbC = &DC[DIR_00P * numberOfLBnodesCoarse];
+ ftC = &DC[DIR_00M * numberOfLBnodesCoarse];
+ fswC = &DC[DIR_PP0 * numberOfLBnodesCoarse];
+ fneC = &DC[DIR_MM0 * numberOfLBnodesCoarse];
+ fnwC = &DC[DIR_PM0 * numberOfLBnodesCoarse];
+ fseC = &DC[DIR_MP0 * numberOfLBnodesCoarse];
+ fbwC = &DC[DIR_P0P * numberOfLBnodesCoarse];
+ fteC = &DC[DIR_M0M * numberOfLBnodesCoarse];
+ ftwC = &DC[DIR_P0M * numberOfLBnodesCoarse];
+ fbeC = &DC[DIR_M0P * numberOfLBnodesCoarse];
+ fbsC = &DC[DIR_0PP * numberOfLBnodesCoarse];
+ ftnC = &DC[DIR_0MM * numberOfLBnodesCoarse];
+ ftsC = &DC[DIR_0PM * numberOfLBnodesCoarse];
+ fbnC = &DC[DIR_0MP * numberOfLBnodesCoarse];
+ fzeroC = &DC[DIR_000 * numberOfLBnodesCoarse];
+ fbswC = &DC[DIR_PPP * numberOfLBnodesCoarse];
+ fbneC = &DC[DIR_MMP * numberOfLBnodesCoarse];
+ fbnwC = &DC[DIR_PMP * numberOfLBnodesCoarse];
+ fbseC = &DC[DIR_MPP * numberOfLBnodesCoarse];
+ ftswC = &DC[DIR_PPM * numberOfLBnodesCoarse];
+ ftneC = &DC[DIR_MMM * numberOfLBnodesCoarse];
+ ftnwC = &DC[DIR_PMM * numberOfLBnodesCoarse];
+ ftseC = &DC[DIR_MPM * numberOfLBnodesCoarse];
}
////////////////////////////////////////////////////////////////////////////////
@@ -17849,8 +17849,8 @@ __global__ void scaleCF_RhoSq_comp_27(real* DC,
unsigned int* neighborFX,
unsigned int* neighborFY,
unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
+ unsigned long long numberOfLBnodesCoarse,
+ unsigned long long numberOfLBnodesFine,
bool isEvenTimestep,
unsigned int* posCSWB,
unsigned int* posFSWB,
@@ -17867,96 +17867,96 @@ __global__ void scaleCF_RhoSq_comp_27(real* DC,
real *feF, *fwF, *fnF, *fsF, *ftF, *fbF, *fneF, *fswF, *fseF, *fnwF, *fteF, *fbwF, *fbeF, *ftwF, *ftnF, *fbsF, *fbnF, *ftsF, *fzeroF,
*ftneF, *ftswF, *ftseF, *ftnwF, *fbneF, *fbswF, *fbseF, *fbnwF;
- feF = &DF[DIR_P00 *size_MatF];
- fwF = &DF[DIR_M00 *size_MatF];
- fnF = &DF[DIR_0P0 *size_MatF];
- fsF = &DF[DIR_0M0 *size_MatF];
- ftF = &DF[DIR_00P *size_MatF];
- fbF = &DF[DIR_00M *size_MatF];
- fneF = &DF[DIR_PP0 *size_MatF];
- fswF = &DF[DIR_MM0 *size_MatF];
- fseF = &DF[DIR_PM0 *size_MatF];
- fnwF = &DF[DIR_MP0 *size_MatF];
- fteF = &DF[DIR_P0P *size_MatF];
- fbwF = &DF[DIR_M0M *size_MatF];
- fbeF = &DF[DIR_P0M *size_MatF];
- ftwF = &DF[DIR_M0P *size_MatF];
- ftnF = &DF[DIR_0PP *size_MatF];
- fbsF = &DF[DIR_0MM *size_MatF];
- fbnF = &DF[DIR_0PM *size_MatF];
- ftsF = &DF[DIR_0MP *size_MatF];
- fzeroF = &DF[DIR_000*size_MatF];
- ftneF = &DF[DIR_PPP *size_MatF];
- ftswF = &DF[DIR_MMP *size_MatF];
- ftseF = &DF[DIR_PMP *size_MatF];
- ftnwF = &DF[DIR_MPP *size_MatF];
- fbneF = &DF[DIR_PPM *size_MatF];
- fbswF = &DF[DIR_MMM *size_MatF];
- fbseF = &DF[DIR_PMM *size_MatF];
- fbnwF = &DF[DIR_MPM *size_MatF];
+ feF = &DF[DIR_P00 * numberOfLBnodesFine];
+ fwF = &DF[DIR_M00 * numberOfLBnodesFine];
+ fnF = &DF[DIR_0P0 * numberOfLBnodesFine];
+ fsF = &DF[DIR_0M0 * numberOfLBnodesFine];
+ ftF = &DF[DIR_00P * numberOfLBnodesFine];
+ fbF = &DF[DIR_00M * numberOfLBnodesFine];
+ fneF = &DF[DIR_PP0 * numberOfLBnodesFine];
+ fswF = &DF[DIR_MM0 * numberOfLBnodesFine];
+ fseF = &DF[DIR_PM0 * numberOfLBnodesFine];
+ fnwF = &DF[DIR_MP0 * numberOfLBnodesFine];
+ fteF = &DF[DIR_P0P * numberOfLBnodesFine];
+ fbwF = &DF[DIR_M0M * numberOfLBnodesFine];
+ fbeF = &DF[DIR_P0M * numberOfLBnodesFine];
+ ftwF = &DF[DIR_M0P * numberOfLBnodesFine];
+ ftnF = &DF[DIR_0PP * numberOfLBnodesFine];
+ fbsF = &DF[DIR_0MM * numberOfLBnodesFine];
+ fbnF = &DF[DIR_0PM * numberOfLBnodesFine];
+ ftsF = &DF[DIR_0MP * numberOfLBnodesFine];
+ fzeroF = &DF[DIR_000 * numberOfLBnodesFine];
+ ftneF = &DF[DIR_PPP * numberOfLBnodesFine];
+ ftswF = &DF[DIR_MMP * numberOfLBnodesFine];
+ ftseF = &DF[DIR_PMP * numberOfLBnodesFine];
+ ftnwF = &DF[DIR_MPP * numberOfLBnodesFine];
+ fbneF = &DF[DIR_PPM * numberOfLBnodesFine];
+ fbswF = &DF[DIR_MMM * numberOfLBnodesFine];
+ fbseF = &DF[DIR_PMM * numberOfLBnodesFine];
+ fbnwF = &DF[DIR_MPM * numberOfLBnodesFine];
real *feC, *fwC, *fnC, *fsC, *ftC, *fbC, *fneC, *fswC, *fseC, *fnwC, *fteC, *fbwC, *fbeC, *ftwC, *ftnC, *fbsC, *fbnC, *ftsC, *fzeroC,
*ftneC, *ftswC, *ftseC, *ftnwC, *fbneC, *fbswC, *fbseC, *fbnwC;
if (isEvenTimestep==true)
{
- feC = &DC[DIR_P00 *size_MatC];
- fwC = &DC[DIR_M00 *size_MatC];
- fnC = &DC[DIR_0P0 *size_MatC];
- fsC = &DC[DIR_0M0 *size_MatC];
- ftC = &DC[DIR_00P *size_MatC];
- fbC = &DC[DIR_00M *size_MatC];
- fneC = &DC[DIR_PP0 *size_MatC];
- fswC = &DC[DIR_MM0 *size_MatC];
- fseC = &DC[DIR_PM0 *size_MatC];
- fnwC = &DC[DIR_MP0 *size_MatC];
- fteC = &DC[DIR_P0P *size_MatC];
- fbwC = &DC[DIR_M0M *size_MatC];
- fbeC = &DC[DIR_P0M *size_MatC];
- ftwC = &DC[DIR_M0P *size_MatC];
- ftnC = &DC[DIR_0PP *size_MatC];
- fbsC = &DC[DIR_0MM *size_MatC];
- fbnC = &DC[DIR_0PM *size_MatC];
- ftsC = &DC[DIR_0MP *size_MatC];
- fzeroC = &DC[DIR_000*size_MatC];
- ftneC = &DC[DIR_PPP *size_MatC];
- ftswC = &DC[DIR_MMP *size_MatC];
- ftseC = &DC[DIR_PMP *size_MatC];
- ftnwC = &DC[DIR_MPP *size_MatC];
- fbneC = &DC[DIR_PPM *size_MatC];
- fbswC = &DC[DIR_MMM *size_MatC];
- fbseC = &DC[DIR_PMM *size_MatC];
- fbnwC = &DC[DIR_MPM *size_MatC];
+ feC = &DC[DIR_P00 * numberOfLBnodesCoarse];
+ fwC = &DC[DIR_M00 * numberOfLBnodesCoarse];
+ fnC = &DC[DIR_0P0 * numberOfLBnodesCoarse];
+ fsC = &DC[DIR_0M0 * numberOfLBnodesCoarse];
+ ftC = &DC[DIR_00P * numberOfLBnodesCoarse];
+ fbC = &DC[DIR_00M * numberOfLBnodesCoarse];
+ fneC = &DC[DIR_PP0 * numberOfLBnodesCoarse];
+ fswC = &DC[DIR_MM0 * numberOfLBnodesCoarse];
+ fseC = &DC[DIR_PM0 * numberOfLBnodesCoarse];
+ fnwC = &DC[DIR_MP0 * numberOfLBnodesCoarse];
+ fteC = &DC[DIR_P0P * numberOfLBnodesCoarse];
+ fbwC = &DC[DIR_M0M * numberOfLBnodesCoarse];
+ fbeC = &DC[DIR_P0M * numberOfLBnodesCoarse];
+ ftwC = &DC[DIR_M0P * numberOfLBnodesCoarse];
+ ftnC = &DC[DIR_0PP * numberOfLBnodesCoarse];
+ fbsC = &DC[DIR_0MM * numberOfLBnodesCoarse];
+ fbnC = &DC[DIR_0PM * numberOfLBnodesCoarse];
+ ftsC = &DC[DIR_0MP * numberOfLBnodesCoarse];
+ fzeroC = &DC[DIR_000 * numberOfLBnodesCoarse];
+ ftneC = &DC[DIR_PPP * numberOfLBnodesCoarse];
+ ftswC = &DC[DIR_MMP * numberOfLBnodesCoarse];
+ ftseC = &DC[DIR_PMP * numberOfLBnodesCoarse];
+ ftnwC = &DC[DIR_MPP * numberOfLBnodesCoarse];
+ fbneC = &DC[DIR_PPM * numberOfLBnodesCoarse];
+ fbswC = &DC[DIR_MMM * numberOfLBnodesCoarse];
+ fbseC = &DC[DIR_PMM * numberOfLBnodesCoarse];
+ fbnwC = &DC[DIR_MPM * numberOfLBnodesCoarse];
}
else
{
- fwC = &DC[DIR_P00 *size_MatC];
- feC = &DC[DIR_M00 *size_MatC];
- fsC = &DC[DIR_0P0 *size_MatC];
- fnC = &DC[DIR_0M0 *size_MatC];
- fbC = &DC[DIR_00P *size_MatC];
- ftC = &DC[DIR_00M *size_MatC];
- fswC = &DC[DIR_PP0 *size_MatC];
- fneC = &DC[DIR_MM0 *size_MatC];
- fnwC = &DC[DIR_PM0 *size_MatC];
- fseC = &DC[DIR_MP0 *size_MatC];
- fbwC = &DC[DIR_P0P *size_MatC];
- fteC = &DC[DIR_M0M *size_MatC];
- ftwC = &DC[DIR_P0M *size_MatC];
- fbeC = &DC[DIR_M0P *size_MatC];
- fbsC = &DC[DIR_0PP *size_MatC];
- ftnC = &DC[DIR_0MM *size_MatC];
- ftsC = &DC[DIR_0PM *size_MatC];
- fbnC = &DC[DIR_0MP *size_MatC];
- fzeroC = &DC[DIR_000*size_MatC];
- fbswC = &DC[DIR_PPP *size_MatC];
- fbneC = &DC[DIR_MMP *size_MatC];
- fbnwC = &DC[DIR_PMP *size_MatC];
- fbseC = &DC[DIR_MPP *size_MatC];
- ftswC = &DC[DIR_PPM *size_MatC];
- ftneC = &DC[DIR_MMM *size_MatC];
- ftnwC = &DC[DIR_PMM *size_MatC];
- ftseC = &DC[DIR_MPM *size_MatC];
+ fwC = &DC[DIR_P00 * numberOfLBnodesCoarse];
+ feC = &DC[DIR_M00 * numberOfLBnodesCoarse];
+ fsC = &DC[DIR_0P0 * numberOfLBnodesCoarse];
+ fnC = &DC[DIR_0M0 * numberOfLBnodesCoarse];
+ fbC = &DC[DIR_00P * numberOfLBnodesCoarse];
+ ftC = &DC[DIR_00M * numberOfLBnodesCoarse];
+ fswC = &DC[DIR_PP0 * numberOfLBnodesCoarse];
+ fneC = &DC[DIR_MM0 * numberOfLBnodesCoarse];
+ fnwC = &DC[DIR_PM0 * numberOfLBnodesCoarse];
+ fseC = &DC[DIR_MP0 * numberOfLBnodesCoarse];
+ fbwC = &DC[DIR_P0P * numberOfLBnodesCoarse];
+ fteC = &DC[DIR_M0M * numberOfLBnodesCoarse];
+ ftwC = &DC[DIR_P0M * numberOfLBnodesCoarse];
+ fbeC = &DC[DIR_M0P * numberOfLBnodesCoarse];
+ fbsC = &DC[DIR_0PP * numberOfLBnodesCoarse];
+ ftnC = &DC[DIR_0MM * numberOfLBnodesCoarse];
+ ftsC = &DC[DIR_0PM * numberOfLBnodesCoarse];
+ fbnC = &DC[DIR_0MP * numberOfLBnodesCoarse];
+ fzeroC = &DC[DIR_000 * numberOfLBnodesCoarse];
+ fbswC = &DC[DIR_PPP * numberOfLBnodesCoarse];
+ fbneC = &DC[DIR_MMP * numberOfLBnodesCoarse];
+ fbnwC = &DC[DIR_PMP * numberOfLBnodesCoarse];
+ fbseC = &DC[DIR_MPP * numberOfLBnodesCoarse];
+ ftswC = &DC[DIR_PPM * numberOfLBnodesCoarse];
+ ftneC = &DC[DIR_MMM * numberOfLBnodesCoarse];
+ ftnwC = &DC[DIR_PMM * numberOfLBnodesCoarse];
+ ftseC = &DC[DIR_MPM * numberOfLBnodesCoarse];
}
////////////////////////////////////////////////////////////////////////////////
@@ -22133,8 +22133,8 @@ __global__ void scaleCF_staggered_time_comp_27( real* DC,
unsigned int* neighborFX,
unsigned int* neighborFY,
unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
+ unsigned long long numberOfLBnodesCoarse,
+ unsigned long long numberOfLBnodesFine,
bool isEvenTimestep,
unsigned int* posCSWB,
unsigned int* posFSWB,
@@ -22151,96 +22151,96 @@ __global__ void scaleCF_staggered_time_comp_27( real* DC,
real *feF, *fwF, *fnF, *fsF, *ftF, *fbF, *fneF, *fswF, *fseF, *fnwF, *fteF, *fbwF, *fbeF, *ftwF, *ftnF, *fbsF, *fbnF, *ftsF, *fzeroF,
*ftneF, *ftswF, *ftseF, *ftnwF, *fbneF, *fbswF, *fbseF, *fbnwF;
- feF = &DF[DIR_P00 *size_MatF];
- fwF = &DF[DIR_M00 *size_MatF];
- fnF = &DF[DIR_0P0 *size_MatF];
- fsF = &DF[DIR_0M0 *size_MatF];
- ftF = &DF[DIR_00P *size_MatF];
- fbF = &DF[DIR_00M *size_MatF];
- fneF = &DF[DIR_PP0 *size_MatF];
- fswF = &DF[DIR_MM0 *size_MatF];
- fseF = &DF[DIR_PM0 *size_MatF];
- fnwF = &DF[DIR_MP0 *size_MatF];
- fteF = &DF[DIR_P0P *size_MatF];
- fbwF = &DF[DIR_M0M *size_MatF];
- fbeF = &DF[DIR_P0M *size_MatF];
- ftwF = &DF[DIR_M0P *size_MatF];
- ftnF = &DF[DIR_0PP *size_MatF];
- fbsF = &DF[DIR_0MM *size_MatF];
- fbnF = &DF[DIR_0PM *size_MatF];
- ftsF = &DF[DIR_0MP *size_MatF];
- fzeroF = &DF[DIR_000*size_MatF];
- ftneF = &DF[DIR_PPP *size_MatF];
- ftswF = &DF[DIR_MMP *size_MatF];
- ftseF = &DF[DIR_PMP *size_MatF];
- ftnwF = &DF[DIR_MPP *size_MatF];
- fbneF = &DF[DIR_PPM *size_MatF];
- fbswF = &DF[DIR_MMM *size_MatF];
- fbseF = &DF[DIR_PMM *size_MatF];
- fbnwF = &DF[DIR_MPM *size_MatF];
+ feF = &DF[DIR_P00 * numberOfLBnodesFine];
+ fwF = &DF[DIR_M00 * numberOfLBnodesFine];
+ fnF = &DF[DIR_0P0 * numberOfLBnodesFine];
+ fsF = &DF[DIR_0M0 * numberOfLBnodesFine];
+ ftF = &DF[DIR_00P * numberOfLBnodesFine];
+ fbF = &DF[DIR_00M * numberOfLBnodesFine];
+ fneF = &DF[DIR_PP0 * numberOfLBnodesFine];
+ fswF = &DF[DIR_MM0 * numberOfLBnodesFine];
+ fseF = &DF[DIR_PM0 * numberOfLBnodesFine];
+ fnwF = &DF[DIR_MP0 * numberOfLBnodesFine];
+ fteF = &DF[DIR_P0P * numberOfLBnodesFine];
+ fbwF = &DF[DIR_M0M * numberOfLBnodesFine];
+ fbeF = &DF[DIR_P0M * numberOfLBnodesFine];
+ ftwF = &DF[DIR_M0P * numberOfLBnodesFine];
+ ftnF = &DF[DIR_0PP * numberOfLBnodesFine];
+ fbsF = &DF[DIR_0MM * numberOfLBnodesFine];
+ fbnF = &DF[DIR_0PM * numberOfLBnodesFine];
+ ftsF = &DF[DIR_0MP * numberOfLBnodesFine];
+ fzeroF = &DF[DIR_000 * numberOfLBnodesFine];
+ ftneF = &DF[DIR_PPP * numberOfLBnodesFine];
+ ftswF = &DF[DIR_MMP * numberOfLBnodesFine];
+ ftseF = &DF[DIR_PMP * numberOfLBnodesFine];
+ ftnwF = &DF[DIR_MPP * numberOfLBnodesFine];
+ fbneF = &DF[DIR_PPM * numberOfLBnodesFine];
+ fbswF = &DF[DIR_MMM * numberOfLBnodesFine];
+ fbseF = &DF[DIR_PMM * numberOfLBnodesFine];
+ fbnwF = &DF[DIR_MPM * numberOfLBnodesFine];
real *feC, *fwC, *fnC, *fsC, *ftC, *fbC, *fneC, *fswC, *fseC, *fnwC, *fteC, *fbwC, *fbeC, *ftwC, *ftnC, *fbsC, *fbnC, *ftsC, *fzeroC,
*ftneC, *ftswC, *ftseC, *ftnwC, *fbneC, *fbswC, *fbseC, *fbnwC;
if (isEvenTimestep==true)
{
- feC = &DC[DIR_P00 *size_MatC];
- fwC = &DC[DIR_M00 *size_MatC];
- fnC = &DC[DIR_0P0 *size_MatC];
- fsC = &DC[DIR_0M0 *size_MatC];
- ftC = &DC[DIR_00P *size_MatC];
- fbC = &DC[DIR_00M *size_MatC];
- fneC = &DC[DIR_PP0 *size_MatC];
- fswC = &DC[DIR_MM0 *size_MatC];
- fseC = &DC[DIR_PM0 *size_MatC];
- fnwC = &DC[DIR_MP0 *size_MatC];
- fteC = &DC[DIR_P0P *size_MatC];
- fbwC = &DC[DIR_M0M *size_MatC];
- fbeC = &DC[DIR_P0M *size_MatC];
- ftwC = &DC[DIR_M0P *size_MatC];
- ftnC = &DC[DIR_0PP *size_MatC];
- fbsC = &DC[DIR_0MM *size_MatC];
- fbnC = &DC[DIR_0PM *size_MatC];
- ftsC = &DC[DIR_0MP *size_MatC];
- fzeroC = &DC[DIR_000*size_MatC];
- ftneC = &DC[DIR_PPP *size_MatC];
- ftswC = &DC[DIR_MMP *size_MatC];
- ftseC = &DC[DIR_PMP *size_MatC];
- ftnwC = &DC[DIR_MPP *size_MatC];
- fbneC = &DC[DIR_PPM *size_MatC];
- fbswC = &DC[DIR_MMM *size_MatC];
- fbseC = &DC[DIR_PMM *size_MatC];
- fbnwC = &DC[DIR_MPM *size_MatC];
+ feC = &DC[DIR_P00 * numberOfLBnodesCoarse];
+ fwC = &DC[DIR_M00 * numberOfLBnodesCoarse];
+ fnC = &DC[DIR_0P0 * numberOfLBnodesCoarse];
+ fsC = &DC[DIR_0M0 * numberOfLBnodesCoarse];
+ ftC = &DC[DIR_00P * numberOfLBnodesCoarse];
+ fbC = &DC[DIR_00M * numberOfLBnodesCoarse];
+ fneC = &DC[DIR_PP0 * numberOfLBnodesCoarse];
+ fswC = &DC[DIR_MM0 * numberOfLBnodesCoarse];
+ fseC = &DC[DIR_PM0 * numberOfLBnodesCoarse];
+ fnwC = &DC[DIR_MP0 * numberOfLBnodesCoarse];
+ fteC = &DC[DIR_P0P * numberOfLBnodesCoarse];
+ fbwC = &DC[DIR_M0M * numberOfLBnodesCoarse];
+ fbeC = &DC[DIR_P0M * numberOfLBnodesCoarse];
+ ftwC = &DC[DIR_M0P * numberOfLBnodesCoarse];
+ ftnC = &DC[DIR_0PP * numberOfLBnodesCoarse];
+ fbsC = &DC[DIR_0MM * numberOfLBnodesCoarse];
+ fbnC = &DC[DIR_0PM * numberOfLBnodesCoarse];
+ ftsC = &DC[DIR_0MP * numberOfLBnodesCoarse];
+ fzeroC = &DC[DIR_000 * numberOfLBnodesCoarse];
+ ftneC = &DC[DIR_PPP * numberOfLBnodesCoarse];
+ ftswC = &DC[DIR_MMP * numberOfLBnodesCoarse];
+ ftseC = &DC[DIR_PMP * numberOfLBnodesCoarse];
+ ftnwC = &DC[DIR_MPP * numberOfLBnodesCoarse];
+ fbneC = &DC[DIR_PPM * numberOfLBnodesCoarse];
+ fbswC = &DC[DIR_MMM * numberOfLBnodesCoarse];
+ fbseC = &DC[DIR_PMM * numberOfLBnodesCoarse];
+ fbnwC = &DC[DIR_MPM * numberOfLBnodesCoarse];
}
else
{
- fwC = &DC[DIR_P00 *size_MatC];
- feC = &DC[DIR_M00 *size_MatC];
- fsC = &DC[DIR_0P0 *size_MatC];
- fnC = &DC[DIR_0M0 *size_MatC];
- fbC = &DC[DIR_00P *size_MatC];
- ftC = &DC[DIR_00M *size_MatC];
- fswC = &DC[DIR_PP0 *size_MatC];
- fneC = &DC[DIR_MM0 *size_MatC];
- fnwC = &DC[DIR_PM0 *size_MatC];
- fseC = &DC[DIR_MP0 *size_MatC];
- fbwC = &DC[DIR_P0P *size_MatC];
- fteC = &DC[DIR_M0M *size_MatC];
- ftwC = &DC[DIR_P0M *size_MatC];
- fbeC = &DC[DIR_M0P *size_MatC];
- fbsC = &DC[DIR_0PP *size_MatC];
- ftnC = &DC[DIR_0MM *size_MatC];
- ftsC = &DC[DIR_0PM *size_MatC];
- fbnC = &DC[DIR_0MP *size_MatC];
- fzeroC = &DC[DIR_000*size_MatC];
- fbswC = &DC[DIR_PPP *size_MatC];
- fbneC = &DC[DIR_MMP *size_MatC];
- fbnwC = &DC[DIR_PMP *size_MatC];
- fbseC = &DC[DIR_MPP *size_MatC];
- ftswC = &DC[DIR_PPM *size_MatC];
- ftneC = &DC[DIR_MMM *size_MatC];
- ftnwC = &DC[DIR_PMM *size_MatC];
- ftseC = &DC[DIR_MPM *size_MatC];
+ fwC = &DC[DIR_P00 * numberOfLBnodesCoarse];
+ feC = &DC[DIR_M00 * numberOfLBnodesCoarse];
+ fsC = &DC[DIR_0P0 * numberOfLBnodesCoarse];
+ fnC = &DC[DIR_0M0 * numberOfLBnodesCoarse];
+ fbC = &DC[DIR_00P * numberOfLBnodesCoarse];
+ ftC = &DC[DIR_00M * numberOfLBnodesCoarse];
+ fswC = &DC[DIR_PP0 * numberOfLBnodesCoarse];
+ fneC = &DC[DIR_MM0 * numberOfLBnodesCoarse];
+ fnwC = &DC[DIR_PM0 * numberOfLBnodesCoarse];
+ fseC = &DC[DIR_MP0 * numberOfLBnodesCoarse];
+ fbwC = &DC[DIR_P0P * numberOfLBnodesCoarse];
+ fteC = &DC[DIR_M0M * numberOfLBnodesCoarse];
+ ftwC = &DC[DIR_P0M * numberOfLBnodesCoarse];
+ fbeC = &DC[DIR_M0P * numberOfLBnodesCoarse];
+ fbsC = &DC[DIR_0PP * numberOfLBnodesCoarse];
+ ftnC = &DC[DIR_0MM * numberOfLBnodesCoarse];
+ ftsC = &DC[DIR_0PM * numberOfLBnodesCoarse];
+ fbnC = &DC[DIR_0MP * numberOfLBnodesCoarse];
+ fzeroC = &DC[DIR_000 * numberOfLBnodesCoarse];
+ fbswC = &DC[DIR_PPP * numberOfLBnodesCoarse];
+ fbneC = &DC[DIR_MMP * numberOfLBnodesCoarse];
+ fbnwC = &DC[DIR_PMP * numberOfLBnodesCoarse];
+ fbseC = &DC[DIR_MPP * numberOfLBnodesCoarse];
+ ftswC = &DC[DIR_PPM * numberOfLBnodesCoarse];
+ ftneC = &DC[DIR_MMM * numberOfLBnodesCoarse];
+ ftnwC = &DC[DIR_PMM * numberOfLBnodesCoarse];
+ ftseC = &DC[DIR_MPM * numberOfLBnodesCoarse];
}
////////////////////////////////////////////////////////////////////////////////
@@ -26369,8 +26369,8 @@ __global__ void scaleCF_Fix_comp_27( real* DC,
unsigned int* neighborFX,
unsigned int* neighborFY,
unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
+ unsigned long long numberOfLBnodesCoarse,
+ unsigned long long numberOfLBnodesFine,
bool isEvenTimestep,
unsigned int* posCSWB,
unsigned int* posFSWB,
@@ -26387,96 +26387,96 @@ __global__ void scaleCF_Fix_comp_27( real* DC,
real *feF, *fwF, *fnF, *fsF, *ftF, *fbF, *fneF, *fswF, *fseF, *fnwF, *fteF, *fbwF, *fbeF, *ftwF, *ftnF, *fbsF, *fbnF, *ftsF, *fzeroF,
*ftneF, *ftswF, *ftseF, *ftnwF, *fbneF, *fbswF, *fbseF, *fbnwF;
- feF = &DF[DIR_P00 *size_MatF];
- fwF = &DF[DIR_M00 *size_MatF];
- fnF = &DF[DIR_0P0 *size_MatF];
- fsF = &DF[DIR_0M0 *size_MatF];
- ftF = &DF[DIR_00P *size_MatF];
- fbF = &DF[DIR_00M *size_MatF];
- fneF = &DF[DIR_PP0 *size_MatF];
- fswF = &DF[DIR_MM0 *size_MatF];
- fseF = &DF[DIR_PM0 *size_MatF];
- fnwF = &DF[DIR_MP0 *size_MatF];
- fteF = &DF[DIR_P0P *size_MatF];
- fbwF = &DF[DIR_M0M *size_MatF];
- fbeF = &DF[DIR_P0M *size_MatF];
- ftwF = &DF[DIR_M0P *size_MatF];
- ftnF = &DF[DIR_0PP *size_MatF];
- fbsF = &DF[DIR_0MM *size_MatF];
- fbnF = &DF[DIR_0PM *size_MatF];
- ftsF = &DF[DIR_0MP *size_MatF];
- fzeroF = &DF[DIR_000*size_MatF];
- ftneF = &DF[DIR_PPP *size_MatF];
- ftswF = &DF[DIR_MMP *size_MatF];
- ftseF = &DF[DIR_PMP *size_MatF];
- ftnwF = &DF[DIR_MPP *size_MatF];
- fbneF = &DF[DIR_PPM *size_MatF];
- fbswF = &DF[DIR_MMM *size_MatF];
- fbseF = &DF[DIR_PMM *size_MatF];
- fbnwF = &DF[DIR_MPM *size_MatF];
+ feF = &DF[DIR_P00 * numberOfLBnodesFine];
+ fwF = &DF[DIR_M00 * numberOfLBnodesFine];
+ fnF = &DF[DIR_0P0 * numberOfLBnodesFine];
+ fsF = &DF[DIR_0M0 * numberOfLBnodesFine];
+ ftF = &DF[DIR_00P * numberOfLBnodesFine];
+ fbF = &DF[DIR_00M * numberOfLBnodesFine];
+ fneF = &DF[DIR_PP0 * numberOfLBnodesFine];
+ fswF = &DF[DIR_MM0 * numberOfLBnodesFine];
+ fseF = &DF[DIR_PM0 * numberOfLBnodesFine];
+ fnwF = &DF[DIR_MP0 * numberOfLBnodesFine];
+ fteF = &DF[DIR_P0P * numberOfLBnodesFine];
+ fbwF = &DF[DIR_M0M * numberOfLBnodesFine];
+ fbeF = &DF[DIR_P0M * numberOfLBnodesFine];
+ ftwF = &DF[DIR_M0P * numberOfLBnodesFine];
+ ftnF = &DF[DIR_0PP * numberOfLBnodesFine];
+ fbsF = &DF[DIR_0MM * numberOfLBnodesFine];
+ fbnF = &DF[DIR_0PM * numberOfLBnodesFine];
+ ftsF = &DF[DIR_0MP * numberOfLBnodesFine];
+ fzeroF = &DF[DIR_000 * numberOfLBnodesFine];
+ ftneF = &DF[DIR_PPP * numberOfLBnodesFine];
+ ftswF = &DF[DIR_MMP * numberOfLBnodesFine];
+ ftseF = &DF[DIR_PMP * numberOfLBnodesFine];
+ ftnwF = &DF[DIR_MPP * numberOfLBnodesFine];
+ fbneF = &DF[DIR_PPM * numberOfLBnodesFine];
+ fbswF = &DF[DIR_MMM * numberOfLBnodesFine];
+ fbseF = &DF[DIR_PMM * numberOfLBnodesFine];
+ fbnwF = &DF[DIR_MPM * numberOfLBnodesFine];
real *feC, *fwC, *fnC, *fsC, *ftC, *fbC, *fneC, *fswC, *fseC, *fnwC, *fteC, *fbwC, *fbeC, *ftwC, *ftnC, *fbsC, *fbnC, *ftsC, *fzeroC,
*ftneC, *ftswC, *ftseC, *ftnwC, *fbneC, *fbswC, *fbseC, *fbnwC;
if (isEvenTimestep==true)
{
- feC = &DC[DIR_P00 *size_MatC];
- fwC = &DC[DIR_M00 *size_MatC];
- fnC = &DC[DIR_0P0 *size_MatC];
- fsC = &DC[DIR_0M0 *size_MatC];
- ftC = &DC[DIR_00P *size_MatC];
- fbC = &DC[DIR_00M *size_MatC];
- fneC = &DC[DIR_PP0 *size_MatC];
- fswC = &DC[DIR_MM0 *size_MatC];
- fseC = &DC[DIR_PM0 *size_MatC];
- fnwC = &DC[DIR_MP0 *size_MatC];
- fteC = &DC[DIR_P0P *size_MatC];
- fbwC = &DC[DIR_M0M *size_MatC];
- fbeC = &DC[DIR_P0M *size_MatC];
- ftwC = &DC[DIR_M0P *size_MatC];
- ftnC = &DC[DIR_0PP *size_MatC];
- fbsC = &DC[DIR_0MM *size_MatC];
- fbnC = &DC[DIR_0PM *size_MatC];
- ftsC = &DC[DIR_0MP *size_MatC];
- fzeroC = &DC[DIR_000*size_MatC];
- ftneC = &DC[DIR_PPP *size_MatC];
- ftswC = &DC[DIR_MMP *size_MatC];
- ftseC = &DC[DIR_PMP *size_MatC];
- ftnwC = &DC[DIR_MPP *size_MatC];
- fbneC = &DC[DIR_PPM *size_MatC];
- fbswC = &DC[DIR_MMM *size_MatC];
- fbseC = &DC[DIR_PMM *size_MatC];
- fbnwC = &DC[DIR_MPM *size_MatC];
+ feC = &DC[DIR_P00 * numberOfLBnodesCoarse];
+ fwC = &DC[DIR_M00 * numberOfLBnodesCoarse];
+ fnC = &DC[DIR_0P0 * numberOfLBnodesCoarse];
+ fsC = &DC[DIR_0M0 * numberOfLBnodesCoarse];
+ ftC = &DC[DIR_00P * numberOfLBnodesCoarse];
+ fbC = &DC[DIR_00M * numberOfLBnodesCoarse];
+ fneC = &DC[DIR_PP0 * numberOfLBnodesCoarse];
+ fswC = &DC[DIR_MM0 * numberOfLBnodesCoarse];
+ fseC = &DC[DIR_PM0 * numberOfLBnodesCoarse];
+ fnwC = &DC[DIR_MP0 * numberOfLBnodesCoarse];
+ fteC = &DC[DIR_P0P * numberOfLBnodesCoarse];
+ fbwC = &DC[DIR_M0M * numberOfLBnodesCoarse];
+ fbeC = &DC[DIR_P0M * numberOfLBnodesCoarse];
+ ftwC = &DC[DIR_M0P * numberOfLBnodesCoarse];
+ ftnC = &DC[DIR_0PP * numberOfLBnodesCoarse];
+ fbsC = &DC[DIR_0MM * numberOfLBnodesCoarse];
+ fbnC = &DC[DIR_0PM * numberOfLBnodesCoarse];
+ ftsC = &DC[DIR_0MP * numberOfLBnodesCoarse];
+ fzeroC = &DC[DIR_000 * numberOfLBnodesCoarse];
+ ftneC = &DC[DIR_PPP * numberOfLBnodesCoarse];
+ ftswC = &DC[DIR_MMP * numberOfLBnodesCoarse];
+ ftseC = &DC[DIR_PMP * numberOfLBnodesCoarse];
+ ftnwC = &DC[DIR_MPP * numberOfLBnodesCoarse];
+ fbneC = &DC[DIR_PPM * numberOfLBnodesCoarse];
+ fbswC = &DC[DIR_MMM * numberOfLBnodesCoarse];
+ fbseC = &DC[DIR_PMM * numberOfLBnodesCoarse];
+ fbnwC = &DC[DIR_MPM * numberOfLBnodesCoarse];
}
else
{
- fwC = &DC[DIR_P00 *size_MatC];
- feC = &DC[DIR_M00 *size_MatC];
- fsC = &DC[DIR_0P0 *size_MatC];
- fnC = &DC[DIR_0M0 *size_MatC];
- fbC = &DC[DIR_00P *size_MatC];
- ftC = &DC[DIR_00M *size_MatC];
- fswC = &DC[DIR_PP0 *size_MatC];
- fneC = &DC[DIR_MM0 *size_MatC];
- fnwC = &DC[DIR_PM0 *size_MatC];
- fseC = &DC[DIR_MP0 *size_MatC];
- fbwC = &DC[DIR_P0P *size_MatC];
- fteC = &DC[DIR_M0M *size_MatC];
- ftwC = &DC[DIR_P0M *size_MatC];
- fbeC = &DC[DIR_M0P *size_MatC];
- fbsC = &DC[DIR_0PP *size_MatC];
- ftnC = &DC[DIR_0MM *size_MatC];
- ftsC = &DC[DIR_0PM *size_MatC];
- fbnC = &DC[DIR_0MP *size_MatC];
- fzeroC = &DC[DIR_000*size_MatC];
- fbswC = &DC[DIR_PPP *size_MatC];
- fbneC = &DC[DIR_MMP *size_MatC];
- fbnwC = &DC[DIR_PMP *size_MatC];
- fbseC = &DC[DIR_MPP *size_MatC];
- ftswC = &DC[DIR_PPM *size_MatC];
- ftneC = &DC[DIR_MMM *size_MatC];
- ftnwC = &DC[DIR_PMM *size_MatC];
- ftseC = &DC[DIR_MPM *size_MatC];
+ fwC = &DC[DIR_P00 * numberOfLBnodesCoarse];
+ feC = &DC[DIR_M00 * numberOfLBnodesCoarse];
+ fsC = &DC[DIR_0P0 * numberOfLBnodesCoarse];
+ fnC = &DC[DIR_0M0 * numberOfLBnodesCoarse];
+ fbC = &DC[DIR_00P * numberOfLBnodesCoarse];
+ ftC = &DC[DIR_00M * numberOfLBnodesCoarse];
+ fswC = &DC[DIR_PP0 * numberOfLBnodesCoarse];
+ fneC = &DC[DIR_MM0 * numberOfLBnodesCoarse];
+ fnwC = &DC[DIR_PM0 * numberOfLBnodesCoarse];
+ fseC = &DC[DIR_MP0 * numberOfLBnodesCoarse];
+ fbwC = &DC[DIR_P0P * numberOfLBnodesCoarse];
+ fteC = &DC[DIR_M0M * numberOfLBnodesCoarse];
+ ftwC = &DC[DIR_P0M * numberOfLBnodesCoarse];
+ fbeC = &DC[DIR_M0P * numberOfLBnodesCoarse];
+ fbsC = &DC[DIR_0PP * numberOfLBnodesCoarse];
+ ftnC = &DC[DIR_0MM * numberOfLBnodesCoarse];
+ ftsC = &DC[DIR_0PM * numberOfLBnodesCoarse];
+ fbnC = &DC[DIR_0MP * numberOfLBnodesCoarse];
+ fzeroC = &DC[DIR_000 * numberOfLBnodesCoarse];
+ fbswC = &DC[DIR_PPP * numberOfLBnodesCoarse];
+ fbneC = &DC[DIR_MMP * numberOfLBnodesCoarse];
+ fbnwC = &DC[DIR_PMP * numberOfLBnodesCoarse];
+ fbseC = &DC[DIR_MPP * numberOfLBnodesCoarse];
+ ftswC = &DC[DIR_PPM * numberOfLBnodesCoarse];
+ ftneC = &DC[DIR_MMM * numberOfLBnodesCoarse];
+ ftnwC = &DC[DIR_PMM * numberOfLBnodesCoarse];
+ ftseC = &DC[DIR_MPM * numberOfLBnodesCoarse];
}
////////////////////////////////////////////////////////////////////////////////
@@ -31136,8 +31136,8 @@ __global__ void scaleCF_NSPress_27( real* DC,
unsigned int* neighborFX,
unsigned int* neighborFY,
unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
+ unsigned long long numberOfLBnodesCoarse,
+ unsigned long long numberOfLBnodesFine,
bool isEvenTimestep,
unsigned int* posCSWB,
unsigned int* posFSWB,
@@ -31154,96 +31154,96 @@ __global__ void scaleCF_NSPress_27( real* DC,
real *feF, *fwF, *fnF, *fsF, *ftF, *fbF, *fneF, *fswF, *fseF, *fnwF, *fteF, *fbwF, *fbeF, *ftwF, *ftnF, *fbsF, *fbnF, *ftsF, *fzeroF,
*ftneF, *ftswF, *ftseF, *ftnwF, *fbneF, *fbswF, *fbseF, *fbnwF;
- feF = &DF[DIR_P00 *size_MatF];
- fwF = &DF[DIR_M00 *size_MatF];
- fnF = &DF[DIR_0P0 *size_MatF];
- fsF = &DF[DIR_0M0 *size_MatF];
- ftF = &DF[DIR_00P *size_MatF];
- fbF = &DF[DIR_00M *size_MatF];
- fneF = &DF[DIR_PP0 *size_MatF];
- fswF = &DF[DIR_MM0 *size_MatF];
- fseF = &DF[DIR_PM0 *size_MatF];
- fnwF = &DF[DIR_MP0 *size_MatF];
- fteF = &DF[DIR_P0P *size_MatF];
- fbwF = &DF[DIR_M0M *size_MatF];
- fbeF = &DF[DIR_P0M *size_MatF];
- ftwF = &DF[DIR_M0P *size_MatF];
- ftnF = &DF[DIR_0PP *size_MatF];
- fbsF = &DF[DIR_0MM *size_MatF];
- fbnF = &DF[DIR_0PM *size_MatF];
- ftsF = &DF[DIR_0MP *size_MatF];
- fzeroF = &DF[DIR_000*size_MatF];
- ftneF = &DF[DIR_PPP *size_MatF];
- ftswF = &DF[DIR_MMP *size_MatF];
- ftseF = &DF[DIR_PMP *size_MatF];
- ftnwF = &DF[DIR_MPP *size_MatF];
- fbneF = &DF[DIR_PPM *size_MatF];
- fbswF = &DF[DIR_MMM *size_MatF];
- fbseF = &DF[DIR_PMM *size_MatF];
- fbnwF = &DF[DIR_MPM *size_MatF];
+ feF = &DF[DIR_P00 * numberOfLBnodesFine];
+ fwF = &DF[DIR_M00 * numberOfLBnodesFine];
+ fnF = &DF[DIR_0P0 * numberOfLBnodesFine];
+ fsF = &DF[DIR_0M0 * numberOfLBnodesFine];
+ ftF = &DF[DIR_00P * numberOfLBnodesFine];
+ fbF = &DF[DIR_00M * numberOfLBnodesFine];
+ fneF = &DF[DIR_PP0 * numberOfLBnodesFine];
+ fswF = &DF[DIR_MM0 * numberOfLBnodesFine];
+ fseF = &DF[DIR_PM0 * numberOfLBnodesFine];
+ fnwF = &DF[DIR_MP0 * numberOfLBnodesFine];
+ fteF = &DF[DIR_P0P * numberOfLBnodesFine];
+ fbwF = &DF[DIR_M0M * numberOfLBnodesFine];
+ fbeF = &DF[DIR_P0M * numberOfLBnodesFine];
+ ftwF = &DF[DIR_M0P * numberOfLBnodesFine];
+ ftnF = &DF[DIR_0PP * numberOfLBnodesFine];
+ fbsF = &DF[DIR_0MM * numberOfLBnodesFine];
+ fbnF = &DF[DIR_0PM * numberOfLBnodesFine];
+ ftsF = &DF[DIR_0MP * numberOfLBnodesFine];
+ fzeroF = &DF[DIR_000 * numberOfLBnodesFine];
+ ftneF = &DF[DIR_PPP * numberOfLBnodesFine];
+ ftswF = &DF[DIR_MMP * numberOfLBnodesFine];
+ ftseF = &DF[DIR_PMP * numberOfLBnodesFine];
+ ftnwF = &DF[DIR_MPP * numberOfLBnodesFine];
+ fbneF = &DF[DIR_PPM * numberOfLBnodesFine];
+ fbswF = &DF[DIR_MMM * numberOfLBnodesFine];
+ fbseF = &DF[DIR_PMM * numberOfLBnodesFine];
+ fbnwF = &DF[DIR_MPM * numberOfLBnodesFine];
real *feC, *fwC, *fnC, *fsC, *ftC, *fbC, *fneC, *fswC, *fseC, *fnwC, *fteC, *fbwC, *fbeC, *ftwC, *ftnC, *fbsC, *fbnC, *ftsC, *fzeroC,
*ftneC, *ftswC, *ftseC, *ftnwC, *fbneC, *fbswC, *fbseC, *fbnwC;
if (isEvenTimestep==true)
{
- feC = &DC[DIR_P00 *size_MatC];
- fwC = &DC[DIR_M00 *size_MatC];
- fnC = &DC[DIR_0P0 *size_MatC];
- fsC = &DC[DIR_0M0 *size_MatC];
- ftC = &DC[DIR_00P *size_MatC];
- fbC = &DC[DIR_00M *size_MatC];
- fneC = &DC[DIR_PP0 *size_MatC];
- fswC = &DC[DIR_MM0 *size_MatC];
- fseC = &DC[DIR_PM0 *size_MatC];
- fnwC = &DC[DIR_MP0 *size_MatC];
- fteC = &DC[DIR_P0P *size_MatC];
- fbwC = &DC[DIR_M0M *size_MatC];
- fbeC = &DC[DIR_P0M *size_MatC];
- ftwC = &DC[DIR_M0P *size_MatC];
- ftnC = &DC[DIR_0PP *size_MatC];
- fbsC = &DC[DIR_0MM *size_MatC];
- fbnC = &DC[DIR_0PM *size_MatC];
- ftsC = &DC[DIR_0MP *size_MatC];
- fzeroC = &DC[DIR_000*size_MatC];
- ftneC = &DC[DIR_PPP *size_MatC];
- ftswC = &DC[DIR_MMP *size_MatC];
- ftseC = &DC[DIR_PMP *size_MatC];
- ftnwC = &DC[DIR_MPP *size_MatC];
- fbneC = &DC[DIR_PPM *size_MatC];
- fbswC = &DC[DIR_MMM *size_MatC];
- fbseC = &DC[DIR_PMM *size_MatC];
- fbnwC = &DC[DIR_MPM *size_MatC];
+ feC = &DC[DIR_P00 * numberOfLBnodesCoarse];
+ fwC = &DC[DIR_M00 * numberOfLBnodesCoarse];
+ fnC = &DC[DIR_0P0 * numberOfLBnodesCoarse];
+ fsC = &DC[DIR_0M0 * numberOfLBnodesCoarse];
+ ftC = &DC[DIR_00P * numberOfLBnodesCoarse];
+ fbC = &DC[DIR_00M * numberOfLBnodesCoarse];
+ fneC = &DC[DIR_PP0 * numberOfLBnodesCoarse];
+ fswC = &DC[DIR_MM0 * numberOfLBnodesCoarse];
+ fseC = &DC[DIR_PM0 * numberOfLBnodesCoarse];
+ fnwC = &DC[DIR_MP0 * numberOfLBnodesCoarse];
+ fteC = &DC[DIR_P0P * numberOfLBnodesCoarse];
+ fbwC = &DC[DIR_M0M * numberOfLBnodesCoarse];
+ fbeC = &DC[DIR_P0M * numberOfLBnodesCoarse];
+ ftwC = &DC[DIR_M0P * numberOfLBnodesCoarse];
+ ftnC = &DC[DIR_0PP * numberOfLBnodesCoarse];
+ fbsC = &DC[DIR_0MM * numberOfLBnodesCoarse];
+ fbnC = &DC[DIR_0PM * numberOfLBnodesCoarse];
+ ftsC = &DC[DIR_0MP * numberOfLBnodesCoarse];
+ fzeroC = &DC[DIR_000 * numberOfLBnodesCoarse];
+ ftneC = &DC[DIR_PPP * numberOfLBnodesCoarse];
+ ftswC = &DC[DIR_MMP * numberOfLBnodesCoarse];
+ ftseC = &DC[DIR_PMP * numberOfLBnodesCoarse];
+ ftnwC = &DC[DIR_MPP * numberOfLBnodesCoarse];
+ fbneC = &DC[DIR_PPM * numberOfLBnodesCoarse];
+ fbswC = &DC[DIR_MMM * numberOfLBnodesCoarse];
+ fbseC = &DC[DIR_PMM * numberOfLBnodesCoarse];
+ fbnwC = &DC[DIR_MPM * numberOfLBnodesCoarse];
}
else
{
- fwC = &DC[DIR_P00 *size_MatC];
- feC = &DC[DIR_M00 *size_MatC];
- fsC = &DC[DIR_0P0 *size_MatC];
- fnC = &DC[DIR_0M0 *size_MatC];
- fbC = &DC[DIR_00P *size_MatC];
- ftC = &DC[DIR_00M *size_MatC];
- fswC = &DC[DIR_PP0 *size_MatC];
- fneC = &DC[DIR_MM0 *size_MatC];
- fnwC = &DC[DIR_PM0 *size_MatC];
- fseC = &DC[DIR_MP0 *size_MatC];
- fbwC = &DC[DIR_P0P *size_MatC];
- fteC = &DC[DIR_M0M *size_MatC];
- ftwC = &DC[DIR_P0M *size_MatC];
- fbeC = &DC[DIR_M0P *size_MatC];
- fbsC = &DC[DIR_0PP *size_MatC];
- ftnC = &DC[DIR_0MM *size_MatC];
- ftsC = &DC[DIR_0PM *size_MatC];
- fbnC = &DC[DIR_0MP *size_MatC];
- fzeroC = &DC[DIR_000*size_MatC];
- fbswC = &DC[DIR_PPP *size_MatC];
- fbneC = &DC[DIR_MMP *size_MatC];
- fbnwC = &DC[DIR_PMP *size_MatC];
- fbseC = &DC[DIR_MPP *size_MatC];
- ftswC = &DC[DIR_PPM *size_MatC];
- ftneC = &DC[DIR_MMM *size_MatC];
- ftnwC = &DC[DIR_PMM *size_MatC];
- ftseC = &DC[DIR_MPM *size_MatC];
+ fwC = &DC[DIR_P00 * numberOfLBnodesCoarse];
+ feC = &DC[DIR_M00 * numberOfLBnodesCoarse];
+ fsC = &DC[DIR_0P0 * numberOfLBnodesCoarse];
+ fnC = &DC[DIR_0M0 * numberOfLBnodesCoarse];
+ fbC = &DC[DIR_00P * numberOfLBnodesCoarse];
+ ftC = &DC[DIR_00M * numberOfLBnodesCoarse];
+ fswC = &DC[DIR_PP0 * numberOfLBnodesCoarse];
+ fneC = &DC[DIR_MM0 * numberOfLBnodesCoarse];
+ fnwC = &DC[DIR_PM0 * numberOfLBnodesCoarse];
+ fseC = &DC[DIR_MP0 * numberOfLBnodesCoarse];
+ fbwC = &DC[DIR_P0P * numberOfLBnodesCoarse];
+ fteC = &DC[DIR_M0M * numberOfLBnodesCoarse];
+ ftwC = &DC[DIR_P0M * numberOfLBnodesCoarse];
+ fbeC = &DC[DIR_M0P * numberOfLBnodesCoarse];
+ fbsC = &DC[DIR_0PP * numberOfLBnodesCoarse];
+ ftnC = &DC[DIR_0MM * numberOfLBnodesCoarse];
+ ftsC = &DC[DIR_0PM * numberOfLBnodesCoarse];
+ fbnC = &DC[DIR_0MP * numberOfLBnodesCoarse];
+ fzeroC = &DC[DIR_000 * numberOfLBnodesCoarse];
+ fbswC = &DC[DIR_PPP * numberOfLBnodesCoarse];
+ fbneC = &DC[DIR_MMP * numberOfLBnodesCoarse];
+ fbnwC = &DC[DIR_PMP * numberOfLBnodesCoarse];
+ fbseC = &DC[DIR_MPP * numberOfLBnodesCoarse];
+ ftswC = &DC[DIR_PPM * numberOfLBnodesCoarse];
+ ftneC = &DC[DIR_MMM * numberOfLBnodesCoarse];
+ ftnwC = &DC[DIR_PMM * numberOfLBnodesCoarse];
+ ftseC = &DC[DIR_MPM * numberOfLBnodesCoarse];
}
////////////////////////////////////////////////////////////////////////////////
@@ -35080,8 +35080,8 @@ __global__ void scaleCF_Fix_27( real* DC,
unsigned int* neighborFX,
unsigned int* neighborFY,
unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
+ unsigned long long numberOfLBnodesCoarse,
+ unsigned long long numberOfLBnodesFine,
bool isEvenTimestep,
unsigned int* posCSWB,
unsigned int* posFSWB,
@@ -35098,96 +35098,96 @@ __global__ void scaleCF_Fix_27( real* DC,
real *feF, *fwF, *fnF, *fsF, *ftF, *fbF, *fneF, *fswF, *fseF, *fnwF, *fteF, *fbwF, *fbeF, *ftwF, *ftnF, *fbsF, *fbnF, *ftsF, *fzeroF,
*ftneF, *ftswF, *ftseF, *ftnwF, *fbneF, *fbswF, *fbseF, *fbnwF;
- feF = &DF[DIR_P00 *size_MatF];
- fwF = &DF[DIR_M00 *size_MatF];
- fnF = &DF[DIR_0P0 *size_MatF];
- fsF = &DF[DIR_0M0 *size_MatF];
- ftF = &DF[DIR_00P *size_MatF];
- fbF = &DF[DIR_00M *size_MatF];
- fneF = &DF[DIR_PP0 *size_MatF];
- fswF = &DF[DIR_MM0 *size_MatF];
- fseF = &DF[DIR_PM0 *size_MatF];
- fnwF = &DF[DIR_MP0 *size_MatF];
- fteF = &DF[DIR_P0P *size_MatF];
- fbwF = &DF[DIR_M0M *size_MatF];
- fbeF = &DF[DIR_P0M *size_MatF];
- ftwF = &DF[DIR_M0P *size_MatF];
- ftnF = &DF[DIR_0PP *size_MatF];
- fbsF = &DF[DIR_0MM *size_MatF];
- fbnF = &DF[DIR_0PM *size_MatF];
- ftsF = &DF[DIR_0MP *size_MatF];
- fzeroF = &DF[DIR_000*size_MatF];
- ftneF = &DF[DIR_PPP *size_MatF];
- ftswF = &DF[DIR_MMP *size_MatF];
- ftseF = &DF[DIR_PMP *size_MatF];
- ftnwF = &DF[DIR_MPP *size_MatF];
- fbneF = &DF[DIR_PPM *size_MatF];
- fbswF = &DF[DIR_MMM *size_MatF];
- fbseF = &DF[DIR_PMM *size_MatF];
- fbnwF = &DF[DIR_MPM *size_MatF];
+ feF = &DF[DIR_P00 * numberOfLBnodesFine];
+ fwF = &DF[DIR_M00 * numberOfLBnodesFine];
+ fnF = &DF[DIR_0P0 * numberOfLBnodesFine];
+ fsF = &DF[DIR_0M0 * numberOfLBnodesFine];
+ ftF = &DF[DIR_00P * numberOfLBnodesFine];
+ fbF = &DF[DIR_00M * numberOfLBnodesFine];
+ fneF = &DF[DIR_PP0 * numberOfLBnodesFine];
+ fswF = &DF[DIR_MM0 * numberOfLBnodesFine];
+ fseF = &DF[DIR_PM0 * numberOfLBnodesFine];
+ fnwF = &DF[DIR_MP0 * numberOfLBnodesFine];
+ fteF = &DF[DIR_P0P * numberOfLBnodesFine];
+ fbwF = &DF[DIR_M0M * numberOfLBnodesFine];
+ fbeF = &DF[DIR_P0M * numberOfLBnodesFine];
+ ftwF = &DF[DIR_M0P * numberOfLBnodesFine];
+ ftnF = &DF[DIR_0PP * numberOfLBnodesFine];
+ fbsF = &DF[DIR_0MM * numberOfLBnodesFine];
+ fbnF = &DF[DIR_0PM * numberOfLBnodesFine];
+ ftsF = &DF[DIR_0MP * numberOfLBnodesFine];
+ fzeroF = &DF[DIR_000 * numberOfLBnodesFine];
+ ftneF = &DF[DIR_PPP * numberOfLBnodesFine];
+ ftswF = &DF[DIR_MMP * numberOfLBnodesFine];
+ ftseF = &DF[DIR_PMP * numberOfLBnodesFine];
+ ftnwF = &DF[DIR_MPP * numberOfLBnodesFine];
+ fbneF = &DF[DIR_PPM * numberOfLBnodesFine];
+ fbswF = &DF[DIR_MMM * numberOfLBnodesFine];
+ fbseF = &DF[DIR_PMM * numberOfLBnodesFine];
+ fbnwF = &DF[DIR_MPM * numberOfLBnodesFine];
real *feC, *fwC, *fnC, *fsC, *ftC, *fbC, *fneC, *fswC, *fseC, *fnwC, *fteC, *fbwC, *fbeC, *ftwC, *ftnC, *fbsC, *fbnC, *ftsC, *fzeroC,
*ftneC, *ftswC, *ftseC, *ftnwC, *fbneC, *fbswC, *fbseC, *fbnwC;
if (isEvenTimestep==true)
{
- feC = &DC[DIR_P00 *size_MatC];
- fwC = &DC[DIR_M00 *size_MatC];
- fnC = &DC[DIR_0P0 *size_MatC];
- fsC = &DC[DIR_0M0 *size_MatC];
- ftC = &DC[DIR_00P *size_MatC];
- fbC = &DC[DIR_00M *size_MatC];
- fneC = &DC[DIR_PP0 *size_MatC];
- fswC = &DC[DIR_MM0 *size_MatC];
- fseC = &DC[DIR_PM0 *size_MatC];
- fnwC = &DC[DIR_MP0 *size_MatC];
- fteC = &DC[DIR_P0P *size_MatC];
- fbwC = &DC[DIR_M0M *size_MatC];
- fbeC = &DC[DIR_P0M *size_MatC];
- ftwC = &DC[DIR_M0P *size_MatC];
- ftnC = &DC[DIR_0PP *size_MatC];
- fbsC = &DC[DIR_0MM *size_MatC];
- fbnC = &DC[DIR_0PM *size_MatC];
- ftsC = &DC[DIR_0MP *size_MatC];
- fzeroC = &DC[DIR_000*size_MatC];
- ftneC = &DC[DIR_PPP *size_MatC];
- ftswC = &DC[DIR_MMP *size_MatC];
- ftseC = &DC[DIR_PMP *size_MatC];
- ftnwC = &DC[DIR_MPP *size_MatC];
- fbneC = &DC[DIR_PPM *size_MatC];
- fbswC = &DC[DIR_MMM *size_MatC];
- fbseC = &DC[DIR_PMM *size_MatC];
- fbnwC = &DC[DIR_MPM *size_MatC];
+ feC = &DC[DIR_P00 * numberOfLBnodesCoarse];
+ fwC = &DC[DIR_M00 * numberOfLBnodesCoarse];
+ fnC = &DC[DIR_0P0 * numberOfLBnodesCoarse];
+ fsC = &DC[DIR_0M0 * numberOfLBnodesCoarse];
+ ftC = &DC[DIR_00P * numberOfLBnodesCoarse];
+ fbC = &DC[DIR_00M * numberOfLBnodesCoarse];
+ fneC = &DC[DIR_PP0 * numberOfLBnodesCoarse];
+ fswC = &DC[DIR_MM0 * numberOfLBnodesCoarse];
+ fseC = &DC[DIR_PM0 * numberOfLBnodesCoarse];
+ fnwC = &DC[DIR_MP0 * numberOfLBnodesCoarse];
+ fteC = &DC[DIR_P0P * numberOfLBnodesCoarse];
+ fbwC = &DC[DIR_M0M * numberOfLBnodesCoarse];
+ fbeC = &DC[DIR_P0M * numberOfLBnodesCoarse];
+ ftwC = &DC[DIR_M0P * numberOfLBnodesCoarse];
+ ftnC = &DC[DIR_0PP * numberOfLBnodesCoarse];
+ fbsC = &DC[DIR_0MM * numberOfLBnodesCoarse];
+ fbnC = &DC[DIR_0PM * numberOfLBnodesCoarse];
+ ftsC = &DC[DIR_0MP * numberOfLBnodesCoarse];
+ fzeroC = &DC[DIR_000 * numberOfLBnodesCoarse];
+ ftneC = &DC[DIR_PPP * numberOfLBnodesCoarse];
+ ftswC = &DC[DIR_MMP * numberOfLBnodesCoarse];
+ ftseC = &DC[DIR_PMP * numberOfLBnodesCoarse];
+ ftnwC = &DC[DIR_MPP * numberOfLBnodesCoarse];
+ fbneC = &DC[DIR_PPM * numberOfLBnodesCoarse];
+ fbswC = &DC[DIR_MMM * numberOfLBnodesCoarse];
+ fbseC = &DC[DIR_PMM * numberOfLBnodesCoarse];
+ fbnwC = &DC[DIR_MPM * numberOfLBnodesCoarse];
}
else
{
- fwC = &DC[DIR_P00 *size_MatC];
- feC = &DC[DIR_M00 *size_MatC];
- fsC = &DC[DIR_0P0 *size_MatC];
- fnC = &DC[DIR_0M0 *size_MatC];
- fbC = &DC[DIR_00P *size_MatC];
- ftC = &DC[DIR_00M *size_MatC];
- fswC = &DC[DIR_PP0 *size_MatC];
- fneC = &DC[DIR_MM0 *size_MatC];
- fnwC = &DC[DIR_PM0 *size_MatC];
- fseC = &DC[DIR_MP0 *size_MatC];
- fbwC = &DC[DIR_P0P *size_MatC];
- fteC = &DC[DIR_M0M *size_MatC];
- ftwC = &DC[DIR_P0M *size_MatC];
- fbeC = &DC[DIR_M0P *size_MatC];
- fbsC = &DC[DIR_0PP *size_MatC];
- ftnC = &DC[DIR_0MM *size_MatC];
- ftsC = &DC[DIR_0PM *size_MatC];
- fbnC = &DC[DIR_0MP *size_MatC];
- fzeroC = &DC[DIR_000*size_MatC];
- fbswC = &DC[DIR_PPP *size_MatC];
- fbneC = &DC[DIR_MMP *size_MatC];
- fbnwC = &DC[DIR_PMP *size_MatC];
- fbseC = &DC[DIR_MPP *size_MatC];
- ftswC = &DC[DIR_PPM *size_MatC];
- ftneC = &DC[DIR_MMM *size_MatC];
- ftnwC = &DC[DIR_PMM *size_MatC];
- ftseC = &DC[DIR_MPM *size_MatC];
+ fwC = &DC[DIR_P00 * numberOfLBnodesCoarse];
+ feC = &DC[DIR_M00 * numberOfLBnodesCoarse];
+ fsC = &DC[DIR_0P0 * numberOfLBnodesCoarse];
+ fnC = &DC[DIR_0M0 * numberOfLBnodesCoarse];
+ fbC = &DC[DIR_00P * numberOfLBnodesCoarse];
+ ftC = &DC[DIR_00M * numberOfLBnodesCoarse];
+ fswC = &DC[DIR_PP0 * numberOfLBnodesCoarse];
+ fneC = &DC[DIR_MM0 * numberOfLBnodesCoarse];
+ fnwC = &DC[DIR_PM0 * numberOfLBnodesCoarse];
+ fseC = &DC[DIR_MP0 * numberOfLBnodesCoarse];
+ fbwC = &DC[DIR_P0P * numberOfLBnodesCoarse];
+ fteC = &DC[DIR_M0M * numberOfLBnodesCoarse];
+ ftwC = &DC[DIR_P0M * numberOfLBnodesCoarse];
+ fbeC = &DC[DIR_M0P * numberOfLBnodesCoarse];
+ fbsC = &DC[DIR_0PP * numberOfLBnodesCoarse];
+ ftnC = &DC[DIR_0MM * numberOfLBnodesCoarse];
+ ftsC = &DC[DIR_0PM * numberOfLBnodesCoarse];
+ fbnC = &DC[DIR_0MP * numberOfLBnodesCoarse];
+ fzeroC = &DC[DIR_000 * numberOfLBnodesCoarse];
+ fbswC = &DC[DIR_PPP * numberOfLBnodesCoarse];
+ fbneC = &DC[DIR_MMP * numberOfLBnodesCoarse];
+ fbnwC = &DC[DIR_PMP * numberOfLBnodesCoarse];
+ fbseC = &DC[DIR_MPP * numberOfLBnodesCoarse];
+ ftswC = &DC[DIR_PPM * numberOfLBnodesCoarse];
+ ftneC = &DC[DIR_MMM * numberOfLBnodesCoarse];
+ ftnwC = &DC[DIR_PMM * numberOfLBnodesCoarse];
+ ftseC = &DC[DIR_MPM * numberOfLBnodesCoarse];
}
////////////////////////////////////////////////////////////////////////////////
@@ -39338,8 +39338,8 @@ __global__ void scaleCFpress27( real* DC,
unsigned int* neighborFX,
unsigned int* neighborFY,
unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
+ unsigned long long numberOfLBnodesCoarse,
+ unsigned long long numberOfLBnodesFine,
bool isEvenTimestep,
unsigned int* posCSWB,
unsigned int* posFSWB,
@@ -39356,96 +39356,96 @@ __global__ void scaleCFpress27( real* DC,
real *feF, *fwF, *fnF, *fsF, *ftF, *fbF, *fneF, *fswF, *fseF, *fnwF, *fteF, *fbwF, *fbeF, *ftwF, *ftnF, *fbsF, *fbnF, *ftsF, *fzeroF,
*ftneF, *ftswF, *ftseF, *ftnwF, *fbneF, *fbswF, *fbseF, *fbnwF;
- feF = &DF[DIR_P00 *size_MatF];
- fwF = &DF[DIR_M00 *size_MatF];
- fnF = &DF[DIR_0P0 *size_MatF];
- fsF = &DF[DIR_0M0 *size_MatF];
- ftF = &DF[DIR_00P *size_MatF];
- fbF = &DF[DIR_00M *size_MatF];
- fneF = &DF[DIR_PP0 *size_MatF];
- fswF = &DF[DIR_MM0 *size_MatF];
- fseF = &DF[DIR_PM0 *size_MatF];
- fnwF = &DF[DIR_MP0 *size_MatF];
- fteF = &DF[DIR_P0P *size_MatF];
- fbwF = &DF[DIR_M0M *size_MatF];
- fbeF = &DF[DIR_P0M *size_MatF];
- ftwF = &DF[DIR_M0P *size_MatF];
- ftnF = &DF[DIR_0PP *size_MatF];
- fbsF = &DF[DIR_0MM *size_MatF];
- fbnF = &DF[DIR_0PM *size_MatF];
- ftsF = &DF[DIR_0MP *size_MatF];
- fzeroF = &DF[DIR_000*size_MatF];
- ftneF = &DF[DIR_PPP *size_MatF];
- ftswF = &DF[DIR_MMP *size_MatF];
- ftseF = &DF[DIR_PMP *size_MatF];
- ftnwF = &DF[DIR_MPP *size_MatF];
- fbneF = &DF[DIR_PPM *size_MatF];
- fbswF = &DF[DIR_MMM *size_MatF];
- fbseF = &DF[DIR_PMM *size_MatF];
- fbnwF = &DF[DIR_MPM *size_MatF];
+ feF = &DF[DIR_P00 * numberOfLBnodesFine];
+ fwF = &DF[DIR_M00 * numberOfLBnodesFine];
+ fnF = &DF[DIR_0P0 * numberOfLBnodesFine];
+ fsF = &DF[DIR_0M0 * numberOfLBnodesFine];
+ ftF = &DF[DIR_00P * numberOfLBnodesFine];
+ fbF = &DF[DIR_00M * numberOfLBnodesFine];
+ fneF = &DF[DIR_PP0 * numberOfLBnodesFine];
+ fswF = &DF[DIR_MM0 * numberOfLBnodesFine];
+ fseF = &DF[DIR_PM0 * numberOfLBnodesFine];
+ fnwF = &DF[DIR_MP0 * numberOfLBnodesFine];
+ fteF = &DF[DIR_P0P * numberOfLBnodesFine];
+ fbwF = &DF[DIR_M0M * numberOfLBnodesFine];
+ fbeF = &DF[DIR_P0M * numberOfLBnodesFine];
+ ftwF = &DF[DIR_M0P * numberOfLBnodesFine];
+ ftnF = &DF[DIR_0PP * numberOfLBnodesFine];
+ fbsF = &DF[DIR_0MM * numberOfLBnodesFine];
+ fbnF = &DF[DIR_0PM * numberOfLBnodesFine];
+ ftsF = &DF[DIR_0MP * numberOfLBnodesFine];
+ fzeroF = &DF[DIR_000 * numberOfLBnodesFine];
+ ftneF = &DF[DIR_PPP * numberOfLBnodesFine];
+ ftswF = &DF[DIR_MMP * numberOfLBnodesFine];
+ ftseF = &DF[DIR_PMP * numberOfLBnodesFine];
+ ftnwF = &DF[DIR_MPP * numberOfLBnodesFine];
+ fbneF = &DF[DIR_PPM * numberOfLBnodesFine];
+ fbswF = &DF[DIR_MMM * numberOfLBnodesFine];
+ fbseF = &DF[DIR_PMM * numberOfLBnodesFine];
+ fbnwF = &DF[DIR_MPM * numberOfLBnodesFine];
real *feC, *fwC, *fnC, *fsC, *ftC, *fbC, *fneC, *fswC, *fseC, *fnwC, *fteC, *fbwC, *fbeC, *ftwC, *ftnC, *fbsC, *fbnC, *ftsC, *fzeroC,
*ftneC, *ftswC, *ftseC, *ftnwC, *fbneC, *fbswC, *fbseC, *fbnwC;
if (isEvenTimestep==true)
{
- feC = &DC[DIR_P00 *size_MatC];
- fwC = &DC[DIR_M00 *size_MatC];
- fnC = &DC[DIR_0P0 *size_MatC];
- fsC = &DC[DIR_0M0 *size_MatC];
- ftC = &DC[DIR_00P *size_MatC];
- fbC = &DC[DIR_00M *size_MatC];
- fneC = &DC[DIR_PP0 *size_MatC];
- fswC = &DC[DIR_MM0 *size_MatC];
- fseC = &DC[DIR_PM0 *size_MatC];
- fnwC = &DC[DIR_MP0 *size_MatC];
- fteC = &DC[DIR_P0P *size_MatC];
- fbwC = &DC[DIR_M0M *size_MatC];
- fbeC = &DC[DIR_P0M *size_MatC];
- ftwC = &DC[DIR_M0P *size_MatC];
- ftnC = &DC[DIR_0PP *size_MatC];
- fbsC = &DC[DIR_0MM *size_MatC];
- fbnC = &DC[DIR_0PM *size_MatC];
- ftsC = &DC[DIR_0MP *size_MatC];
- fzeroC = &DC[DIR_000*size_MatC];
- ftneC = &DC[DIR_PPP *size_MatC];
- ftswC = &DC[DIR_MMP *size_MatC];
- ftseC = &DC[DIR_PMP *size_MatC];
- ftnwC = &DC[DIR_MPP *size_MatC];
- fbneC = &DC[DIR_PPM *size_MatC];
- fbswC = &DC[DIR_MMM *size_MatC];
- fbseC = &DC[DIR_PMM *size_MatC];
- fbnwC = &DC[DIR_MPM *size_MatC];
+ feC = &DC[DIR_P00 * numberOfLBnodesCoarse];
+ fwC = &DC[DIR_M00 * numberOfLBnodesCoarse];
+ fnC = &DC[DIR_0P0 * numberOfLBnodesCoarse];
+ fsC = &DC[DIR_0M0 * numberOfLBnodesCoarse];
+ ftC = &DC[DIR_00P * numberOfLBnodesCoarse];
+ fbC = &DC[DIR_00M * numberOfLBnodesCoarse];
+ fneC = &DC[DIR_PP0 * numberOfLBnodesCoarse];
+ fswC = &DC[DIR_MM0 * numberOfLBnodesCoarse];
+ fseC = &DC[DIR_PM0 * numberOfLBnodesCoarse];
+ fnwC = &DC[DIR_MP0 * numberOfLBnodesCoarse];
+ fteC = &DC[DIR_P0P * numberOfLBnodesCoarse];
+ fbwC = &DC[DIR_M0M * numberOfLBnodesCoarse];
+ fbeC = &DC[DIR_P0M * numberOfLBnodesCoarse];
+ ftwC = &DC[DIR_M0P * numberOfLBnodesCoarse];
+ ftnC = &DC[DIR_0PP * numberOfLBnodesCoarse];
+ fbsC = &DC[DIR_0MM * numberOfLBnodesCoarse];
+ fbnC = &DC[DIR_0PM * numberOfLBnodesCoarse];
+ ftsC = &DC[DIR_0MP * numberOfLBnodesCoarse];
+ fzeroC = &DC[DIR_000 * numberOfLBnodesCoarse];
+ ftneC = &DC[DIR_PPP * numberOfLBnodesCoarse];
+ ftswC = &DC[DIR_MMP * numberOfLBnodesCoarse];
+ ftseC = &DC[DIR_PMP * numberOfLBnodesCoarse];
+ ftnwC = &DC[DIR_MPP * numberOfLBnodesCoarse];
+ fbneC = &DC[DIR_PPM * numberOfLBnodesCoarse];
+ fbswC = &DC[DIR_MMM * numberOfLBnodesCoarse];
+ fbseC = &DC[DIR_PMM * numberOfLBnodesCoarse];
+ fbnwC = &DC[DIR_MPM * numberOfLBnodesCoarse];
}
else
{
- fwC = &DC[DIR_P00 *size_MatC];
- feC = &DC[DIR_M00 *size_MatC];
- fsC = &DC[DIR_0P0 *size_MatC];
- fnC = &DC[DIR_0M0 *size_MatC];
- fbC = &DC[DIR_00P *size_MatC];
- ftC = &DC[DIR_00M *size_MatC];
- fswC = &DC[DIR_PP0 *size_MatC];
- fneC = &DC[DIR_MM0 *size_MatC];
- fnwC = &DC[DIR_PM0 *size_MatC];
- fseC = &DC[DIR_MP0 *size_MatC];
- fbwC = &DC[DIR_P0P *size_MatC];
- fteC = &DC[DIR_M0M *size_MatC];
- ftwC = &DC[DIR_P0M *size_MatC];
- fbeC = &DC[DIR_M0P *size_MatC];
- fbsC = &DC[DIR_0PP *size_MatC];
- ftnC = &DC[DIR_0MM *size_MatC];
- ftsC = &DC[DIR_0PM *size_MatC];
- fbnC = &DC[DIR_0MP *size_MatC];
- fzeroC = &DC[DIR_000*size_MatC];
- fbswC = &DC[DIR_PPP *size_MatC];
- fbneC = &DC[DIR_MMP *size_MatC];
- fbnwC = &DC[DIR_PMP *size_MatC];
- fbseC = &DC[DIR_MPP *size_MatC];
- ftswC = &DC[DIR_PPM *size_MatC];
- ftneC = &DC[DIR_MMM *size_MatC];
- ftnwC = &DC[DIR_PMM *size_MatC];
- ftseC = &DC[DIR_MPM *size_MatC];
+ fwC = &DC[DIR_P00 * numberOfLBnodesCoarse];
+ feC = &DC[DIR_M00 * numberOfLBnodesCoarse];
+ fsC = &DC[DIR_0P0 * numberOfLBnodesCoarse];
+ fnC = &DC[DIR_0M0 * numberOfLBnodesCoarse];
+ fbC = &DC[DIR_00P * numberOfLBnodesCoarse];
+ ftC = &DC[DIR_00M * numberOfLBnodesCoarse];
+ fswC = &DC[DIR_PP0 * numberOfLBnodesCoarse];
+ fneC = &DC[DIR_MM0 * numberOfLBnodesCoarse];
+ fnwC = &DC[DIR_PM0 * numberOfLBnodesCoarse];
+ fseC = &DC[DIR_MP0 * numberOfLBnodesCoarse];
+ fbwC = &DC[DIR_P0P * numberOfLBnodesCoarse];
+ fteC = &DC[DIR_M0M * numberOfLBnodesCoarse];
+ ftwC = &DC[DIR_P0M * numberOfLBnodesCoarse];
+ fbeC = &DC[DIR_M0P * numberOfLBnodesCoarse];
+ fbsC = &DC[DIR_0PP * numberOfLBnodesCoarse];
+ ftnC = &DC[DIR_0MM * numberOfLBnodesCoarse];
+ ftsC = &DC[DIR_0PM * numberOfLBnodesCoarse];
+ fbnC = &DC[DIR_0MP * numberOfLBnodesCoarse];
+ fzeroC = &DC[DIR_000 * numberOfLBnodesCoarse];
+ fbswC = &DC[DIR_PPP * numberOfLBnodesCoarse];
+ fbneC = &DC[DIR_MMP * numberOfLBnodesCoarse];
+ fbnwC = &DC[DIR_PMP * numberOfLBnodesCoarse];
+ fbseC = &DC[DIR_MPP * numberOfLBnodesCoarse];
+ ftswC = &DC[DIR_PPM * numberOfLBnodesCoarse];
+ ftneC = &DC[DIR_MMM * numberOfLBnodesCoarse];
+ ftnwC = &DC[DIR_PMM * numberOfLBnodesCoarse];
+ ftseC = &DC[DIR_MPM * numberOfLBnodesCoarse];
}
////////////////////////////////////////////////////////////////////////////////
@@ -41012,8 +41012,8 @@ __global__ void scaleCFLast27( real* DC,
unsigned int* neighborFX,
unsigned int* neighborFY,
unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
+ unsigned long long numberOfLBnodesCoarse,
+ unsigned long long numberOfLBnodesFine,
bool isEvenTimestep,
unsigned int* posCSWB,
unsigned int* posFSWB,
@@ -41030,96 +41030,96 @@ __global__ void scaleCFLast27( real* DC,
real *feF, *fwF, *fnF, *fsF, *ftF, *fbF, *fneF, *fswF, *fseF, *fnwF, *fteF, *fbwF, *fbeF, *ftwF, *ftnF, *fbsF, *fbnF, *ftsF, *fzeroF,
*ftneF, *ftswF, *ftseF, *ftnwF, *fbneF, *fbswF, *fbseF, *fbnwF;
- feF = &DF[DIR_P00 *size_MatF];
- fwF = &DF[DIR_M00 *size_MatF];
- fnF = &DF[DIR_0P0 *size_MatF];
- fsF = &DF[DIR_0M0 *size_MatF];
- ftF = &DF[DIR_00P *size_MatF];
- fbF = &DF[DIR_00M *size_MatF];
- fneF = &DF[DIR_PP0 *size_MatF];
- fswF = &DF[DIR_MM0 *size_MatF];
- fseF = &DF[DIR_PM0 *size_MatF];
- fnwF = &DF[DIR_MP0 *size_MatF];
- fteF = &DF[DIR_P0P *size_MatF];
- fbwF = &DF[DIR_M0M *size_MatF];
- fbeF = &DF[DIR_P0M *size_MatF];
- ftwF = &DF[DIR_M0P *size_MatF];
- ftnF = &DF[DIR_0PP *size_MatF];
- fbsF = &DF[DIR_0MM *size_MatF];
- fbnF = &DF[DIR_0PM *size_MatF];
- ftsF = &DF[DIR_0MP *size_MatF];
- fzeroF = &DF[DIR_000*size_MatF];
- ftneF = &DF[DIR_PPP *size_MatF];
- ftswF = &DF[DIR_MMP *size_MatF];
- ftseF = &DF[DIR_PMP *size_MatF];
- ftnwF = &DF[DIR_MPP *size_MatF];
- fbneF = &DF[DIR_PPM *size_MatF];
- fbswF = &DF[DIR_MMM *size_MatF];
- fbseF = &DF[DIR_PMM *size_MatF];
- fbnwF = &DF[DIR_MPM *size_MatF];
+ feF = &DF[DIR_P00 * numberOfLBnodesFine];
+ fwF = &DF[DIR_M00 * numberOfLBnodesFine];
+ fnF = &DF[DIR_0P0 * numberOfLBnodesFine];
+ fsF = &DF[DIR_0M0 * numberOfLBnodesFine];
+ ftF = &DF[DIR_00P * numberOfLBnodesFine];
+ fbF = &DF[DIR_00M * numberOfLBnodesFine];
+ fneF = &DF[DIR_PP0 * numberOfLBnodesFine];
+ fswF = &DF[DIR_MM0 * numberOfLBnodesFine];
+ fseF = &DF[DIR_PM0 * numberOfLBnodesFine];
+ fnwF = &DF[DIR_MP0 * numberOfLBnodesFine];
+ fteF = &DF[DIR_P0P * numberOfLBnodesFine];
+ fbwF = &DF[DIR_M0M * numberOfLBnodesFine];
+ fbeF = &DF[DIR_P0M * numberOfLBnodesFine];
+ ftwF = &DF[DIR_M0P * numberOfLBnodesFine];
+ ftnF = &DF[DIR_0PP * numberOfLBnodesFine];
+ fbsF = &DF[DIR_0MM * numberOfLBnodesFine];
+ fbnF = &DF[DIR_0PM * numberOfLBnodesFine];
+ ftsF = &DF[DIR_0MP * numberOfLBnodesFine];
+ fzeroF = &DF[DIR_000 * numberOfLBnodesFine];
+ ftneF = &DF[DIR_PPP * numberOfLBnodesFine];
+ ftswF = &DF[DIR_MMP * numberOfLBnodesFine];
+ ftseF = &DF[DIR_PMP * numberOfLBnodesFine];
+ ftnwF = &DF[DIR_MPP * numberOfLBnodesFine];
+ fbneF = &DF[DIR_PPM * numberOfLBnodesFine];
+ fbswF = &DF[DIR_MMM * numberOfLBnodesFine];
+ fbseF = &DF[DIR_PMM * numberOfLBnodesFine];
+ fbnwF = &DF[DIR_MPM * numberOfLBnodesFine];
real *feC, *fwC, *fnC, *fsC, *ftC, *fbC, *fneC, *fswC, *fseC, *fnwC, *fteC, *fbwC, *fbeC, *ftwC, *ftnC, *fbsC, *fbnC, *ftsC, *fzeroC,
*ftneC, *ftswC, *ftseC, *ftnwC, *fbneC, *fbswC, *fbseC, *fbnwC;
if (isEvenTimestep==true)
{
- feC = &DC[DIR_P00 *size_MatC];
- fwC = &DC[DIR_M00 *size_MatC];
- fnC = &DC[DIR_0P0 *size_MatC];
- fsC = &DC[DIR_0M0 *size_MatC];
- ftC = &DC[DIR_00P *size_MatC];
- fbC = &DC[DIR_00M *size_MatC];
- fneC = &DC[DIR_PP0 *size_MatC];
- fswC = &DC[DIR_MM0 *size_MatC];
- fseC = &DC[DIR_PM0 *size_MatC];
- fnwC = &DC[DIR_MP0 *size_MatC];
- fteC = &DC[DIR_P0P *size_MatC];
- fbwC = &DC[DIR_M0M *size_MatC];
- fbeC = &DC[DIR_P0M *size_MatC];
- ftwC = &DC[DIR_M0P *size_MatC];
- ftnC = &DC[DIR_0PP *size_MatC];
- fbsC = &DC[DIR_0MM *size_MatC];
- fbnC = &DC[DIR_0PM *size_MatC];
- ftsC = &DC[DIR_0MP *size_MatC];
- fzeroC = &DC[DIR_000*size_MatC];
- ftneC = &DC[DIR_PPP *size_MatC];
- ftswC = &DC[DIR_MMP *size_MatC];
- ftseC = &DC[DIR_PMP *size_MatC];
- ftnwC = &DC[DIR_MPP *size_MatC];
- fbneC = &DC[DIR_PPM *size_MatC];
- fbswC = &DC[DIR_MMM *size_MatC];
- fbseC = &DC[DIR_PMM *size_MatC];
- fbnwC = &DC[DIR_MPM *size_MatC];
+ feC = &DC[DIR_P00 * numberOfLBnodesCoarse];
+ fwC = &DC[DIR_M00 * numberOfLBnodesCoarse];
+ fnC = &DC[DIR_0P0 * numberOfLBnodesCoarse];
+ fsC = &DC[DIR_0M0 * numberOfLBnodesCoarse];
+ ftC = &DC[DIR_00P * numberOfLBnodesCoarse];
+ fbC = &DC[DIR_00M * numberOfLBnodesCoarse];
+ fneC = &DC[DIR_PP0 * numberOfLBnodesCoarse];
+ fswC = &DC[DIR_MM0 * numberOfLBnodesCoarse];
+ fseC = &DC[DIR_PM0 * numberOfLBnodesCoarse];
+ fnwC = &DC[DIR_MP0 * numberOfLBnodesCoarse];
+ fteC = &DC[DIR_P0P * numberOfLBnodesCoarse];
+ fbwC = &DC[DIR_M0M * numberOfLBnodesCoarse];
+ fbeC = &DC[DIR_P0M * numberOfLBnodesCoarse];
+ ftwC = &DC[DIR_M0P * numberOfLBnodesCoarse];
+ ftnC = &DC[DIR_0PP * numberOfLBnodesCoarse];
+ fbsC = &DC[DIR_0MM * numberOfLBnodesCoarse];
+ fbnC = &DC[DIR_0PM * numberOfLBnodesCoarse];
+ ftsC = &DC[DIR_0MP * numberOfLBnodesCoarse];
+ fzeroC = &DC[DIR_000 * numberOfLBnodesCoarse];
+ ftneC = &DC[DIR_PPP * numberOfLBnodesCoarse];
+ ftswC = &DC[DIR_MMP * numberOfLBnodesCoarse];
+ ftseC = &DC[DIR_PMP * numberOfLBnodesCoarse];
+ ftnwC = &DC[DIR_MPP * numberOfLBnodesCoarse];
+ fbneC = &DC[DIR_PPM * numberOfLBnodesCoarse];
+ fbswC = &DC[DIR_MMM * numberOfLBnodesCoarse];
+ fbseC = &DC[DIR_PMM * numberOfLBnodesCoarse];
+ fbnwC = &DC[DIR_MPM * numberOfLBnodesCoarse];
}
else
{
- fwC = &DC[DIR_P00 *size_MatC];
- feC = &DC[DIR_M00 *size_MatC];
- fsC = &DC[DIR_0P0 *size_MatC];
- fnC = &DC[DIR_0M0 *size_MatC];
- fbC = &DC[DIR_00P *size_MatC];
- ftC = &DC[DIR_00M *size_MatC];
- fswC = &DC[DIR_PP0 *size_MatC];
- fneC = &DC[DIR_MM0 *size_MatC];
- fnwC = &DC[DIR_PM0 *size_MatC];
- fseC = &DC[DIR_MP0 *size_MatC];
- fbwC = &DC[DIR_P0P *size_MatC];
- fteC = &DC[DIR_M0M *size_MatC];
- ftwC = &DC[DIR_P0M *size_MatC];
- fbeC = &DC[DIR_M0P *size_MatC];
- fbsC = &DC[DIR_0PP *size_MatC];
- ftnC = &DC[DIR_0MM *size_MatC];
- ftsC = &DC[DIR_0PM *size_MatC];
- fbnC = &DC[DIR_0MP *size_MatC];
- fzeroC = &DC[DIR_000*size_MatC];
- fbswC = &DC[DIR_PPP *size_MatC];
- fbneC = &DC[DIR_MMP *size_MatC];
- fbnwC = &DC[DIR_PMP *size_MatC];
- fbseC = &DC[DIR_MPP *size_MatC];
- ftswC = &DC[DIR_PPM *size_MatC];
- ftneC = &DC[DIR_MMM *size_MatC];
- ftnwC = &DC[DIR_PMM *size_MatC];
- ftseC = &DC[DIR_MPM *size_MatC];
+ fwC = &DC[DIR_P00 * numberOfLBnodesCoarse];
+ feC = &DC[DIR_M00 * numberOfLBnodesCoarse];
+ fsC = &DC[DIR_0P0 * numberOfLBnodesCoarse];
+ fnC = &DC[DIR_0M0 * numberOfLBnodesCoarse];
+ fbC = &DC[DIR_00P * numberOfLBnodesCoarse];
+ ftC = &DC[DIR_00M * numberOfLBnodesCoarse];
+ fswC = &DC[DIR_PP0 * numberOfLBnodesCoarse];
+ fneC = &DC[DIR_MM0 * numberOfLBnodesCoarse];
+ fnwC = &DC[DIR_PM0 * numberOfLBnodesCoarse];
+ fseC = &DC[DIR_MP0 * numberOfLBnodesCoarse];
+ fbwC = &DC[DIR_P0P * numberOfLBnodesCoarse];
+ fteC = &DC[DIR_M0M * numberOfLBnodesCoarse];
+ ftwC = &DC[DIR_P0M * numberOfLBnodesCoarse];
+ fbeC = &DC[DIR_M0P * numberOfLBnodesCoarse];
+ fbsC = &DC[DIR_0PP * numberOfLBnodesCoarse];
+ ftnC = &DC[DIR_0MM * numberOfLBnodesCoarse];
+ ftsC = &DC[DIR_0PM * numberOfLBnodesCoarse];
+ fbnC = &DC[DIR_0MP * numberOfLBnodesCoarse];
+ fzeroC = &DC[DIR_000 * numberOfLBnodesCoarse];
+ fbswC = &DC[DIR_PPP * numberOfLBnodesCoarse];
+ fbneC = &DC[DIR_MMP * numberOfLBnodesCoarse];
+ fbnwC = &DC[DIR_PMP * numberOfLBnodesCoarse];
+ fbseC = &DC[DIR_MPP * numberOfLBnodesCoarse];
+ ftswC = &DC[DIR_PPM * numberOfLBnodesCoarse];
+ ftneC = &DC[DIR_MMM * numberOfLBnodesCoarse];
+ ftnwC = &DC[DIR_PMM * numberOfLBnodesCoarse];
+ ftseC = &DC[DIR_MPM * numberOfLBnodesCoarse];
}
////////////////////////////////////////////////////////////////////////////////
@@ -43249,8 +43249,8 @@ __global__ void scaleCFThSMG7( real* DC,
unsigned int* neighborFX,
unsigned int* neighborFY,
unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
+ unsigned long long numberOfLBnodesCoarse,
+ unsigned long long numberOfLBnodesFine,
bool isEvenTimestep,
unsigned int* posCSWB,
unsigned int* posFSWB,
@@ -43261,128 +43261,128 @@ __global__ void scaleCFThSMG7( real* DC,
{
real *feF, *fwF, *fnF, *fsF, *ftF, *fbF, *fneF, *fswF, *fseF, *fnwF, *fteF, *fbwF, *fbeF, *ftwF, *ftnF, *fbsF, *fbnF, *ftsF, /**fzeroF,*/ *ftneF, *ftswF, *ftseF, *ftnwF, *fbneF, *fbswF, *fbseF, *fbnwF;
- feF = &DF[DIR_P00 *size_MatF];
- fwF = &DF[DIR_M00 *size_MatF];
- fnF = &DF[DIR_0P0 *size_MatF];
- fsF = &DF[DIR_0M0 *size_MatF];
- ftF = &DF[DIR_00P *size_MatF];
- fbF = &DF[DIR_00M *size_MatF];
- fneF = &DF[DIR_PP0 *size_MatF];
- fswF = &DF[DIR_MM0 *size_MatF];
- fseF = &DF[DIR_PM0 *size_MatF];
- fnwF = &DF[DIR_MP0 *size_MatF];
- fteF = &DF[DIR_P0P *size_MatF];
- fbwF = &DF[DIR_M0M *size_MatF];
- fbeF = &DF[DIR_P0M *size_MatF];
- ftwF = &DF[DIR_M0P *size_MatF];
- ftnF = &DF[DIR_0PP *size_MatF];
- fbsF = &DF[DIR_0MM *size_MatF];
- fbnF = &DF[DIR_0PM *size_MatF];
- ftsF = &DF[DIR_0MP *size_MatF];
- //fzeroF = &DF[DIR_000*size_MatF];
- ftneF = &DF[DIR_PPP *size_MatF];
- ftswF = &DF[DIR_MMP *size_MatF];
- ftseF = &DF[DIR_PMP *size_MatF];
- ftnwF = &DF[DIR_MPP *size_MatF];
- fbneF = &DF[DIR_PPM *size_MatF];
- fbswF = &DF[DIR_MMM *size_MatF];
- fbseF = &DF[DIR_PMM *size_MatF];
- fbnwF = &DF[DIR_MPM *size_MatF];
+ feF = &DF[DIR_P00 * numberOfLBnodesFine];
+ fwF = &DF[DIR_M00 * numberOfLBnodesFine];
+ fnF = &DF[DIR_0P0 * numberOfLBnodesFine];
+ fsF = &DF[DIR_0M0 * numberOfLBnodesFine];
+ ftF = &DF[DIR_00P * numberOfLBnodesFine];
+ fbF = &DF[DIR_00M * numberOfLBnodesFine];
+ fneF = &DF[DIR_PP0 * numberOfLBnodesFine];
+ fswF = &DF[DIR_MM0 * numberOfLBnodesFine];
+ fseF = &DF[DIR_PM0 * numberOfLBnodesFine];
+ fnwF = &DF[DIR_MP0 * numberOfLBnodesFine];
+ fteF = &DF[DIR_P0P * numberOfLBnodesFine];
+ fbwF = &DF[DIR_M0M * numberOfLBnodesFine];
+ fbeF = &DF[DIR_P0M * numberOfLBnodesFine];
+ ftwF = &DF[DIR_M0P * numberOfLBnodesFine];
+ ftnF = &DF[DIR_0PP * numberOfLBnodesFine];
+ fbsF = &DF[DIR_0MM * numberOfLBnodesFine];
+ fbnF = &DF[DIR_0PM * numberOfLBnodesFine];
+ ftsF = &DF[DIR_0MP * numberOfLBnodesFine];
+ //fzeroF = &DF[DIR_000 * size_MatF];
+ ftneF = &DF[DIR_PPP * numberOfLBnodesFine];
+ ftswF = &DF[DIR_MMP * numberOfLBnodesFine];
+ ftseF = &DF[DIR_PMP * numberOfLBnodesFine];
+ ftnwF = &DF[DIR_MPP * numberOfLBnodesFine];
+ fbneF = &DF[DIR_PPM * numberOfLBnodesFine];
+ fbswF = &DF[DIR_MMM * numberOfLBnodesFine];
+ fbseF = &DF[DIR_PMM * numberOfLBnodesFine];
+ fbnwF = &DF[DIR_MPM * numberOfLBnodesFine];
real *feC, *fwC, *fnC, *fsC, *ftC, *fbC, *fneC, *fswC, *fseC, *fnwC, *fteC, *fbwC, *fbeC, *ftwC, *ftnC, *fbsC, *fbnC, *ftsC, //*fzeroC,
*ftneC, *ftswC, *ftseC, *ftnwC, *fbneC, *fbswC, *fbseC, *fbnwC;
if (isEvenTimestep==true)
{
- feC = &DC[DIR_P00 *size_MatC];
- fwC = &DC[DIR_M00 *size_MatC];
- fnC = &DC[DIR_0P0 *size_MatC];
- fsC = &DC[DIR_0M0 *size_MatC];
- ftC = &DC[DIR_00P *size_MatC];
- fbC = &DC[DIR_00M *size_MatC];
- fneC = &DC[DIR_PP0 *size_MatC];
- fswC = &DC[DIR_MM0 *size_MatC];
- fseC = &DC[DIR_PM0 *size_MatC];
- fnwC = &DC[DIR_MP0 *size_MatC];
- fteC = &DC[DIR_P0P *size_MatC];
- fbwC = &DC[DIR_M0M *size_MatC];
- fbeC = &DC[DIR_P0M *size_MatC];
- ftwC = &DC[DIR_M0P *size_MatC];
- ftnC = &DC[DIR_0PP *size_MatC];
- fbsC = &DC[DIR_0MM *size_MatC];
- fbnC = &DC[DIR_0PM *size_MatC];
- ftsC = &DC[DIR_0MP *size_MatC];
- //fzeroC = &DC[DIR_000*size_MatC];
- ftneC = &DC[DIR_PPP *size_MatC];
- ftswC = &DC[DIR_MMP *size_MatC];
- ftseC = &DC[DIR_PMP *size_MatC];
- ftnwC = &DC[DIR_MPP *size_MatC];
- fbneC = &DC[DIR_PPM *size_MatC];
- fbswC = &DC[DIR_MMM *size_MatC];
- fbseC = &DC[DIR_PMM *size_MatC];
- fbnwC = &DC[DIR_MPM *size_MatC];
+ feC = &DC[DIR_P00 * numberOfLBnodesCoarse];
+ fwC = &DC[DIR_M00 * numberOfLBnodesCoarse];
+ fnC = &DC[DIR_0P0 * numberOfLBnodesCoarse];
+ fsC = &DC[DIR_0M0 * numberOfLBnodesCoarse];
+ ftC = &DC[DIR_00P * numberOfLBnodesCoarse];
+ fbC = &DC[DIR_00M * numberOfLBnodesCoarse];
+ fneC = &DC[DIR_PP0 * numberOfLBnodesCoarse];
+ fswC = &DC[DIR_MM0 * numberOfLBnodesCoarse];
+ fseC = &DC[DIR_PM0 * numberOfLBnodesCoarse];
+ fnwC = &DC[DIR_MP0 * numberOfLBnodesCoarse];
+ fteC = &DC[DIR_P0P * numberOfLBnodesCoarse];
+ fbwC = &DC[DIR_M0M * numberOfLBnodesCoarse];
+ fbeC = &DC[DIR_P0M * numberOfLBnodesCoarse];
+ ftwC = &DC[DIR_M0P * numberOfLBnodesCoarse];
+ ftnC = &DC[DIR_0PP * numberOfLBnodesCoarse];
+ fbsC = &DC[DIR_0MM * numberOfLBnodesCoarse];
+ fbnC = &DC[DIR_0PM * numberOfLBnodesCoarse];
+ ftsC = &DC[DIR_0MP * numberOfLBnodesCoarse];
+ //fzeroC = &DC[DIR_000 * size_MatC];
+ ftneC = &DC[DIR_PPP * numberOfLBnodesCoarse];
+ ftswC = &DC[DIR_MMP * numberOfLBnodesCoarse];
+ ftseC = &DC[DIR_PMP * numberOfLBnodesCoarse];
+ ftnwC = &DC[DIR_MPP * numberOfLBnodesCoarse];
+ fbneC = &DC[DIR_PPM * numberOfLBnodesCoarse];
+ fbswC = &DC[DIR_MMM * numberOfLBnodesCoarse];
+ fbseC = &DC[DIR_PMM * numberOfLBnodesCoarse];
+ fbnwC = &DC[DIR_MPM * numberOfLBnodesCoarse];
}
else
{
- fwC = &DC[DIR_P00 *size_MatC];
- feC = &DC[DIR_M00 *size_MatC];
- fsC = &DC[DIR_0P0 *size_MatC];
- fnC = &DC[DIR_0M0 *size_MatC];
- fbC = &DC[DIR_00P *size_MatC];
- ftC = &DC[DIR_00M *size_MatC];
- fswC = &DC[DIR_PP0 *size_MatC];
- fneC = &DC[DIR_MM0 *size_MatC];
- fnwC = &DC[DIR_PM0 *size_MatC];
- fseC = &DC[DIR_MP0 *size_MatC];
- fbwC = &DC[DIR_P0P *size_MatC];
- fteC = &DC[DIR_M0M *size_MatC];
- ftwC = &DC[DIR_P0M *size_MatC];
- fbeC = &DC[DIR_M0P *size_MatC];
- fbsC = &DC[DIR_0PP *size_MatC];
- ftnC = &DC[DIR_0MM *size_MatC];
- ftsC = &DC[DIR_0PM *size_MatC];
- fbnC = &DC[DIR_0MP *size_MatC];
- //fzeroC = &DC[DIR_000*size_MatC];
- fbswC = &DC[DIR_PPP *size_MatC];
- fbneC = &DC[DIR_MMP *size_MatC];
- fbnwC = &DC[DIR_PMP *size_MatC];
- fbseC = &DC[DIR_MPP *size_MatC];
- ftswC = &DC[DIR_PPM *size_MatC];
- ftneC = &DC[DIR_MMM *size_MatC];
- ftnwC = &DC[DIR_PMM *size_MatC];
- ftseC = &DC[DIR_MPM *size_MatC];
+ fwC = &DC[DIR_P00 * numberOfLBnodesCoarse];
+ feC = &DC[DIR_M00 * numberOfLBnodesCoarse];
+ fsC = &DC[DIR_0P0 * numberOfLBnodesCoarse];
+ fnC = &DC[DIR_0M0 * numberOfLBnodesCoarse];
+ fbC = &DC[DIR_00P * numberOfLBnodesCoarse];
+ ftC = &DC[DIR_00M * numberOfLBnodesCoarse];
+ fswC = &DC[DIR_PP0 * numberOfLBnodesCoarse];
+ fneC = &DC[DIR_MM0 * numberOfLBnodesCoarse];
+ fnwC = &DC[DIR_PM0 * numberOfLBnodesCoarse];
+ fseC = &DC[DIR_MP0 * numberOfLBnodesCoarse];
+ fbwC = &DC[DIR_P0P * numberOfLBnodesCoarse];
+ fteC = &DC[DIR_M0M * numberOfLBnodesCoarse];
+ ftwC = &DC[DIR_P0M * numberOfLBnodesCoarse];
+ fbeC = &DC[DIR_M0P * numberOfLBnodesCoarse];
+ fbsC = &DC[DIR_0PP * numberOfLBnodesCoarse];
+ ftnC = &DC[DIR_0MM * numberOfLBnodesCoarse];
+ ftsC = &DC[DIR_0PM * numberOfLBnodesCoarse];
+ fbnC = &DC[DIR_0MP * numberOfLBnodesCoarse];
+ //fzeroC = &DC[DIR_000 * size_MatC];
+ fbswC = &DC[DIR_PPP * numberOfLBnodesCoarse];
+ fbneC = &DC[DIR_MMP * numberOfLBnodesCoarse];
+ fbnwC = &DC[DIR_PMP * numberOfLBnodesCoarse];
+ fbseC = &DC[DIR_MPP * numberOfLBnodesCoarse];
+ ftswC = &DC[DIR_PPM * numberOfLBnodesCoarse];
+ ftneC = &DC[DIR_MMM * numberOfLBnodesCoarse];
+ ftnwC = &DC[DIR_PMM * numberOfLBnodesCoarse];
+ ftseC = &DC[DIR_MPM * numberOfLBnodesCoarse];
}
Distributions7 D7F;
- D7F.f[0] = &DD7F[0*size_MatF];
- D7F.f[1] = &DD7F[1*size_MatF];
- D7F.f[2] = &DD7F[2*size_MatF];
- D7F.f[3] = &DD7F[3*size_MatF];
- D7F.f[4] = &DD7F[4*size_MatF];
- D7F.f[5] = &DD7F[5*size_MatF];
- D7F.f[6] = &DD7F[6*size_MatF];
+ D7F.f[0] = &DD7F[0*numberOfLBnodesFine];
+ D7F.f[1] = &DD7F[1*numberOfLBnodesFine];
+ D7F.f[2] = &DD7F[2*numberOfLBnodesFine];
+ D7F.f[3] = &DD7F[3*numberOfLBnodesFine];
+ D7F.f[4] = &DD7F[4*numberOfLBnodesFine];
+ D7F.f[5] = &DD7F[5*numberOfLBnodesFine];
+ D7F.f[6] = &DD7F[6*numberOfLBnodesFine];
Distributions7 D7C;
if (isEvenTimestep==true)
{
- D7C.f[0] = &DD7C[0*size_MatC];
- D7C.f[1] = &DD7C[1*size_MatC];
- D7C.f[2] = &DD7C[2*size_MatC];
- D7C.f[3] = &DD7C[3*size_MatC];
- D7C.f[4] = &DD7C[4*size_MatC];
- D7C.f[5] = &DD7C[5*size_MatC];
- D7C.f[6] = &DD7C[6*size_MatC];
+ D7C.f[0] = &DD7C[0*numberOfLBnodesCoarse];
+ D7C.f[1] = &DD7C[1*numberOfLBnodesCoarse];
+ D7C.f[2] = &DD7C[2*numberOfLBnodesCoarse];
+ D7C.f[3] = &DD7C[3*numberOfLBnodesCoarse];
+ D7C.f[4] = &DD7C[4*numberOfLBnodesCoarse];
+ D7C.f[5] = &DD7C[5*numberOfLBnodesCoarse];
+ D7C.f[6] = &DD7C[6*numberOfLBnodesCoarse];
}
else
{
- D7C.f[0] = &DD7C[0*size_MatC];
- D7C.f[2] = &DD7C[1*size_MatC];
- D7C.f[1] = &DD7C[2*size_MatC];
- D7C.f[4] = &DD7C[3*size_MatC];
- D7C.f[3] = &DD7C[4*size_MatC];
- D7C.f[6] = &DD7C[5*size_MatC];
- D7C.f[5] = &DD7C[6*size_MatC];
+ D7C.f[0] = &DD7C[0*numberOfLBnodesCoarse];
+ D7C.f[2] = &DD7C[1*numberOfLBnodesCoarse];
+ D7C.f[1] = &DD7C[2*numberOfLBnodesCoarse];
+ D7C.f[4] = &DD7C[3*numberOfLBnodesCoarse];
+ D7C.f[3] = &DD7C[4*numberOfLBnodesCoarse];
+ D7C.f[6] = &DD7C[5*numberOfLBnodesCoarse];
+ D7C.f[5] = &DD7C[6*numberOfLBnodesCoarse];
}
////////////////////////////////////////////////////////////////////////////////
@@ -44476,8 +44476,8 @@ __global__ void scaleCFThS7( real* DC,
unsigned int* neighborFX,
unsigned int* neighborFY,
unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
+ unsigned long long numberOfLBnodesCoarse,
+ unsigned long long numberOfLBnodesFine,
bool isEvenTimestep,
unsigned int* posCSWB,
unsigned int* posFSWB,
@@ -44487,128 +44487,128 @@ __global__ void scaleCFThS7( real* DC,
{
real *feF, *fwF, *fnF, *fsF, *ftF, *fbF, *fneF, *fswF, *fseF, *fnwF, *fteF, *fbwF, *fbeF, *ftwF, *ftnF, *fbsF, *fbnF, *ftsF, /**fzeroF,*/ *ftneF, *ftswF, *ftseF, *ftnwF, *fbneF, *fbswF, *fbseF, *fbnwF;
- feF = &DF[DIR_P00 *size_MatF];
- fwF = &DF[DIR_M00 *size_MatF];
- fnF = &DF[DIR_0P0 *size_MatF];
- fsF = &DF[DIR_0M0 *size_MatF];
- ftF = &DF[DIR_00P *size_MatF];
- fbF = &DF[DIR_00M *size_MatF];
- fneF = &DF[DIR_PP0 *size_MatF];
- fswF = &DF[DIR_MM0 *size_MatF];
- fseF = &DF[DIR_PM0 *size_MatF];
- fnwF = &DF[DIR_MP0 *size_MatF];
- fteF = &DF[DIR_P0P *size_MatF];
- fbwF = &DF[DIR_M0M *size_MatF];
- fbeF = &DF[DIR_P0M *size_MatF];
- ftwF = &DF[DIR_M0P *size_MatF];
- ftnF = &DF[DIR_0PP *size_MatF];
- fbsF = &DF[DIR_0MM *size_MatF];
- fbnF = &DF[DIR_0PM *size_MatF];
- ftsF = &DF[DIR_0MP *size_MatF];
- //fzeroF = &DF[DIR_000*size_MatF];
- ftneF = &DF[DIR_PPP *size_MatF];
- ftswF = &DF[DIR_MMP *size_MatF];
- ftseF = &DF[DIR_PMP *size_MatF];
- ftnwF = &DF[DIR_MPP *size_MatF];
- fbneF = &DF[DIR_PPM *size_MatF];
- fbswF = &DF[DIR_MMM *size_MatF];
- fbseF = &DF[DIR_PMM *size_MatF];
- fbnwF = &DF[DIR_MPM *size_MatF];
+ feF = &DF[DIR_P00 * numberOfLBnodesFine];
+ fwF = &DF[DIR_M00 * numberOfLBnodesFine];
+ fnF = &DF[DIR_0P0 * numberOfLBnodesFine];
+ fsF = &DF[DIR_0M0 * numberOfLBnodesFine];
+ ftF = &DF[DIR_00P * numberOfLBnodesFine];
+ fbF = &DF[DIR_00M * numberOfLBnodesFine];
+ fneF = &DF[DIR_PP0 * numberOfLBnodesFine];
+ fswF = &DF[DIR_MM0 * numberOfLBnodesFine];
+ fseF = &DF[DIR_PM0 * numberOfLBnodesFine];
+ fnwF = &DF[DIR_MP0 * numberOfLBnodesFine];
+ fteF = &DF[DIR_P0P * numberOfLBnodesFine];
+ fbwF = &DF[DIR_M0M * numberOfLBnodesFine];
+ fbeF = &DF[DIR_P0M * numberOfLBnodesFine];
+ ftwF = &DF[DIR_M0P * numberOfLBnodesFine];
+ ftnF = &DF[DIR_0PP * numberOfLBnodesFine];
+ fbsF = &DF[DIR_0MM * numberOfLBnodesFine];
+ fbnF = &DF[DIR_0PM * numberOfLBnodesFine];
+ ftsF = &DF[DIR_0MP * numberOfLBnodesFine];
+ //fzeroF = &DF[DIR_000 * size_MatF];
+ ftneF = &DF[DIR_PPP * numberOfLBnodesFine];
+ ftswF = &DF[DIR_MMP * numberOfLBnodesFine];
+ ftseF = &DF[DIR_PMP * numberOfLBnodesFine];
+ ftnwF = &DF[DIR_MPP * numberOfLBnodesFine];
+ fbneF = &DF[DIR_PPM * numberOfLBnodesFine];
+ fbswF = &DF[DIR_MMM * numberOfLBnodesFine];
+ fbseF = &DF[DIR_PMM * numberOfLBnodesFine];
+ fbnwF = &DF[DIR_MPM * numberOfLBnodesFine];
real *feC, *fwC, *fnC, *fsC, *ftC, *fbC, *fneC, *fswC, *fseC, *fnwC, *fteC, *fbwC, *fbeC, *ftwC, *ftnC, *fbsC, *fbnC, *ftsC, //*fzeroC,
*ftneC, *ftswC, *ftseC, *ftnwC, *fbneC, *fbswC, *fbseC, *fbnwC;
if (isEvenTimestep==true)
{
- feC = &DC[DIR_P00 *size_MatC];
- fwC = &DC[DIR_M00 *size_MatC];
- fnC = &DC[DIR_0P0 *size_MatC];
- fsC = &DC[DIR_0M0 *size_MatC];
- ftC = &DC[DIR_00P *size_MatC];
- fbC = &DC[DIR_00M *size_MatC];
- fneC = &DC[DIR_PP0 *size_MatC];
- fswC = &DC[DIR_MM0 *size_MatC];
- fseC = &DC[DIR_PM0 *size_MatC];
- fnwC = &DC[DIR_MP0 *size_MatC];
- fteC = &DC[DIR_P0P *size_MatC];
- fbwC = &DC[DIR_M0M *size_MatC];
- fbeC = &DC[DIR_P0M *size_MatC];
- ftwC = &DC[DIR_M0P *size_MatC];
- ftnC = &DC[DIR_0PP *size_MatC];
- fbsC = &DC[DIR_0MM *size_MatC];
- fbnC = &DC[DIR_0PM *size_MatC];
- ftsC = &DC[DIR_0MP *size_MatC];
- //fzeroC = &DC[DIR_000*size_MatC];
- ftneC = &DC[DIR_PPP *size_MatC];
- ftswC = &DC[DIR_MMP *size_MatC];
- ftseC = &DC[DIR_PMP *size_MatC];
- ftnwC = &DC[DIR_MPP *size_MatC];
- fbneC = &DC[DIR_PPM *size_MatC];
- fbswC = &DC[DIR_MMM *size_MatC];
- fbseC = &DC[DIR_PMM *size_MatC];
- fbnwC = &DC[DIR_MPM *size_MatC];
+ feC = &DC[DIR_P00 * numberOfLBnodesCoarse];
+ fwC = &DC[DIR_M00 * numberOfLBnodesCoarse];
+ fnC = &DC[DIR_0P0 * numberOfLBnodesCoarse];
+ fsC = &DC[DIR_0M0 * numberOfLBnodesCoarse];
+ ftC = &DC[DIR_00P * numberOfLBnodesCoarse];
+ fbC = &DC[DIR_00M * numberOfLBnodesCoarse];
+ fneC = &DC[DIR_PP0 * numberOfLBnodesCoarse];
+ fswC = &DC[DIR_MM0 * numberOfLBnodesCoarse];
+ fseC = &DC[DIR_PM0 * numberOfLBnodesCoarse];
+ fnwC = &DC[DIR_MP0 * numberOfLBnodesCoarse];
+ fteC = &DC[DIR_P0P * numberOfLBnodesCoarse];
+ fbwC = &DC[DIR_M0M * numberOfLBnodesCoarse];
+ fbeC = &DC[DIR_P0M * numberOfLBnodesCoarse];
+ ftwC = &DC[DIR_M0P * numberOfLBnodesCoarse];
+ ftnC = &DC[DIR_0PP * numberOfLBnodesCoarse];
+ fbsC = &DC[DIR_0MM * numberOfLBnodesCoarse];
+ fbnC = &DC[DIR_0PM * numberOfLBnodesCoarse];
+ ftsC = &DC[DIR_0MP * numberOfLBnodesCoarse];
+ //fzeroC = &DC[DIR_000 * size_MatC];
+ ftneC = &DC[DIR_PPP * numberOfLBnodesCoarse];
+ ftswC = &DC[DIR_MMP * numberOfLBnodesCoarse];
+ ftseC = &DC[DIR_PMP * numberOfLBnodesCoarse];
+ ftnwC = &DC[DIR_MPP * numberOfLBnodesCoarse];
+ fbneC = &DC[DIR_PPM * numberOfLBnodesCoarse];
+ fbswC = &DC[DIR_MMM * numberOfLBnodesCoarse];
+ fbseC = &DC[DIR_PMM * numberOfLBnodesCoarse];
+ fbnwC = &DC[DIR_MPM * numberOfLBnodesCoarse];
}
else
{
- fwC = &DC[DIR_P00 *size_MatC];
- feC = &DC[DIR_M00 *size_MatC];
- fsC = &DC[DIR_0P0 *size_MatC];
- fnC = &DC[DIR_0M0 *size_MatC];
- fbC = &DC[DIR_00P *size_MatC];
- ftC = &DC[DIR_00M *size_MatC];
- fswC = &DC[DIR_PP0 *size_MatC];
- fneC = &DC[DIR_MM0 *size_MatC];
- fnwC = &DC[DIR_PM0 *size_MatC];
- fseC = &DC[DIR_MP0 *size_MatC];
- fbwC = &DC[DIR_P0P *size_MatC];
- fteC = &DC[DIR_M0M *size_MatC];
- ftwC = &DC[DIR_P0M *size_MatC];
- fbeC = &DC[DIR_M0P *size_MatC];
- fbsC = &DC[DIR_0PP *size_MatC];
- ftnC = &DC[DIR_0MM *size_MatC];
- ftsC = &DC[DIR_0PM *size_MatC];
- fbnC = &DC[DIR_0MP *size_MatC];
- //fzeroC = &DC[DIR_000*size_MatC];
- fbswC = &DC[DIR_PPP *size_MatC];
- fbneC = &DC[DIR_MMP *size_MatC];
- fbnwC = &DC[DIR_PMP *size_MatC];
- fbseC = &DC[DIR_MPP *size_MatC];
- ftswC = &DC[DIR_PPM *size_MatC];
- ftneC = &DC[DIR_MMM *size_MatC];
- ftnwC = &DC[DIR_PMM *size_MatC];
- ftseC = &DC[DIR_MPM *size_MatC];
+ fwC = &DC[DIR_P00 * numberOfLBnodesCoarse];
+ feC = &DC[DIR_M00 * numberOfLBnodesCoarse];
+ fsC = &DC[DIR_0P0 * numberOfLBnodesCoarse];
+ fnC = &DC[DIR_0M0 * numberOfLBnodesCoarse];
+ fbC = &DC[DIR_00P * numberOfLBnodesCoarse];
+ ftC = &DC[DIR_00M * numberOfLBnodesCoarse];
+ fswC = &DC[DIR_PP0 * numberOfLBnodesCoarse];
+ fneC = &DC[DIR_MM0 * numberOfLBnodesCoarse];
+ fnwC = &DC[DIR_PM0 * numberOfLBnodesCoarse];
+ fseC = &DC[DIR_MP0 * numberOfLBnodesCoarse];
+ fbwC = &DC[DIR_P0P * numberOfLBnodesCoarse];
+ fteC = &DC[DIR_M0M * numberOfLBnodesCoarse];
+ ftwC = &DC[DIR_P0M * numberOfLBnodesCoarse];
+ fbeC = &DC[DIR_M0P * numberOfLBnodesCoarse];
+ fbsC = &DC[DIR_0PP * numberOfLBnodesCoarse];
+ ftnC = &DC[DIR_0MM * numberOfLBnodesCoarse];
+ ftsC = &DC[DIR_0PM * numberOfLBnodesCoarse];
+ fbnC = &DC[DIR_0MP * numberOfLBnodesCoarse];
+ //fzeroC = &DC[DIR_000 * size_MatC];
+ fbswC = &DC[DIR_PPP * numberOfLBnodesCoarse];
+ fbneC = &DC[DIR_MMP * numberOfLBnodesCoarse];
+ fbnwC = &DC[DIR_PMP * numberOfLBnodesCoarse];
+ fbseC = &DC[DIR_MPP * numberOfLBnodesCoarse];
+ ftswC = &DC[DIR_PPM * numberOfLBnodesCoarse];
+ ftneC = &DC[DIR_MMM * numberOfLBnodesCoarse];
+ ftnwC = &DC[DIR_PMM * numberOfLBnodesCoarse];
+ ftseC = &DC[DIR_MPM * numberOfLBnodesCoarse];
}
Distributions7 D7F;
- D7F.f[0] = &DD7F[0*size_MatF];
- D7F.f[1] = &DD7F[1*size_MatF];
- D7F.f[2] = &DD7F[2*size_MatF];
- D7F.f[3] = &DD7F[3*size_MatF];
- D7F.f[4] = &DD7F[4*size_MatF];
- D7F.f[5] = &DD7F[5*size_MatF];
- D7F.f[6] = &DD7F[6*size_MatF];
+ D7F.f[0] = &DD7F[0*numberOfLBnodesFine];
+ D7F.f[1] = &DD7F[1*numberOfLBnodesFine];
+ D7F.f[2] = &DD7F[2*numberOfLBnodesFine];
+ D7F.f[3] = &DD7F[3*numberOfLBnodesFine];
+ D7F.f[4] = &DD7F[4*numberOfLBnodesFine];
+ D7F.f[5] = &DD7F[5*numberOfLBnodesFine];
+ D7F.f[6] = &DD7F[6*numberOfLBnodesFine];
Distributions7 D7C;
if (isEvenTimestep==true)
{
- D7C.f[0] = &DD7C[0*size_MatC];
- D7C.f[1] = &DD7C[1*size_MatC];
- D7C.f[2] = &DD7C[2*size_MatC];
- D7C.f[3] = &DD7C[3*size_MatC];
- D7C.f[4] = &DD7C[4*size_MatC];
- D7C.f[5] = &DD7C[5*size_MatC];
- D7C.f[6] = &DD7C[6*size_MatC];
+ D7C.f[0] = &DD7C[0*numberOfLBnodesCoarse];
+ D7C.f[1] = &DD7C[1*numberOfLBnodesCoarse];
+ D7C.f[2] = &DD7C[2*numberOfLBnodesCoarse];
+ D7C.f[3] = &DD7C[3*numberOfLBnodesCoarse];
+ D7C.f[4] = &DD7C[4*numberOfLBnodesCoarse];
+ D7C.f[5] = &DD7C[5*numberOfLBnodesCoarse];
+ D7C.f[6] = &DD7C[6*numberOfLBnodesCoarse];
}
else
{
- D7C.f[0] = &DD7C[0*size_MatC];
- D7C.f[2] = &DD7C[1*size_MatC];
- D7C.f[1] = &DD7C[2*size_MatC];
- D7C.f[4] = &DD7C[3*size_MatC];
- D7C.f[3] = &DD7C[4*size_MatC];
- D7C.f[6] = &DD7C[5*size_MatC];
- D7C.f[5] = &DD7C[6*size_MatC];
+ D7C.f[0] = &DD7C[0*numberOfLBnodesCoarse];
+ D7C.f[2] = &DD7C[1*numberOfLBnodesCoarse];
+ D7C.f[1] = &DD7C[2*numberOfLBnodesCoarse];
+ D7C.f[4] = &DD7C[3*numberOfLBnodesCoarse];
+ D7C.f[3] = &DD7C[4*numberOfLBnodesCoarse];
+ D7C.f[6] = &DD7C[5*numberOfLBnodesCoarse];
+ D7C.f[5] = &DD7C[6*numberOfLBnodesCoarse];
}
////////////////////////////////////////////////////////////////////////////////
@@ -45599,8 +45599,8 @@ __global__ void scaleCFThS27( real* DC,
unsigned int* neighborFX,
unsigned int* neighborFY,
unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
+ unsigned long long numberOfLBnodesCoarse,
+ unsigned long long numberOfLBnodesFine,
bool isEvenTimestep,
unsigned int* posCSWB,
unsigned int* posFSWB,
@@ -45611,188 +45611,188 @@ __global__ void scaleCFThS27( real* DC,
{
real *feF, *fwF, *fnF, *fsF, *ftF, *fbF, *fneF, *fswF, *fseF, *fnwF, *fteF, *fbwF, *fbeF, *ftwF, *ftnF, *fbsF, *fbnF, *ftsF, /**fzeroF,*/ *ftneF, *ftswF, *ftseF, *ftnwF, *fbneF, *fbswF, *fbseF, *fbnwF;
- feF = &DF[DIR_P00 *size_MatF];
- fwF = &DF[DIR_M00 *size_MatF];
- fnF = &DF[DIR_0P0 *size_MatF];
- fsF = &DF[DIR_0M0 *size_MatF];
- ftF = &DF[DIR_00P *size_MatF];
- fbF = &DF[DIR_00M *size_MatF];
- fneF = &DF[DIR_PP0 *size_MatF];
- fswF = &DF[DIR_MM0 *size_MatF];
- fseF = &DF[DIR_PM0 *size_MatF];
- fnwF = &DF[DIR_MP0 *size_MatF];
- fteF = &DF[DIR_P0P *size_MatF];
- fbwF = &DF[DIR_M0M *size_MatF];
- fbeF = &DF[DIR_P0M *size_MatF];
- ftwF = &DF[DIR_M0P *size_MatF];
- ftnF = &DF[DIR_0PP *size_MatF];
- fbsF = &DF[DIR_0MM *size_MatF];
- fbnF = &DF[DIR_0PM *size_MatF];
- ftsF = &DF[DIR_0MP *size_MatF];
- //fzeroF = &DF[DIR_000*size_MatF];
- ftneF = &DF[DIR_PPP *size_MatF];
- ftswF = &DF[DIR_MMP *size_MatF];
- ftseF = &DF[DIR_PMP *size_MatF];
- ftnwF = &DF[DIR_MPP *size_MatF];
- fbneF = &DF[DIR_PPM *size_MatF];
- fbswF = &DF[DIR_MMM *size_MatF];
- fbseF = &DF[DIR_PMM *size_MatF];
- fbnwF = &DF[DIR_MPM *size_MatF];
+ feF = &DF[DIR_P00 * numberOfLBnodesFine];
+ fwF = &DF[DIR_M00 * numberOfLBnodesFine];
+ fnF = &DF[DIR_0P0 * numberOfLBnodesFine];
+ fsF = &DF[DIR_0M0 * numberOfLBnodesFine];
+ ftF = &DF[DIR_00P * numberOfLBnodesFine];
+ fbF = &DF[DIR_00M * numberOfLBnodesFine];
+ fneF = &DF[DIR_PP0 * numberOfLBnodesFine];
+ fswF = &DF[DIR_MM0 * numberOfLBnodesFine];
+ fseF = &DF[DIR_PM0 * numberOfLBnodesFine];
+ fnwF = &DF[DIR_MP0 * numberOfLBnodesFine];
+ fteF = &DF[DIR_P0P * numberOfLBnodesFine];
+ fbwF = &DF[DIR_M0M * numberOfLBnodesFine];
+ fbeF = &DF[DIR_P0M * numberOfLBnodesFine];
+ ftwF = &DF[DIR_M0P * numberOfLBnodesFine];
+ ftnF = &DF[DIR_0PP * numberOfLBnodesFine];
+ fbsF = &DF[DIR_0MM * numberOfLBnodesFine];
+ fbnF = &DF[DIR_0PM * numberOfLBnodesFine];
+ ftsF = &DF[DIR_0MP * numberOfLBnodesFine];
+ //fzeroF = &DF[DIR_000 * size_MatF];
+ ftneF = &DF[DIR_PPP * numberOfLBnodesFine];
+ ftswF = &DF[DIR_MMP * numberOfLBnodesFine];
+ ftseF = &DF[DIR_PMP * numberOfLBnodesFine];
+ ftnwF = &DF[DIR_MPP * numberOfLBnodesFine];
+ fbneF = &DF[DIR_PPM * numberOfLBnodesFine];
+ fbswF = &DF[DIR_MMM * numberOfLBnodesFine];
+ fbseF = &DF[DIR_PMM * numberOfLBnodesFine];
+ fbnwF = &DF[DIR_MPM * numberOfLBnodesFine];
real *feC, *fwC, *fnC, *fsC, *ftC, *fbC, *fneC, *fswC, *fseC, *fnwC, *fteC, *fbwC, *fbeC, *ftwC, *ftnC, *fbsC, *fbnC, *ftsC, //*fzeroC,
*ftneC, *ftswC, *ftseC, *ftnwC, *fbneC, *fbswC, *fbseC, *fbnwC;
if (isEvenTimestep==true)
{
- feC = &DC[DIR_P00 *size_MatC];
- fwC = &DC[DIR_M00 *size_MatC];
- fnC = &DC[DIR_0P0 *size_MatC];
- fsC = &DC[DIR_0M0 *size_MatC];
- ftC = &DC[DIR_00P *size_MatC];
- fbC = &DC[DIR_00M *size_MatC];
- fneC = &DC[DIR_PP0 *size_MatC];
- fswC = &DC[DIR_MM0 *size_MatC];
- fseC = &DC[DIR_PM0 *size_MatC];
- fnwC = &DC[DIR_MP0 *size_MatC];
- fteC = &DC[DIR_P0P *size_MatC];
- fbwC = &DC[DIR_M0M *size_MatC];
- fbeC = &DC[DIR_P0M *size_MatC];
- ftwC = &DC[DIR_M0P *size_MatC];
- ftnC = &DC[DIR_0PP *size_MatC];
- fbsC = &DC[DIR_0MM *size_MatC];
- fbnC = &DC[DIR_0PM *size_MatC];
- ftsC = &DC[DIR_0MP *size_MatC];
- //fzeroC = &DC[DIR_000*size_MatC];
- ftneC = &DC[DIR_PPP *size_MatC];
- ftswC = &DC[DIR_MMP *size_MatC];
- ftseC = &DC[DIR_PMP *size_MatC];
- ftnwC = &DC[DIR_MPP *size_MatC];
- fbneC = &DC[DIR_PPM *size_MatC];
- fbswC = &DC[DIR_MMM *size_MatC];
- fbseC = &DC[DIR_PMM *size_MatC];
- fbnwC = &DC[DIR_MPM *size_MatC];
+ feC = &DC[DIR_P00 * numberOfLBnodesCoarse];
+ fwC = &DC[DIR_M00 * numberOfLBnodesCoarse];
+ fnC = &DC[DIR_0P0 * numberOfLBnodesCoarse];
+ fsC = &DC[DIR_0M0 * numberOfLBnodesCoarse];
+ ftC = &DC[DIR_00P * numberOfLBnodesCoarse];
+ fbC = &DC[DIR_00M * numberOfLBnodesCoarse];
+ fneC = &DC[DIR_PP0 * numberOfLBnodesCoarse];
+ fswC = &DC[DIR_MM0 * numberOfLBnodesCoarse];
+ fseC = &DC[DIR_PM0 * numberOfLBnodesCoarse];
+ fnwC = &DC[DIR_MP0 * numberOfLBnodesCoarse];
+ fteC = &DC[DIR_P0P * numberOfLBnodesCoarse];
+ fbwC = &DC[DIR_M0M * numberOfLBnodesCoarse];
+ fbeC = &DC[DIR_P0M * numberOfLBnodesCoarse];
+ ftwC = &DC[DIR_M0P * numberOfLBnodesCoarse];
+ ftnC = &DC[DIR_0PP * numberOfLBnodesCoarse];
+ fbsC = &DC[DIR_0MM * numberOfLBnodesCoarse];
+ fbnC = &DC[DIR_0PM * numberOfLBnodesCoarse];
+ ftsC = &DC[DIR_0MP * numberOfLBnodesCoarse];
+ //fzeroC = &DC[DIR_000 * size_MatC];
+ ftneC = &DC[DIR_PPP * numberOfLBnodesCoarse];
+ ftswC = &DC[DIR_MMP * numberOfLBnodesCoarse];
+ ftseC = &DC[DIR_PMP * numberOfLBnodesCoarse];
+ ftnwC = &DC[DIR_MPP * numberOfLBnodesCoarse];
+ fbneC = &DC[DIR_PPM * numberOfLBnodesCoarse];
+ fbswC = &DC[DIR_MMM * numberOfLBnodesCoarse];
+ fbseC = &DC[DIR_PMM * numberOfLBnodesCoarse];
+ fbnwC = &DC[DIR_MPM * numberOfLBnodesCoarse];
}
else
{
- fwC = &DC[DIR_P00 *size_MatC];
- feC = &DC[DIR_M00 *size_MatC];
- fsC = &DC[DIR_0P0 *size_MatC];
- fnC = &DC[DIR_0M0 *size_MatC];
- fbC = &DC[DIR_00P *size_MatC];
- ftC = &DC[DIR_00M *size_MatC];
- fswC = &DC[DIR_PP0 *size_MatC];
- fneC = &DC[DIR_MM0 *size_MatC];
- fnwC = &DC[DIR_PM0 *size_MatC];
- fseC = &DC[DIR_MP0 *size_MatC];
- fbwC = &DC[DIR_P0P *size_MatC];
- fteC = &DC[DIR_M0M *size_MatC];
- ftwC = &DC[DIR_P0M *size_MatC];
- fbeC = &DC[DIR_M0P *size_MatC];
- fbsC = &DC[DIR_0PP *size_MatC];
- ftnC = &DC[DIR_0MM *size_MatC];
- ftsC = &DC[DIR_0PM *size_MatC];
- fbnC = &DC[DIR_0MP *size_MatC];
- //fzeroC = &DC[DIR_000*size_MatC];
- fbswC = &DC[DIR_PPP *size_MatC];
- fbneC = &DC[DIR_MMP *size_MatC];
- fbnwC = &DC[DIR_PMP *size_MatC];
- fbseC = &DC[DIR_MPP *size_MatC];
- ftswC = &DC[DIR_PPM *size_MatC];
- ftneC = &DC[DIR_MMM *size_MatC];
- ftnwC = &DC[DIR_PMM *size_MatC];
- ftseC = &DC[DIR_MPM *size_MatC];
+ fwC = &DC[DIR_P00 * numberOfLBnodesCoarse];
+ feC = &DC[DIR_M00 * numberOfLBnodesCoarse];
+ fsC = &DC[DIR_0P0 * numberOfLBnodesCoarse];
+ fnC = &DC[DIR_0M0 * numberOfLBnodesCoarse];
+ fbC = &DC[DIR_00P * numberOfLBnodesCoarse];
+ ftC = &DC[DIR_00M * numberOfLBnodesCoarse];
+ fswC = &DC[DIR_PP0 * numberOfLBnodesCoarse];
+ fneC = &DC[DIR_MM0 * numberOfLBnodesCoarse];
+ fnwC = &DC[DIR_PM0 * numberOfLBnodesCoarse];
+ fseC = &DC[DIR_MP0 * numberOfLBnodesCoarse];
+ fbwC = &DC[DIR_P0P * numberOfLBnodesCoarse];
+ fteC = &DC[DIR_M0M * numberOfLBnodesCoarse];
+ ftwC = &DC[DIR_P0M * numberOfLBnodesCoarse];
+ fbeC = &DC[DIR_M0P * numberOfLBnodesCoarse];
+ fbsC = &DC[DIR_0PP * numberOfLBnodesCoarse];
+ ftnC = &DC[DIR_0MM * numberOfLBnodesCoarse];
+ ftsC = &DC[DIR_0PM * numberOfLBnodesCoarse];
+ fbnC = &DC[DIR_0MP * numberOfLBnodesCoarse];
+ //fzeroC = &DC[DIR_000 * size_MatC];
+ fbswC = &DC[DIR_PPP * numberOfLBnodesCoarse];
+ fbneC = &DC[DIR_MMP * numberOfLBnodesCoarse];
+ fbnwC = &DC[DIR_PMP * numberOfLBnodesCoarse];
+ fbseC = &DC[DIR_MPP * numberOfLBnodesCoarse];
+ ftswC = &DC[DIR_PPM * numberOfLBnodesCoarse];
+ ftneC = &DC[DIR_MMM * numberOfLBnodesCoarse];
+ ftnwC = &DC[DIR_PMM * numberOfLBnodesCoarse];
+ ftseC = &DC[DIR_MPM * numberOfLBnodesCoarse];
}
Distributions27 D27F;
- D27F.f[DIR_P00 ] = &DD27F[DIR_P00 *size_MatF];
- D27F.f[DIR_M00 ] = &DD27F[DIR_M00 *size_MatF];
- D27F.f[DIR_0P0 ] = &DD27F[DIR_0P0 *size_MatF];
- D27F.f[DIR_0M0 ] = &DD27F[DIR_0M0 *size_MatF];
- D27F.f[DIR_00P ] = &DD27F[DIR_00P *size_MatF];
- D27F.f[DIR_00M ] = &DD27F[DIR_00M *size_MatF];
- D27F.f[DIR_PP0 ] = &DD27F[DIR_PP0 *size_MatF];
- D27F.f[DIR_MM0 ] = &DD27F[DIR_MM0 *size_MatF];
- D27F.f[DIR_PM0 ] = &DD27F[DIR_PM0 *size_MatF];
- D27F.f[DIR_MP0 ] = &DD27F[DIR_MP0 *size_MatF];
- D27F.f[DIR_P0P ] = &DD27F[DIR_P0P *size_MatF];
- D27F.f[DIR_M0M ] = &DD27F[DIR_M0M *size_MatF];
- D27F.f[DIR_P0M ] = &DD27F[DIR_P0M *size_MatF];
- D27F.f[DIR_M0P ] = &DD27F[DIR_M0P *size_MatF];
- D27F.f[DIR_0PP ] = &DD27F[DIR_0PP *size_MatF];
- D27F.f[DIR_0MM ] = &DD27F[DIR_0MM *size_MatF];
- D27F.f[DIR_0PM ] = &DD27F[DIR_0PM *size_MatF];
- D27F.f[DIR_0MP ] = &DD27F[DIR_0MP *size_MatF];
- D27F.f[DIR_000] = &DD27F[DIR_000*size_MatF];
- D27F.f[DIR_PPP ] = &DD27F[DIR_PPP *size_MatF];
- D27F.f[DIR_MMP ] = &DD27F[DIR_MMP *size_MatF];
- D27F.f[DIR_PMP ] = &DD27F[DIR_PMP *size_MatF];
- D27F.f[DIR_MPP ] = &DD27F[DIR_MPP *size_MatF];
- D27F.f[DIR_PPM ] = &DD27F[DIR_PPM *size_MatF];
- D27F.f[DIR_MMM ] = &DD27F[DIR_MMM *size_MatF];
- D27F.f[DIR_PMM ] = &DD27F[DIR_PMM *size_MatF];
- D27F.f[DIR_MPM ] = &DD27F[DIR_MPM *size_MatF];
+ D27F.f[DIR_P00] = &DD27F[DIR_P00 * numberOfLBnodesFine];
+ D27F.f[DIR_M00] = &DD27F[DIR_M00 * numberOfLBnodesFine];
+ D27F.f[DIR_0P0] = &DD27F[DIR_0P0 * numberOfLBnodesFine];
+ D27F.f[DIR_0M0] = &DD27F[DIR_0M0 * numberOfLBnodesFine];
+ D27F.f[DIR_00P] = &DD27F[DIR_00P * numberOfLBnodesFine];
+ D27F.f[DIR_00M] = &DD27F[DIR_00M * numberOfLBnodesFine];
+ D27F.f[DIR_PP0] = &DD27F[DIR_PP0 * numberOfLBnodesFine];
+ D27F.f[DIR_MM0] = &DD27F[DIR_MM0 * numberOfLBnodesFine];
+ D27F.f[DIR_PM0] = &DD27F[DIR_PM0 * numberOfLBnodesFine];
+ D27F.f[DIR_MP0] = &DD27F[DIR_MP0 * numberOfLBnodesFine];
+ D27F.f[DIR_P0P] = &DD27F[DIR_P0P * numberOfLBnodesFine];
+ D27F.f[DIR_M0M] = &DD27F[DIR_M0M * numberOfLBnodesFine];
+ D27F.f[DIR_P0M] = &DD27F[DIR_P0M * numberOfLBnodesFine];
+ D27F.f[DIR_M0P] = &DD27F[DIR_M0P * numberOfLBnodesFine];
+ D27F.f[DIR_0PP] = &DD27F[DIR_0PP * numberOfLBnodesFine];
+ D27F.f[DIR_0MM] = &DD27F[DIR_0MM * numberOfLBnodesFine];
+ D27F.f[DIR_0PM] = &DD27F[DIR_0PM * numberOfLBnodesFine];
+ D27F.f[DIR_0MP] = &DD27F[DIR_0MP * numberOfLBnodesFine];
+ D27F.f[DIR_000] = &DD27F[DIR_000 * numberOfLBnodesFine];
+ D27F.f[DIR_PPP] = &DD27F[DIR_PPP * numberOfLBnodesFine];
+ D27F.f[DIR_MMP] = &DD27F[DIR_MMP * numberOfLBnodesFine];
+ D27F.f[DIR_PMP] = &DD27F[DIR_PMP * numberOfLBnodesFine];
+ D27F.f[DIR_MPP] = &DD27F[DIR_MPP * numberOfLBnodesFine];
+ D27F.f[DIR_PPM] = &DD27F[DIR_PPM * numberOfLBnodesFine];
+ D27F.f[DIR_MMM] = &DD27F[DIR_MMM * numberOfLBnodesFine];
+ D27F.f[DIR_PMM] = &DD27F[DIR_PMM * numberOfLBnodesFine];
+ D27F.f[DIR_MPM] = &DD27F[DIR_MPM * numberOfLBnodesFine];
Distributions27 D27C;
if (isEvenTimestep==true)
{
- D27C.f[DIR_P00 ] = &DD27C[DIR_P00 *size_MatC];
- D27C.f[DIR_M00 ] = &DD27C[DIR_M00 *size_MatC];
- D27C.f[DIR_0P0 ] = &DD27C[DIR_0P0 *size_MatC];
- D27C.f[DIR_0M0 ] = &DD27C[DIR_0M0 *size_MatC];
- D27C.f[DIR_00P ] = &DD27C[DIR_00P *size_MatC];
- D27C.f[DIR_00M ] = &DD27C[DIR_00M *size_MatC];
- D27C.f[DIR_PP0 ] = &DD27C[DIR_PP0 *size_MatC];
- D27C.f[DIR_MM0 ] = &DD27C[DIR_MM0 *size_MatC];
- D27C.f[DIR_PM0 ] = &DD27C[DIR_PM0 *size_MatC];
- D27C.f[DIR_MP0 ] = &DD27C[DIR_MP0 *size_MatC];
- D27C.f[DIR_P0P ] = &DD27C[DIR_P0P *size_MatC];
- D27C.f[DIR_M0M ] = &DD27C[DIR_M0M *size_MatC];
- D27C.f[DIR_P0M ] = &DD27C[DIR_P0M *size_MatC];
- D27C.f[DIR_M0P ] = &DD27C[DIR_M0P *size_MatC];
- D27C.f[DIR_0PP ] = &DD27C[DIR_0PP *size_MatC];
- D27C.f[DIR_0MM ] = &DD27C[DIR_0MM *size_MatC];
- D27C.f[DIR_0PM ] = &DD27C[DIR_0PM *size_MatC];
- D27C.f[DIR_0MP ] = &DD27C[DIR_0MP *size_MatC];
- D27C.f[DIR_000] = &DD27C[DIR_000*size_MatC];
- D27C.f[DIR_PPP ] = &DD27C[DIR_PPP *size_MatC];
- D27C.f[DIR_MMP ] = &DD27C[DIR_MMP *size_MatC];
- D27C.f[DIR_PMP ] = &DD27C[DIR_PMP *size_MatC];
- D27C.f[DIR_MPP ] = &DD27C[DIR_MPP *size_MatC];
- D27C.f[DIR_PPM ] = &DD27C[DIR_PPM *size_MatC];
- D27C.f[DIR_MMM ] = &DD27C[DIR_MMM *size_MatC];
- D27C.f[DIR_PMM ] = &DD27C[DIR_PMM *size_MatC];
- D27C.f[DIR_MPM ] = &DD27C[DIR_MPM *size_MatC];
+ D27C.f[DIR_P00] = &DD27C[DIR_P00 * numberOfLBnodesCoarse];
+ D27C.f[DIR_M00] = &DD27C[DIR_M00 * numberOfLBnodesCoarse];
+ D27C.f[DIR_0P0] = &DD27C[DIR_0P0 * numberOfLBnodesCoarse];
+ D27C.f[DIR_0M0] = &DD27C[DIR_0M0 * numberOfLBnodesCoarse];
+ D27C.f[DIR_00P] = &DD27C[DIR_00P * numberOfLBnodesCoarse];
+ D27C.f[DIR_00M] = &DD27C[DIR_00M * numberOfLBnodesCoarse];
+ D27C.f[DIR_PP0] = &DD27C[DIR_PP0 * numberOfLBnodesCoarse];
+ D27C.f[DIR_MM0] = &DD27C[DIR_MM0 * numberOfLBnodesCoarse];
+ D27C.f[DIR_PM0] = &DD27C[DIR_PM0 * numberOfLBnodesCoarse];
+ D27C.f[DIR_MP0] = &DD27C[DIR_MP0 * numberOfLBnodesCoarse];
+ D27C.f[DIR_P0P] = &DD27C[DIR_P0P * numberOfLBnodesCoarse];
+ D27C.f[DIR_M0M] = &DD27C[DIR_M0M * numberOfLBnodesCoarse];
+ D27C.f[DIR_P0M] = &DD27C[DIR_P0M * numberOfLBnodesCoarse];
+ D27C.f[DIR_M0P] = &DD27C[DIR_M0P * numberOfLBnodesCoarse];
+ D27C.f[DIR_0PP] = &DD27C[DIR_0PP * numberOfLBnodesCoarse];
+ D27C.f[DIR_0MM] = &DD27C[DIR_0MM * numberOfLBnodesCoarse];
+ D27C.f[DIR_0PM] = &DD27C[DIR_0PM * numberOfLBnodesCoarse];
+ D27C.f[DIR_0MP] = &DD27C[DIR_0MP * numberOfLBnodesCoarse];
+ D27C.f[DIR_000] = &DD27C[DIR_000 * numberOfLBnodesCoarse];
+ D27C.f[DIR_PPP] = &DD27C[DIR_PPP * numberOfLBnodesCoarse];
+ D27C.f[DIR_MMP] = &DD27C[DIR_MMP * numberOfLBnodesCoarse];
+ D27C.f[DIR_PMP] = &DD27C[DIR_PMP * numberOfLBnodesCoarse];
+ D27C.f[DIR_MPP] = &DD27C[DIR_MPP * numberOfLBnodesCoarse];
+ D27C.f[DIR_PPM] = &DD27C[DIR_PPM * numberOfLBnodesCoarse];
+ D27C.f[DIR_MMM] = &DD27C[DIR_MMM * numberOfLBnodesCoarse];
+ D27C.f[DIR_PMM] = &DD27C[DIR_PMM * numberOfLBnodesCoarse];
+ D27C.f[DIR_MPM] = &DD27C[DIR_MPM * numberOfLBnodesCoarse];
}
else
{
- D27C.f[DIR_M00 ] = &DD27C[DIR_P00 *size_MatC];
- D27C.f[DIR_P00 ] = &DD27C[DIR_M00 *size_MatC];
- D27C.f[DIR_0M0 ] = &DD27C[DIR_0P0 *size_MatC];
- D27C.f[DIR_0P0 ] = &DD27C[DIR_0M0 *size_MatC];
- D27C.f[DIR_00M ] = &DD27C[DIR_00P *size_MatC];
- D27C.f[DIR_00P ] = &DD27C[DIR_00M *size_MatC];
- D27C.f[DIR_MM0 ] = &DD27C[DIR_PP0 *size_MatC];
- D27C.f[DIR_PP0 ] = &DD27C[DIR_MM0 *size_MatC];
- D27C.f[DIR_MP0 ] = &DD27C[DIR_PM0 *size_MatC];
- D27C.f[DIR_PM0 ] = &DD27C[DIR_MP0 *size_MatC];
- D27C.f[DIR_M0M ] = &DD27C[DIR_P0P *size_MatC];
- D27C.f[DIR_P0P ] = &DD27C[DIR_M0M *size_MatC];
- D27C.f[DIR_M0P ] = &DD27C[DIR_P0M *size_MatC];
- D27C.f[DIR_P0M ] = &DD27C[DIR_M0P *size_MatC];
- D27C.f[DIR_0MM ] = &DD27C[DIR_0PP *size_MatC];
- D27C.f[DIR_0PP ] = &DD27C[DIR_0MM *size_MatC];
- D27C.f[DIR_0MP ] = &DD27C[DIR_0PM *size_MatC];
- D27C.f[DIR_0PM ] = &DD27C[DIR_0MP *size_MatC];
- D27C.f[DIR_000] = &DD27C[DIR_000*size_MatC];
- D27C.f[DIR_MMM ] = &DD27C[DIR_PPP *size_MatC];
- D27C.f[DIR_PPM ] = &DD27C[DIR_MMP *size_MatC];
- D27C.f[DIR_MPM ] = &DD27C[DIR_PMP *size_MatC];
- D27C.f[DIR_PMM ] = &DD27C[DIR_MPP *size_MatC];
- D27C.f[DIR_MMP ] = &DD27C[DIR_PPM *size_MatC];
- D27C.f[DIR_PPP ] = &DD27C[DIR_MMM *size_MatC];
- D27C.f[DIR_MPP ] = &DD27C[DIR_PMM *size_MatC];
- D27C.f[DIR_PMP ] = &DD27C[DIR_MPM *size_MatC];
+ D27C.f[DIR_M00] = &DD27C[DIR_P00 * numberOfLBnodesCoarse];
+ D27C.f[DIR_P00] = &DD27C[DIR_M00 * numberOfLBnodesCoarse];
+ D27C.f[DIR_0M0] = &DD27C[DIR_0P0 * numberOfLBnodesCoarse];
+ D27C.f[DIR_0P0] = &DD27C[DIR_0M0 * numberOfLBnodesCoarse];
+ D27C.f[DIR_00M] = &DD27C[DIR_00P * numberOfLBnodesCoarse];
+ D27C.f[DIR_00P] = &DD27C[DIR_00M * numberOfLBnodesCoarse];
+ D27C.f[DIR_MM0] = &DD27C[DIR_PP0 * numberOfLBnodesCoarse];
+ D27C.f[DIR_PP0] = &DD27C[DIR_MM0 * numberOfLBnodesCoarse];
+ D27C.f[DIR_MP0] = &DD27C[DIR_PM0 * numberOfLBnodesCoarse];
+ D27C.f[DIR_PM0] = &DD27C[DIR_MP0 * numberOfLBnodesCoarse];
+ D27C.f[DIR_M0M] = &DD27C[DIR_P0P * numberOfLBnodesCoarse];
+ D27C.f[DIR_P0P] = &DD27C[DIR_M0M * numberOfLBnodesCoarse];
+ D27C.f[DIR_M0P] = &DD27C[DIR_P0M * numberOfLBnodesCoarse];
+ D27C.f[DIR_P0M] = &DD27C[DIR_M0P * numberOfLBnodesCoarse];
+ D27C.f[DIR_0MM] = &DD27C[DIR_0PP * numberOfLBnodesCoarse];
+ D27C.f[DIR_0PP] = &DD27C[DIR_0MM * numberOfLBnodesCoarse];
+ D27C.f[DIR_0MP] = &DD27C[DIR_0PM * numberOfLBnodesCoarse];
+ D27C.f[DIR_0PM] = &DD27C[DIR_0MP * numberOfLBnodesCoarse];
+ D27C.f[DIR_000] = &DD27C[DIR_000 * numberOfLBnodesCoarse];
+ D27C.f[DIR_MMM] = &DD27C[DIR_PPP * numberOfLBnodesCoarse];
+ D27C.f[DIR_PPM] = &DD27C[DIR_MMP * numberOfLBnodesCoarse];
+ D27C.f[DIR_MPM] = &DD27C[DIR_PMP * numberOfLBnodesCoarse];
+ D27C.f[DIR_PMM] = &DD27C[DIR_MPP * numberOfLBnodesCoarse];
+ D27C.f[DIR_MMP] = &DD27C[DIR_PPM * numberOfLBnodesCoarse];
+ D27C.f[DIR_PPP] = &DD27C[DIR_MMM * numberOfLBnodesCoarse];
+ D27C.f[DIR_MPP] = &DD27C[DIR_PMM * numberOfLBnodesCoarse];
+ D27C.f[DIR_PMP] = &DD27C[DIR_MPM * numberOfLBnodesCoarse];
}
////////////////////////////////////////////////////////////////////////////////
@@ -45892,33 +45892,33 @@ __global__ void scaleCFThS27( real* DC,
f_BSE = fbseC[kbs];
f_BNW = fbnwC[kbw];
//////////////////////////////////////////////////////////////////////////////////
- f27E = (D27C.f[DIR_P00 ])[kzero];//ke
- f27W = (D27C.f[DIR_M00 ])[kw ];
- f27N = (D27C.f[DIR_0P0 ])[kzero];//kn
- f27S = (D27C.f[DIR_0M0 ])[ks ];
- f27T = (D27C.f[DIR_00P ])[kzero];//kt
- f27B = (D27C.f[DIR_00M ])[kb ];
- f27NE = (D27C.f[DIR_PP0 ])[kzero];//kne
- f27SW = (D27C.f[DIR_MM0 ])[ksw ];
- f27SE = (D27C.f[DIR_PM0 ])[ks ];//kse
- f27NW = (D27C.f[DIR_MP0 ])[kw ];//knw
- f27TE = (D27C.f[DIR_P0P ])[kzero];//kte
- f27BW = (D27C.f[DIR_M0M ])[kbw ];
- f27BE = (D27C.f[DIR_P0M ])[kb ];//kbe
- f27TW = (D27C.f[DIR_M0P ])[kw ];//ktw
- f27TN = (D27C.f[DIR_0PP ])[kzero];//ktn
- f27BS = (D27C.f[DIR_0MM ])[kbs ];
- f27BN = (D27C.f[DIR_0PM ])[kb ];//kbn
- f27TS = (D27C.f[DIR_0MP ])[ks ];//kts
+ f27E = (D27C.f[DIR_P00])[kzero];//ke
+ f27W = (D27C.f[DIR_M00])[kw ];
+ f27N = (D27C.f[DIR_0P0])[kzero];//kn
+ f27S = (D27C.f[DIR_0M0])[ks ];
+ f27T = (D27C.f[DIR_00P])[kzero];//kt
+ f27B = (D27C.f[DIR_00M])[kb ];
+ f27NE = (D27C.f[DIR_PP0])[kzero];//kne
+ f27SW = (D27C.f[DIR_MM0])[ksw ];
+ f27SE = (D27C.f[DIR_PM0])[ks ];//kse
+ f27NW = (D27C.f[DIR_MP0])[kw ];//knw
+ f27TE = (D27C.f[DIR_P0P])[kzero];//kte
+ f27BW = (D27C.f[DIR_M0M])[kbw ];
+ f27BE = (D27C.f[DIR_P0M])[kb ];//kbe
+ f27TW = (D27C.f[DIR_M0P])[kw ];//ktw
+ f27TN = (D27C.f[DIR_0PP])[kzero];//ktn
+ f27BS = (D27C.f[DIR_0MM])[kbs ];
+ f27BN = (D27C.f[DIR_0PM])[kb ];//kbn
+ f27TS = (D27C.f[DIR_0MP])[ks ];//kts
f27ZERO = (D27C.f[DIR_000])[kzero];//kzero
- f27TNE = (D27C.f[DIR_PPP ])[kzero];//ktne
- f27TSW = (D27C.f[DIR_MMP ])[ksw ];//ktsw
- f27TSE = (D27C.f[DIR_PMP ])[ks ];//ktse
- f27TNW = (D27C.f[DIR_MPP ])[kw ];//ktnw
- f27BNE = (D27C.f[DIR_PPM ])[kb ];//kbne
- f27BSW = (D27C.f[DIR_MMM ])[kbsw ];
- f27BSE = (D27C.f[DIR_PMM ])[kbs ];//kbse
- f27BNW = (D27C.f[DIR_MPM ])[kbw ];//kbnw
+ f27TNE = (D27C.f[DIR_PPP])[kzero];//ktne
+ f27TSW = (D27C.f[DIR_MMP])[ksw ];//ktsw
+ f27TSE = (D27C.f[DIR_PMP])[ks ];//ktse
+ f27TNW = (D27C.f[DIR_MPP])[kw ];//ktnw
+ f27BNE = (D27C.f[DIR_PPM])[kb ];//kbne
+ f27BSW = (D27C.f[DIR_MMM])[kbsw ];
+ f27BSE = (D27C.f[DIR_PMM])[kbs ];//kbse
+ f27BNW = (D27C.f[DIR_MPM])[kbw ];//kbnw
Conc_C_SWB = f27E + f27W + f27N + f27S + f27T + f27B + f27NE + f27SW + f27SE + f27NW +
f27TE + f27BW + f27BE + f27TW + f27TN + f27BS + f27BN + f27TS + f27ZERO +
@@ -45979,33 +45979,33 @@ __global__ void scaleCFThS27( real* DC,
f_BSE = fbseC[kbs];
f_BNW = fbnwC[kbw];
//////////////////////////////////////////////////////////////////////////////////
- f27E = (D27C.f[DIR_P00 ])[kzero];//ke
- f27W = (D27C.f[DIR_M00 ])[kw ];
- f27N = (D27C.f[DIR_0P0 ])[kzero];//kn
- f27S = (D27C.f[DIR_0M0 ])[ks ];
- f27T = (D27C.f[DIR_00P ])[kzero];//kt
- f27B = (D27C.f[DIR_00M ])[kb ];
- f27NE = (D27C.f[DIR_PP0 ])[kzero];//kne
- f27SW = (D27C.f[DIR_MM0 ])[ksw ];
- f27SE = (D27C.f[DIR_PM0 ])[ks ];//kse
- f27NW = (D27C.f[DIR_MP0 ])[kw ];//knw
- f27TE = (D27C.f[DIR_P0P ])[kzero];//kte
- f27BW = (D27C.f[DIR_M0M ])[kbw ];
- f27BE = (D27C.f[DIR_P0M ])[kb ];//kbe
- f27TW = (D27C.f[DIR_M0P ])[kw ];//ktw
- f27TN = (D27C.f[DIR_0PP ])[kzero];//ktn
- f27BS = (D27C.f[DIR_0MM ])[kbs ];
- f27BN = (D27C.f[DIR_0PM ])[kb ];//kbn
- f27TS = (D27C.f[DIR_0MP ])[ks ];//kts
+ f27E = (D27C.f[DIR_P00])[kzero];//ke
+ f27W = (D27C.f[DIR_M00])[kw ];
+ f27N = (D27C.f[DIR_0P0])[kzero];//kn
+ f27S = (D27C.f[DIR_0M0])[ks ];
+ f27T = (D27C.f[DIR_00P])[kzero];//kt
+ f27B = (D27C.f[DIR_00M])[kb ];
+ f27NE = (D27C.f[DIR_PP0])[kzero];//kne
+ f27SW = (D27C.f[DIR_MM0])[ksw ];
+ f27SE = (D27C.f[DIR_PM0])[ks ];//kse
+ f27NW = (D27C.f[DIR_MP0])[kw ];//knw
+ f27TE = (D27C.f[DIR_P0P])[kzero];//kte
+ f27BW = (D27C.f[DIR_M0M])[kbw ];
+ f27BE = (D27C.f[DIR_P0M])[kb ];//kbe
+ f27TW = (D27C.f[DIR_M0P])[kw ];//ktw
+ f27TN = (D27C.f[DIR_0PP])[kzero];//ktn
+ f27BS = (D27C.f[DIR_0MM])[kbs ];
+ f27BN = (D27C.f[DIR_0PM])[kb ];//kbn
+ f27TS = (D27C.f[DIR_0MP])[ks ];//kts
f27ZERO = (D27C.f[DIR_000])[kzero];//kzero
- f27TNE = (D27C.f[DIR_PPP ])[kzero];//ktne
- f27TSW = (D27C.f[DIR_MMP ])[ksw ];//ktsw
- f27TSE = (D27C.f[DIR_PMP ])[ks ];//ktse
- f27TNW = (D27C.f[DIR_MPP ])[kw ];//ktnw
- f27BNE = (D27C.f[DIR_PPM ])[kb ];//kbne
- f27BSW = (D27C.f[DIR_MMM ])[kbsw ];
- f27BSE = (D27C.f[DIR_PMM ])[kbs ];//kbse
- f27BNW = (D27C.f[DIR_MPM ])[kbw ];//kbnw
+ f27TNE = (D27C.f[DIR_PPP])[kzero];//ktne
+ f27TSW = (D27C.f[DIR_MMP])[ksw ];//ktsw
+ f27TSE = (D27C.f[DIR_PMP])[ks ];//ktse
+ f27TNW = (D27C.f[DIR_MPP])[kw ];//ktnw
+ f27BNE = (D27C.f[DIR_PPM])[kb ];//kbne
+ f27BSW = (D27C.f[DIR_MMM])[kbsw ];
+ f27BSE = (D27C.f[DIR_PMM])[kbs ];//kbse
+ f27BNW = (D27C.f[DIR_MPM])[kbw ];//kbnw
Conc_C_SWT = f27E + f27W + f27N + f27S + f27T + f27B + f27NE + f27SW + f27SE + f27NW +
f27TE + f27BW + f27BE + f27TW + f27TN + f27BS + f27BN + f27TS + f27ZERO +
@@ -46066,33 +46066,33 @@ __global__ void scaleCFThS27( real* DC,
f_BSE = fbseC[kbs];
f_BNW = fbnwC[kbw];
//////////////////////////////////////////////////////////////////////////////////
- f27E = (D27C.f[DIR_P00 ])[kzero];//ke
- f27W = (D27C.f[DIR_M00 ])[kw ];
- f27N = (D27C.f[DIR_0P0 ])[kzero];//kn
- f27S = (D27C.f[DIR_0M0 ])[ks ];
- f27T = (D27C.f[DIR_00P ])[kzero];//kt
- f27B = (D27C.f[DIR_00M ])[kb ];
- f27NE = (D27C.f[DIR_PP0 ])[kzero];//kne
- f27SW = (D27C.f[DIR_MM0 ])[ksw ];
- f27SE = (D27C.f[DIR_PM0 ])[ks ];//kse
- f27NW = (D27C.f[DIR_MP0 ])[kw ];//knw
- f27TE = (D27C.f[DIR_P0P ])[kzero];//kte
- f27BW = (D27C.f[DIR_M0M ])[kbw ];
- f27BE = (D27C.f[DIR_P0M ])[kb ];//kbe
- f27TW = (D27C.f[DIR_M0P ])[kw ];//ktw
- f27TN = (D27C.f[DIR_0PP ])[kzero];//ktn
- f27BS = (D27C.f[DIR_0MM ])[kbs ];
- f27BN = (D27C.f[DIR_0PM ])[kb ];//kbn
- f27TS = (D27C.f[DIR_0MP ])[ks ];//kts
+ f27E = (D27C.f[DIR_P00])[kzero];//ke
+ f27W = (D27C.f[DIR_M00])[kw ];
+ f27N = (D27C.f[DIR_0P0])[kzero];//kn
+ f27S = (D27C.f[DIR_0M0])[ks ];
+ f27T = (D27C.f[DIR_00P])[kzero];//kt
+ f27B = (D27C.f[DIR_00M])[kb ];
+ f27NE = (D27C.f[DIR_PP0])[kzero];//kne
+ f27SW = (D27C.f[DIR_MM0])[ksw ];
+ f27SE = (D27C.f[DIR_PM0])[ks ];//kse
+ f27NW = (D27C.f[DIR_MP0])[kw ];//knw
+ f27TE = (D27C.f[DIR_P0P])[kzero];//kte
+ f27BW = (D27C.f[DIR_M0M])[kbw ];
+ f27BE = (D27C.f[DIR_P0M])[kb ];//kbe
+ f27TW = (D27C.f[DIR_M0P])[kw ];//ktw
+ f27TN = (D27C.f[DIR_0PP])[kzero];//ktn
+ f27BS = (D27C.f[DIR_0MM])[kbs ];
+ f27BN = (D27C.f[DIR_0PM])[kb ];//kbn
+ f27TS = (D27C.f[DIR_0MP])[ks ];//kts
f27ZERO = (D27C.f[DIR_000])[kzero];//kzero
- f27TNE = (D27C.f[DIR_PPP ])[kzero];//ktne
- f27TSW = (D27C.f[DIR_MMP ])[ksw ];//ktsw
- f27TSE = (D27C.f[DIR_PMP ])[ks ];//ktse
- f27TNW = (D27C.f[DIR_MPP ])[kw ];//ktnw
- f27BNE = (D27C.f[DIR_PPM ])[kb ];//kbne
- f27BSW = (D27C.f[DIR_MMM ])[kbsw ];
- f27BSE = (D27C.f[DIR_PMM ])[kbs ];//kbse
- f27BNW = (D27C.f[DIR_MPM ])[kbw ];//kbnw
+ f27TNE = (D27C.f[DIR_PPP])[kzero];//ktne
+ f27TSW = (D27C.f[DIR_MMP])[ksw ];//ktsw
+ f27TSE = (D27C.f[DIR_PMP])[ks ];//ktse
+ f27TNW = (D27C.f[DIR_MPP])[kw ];//ktnw
+ f27BNE = (D27C.f[DIR_PPM])[kb ];//kbne
+ f27BSW = (D27C.f[DIR_MMM])[kbsw ];
+ f27BSE = (D27C.f[DIR_PMM])[kbs ];//kbse
+ f27BNW = (D27C.f[DIR_MPM])[kbw ];//kbnw
Conc_C_SET = f27E + f27W + f27N + f27S + f27T + f27B + f27NE + f27SW + f27SE + f27NW +
f27TE + f27BW + f27BE + f27TW + f27TN + f27BS + f27BN + f27TS + f27ZERO +
@@ -46153,33 +46153,33 @@ __global__ void scaleCFThS27( real* DC,
f_BSE = fbseC[kbs];
f_BNW = fbnwC[kbw];
//////////////////////////////////////////////////////////////////////////////////
- f27E = (D27C.f[DIR_P00 ])[kzero];//ke
- f27W = (D27C.f[DIR_M00 ])[kw ];
- f27N = (D27C.f[DIR_0P0 ])[kzero];//kn
- f27S = (D27C.f[DIR_0M0 ])[ks ];
- f27T = (D27C.f[DIR_00P ])[kzero];//kt
- f27B = (D27C.f[DIR_00M ])[kb ];
- f27NE = (D27C.f[DIR_PP0 ])[kzero];//kne
- f27SW = (D27C.f[DIR_MM0 ])[ksw ];
- f27SE = (D27C.f[DIR_PM0 ])[ks ];//kse
- f27NW = (D27C.f[DIR_MP0 ])[kw ];//knw
- f27TE = (D27C.f[DIR_P0P ])[kzero];//kte
- f27BW = (D27C.f[DIR_M0M ])[kbw ];
- f27BE = (D27C.f[DIR_P0M ])[kb ];//kbe
- f27TW = (D27C.f[DIR_M0P ])[kw ];//ktw
- f27TN = (D27C.f[DIR_0PP ])[kzero];//ktn
- f27BS = (D27C.f[DIR_0MM ])[kbs ];
- f27BN = (D27C.f[DIR_0PM ])[kb ];//kbn
- f27TS = (D27C.f[DIR_0MP ])[ks ];//kts
+ f27E = (D27C.f[DIR_P00])[kzero];//ke
+ f27W = (D27C.f[DIR_M00])[kw ];
+ f27N = (D27C.f[DIR_0P0])[kzero];//kn
+ f27S = (D27C.f[DIR_0M0])[ks ];
+ f27T = (D27C.f[DIR_00P])[kzero];//kt
+ f27B = (D27C.f[DIR_00M])[kb ];
+ f27NE = (D27C.f[DIR_PP0])[kzero];//kne
+ f27SW = (D27C.f[DIR_MM0])[ksw ];
+ f27SE = (D27C.f[DIR_PM0])[ks ];//kse
+ f27NW = (D27C.f[DIR_MP0])[kw ];//knw
+ f27TE = (D27C.f[DIR_P0P])[kzero];//kte
+ f27BW = (D27C.f[DIR_M0M])[kbw ];
+ f27BE = (D27C.f[DIR_P0M])[kb ];//kbe
+ f27TW = (D27C.f[DIR_M0P])[kw ];//ktw
+ f27TN = (D27C.f[DIR_0PP])[kzero];//ktn
+ f27BS = (D27C.f[DIR_0MM])[kbs ];
+ f27BN = (D27C.f[DIR_0PM])[kb ];//kbn
+ f27TS = (D27C.f[DIR_0MP])[ks ];//kts
f27ZERO = (D27C.f[DIR_000])[kzero];//kzero
- f27TNE = (D27C.f[DIR_PPP ])[kzero];//ktne
- f27TSW = (D27C.f[DIR_MMP ])[ksw ];//ktsw
- f27TSE = (D27C.f[DIR_PMP ])[ks ];//ktse
- f27TNW = (D27C.f[DIR_MPP ])[kw ];//ktnw
- f27BNE = (D27C.f[DIR_PPM ])[kb ];//kbne
- f27BSW = (D27C.f[DIR_MMM ])[kbsw ];
- f27BSE = (D27C.f[DIR_PMM ])[kbs ];//kbse
- f27BNW = (D27C.f[DIR_MPM ])[kbw ];//kbnw
+ f27TNE = (D27C.f[DIR_PPP])[kzero];//ktne
+ f27TSW = (D27C.f[DIR_MMP])[ksw ];//ktsw
+ f27TSE = (D27C.f[DIR_PMP])[ks ];//ktse
+ f27TNW = (D27C.f[DIR_MPP])[kw ];//ktnw
+ f27BNE = (D27C.f[DIR_PPM])[kb ];//kbne
+ f27BSW = (D27C.f[DIR_MMM])[kbsw ];
+ f27BSE = (D27C.f[DIR_PMM])[kbs ];//kbse
+ f27BNW = (D27C.f[DIR_MPM])[kbw ];//kbnw
Conc_C_SEB = f27E + f27W + f27N + f27S + f27T + f27B + f27NE + f27SW + f27SE + f27NW +
f27TE + f27BW + f27BE + f27TW + f27TN + f27BS + f27BN + f27TS + f27ZERO +
@@ -46250,33 +46250,33 @@ __global__ void scaleCFThS27( real* DC,
f_BSE = fbseC[kbs];
f_BNW = fbnwC[kbw];
//////////////////////////////////////////////////////////////////////////////////
- f27E = (D27C.f[DIR_P00 ])[kzero];//ke
- f27W = (D27C.f[DIR_M00 ])[kw ];
- f27N = (D27C.f[DIR_0P0 ])[kzero];//kn
- f27S = (D27C.f[DIR_0M0 ])[ks ];
- f27T = (D27C.f[DIR_00P ])[kzero];//kt
- f27B = (D27C.f[DIR_00M ])[kb ];
- f27NE = (D27C.f[DIR_PP0 ])[kzero];//kne
- f27SW = (D27C.f[DIR_MM0 ])[ksw ];
- f27SE = (D27C.f[DIR_PM0 ])[ks ];//kse
- f27NW = (D27C.f[DIR_MP0 ])[kw ];//knw
- f27TE = (D27C.f[DIR_P0P ])[kzero];//kte
- f27BW = (D27C.f[DIR_M0M ])[kbw ];
- f27BE = (D27C.f[DIR_P0M ])[kb ];//kbe
- f27TW = (D27C.f[DIR_M0P ])[kw ];//ktw
- f27TN = (D27C.f[DIR_0PP ])[kzero];//ktn
- f27BS = (D27C.f[DIR_0MM ])[kbs ];
- f27BN = (D27C.f[DIR_0PM ])[kb ];//kbn
- f27TS = (D27C.f[DIR_0MP ])[ks ];//kts
+ f27E = (D27C.f[DIR_P00])[kzero];//ke
+ f27W = (D27C.f[DIR_M00])[kw ];
+ f27N = (D27C.f[DIR_0P0])[kzero];//kn
+ f27S = (D27C.f[DIR_0M0])[ks ];
+ f27T = (D27C.f[DIR_00P])[kzero];//kt
+ f27B = (D27C.f[DIR_00M])[kb ];
+ f27NE = (D27C.f[DIR_PP0])[kzero];//kne
+ f27SW = (D27C.f[DIR_MM0])[ksw ];
+ f27SE = (D27C.f[DIR_PM0])[ks ];//kse
+ f27NW = (D27C.f[DIR_MP0])[kw ];//knw
+ f27TE = (D27C.f[DIR_P0P])[kzero];//kte
+ f27BW = (D27C.f[DIR_M0M])[kbw ];
+ f27BE = (D27C.f[DIR_P0M])[kb ];//kbe
+ f27TW = (D27C.f[DIR_M0P])[kw ];//ktw
+ f27TN = (D27C.f[DIR_0PP])[kzero];//ktn
+ f27BS = (D27C.f[DIR_0MM])[kbs ];
+ f27BN = (D27C.f[DIR_0PM])[kb ];//kbn
+ f27TS = (D27C.f[DIR_0MP])[ks ];//kts
f27ZERO = (D27C.f[DIR_000])[kzero];//kzero
- f27TNE = (D27C.f[DIR_PPP ])[kzero];//ktne
- f27TSW = (D27C.f[DIR_MMP ])[ksw ];//ktsw
- f27TSE = (D27C.f[DIR_PMP ])[ks ];//ktse
- f27TNW = (D27C.f[DIR_MPP ])[kw ];//ktnw
- f27BNE = (D27C.f[DIR_PPM ])[kb ];//kbne
- f27BSW = (D27C.f[DIR_MMM ])[kbsw ];
- f27BSE = (D27C.f[DIR_PMM ])[kbs ];//kbse
- f27BNW = (D27C.f[DIR_MPM ])[kbw ];//kbnw
+ f27TNE = (D27C.f[DIR_PPP])[kzero];//ktne
+ f27TSW = (D27C.f[DIR_MMP])[ksw ];//ktsw
+ f27TSE = (D27C.f[DIR_PMP])[ks ];//ktse
+ f27TNW = (D27C.f[DIR_MPP])[kw ];//ktnw
+ f27BNE = (D27C.f[DIR_PPM])[kb ];//kbne
+ f27BSW = (D27C.f[DIR_MMM])[kbsw ];
+ f27BSE = (D27C.f[DIR_PMM])[kbs ];//kbse
+ f27BNW = (D27C.f[DIR_MPM])[kbw ];//kbnw
Conc_C_NWB = f27E + f27W + f27N + f27S + f27T + f27B + f27NE + f27SW + f27SE + f27NW +
f27TE + f27BW + f27BE + f27TW + f27TN + f27BS + f27BN + f27TS + f27ZERO +
@@ -46337,33 +46337,33 @@ __global__ void scaleCFThS27( real* DC,
f_BSE = fbseC[kbs];
f_BNW = fbnwC[kbw];
//////////////////////////////////////////////////////////////////////////////////
- f27E = (D27C.f[DIR_P00 ])[kzero];//ke
- f27W = (D27C.f[DIR_M00 ])[kw ];
- f27N = (D27C.f[DIR_0P0 ])[kzero];//kn
- f27S = (D27C.f[DIR_0M0 ])[ks ];
- f27T = (D27C.f[DIR_00P ])[kzero];//kt
- f27B = (D27C.f[DIR_00M ])[kb ];
- f27NE = (D27C.f[DIR_PP0 ])[kzero];//kne
- f27SW = (D27C.f[DIR_MM0 ])[ksw ];
- f27SE = (D27C.f[DIR_PM0 ])[ks ];//kse
- f27NW = (D27C.f[DIR_MP0 ])[kw ];//knw
- f27TE = (D27C.f[DIR_P0P ])[kzero];//kte
- f27BW = (D27C.f[DIR_M0M ])[kbw ];
- f27BE = (D27C.f[DIR_P0M ])[kb ];//kbe
- f27TW = (D27C.f[DIR_M0P ])[kw ];//ktw
- f27TN = (D27C.f[DIR_0PP ])[kzero];//ktn
- f27BS = (D27C.f[DIR_0MM ])[kbs ];
- f27BN = (D27C.f[DIR_0PM ])[kb ];//kbn
- f27TS = (D27C.f[DIR_0MP ])[ks ];//kts
+ f27E = (D27C.f[DIR_P00])[kzero];//ke
+ f27W = (D27C.f[DIR_M00])[kw ];
+ f27N = (D27C.f[DIR_0P0])[kzero];//kn
+ f27S = (D27C.f[DIR_0M0])[ks ];
+ f27T = (D27C.f[DIR_00P])[kzero];//kt
+ f27B = (D27C.f[DIR_00M])[kb ];
+ f27NE = (D27C.f[DIR_PP0])[kzero];//kne
+ f27SW = (D27C.f[DIR_MM0])[ksw ];
+ f27SE = (D27C.f[DIR_PM0])[ks ];//kse
+ f27NW = (D27C.f[DIR_MP0])[kw ];//knw
+ f27TE = (D27C.f[DIR_P0P])[kzero];//kte
+ f27BW = (D27C.f[DIR_M0M])[kbw ];
+ f27BE = (D27C.f[DIR_P0M])[kb ];//kbe
+ f27TW = (D27C.f[DIR_M0P])[kw ];//ktw
+ f27TN = (D27C.f[DIR_0PP])[kzero];//ktn
+ f27BS = (D27C.f[DIR_0MM])[kbs ];
+ f27BN = (D27C.f[DIR_0PM])[kb ];//kbn
+ f27TS = (D27C.f[DIR_0MP])[ks ];//kts
f27ZERO = (D27C.f[DIR_000])[kzero];//kzero
- f27TNE = (D27C.f[DIR_PPP ])[kzero];//ktne
- f27TSW = (D27C.f[DIR_MMP ])[ksw ];//ktsw
- f27TSE = (D27C.f[DIR_PMP ])[ks ];//ktse
- f27TNW = (D27C.f[DIR_MPP ])[kw ];//ktnw
- f27BNE = (D27C.f[DIR_PPM ])[kb ];//kbne
- f27BSW = (D27C.f[DIR_MMM ])[kbsw ];
- f27BSE = (D27C.f[DIR_PMM ])[kbs ];//kbse
- f27BNW = (D27C.f[DIR_MPM ])[kbw ];//kbnw
+ f27TNE = (D27C.f[DIR_PPP])[kzero];//ktne
+ f27TSW = (D27C.f[DIR_MMP])[ksw ];//ktsw
+ f27TSE = (D27C.f[DIR_PMP])[ks ];//ktse
+ f27TNW = (D27C.f[DIR_MPP])[kw ];//ktnw
+ f27BNE = (D27C.f[DIR_PPM])[kb ];//kbne
+ f27BSW = (D27C.f[DIR_MMM])[kbsw ];
+ f27BSE = (D27C.f[DIR_PMM])[kbs ];//kbse
+ f27BNW = (D27C.f[DIR_MPM])[kbw ];//kbnw
Conc_C_NWT = f27E + f27W + f27N + f27S + f27T + f27B + f27NE + f27SW + f27SE + f27NW +
f27TE + f27BW + f27BE + f27TW + f27TN + f27BS + f27BN + f27TS + f27ZERO +
@@ -46424,33 +46424,33 @@ __global__ void scaleCFThS27( real* DC,
f_BSE = fbseC[kbs];
f_BNW = fbnwC[kbw];
//////////////////////////////////////////////////////////////////////////////////
- f27E = (D27C.f[DIR_P00 ])[kzero];//ke
- f27W = (D27C.f[DIR_M00 ])[kw ];
- f27N = (D27C.f[DIR_0P0 ])[kzero];//kn
- f27S = (D27C.f[DIR_0M0 ])[ks ];
- f27T = (D27C.f[DIR_00P ])[kzero];//kt
- f27B = (D27C.f[DIR_00M ])[kb ];
- f27NE = (D27C.f[DIR_PP0 ])[kzero];//kne
- f27SW = (D27C.f[DIR_MM0 ])[ksw ];
- f27SE = (D27C.f[DIR_PM0 ])[ks ];//kse
- f27NW = (D27C.f[DIR_MP0 ])[kw ];//knw
- f27TE = (D27C.f[DIR_P0P ])[kzero];//kte
- f27BW = (D27C.f[DIR_M0M ])[kbw ];
- f27BE = (D27C.f[DIR_P0M ])[kb ];//kbe
- f27TW = (D27C.f[DIR_M0P ])[kw ];//ktw
- f27TN = (D27C.f[DIR_0PP ])[kzero];//ktn
- f27BS = (D27C.f[DIR_0MM ])[kbs ];
- f27BN = (D27C.f[DIR_0PM ])[kb ];//kbn
- f27TS = (D27C.f[DIR_0MP ])[ks ];//kts
+ f27E = (D27C.f[DIR_P00])[kzero];//ke
+ f27W = (D27C.f[DIR_M00])[kw ];
+ f27N = (D27C.f[DIR_0P0])[kzero];//kn
+ f27S = (D27C.f[DIR_0M0])[ks ];
+ f27T = (D27C.f[DIR_00P])[kzero];//kt
+ f27B = (D27C.f[DIR_00M])[kb ];
+ f27NE = (D27C.f[DIR_PP0])[kzero];//kne
+ f27SW = (D27C.f[DIR_MM0])[ksw ];
+ f27SE = (D27C.f[DIR_PM0])[ks ];//kse
+ f27NW = (D27C.f[DIR_MP0])[kw ];//knw
+ f27TE = (D27C.f[DIR_P0P])[kzero];//kte
+ f27BW = (D27C.f[DIR_M0M])[kbw ];
+ f27BE = (D27C.f[DIR_P0M])[kb ];//kbe
+ f27TW = (D27C.f[DIR_M0P])[kw ];//ktw
+ f27TN = (D27C.f[DIR_0PP])[kzero];//ktn
+ f27BS = (D27C.f[DIR_0MM])[kbs ];
+ f27BN = (D27C.f[DIR_0PM])[kb ];//kbn
+ f27TS = (D27C.f[DIR_0MP])[ks ];//kts
f27ZERO = (D27C.f[DIR_000])[kzero];//kzero
- f27TNE = (D27C.f[DIR_PPP ])[kzero];//ktne
- f27TSW = (D27C.f[DIR_MMP ])[ksw ];//ktsw
- f27TSE = (D27C.f[DIR_PMP ])[ks ];//ktse
- f27TNW = (D27C.f[DIR_MPP ])[kw ];//ktnw
- f27BNE = (D27C.f[DIR_PPM ])[kb ];//kbne
- f27BSW = (D27C.f[DIR_MMM ])[kbsw ];
- f27BSE = (D27C.f[DIR_PMM ])[kbs ];//kbse
- f27BNW = (D27C.f[DIR_MPM ])[kbw ];//kbnw
+ f27TNE = (D27C.f[DIR_PPP])[kzero];//ktne
+ f27TSW = (D27C.f[DIR_MMP])[ksw ];//ktsw
+ f27TSE = (D27C.f[DIR_PMP])[ks ];//ktse
+ f27TNW = (D27C.f[DIR_MPP])[kw ];//ktnw
+ f27BNE = (D27C.f[DIR_PPM])[kb ];//kbne
+ f27BSW = (D27C.f[DIR_MMM])[kbsw ];
+ f27BSE = (D27C.f[DIR_PMM])[kbs ];//kbse
+ f27BNW = (D27C.f[DIR_MPM])[kbw ];//kbnw
Conc_C_NET = f27E + f27W + f27N + f27S + f27T + f27B + f27NE + f27SW + f27SE + f27NW +
f27TE + f27BW + f27BE + f27TW + f27TN + f27BS + f27BN + f27TS + f27ZERO +
@@ -46511,33 +46511,33 @@ __global__ void scaleCFThS27( real* DC,
f_BSE = fbseC[kbs];
f_BNW = fbnwC[kbw];
////////////////////////////////////////////////////////////////////////////////
- f27E = (D27C.f[DIR_P00 ])[kzero];//ke
- f27W = (D27C.f[DIR_M00 ])[kw ];
- f27N = (D27C.f[DIR_0P0 ])[kzero];//kn
- f27S = (D27C.f[DIR_0M0 ])[ks ];
- f27T = (D27C.f[DIR_00P ])[kzero];//kt
- f27B = (D27C.f[DIR_00M ])[kb ];
- f27NE = (D27C.f[DIR_PP0 ])[kzero];//kne
- f27SW = (D27C.f[DIR_MM0 ])[ksw ];
- f27SE = (D27C.f[DIR_PM0 ])[ks ];//kse
- f27NW = (D27C.f[DIR_MP0 ])[kw ];//knw
- f27TE = (D27C.f[DIR_P0P ])[kzero];//kte
- f27BW = (D27C.f[DIR_M0M ])[kbw ];
- f27BE = (D27C.f[DIR_P0M ])[kb ];//kbe
- f27TW = (D27C.f[DIR_M0P ])[kw ];//ktw
- f27TN = (D27C.f[DIR_0PP ])[kzero];//ktn
- f27BS = (D27C.f[DIR_0MM ])[kbs ];
- f27BN = (D27C.f[DIR_0PM ])[kb ];//kbn
- f27TS = (D27C.f[DIR_0MP ])[ks ];//kts
+ f27E = (D27C.f[DIR_P00])[kzero];//ke
+ f27W = (D27C.f[DIR_M00])[kw ];
+ f27N = (D27C.f[DIR_0P0])[kzero];//kn
+ f27S = (D27C.f[DIR_0M0])[ks ];
+ f27T = (D27C.f[DIR_00P])[kzero];//kt
+ f27B = (D27C.f[DIR_00M])[kb ];
+ f27NE = (D27C.f[DIR_PP0])[kzero];//kne
+ f27SW = (D27C.f[DIR_MM0])[ksw ];
+ f27SE = (D27C.f[DIR_PM0])[ks ];//kse
+ f27NW = (D27C.f[DIR_MP0])[kw ];//knw
+ f27TE = (D27C.f[DIR_P0P])[kzero];//kte
+ f27BW = (D27C.f[DIR_M0M])[kbw ];
+ f27BE = (D27C.f[DIR_P0M])[kb ];//kbe
+ f27TW = (D27C.f[DIR_M0P])[kw ];//ktw
+ f27TN = (D27C.f[DIR_0PP])[kzero];//ktn
+ f27BS = (D27C.f[DIR_0MM])[kbs ];
+ f27BN = (D27C.f[DIR_0PM])[kb ];//kbn
+ f27TS = (D27C.f[DIR_0MP])[ks ];//kts
f27ZERO = (D27C.f[DIR_000])[kzero];//kzero
- f27TNE = (D27C.f[DIR_PPP ])[kzero];//ktne
- f27TSW = (D27C.f[DIR_MMP ])[ksw ];//ktsw
- f27TSE = (D27C.f[DIR_PMP ])[ks ];//ktse
- f27TNW = (D27C.f[DIR_MPP ])[kw ];//ktnw
- f27BNE = (D27C.f[DIR_PPM ])[kb ];//kbne
- f27BSW = (D27C.f[DIR_MMM ])[kbsw ];
- f27BSE = (D27C.f[DIR_PMM ])[kbs ];//kbse
- f27BNW = (D27C.f[DIR_MPM ])[kbw ];//kbnw
+ f27TNE = (D27C.f[DIR_PPP])[kzero];//ktne
+ f27TSW = (D27C.f[DIR_MMP])[ksw ];//ktsw
+ f27TSE = (D27C.f[DIR_PMP])[ks ];//ktse
+ f27TNW = (D27C.f[DIR_MPP])[kw ];//ktnw
+ f27BNE = (D27C.f[DIR_PPM])[kb ];//kbne
+ f27BSW = (D27C.f[DIR_MMM])[kbsw ];
+ f27BSE = (D27C.f[DIR_PMM])[kbs ];//kbse
+ f27BNW = (D27C.f[DIR_MPM])[kbw ];//kbnw
Conc_C_NEB = f27E + f27W + f27N + f27S + f27T + f27B + f27NE + f27SW + f27SE + f27NW +
f27TE + f27BW + f27BE + f27TW + f27TN + f27BS + f27BN + f27TS + f27ZERO +
@@ -46656,32 +46656,32 @@ __global__ void scaleCFThS27( real* DC,
cu_sq=c3o2*(vx1*vx1+vx2*vx2+vx3*vx3);
(D27F.f[DIR_000])[kzero] = c8o27* Conc_F*(c1o1-cu_sq);
- (D27F.f[DIR_P00 ])[kzero] = c2o27* (c3o1*( Mx )+Conc_F*(c1o1+c9o2*( vx1 )*( vx1 )-cu_sq));
- (D27F.f[DIR_M00 ])[kw ] = c2o27* (c3o1*(-Mx )+Conc_F*(c1o1+c9o2*(-vx1 )*(-vx1 )-cu_sq));
- (D27F.f[DIR_0P0 ])[kzero] = c2o27* (c3o1*( My )+Conc_F*(c1o1+c9o2*( vx2 )*( vx2 )-cu_sq));
- (D27F.f[DIR_0M0 ])[ks ] = c2o27* (c3o1*( -My )+Conc_F*(c1o1+c9o2*( -vx2 )*( -vx2 )-cu_sq));
- (D27F.f[DIR_00P ])[kzero] = c2o27* (c3o1*( Mz)+Conc_F*(c1o1+c9o2*( vx3)*( vx3)-cu_sq));
- (D27F.f[DIR_00M ])[kb ] = c2o27* (c3o1*( -Mz)+Conc_F*(c1o1+c9o2*( -vx3)*( -vx3)-cu_sq));
- (D27F.f[DIR_PP0 ])[kzero] = c1o54* (c3o1*( Mx +My )+Conc_F*(c1o1+c9o2*( vx1+vx2 )*( vx1+vx2 )-cu_sq));
- (D27F.f[DIR_MM0 ])[ksw ] = c1o54* (c3o1*(-Mx -My )+Conc_F*(c1o1+c9o2*(-vx1-vx2 )*(-vx1-vx2 )-cu_sq));
- (D27F.f[DIR_PM0 ])[ks ] = c1o54* (c3o1*( Mx -My )+Conc_F*(c1o1+c9o2*( vx1-vx2 )*( vx1-vx2 )-cu_sq));
- (D27F.f[DIR_MP0 ])[kw ] = c1o54* (c3o1*(-Mx +My )+Conc_F*(c1o1+c9o2*(-vx1+vx2 )*(-vx1+vx2 )-cu_sq));
- (D27F.f[DIR_P0P ])[kzero] = c1o54* (c3o1*( Mx +Mz)+Conc_F*(c1o1+c9o2*( vx1 +vx3)*( vx1 +vx3)-cu_sq));
- (D27F.f[DIR_M0M ])[kbw ] = c1o54* (c3o1*(-Mx -Mz)+Conc_F*(c1o1+c9o2*(-vx1 -vx3)*(-vx1 -vx3)-cu_sq));
- (D27F.f[DIR_P0M ])[kb ] = c1o54* (c3o1*( Mx -Mz)+Conc_F*(c1o1+c9o2*( vx1 -vx3)*( vx1 -vx3)-cu_sq));
- (D27F.f[DIR_M0P ])[kw ] = c1o54* (c3o1*(-Mx +Mz)+Conc_F*(c1o1+c9o2*(-vx1 +vx3)*(-vx1 +vx3)-cu_sq));
- (D27F.f[DIR_0PP ])[kzero] = c1o54* (c3o1*( My +Mz)+Conc_F*(c1o1+c9o2*( vx2+vx3)*( vx2+vx3)-cu_sq));
- (D27F.f[DIR_0MM ])[kbs ] = c1o54* (c3o1*( -My -Mz)+Conc_F*(c1o1+c9o2*( -vx2-vx3)*( -vx2-vx3)-cu_sq));
- (D27F.f[DIR_0PM ])[kb ] = c1o54* (c3o1*( My -Mz)+Conc_F*(c1o1+c9o2*( vx2-vx3)*( vx2-vx3)-cu_sq));
- (D27F.f[DIR_0MP ])[ks ] = c1o54* (c3o1*( -My +Mz)+Conc_F*(c1o1+c9o2*( -vx2+vx3)*( -vx2+vx3)-cu_sq));
- (D27F.f[DIR_PPP ])[kzero] = c1o216*(c3o1*( Mx +My +Mz)+Conc_F*(c1o1+c9o2*( vx1+vx2+vx3)*( vx1+vx2+vx3)-cu_sq));
- (D27F.f[DIR_MMM ])[kbsw ] = c1o216*(c3o1*(-Mx -My -Mz)+Conc_F*(c1o1+c9o2*(-vx1-vx2-vx3)*(-vx1-vx2-vx3)-cu_sq));
- (D27F.f[DIR_PPM ])[kb ] = c1o216*(c3o1*( Mx +My -Mz)+Conc_F*(c1o1+c9o2*( vx1+vx2-vx3)*( vx1+vx2-vx3)-cu_sq));
- (D27F.f[DIR_MMP ])[ksw ] = c1o216*(c3o1*(-Mx -My +Mz)+Conc_F*(c1o1+c9o2*(-vx1-vx2+vx3)*(-vx1-vx2+vx3)-cu_sq));
- (D27F.f[DIR_PMP ])[ks ] = c1o216*(c3o1*( Mx -My +Mz)+Conc_F*(c1o1+c9o2*( vx1-vx2+vx3)*( vx1-vx2+vx3)-cu_sq));
- (D27F.f[DIR_MPM ])[kbw ] = c1o216*(c3o1*(-Mx +My -Mz)+Conc_F*(c1o1+c9o2*(-vx1+vx2-vx3)*(-vx1+vx2-vx3)-cu_sq));
- (D27F.f[DIR_PMM ])[kbs ] = c1o216*(c3o1*( Mx -My -Mz)+Conc_F*(c1o1+c9o2*( vx1-vx2-vx3)*( vx1-vx2-vx3)-cu_sq));
- (D27F.f[DIR_MPP ])[kw ] = c1o216*(c3o1*(-Mx +My +Mz)+Conc_F*(c1o1+c9o2*(-vx1+vx2+vx3)*(-vx1+vx2+vx3)-cu_sq));
+ (D27F.f[DIR_P00])[kzero] = c2o27* (c3o1*( Mx )+Conc_F*(c1o1+c9o2*( vx1 )*( vx1 )-cu_sq));
+ (D27F.f[DIR_M00])[kw ] = c2o27* (c3o1*(-Mx )+Conc_F*(c1o1+c9o2*(-vx1 )*(-vx1 )-cu_sq));
+ (D27F.f[DIR_0P0])[kzero] = c2o27* (c3o1*( My )+Conc_F*(c1o1+c9o2*( vx2 )*( vx2 )-cu_sq));
+ (D27F.f[DIR_0M0])[ks ] = c2o27* (c3o1*( -My )+Conc_F*(c1o1+c9o2*( -vx2 )*( -vx2 )-cu_sq));
+ (D27F.f[DIR_00P])[kzero] = c2o27* (c3o1*( Mz)+Conc_F*(c1o1+c9o2*( vx3)*( vx3)-cu_sq));
+ (D27F.f[DIR_00M])[kb ] = c2o27* (c3o1*( -Mz)+Conc_F*(c1o1+c9o2*( -vx3)*( -vx3)-cu_sq));
+ (D27F.f[DIR_PP0])[kzero] = c1o54* (c3o1*( Mx +My )+Conc_F*(c1o1+c9o2*( vx1+vx2 )*( vx1+vx2 )-cu_sq));
+ (D27F.f[DIR_MM0])[ksw ] = c1o54* (c3o1*(-Mx -My )+Conc_F*(c1o1+c9o2*(-vx1-vx2 )*(-vx1-vx2 )-cu_sq));
+ (D27F.f[DIR_PM0])[ks ] = c1o54* (c3o1*( Mx -My )+Conc_F*(c1o1+c9o2*( vx1-vx2 )*( vx1-vx2 )-cu_sq));
+ (D27F.f[DIR_MP0])[kw ] = c1o54* (c3o1*(-Mx +My )+Conc_F*(c1o1+c9o2*(-vx1+vx2 )*(-vx1+vx2 )-cu_sq));
+ (D27F.f[DIR_P0P])[kzero] = c1o54* (c3o1*( Mx +Mz)+Conc_F*(c1o1+c9o2*( vx1 +vx3)*( vx1 +vx3)-cu_sq));
+ (D27F.f[DIR_M0M])[kbw ] = c1o54* (c3o1*(-Mx -Mz)+Conc_F*(c1o1+c9o2*(-vx1 -vx3)*(-vx1 -vx3)-cu_sq));
+ (D27F.f[DIR_P0M])[kb ] = c1o54* (c3o1*( Mx -Mz)+Conc_F*(c1o1+c9o2*( vx1 -vx3)*( vx1 -vx3)-cu_sq));
+ (D27F.f[DIR_M0P])[kw ] = c1o54* (c3o1*(-Mx +Mz)+Conc_F*(c1o1+c9o2*(-vx1 +vx3)*(-vx1 +vx3)-cu_sq));
+ (D27F.f[DIR_0PP])[kzero] = c1o54* (c3o1*( My +Mz)+Conc_F*(c1o1+c9o2*( vx2+vx3)*( vx2+vx3)-cu_sq));
+ (D27F.f[DIR_0MM])[kbs ] = c1o54* (c3o1*( -My -Mz)+Conc_F*(c1o1+c9o2*( -vx2-vx3)*( -vx2-vx3)-cu_sq));
+ (D27F.f[DIR_0PM])[kb ] = c1o54* (c3o1*( My -Mz)+Conc_F*(c1o1+c9o2*( vx2-vx3)*( vx2-vx3)-cu_sq));
+ (D27F.f[DIR_0MP])[ks ] = c1o54* (c3o1*( -My +Mz)+Conc_F*(c1o1+c9o2*( -vx2+vx3)*( -vx2+vx3)-cu_sq));
+ (D27F.f[DIR_PPP])[kzero] = c1o216*(c3o1*( Mx +My +Mz)+Conc_F*(c1o1+c9o2*( vx1+vx2+vx3)*( vx1+vx2+vx3)-cu_sq));
+ (D27F.f[DIR_MMM])[kbsw ] = c1o216*(c3o1*(-Mx -My -Mz)+Conc_F*(c1o1+c9o2*(-vx1-vx2-vx3)*(-vx1-vx2-vx3)-cu_sq));
+ (D27F.f[DIR_PPM])[kb ] = c1o216*(c3o1*( Mx +My -Mz)+Conc_F*(c1o1+c9o2*( vx1+vx2-vx3)*( vx1+vx2-vx3)-cu_sq));
+ (D27F.f[DIR_MMP])[ksw ] = c1o216*(c3o1*(-Mx -My +Mz)+Conc_F*(c1o1+c9o2*(-vx1-vx2+vx3)*(-vx1-vx2+vx3)-cu_sq));
+ (D27F.f[DIR_PMP])[ks ] = c1o216*(c3o1*( Mx -My +Mz)+Conc_F*(c1o1+c9o2*( vx1-vx2+vx3)*( vx1-vx2+vx3)-cu_sq));
+ (D27F.f[DIR_MPM])[kbw ] = c1o216*(c3o1*(-Mx +My -Mz)+Conc_F*(c1o1+c9o2*(-vx1+vx2-vx3)*(-vx1+vx2-vx3)-cu_sq));
+ (D27F.f[DIR_PMM])[kbs ] = c1o216*(c3o1*( Mx -My -Mz)+Conc_F*(c1o1+c9o2*( vx1-vx2-vx3)*( vx1-vx2-vx3)-cu_sq));
+ (D27F.f[DIR_MPP])[kw ] = c1o216*(c3o1*(-Mx +My +Mz)+Conc_F*(c1o1+c9o2*(-vx1+vx2+vx3)*(-vx1+vx2+vx3)-cu_sq));
@@ -46734,32 +46734,32 @@ __global__ void scaleCFThS27( real* DC,
cu_sq=c3o2*(vx1*vx1+vx2*vx2+vx3*vx3);
(D27F.f[DIR_000])[kzero] = c8o27* Conc_F*(c1o1-cu_sq);
- (D27F.f[DIR_P00 ])[kzero] = c2o27* (c3o1*( Mx )+Conc_F*(c1o1+c9o2*( vx1 )*( vx1 )-cu_sq));
- (D27F.f[DIR_M00 ])[kw ] = c2o27* (c3o1*(-Mx )+Conc_F*(c1o1+c9o2*(-vx1 )*(-vx1 )-cu_sq));
- (D27F.f[DIR_0P0 ])[kzero] = c2o27* (c3o1*( My )+Conc_F*(c1o1+c9o2*( vx2 )*( vx2 )-cu_sq));
- (D27F.f[DIR_0M0 ])[ks ] = c2o27* (c3o1*( -My )+Conc_F*(c1o1+c9o2*( -vx2 )*( -vx2 )-cu_sq));
- (D27F.f[DIR_00P ])[kzero] = c2o27* (c3o1*( Mz)+Conc_F*(c1o1+c9o2*( vx3)*( vx3)-cu_sq));
- (D27F.f[DIR_00M ])[kb ] = c2o27* (c3o1*( -Mz)+Conc_F*(c1o1+c9o2*( -vx3)*( -vx3)-cu_sq));
- (D27F.f[DIR_PP0 ])[kzero] = c1o54* (c3o1*( Mx +My )+Conc_F*(c1o1+c9o2*( vx1+vx2 )*( vx1+vx2 )-cu_sq));
- (D27F.f[DIR_MM0 ])[ksw ] = c1o54* (c3o1*(-Mx -My )+Conc_F*(c1o1+c9o2*(-vx1-vx2 )*(-vx1-vx2 )-cu_sq));
- (D27F.f[DIR_PM0 ])[ks ] = c1o54* (c3o1*( Mx -My )+Conc_F*(c1o1+c9o2*( vx1-vx2 )*( vx1-vx2 )-cu_sq));
- (D27F.f[DIR_MP0 ])[kw ] = c1o54* (c3o1*(-Mx +My )+Conc_F*(c1o1+c9o2*(-vx1+vx2 )*(-vx1+vx2 )-cu_sq));
- (D27F.f[DIR_P0P ])[kzero] = c1o54* (c3o1*( Mx +Mz)+Conc_F*(c1o1+c9o2*( vx1 +vx3)*( vx1 +vx3)-cu_sq));
- (D27F.f[DIR_M0M ])[kbw ] = c1o54* (c3o1*(-Mx -Mz)+Conc_F*(c1o1+c9o2*(-vx1 -vx3)*(-vx1 -vx3)-cu_sq));
- (D27F.f[DIR_P0M ])[kb ] = c1o54* (c3o1*( Mx -Mz)+Conc_F*(c1o1+c9o2*( vx1 -vx3)*( vx1 -vx3)-cu_sq));
- (D27F.f[DIR_M0P ])[kw ] = c1o54* (c3o1*(-Mx +Mz)+Conc_F*(c1o1+c9o2*(-vx1 +vx3)*(-vx1 +vx3)-cu_sq));
- (D27F.f[DIR_0PP ])[kzero] = c1o54* (c3o1*( My +Mz)+Conc_F*(c1o1+c9o2*( vx2+vx3)*( vx2+vx3)-cu_sq));
- (D27F.f[DIR_0MM ])[kbs ] = c1o54* (c3o1*( -My -Mz)+Conc_F*(c1o1+c9o2*( -vx2-vx3)*( -vx2-vx3)-cu_sq));
- (D27F.f[DIR_0PM ])[kb ] = c1o54* (c3o1*( My -Mz)+Conc_F*(c1o1+c9o2*( vx2-vx3)*( vx2-vx3)-cu_sq));
- (D27F.f[DIR_0MP ])[ks ] = c1o54* (c3o1*( -My +Mz)+Conc_F*(c1o1+c9o2*( -vx2+vx3)*( -vx2+vx3)-cu_sq));
- (D27F.f[DIR_PPP ])[kzero] = c1o216*(c3o1*( Mx +My +Mz)+Conc_F*(c1o1+c9o2*( vx1+vx2+vx3)*( vx1+vx2+vx3)-cu_sq));
- (D27F.f[DIR_MMM ])[kbsw ] = c1o216*(c3o1*(-Mx -My -Mz)+Conc_F*(c1o1+c9o2*(-vx1-vx2-vx3)*(-vx1-vx2-vx3)-cu_sq));
- (D27F.f[DIR_PPM ])[kb ] = c1o216*(c3o1*( Mx +My -Mz)+Conc_F*(c1o1+c9o2*( vx1+vx2-vx3)*( vx1+vx2-vx3)-cu_sq));
- (D27F.f[DIR_MMP ])[ksw ] = c1o216*(c3o1*(-Mx -My +Mz)+Conc_F*(c1o1+c9o2*(-vx1-vx2+vx3)*(-vx1-vx2+vx3)-cu_sq));
- (D27F.f[DIR_PMP ])[ks ] = c1o216*(c3o1*( Mx -My +Mz)+Conc_F*(c1o1+c9o2*( vx1-vx2+vx3)*( vx1-vx2+vx3)-cu_sq));
- (D27F.f[DIR_MPM ])[kbw ] = c1o216*(c3o1*(-Mx +My -Mz)+Conc_F*(c1o1+c9o2*(-vx1+vx2-vx3)*(-vx1+vx2-vx3)-cu_sq));
- (D27F.f[DIR_PMM ])[kbs ] = c1o216*(c3o1*( Mx -My -Mz)+Conc_F*(c1o1+c9o2*( vx1-vx2-vx3)*( vx1-vx2-vx3)-cu_sq));
- (D27F.f[DIR_MPP ])[kw ] = c1o216*(c3o1*(-Mx +My +Mz)+Conc_F*(c1o1+c9o2*(-vx1+vx2+vx3)*(-vx1+vx2+vx3)-cu_sq));
+ (D27F.f[DIR_P00])[kzero] = c2o27* (c3o1*( Mx )+Conc_F*(c1o1+c9o2*( vx1 )*( vx1 )-cu_sq));
+ (D27F.f[DIR_M00])[kw ] = c2o27* (c3o1*(-Mx )+Conc_F*(c1o1+c9o2*(-vx1 )*(-vx1 )-cu_sq));
+ (D27F.f[DIR_0P0])[kzero] = c2o27* (c3o1*( My )+Conc_F*(c1o1+c9o2*( vx2 )*( vx2 )-cu_sq));
+ (D27F.f[DIR_0M0])[ks ] = c2o27* (c3o1*( -My )+Conc_F*(c1o1+c9o2*( -vx2 )*( -vx2 )-cu_sq));
+ (D27F.f[DIR_00P])[kzero] = c2o27* (c3o1*( Mz)+Conc_F*(c1o1+c9o2*( vx3)*( vx3)-cu_sq));
+ (D27F.f[DIR_00M])[kb ] = c2o27* (c3o1*( -Mz)+Conc_F*(c1o1+c9o2*( -vx3)*( -vx3)-cu_sq));
+ (D27F.f[DIR_PP0])[kzero] = c1o54* (c3o1*( Mx +My )+Conc_F*(c1o1+c9o2*( vx1+vx2 )*( vx1+vx2 )-cu_sq));
+ (D27F.f[DIR_MM0])[ksw ] = c1o54* (c3o1*(-Mx -My )+Conc_F*(c1o1+c9o2*(-vx1-vx2 )*(-vx1-vx2 )-cu_sq));
+ (D27F.f[DIR_PM0])[ks ] = c1o54* (c3o1*( Mx -My )+Conc_F*(c1o1+c9o2*( vx1-vx2 )*( vx1-vx2 )-cu_sq));
+ (D27F.f[DIR_MP0])[kw ] = c1o54* (c3o1*(-Mx +My )+Conc_F*(c1o1+c9o2*(-vx1+vx2 )*(-vx1+vx2 )-cu_sq));
+ (D27F.f[DIR_P0P])[kzero] = c1o54* (c3o1*( Mx +Mz)+Conc_F*(c1o1+c9o2*( vx1 +vx3)*( vx1 +vx3)-cu_sq));
+ (D27F.f[DIR_M0M])[kbw ] = c1o54* (c3o1*(-Mx -Mz)+Conc_F*(c1o1+c9o2*(-vx1 -vx3)*(-vx1 -vx3)-cu_sq));
+ (D27F.f[DIR_P0M])[kb ] = c1o54* (c3o1*( Mx -Mz)+Conc_F*(c1o1+c9o2*( vx1 -vx3)*( vx1 -vx3)-cu_sq));
+ (D27F.f[DIR_M0P])[kw ] = c1o54* (c3o1*(-Mx +Mz)+Conc_F*(c1o1+c9o2*(-vx1 +vx3)*(-vx1 +vx3)-cu_sq));
+ (D27F.f[DIR_0PP])[kzero] = c1o54* (c3o1*( My +Mz)+Conc_F*(c1o1+c9o2*( vx2+vx3)*( vx2+vx3)-cu_sq));
+ (D27F.f[DIR_0MM])[kbs ] = c1o54* (c3o1*( -My -Mz)+Conc_F*(c1o1+c9o2*( -vx2-vx3)*( -vx2-vx3)-cu_sq));
+ (D27F.f[DIR_0PM])[kb ] = c1o54* (c3o1*( My -Mz)+Conc_F*(c1o1+c9o2*( vx2-vx3)*( vx2-vx3)-cu_sq));
+ (D27F.f[DIR_0MP])[ks ] = c1o54* (c3o1*( -My +Mz)+Conc_F*(c1o1+c9o2*( -vx2+vx3)*( -vx2+vx3)-cu_sq));
+ (D27F.f[DIR_PPP])[kzero] = c1o216*(c3o1*( Mx +My +Mz)+Conc_F*(c1o1+c9o2*( vx1+vx2+vx3)*( vx1+vx2+vx3)-cu_sq));
+ (D27F.f[DIR_MMM])[kbsw ] = c1o216*(c3o1*(-Mx -My -Mz)+Conc_F*(c1o1+c9o2*(-vx1-vx2-vx3)*(-vx1-vx2-vx3)-cu_sq));
+ (D27F.f[DIR_PPM])[kb ] = c1o216*(c3o1*( Mx +My -Mz)+Conc_F*(c1o1+c9o2*( vx1+vx2-vx3)*( vx1+vx2-vx3)-cu_sq));
+ (D27F.f[DIR_MMP])[ksw ] = c1o216*(c3o1*(-Mx -My +Mz)+Conc_F*(c1o1+c9o2*(-vx1-vx2+vx3)*(-vx1-vx2+vx3)-cu_sq));
+ (D27F.f[DIR_PMP])[ks ] = c1o216*(c3o1*( Mx -My +Mz)+Conc_F*(c1o1+c9o2*( vx1-vx2+vx3)*( vx1-vx2+vx3)-cu_sq));
+ (D27F.f[DIR_MPM])[kbw ] = c1o216*(c3o1*(-Mx +My -Mz)+Conc_F*(c1o1+c9o2*(-vx1+vx2-vx3)*(-vx1+vx2-vx3)-cu_sq));
+ (D27F.f[DIR_PMM])[kbs ] = c1o216*(c3o1*( Mx -My -Mz)+Conc_F*(c1o1+c9o2*( vx1-vx2-vx3)*( vx1-vx2-vx3)-cu_sq));
+ (D27F.f[DIR_MPP])[kw ] = c1o216*(c3o1*(-Mx +My +Mz)+Conc_F*(c1o1+c9o2*(-vx1+vx2+vx3)*(-vx1+vx2+vx3)-cu_sq));
@@ -46812,32 +46812,32 @@ __global__ void scaleCFThS27( real* DC,
cu_sq=c3o2*(vx1*vx1+vx2*vx2+vx3*vx3);
(D27F.f[DIR_000])[kzero] = c8o27* Conc_F*(c1o1-cu_sq);
- (D27F.f[DIR_P00 ])[kzero] = c2o27* (c3o1*( Mx )+Conc_F*(c1o1+c9o2*( vx1 )*( vx1 )-cu_sq));
- (D27F.f[DIR_M00 ])[kw ] = c2o27* (c3o1*(-Mx )+Conc_F*(c1o1+c9o2*(-vx1 )*(-vx1 )-cu_sq));
- (D27F.f[DIR_0P0 ])[kzero] = c2o27* (c3o1*( My )+Conc_F*(c1o1+c9o2*( vx2 )*( vx2 )-cu_sq));
- (D27F.f[DIR_0M0 ])[ks ] = c2o27* (c3o1*( -My )+Conc_F*(c1o1+c9o2*( -vx2 )*( -vx2 )-cu_sq));
- (D27F.f[DIR_00P ])[kzero] = c2o27* (c3o1*( Mz)+Conc_F*(c1o1+c9o2*( vx3)*( vx3)-cu_sq));
- (D27F.f[DIR_00M ])[kb ] = c2o27* (c3o1*( -Mz)+Conc_F*(c1o1+c9o2*( -vx3)*( -vx3)-cu_sq));
- (D27F.f[DIR_PP0 ])[kzero] = c1o54* (c3o1*( Mx +My )+Conc_F*(c1o1+c9o2*( vx1+vx2 )*( vx1+vx2 )-cu_sq));
- (D27F.f[DIR_MM0 ])[ksw ] = c1o54* (c3o1*(-Mx -My )+Conc_F*(c1o1+c9o2*(-vx1-vx2 )*(-vx1-vx2 )-cu_sq));
- (D27F.f[DIR_PM0 ])[ks ] = c1o54* (c3o1*( Mx -My )+Conc_F*(c1o1+c9o2*( vx1-vx2 )*( vx1-vx2 )-cu_sq));
- (D27F.f[DIR_MP0 ])[kw ] = c1o54* (c3o1*(-Mx +My )+Conc_F*(c1o1+c9o2*(-vx1+vx2 )*(-vx1+vx2 )-cu_sq));
- (D27F.f[DIR_P0P ])[kzero] = c1o54* (c3o1*( Mx +Mz)+Conc_F*(c1o1+c9o2*( vx1 +vx3)*( vx1 +vx3)-cu_sq));
- (D27F.f[DIR_M0M ])[kbw ] = c1o54* (c3o1*(-Mx -Mz)+Conc_F*(c1o1+c9o2*(-vx1 -vx3)*(-vx1 -vx3)-cu_sq));
- (D27F.f[DIR_P0M ])[kb ] = c1o54* (c3o1*( Mx -Mz)+Conc_F*(c1o1+c9o2*( vx1 -vx3)*( vx1 -vx3)-cu_sq));
- (D27F.f[DIR_M0P ])[kw ] = c1o54* (c3o1*(-Mx +Mz)+Conc_F*(c1o1+c9o2*(-vx1 +vx3)*(-vx1 +vx3)-cu_sq));
- (D27F.f[DIR_0PP ])[kzero] = c1o54* (c3o1*( My +Mz)+Conc_F*(c1o1+c9o2*( vx2+vx3)*( vx2+vx3)-cu_sq));
- (D27F.f[DIR_0MM ])[kbs ] = c1o54* (c3o1*( -My -Mz)+Conc_F*(c1o1+c9o2*( -vx2-vx3)*( -vx2-vx3)-cu_sq));
- (D27F.f[DIR_0PM ])[kb ] = c1o54* (c3o1*( My -Mz)+Conc_F*(c1o1+c9o2*( vx2-vx3)*( vx2-vx3)-cu_sq));
- (D27F.f[DIR_0MP ])[ks ] = c1o54* (c3o1*( -My +Mz)+Conc_F*(c1o1+c9o2*( -vx2+vx3)*( -vx2+vx3)-cu_sq));
- (D27F.f[DIR_PPP ])[kzero] = c1o216*(c3o1*( Mx +My +Mz)+Conc_F*(c1o1+c9o2*( vx1+vx2+vx3)*( vx1+vx2+vx3)-cu_sq));
- (D27F.f[DIR_MMM ])[kbsw ] = c1o216*(c3o1*(-Mx -My -Mz)+Conc_F*(c1o1+c9o2*(-vx1-vx2-vx3)*(-vx1-vx2-vx3)-cu_sq));
- (D27F.f[DIR_PPM ])[kb ] = c1o216*(c3o1*( Mx +My -Mz)+Conc_F*(c1o1+c9o2*( vx1+vx2-vx3)*( vx1+vx2-vx3)-cu_sq));
- (D27F.f[DIR_MMP ])[ksw ] = c1o216*(c3o1*(-Mx -My +Mz)+Conc_F*(c1o1+c9o2*(-vx1-vx2+vx3)*(-vx1-vx2+vx3)-cu_sq));
- (D27F.f[DIR_PMP ])[ks ] = c1o216*(c3o1*( Mx -My +Mz)+Conc_F*(c1o1+c9o2*( vx1-vx2+vx3)*( vx1-vx2+vx3)-cu_sq));
- (D27F.f[DIR_MPM ])[kbw ] = c1o216*(c3o1*(-Mx +My -Mz)+Conc_F*(c1o1+c9o2*(-vx1+vx2-vx3)*(-vx1+vx2-vx3)-cu_sq));
- (D27F.f[DIR_PMM ])[kbs ] = c1o216*(c3o1*( Mx -My -Mz)+Conc_F*(c1o1+c9o2*( vx1-vx2-vx3)*( vx1-vx2-vx3)-cu_sq));
- (D27F.f[DIR_MPP ])[kw ] = c1o216*(c3o1*(-Mx +My +Mz)+Conc_F*(c1o1+c9o2*(-vx1+vx2+vx3)*(-vx1+vx2+vx3)-cu_sq));
+ (D27F.f[DIR_P00])[kzero] = c2o27* (c3o1*( Mx )+Conc_F*(c1o1+c9o2*( vx1 )*( vx1 )-cu_sq));
+ (D27F.f[DIR_M00])[kw ] = c2o27* (c3o1*(-Mx )+Conc_F*(c1o1+c9o2*(-vx1 )*(-vx1 )-cu_sq));
+ (D27F.f[DIR_0P0])[kzero] = c2o27* (c3o1*( My )+Conc_F*(c1o1+c9o2*( vx2 )*( vx2 )-cu_sq));
+ (D27F.f[DIR_0M0])[ks ] = c2o27* (c3o1*( -My )+Conc_F*(c1o1+c9o2*( -vx2 )*( -vx2 )-cu_sq));
+ (D27F.f[DIR_00P])[kzero] = c2o27* (c3o1*( Mz)+Conc_F*(c1o1+c9o2*( vx3)*( vx3)-cu_sq));
+ (D27F.f[DIR_00M])[kb ] = c2o27* (c3o1*( -Mz)+Conc_F*(c1o1+c9o2*( -vx3)*( -vx3)-cu_sq));
+ (D27F.f[DIR_PP0])[kzero] = c1o54* (c3o1*( Mx +My )+Conc_F*(c1o1+c9o2*( vx1+vx2 )*( vx1+vx2 )-cu_sq));
+ (D27F.f[DIR_MM0])[ksw ] = c1o54* (c3o1*(-Mx -My )+Conc_F*(c1o1+c9o2*(-vx1-vx2 )*(-vx1-vx2 )-cu_sq));
+ (D27F.f[DIR_PM0])[ks ] = c1o54* (c3o1*( Mx -My )+Conc_F*(c1o1+c9o2*( vx1-vx2 )*( vx1-vx2 )-cu_sq));
+ (D27F.f[DIR_MP0])[kw ] = c1o54* (c3o1*(-Mx +My )+Conc_F*(c1o1+c9o2*(-vx1+vx2 )*(-vx1+vx2 )-cu_sq));
+ (D27F.f[DIR_P0P])[kzero] = c1o54* (c3o1*( Mx +Mz)+Conc_F*(c1o1+c9o2*( vx1 +vx3)*( vx1 +vx3)-cu_sq));
+ (D27F.f[DIR_M0M])[kbw ] = c1o54* (c3o1*(-Mx -Mz)+Conc_F*(c1o1+c9o2*(-vx1 -vx3)*(-vx1 -vx3)-cu_sq));
+ (D27F.f[DIR_P0M])[kb ] = c1o54* (c3o1*( Mx -Mz)+Conc_F*(c1o1+c9o2*( vx1 -vx3)*( vx1 -vx3)-cu_sq));
+ (D27F.f[DIR_M0P])[kw ] = c1o54* (c3o1*(-Mx +Mz)+Conc_F*(c1o1+c9o2*(-vx1 +vx3)*(-vx1 +vx3)-cu_sq));
+ (D27F.f[DIR_0PP])[kzero] = c1o54* (c3o1*( My +Mz)+Conc_F*(c1o1+c9o2*( vx2+vx3)*( vx2+vx3)-cu_sq));
+ (D27F.f[DIR_0MM])[kbs ] = c1o54* (c3o1*( -My -Mz)+Conc_F*(c1o1+c9o2*( -vx2-vx3)*( -vx2-vx3)-cu_sq));
+ (D27F.f[DIR_0PM])[kb ] = c1o54* (c3o1*( My -Mz)+Conc_F*(c1o1+c9o2*( vx2-vx3)*( vx2-vx3)-cu_sq));
+ (D27F.f[DIR_0MP])[ks ] = c1o54* (c3o1*( -My +Mz)+Conc_F*(c1o1+c9o2*( -vx2+vx3)*( -vx2+vx3)-cu_sq));
+ (D27F.f[DIR_PPP])[kzero] = c1o216*(c3o1*( Mx +My +Mz)+Conc_F*(c1o1+c9o2*( vx1+vx2+vx3)*( vx1+vx2+vx3)-cu_sq));
+ (D27F.f[DIR_MMM])[kbsw ] = c1o216*(c3o1*(-Mx -My -Mz)+Conc_F*(c1o1+c9o2*(-vx1-vx2-vx3)*(-vx1-vx2-vx3)-cu_sq));
+ (D27F.f[DIR_PPM])[kb ] = c1o216*(c3o1*( Mx +My -Mz)+Conc_F*(c1o1+c9o2*( vx1+vx2-vx3)*( vx1+vx2-vx3)-cu_sq));
+ (D27F.f[DIR_MMP])[ksw ] = c1o216*(c3o1*(-Mx -My +Mz)+Conc_F*(c1o1+c9o2*(-vx1-vx2+vx3)*(-vx1-vx2+vx3)-cu_sq));
+ (D27F.f[DIR_PMP])[ks ] = c1o216*(c3o1*( Mx -My +Mz)+Conc_F*(c1o1+c9o2*( vx1-vx2+vx3)*( vx1-vx2+vx3)-cu_sq));
+ (D27F.f[DIR_MPM])[kbw ] = c1o216*(c3o1*(-Mx +My -Mz)+Conc_F*(c1o1+c9o2*(-vx1+vx2-vx3)*(-vx1+vx2-vx3)-cu_sq));
+ (D27F.f[DIR_PMM])[kbs ] = c1o216*(c3o1*( Mx -My -Mz)+Conc_F*(c1o1+c9o2*( vx1-vx2-vx3)*( vx1-vx2-vx3)-cu_sq));
+ (D27F.f[DIR_MPP])[kw ] = c1o216*(c3o1*(-Mx +My +Mz)+Conc_F*(c1o1+c9o2*(-vx1+vx2+vx3)*(-vx1+vx2+vx3)-cu_sq));
@@ -46890,32 +46890,32 @@ __global__ void scaleCFThS27( real* DC,
cu_sq=c3o2*(vx1*vx1+vx2*vx2+vx3*vx3);
(D27F.f[DIR_000])[kzero] = c8o27* Conc_F*(c1o1-cu_sq);
- (D27F.f[DIR_P00 ])[kzero] = c2o27* (c3o1*( Mx )+Conc_F*(c1o1+c9o2*( vx1 )*( vx1 )-cu_sq));
- (D27F.f[DIR_M00 ])[kw ] = c2o27* (c3o1*(-Mx )+Conc_F*(c1o1+c9o2*(-vx1 )*(-vx1 )-cu_sq));
- (D27F.f[DIR_0P0 ])[kzero] = c2o27* (c3o1*( My )+Conc_F*(c1o1+c9o2*( vx2 )*( vx2 )-cu_sq));
- (D27F.f[DIR_0M0 ])[ks ] = c2o27* (c3o1*( -My )+Conc_F*(c1o1+c9o2*( -vx2 )*( -vx2 )-cu_sq));
- (D27F.f[DIR_00P ])[kzero] = c2o27* (c3o1*( Mz)+Conc_F*(c1o1+c9o2*( vx3)*( vx3)-cu_sq));
- (D27F.f[DIR_00M ])[kb ] = c2o27* (c3o1*( -Mz)+Conc_F*(c1o1+c9o2*( -vx3)*( -vx3)-cu_sq));
- (D27F.f[DIR_PP0 ])[kzero] = c1o54* (c3o1*( Mx +My )+Conc_F*(c1o1+c9o2*( vx1+vx2 )*( vx1+vx2 )-cu_sq));
- (D27F.f[DIR_MM0 ])[ksw ] = c1o54* (c3o1*(-Mx -My )+Conc_F*(c1o1+c9o2*(-vx1-vx2 )*(-vx1-vx2 )-cu_sq));
- (D27F.f[DIR_PM0 ])[ks ] = c1o54* (c3o1*( Mx -My )+Conc_F*(c1o1+c9o2*( vx1-vx2 )*( vx1-vx2 )-cu_sq));
- (D27F.f[DIR_MP0 ])[kw ] = c1o54* (c3o1*(-Mx +My )+Conc_F*(c1o1+c9o2*(-vx1+vx2 )*(-vx1+vx2 )-cu_sq));
- (D27F.f[DIR_P0P ])[kzero] = c1o54* (c3o1*( Mx +Mz)+Conc_F*(c1o1+c9o2*( vx1 +vx3)*( vx1 +vx3)-cu_sq));
- (D27F.f[DIR_M0M ])[kbw ] = c1o54* (c3o1*(-Mx -Mz)+Conc_F*(c1o1+c9o2*(-vx1 -vx3)*(-vx1 -vx3)-cu_sq));
- (D27F.f[DIR_P0M ])[kb ] = c1o54* (c3o1*( Mx -Mz)+Conc_F*(c1o1+c9o2*( vx1 -vx3)*( vx1 -vx3)-cu_sq));
- (D27F.f[DIR_M0P ])[kw ] = c1o54* (c3o1*(-Mx +Mz)+Conc_F*(c1o1+c9o2*(-vx1 +vx3)*(-vx1 +vx3)-cu_sq));
- (D27F.f[DIR_0PP ])[kzero] = c1o54* (c3o1*( My +Mz)+Conc_F*(c1o1+c9o2*( vx2+vx3)*( vx2+vx3)-cu_sq));
- (D27F.f[DIR_0MM ])[kbs ] = c1o54* (c3o1*( -My -Mz)+Conc_F*(c1o1+c9o2*( -vx2-vx3)*( -vx2-vx3)-cu_sq));
- (D27F.f[DIR_0PM ])[kb ] = c1o54* (c3o1*( My -Mz)+Conc_F*(c1o1+c9o2*( vx2-vx3)*( vx2-vx3)-cu_sq));
- (D27F.f[DIR_0MP ])[ks ] = c1o54* (c3o1*( -My +Mz)+Conc_F*(c1o1+c9o2*( -vx2+vx3)*( -vx2+vx3)-cu_sq));
- (D27F.f[DIR_PPP ])[kzero] = c1o216*(c3o1*( Mx +My +Mz)+Conc_F*(c1o1+c9o2*( vx1+vx2+vx3)*( vx1+vx2+vx3)-cu_sq));
- (D27F.f[DIR_MMM ])[kbsw ] = c1o216*(c3o1*(-Mx -My -Mz)+Conc_F*(c1o1+c9o2*(-vx1-vx2-vx3)*(-vx1-vx2-vx3)-cu_sq));
- (D27F.f[DIR_PPM ])[kb ] = c1o216*(c3o1*( Mx +My -Mz)+Conc_F*(c1o1+c9o2*( vx1+vx2-vx3)*( vx1+vx2-vx3)-cu_sq));
- (D27F.f[DIR_MMP ])[ksw ] = c1o216*(c3o1*(-Mx -My +Mz)+Conc_F*(c1o1+c9o2*(-vx1-vx2+vx3)*(-vx1-vx2+vx3)-cu_sq));
- (D27F.f[DIR_PMP ])[ks ] = c1o216*(c3o1*( Mx -My +Mz)+Conc_F*(c1o1+c9o2*( vx1-vx2+vx3)*( vx1-vx2+vx3)-cu_sq));
- (D27F.f[DIR_MPM ])[kbw ] = c1o216*(c3o1*(-Mx +My -Mz)+Conc_F*(c1o1+c9o2*(-vx1+vx2-vx3)*(-vx1+vx2-vx3)-cu_sq));
- (D27F.f[DIR_PMM ])[kbs ] = c1o216*(c3o1*( Mx -My -Mz)+Conc_F*(c1o1+c9o2*( vx1-vx2-vx3)*( vx1-vx2-vx3)-cu_sq));
- (D27F.f[DIR_MPP ])[kw ] = c1o216*(c3o1*(-Mx +My +Mz)+Conc_F*(c1o1+c9o2*(-vx1+vx2+vx3)*(-vx1+vx2+vx3)-cu_sq));
+ (D27F.f[DIR_P00])[kzero] = c2o27* (c3o1*( Mx )+Conc_F*(c1o1+c9o2*( vx1 )*( vx1 )-cu_sq));
+ (D27F.f[DIR_M00])[kw ] = c2o27* (c3o1*(-Mx )+Conc_F*(c1o1+c9o2*(-vx1 )*(-vx1 )-cu_sq));
+ (D27F.f[DIR_0P0])[kzero] = c2o27* (c3o1*( My )+Conc_F*(c1o1+c9o2*( vx2 )*( vx2 )-cu_sq));
+ (D27F.f[DIR_0M0])[ks ] = c2o27* (c3o1*( -My )+Conc_F*(c1o1+c9o2*( -vx2 )*( -vx2 )-cu_sq));
+ (D27F.f[DIR_00P])[kzero] = c2o27* (c3o1*( Mz)+Conc_F*(c1o1+c9o2*( vx3)*( vx3)-cu_sq));
+ (D27F.f[DIR_00M])[kb ] = c2o27* (c3o1*( -Mz)+Conc_F*(c1o1+c9o2*( -vx3)*( -vx3)-cu_sq));
+ (D27F.f[DIR_PP0])[kzero] = c1o54* (c3o1*( Mx +My )+Conc_F*(c1o1+c9o2*( vx1+vx2 )*( vx1+vx2 )-cu_sq));
+ (D27F.f[DIR_MM0])[ksw ] = c1o54* (c3o1*(-Mx -My )+Conc_F*(c1o1+c9o2*(-vx1-vx2 )*(-vx1-vx2 )-cu_sq));
+ (D27F.f[DIR_PM0])[ks ] = c1o54* (c3o1*( Mx -My )+Conc_F*(c1o1+c9o2*( vx1-vx2 )*( vx1-vx2 )-cu_sq));
+ (D27F.f[DIR_MP0])[kw ] = c1o54* (c3o1*(-Mx +My )+Conc_F*(c1o1+c9o2*(-vx1+vx2 )*(-vx1+vx2 )-cu_sq));
+ (D27F.f[DIR_P0P])[kzero] = c1o54* (c3o1*( Mx +Mz)+Conc_F*(c1o1+c9o2*( vx1 +vx3)*( vx1 +vx3)-cu_sq));
+ (D27F.f[DIR_M0M])[kbw ] = c1o54* (c3o1*(-Mx -Mz)+Conc_F*(c1o1+c9o2*(-vx1 -vx3)*(-vx1 -vx3)-cu_sq));
+ (D27F.f[DIR_P0M])[kb ] = c1o54* (c3o1*( Mx -Mz)+Conc_F*(c1o1+c9o2*( vx1 -vx3)*( vx1 -vx3)-cu_sq));
+ (D27F.f[DIR_M0P])[kw ] = c1o54* (c3o1*(-Mx +Mz)+Conc_F*(c1o1+c9o2*(-vx1 +vx3)*(-vx1 +vx3)-cu_sq));
+ (D27F.f[DIR_0PP])[kzero] = c1o54* (c3o1*( My +Mz)+Conc_F*(c1o1+c9o2*( vx2+vx3)*( vx2+vx3)-cu_sq));
+ (D27F.f[DIR_0MM])[kbs ] = c1o54* (c3o1*( -My -Mz)+Conc_F*(c1o1+c9o2*( -vx2-vx3)*( -vx2-vx3)-cu_sq));
+ (D27F.f[DIR_0PM])[kb ] = c1o54* (c3o1*( My -Mz)+Conc_F*(c1o1+c9o2*( vx2-vx3)*( vx2-vx3)-cu_sq));
+ (D27F.f[DIR_0MP])[ks ] = c1o54* (c3o1*( -My +Mz)+Conc_F*(c1o1+c9o2*( -vx2+vx3)*( -vx2+vx3)-cu_sq));
+ (D27F.f[DIR_PPP])[kzero] = c1o216*(c3o1*( Mx +My +Mz)+Conc_F*(c1o1+c9o2*( vx1+vx2+vx3)*( vx1+vx2+vx3)-cu_sq));
+ (D27F.f[DIR_MMM])[kbsw ] = c1o216*(c3o1*(-Mx -My -Mz)+Conc_F*(c1o1+c9o2*(-vx1-vx2-vx3)*(-vx1-vx2-vx3)-cu_sq));
+ (D27F.f[DIR_PPM])[kb ] = c1o216*(c3o1*( Mx +My -Mz)+Conc_F*(c1o1+c9o2*( vx1+vx2-vx3)*( vx1+vx2-vx3)-cu_sq));
+ (D27F.f[DIR_MMP])[ksw ] = c1o216*(c3o1*(-Mx -My +Mz)+Conc_F*(c1o1+c9o2*(-vx1-vx2+vx3)*(-vx1-vx2+vx3)-cu_sq));
+ (D27F.f[DIR_PMP])[ks ] = c1o216*(c3o1*( Mx -My +Mz)+Conc_F*(c1o1+c9o2*( vx1-vx2+vx3)*( vx1-vx2+vx3)-cu_sq));
+ (D27F.f[DIR_MPM])[kbw ] = c1o216*(c3o1*(-Mx +My -Mz)+Conc_F*(c1o1+c9o2*(-vx1+vx2-vx3)*(-vx1+vx2-vx3)-cu_sq));
+ (D27F.f[DIR_PMM])[kbs ] = c1o216*(c3o1*( Mx -My -Mz)+Conc_F*(c1o1+c9o2*( vx1-vx2-vx3)*( vx1-vx2-vx3)-cu_sq));
+ (D27F.f[DIR_MPP])[kw ] = c1o216*(c3o1*(-Mx +My +Mz)+Conc_F*(c1o1+c9o2*(-vx1+vx2+vx3)*(-vx1+vx2+vx3)-cu_sq));
@@ -46978,32 +46978,32 @@ __global__ void scaleCFThS27( real* DC,
cu_sq=c3o2*(vx1*vx1+vx2*vx2+vx3*vx3);
(D27F.f[DIR_000])[kzero] = c8o27* Conc_F*(c1o1-cu_sq);
- (D27F.f[DIR_P00 ])[kzero] = c2o27* (c3o1*( Mx )+Conc_F*(c1o1+c9o2*( vx1 )*( vx1 )-cu_sq));
- (D27F.f[DIR_M00 ])[kw ] = c2o27* (c3o1*(-Mx )+Conc_F*(c1o1+c9o2*(-vx1 )*(-vx1 )-cu_sq));
- (D27F.f[DIR_0P0 ])[kzero] = c2o27* (c3o1*( My )+Conc_F*(c1o1+c9o2*( vx2 )*( vx2 )-cu_sq));
- (D27F.f[DIR_0M0 ])[ks ] = c2o27* (c3o1*( -My )+Conc_F*(c1o1+c9o2*( -vx2 )*( -vx2 )-cu_sq));
- (D27F.f[DIR_00P ])[kzero] = c2o27* (c3o1*( Mz)+Conc_F*(c1o1+c9o2*( vx3)*( vx3)-cu_sq));
- (D27F.f[DIR_00M ])[kb ] = c2o27* (c3o1*( -Mz)+Conc_F*(c1o1+c9o2*( -vx3)*( -vx3)-cu_sq));
- (D27F.f[DIR_PP0 ])[kzero] = c1o54* (c3o1*( Mx +My )+Conc_F*(c1o1+c9o2*( vx1+vx2 )*( vx1+vx2 )-cu_sq));
- (D27F.f[DIR_MM0 ])[ksw ] = c1o54* (c3o1*(-Mx -My )+Conc_F*(c1o1+c9o2*(-vx1-vx2 )*(-vx1-vx2 )-cu_sq));
- (D27F.f[DIR_PM0 ])[ks ] = c1o54* (c3o1*( Mx -My )+Conc_F*(c1o1+c9o2*( vx1-vx2 )*( vx1-vx2 )-cu_sq));
- (D27F.f[DIR_MP0 ])[kw ] = c1o54* (c3o1*(-Mx +My )+Conc_F*(c1o1+c9o2*(-vx1+vx2 )*(-vx1+vx2 )-cu_sq));
- (D27F.f[DIR_P0P ])[kzero] = c1o54* (c3o1*( Mx +Mz)+Conc_F*(c1o1+c9o2*( vx1 +vx3)*( vx1 +vx3)-cu_sq));
- (D27F.f[DIR_M0M ])[kbw ] = c1o54* (c3o1*(-Mx -Mz)+Conc_F*(c1o1+c9o2*(-vx1 -vx3)*(-vx1 -vx3)-cu_sq));
- (D27F.f[DIR_P0M ])[kb ] = c1o54* (c3o1*( Mx -Mz)+Conc_F*(c1o1+c9o2*( vx1 -vx3)*( vx1 -vx3)-cu_sq));
- (D27F.f[DIR_M0P ])[kw ] = c1o54* (c3o1*(-Mx +Mz)+Conc_F*(c1o1+c9o2*(-vx1 +vx3)*(-vx1 +vx3)-cu_sq));
- (D27F.f[DIR_0PP ])[kzero] = c1o54* (c3o1*( My +Mz)+Conc_F*(c1o1+c9o2*( vx2+vx3)*( vx2+vx3)-cu_sq));
- (D27F.f[DIR_0MM ])[kbs ] = c1o54* (c3o1*( -My -Mz)+Conc_F*(c1o1+c9o2*( -vx2-vx3)*( -vx2-vx3)-cu_sq));
- (D27F.f[DIR_0PM ])[kb ] = c1o54* (c3o1*( My -Mz)+Conc_F*(c1o1+c9o2*( vx2-vx3)*( vx2-vx3)-cu_sq));
- (D27F.f[DIR_0MP ])[ks ] = c1o54* (c3o1*( -My +Mz)+Conc_F*(c1o1+c9o2*( -vx2+vx3)*( -vx2+vx3)-cu_sq));
- (D27F.f[DIR_PPP ])[kzero] = c1o216*(c3o1*( Mx +My +Mz)+Conc_F*(c1o1+c9o2*( vx1+vx2+vx3)*( vx1+vx2+vx3)-cu_sq));
- (D27F.f[DIR_MMM ])[kbsw ] = c1o216*(c3o1*(-Mx -My -Mz)+Conc_F*(c1o1+c9o2*(-vx1-vx2-vx3)*(-vx1-vx2-vx3)-cu_sq));
- (D27F.f[DIR_PPM ])[kb ] = c1o216*(c3o1*( Mx +My -Mz)+Conc_F*(c1o1+c9o2*( vx1+vx2-vx3)*( vx1+vx2-vx3)-cu_sq));
- (D27F.f[DIR_MMP ])[ksw ] = c1o216*(c3o1*(-Mx -My +Mz)+Conc_F*(c1o1+c9o2*(-vx1-vx2+vx3)*(-vx1-vx2+vx3)-cu_sq));
- (D27F.f[DIR_PMP ])[ks ] = c1o216*(c3o1*( Mx -My +Mz)+Conc_F*(c1o1+c9o2*( vx1-vx2+vx3)*( vx1-vx2+vx3)-cu_sq));
- (D27F.f[DIR_MPM ])[kbw ] = c1o216*(c3o1*(-Mx +My -Mz)+Conc_F*(c1o1+c9o2*(-vx1+vx2-vx3)*(-vx1+vx2-vx3)-cu_sq));
- (D27F.f[DIR_PMM ])[kbs ] = c1o216*(c3o1*( Mx -My -Mz)+Conc_F*(c1o1+c9o2*( vx1-vx2-vx3)*( vx1-vx2-vx3)-cu_sq));
- (D27F.f[DIR_MPP ])[kw ] = c1o216*(c3o1*(-Mx +My +Mz)+Conc_F*(c1o1+c9o2*(-vx1+vx2+vx3)*(-vx1+vx2+vx3)-cu_sq));
+ (D27F.f[DIR_P00])[kzero] = c2o27* (c3o1*( Mx )+Conc_F*(c1o1+c9o2*( vx1 )*( vx1 )-cu_sq));
+ (D27F.f[DIR_M00])[kw ] = c2o27* (c3o1*(-Mx )+Conc_F*(c1o1+c9o2*(-vx1 )*(-vx1 )-cu_sq));
+ (D27F.f[DIR_0P0])[kzero] = c2o27* (c3o1*( My )+Conc_F*(c1o1+c9o2*( vx2 )*( vx2 )-cu_sq));
+ (D27F.f[DIR_0M0])[ks ] = c2o27* (c3o1*( -My )+Conc_F*(c1o1+c9o2*( -vx2 )*( -vx2 )-cu_sq));
+ (D27F.f[DIR_00P])[kzero] = c2o27* (c3o1*( Mz)+Conc_F*(c1o1+c9o2*( vx3)*( vx3)-cu_sq));
+ (D27F.f[DIR_00M])[kb ] = c2o27* (c3o1*( -Mz)+Conc_F*(c1o1+c9o2*( -vx3)*( -vx3)-cu_sq));
+ (D27F.f[DIR_PP0])[kzero] = c1o54* (c3o1*( Mx +My )+Conc_F*(c1o1+c9o2*( vx1+vx2 )*( vx1+vx2 )-cu_sq));
+ (D27F.f[DIR_MM0])[ksw ] = c1o54* (c3o1*(-Mx -My )+Conc_F*(c1o1+c9o2*(-vx1-vx2 )*(-vx1-vx2 )-cu_sq));
+ (D27F.f[DIR_PM0])[ks ] = c1o54* (c3o1*( Mx -My )+Conc_F*(c1o1+c9o2*( vx1-vx2 )*( vx1-vx2 )-cu_sq));
+ (D27F.f[DIR_MP0])[kw ] = c1o54* (c3o1*(-Mx +My )+Conc_F*(c1o1+c9o2*(-vx1+vx2 )*(-vx1+vx2 )-cu_sq));
+ (D27F.f[DIR_P0P])[kzero] = c1o54* (c3o1*( Mx +Mz)+Conc_F*(c1o1+c9o2*( vx1 +vx3)*( vx1 +vx3)-cu_sq));
+ (D27F.f[DIR_M0M])[kbw ] = c1o54* (c3o1*(-Mx -Mz)+Conc_F*(c1o1+c9o2*(-vx1 -vx3)*(-vx1 -vx3)-cu_sq));
+ (D27F.f[DIR_P0M])[kb ] = c1o54* (c3o1*( Mx -Mz)+Conc_F*(c1o1+c9o2*( vx1 -vx3)*( vx1 -vx3)-cu_sq));
+ (D27F.f[DIR_M0P])[kw ] = c1o54* (c3o1*(-Mx +Mz)+Conc_F*(c1o1+c9o2*(-vx1 +vx3)*(-vx1 +vx3)-cu_sq));
+ (D27F.f[DIR_0PP])[kzero] = c1o54* (c3o1*( My +Mz)+Conc_F*(c1o1+c9o2*( vx2+vx3)*( vx2+vx3)-cu_sq));
+ (D27F.f[DIR_0MM])[kbs ] = c1o54* (c3o1*( -My -Mz)+Conc_F*(c1o1+c9o2*( -vx2-vx3)*( -vx2-vx3)-cu_sq));
+ (D27F.f[DIR_0PM])[kb ] = c1o54* (c3o1*( My -Mz)+Conc_F*(c1o1+c9o2*( vx2-vx3)*( vx2-vx3)-cu_sq));
+ (D27F.f[DIR_0MP])[ks ] = c1o54* (c3o1*( -My +Mz)+Conc_F*(c1o1+c9o2*( -vx2+vx3)*( -vx2+vx3)-cu_sq));
+ (D27F.f[DIR_PPP])[kzero] = c1o216*(c3o1*( Mx +My +Mz)+Conc_F*(c1o1+c9o2*( vx1+vx2+vx3)*( vx1+vx2+vx3)-cu_sq));
+ (D27F.f[DIR_MMM])[kbsw ] = c1o216*(c3o1*(-Mx -My -Mz)+Conc_F*(c1o1+c9o2*(-vx1-vx2-vx3)*(-vx1-vx2-vx3)-cu_sq));
+ (D27F.f[DIR_PPM])[kb ] = c1o216*(c3o1*( Mx +My -Mz)+Conc_F*(c1o1+c9o2*( vx1+vx2-vx3)*( vx1+vx2-vx3)-cu_sq));
+ (D27F.f[DIR_MMP])[ksw ] = c1o216*(c3o1*(-Mx -My +Mz)+Conc_F*(c1o1+c9o2*(-vx1-vx2+vx3)*(-vx1-vx2+vx3)-cu_sq));
+ (D27F.f[DIR_PMP])[ks ] = c1o216*(c3o1*( Mx -My +Mz)+Conc_F*(c1o1+c9o2*( vx1-vx2+vx3)*( vx1-vx2+vx3)-cu_sq));
+ (D27F.f[DIR_MPM])[kbw ] = c1o216*(c3o1*(-Mx +My -Mz)+Conc_F*(c1o1+c9o2*(-vx1+vx2-vx3)*(-vx1+vx2-vx3)-cu_sq));
+ (D27F.f[DIR_PMM])[kbs ] = c1o216*(c3o1*( Mx -My -Mz)+Conc_F*(c1o1+c9o2*( vx1-vx2-vx3)*( vx1-vx2-vx3)-cu_sq));
+ (D27F.f[DIR_MPP])[kw ] = c1o216*(c3o1*(-Mx +My +Mz)+Conc_F*(c1o1+c9o2*(-vx1+vx2+vx3)*(-vx1+vx2+vx3)-cu_sq));
@@ -47056,32 +47056,32 @@ __global__ void scaleCFThS27( real* DC,
cu_sq=c3o2*(vx1*vx1+vx2*vx2+vx3*vx3);
(D27F.f[DIR_000])[kzero] = c8o27* Conc_F*(c1o1-cu_sq);
- (D27F.f[DIR_P00 ])[kzero] = c2o27* (c3o1*( Mx )+Conc_F*(c1o1+c9o2*( vx1 )*( vx1 )-cu_sq));
- (D27F.f[DIR_M00 ])[kw ] = c2o27* (c3o1*(-Mx )+Conc_F*(c1o1+c9o2*(-vx1 )*(-vx1 )-cu_sq));
- (D27F.f[DIR_0P0 ])[kzero] = c2o27* (c3o1*( My )+Conc_F*(c1o1+c9o2*( vx2 )*( vx2 )-cu_sq));
- (D27F.f[DIR_0M0 ])[ks ] = c2o27* (c3o1*( -My )+Conc_F*(c1o1+c9o2*( -vx2 )*( -vx2 )-cu_sq));
- (D27F.f[DIR_00P ])[kzero] = c2o27* (c3o1*( Mz)+Conc_F*(c1o1+c9o2*( vx3)*( vx3)-cu_sq));
- (D27F.f[DIR_00M ])[kb ] = c2o27* (c3o1*( -Mz)+Conc_F*(c1o1+c9o2*( -vx3)*( -vx3)-cu_sq));
- (D27F.f[DIR_PP0 ])[kzero] = c1o54* (c3o1*( Mx +My )+Conc_F*(c1o1+c9o2*( vx1+vx2 )*( vx1+vx2 )-cu_sq));
- (D27F.f[DIR_MM0 ])[ksw ] = c1o54* (c3o1*(-Mx -My )+Conc_F*(c1o1+c9o2*(-vx1-vx2 )*(-vx1-vx2 )-cu_sq));
- (D27F.f[DIR_PM0 ])[ks ] = c1o54* (c3o1*( Mx -My )+Conc_F*(c1o1+c9o2*( vx1-vx2 )*( vx1-vx2 )-cu_sq));
- (D27F.f[DIR_MP0 ])[kw ] = c1o54* (c3o1*(-Mx +My )+Conc_F*(c1o1+c9o2*(-vx1+vx2 )*(-vx1+vx2 )-cu_sq));
- (D27F.f[DIR_P0P ])[kzero] = c1o54* (c3o1*( Mx +Mz)+Conc_F*(c1o1+c9o2*( vx1 +vx3)*( vx1 +vx3)-cu_sq));
- (D27F.f[DIR_M0M ])[kbw ] = c1o54* (c3o1*(-Mx -Mz)+Conc_F*(c1o1+c9o2*(-vx1 -vx3)*(-vx1 -vx3)-cu_sq));
- (D27F.f[DIR_P0M ])[kb ] = c1o54* (c3o1*( Mx -Mz)+Conc_F*(c1o1+c9o2*( vx1 -vx3)*( vx1 -vx3)-cu_sq));
- (D27F.f[DIR_M0P ])[kw ] = c1o54* (c3o1*(-Mx +Mz)+Conc_F*(c1o1+c9o2*(-vx1 +vx3)*(-vx1 +vx3)-cu_sq));
- (D27F.f[DIR_0PP ])[kzero] = c1o54* (c3o1*( My +Mz)+Conc_F*(c1o1+c9o2*( vx2+vx3)*( vx2+vx3)-cu_sq));
- (D27F.f[DIR_0MM ])[kbs ] = c1o54* (c3o1*( -My -Mz)+Conc_F*(c1o1+c9o2*( -vx2-vx3)*( -vx2-vx3)-cu_sq));
- (D27F.f[DIR_0PM ])[kb ] = c1o54* (c3o1*( My -Mz)+Conc_F*(c1o1+c9o2*( vx2-vx3)*( vx2-vx3)-cu_sq));
- (D27F.f[DIR_0MP ])[ks ] = c1o54* (c3o1*( -My +Mz)+Conc_F*(c1o1+c9o2*( -vx2+vx3)*( -vx2+vx3)-cu_sq));
- (D27F.f[DIR_PPP ])[kzero] = c1o216*(c3o1*( Mx +My +Mz)+Conc_F*(c1o1+c9o2*( vx1+vx2+vx3)*( vx1+vx2+vx3)-cu_sq));
- (D27F.f[DIR_MMM ])[kbsw ] = c1o216*(c3o1*(-Mx -My -Mz)+Conc_F*(c1o1+c9o2*(-vx1-vx2-vx3)*(-vx1-vx2-vx3)-cu_sq));
- (D27F.f[DIR_PPM ])[kb ] = c1o216*(c3o1*( Mx +My -Mz)+Conc_F*(c1o1+c9o2*( vx1+vx2-vx3)*( vx1+vx2-vx3)-cu_sq));
- (D27F.f[DIR_MMP ])[ksw ] = c1o216*(c3o1*(-Mx -My +Mz)+Conc_F*(c1o1+c9o2*(-vx1-vx2+vx3)*(-vx1-vx2+vx3)-cu_sq));
- (D27F.f[DIR_PMP ])[ks ] = c1o216*(c3o1*( Mx -My +Mz)+Conc_F*(c1o1+c9o2*( vx1-vx2+vx3)*( vx1-vx2+vx3)-cu_sq));
- (D27F.f[DIR_MPM ])[kbw ] = c1o216*(c3o1*(-Mx +My -Mz)+Conc_F*(c1o1+c9o2*(-vx1+vx2-vx3)*(-vx1+vx2-vx3)-cu_sq));
- (D27F.f[DIR_PMM ])[kbs ] = c1o216*(c3o1*( Mx -My -Mz)+Conc_F*(c1o1+c9o2*( vx1-vx2-vx3)*( vx1-vx2-vx3)-cu_sq));
- (D27F.f[DIR_MPP ])[kw ] = c1o216*(c3o1*(-Mx +My +Mz)+Conc_F*(c1o1+c9o2*(-vx1+vx2+vx3)*(-vx1+vx2+vx3)-cu_sq));
+ (D27F.f[DIR_P00])[kzero] = c2o27* (c3o1*( Mx )+Conc_F*(c1o1+c9o2*( vx1 )*( vx1 )-cu_sq));
+ (D27F.f[DIR_M00])[kw ] = c2o27* (c3o1*(-Mx )+Conc_F*(c1o1+c9o2*(-vx1 )*(-vx1 )-cu_sq));
+ (D27F.f[DIR_0P0])[kzero] = c2o27* (c3o1*( My )+Conc_F*(c1o1+c9o2*( vx2 )*( vx2 )-cu_sq));
+ (D27F.f[DIR_0M0])[ks ] = c2o27* (c3o1*( -My )+Conc_F*(c1o1+c9o2*( -vx2 )*( -vx2 )-cu_sq));
+ (D27F.f[DIR_00P])[kzero] = c2o27* (c3o1*( Mz)+Conc_F*(c1o1+c9o2*( vx3)*( vx3)-cu_sq));
+ (D27F.f[DIR_00M])[kb ] = c2o27* (c3o1*( -Mz)+Conc_F*(c1o1+c9o2*( -vx3)*( -vx3)-cu_sq));
+ (D27F.f[DIR_PP0])[kzero] = c1o54* (c3o1*( Mx +My )+Conc_F*(c1o1+c9o2*( vx1+vx2 )*( vx1+vx2 )-cu_sq));
+ (D27F.f[DIR_MM0])[ksw ] = c1o54* (c3o1*(-Mx -My )+Conc_F*(c1o1+c9o2*(-vx1-vx2 )*(-vx1-vx2 )-cu_sq));
+ (D27F.f[DIR_PM0])[ks ] = c1o54* (c3o1*( Mx -My )+Conc_F*(c1o1+c9o2*( vx1-vx2 )*( vx1-vx2 )-cu_sq));
+ (D27F.f[DIR_MP0])[kw ] = c1o54* (c3o1*(-Mx +My )+Conc_F*(c1o1+c9o2*(-vx1+vx2 )*(-vx1+vx2 )-cu_sq));
+ (D27F.f[DIR_P0P])[kzero] = c1o54* (c3o1*( Mx +Mz)+Conc_F*(c1o1+c9o2*( vx1 +vx3)*( vx1 +vx3)-cu_sq));
+ (D27F.f[DIR_M0M])[kbw ] = c1o54* (c3o1*(-Mx -Mz)+Conc_F*(c1o1+c9o2*(-vx1 -vx3)*(-vx1 -vx3)-cu_sq));
+ (D27F.f[DIR_P0M])[kb ] = c1o54* (c3o1*( Mx -Mz)+Conc_F*(c1o1+c9o2*( vx1 -vx3)*( vx1 -vx3)-cu_sq));
+ (D27F.f[DIR_M0P])[kw ] = c1o54* (c3o1*(-Mx +Mz)+Conc_F*(c1o1+c9o2*(-vx1 +vx3)*(-vx1 +vx3)-cu_sq));
+ (D27F.f[DIR_0PP])[kzero] = c1o54* (c3o1*( My +Mz)+Conc_F*(c1o1+c9o2*( vx2+vx3)*( vx2+vx3)-cu_sq));
+ (D27F.f[DIR_0MM])[kbs ] = c1o54* (c3o1*( -My -Mz)+Conc_F*(c1o1+c9o2*( -vx2-vx3)*( -vx2-vx3)-cu_sq));
+ (D27F.f[DIR_0PM])[kb ] = c1o54* (c3o1*( My -Mz)+Conc_F*(c1o1+c9o2*( vx2-vx3)*( vx2-vx3)-cu_sq));
+ (D27F.f[DIR_0MP])[ks ] = c1o54* (c3o1*( -My +Mz)+Conc_F*(c1o1+c9o2*( -vx2+vx3)*( -vx2+vx3)-cu_sq));
+ (D27F.f[DIR_PPP])[kzero] = c1o216*(c3o1*( Mx +My +Mz)+Conc_F*(c1o1+c9o2*( vx1+vx2+vx3)*( vx1+vx2+vx3)-cu_sq));
+ (D27F.f[DIR_MMM])[kbsw ] = c1o216*(c3o1*(-Mx -My -Mz)+Conc_F*(c1o1+c9o2*(-vx1-vx2-vx3)*(-vx1-vx2-vx3)-cu_sq));
+ (D27F.f[DIR_PPM])[kb ] = c1o216*(c3o1*( Mx +My -Mz)+Conc_F*(c1o1+c9o2*( vx1+vx2-vx3)*( vx1+vx2-vx3)-cu_sq));
+ (D27F.f[DIR_MMP])[ksw ] = c1o216*(c3o1*(-Mx -My +Mz)+Conc_F*(c1o1+c9o2*(-vx1-vx2+vx3)*(-vx1-vx2+vx3)-cu_sq));
+ (D27F.f[DIR_PMP])[ks ] = c1o216*(c3o1*( Mx -My +Mz)+Conc_F*(c1o1+c9o2*( vx1-vx2+vx3)*( vx1-vx2+vx3)-cu_sq));
+ (D27F.f[DIR_MPM])[kbw ] = c1o216*(c3o1*(-Mx +My -Mz)+Conc_F*(c1o1+c9o2*(-vx1+vx2-vx3)*(-vx1+vx2-vx3)-cu_sq));
+ (D27F.f[DIR_PMM])[kbs ] = c1o216*(c3o1*( Mx -My -Mz)+Conc_F*(c1o1+c9o2*( vx1-vx2-vx3)*( vx1-vx2-vx3)-cu_sq));
+ (D27F.f[DIR_MPP])[kw ] = c1o216*(c3o1*(-Mx +My +Mz)+Conc_F*(c1o1+c9o2*(-vx1+vx2+vx3)*(-vx1+vx2+vx3)-cu_sq));
@@ -47134,32 +47134,32 @@ __global__ void scaleCFThS27( real* DC,
cu_sq=c3o2*(vx1*vx1+vx2*vx2+vx3*vx3);
(D27F.f[DIR_000])[kzero] = c8o27* Conc_F*(c1o1-cu_sq);
- (D27F.f[DIR_P00 ])[kzero] = c2o27* (c3o1*( Mx )+Conc_F*(c1o1+c9o2*( vx1 )*( vx1 )-cu_sq));
- (D27F.f[DIR_M00 ])[kw ] = c2o27* (c3o1*(-Mx )+Conc_F*(c1o1+c9o2*(-vx1 )*(-vx1 )-cu_sq));
- (D27F.f[DIR_0P0 ])[kzero] = c2o27* (c3o1*( My )+Conc_F*(c1o1+c9o2*( vx2 )*( vx2 )-cu_sq));
- (D27F.f[DIR_0M0 ])[ks ] = c2o27* (c3o1*( -My )+Conc_F*(c1o1+c9o2*( -vx2 )*( -vx2 )-cu_sq));
- (D27F.f[DIR_00P ])[kzero] = c2o27* (c3o1*( Mz)+Conc_F*(c1o1+c9o2*( vx3)*( vx3)-cu_sq));
- (D27F.f[DIR_00M ])[kb ] = c2o27* (c3o1*( -Mz)+Conc_F*(c1o1+c9o2*( -vx3)*( -vx3)-cu_sq));
- (D27F.f[DIR_PP0 ])[kzero] = c1o54* (c3o1*( Mx +My )+Conc_F*(c1o1+c9o2*( vx1+vx2 )*( vx1+vx2 )-cu_sq));
- (D27F.f[DIR_MM0 ])[ksw ] = c1o54* (c3o1*(-Mx -My )+Conc_F*(c1o1+c9o2*(-vx1-vx2 )*(-vx1-vx2 )-cu_sq));
- (D27F.f[DIR_PM0 ])[ks ] = c1o54* (c3o1*( Mx -My )+Conc_F*(c1o1+c9o2*( vx1-vx2 )*( vx1-vx2 )-cu_sq));
- (D27F.f[DIR_MP0 ])[kw ] = c1o54* (c3o1*(-Mx +My )+Conc_F*(c1o1+c9o2*(-vx1+vx2 )*(-vx1+vx2 )-cu_sq));
- (D27F.f[DIR_P0P ])[kzero] = c1o54* (c3o1*( Mx +Mz)+Conc_F*(c1o1+c9o2*( vx1 +vx3)*( vx1 +vx3)-cu_sq));
- (D27F.f[DIR_M0M ])[kbw ] = c1o54* (c3o1*(-Mx -Mz)+Conc_F*(c1o1+c9o2*(-vx1 -vx3)*(-vx1 -vx3)-cu_sq));
- (D27F.f[DIR_P0M ])[kb ] = c1o54* (c3o1*( Mx -Mz)+Conc_F*(c1o1+c9o2*( vx1 -vx3)*( vx1 -vx3)-cu_sq));
- (D27F.f[DIR_M0P ])[kw ] = c1o54* (c3o1*(-Mx +Mz)+Conc_F*(c1o1+c9o2*(-vx1 +vx3)*(-vx1 +vx3)-cu_sq));
- (D27F.f[DIR_0PP ])[kzero] = c1o54* (c3o1*( My +Mz)+Conc_F*(c1o1+c9o2*( vx2+vx3)*( vx2+vx3)-cu_sq));
- (D27F.f[DIR_0MM ])[kbs ] = c1o54* (c3o1*( -My -Mz)+Conc_F*(c1o1+c9o2*( -vx2-vx3)*( -vx2-vx3)-cu_sq));
- (D27F.f[DIR_0PM ])[kb ] = c1o54* (c3o1*( My -Mz)+Conc_F*(c1o1+c9o2*( vx2-vx3)*( vx2-vx3)-cu_sq));
- (D27F.f[DIR_0MP ])[ks ] = c1o54* (c3o1*( -My +Mz)+Conc_F*(c1o1+c9o2*( -vx2+vx3)*( -vx2+vx3)-cu_sq));
- (D27F.f[DIR_PPP ])[kzero] = c1o216*(c3o1*( Mx +My +Mz)+Conc_F*(c1o1+c9o2*( vx1+vx2+vx3)*( vx1+vx2+vx3)-cu_sq));
- (D27F.f[DIR_MMM ])[kbsw ] = c1o216*(c3o1*(-Mx -My -Mz)+Conc_F*(c1o1+c9o2*(-vx1-vx2-vx3)*(-vx1-vx2-vx3)-cu_sq));
- (D27F.f[DIR_PPM ])[kb ] = c1o216*(c3o1*( Mx +My -Mz)+Conc_F*(c1o1+c9o2*( vx1+vx2-vx3)*( vx1+vx2-vx3)-cu_sq));
- (D27F.f[DIR_MMP ])[ksw ] = c1o216*(c3o1*(-Mx -My +Mz)+Conc_F*(c1o1+c9o2*(-vx1-vx2+vx3)*(-vx1-vx2+vx3)-cu_sq));
- (D27F.f[DIR_PMP ])[ks ] = c1o216*(c3o1*( Mx -My +Mz)+Conc_F*(c1o1+c9o2*( vx1-vx2+vx3)*( vx1-vx2+vx3)-cu_sq));
- (D27F.f[DIR_MPM ])[kbw ] = c1o216*(c3o1*(-Mx +My -Mz)+Conc_F*(c1o1+c9o2*(-vx1+vx2-vx3)*(-vx1+vx2-vx3)-cu_sq));
- (D27F.f[DIR_PMM ])[kbs ] = c1o216*(c3o1*( Mx -My -Mz)+Conc_F*(c1o1+c9o2*( vx1-vx2-vx3)*( vx1-vx2-vx3)-cu_sq));
- (D27F.f[DIR_MPP ])[kw ] = c1o216*(c3o1*(-Mx +My +Mz)+Conc_F*(c1o1+c9o2*(-vx1+vx2+vx3)*(-vx1+vx2+vx3)-cu_sq));
+ (D27F.f[DIR_P00])[kzero] = c2o27* (c3o1*( Mx )+Conc_F*(c1o1+c9o2*( vx1 )*( vx1 )-cu_sq));
+ (D27F.f[DIR_M00])[kw ] = c2o27* (c3o1*(-Mx )+Conc_F*(c1o1+c9o2*(-vx1 )*(-vx1 )-cu_sq));
+ (D27F.f[DIR_0P0])[kzero] = c2o27* (c3o1*( My )+Conc_F*(c1o1+c9o2*( vx2 )*( vx2 )-cu_sq));
+ (D27F.f[DIR_0M0])[ks ] = c2o27* (c3o1*( -My )+Conc_F*(c1o1+c9o2*( -vx2 )*( -vx2 )-cu_sq));
+ (D27F.f[DIR_00P])[kzero] = c2o27* (c3o1*( Mz)+Conc_F*(c1o1+c9o2*( vx3)*( vx3)-cu_sq));
+ (D27F.f[DIR_00M])[kb ] = c2o27* (c3o1*( -Mz)+Conc_F*(c1o1+c9o2*( -vx3)*( -vx3)-cu_sq));
+ (D27F.f[DIR_PP0])[kzero] = c1o54* (c3o1*( Mx +My )+Conc_F*(c1o1+c9o2*( vx1+vx2 )*( vx1+vx2 )-cu_sq));
+ (D27F.f[DIR_MM0])[ksw ] = c1o54* (c3o1*(-Mx -My )+Conc_F*(c1o1+c9o2*(-vx1-vx2 )*(-vx1-vx2 )-cu_sq));
+ (D27F.f[DIR_PM0])[ks ] = c1o54* (c3o1*( Mx -My )+Conc_F*(c1o1+c9o2*( vx1-vx2 )*( vx1-vx2 )-cu_sq));
+ (D27F.f[DIR_MP0])[kw ] = c1o54* (c3o1*(-Mx +My )+Conc_F*(c1o1+c9o2*(-vx1+vx2 )*(-vx1+vx2 )-cu_sq));
+ (D27F.f[DIR_P0P])[kzero] = c1o54* (c3o1*( Mx +Mz)+Conc_F*(c1o1+c9o2*( vx1 +vx3)*( vx1 +vx3)-cu_sq));
+ (D27F.f[DIR_M0M])[kbw ] = c1o54* (c3o1*(-Mx -Mz)+Conc_F*(c1o1+c9o2*(-vx1 -vx3)*(-vx1 -vx3)-cu_sq));
+ (D27F.f[DIR_P0M])[kb ] = c1o54* (c3o1*( Mx -Mz)+Conc_F*(c1o1+c9o2*( vx1 -vx3)*( vx1 -vx3)-cu_sq));
+ (D27F.f[DIR_M0P])[kw ] = c1o54* (c3o1*(-Mx +Mz)+Conc_F*(c1o1+c9o2*(-vx1 +vx3)*(-vx1 +vx3)-cu_sq));
+ (D27F.f[DIR_0PP])[kzero] = c1o54* (c3o1*( My +Mz)+Conc_F*(c1o1+c9o2*( vx2+vx3)*( vx2+vx3)-cu_sq));
+ (D27F.f[DIR_0MM])[kbs ] = c1o54* (c3o1*( -My -Mz)+Conc_F*(c1o1+c9o2*( -vx2-vx3)*( -vx2-vx3)-cu_sq));
+ (D27F.f[DIR_0PM])[kb ] = c1o54* (c3o1*( My -Mz)+Conc_F*(c1o1+c9o2*( vx2-vx3)*( vx2-vx3)-cu_sq));
+ (D27F.f[DIR_0MP])[ks ] = c1o54* (c3o1*( -My +Mz)+Conc_F*(c1o1+c9o2*( -vx2+vx3)*( -vx2+vx3)-cu_sq));
+ (D27F.f[DIR_PPP])[kzero] = c1o216*(c3o1*( Mx +My +Mz)+Conc_F*(c1o1+c9o2*( vx1+vx2+vx3)*( vx1+vx2+vx3)-cu_sq));
+ (D27F.f[DIR_MMM])[kbsw ] = c1o216*(c3o1*(-Mx -My -Mz)+Conc_F*(c1o1+c9o2*(-vx1-vx2-vx3)*(-vx1-vx2-vx3)-cu_sq));
+ (D27F.f[DIR_PPM])[kb ] = c1o216*(c3o1*( Mx +My -Mz)+Conc_F*(c1o1+c9o2*( vx1+vx2-vx3)*( vx1+vx2-vx3)-cu_sq));
+ (D27F.f[DIR_MMP])[ksw ] = c1o216*(c3o1*(-Mx -My +Mz)+Conc_F*(c1o1+c9o2*(-vx1-vx2+vx3)*(-vx1-vx2+vx3)-cu_sq));
+ (D27F.f[DIR_PMP])[ks ] = c1o216*(c3o1*( Mx -My +Mz)+Conc_F*(c1o1+c9o2*( vx1-vx2+vx3)*( vx1-vx2+vx3)-cu_sq));
+ (D27F.f[DIR_MPM])[kbw ] = c1o216*(c3o1*(-Mx +My -Mz)+Conc_F*(c1o1+c9o2*(-vx1+vx2-vx3)*(-vx1+vx2-vx3)-cu_sq));
+ (D27F.f[DIR_PMM])[kbs ] = c1o216*(c3o1*( Mx -My -Mz)+Conc_F*(c1o1+c9o2*( vx1-vx2-vx3)*( vx1-vx2-vx3)-cu_sq));
+ (D27F.f[DIR_MPP])[kw ] = c1o216*(c3o1*(-Mx +My +Mz)+Conc_F*(c1o1+c9o2*(-vx1+vx2+vx3)*(-vx1+vx2+vx3)-cu_sq));
@@ -47212,32 +47212,32 @@ __global__ void scaleCFThS27( real* DC,
cu_sq=c3o2*(vx1*vx1+vx2*vx2+vx3*vx3);
(D27F.f[DIR_000])[kzero] = c8o27* Conc_F*(c1o1-cu_sq);
- (D27F.f[DIR_P00 ])[kzero] = c2o27* (c3o1*( Mx )+Conc_F*(c1o1+c9o2*( vx1 )*( vx1 )-cu_sq));
- (D27F.f[DIR_M00 ])[kw ] = c2o27* (c3o1*(-Mx )+Conc_F*(c1o1+c9o2*(-vx1 )*(-vx1 )-cu_sq));
- (D27F.f[DIR_0P0 ])[kzero] = c2o27* (c3o1*( My )+Conc_F*(c1o1+c9o2*( vx2 )*( vx2 )-cu_sq));
- (D27F.f[DIR_0M0 ])[ks ] = c2o27* (c3o1*( -My )+Conc_F*(c1o1+c9o2*( -vx2 )*( -vx2 )-cu_sq));
- (D27F.f[DIR_00P ])[kzero] = c2o27* (c3o1*( Mz)+Conc_F*(c1o1+c9o2*( vx3)*( vx3)-cu_sq));
- (D27F.f[DIR_00M ])[kb ] = c2o27* (c3o1*( -Mz)+Conc_F*(c1o1+c9o2*( -vx3)*( -vx3)-cu_sq));
- (D27F.f[DIR_PP0 ])[kzero] = c1o54* (c3o1*( Mx +My )+Conc_F*(c1o1+c9o2*( vx1+vx2 )*( vx1+vx2 )-cu_sq));
- (D27F.f[DIR_MM0 ])[ksw ] = c1o54* (c3o1*(-Mx -My )+Conc_F*(c1o1+c9o2*(-vx1-vx2 )*(-vx1-vx2 )-cu_sq));
- (D27F.f[DIR_PM0 ])[ks ] = c1o54* (c3o1*( Mx -My )+Conc_F*(c1o1+c9o2*( vx1-vx2 )*( vx1-vx2 )-cu_sq));
- (D27F.f[DIR_MP0 ])[kw ] = c1o54* (c3o1*(-Mx +My )+Conc_F*(c1o1+c9o2*(-vx1+vx2 )*(-vx1+vx2 )-cu_sq));
- (D27F.f[DIR_P0P ])[kzero] = c1o54* (c3o1*( Mx +Mz)+Conc_F*(c1o1+c9o2*( vx1 +vx3)*( vx1 +vx3)-cu_sq));
- (D27F.f[DIR_M0M ])[kbw ] = c1o54* (c3o1*(-Mx -Mz)+Conc_F*(c1o1+c9o2*(-vx1 -vx3)*(-vx1 -vx3)-cu_sq));
- (D27F.f[DIR_P0M ])[kb ] = c1o54* (c3o1*( Mx -Mz)+Conc_F*(c1o1+c9o2*( vx1 -vx3)*( vx1 -vx3)-cu_sq));
- (D27F.f[DIR_M0P ])[kw ] = c1o54* (c3o1*(-Mx +Mz)+Conc_F*(c1o1+c9o2*(-vx1 +vx3)*(-vx1 +vx3)-cu_sq));
- (D27F.f[DIR_0PP ])[kzero] = c1o54* (c3o1*( My +Mz)+Conc_F*(c1o1+c9o2*( vx2+vx3)*( vx2+vx3)-cu_sq));
- (D27F.f[DIR_0MM ])[kbs ] = c1o54* (c3o1*( -My -Mz)+Conc_F*(c1o1+c9o2*( -vx2-vx3)*( -vx2-vx3)-cu_sq));
- (D27F.f[DIR_0PM ])[kb ] = c1o54* (c3o1*( My -Mz)+Conc_F*(c1o1+c9o2*( vx2-vx3)*( vx2-vx3)-cu_sq));
- (D27F.f[DIR_0MP ])[ks ] = c1o54* (c3o1*( -My +Mz)+Conc_F*(c1o1+c9o2*( -vx2+vx3)*( -vx2+vx3)-cu_sq));
- (D27F.f[DIR_PPP ])[kzero] = c1o216*(c3o1*( Mx +My +Mz)+Conc_F*(c1o1+c9o2*( vx1+vx2+vx3)*( vx1+vx2+vx3)-cu_sq));
- (D27F.f[DIR_MMM ])[kbsw ] = c1o216*(c3o1*(-Mx -My -Mz)+Conc_F*(c1o1+c9o2*(-vx1-vx2-vx3)*(-vx1-vx2-vx3)-cu_sq));
- (D27F.f[DIR_PPM ])[kb ] = c1o216*(c3o1*( Mx +My -Mz)+Conc_F*(c1o1+c9o2*( vx1+vx2-vx3)*( vx1+vx2-vx3)-cu_sq));
- (D27F.f[DIR_MMP ])[ksw ] = c1o216*(c3o1*(-Mx -My +Mz)+Conc_F*(c1o1+c9o2*(-vx1-vx2+vx3)*(-vx1-vx2+vx3)-cu_sq));
- (D27F.f[DIR_PMP ])[ks ] = c1o216*(c3o1*( Mx -My +Mz)+Conc_F*(c1o1+c9o2*( vx1-vx2+vx3)*( vx1-vx2+vx3)-cu_sq));
- (D27F.f[DIR_MPM ])[kbw ] = c1o216*(c3o1*(-Mx +My -Mz)+Conc_F*(c1o1+c9o2*(-vx1+vx2-vx3)*(-vx1+vx2-vx3)-cu_sq));
- (D27F.f[DIR_PMM ])[kbs ] = c1o216*(c3o1*( Mx -My -Mz)+Conc_F*(c1o1+c9o2*( vx1-vx2-vx3)*( vx1-vx2-vx3)-cu_sq));
- (D27F.f[DIR_MPP ])[kw ] = c1o216*(c3o1*(-Mx +My +Mz)+Conc_F*(c1o1+c9o2*(-vx1+vx2+vx3)*(-vx1+vx2+vx3)-cu_sq));
+ (D27F.f[DIR_P00])[kzero] = c2o27* (c3o1*( Mx )+Conc_F*(c1o1+c9o2*( vx1 )*( vx1 )-cu_sq));
+ (D27F.f[DIR_M00])[kw ] = c2o27* (c3o1*(-Mx )+Conc_F*(c1o1+c9o2*(-vx1 )*(-vx1 )-cu_sq));
+ (D27F.f[DIR_0P0])[kzero] = c2o27* (c3o1*( My )+Conc_F*(c1o1+c9o2*( vx2 )*( vx2 )-cu_sq));
+ (D27F.f[DIR_0M0])[ks ] = c2o27* (c3o1*( -My )+Conc_F*(c1o1+c9o2*( -vx2 )*( -vx2 )-cu_sq));
+ (D27F.f[DIR_00P])[kzero] = c2o27* (c3o1*( Mz)+Conc_F*(c1o1+c9o2*( vx3)*( vx3)-cu_sq));
+ (D27F.f[DIR_00M])[kb ] = c2o27* (c3o1*( -Mz)+Conc_F*(c1o1+c9o2*( -vx3)*( -vx3)-cu_sq));
+ (D27F.f[DIR_PP0])[kzero] = c1o54* (c3o1*( Mx +My )+Conc_F*(c1o1+c9o2*( vx1+vx2 )*( vx1+vx2 )-cu_sq));
+ (D27F.f[DIR_MM0])[ksw ] = c1o54* (c3o1*(-Mx -My )+Conc_F*(c1o1+c9o2*(-vx1-vx2 )*(-vx1-vx2 )-cu_sq));
+ (D27F.f[DIR_PM0])[ks ] = c1o54* (c3o1*( Mx -My )+Conc_F*(c1o1+c9o2*( vx1-vx2 )*( vx1-vx2 )-cu_sq));
+ (D27F.f[DIR_MP0])[kw ] = c1o54* (c3o1*(-Mx +My )+Conc_F*(c1o1+c9o2*(-vx1+vx2 )*(-vx1+vx2 )-cu_sq));
+ (D27F.f[DIR_P0P])[kzero] = c1o54* (c3o1*( Mx +Mz)+Conc_F*(c1o1+c9o2*( vx1 +vx3)*( vx1 +vx3)-cu_sq));
+ (D27F.f[DIR_M0M])[kbw ] = c1o54* (c3o1*(-Mx -Mz)+Conc_F*(c1o1+c9o2*(-vx1 -vx3)*(-vx1 -vx3)-cu_sq));
+ (D27F.f[DIR_P0M])[kb ] = c1o54* (c3o1*( Mx -Mz)+Conc_F*(c1o1+c9o2*( vx1 -vx3)*( vx1 -vx3)-cu_sq));
+ (D27F.f[DIR_M0P])[kw ] = c1o54* (c3o1*(-Mx +Mz)+Conc_F*(c1o1+c9o2*(-vx1 +vx3)*(-vx1 +vx3)-cu_sq));
+ (D27F.f[DIR_0PP])[kzero] = c1o54* (c3o1*( My +Mz)+Conc_F*(c1o1+c9o2*( vx2+vx3)*( vx2+vx3)-cu_sq));
+ (D27F.f[DIR_0MM])[kbs ] = c1o54* (c3o1*( -My -Mz)+Conc_F*(c1o1+c9o2*( -vx2-vx3)*( -vx2-vx3)-cu_sq));
+ (D27F.f[DIR_0PM])[kb ] = c1o54* (c3o1*( My -Mz)+Conc_F*(c1o1+c9o2*( vx2-vx3)*( vx2-vx3)-cu_sq));
+ (D27F.f[DIR_0MP])[ks ] = c1o54* (c3o1*( -My +Mz)+Conc_F*(c1o1+c9o2*( -vx2+vx3)*( -vx2+vx3)-cu_sq));
+ (D27F.f[DIR_PPP])[kzero] = c1o216*(c3o1*( Mx +My +Mz)+Conc_F*(c1o1+c9o2*( vx1+vx2+vx3)*( vx1+vx2+vx3)-cu_sq));
+ (D27F.f[DIR_MMM])[kbsw ] = c1o216*(c3o1*(-Mx -My -Mz)+Conc_F*(c1o1+c9o2*(-vx1-vx2-vx3)*(-vx1-vx2-vx3)-cu_sq));
+ (D27F.f[DIR_PPM])[kb ] = c1o216*(c3o1*( Mx +My -Mz)+Conc_F*(c1o1+c9o2*( vx1+vx2-vx3)*( vx1+vx2-vx3)-cu_sq));
+ (D27F.f[DIR_MMP])[ksw ] = c1o216*(c3o1*(-Mx -My +Mz)+Conc_F*(c1o1+c9o2*(-vx1-vx2+vx3)*(-vx1-vx2+vx3)-cu_sq));
+ (D27F.f[DIR_PMP])[ks ] = c1o216*(c3o1*( Mx -My +Mz)+Conc_F*(c1o1+c9o2*( vx1-vx2+vx3)*( vx1-vx2+vx3)-cu_sq));
+ (D27F.f[DIR_MPM])[kbw ] = c1o216*(c3o1*(-Mx +My -Mz)+Conc_F*(c1o1+c9o2*(-vx1+vx2-vx3)*(-vx1+vx2-vx3)-cu_sq));
+ (D27F.f[DIR_PMM])[kbs ] = c1o216*(c3o1*( Mx -My -Mz)+Conc_F*(c1o1+c9o2*( vx1-vx2-vx3)*( vx1-vx2-vx3)-cu_sq));
+ (D27F.f[DIR_MPP])[kw ] = c1o216*(c3o1*(-Mx +My +Mz)+Conc_F*(c1o1+c9o2*(-vx1+vx2+vx3)*(-vx1+vx2+vx3)-cu_sq));
}
}
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
@@ -47287,8 +47287,8 @@ __global__ void scaleCFEff27(real* DC,
unsigned int* neighborFX,
unsigned int* neighborFY,
unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
+ unsigned long long numberOfLBnodesCoarse,
+ unsigned long long numberOfLBnodesFine,
bool isEvenTimestep,
unsigned int* posCSWB,
unsigned int* posFSWB,
@@ -47305,96 +47305,96 @@ __global__ void scaleCFEff27(real* DC,
real *feF, *fwF, *fnF, *fsF, *ftF, *fbF, *fneF, *fswF, *fseF, *fnwF, *fteF, *fbwF, *fbeF, *ftwF, *ftnF, *fbsF, *fbnF, *ftsF, *fzeroF,
*ftneF, *ftswF, *ftseF, *ftnwF, *fbneF, *fbswF, *fbseF, *fbnwF;
- feF = &DF[DIR_P00 *size_MatF];
- fwF = &DF[DIR_M00 *size_MatF];
- fnF = &DF[DIR_0P0 *size_MatF];
- fsF = &DF[DIR_0M0 *size_MatF];
- ftF = &DF[DIR_00P *size_MatF];
- fbF = &DF[DIR_00M *size_MatF];
- fneF = &DF[DIR_PP0 *size_MatF];
- fswF = &DF[DIR_MM0 *size_MatF];
- fseF = &DF[DIR_PM0 *size_MatF];
- fnwF = &DF[DIR_MP0 *size_MatF];
- fteF = &DF[DIR_P0P *size_MatF];
- fbwF = &DF[DIR_M0M *size_MatF];
- fbeF = &DF[DIR_P0M *size_MatF];
- ftwF = &DF[DIR_M0P *size_MatF];
- ftnF = &DF[DIR_0PP *size_MatF];
- fbsF = &DF[DIR_0MM *size_MatF];
- fbnF = &DF[DIR_0PM *size_MatF];
- ftsF = &DF[DIR_0MP *size_MatF];
- fzeroF = &DF[DIR_000*size_MatF];
- ftneF = &DF[DIR_PPP *size_MatF];
- ftswF = &DF[DIR_MMP *size_MatF];
- ftseF = &DF[DIR_PMP *size_MatF];
- ftnwF = &DF[DIR_MPP *size_MatF];
- fbneF = &DF[DIR_PPM *size_MatF];
- fbswF = &DF[DIR_MMM *size_MatF];
- fbseF = &DF[DIR_PMM *size_MatF];
- fbnwF = &DF[DIR_MPM *size_MatF];
+ feF = &DF[DIR_P00 * numberOfLBnodesFine];
+ fwF = &DF[DIR_M00 * numberOfLBnodesFine];
+ fnF = &DF[DIR_0P0 * numberOfLBnodesFine];
+ fsF = &DF[DIR_0M0 * numberOfLBnodesFine];
+ ftF = &DF[DIR_00P * numberOfLBnodesFine];
+ fbF = &DF[DIR_00M * numberOfLBnodesFine];
+ fneF = &DF[DIR_PP0 * numberOfLBnodesFine];
+ fswF = &DF[DIR_MM0 * numberOfLBnodesFine];
+ fseF = &DF[DIR_PM0 * numberOfLBnodesFine];
+ fnwF = &DF[DIR_MP0 * numberOfLBnodesFine];
+ fteF = &DF[DIR_P0P * numberOfLBnodesFine];
+ fbwF = &DF[DIR_M0M * numberOfLBnodesFine];
+ fbeF = &DF[DIR_P0M * numberOfLBnodesFine];
+ ftwF = &DF[DIR_M0P * numberOfLBnodesFine];
+ ftnF = &DF[DIR_0PP * numberOfLBnodesFine];
+ fbsF = &DF[DIR_0MM * numberOfLBnodesFine];
+ fbnF = &DF[DIR_0PM * numberOfLBnodesFine];
+ ftsF = &DF[DIR_0MP * numberOfLBnodesFine];
+ fzeroF = &DF[DIR_000 * numberOfLBnodesFine];
+ ftneF = &DF[DIR_PPP * numberOfLBnodesFine];
+ ftswF = &DF[DIR_MMP * numberOfLBnodesFine];
+ ftseF = &DF[DIR_PMP * numberOfLBnodesFine];
+ ftnwF = &DF[DIR_MPP * numberOfLBnodesFine];
+ fbneF = &DF[DIR_PPM * numberOfLBnodesFine];
+ fbswF = &DF[DIR_MMM * numberOfLBnodesFine];
+ fbseF = &DF[DIR_PMM * numberOfLBnodesFine];
+ fbnwF = &DF[DIR_MPM * numberOfLBnodesFine];
real *feC, *fwC, *fnC, *fsC, *ftC, *fbC, *fneC, *fswC, *fseC, *fnwC, *fteC, *fbwC, *fbeC, *ftwC, *ftnC, *fbsC, *fbnC, *ftsC, *fzeroC,
*ftneC, *ftswC, *ftseC, *ftnwC, *fbneC, *fbswC, *fbseC, *fbnwC;
if (isEvenTimestep==true)
{
- feC = &DC[DIR_P00 *size_MatC];
- fwC = &DC[DIR_M00 *size_MatC];
- fnC = &DC[DIR_0P0 *size_MatC];
- fsC = &DC[DIR_0M0 *size_MatC];
- ftC = &DC[DIR_00P *size_MatC];
- fbC = &DC[DIR_00M *size_MatC];
- fneC = &DC[DIR_PP0 *size_MatC];
- fswC = &DC[DIR_MM0 *size_MatC];
- fseC = &DC[DIR_PM0 *size_MatC];
- fnwC = &DC[DIR_MP0 *size_MatC];
- fteC = &DC[DIR_P0P *size_MatC];
- fbwC = &DC[DIR_M0M *size_MatC];
- fbeC = &DC[DIR_P0M *size_MatC];
- ftwC = &DC[DIR_M0P *size_MatC];
- ftnC = &DC[DIR_0PP *size_MatC];
- fbsC = &DC[DIR_0MM *size_MatC];
- fbnC = &DC[DIR_0PM *size_MatC];
- ftsC = &DC[DIR_0MP *size_MatC];
- fzeroC = &DC[DIR_000*size_MatC];
- ftneC = &DC[DIR_PPP *size_MatC];
- ftswC = &DC[DIR_MMP *size_MatC];
- ftseC = &DC[DIR_PMP *size_MatC];
- ftnwC = &DC[DIR_MPP *size_MatC];
- fbneC = &DC[DIR_PPM *size_MatC];
- fbswC = &DC[DIR_MMM *size_MatC];
- fbseC = &DC[DIR_PMM *size_MatC];
- fbnwC = &DC[DIR_MPM *size_MatC];
+ feC = &DC[DIR_P00 * numberOfLBnodesCoarse];
+ fwC = &DC[DIR_M00 * numberOfLBnodesCoarse];
+ fnC = &DC[DIR_0P0 * numberOfLBnodesCoarse];
+ fsC = &DC[DIR_0M0 * numberOfLBnodesCoarse];
+ ftC = &DC[DIR_00P * numberOfLBnodesCoarse];
+ fbC = &DC[DIR_00M * numberOfLBnodesCoarse];
+ fneC = &DC[DIR_PP0 * numberOfLBnodesCoarse];
+ fswC = &DC[DIR_MM0 * numberOfLBnodesCoarse];
+ fseC = &DC[DIR_PM0 * numberOfLBnodesCoarse];
+ fnwC = &DC[DIR_MP0 * numberOfLBnodesCoarse];
+ fteC = &DC[DIR_P0P * numberOfLBnodesCoarse];
+ fbwC = &DC[DIR_M0M * numberOfLBnodesCoarse];
+ fbeC = &DC[DIR_P0M * numberOfLBnodesCoarse];
+ ftwC = &DC[DIR_M0P * numberOfLBnodesCoarse];
+ ftnC = &DC[DIR_0PP * numberOfLBnodesCoarse];
+ fbsC = &DC[DIR_0MM * numberOfLBnodesCoarse];
+ fbnC = &DC[DIR_0PM * numberOfLBnodesCoarse];
+ ftsC = &DC[DIR_0MP * numberOfLBnodesCoarse];
+ fzeroC = &DC[DIR_000 * numberOfLBnodesCoarse];
+ ftneC = &DC[DIR_PPP * numberOfLBnodesCoarse];
+ ftswC = &DC[DIR_MMP * numberOfLBnodesCoarse];
+ ftseC = &DC[DIR_PMP * numberOfLBnodesCoarse];
+ ftnwC = &DC[DIR_MPP * numberOfLBnodesCoarse];
+ fbneC = &DC[DIR_PPM * numberOfLBnodesCoarse];
+ fbswC = &DC[DIR_MMM * numberOfLBnodesCoarse];
+ fbseC = &DC[DIR_PMM * numberOfLBnodesCoarse];
+ fbnwC = &DC[DIR_MPM * numberOfLBnodesCoarse];
}
else
{
- fwC = &DC[DIR_P00 *size_MatC];
- feC = &DC[DIR_M00 *size_MatC];
- fsC = &DC[DIR_0P0 *size_MatC];
- fnC = &DC[DIR_0M0 *size_MatC];
- fbC = &DC[DIR_00P *size_MatC];
- ftC = &DC[DIR_00M *size_MatC];
- fswC = &DC[DIR_PP0 *size_MatC];
- fneC = &DC[DIR_MM0 *size_MatC];
- fnwC = &DC[DIR_PM0 *size_MatC];
- fseC = &DC[DIR_MP0 *size_MatC];
- fbwC = &DC[DIR_P0P *size_MatC];
- fteC = &DC[DIR_M0M *size_MatC];
- ftwC = &DC[DIR_P0M *size_MatC];
- fbeC = &DC[DIR_M0P *size_MatC];
- fbsC = &DC[DIR_0PP *size_MatC];
- ftnC = &DC[DIR_0MM *size_MatC];
- ftsC = &DC[DIR_0PM *size_MatC];
- fbnC = &DC[DIR_0MP *size_MatC];
- fzeroC = &DC[DIR_000*size_MatC];
- fbswC = &DC[DIR_PPP *size_MatC];
- fbneC = &DC[DIR_MMP *size_MatC];
- fbnwC = &DC[DIR_PMP *size_MatC];
- fbseC = &DC[DIR_MPP *size_MatC];
- ftswC = &DC[DIR_PPM *size_MatC];
- ftneC = &DC[DIR_MMM *size_MatC];
- ftnwC = &DC[DIR_PMM *size_MatC];
- ftseC = &DC[DIR_MPM *size_MatC];
+ fwC = &DC[DIR_P00 * numberOfLBnodesCoarse];
+ feC = &DC[DIR_M00 * numberOfLBnodesCoarse];
+ fsC = &DC[DIR_0P0 * numberOfLBnodesCoarse];
+ fnC = &DC[DIR_0M0 * numberOfLBnodesCoarse];
+ fbC = &DC[DIR_00P * numberOfLBnodesCoarse];
+ ftC = &DC[DIR_00M * numberOfLBnodesCoarse];
+ fswC = &DC[DIR_PP0 * numberOfLBnodesCoarse];
+ fneC = &DC[DIR_MM0 * numberOfLBnodesCoarse];
+ fnwC = &DC[DIR_PM0 * numberOfLBnodesCoarse];
+ fseC = &DC[DIR_MP0 * numberOfLBnodesCoarse];
+ fbwC = &DC[DIR_P0P * numberOfLBnodesCoarse];
+ fteC = &DC[DIR_M0M * numberOfLBnodesCoarse];
+ ftwC = &DC[DIR_P0M * numberOfLBnodesCoarse];
+ fbeC = &DC[DIR_M0P * numberOfLBnodesCoarse];
+ fbsC = &DC[DIR_0PP * numberOfLBnodesCoarse];
+ ftnC = &DC[DIR_0MM * numberOfLBnodesCoarse];
+ ftsC = &DC[DIR_0PM * numberOfLBnodesCoarse];
+ fbnC = &DC[DIR_0MP * numberOfLBnodesCoarse];
+ fzeroC = &DC[DIR_000 * numberOfLBnodesCoarse];
+ fbswC = &DC[DIR_PPP * numberOfLBnodesCoarse];
+ fbneC = &DC[DIR_MMP * numberOfLBnodesCoarse];
+ fbnwC = &DC[DIR_PMP * numberOfLBnodesCoarse];
+ fbseC = &DC[DIR_MPP * numberOfLBnodesCoarse];
+ ftswC = &DC[DIR_PPM * numberOfLBnodesCoarse];
+ ftneC = &DC[DIR_MMM * numberOfLBnodesCoarse];
+ ftnwC = &DC[DIR_PMM * numberOfLBnodesCoarse];
+ ftseC = &DC[DIR_MPM * numberOfLBnodesCoarse];
}
////////////////////////////////////////////////////////////////////////////////
@@ -48997,8 +48997,8 @@ __global__ void scaleCF27(real* DC,
unsigned int* neighborFX,
unsigned int* neighborFY,
unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
+ unsigned long long numberOfLBnodesCoarse,
+ unsigned long long numberOfLBnodesFine,
bool isEvenTimestep,
unsigned int* posCSWB,
unsigned int* posFSWB,
@@ -49014,96 +49014,96 @@ __global__ void scaleCF27(real* DC,
real *feF, *fwF, *fnF, *fsF, *ftF, *fbF, *fneF, *fswF, *fseF, *fnwF, *fteF, *fbwF, *fbeF, *ftwF, *ftnF, *fbsF, *fbnF, *ftsF, *fzeroF,
*ftneF, *ftswF, *ftseF, *ftnwF, *fbneF, *fbswF, *fbseF, *fbnwF;
- feF = &DF[DIR_P00 *size_MatF];
- fwF = &DF[DIR_M00 *size_MatF];
- fnF = &DF[DIR_0P0 *size_MatF];
- fsF = &DF[DIR_0M0 *size_MatF];
- ftF = &DF[DIR_00P *size_MatF];
- fbF = &DF[DIR_00M *size_MatF];
- fneF = &DF[DIR_PP0 *size_MatF];
- fswF = &DF[DIR_MM0 *size_MatF];
- fseF = &DF[DIR_PM0 *size_MatF];
- fnwF = &DF[DIR_MP0 *size_MatF];
- fteF = &DF[DIR_P0P *size_MatF];
- fbwF = &DF[DIR_M0M *size_MatF];
- fbeF = &DF[DIR_P0M *size_MatF];
- ftwF = &DF[DIR_M0P *size_MatF];
- ftnF = &DF[DIR_0PP *size_MatF];
- fbsF = &DF[DIR_0MM *size_MatF];
- fbnF = &DF[DIR_0PM *size_MatF];
- ftsF = &DF[DIR_0MP *size_MatF];
- fzeroF = &DF[DIR_000*size_MatF];
- ftneF = &DF[DIR_PPP *size_MatF];
- ftswF = &DF[DIR_MMP *size_MatF];
- ftseF = &DF[DIR_PMP *size_MatF];
- ftnwF = &DF[DIR_MPP *size_MatF];
- fbneF = &DF[DIR_PPM *size_MatF];
- fbswF = &DF[DIR_MMM *size_MatF];
- fbseF = &DF[DIR_PMM *size_MatF];
- fbnwF = &DF[DIR_MPM *size_MatF];
+ feF = &DF[DIR_P00 * numberOfLBnodesFine];
+ fwF = &DF[DIR_M00 * numberOfLBnodesFine];
+ fnF = &DF[DIR_0P0 * numberOfLBnodesFine];
+ fsF = &DF[DIR_0M0 * numberOfLBnodesFine];
+ ftF = &DF[DIR_00P * numberOfLBnodesFine];
+ fbF = &DF[DIR_00M * numberOfLBnodesFine];
+ fneF = &DF[DIR_PP0 * numberOfLBnodesFine];
+ fswF = &DF[DIR_MM0 * numberOfLBnodesFine];
+ fseF = &DF[DIR_PM0 * numberOfLBnodesFine];
+ fnwF = &DF[DIR_MP0 * numberOfLBnodesFine];
+ fteF = &DF[DIR_P0P * numberOfLBnodesFine];
+ fbwF = &DF[DIR_M0M * numberOfLBnodesFine];
+ fbeF = &DF[DIR_P0M * numberOfLBnodesFine];
+ ftwF = &DF[DIR_M0P * numberOfLBnodesFine];
+ ftnF = &DF[DIR_0PP * numberOfLBnodesFine];
+ fbsF = &DF[DIR_0MM * numberOfLBnodesFine];
+ fbnF = &DF[DIR_0PM * numberOfLBnodesFine];
+ ftsF = &DF[DIR_0MP * numberOfLBnodesFine];
+ fzeroF = &DF[DIR_000 * numberOfLBnodesFine];
+ ftneF = &DF[DIR_PPP * numberOfLBnodesFine];
+ ftswF = &DF[DIR_MMP * numberOfLBnodesFine];
+ ftseF = &DF[DIR_PMP * numberOfLBnodesFine];
+ ftnwF = &DF[DIR_MPP * numberOfLBnodesFine];
+ fbneF = &DF[DIR_PPM * numberOfLBnodesFine];
+ fbswF = &DF[DIR_MMM * numberOfLBnodesFine];
+ fbseF = &DF[DIR_PMM * numberOfLBnodesFine];
+ fbnwF = &DF[DIR_MPM * numberOfLBnodesFine];
real *feC, *fwC, *fnC, *fsC, *ftC, *fbC, *fneC, *fswC, *fseC, *fnwC, *fteC, *fbwC, *fbeC, *ftwC, *ftnC, *fbsC, *fbnC, *ftsC, *fzeroC,
*ftneC, *ftswC, *ftseC, *ftnwC, *fbneC, *fbswC, *fbseC, *fbnwC;
if (isEvenTimestep==true)
{
- feC = &DC[DIR_P00 *size_MatC];
- fwC = &DC[DIR_M00 *size_MatC];
- fnC = &DC[DIR_0P0 *size_MatC];
- fsC = &DC[DIR_0M0 *size_MatC];
- ftC = &DC[DIR_00P *size_MatC];
- fbC = &DC[DIR_00M *size_MatC];
- fneC = &DC[DIR_PP0 *size_MatC];
- fswC = &DC[DIR_MM0 *size_MatC];
- fseC = &DC[DIR_PM0 *size_MatC];
- fnwC = &DC[DIR_MP0 *size_MatC];
- fteC = &DC[DIR_P0P *size_MatC];
- fbwC = &DC[DIR_M0M *size_MatC];
- fbeC = &DC[DIR_P0M *size_MatC];
- ftwC = &DC[DIR_M0P *size_MatC];
- ftnC = &DC[DIR_0PP *size_MatC];
- fbsC = &DC[DIR_0MM *size_MatC];
- fbnC = &DC[DIR_0PM *size_MatC];
- ftsC = &DC[DIR_0MP *size_MatC];
- fzeroC = &DC[DIR_000*size_MatC];
- ftneC = &DC[DIR_PPP *size_MatC];
- ftswC = &DC[DIR_MMP *size_MatC];
- ftseC = &DC[DIR_PMP *size_MatC];
- ftnwC = &DC[DIR_MPP *size_MatC];
- fbneC = &DC[DIR_PPM *size_MatC];
- fbswC = &DC[DIR_MMM *size_MatC];
- fbseC = &DC[DIR_PMM *size_MatC];
- fbnwC = &DC[DIR_MPM *size_MatC];
+ feC = &DC[DIR_P00 * numberOfLBnodesCoarse];
+ fwC = &DC[DIR_M00 * numberOfLBnodesCoarse];
+ fnC = &DC[DIR_0P0 * numberOfLBnodesCoarse];
+ fsC = &DC[DIR_0M0 * numberOfLBnodesCoarse];
+ ftC = &DC[DIR_00P * numberOfLBnodesCoarse];
+ fbC = &DC[DIR_00M * numberOfLBnodesCoarse];
+ fneC = &DC[DIR_PP0 * numberOfLBnodesCoarse];
+ fswC = &DC[DIR_MM0 * numberOfLBnodesCoarse];
+ fseC = &DC[DIR_PM0 * numberOfLBnodesCoarse];
+ fnwC = &DC[DIR_MP0 * numberOfLBnodesCoarse];
+ fteC = &DC[DIR_P0P * numberOfLBnodesCoarse];
+ fbwC = &DC[DIR_M0M * numberOfLBnodesCoarse];
+ fbeC = &DC[DIR_P0M * numberOfLBnodesCoarse];
+ ftwC = &DC[DIR_M0P * numberOfLBnodesCoarse];
+ ftnC = &DC[DIR_0PP * numberOfLBnodesCoarse];
+ fbsC = &DC[DIR_0MM * numberOfLBnodesCoarse];
+ fbnC = &DC[DIR_0PM * numberOfLBnodesCoarse];
+ ftsC = &DC[DIR_0MP * numberOfLBnodesCoarse];
+ fzeroC = &DC[DIR_000 * numberOfLBnodesCoarse];
+ ftneC = &DC[DIR_PPP * numberOfLBnodesCoarse];
+ ftswC = &DC[DIR_MMP * numberOfLBnodesCoarse];
+ ftseC = &DC[DIR_PMP * numberOfLBnodesCoarse];
+ ftnwC = &DC[DIR_MPP * numberOfLBnodesCoarse];
+ fbneC = &DC[DIR_PPM * numberOfLBnodesCoarse];
+ fbswC = &DC[DIR_MMM * numberOfLBnodesCoarse];
+ fbseC = &DC[DIR_PMM * numberOfLBnodesCoarse];
+ fbnwC = &DC[DIR_MPM * numberOfLBnodesCoarse];
}
else
{
- fwC = &DC[DIR_P00 *size_MatC];
- feC = &DC[DIR_M00 *size_MatC];
- fsC = &DC[DIR_0P0 *size_MatC];
- fnC = &DC[DIR_0M0 *size_MatC];
- fbC = &DC[DIR_00P *size_MatC];
- ftC = &DC[DIR_00M *size_MatC];
- fswC = &DC[DIR_PP0 *size_MatC];
- fneC = &DC[DIR_MM0 *size_MatC];
- fnwC = &DC[DIR_PM0 *size_MatC];
- fseC = &DC[DIR_MP0 *size_MatC];
- fbwC = &DC[DIR_P0P *size_MatC];
- fteC = &DC[DIR_M0M *size_MatC];
- ftwC = &DC[DIR_P0M *size_MatC];
- fbeC = &DC[DIR_M0P *size_MatC];
- fbsC = &DC[DIR_0PP *size_MatC];
- ftnC = &DC[DIR_0MM *size_MatC];
- ftsC = &DC[DIR_0PM *size_MatC];
- fbnC = &DC[DIR_0MP *size_MatC];
- fzeroC = &DC[DIR_000*size_MatC];
- fbswC = &DC[DIR_PPP *size_MatC];
- fbneC = &DC[DIR_MMP *size_MatC];
- fbnwC = &DC[DIR_PMP *size_MatC];
- fbseC = &DC[DIR_MPP *size_MatC];
- ftswC = &DC[DIR_PPM *size_MatC];
- ftneC = &DC[DIR_MMM *size_MatC];
- ftnwC = &DC[DIR_PMM *size_MatC];
- ftseC = &DC[DIR_MPM *size_MatC];
+ fwC = &DC[DIR_P00 * numberOfLBnodesCoarse];
+ feC = &DC[DIR_M00 * numberOfLBnodesCoarse];
+ fsC = &DC[DIR_0P0 * numberOfLBnodesCoarse];
+ fnC = &DC[DIR_0M0 * numberOfLBnodesCoarse];
+ fbC = &DC[DIR_00P * numberOfLBnodesCoarse];
+ ftC = &DC[DIR_00M * numberOfLBnodesCoarse];
+ fswC = &DC[DIR_PP0 * numberOfLBnodesCoarse];
+ fneC = &DC[DIR_MM0 * numberOfLBnodesCoarse];
+ fnwC = &DC[DIR_PM0 * numberOfLBnodesCoarse];
+ fseC = &DC[DIR_MP0 * numberOfLBnodesCoarse];
+ fbwC = &DC[DIR_P0P * numberOfLBnodesCoarse];
+ fteC = &DC[DIR_M0M * numberOfLBnodesCoarse];
+ ftwC = &DC[DIR_P0M * numberOfLBnodesCoarse];
+ fbeC = &DC[DIR_M0P * numberOfLBnodesCoarse];
+ fbsC = &DC[DIR_0PP * numberOfLBnodesCoarse];
+ ftnC = &DC[DIR_0MM * numberOfLBnodesCoarse];
+ ftsC = &DC[DIR_0PM * numberOfLBnodesCoarse];
+ fbnC = &DC[DIR_0MP * numberOfLBnodesCoarse];
+ fzeroC = &DC[DIR_000 * numberOfLBnodesCoarse];
+ fbswC = &DC[DIR_PPP * numberOfLBnodesCoarse];
+ fbneC = &DC[DIR_MMP * numberOfLBnodesCoarse];
+ fbnwC = &DC[DIR_PMP * numberOfLBnodesCoarse];
+ fbseC = &DC[DIR_MPP * numberOfLBnodesCoarse];
+ ftswC = &DC[DIR_PPM * numberOfLBnodesCoarse];
+ ftneC = &DC[DIR_MMM * numberOfLBnodesCoarse];
+ ftnwC = &DC[DIR_PMM * numberOfLBnodesCoarse];
+ ftseC = &DC[DIR_MPM * numberOfLBnodesCoarse];
}
////////////////////////////////////////////////////////////////////////////////
diff --git a/src/gpu/VirtualFluids_GPU/GPU/GridScaling/ScaleCF_F3_27.cu b/src/gpu/VirtualFluids_GPU/GPU/GridScaling/ScaleCF_F3_27.cu
index cb8bd2a322cc9176cd0aa31625ee386e1f62d63d..386493280fd71fff93c117483e754a248bb0830d 100644
--- a/src/gpu/VirtualFluids_GPU/GPU/GridScaling/ScaleCF_F3_27.cu
+++ b/src/gpu/VirtualFluids_GPU/GPU/GridScaling/ScaleCF_F3_27.cu
@@ -23,8 +23,8 @@ __global__ void scaleCF_comp_D3Q27F3_2018(real* DC,
unsigned int* neighborFX,
unsigned int* neighborFY,
unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
+ unsigned long long numberOfLBnodesCoarse,
+ unsigned long long numberOfLBnodesFine,
bool isEvenTimestep,
unsigned int* posCSWB,
unsigned int* posFSWB,
@@ -44,33 +44,33 @@ __global__ void scaleCF_comp_D3Q27F3_2018(real* DC,
*f000dest, *fMMMdest, *fMMPdest, *fMPPdest, *fMPMdest, *fPPMdest, *fPPPdest, *fPMPdest, *fPMMdest;
- fP00dest = &DF[DIR_P00 *size_MatF];
- fM00dest = &DF[DIR_M00 *size_MatF];
- f0P0dest = &DF[DIR_0P0 *size_MatF];
- f0M0dest = &DF[DIR_0M0 *size_MatF];
- f00Pdest = &DF[DIR_00P *size_MatF];
- f00Mdest = &DF[DIR_00M *size_MatF];
- fPP0dest = &DF[DIR_PP0 *size_MatF];
- fMM0dest = &DF[DIR_MM0 *size_MatF];
- fPM0dest = &DF[DIR_PM0 *size_MatF];
- fMP0dest = &DF[DIR_MP0 *size_MatF];
- fP0Pdest = &DF[DIR_P0P *size_MatF];
- fM0Mdest = &DF[DIR_M0M *size_MatF];
- fP0Mdest = &DF[DIR_P0M *size_MatF];
- fM0Pdest = &DF[DIR_M0P *size_MatF];
- f0PPdest = &DF[DIR_0PP *size_MatF];
- f0MMdest = &DF[DIR_0MM *size_MatF];
- f0PMdest = &DF[DIR_0PM *size_MatF];
- f0MPdest = &DF[DIR_0MP *size_MatF];
- f000dest = &DF[DIR_000*size_MatF];
- fMMMdest = &DF[DIR_MMM *size_MatF];
- fMMPdest = &DF[DIR_MMP *size_MatF];
- fMPPdest = &DF[DIR_MPP *size_MatF];
- fMPMdest = &DF[DIR_MPM *size_MatF];
- fPPMdest = &DF[DIR_PPM *size_MatF];
- fPPPdest = &DF[DIR_PPP *size_MatF];
- fPMPdest = &DF[DIR_PMP *size_MatF];
- fPMMdest = &DF[DIR_PMM *size_MatF];
+ fP00dest = &DF[DIR_P00 * numberOfLBnodesFine];
+ fM00dest = &DF[DIR_M00 * numberOfLBnodesFine];
+ f0P0dest = &DF[DIR_0P0 * numberOfLBnodesFine];
+ f0M0dest = &DF[DIR_0M0 * numberOfLBnodesFine];
+ f00Pdest = &DF[DIR_00P * numberOfLBnodesFine];
+ f00Mdest = &DF[DIR_00M * numberOfLBnodesFine];
+ fPP0dest = &DF[DIR_PP0 * numberOfLBnodesFine];
+ fMM0dest = &DF[DIR_MM0 * numberOfLBnodesFine];
+ fPM0dest = &DF[DIR_PM0 * numberOfLBnodesFine];
+ fMP0dest = &DF[DIR_MP0 * numberOfLBnodesFine];
+ fP0Pdest = &DF[DIR_P0P * numberOfLBnodesFine];
+ fM0Mdest = &DF[DIR_M0M * numberOfLBnodesFine];
+ fP0Mdest = &DF[DIR_P0M * numberOfLBnodesFine];
+ fM0Pdest = &DF[DIR_M0P * numberOfLBnodesFine];
+ f0PPdest = &DF[DIR_0PP * numberOfLBnodesFine];
+ f0MMdest = &DF[DIR_0MM * numberOfLBnodesFine];
+ f0PMdest = &DF[DIR_0PM * numberOfLBnodesFine];
+ f0MPdest = &DF[DIR_0MP * numberOfLBnodesFine];
+ f000dest = &DF[DIR_000 * numberOfLBnodesFine];
+ fMMMdest = &DF[DIR_MMM * numberOfLBnodesFine];
+ fMMPdest = &DF[DIR_MMP * numberOfLBnodesFine];
+ fMPPdest = &DF[DIR_MPP * numberOfLBnodesFine];
+ fMPMdest = &DF[DIR_MPM * numberOfLBnodesFine];
+ fPPMdest = &DF[DIR_PPM * numberOfLBnodesFine];
+ fPPPdest = &DF[DIR_PPP * numberOfLBnodesFine];
+ fPMPdest = &DF[DIR_PMP * numberOfLBnodesFine];
+ fPMMdest = &DF[DIR_PMM * numberOfLBnodesFine];
real
*fP00source, *fM00source, *f0P0source, *f0M0source, *f00Psource, *f00Msource, *fPP0source, *fMM0source, *fPM0source,
@@ -79,72 +79,72 @@ __global__ void scaleCF_comp_D3Q27F3_2018(real* DC,
if (isEvenTimestep == true)
{
- fP00source = &DC[DIR_P00 *size_MatC];
- fM00source = &DC[DIR_M00 *size_MatC];
- f0P0source = &DC[DIR_0P0 *size_MatC];
- f0M0source = &DC[DIR_0M0 *size_MatC];
- f00Psource = &DC[DIR_00P *size_MatC];
- f00Msource = &DC[DIR_00M *size_MatC];
- fPP0source = &DC[DIR_PP0 *size_MatC];
- fMM0source = &DC[DIR_MM0 *size_MatC];
- fPM0source = &DC[DIR_PM0 *size_MatC];
- fMP0source = &DC[DIR_MP0 *size_MatC];
- fP0Psource = &DC[DIR_P0P *size_MatC];
- fM0Msource = &DC[DIR_M0M *size_MatC];
- fP0Msource = &DC[DIR_P0M *size_MatC];
- fM0Psource = &DC[DIR_M0P *size_MatC];
- f0PPsource = &DC[DIR_0PP *size_MatC];
- f0MMsource = &DC[DIR_0MM *size_MatC];
- f0PMsource = &DC[DIR_0PM *size_MatC];
- f0MPsource = &DC[DIR_0MP *size_MatC];
- f000source = &DC[DIR_000*size_MatC];
- fMMMsource = &DC[DIR_MMM *size_MatC];
- fMMPsource = &DC[DIR_MMP *size_MatC];
- fMPPsource = &DC[DIR_MPP *size_MatC];
- fMPMsource = &DC[DIR_MPM *size_MatC];
- fPPMsource = &DC[DIR_PPM *size_MatC];
- fPPPsource = &DC[DIR_PPP *size_MatC];
- fPMPsource = &DC[DIR_PMP *size_MatC];
- fPMMsource = &DC[DIR_PMM *size_MatC];
+ fP00source = &DC[DIR_P00 * numberOfLBnodesCoarse];
+ fM00source = &DC[DIR_M00 * numberOfLBnodesCoarse];
+ f0P0source = &DC[DIR_0P0 * numberOfLBnodesCoarse];
+ f0M0source = &DC[DIR_0M0 * numberOfLBnodesCoarse];
+ f00Psource = &DC[DIR_00P * numberOfLBnodesCoarse];
+ f00Msource = &DC[DIR_00M * numberOfLBnodesCoarse];
+ fPP0source = &DC[DIR_PP0 * numberOfLBnodesCoarse];
+ fMM0source = &DC[DIR_MM0 * numberOfLBnodesCoarse];
+ fPM0source = &DC[DIR_PM0 * numberOfLBnodesCoarse];
+ fMP0source = &DC[DIR_MP0 * numberOfLBnodesCoarse];
+ fP0Psource = &DC[DIR_P0P * numberOfLBnodesCoarse];
+ fM0Msource = &DC[DIR_M0M * numberOfLBnodesCoarse];
+ fP0Msource = &DC[DIR_P0M * numberOfLBnodesCoarse];
+ fM0Psource = &DC[DIR_M0P * numberOfLBnodesCoarse];
+ f0PPsource = &DC[DIR_0PP * numberOfLBnodesCoarse];
+ f0MMsource = &DC[DIR_0MM * numberOfLBnodesCoarse];
+ f0PMsource = &DC[DIR_0PM * numberOfLBnodesCoarse];
+ f0MPsource = &DC[DIR_0MP * numberOfLBnodesCoarse];
+ f000source = &DC[DIR_000 * numberOfLBnodesCoarse];
+ fMMMsource = &DC[DIR_MMM * numberOfLBnodesCoarse];
+ fMMPsource = &DC[DIR_MMP * numberOfLBnodesCoarse];
+ fMPPsource = &DC[DIR_MPP * numberOfLBnodesCoarse];
+ fMPMsource = &DC[DIR_MPM * numberOfLBnodesCoarse];
+ fPPMsource = &DC[DIR_PPM * numberOfLBnodesCoarse];
+ fPPPsource = &DC[DIR_PPP * numberOfLBnodesCoarse];
+ fPMPsource = &DC[DIR_PMP * numberOfLBnodesCoarse];
+ fPMMsource = &DC[DIR_PMM * numberOfLBnodesCoarse];
}
else
{
- fP00source = &DC[DIR_M00 *size_MatC];
- fM00source = &DC[DIR_P00 *size_MatC];
- f0P0source = &DC[DIR_0M0 *size_MatC];
- f0M0source = &DC[DIR_0P0 *size_MatC];
- f00Psource = &DC[DIR_00M *size_MatC];
- f00Msource = &DC[DIR_00P *size_MatC];
- fPP0source = &DC[DIR_MM0 *size_MatC];
- fMM0source = &DC[DIR_PP0 *size_MatC];
- fPM0source = &DC[DIR_MP0 *size_MatC];
- fMP0source = &DC[DIR_PM0 *size_MatC];
- fP0Psource = &DC[DIR_M0M *size_MatC];
- fM0Msource = &DC[DIR_P0P *size_MatC];
- fP0Msource = &DC[DIR_M0P *size_MatC];
- fM0Psource = &DC[DIR_P0M *size_MatC];
- f0PPsource = &DC[DIR_0MM *size_MatC];
- f0MMsource = &DC[DIR_0PP *size_MatC];
- f0PMsource = &DC[DIR_0MP *size_MatC];
- f0MPsource = &DC[DIR_0PM *size_MatC];
- f000source = &DC[DIR_000*size_MatC];
- fMMMsource = &DC[DIR_PPP *size_MatC];
- fMMPsource = &DC[DIR_PPM *size_MatC];
- fMPPsource = &DC[DIR_PMM *size_MatC];
- fMPMsource = &DC[DIR_PMP *size_MatC];
- fPPMsource = &DC[DIR_MMP *size_MatC];
- fPPPsource = &DC[DIR_MMM *size_MatC];
- fPMPsource = &DC[DIR_MPM *size_MatC];
- fPMMsource = &DC[DIR_MPP *size_MatC];
+ fP00source = &DC[DIR_M00 * numberOfLBnodesCoarse];
+ fM00source = &DC[DIR_P00 * numberOfLBnodesCoarse];
+ f0P0source = &DC[DIR_0M0 * numberOfLBnodesCoarse];
+ f0M0source = &DC[DIR_0P0 * numberOfLBnodesCoarse];
+ f00Psource = &DC[DIR_00M * numberOfLBnodesCoarse];
+ f00Msource = &DC[DIR_00P * numberOfLBnodesCoarse];
+ fPP0source = &DC[DIR_MM0 * numberOfLBnodesCoarse];
+ fMM0source = &DC[DIR_PP0 * numberOfLBnodesCoarse];
+ fPM0source = &DC[DIR_MP0 * numberOfLBnodesCoarse];
+ fMP0source = &DC[DIR_PM0 * numberOfLBnodesCoarse];
+ fP0Psource = &DC[DIR_M0M * numberOfLBnodesCoarse];
+ fM0Msource = &DC[DIR_P0P * numberOfLBnodesCoarse];
+ fP0Msource = &DC[DIR_M0P * numberOfLBnodesCoarse];
+ fM0Psource = &DC[DIR_P0M * numberOfLBnodesCoarse];
+ f0PPsource = &DC[DIR_0MM * numberOfLBnodesCoarse];
+ f0MMsource = &DC[DIR_0PP * numberOfLBnodesCoarse];
+ f0PMsource = &DC[DIR_0MP * numberOfLBnodesCoarse];
+ f0MPsource = &DC[DIR_0PM * numberOfLBnodesCoarse];
+ f000source = &DC[DIR_000 * numberOfLBnodesCoarse];
+ fMMMsource = &DC[DIR_PPP * numberOfLBnodesCoarse];
+ fMMPsource = &DC[DIR_PPM * numberOfLBnodesCoarse];
+ fMPPsource = &DC[DIR_PMM * numberOfLBnodesCoarse];
+ fMPMsource = &DC[DIR_PMP * numberOfLBnodesCoarse];
+ fPPMsource = &DC[DIR_MMP * numberOfLBnodesCoarse];
+ fPPPsource = &DC[DIR_MMM * numberOfLBnodesCoarse];
+ fPMPsource = &DC[DIR_MPM * numberOfLBnodesCoarse];
+ fPMMsource = &DC[DIR_MPP * numberOfLBnodesCoarse];
}
Distributions6 G;
- G.g[DIR_P00] = &G6[DIR_P00 *size_MatF];
- G.g[DIR_M00] = &G6[DIR_M00 *size_MatF];
- G.g[DIR_0P0] = &G6[DIR_0P0 *size_MatF];
- G.g[DIR_0M0] = &G6[DIR_0M0 *size_MatF];
- G.g[DIR_00P] = &G6[DIR_00P *size_MatF];
- G.g[DIR_00M] = &G6[DIR_00M *size_MatF];
+ G.g[DIR_P00] = &G6[DIR_P00 * numberOfLBnodesFine];
+ G.g[DIR_M00] = &G6[DIR_M00 * numberOfLBnodesFine];
+ G.g[DIR_0P0] = &G6[DIR_0P0 * numberOfLBnodesFine];
+ G.g[DIR_0M0] = &G6[DIR_0M0 * numberOfLBnodesFine];
+ G.g[DIR_00P] = &G6[DIR_00P * numberOfLBnodesFine];
+ G.g[DIR_00M] = &G6[DIR_00M * numberOfLBnodesFine];
////////////////////////////////////////////////////////////////////////////////
const unsigned ix = threadIdx.x; // Globaler x-Index
@@ -4370,8 +4370,8 @@ __global__ void scaleCF_comp_D3Q27F3( real* DC,
unsigned int* neighborFX,
unsigned int* neighborFY,
unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
+ unsigned long long numberOfLBnodesCoarse,
+ unsigned long long numberOfLBnodesFine,
bool isEvenTimestep,
unsigned int* posCSWB,
unsigned int* posFSWB,
@@ -4391,33 +4391,33 @@ __global__ void scaleCF_comp_D3Q27F3( real* DC,
*f000dest, *fMMMdest, *fMMPdest, *fMPPdest, *fMPMdest, *fPPMdest, *fPPPdest, *fPMPdest, *fPMMdest;
- fP00dest = &DF[DIR_P00 *size_MatF];
- fM00dest = &DF[DIR_M00 *size_MatF];
- f0P0dest = &DF[DIR_0P0 *size_MatF];
- f0M0dest = &DF[DIR_0M0 *size_MatF];
- f00Pdest = &DF[DIR_00P *size_MatF];
- f00Mdest = &DF[DIR_00M *size_MatF];
- fPP0dest = &DF[DIR_PP0 *size_MatF];
- fMM0dest = &DF[DIR_MM0 *size_MatF];
- fPM0dest = &DF[DIR_PM0 *size_MatF];
- fMP0dest = &DF[DIR_MP0 *size_MatF];
- fP0Pdest = &DF[DIR_P0P *size_MatF];
- fM0Mdest = &DF[DIR_M0M *size_MatF];
- fP0Mdest = &DF[DIR_P0M *size_MatF];
- fM0Pdest = &DF[DIR_M0P *size_MatF];
- f0PPdest = &DF[DIR_0PP *size_MatF];
- f0MMdest = &DF[DIR_0MM *size_MatF];
- f0PMdest = &DF[DIR_0PM *size_MatF];
- f0MPdest = &DF[DIR_0MP *size_MatF];
- f000dest = &DF[DIR_000*size_MatF];
- fMMMdest = &DF[DIR_MMM *size_MatF];
- fMMPdest = &DF[DIR_MMP *size_MatF];
- fMPPdest = &DF[DIR_MPP *size_MatF];
- fMPMdest = &DF[DIR_MPM *size_MatF];
- fPPMdest = &DF[DIR_PPM *size_MatF];
- fPPPdest = &DF[DIR_PPP *size_MatF];
- fPMPdest = &DF[DIR_PMP *size_MatF];
- fPMMdest = &DF[DIR_PMM *size_MatF];
+ fP00dest = &DF[DIR_P00 * numberOfLBnodesFine];
+ fM00dest = &DF[DIR_M00 * numberOfLBnodesFine];
+ f0P0dest = &DF[DIR_0P0 * numberOfLBnodesFine];
+ f0M0dest = &DF[DIR_0M0 * numberOfLBnodesFine];
+ f00Pdest = &DF[DIR_00P * numberOfLBnodesFine];
+ f00Mdest = &DF[DIR_00M * numberOfLBnodesFine];
+ fPP0dest = &DF[DIR_PP0 * numberOfLBnodesFine];
+ fMM0dest = &DF[DIR_MM0 * numberOfLBnodesFine];
+ fPM0dest = &DF[DIR_PM0 * numberOfLBnodesFine];
+ fMP0dest = &DF[DIR_MP0 * numberOfLBnodesFine];
+ fP0Pdest = &DF[DIR_P0P * numberOfLBnodesFine];
+ fM0Mdest = &DF[DIR_M0M * numberOfLBnodesFine];
+ fP0Mdest = &DF[DIR_P0M * numberOfLBnodesFine];
+ fM0Pdest = &DF[DIR_M0P * numberOfLBnodesFine];
+ f0PPdest = &DF[DIR_0PP * numberOfLBnodesFine];
+ f0MMdest = &DF[DIR_0MM * numberOfLBnodesFine];
+ f0PMdest = &DF[DIR_0PM * numberOfLBnodesFine];
+ f0MPdest = &DF[DIR_0MP * numberOfLBnodesFine];
+ f000dest = &DF[DIR_000 * numberOfLBnodesFine];
+ fMMMdest = &DF[DIR_MMM * numberOfLBnodesFine];
+ fMMPdest = &DF[DIR_MMP * numberOfLBnodesFine];
+ fMPPdest = &DF[DIR_MPP * numberOfLBnodesFine];
+ fMPMdest = &DF[DIR_MPM * numberOfLBnodesFine];
+ fPPMdest = &DF[DIR_PPM * numberOfLBnodesFine];
+ fPPPdest = &DF[DIR_PPP * numberOfLBnodesFine];
+ fPMPdest = &DF[DIR_PMP * numberOfLBnodesFine];
+ fPMMdest = &DF[DIR_PMM * numberOfLBnodesFine];
real
*fP00source, *fM00source, *f0P0source, *f0M0source, *f00Psource, *f00Msource, *fPP0source, *fMM0source, *fPM0source,
@@ -4426,72 +4426,72 @@ __global__ void scaleCF_comp_D3Q27F3( real* DC,
if (isEvenTimestep == true)
{
- fP00source = &DC[DIR_P00 *size_MatC];
- fM00source = &DC[DIR_M00 *size_MatC];
- f0P0source = &DC[DIR_0P0 *size_MatC];
- f0M0source = &DC[DIR_0M0 *size_MatC];
- f00Psource = &DC[DIR_00P *size_MatC];
- f00Msource = &DC[DIR_00M *size_MatC];
- fPP0source = &DC[DIR_PP0 *size_MatC];
- fMM0source = &DC[DIR_MM0 *size_MatC];
- fPM0source = &DC[DIR_PM0 *size_MatC];
- fMP0source = &DC[DIR_MP0 *size_MatC];
- fP0Psource = &DC[DIR_P0P *size_MatC];
- fM0Msource = &DC[DIR_M0M *size_MatC];
- fP0Msource = &DC[DIR_P0M *size_MatC];
- fM0Psource = &DC[DIR_M0P *size_MatC];
- f0PPsource = &DC[DIR_0PP *size_MatC];
- f0MMsource = &DC[DIR_0MM *size_MatC];
- f0PMsource = &DC[DIR_0PM *size_MatC];
- f0MPsource = &DC[DIR_0MP *size_MatC];
- f000source = &DC[DIR_000*size_MatC];
- fMMMsource = &DC[DIR_MMM *size_MatC];
- fMMPsource = &DC[DIR_MMP *size_MatC];
- fMPPsource = &DC[DIR_MPP *size_MatC];
- fMPMsource = &DC[DIR_MPM *size_MatC];
- fPPMsource = &DC[DIR_PPM *size_MatC];
- fPPPsource = &DC[DIR_PPP *size_MatC];
- fPMPsource = &DC[DIR_PMP *size_MatC];
- fPMMsource = &DC[DIR_PMM *size_MatC];
+ fP00source = &DC[DIR_P00 * numberOfLBnodesCoarse];
+ fM00source = &DC[DIR_M00 * numberOfLBnodesCoarse];
+ f0P0source = &DC[DIR_0P0 * numberOfLBnodesCoarse];
+ f0M0source = &DC[DIR_0M0 * numberOfLBnodesCoarse];
+ f00Psource = &DC[DIR_00P * numberOfLBnodesCoarse];
+ f00Msource = &DC[DIR_00M * numberOfLBnodesCoarse];
+ fPP0source = &DC[DIR_PP0 * numberOfLBnodesCoarse];
+ fMM0source = &DC[DIR_MM0 * numberOfLBnodesCoarse];
+ fPM0source = &DC[DIR_PM0 * numberOfLBnodesCoarse];
+ fMP0source = &DC[DIR_MP0 * numberOfLBnodesCoarse];
+ fP0Psource = &DC[DIR_P0P * numberOfLBnodesCoarse];
+ fM0Msource = &DC[DIR_M0M * numberOfLBnodesCoarse];
+ fP0Msource = &DC[DIR_P0M * numberOfLBnodesCoarse];
+ fM0Psource = &DC[DIR_M0P * numberOfLBnodesCoarse];
+ f0PPsource = &DC[DIR_0PP * numberOfLBnodesCoarse];
+ f0MMsource = &DC[DIR_0MM * numberOfLBnodesCoarse];
+ f0PMsource = &DC[DIR_0PM * numberOfLBnodesCoarse];
+ f0MPsource = &DC[DIR_0MP * numberOfLBnodesCoarse];
+ f000source = &DC[DIR_000 * numberOfLBnodesCoarse];
+ fMMMsource = &DC[DIR_MMM * numberOfLBnodesCoarse];
+ fMMPsource = &DC[DIR_MMP * numberOfLBnodesCoarse];
+ fMPPsource = &DC[DIR_MPP * numberOfLBnodesCoarse];
+ fMPMsource = &DC[DIR_MPM * numberOfLBnodesCoarse];
+ fPPMsource = &DC[DIR_PPM * numberOfLBnodesCoarse];
+ fPPPsource = &DC[DIR_PPP * numberOfLBnodesCoarse];
+ fPMPsource = &DC[DIR_PMP * numberOfLBnodesCoarse];
+ fPMMsource = &DC[DIR_PMM * numberOfLBnodesCoarse];
}
else
{
- fP00source = &DC[DIR_M00 *size_MatC];
- fM00source = &DC[DIR_P00 *size_MatC];
- f0P0source = &DC[DIR_0M0 *size_MatC];
- f0M0source = &DC[DIR_0P0 *size_MatC];
- f00Psource = &DC[DIR_00M *size_MatC];
- f00Msource = &DC[DIR_00P *size_MatC];
- fPP0source = &DC[DIR_MM0 *size_MatC];
- fMM0source = &DC[DIR_PP0 *size_MatC];
- fPM0source = &DC[DIR_MP0 *size_MatC];
- fMP0source = &DC[DIR_PM0 *size_MatC];
- fP0Psource = &DC[DIR_M0M *size_MatC];
- fM0Msource = &DC[DIR_P0P *size_MatC];
- fP0Msource = &DC[DIR_M0P *size_MatC];
- fM0Psource = &DC[DIR_P0M *size_MatC];
- f0PPsource = &DC[DIR_0MM *size_MatC];
- f0MMsource = &DC[DIR_0PP *size_MatC];
- f0PMsource = &DC[DIR_0MP *size_MatC];
- f0MPsource = &DC[DIR_0PM *size_MatC];
- f000source = &DC[DIR_000*size_MatC];
- fMMMsource = &DC[DIR_PPP *size_MatC];
- fMMPsource = &DC[DIR_PPM *size_MatC];
- fMPPsource = &DC[DIR_PMM *size_MatC];
- fMPMsource = &DC[DIR_PMP *size_MatC];
- fPPMsource = &DC[DIR_MMP *size_MatC];
- fPPPsource = &DC[DIR_MMM *size_MatC];
- fPMPsource = &DC[DIR_MPM *size_MatC];
- fPMMsource = &DC[DIR_MPP *size_MatC];
+ fP00source = &DC[DIR_M00 * numberOfLBnodesCoarse];
+ fM00source = &DC[DIR_P00 * numberOfLBnodesCoarse];
+ f0P0source = &DC[DIR_0M0 * numberOfLBnodesCoarse];
+ f0M0source = &DC[DIR_0P0 * numberOfLBnodesCoarse];
+ f00Psource = &DC[DIR_00M * numberOfLBnodesCoarse];
+ f00Msource = &DC[DIR_00P * numberOfLBnodesCoarse];
+ fPP0source = &DC[DIR_MM0 * numberOfLBnodesCoarse];
+ fMM0source = &DC[DIR_PP0 * numberOfLBnodesCoarse];
+ fPM0source = &DC[DIR_MP0 * numberOfLBnodesCoarse];
+ fMP0source = &DC[DIR_PM0 * numberOfLBnodesCoarse];
+ fP0Psource = &DC[DIR_M0M * numberOfLBnodesCoarse];
+ fM0Msource = &DC[DIR_P0P * numberOfLBnodesCoarse];
+ fP0Msource = &DC[DIR_M0P * numberOfLBnodesCoarse];
+ fM0Psource = &DC[DIR_P0M * numberOfLBnodesCoarse];
+ f0PPsource = &DC[DIR_0MM * numberOfLBnodesCoarse];
+ f0MMsource = &DC[DIR_0PP * numberOfLBnodesCoarse];
+ f0PMsource = &DC[DIR_0MP * numberOfLBnodesCoarse];
+ f0MPsource = &DC[DIR_0PM * numberOfLBnodesCoarse];
+ f000source = &DC[DIR_000 * numberOfLBnodesCoarse];
+ fMMMsource = &DC[DIR_PPP * numberOfLBnodesCoarse];
+ fMMPsource = &DC[DIR_PPM * numberOfLBnodesCoarse];
+ fMPPsource = &DC[DIR_PMM * numberOfLBnodesCoarse];
+ fMPMsource = &DC[DIR_PMP * numberOfLBnodesCoarse];
+ fPPMsource = &DC[DIR_MMP * numberOfLBnodesCoarse];
+ fPPPsource = &DC[DIR_MMM * numberOfLBnodesCoarse];
+ fPMPsource = &DC[DIR_MPM * numberOfLBnodesCoarse];
+ fPMMsource = &DC[DIR_MPP * numberOfLBnodesCoarse];
}
Distributions6 G;
- G.g[DIR_P00] = &G6[DIR_P00 *size_MatF];
- G.g[DIR_M00] = &G6[DIR_M00 *size_MatF];
- G.g[DIR_0P0] = &G6[DIR_0P0 *size_MatF];
- G.g[DIR_0M0] = &G6[DIR_0M0 *size_MatF];
- G.g[DIR_00P] = &G6[DIR_00P *size_MatF];
- G.g[DIR_00M] = &G6[DIR_00M *size_MatF];
+ G.g[DIR_P00] = &G6[DIR_P00 * numberOfLBnodesFine];
+ G.g[DIR_M00] = &G6[DIR_M00 * numberOfLBnodesFine];
+ G.g[DIR_0P0] = &G6[DIR_0P0 * numberOfLBnodesFine];
+ G.g[DIR_0M0] = &G6[DIR_0M0 * numberOfLBnodesFine];
+ G.g[DIR_00P] = &G6[DIR_00P * numberOfLBnodesFine];
+ G.g[DIR_00M] = &G6[DIR_00M * numberOfLBnodesFine];
////////////////////////////////////////////////////////////////////////////////
const unsigned ix = threadIdx.x; // Globaler x-Index
diff --git a/src/gpu/VirtualFluids_GPU/GPU/GridScaling/ScaleFC27.cu b/src/gpu/VirtualFluids_GPU/GPU/GridScaling/ScaleFC27.cu
index f2a66876cf39e3519e22fc2b0e236514f05ce85a..b37ab44d81d15fbbde46c875c860acd7198b8041 100644
--- a/src/gpu/VirtualFluids_GPU/GPU/GridScaling/ScaleFC27.cu
+++ b/src/gpu/VirtualFluids_GPU/GPU/GridScaling/ScaleFC27.cu
@@ -22,8 +22,8 @@ __global__ void scaleFC_0817_comp_27( real* DC,
unsigned int* neighborFX,
unsigned int* neighborFY,
unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
+ unsigned long long numberOfLBnodesCoarse,
+ unsigned long long numberOfLBnodesFine,
bool isEvenTimestep,
unsigned int* posC,
unsigned int* posFSWB,
@@ -43,33 +43,33 @@ __global__ void scaleFC_0817_comp_27( real* DC,
*f000source, *fMMMsource, *fMMPsource, *fMPPsource, *fMPMsource, *fPPMsource, *fPPPsource, *fPMPsource, *fPMMsource;
- fP00source = &DF[DIR_P00 *size_MatF];
- fM00source = &DF[DIR_M00 *size_MatF];
- f0P0source = &DF[DIR_0P0 *size_MatF];
- f0M0source = &DF[DIR_0M0 *size_MatF];
- f00Psource = &DF[DIR_00P *size_MatF];
- f00Msource = &DF[DIR_00M *size_MatF];
- fPP0source = &DF[DIR_PP0 *size_MatF];
- fMM0source = &DF[DIR_MM0 *size_MatF];
- fPM0source = &DF[DIR_PM0 *size_MatF];
- fMP0source = &DF[DIR_MP0 *size_MatF];
- fP0Psource = &DF[DIR_P0P *size_MatF];
- fM0Msource = &DF[DIR_M0M *size_MatF];
- fP0Msource = &DF[DIR_P0M *size_MatF];
- fM0Psource = &DF[DIR_M0P *size_MatF];
- f0PPsource = &DF[DIR_0PP *size_MatF];
- f0MMsource = &DF[DIR_0MM *size_MatF];
- f0PMsource = &DF[DIR_0PM *size_MatF];
- f0MPsource = &DF[DIR_0MP *size_MatF];
- f000source = &DF[DIR_000*size_MatF];
- fMMMsource = &DF[DIR_MMM *size_MatF];
- fMMPsource = &DF[DIR_MMP *size_MatF];
- fMPPsource = &DF[DIR_MPP *size_MatF];
- fMPMsource = &DF[DIR_MPM *size_MatF];
- fPPMsource = &DF[DIR_PPM *size_MatF];
- fPPPsource = &DF[DIR_PPP *size_MatF];
- fPMPsource = &DF[DIR_PMP *size_MatF];
- fPMMsource = &DF[DIR_PMM *size_MatF];
+ fP00source = &DF[DIR_P00 * numberOfLBnodesFine];
+ fM00source = &DF[DIR_M00 * numberOfLBnodesFine];
+ f0P0source = &DF[DIR_0P0 * numberOfLBnodesFine];
+ f0M0source = &DF[DIR_0M0 * numberOfLBnodesFine];
+ f00Psource = &DF[DIR_00P * numberOfLBnodesFine];
+ f00Msource = &DF[DIR_00M * numberOfLBnodesFine];
+ fPP0source = &DF[DIR_PP0 * numberOfLBnodesFine];
+ fMM0source = &DF[DIR_MM0 * numberOfLBnodesFine];
+ fPM0source = &DF[DIR_PM0 * numberOfLBnodesFine];
+ fMP0source = &DF[DIR_MP0 * numberOfLBnodesFine];
+ fP0Psource = &DF[DIR_P0P * numberOfLBnodesFine];
+ fM0Msource = &DF[DIR_M0M * numberOfLBnodesFine];
+ fP0Msource = &DF[DIR_P0M * numberOfLBnodesFine];
+ fM0Psource = &DF[DIR_M0P * numberOfLBnodesFine];
+ f0PPsource = &DF[DIR_0PP * numberOfLBnodesFine];
+ f0MMsource = &DF[DIR_0MM * numberOfLBnodesFine];
+ f0PMsource = &DF[DIR_0PM * numberOfLBnodesFine];
+ f0MPsource = &DF[DIR_0MP * numberOfLBnodesFine];
+ f000source = &DF[DIR_000 * numberOfLBnodesFine];
+ fMMMsource = &DF[DIR_MMM * numberOfLBnodesFine];
+ fMMPsource = &DF[DIR_MMP * numberOfLBnodesFine];
+ fMPPsource = &DF[DIR_MPP * numberOfLBnodesFine];
+ fMPMsource = &DF[DIR_MPM * numberOfLBnodesFine];
+ fPPMsource = &DF[DIR_PPM * numberOfLBnodesFine];
+ fPPPsource = &DF[DIR_PPP * numberOfLBnodesFine];
+ fPMPsource = &DF[DIR_PMP * numberOfLBnodesFine];
+ fPMMsource = &DF[DIR_PMM * numberOfLBnodesFine];
real
*fP00dest, *fM00dest, *f0P0dest, *f0M0dest, *f00Pdest, *f00Mdest, *fPP0dest, *fMM0dest, *fPM0dest,
@@ -78,63 +78,63 @@ __global__ void scaleFC_0817_comp_27( real* DC,
if (isEvenTimestep==true)
{
- fP00dest = &DC[DIR_P00 *size_MatC];
- fM00dest = &DC[DIR_M00 *size_MatC];
- f0P0dest = &DC[DIR_0P0 *size_MatC];
- f0M0dest = &DC[DIR_0M0 *size_MatC];
- f00Pdest = &DC[DIR_00P *size_MatC];
- f00Mdest = &DC[DIR_00M *size_MatC];
- fPP0dest = &DC[DIR_PP0 *size_MatC];
- fMM0dest = &DC[DIR_MM0 *size_MatC];
- fPM0dest = &DC[DIR_PM0 *size_MatC];
- fMP0dest = &DC[DIR_MP0 *size_MatC];
- fP0Pdest = &DC[DIR_P0P *size_MatC];
- fM0Mdest = &DC[DIR_M0M *size_MatC];
- fP0Mdest = &DC[DIR_P0M *size_MatC];
- fM0Pdest = &DC[DIR_M0P *size_MatC];
- f0PPdest = &DC[DIR_0PP *size_MatC];
- f0MMdest = &DC[DIR_0MM *size_MatC];
- f0PMdest = &DC[DIR_0PM *size_MatC];
- f0MPdest = &DC[DIR_0MP *size_MatC];
- f000dest = &DC[DIR_000*size_MatC];
- fMMMdest = &DC[DIR_MMM *size_MatC];
- fMMPdest = &DC[DIR_MMP *size_MatC];
- fMPPdest = &DC[DIR_MPP *size_MatC];
- fMPMdest = &DC[DIR_MPM *size_MatC];
- fPPMdest = &DC[DIR_PPM *size_MatC];
- fPPPdest = &DC[DIR_PPP *size_MatC];
- fPMPdest = &DC[DIR_PMP *size_MatC];
- fPMMdest = &DC[DIR_PMM *size_MatC];
+ fP00dest = &DC[DIR_P00 * numberOfLBnodesCoarse];
+ fM00dest = &DC[DIR_M00 * numberOfLBnodesCoarse];
+ f0P0dest = &DC[DIR_0P0 * numberOfLBnodesCoarse];
+ f0M0dest = &DC[DIR_0M0 * numberOfLBnodesCoarse];
+ f00Pdest = &DC[DIR_00P * numberOfLBnodesCoarse];
+ f00Mdest = &DC[DIR_00M * numberOfLBnodesCoarse];
+ fPP0dest = &DC[DIR_PP0 * numberOfLBnodesCoarse];
+ fMM0dest = &DC[DIR_MM0 * numberOfLBnodesCoarse];
+ fPM0dest = &DC[DIR_PM0 * numberOfLBnodesCoarse];
+ fMP0dest = &DC[DIR_MP0 * numberOfLBnodesCoarse];
+ fP0Pdest = &DC[DIR_P0P * numberOfLBnodesCoarse];
+ fM0Mdest = &DC[DIR_M0M * numberOfLBnodesCoarse];
+ fP0Mdest = &DC[DIR_P0M * numberOfLBnodesCoarse];
+ fM0Pdest = &DC[DIR_M0P * numberOfLBnodesCoarse];
+ f0PPdest = &DC[DIR_0PP * numberOfLBnodesCoarse];
+ f0MMdest = &DC[DIR_0MM * numberOfLBnodesCoarse];
+ f0PMdest = &DC[DIR_0PM * numberOfLBnodesCoarse];
+ f0MPdest = &DC[DIR_0MP * numberOfLBnodesCoarse];
+ f000dest = &DC[DIR_000 * numberOfLBnodesCoarse];
+ fMMMdest = &DC[DIR_MMM * numberOfLBnodesCoarse];
+ fMMPdest = &DC[DIR_MMP * numberOfLBnodesCoarse];
+ fMPPdest = &DC[DIR_MPP * numberOfLBnodesCoarse];
+ fMPMdest = &DC[DIR_MPM * numberOfLBnodesCoarse];
+ fPPMdest = &DC[DIR_PPM * numberOfLBnodesCoarse];
+ fPPPdest = &DC[DIR_PPP * numberOfLBnodesCoarse];
+ fPMPdest = &DC[DIR_PMP * numberOfLBnodesCoarse];
+ fPMMdest = &DC[DIR_PMM * numberOfLBnodesCoarse];
}
else
{
- fP00dest = &DC[DIR_M00 *size_MatC];
- fM00dest = &DC[DIR_P00 *size_MatC];
- f0P0dest = &DC[DIR_0M0 *size_MatC];
- f0M0dest = &DC[DIR_0P0 *size_MatC];
- f00Pdest = &DC[DIR_00M *size_MatC];
- f00Mdest = &DC[DIR_00P *size_MatC];
- fPP0dest = &DC[DIR_MM0 *size_MatC];
- fMM0dest = &DC[DIR_PP0 *size_MatC];
- fPM0dest = &DC[DIR_MP0 *size_MatC];
- fMP0dest = &DC[DIR_PM0 *size_MatC];
- fP0Pdest = &DC[DIR_M0M *size_MatC];
- fM0Mdest = &DC[DIR_P0P *size_MatC];
- fP0Mdest = &DC[DIR_M0P *size_MatC];
- fM0Pdest = &DC[DIR_P0M *size_MatC];
- f0PPdest = &DC[DIR_0MM *size_MatC];
- f0MMdest = &DC[DIR_0PP *size_MatC];
- f0PMdest = &DC[DIR_0MP *size_MatC];
- f0MPdest = &DC[DIR_0PM *size_MatC];
- f000dest = &DC[DIR_000*size_MatC];
- fMMMdest = &DC[DIR_PPP *size_MatC];
- fMMPdest = &DC[DIR_PPM *size_MatC];
- fMPPdest = &DC[DIR_PMM *size_MatC];
- fMPMdest = &DC[DIR_PMP *size_MatC];
- fPPMdest = &DC[DIR_MMP *size_MatC];
- fPPPdest = &DC[DIR_MMM *size_MatC];
- fPMPdest = &DC[DIR_MPM *size_MatC];
- fPMMdest = &DC[DIR_MPP *size_MatC];
+ fP00dest = &DC[DIR_M00 * numberOfLBnodesCoarse];
+ fM00dest = &DC[DIR_P00 * numberOfLBnodesCoarse];
+ f0P0dest = &DC[DIR_0M0 * numberOfLBnodesCoarse];
+ f0M0dest = &DC[DIR_0P0 * numberOfLBnodesCoarse];
+ f00Pdest = &DC[DIR_00M * numberOfLBnodesCoarse];
+ f00Mdest = &DC[DIR_00P * numberOfLBnodesCoarse];
+ fPP0dest = &DC[DIR_MM0 * numberOfLBnodesCoarse];
+ fMM0dest = &DC[DIR_PP0 * numberOfLBnodesCoarse];
+ fPM0dest = &DC[DIR_MP0 * numberOfLBnodesCoarse];
+ fMP0dest = &DC[DIR_PM0 * numberOfLBnodesCoarse];
+ fP0Pdest = &DC[DIR_M0M * numberOfLBnodesCoarse];
+ fM0Mdest = &DC[DIR_P0P * numberOfLBnodesCoarse];
+ fP0Mdest = &DC[DIR_M0P * numberOfLBnodesCoarse];
+ fM0Pdest = &DC[DIR_P0M * numberOfLBnodesCoarse];
+ f0PPdest = &DC[DIR_0MM * numberOfLBnodesCoarse];
+ f0MMdest = &DC[DIR_0PP * numberOfLBnodesCoarse];
+ f0PMdest = &DC[DIR_0MP * numberOfLBnodesCoarse];
+ f0MPdest = &DC[DIR_0PM * numberOfLBnodesCoarse];
+ f000dest = &DC[DIR_000 * numberOfLBnodesCoarse];
+ fMMMdest = &DC[DIR_PPP * numberOfLBnodesCoarse];
+ fMMPdest = &DC[DIR_PPM * numberOfLBnodesCoarse];
+ fMPPdest = &DC[DIR_PMM * numberOfLBnodesCoarse];
+ fMPMdest = &DC[DIR_PMP * numberOfLBnodesCoarse];
+ fPPMdest = &DC[DIR_MMP * numberOfLBnodesCoarse];
+ fPPPdest = &DC[DIR_MMM * numberOfLBnodesCoarse];
+ fPMPdest = &DC[DIR_MPM * numberOfLBnodesCoarse];
+ fPMMdest = &DC[DIR_MPP * numberOfLBnodesCoarse];
}
////////////////////////////////////////////////////////////////////////////////
const unsigned ix = threadIdx.x; // Globaler x-Index
@@ -1218,8 +1218,8 @@ __global__ void scaleFC_AA2016_comp_27(real* DC,
unsigned int* neighborFX,
unsigned int* neighborFY,
unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
+ unsigned long long numberOfLBnodesCoarse,
+ unsigned long long numberOfLBnodesFine,
bool isEvenTimestep,
unsigned int* posC,
unsigned int* posFSWB,
@@ -1236,96 +1236,96 @@ __global__ void scaleFC_AA2016_comp_27(real* DC,
real *feF, *fwF, *fnF, *fsF, *ftF, *fbF, *fneF, *fswF, *fseF, *fnwF, *fteF, *fbwF, *fbeF, *ftwF, *ftnF, *fbsF, *fbnF, *ftsF, *fzeroF,
*ftneF, *ftswF, *ftseF, *ftnwF, *fbneF, *fbswF, *fbseF, *fbnwF;
- feF = &DF[DIR_P00 *size_MatF];
- fwF = &DF[DIR_M00 *size_MatF];
- fnF = &DF[DIR_0P0 *size_MatF];
- fsF = &DF[DIR_0M0 *size_MatF];
- ftF = &DF[DIR_00P *size_MatF];
- fbF = &DF[DIR_00M *size_MatF];
- fneF = &DF[DIR_PP0 *size_MatF];
- fswF = &DF[DIR_MM0 *size_MatF];
- fseF = &DF[DIR_PM0 *size_MatF];
- fnwF = &DF[DIR_MP0 *size_MatF];
- fteF = &DF[DIR_P0P *size_MatF];
- fbwF = &DF[DIR_M0M *size_MatF];
- fbeF = &DF[DIR_P0M *size_MatF];
- ftwF = &DF[DIR_M0P *size_MatF];
- ftnF = &DF[DIR_0PP *size_MatF];
- fbsF = &DF[DIR_0MM *size_MatF];
- fbnF = &DF[DIR_0PM *size_MatF];
- ftsF = &DF[DIR_0MP *size_MatF];
- fzeroF = &DF[DIR_000*size_MatF];
- ftneF = &DF[DIR_PPP *size_MatF];
- ftswF = &DF[DIR_MMP *size_MatF];
- ftseF = &DF[DIR_PMP *size_MatF];
- ftnwF = &DF[DIR_MPP *size_MatF];
- fbneF = &DF[DIR_PPM *size_MatF];
- fbswF = &DF[DIR_MMM *size_MatF];
- fbseF = &DF[DIR_PMM *size_MatF];
- fbnwF = &DF[DIR_MPM *size_MatF];
+ feF = &DF[DIR_P00 * numberOfLBnodesFine];
+ fwF = &DF[DIR_M00 * numberOfLBnodesFine];
+ fnF = &DF[DIR_0P0 * numberOfLBnodesFine];
+ fsF = &DF[DIR_0M0 * numberOfLBnodesFine];
+ ftF = &DF[DIR_00P * numberOfLBnodesFine];
+ fbF = &DF[DIR_00M * numberOfLBnodesFine];
+ fneF = &DF[DIR_PP0 * numberOfLBnodesFine];
+ fswF = &DF[DIR_MM0 * numberOfLBnodesFine];
+ fseF = &DF[DIR_PM0 * numberOfLBnodesFine];
+ fnwF = &DF[DIR_MP0 * numberOfLBnodesFine];
+ fteF = &DF[DIR_P0P * numberOfLBnodesFine];
+ fbwF = &DF[DIR_M0M * numberOfLBnodesFine];
+ fbeF = &DF[DIR_P0M * numberOfLBnodesFine];
+ ftwF = &DF[DIR_M0P * numberOfLBnodesFine];
+ ftnF = &DF[DIR_0PP * numberOfLBnodesFine];
+ fbsF = &DF[DIR_0MM * numberOfLBnodesFine];
+ fbnF = &DF[DIR_0PM * numberOfLBnodesFine];
+ ftsF = &DF[DIR_0MP * numberOfLBnodesFine];
+ fzeroF = &DF[DIR_000 * numberOfLBnodesFine];
+ ftneF = &DF[DIR_PPP * numberOfLBnodesFine];
+ ftswF = &DF[DIR_MMP * numberOfLBnodesFine];
+ ftseF = &DF[DIR_PMP * numberOfLBnodesFine];
+ ftnwF = &DF[DIR_MPP * numberOfLBnodesFine];
+ fbneF = &DF[DIR_PPM * numberOfLBnodesFine];
+ fbswF = &DF[DIR_MMM * numberOfLBnodesFine];
+ fbseF = &DF[DIR_PMM * numberOfLBnodesFine];
+ fbnwF = &DF[DIR_MPM * numberOfLBnodesFine];
real *feC, *fwC, *fnC, *fsC, *ftC, *fbC, *fneC, *fswC, *fseC, *fnwC, *fteC, *fbwC, *fbeC, *ftwC, *ftnC, *fbsC, *fbnC, *ftsC, *fzeroC,
*ftneC, *ftswC, *ftseC, *ftnwC, *fbneC, *fbswC, *fbseC, *fbnwC;
if (isEvenTimestep==true)
{
- feC = &DC[DIR_P00 *size_MatC];
- fwC = &DC[DIR_M00 *size_MatC];
- fnC = &DC[DIR_0P0 *size_MatC];
- fsC = &DC[DIR_0M0 *size_MatC];
- ftC = &DC[DIR_00P *size_MatC];
- fbC = &DC[DIR_00M *size_MatC];
- fneC = &DC[DIR_PP0 *size_MatC];
- fswC = &DC[DIR_MM0 *size_MatC];
- fseC = &DC[DIR_PM0 *size_MatC];
- fnwC = &DC[DIR_MP0 *size_MatC];
- fteC = &DC[DIR_P0P *size_MatC];
- fbwC = &DC[DIR_M0M *size_MatC];
- fbeC = &DC[DIR_P0M *size_MatC];
- ftwC = &DC[DIR_M0P *size_MatC];
- ftnC = &DC[DIR_0PP *size_MatC];
- fbsC = &DC[DIR_0MM *size_MatC];
- fbnC = &DC[DIR_0PM *size_MatC];
- ftsC = &DC[DIR_0MP *size_MatC];
- fzeroC = &DC[DIR_000*size_MatC];
- ftneC = &DC[DIR_PPP *size_MatC];
- ftswC = &DC[DIR_MMP *size_MatC];
- ftseC = &DC[DIR_PMP *size_MatC];
- ftnwC = &DC[DIR_MPP *size_MatC];
- fbneC = &DC[DIR_PPM *size_MatC];
- fbswC = &DC[DIR_MMM *size_MatC];
- fbseC = &DC[DIR_PMM *size_MatC];
- fbnwC = &DC[DIR_MPM *size_MatC];
+ feC = &DC[DIR_P00 * numberOfLBnodesCoarse];
+ fwC = &DC[DIR_M00 * numberOfLBnodesCoarse];
+ fnC = &DC[DIR_0P0 * numberOfLBnodesCoarse];
+ fsC = &DC[DIR_0M0 * numberOfLBnodesCoarse];
+ ftC = &DC[DIR_00P * numberOfLBnodesCoarse];
+ fbC = &DC[DIR_00M * numberOfLBnodesCoarse];
+ fneC = &DC[DIR_PP0 * numberOfLBnodesCoarse];
+ fswC = &DC[DIR_MM0 * numberOfLBnodesCoarse];
+ fseC = &DC[DIR_PM0 * numberOfLBnodesCoarse];
+ fnwC = &DC[DIR_MP0 * numberOfLBnodesCoarse];
+ fteC = &DC[DIR_P0P * numberOfLBnodesCoarse];
+ fbwC = &DC[DIR_M0M * numberOfLBnodesCoarse];
+ fbeC = &DC[DIR_P0M * numberOfLBnodesCoarse];
+ ftwC = &DC[DIR_M0P * numberOfLBnodesCoarse];
+ ftnC = &DC[DIR_0PP * numberOfLBnodesCoarse];
+ fbsC = &DC[DIR_0MM * numberOfLBnodesCoarse];
+ fbnC = &DC[DIR_0PM * numberOfLBnodesCoarse];
+ ftsC = &DC[DIR_0MP * numberOfLBnodesCoarse];
+ fzeroC = &DC[DIR_000 * numberOfLBnodesCoarse];
+ ftneC = &DC[DIR_PPP * numberOfLBnodesCoarse];
+ ftswC = &DC[DIR_MMP * numberOfLBnodesCoarse];
+ ftseC = &DC[DIR_PMP * numberOfLBnodesCoarse];
+ ftnwC = &DC[DIR_MPP * numberOfLBnodesCoarse];
+ fbneC = &DC[DIR_PPM * numberOfLBnodesCoarse];
+ fbswC = &DC[DIR_MMM * numberOfLBnodesCoarse];
+ fbseC = &DC[DIR_PMM * numberOfLBnodesCoarse];
+ fbnwC = &DC[DIR_MPM * numberOfLBnodesCoarse];
}
else
{
- fwC = &DC[DIR_P00 *size_MatC];
- feC = &DC[DIR_M00 *size_MatC];
- fsC = &DC[DIR_0P0 *size_MatC];
- fnC = &DC[DIR_0M0 *size_MatC];
- fbC = &DC[DIR_00P *size_MatC];
- ftC = &DC[DIR_00M *size_MatC];
- fswC = &DC[DIR_PP0 *size_MatC];
- fneC = &DC[DIR_MM0 *size_MatC];
- fnwC = &DC[DIR_PM0 *size_MatC];
- fseC = &DC[DIR_MP0 *size_MatC];
- fbwC = &DC[DIR_P0P *size_MatC];
- fteC = &DC[DIR_M0M *size_MatC];
- ftwC = &DC[DIR_P0M *size_MatC];
- fbeC = &DC[DIR_M0P *size_MatC];
- fbsC = &DC[DIR_0PP *size_MatC];
- ftnC = &DC[DIR_0MM *size_MatC];
- ftsC = &DC[DIR_0PM *size_MatC];
- fbnC = &DC[DIR_0MP *size_MatC];
- fzeroC = &DC[DIR_000*size_MatC];
- fbswC = &DC[DIR_PPP *size_MatC];
- fbneC = &DC[DIR_MMP *size_MatC];
- fbnwC = &DC[DIR_PMP *size_MatC];
- fbseC = &DC[DIR_MPP *size_MatC];
- ftswC = &DC[DIR_PPM *size_MatC];
- ftneC = &DC[DIR_MMM *size_MatC];
- ftnwC = &DC[DIR_PMM *size_MatC];
- ftseC = &DC[DIR_MPM *size_MatC];
+ fwC = &DC[DIR_P00 * numberOfLBnodesCoarse];
+ feC = &DC[DIR_M00 * numberOfLBnodesCoarse];
+ fsC = &DC[DIR_0P0 * numberOfLBnodesCoarse];
+ fnC = &DC[DIR_0M0 * numberOfLBnodesCoarse];
+ fbC = &DC[DIR_00P * numberOfLBnodesCoarse];
+ ftC = &DC[DIR_00M * numberOfLBnodesCoarse];
+ fswC = &DC[DIR_PP0 * numberOfLBnodesCoarse];
+ fneC = &DC[DIR_MM0 * numberOfLBnodesCoarse];
+ fnwC = &DC[DIR_PM0 * numberOfLBnodesCoarse];
+ fseC = &DC[DIR_MP0 * numberOfLBnodesCoarse];
+ fbwC = &DC[DIR_P0P * numberOfLBnodesCoarse];
+ fteC = &DC[DIR_M0M * numberOfLBnodesCoarse];
+ ftwC = &DC[DIR_P0M * numberOfLBnodesCoarse];
+ fbeC = &DC[DIR_M0P * numberOfLBnodesCoarse];
+ fbsC = &DC[DIR_0PP * numberOfLBnodesCoarse];
+ ftnC = &DC[DIR_0MM * numberOfLBnodesCoarse];
+ ftsC = &DC[DIR_0PM * numberOfLBnodesCoarse];
+ fbnC = &DC[DIR_0MP * numberOfLBnodesCoarse];
+ fzeroC = &DC[DIR_000 * numberOfLBnodesCoarse];
+ fbswC = &DC[DIR_PPP * numberOfLBnodesCoarse];
+ fbneC = &DC[DIR_MMP * numberOfLBnodesCoarse];
+ fbnwC = &DC[DIR_PMP * numberOfLBnodesCoarse];
+ fbseC = &DC[DIR_MPP * numberOfLBnodesCoarse];
+ ftswC = &DC[DIR_PPM * numberOfLBnodesCoarse];
+ ftneC = &DC[DIR_MMM * numberOfLBnodesCoarse];
+ ftnwC = &DC[DIR_PMM * numberOfLBnodesCoarse];
+ ftseC = &DC[DIR_MPM * numberOfLBnodesCoarse];
}
////////////////////////////////////////////////////////////////////////////////
const unsigned ix = threadIdx.x; // Globaler x-Index
@@ -5407,8 +5407,8 @@ __global__ void scaleFC_RhoSq_3rdMom_comp_27(real* DC,
unsigned int* neighborFX,
unsigned int* neighborFY,
unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
+ unsigned long long numberOfLBnodesCoarse,
+ unsigned long long numberOfLBnodesFine,
bool isEvenTimestep,
unsigned int* posC,
unsigned int* posFSWB,
@@ -5425,96 +5425,96 @@ __global__ void scaleFC_RhoSq_3rdMom_comp_27(real* DC,
real *feF, *fwF, *fnF, *fsF, *ftF, *fbF, *fneF, *fswF, *fseF, *fnwF, *fteF, *fbwF, *fbeF, *ftwF, *ftnF, *fbsF, *fbnF, *ftsF, *fzeroF,
*ftneF, *ftswF, *ftseF, *ftnwF, *fbneF, *fbswF, *fbseF, *fbnwF;
- feF = &DF[DIR_P00 *size_MatF];
- fwF = &DF[DIR_M00 *size_MatF];
- fnF = &DF[DIR_0P0 *size_MatF];
- fsF = &DF[DIR_0M0 *size_MatF];
- ftF = &DF[DIR_00P *size_MatF];
- fbF = &DF[DIR_00M *size_MatF];
- fneF = &DF[DIR_PP0 *size_MatF];
- fswF = &DF[DIR_MM0 *size_MatF];
- fseF = &DF[DIR_PM0 *size_MatF];
- fnwF = &DF[DIR_MP0 *size_MatF];
- fteF = &DF[DIR_P0P *size_MatF];
- fbwF = &DF[DIR_M0M *size_MatF];
- fbeF = &DF[DIR_P0M *size_MatF];
- ftwF = &DF[DIR_M0P *size_MatF];
- ftnF = &DF[DIR_0PP *size_MatF];
- fbsF = &DF[DIR_0MM *size_MatF];
- fbnF = &DF[DIR_0PM *size_MatF];
- ftsF = &DF[DIR_0MP *size_MatF];
- fzeroF = &DF[DIR_000*size_MatF];
- ftneF = &DF[DIR_PPP *size_MatF];
- ftswF = &DF[DIR_MMP *size_MatF];
- ftseF = &DF[DIR_PMP *size_MatF];
- ftnwF = &DF[DIR_MPP *size_MatF];
- fbneF = &DF[DIR_PPM *size_MatF];
- fbswF = &DF[DIR_MMM *size_MatF];
- fbseF = &DF[DIR_PMM *size_MatF];
- fbnwF = &DF[DIR_MPM *size_MatF];
+ feF = &DF[DIR_P00 * numberOfLBnodesFine];
+ fwF = &DF[DIR_M00 * numberOfLBnodesFine];
+ fnF = &DF[DIR_0P0 * numberOfLBnodesFine];
+ fsF = &DF[DIR_0M0 * numberOfLBnodesFine];
+ ftF = &DF[DIR_00P * numberOfLBnodesFine];
+ fbF = &DF[DIR_00M * numberOfLBnodesFine];
+ fneF = &DF[DIR_PP0 * numberOfLBnodesFine];
+ fswF = &DF[DIR_MM0 * numberOfLBnodesFine];
+ fseF = &DF[DIR_PM0 * numberOfLBnodesFine];
+ fnwF = &DF[DIR_MP0 * numberOfLBnodesFine];
+ fteF = &DF[DIR_P0P * numberOfLBnodesFine];
+ fbwF = &DF[DIR_M0M * numberOfLBnodesFine];
+ fbeF = &DF[DIR_P0M * numberOfLBnodesFine];
+ ftwF = &DF[DIR_M0P * numberOfLBnodesFine];
+ ftnF = &DF[DIR_0PP * numberOfLBnodesFine];
+ fbsF = &DF[DIR_0MM * numberOfLBnodesFine];
+ fbnF = &DF[DIR_0PM * numberOfLBnodesFine];
+ ftsF = &DF[DIR_0MP * numberOfLBnodesFine];
+ fzeroF = &DF[DIR_000 * numberOfLBnodesFine];
+ ftneF = &DF[DIR_PPP * numberOfLBnodesFine];
+ ftswF = &DF[DIR_MMP * numberOfLBnodesFine];
+ ftseF = &DF[DIR_PMP * numberOfLBnodesFine];
+ ftnwF = &DF[DIR_MPP * numberOfLBnodesFine];
+ fbneF = &DF[DIR_PPM * numberOfLBnodesFine];
+ fbswF = &DF[DIR_MMM * numberOfLBnodesFine];
+ fbseF = &DF[DIR_PMM * numberOfLBnodesFine];
+ fbnwF = &DF[DIR_MPM * numberOfLBnodesFine];
real *feC, *fwC, *fnC, *fsC, *ftC, *fbC, *fneC, *fswC, *fseC, *fnwC, *fteC, *fbwC, *fbeC, *ftwC, *ftnC, *fbsC, *fbnC, *ftsC, *fzeroC,
*ftneC, *ftswC, *ftseC, *ftnwC, *fbneC, *fbswC, *fbseC, *fbnwC;
if (isEvenTimestep==true)
{
- feC = &DC[DIR_P00 *size_MatC];
- fwC = &DC[DIR_M00 *size_MatC];
- fnC = &DC[DIR_0P0 *size_MatC];
- fsC = &DC[DIR_0M0 *size_MatC];
- ftC = &DC[DIR_00P *size_MatC];
- fbC = &DC[DIR_00M *size_MatC];
- fneC = &DC[DIR_PP0 *size_MatC];
- fswC = &DC[DIR_MM0 *size_MatC];
- fseC = &DC[DIR_PM0 *size_MatC];
- fnwC = &DC[DIR_MP0 *size_MatC];
- fteC = &DC[DIR_P0P *size_MatC];
- fbwC = &DC[DIR_M0M *size_MatC];
- fbeC = &DC[DIR_P0M *size_MatC];
- ftwC = &DC[DIR_M0P *size_MatC];
- ftnC = &DC[DIR_0PP *size_MatC];
- fbsC = &DC[DIR_0MM *size_MatC];
- fbnC = &DC[DIR_0PM *size_MatC];
- ftsC = &DC[DIR_0MP *size_MatC];
- fzeroC = &DC[DIR_000*size_MatC];
- ftneC = &DC[DIR_PPP *size_MatC];
- ftswC = &DC[DIR_MMP *size_MatC];
- ftseC = &DC[DIR_PMP *size_MatC];
- ftnwC = &DC[DIR_MPP *size_MatC];
- fbneC = &DC[DIR_PPM *size_MatC];
- fbswC = &DC[DIR_MMM *size_MatC];
- fbseC = &DC[DIR_PMM *size_MatC];
- fbnwC = &DC[DIR_MPM *size_MatC];
+ feC = &DC[DIR_P00 * numberOfLBnodesCoarse];
+ fwC = &DC[DIR_M00 * numberOfLBnodesCoarse];
+ fnC = &DC[DIR_0P0 * numberOfLBnodesCoarse];
+ fsC = &DC[DIR_0M0 * numberOfLBnodesCoarse];
+ ftC = &DC[DIR_00P * numberOfLBnodesCoarse];
+ fbC = &DC[DIR_00M * numberOfLBnodesCoarse];
+ fneC = &DC[DIR_PP0 * numberOfLBnodesCoarse];
+ fswC = &DC[DIR_MM0 * numberOfLBnodesCoarse];
+ fseC = &DC[DIR_PM0 * numberOfLBnodesCoarse];
+ fnwC = &DC[DIR_MP0 * numberOfLBnodesCoarse];
+ fteC = &DC[DIR_P0P * numberOfLBnodesCoarse];
+ fbwC = &DC[DIR_M0M * numberOfLBnodesCoarse];
+ fbeC = &DC[DIR_P0M * numberOfLBnodesCoarse];
+ ftwC = &DC[DIR_M0P * numberOfLBnodesCoarse];
+ ftnC = &DC[DIR_0PP * numberOfLBnodesCoarse];
+ fbsC = &DC[DIR_0MM * numberOfLBnodesCoarse];
+ fbnC = &DC[DIR_0PM * numberOfLBnodesCoarse];
+ ftsC = &DC[DIR_0MP * numberOfLBnodesCoarse];
+ fzeroC = &DC[DIR_000 * numberOfLBnodesCoarse];
+ ftneC = &DC[DIR_PPP * numberOfLBnodesCoarse];
+ ftswC = &DC[DIR_MMP * numberOfLBnodesCoarse];
+ ftseC = &DC[DIR_PMP * numberOfLBnodesCoarse];
+ ftnwC = &DC[DIR_MPP * numberOfLBnodesCoarse];
+ fbneC = &DC[DIR_PPM * numberOfLBnodesCoarse];
+ fbswC = &DC[DIR_MMM * numberOfLBnodesCoarse];
+ fbseC = &DC[DIR_PMM * numberOfLBnodesCoarse];
+ fbnwC = &DC[DIR_MPM * numberOfLBnodesCoarse];
}
else
{
- fwC = &DC[DIR_P00 *size_MatC];
- feC = &DC[DIR_M00 *size_MatC];
- fsC = &DC[DIR_0P0 *size_MatC];
- fnC = &DC[DIR_0M0 *size_MatC];
- fbC = &DC[DIR_00P *size_MatC];
- ftC = &DC[DIR_00M *size_MatC];
- fswC = &DC[DIR_PP0 *size_MatC];
- fneC = &DC[DIR_MM0 *size_MatC];
- fnwC = &DC[DIR_PM0 *size_MatC];
- fseC = &DC[DIR_MP0 *size_MatC];
- fbwC = &DC[DIR_P0P *size_MatC];
- fteC = &DC[DIR_M0M *size_MatC];
- ftwC = &DC[DIR_P0M *size_MatC];
- fbeC = &DC[DIR_M0P *size_MatC];
- fbsC = &DC[DIR_0PP *size_MatC];
- ftnC = &DC[DIR_0MM *size_MatC];
- ftsC = &DC[DIR_0PM *size_MatC];
- fbnC = &DC[DIR_0MP *size_MatC];
- fzeroC = &DC[DIR_000*size_MatC];
- fbswC = &DC[DIR_PPP *size_MatC];
- fbneC = &DC[DIR_MMP *size_MatC];
- fbnwC = &DC[DIR_PMP *size_MatC];
- fbseC = &DC[DIR_MPP *size_MatC];
- ftswC = &DC[DIR_PPM *size_MatC];
- ftneC = &DC[DIR_MMM *size_MatC];
- ftnwC = &DC[DIR_PMM *size_MatC];
- ftseC = &DC[DIR_MPM *size_MatC];
+ fwC = &DC[DIR_P00 * numberOfLBnodesCoarse];
+ feC = &DC[DIR_M00 * numberOfLBnodesCoarse];
+ fsC = &DC[DIR_0P0 * numberOfLBnodesCoarse];
+ fnC = &DC[DIR_0M0 * numberOfLBnodesCoarse];
+ fbC = &DC[DIR_00P * numberOfLBnodesCoarse];
+ ftC = &DC[DIR_00M * numberOfLBnodesCoarse];
+ fswC = &DC[DIR_PP0 * numberOfLBnodesCoarse];
+ fneC = &DC[DIR_MM0 * numberOfLBnodesCoarse];
+ fnwC = &DC[DIR_PM0 * numberOfLBnodesCoarse];
+ fseC = &DC[DIR_MP0 * numberOfLBnodesCoarse];
+ fbwC = &DC[DIR_P0P * numberOfLBnodesCoarse];
+ fteC = &DC[DIR_M0M * numberOfLBnodesCoarse];
+ ftwC = &DC[DIR_P0M * numberOfLBnodesCoarse];
+ fbeC = &DC[DIR_M0P * numberOfLBnodesCoarse];
+ fbsC = &DC[DIR_0PP * numberOfLBnodesCoarse];
+ ftnC = &DC[DIR_0MM * numberOfLBnodesCoarse];
+ ftsC = &DC[DIR_0PM * numberOfLBnodesCoarse];
+ fbnC = &DC[DIR_0MP * numberOfLBnodesCoarse];
+ fzeroC = &DC[DIR_000 * numberOfLBnodesCoarse];
+ fbswC = &DC[DIR_PPP * numberOfLBnodesCoarse];
+ fbneC = &DC[DIR_MMP * numberOfLBnodesCoarse];
+ fbnwC = &DC[DIR_PMP * numberOfLBnodesCoarse];
+ fbseC = &DC[DIR_MPP * numberOfLBnodesCoarse];
+ ftswC = &DC[DIR_PPM * numberOfLBnodesCoarse];
+ ftneC = &DC[DIR_MMM * numberOfLBnodesCoarse];
+ ftnwC = &DC[DIR_PMM * numberOfLBnodesCoarse];
+ ftseC = &DC[DIR_MPM * numberOfLBnodesCoarse];
}
////////////////////////////////////////////////////////////////////////////////
const unsigned ix = threadIdx.x; // Globaler x-Index
@@ -9587,103 +9587,120 @@ __global__ void scaleFC_RhoSq_3rdMom_comp_27(real* DC,
//////////////////////////////////////////////////////////////////////////
-__device__ void scaleFC_RhoSq_comp_27_Calculation(real *DC, real *DF, unsigned int *neighborCX, unsigned int *neighborCY,
- unsigned int *neighborCZ, unsigned int *neighborFX, unsigned int *neighborFY,
- unsigned int *neighborFZ, unsigned int size_MatC, unsigned int size_MatF,
- bool isEvenTimestep, unsigned int *posC, unsigned int *posFSWB, unsigned int kFC,
- real omCoarse, real omFine, real nu, unsigned int nxC, unsigned int nyC,
- unsigned int nxF, unsigned int nyF, OffFC offFC, const unsigned k)
+__device__ void scaleFC_RhoSq_comp_27_Calculation(
+ real *DC, real *DF,
+ unsigned int *neighborCX,
+ unsigned int *neighborCY,
+ unsigned int *neighborCZ,
+ unsigned int *neighborFX,
+ unsigned int *neighborFY,
+ unsigned int *neighborFZ,
+ unsigned long long numberOfLBnodesCoarse,
+ unsigned long long numberOfLBnodesFine,
+ bool isEvenTimestep,
+ unsigned int *posC,
+ unsigned int *posFSWB,
+ unsigned int kFC,
+ real omCoarse,
+ real omFine,
+ real nu,
+ unsigned int nxC,
+ unsigned int nyC,
+ unsigned int nxF,
+ unsigned int nyF,
+ OffFC offFC,
+ const unsigned k)
{
real *feF, *fwF, *fnF, *fsF, *ftF, *fbF, *fneF, *fswF, *fseF, *fnwF, *fteF, *fbwF, *fbeF, *ftwF, *ftnF, *fbsF,
*fbnF, *ftsF, *fzeroF, *ftneF, *ftswF, *ftseF, *ftnwF, *fbneF, *fbswF, *fbseF, *fbnwF;
- feF = &DF[DIR_P00 * size_MatF];
- fwF = &DF[DIR_M00 * size_MatF];
- fnF = &DF[DIR_0P0 * size_MatF];
- fsF = &DF[DIR_0M0 * size_MatF];
- ftF = &DF[DIR_00P * size_MatF];
- fbF = &DF[DIR_00M * size_MatF];
- fneF = &DF[DIR_PP0 * size_MatF];
- fswF = &DF[DIR_MM0 * size_MatF];
- fseF = &DF[DIR_PM0 * size_MatF];
- fnwF = &DF[DIR_MP0 * size_MatF];
- fteF = &DF[DIR_P0P * size_MatF];
- fbwF = &DF[DIR_M0M * size_MatF];
- fbeF = &DF[DIR_P0M * size_MatF];
- ftwF = &DF[DIR_M0P * size_MatF];
- ftnF = &DF[DIR_0PP * size_MatF];
- fbsF = &DF[DIR_0MM * size_MatF];
- fbnF = &DF[DIR_0PM * size_MatF];
- ftsF = &DF[DIR_0MP * size_MatF];
- fzeroF = &DF[DIR_000 * size_MatF];
- ftneF = &DF[DIR_PPP * size_MatF];
- ftswF = &DF[DIR_MMP * size_MatF];
- ftseF = &DF[DIR_PMP * size_MatF];
- ftnwF = &DF[DIR_MPP * size_MatF];
- fbneF = &DF[DIR_PPM * size_MatF];
- fbswF = &DF[DIR_MMM * size_MatF];
- fbseF = &DF[DIR_PMM * size_MatF];
- fbnwF = &DF[DIR_MPM * size_MatF];
+ feF = &DF[DIR_P00 * numberOfLBnodesFine];
+ fwF = &DF[DIR_M00 * numberOfLBnodesFine];
+ fnF = &DF[DIR_0P0 * numberOfLBnodesFine];
+ fsF = &DF[DIR_0M0 * numberOfLBnodesFine];
+ ftF = &DF[DIR_00P * numberOfLBnodesFine];
+ fbF = &DF[DIR_00M * numberOfLBnodesFine];
+ fneF = &DF[DIR_PP0 * numberOfLBnodesFine];
+ fswF = &DF[DIR_MM0 * numberOfLBnodesFine];
+ fseF = &DF[DIR_PM0 * numberOfLBnodesFine];
+ fnwF = &DF[DIR_MP0 * numberOfLBnodesFine];
+ fteF = &DF[DIR_P0P * numberOfLBnodesFine];
+ fbwF = &DF[DIR_M0M * numberOfLBnodesFine];
+ fbeF = &DF[DIR_P0M * numberOfLBnodesFine];
+ ftwF = &DF[DIR_M0P * numberOfLBnodesFine];
+ ftnF = &DF[DIR_0PP * numberOfLBnodesFine];
+ fbsF = &DF[DIR_0MM * numberOfLBnodesFine];
+ fbnF = &DF[DIR_0PM * numberOfLBnodesFine];
+ ftsF = &DF[DIR_0MP * numberOfLBnodesFine];
+ fzeroF = &DF[DIR_000 * numberOfLBnodesFine];
+ ftneF = &DF[DIR_PPP * numberOfLBnodesFine];
+ ftswF = &DF[DIR_MMP * numberOfLBnodesFine];
+ ftseF = &DF[DIR_PMP * numberOfLBnodesFine];
+ ftnwF = &DF[DIR_MPP * numberOfLBnodesFine];
+ fbneF = &DF[DIR_PPM * numberOfLBnodesFine];
+ fbswF = &DF[DIR_MMM * numberOfLBnodesFine];
+ fbseF = &DF[DIR_PMM * numberOfLBnodesFine];
+ fbnwF = &DF[DIR_MPM * numberOfLBnodesFine];
real *feC, *fwC, *fnC, *fsC, *ftC, *fbC, *fneC, *fswC, *fseC, *fnwC, *fteC, *fbwC, *fbeC, *ftwC, *ftnC, *fbsC,
*fbnC, *ftsC, *fzeroC, *ftneC, *ftswC, *ftseC, *ftnwC, *fbneC, *fbswC, *fbseC, *fbnwC;
if (isEvenTimestep == true) {
- feC = &DC[DIR_P00 * size_MatC];
- fwC = &DC[DIR_M00 * size_MatC];
- fnC = &DC[DIR_0P0 * size_MatC];
- fsC = &DC[DIR_0M0 * size_MatC];
- ftC = &DC[DIR_00P * size_MatC];
- fbC = &DC[DIR_00M * size_MatC];
- fneC = &DC[DIR_PP0 * size_MatC];
- fswC = &DC[DIR_MM0 * size_MatC];
- fseC = &DC[DIR_PM0 * size_MatC];
- fnwC = &DC[DIR_MP0 * size_MatC];
- fteC = &DC[DIR_P0P * size_MatC];
- fbwC = &DC[DIR_M0M * size_MatC];
- fbeC = &DC[DIR_P0M * size_MatC];
- ftwC = &DC[DIR_M0P * size_MatC];
- ftnC = &DC[DIR_0PP * size_MatC];
- fbsC = &DC[DIR_0MM * size_MatC];
- fbnC = &DC[DIR_0PM * size_MatC];
- ftsC = &DC[DIR_0MP * size_MatC];
- fzeroC = &DC[DIR_000 * size_MatC];
- ftneC = &DC[DIR_PPP * size_MatC];
- ftswC = &DC[DIR_MMP * size_MatC];
- ftseC = &DC[DIR_PMP * size_MatC];
- ftnwC = &DC[DIR_MPP * size_MatC];
- fbneC = &DC[DIR_PPM * size_MatC];
- fbswC = &DC[DIR_MMM * size_MatC];
- fbseC = &DC[DIR_PMM * size_MatC];
- fbnwC = &DC[DIR_MPM * size_MatC];
+ feC = &DC[DIR_P00 * numberOfLBnodesCoarse];
+ fwC = &DC[DIR_M00 * numberOfLBnodesCoarse];
+ fnC = &DC[DIR_0P0 * numberOfLBnodesCoarse];
+ fsC = &DC[DIR_0M0 * numberOfLBnodesCoarse];
+ ftC = &DC[DIR_00P * numberOfLBnodesCoarse];
+ fbC = &DC[DIR_00M * numberOfLBnodesCoarse];
+ fneC = &DC[DIR_PP0 * numberOfLBnodesCoarse];
+ fswC = &DC[DIR_MM0 * numberOfLBnodesCoarse];
+ fseC = &DC[DIR_PM0 * numberOfLBnodesCoarse];
+ fnwC = &DC[DIR_MP0 * numberOfLBnodesCoarse];
+ fteC = &DC[DIR_P0P * numberOfLBnodesCoarse];
+ fbwC = &DC[DIR_M0M * numberOfLBnodesCoarse];
+ fbeC = &DC[DIR_P0M * numberOfLBnodesCoarse];
+ ftwC = &DC[DIR_M0P * numberOfLBnodesCoarse];
+ ftnC = &DC[DIR_0PP * numberOfLBnodesCoarse];
+ fbsC = &DC[DIR_0MM * numberOfLBnodesCoarse];
+ fbnC = &DC[DIR_0PM * numberOfLBnodesCoarse];
+ ftsC = &DC[DIR_0MP * numberOfLBnodesCoarse];
+ fzeroC = &DC[DIR_000 * numberOfLBnodesCoarse];
+ ftneC = &DC[DIR_PPP * numberOfLBnodesCoarse];
+ ftswC = &DC[DIR_MMP * numberOfLBnodesCoarse];
+ ftseC = &DC[DIR_PMP * numberOfLBnodesCoarse];
+ ftnwC = &DC[DIR_MPP * numberOfLBnodesCoarse];
+ fbneC = &DC[DIR_PPM * numberOfLBnodesCoarse];
+ fbswC = &DC[DIR_MMM * numberOfLBnodesCoarse];
+ fbseC = &DC[DIR_PMM * numberOfLBnodesCoarse];
+ fbnwC = &DC[DIR_MPM * numberOfLBnodesCoarse];
} else {
- fwC = &DC[DIR_P00 * size_MatC];
- feC = &DC[DIR_M00 * size_MatC];
- fsC = &DC[DIR_0P0 * size_MatC];
- fnC = &DC[DIR_0M0 * size_MatC];
- fbC = &DC[DIR_00P * size_MatC];
- ftC = &DC[DIR_00M * size_MatC];
- fswC = &DC[DIR_PP0 * size_MatC];
- fneC = &DC[DIR_MM0 * size_MatC];
- fnwC = &DC[DIR_PM0 * size_MatC];
- fseC = &DC[DIR_MP0 * size_MatC];
- fbwC = &DC[DIR_P0P * size_MatC];
- fteC = &DC[DIR_M0M * size_MatC];
- ftwC = &DC[DIR_P0M * size_MatC];
- fbeC = &DC[DIR_M0P * size_MatC];
- fbsC = &DC[DIR_0PP * size_MatC];
- ftnC = &DC[DIR_0MM * size_MatC];
- ftsC = &DC[DIR_0PM * size_MatC];
- fbnC = &DC[DIR_0MP * size_MatC];
- fzeroC = &DC[DIR_000 * size_MatC];
- fbswC = &DC[DIR_PPP * size_MatC];
- fbneC = &DC[DIR_MMP * size_MatC];
- fbnwC = &DC[DIR_PMP * size_MatC];
- fbseC = &DC[DIR_MPP * size_MatC];
- ftswC = &DC[DIR_PPM * size_MatC];
- ftneC = &DC[DIR_MMM * size_MatC];
- ftnwC = &DC[DIR_PMM * size_MatC];
- ftseC = &DC[DIR_MPM * size_MatC];
+ fwC = &DC[DIR_P00 * numberOfLBnodesCoarse];
+ feC = &DC[DIR_M00 * numberOfLBnodesCoarse];
+ fsC = &DC[DIR_0P0 * numberOfLBnodesCoarse];
+ fnC = &DC[DIR_0M0 * numberOfLBnodesCoarse];
+ fbC = &DC[DIR_00P * numberOfLBnodesCoarse];
+ ftC = &DC[DIR_00M * numberOfLBnodesCoarse];
+ fswC = &DC[DIR_PP0 * numberOfLBnodesCoarse];
+ fneC = &DC[DIR_MM0 * numberOfLBnodesCoarse];
+ fnwC = &DC[DIR_PM0 * numberOfLBnodesCoarse];
+ fseC = &DC[DIR_MP0 * numberOfLBnodesCoarse];
+ fbwC = &DC[DIR_P0P * numberOfLBnodesCoarse];
+ fteC = &DC[DIR_M0M * numberOfLBnodesCoarse];
+ ftwC = &DC[DIR_P0M * numberOfLBnodesCoarse];
+ fbeC = &DC[DIR_M0P * numberOfLBnodesCoarse];
+ fbsC = &DC[DIR_0PP * numberOfLBnodesCoarse];
+ ftnC = &DC[DIR_0MM * numberOfLBnodesCoarse];
+ ftsC = &DC[DIR_0PM * numberOfLBnodesCoarse];
+ fbnC = &DC[DIR_0MP * numberOfLBnodesCoarse];
+ fzeroC = &DC[DIR_000 * numberOfLBnodesCoarse];
+ fbswC = &DC[DIR_PPP * numberOfLBnodesCoarse];
+ fbneC = &DC[DIR_MMP * numberOfLBnodesCoarse];
+ fbnwC = &DC[DIR_PMP * numberOfLBnodesCoarse];
+ fbseC = &DC[DIR_MPP * numberOfLBnodesCoarse];
+ ftswC = &DC[DIR_PPM * numberOfLBnodesCoarse];
+ ftneC = &DC[DIR_MMM * numberOfLBnodesCoarse];
+ ftnwC = &DC[DIR_PMM * numberOfLBnodesCoarse];
+ ftseC = &DC[DIR_MPM * numberOfLBnodesCoarse];
}
////////////////////////////////////////////////////////////////////////////////
@@ -11064,8 +11081,8 @@ __global__ void scaleFC_RhoSq_comp_27(real* DC,
unsigned int* neighborFX,
unsigned int* neighborFY,
unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
+ unsigned long long numberOfLBnodesCoarse,
+ unsigned long long numberOfLBnodesFine,
bool isEvenTimestep,
unsigned int* posC,
unsigned int* posFSWB,
@@ -11091,7 +11108,7 @@ __global__ void scaleFC_RhoSq_comp_27(real* DC,
//////////////////////////////////////////////////////////////////////////
scaleFC_RhoSq_comp_27_Calculation(DC, DF, neighborCX, neighborCY, neighborCZ, neighborFX, neighborFY, neighborFZ,
- size_MatC, size_MatF, isEvenTimestep, posC, posFSWB, kFC, omCoarse, omFine, nu, nxC,
+ numberOfLBnodesCoarse, numberOfLBnodesFine, isEvenTimestep, posC, posFSWB, kFC, omCoarse, omFine, nu, nxC,
nyC, nxF, nyF, offFC, k);
}
@@ -11157,8 +11174,8 @@ __global__ void scaleFC_staggered_time_comp_27( real* DC,
unsigned int* neighborFX,
unsigned int* neighborFY,
unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
+ unsigned long long numberOfLBnodesCoarse,
+ unsigned long long numberOfLBnodesFine,
bool isEvenTimestep,
unsigned int* posC,
unsigned int* posFSWB,
@@ -11175,96 +11192,96 @@ __global__ void scaleFC_staggered_time_comp_27( real* DC,
real *feF, *fwF, *fnF, *fsF, *ftF, *fbF, *fneF, *fswF, *fseF, *fnwF, *fteF, *fbwF, *fbeF, *ftwF, *ftnF, *fbsF, *fbnF, *ftsF, *fzeroF,
*ftneF, *ftswF, *ftseF, *ftnwF, *fbneF, *fbswF, *fbseF, *fbnwF;
- feF = &DF[DIR_P00 *size_MatF];
- fwF = &DF[DIR_M00 *size_MatF];
- fnF = &DF[DIR_0P0 *size_MatF];
- fsF = &DF[DIR_0M0 *size_MatF];
- ftF = &DF[DIR_00P *size_MatF];
- fbF = &DF[DIR_00M *size_MatF];
- fneF = &DF[DIR_PP0 *size_MatF];
- fswF = &DF[DIR_MM0 *size_MatF];
- fseF = &DF[DIR_PM0 *size_MatF];
- fnwF = &DF[DIR_MP0 *size_MatF];
- fteF = &DF[DIR_P0P *size_MatF];
- fbwF = &DF[DIR_M0M *size_MatF];
- fbeF = &DF[DIR_P0M *size_MatF];
- ftwF = &DF[DIR_M0P *size_MatF];
- ftnF = &DF[DIR_0PP *size_MatF];
- fbsF = &DF[DIR_0MM *size_MatF];
- fbnF = &DF[DIR_0PM *size_MatF];
- ftsF = &DF[DIR_0MP *size_MatF];
- fzeroF = &DF[DIR_000*size_MatF];
- ftneF = &DF[DIR_PPP *size_MatF];
- ftswF = &DF[DIR_MMP *size_MatF];
- ftseF = &DF[DIR_PMP *size_MatF];
- ftnwF = &DF[DIR_MPP *size_MatF];
- fbneF = &DF[DIR_PPM *size_MatF];
- fbswF = &DF[DIR_MMM *size_MatF];
- fbseF = &DF[DIR_PMM *size_MatF];
- fbnwF = &DF[DIR_MPM *size_MatF];
+ feF = &DF[DIR_P00 * numberOfLBnodesFine];
+ fwF = &DF[DIR_M00 * numberOfLBnodesFine];
+ fnF = &DF[DIR_0P0 * numberOfLBnodesFine];
+ fsF = &DF[DIR_0M0 * numberOfLBnodesFine];
+ ftF = &DF[DIR_00P * numberOfLBnodesFine];
+ fbF = &DF[DIR_00M * numberOfLBnodesFine];
+ fneF = &DF[DIR_PP0 * numberOfLBnodesFine];
+ fswF = &DF[DIR_MM0 * numberOfLBnodesFine];
+ fseF = &DF[DIR_PM0 * numberOfLBnodesFine];
+ fnwF = &DF[DIR_MP0 * numberOfLBnodesFine];
+ fteF = &DF[DIR_P0P * numberOfLBnodesFine];
+ fbwF = &DF[DIR_M0M * numberOfLBnodesFine];
+ fbeF = &DF[DIR_P0M * numberOfLBnodesFine];
+ ftwF = &DF[DIR_M0P * numberOfLBnodesFine];
+ ftnF = &DF[DIR_0PP * numberOfLBnodesFine];
+ fbsF = &DF[DIR_0MM * numberOfLBnodesFine];
+ fbnF = &DF[DIR_0PM * numberOfLBnodesFine];
+ ftsF = &DF[DIR_0MP * numberOfLBnodesFine];
+ fzeroF = &DF[DIR_000 * numberOfLBnodesFine];
+ ftneF = &DF[DIR_PPP * numberOfLBnodesFine];
+ ftswF = &DF[DIR_MMP * numberOfLBnodesFine];
+ ftseF = &DF[DIR_PMP * numberOfLBnodesFine];
+ ftnwF = &DF[DIR_MPP * numberOfLBnodesFine];
+ fbneF = &DF[DIR_PPM * numberOfLBnodesFine];
+ fbswF = &DF[DIR_MMM * numberOfLBnodesFine];
+ fbseF = &DF[DIR_PMM * numberOfLBnodesFine];
+ fbnwF = &DF[DIR_MPM * numberOfLBnodesFine];
real *feC, *fwC, *fnC, *fsC, *ftC, *fbC, *fneC, *fswC, *fseC, *fnwC, *fteC, *fbwC, *fbeC, *ftwC, *ftnC, *fbsC, *fbnC, *ftsC, *fzeroC,
*ftneC, *ftswC, *ftseC, *ftnwC, *fbneC, *fbswC, *fbseC, *fbnwC;
if (isEvenTimestep==true)
{
- feC = &DC[DIR_P00 *size_MatC];
- fwC = &DC[DIR_M00 *size_MatC];
- fnC = &DC[DIR_0P0 *size_MatC];
- fsC = &DC[DIR_0M0 *size_MatC];
- ftC = &DC[DIR_00P *size_MatC];
- fbC = &DC[DIR_00M *size_MatC];
- fneC = &DC[DIR_PP0 *size_MatC];
- fswC = &DC[DIR_MM0 *size_MatC];
- fseC = &DC[DIR_PM0 *size_MatC];
- fnwC = &DC[DIR_MP0 *size_MatC];
- fteC = &DC[DIR_P0P *size_MatC];
- fbwC = &DC[DIR_M0M *size_MatC];
- fbeC = &DC[DIR_P0M *size_MatC];
- ftwC = &DC[DIR_M0P *size_MatC];
- ftnC = &DC[DIR_0PP *size_MatC];
- fbsC = &DC[DIR_0MM *size_MatC];
- fbnC = &DC[DIR_0PM *size_MatC];
- ftsC = &DC[DIR_0MP *size_MatC];
- fzeroC = &DC[DIR_000*size_MatC];
- ftneC = &DC[DIR_PPP *size_MatC];
- ftswC = &DC[DIR_MMP *size_MatC];
- ftseC = &DC[DIR_PMP *size_MatC];
- ftnwC = &DC[DIR_MPP *size_MatC];
- fbneC = &DC[DIR_PPM *size_MatC];
- fbswC = &DC[DIR_MMM *size_MatC];
- fbseC = &DC[DIR_PMM *size_MatC];
- fbnwC = &DC[DIR_MPM *size_MatC];
+ feC = &DC[DIR_P00 * numberOfLBnodesCoarse];
+ fwC = &DC[DIR_M00 * numberOfLBnodesCoarse];
+ fnC = &DC[DIR_0P0 * numberOfLBnodesCoarse];
+ fsC = &DC[DIR_0M0 * numberOfLBnodesCoarse];
+ ftC = &DC[DIR_00P * numberOfLBnodesCoarse];
+ fbC = &DC[DIR_00M * numberOfLBnodesCoarse];
+ fneC = &DC[DIR_PP0 * numberOfLBnodesCoarse];
+ fswC = &DC[DIR_MM0 * numberOfLBnodesCoarse];
+ fseC = &DC[DIR_PM0 * numberOfLBnodesCoarse];
+ fnwC = &DC[DIR_MP0 * numberOfLBnodesCoarse];
+ fteC = &DC[DIR_P0P * numberOfLBnodesCoarse];
+ fbwC = &DC[DIR_M0M * numberOfLBnodesCoarse];
+ fbeC = &DC[DIR_P0M * numberOfLBnodesCoarse];
+ ftwC = &DC[DIR_M0P * numberOfLBnodesCoarse];
+ ftnC = &DC[DIR_0PP * numberOfLBnodesCoarse];
+ fbsC = &DC[DIR_0MM * numberOfLBnodesCoarse];
+ fbnC = &DC[DIR_0PM * numberOfLBnodesCoarse];
+ ftsC = &DC[DIR_0MP * numberOfLBnodesCoarse];
+ fzeroC = &DC[DIR_000 * numberOfLBnodesCoarse];
+ ftneC = &DC[DIR_PPP * numberOfLBnodesCoarse];
+ ftswC = &DC[DIR_MMP * numberOfLBnodesCoarse];
+ ftseC = &DC[DIR_PMP * numberOfLBnodesCoarse];
+ ftnwC = &DC[DIR_MPP * numberOfLBnodesCoarse];
+ fbneC = &DC[DIR_PPM * numberOfLBnodesCoarse];
+ fbswC = &DC[DIR_MMM * numberOfLBnodesCoarse];
+ fbseC = &DC[DIR_PMM * numberOfLBnodesCoarse];
+ fbnwC = &DC[DIR_MPM * numberOfLBnodesCoarse];
}
else
{
- fwC = &DC[DIR_P00 *size_MatC];
- feC = &DC[DIR_M00 *size_MatC];
- fsC = &DC[DIR_0P0 *size_MatC];
- fnC = &DC[DIR_0M0 *size_MatC];
- fbC = &DC[DIR_00P *size_MatC];
- ftC = &DC[DIR_00M *size_MatC];
- fswC = &DC[DIR_PP0 *size_MatC];
- fneC = &DC[DIR_MM0 *size_MatC];
- fnwC = &DC[DIR_PM0 *size_MatC];
- fseC = &DC[DIR_MP0 *size_MatC];
- fbwC = &DC[DIR_P0P *size_MatC];
- fteC = &DC[DIR_M0M *size_MatC];
- ftwC = &DC[DIR_P0M *size_MatC];
- fbeC = &DC[DIR_M0P *size_MatC];
- fbsC = &DC[DIR_0PP *size_MatC];
- ftnC = &DC[DIR_0MM *size_MatC];
- ftsC = &DC[DIR_0PM *size_MatC];
- fbnC = &DC[DIR_0MP *size_MatC];
- fzeroC = &DC[DIR_000*size_MatC];
- fbswC = &DC[DIR_PPP *size_MatC];
- fbneC = &DC[DIR_MMP *size_MatC];
- fbnwC = &DC[DIR_PMP *size_MatC];
- fbseC = &DC[DIR_MPP *size_MatC];
- ftswC = &DC[DIR_PPM *size_MatC];
- ftneC = &DC[DIR_MMM *size_MatC];
- ftnwC = &DC[DIR_PMM *size_MatC];
- ftseC = &DC[DIR_MPM *size_MatC];
+ fwC = &DC[DIR_P00 * numberOfLBnodesCoarse];
+ feC = &DC[DIR_M00 * numberOfLBnodesCoarse];
+ fsC = &DC[DIR_0P0 * numberOfLBnodesCoarse];
+ fnC = &DC[DIR_0M0 * numberOfLBnodesCoarse];
+ fbC = &DC[DIR_00P * numberOfLBnodesCoarse];
+ ftC = &DC[DIR_00M * numberOfLBnodesCoarse];
+ fswC = &DC[DIR_PP0 * numberOfLBnodesCoarse];
+ fneC = &DC[DIR_MM0 * numberOfLBnodesCoarse];
+ fnwC = &DC[DIR_PM0 * numberOfLBnodesCoarse];
+ fseC = &DC[DIR_MP0 * numberOfLBnodesCoarse];
+ fbwC = &DC[DIR_P0P * numberOfLBnodesCoarse];
+ fteC = &DC[DIR_M0M * numberOfLBnodesCoarse];
+ ftwC = &DC[DIR_P0M * numberOfLBnodesCoarse];
+ fbeC = &DC[DIR_M0P * numberOfLBnodesCoarse];
+ fbsC = &DC[DIR_0PP * numberOfLBnodesCoarse];
+ ftnC = &DC[DIR_0MM * numberOfLBnodesCoarse];
+ ftsC = &DC[DIR_0PM * numberOfLBnodesCoarse];
+ fbnC = &DC[DIR_0MP * numberOfLBnodesCoarse];
+ fzeroC = &DC[DIR_000 * numberOfLBnodesCoarse];
+ fbswC = &DC[DIR_PPP * numberOfLBnodesCoarse];
+ fbneC = &DC[DIR_MMP * numberOfLBnodesCoarse];
+ fbnwC = &DC[DIR_PMP * numberOfLBnodesCoarse];
+ fbseC = &DC[DIR_MPP * numberOfLBnodesCoarse];
+ ftswC = &DC[DIR_PPM * numberOfLBnodesCoarse];
+ ftneC = &DC[DIR_MMM * numberOfLBnodesCoarse];
+ ftnwC = &DC[DIR_PMM * numberOfLBnodesCoarse];
+ ftseC = &DC[DIR_MPM * numberOfLBnodesCoarse];
}
////////////////////////////////////////////////////////////////////////////////
const unsigned ix = threadIdx.x; // Globaler x-Index
@@ -11755,827 +11772,6 @@ __global__ void scaleFC_staggered_time_comp_27( real* DC,
kxxMyyFromfcNEQ_NEB = -c3o2*omegaS *((f_BW+f_W+f_TW-f_BS-f_S-f_TS-f_BN-f_N-f_TN+f_BE+f_E+f_TE ) / (c1o1 + drho_NEB) - ((vx1_NEB*vx1_NEB-vx2_NEB*vx2_NEB)));
kxxMzzFromfcNEQ_NEB = -c3o2*omegaS *((f_SW+f_W+f_NW-f_BS-f_TS-f_B-f_T-f_BN-f_TN+f_SE+f_E+f_NE ) / (c1o1 + drho_NEB) - ((vx1_NEB*vx1_NEB-vx3_NEB*vx3_NEB)));
- // //////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
- ////pointertausch
- // if (isEvenTimestep==false)
- // {
- // feC = &DC[DIR_P00 *size_MatC];
- // fwC = &DC[DIR_M00 *size_MatC];
- // fnC = &DC[DIR_0P0 *size_MatC];
- // fsC = &DC[DIR_0M0 *size_MatC];
- // ftC = &DC[DIR_00P *size_MatC];
- // fbC = &DC[DIR_00M *size_MatC];
- // fneC = &DC[DIR_PP0 *size_MatC];
- // fswC = &DC[DIR_MM0 *size_MatC];
- // fseC = &DC[DIR_PM0 *size_MatC];
- // fnwC = &DC[DIR_MP0 *size_MatC];
- // fteC = &DC[DIR_P0P *size_MatC];
- // fbwC = &DC[DIR_M0M *size_MatC];
- // fbeC = &DC[DIR_P0M *size_MatC];
- // ftwC = &DC[DIR_M0P *size_MatC];
- // ftnC = &DC[DIR_0PP *size_MatC];
- // fbsC = &DC[DIR_0MM *size_MatC];
- // fbnC = &DC[DIR_0PM *size_MatC];
- // ftsC = &DC[DIR_0MP *size_MatC];
- // fzeroC = &DC[DIR_000*size_MatC];
- // ftneC = &DC[DIR_PPP *size_MatC];
- // ftswC = &DC[DIR_MMP *size_MatC];
- // ftseC = &DC[DIR_PMP *size_MatC];
- // ftnwC = &DC[DIR_MPP *size_MatC];
- // fbneC = &DC[DIR_PPM *size_MatC];
- // fbswC = &DC[DIR_MMM *size_MatC];
- // fbseC = &DC[DIR_PMM *size_MatC];
- // fbnwC = &DC[DIR_MPM *size_MatC];
- // }
- // else
- // {
- // fwC = &DC[DIR_P00 *size_MatC];
- // feC = &DC[DIR_M00 *size_MatC];
- // fsC = &DC[DIR_0P0 *size_MatC];
- // fnC = &DC[DIR_0M0 *size_MatC];
- // fbC = &DC[DIR_00P *size_MatC];
- // ftC = &DC[DIR_00M *size_MatC];
- // fswC = &DC[DIR_PP0 *size_MatC];
- // fneC = &DC[DIR_MM0 *size_MatC];
- // fnwC = &DC[DIR_PM0 *size_MatC];
- // fseC = &DC[DIR_MP0 *size_MatC];
- // fbwC = &DC[DIR_P0P *size_MatC];
- // fteC = &DC[DIR_M0M *size_MatC];
- // ftwC = &DC[DIR_P0M *size_MatC];
- // fbeC = &DC[DIR_M0P *size_MatC];
- // fbsC = &DC[DIR_0PP *size_MatC];
- // ftnC = &DC[DIR_0MM *size_MatC];
- // ftsC = &DC[DIR_0PM *size_MatC];
- // fbnC = &DC[DIR_0MP *size_MatC];
- // fzeroC = &DC[DIR_000*size_MatC];
- // fbswC = &DC[DIR_PPP *size_MatC];
- // fbneC = &DC[DIR_MMP *size_MatC];
- // fbnwC = &DC[DIR_PMP *size_MatC];
- // fbseC = &DC[DIR_MPP *size_MatC];
- // ftswC = &DC[DIR_PPM *size_MatC];
- // ftneC = &DC[DIR_MMM *size_MatC];
- // ftnwC = &DC[DIR_PMM *size_MatC];
- // ftseC = &DC[DIR_MPM *size_MatC];
- // }
-
- // real rho_tmp;
- //real vx1_tmp;
- //real vx2_tmp;
- //real vx3_tmp;
-
- // //////////////////////////////////////////////////////////////////////////
- // xoff = offFC.xOffFC[k];
- // yoff = offFC.yOffFC[k];
- // zoff = offFC.zOffFC[k];
- // xoff_sq = xoff * xoff;
- // yoff_sq = yoff * yoff;
- // zoff_sq = zoff * zoff;
- // //////////////////////////////////////////////////////////////////////////
- // //SWB//
- // //////////////////////////////////////////////////////////////////////////
- // //index 0
- // k0zero= posFSWB[k];
- // k0w = neighborFX[k0zero];
- // k0s = neighborFY[k0zero];
- // k0b = neighborFZ[k0zero];
- // k0sw = neighborFY[k0w];
- // k0bw = neighborFZ[k0w];
- // k0bs = neighborFZ[k0s];
- // k0bsw = neighborFZ[k0sw];
- // //////////////////////////////////////////////////////////////////////////
- // //index
- // kzero= k0zero;
- // kw = k0w;
- // ks = k0s;
- // kb = k0b;
- // ksw = k0sw;
- // kbw = k0bw;
- // kbs = k0bs;
- // kbsw = k0bsw;
- // ////////////////////////////////////////////////////////////////////////////////
- // f_E = fwF[kw];
- // f_W = feF[kzero];
- // f_N = fsF[ks];
- // f_S = fnF[kzero];
- // f_T = fbF[kb];
- // f_B = ftF[kzero];
- // f_NE = fswF[ksw];
- // f_SW = fneF[kzero];
- // f_SE = fnwF[kw];
- // f_NW = fseF[ks];
- // f_TE = fbwF[kbw];
- // f_BW = fteF[kzero];
- // f_BE = ftwF[kw];
- // f_TW = fbeF[kb];
- // f_TN = fbsF[kbs];
- // f_BS = ftnF[kzero];
- // f_BN = ftsF[ks];
- // f_TS = fbnF[kb];
- // f_ZERO = fzeroF[kzero];
- // f_TNE = fbswF[kbsw];
- // f_TSW = fbneF[kb];
- // f_TSE = fbnwF[kbw];
- // f_TNW = fbseF[kbs];
- // f_BNE = ftswF[ksw];
- // f_BSW = ftneF[kzero];
- // f_BSE = ftnwF[kw];
- // f_BNW = ftseF[ks];
-
- // //drho_SWB = f_E+f_W+f_N+f_S+f_T+f_B+f_NE+f_SW+f_SE+f_NW+f_TE+f_BW+f_BE+f_TW+f_TN+f_BS+f_BN+f_TS+f_ZERO+f_TNE+f_TSW+f_TSE+f_TNW+f_BNE+f_BSW+f_BSE+f_BNW;
- // //vx1_SWB = (((f_TNE-f_BSW)+(f_TSE-f_BNW)+(f_BNE-f_TSW)+(f_BSE-f_TNW)) + (((f_NE-f_SW)+(f_TE-f_BW))+((f_SE-f_NW)+(f_BE-f_TW))) + (f_E-f_W))/(one + drho_SWB);
- ////vx2_SWB = (((f_TNE-f_BSW)+(f_TNW-f_BSE)+(f_BNE-f_TSW)+(f_BNW-f_TSE)) + (((f_NE-f_SW)+(f_TN-f_BS))+((f_BN-f_TS)+(f_NW-f_SE))) + (f_N-f_S))/(one + drho_SWB);
- ////vx3_SWB = (((f_TNE-f_BSW)+(f_TNW-f_BSE)+(f_TSE-f_BNW)+(f_TSW-f_BNE)) + (((f_TE-f_BW)+(f_TN-f_BS))+((f_TW-f_BE)+(f_TS-f_BN))) + (f_T-f_B))/(one + drho_SWB);
-
- // //kxyFromfcNEQ_SWB = -three*omegaS/(one-omegaS)*((f_SW+f_BSW+f_TSW-f_NW-f_BNW-f_TNW-f_SE-f_BSE-f_TSE+f_NE+f_BNE+f_TNE ) / (one + drho_SWB) - ((vx1_SWB*vx2_SWB)));
- // //kyzFromfcNEQ_SWB = -three*omegaS/(one-omegaS)*((f_BS+f_BSE+f_BSW-f_TS-f_TSE-f_TSW-f_BN-f_BNE-f_BNW+f_TN+f_TNE+f_TNW ) / (one + drho_SWB) - ((vx2_SWB*vx3_SWB)));
- // //kxzFromfcNEQ_SWB = -three*omegaS/(one-omegaS)*((f_BW+f_BSW+f_BNW-f_TW-f_TSW-f_TNW-f_BE-f_BSE-f_BNE+f_TE+f_TSE+f_TNE ) / (one + drho_SWB) - ((vx1_SWB*vx3_SWB)));
- // //kxxMyyFromfcNEQ_SWB = -c3o2*omegaS/(one-omegaS) *((f_BW+f_W+f_TW-f_BS-f_S-f_TS-f_BN-f_N-f_TN+f_BE+f_E+f_TE ) / (one + drho_SWB) - ((vx1_SWB*vx1_SWB-vx2_SWB*vx2_SWB)));
- // //kxxMzzFromfcNEQ_SWB = -c3o2*omegaS/(one-omegaS) *((f_SW+f_W+f_NW-f_BS-f_TS-f_B-f_T-f_BN-f_TN+f_SE+f_E+f_NE ) / (one + drho_SWB) - ((vx1_SWB*vx1_SWB-vx3_SWB*vx3_SWB)));
-
- //rho_tmp = f_E+f_W+f_N+f_S+f_T+f_B+f_NE+f_SW+f_SE+f_NW+f_TE+f_BW+f_BE+f_TW+f_TN+f_BS+f_BN+f_TS+f_ZERO+f_TNE+f_TSW+f_TSE+f_TNW+f_BNE+f_BSW+f_BSE+f_BNW;
- //
- //drho_SWB += rho_tmp;
-
- //vx1_tmp = (((f_TNE-f_BSW)+(f_TSE-f_BNW)+(f_BNE-f_TSW)+(f_BSE-f_TNW)) + (((f_NE-f_SW)+(f_TE-f_BW))+((f_SE-f_NW)+(f_BE-f_TW))) + (f_E-f_W))/(one + rho_tmp);
- //vx2_tmp = (((f_TNE-f_BSW)+(f_TNW-f_BSE)+(f_BNE-f_TSW)+(f_BNW-f_TSE)) + (((f_NE-f_SW)+(f_TN-f_BS))+((f_BN-f_TS)+(f_NW-f_SE))) + (f_N-f_S))/(one + rho_tmp);
- //vx3_tmp = (((f_TNE-f_BSW)+(f_TNW-f_BSE)+(f_TSE-f_BNW)+(f_TSW-f_BNE)) + (((f_TE-f_BW)+(f_TN-f_BS))+((f_TW-f_BE)+(f_TS-f_BN))) + (f_T-f_B))/(one + rho_tmp);
-
- // vx1_SWB += vx1_tmp;
- //vx2_SWB += vx2_tmp;
- //vx3_SWB += vx3_tmp;
-
- // drho_SWB *= c1o2;
- // vx1_SWB *= c1o2;
- //vx2_SWB *= c1o2;
- //vx3_SWB *= c1o2;
-
- // kxyFromfcNEQ_SWB += -three*omegaS/(one-omegaS)*((f_SW+f_BSW+f_TSW-f_NW-f_BNW-f_TNW-f_SE-f_BSE-f_TSE+f_NE+f_BNE+f_TNE ) / (one + rho_tmp) - ((vx1_tmp*vx2_tmp)));
- // kyzFromfcNEQ_SWB += -three*omegaS/(one-omegaS)*((f_BS+f_BSE+f_BSW-f_TS-f_TSE-f_TSW-f_BN-f_BNE-f_BNW+f_TN+f_TNE+f_TNW ) / (one + rho_tmp) - ((vx2_tmp*vx3_tmp)));
- // kxzFromfcNEQ_SWB += -three*omegaS/(one-omegaS)*((f_BW+f_BSW+f_BNW-f_TW-f_TSW-f_TNW-f_BE-f_BSE-f_BNE+f_TE+f_TSE+f_TNE ) / (one + rho_tmp) - ((vx1_tmp*vx3_tmp)));
- // kxxMyyFromfcNEQ_SWB += -c3o2*omegaS/(one-omegaS) *((f_BW+f_W+f_TW-f_BS-f_S-f_TS-f_BN-f_N-f_TN+f_BE+f_E+f_TE ) / (one + rho_tmp) - ((vx1_tmp*vx1_tmp-vx2_tmp*vx2_tmp)));
- // kxxMzzFromfcNEQ_SWB += -c3o2*omegaS/(one-omegaS) *((f_SW+f_W+f_NW-f_BS-f_TS-f_B-f_T-f_BN-f_TN+f_SE+f_E+f_NE ) / (one + rho_tmp) - ((vx1_tmp*vx1_tmp-vx3_tmp*vx3_tmp)));
-
- //kxyFromfcNEQ_SWB *= c1o2;
- //kyzFromfcNEQ_SWB *= c1o2;
- //kxzFromfcNEQ_SWB *= c1o2;
- //kxxMyyFromfcNEQ_SWB *= c1o2;
- //kxxMzzFromfcNEQ_SWB *= c1o2;
-
- // //////////////////////////////////////////////////////////////////////////
- // //SWT//
- // //////////////////////////////////////////////////////////////////////////
- // //index
- // kzero= kb;
- // kw = kbw;
- // ks = kbs;
- // kb = neighborFZ[kb];
- // ksw = kbsw;
- // kbw = neighborFZ[kbw];
- // kbs = neighborFZ[kbs];
- // kbsw = neighborFZ[kbsw];
- // ////////////////////////////////////////////////////////////////////////////////
- // f_E = fwF[kw];
- // f_W = feF[kzero];
- // f_N = fsF[ks];
- // f_S = fnF[kzero];
- // f_T = fbF[kb];
- // f_B = ftF[kzero];
- // f_NE = fswF[ksw];
- // f_SW = fneF[kzero];
- // f_SE = fnwF[kw];
- // f_NW = fseF[ks];
- // f_TE = fbwF[kbw];
- // f_BW = fteF[kzero];
- // f_BE = ftwF[kw];
- // f_TW = fbeF[kb];
- // f_TN = fbsF[kbs];
- // f_BS = ftnF[kzero];
- // f_BN = ftsF[ks];
- // f_TS = fbnF[kb];
- // f_ZERO = fzeroF[kzero];
- // f_TNE = fbswF[kbsw];
- // f_TSW = fbneF[kb];
- // f_TSE = fbnwF[kbw];
- // f_TNW = fbseF[kbs];
- // f_BNE = ftswF[ksw];
- // f_BSW = ftneF[kzero];
- // f_BSE = ftnwF[kw];
- // f_BNW = ftseF[ks];
-
- // //drho_SWT = f_E+f_W+f_N+f_S+f_T+f_B+f_NE+f_SW+f_SE+f_NW+f_TE+f_BW+f_BE+f_TW+f_TN+f_BS+f_BN+f_TS+f_ZERO+f_TNE+f_TSW+f_TSE+f_TNW+f_BNE+f_BSW+f_BSE+f_BNW;
- // //vx1_SWT = (((f_TNE-f_BSW)+(f_TSE-f_BNW)+(f_BNE-f_TSW)+(f_BSE-f_TNW)) + (((f_NE-f_SW)+(f_TE-f_BW))+((f_SE-f_NW)+(f_BE-f_TW))) + (f_E-f_W))/(one + drho_SWT);
- ////vx2_SWT = (((f_TNE-f_BSW)+(f_TNW-f_BSE)+(f_BNE-f_TSW)+(f_BNW-f_TSE)) + (((f_NE-f_SW)+(f_TN-f_BS))+((f_BN-f_TS)+(f_NW-f_SE))) + (f_N-f_S))/(one + drho_SWT);
- ////vx3_SWT = (((f_TNE-f_BSW)+(f_TNW-f_BSE)+(f_TSE-f_BNW)+(f_TSW-f_BNE)) + (((f_TE-f_BW)+(f_TN-f_BS))+((f_TW-f_BE)+(f_TS-f_BN))) + (f_T-f_B))/(one + drho_SWT);
-
- // //kxyFromfcNEQ_SWT = -three*omegaS/(one-omegaS)*((f_SW+f_BSW+f_TSW-f_NW-f_BNW-f_TNW-f_SE-f_BSE-f_TSE+f_NE+f_BNE+f_TNE ) / (one + drho_SWT) - ((vx1_SWT*vx2_SWT)));
- // //kyzFromfcNEQ_SWT = -three*omegaS/(one-omegaS)*((f_BS+f_BSE+f_BSW-f_TS-f_TSE-f_TSW-f_BN-f_BNE-f_BNW+f_TN+f_TNE+f_TNW ) / (one + drho_SWT) - ((vx2_SWT*vx3_SWT)));
- // //kxzFromfcNEQ_SWT = -three*omegaS/(one-omegaS)*((f_BW+f_BSW+f_BNW-f_TW-f_TSW-f_TNW-f_BE-f_BSE-f_BNE+f_TE+f_TSE+f_TNE ) / (one + drho_SWT) - ((vx1_SWT*vx3_SWT)));
- // //kxxMyyFromfcNEQ_SWT = -c3o2*omegaS/(one-omegaS) *((f_BW+f_W+f_TW-f_BS-f_S-f_TS-f_BN-f_N-f_TN+f_BE+f_E+f_TE ) / (one + drho_SWT) - ((vx1_SWT*vx1_SWT-vx2_SWT*vx2_SWT)));
- // //kxxMzzFromfcNEQ_SWT = -c3o2*omegaS/(one-omegaS) *((f_SW+f_W+f_NW-f_BS-f_TS-f_B-f_T-f_BN-f_TN+f_SE+f_E+f_NE ) / (one + drho_SWT) - ((vx1_SWT*vx1_SWT-vx3_SWT*vx3_SWT)));
-
- //rho_tmp = f_E+f_W+f_N+f_S+f_T+f_B+f_NE+f_SW+f_SE+f_NW+f_TE+f_BW+f_BE+f_TW+f_TN+f_BS+f_BN+f_TS+f_ZERO+f_TNE+f_TSW+f_TSE+f_TNW+f_BNE+f_BSW+f_BSE+f_BNW;
- //
- //drho_SWT += rho_tmp;
-
- //vx1_tmp = (((f_TNE-f_BSW)+(f_TSE-f_BNW)+(f_BNE-f_TSW)+(f_BSE-f_TNW)) + (((f_NE-f_SW)+(f_TE-f_BW))+((f_SE-f_NW)+(f_BE-f_TW))) + (f_E-f_W))/(one + rho_tmp);
- //vx2_tmp = (((f_TNE-f_BSW)+(f_TNW-f_BSE)+(f_BNE-f_TSW)+(f_BNW-f_TSE)) + (((f_NE-f_SW)+(f_TN-f_BS))+((f_BN-f_TS)+(f_NW-f_SE))) + (f_N-f_S))/(one + rho_tmp);
- //vx3_tmp = (((f_TNE-f_BSW)+(f_TNW-f_BSE)+(f_TSE-f_BNW)+(f_TSW-f_BNE)) + (((f_TE-f_BW)+(f_TN-f_BS))+((f_TW-f_BE)+(f_TS-f_BN))) + (f_T-f_B))/(one + rho_tmp);
-
- // vx1_SWT += vx1_tmp;
- //vx2_SWT += vx2_tmp;
- //vx3_SWT += vx3_tmp;
-
- // drho_SWT *= c1o2;
- // vx1_SWT *= c1o2;
- //vx2_SWT *= c1o2;
- //vx3_SWT *= c1o2;
-
- // kxyFromfcNEQ_SWT += -three*omegaS/(one-omegaS)*((f_SW+f_BSW+f_TSW-f_NW-f_BNW-f_TNW-f_SE-f_BSE-f_TSE+f_NE+f_BNE+f_TNE ) / (one + rho_tmp) - ((vx1_tmp*vx2_tmp)));
- // kyzFromfcNEQ_SWT += -three*omegaS/(one-omegaS)*((f_BS+f_BSE+f_BSW-f_TS-f_TSE-f_TSW-f_BN-f_BNE-f_BNW+f_TN+f_TNE+f_TNW ) / (one + rho_tmp) - ((vx2_tmp*vx3_tmp)));
- // kxzFromfcNEQ_SWT += -three*omegaS/(one-omegaS)*((f_BW+f_BSW+f_BNW-f_TW-f_TSW-f_TNW-f_BE-f_BSE-f_BNE+f_TE+f_TSE+f_TNE ) / (one + rho_tmp) - ((vx1_tmp*vx3_tmp)));
- // kxxMyyFromfcNEQ_SWT += -c3o2*omegaS/(one-omegaS) *((f_BW+f_W+f_TW-f_BS-f_S-f_TS-f_BN-f_N-f_TN+f_BE+f_E+f_TE ) / (one + rho_tmp) - ((vx1_tmp*vx1_tmp-vx2_tmp*vx2_tmp)));
- // kxxMzzFromfcNEQ_SWT += -c3o2*omegaS/(one-omegaS) *((f_SW+f_W+f_NW-f_BS-f_TS-f_B-f_T-f_BN-f_TN+f_SE+f_E+f_NE ) / (one + rho_tmp) - ((vx1_tmp*vx1_tmp-vx3_tmp*vx3_tmp)));
-
- //kxyFromfcNEQ_SWT *= c1o2;
- //kyzFromfcNEQ_SWT *= c1o2;
- //kxzFromfcNEQ_SWT *= c1o2;
- //kxxMyyFromfcNEQ_SWT *= c1o2;
- //kxxMzzFromfcNEQ_SWT *= c1o2;
-
- // //////////////////////////////////////////////////////////////////////////
- // //SET//
- // //////////////////////////////////////////////////////////////////////////
- // //index
- // kzero= kw;
- // kw = neighborFX[kw];
- // ks = ksw;
- // kb = kbw;
- // ksw = neighborFX[ksw];
- // kbw = neighborFX[kbw];
- // kbs = kbsw;
- // kbsw = neighborFX[kbsw];
- // ////////////////////////////////////////////////////////////////////////////////
- // f_E = fwF[kw];
- // f_W = feF[kzero];
- // f_N = fsF[ks];
- // f_S = fnF[kzero];
- // f_T = fbF[kb];
- // f_B = ftF[kzero];
- // f_NE = fswF[ksw];
- // f_SW = fneF[kzero];
- // f_SE = fnwF[kw];
- // f_NW = fseF[ks];
- // f_TE = fbwF[kbw];
- // f_BW = fteF[kzero];
- // f_BE = ftwF[kw];
- // f_TW = fbeF[kb];
- // f_TN = fbsF[kbs];
- // f_BS = ftnF[kzero];
- // f_BN = ftsF[ks];
- // f_TS = fbnF[kb];
- // f_ZERO = fzeroF[kzero];
- // f_TNE = fbswF[kbsw];
- // f_TSW = fbneF[kb];
- // f_TSE = fbnwF[kbw];
- // f_TNW = fbseF[kbs];
- // f_BNE = ftswF[ksw];
- // f_BSW = ftneF[kzero];
- // f_BSE = ftnwF[kw];
- // f_BNW = ftseF[ks];
-
- // //drho_SET = f_E+f_W+f_N+f_S+f_T+f_B+f_NE+f_SW+f_SE+f_NW+f_TE+f_BW+f_BE+f_TW+f_TN+f_BS+f_BN+f_TS+f_ZERO+f_TNE+f_TSW+f_TSE+f_TNW+f_BNE+f_BSW+f_BSE+f_BNW;
- // //vx1_SET = (((f_TNE-f_BSW)+(f_TSE-f_BNW)+(f_BNE-f_TSW)+(f_BSE-f_TNW)) + (((f_NE-f_SW)+(f_TE-f_BW))+((f_SE-f_NW)+(f_BE-f_TW))) + (f_E-f_W))/(one + drho_SET);
- ////vx2_SET = (((f_TNE-f_BSW)+(f_TNW-f_BSE)+(f_BNE-f_TSW)+(f_BNW-f_TSE)) + (((f_NE-f_SW)+(f_TN-f_BS))+((f_BN-f_TS)+(f_NW-f_SE))) + (f_N-f_S))/(one + drho_SET);
- ////vx3_SET = (((f_TNE-f_BSW)+(f_TNW-f_BSE)+(f_TSE-f_BNW)+(f_TSW-f_BNE)) + (((f_TE-f_BW)+(f_TN-f_BS))+((f_TW-f_BE)+(f_TS-f_BN))) + (f_T-f_B))/(one + drho_SET);
-
- // //kxyFromfcNEQ_SET = -three*omegaS/(one-omegaS)*((f_SW+f_BSW+f_TSW-f_NW-f_BNW-f_TNW-f_SE-f_BSE-f_TSE+f_NE+f_BNE+f_TNE ) / (one + drho_SET) - ((vx1_SET*vx2_SET)));
- // //kyzFromfcNEQ_SET = -three*omegaS/(one-omegaS)*((f_BS+f_BSE+f_BSW-f_TS-f_TSE-f_TSW-f_BN-f_BNE-f_BNW+f_TN+f_TNE+f_TNW ) / (one + drho_SET) - ((vx2_SET*vx3_SET)));
- // //kxzFromfcNEQ_SET = -three*omegaS/(one-omegaS)*((f_BW+f_BSW+f_BNW-f_TW-f_TSW-f_TNW-f_BE-f_BSE-f_BNE+f_TE+f_TSE+f_TNE ) / (one + drho_SET) - ((vx1_SET*vx3_SET)));
- // //kxxMyyFromfcNEQ_SET = -c3o2*omegaS/(one-omegaS) *((f_BW+f_W+f_TW-f_BS-f_S-f_TS-f_BN-f_N-f_TN+f_BE+f_E+f_TE ) / (one + drho_SET) - ((vx1_SET*vx1_SET-vx2_SET*vx2_SET)));
- // //kxxMzzFromfcNEQ_SET = -c3o2*omegaS/(one-omegaS) *((f_SW+f_W+f_NW-f_BS-f_TS-f_B-f_T-f_BN-f_TN+f_SE+f_E+f_NE ) / (one + drho_SET) - ((vx1_SET*vx1_SET-vx3_SET*vx3_SET)));
-
- //rho_tmp = f_E+f_W+f_N+f_S+f_T+f_B+f_NE+f_SW+f_SE+f_NW+f_TE+f_BW+f_BE+f_TW+f_TN+f_BS+f_BN+f_TS+f_ZERO+f_TNE+f_TSW+f_TSE+f_TNW+f_BNE+f_BSW+f_BSE+f_BNW;
- //
- //drho_SET += rho_tmp;
-
- //vx1_tmp = (((f_TNE-f_BSW)+(f_TSE-f_BNW)+(f_BNE-f_TSW)+(f_BSE-f_TNW)) + (((f_NE-f_SW)+(f_TE-f_BW))+((f_SE-f_NW)+(f_BE-f_TW))) + (f_E-f_W))/(one + rho_tmp);
- //vx2_tmp = (((f_TNE-f_BSW)+(f_TNW-f_BSE)+(f_BNE-f_TSW)+(f_BNW-f_TSE)) + (((f_NE-f_SW)+(f_TN-f_BS))+((f_BN-f_TS)+(f_NW-f_SE))) + (f_N-f_S))/(one + rho_tmp);
- //vx3_tmp = (((f_TNE-f_BSW)+(f_TNW-f_BSE)+(f_TSE-f_BNW)+(f_TSW-f_BNE)) + (((f_TE-f_BW)+(f_TN-f_BS))+((f_TW-f_BE)+(f_TS-f_BN))) + (f_T-f_B))/(one + rho_tmp);
-
- // vx1_SET += vx1_tmp;
- //vx2_SET += vx2_tmp;
- //vx3_SET += vx3_tmp;
-
- // drho_SET *= c1o2;
- // vx1_SET *= c1o2;
- //vx2_SET *= c1o2;
- //vx3_SET *= c1o2;
-
- // kxyFromfcNEQ_SET += -three*omegaS/(one-omegaS)*((f_SW+f_BSW+f_TSW-f_NW-f_BNW-f_TNW-f_SE-f_BSE-f_TSE+f_NE+f_BNE+f_TNE ) / (one + rho_tmp) - ((vx1_tmp*vx2_tmp)));
- // kyzFromfcNEQ_SET += -three*omegaS/(one-omegaS)*((f_BS+f_BSE+f_BSW-f_TS-f_TSE-f_TSW-f_BN-f_BNE-f_BNW+f_TN+f_TNE+f_TNW ) / (one + rho_tmp) - ((vx2_tmp*vx3_tmp)));
- // kxzFromfcNEQ_SET += -three*omegaS/(one-omegaS)*((f_BW+f_BSW+f_BNW-f_TW-f_TSW-f_TNW-f_BE-f_BSE-f_BNE+f_TE+f_TSE+f_TNE ) / (one + rho_tmp) - ((vx1_tmp*vx3_tmp)));
- // kxxMyyFromfcNEQ_SET += -c3o2*omegaS/(one-omegaS) *((f_BW+f_W+f_TW-f_BS-f_S-f_TS-f_BN-f_N-f_TN+f_BE+f_E+f_TE ) / (one + rho_tmp) - ((vx1_tmp*vx1_tmp-vx2_tmp*vx2_tmp)));
- // kxxMzzFromfcNEQ_SET += -c3o2*omegaS/(one-omegaS) *((f_SW+f_W+f_NW-f_BS-f_TS-f_B-f_T-f_BN-f_TN+f_SE+f_E+f_NE ) / (one + rho_tmp) - ((vx1_tmp*vx1_tmp-vx3_tmp*vx3_tmp)));
-
- //kxyFromfcNEQ_SET *= c1o2;
- //kyzFromfcNEQ_SET *= c1o2;
- //kxzFromfcNEQ_SET *= c1o2;
- //kxxMyyFromfcNEQ_SET *= c1o2;
- //kxxMzzFromfcNEQ_SET *= c1o2;
-
- // //////////////////////////////////////////////////////////////////////////
- // //SEB//
- // //////////////////////////////////////////////////////////////////////////
- // //index
- // kb = kzero;
- // kbw = kw;
- // kbs = ks;
- // kbsw = ksw;
- // kzero= k0w;
- // kw = neighborFX[k0w];
- // ks = k0sw;
- // ksw = neighborFX[k0sw];
- // ////////////////////////////////////////////////////////////////////////////////
- // f_E = fwF[kw];
- // f_W = feF[kzero];
- // f_N = fsF[ks];
- // f_S = fnF[kzero];
- // f_T = fbF[kb];
- // f_B = ftF[kzero];
- // f_NE = fswF[ksw];
- // f_SW = fneF[kzero];
- // f_SE = fnwF[kw];
- // f_NW = fseF[ks];
- // f_TE = fbwF[kbw];
- // f_BW = fteF[kzero];
- // f_BE = ftwF[kw];
- // f_TW = fbeF[kb];
- // f_TN = fbsF[kbs];
- // f_BS = ftnF[kzero];
- // f_BN = ftsF[ks];
- // f_TS = fbnF[kb];
- // f_ZERO = fzeroF[kzero];
- // f_TNE = fbswF[kbsw];
- // f_TSW = fbneF[kb];
- // f_TSE = fbnwF[kbw];
- // f_TNW = fbseF[kbs];
- // f_BNE = ftswF[ksw];
- // f_BSW = ftneF[kzero];
- // f_BSE = ftnwF[kw];
- // f_BNW = ftseF[ks];
-
- // //drho_SEB = f_E+f_W+f_N+f_S+f_T+f_B+f_NE+f_SW+f_SE+f_NW+f_TE+f_BW+f_BE+f_TW+f_TN+f_BS+f_BN+f_TS+f_ZERO+f_TNE+f_TSW+f_TSE+f_TNW+f_BNE+f_BSW+f_BSE+f_BNW;
- // //vx1_SEB = (((f_TNE-f_BSW)+(f_TSE-f_BNW)+(f_BNE-f_TSW)+(f_BSE-f_TNW)) + (((f_NE-f_SW)+(f_TE-f_BW))+((f_SE-f_NW)+(f_BE-f_TW))) + (f_E-f_W))/(one + drho_SEB);
- ////vx2_SEB = (((f_TNE-f_BSW)+(f_TNW-f_BSE)+(f_BNE-f_TSW)+(f_BNW-f_TSE)) + (((f_NE-f_SW)+(f_TN-f_BS))+((f_BN-f_TS)+(f_NW-f_SE))) + (f_N-f_S))/(one + drho_SEB);
- ////vx3_SEB = (((f_TNE-f_BSW)+(f_TNW-f_BSE)+(f_TSE-f_BNW)+(f_TSW-f_BNE)) + (((f_TE-f_BW)+(f_TN-f_BS))+((f_TW-f_BE)+(f_TS-f_BN))) + (f_T-f_B))/(one + drho_SEB);
-
- // //kxyFromfcNEQ_SEB = -three*omegaS/(one-omegaS)*((f_SW+f_BSW+f_TSW-f_NW-f_BNW-f_TNW-f_SE-f_BSE-f_TSE+f_NE+f_BNE+f_TNE ) / (one + drho_SEB) - ((vx1_SEB*vx2_SEB)));
- // //kyzFromfcNEQ_SEB = -three*omegaS/(one-omegaS)*((f_BS+f_BSE+f_BSW-f_TS-f_TSE-f_TSW-f_BN-f_BNE-f_BNW+f_TN+f_TNE+f_TNW ) / (one + drho_SEB) - ((vx2_SEB*vx3_SEB)));
- // //kxzFromfcNEQ_SEB = -three*omegaS/(one-omegaS)*((f_BW+f_BSW+f_BNW-f_TW-f_TSW-f_TNW-f_BE-f_BSE-f_BNE+f_TE+f_TSE+f_TNE ) / (one + drho_SEB) - ((vx1_SEB*vx3_SEB)));
- // //kxxMyyFromfcNEQ_SEB = -c3o2*omegaS/(one-omegaS) *((f_BW+f_W+f_TW-f_BS-f_S-f_TS-f_BN-f_N-f_TN+f_BE+f_E+f_TE ) / (one + drho_SEB) - ((vx1_SEB*vx1_SEB-vx2_SEB*vx2_SEB)));
- // //kxxMzzFromfcNEQ_SEB = -c3o2*omegaS/(one-omegaS) *((f_SW+f_W+f_NW-f_BS-f_TS-f_B-f_T-f_BN-f_TN+f_SE+f_E+f_NE ) / (one + drho_SEB) - ((vx1_SEB*vx1_SEB-vx3_SEB*vx3_SEB)));
-
- //rho_tmp = f_E+f_W+f_N+f_S+f_T+f_B+f_NE+f_SW+f_SE+f_NW+f_TE+f_BW+f_BE+f_TW+f_TN+f_BS+f_BN+f_TS+f_ZERO+f_TNE+f_TSW+f_TSE+f_TNW+f_BNE+f_BSW+f_BSE+f_BNW;
- //
- //drho_SEB += rho_tmp;
-
- //vx1_tmp = (((f_TNE-f_BSW)+(f_TSE-f_BNW)+(f_BNE-f_TSW)+(f_BSE-f_TNW)) + (((f_NE-f_SW)+(f_TE-f_BW))+((f_SE-f_NW)+(f_BE-f_TW))) + (f_E-f_W))/(one + rho_tmp);
- //vx2_tmp = (((f_TNE-f_BSW)+(f_TNW-f_BSE)+(f_BNE-f_TSW)+(f_BNW-f_TSE)) + (((f_NE-f_SW)+(f_TN-f_BS))+((f_BN-f_TS)+(f_NW-f_SE))) + (f_N-f_S))/(one + rho_tmp);
- //vx3_tmp = (((f_TNE-f_BSW)+(f_TNW-f_BSE)+(f_TSE-f_BNW)+(f_TSW-f_BNE)) + (((f_TE-f_BW)+(f_TN-f_BS))+((f_TW-f_BE)+(f_TS-f_BN))) + (f_T-f_B))/(one + rho_tmp);
-
- // vx1_SEB += vx1_tmp;
- //vx2_SEB += vx2_tmp;
- //vx3_SEB += vx3_tmp;
-
- // drho_SEB *= c1o2;
- // vx1_SEB *= c1o2;
- //vx2_SEB *= c1o2;
- //vx3_SEB *= c1o2;
-
- // kxyFromfcNEQ_SEB += -three*omegaS/(one-omegaS)*((f_SW+f_BSW+f_TSW-f_NW-f_BNW-f_TNW-f_SE-f_BSE-f_TSE+f_NE+f_BNE+f_TNE ) / (one + rho_tmp) - ((vx1_tmp*vx2_tmp)));
- // kyzFromfcNEQ_SEB += -three*omegaS/(one-omegaS)*((f_BS+f_BSE+f_BSW-f_TS-f_TSE-f_TSW-f_BN-f_BNE-f_BNW+f_TN+f_TNE+f_TNW ) / (one + rho_tmp) - ((vx2_tmp*vx3_tmp)));
- // kxzFromfcNEQ_SEB += -three*omegaS/(one-omegaS)*((f_BW+f_BSW+f_BNW-f_TW-f_TSW-f_TNW-f_BE-f_BSE-f_BNE+f_TE+f_TSE+f_TNE ) / (one + rho_tmp) - ((vx1_tmp*vx3_tmp)));
- // kxxMyyFromfcNEQ_SEB += -c3o2*omegaS/(one-omegaS) *((f_BW+f_W+f_TW-f_BS-f_S-f_TS-f_BN-f_N-f_TN+f_BE+f_E+f_TE ) / (one + rho_tmp) - ((vx1_tmp*vx1_tmp-vx2_tmp*vx2_tmp)));
- // kxxMzzFromfcNEQ_SEB += -c3o2*omegaS/(one-omegaS) *((f_SW+f_W+f_NW-f_BS-f_TS-f_B-f_T-f_BN-f_TN+f_SE+f_E+f_NE ) / (one + rho_tmp) - ((vx1_tmp*vx1_tmp-vx3_tmp*vx3_tmp)));
-
- //kxyFromfcNEQ_SEB *= c1o2;
- //kyzFromfcNEQ_SEB *= c1o2;
- //kxzFromfcNEQ_SEB *= c1o2;
- //kxxMyyFromfcNEQ_SEB *= c1o2;
- //kxxMzzFromfcNEQ_SEB *= c1o2;
-
- // //////////////////////////////////////////////////////////////////////////
- // //NWB//
- // //////////////////////////////////////////////////////////////////////////
- // //index 0
- // k0zero= k0s;
- // k0w = k0sw;
- // k0s = neighborFY[k0s];
- // k0b = k0bs;
- // k0sw = neighborFY[k0sw];
- // k0bw = k0bsw;
- // k0bs = neighborFY[k0bs];
- // k0bsw = neighborFY[k0bsw];
- // //////////////////////////////////////////////////////////////////////////
- // //index
- // kzero= k0zero;
- // kw = k0w;
- // ks = k0s;
- // kb = k0b;
- // ksw = k0sw;
- // kbw = k0bw;
- // kbs = k0bs;
- // kbsw = k0bsw;
- // ////////////////////////////////////////////////////////////////////////////////
- // f_E = fwF[kw];
- // f_W = feF[kzero];
- // f_N = fsF[ks];
- // f_S = fnF[kzero];
- // f_T = fbF[kb];
- // f_B = ftF[kzero];
- // f_NE = fswF[ksw];
- // f_SW = fneF[kzero];
- // f_SE = fnwF[kw];
- // f_NW = fseF[ks];
- // f_TE = fbwF[kbw];
- // f_BW = fteF[kzero];
- // f_BE = ftwF[kw];
- // f_TW = fbeF[kb];
- // f_TN = fbsF[kbs];
- // f_BS = ftnF[kzero];
- // f_BN = ftsF[ks];
- // f_TS = fbnF[kb];
- // f_ZERO = fzeroF[kzero];
- // f_TNE = fbswF[kbsw];
- // f_TSW = fbneF[kb];
- // f_TSE = fbnwF[kbw];
- // f_TNW = fbseF[kbs];
- // f_BNE = ftswF[ksw];
- // f_BSW = ftneF[kzero];
- // f_BSE = ftnwF[kw];
- // f_BNW = ftseF[ks];
-
- // //drho_NWB = f_E+f_W+f_N+f_S+f_T+f_B+f_NE+f_SW+f_SE+f_NW+f_TE+f_BW+f_BE+f_TW+f_TN+f_BS+f_BN+f_TS+f_ZERO+f_TNE+f_TSW+f_TSE+f_TNW+f_BNE+f_BSW+f_BSE+f_BNW;
- // //vx1_NWB = (((f_TNE-f_BSW)+(f_TSE-f_BNW)+(f_BNE-f_TSW)+(f_BSE-f_TNW)) + (((f_NE-f_SW)+(f_TE-f_BW))+((f_SE-f_NW)+(f_BE-f_TW))) + (f_E-f_W))/(one + drho_NWB);
- ////vx2_NWB = (((f_TNE-f_BSW)+(f_TNW-f_BSE)+(f_BNE-f_TSW)+(f_BNW-f_TSE)) + (((f_NE-f_SW)+(f_TN-f_BS))+((f_BN-f_TS)+(f_NW-f_SE))) + (f_N-f_S))/(one + drho_NWB);
- ////vx3_NWB = (((f_TNE-f_BSW)+(f_TNW-f_BSE)+(f_TSE-f_BNW)+(f_TSW-f_BNE)) + (((f_TE-f_BW)+(f_TN-f_BS))+((f_TW-f_BE)+(f_TS-f_BN))) + (f_T-f_B))/(one + drho_NWB);
-
- // //kxyFromfcNEQ_NWB = -three*omegaS/(one-omegaS)*((f_SW+f_BSW+f_TSW-f_NW-f_BNW-f_TNW-f_SE-f_BSE-f_TSE+f_NE+f_BNE+f_TNE ) / (one + drho_NWB) - ((vx1_NWB*vx2_NWB)));
- // //kyzFromfcNEQ_NWB = -three*omegaS/(one-omegaS)*((f_BS+f_BSE+f_BSW-f_TS-f_TSE-f_TSW-f_BN-f_BNE-f_BNW+f_TN+f_TNE+f_TNW ) / (one + drho_NWB) - ((vx2_NWB*vx3_NWB)));
- // //kxzFromfcNEQ_NWB = -three*omegaS/(one-omegaS)*((f_BW+f_BSW+f_BNW-f_TW-f_TSW-f_TNW-f_BE-f_BSE-f_BNE+f_TE+f_TSE+f_TNE ) / (one + drho_NWB) - ((vx1_NWB*vx3_NWB)));
- // //kxxMyyFromfcNEQ_NWB = -c3o2*omegaS/(one-omegaS) *((f_BW+f_W+f_TW-f_BS-f_S-f_TS-f_BN-f_N-f_TN+f_BE+f_E+f_TE ) / (one + drho_NWB) - ((vx1_NWB*vx1_NWB-vx2_NWB*vx2_NWB)));
- // //kxxMzzFromfcNEQ_NWB = -c3o2*omegaS/(one-omegaS) *((f_SW+f_W+f_NW-f_BS-f_TS-f_B-f_T-f_BN-f_TN+f_SE+f_E+f_NE ) / (one + drho_NWB) - ((vx1_NWB*vx1_NWB-vx3_NWB*vx3_NWB)));
-
- //rho_tmp = f_E+f_W+f_N+f_S+f_T+f_B+f_NE+f_SW+f_SE+f_NW+f_TE+f_BW+f_BE+f_TW+f_TN+f_BS+f_BN+f_TS+f_ZERO+f_TNE+f_TSW+f_TSE+f_TNW+f_BNE+f_BSW+f_BSE+f_BNW;
- //
- //drho_NWB += rho_tmp;
-
- //vx1_tmp = (((f_TNE-f_BSW)+(f_TSE-f_BNW)+(f_BNE-f_TSW)+(f_BSE-f_TNW)) + (((f_NE-f_SW)+(f_TE-f_BW))+((f_SE-f_NW)+(f_BE-f_TW))) + (f_E-f_W))/(one + rho_tmp);
- //vx2_tmp = (((f_TNE-f_BSW)+(f_TNW-f_BSE)+(f_BNE-f_TSW)+(f_BNW-f_TSE)) + (((f_NE-f_SW)+(f_TN-f_BS))+((f_BN-f_TS)+(f_NW-f_SE))) + (f_N-f_S))/(one + rho_tmp);
- //vx3_tmp = (((f_TNE-f_BSW)+(f_TNW-f_BSE)+(f_TSE-f_BNW)+(f_TSW-f_BNE)) + (((f_TE-f_BW)+(f_TN-f_BS))+((f_TW-f_BE)+(f_TS-f_BN))) + (f_T-f_B))/(one + rho_tmp);
-
- // vx1_NWB += vx1_tmp;
- //vx2_NWB += vx2_tmp;
- //vx3_NWB += vx3_tmp;
-
- // drho_NWB *= c1o2;
- // vx1_NWB *= c1o2;
- //vx2_NWB *= c1o2;
- //vx3_NWB *= c1o2;
-
- // kxyFromfcNEQ_NWB += -three*omegaS/(one-omegaS)*((f_SW+f_BSW+f_TSW-f_NW-f_BNW-f_TNW-f_SE-f_BSE-f_TSE+f_NE+f_BNE+f_TNE ) / (one + rho_tmp) - ((vx1_tmp*vx2_tmp)));
- // kyzFromfcNEQ_NWB += -three*omegaS/(one-omegaS)*((f_BS+f_BSE+f_BSW-f_TS-f_TSE-f_TSW-f_BN-f_BNE-f_BNW+f_TN+f_TNE+f_TNW ) / (one + rho_tmp) - ((vx2_tmp*vx3_tmp)));
- // kxzFromfcNEQ_NWB += -three*omegaS/(one-omegaS)*((f_BW+f_BSW+f_BNW-f_TW-f_TSW-f_TNW-f_BE-f_BSE-f_BNE+f_TE+f_TSE+f_TNE ) / (one + rho_tmp) - ((vx1_tmp*vx3_tmp)));
- // kxxMyyFromfcNEQ_NWB += -c3o2*omegaS/(one-omegaS) *((f_BW+f_W+f_TW-f_BS-f_S-f_TS-f_BN-f_N-f_TN+f_BE+f_E+f_TE ) / (one + rho_tmp) - ((vx1_tmp*vx1_tmp-vx2_tmp*vx2_tmp)));
- // kxxMzzFromfcNEQ_NWB += -c3o2*omegaS/(one-omegaS) *((f_SW+f_W+f_NW-f_BS-f_TS-f_B-f_T-f_BN-f_TN+f_SE+f_E+f_NE ) / (one + rho_tmp) - ((vx1_tmp*vx1_tmp-vx3_tmp*vx3_tmp)));
-
- //kxyFromfcNEQ_NWB *= c1o2;
- //kyzFromfcNEQ_NWB *= c1o2;
- //kxzFromfcNEQ_NWB *= c1o2;
- //kxxMyyFromfcNEQ_NWB *= c1o2;
- //kxxMzzFromfcNEQ_NWB *= c1o2;
-
- // //////////////////////////////////////////////////////////////////////////
- // //NWT//
- // //////////////////////////////////////////////////////////////////////////
- // //index
- // kzero= kb;
- // kw = kbw;
- // ks = kbs;
- // kb = neighborFZ[kb];
- // ksw = kbsw;
- // kbw = neighborFZ[kbw];
- // kbs = neighborFZ[kbs];
- // kbsw = neighborFZ[kbsw];
- // ////////////////////////////////////////////////////////////////////////////////
- // f_E = fwF[kw];
- // f_W = feF[kzero];
- // f_N = fsF[ks];
- // f_S = fnF[kzero];
- // f_T = fbF[kb];
- // f_B = ftF[kzero];
- // f_NE = fswF[ksw];
- // f_SW = fneF[kzero];
- // f_SE = fnwF[kw];
- // f_NW = fseF[ks];
- // f_TE = fbwF[kbw];
- // f_BW = fteF[kzero];
- // f_BE = ftwF[kw];
- // f_TW = fbeF[kb];
- // f_TN = fbsF[kbs];
- // f_BS = ftnF[kzero];
- // f_BN = ftsF[ks];
- // f_TS = fbnF[kb];
- // f_ZERO = fzeroF[kzero];
- // f_TNE = fbswF[kbsw];
- // f_TSW = fbneF[kb];
- // f_TSE = fbnwF[kbw];
- // f_TNW = fbseF[kbs];
- // f_BNE = ftswF[ksw];
- // f_BSW = ftneF[kzero];
- // f_BSE = ftnwF[kw];
- // f_BNW = ftseF[ks];
-
- // //drho_NWT = f_E+f_W+f_N+f_S+f_T+f_B+f_NE+f_SW+f_SE+f_NW+f_TE+f_BW+f_BE+f_TW+f_TN+f_BS+f_BN+f_TS+f_ZERO+f_TNE+f_TSW+f_TSE+f_TNW+f_BNE+f_BSW+f_BSE+f_BNW;
- // //vx1_NWT = (((f_TNE-f_BSW)+(f_TSE-f_BNW)+(f_BNE-f_TSW)+(f_BSE-f_TNW)) + (((f_NE-f_SW)+(f_TE-f_BW))+((f_SE-f_NW)+(f_BE-f_TW))) + (f_E-f_W))/(one + drho_NWT);
- ////vx2_NWT = (((f_TNE-f_BSW)+(f_TNW-f_BSE)+(f_BNE-f_TSW)+(f_BNW-f_TSE)) + (((f_NE-f_SW)+(f_TN-f_BS))+((f_BN-f_TS)+(f_NW-f_SE))) + (f_N-f_S))/(one + drho_NWT);
- ////vx3_NWT = (((f_TNE-f_BSW)+(f_TNW-f_BSE)+(f_TSE-f_BNW)+(f_TSW-f_BNE)) + (((f_TE-f_BW)+(f_TN-f_BS))+((f_TW-f_BE)+(f_TS-f_BN))) + (f_T-f_B))/(one + drho_NWT);
-
- // //kxyFromfcNEQ_NWT = -three*omegaS/(one-omegaS)*((f_SW+f_BSW+f_TSW-f_NW-f_BNW-f_TNW-f_SE-f_BSE-f_TSE+f_NE+f_BNE+f_TNE ) / (one + drho_NWT) - ((vx1_NWT*vx2_NWT)));
- // //kyzFromfcNEQ_NWT = -three*omegaS/(one-omegaS)*((f_BS+f_BSE+f_BSW-f_TS-f_TSE-f_TSW-f_BN-f_BNE-f_BNW+f_TN+f_TNE+f_TNW ) / (one + drho_NWT) - ((vx2_NWT*vx3_NWT)));
- // //kxzFromfcNEQ_NWT = -three*omegaS/(one-omegaS)*((f_BW+f_BSW+f_BNW-f_TW-f_TSW-f_TNW-f_BE-f_BSE-f_BNE+f_TE+f_TSE+f_TNE ) / (one + drho_NWT) - ((vx1_NWT*vx3_NWT)));
- // //kxxMyyFromfcNEQ_NWT = -c3o2*omegaS/(one-omegaS) *((f_BW+f_W+f_TW-f_BS-f_S-f_TS-f_BN-f_N-f_TN+f_BE+f_E+f_TE ) / (one + drho_NWT) - ((vx1_NWT*vx1_NWT-vx2_NWT*vx2_NWT)));
- // //kxxMzzFromfcNEQ_NWT = -c3o2*omegaS/(one-omegaS) *((f_SW+f_W+f_NW-f_BS-f_TS-f_B-f_T-f_BN-f_TN+f_SE+f_E+f_NE ) / (one + drho_NWT) - ((vx1_NWT*vx1_NWT-vx3_NWT*vx3_NWT)));
-
- //rho_tmp = f_E+f_W+f_N+f_S+f_T+f_B+f_NE+f_SW+f_SE+f_NW+f_TE+f_BW+f_BE+f_TW+f_TN+f_BS+f_BN+f_TS+f_ZERO+f_TNE+f_TSW+f_TSE+f_TNW+f_BNE+f_BSW+f_BSE+f_BNW;
- //
- //drho_NWT += rho_tmp;
-
- //vx1_tmp = (((f_TNE-f_BSW)+(f_TSE-f_BNW)+(f_BNE-f_TSW)+(f_BSE-f_TNW)) + (((f_NE-f_SW)+(f_TE-f_BW))+((f_SE-f_NW)+(f_BE-f_TW))) + (f_E-f_W))/(one + rho_tmp);
- //vx2_tmp = (((f_TNE-f_BSW)+(f_TNW-f_BSE)+(f_BNE-f_TSW)+(f_BNW-f_TSE)) + (((f_NE-f_SW)+(f_TN-f_BS))+((f_BN-f_TS)+(f_NW-f_SE))) + (f_N-f_S))/(one + rho_tmp);
- //vx3_tmp = (((f_TNE-f_BSW)+(f_TNW-f_BSE)+(f_TSE-f_BNW)+(f_TSW-f_BNE)) + (((f_TE-f_BW)+(f_TN-f_BS))+((f_TW-f_BE)+(f_TS-f_BN))) + (f_T-f_B))/(one + rho_tmp);
-
- // vx1_NWT += vx1_tmp;
- //vx2_NWT += vx2_tmp;
- //vx3_NWT += vx3_tmp;
-
- // drho_NWT *= c1o2;
- // vx1_NWT *= c1o2;
- //vx2_NWT *= c1o2;
- //vx3_NWT *= c1o2;
-
- // kxyFromfcNEQ_NWT += -three*omegaS/(one-omegaS)*((f_SW+f_BSW+f_TSW-f_NW-f_BNW-f_TNW-f_SE-f_BSE-f_TSE+f_NE+f_BNE+f_TNE ) / (one + rho_tmp) - ((vx1_tmp*vx2_tmp)));
- // kyzFromfcNEQ_NWT += -three*omegaS/(one-omegaS)*((f_BS+f_BSE+f_BSW-f_TS-f_TSE-f_TSW-f_BN-f_BNE-f_BNW+f_TN+f_TNE+f_TNW ) / (one + rho_tmp) - ((vx2_tmp*vx3_tmp)));
- // kxzFromfcNEQ_NWT += -three*omegaS/(one-omegaS)*((f_BW+f_BSW+f_BNW-f_TW-f_TSW-f_TNW-f_BE-f_BSE-f_BNE+f_TE+f_TSE+f_TNE ) / (one + rho_tmp) - ((vx1_tmp*vx3_tmp)));
- // kxxMyyFromfcNEQ_NWT += -c3o2*omegaS/(one-omegaS) *((f_BW+f_W+f_TW-f_BS-f_S-f_TS-f_BN-f_N-f_TN+f_BE+f_E+f_TE ) / (one + rho_tmp) - ((vx1_tmp*vx1_tmp-vx2_tmp*vx2_tmp)));
- // kxxMzzFromfcNEQ_NWT += -c3o2*omegaS/(one-omegaS) *((f_SW+f_W+f_NW-f_BS-f_TS-f_B-f_T-f_BN-f_TN+f_SE+f_E+f_NE ) / (one + rho_tmp) - ((vx1_tmp*vx1_tmp-vx3_tmp*vx3_tmp)));
-
- //kxyFromfcNEQ_NWT *= c1o2;
- //kyzFromfcNEQ_NWT *= c1o2;
- //kxzFromfcNEQ_NWT *= c1o2;
- //kxxMyyFromfcNEQ_NWT *= c1o2;
- //kxxMzzFromfcNEQ_NWT *= c1o2;
-
- // //////////////////////////////////////////////////////////////////////////
- // //NET//
- // //////////////////////////////////////////////////////////////////////////
- // //index
- // kzero= kw;
- // kw = neighborFX[kw];
- // ks = ksw;
- // kb = kbw;
- // ksw = neighborFX[ksw];
- // kbw = neighborFX[kbw];
- // kbs = kbsw;
- // kbsw = neighborFX[kbsw];
- // ////////////////////////////////////////////////////////////////////////////////
- // f_E = fwF[kw];
- // f_W = feF[kzero];
- // f_N = fsF[ks];
- // f_S = fnF[kzero];
- // f_T = fbF[kb];
- // f_B = ftF[kzero];
- // f_NE = fswF[ksw];
- // f_SW = fneF[kzero];
- // f_SE = fnwF[kw];
- // f_NW = fseF[ks];
- // f_TE = fbwF[kbw];
- // f_BW = fteF[kzero];
- // f_BE = ftwF[kw];
- // f_TW = fbeF[kb];
- // f_TN = fbsF[kbs];
- // f_BS = ftnF[kzero];
- // f_BN = ftsF[ks];
- // f_TS = fbnF[kb];
- // f_ZERO = fzeroF[kzero];
- // f_TNE = fbswF[kbsw];
- // f_TSW = fbneF[kb];
- // f_TSE = fbnwF[kbw];
- // f_TNW = fbseF[kbs];
- // f_BNE = ftswF[ksw];
- // f_BSW = ftneF[kzero];
- // f_BSE = ftnwF[kw];
- // f_BNW = ftseF[ks];
-
- // //drho_NET = f_E+f_W+f_N+f_S+f_T+f_B+f_NE+f_SW+f_SE+f_NW+f_TE+f_BW+f_BE+f_TW+f_TN+f_BS+f_BN+f_TS+f_ZERO+f_TNE+f_TSW+f_TSE+f_TNW+f_BNE+f_BSW+f_BSE+f_BNW;
- // //vx1_NET = (((f_TNE-f_BSW)+(f_TSE-f_BNW)+(f_BNE-f_TSW)+(f_BSE-f_TNW)) + (((f_NE-f_SW)+(f_TE-f_BW))+((f_SE-f_NW)+(f_BE-f_TW))) + (f_E-f_W))/(one + drho_NET);
- ////vx2_NET = (((f_TNE-f_BSW)+(f_TNW-f_BSE)+(f_BNE-f_TSW)+(f_BNW-f_TSE)) + (((f_NE-f_SW)+(f_TN-f_BS))+((f_BN-f_TS)+(f_NW-f_SE))) + (f_N-f_S))/(one + drho_NET);
- ////vx3_NET = (((f_TNE-f_BSW)+(f_TNW-f_BSE)+(f_TSE-f_BNW)+(f_TSW-f_BNE)) + (((f_TE-f_BW)+(f_TN-f_BS))+((f_TW-f_BE)+(f_TS-f_BN))) + (f_T-f_B))/(one + drho_NET);
-
- // //kxyFromfcNEQ_NET = -three*omegaS/(one-omegaS)*((f_SW+f_BSW+f_TSW-f_NW-f_BNW-f_TNW-f_SE-f_BSE-f_TSE+f_NE+f_BNE+f_TNE ) / (one + drho_NET) - ((vx1_NET*vx2_NET)));
- // //kyzFromfcNEQ_NET = -three*omegaS/(one-omegaS)*((f_BS+f_BSE+f_BSW-f_TS-f_TSE-f_TSW-f_BN-f_BNE-f_BNW+f_TN+f_TNE+f_TNW ) / (one + drho_NET) - ((vx2_NET*vx3_NET)));
- // //kxzFromfcNEQ_NET = -three*omegaS/(one-omegaS)*((f_BW+f_BSW+f_BNW-f_TW-f_TSW-f_TNW-f_BE-f_BSE-f_BNE+f_TE+f_TSE+f_TNE ) / (one + drho_NET) - ((vx1_NET*vx3_NET)));
- // //kxxMyyFromfcNEQ_NET = -c3o2*omegaS/(one-omegaS) *((f_BW+f_W+f_TW-f_BS-f_S-f_TS-f_BN-f_N-f_TN+f_BE+f_E+f_TE ) / (one + drho_NET) - ((vx1_NET*vx1_NET-vx2_NET*vx2_NET)));
- // //kxxMzzFromfcNEQ_NET = -c3o2*omegaS/(one-omegaS) *((f_SW+f_W+f_NW-f_BS-f_TS-f_B-f_T-f_BN-f_TN+f_SE+f_E+f_NE ) / (one + drho_NET) - ((vx1_NET*vx1_NET-vx3_NET*vx3_NET)));
-
- //rho_tmp = f_E+f_W+f_N+f_S+f_T+f_B+f_NE+f_SW+f_SE+f_NW+f_TE+f_BW+f_BE+f_TW+f_TN+f_BS+f_BN+f_TS+f_ZERO+f_TNE+f_TSW+f_TSE+f_TNW+f_BNE+f_BSW+f_BSE+f_BNW;
- //
- //drho_NET += rho_tmp;
-
- //vx1_tmp = (((f_TNE-f_BSW)+(f_TSE-f_BNW)+(f_BNE-f_TSW)+(f_BSE-f_TNW)) + (((f_NE-f_SW)+(f_TE-f_BW))+((f_SE-f_NW)+(f_BE-f_TW))) + (f_E-f_W))/(one + rho_tmp);
- //vx2_tmp = (((f_TNE-f_BSW)+(f_TNW-f_BSE)+(f_BNE-f_TSW)+(f_BNW-f_TSE)) + (((f_NE-f_SW)+(f_TN-f_BS))+((f_BN-f_TS)+(f_NW-f_SE))) + (f_N-f_S))/(one + rho_tmp);
- //vx3_tmp = (((f_TNE-f_BSW)+(f_TNW-f_BSE)+(f_TSE-f_BNW)+(f_TSW-f_BNE)) + (((f_TE-f_BW)+(f_TN-f_BS))+((f_TW-f_BE)+(f_TS-f_BN))) + (f_T-f_B))/(one + rho_tmp);
-
- // vx1_NET += vx1_tmp;
- //vx2_NET += vx2_tmp;
- //vx3_NET += vx3_tmp;
-
- // drho_NET *= c1o2;
- // vx1_NET *= c1o2;
- //vx2_NET *= c1o2;
- //vx3_NET *= c1o2;
-
- // kxyFromfcNEQ_NET += -three*omegaS/(one-omegaS)*((f_SW+f_BSW+f_TSW-f_NW-f_BNW-f_TNW-f_SE-f_BSE-f_TSE+f_NE+f_BNE+f_TNE ) / (one + rho_tmp) - ((vx1_tmp*vx2_tmp)));
- // kyzFromfcNEQ_NET += -three*omegaS/(one-omegaS)*((f_BS+f_BSE+f_BSW-f_TS-f_TSE-f_TSW-f_BN-f_BNE-f_BNW+f_TN+f_TNE+f_TNW ) / (one + rho_tmp) - ((vx2_tmp*vx3_tmp)));
- // kxzFromfcNEQ_NET += -three*omegaS/(one-omegaS)*((f_BW+f_BSW+f_BNW-f_TW-f_TSW-f_TNW-f_BE-f_BSE-f_BNE+f_TE+f_TSE+f_TNE ) / (one + rho_tmp) - ((vx1_tmp*vx3_tmp)));
- // kxxMyyFromfcNEQ_NET += -c3o2*omegaS/(one-omegaS) *((f_BW+f_W+f_TW-f_BS-f_S-f_TS-f_BN-f_N-f_TN+f_BE+f_E+f_TE ) / (one + rho_tmp) - ((vx1_tmp*vx1_tmp-vx2_tmp*vx2_tmp)));
- // kxxMzzFromfcNEQ_NET += -c3o2*omegaS/(one-omegaS) *((f_SW+f_W+f_NW-f_BS-f_TS-f_B-f_T-f_BN-f_TN+f_SE+f_E+f_NE ) / (one + rho_tmp) - ((vx1_tmp*vx1_tmp-vx3_tmp*vx3_tmp)));
-
- //kxyFromfcNEQ_NET *= c1o2;
- //kyzFromfcNEQ_NET *= c1o2;
- //kxzFromfcNEQ_NET *= c1o2;
- //kxxMyyFromfcNEQ_NET *= c1o2;
- //kxxMzzFromfcNEQ_NET *= c1o2;
-
- // //////////////////////////////////////////////////////////////////////////
- // //NEB//
- // //////////////////////////////////////////////////////////////////////////
- // //index
- // kb = kzero;
- // kbw = kw;
- // kbs = ks;
- // kbsw = ksw;
- // kzero= k0w;
- // kw = neighborFX[k0w];
- // ks = k0sw;
- // ksw = neighborFX[k0sw];
- // ////////////////////////////////////////////////////////////////////////////////
- // f_E = fwF[kw];
- // f_W = feF[kzero];
- // f_N = fsF[ks];
- // f_S = fnF[kzero];
- // f_T = fbF[kb];
- // f_B = ftF[kzero];
- // f_NE = fswF[ksw];
- // f_SW = fneF[kzero];
- // f_SE = fnwF[kw];
- // f_NW = fseF[ks];
- // f_TE = fbwF[kbw];
- // f_BW = fteF[kzero];
- // f_BE = ftwF[kw];
- // f_TW = fbeF[kb];
- // f_TN = fbsF[kbs];
- // f_BS = ftnF[kzero];
- // f_BN = ftsF[ks];
- // f_TS = fbnF[kb];
- // f_ZERO = fzeroF[kzero];
- // f_TNE = fbswF[kbsw];
- // f_TSW = fbneF[kb];
- // f_TSE = fbnwF[kbw];
- // f_TNW = fbseF[kbs];
- // f_BNE = ftswF[ksw];
- // f_BSW = ftneF[kzero];
- // f_BSE = ftnwF[kw];
- // f_BNW = ftseF[ks];
-
- // //drho_NEB = f_E+f_W+f_N+f_S+f_T+f_B+f_NE+f_SW+f_SE+f_NW+f_TE+f_BW+f_BE+f_TW+f_TN+f_BS+f_BN+f_TS+f_ZERO+f_TNE+f_TSW+f_TSE+f_TNW+f_BNE+f_BSW+f_BSE+f_BNW;
- // //vx1_NEB = (((f_TNE-f_BSW)+(f_TSE-f_BNW)+(f_BNE-f_TSW)+(f_BSE-f_TNW)) + (((f_NE-f_SW)+(f_TE-f_BW))+((f_SE-f_NW)+(f_BE-f_TW))) + (f_E-f_W))/(one + drho_NEB);
- ////vx2_NEB = (((f_TNE-f_BSW)+(f_TNW-f_BSE)+(f_BNE-f_TSW)+(f_BNW-f_TSE)) + (((f_NE-f_SW)+(f_TN-f_BS))+((f_BN-f_TS)+(f_NW-f_SE))) + (f_N-f_S))/(one + drho_NEB);
- ////vx3_NEB = (((f_TNE-f_BSW)+(f_TNW-f_BSE)+(f_TSE-f_BNW)+(f_TSW-f_BNE)) + (((f_TE-f_BW)+(f_TN-f_BS))+((f_TW-f_BE)+(f_TS-f_BN))) + (f_T-f_B))/(one + drho_NEB);
-
- // //kxyFromfcNEQ_NEB = -three*omegaS/(one-omegaS)*((f_SW+f_BSW+f_TSW-f_NW-f_BNW-f_TNW-f_SE-f_BSE-f_TSE+f_NE+f_BNE+f_TNE ) / (one + drho_NEB) - ((vx1_NEB*vx2_NEB)));
- // //kyzFromfcNEQ_NEB = -three*omegaS/(one-omegaS)*((f_BS+f_BSE+f_BSW-f_TS-f_TSE-f_TSW-f_BN-f_BNE-f_BNW+f_TN+f_TNE+f_TNW ) / (one + drho_NEB) - ((vx2_NEB*vx3_NEB)));
- // //kxzFromfcNEQ_NEB = -three*omegaS/(one-omegaS)*((f_BW+f_BSW+f_BNW-f_TW-f_TSW-f_TNW-f_BE-f_BSE-f_BNE+f_TE+f_TSE+f_TNE ) / (one + drho_NEB) - ((vx1_NEB*vx3_NEB)));
- // //kxxMyyFromfcNEQ_NEB = -c3o2*omegaS/(one-omegaS) *((f_BW+f_W+f_TW-f_BS-f_S-f_TS-f_BN-f_N-f_TN+f_BE+f_E+f_TE ) / (one + drho_NEB) - ((vx1_NEB*vx1_NEB-vx2_NEB*vx2_NEB)));
- // //kxxMzzFromfcNEQ_NEB = -c3o2*omegaS/(one-omegaS) *((f_SW+f_W+f_NW-f_BS-f_TS-f_B-f_T-f_BN-f_TN+f_SE+f_E+f_NE ) / (one + drho_NEB) - ((vx1_NEB*vx1_NEB-vx3_NEB*vx3_NEB)));
-
- //rho_tmp = f_E+f_W+f_N+f_S+f_T+f_B+f_NE+f_SW+f_SE+f_NW+f_TE+f_BW+f_BE+f_TW+f_TN+f_BS+f_BN+f_TS+f_ZERO+f_TNE+f_TSW+f_TSE+f_TNW+f_BNE+f_BSW+f_BSE+f_BNW;
- //
- //drho_NEB += rho_tmp;
-
- //vx1_tmp = (((f_TNE-f_BSW)+(f_TSE-f_BNW)+(f_BNE-f_TSW)+(f_BSE-f_TNW)) + (((f_NE-f_SW)+(f_TE-f_BW))+((f_SE-f_NW)+(f_BE-f_TW))) + (f_E-f_W))/(one + rho_tmp);
- //vx2_tmp = (((f_TNE-f_BSW)+(f_TNW-f_BSE)+(f_BNE-f_TSW)+(f_BNW-f_TSE)) + (((f_NE-f_SW)+(f_TN-f_BS))+((f_BN-f_TS)+(f_NW-f_SE))) + (f_N-f_S))/(one + rho_tmp);
- //vx3_tmp = (((f_TNE-f_BSW)+(f_TNW-f_BSE)+(f_TSE-f_BNW)+(f_TSW-f_BNE)) + (((f_TE-f_BW)+(f_TN-f_BS))+((f_TW-f_BE)+(f_TS-f_BN))) + (f_T-f_B))/(one + rho_tmp);
-
- // vx1_NEB += vx1_tmp;
- //vx2_NEB += vx2_tmp;
- //vx3_NEB += vx3_tmp;
-
- // drho_NEB *= c1o2;
- // vx1_NEB *= c1o2;
- //vx2_NEB *= c1o2;
- //vx3_NEB *= c1o2;
-
- // kxyFromfcNEQ_NEB += -three*omegaS/(one-omegaS)*((f_SW+f_BSW+f_TSW-f_NW-f_BNW-f_TNW-f_SE-f_BSE-f_TSE+f_NE+f_BNE+f_TNE ) / (one + rho_tmp) - ((vx1_tmp*vx2_tmp)));
- // kyzFromfcNEQ_NEB += -three*omegaS/(one-omegaS)*((f_BS+f_BSE+f_BSW-f_TS-f_TSE-f_TSW-f_BN-f_BNE-f_BNW+f_TN+f_TNE+f_TNW ) / (one + rho_tmp) - ((vx2_tmp*vx3_tmp)));
- // kxzFromfcNEQ_NEB += -three*omegaS/(one-omegaS)*((f_BW+f_BSW+f_BNW-f_TW-f_TSW-f_TNW-f_BE-f_BSE-f_BNE+f_TE+f_TSE+f_TNE ) / (one + rho_tmp) - ((vx1_tmp*vx3_tmp)));
- // kxxMyyFromfcNEQ_NEB += -c3o2*omegaS/(one-omegaS) *((f_BW+f_W+f_TW-f_BS-f_S-f_TS-f_BN-f_N-f_TN+f_BE+f_E+f_TE ) / (one + rho_tmp) - ((vx1_tmp*vx1_tmp-vx2_tmp*vx2_tmp)));
- // kxxMzzFromfcNEQ_NEB += -c3o2*omegaS/(one-omegaS) *((f_SW+f_W+f_NW-f_BS-f_TS-f_B-f_T-f_BN-f_TN+f_SE+f_E+f_NE ) / (one + rho_tmp) - ((vx1_tmp*vx1_tmp-vx3_tmp*vx3_tmp)));
-
- //kxyFromfcNEQ_NEB *= c1o2;
- //kyzFromfcNEQ_NEB *= c1o2;
- //kxzFromfcNEQ_NEB *= c1o2;
- //kxxMyyFromfcNEQ_NEB *= c1o2;
- //kxxMzzFromfcNEQ_NEB *= c1o2;
- //
- //
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
- //////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
- //kxyFromfcNEQ_SWB = zero;
- //kyzFromfcNEQ_SWB = zero;
- //kxzFromfcNEQ_SWB = zero;
- //kxxMyyFromfcNEQ_SWB = zero;
- //kxxMzzFromfcNEQ_SWB = zero;
- //kxyFromfcNEQ_SWT = zero;
- //kyzFromfcNEQ_SWT = zero;
- //kxzFromfcNEQ_SWT = zero;
- //kxxMyyFromfcNEQ_SWT = zero;
- //kxxMzzFromfcNEQ_SWT = zero;
- //kxyFromfcNEQ_SET = zero;
- //kyzFromfcNEQ_SET = zero;
- //kxzFromfcNEQ_SET = zero;
- //kxxMyyFromfcNEQ_SET = zero;
- //kxxMzzFromfcNEQ_SET = zero;
- //kxyFromfcNEQ_SEB = zero;
- //kyzFromfcNEQ_SEB = zero;
- //kxzFromfcNEQ_SEB = zero;
- //kxxMyyFromfcNEQ_SEB = zero;
- //kxxMzzFromfcNEQ_SEB = zero;
- //kxyFromfcNEQ_NWB = zero;
- //kyzFromfcNEQ_NWB = zero;
- //kxzFromfcNEQ_NWB = zero;
- //kxxMyyFromfcNEQ_NWB = zero;
- //kxxMzzFromfcNEQ_NWB = zero;
- //kxyFromfcNEQ_NWT = zero;
- //kyzFromfcNEQ_NWT = zero;
- //kxzFromfcNEQ_NWT = zero;
- //kxxMyyFromfcNEQ_NWT = zero;
- //kxxMzzFromfcNEQ_NWT = zero;
- //kxyFromfcNEQ_NET = zero;
- //kyzFromfcNEQ_NET = zero;
- //kxzFromfcNEQ_NET = zero;
- //kxxMyyFromfcNEQ_NET = zero;
- //kxxMzzFromfcNEQ_NET = zero;
- //kxyFromfcNEQ_NEB = zero;
- //kyzFromfcNEQ_NEB = zero;
- //kxzFromfcNEQ_NEB = zero;
- //kxxMyyFromfcNEQ_NEB = zero;
- //kxxMzzFromfcNEQ_NEB = zero;
//////////////////////////////////////////////////////////////////////////
//3
//////////////////////////////////////////////////////////////////////////
@@ -13278,8 +12474,8 @@ __global__ void scaleFC_Fix_comp_27( real* DC,
unsigned int* neighborFX,
unsigned int* neighborFY,
unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
+ unsigned long long numberOfLBnodesCoarse,
+ unsigned long long numberOfLBnodesFine,
bool isEvenTimestep,
unsigned int* posC,
unsigned int* posFSWB,
@@ -13296,96 +12492,96 @@ __global__ void scaleFC_Fix_comp_27( real* DC,
real *feF, *fwF, *fnF, *fsF, *ftF, *fbF, *fneF, *fswF, *fseF, *fnwF, *fteF, *fbwF, *fbeF, *ftwF, *ftnF, *fbsF, *fbnF, *ftsF, *fzeroF,
*ftneF, *ftswF, *ftseF, *ftnwF, *fbneF, *fbswF, *fbseF, *fbnwF;
- feF = &DF[DIR_P00 *size_MatF];
- fwF = &DF[DIR_M00 *size_MatF];
- fnF = &DF[DIR_0P0 *size_MatF];
- fsF = &DF[DIR_0M0 *size_MatF];
- ftF = &DF[DIR_00P *size_MatF];
- fbF = &DF[DIR_00M *size_MatF];
- fneF = &DF[DIR_PP0 *size_MatF];
- fswF = &DF[DIR_MM0 *size_MatF];
- fseF = &DF[DIR_PM0 *size_MatF];
- fnwF = &DF[DIR_MP0 *size_MatF];
- fteF = &DF[DIR_P0P *size_MatF];
- fbwF = &DF[DIR_M0M *size_MatF];
- fbeF = &DF[DIR_P0M *size_MatF];
- ftwF = &DF[DIR_M0P *size_MatF];
- ftnF = &DF[DIR_0PP *size_MatF];
- fbsF = &DF[DIR_0MM *size_MatF];
- fbnF = &DF[DIR_0PM *size_MatF];
- ftsF = &DF[DIR_0MP *size_MatF];
- fzeroF = &DF[DIR_000*size_MatF];
- ftneF = &DF[DIR_PPP *size_MatF];
- ftswF = &DF[DIR_MMP *size_MatF];
- ftseF = &DF[DIR_PMP *size_MatF];
- ftnwF = &DF[DIR_MPP *size_MatF];
- fbneF = &DF[DIR_PPM *size_MatF];
- fbswF = &DF[DIR_MMM *size_MatF];
- fbseF = &DF[DIR_PMM *size_MatF];
- fbnwF = &DF[DIR_MPM *size_MatF];
+ feF = &DF[DIR_P00 * numberOfLBnodesFine];
+ fwF = &DF[DIR_M00 * numberOfLBnodesFine];
+ fnF = &DF[DIR_0P0 * numberOfLBnodesFine];
+ fsF = &DF[DIR_0M0 * numberOfLBnodesFine];
+ ftF = &DF[DIR_00P * numberOfLBnodesFine];
+ fbF = &DF[DIR_00M * numberOfLBnodesFine];
+ fneF = &DF[DIR_PP0 * numberOfLBnodesFine];
+ fswF = &DF[DIR_MM0 * numberOfLBnodesFine];
+ fseF = &DF[DIR_PM0 * numberOfLBnodesFine];
+ fnwF = &DF[DIR_MP0 * numberOfLBnodesFine];
+ fteF = &DF[DIR_P0P * numberOfLBnodesFine];
+ fbwF = &DF[DIR_M0M * numberOfLBnodesFine];
+ fbeF = &DF[DIR_P0M * numberOfLBnodesFine];
+ ftwF = &DF[DIR_M0P * numberOfLBnodesFine];
+ ftnF = &DF[DIR_0PP * numberOfLBnodesFine];
+ fbsF = &DF[DIR_0MM * numberOfLBnodesFine];
+ fbnF = &DF[DIR_0PM * numberOfLBnodesFine];
+ ftsF = &DF[DIR_0MP * numberOfLBnodesFine];
+ fzeroF = &DF[DIR_000 * numberOfLBnodesFine];
+ ftneF = &DF[DIR_PPP * numberOfLBnodesFine];
+ ftswF = &DF[DIR_MMP * numberOfLBnodesFine];
+ ftseF = &DF[DIR_PMP * numberOfLBnodesFine];
+ ftnwF = &DF[DIR_MPP * numberOfLBnodesFine];
+ fbneF = &DF[DIR_PPM * numberOfLBnodesFine];
+ fbswF = &DF[DIR_MMM * numberOfLBnodesFine];
+ fbseF = &DF[DIR_PMM * numberOfLBnodesFine];
+ fbnwF = &DF[DIR_MPM * numberOfLBnodesFine];
real *feC, *fwC, *fnC, *fsC, *ftC, *fbC, *fneC, *fswC, *fseC, *fnwC, *fteC, *fbwC, *fbeC, *ftwC, *ftnC, *fbsC, *fbnC, *ftsC, *fzeroC,
*ftneC, *ftswC, *ftseC, *ftnwC, *fbneC, *fbswC, *fbseC, *fbnwC;
if (isEvenTimestep==true)
{
- feC = &DC[DIR_P00 *size_MatC];
- fwC = &DC[DIR_M00 *size_MatC];
- fnC = &DC[DIR_0P0 *size_MatC];
- fsC = &DC[DIR_0M0 *size_MatC];
- ftC = &DC[DIR_00P *size_MatC];
- fbC = &DC[DIR_00M *size_MatC];
- fneC = &DC[DIR_PP0 *size_MatC];
- fswC = &DC[DIR_MM0 *size_MatC];
- fseC = &DC[DIR_PM0 *size_MatC];
- fnwC = &DC[DIR_MP0 *size_MatC];
- fteC = &DC[DIR_P0P *size_MatC];
- fbwC = &DC[DIR_M0M *size_MatC];
- fbeC = &DC[DIR_P0M *size_MatC];
- ftwC = &DC[DIR_M0P *size_MatC];
- ftnC = &DC[DIR_0PP *size_MatC];
- fbsC = &DC[DIR_0MM *size_MatC];
- fbnC = &DC[DIR_0PM *size_MatC];
- ftsC = &DC[DIR_0MP *size_MatC];
- fzeroC = &DC[DIR_000*size_MatC];
- ftneC = &DC[DIR_PPP *size_MatC];
- ftswC = &DC[DIR_MMP *size_MatC];
- ftseC = &DC[DIR_PMP *size_MatC];
- ftnwC = &DC[DIR_MPP *size_MatC];
- fbneC = &DC[DIR_PPM *size_MatC];
- fbswC = &DC[DIR_MMM *size_MatC];
- fbseC = &DC[DIR_PMM *size_MatC];
- fbnwC = &DC[DIR_MPM *size_MatC];
+ feC = &DC[DIR_P00 * numberOfLBnodesCoarse];
+ fwC = &DC[DIR_M00 * numberOfLBnodesCoarse];
+ fnC = &DC[DIR_0P0 * numberOfLBnodesCoarse];
+ fsC = &DC[DIR_0M0 * numberOfLBnodesCoarse];
+ ftC = &DC[DIR_00P * numberOfLBnodesCoarse];
+ fbC = &DC[DIR_00M * numberOfLBnodesCoarse];
+ fneC = &DC[DIR_PP0 * numberOfLBnodesCoarse];
+ fswC = &DC[DIR_MM0 * numberOfLBnodesCoarse];
+ fseC = &DC[DIR_PM0 * numberOfLBnodesCoarse];
+ fnwC = &DC[DIR_MP0 * numberOfLBnodesCoarse];
+ fteC = &DC[DIR_P0P * numberOfLBnodesCoarse];
+ fbwC = &DC[DIR_M0M * numberOfLBnodesCoarse];
+ fbeC = &DC[DIR_P0M * numberOfLBnodesCoarse];
+ ftwC = &DC[DIR_M0P * numberOfLBnodesCoarse];
+ ftnC = &DC[DIR_0PP * numberOfLBnodesCoarse];
+ fbsC = &DC[DIR_0MM * numberOfLBnodesCoarse];
+ fbnC = &DC[DIR_0PM * numberOfLBnodesCoarse];
+ ftsC = &DC[DIR_0MP * numberOfLBnodesCoarse];
+ fzeroC = &DC[DIR_000 * numberOfLBnodesCoarse];
+ ftneC = &DC[DIR_PPP * numberOfLBnodesCoarse];
+ ftswC = &DC[DIR_MMP * numberOfLBnodesCoarse];
+ ftseC = &DC[DIR_PMP * numberOfLBnodesCoarse];
+ ftnwC = &DC[DIR_MPP * numberOfLBnodesCoarse];
+ fbneC = &DC[DIR_PPM * numberOfLBnodesCoarse];
+ fbswC = &DC[DIR_MMM * numberOfLBnodesCoarse];
+ fbseC = &DC[DIR_PMM * numberOfLBnodesCoarse];
+ fbnwC = &DC[DIR_MPM * numberOfLBnodesCoarse];
}
else
{
- fwC = &DC[DIR_P00 *size_MatC];
- feC = &DC[DIR_M00 *size_MatC];
- fsC = &DC[DIR_0P0 *size_MatC];
- fnC = &DC[DIR_0M0 *size_MatC];
- fbC = &DC[DIR_00P *size_MatC];
- ftC = &DC[DIR_00M *size_MatC];
- fswC = &DC[DIR_PP0 *size_MatC];
- fneC = &DC[DIR_MM0 *size_MatC];
- fnwC = &DC[DIR_PM0 *size_MatC];
- fseC = &DC[DIR_MP0 *size_MatC];
- fbwC = &DC[DIR_P0P *size_MatC];
- fteC = &DC[DIR_M0M *size_MatC];
- ftwC = &DC[DIR_P0M *size_MatC];
- fbeC = &DC[DIR_M0P *size_MatC];
- fbsC = &DC[DIR_0PP *size_MatC];
- ftnC = &DC[DIR_0MM *size_MatC];
- ftsC = &DC[DIR_0PM *size_MatC];
- fbnC = &DC[DIR_0MP *size_MatC];
- fzeroC = &DC[DIR_000*size_MatC];
- fbswC = &DC[DIR_PPP *size_MatC];
- fbneC = &DC[DIR_MMP *size_MatC];
- fbnwC = &DC[DIR_PMP *size_MatC];
- fbseC = &DC[DIR_MPP *size_MatC];
- ftswC = &DC[DIR_PPM *size_MatC];
- ftneC = &DC[DIR_MMM *size_MatC];
- ftnwC = &DC[DIR_PMM *size_MatC];
- ftseC = &DC[DIR_MPM *size_MatC];
+ fwC = &DC[DIR_P00 * numberOfLBnodesCoarse];
+ feC = &DC[DIR_M00 * numberOfLBnodesCoarse];
+ fsC = &DC[DIR_0P0 * numberOfLBnodesCoarse];
+ fnC = &DC[DIR_0M0 * numberOfLBnodesCoarse];
+ fbC = &DC[DIR_00P * numberOfLBnodesCoarse];
+ ftC = &DC[DIR_00M * numberOfLBnodesCoarse];
+ fswC = &DC[DIR_PP0 * numberOfLBnodesCoarse];
+ fneC = &DC[DIR_MM0 * numberOfLBnodesCoarse];
+ fnwC = &DC[DIR_PM0 * numberOfLBnodesCoarse];
+ fseC = &DC[DIR_MP0 * numberOfLBnodesCoarse];
+ fbwC = &DC[DIR_P0P * numberOfLBnodesCoarse];
+ fteC = &DC[DIR_M0M * numberOfLBnodesCoarse];
+ ftwC = &DC[DIR_P0M * numberOfLBnodesCoarse];
+ fbeC = &DC[DIR_M0P * numberOfLBnodesCoarse];
+ fbsC = &DC[DIR_0PP * numberOfLBnodesCoarse];
+ ftnC = &DC[DIR_0MM * numberOfLBnodesCoarse];
+ ftsC = &DC[DIR_0PM * numberOfLBnodesCoarse];
+ fbnC = &DC[DIR_0MP * numberOfLBnodesCoarse];
+ fzeroC = &DC[DIR_000 * numberOfLBnodesCoarse];
+ fbswC = &DC[DIR_PPP * numberOfLBnodesCoarse];
+ fbneC = &DC[DIR_MMP * numberOfLBnodesCoarse];
+ fbnwC = &DC[DIR_PMP * numberOfLBnodesCoarse];
+ fbseC = &DC[DIR_MPP * numberOfLBnodesCoarse];
+ ftswC = &DC[DIR_PPM * numberOfLBnodesCoarse];
+ ftneC = &DC[DIR_MMM * numberOfLBnodesCoarse];
+ ftnwC = &DC[DIR_PMM * numberOfLBnodesCoarse];
+ ftseC = &DC[DIR_MPM * numberOfLBnodesCoarse];
}
////////////////////////////////////////////////////////////////////////////////
const unsigned ix = threadIdx.x; // Globaler x-Index
@@ -15138,8 +14334,8 @@ __global__ void scaleFC_NSPress_27( real* DC,
unsigned int* neighborFX,
unsigned int* neighborFY,
unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
+ unsigned long long numberOfLBnodesCoarse,
+ unsigned long long numberOfLBnodesFine,
bool isEvenTimestep,
unsigned int* posC,
unsigned int* posFSWB,
@@ -15156,96 +14352,96 @@ __global__ void scaleFC_NSPress_27( real* DC,
real *feF, *fwF, *fnF, *fsF, *ftF, *fbF, *fneF, *fswF, *fseF, *fnwF, *fteF, *fbwF, *fbeF, *ftwF, *ftnF, *fbsF, *fbnF, *ftsF, *fzeroF,
*ftneF, *ftswF, *ftseF, *ftnwF, *fbneF, *fbswF, *fbseF, *fbnwF;
- feF = &DF[DIR_P00 *size_MatF];
- fwF = &DF[DIR_M00 *size_MatF];
- fnF = &DF[DIR_0P0 *size_MatF];
- fsF = &DF[DIR_0M0 *size_MatF];
- ftF = &DF[DIR_00P *size_MatF];
- fbF = &DF[DIR_00M *size_MatF];
- fneF = &DF[DIR_PP0 *size_MatF];
- fswF = &DF[DIR_MM0 *size_MatF];
- fseF = &DF[DIR_PM0 *size_MatF];
- fnwF = &DF[DIR_MP0 *size_MatF];
- fteF = &DF[DIR_P0P *size_MatF];
- fbwF = &DF[DIR_M0M *size_MatF];
- fbeF = &DF[DIR_P0M *size_MatF];
- ftwF = &DF[DIR_M0P *size_MatF];
- ftnF = &DF[DIR_0PP *size_MatF];
- fbsF = &DF[DIR_0MM *size_MatF];
- fbnF = &DF[DIR_0PM *size_MatF];
- ftsF = &DF[DIR_0MP *size_MatF];
- fzeroF = &DF[DIR_000*size_MatF];
- ftneF = &DF[DIR_PPP *size_MatF];
- ftswF = &DF[DIR_MMP *size_MatF];
- ftseF = &DF[DIR_PMP *size_MatF];
- ftnwF = &DF[DIR_MPP *size_MatF];
- fbneF = &DF[DIR_PPM *size_MatF];
- fbswF = &DF[DIR_MMM *size_MatF];
- fbseF = &DF[DIR_PMM *size_MatF];
- fbnwF = &DF[DIR_MPM *size_MatF];
+ feF = &DF[DIR_P00 * numberOfLBnodesFine];
+ fwF = &DF[DIR_M00 * numberOfLBnodesFine];
+ fnF = &DF[DIR_0P0 * numberOfLBnodesFine];
+ fsF = &DF[DIR_0M0 * numberOfLBnodesFine];
+ ftF = &DF[DIR_00P * numberOfLBnodesFine];
+ fbF = &DF[DIR_00M * numberOfLBnodesFine];
+ fneF = &DF[DIR_PP0 * numberOfLBnodesFine];
+ fswF = &DF[DIR_MM0 * numberOfLBnodesFine];
+ fseF = &DF[DIR_PM0 * numberOfLBnodesFine];
+ fnwF = &DF[DIR_MP0 * numberOfLBnodesFine];
+ fteF = &DF[DIR_P0P * numberOfLBnodesFine];
+ fbwF = &DF[DIR_M0M * numberOfLBnodesFine];
+ fbeF = &DF[DIR_P0M * numberOfLBnodesFine];
+ ftwF = &DF[DIR_M0P * numberOfLBnodesFine];
+ ftnF = &DF[DIR_0PP * numberOfLBnodesFine];
+ fbsF = &DF[DIR_0MM * numberOfLBnodesFine];
+ fbnF = &DF[DIR_0PM * numberOfLBnodesFine];
+ ftsF = &DF[DIR_0MP * numberOfLBnodesFine];
+ fzeroF = &DF[DIR_000 * numberOfLBnodesFine];
+ ftneF = &DF[DIR_PPP * numberOfLBnodesFine];
+ ftswF = &DF[DIR_MMP * numberOfLBnodesFine];
+ ftseF = &DF[DIR_PMP * numberOfLBnodesFine];
+ ftnwF = &DF[DIR_MPP * numberOfLBnodesFine];
+ fbneF = &DF[DIR_PPM * numberOfLBnodesFine];
+ fbswF = &DF[DIR_MMM * numberOfLBnodesFine];
+ fbseF = &DF[DIR_PMM * numberOfLBnodesFine];
+ fbnwF = &DF[DIR_MPM * numberOfLBnodesFine];
real *feC, *fwC, *fnC, *fsC, *ftC, *fbC, *fneC, *fswC, *fseC, *fnwC, *fteC, *fbwC, *fbeC, *ftwC, *ftnC, *fbsC, *fbnC, *ftsC, *fzeroC,
*ftneC, *ftswC, *ftseC, *ftnwC, *fbneC, *fbswC, *fbseC, *fbnwC;
if (isEvenTimestep==true)
{
- feC = &DC[DIR_P00 *size_MatC];
- fwC = &DC[DIR_M00 *size_MatC];
- fnC = &DC[DIR_0P0 *size_MatC];
- fsC = &DC[DIR_0M0 *size_MatC];
- ftC = &DC[DIR_00P *size_MatC];
- fbC = &DC[DIR_00M *size_MatC];
- fneC = &DC[DIR_PP0 *size_MatC];
- fswC = &DC[DIR_MM0 *size_MatC];
- fseC = &DC[DIR_PM0 *size_MatC];
- fnwC = &DC[DIR_MP0 *size_MatC];
- fteC = &DC[DIR_P0P *size_MatC];
- fbwC = &DC[DIR_M0M *size_MatC];
- fbeC = &DC[DIR_P0M *size_MatC];
- ftwC = &DC[DIR_M0P *size_MatC];
- ftnC = &DC[DIR_0PP *size_MatC];
- fbsC = &DC[DIR_0MM *size_MatC];
- fbnC = &DC[DIR_0PM *size_MatC];
- ftsC = &DC[DIR_0MP *size_MatC];
- fzeroC = &DC[DIR_000*size_MatC];
- ftneC = &DC[DIR_PPP *size_MatC];
- ftswC = &DC[DIR_MMP *size_MatC];
- ftseC = &DC[DIR_PMP *size_MatC];
- ftnwC = &DC[DIR_MPP *size_MatC];
- fbneC = &DC[DIR_PPM *size_MatC];
- fbswC = &DC[DIR_MMM *size_MatC];
- fbseC = &DC[DIR_PMM *size_MatC];
- fbnwC = &DC[DIR_MPM *size_MatC];
+ feC = &DC[DIR_P00 * numberOfLBnodesCoarse];
+ fwC = &DC[DIR_M00 * numberOfLBnodesCoarse];
+ fnC = &DC[DIR_0P0 * numberOfLBnodesCoarse];
+ fsC = &DC[DIR_0M0 * numberOfLBnodesCoarse];
+ ftC = &DC[DIR_00P * numberOfLBnodesCoarse];
+ fbC = &DC[DIR_00M * numberOfLBnodesCoarse];
+ fneC = &DC[DIR_PP0 * numberOfLBnodesCoarse];
+ fswC = &DC[DIR_MM0 * numberOfLBnodesCoarse];
+ fseC = &DC[DIR_PM0 * numberOfLBnodesCoarse];
+ fnwC = &DC[DIR_MP0 * numberOfLBnodesCoarse];
+ fteC = &DC[DIR_P0P * numberOfLBnodesCoarse];
+ fbwC = &DC[DIR_M0M * numberOfLBnodesCoarse];
+ fbeC = &DC[DIR_P0M * numberOfLBnodesCoarse];
+ ftwC = &DC[DIR_M0P * numberOfLBnodesCoarse];
+ ftnC = &DC[DIR_0PP * numberOfLBnodesCoarse];
+ fbsC = &DC[DIR_0MM * numberOfLBnodesCoarse];
+ fbnC = &DC[DIR_0PM * numberOfLBnodesCoarse];
+ ftsC = &DC[DIR_0MP * numberOfLBnodesCoarse];
+ fzeroC = &DC[DIR_000 * numberOfLBnodesCoarse];
+ ftneC = &DC[DIR_PPP * numberOfLBnodesCoarse];
+ ftswC = &DC[DIR_MMP * numberOfLBnodesCoarse];
+ ftseC = &DC[DIR_PMP * numberOfLBnodesCoarse];
+ ftnwC = &DC[DIR_MPP * numberOfLBnodesCoarse];
+ fbneC = &DC[DIR_PPM * numberOfLBnodesCoarse];
+ fbswC = &DC[DIR_MMM * numberOfLBnodesCoarse];
+ fbseC = &DC[DIR_PMM * numberOfLBnodesCoarse];
+ fbnwC = &DC[DIR_MPM * numberOfLBnodesCoarse];
}
else
{
- fwC = &DC[DIR_P00 *size_MatC];
- feC = &DC[DIR_M00 *size_MatC];
- fsC = &DC[DIR_0P0 *size_MatC];
- fnC = &DC[DIR_0M0 *size_MatC];
- fbC = &DC[DIR_00P *size_MatC];
- ftC = &DC[DIR_00M *size_MatC];
- fswC = &DC[DIR_PP0 *size_MatC];
- fneC = &DC[DIR_MM0 *size_MatC];
- fnwC = &DC[DIR_PM0 *size_MatC];
- fseC = &DC[DIR_MP0 *size_MatC];
- fbwC = &DC[DIR_P0P *size_MatC];
- fteC = &DC[DIR_M0M *size_MatC];
- ftwC = &DC[DIR_P0M *size_MatC];
- fbeC = &DC[DIR_M0P *size_MatC];
- fbsC = &DC[DIR_0PP *size_MatC];
- ftnC = &DC[DIR_0MM *size_MatC];
- ftsC = &DC[DIR_0PM *size_MatC];
- fbnC = &DC[DIR_0MP *size_MatC];
- fzeroC = &DC[DIR_000*size_MatC];
- fbswC = &DC[DIR_PPP *size_MatC];
- fbneC = &DC[DIR_MMP *size_MatC];
- fbnwC = &DC[DIR_PMP *size_MatC];
- fbseC = &DC[DIR_MPP *size_MatC];
- ftswC = &DC[DIR_PPM *size_MatC];
- ftneC = &DC[DIR_MMM *size_MatC];
- ftnwC = &DC[DIR_PMM *size_MatC];
- ftseC = &DC[DIR_MPM *size_MatC];
+ fwC = &DC[DIR_P00 * numberOfLBnodesCoarse];
+ feC = &DC[DIR_M00 * numberOfLBnodesCoarse];
+ fsC = &DC[DIR_0P0 * numberOfLBnodesCoarse];
+ fnC = &DC[DIR_0M0 * numberOfLBnodesCoarse];
+ fbC = &DC[DIR_00P * numberOfLBnodesCoarse];
+ ftC = &DC[DIR_00M * numberOfLBnodesCoarse];
+ fswC = &DC[DIR_PP0 * numberOfLBnodesCoarse];
+ fneC = &DC[DIR_MM0 * numberOfLBnodesCoarse];
+ fnwC = &DC[DIR_PM0 * numberOfLBnodesCoarse];
+ fseC = &DC[DIR_MP0 * numberOfLBnodesCoarse];
+ fbwC = &DC[DIR_P0P * numberOfLBnodesCoarse];
+ fteC = &DC[DIR_M0M * numberOfLBnodesCoarse];
+ ftwC = &DC[DIR_P0M * numberOfLBnodesCoarse];
+ fbeC = &DC[DIR_M0P * numberOfLBnodesCoarse];
+ fbsC = &DC[DIR_0PP * numberOfLBnodesCoarse];
+ ftnC = &DC[DIR_0MM * numberOfLBnodesCoarse];
+ ftsC = &DC[DIR_0PM * numberOfLBnodesCoarse];
+ fbnC = &DC[DIR_0MP * numberOfLBnodesCoarse];
+ fzeroC = &DC[DIR_000 * numberOfLBnodesCoarse];
+ fbswC = &DC[DIR_PPP * numberOfLBnodesCoarse];
+ fbneC = &DC[DIR_MMP * numberOfLBnodesCoarse];
+ fbnwC = &DC[DIR_PMP * numberOfLBnodesCoarse];
+ fbseC = &DC[DIR_MPP * numberOfLBnodesCoarse];
+ ftswC = &DC[DIR_PPM * numberOfLBnodesCoarse];
+ ftneC = &DC[DIR_MMM * numberOfLBnodesCoarse];
+ ftnwC = &DC[DIR_PMM * numberOfLBnodesCoarse];
+ ftseC = &DC[DIR_MPM * numberOfLBnodesCoarse];
}
////////////////////////////////////////////////////////////////////////////////
const unsigned ix = threadIdx.x; // Globaler x-Index
@@ -16344,8 +15540,8 @@ __global__ void scaleFC_Fix_27( real* DC,
unsigned int* neighborFX,
unsigned int* neighborFY,
unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
+ unsigned long long numberOfLBnodesCoarse,
+ unsigned long long numberOfLBnodesFine,
bool isEvenTimestep,
unsigned int* posC,
unsigned int* posFSWB,
@@ -16362,96 +15558,96 @@ __global__ void scaleFC_Fix_27( real* DC,
real *feF, *fwF, *fnF, *fsF, *ftF, *fbF, *fneF, *fswF, *fseF, *fnwF, *fteF, *fbwF, *fbeF, *ftwF, *ftnF, *fbsF, *fbnF, *ftsF, *fzeroF,
*ftneF, *ftswF, *ftseF, *ftnwF, *fbneF, *fbswF, *fbseF, *fbnwF;
- feF = &DF[DIR_P00 *size_MatF];
- fwF = &DF[DIR_M00 *size_MatF];
- fnF = &DF[DIR_0P0 *size_MatF];
- fsF = &DF[DIR_0M0 *size_MatF];
- ftF = &DF[DIR_00P *size_MatF];
- fbF = &DF[DIR_00M *size_MatF];
- fneF = &DF[DIR_PP0 *size_MatF];
- fswF = &DF[DIR_MM0 *size_MatF];
- fseF = &DF[DIR_PM0 *size_MatF];
- fnwF = &DF[DIR_MP0 *size_MatF];
- fteF = &DF[DIR_P0P *size_MatF];
- fbwF = &DF[DIR_M0M *size_MatF];
- fbeF = &DF[DIR_P0M *size_MatF];
- ftwF = &DF[DIR_M0P *size_MatF];
- ftnF = &DF[DIR_0PP *size_MatF];
- fbsF = &DF[DIR_0MM *size_MatF];
- fbnF = &DF[DIR_0PM *size_MatF];
- ftsF = &DF[DIR_0MP *size_MatF];
- fzeroF = &DF[DIR_000*size_MatF];
- ftneF = &DF[DIR_PPP *size_MatF];
- ftswF = &DF[DIR_MMP *size_MatF];
- ftseF = &DF[DIR_PMP *size_MatF];
- ftnwF = &DF[DIR_MPP *size_MatF];
- fbneF = &DF[DIR_PPM *size_MatF];
- fbswF = &DF[DIR_MMM *size_MatF];
- fbseF = &DF[DIR_PMM *size_MatF];
- fbnwF = &DF[DIR_MPM *size_MatF];
+ feF = &DF[DIR_P00 * numberOfLBnodesFine];
+ fwF = &DF[DIR_M00 * numberOfLBnodesFine];
+ fnF = &DF[DIR_0P0 * numberOfLBnodesFine];
+ fsF = &DF[DIR_0M0 * numberOfLBnodesFine];
+ ftF = &DF[DIR_00P * numberOfLBnodesFine];
+ fbF = &DF[DIR_00M * numberOfLBnodesFine];
+ fneF = &DF[DIR_PP0 * numberOfLBnodesFine];
+ fswF = &DF[DIR_MM0 * numberOfLBnodesFine];
+ fseF = &DF[DIR_PM0 * numberOfLBnodesFine];
+ fnwF = &DF[DIR_MP0 * numberOfLBnodesFine];
+ fteF = &DF[DIR_P0P * numberOfLBnodesFine];
+ fbwF = &DF[DIR_M0M * numberOfLBnodesFine];
+ fbeF = &DF[DIR_P0M * numberOfLBnodesFine];
+ ftwF = &DF[DIR_M0P * numberOfLBnodesFine];
+ ftnF = &DF[DIR_0PP * numberOfLBnodesFine];
+ fbsF = &DF[DIR_0MM * numberOfLBnodesFine];
+ fbnF = &DF[DIR_0PM * numberOfLBnodesFine];
+ ftsF = &DF[DIR_0MP * numberOfLBnodesFine];
+ fzeroF = &DF[DIR_000 * numberOfLBnodesFine];
+ ftneF = &DF[DIR_PPP * numberOfLBnodesFine];
+ ftswF = &DF[DIR_MMP * numberOfLBnodesFine];
+ ftseF = &DF[DIR_PMP * numberOfLBnodesFine];
+ ftnwF = &DF[DIR_MPP * numberOfLBnodesFine];
+ fbneF = &DF[DIR_PPM * numberOfLBnodesFine];
+ fbswF = &DF[DIR_MMM * numberOfLBnodesFine];
+ fbseF = &DF[DIR_PMM * numberOfLBnodesFine];
+ fbnwF = &DF[DIR_MPM * numberOfLBnodesFine];
real *feC, *fwC, *fnC, *fsC, *ftC, *fbC, *fneC, *fswC, *fseC, *fnwC, *fteC, *fbwC, *fbeC, *ftwC, *ftnC, *fbsC, *fbnC, *ftsC, *fzeroC,
*ftneC, *ftswC, *ftseC, *ftnwC, *fbneC, *fbswC, *fbseC, *fbnwC;
if (isEvenTimestep==true)
{
- feC = &DC[DIR_P00 *size_MatC];
- fwC = &DC[DIR_M00 *size_MatC];
- fnC = &DC[DIR_0P0 *size_MatC];
- fsC = &DC[DIR_0M0 *size_MatC];
- ftC = &DC[DIR_00P *size_MatC];
- fbC = &DC[DIR_00M *size_MatC];
- fneC = &DC[DIR_PP0 *size_MatC];
- fswC = &DC[DIR_MM0 *size_MatC];
- fseC = &DC[DIR_PM0 *size_MatC];
- fnwC = &DC[DIR_MP0 *size_MatC];
- fteC = &DC[DIR_P0P *size_MatC];
- fbwC = &DC[DIR_M0M *size_MatC];
- fbeC = &DC[DIR_P0M *size_MatC];
- ftwC = &DC[DIR_M0P *size_MatC];
- ftnC = &DC[DIR_0PP *size_MatC];
- fbsC = &DC[DIR_0MM *size_MatC];
- fbnC = &DC[DIR_0PM *size_MatC];
- ftsC = &DC[DIR_0MP *size_MatC];
- fzeroC = &DC[DIR_000*size_MatC];
- ftneC = &DC[DIR_PPP *size_MatC];
- ftswC = &DC[DIR_MMP *size_MatC];
- ftseC = &DC[DIR_PMP *size_MatC];
- ftnwC = &DC[DIR_MPP *size_MatC];
- fbneC = &DC[DIR_PPM *size_MatC];
- fbswC = &DC[DIR_MMM *size_MatC];
- fbseC = &DC[DIR_PMM *size_MatC];
- fbnwC = &DC[DIR_MPM *size_MatC];
+ feC = &DC[DIR_P00 * numberOfLBnodesCoarse];
+ fwC = &DC[DIR_M00 * numberOfLBnodesCoarse];
+ fnC = &DC[DIR_0P0 * numberOfLBnodesCoarse];
+ fsC = &DC[DIR_0M0 * numberOfLBnodesCoarse];
+ ftC = &DC[DIR_00P * numberOfLBnodesCoarse];
+ fbC = &DC[DIR_00M * numberOfLBnodesCoarse];
+ fneC = &DC[DIR_PP0 * numberOfLBnodesCoarse];
+ fswC = &DC[DIR_MM0 * numberOfLBnodesCoarse];
+ fseC = &DC[DIR_PM0 * numberOfLBnodesCoarse];
+ fnwC = &DC[DIR_MP0 * numberOfLBnodesCoarse];
+ fteC = &DC[DIR_P0P * numberOfLBnodesCoarse];
+ fbwC = &DC[DIR_M0M * numberOfLBnodesCoarse];
+ fbeC = &DC[DIR_P0M * numberOfLBnodesCoarse];
+ ftwC = &DC[DIR_M0P * numberOfLBnodesCoarse];
+ ftnC = &DC[DIR_0PP * numberOfLBnodesCoarse];
+ fbsC = &DC[DIR_0MM * numberOfLBnodesCoarse];
+ fbnC = &DC[DIR_0PM * numberOfLBnodesCoarse];
+ ftsC = &DC[DIR_0MP * numberOfLBnodesCoarse];
+ fzeroC = &DC[DIR_000 * numberOfLBnodesCoarse];
+ ftneC = &DC[DIR_PPP * numberOfLBnodesCoarse];
+ ftswC = &DC[DIR_MMP * numberOfLBnodesCoarse];
+ ftseC = &DC[DIR_PMP * numberOfLBnodesCoarse];
+ ftnwC = &DC[DIR_MPP * numberOfLBnodesCoarse];
+ fbneC = &DC[DIR_PPM * numberOfLBnodesCoarse];
+ fbswC = &DC[DIR_MMM * numberOfLBnodesCoarse];
+ fbseC = &DC[DIR_PMM * numberOfLBnodesCoarse];
+ fbnwC = &DC[DIR_MPM * numberOfLBnodesCoarse];
}
else
{
- fwC = &DC[DIR_P00 *size_MatC];
- feC = &DC[DIR_M00 *size_MatC];
- fsC = &DC[DIR_0P0 *size_MatC];
- fnC = &DC[DIR_0M0 *size_MatC];
- fbC = &DC[DIR_00P *size_MatC];
- ftC = &DC[DIR_00M *size_MatC];
- fswC = &DC[DIR_PP0 *size_MatC];
- fneC = &DC[DIR_MM0 *size_MatC];
- fnwC = &DC[DIR_PM0 *size_MatC];
- fseC = &DC[DIR_MP0 *size_MatC];
- fbwC = &DC[DIR_P0P *size_MatC];
- fteC = &DC[DIR_M0M *size_MatC];
- ftwC = &DC[DIR_P0M *size_MatC];
- fbeC = &DC[DIR_M0P *size_MatC];
- fbsC = &DC[DIR_0PP *size_MatC];
- ftnC = &DC[DIR_0MM *size_MatC];
- ftsC = &DC[DIR_0PM *size_MatC];
- fbnC = &DC[DIR_0MP *size_MatC];
- fzeroC = &DC[DIR_000*size_MatC];
- fbswC = &DC[DIR_PPP *size_MatC];
- fbneC = &DC[DIR_MMP *size_MatC];
- fbnwC = &DC[DIR_PMP *size_MatC];
- fbseC = &DC[DIR_MPP *size_MatC];
- ftswC = &DC[DIR_PPM *size_MatC];
- ftneC = &DC[DIR_MMM *size_MatC];
- ftnwC = &DC[DIR_PMM *size_MatC];
- ftseC = &DC[DIR_MPM *size_MatC];
+ fwC = &DC[DIR_P00 * numberOfLBnodesCoarse];
+ feC = &DC[DIR_M00 * numberOfLBnodesCoarse];
+ fsC = &DC[DIR_0P0 * numberOfLBnodesCoarse];
+ fnC = &DC[DIR_0M0 * numberOfLBnodesCoarse];
+ fbC = &DC[DIR_00P * numberOfLBnodesCoarse];
+ ftC = &DC[DIR_00M * numberOfLBnodesCoarse];
+ fswC = &DC[DIR_PP0 * numberOfLBnodesCoarse];
+ fneC = &DC[DIR_MM0 * numberOfLBnodesCoarse];
+ fnwC = &DC[DIR_PM0 * numberOfLBnodesCoarse];
+ fseC = &DC[DIR_MP0 * numberOfLBnodesCoarse];
+ fbwC = &DC[DIR_P0P * numberOfLBnodesCoarse];
+ fteC = &DC[DIR_M0M * numberOfLBnodesCoarse];
+ ftwC = &DC[DIR_P0M * numberOfLBnodesCoarse];
+ fbeC = &DC[DIR_M0P * numberOfLBnodesCoarse];
+ fbsC = &DC[DIR_0PP * numberOfLBnodesCoarse];
+ ftnC = &DC[DIR_0MM * numberOfLBnodesCoarse];
+ ftsC = &DC[DIR_0PM * numberOfLBnodesCoarse];
+ fbnC = &DC[DIR_0MP * numberOfLBnodesCoarse];
+ fzeroC = &DC[DIR_000 * numberOfLBnodesCoarse];
+ fbswC = &DC[DIR_PPP * numberOfLBnodesCoarse];
+ fbneC = &DC[DIR_MMP * numberOfLBnodesCoarse];
+ fbnwC = &DC[DIR_PMP * numberOfLBnodesCoarse];
+ fbseC = &DC[DIR_MPP * numberOfLBnodesCoarse];
+ ftswC = &DC[DIR_PPM * numberOfLBnodesCoarse];
+ ftneC = &DC[DIR_MMM * numberOfLBnodesCoarse];
+ ftnwC = &DC[DIR_PMM * numberOfLBnodesCoarse];
+ ftseC = &DC[DIR_MPM * numberOfLBnodesCoarse];
}
////////////////////////////////////////////////////////////////////////////////
const unsigned ix = threadIdx.x; // Globaler x-Index
@@ -17704,8 +16900,8 @@ __global__ void scaleFCpress27(real* DC,
unsigned int* neighborFX,
unsigned int* neighborFY,
unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
+ unsigned long long numberOfLBnodesCoarse,
+ unsigned long long numberOfLBnodesFine,
bool isEvenTimestep,
unsigned int* posC,
unsigned int* posFSWB,
@@ -17722,96 +16918,96 @@ __global__ void scaleFCpress27(real* DC,
real *feF, *fwF, *fnF, *fsF, *ftF, *fbF, *fneF, *fswF, *fseF, *fnwF, *fteF, *fbwF, *fbeF, *ftwF, *ftnF, *fbsF, *fbnF, *ftsF, *fzeroF,
*ftneF, *ftswF, *ftseF, *ftnwF, *fbneF, *fbswF, *fbseF, *fbnwF;
- feF = &DF[DIR_P00 *size_MatF];
- fwF = &DF[DIR_M00 *size_MatF];
- fnF = &DF[DIR_0P0 *size_MatF];
- fsF = &DF[DIR_0M0 *size_MatF];
- ftF = &DF[DIR_00P *size_MatF];
- fbF = &DF[DIR_00M *size_MatF];
- fneF = &DF[DIR_PP0 *size_MatF];
- fswF = &DF[DIR_MM0 *size_MatF];
- fseF = &DF[DIR_PM0 *size_MatF];
- fnwF = &DF[DIR_MP0 *size_MatF];
- fteF = &DF[DIR_P0P *size_MatF];
- fbwF = &DF[DIR_M0M *size_MatF];
- fbeF = &DF[DIR_P0M *size_MatF];
- ftwF = &DF[DIR_M0P *size_MatF];
- ftnF = &DF[DIR_0PP *size_MatF];
- fbsF = &DF[DIR_0MM *size_MatF];
- fbnF = &DF[DIR_0PM *size_MatF];
- ftsF = &DF[DIR_0MP *size_MatF];
- fzeroF = &DF[DIR_000*size_MatF];
- ftneF = &DF[DIR_PPP *size_MatF];
- ftswF = &DF[DIR_MMP *size_MatF];
- ftseF = &DF[DIR_PMP *size_MatF];
- ftnwF = &DF[DIR_MPP *size_MatF];
- fbneF = &DF[DIR_PPM *size_MatF];
- fbswF = &DF[DIR_MMM *size_MatF];
- fbseF = &DF[DIR_PMM *size_MatF];
- fbnwF = &DF[DIR_MPM *size_MatF];
+ feF = &DF[DIR_P00 * numberOfLBnodesFine];
+ fwF = &DF[DIR_M00 * numberOfLBnodesFine];
+ fnF = &DF[DIR_0P0 * numberOfLBnodesFine];
+ fsF = &DF[DIR_0M0 * numberOfLBnodesFine];
+ ftF = &DF[DIR_00P * numberOfLBnodesFine];
+ fbF = &DF[DIR_00M * numberOfLBnodesFine];
+ fneF = &DF[DIR_PP0 * numberOfLBnodesFine];
+ fswF = &DF[DIR_MM0 * numberOfLBnodesFine];
+ fseF = &DF[DIR_PM0 * numberOfLBnodesFine];
+ fnwF = &DF[DIR_MP0 * numberOfLBnodesFine];
+ fteF = &DF[DIR_P0P * numberOfLBnodesFine];
+ fbwF = &DF[DIR_M0M * numberOfLBnodesFine];
+ fbeF = &DF[DIR_P0M * numberOfLBnodesFine];
+ ftwF = &DF[DIR_M0P * numberOfLBnodesFine];
+ ftnF = &DF[DIR_0PP * numberOfLBnodesFine];
+ fbsF = &DF[DIR_0MM * numberOfLBnodesFine];
+ fbnF = &DF[DIR_0PM * numberOfLBnodesFine];
+ ftsF = &DF[DIR_0MP * numberOfLBnodesFine];
+ fzeroF = &DF[DIR_000 * numberOfLBnodesFine];
+ ftneF = &DF[DIR_PPP * numberOfLBnodesFine];
+ ftswF = &DF[DIR_MMP * numberOfLBnodesFine];
+ ftseF = &DF[DIR_PMP * numberOfLBnodesFine];
+ ftnwF = &DF[DIR_MPP * numberOfLBnodesFine];
+ fbneF = &DF[DIR_PPM * numberOfLBnodesFine];
+ fbswF = &DF[DIR_MMM * numberOfLBnodesFine];
+ fbseF = &DF[DIR_PMM * numberOfLBnodesFine];
+ fbnwF = &DF[DIR_MPM * numberOfLBnodesFine];
real *feC, *fwC, *fnC, *fsC, *ftC, *fbC, *fneC, *fswC, *fseC, *fnwC, *fteC, *fbwC, *fbeC, *ftwC, *ftnC, *fbsC, *fbnC, *ftsC, *fzeroC,
*ftneC, *ftswC, *ftseC, *ftnwC, *fbneC, *fbswC, *fbseC, *fbnwC;
if (isEvenTimestep==true)
{
- feC = &DC[DIR_P00 *size_MatC];
- fwC = &DC[DIR_M00 *size_MatC];
- fnC = &DC[DIR_0P0 *size_MatC];
- fsC = &DC[DIR_0M0 *size_MatC];
- ftC = &DC[DIR_00P *size_MatC];
- fbC = &DC[DIR_00M *size_MatC];
- fneC = &DC[DIR_PP0 *size_MatC];
- fswC = &DC[DIR_MM0 *size_MatC];
- fseC = &DC[DIR_PM0 *size_MatC];
- fnwC = &DC[DIR_MP0 *size_MatC];
- fteC = &DC[DIR_P0P *size_MatC];
- fbwC = &DC[DIR_M0M *size_MatC];
- fbeC = &DC[DIR_P0M *size_MatC];
- ftwC = &DC[DIR_M0P *size_MatC];
- ftnC = &DC[DIR_0PP *size_MatC];
- fbsC = &DC[DIR_0MM *size_MatC];
- fbnC = &DC[DIR_0PM *size_MatC];
- ftsC = &DC[DIR_0MP *size_MatC];
- fzeroC = &DC[DIR_000*size_MatC];
- ftneC = &DC[DIR_PPP *size_MatC];
- ftswC = &DC[DIR_MMP *size_MatC];
- ftseC = &DC[DIR_PMP *size_MatC];
- ftnwC = &DC[DIR_MPP *size_MatC];
- fbneC = &DC[DIR_PPM *size_MatC];
- fbswC = &DC[DIR_MMM *size_MatC];
- fbseC = &DC[DIR_PMM *size_MatC];
- fbnwC = &DC[DIR_MPM *size_MatC];
+ feC = &DC[DIR_P00 * numberOfLBnodesCoarse];
+ fwC = &DC[DIR_M00 * numberOfLBnodesCoarse];
+ fnC = &DC[DIR_0P0 * numberOfLBnodesCoarse];
+ fsC = &DC[DIR_0M0 * numberOfLBnodesCoarse];
+ ftC = &DC[DIR_00P * numberOfLBnodesCoarse];
+ fbC = &DC[DIR_00M * numberOfLBnodesCoarse];
+ fneC = &DC[DIR_PP0 * numberOfLBnodesCoarse];
+ fswC = &DC[DIR_MM0 * numberOfLBnodesCoarse];
+ fseC = &DC[DIR_PM0 * numberOfLBnodesCoarse];
+ fnwC = &DC[DIR_MP0 * numberOfLBnodesCoarse];
+ fteC = &DC[DIR_P0P * numberOfLBnodesCoarse];
+ fbwC = &DC[DIR_M0M * numberOfLBnodesCoarse];
+ fbeC = &DC[DIR_P0M * numberOfLBnodesCoarse];
+ ftwC = &DC[DIR_M0P * numberOfLBnodesCoarse];
+ ftnC = &DC[DIR_0PP * numberOfLBnodesCoarse];
+ fbsC = &DC[DIR_0MM * numberOfLBnodesCoarse];
+ fbnC = &DC[DIR_0PM * numberOfLBnodesCoarse];
+ ftsC = &DC[DIR_0MP * numberOfLBnodesCoarse];
+ fzeroC = &DC[DIR_000 * numberOfLBnodesCoarse];
+ ftneC = &DC[DIR_PPP * numberOfLBnodesCoarse];
+ ftswC = &DC[DIR_MMP * numberOfLBnodesCoarse];
+ ftseC = &DC[DIR_PMP * numberOfLBnodesCoarse];
+ ftnwC = &DC[DIR_MPP * numberOfLBnodesCoarse];
+ fbneC = &DC[DIR_PPM * numberOfLBnodesCoarse];
+ fbswC = &DC[DIR_MMM * numberOfLBnodesCoarse];
+ fbseC = &DC[DIR_PMM * numberOfLBnodesCoarse];
+ fbnwC = &DC[DIR_MPM * numberOfLBnodesCoarse];
}
else
{
- fwC = &DC[DIR_P00 *size_MatC];
- feC = &DC[DIR_M00 *size_MatC];
- fsC = &DC[DIR_0P0 *size_MatC];
- fnC = &DC[DIR_0M0 *size_MatC];
- fbC = &DC[DIR_00P *size_MatC];
- ftC = &DC[DIR_00M *size_MatC];
- fswC = &DC[DIR_PP0 *size_MatC];
- fneC = &DC[DIR_MM0 *size_MatC];
- fnwC = &DC[DIR_PM0 *size_MatC];
- fseC = &DC[DIR_MP0 *size_MatC];
- fbwC = &DC[DIR_P0P *size_MatC];
- fteC = &DC[DIR_M0M *size_MatC];
- ftwC = &DC[DIR_P0M *size_MatC];
- fbeC = &DC[DIR_M0P *size_MatC];
- fbsC = &DC[DIR_0PP *size_MatC];
- ftnC = &DC[DIR_0MM *size_MatC];
- ftsC = &DC[DIR_0PM *size_MatC];
- fbnC = &DC[DIR_0MP *size_MatC];
- fzeroC = &DC[DIR_000*size_MatC];
- fbswC = &DC[DIR_PPP *size_MatC];
- fbneC = &DC[DIR_MMP *size_MatC];
- fbnwC = &DC[DIR_PMP *size_MatC];
- fbseC = &DC[DIR_MPP *size_MatC];
- ftswC = &DC[DIR_PPM *size_MatC];
- ftneC = &DC[DIR_MMM *size_MatC];
- ftnwC = &DC[DIR_PMM *size_MatC];
- ftseC = &DC[DIR_MPM *size_MatC];
+ fwC = &DC[DIR_P00 * numberOfLBnodesCoarse];
+ feC = &DC[DIR_M00 * numberOfLBnodesCoarse];
+ fsC = &DC[DIR_0P0 * numberOfLBnodesCoarse];
+ fnC = &DC[DIR_0M0 * numberOfLBnodesCoarse];
+ fbC = &DC[DIR_00P * numberOfLBnodesCoarse];
+ ftC = &DC[DIR_00M * numberOfLBnodesCoarse];
+ fswC = &DC[DIR_PP0 * numberOfLBnodesCoarse];
+ fneC = &DC[DIR_MM0 * numberOfLBnodesCoarse];
+ fnwC = &DC[DIR_PM0 * numberOfLBnodesCoarse];
+ fseC = &DC[DIR_MP0 * numberOfLBnodesCoarse];
+ fbwC = &DC[DIR_P0P * numberOfLBnodesCoarse];
+ fteC = &DC[DIR_M0M * numberOfLBnodesCoarse];
+ ftwC = &DC[DIR_P0M * numberOfLBnodesCoarse];
+ fbeC = &DC[DIR_M0P * numberOfLBnodesCoarse];
+ fbsC = &DC[DIR_0PP * numberOfLBnodesCoarse];
+ ftnC = &DC[DIR_0MM * numberOfLBnodesCoarse];
+ ftsC = &DC[DIR_0PM * numberOfLBnodesCoarse];
+ fbnC = &DC[DIR_0MP * numberOfLBnodesCoarse];
+ fzeroC = &DC[DIR_000 * numberOfLBnodesCoarse];
+ fbswC = &DC[DIR_PPP * numberOfLBnodesCoarse];
+ fbneC = &DC[DIR_MMP * numberOfLBnodesCoarse];
+ fbnwC = &DC[DIR_PMP * numberOfLBnodesCoarse];
+ fbseC = &DC[DIR_MPP * numberOfLBnodesCoarse];
+ ftswC = &DC[DIR_PPM * numberOfLBnodesCoarse];
+ ftneC = &DC[DIR_MMM * numberOfLBnodesCoarse];
+ ftnwC = &DC[DIR_PMM * numberOfLBnodesCoarse];
+ ftseC = &DC[DIR_MPM * numberOfLBnodesCoarse];
}
////////////////////////////////////////////////////////////////////////////////
const unsigned ix = threadIdx.x; // Globaler x-Index
@@ -18629,8 +17825,8 @@ __global__ void scaleFCLast27( real* DC,
unsigned int* neighborFX,
unsigned int* neighborFY,
unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
+ unsigned long long numberOfLBnodesCoarse,
+ unsigned long long numberOfLBnodesFine,
bool isEvenTimestep,
unsigned int* posC,
unsigned int* posFSWB,
@@ -18647,96 +17843,96 @@ __global__ void scaleFCLast27( real* DC,
real *feF, *fwF, *fnF, *fsF, *ftF, *fbF, *fneF, *fswF, *fseF, *fnwF, *fteF, *fbwF, *fbeF, *ftwF, *ftnF, *fbsF, *fbnF, *ftsF, *fzeroF,
*ftneF, *ftswF, *ftseF, *ftnwF, *fbneF, *fbswF, *fbseF, *fbnwF;
- feF = &DF[DIR_P00 *size_MatF];
- fwF = &DF[DIR_M00 *size_MatF];
- fnF = &DF[DIR_0P0 *size_MatF];
- fsF = &DF[DIR_0M0 *size_MatF];
- ftF = &DF[DIR_00P *size_MatF];
- fbF = &DF[DIR_00M *size_MatF];
- fneF = &DF[DIR_PP0 *size_MatF];
- fswF = &DF[DIR_MM0 *size_MatF];
- fseF = &DF[DIR_PM0 *size_MatF];
- fnwF = &DF[DIR_MP0 *size_MatF];
- fteF = &DF[DIR_P0P *size_MatF];
- fbwF = &DF[DIR_M0M *size_MatF];
- fbeF = &DF[DIR_P0M *size_MatF];
- ftwF = &DF[DIR_M0P *size_MatF];
- ftnF = &DF[DIR_0PP *size_MatF];
- fbsF = &DF[DIR_0MM *size_MatF];
- fbnF = &DF[DIR_0PM *size_MatF];
- ftsF = &DF[DIR_0MP *size_MatF];
- fzeroF = &DF[DIR_000*size_MatF];
- ftneF = &DF[DIR_PPP *size_MatF];
- ftswF = &DF[DIR_MMP *size_MatF];
- ftseF = &DF[DIR_PMP *size_MatF];
- ftnwF = &DF[DIR_MPP *size_MatF];
- fbneF = &DF[DIR_PPM *size_MatF];
- fbswF = &DF[DIR_MMM *size_MatF];
- fbseF = &DF[DIR_PMM *size_MatF];
- fbnwF = &DF[DIR_MPM *size_MatF];
+ feF = &DF[DIR_P00 * numberOfLBnodesFine];
+ fwF = &DF[DIR_M00 * numberOfLBnodesFine];
+ fnF = &DF[DIR_0P0 * numberOfLBnodesFine];
+ fsF = &DF[DIR_0M0 * numberOfLBnodesFine];
+ ftF = &DF[DIR_00P * numberOfLBnodesFine];
+ fbF = &DF[DIR_00M * numberOfLBnodesFine];
+ fneF = &DF[DIR_PP0 * numberOfLBnodesFine];
+ fswF = &DF[DIR_MM0 * numberOfLBnodesFine];
+ fseF = &DF[DIR_PM0 * numberOfLBnodesFine];
+ fnwF = &DF[DIR_MP0 * numberOfLBnodesFine];
+ fteF = &DF[DIR_P0P * numberOfLBnodesFine];
+ fbwF = &DF[DIR_M0M * numberOfLBnodesFine];
+ fbeF = &DF[DIR_P0M * numberOfLBnodesFine];
+ ftwF = &DF[DIR_M0P * numberOfLBnodesFine];
+ ftnF = &DF[DIR_0PP * numberOfLBnodesFine];
+ fbsF = &DF[DIR_0MM * numberOfLBnodesFine];
+ fbnF = &DF[DIR_0PM * numberOfLBnodesFine];
+ ftsF = &DF[DIR_0MP * numberOfLBnodesFine];
+ fzeroF = &DF[DIR_000 * numberOfLBnodesFine];
+ ftneF = &DF[DIR_PPP * numberOfLBnodesFine];
+ ftswF = &DF[DIR_MMP * numberOfLBnodesFine];
+ ftseF = &DF[DIR_PMP * numberOfLBnodesFine];
+ ftnwF = &DF[DIR_MPP * numberOfLBnodesFine];
+ fbneF = &DF[DIR_PPM * numberOfLBnodesFine];
+ fbswF = &DF[DIR_MMM * numberOfLBnodesFine];
+ fbseF = &DF[DIR_PMM * numberOfLBnodesFine];
+ fbnwF = &DF[DIR_MPM * numberOfLBnodesFine];
real *feC, *fwC, *fnC, *fsC, *ftC, *fbC, *fneC, *fswC, *fseC, *fnwC, *fteC, *fbwC, *fbeC, *ftwC, *ftnC, *fbsC, *fbnC, *ftsC, *fzeroC,
*ftneC, *ftswC, *ftseC, *ftnwC, *fbneC, *fbswC, *fbseC, *fbnwC;
if (isEvenTimestep==true)
{
- feC = &DC[DIR_P00 *size_MatC];
- fwC = &DC[DIR_M00 *size_MatC];
- fnC = &DC[DIR_0P0 *size_MatC];
- fsC = &DC[DIR_0M0 *size_MatC];
- ftC = &DC[DIR_00P *size_MatC];
- fbC = &DC[DIR_00M *size_MatC];
- fneC = &DC[DIR_PP0 *size_MatC];
- fswC = &DC[DIR_MM0 *size_MatC];
- fseC = &DC[DIR_PM0 *size_MatC];
- fnwC = &DC[DIR_MP0 *size_MatC];
- fteC = &DC[DIR_P0P *size_MatC];
- fbwC = &DC[DIR_M0M *size_MatC];
- fbeC = &DC[DIR_P0M *size_MatC];
- ftwC = &DC[DIR_M0P *size_MatC];
- ftnC = &DC[DIR_0PP *size_MatC];
- fbsC = &DC[DIR_0MM *size_MatC];
- fbnC = &DC[DIR_0PM *size_MatC];
- ftsC = &DC[DIR_0MP *size_MatC];
- fzeroC = &DC[DIR_000*size_MatC];
- ftneC = &DC[DIR_PPP *size_MatC];
- ftswC = &DC[DIR_MMP *size_MatC];
- ftseC = &DC[DIR_PMP *size_MatC];
- ftnwC = &DC[DIR_MPP *size_MatC];
- fbneC = &DC[DIR_PPM *size_MatC];
- fbswC = &DC[DIR_MMM *size_MatC];
- fbseC = &DC[DIR_PMM *size_MatC];
- fbnwC = &DC[DIR_MPM *size_MatC];
+ feC = &DC[DIR_P00 * numberOfLBnodesCoarse];
+ fwC = &DC[DIR_M00 * numberOfLBnodesCoarse];
+ fnC = &DC[DIR_0P0 * numberOfLBnodesCoarse];
+ fsC = &DC[DIR_0M0 * numberOfLBnodesCoarse];
+ ftC = &DC[DIR_00P * numberOfLBnodesCoarse];
+ fbC = &DC[DIR_00M * numberOfLBnodesCoarse];
+ fneC = &DC[DIR_PP0 * numberOfLBnodesCoarse];
+ fswC = &DC[DIR_MM0 * numberOfLBnodesCoarse];
+ fseC = &DC[DIR_PM0 * numberOfLBnodesCoarse];
+ fnwC = &DC[DIR_MP0 * numberOfLBnodesCoarse];
+ fteC = &DC[DIR_P0P * numberOfLBnodesCoarse];
+ fbwC = &DC[DIR_M0M * numberOfLBnodesCoarse];
+ fbeC = &DC[DIR_P0M * numberOfLBnodesCoarse];
+ ftwC = &DC[DIR_M0P * numberOfLBnodesCoarse];
+ ftnC = &DC[DIR_0PP * numberOfLBnodesCoarse];
+ fbsC = &DC[DIR_0MM * numberOfLBnodesCoarse];
+ fbnC = &DC[DIR_0PM * numberOfLBnodesCoarse];
+ ftsC = &DC[DIR_0MP * numberOfLBnodesCoarse];
+ fzeroC = &DC[DIR_000 * numberOfLBnodesCoarse];
+ ftneC = &DC[DIR_PPP * numberOfLBnodesCoarse];
+ ftswC = &DC[DIR_MMP * numberOfLBnodesCoarse];
+ ftseC = &DC[DIR_PMP * numberOfLBnodesCoarse];
+ ftnwC = &DC[DIR_MPP * numberOfLBnodesCoarse];
+ fbneC = &DC[DIR_PPM * numberOfLBnodesCoarse];
+ fbswC = &DC[DIR_MMM * numberOfLBnodesCoarse];
+ fbseC = &DC[DIR_PMM * numberOfLBnodesCoarse];
+ fbnwC = &DC[DIR_MPM * numberOfLBnodesCoarse];
}
else
{
- fwC = &DC[DIR_P00 *size_MatC];
- feC = &DC[DIR_M00 *size_MatC];
- fsC = &DC[DIR_0P0 *size_MatC];
- fnC = &DC[DIR_0M0 *size_MatC];
- fbC = &DC[DIR_00P *size_MatC];
- ftC = &DC[DIR_00M *size_MatC];
- fswC = &DC[DIR_PP0 *size_MatC];
- fneC = &DC[DIR_MM0 *size_MatC];
- fnwC = &DC[DIR_PM0 *size_MatC];
- fseC = &DC[DIR_MP0 *size_MatC];
- fbwC = &DC[DIR_P0P *size_MatC];
- fteC = &DC[DIR_M0M *size_MatC];
- ftwC = &DC[DIR_P0M *size_MatC];
- fbeC = &DC[DIR_M0P *size_MatC];
- fbsC = &DC[DIR_0PP *size_MatC];
- ftnC = &DC[DIR_0MM *size_MatC];
- ftsC = &DC[DIR_0PM *size_MatC];
- fbnC = &DC[DIR_0MP *size_MatC];
- fzeroC = &DC[DIR_000*size_MatC];
- fbswC = &DC[DIR_PPP *size_MatC];
- fbneC = &DC[DIR_MMP *size_MatC];
- fbnwC = &DC[DIR_PMP *size_MatC];
- fbseC = &DC[DIR_MPP *size_MatC];
- ftswC = &DC[DIR_PPM *size_MatC];
- ftneC = &DC[DIR_MMM *size_MatC];
- ftnwC = &DC[DIR_PMM *size_MatC];
- ftseC = &DC[DIR_MPM *size_MatC];
+ fwC = &DC[DIR_P00 * numberOfLBnodesCoarse];
+ feC = &DC[DIR_M00 * numberOfLBnodesCoarse];
+ fsC = &DC[DIR_0P0 * numberOfLBnodesCoarse];
+ fnC = &DC[DIR_0M0 * numberOfLBnodesCoarse];
+ fbC = &DC[DIR_00P * numberOfLBnodesCoarse];
+ ftC = &DC[DIR_00M * numberOfLBnodesCoarse];
+ fswC = &DC[DIR_PP0 * numberOfLBnodesCoarse];
+ fneC = &DC[DIR_MM0 * numberOfLBnodesCoarse];
+ fnwC = &DC[DIR_PM0 * numberOfLBnodesCoarse];
+ fseC = &DC[DIR_MP0 * numberOfLBnodesCoarse];
+ fbwC = &DC[DIR_P0P * numberOfLBnodesCoarse];
+ fteC = &DC[DIR_M0M * numberOfLBnodesCoarse];
+ ftwC = &DC[DIR_P0M * numberOfLBnodesCoarse];
+ fbeC = &DC[DIR_M0P * numberOfLBnodesCoarse];
+ fbsC = &DC[DIR_0PP * numberOfLBnodesCoarse];
+ ftnC = &DC[DIR_0MM * numberOfLBnodesCoarse];
+ ftsC = &DC[DIR_0PM * numberOfLBnodesCoarse];
+ fbnC = &DC[DIR_0MP * numberOfLBnodesCoarse];
+ fzeroC = &DC[DIR_000 * numberOfLBnodesCoarse];
+ fbswC = &DC[DIR_PPP * numberOfLBnodesCoarse];
+ fbneC = &DC[DIR_MMP * numberOfLBnodesCoarse];
+ fbnwC = &DC[DIR_PMP * numberOfLBnodesCoarse];
+ fbseC = &DC[DIR_MPP * numberOfLBnodesCoarse];
+ ftswC = &DC[DIR_PPM * numberOfLBnodesCoarse];
+ ftneC = &DC[DIR_MMM * numberOfLBnodesCoarse];
+ ftnwC = &DC[DIR_PMM * numberOfLBnodesCoarse];
+ ftseC = &DC[DIR_MPM * numberOfLBnodesCoarse];
}
////////////////////////////////////////////////////////////////////////////////
const unsigned ix = threadIdx.x; // Globaler x-Index
@@ -20027,8 +19223,8 @@ __global__ void scaleFCThSMG7( real* DC,
unsigned int* neighborFX,
unsigned int* neighborFY,
unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
+ unsigned long long numberOfLBnodesCoarse,
+ unsigned long long numberOfLBnodesFine,
bool isEvenTimestep,
unsigned int* posC,
unsigned int* posFSWB,
@@ -20040,127 +19236,124 @@ __global__ void scaleFCThSMG7( real* DC,
real *feF, *fwF, *fnF, *fsF, *ftF, *fbF, *fneF, *fswF, *fseF, *fnwF, *fteF, *fbwF, *fbeF, *ftwF, *ftnF, *fbsF, *fbnF, *ftsF, //*fzeroF,
*ftneF, *ftswF, *ftseF, *ftnwF, *fbneF, *fbswF, *fbseF, *fbnwF;
- feF = &DF[DIR_P00 *size_MatF];
- fwF = &DF[DIR_M00 *size_MatF];
- fnF = &DF[DIR_0P0 *size_MatF];
- fsF = &DF[DIR_0M0 *size_MatF];
- ftF = &DF[DIR_00P *size_MatF];
- fbF = &DF[DIR_00M *size_MatF];
- fneF = &DF[DIR_PP0 *size_MatF];
- fswF = &DF[DIR_MM0 *size_MatF];
- fseF = &DF[DIR_PM0 *size_MatF];
- fnwF = &DF[DIR_MP0 *size_MatF];
- fteF = &DF[DIR_P0P *size_MatF];
- fbwF = &DF[DIR_M0M *size_MatF];
- fbeF = &DF[DIR_P0M *size_MatF];
- ftwF = &DF[DIR_M0P *size_MatF];
- ftnF = &DF[DIR_0PP *size_MatF];
- fbsF = &DF[DIR_0MM *size_MatF];
- fbnF = &DF[DIR_0PM *size_MatF];
- ftsF = &DF[DIR_0MP *size_MatF];
- //fzeroF = &DF[DIR_000*size_MatF];
- ftneF = &DF[DIR_PPP *size_MatF];
- ftswF = &DF[DIR_MMP *size_MatF];
- ftseF = &DF[DIR_PMP *size_MatF];
- ftnwF = &DF[DIR_MPP *size_MatF];
- fbneF = &DF[DIR_PPM *size_MatF];
- fbswF = &DF[DIR_MMM *size_MatF];
- fbseF = &DF[DIR_PMM *size_MatF];
- fbnwF = &DF[DIR_MPM *size_MatF];
+ feF = &DF[DIR_P00 * numberOfLBnodesFine];
+ fwF = &DF[DIR_M00 * numberOfLBnodesFine];
+ fnF = &DF[DIR_0P0 * numberOfLBnodesFine];
+ fsF = &DF[DIR_0M0 * numberOfLBnodesFine];
+ ftF = &DF[DIR_00P * numberOfLBnodesFine];
+ fbF = &DF[DIR_00M * numberOfLBnodesFine];
+ fneF = &DF[DIR_PP0 * numberOfLBnodesFine];
+ fswF = &DF[DIR_MM0 * numberOfLBnodesFine];
+ fseF = &DF[DIR_PM0 * numberOfLBnodesFine];
+ fnwF = &DF[DIR_MP0 * numberOfLBnodesFine];
+ fteF = &DF[DIR_P0P * numberOfLBnodesFine];
+ fbwF = &DF[DIR_M0M * numberOfLBnodesFine];
+ fbeF = &DF[DIR_P0M * numberOfLBnodesFine];
+ ftwF = &DF[DIR_M0P * numberOfLBnodesFine];
+ ftnF = &DF[DIR_0PP * numberOfLBnodesFine];
+ fbsF = &DF[DIR_0MM * numberOfLBnodesFine];
+ fbnF = &DF[DIR_0PM * numberOfLBnodesFine];
+ ftsF = &DF[DIR_0MP * numberOfLBnodesFine];
+ ftneF = &DF[DIR_PPP * numberOfLBnodesFine];
+ ftswF = &DF[DIR_MMP * numberOfLBnodesFine];
+ ftseF = &DF[DIR_PMP * numberOfLBnodesFine];
+ ftnwF = &DF[DIR_MPP * numberOfLBnodesFine];
+ fbneF = &DF[DIR_PPM * numberOfLBnodesFine];
+ fbswF = &DF[DIR_MMM * numberOfLBnodesFine];
+ fbseF = &DF[DIR_PMM * numberOfLBnodesFine];
+ fbnwF = &DF[DIR_MPM * numberOfLBnodesFine];
real *feC, *fwC, *fnC, *fsC, *ftC, *fbC, *fneC, *fswC, *fseC, *fnwC, *fteC, *fbwC, *fbeC, *ftwC, *ftnC, *fbsC, *fbnC, *ftsC, //*fzeroC,
*ftneC, *ftswC, *ftseC, *ftnwC, *fbneC, *fbswC, *fbseC, *fbnwC;
if (isEvenTimestep==true)
{
- feC = &DC[DIR_P00 *size_MatC];
- fwC = &DC[DIR_M00 *size_MatC];
- fnC = &DC[DIR_0P0 *size_MatC];
- fsC = &DC[DIR_0M0 *size_MatC];
- ftC = &DC[DIR_00P *size_MatC];
- fbC = &DC[DIR_00M *size_MatC];
- fneC = &DC[DIR_PP0 *size_MatC];
- fswC = &DC[DIR_MM0 *size_MatC];
- fseC = &DC[DIR_PM0 *size_MatC];
- fnwC = &DC[DIR_MP0 *size_MatC];
- fteC = &DC[DIR_P0P *size_MatC];
- fbwC = &DC[DIR_M0M *size_MatC];
- fbeC = &DC[DIR_P0M *size_MatC];
- ftwC = &DC[DIR_M0P *size_MatC];
- ftnC = &DC[DIR_0PP *size_MatC];
- fbsC = &DC[DIR_0MM *size_MatC];
- fbnC = &DC[DIR_0PM *size_MatC];
- ftsC = &DC[DIR_0MP *size_MatC];
- //fzeroC = &DC[DIR_000*size_MatC];
- ftneC = &DC[DIR_PPP *size_MatC];
- ftswC = &DC[DIR_MMP *size_MatC];
- ftseC = &DC[DIR_PMP *size_MatC];
- ftnwC = &DC[DIR_MPP *size_MatC];
- fbneC = &DC[DIR_PPM *size_MatC];
- fbswC = &DC[DIR_MMM *size_MatC];
- fbseC = &DC[DIR_PMM *size_MatC];
- fbnwC = &DC[DIR_MPM *size_MatC];
+ feC = &DC[DIR_P00 * numberOfLBnodesCoarse];
+ fwC = &DC[DIR_M00 * numberOfLBnodesCoarse];
+ fnC = &DC[DIR_0P0 * numberOfLBnodesCoarse];
+ fsC = &DC[DIR_0M0 * numberOfLBnodesCoarse];
+ ftC = &DC[DIR_00P * numberOfLBnodesCoarse];
+ fbC = &DC[DIR_00M * numberOfLBnodesCoarse];
+ fneC = &DC[DIR_PP0 * numberOfLBnodesCoarse];
+ fswC = &DC[DIR_MM0 * numberOfLBnodesCoarse];
+ fseC = &DC[DIR_PM0 * numberOfLBnodesCoarse];
+ fnwC = &DC[DIR_MP0 * numberOfLBnodesCoarse];
+ fteC = &DC[DIR_P0P * numberOfLBnodesCoarse];
+ fbwC = &DC[DIR_M0M * numberOfLBnodesCoarse];
+ fbeC = &DC[DIR_P0M * numberOfLBnodesCoarse];
+ ftwC = &DC[DIR_M0P * numberOfLBnodesCoarse];
+ ftnC = &DC[DIR_0PP * numberOfLBnodesCoarse];
+ fbsC = &DC[DIR_0MM * numberOfLBnodesCoarse];
+ fbnC = &DC[DIR_0PM * numberOfLBnodesCoarse];
+ ftsC = &DC[DIR_0MP * numberOfLBnodesCoarse];
+ ftneC = &DC[DIR_PPP * numberOfLBnodesCoarse];
+ ftswC = &DC[DIR_MMP * numberOfLBnodesCoarse];
+ ftseC = &DC[DIR_PMP * numberOfLBnodesCoarse];
+ ftnwC = &DC[DIR_MPP * numberOfLBnodesCoarse];
+ fbneC = &DC[DIR_PPM * numberOfLBnodesCoarse];
+ fbswC = &DC[DIR_MMM * numberOfLBnodesCoarse];
+ fbseC = &DC[DIR_PMM * numberOfLBnodesCoarse];
+ fbnwC = &DC[DIR_MPM * numberOfLBnodesCoarse];
}
else
{
- fwC = &DC[DIR_P00 *size_MatC];
- feC = &DC[DIR_M00 *size_MatC];
- fsC = &DC[DIR_0P0 *size_MatC];
- fnC = &DC[DIR_0M0 *size_MatC];
- fbC = &DC[DIR_00P *size_MatC];
- ftC = &DC[DIR_00M *size_MatC];
- fswC = &DC[DIR_PP0 *size_MatC];
- fneC = &DC[DIR_MM0 *size_MatC];
- fnwC = &DC[DIR_PM0 *size_MatC];
- fseC = &DC[DIR_MP0 *size_MatC];
- fbwC = &DC[DIR_P0P *size_MatC];
- fteC = &DC[DIR_M0M *size_MatC];
- ftwC = &DC[DIR_P0M *size_MatC];
- fbeC = &DC[DIR_M0P *size_MatC];
- fbsC = &DC[DIR_0PP *size_MatC];
- ftnC = &DC[DIR_0MM *size_MatC];
- ftsC = &DC[DIR_0PM *size_MatC];
- fbnC = &DC[DIR_0MP *size_MatC];
- //fzeroC = &DC[DIR_000*size_MatC];
- fbswC = &DC[DIR_PPP *size_MatC];
- fbneC = &DC[DIR_MMP *size_MatC];
- fbnwC = &DC[DIR_PMP *size_MatC];
- fbseC = &DC[DIR_MPP *size_MatC];
- ftswC = &DC[DIR_PPM *size_MatC];
- ftneC = &DC[DIR_MMM *size_MatC];
- ftnwC = &DC[DIR_PMM *size_MatC];
- ftseC = &DC[DIR_MPM *size_MatC];
+ fwC = &DC[DIR_P00 * numberOfLBnodesCoarse];
+ feC = &DC[DIR_M00 * numberOfLBnodesCoarse];
+ fsC = &DC[DIR_0P0 * numberOfLBnodesCoarse];
+ fnC = &DC[DIR_0M0 * numberOfLBnodesCoarse];
+ fbC = &DC[DIR_00P * numberOfLBnodesCoarse];
+ ftC = &DC[DIR_00M * numberOfLBnodesCoarse];
+ fswC = &DC[DIR_PP0 * numberOfLBnodesCoarse];
+ fneC = &DC[DIR_MM0 * numberOfLBnodesCoarse];
+ fnwC = &DC[DIR_PM0 * numberOfLBnodesCoarse];
+ fseC = &DC[DIR_MP0 * numberOfLBnodesCoarse];
+ fbwC = &DC[DIR_P0P * numberOfLBnodesCoarse];
+ fteC = &DC[DIR_M0M * numberOfLBnodesCoarse];
+ ftwC = &DC[DIR_P0M * numberOfLBnodesCoarse];
+ fbeC = &DC[DIR_M0P * numberOfLBnodesCoarse];
+ fbsC = &DC[DIR_0PP * numberOfLBnodesCoarse];
+ ftnC = &DC[DIR_0MM * numberOfLBnodesCoarse];
+ ftsC = &DC[DIR_0PM * numberOfLBnodesCoarse];
+ fbnC = &DC[DIR_0MP * numberOfLBnodesCoarse];
+ fbswC = &DC[DIR_PPP * numberOfLBnodesCoarse];
+ fbneC = &DC[DIR_MMP * numberOfLBnodesCoarse];
+ fbnwC = &DC[DIR_PMP * numberOfLBnodesCoarse];
+ fbseC = &DC[DIR_MPP * numberOfLBnodesCoarse];
+ ftswC = &DC[DIR_PPM * numberOfLBnodesCoarse];
+ ftneC = &DC[DIR_MMM * numberOfLBnodesCoarse];
+ ftnwC = &DC[DIR_PMM * numberOfLBnodesCoarse];
+ ftseC = &DC[DIR_MPM * numberOfLBnodesCoarse];
}
Distributions7 D7F;
- D7F.f[0] = &DD7F[0*size_MatF];
- D7F.f[1] = &DD7F[1*size_MatF];
- D7F.f[2] = &DD7F[2*size_MatF];
- D7F.f[3] = &DD7F[3*size_MatF];
- D7F.f[4] = &DD7F[4*size_MatF];
- D7F.f[5] = &DD7F[5*size_MatF];
- D7F.f[6] = &DD7F[6*size_MatF];
+ D7F.f[0] = &DD7F[0*numberOfLBnodesFine];
+ D7F.f[1] = &DD7F[1*numberOfLBnodesFine];
+ D7F.f[2] = &DD7F[2*numberOfLBnodesFine];
+ D7F.f[3] = &DD7F[3*numberOfLBnodesFine];
+ D7F.f[4] = &DD7F[4*numberOfLBnodesFine];
+ D7F.f[5] = &DD7F[5*numberOfLBnodesFine];
+ D7F.f[6] = &DD7F[6*numberOfLBnodesFine];
Distributions7 D7C;
if (isEvenTimestep==true)
{
- D7C.f[0] = &DD7C[0*size_MatC];
- D7C.f[1] = &DD7C[1*size_MatC];
- D7C.f[2] = &DD7C[2*size_MatC];
- D7C.f[3] = &DD7C[3*size_MatC];
- D7C.f[4] = &DD7C[4*size_MatC];
- D7C.f[5] = &DD7C[5*size_MatC];
- D7C.f[6] = &DD7C[6*size_MatC];
+ D7C.f[0] = &DD7C[0*numberOfLBnodesCoarse];
+ D7C.f[1] = &DD7C[1*numberOfLBnodesCoarse];
+ D7C.f[2] = &DD7C[2*numberOfLBnodesCoarse];
+ D7C.f[3] = &DD7C[3*numberOfLBnodesCoarse];
+ D7C.f[4] = &DD7C[4*numberOfLBnodesCoarse];
+ D7C.f[5] = &DD7C[5*numberOfLBnodesCoarse];
+ D7C.f[6] = &DD7C[6*numberOfLBnodesCoarse];
}
else
{
- D7C.f[0] = &DD7C[0*size_MatC];
- D7C.f[2] = &DD7C[1*size_MatC];
- D7C.f[1] = &DD7C[2*size_MatC];
- D7C.f[4] = &DD7C[3*size_MatC];
- D7C.f[3] = &DD7C[4*size_MatC];
- D7C.f[6] = &DD7C[5*size_MatC];
- D7C.f[5] = &DD7C[6*size_MatC];
+ D7C.f[0] = &DD7C[0*numberOfLBnodesCoarse];
+ D7C.f[2] = &DD7C[1*numberOfLBnodesCoarse];
+ D7C.f[1] = &DD7C[2*numberOfLBnodesCoarse];
+ D7C.f[4] = &DD7C[3*numberOfLBnodesCoarse];
+ D7C.f[3] = &DD7C[4*numberOfLBnodesCoarse];
+ D7C.f[6] = &DD7C[5*numberOfLBnodesCoarse];
+ D7C.f[5] = &DD7C[6*numberOfLBnodesCoarse];
}
////////////////////////////////////////////////////////////////////////////////
@@ -20900,8 +20093,8 @@ __global__ void scaleFCThS7( real* DC,
unsigned int* neighborFX,
unsigned int* neighborFY,
unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
+ unsigned long long numberOfLBnodesCoarse,
+ unsigned long long numberOfLBnodesFine,
bool isEvenTimestep,
unsigned int* posC,
unsigned int* posFSWB,
@@ -20912,127 +20105,124 @@ __global__ void scaleFCThS7( real* DC,
real *feF, *fwF, *fnF, *fsF, *ftF, *fbF, *fneF, *fswF, *fseF, *fnwF, *fteF, *fbwF, *fbeF, *ftwF, *ftnF, *fbsF, *fbnF, *ftsF, //*fzeroF,
*ftneF, *ftswF, *ftseF, *ftnwF, *fbneF, *fbswF, *fbseF, *fbnwF;
- feF = &DF[DIR_P00 *size_MatF];
- fwF = &DF[DIR_M00 *size_MatF];
- fnF = &DF[DIR_0P0 *size_MatF];
- fsF = &DF[DIR_0M0 *size_MatF];
- ftF = &DF[DIR_00P *size_MatF];
- fbF = &DF[DIR_00M *size_MatF];
- fneF = &DF[DIR_PP0 *size_MatF];
- fswF = &DF[DIR_MM0 *size_MatF];
- fseF = &DF[DIR_PM0 *size_MatF];
- fnwF = &DF[DIR_MP0 *size_MatF];
- fteF = &DF[DIR_P0P *size_MatF];
- fbwF = &DF[DIR_M0M *size_MatF];
- fbeF = &DF[DIR_P0M *size_MatF];
- ftwF = &DF[DIR_M0P *size_MatF];
- ftnF = &DF[DIR_0PP *size_MatF];
- fbsF = &DF[DIR_0MM *size_MatF];
- fbnF = &DF[DIR_0PM *size_MatF];
- ftsF = &DF[DIR_0MP *size_MatF];
- //fzeroF = &DF[DIR_000*size_MatF];
- ftneF = &DF[DIR_PPP *size_MatF];
- ftswF = &DF[DIR_MMP *size_MatF];
- ftseF = &DF[DIR_PMP *size_MatF];
- ftnwF = &DF[DIR_MPP *size_MatF];
- fbneF = &DF[DIR_PPM *size_MatF];
- fbswF = &DF[DIR_MMM *size_MatF];
- fbseF = &DF[DIR_PMM *size_MatF];
- fbnwF = &DF[DIR_MPM *size_MatF];
+ feF = &DF[DIR_P00 * numberOfLBnodesFine];
+ fwF = &DF[DIR_M00 * numberOfLBnodesFine];
+ fnF = &DF[DIR_0P0 * numberOfLBnodesFine];
+ fsF = &DF[DIR_0M0 * numberOfLBnodesFine];
+ ftF = &DF[DIR_00P * numberOfLBnodesFine];
+ fbF = &DF[DIR_00M * numberOfLBnodesFine];
+ fneF = &DF[DIR_PP0 * numberOfLBnodesFine];
+ fswF = &DF[DIR_MM0 * numberOfLBnodesFine];
+ fseF = &DF[DIR_PM0 * numberOfLBnodesFine];
+ fnwF = &DF[DIR_MP0 * numberOfLBnodesFine];
+ fteF = &DF[DIR_P0P * numberOfLBnodesFine];
+ fbwF = &DF[DIR_M0M * numberOfLBnodesFine];
+ fbeF = &DF[DIR_P0M * numberOfLBnodesFine];
+ ftwF = &DF[DIR_M0P * numberOfLBnodesFine];
+ ftnF = &DF[DIR_0PP * numberOfLBnodesFine];
+ fbsF = &DF[DIR_0MM * numberOfLBnodesFine];
+ fbnF = &DF[DIR_0PM * numberOfLBnodesFine];
+ ftsF = &DF[DIR_0MP * numberOfLBnodesFine];
+ ftneF = &DF[DIR_PPP * numberOfLBnodesFine];
+ ftswF = &DF[DIR_MMP * numberOfLBnodesFine];
+ ftseF = &DF[DIR_PMP * numberOfLBnodesFine];
+ ftnwF = &DF[DIR_MPP * numberOfLBnodesFine];
+ fbneF = &DF[DIR_PPM * numberOfLBnodesFine];
+ fbswF = &DF[DIR_MMM * numberOfLBnodesFine];
+ fbseF = &DF[DIR_PMM * numberOfLBnodesFine];
+ fbnwF = &DF[DIR_MPM * numberOfLBnodesFine];
real *feC, *fwC, *fnC, *fsC, *ftC, *fbC, *fneC, *fswC, *fseC, *fnwC, *fteC, *fbwC, *fbeC, *ftwC, *ftnC, *fbsC, *fbnC, *ftsC, //*fzeroC,
*ftneC, *ftswC, *ftseC, *ftnwC, *fbneC, *fbswC, *fbseC, *fbnwC;
if (isEvenTimestep==true)
{
- feC = &DC[DIR_P00 *size_MatC];
- fwC = &DC[DIR_M00 *size_MatC];
- fnC = &DC[DIR_0P0 *size_MatC];
- fsC = &DC[DIR_0M0 *size_MatC];
- ftC = &DC[DIR_00P *size_MatC];
- fbC = &DC[DIR_00M *size_MatC];
- fneC = &DC[DIR_PP0 *size_MatC];
- fswC = &DC[DIR_MM0 *size_MatC];
- fseC = &DC[DIR_PM0 *size_MatC];
- fnwC = &DC[DIR_MP0 *size_MatC];
- fteC = &DC[DIR_P0P *size_MatC];
- fbwC = &DC[DIR_M0M *size_MatC];
- fbeC = &DC[DIR_P0M *size_MatC];
- ftwC = &DC[DIR_M0P *size_MatC];
- ftnC = &DC[DIR_0PP *size_MatC];
- fbsC = &DC[DIR_0MM *size_MatC];
- fbnC = &DC[DIR_0PM *size_MatC];
- ftsC = &DC[DIR_0MP *size_MatC];
- //fzeroC = &DC[DIR_000*size_MatC];
- ftneC = &DC[DIR_PPP *size_MatC];
- ftswC = &DC[DIR_MMP *size_MatC];
- ftseC = &DC[DIR_PMP *size_MatC];
- ftnwC = &DC[DIR_MPP *size_MatC];
- fbneC = &DC[DIR_PPM *size_MatC];
- fbswC = &DC[DIR_MMM *size_MatC];
- fbseC = &DC[DIR_PMM *size_MatC];
- fbnwC = &DC[DIR_MPM *size_MatC];
+ feC = &DC[DIR_P00 * numberOfLBnodesCoarse];
+ fwC = &DC[DIR_M00 * numberOfLBnodesCoarse];
+ fnC = &DC[DIR_0P0 * numberOfLBnodesCoarse];
+ fsC = &DC[DIR_0M0 * numberOfLBnodesCoarse];
+ ftC = &DC[DIR_00P * numberOfLBnodesCoarse];
+ fbC = &DC[DIR_00M * numberOfLBnodesCoarse];
+ fneC = &DC[DIR_PP0 * numberOfLBnodesCoarse];
+ fswC = &DC[DIR_MM0 * numberOfLBnodesCoarse];
+ fseC = &DC[DIR_PM0 * numberOfLBnodesCoarse];
+ fnwC = &DC[DIR_MP0 * numberOfLBnodesCoarse];
+ fteC = &DC[DIR_P0P * numberOfLBnodesCoarse];
+ fbwC = &DC[DIR_M0M * numberOfLBnodesCoarse];
+ fbeC = &DC[DIR_P0M * numberOfLBnodesCoarse];
+ ftwC = &DC[DIR_M0P * numberOfLBnodesCoarse];
+ ftnC = &DC[DIR_0PP * numberOfLBnodesCoarse];
+ fbsC = &DC[DIR_0MM * numberOfLBnodesCoarse];
+ fbnC = &DC[DIR_0PM * numberOfLBnodesCoarse];
+ ftsC = &DC[DIR_0MP * numberOfLBnodesCoarse];
+ ftneC = &DC[DIR_PPP * numberOfLBnodesCoarse];
+ ftswC = &DC[DIR_MMP * numberOfLBnodesCoarse];
+ ftseC = &DC[DIR_PMP * numberOfLBnodesCoarse];
+ ftnwC = &DC[DIR_MPP * numberOfLBnodesCoarse];
+ fbneC = &DC[DIR_PPM * numberOfLBnodesCoarse];
+ fbswC = &DC[DIR_MMM * numberOfLBnodesCoarse];
+ fbseC = &DC[DIR_PMM * numberOfLBnodesCoarse];
+ fbnwC = &DC[DIR_MPM * numberOfLBnodesCoarse];
}
else
{
- fwC = &DC[DIR_P00 *size_MatC];
- feC = &DC[DIR_M00 *size_MatC];
- fsC = &DC[DIR_0P0 *size_MatC];
- fnC = &DC[DIR_0M0 *size_MatC];
- fbC = &DC[DIR_00P *size_MatC];
- ftC = &DC[DIR_00M *size_MatC];
- fswC = &DC[DIR_PP0 *size_MatC];
- fneC = &DC[DIR_MM0 *size_MatC];
- fnwC = &DC[DIR_PM0 *size_MatC];
- fseC = &DC[DIR_MP0 *size_MatC];
- fbwC = &DC[DIR_P0P *size_MatC];
- fteC = &DC[DIR_M0M *size_MatC];
- ftwC = &DC[DIR_P0M *size_MatC];
- fbeC = &DC[DIR_M0P *size_MatC];
- fbsC = &DC[DIR_0PP *size_MatC];
- ftnC = &DC[DIR_0MM *size_MatC];
- ftsC = &DC[DIR_0PM *size_MatC];
- fbnC = &DC[DIR_0MP *size_MatC];
- //fzeroC = &DC[DIR_000*size_MatC];
- fbswC = &DC[DIR_PPP *size_MatC];
- fbneC = &DC[DIR_MMP *size_MatC];
- fbnwC = &DC[DIR_PMP *size_MatC];
- fbseC = &DC[DIR_MPP *size_MatC];
- ftswC = &DC[DIR_PPM *size_MatC];
- ftneC = &DC[DIR_MMM *size_MatC];
- ftnwC = &DC[DIR_PMM *size_MatC];
- ftseC = &DC[DIR_MPM *size_MatC];
+ fwC = &DC[DIR_P00 * numberOfLBnodesCoarse];
+ feC = &DC[DIR_M00 * numberOfLBnodesCoarse];
+ fsC = &DC[DIR_0P0 * numberOfLBnodesCoarse];
+ fnC = &DC[DIR_0M0 * numberOfLBnodesCoarse];
+ fbC = &DC[DIR_00P * numberOfLBnodesCoarse];
+ ftC = &DC[DIR_00M * numberOfLBnodesCoarse];
+ fswC = &DC[DIR_PP0 * numberOfLBnodesCoarse];
+ fneC = &DC[DIR_MM0 * numberOfLBnodesCoarse];
+ fnwC = &DC[DIR_PM0 * numberOfLBnodesCoarse];
+ fseC = &DC[DIR_MP0 * numberOfLBnodesCoarse];
+ fbwC = &DC[DIR_P0P * numberOfLBnodesCoarse];
+ fteC = &DC[DIR_M0M * numberOfLBnodesCoarse];
+ ftwC = &DC[DIR_P0M * numberOfLBnodesCoarse];
+ fbeC = &DC[DIR_M0P * numberOfLBnodesCoarse];
+ fbsC = &DC[DIR_0PP * numberOfLBnodesCoarse];
+ ftnC = &DC[DIR_0MM * numberOfLBnodesCoarse];
+ ftsC = &DC[DIR_0PM * numberOfLBnodesCoarse];
+ fbnC = &DC[DIR_0MP * numberOfLBnodesCoarse];
+ fbswC = &DC[DIR_PPP * numberOfLBnodesCoarse];
+ fbneC = &DC[DIR_MMP * numberOfLBnodesCoarse];
+ fbnwC = &DC[DIR_PMP * numberOfLBnodesCoarse];
+ fbseC = &DC[DIR_MPP * numberOfLBnodesCoarse];
+ ftswC = &DC[DIR_PPM * numberOfLBnodesCoarse];
+ ftneC = &DC[DIR_MMM * numberOfLBnodesCoarse];
+ ftnwC = &DC[DIR_PMM * numberOfLBnodesCoarse];
+ ftseC = &DC[DIR_MPM * numberOfLBnodesCoarse];
}
Distributions7 D7F;
- D7F.f[0] = &DD7F[0*size_MatF];
- D7F.f[1] = &DD7F[1*size_MatF];
- D7F.f[2] = &DD7F[2*size_MatF];
- D7F.f[3] = &DD7F[3*size_MatF];
- D7F.f[4] = &DD7F[4*size_MatF];
- D7F.f[5] = &DD7F[5*size_MatF];
- D7F.f[6] = &DD7F[6*size_MatF];
+ D7F.f[0] = &DD7F[0*numberOfLBnodesFine];
+ D7F.f[1] = &DD7F[1*numberOfLBnodesFine];
+ D7F.f[2] = &DD7F[2*numberOfLBnodesFine];
+ D7F.f[3] = &DD7F[3*numberOfLBnodesFine];
+ D7F.f[4] = &DD7F[4*numberOfLBnodesFine];
+ D7F.f[5] = &DD7F[5*numberOfLBnodesFine];
+ D7F.f[6] = &DD7F[6*numberOfLBnodesFine];
Distributions7 D7C;
if (isEvenTimestep==true)
{
- D7C.f[0] = &DD7C[0*size_MatC];
- D7C.f[1] = &DD7C[1*size_MatC];
- D7C.f[2] = &DD7C[2*size_MatC];
- D7C.f[3] = &DD7C[3*size_MatC];
- D7C.f[4] = &DD7C[4*size_MatC];
- D7C.f[5] = &DD7C[5*size_MatC];
- D7C.f[6] = &DD7C[6*size_MatC];
+ D7C.f[0] = &DD7C[0*numberOfLBnodesCoarse];
+ D7C.f[1] = &DD7C[1*numberOfLBnodesCoarse];
+ D7C.f[2] = &DD7C[2*numberOfLBnodesCoarse];
+ D7C.f[3] = &DD7C[3*numberOfLBnodesCoarse];
+ D7C.f[4] = &DD7C[4*numberOfLBnodesCoarse];
+ D7C.f[5] = &DD7C[5*numberOfLBnodesCoarse];
+ D7C.f[6] = &DD7C[6*numberOfLBnodesCoarse];
}
else
{
- D7C.f[0] = &DD7C[0*size_MatC];
- D7C.f[2] = &DD7C[1*size_MatC];
- D7C.f[1] = &DD7C[2*size_MatC];
- D7C.f[4] = &DD7C[3*size_MatC];
- D7C.f[3] = &DD7C[4*size_MatC];
- D7C.f[6] = &DD7C[5*size_MatC];
- D7C.f[5] = &DD7C[6*size_MatC];
+ D7C.f[0] = &DD7C[0*numberOfLBnodesCoarse];
+ D7C.f[2] = &DD7C[1*numberOfLBnodesCoarse];
+ D7C.f[1] = &DD7C[2*numberOfLBnodesCoarse];
+ D7C.f[4] = &DD7C[3*numberOfLBnodesCoarse];
+ D7C.f[3] = &DD7C[4*numberOfLBnodesCoarse];
+ D7C.f[6] = &DD7C[5*numberOfLBnodesCoarse];
+ D7C.f[5] = &DD7C[6*numberOfLBnodesCoarse];
}
////////////////////////////////////////////////////////////////////////////////
@@ -21691,8 +20881,8 @@ __global__ void scaleFCThS27( real* DC,
unsigned int* neighborFX,
unsigned int* neighborFY,
unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
+ unsigned long long numberOfLBnodesCoarse,
+ unsigned long long numberOfLBnodesFine,
bool isEvenTimestep,
unsigned int* posC,
unsigned int* posFSWB,
@@ -21704,187 +20894,184 @@ __global__ void scaleFCThS27( real* DC,
real *feF, *fwF, *fnF, *fsF, *ftF, *fbF, *fneF, *fswF, *fseF, *fnwF, *fteF, *fbwF, *fbeF, *ftwF, *ftnF, *fbsF, *fbnF, *ftsF, //*fzeroF,
*ftneF, *ftswF, *ftseF, *ftnwF, *fbneF, *fbswF, *fbseF, *fbnwF;
- feF = &DF[DIR_P00 *size_MatF];
- fwF = &DF[DIR_M00 *size_MatF];
- fnF = &DF[DIR_0P0 *size_MatF];
- fsF = &DF[DIR_0M0 *size_MatF];
- ftF = &DF[DIR_00P *size_MatF];
- fbF = &DF[DIR_00M *size_MatF];
- fneF = &DF[DIR_PP0 *size_MatF];
- fswF = &DF[DIR_MM0 *size_MatF];
- fseF = &DF[DIR_PM0 *size_MatF];
- fnwF = &DF[DIR_MP0 *size_MatF];
- fteF = &DF[DIR_P0P *size_MatF];
- fbwF = &DF[DIR_M0M *size_MatF];
- fbeF = &DF[DIR_P0M *size_MatF];
- ftwF = &DF[DIR_M0P *size_MatF];
- ftnF = &DF[DIR_0PP *size_MatF];
- fbsF = &DF[DIR_0MM *size_MatF];
- fbnF = &DF[DIR_0PM *size_MatF];
- ftsF = &DF[DIR_0MP *size_MatF];
- //fzeroF = &DF[DIR_000*size_MatF];
- ftneF = &DF[DIR_PPP *size_MatF];
- ftswF = &DF[DIR_MMP *size_MatF];
- ftseF = &DF[DIR_PMP *size_MatF];
- ftnwF = &DF[DIR_MPP *size_MatF];
- fbneF = &DF[DIR_PPM *size_MatF];
- fbswF = &DF[DIR_MMM *size_MatF];
- fbseF = &DF[DIR_PMM *size_MatF];
- fbnwF = &DF[DIR_MPM *size_MatF];
+ feF = &DF[DIR_P00 * numberOfLBnodesFine];
+ fwF = &DF[DIR_M00 * numberOfLBnodesFine];
+ fnF = &DF[DIR_0P0 * numberOfLBnodesFine];
+ fsF = &DF[DIR_0M0 * numberOfLBnodesFine];
+ ftF = &DF[DIR_00P * numberOfLBnodesFine];
+ fbF = &DF[DIR_00M * numberOfLBnodesFine];
+ fneF = &DF[DIR_PP0 * numberOfLBnodesFine];
+ fswF = &DF[DIR_MM0 * numberOfLBnodesFine];
+ fseF = &DF[DIR_PM0 * numberOfLBnodesFine];
+ fnwF = &DF[DIR_MP0 * numberOfLBnodesFine];
+ fteF = &DF[DIR_P0P * numberOfLBnodesFine];
+ fbwF = &DF[DIR_M0M * numberOfLBnodesFine];
+ fbeF = &DF[DIR_P0M * numberOfLBnodesFine];
+ ftwF = &DF[DIR_M0P * numberOfLBnodesFine];
+ ftnF = &DF[DIR_0PP * numberOfLBnodesFine];
+ fbsF = &DF[DIR_0MM * numberOfLBnodesFine];
+ fbnF = &DF[DIR_0PM * numberOfLBnodesFine];
+ ftsF = &DF[DIR_0MP * numberOfLBnodesFine];
+ ftneF = &DF[DIR_PPP * numberOfLBnodesFine];
+ ftswF = &DF[DIR_MMP * numberOfLBnodesFine];
+ ftseF = &DF[DIR_PMP * numberOfLBnodesFine];
+ ftnwF = &DF[DIR_MPP * numberOfLBnodesFine];
+ fbneF = &DF[DIR_PPM * numberOfLBnodesFine];
+ fbswF = &DF[DIR_MMM * numberOfLBnodesFine];
+ fbseF = &DF[DIR_PMM * numberOfLBnodesFine];
+ fbnwF = &DF[DIR_MPM * numberOfLBnodesFine];
real *feC, *fwC, *fnC, *fsC, *ftC, *fbC, *fneC, *fswC, *fseC, *fnwC, *fteC, *fbwC, *fbeC, *ftwC, *ftnC, *fbsC, *fbnC, *ftsC, //*fzeroC,
*ftneC, *ftswC, *ftseC, *ftnwC, *fbneC, *fbswC, *fbseC, *fbnwC;
if (isEvenTimestep==true)
{
- feC = &DC[DIR_P00 *size_MatC];
- fwC = &DC[DIR_M00 *size_MatC];
- fnC = &DC[DIR_0P0 *size_MatC];
- fsC = &DC[DIR_0M0 *size_MatC];
- ftC = &DC[DIR_00P *size_MatC];
- fbC = &DC[DIR_00M *size_MatC];
- fneC = &DC[DIR_PP0 *size_MatC];
- fswC = &DC[DIR_MM0 *size_MatC];
- fseC = &DC[DIR_PM0 *size_MatC];
- fnwC = &DC[DIR_MP0 *size_MatC];
- fteC = &DC[DIR_P0P *size_MatC];
- fbwC = &DC[DIR_M0M *size_MatC];
- fbeC = &DC[DIR_P0M *size_MatC];
- ftwC = &DC[DIR_M0P *size_MatC];
- ftnC = &DC[DIR_0PP *size_MatC];
- fbsC = &DC[DIR_0MM *size_MatC];
- fbnC = &DC[DIR_0PM *size_MatC];
- ftsC = &DC[DIR_0MP *size_MatC];
- //fzeroC = &DC[DIR_000*size_MatC];
- ftneC = &DC[DIR_PPP *size_MatC];
- ftswC = &DC[DIR_MMP *size_MatC];
- ftseC = &DC[DIR_PMP *size_MatC];
- ftnwC = &DC[DIR_MPP *size_MatC];
- fbneC = &DC[DIR_PPM *size_MatC];
- fbswC = &DC[DIR_MMM *size_MatC];
- fbseC = &DC[DIR_PMM *size_MatC];
- fbnwC = &DC[DIR_MPM *size_MatC];
+ feC = &DC[DIR_P00 * numberOfLBnodesCoarse];
+ fwC = &DC[DIR_M00 * numberOfLBnodesCoarse];
+ fnC = &DC[DIR_0P0 * numberOfLBnodesCoarse];
+ fsC = &DC[DIR_0M0 * numberOfLBnodesCoarse];
+ ftC = &DC[DIR_00P * numberOfLBnodesCoarse];
+ fbC = &DC[DIR_00M * numberOfLBnodesCoarse];
+ fneC = &DC[DIR_PP0 * numberOfLBnodesCoarse];
+ fswC = &DC[DIR_MM0 * numberOfLBnodesCoarse];
+ fseC = &DC[DIR_PM0 * numberOfLBnodesCoarse];
+ fnwC = &DC[DIR_MP0 * numberOfLBnodesCoarse];
+ fteC = &DC[DIR_P0P * numberOfLBnodesCoarse];
+ fbwC = &DC[DIR_M0M * numberOfLBnodesCoarse];
+ fbeC = &DC[DIR_P0M * numberOfLBnodesCoarse];
+ ftwC = &DC[DIR_M0P * numberOfLBnodesCoarse];
+ ftnC = &DC[DIR_0PP * numberOfLBnodesCoarse];
+ fbsC = &DC[DIR_0MM * numberOfLBnodesCoarse];
+ fbnC = &DC[DIR_0PM * numberOfLBnodesCoarse];
+ ftsC = &DC[DIR_0MP * numberOfLBnodesCoarse];
+ ftneC = &DC[DIR_PPP * numberOfLBnodesCoarse];
+ ftswC = &DC[DIR_MMP * numberOfLBnodesCoarse];
+ ftseC = &DC[DIR_PMP * numberOfLBnodesCoarse];
+ ftnwC = &DC[DIR_MPP * numberOfLBnodesCoarse];
+ fbneC = &DC[DIR_PPM * numberOfLBnodesCoarse];
+ fbswC = &DC[DIR_MMM * numberOfLBnodesCoarse];
+ fbseC = &DC[DIR_PMM * numberOfLBnodesCoarse];
+ fbnwC = &DC[DIR_MPM * numberOfLBnodesCoarse];
}
else
{
- fwC = &DC[DIR_P00 *size_MatC];
- feC = &DC[DIR_M00 *size_MatC];
- fsC = &DC[DIR_0P0 *size_MatC];
- fnC = &DC[DIR_0M0 *size_MatC];
- fbC = &DC[DIR_00P *size_MatC];
- ftC = &DC[DIR_00M *size_MatC];
- fswC = &DC[DIR_PP0 *size_MatC];
- fneC = &DC[DIR_MM0 *size_MatC];
- fnwC = &DC[DIR_PM0 *size_MatC];
- fseC = &DC[DIR_MP0 *size_MatC];
- fbwC = &DC[DIR_P0P *size_MatC];
- fteC = &DC[DIR_M0M *size_MatC];
- ftwC = &DC[DIR_P0M *size_MatC];
- fbeC = &DC[DIR_M0P *size_MatC];
- fbsC = &DC[DIR_0PP *size_MatC];
- ftnC = &DC[DIR_0MM *size_MatC];
- ftsC = &DC[DIR_0PM *size_MatC];
- fbnC = &DC[DIR_0MP *size_MatC];
- //fzeroC = &DC[DIR_000*size_MatC];
- fbswC = &DC[DIR_PPP *size_MatC];
- fbneC = &DC[DIR_MMP *size_MatC];
- fbnwC = &DC[DIR_PMP *size_MatC];
- fbseC = &DC[DIR_MPP *size_MatC];
- ftswC = &DC[DIR_PPM *size_MatC];
- ftneC = &DC[DIR_MMM *size_MatC];
- ftnwC = &DC[DIR_PMM *size_MatC];
- ftseC = &DC[DIR_MPM *size_MatC];
+ fwC = &DC[DIR_P00 * numberOfLBnodesCoarse];
+ feC = &DC[DIR_M00 * numberOfLBnodesCoarse];
+ fsC = &DC[DIR_0P0 * numberOfLBnodesCoarse];
+ fnC = &DC[DIR_0M0 * numberOfLBnodesCoarse];
+ fbC = &DC[DIR_00P * numberOfLBnodesCoarse];
+ ftC = &DC[DIR_00M * numberOfLBnodesCoarse];
+ fswC = &DC[DIR_PP0 * numberOfLBnodesCoarse];
+ fneC = &DC[DIR_MM0 * numberOfLBnodesCoarse];
+ fnwC = &DC[DIR_PM0 * numberOfLBnodesCoarse];
+ fseC = &DC[DIR_MP0 * numberOfLBnodesCoarse];
+ fbwC = &DC[DIR_P0P * numberOfLBnodesCoarse];
+ fteC = &DC[DIR_M0M * numberOfLBnodesCoarse];
+ ftwC = &DC[DIR_P0M * numberOfLBnodesCoarse];
+ fbeC = &DC[DIR_M0P * numberOfLBnodesCoarse];
+ fbsC = &DC[DIR_0PP * numberOfLBnodesCoarse];
+ ftnC = &DC[DIR_0MM * numberOfLBnodesCoarse];
+ ftsC = &DC[DIR_0PM * numberOfLBnodesCoarse];
+ fbnC = &DC[DIR_0MP * numberOfLBnodesCoarse];
+ fbswC = &DC[DIR_PPP * numberOfLBnodesCoarse];
+ fbneC = &DC[DIR_MMP * numberOfLBnodesCoarse];
+ fbnwC = &DC[DIR_PMP * numberOfLBnodesCoarse];
+ fbseC = &DC[DIR_MPP * numberOfLBnodesCoarse];
+ ftswC = &DC[DIR_PPM * numberOfLBnodesCoarse];
+ ftneC = &DC[DIR_MMM * numberOfLBnodesCoarse];
+ ftnwC = &DC[DIR_PMM * numberOfLBnodesCoarse];
+ ftseC = &DC[DIR_MPM * numberOfLBnodesCoarse];
}
Distributions27 D27F;
- D27F.f[DIR_P00 ] = &DD27F[DIR_P00 *size_MatF];
- D27F.f[DIR_M00 ] = &DD27F[DIR_M00 *size_MatF];
- D27F.f[DIR_0P0 ] = &DD27F[DIR_0P0 *size_MatF];
- D27F.f[DIR_0M0 ] = &DD27F[DIR_0M0 *size_MatF];
- D27F.f[DIR_00P ] = &DD27F[DIR_00P *size_MatF];
- D27F.f[DIR_00M ] = &DD27F[DIR_00M *size_MatF];
- D27F.f[DIR_PP0 ] = &DD27F[DIR_PP0 *size_MatF];
- D27F.f[DIR_MM0 ] = &DD27F[DIR_MM0 *size_MatF];
- D27F.f[DIR_PM0 ] = &DD27F[DIR_PM0 *size_MatF];
- D27F.f[DIR_MP0 ] = &DD27F[DIR_MP0 *size_MatF];
- D27F.f[DIR_P0P ] = &DD27F[DIR_P0P *size_MatF];
- D27F.f[DIR_M0M ] = &DD27F[DIR_M0M *size_MatF];
- D27F.f[DIR_P0M ] = &DD27F[DIR_P0M *size_MatF];
- D27F.f[DIR_M0P ] = &DD27F[DIR_M0P *size_MatF];
- D27F.f[DIR_0PP ] = &DD27F[DIR_0PP *size_MatF];
- D27F.f[DIR_0MM ] = &DD27F[DIR_0MM *size_MatF];
- D27F.f[DIR_0PM ] = &DD27F[DIR_0PM *size_MatF];
- D27F.f[DIR_0MP ] = &DD27F[DIR_0MP *size_MatF];
- D27F.f[DIR_000] = &DD27F[DIR_000*size_MatF];
- D27F.f[DIR_PPP ] = &DD27F[DIR_PPP *size_MatF];
- D27F.f[DIR_MMP ] = &DD27F[DIR_MMP *size_MatF];
- D27F.f[DIR_PMP ] = &DD27F[DIR_PMP *size_MatF];
- D27F.f[DIR_MPP ] = &DD27F[DIR_MPP *size_MatF];
- D27F.f[DIR_PPM ] = &DD27F[DIR_PPM *size_MatF];
- D27F.f[DIR_MMM ] = &DD27F[DIR_MMM *size_MatF];
- D27F.f[DIR_PMM ] = &DD27F[DIR_PMM *size_MatF];
- D27F.f[DIR_MPM ] = &DD27F[DIR_MPM *size_MatF];
+ D27F.f[DIR_P00] = &DD27F[DIR_P00 * numberOfLBnodesFine];
+ D27F.f[DIR_M00] = &DD27F[DIR_M00 * numberOfLBnodesFine];
+ D27F.f[DIR_0P0] = &DD27F[DIR_0P0 * numberOfLBnodesFine];
+ D27F.f[DIR_0M0] = &DD27F[DIR_0M0 * numberOfLBnodesFine];
+ D27F.f[DIR_00P] = &DD27F[DIR_00P * numberOfLBnodesFine];
+ D27F.f[DIR_00M] = &DD27F[DIR_00M * numberOfLBnodesFine];
+ D27F.f[DIR_PP0] = &DD27F[DIR_PP0 * numberOfLBnodesFine];
+ D27F.f[DIR_MM0] = &DD27F[DIR_MM0 * numberOfLBnodesFine];
+ D27F.f[DIR_PM0] = &DD27F[DIR_PM0 * numberOfLBnodesFine];
+ D27F.f[DIR_MP0] = &DD27F[DIR_MP0 * numberOfLBnodesFine];
+ D27F.f[DIR_P0P] = &DD27F[DIR_P0P * numberOfLBnodesFine];
+ D27F.f[DIR_M0M] = &DD27F[DIR_M0M * numberOfLBnodesFine];
+ D27F.f[DIR_P0M] = &DD27F[DIR_P0M * numberOfLBnodesFine];
+ D27F.f[DIR_M0P] = &DD27F[DIR_M0P * numberOfLBnodesFine];
+ D27F.f[DIR_0PP] = &DD27F[DIR_0PP * numberOfLBnodesFine];
+ D27F.f[DIR_0MM] = &DD27F[DIR_0MM * numberOfLBnodesFine];
+ D27F.f[DIR_0PM] = &DD27F[DIR_0PM * numberOfLBnodesFine];
+ D27F.f[DIR_0MP] = &DD27F[DIR_0MP * numberOfLBnodesFine];
+ D27F.f[DIR_000] = &DD27F[DIR_000 * numberOfLBnodesFine];
+ D27F.f[DIR_PPP] = &DD27F[DIR_PPP * numberOfLBnodesFine];
+ D27F.f[DIR_MMP] = &DD27F[DIR_MMP * numberOfLBnodesFine];
+ D27F.f[DIR_PMP] = &DD27F[DIR_PMP * numberOfLBnodesFine];
+ D27F.f[DIR_MPP] = &DD27F[DIR_MPP * numberOfLBnodesFine];
+ D27F.f[DIR_PPM] = &DD27F[DIR_PPM * numberOfLBnodesFine];
+ D27F.f[DIR_MMM] = &DD27F[DIR_MMM * numberOfLBnodesFine];
+ D27F.f[DIR_PMM] = &DD27F[DIR_PMM * numberOfLBnodesFine];
+ D27F.f[DIR_MPM] = &DD27F[DIR_MPM * numberOfLBnodesFine];
Distributions27 D27C;
if (isEvenTimestep==true)
{
- D27C.f[DIR_P00 ] = &DD27C[DIR_P00 *size_MatC];
- D27C.f[DIR_M00 ] = &DD27C[DIR_M00 *size_MatC];
- D27C.f[DIR_0P0 ] = &DD27C[DIR_0P0 *size_MatC];
- D27C.f[DIR_0M0 ] = &DD27C[DIR_0M0 *size_MatC];
- D27C.f[DIR_00P ] = &DD27C[DIR_00P *size_MatC];
- D27C.f[DIR_00M ] = &DD27C[DIR_00M *size_MatC];
- D27C.f[DIR_PP0 ] = &DD27C[DIR_PP0 *size_MatC];
- D27C.f[DIR_MM0 ] = &DD27C[DIR_MM0 *size_MatC];
- D27C.f[DIR_PM0 ] = &DD27C[DIR_PM0 *size_MatC];
- D27C.f[DIR_MP0 ] = &DD27C[DIR_MP0 *size_MatC];
- D27C.f[DIR_P0P ] = &DD27C[DIR_P0P *size_MatC];
- D27C.f[DIR_M0M ] = &DD27C[DIR_M0M *size_MatC];
- D27C.f[DIR_P0M ] = &DD27C[DIR_P0M *size_MatC];
- D27C.f[DIR_M0P ] = &DD27C[DIR_M0P *size_MatC];
- D27C.f[DIR_0PP ] = &DD27C[DIR_0PP *size_MatC];
- D27C.f[DIR_0MM ] = &DD27C[DIR_0MM *size_MatC];
- D27C.f[DIR_0PM ] = &DD27C[DIR_0PM *size_MatC];
- D27C.f[DIR_0MP ] = &DD27C[DIR_0MP *size_MatC];
- D27C.f[DIR_000] = &DD27C[DIR_000*size_MatC];
- D27C.f[DIR_PPP ] = &DD27C[DIR_PPP *size_MatC];
- D27C.f[DIR_MMP ] = &DD27C[DIR_MMP *size_MatC];
- D27C.f[DIR_PMP ] = &DD27C[DIR_PMP *size_MatC];
- D27C.f[DIR_MPP ] = &DD27C[DIR_MPP *size_MatC];
- D27C.f[DIR_PPM ] = &DD27C[DIR_PPM *size_MatC];
- D27C.f[DIR_MMM ] = &DD27C[DIR_MMM *size_MatC];
- D27C.f[DIR_PMM ] = &DD27C[DIR_PMM *size_MatC];
- D27C.f[DIR_MPM ] = &DD27C[DIR_MPM *size_MatC];
+ D27C.f[DIR_P00] = &DD27C[DIR_P00 * numberOfLBnodesCoarse];
+ D27C.f[DIR_M00] = &DD27C[DIR_M00 * numberOfLBnodesCoarse];
+ D27C.f[DIR_0P0] = &DD27C[DIR_0P0 * numberOfLBnodesCoarse];
+ D27C.f[DIR_0M0] = &DD27C[DIR_0M0 * numberOfLBnodesCoarse];
+ D27C.f[DIR_00P] = &DD27C[DIR_00P * numberOfLBnodesCoarse];
+ D27C.f[DIR_00M] = &DD27C[DIR_00M * numberOfLBnodesCoarse];
+ D27C.f[DIR_PP0] = &DD27C[DIR_PP0 * numberOfLBnodesCoarse];
+ D27C.f[DIR_MM0] = &DD27C[DIR_MM0 * numberOfLBnodesCoarse];
+ D27C.f[DIR_PM0] = &DD27C[DIR_PM0 * numberOfLBnodesCoarse];
+ D27C.f[DIR_MP0] = &DD27C[DIR_MP0 * numberOfLBnodesCoarse];
+ D27C.f[DIR_P0P] = &DD27C[DIR_P0P * numberOfLBnodesCoarse];
+ D27C.f[DIR_M0M] = &DD27C[DIR_M0M * numberOfLBnodesCoarse];
+ D27C.f[DIR_P0M] = &DD27C[DIR_P0M * numberOfLBnodesCoarse];
+ D27C.f[DIR_M0P] = &DD27C[DIR_M0P * numberOfLBnodesCoarse];
+ D27C.f[DIR_0PP] = &DD27C[DIR_0PP * numberOfLBnodesCoarse];
+ D27C.f[DIR_0MM] = &DD27C[DIR_0MM * numberOfLBnodesCoarse];
+ D27C.f[DIR_0PM] = &DD27C[DIR_0PM * numberOfLBnodesCoarse];
+ D27C.f[DIR_0MP] = &DD27C[DIR_0MP * numberOfLBnodesCoarse];
+ D27C.f[DIR_000] = &DD27C[DIR_000 * numberOfLBnodesCoarse];
+ D27C.f[DIR_PPP] = &DD27C[DIR_PPP * numberOfLBnodesCoarse];
+ D27C.f[DIR_MMP] = &DD27C[DIR_MMP * numberOfLBnodesCoarse];
+ D27C.f[DIR_PMP] = &DD27C[DIR_PMP * numberOfLBnodesCoarse];
+ D27C.f[DIR_MPP] = &DD27C[DIR_MPP * numberOfLBnodesCoarse];
+ D27C.f[DIR_PPM] = &DD27C[DIR_PPM * numberOfLBnodesCoarse];
+ D27C.f[DIR_MMM] = &DD27C[DIR_MMM * numberOfLBnodesCoarse];
+ D27C.f[DIR_PMM] = &DD27C[DIR_PMM * numberOfLBnodesCoarse];
+ D27C.f[DIR_MPM] = &DD27C[DIR_MPM * numberOfLBnodesCoarse];
}
else
{
- D27C.f[DIR_M00 ] = &DD27C[DIR_P00 *size_MatC];
- D27C.f[DIR_P00 ] = &DD27C[DIR_M00 *size_MatC];
- D27C.f[DIR_0M0 ] = &DD27C[DIR_0P0 *size_MatC];
- D27C.f[DIR_0P0 ] = &DD27C[DIR_0M0 *size_MatC];
- D27C.f[DIR_00M ] = &DD27C[DIR_00P *size_MatC];
- D27C.f[DIR_00P ] = &DD27C[DIR_00M *size_MatC];
- D27C.f[DIR_MM0 ] = &DD27C[DIR_PP0 *size_MatC];
- D27C.f[DIR_PP0 ] = &DD27C[DIR_MM0 *size_MatC];
- D27C.f[DIR_MP0 ] = &DD27C[DIR_PM0 *size_MatC];
- D27C.f[DIR_PM0 ] = &DD27C[DIR_MP0 *size_MatC];
- D27C.f[DIR_M0M ] = &DD27C[DIR_P0P *size_MatC];
- D27C.f[DIR_P0P ] = &DD27C[DIR_M0M *size_MatC];
- D27C.f[DIR_M0P ] = &DD27C[DIR_P0M *size_MatC];
- D27C.f[DIR_P0M ] = &DD27C[DIR_M0P *size_MatC];
- D27C.f[DIR_0MM ] = &DD27C[DIR_0PP *size_MatC];
- D27C.f[DIR_0PP ] = &DD27C[DIR_0MM *size_MatC];
- D27C.f[DIR_0MP ] = &DD27C[DIR_0PM *size_MatC];
- D27C.f[DIR_0PM ] = &DD27C[DIR_0MP *size_MatC];
- D27C.f[DIR_000] = &DD27C[DIR_000*size_MatC];
- D27C.f[DIR_MMM ] = &DD27C[DIR_PPP *size_MatC];
- D27C.f[DIR_PPM ] = &DD27C[DIR_MMP *size_MatC];
- D27C.f[DIR_MPM ] = &DD27C[DIR_PMP *size_MatC];
- D27C.f[DIR_PMM ] = &DD27C[DIR_MPP *size_MatC];
- D27C.f[DIR_MMP ] = &DD27C[DIR_PPM *size_MatC];
- D27C.f[DIR_PPP ] = &DD27C[DIR_MMM *size_MatC];
- D27C.f[DIR_MPP ] = &DD27C[DIR_PMM *size_MatC];
- D27C.f[DIR_PMP ] = &DD27C[DIR_MPM *size_MatC];
+ D27C.f[DIR_M00] = &DD27C[DIR_P00 * numberOfLBnodesCoarse];
+ D27C.f[DIR_P00] = &DD27C[DIR_M00 * numberOfLBnodesCoarse];
+ D27C.f[DIR_0M0] = &DD27C[DIR_0P0 * numberOfLBnodesCoarse];
+ D27C.f[DIR_0P0] = &DD27C[DIR_0M0 * numberOfLBnodesCoarse];
+ D27C.f[DIR_00M] = &DD27C[DIR_00P * numberOfLBnodesCoarse];
+ D27C.f[DIR_00P] = &DD27C[DIR_00M * numberOfLBnodesCoarse];
+ D27C.f[DIR_MM0] = &DD27C[DIR_PP0 * numberOfLBnodesCoarse];
+ D27C.f[DIR_PP0] = &DD27C[DIR_MM0 * numberOfLBnodesCoarse];
+ D27C.f[DIR_MP0] = &DD27C[DIR_PM0 * numberOfLBnodesCoarse];
+ D27C.f[DIR_PM0] = &DD27C[DIR_MP0 * numberOfLBnodesCoarse];
+ D27C.f[DIR_M0M] = &DD27C[DIR_P0P * numberOfLBnodesCoarse];
+ D27C.f[DIR_P0P] = &DD27C[DIR_M0M * numberOfLBnodesCoarse];
+ D27C.f[DIR_M0P] = &DD27C[DIR_P0M * numberOfLBnodesCoarse];
+ D27C.f[DIR_P0M] = &DD27C[DIR_M0P * numberOfLBnodesCoarse];
+ D27C.f[DIR_0MM] = &DD27C[DIR_0PP * numberOfLBnodesCoarse];
+ D27C.f[DIR_0PP] = &DD27C[DIR_0MM * numberOfLBnodesCoarse];
+ D27C.f[DIR_0MP] = &DD27C[DIR_0PM * numberOfLBnodesCoarse];
+ D27C.f[DIR_0PM] = &DD27C[DIR_0MP * numberOfLBnodesCoarse];
+ D27C.f[DIR_000] = &DD27C[DIR_000 * numberOfLBnodesCoarse];
+ D27C.f[DIR_MMM] = &DD27C[DIR_PPP * numberOfLBnodesCoarse];
+ D27C.f[DIR_PPM] = &DD27C[DIR_MMP * numberOfLBnodesCoarse];
+ D27C.f[DIR_MPM] = &DD27C[DIR_PMP * numberOfLBnodesCoarse];
+ D27C.f[DIR_PMM] = &DD27C[DIR_MPP * numberOfLBnodesCoarse];
+ D27C.f[DIR_MMP] = &DD27C[DIR_PPM * numberOfLBnodesCoarse];
+ D27C.f[DIR_PPP] = &DD27C[DIR_MMM * numberOfLBnodesCoarse];
+ D27C.f[DIR_MPP] = &DD27C[DIR_PMM * numberOfLBnodesCoarse];
+ D27C.f[DIR_PMP] = &DD27C[DIR_MPM * numberOfLBnodesCoarse];
}
////////////////////////////////////////////////////////////////////////////////
@@ -21980,33 +21167,33 @@ __global__ void scaleFCThS27( real* DC,
f_BSE = fbseF[kbs];
f_BNW = fbnwF[kbw];
//////////////////////////////////////////////////////////////////////////////////
- f27E = (D27F.f[DIR_P00 ])[kzero];//ke
- f27W = (D27F.f[DIR_M00 ])[kw ];
- f27N = (D27F.f[DIR_0P0 ])[kzero];//kn
- f27S = (D27F.f[DIR_0M0 ])[ks ];
- f27T = (D27F.f[DIR_00P ])[kzero];//kt
- f27B = (D27F.f[DIR_00M ])[kb ];
- f27NE = (D27F.f[DIR_PP0 ])[kzero];//kne
- f27SW = (D27F.f[DIR_MM0 ])[ksw ];
- f27SE = (D27F.f[DIR_PM0 ])[ks ];//kse
- f27NW = (D27F.f[DIR_MP0 ])[kw ];//knw
- f27TE = (D27F.f[DIR_P0P ])[kzero];//kte
- f27BW = (D27F.f[DIR_M0M ])[kbw ];
- f27BE = (D27F.f[DIR_P0M ])[kb ];//kbe
- f27TW = (D27F.f[DIR_M0P ])[kw ];//ktw
- f27TN = (D27F.f[DIR_0PP ])[kzero];//ktn
- f27BS = (D27F.f[DIR_0MM ])[kbs ];
- f27BN = (D27F.f[DIR_0PM ])[kb ];//kbn
- f27TS = (D27F.f[DIR_0MP ])[ks ];//kts
+ f27E = (D27F.f[DIR_P00])[kzero];//ke
+ f27W = (D27F.f[DIR_M00])[kw ];
+ f27N = (D27F.f[DIR_0P0])[kzero];//kn
+ f27S = (D27F.f[DIR_0M0])[ks ];
+ f27T = (D27F.f[DIR_00P])[kzero];//kt
+ f27B = (D27F.f[DIR_00M])[kb ];
+ f27NE = (D27F.f[DIR_PP0])[kzero];//kne
+ f27SW = (D27F.f[DIR_MM0])[ksw ];
+ f27SE = (D27F.f[DIR_PM0])[ks ];//kse
+ f27NW = (D27F.f[DIR_MP0])[kw ];//knw
+ f27TE = (D27F.f[DIR_P0P])[kzero];//kte
+ f27BW = (D27F.f[DIR_M0M])[kbw ];
+ f27BE = (D27F.f[DIR_P0M])[kb ];//kbe
+ f27TW = (D27F.f[DIR_M0P])[kw ];//ktw
+ f27TN = (D27F.f[DIR_0PP])[kzero];//ktn
+ f27BS = (D27F.f[DIR_0MM])[kbs ];
+ f27BN = (D27F.f[DIR_0PM])[kb ];//kbn
+ f27TS = (D27F.f[DIR_0MP])[ks ];//kts
f27ZERO = (D27F.f[DIR_000])[kzero];//kzero
- f27TNE = (D27F.f[DIR_PPP ])[kzero];//ktne
- f27TSW = (D27F.f[DIR_MMP ])[ksw ];//ktsw
- f27TSE = (D27F.f[DIR_PMP ])[ks ];//ktse
- f27TNW = (D27F.f[DIR_MPP ])[kw ];//ktnw
- f27BNE = (D27F.f[DIR_PPM ])[kb ];//kbne
- f27BSW = (D27F.f[DIR_MMM ])[kbsw ];
- f27BSE = (D27F.f[DIR_PMM ])[kbs ];//kbse
- f27BNW = (D27F.f[DIR_MPM ])[kbw ];//kbnw
+ f27TNE = (D27F.f[DIR_PPP])[kzero];//ktne
+ f27TSW = (D27F.f[DIR_MMP])[ksw ];//ktsw
+ f27TSE = (D27F.f[DIR_PMP])[ks ];//ktse
+ f27TNW = (D27F.f[DIR_MPP])[kw ];//ktnw
+ f27BNE = (D27F.f[DIR_PPM])[kb ];//kbne
+ f27BSW = (D27F.f[DIR_MMM])[kbsw ];
+ f27BSE = (D27F.f[DIR_PMM])[kbs ];//kbse
+ f27BNW = (D27F.f[DIR_MPM])[kbw ];//kbnw
Conc_F_SWB = f27E + f27W + f27N + f27S + f27T + f27B + f27NE + f27SW + f27SE + f27NW +
f27TE + f27BW + f27BE + f27TW + f27TN + f27BS + f27BN + f27TS + f27ZERO +
@@ -22067,33 +21254,33 @@ __global__ void scaleFCThS27( real* DC,
f_BSE = fbseF[kbs];
f_BNW = fbnwF[kbw];
//////////////////////////////////////////////////////////////////////////////////
- f27E = (D27F.f[DIR_P00 ])[kzero];//ke
- f27W = (D27F.f[DIR_M00 ])[kw ];
- f27N = (D27F.f[DIR_0P0 ])[kzero];//kn
- f27S = (D27F.f[DIR_0M0 ])[ks ];
- f27T = (D27F.f[DIR_00P ])[kzero];//kt
- f27B = (D27F.f[DIR_00M ])[kb ];
- f27NE = (D27F.f[DIR_PP0 ])[kzero];//kne
- f27SW = (D27F.f[DIR_MM0 ])[ksw ];
- f27SE = (D27F.f[DIR_PM0 ])[ks ];//kse
- f27NW = (D27F.f[DIR_MP0 ])[kw ];//knw
- f27TE = (D27F.f[DIR_P0P ])[kzero];//kte
- f27BW = (D27F.f[DIR_M0M ])[kbw ];
- f27BE = (D27F.f[DIR_P0M ])[kb ];//kbe
- f27TW = (D27F.f[DIR_M0P ])[kw ];//ktw
- f27TN = (D27F.f[DIR_0PP ])[kzero];//ktn
- f27BS = (D27F.f[DIR_0MM ])[kbs ];
- f27BN = (D27F.f[DIR_0PM ])[kb ];//kbn
- f27TS = (D27F.f[DIR_0MP ])[ks ];//kts
+ f27E = (D27F.f[DIR_P00])[kzero];//ke
+ f27W = (D27F.f[DIR_M00])[kw ];
+ f27N = (D27F.f[DIR_0P0])[kzero];//kn
+ f27S = (D27F.f[DIR_0M0])[ks ];
+ f27T = (D27F.f[DIR_00P])[kzero];//kt
+ f27B = (D27F.f[DIR_00M])[kb ];
+ f27NE = (D27F.f[DIR_PP0])[kzero];//kne
+ f27SW = (D27F.f[DIR_MM0])[ksw ];
+ f27SE = (D27F.f[DIR_PM0])[ks ];//kse
+ f27NW = (D27F.f[DIR_MP0])[kw ];//knw
+ f27TE = (D27F.f[DIR_P0P])[kzero];//kte
+ f27BW = (D27F.f[DIR_M0M])[kbw ];
+ f27BE = (D27F.f[DIR_P0M])[kb ];//kbe
+ f27TW = (D27F.f[DIR_M0P])[kw ];//ktw
+ f27TN = (D27F.f[DIR_0PP])[kzero];//ktn
+ f27BS = (D27F.f[DIR_0MM])[kbs ];
+ f27BN = (D27F.f[DIR_0PM])[kb ];//kbn
+ f27TS = (D27F.f[DIR_0MP])[ks ];//kts
f27ZERO = (D27F.f[DIR_000])[kzero];//kzero
- f27TNE = (D27F.f[DIR_PPP ])[kzero];//ktne
- f27TSW = (D27F.f[DIR_MMP ])[ksw ];//ktsw
- f27TSE = (D27F.f[DIR_PMP ])[ks ];//ktse
- f27TNW = (D27F.f[DIR_MPP ])[kw ];//ktnw
- f27BNE = (D27F.f[DIR_PPM ])[kb ];//kbne
- f27BSW = (D27F.f[DIR_MMM ])[kbsw ];
- f27BSE = (D27F.f[DIR_PMM ])[kbs ];//kbse
- f27BNW = (D27F.f[DIR_MPM ])[kbw ];//kbnw
+ f27TNE = (D27F.f[DIR_PPP])[kzero];//ktne
+ f27TSW = (D27F.f[DIR_MMP])[ksw ];//ktsw
+ f27TSE = (D27F.f[DIR_PMP])[ks ];//ktse
+ f27TNW = (D27F.f[DIR_MPP])[kw ];//ktnw
+ f27BNE = (D27F.f[DIR_PPM])[kb ];//kbne
+ f27BSW = (D27F.f[DIR_MMM])[kbsw ];
+ f27BSE = (D27F.f[DIR_PMM])[kbs ];//kbse
+ f27BNW = (D27F.f[DIR_MPM])[kbw ];//kbnw
Conc_F_SWT = f27E + f27W + f27N + f27S + f27T + f27B + f27NE + f27SW + f27SE + f27NW +
f27TE + f27BW + f27BE + f27TW + f27TN + f27BS + f27BN + f27TS + f27ZERO +
@@ -22154,33 +21341,33 @@ __global__ void scaleFCThS27( real* DC,
f_BSE = fbseF[kbs];
f_BNW = fbnwF[kbw];
//////////////////////////////////////////////////////////////////////////////////
- f27E = (D27F.f[DIR_P00 ])[kzero];//ke
- f27W = (D27F.f[DIR_M00 ])[kw ];
- f27N = (D27F.f[DIR_0P0 ])[kzero];//kn
- f27S = (D27F.f[DIR_0M0 ])[ks ];
- f27T = (D27F.f[DIR_00P ])[kzero];//kt
- f27B = (D27F.f[DIR_00M ])[kb ];
- f27NE = (D27F.f[DIR_PP0 ])[kzero];//kne
- f27SW = (D27F.f[DIR_MM0 ])[ksw ];
- f27SE = (D27F.f[DIR_PM0 ])[ks ];//kse
- f27NW = (D27F.f[DIR_MP0 ])[kw ];//knw
- f27TE = (D27F.f[DIR_P0P ])[kzero];//kte
- f27BW = (D27F.f[DIR_M0M ])[kbw ];
- f27BE = (D27F.f[DIR_P0M ])[kb ];//kbe
- f27TW = (D27F.f[DIR_M0P ])[kw ];//ktw
- f27TN = (D27F.f[DIR_0PP ])[kzero];//ktn
- f27BS = (D27F.f[DIR_0MM ])[kbs ];
- f27BN = (D27F.f[DIR_0PM ])[kb ];//kbn
- f27TS = (D27F.f[DIR_0MP ])[ks ];//kts
+ f27E = (D27F.f[DIR_P00])[kzero];//ke
+ f27W = (D27F.f[DIR_M00])[kw ];
+ f27N = (D27F.f[DIR_0P0])[kzero];//kn
+ f27S = (D27F.f[DIR_0M0])[ks ];
+ f27T = (D27F.f[DIR_00P])[kzero];//kt
+ f27B = (D27F.f[DIR_00M])[kb ];
+ f27NE = (D27F.f[DIR_PP0])[kzero];//kne
+ f27SW = (D27F.f[DIR_MM0])[ksw ];
+ f27SE = (D27F.f[DIR_PM0])[ks ];//kse
+ f27NW = (D27F.f[DIR_MP0])[kw ];//knw
+ f27TE = (D27F.f[DIR_P0P])[kzero];//kte
+ f27BW = (D27F.f[DIR_M0M])[kbw ];
+ f27BE = (D27F.f[DIR_P0M])[kb ];//kbe
+ f27TW = (D27F.f[DIR_M0P])[kw ];//ktw
+ f27TN = (D27F.f[DIR_0PP])[kzero];//ktn
+ f27BS = (D27F.f[DIR_0MM])[kbs ];
+ f27BN = (D27F.f[DIR_0PM])[kb ];//kbn
+ f27TS = (D27F.f[DIR_0MP])[ks ];//kts
f27ZERO = (D27F.f[DIR_000])[kzero];//kzero
- f27TNE = (D27F.f[DIR_PPP ])[kzero];//ktne
- f27TSW = (D27F.f[DIR_MMP ])[ksw ];//ktsw
- f27TSE = (D27F.f[DIR_PMP ])[ks ];//ktse
- f27TNW = (D27F.f[DIR_MPP ])[kw ];//ktnw
- f27BNE = (D27F.f[DIR_PPM ])[kb ];//kbne
- f27BSW = (D27F.f[DIR_MMM ])[kbsw ];
- f27BSE = (D27F.f[DIR_PMM ])[kbs ];//kbse
- f27BNW = (D27F.f[DIR_MPM ])[kbw ];//kbnw
+ f27TNE = (D27F.f[DIR_PPP])[kzero];//ktne
+ f27TSW = (D27F.f[DIR_MMP])[ksw ];//ktsw
+ f27TSE = (D27F.f[DIR_PMP])[ks ];//ktse
+ f27TNW = (D27F.f[DIR_MPP])[kw ];//ktnw
+ f27BNE = (D27F.f[DIR_PPM])[kb ];//kbne
+ f27BSW = (D27F.f[DIR_MMM])[kbsw ];
+ f27BSE = (D27F.f[DIR_PMM])[kbs ];//kbse
+ f27BNW = (D27F.f[DIR_MPM])[kbw ];//kbnw
Conc_F_SET = f27E + f27W + f27N + f27S + f27T + f27B + f27NE + f27SW + f27SE + f27NW +
f27TE + f27BW + f27BE + f27TW + f27TN + f27BS + f27BN + f27TS + f27ZERO +
@@ -22241,33 +21428,33 @@ __global__ void scaleFCThS27( real* DC,
f_BSE = fbseF[kbs];
f_BNW = fbnwF[kbw];
//////////////////////////////////////////////////////////////////////////////////
- f27E = (D27F.f[DIR_P00 ])[kzero];//ke
- f27W = (D27F.f[DIR_M00 ])[kw ];
- f27N = (D27F.f[DIR_0P0 ])[kzero];//kn
- f27S = (D27F.f[DIR_0M0 ])[ks ];
- f27T = (D27F.f[DIR_00P ])[kzero];//kt
- f27B = (D27F.f[DIR_00M ])[kb ];
- f27NE = (D27F.f[DIR_PP0 ])[kzero];//kne
- f27SW = (D27F.f[DIR_MM0 ])[ksw ];
- f27SE = (D27F.f[DIR_PM0 ])[ks ];//kse
- f27NW = (D27F.f[DIR_MP0 ])[kw ];//knw
- f27TE = (D27F.f[DIR_P0P ])[kzero];//kte
- f27BW = (D27F.f[DIR_M0M ])[kbw ];
- f27BE = (D27F.f[DIR_P0M ])[kb ];//kbe
- f27TW = (D27F.f[DIR_M0P ])[kw ];//ktw
- f27TN = (D27F.f[DIR_0PP ])[kzero];//ktn
- f27BS = (D27F.f[DIR_0MM ])[kbs ];
- f27BN = (D27F.f[DIR_0PM ])[kb ];//kbn
- f27TS = (D27F.f[DIR_0MP ])[ks ];//kts
+ f27E = (D27F.f[DIR_P00])[kzero];//ke
+ f27W = (D27F.f[DIR_M00])[kw ];
+ f27N = (D27F.f[DIR_0P0])[kzero];//kn
+ f27S = (D27F.f[DIR_0M0])[ks ];
+ f27T = (D27F.f[DIR_00P])[kzero];//kt
+ f27B = (D27F.f[DIR_00M])[kb ];
+ f27NE = (D27F.f[DIR_PP0])[kzero];//kne
+ f27SW = (D27F.f[DIR_MM0])[ksw ];
+ f27SE = (D27F.f[DIR_PM0])[ks ];//kse
+ f27NW = (D27F.f[DIR_MP0])[kw ];//knw
+ f27TE = (D27F.f[DIR_P0P])[kzero];//kte
+ f27BW = (D27F.f[DIR_M0M])[kbw ];
+ f27BE = (D27F.f[DIR_P0M])[kb ];//kbe
+ f27TW = (D27F.f[DIR_M0P])[kw ];//ktw
+ f27TN = (D27F.f[DIR_0PP])[kzero];//ktn
+ f27BS = (D27F.f[DIR_0MM])[kbs ];
+ f27BN = (D27F.f[DIR_0PM])[kb ];//kbn
+ f27TS = (D27F.f[DIR_0MP])[ks ];//kts
f27ZERO = (D27F.f[DIR_000])[kzero];//kzero
- f27TNE = (D27F.f[DIR_PPP ])[kzero];//ktne
- f27TSW = (D27F.f[DIR_MMP ])[ksw ];//ktsw
- f27TSE = (D27F.f[DIR_PMP ])[ks ];//ktse
- f27TNW = (D27F.f[DIR_MPP ])[kw ];//ktnw
- f27BNE = (D27F.f[DIR_PPM ])[kb ];//kbne
- f27BSW = (D27F.f[DIR_MMM ])[kbsw ];
- f27BSE = (D27F.f[DIR_PMM ])[kbs ];//kbse
- f27BNW = (D27F.f[DIR_MPM ])[kbw ];//kbnw
+ f27TNE = (D27F.f[DIR_PPP])[kzero];//ktne
+ f27TSW = (D27F.f[DIR_MMP])[ksw ];//ktsw
+ f27TSE = (D27F.f[DIR_PMP])[ks ];//ktse
+ f27TNW = (D27F.f[DIR_MPP])[kw ];//ktnw
+ f27BNE = (D27F.f[DIR_PPM])[kb ];//kbne
+ f27BSW = (D27F.f[DIR_MMM])[kbsw ];
+ f27BSE = (D27F.f[DIR_PMM])[kbs ];//kbse
+ f27BNW = (D27F.f[DIR_MPM])[kbw ];//kbnw
Conc_F_SEB = f27E + f27W + f27N + f27S + f27T + f27B + f27NE + f27SW + f27SE + f27NW +
f27TE + f27BW + f27BE + f27TW + f27TN + f27BS + f27BN + f27TS + f27ZERO +
@@ -22338,33 +21525,33 @@ __global__ void scaleFCThS27( real* DC,
f_BSE = fbseF[kbs];
f_BNW = fbnwF[kbw];
//////////////////////////////////////////////////////////////////////////////////
- f27E = (D27F.f[DIR_P00 ])[kzero];//ke
- f27W = (D27F.f[DIR_M00 ])[kw ];
- f27N = (D27F.f[DIR_0P0 ])[kzero];//kn
- f27S = (D27F.f[DIR_0M0 ])[ks ];
- f27T = (D27F.f[DIR_00P ])[kzero];//kt
- f27B = (D27F.f[DIR_00M ])[kb ];
- f27NE = (D27F.f[DIR_PP0 ])[kzero];//kne
- f27SW = (D27F.f[DIR_MM0 ])[ksw ];
- f27SE = (D27F.f[DIR_PM0 ])[ks ];//kse
- f27NW = (D27F.f[DIR_MP0 ])[kw ];//knw
- f27TE = (D27F.f[DIR_P0P ])[kzero];//kte
- f27BW = (D27F.f[DIR_M0M ])[kbw ];
- f27BE = (D27F.f[DIR_P0M ])[kb ];//kbe
- f27TW = (D27F.f[DIR_M0P ])[kw ];//ktw
- f27TN = (D27F.f[DIR_0PP ])[kzero];//ktn
- f27BS = (D27F.f[DIR_0MM ])[kbs ];
- f27BN = (D27F.f[DIR_0PM ])[kb ];//kbn
- f27TS = (D27F.f[DIR_0MP ])[ks ];//kts
+ f27E = (D27F.f[DIR_P00])[kzero];//ke
+ f27W = (D27F.f[DIR_M00])[kw ];
+ f27N = (D27F.f[DIR_0P0])[kzero];//kn
+ f27S = (D27F.f[DIR_0M0])[ks ];
+ f27T = (D27F.f[DIR_00P])[kzero];//kt
+ f27B = (D27F.f[DIR_00M])[kb ];
+ f27NE = (D27F.f[DIR_PP0])[kzero];//kne
+ f27SW = (D27F.f[DIR_MM0])[ksw ];
+ f27SE = (D27F.f[DIR_PM0])[ks ];//kse
+ f27NW = (D27F.f[DIR_MP0])[kw ];//knw
+ f27TE = (D27F.f[DIR_P0P])[kzero];//kte
+ f27BW = (D27F.f[DIR_M0M])[kbw ];
+ f27BE = (D27F.f[DIR_P0M])[kb ];//kbe
+ f27TW = (D27F.f[DIR_M0P])[kw ];//ktw
+ f27TN = (D27F.f[DIR_0PP])[kzero];//ktn
+ f27BS = (D27F.f[DIR_0MM])[kbs ];
+ f27BN = (D27F.f[DIR_0PM])[kb ];//kbn
+ f27TS = (D27F.f[DIR_0MP])[ks ];//kts
f27ZERO = (D27F.f[DIR_000])[kzero];//kzero
- f27TNE = (D27F.f[DIR_PPP ])[kzero];//ktne
- f27TSW = (D27F.f[DIR_MMP ])[ksw ];//ktsw
- f27TSE = (D27F.f[DIR_PMP ])[ks ];//ktse
- f27TNW = (D27F.f[DIR_MPP ])[kw ];//ktnw
- f27BNE = (D27F.f[DIR_PPM ])[kb ];//kbne
- f27BSW = (D27F.f[DIR_MMM ])[kbsw ];
- f27BSE = (D27F.f[DIR_PMM ])[kbs ];//kbse
- f27BNW = (D27F.f[DIR_MPM ])[kbw ];//kbnw
+ f27TNE = (D27F.f[DIR_PPP])[kzero];//ktne
+ f27TSW = (D27F.f[DIR_MMP])[ksw ];//ktsw
+ f27TSE = (D27F.f[DIR_PMP])[ks ];//ktse
+ f27TNW = (D27F.f[DIR_MPP])[kw ];//ktnw
+ f27BNE = (D27F.f[DIR_PPM])[kb ];//kbne
+ f27BSW = (D27F.f[DIR_MMM])[kbsw ];
+ f27BSE = (D27F.f[DIR_PMM])[kbs ];//kbse
+ f27BNW = (D27F.f[DIR_MPM])[kbw ];//kbnw
Conc_F_NWB = f27E + f27W + f27N + f27S + f27T + f27B + f27NE + f27SW + f27SE + f27NW +
f27TE + f27BW + f27BE + f27TW + f27TN + f27BS + f27BN + f27TS + f27ZERO +
@@ -22425,33 +21612,33 @@ __global__ void scaleFCThS27( real* DC,
f_BSE = fbseF[kbs];
f_BNW = fbnwF[kbw];
//////////////////////////////////////////////////////////////////////////////////
- f27E = (D27F.f[DIR_P00 ])[kzero];//ke
- f27W = (D27F.f[DIR_M00 ])[kw ];
- f27N = (D27F.f[DIR_0P0 ])[kzero];//kn
- f27S = (D27F.f[DIR_0M0 ])[ks ];
- f27T = (D27F.f[DIR_00P ])[kzero];//kt
- f27B = (D27F.f[DIR_00M ])[kb ];
- f27NE = (D27F.f[DIR_PP0 ])[kzero];//kne
- f27SW = (D27F.f[DIR_MM0 ])[ksw ];
- f27SE = (D27F.f[DIR_PM0 ])[ks ];//kse
- f27NW = (D27F.f[DIR_MP0 ])[kw ];//knw
- f27TE = (D27F.f[DIR_P0P ])[kzero];//kte
- f27BW = (D27F.f[DIR_M0M ])[kbw ];
- f27BE = (D27F.f[DIR_P0M ])[kb ];//kbe
- f27TW = (D27F.f[DIR_M0P ])[kw ];//ktw
- f27TN = (D27F.f[DIR_0PP ])[kzero];//ktn
- f27BS = (D27F.f[DIR_0MM ])[kbs ];
- f27BN = (D27F.f[DIR_0PM ])[kb ];//kbn
- f27TS = (D27F.f[DIR_0MP ])[ks ];//kts
+ f27E = (D27F.f[DIR_P00])[kzero];//ke
+ f27W = (D27F.f[DIR_M00])[kw ];
+ f27N = (D27F.f[DIR_0P0])[kzero];//kn
+ f27S = (D27F.f[DIR_0M0])[ks ];
+ f27T = (D27F.f[DIR_00P])[kzero];//kt
+ f27B = (D27F.f[DIR_00M])[kb ];
+ f27NE = (D27F.f[DIR_PP0])[kzero];//kne
+ f27SW = (D27F.f[DIR_MM0])[ksw ];
+ f27SE = (D27F.f[DIR_PM0])[ks ];//kse
+ f27NW = (D27F.f[DIR_MP0])[kw ];//knw
+ f27TE = (D27F.f[DIR_P0P])[kzero];//kte
+ f27BW = (D27F.f[DIR_M0M])[kbw ];
+ f27BE = (D27F.f[DIR_P0M])[kb ];//kbe
+ f27TW = (D27F.f[DIR_M0P])[kw ];//ktw
+ f27TN = (D27F.f[DIR_0PP])[kzero];//ktn
+ f27BS = (D27F.f[DIR_0MM])[kbs ];
+ f27BN = (D27F.f[DIR_0PM])[kb ];//kbn
+ f27TS = (D27F.f[DIR_0MP])[ks ];//kts
f27ZERO = (D27F.f[DIR_000])[kzero];//kzero
- f27TNE = (D27F.f[DIR_PPP ])[kzero];//ktne
- f27TSW = (D27F.f[DIR_MMP ])[ksw ];//ktsw
- f27TSE = (D27F.f[DIR_PMP ])[ks ];//ktse
- f27TNW = (D27F.f[DIR_MPP ])[kw ];//ktnw
- f27BNE = (D27F.f[DIR_PPM ])[kb ];//kbne
- f27BSW = (D27F.f[DIR_MMM ])[kbsw ];
- f27BSE = (D27F.f[DIR_PMM ])[kbs ];//kbse
- f27BNW = (D27F.f[DIR_MPM ])[kbw ];//kbnw
+ f27TNE = (D27F.f[DIR_PPP])[kzero];//ktne
+ f27TSW = (D27F.f[DIR_MMP])[ksw ];//ktsw
+ f27TSE = (D27F.f[DIR_PMP])[ks ];//ktse
+ f27TNW = (D27F.f[DIR_MPP])[kw ];//ktnw
+ f27BNE = (D27F.f[DIR_PPM])[kb ];//kbne
+ f27BSW = (D27F.f[DIR_MMM])[kbsw ];
+ f27BSE = (D27F.f[DIR_PMM])[kbs ];//kbse
+ f27BNW = (D27F.f[DIR_MPM])[kbw ];//kbnw
Conc_F_NWT = f27E + f27W + f27N + f27S + f27T + f27B + f27NE + f27SW + f27SE + f27NW +
f27TE + f27BW + f27BE + f27TW + f27TN + f27BS + f27BN + f27TS + f27ZERO +
@@ -22512,33 +21699,33 @@ __global__ void scaleFCThS27( real* DC,
f_BSE = fbseF[kbs];
f_BNW = fbnwF[kbw];
//////////////////////////////////////////////////////////////////////////////////
- f27E = (D27F.f[DIR_P00 ])[kzero];//ke
- f27W = (D27F.f[DIR_M00 ])[kw ];
- f27N = (D27F.f[DIR_0P0 ])[kzero];//kn
- f27S = (D27F.f[DIR_0M0 ])[ks ];
- f27T = (D27F.f[DIR_00P ])[kzero];//kt
- f27B = (D27F.f[DIR_00M ])[kb ];
- f27NE = (D27F.f[DIR_PP0 ])[kzero];//kne
- f27SW = (D27F.f[DIR_MM0 ])[ksw ];
- f27SE = (D27F.f[DIR_PM0 ])[ks ];//kse
- f27NW = (D27F.f[DIR_MP0 ])[kw ];//knw
- f27TE = (D27F.f[DIR_P0P ])[kzero];//kte
- f27BW = (D27F.f[DIR_M0M ])[kbw ];
- f27BE = (D27F.f[DIR_P0M ])[kb ];//kbe
- f27TW = (D27F.f[DIR_M0P ])[kw ];//ktw
- f27TN = (D27F.f[DIR_0PP ])[kzero];//ktn
- f27BS = (D27F.f[DIR_0MM ])[kbs ];
- f27BN = (D27F.f[DIR_0PM ])[kb ];//kbn
- f27TS = (D27F.f[DIR_0MP ])[ks ];//kts
+ f27E = (D27F.f[DIR_P00])[kzero];//ke
+ f27W = (D27F.f[DIR_M00])[kw ];
+ f27N = (D27F.f[DIR_0P0])[kzero];//kn
+ f27S = (D27F.f[DIR_0M0])[ks ];
+ f27T = (D27F.f[DIR_00P])[kzero];//kt
+ f27B = (D27F.f[DIR_00M])[kb ];
+ f27NE = (D27F.f[DIR_PP0])[kzero];//kne
+ f27SW = (D27F.f[DIR_MM0])[ksw ];
+ f27SE = (D27F.f[DIR_PM0])[ks ];//kse
+ f27NW = (D27F.f[DIR_MP0])[kw ];//knw
+ f27TE = (D27F.f[DIR_P0P])[kzero];//kte
+ f27BW = (D27F.f[DIR_M0M])[kbw ];
+ f27BE = (D27F.f[DIR_P0M])[kb ];//kbe
+ f27TW = (D27F.f[DIR_M0P])[kw ];//ktw
+ f27TN = (D27F.f[DIR_0PP])[kzero];//ktn
+ f27BS = (D27F.f[DIR_0MM])[kbs ];
+ f27BN = (D27F.f[DIR_0PM])[kb ];//kbn
+ f27TS = (D27F.f[DIR_0MP])[ks ];//kts
f27ZERO = (D27F.f[DIR_000])[kzero];//kzero
- f27TNE = (D27F.f[DIR_PPP ])[kzero];//ktne
- f27TSW = (D27F.f[DIR_MMP ])[ksw ];//ktsw
- f27TSE = (D27F.f[DIR_PMP ])[ks ];//ktse
- f27TNW = (D27F.f[DIR_MPP ])[kw ];//ktnw
- f27BNE = (D27F.f[DIR_PPM ])[kb ];//kbne
- f27BSW = (D27F.f[DIR_MMM ])[kbsw ];
- f27BSE = (D27F.f[DIR_PMM ])[kbs ];//kbse
- f27BNW = (D27F.f[DIR_MPM ])[kbw ];//kbnw
+ f27TNE = (D27F.f[DIR_PPP])[kzero];//ktne
+ f27TSW = (D27F.f[DIR_MMP])[ksw ];//ktsw
+ f27TSE = (D27F.f[DIR_PMP])[ks ];//ktse
+ f27TNW = (D27F.f[DIR_MPP])[kw ];//ktnw
+ f27BNE = (D27F.f[DIR_PPM])[kb ];//kbne
+ f27BSW = (D27F.f[DIR_MMM])[kbsw ];
+ f27BSE = (D27F.f[DIR_PMM])[kbs ];//kbse
+ f27BNW = (D27F.f[DIR_MPM])[kbw ];//kbnw
Conc_F_NET = f27E + f27W + f27N + f27S + f27T + f27B + f27NE + f27SW + f27SE + f27NW +
f27TE + f27BW + f27BE + f27TW + f27TN + f27BS + f27BN + f27TS + f27ZERO +
@@ -22599,33 +21786,33 @@ __global__ void scaleFCThS27( real* DC,
f_BSE = fbseF[kbs];
f_BNW = fbnwF[kbw];
//////////////////////////////////////////////////////////////////////////////////
- f27E = (D27F.f[DIR_P00 ])[kzero];//ke
- f27W = (D27F.f[DIR_M00 ])[kw ];
- f27N = (D27F.f[DIR_0P0 ])[kzero];//kn
- f27S = (D27F.f[DIR_0M0 ])[ks ];
- f27T = (D27F.f[DIR_00P ])[kzero];//kt
- f27B = (D27F.f[DIR_00M ])[kb ];
- f27NE = (D27F.f[DIR_PP0 ])[kzero];//kne
- f27SW = (D27F.f[DIR_MM0 ])[ksw ];
- f27SE = (D27F.f[DIR_PM0 ])[ks ];//kse
- f27NW = (D27F.f[DIR_MP0 ])[kw ];//knw
- f27TE = (D27F.f[DIR_P0P ])[kzero];//kte
- f27BW = (D27F.f[DIR_M0M ])[kbw ];
- f27BE = (D27F.f[DIR_P0M ])[kb ];//kbe
- f27TW = (D27F.f[DIR_M0P ])[kw ];//ktw
- f27TN = (D27F.f[DIR_0PP ])[kzero];//ktn
- f27BS = (D27F.f[DIR_0MM ])[kbs ];
- f27BN = (D27F.f[DIR_0PM ])[kb ];//kbn
- f27TS = (D27F.f[DIR_0MP ])[ks ];//kts
+ f27E = (D27F.f[DIR_P00])[kzero];//ke
+ f27W = (D27F.f[DIR_M00])[kw ];
+ f27N = (D27F.f[DIR_0P0])[kzero];//kn
+ f27S = (D27F.f[DIR_0M0])[ks ];
+ f27T = (D27F.f[DIR_00P])[kzero];//kt
+ f27B = (D27F.f[DIR_00M])[kb ];
+ f27NE = (D27F.f[DIR_PP0])[kzero];//kne
+ f27SW = (D27F.f[DIR_MM0])[ksw ];
+ f27SE = (D27F.f[DIR_PM0])[ks ];//kse
+ f27NW = (D27F.f[DIR_MP0])[kw ];//knw
+ f27TE = (D27F.f[DIR_P0P])[kzero];//kte
+ f27BW = (D27F.f[DIR_M0M])[kbw ];
+ f27BE = (D27F.f[DIR_P0M])[kb ];//kbe
+ f27TW = (D27F.f[DIR_M0P])[kw ];//ktw
+ f27TN = (D27F.f[DIR_0PP])[kzero];//ktn
+ f27BS = (D27F.f[DIR_0MM])[kbs ];
+ f27BN = (D27F.f[DIR_0PM])[kb ];//kbn
+ f27TS = (D27F.f[DIR_0MP])[ks ];//kts
f27ZERO = (D27F.f[DIR_000])[kzero];//kzero
- f27TNE = (D27F.f[DIR_PPP ])[kzero];//ktne
- f27TSW = (D27F.f[DIR_MMP ])[ksw ];//ktsw
- f27TSE = (D27F.f[DIR_PMP ])[ks ];//ktse
- f27TNW = (D27F.f[DIR_MPP ])[kw ];//ktnw
- f27BNE = (D27F.f[DIR_PPM ])[kb ];//kbne
- f27BSW = (D27F.f[DIR_MMM ])[kbsw ];
- f27BSE = (D27F.f[DIR_PMM ])[kbs ];//kbse
- f27BNW = (D27F.f[DIR_MPM ])[kbw ];//kbnw
+ f27TNE = (D27F.f[DIR_PPP])[kzero];//ktne
+ f27TSW = (D27F.f[DIR_MMP])[ksw ];//ktsw
+ f27TSE = (D27F.f[DIR_PMP])[ks ];//ktse
+ f27TNW = (D27F.f[DIR_MPP])[kw ];//ktnw
+ f27BNE = (D27F.f[DIR_PPM])[kb ];//kbne
+ f27BSW = (D27F.f[DIR_MMM])[kbsw ];
+ f27BSE = (D27F.f[DIR_PMM])[kbs ];//kbse
+ f27BNW = (D27F.f[DIR_MPM])[kbw ];//kbnw
Conc_F_NEB = f27E + f27W + f27N + f27S + f27T + f27B + f27NE + f27SW + f27SE + f27NW +
f27TE + f27BW + f27BE + f27TW + f27TN + f27BS + f27BN + f27TS + f27ZERO +
@@ -22739,32 +21926,32 @@ __global__ void scaleFCThS27( real* DC,
cu_sq=c3o2*(vx1*vx1+vx2*vx2+vx3*vx3);
(D27C.f[DIR_000])[kzero] = c8o27* Conc_C*(c1o1-cu_sq);
- (D27C.f[DIR_P00 ])[kzero] = c2o27* (c3o1*( Mx )+Conc_C*(c1o1+c9o2*( vx1 )*( vx1 )-cu_sq));
- (D27C.f[DIR_M00 ])[kw ] = c2o27* (c3o1*(-Mx )+Conc_C*(c1o1+c9o2*(-vx1 )*(-vx1 )-cu_sq));
- (D27C.f[DIR_0P0 ])[kzero] = c2o27* (c3o1*( My )+Conc_C*(c1o1+c9o2*( vx2 )*( vx2 )-cu_sq));
- (D27C.f[DIR_0M0 ])[ks ] = c2o27* (c3o1*( -My )+Conc_C*(c1o1+c9o2*( -vx2 )*( -vx2 )-cu_sq));
- (D27C.f[DIR_00P ])[kzero] = c2o27* (c3o1*( Mz)+Conc_C*(c1o1+c9o2*( vx3)*( vx3)-cu_sq));
- (D27C.f[DIR_00M ])[kb ] = c2o27* (c3o1*( -Mz)+Conc_C*(c1o1+c9o2*( -vx3)*( -vx3)-cu_sq));
- (D27C.f[DIR_PP0 ])[kzero] = c1o54* (c3o1*( Mx +My )+Conc_C*(c1o1+c9o2*( vx1+vx2 )*( vx1+vx2 )-cu_sq));
- (D27C.f[DIR_MM0 ])[ksw ] = c1o54* (c3o1*(-Mx -My )+Conc_C*(c1o1+c9o2*(-vx1-vx2 )*(-vx1-vx2 )-cu_sq));
- (D27C.f[DIR_PM0 ])[ks ] = c1o54* (c3o1*( Mx -My )+Conc_C*(c1o1+c9o2*( vx1-vx2 )*( vx1-vx2 )-cu_sq));
- (D27C.f[DIR_MP0 ])[kw ] = c1o54* (c3o1*(-Mx +My )+Conc_C*(c1o1+c9o2*(-vx1+vx2 )*(-vx1+vx2 )-cu_sq));
- (D27C.f[DIR_P0P ])[kzero] = c1o54* (c3o1*( Mx +Mz)+Conc_C*(c1o1+c9o2*( vx1 +vx3)*( vx1 +vx3)-cu_sq));
- (D27C.f[DIR_M0M ])[kbw ] = c1o54* (c3o1*(-Mx -Mz)+Conc_C*(c1o1+c9o2*(-vx1 -vx3)*(-vx1 -vx3)-cu_sq));
- (D27C.f[DIR_P0M ])[kb ] = c1o54* (c3o1*( Mx -Mz)+Conc_C*(c1o1+c9o2*( vx1 -vx3)*( vx1 -vx3)-cu_sq));
- (D27C.f[DIR_M0P ])[kw ] = c1o54* (c3o1*(-Mx +Mz)+Conc_C*(c1o1+c9o2*(-vx1 +vx3)*(-vx1 +vx3)-cu_sq));
- (D27C.f[DIR_0PP ])[kzero] = c1o54* (c3o1*( My +Mz)+Conc_C*(c1o1+c9o2*( vx2+vx3)*( vx2+vx3)-cu_sq));
- (D27C.f[DIR_0MM ])[kbs ] = c1o54* (c3o1*( -My -Mz)+Conc_C*(c1o1+c9o2*( -vx2-vx3)*( -vx2-vx3)-cu_sq));
- (D27C.f[DIR_0PM ])[kb ] = c1o54* (c3o1*( My -Mz)+Conc_C*(c1o1+c9o2*( vx2-vx3)*( vx2-vx3)-cu_sq));
- (D27C.f[DIR_0MP ])[ks ] = c1o54* (c3o1*( -My +Mz)+Conc_C*(c1o1+c9o2*( -vx2+vx3)*( -vx2+vx3)-cu_sq));
- (D27C.f[DIR_PPP ])[kzero] = c1o216*(c3o1*( Mx +My +Mz)+Conc_C*(c1o1+c9o2*( vx1+vx2+vx3)*( vx1+vx2+vx3)-cu_sq));
- (D27C.f[DIR_MMM ])[kbsw ] = c1o216*(c3o1*(-Mx -My -Mz)+Conc_C*(c1o1+c9o2*(-vx1-vx2-vx3)*(-vx1-vx2-vx3)-cu_sq));
- (D27C.f[DIR_PPM ])[kb ] = c1o216*(c3o1*( Mx +My -Mz)+Conc_C*(c1o1+c9o2*( vx1+vx2-vx3)*( vx1+vx2-vx3)-cu_sq));
- (D27C.f[DIR_MMP ])[ksw ] = c1o216*(c3o1*(-Mx -My +Mz)+Conc_C*(c1o1+c9o2*(-vx1-vx2+vx3)*(-vx1-vx2+vx3)-cu_sq));
- (D27C.f[DIR_PMP ])[ks ] = c1o216*(c3o1*( Mx -My +Mz)+Conc_C*(c1o1+c9o2*( vx1-vx2+vx3)*( vx1-vx2+vx3)-cu_sq));
- (D27C.f[DIR_MPM ])[kbw ] = c1o216*(c3o1*(-Mx +My -Mz)+Conc_C*(c1o1+c9o2*(-vx1+vx2-vx3)*(-vx1+vx2-vx3)-cu_sq));
- (D27C.f[DIR_PMM ])[kbs ] = c1o216*(c3o1*( Mx -My -Mz)+Conc_C*(c1o1+c9o2*( vx1-vx2-vx3)*( vx1-vx2-vx3)-cu_sq));
- (D27C.f[DIR_MPP ])[kw ] = c1o216*(c3o1*(-Mx +My +Mz)+Conc_C*(c1o1+c9o2*(-vx1+vx2+vx3)*(-vx1+vx2+vx3)-cu_sq));
+ (D27C.f[DIR_P00])[kzero] = c2o27* (c3o1*( Mx )+Conc_C*(c1o1+c9o2*( vx1 )*( vx1 )-cu_sq));
+ (D27C.f[DIR_M00])[kw ] = c2o27* (c3o1*(-Mx )+Conc_C*(c1o1+c9o2*(-vx1 )*(-vx1 )-cu_sq));
+ (D27C.f[DIR_0P0])[kzero] = c2o27* (c3o1*( My )+Conc_C*(c1o1+c9o2*( vx2 )*( vx2 )-cu_sq));
+ (D27C.f[DIR_0M0])[ks ] = c2o27* (c3o1*( -My )+Conc_C*(c1o1+c9o2*( -vx2 )*( -vx2 )-cu_sq));
+ (D27C.f[DIR_00P])[kzero] = c2o27* (c3o1*( Mz)+Conc_C*(c1o1+c9o2*( vx3)*( vx3)-cu_sq));
+ (D27C.f[DIR_00M])[kb ] = c2o27* (c3o1*( -Mz)+Conc_C*(c1o1+c9o2*( -vx3)*( -vx3)-cu_sq));
+ (D27C.f[DIR_PP0])[kzero] = c1o54* (c3o1*( Mx +My )+Conc_C*(c1o1+c9o2*( vx1+vx2 )*( vx1+vx2 )-cu_sq));
+ (D27C.f[DIR_MM0])[ksw ] = c1o54* (c3o1*(-Mx -My )+Conc_C*(c1o1+c9o2*(-vx1-vx2 )*(-vx1-vx2 )-cu_sq));
+ (D27C.f[DIR_PM0])[ks ] = c1o54* (c3o1*( Mx -My )+Conc_C*(c1o1+c9o2*( vx1-vx2 )*( vx1-vx2 )-cu_sq));
+ (D27C.f[DIR_MP0])[kw ] = c1o54* (c3o1*(-Mx +My )+Conc_C*(c1o1+c9o2*(-vx1+vx2 )*(-vx1+vx2 )-cu_sq));
+ (D27C.f[DIR_P0P])[kzero] = c1o54* (c3o1*( Mx +Mz)+Conc_C*(c1o1+c9o2*( vx1 +vx3)*( vx1 +vx3)-cu_sq));
+ (D27C.f[DIR_M0M])[kbw ] = c1o54* (c3o1*(-Mx -Mz)+Conc_C*(c1o1+c9o2*(-vx1 -vx3)*(-vx1 -vx3)-cu_sq));
+ (D27C.f[DIR_P0M])[kb ] = c1o54* (c3o1*( Mx -Mz)+Conc_C*(c1o1+c9o2*( vx1 -vx3)*( vx1 -vx3)-cu_sq));
+ (D27C.f[DIR_M0P])[kw ] = c1o54* (c3o1*(-Mx +Mz)+Conc_C*(c1o1+c9o2*(-vx1 +vx3)*(-vx1 +vx3)-cu_sq));
+ (D27C.f[DIR_0PP])[kzero] = c1o54* (c3o1*( My +Mz)+Conc_C*(c1o1+c9o2*( vx2+vx3)*( vx2+vx3)-cu_sq));
+ (D27C.f[DIR_0MM])[kbs ] = c1o54* (c3o1*( -My -Mz)+Conc_C*(c1o1+c9o2*( -vx2-vx3)*( -vx2-vx3)-cu_sq));
+ (D27C.f[DIR_0PM])[kb ] = c1o54* (c3o1*( My -Mz)+Conc_C*(c1o1+c9o2*( vx2-vx3)*( vx2-vx3)-cu_sq));
+ (D27C.f[DIR_0MP])[ks ] = c1o54* (c3o1*( -My +Mz)+Conc_C*(c1o1+c9o2*( -vx2+vx3)*( -vx2+vx3)-cu_sq));
+ (D27C.f[DIR_PPP])[kzero] = c1o216*(c3o1*( Mx +My +Mz)+Conc_C*(c1o1+c9o2*( vx1+vx2+vx3)*( vx1+vx2+vx3)-cu_sq));
+ (D27C.f[DIR_MMM])[kbsw ] = c1o216*(c3o1*(-Mx -My -Mz)+Conc_C*(c1o1+c9o2*(-vx1-vx2-vx3)*(-vx1-vx2-vx3)-cu_sq));
+ (D27C.f[DIR_PPM])[kb ] = c1o216*(c3o1*( Mx +My -Mz)+Conc_C*(c1o1+c9o2*( vx1+vx2-vx3)*( vx1+vx2-vx3)-cu_sq));
+ (D27C.f[DIR_MMP])[ksw ] = c1o216*(c3o1*(-Mx -My +Mz)+Conc_C*(c1o1+c9o2*(-vx1-vx2+vx3)*(-vx1-vx2+vx3)-cu_sq));
+ (D27C.f[DIR_PMP])[ks ] = c1o216*(c3o1*( Mx -My +Mz)+Conc_C*(c1o1+c9o2*( vx1-vx2+vx3)*( vx1-vx2+vx3)-cu_sq));
+ (D27C.f[DIR_MPM])[kbw ] = c1o216*(c3o1*(-Mx +My -Mz)+Conc_C*(c1o1+c9o2*(-vx1+vx2-vx3)*(-vx1+vx2-vx3)-cu_sq));
+ (D27C.f[DIR_PMM])[kbs ] = c1o216*(c3o1*( Mx -My -Mz)+Conc_C*(c1o1+c9o2*( vx1-vx2-vx3)*( vx1-vx2-vx3)-cu_sq));
+ (D27C.f[DIR_MPP])[kw ] = c1o216*(c3o1*(-Mx +My +Mz)+Conc_C*(c1o1+c9o2*(-vx1+vx2+vx3)*(-vx1+vx2+vx3)-cu_sq));
}
}
@@ -22812,8 +21999,8 @@ __global__ void scaleFCEff27(real* DC,
unsigned int* neighborFX,
unsigned int* neighborFY,
unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
+ unsigned long long numberOfLBnodesCoarse,
+ unsigned long long numberOfLBnodesFine,
bool isEvenTimestep,
unsigned int* posC,
unsigned int* posFSWB,
@@ -22830,96 +22017,96 @@ __global__ void scaleFCEff27(real* DC,
real *feF, *fwF, *fnF, *fsF, *ftF, *fbF, *fneF, *fswF, *fseF, *fnwF, *fteF, *fbwF, *fbeF, *ftwF, *ftnF, *fbsF, *fbnF, *ftsF, *fzeroF,
*ftneF, *ftswF, *ftseF, *ftnwF, *fbneF, *fbswF, *fbseF, *fbnwF;
- feF = &DF[DIR_P00 *size_MatF];
- fwF = &DF[DIR_M00 *size_MatF];
- fnF = &DF[DIR_0P0 *size_MatF];
- fsF = &DF[DIR_0M0 *size_MatF];
- ftF = &DF[DIR_00P *size_MatF];
- fbF = &DF[DIR_00M *size_MatF];
- fneF = &DF[DIR_PP0 *size_MatF];
- fswF = &DF[DIR_MM0 *size_MatF];
- fseF = &DF[DIR_PM0 *size_MatF];
- fnwF = &DF[DIR_MP0 *size_MatF];
- fteF = &DF[DIR_P0P *size_MatF];
- fbwF = &DF[DIR_M0M *size_MatF];
- fbeF = &DF[DIR_P0M *size_MatF];
- ftwF = &DF[DIR_M0P *size_MatF];
- ftnF = &DF[DIR_0PP *size_MatF];
- fbsF = &DF[DIR_0MM *size_MatF];
- fbnF = &DF[DIR_0PM *size_MatF];
- ftsF = &DF[DIR_0MP *size_MatF];
- fzeroF = &DF[DIR_000*size_MatF];
- ftneF = &DF[DIR_PPP *size_MatF];
- ftswF = &DF[DIR_MMP *size_MatF];
- ftseF = &DF[DIR_PMP *size_MatF];
- ftnwF = &DF[DIR_MPP *size_MatF];
- fbneF = &DF[DIR_PPM *size_MatF];
- fbswF = &DF[DIR_MMM *size_MatF];
- fbseF = &DF[DIR_PMM *size_MatF];
- fbnwF = &DF[DIR_MPM *size_MatF];
+ feF = &DF[DIR_P00 * numberOfLBnodesFine];
+ fwF = &DF[DIR_M00 * numberOfLBnodesFine];
+ fnF = &DF[DIR_0P0 * numberOfLBnodesFine];
+ fsF = &DF[DIR_0M0 * numberOfLBnodesFine];
+ ftF = &DF[DIR_00P * numberOfLBnodesFine];
+ fbF = &DF[DIR_00M * numberOfLBnodesFine];
+ fneF = &DF[DIR_PP0 * numberOfLBnodesFine];
+ fswF = &DF[DIR_MM0 * numberOfLBnodesFine];
+ fseF = &DF[DIR_PM0 * numberOfLBnodesFine];
+ fnwF = &DF[DIR_MP0 * numberOfLBnodesFine];
+ fteF = &DF[DIR_P0P * numberOfLBnodesFine];
+ fbwF = &DF[DIR_M0M * numberOfLBnodesFine];
+ fbeF = &DF[DIR_P0M * numberOfLBnodesFine];
+ ftwF = &DF[DIR_M0P * numberOfLBnodesFine];
+ ftnF = &DF[DIR_0PP * numberOfLBnodesFine];
+ fbsF = &DF[DIR_0MM * numberOfLBnodesFine];
+ fbnF = &DF[DIR_0PM * numberOfLBnodesFine];
+ ftsF = &DF[DIR_0MP * numberOfLBnodesFine];
+ fzeroF = &DF[DIR_000 * numberOfLBnodesFine];
+ ftneF = &DF[DIR_PPP * numberOfLBnodesFine];
+ ftswF = &DF[DIR_MMP * numberOfLBnodesFine];
+ ftseF = &DF[DIR_PMP * numberOfLBnodesFine];
+ ftnwF = &DF[DIR_MPP * numberOfLBnodesFine];
+ fbneF = &DF[DIR_PPM * numberOfLBnodesFine];
+ fbswF = &DF[DIR_MMM * numberOfLBnodesFine];
+ fbseF = &DF[DIR_PMM * numberOfLBnodesFine];
+ fbnwF = &DF[DIR_MPM * numberOfLBnodesFine];
real *feC, *fwC, *fnC, *fsC, *ftC, *fbC, *fneC, *fswC, *fseC, *fnwC, *fteC, *fbwC, *fbeC, *ftwC, *ftnC, *fbsC, *fbnC, *ftsC, *fzeroC,
*ftneC, *ftswC, *ftseC, *ftnwC, *fbneC, *fbswC, *fbseC, *fbnwC;
if (isEvenTimestep==true)
{
- feC = &DC[DIR_P00 *size_MatC];
- fwC = &DC[DIR_M00 *size_MatC];
- fnC = &DC[DIR_0P0 *size_MatC];
- fsC = &DC[DIR_0M0 *size_MatC];
- ftC = &DC[DIR_00P *size_MatC];
- fbC = &DC[DIR_00M *size_MatC];
- fneC = &DC[DIR_PP0 *size_MatC];
- fswC = &DC[DIR_MM0 *size_MatC];
- fseC = &DC[DIR_PM0 *size_MatC];
- fnwC = &DC[DIR_MP0 *size_MatC];
- fteC = &DC[DIR_P0P *size_MatC];
- fbwC = &DC[DIR_M0M *size_MatC];
- fbeC = &DC[DIR_P0M *size_MatC];
- ftwC = &DC[DIR_M0P *size_MatC];
- ftnC = &DC[DIR_0PP *size_MatC];
- fbsC = &DC[DIR_0MM *size_MatC];
- fbnC = &DC[DIR_0PM *size_MatC];
- ftsC = &DC[DIR_0MP *size_MatC];
- fzeroC = &DC[DIR_000*size_MatC];
- ftneC = &DC[DIR_PPP *size_MatC];
- ftswC = &DC[DIR_MMP *size_MatC];
- ftseC = &DC[DIR_PMP *size_MatC];
- ftnwC = &DC[DIR_MPP *size_MatC];
- fbneC = &DC[DIR_PPM *size_MatC];
- fbswC = &DC[DIR_MMM *size_MatC];
- fbseC = &DC[DIR_PMM *size_MatC];
- fbnwC = &DC[DIR_MPM *size_MatC];
+ feC = &DC[DIR_P00 * numberOfLBnodesCoarse];
+ fwC = &DC[DIR_M00 * numberOfLBnodesCoarse];
+ fnC = &DC[DIR_0P0 * numberOfLBnodesCoarse];
+ fsC = &DC[DIR_0M0 * numberOfLBnodesCoarse];
+ ftC = &DC[DIR_00P * numberOfLBnodesCoarse];
+ fbC = &DC[DIR_00M * numberOfLBnodesCoarse];
+ fneC = &DC[DIR_PP0 * numberOfLBnodesCoarse];
+ fswC = &DC[DIR_MM0 * numberOfLBnodesCoarse];
+ fseC = &DC[DIR_PM0 * numberOfLBnodesCoarse];
+ fnwC = &DC[DIR_MP0 * numberOfLBnodesCoarse];
+ fteC = &DC[DIR_P0P * numberOfLBnodesCoarse];
+ fbwC = &DC[DIR_M0M * numberOfLBnodesCoarse];
+ fbeC = &DC[DIR_P0M * numberOfLBnodesCoarse];
+ ftwC = &DC[DIR_M0P * numberOfLBnodesCoarse];
+ ftnC = &DC[DIR_0PP * numberOfLBnodesCoarse];
+ fbsC = &DC[DIR_0MM * numberOfLBnodesCoarse];
+ fbnC = &DC[DIR_0PM * numberOfLBnodesCoarse];
+ ftsC = &DC[DIR_0MP * numberOfLBnodesCoarse];
+ fzeroC = &DC[DIR_000 * numberOfLBnodesCoarse];
+ ftneC = &DC[DIR_PPP * numberOfLBnodesCoarse];
+ ftswC = &DC[DIR_MMP * numberOfLBnodesCoarse];
+ ftseC = &DC[DIR_PMP * numberOfLBnodesCoarse];
+ ftnwC = &DC[DIR_MPP * numberOfLBnodesCoarse];
+ fbneC = &DC[DIR_PPM * numberOfLBnodesCoarse];
+ fbswC = &DC[DIR_MMM * numberOfLBnodesCoarse];
+ fbseC = &DC[DIR_PMM * numberOfLBnodesCoarse];
+ fbnwC = &DC[DIR_MPM * numberOfLBnodesCoarse];
}
else
{
- fwC = &DC[DIR_P00 *size_MatC];
- feC = &DC[DIR_M00 *size_MatC];
- fsC = &DC[DIR_0P0 *size_MatC];
- fnC = &DC[DIR_0M0 *size_MatC];
- fbC = &DC[DIR_00P *size_MatC];
- ftC = &DC[DIR_00M *size_MatC];
- fswC = &DC[DIR_PP0 *size_MatC];
- fneC = &DC[DIR_MM0 *size_MatC];
- fnwC = &DC[DIR_PM0 *size_MatC];
- fseC = &DC[DIR_MP0 *size_MatC];
- fbwC = &DC[DIR_P0P *size_MatC];
- fteC = &DC[DIR_M0M *size_MatC];
- ftwC = &DC[DIR_P0M *size_MatC];
- fbeC = &DC[DIR_M0P *size_MatC];
- fbsC = &DC[DIR_0PP *size_MatC];
- ftnC = &DC[DIR_0MM *size_MatC];
- ftsC = &DC[DIR_0PM *size_MatC];
- fbnC = &DC[DIR_0MP *size_MatC];
- fzeroC = &DC[DIR_000*size_MatC];
- fbswC = &DC[DIR_PPP *size_MatC];
- fbneC = &DC[DIR_MMP *size_MatC];
- fbnwC = &DC[DIR_PMP *size_MatC];
- fbseC = &DC[DIR_MPP *size_MatC];
- ftswC = &DC[DIR_PPM *size_MatC];
- ftneC = &DC[DIR_MMM *size_MatC];
- ftnwC = &DC[DIR_PMM *size_MatC];
- ftseC = &DC[DIR_MPM *size_MatC];
+ fwC = &DC[DIR_P00 * numberOfLBnodesCoarse];
+ feC = &DC[DIR_M00 * numberOfLBnodesCoarse];
+ fsC = &DC[DIR_0P0 * numberOfLBnodesCoarse];
+ fnC = &DC[DIR_0M0 * numberOfLBnodesCoarse];
+ fbC = &DC[DIR_00P * numberOfLBnodesCoarse];
+ ftC = &DC[DIR_00M * numberOfLBnodesCoarse];
+ fswC = &DC[DIR_PP0 * numberOfLBnodesCoarse];
+ fneC = &DC[DIR_MM0 * numberOfLBnodesCoarse];
+ fnwC = &DC[DIR_PM0 * numberOfLBnodesCoarse];
+ fseC = &DC[DIR_MP0 * numberOfLBnodesCoarse];
+ fbwC = &DC[DIR_P0P * numberOfLBnodesCoarse];
+ fteC = &DC[DIR_M0M * numberOfLBnodesCoarse];
+ ftwC = &DC[DIR_P0M * numberOfLBnodesCoarse];
+ fbeC = &DC[DIR_M0P * numberOfLBnodesCoarse];
+ fbsC = &DC[DIR_0PP * numberOfLBnodesCoarse];
+ ftnC = &DC[DIR_0MM * numberOfLBnodesCoarse];
+ ftsC = &DC[DIR_0PM * numberOfLBnodesCoarse];
+ fbnC = &DC[DIR_0MP * numberOfLBnodesCoarse];
+ fzeroC = &DC[DIR_000 * numberOfLBnodesCoarse];
+ fbswC = &DC[DIR_PPP * numberOfLBnodesCoarse];
+ fbneC = &DC[DIR_MMP * numberOfLBnodesCoarse];
+ fbnwC = &DC[DIR_PMP * numberOfLBnodesCoarse];
+ fbseC = &DC[DIR_MPP * numberOfLBnodesCoarse];
+ ftswC = &DC[DIR_PPM * numberOfLBnodesCoarse];
+ ftneC = &DC[DIR_MMM * numberOfLBnodesCoarse];
+ ftnwC = &DC[DIR_PMM * numberOfLBnodesCoarse];
+ ftseC = &DC[DIR_MPM * numberOfLBnodesCoarse];
}
////////////////////////////////////////////////////////////////////////////////
const unsigned ix = threadIdx.x; // Globaler x-Index
@@ -23791,8 +22978,8 @@ __global__ void scaleFC27(real* DC,
unsigned int* neighborFX,
unsigned int* neighborFY,
unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
+ unsigned long long numberOfLBnodesCoarse,
+ unsigned long long numberOfLBnodesFine,
bool isEvenTimestep,
unsigned int* posC,
unsigned int* posFSWB,
@@ -23808,96 +22995,96 @@ __global__ void scaleFC27(real* DC,
real *feF, *fwF, *fnF, *fsF, *ftF, *fbF, *fneF, *fswF, *fseF, *fnwF, *fteF, *fbwF, *fbeF, *ftwF, *ftnF, *fbsF, *fbnF, *ftsF, *fzeroF,
*ftneF, *ftswF, *ftseF, *ftnwF, *fbneF, *fbswF, *fbseF, *fbnwF;
- feF = &DF[DIR_P00 *size_MatF];
- fwF = &DF[DIR_M00 *size_MatF];
- fnF = &DF[DIR_0P0 *size_MatF];
- fsF = &DF[DIR_0M0 *size_MatF];
- ftF = &DF[DIR_00P *size_MatF];
- fbF = &DF[DIR_00M *size_MatF];
- fneF = &DF[DIR_PP0 *size_MatF];
- fswF = &DF[DIR_MM0 *size_MatF];
- fseF = &DF[DIR_PM0 *size_MatF];
- fnwF = &DF[DIR_MP0 *size_MatF];
- fteF = &DF[DIR_P0P *size_MatF];
- fbwF = &DF[DIR_M0M *size_MatF];
- fbeF = &DF[DIR_P0M *size_MatF];
- ftwF = &DF[DIR_M0P *size_MatF];
- ftnF = &DF[DIR_0PP *size_MatF];
- fbsF = &DF[DIR_0MM *size_MatF];
- fbnF = &DF[DIR_0PM *size_MatF];
- ftsF = &DF[DIR_0MP *size_MatF];
- fzeroF = &DF[DIR_000*size_MatF];
- ftneF = &DF[DIR_PPP *size_MatF];
- ftswF = &DF[DIR_MMP *size_MatF];
- ftseF = &DF[DIR_PMP *size_MatF];
- ftnwF = &DF[DIR_MPP *size_MatF];
- fbneF = &DF[DIR_PPM *size_MatF];
- fbswF = &DF[DIR_MMM *size_MatF];
- fbseF = &DF[DIR_PMM *size_MatF];
- fbnwF = &DF[DIR_MPM *size_MatF];
+ feF = &DF[DIR_P00 * numberOfLBnodesFine];
+ fwF = &DF[DIR_M00 * numberOfLBnodesFine];
+ fnF = &DF[DIR_0P0 * numberOfLBnodesFine];
+ fsF = &DF[DIR_0M0 * numberOfLBnodesFine];
+ ftF = &DF[DIR_00P * numberOfLBnodesFine];
+ fbF = &DF[DIR_00M * numberOfLBnodesFine];
+ fneF = &DF[DIR_PP0 * numberOfLBnodesFine];
+ fswF = &DF[DIR_MM0 * numberOfLBnodesFine];
+ fseF = &DF[DIR_PM0 * numberOfLBnodesFine];
+ fnwF = &DF[DIR_MP0 * numberOfLBnodesFine];
+ fteF = &DF[DIR_P0P * numberOfLBnodesFine];
+ fbwF = &DF[DIR_M0M * numberOfLBnodesFine];
+ fbeF = &DF[DIR_P0M * numberOfLBnodesFine];
+ ftwF = &DF[DIR_M0P * numberOfLBnodesFine];
+ ftnF = &DF[DIR_0PP * numberOfLBnodesFine];
+ fbsF = &DF[DIR_0MM * numberOfLBnodesFine];
+ fbnF = &DF[DIR_0PM * numberOfLBnodesFine];
+ ftsF = &DF[DIR_0MP * numberOfLBnodesFine];
+ fzeroF = &DF[DIR_000 * numberOfLBnodesFine];
+ ftneF = &DF[DIR_PPP * numberOfLBnodesFine];
+ ftswF = &DF[DIR_MMP * numberOfLBnodesFine];
+ ftseF = &DF[DIR_PMP * numberOfLBnodesFine];
+ ftnwF = &DF[DIR_MPP * numberOfLBnodesFine];
+ fbneF = &DF[DIR_PPM * numberOfLBnodesFine];
+ fbswF = &DF[DIR_MMM * numberOfLBnodesFine];
+ fbseF = &DF[DIR_PMM * numberOfLBnodesFine];
+ fbnwF = &DF[DIR_MPM * numberOfLBnodesFine];
real *feC, *fwC, *fnC, *fsC, *ftC, *fbC, *fneC, *fswC, *fseC, *fnwC, *fteC, *fbwC, *fbeC, *ftwC, *ftnC, *fbsC, *fbnC, *ftsC, *fzeroC,
*ftneC, *ftswC, *ftseC, *ftnwC, *fbneC, *fbswC, *fbseC, *fbnwC;
if (isEvenTimestep==true)
{
- feC = &DC[DIR_P00 *size_MatC];
- fwC = &DC[DIR_M00 *size_MatC];
- fnC = &DC[DIR_0P0 *size_MatC];
- fsC = &DC[DIR_0M0 *size_MatC];
- ftC = &DC[DIR_00P *size_MatC];
- fbC = &DC[DIR_00M *size_MatC];
- fneC = &DC[DIR_PP0 *size_MatC];
- fswC = &DC[DIR_MM0 *size_MatC];
- fseC = &DC[DIR_PM0 *size_MatC];
- fnwC = &DC[DIR_MP0 *size_MatC];
- fteC = &DC[DIR_P0P *size_MatC];
- fbwC = &DC[DIR_M0M *size_MatC];
- fbeC = &DC[DIR_P0M *size_MatC];
- ftwC = &DC[DIR_M0P *size_MatC];
- ftnC = &DC[DIR_0PP *size_MatC];
- fbsC = &DC[DIR_0MM *size_MatC];
- fbnC = &DC[DIR_0PM *size_MatC];
- ftsC = &DC[DIR_0MP *size_MatC];
- fzeroC = &DC[DIR_000*size_MatC];
- ftneC = &DC[DIR_PPP *size_MatC];
- ftswC = &DC[DIR_MMP *size_MatC];
- ftseC = &DC[DIR_PMP *size_MatC];
- ftnwC = &DC[DIR_MPP *size_MatC];
- fbneC = &DC[DIR_PPM *size_MatC];
- fbswC = &DC[DIR_MMM *size_MatC];
- fbseC = &DC[DIR_PMM *size_MatC];
- fbnwC = &DC[DIR_MPM *size_MatC];
+ feC = &DC[DIR_P00 * numberOfLBnodesCoarse];
+ fwC = &DC[DIR_M00 * numberOfLBnodesCoarse];
+ fnC = &DC[DIR_0P0 * numberOfLBnodesCoarse];
+ fsC = &DC[DIR_0M0 * numberOfLBnodesCoarse];
+ ftC = &DC[DIR_00P * numberOfLBnodesCoarse];
+ fbC = &DC[DIR_00M * numberOfLBnodesCoarse];
+ fneC = &DC[DIR_PP0 * numberOfLBnodesCoarse];
+ fswC = &DC[DIR_MM0 * numberOfLBnodesCoarse];
+ fseC = &DC[DIR_PM0 * numberOfLBnodesCoarse];
+ fnwC = &DC[DIR_MP0 * numberOfLBnodesCoarse];
+ fteC = &DC[DIR_P0P * numberOfLBnodesCoarse];
+ fbwC = &DC[DIR_M0M * numberOfLBnodesCoarse];
+ fbeC = &DC[DIR_P0M * numberOfLBnodesCoarse];
+ ftwC = &DC[DIR_M0P * numberOfLBnodesCoarse];
+ ftnC = &DC[DIR_0PP * numberOfLBnodesCoarse];
+ fbsC = &DC[DIR_0MM * numberOfLBnodesCoarse];
+ fbnC = &DC[DIR_0PM * numberOfLBnodesCoarse];
+ ftsC = &DC[DIR_0MP * numberOfLBnodesCoarse];
+ fzeroC = &DC[DIR_000 * numberOfLBnodesCoarse];
+ ftneC = &DC[DIR_PPP * numberOfLBnodesCoarse];
+ ftswC = &DC[DIR_MMP * numberOfLBnodesCoarse];
+ ftseC = &DC[DIR_PMP * numberOfLBnodesCoarse];
+ ftnwC = &DC[DIR_MPP * numberOfLBnodesCoarse];
+ fbneC = &DC[DIR_PPM * numberOfLBnodesCoarse];
+ fbswC = &DC[DIR_MMM * numberOfLBnodesCoarse];
+ fbseC = &DC[DIR_PMM * numberOfLBnodesCoarse];
+ fbnwC = &DC[DIR_MPM * numberOfLBnodesCoarse];
}
else
{
- fwC = &DC[DIR_P00 *size_MatC];
- feC = &DC[DIR_M00 *size_MatC];
- fsC = &DC[DIR_0P0 *size_MatC];
- fnC = &DC[DIR_0M0 *size_MatC];
- fbC = &DC[DIR_00P *size_MatC];
- ftC = &DC[DIR_00M *size_MatC];
- fswC = &DC[DIR_PP0 *size_MatC];
- fneC = &DC[DIR_MM0 *size_MatC];
- fnwC = &DC[DIR_PM0 *size_MatC];
- fseC = &DC[DIR_MP0 *size_MatC];
- fbwC = &DC[DIR_P0P *size_MatC];
- fteC = &DC[DIR_M0M *size_MatC];
- ftwC = &DC[DIR_P0M *size_MatC];
- fbeC = &DC[DIR_M0P *size_MatC];
- fbsC = &DC[DIR_0PP *size_MatC];
- ftnC = &DC[DIR_0MM *size_MatC];
- ftsC = &DC[DIR_0PM *size_MatC];
- fbnC = &DC[DIR_0MP *size_MatC];
- fzeroC = &DC[DIR_000*size_MatC];
- fbswC = &DC[DIR_PPP *size_MatC];
- fbneC = &DC[DIR_MMP *size_MatC];
- fbnwC = &DC[DIR_PMP *size_MatC];
- fbseC = &DC[DIR_MPP *size_MatC];
- ftswC = &DC[DIR_PPM *size_MatC];
- ftneC = &DC[DIR_MMM *size_MatC];
- ftnwC = &DC[DIR_PMM *size_MatC];
- ftseC = &DC[DIR_MPM *size_MatC];
+ fwC = &DC[DIR_P00 * numberOfLBnodesCoarse];
+ feC = &DC[DIR_M00 * numberOfLBnodesCoarse];
+ fsC = &DC[DIR_0P0 * numberOfLBnodesCoarse];
+ fnC = &DC[DIR_0M0 * numberOfLBnodesCoarse];
+ fbC = &DC[DIR_00P * numberOfLBnodesCoarse];
+ ftC = &DC[DIR_00M * numberOfLBnodesCoarse];
+ fswC = &DC[DIR_PP0 * numberOfLBnodesCoarse];
+ fneC = &DC[DIR_MM0 * numberOfLBnodesCoarse];
+ fnwC = &DC[DIR_PM0 * numberOfLBnodesCoarse];
+ fseC = &DC[DIR_MP0 * numberOfLBnodesCoarse];
+ fbwC = &DC[DIR_P0P * numberOfLBnodesCoarse];
+ fteC = &DC[DIR_M0M * numberOfLBnodesCoarse];
+ ftwC = &DC[DIR_P0M * numberOfLBnodesCoarse];
+ fbeC = &DC[DIR_M0P * numberOfLBnodesCoarse];
+ fbsC = &DC[DIR_0PP * numberOfLBnodesCoarse];
+ ftnC = &DC[DIR_0MM * numberOfLBnodesCoarse];
+ ftsC = &DC[DIR_0PM * numberOfLBnodesCoarse];
+ fbnC = &DC[DIR_0MP * numberOfLBnodesCoarse];
+ fzeroC = &DC[DIR_000 * numberOfLBnodesCoarse];
+ fbswC = &DC[DIR_PPP * numberOfLBnodesCoarse];
+ fbneC = &DC[DIR_MMP * numberOfLBnodesCoarse];
+ fbnwC = &DC[DIR_PMP * numberOfLBnodesCoarse];
+ fbseC = &DC[DIR_MPP * numberOfLBnodesCoarse];
+ ftswC = &DC[DIR_PPM * numberOfLBnodesCoarse];
+ ftneC = &DC[DIR_MMM * numberOfLBnodesCoarse];
+ ftnwC = &DC[DIR_PMM * numberOfLBnodesCoarse];
+ ftseC = &DC[DIR_MPM * numberOfLBnodesCoarse];
}
////////////////////////////////////////////////////////////////////////////////
const unsigned ix = threadIdx.x; // Globaler x-Index
diff --git a/src/gpu/VirtualFluids_GPU/GPU/GridScaling/ScaleFC_F3_27.cu b/src/gpu/VirtualFluids_GPU/GPU/GridScaling/ScaleFC_F3_27.cu
index e7fe8b50637e97b9c8cc34025216f4d02e684c55..3b108ad4ae43bd63698f3516a207630214695797 100644
--- a/src/gpu/VirtualFluids_GPU/GPU/GridScaling/ScaleFC_F3_27.cu
+++ b/src/gpu/VirtualFluids_GPU/GPU/GridScaling/ScaleFC_F3_27.cu
@@ -23,8 +23,8 @@ __global__ void scaleFC_comp_D3Q27F3_2018(real* DC,
unsigned int* neighborFX,
unsigned int* neighborFY,
unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
+ unsigned long long numberOfLBnodesCoarse,
+ unsigned long long numberOfLBnodesFine,
bool isEvenTimestep,
unsigned int* posC,
unsigned int* posFSWB,
@@ -44,33 +44,33 @@ __global__ void scaleFC_comp_D3Q27F3_2018(real* DC,
*f000source, *fMMMsource, *fMMPsource, *fMPPsource, *fMPMsource, *fPPMsource, *fPPPsource, *fPMPsource, *fPMMsource;
- fP00source = &DF[DIR_P00 *size_MatF];
- fM00source = &DF[DIR_M00 *size_MatF];
- f0P0source = &DF[DIR_0P0 *size_MatF];
- f0M0source = &DF[DIR_0M0 *size_MatF];
- f00Psource = &DF[DIR_00P *size_MatF];
- f00Msource = &DF[DIR_00M *size_MatF];
- fPP0source = &DF[DIR_PP0 *size_MatF];
- fMM0source = &DF[DIR_MM0 *size_MatF];
- fPM0source = &DF[DIR_PM0 *size_MatF];
- fMP0source = &DF[DIR_MP0 *size_MatF];
- fP0Psource = &DF[DIR_P0P *size_MatF];
- fM0Msource = &DF[DIR_M0M *size_MatF];
- fP0Msource = &DF[DIR_P0M *size_MatF];
- fM0Psource = &DF[DIR_M0P *size_MatF];
- f0PPsource = &DF[DIR_0PP *size_MatF];
- f0MMsource = &DF[DIR_0MM *size_MatF];
- f0PMsource = &DF[DIR_0PM *size_MatF];
- f0MPsource = &DF[DIR_0MP *size_MatF];
- f000source = &DF[DIR_000*size_MatF];
- fMMMsource = &DF[DIR_MMM *size_MatF];
- fMMPsource = &DF[DIR_MMP *size_MatF];
- fMPPsource = &DF[DIR_MPP *size_MatF];
- fMPMsource = &DF[DIR_MPM *size_MatF];
- fPPMsource = &DF[DIR_PPM *size_MatF];
- fPPPsource = &DF[DIR_PPP *size_MatF];
- fPMPsource = &DF[DIR_PMP *size_MatF];
- fPMMsource = &DF[DIR_PMM *size_MatF];
+ fP00source = &DF[DIR_P00 * numberOfLBnodesFine];
+ fM00source = &DF[DIR_M00 * numberOfLBnodesFine];
+ f0P0source = &DF[DIR_0P0 * numberOfLBnodesFine];
+ f0M0source = &DF[DIR_0M0 * numberOfLBnodesFine];
+ f00Psource = &DF[DIR_00P * numberOfLBnodesFine];
+ f00Msource = &DF[DIR_00M * numberOfLBnodesFine];
+ fPP0source = &DF[DIR_PP0 * numberOfLBnodesFine];
+ fMM0source = &DF[DIR_MM0 * numberOfLBnodesFine];
+ fPM0source = &DF[DIR_PM0 * numberOfLBnodesFine];
+ fMP0source = &DF[DIR_MP0 * numberOfLBnodesFine];
+ fP0Psource = &DF[DIR_P0P * numberOfLBnodesFine];
+ fM0Msource = &DF[DIR_M0M * numberOfLBnodesFine];
+ fP0Msource = &DF[DIR_P0M * numberOfLBnodesFine];
+ fM0Psource = &DF[DIR_M0P * numberOfLBnodesFine];
+ f0PPsource = &DF[DIR_0PP * numberOfLBnodesFine];
+ f0MMsource = &DF[DIR_0MM * numberOfLBnodesFine];
+ f0PMsource = &DF[DIR_0PM * numberOfLBnodesFine];
+ f0MPsource = &DF[DIR_0MP * numberOfLBnodesFine];
+ f000source = &DF[DIR_000 * numberOfLBnodesFine];
+ fMMMsource = &DF[DIR_MMM * numberOfLBnodesFine];
+ fMMPsource = &DF[DIR_MMP * numberOfLBnodesFine];
+ fMPPsource = &DF[DIR_MPP * numberOfLBnodesFine];
+ fMPMsource = &DF[DIR_MPM * numberOfLBnodesFine];
+ fPPMsource = &DF[DIR_PPM * numberOfLBnodesFine];
+ fPPPsource = &DF[DIR_PPP * numberOfLBnodesFine];
+ fPMPsource = &DF[DIR_PMP * numberOfLBnodesFine];
+ fPMMsource = &DF[DIR_PMM * numberOfLBnodesFine];
real
*fP00dest, *fM00dest, *f0P0dest, *f0M0dest, *f00Pdest, *f00Mdest, *fPP0dest, *fMM0dest, *fPM0dest,
@@ -79,83 +79,83 @@ __global__ void scaleFC_comp_D3Q27F3_2018(real* DC,
if (isEvenTimestep==true)
{
- fP00dest = &DC[DIR_P00 *size_MatC];
- fM00dest = &DC[DIR_M00 *size_MatC];
- f0P0dest = &DC[DIR_0P0 *size_MatC];
- f0M0dest = &DC[DIR_0M0 *size_MatC];
- f00Pdest = &DC[DIR_00P *size_MatC];
- f00Mdest = &DC[DIR_00M *size_MatC];
- fPP0dest = &DC[DIR_PP0 *size_MatC];
- fMM0dest = &DC[DIR_MM0 *size_MatC];
- fPM0dest = &DC[DIR_PM0 *size_MatC];
- fMP0dest = &DC[DIR_MP0 *size_MatC];
- fP0Pdest = &DC[DIR_P0P *size_MatC];
- fM0Mdest = &DC[DIR_M0M *size_MatC];
- fP0Mdest = &DC[DIR_P0M *size_MatC];
- fM0Pdest = &DC[DIR_M0P *size_MatC];
- f0PPdest = &DC[DIR_0PP *size_MatC];
- f0MMdest = &DC[DIR_0MM *size_MatC];
- f0PMdest = &DC[DIR_0PM *size_MatC];
- f0MPdest = &DC[DIR_0MP *size_MatC];
- f000dest = &DC[DIR_000*size_MatC];
- fMMMdest = &DC[DIR_MMM *size_MatC];
- fMMPdest = &DC[DIR_MMP *size_MatC];
- fMPPdest = &DC[DIR_MPP *size_MatC];
- fMPMdest = &DC[DIR_MPM *size_MatC];
- fPPMdest = &DC[DIR_PPM *size_MatC];
- fPPPdest = &DC[DIR_PPP *size_MatC];
- fPMPdest = &DC[DIR_PMP *size_MatC];
- fPMMdest = &DC[DIR_PMM *size_MatC];
+ fP00dest = &DC[DIR_P00 * numberOfLBnodesCoarse];
+ fM00dest = &DC[DIR_M00 * numberOfLBnodesCoarse];
+ f0P0dest = &DC[DIR_0P0 * numberOfLBnodesCoarse];
+ f0M0dest = &DC[DIR_0M0 * numberOfLBnodesCoarse];
+ f00Pdest = &DC[DIR_00P * numberOfLBnodesCoarse];
+ f00Mdest = &DC[DIR_00M * numberOfLBnodesCoarse];
+ fPP0dest = &DC[DIR_PP0 * numberOfLBnodesCoarse];
+ fMM0dest = &DC[DIR_MM0 * numberOfLBnodesCoarse];
+ fPM0dest = &DC[DIR_PM0 * numberOfLBnodesCoarse];
+ fMP0dest = &DC[DIR_MP0 * numberOfLBnodesCoarse];
+ fP0Pdest = &DC[DIR_P0P * numberOfLBnodesCoarse];
+ fM0Mdest = &DC[DIR_M0M * numberOfLBnodesCoarse];
+ fP0Mdest = &DC[DIR_P0M * numberOfLBnodesCoarse];
+ fM0Pdest = &DC[DIR_M0P * numberOfLBnodesCoarse];
+ f0PPdest = &DC[DIR_0PP * numberOfLBnodesCoarse];
+ f0MMdest = &DC[DIR_0MM * numberOfLBnodesCoarse];
+ f0PMdest = &DC[DIR_0PM * numberOfLBnodesCoarse];
+ f0MPdest = &DC[DIR_0MP * numberOfLBnodesCoarse];
+ f000dest = &DC[DIR_000 * numberOfLBnodesCoarse];
+ fMMMdest = &DC[DIR_MMM * numberOfLBnodesCoarse];
+ fMMPdest = &DC[DIR_MMP * numberOfLBnodesCoarse];
+ fMPPdest = &DC[DIR_MPP * numberOfLBnodesCoarse];
+ fMPMdest = &DC[DIR_MPM * numberOfLBnodesCoarse];
+ fPPMdest = &DC[DIR_PPM * numberOfLBnodesCoarse];
+ fPPPdest = &DC[DIR_PPP * numberOfLBnodesCoarse];
+ fPMPdest = &DC[DIR_PMP * numberOfLBnodesCoarse];
+ fPMMdest = &DC[DIR_PMM * numberOfLBnodesCoarse];
}
else
{
- fP00dest = &DC[DIR_M00 *size_MatC];
- fM00dest = &DC[DIR_P00 *size_MatC];
- f0P0dest = &DC[DIR_0M0 *size_MatC];
- f0M0dest = &DC[DIR_0P0 *size_MatC];
- f00Pdest = &DC[DIR_00M *size_MatC];
- f00Mdest = &DC[DIR_00P *size_MatC];
- fPP0dest = &DC[DIR_MM0 *size_MatC];
- fMM0dest = &DC[DIR_PP0 *size_MatC];
- fPM0dest = &DC[DIR_MP0 *size_MatC];
- fMP0dest = &DC[DIR_PM0 *size_MatC];
- fP0Pdest = &DC[DIR_M0M *size_MatC];
- fM0Mdest = &DC[DIR_P0P *size_MatC];
- fP0Mdest = &DC[DIR_M0P *size_MatC];
- fM0Pdest = &DC[DIR_P0M *size_MatC];
- f0PPdest = &DC[DIR_0MM *size_MatC];
- f0MMdest = &DC[DIR_0PP *size_MatC];
- f0PMdest = &DC[DIR_0MP *size_MatC];
- f0MPdest = &DC[DIR_0PM *size_MatC];
- f000dest = &DC[DIR_000*size_MatC];
- fMMMdest = &DC[DIR_PPP *size_MatC];
- fMMPdest = &DC[DIR_PPM *size_MatC];
- fMPPdest = &DC[DIR_PMM *size_MatC];
- fMPMdest = &DC[DIR_PMP *size_MatC];
- fPPMdest = &DC[DIR_MMP *size_MatC];
- fPPPdest = &DC[DIR_MMM *size_MatC];
- fPMPdest = &DC[DIR_MPM *size_MatC];
- fPMMdest = &DC[DIR_MPP *size_MatC];
+ fP00dest = &DC[DIR_M00 * numberOfLBnodesCoarse];
+ fM00dest = &DC[DIR_P00 * numberOfLBnodesCoarse];
+ f0P0dest = &DC[DIR_0M0 * numberOfLBnodesCoarse];
+ f0M0dest = &DC[DIR_0P0 * numberOfLBnodesCoarse];
+ f00Pdest = &DC[DIR_00M * numberOfLBnodesCoarse];
+ f00Mdest = &DC[DIR_00P * numberOfLBnodesCoarse];
+ fPP0dest = &DC[DIR_MM0 * numberOfLBnodesCoarse];
+ fMM0dest = &DC[DIR_PP0 * numberOfLBnodesCoarse];
+ fPM0dest = &DC[DIR_MP0 * numberOfLBnodesCoarse];
+ fMP0dest = &DC[DIR_PM0 * numberOfLBnodesCoarse];
+ fP0Pdest = &DC[DIR_M0M * numberOfLBnodesCoarse];
+ fM0Mdest = &DC[DIR_P0P * numberOfLBnodesCoarse];
+ fP0Mdest = &DC[DIR_M0P * numberOfLBnodesCoarse];
+ fM0Pdest = &DC[DIR_P0M * numberOfLBnodesCoarse];
+ f0PPdest = &DC[DIR_0MM * numberOfLBnodesCoarse];
+ f0MMdest = &DC[DIR_0PP * numberOfLBnodesCoarse];
+ f0PMdest = &DC[DIR_0MP * numberOfLBnodesCoarse];
+ f0MPdest = &DC[DIR_0PM * numberOfLBnodesCoarse];
+ f000dest = &DC[DIR_000 * numberOfLBnodesCoarse];
+ fMMMdest = &DC[DIR_PPP * numberOfLBnodesCoarse];
+ fMMPdest = &DC[DIR_PPM * numberOfLBnodesCoarse];
+ fMPPdest = &DC[DIR_PMM * numberOfLBnodesCoarse];
+ fMPMdest = &DC[DIR_PMP * numberOfLBnodesCoarse];
+ fPPMdest = &DC[DIR_MMP * numberOfLBnodesCoarse];
+ fPPPdest = &DC[DIR_MMM * numberOfLBnodesCoarse];
+ fPMPdest = &DC[DIR_MPM * numberOfLBnodesCoarse];
+ fPMMdest = &DC[DIR_MPP * numberOfLBnodesCoarse];
}
Distributions6 G;
if (isEvenTimestep == true)
{
- G.g[DIR_P00] = &G6[DIR_P00 *size_MatC];
- G.g[DIR_M00] = &G6[DIR_M00 *size_MatC];
- G.g[DIR_0P0] = &G6[DIR_0P0 *size_MatC];
- G.g[DIR_0M0] = &G6[DIR_0M0 *size_MatC];
- G.g[DIR_00P] = &G6[DIR_00P *size_MatC];
- G.g[DIR_00M] = &G6[DIR_00M *size_MatC];
+ G.g[DIR_P00] = &G6[DIR_P00 * numberOfLBnodesCoarse];
+ G.g[DIR_M00] = &G6[DIR_M00 * numberOfLBnodesCoarse];
+ G.g[DIR_0P0] = &G6[DIR_0P0 * numberOfLBnodesCoarse];
+ G.g[DIR_0M0] = &G6[DIR_0M0 * numberOfLBnodesCoarse];
+ G.g[DIR_00P] = &G6[DIR_00P * numberOfLBnodesCoarse];
+ G.g[DIR_00M] = &G6[DIR_00M * numberOfLBnodesCoarse];
}
else
{
- G.g[DIR_M00] = &G6[DIR_P00 *size_MatC];
- G.g[DIR_P00] = &G6[DIR_M00 *size_MatC];
- G.g[DIR_0M0] = &G6[DIR_0P0 *size_MatC];
- G.g[DIR_0P0] = &G6[DIR_0M0 *size_MatC];
- G.g[DIR_00M] = &G6[DIR_00P *size_MatC];
- G.g[DIR_00P] = &G6[DIR_00M *size_MatC];
+ G.g[DIR_M00] = &G6[DIR_P00 * numberOfLBnodesCoarse];
+ G.g[DIR_P00] = &G6[DIR_M00 * numberOfLBnodesCoarse];
+ G.g[DIR_0M0] = &G6[DIR_0P0 * numberOfLBnodesCoarse];
+ G.g[DIR_0P0] = &G6[DIR_0M0 * numberOfLBnodesCoarse];
+ G.g[DIR_00M] = &G6[DIR_00P * numberOfLBnodesCoarse];
+ G.g[DIR_00P] = &G6[DIR_00M * numberOfLBnodesCoarse];
}
////////////////////////////////////////////////////////////////////////////////
@@ -1270,8 +1270,8 @@ __global__ void scaleFC_comp_D3Q27F3( real* DC,
unsigned int* neighborFX,
unsigned int* neighborFY,
unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
+ unsigned long long numberOfLBnodesCoarse,
+ unsigned long long numberOfLBnodesFine,
bool isEvenTimestep,
unsigned int* posC,
unsigned int* posFSWB,
@@ -1291,33 +1291,33 @@ __global__ void scaleFC_comp_D3Q27F3( real* DC,
*f000source, *fMMMsource, *fMMPsource, *fMPPsource, *fMPMsource, *fPPMsource, *fPPPsource, *fPMPsource, *fPMMsource;
- fP00source = &DF[DIR_P00 *size_MatF];
- fM00source = &DF[DIR_M00 *size_MatF];
- f0P0source = &DF[DIR_0P0 *size_MatF];
- f0M0source = &DF[DIR_0M0 *size_MatF];
- f00Psource = &DF[DIR_00P *size_MatF];
- f00Msource = &DF[DIR_00M *size_MatF];
- fPP0source = &DF[DIR_PP0 *size_MatF];
- fMM0source = &DF[DIR_MM0 *size_MatF];
- fPM0source = &DF[DIR_PM0 *size_MatF];
- fMP0source = &DF[DIR_MP0 *size_MatF];
- fP0Psource = &DF[DIR_P0P *size_MatF];
- fM0Msource = &DF[DIR_M0M *size_MatF];
- fP0Msource = &DF[DIR_P0M *size_MatF];
- fM0Psource = &DF[DIR_M0P *size_MatF];
- f0PPsource = &DF[DIR_0PP *size_MatF];
- f0MMsource = &DF[DIR_0MM *size_MatF];
- f0PMsource = &DF[DIR_0PM *size_MatF];
- f0MPsource = &DF[DIR_0MP *size_MatF];
- f000source = &DF[DIR_000*size_MatF];
- fMMMsource = &DF[DIR_MMM *size_MatF];
- fMMPsource = &DF[DIR_MMP *size_MatF];
- fMPPsource = &DF[DIR_MPP *size_MatF];
- fMPMsource = &DF[DIR_MPM *size_MatF];
- fPPMsource = &DF[DIR_PPM *size_MatF];
- fPPPsource = &DF[DIR_PPP *size_MatF];
- fPMPsource = &DF[DIR_PMP *size_MatF];
- fPMMsource = &DF[DIR_PMM *size_MatF];
+ fP00source = &DF[DIR_P00 * numberOfLBnodesFine];
+ fM00source = &DF[DIR_M00 * numberOfLBnodesFine];
+ f0P0source = &DF[DIR_0P0 * numberOfLBnodesFine];
+ f0M0source = &DF[DIR_0M0 * numberOfLBnodesFine];
+ f00Psource = &DF[DIR_00P * numberOfLBnodesFine];
+ f00Msource = &DF[DIR_00M * numberOfLBnodesFine];
+ fPP0source = &DF[DIR_PP0 * numberOfLBnodesFine];
+ fMM0source = &DF[DIR_MM0 * numberOfLBnodesFine];
+ fPM0source = &DF[DIR_PM0 * numberOfLBnodesFine];
+ fMP0source = &DF[DIR_MP0 * numberOfLBnodesFine];
+ fP0Psource = &DF[DIR_P0P * numberOfLBnodesFine];
+ fM0Msource = &DF[DIR_M0M * numberOfLBnodesFine];
+ fP0Msource = &DF[DIR_P0M * numberOfLBnodesFine];
+ fM0Psource = &DF[DIR_M0P * numberOfLBnodesFine];
+ f0PPsource = &DF[DIR_0PP * numberOfLBnodesFine];
+ f0MMsource = &DF[DIR_0MM * numberOfLBnodesFine];
+ f0PMsource = &DF[DIR_0PM * numberOfLBnodesFine];
+ f0MPsource = &DF[DIR_0MP * numberOfLBnodesFine];
+ f000source = &DF[DIR_000 * numberOfLBnodesFine];
+ fMMMsource = &DF[DIR_MMM * numberOfLBnodesFine];
+ fMMPsource = &DF[DIR_MMP * numberOfLBnodesFine];
+ fMPPsource = &DF[DIR_MPP * numberOfLBnodesFine];
+ fMPMsource = &DF[DIR_MPM * numberOfLBnodesFine];
+ fPPMsource = &DF[DIR_PPM * numberOfLBnodesFine];
+ fPPPsource = &DF[DIR_PPP * numberOfLBnodesFine];
+ fPMPsource = &DF[DIR_PMP * numberOfLBnodesFine];
+ fPMMsource = &DF[DIR_PMM * numberOfLBnodesFine];
real
*fP00dest, *fM00dest, *f0P0dest, *f0M0dest, *f00Pdest, *f00Mdest, *fPP0dest, *fMM0dest, *fPM0dest,
@@ -1326,83 +1326,83 @@ __global__ void scaleFC_comp_D3Q27F3( real* DC,
if (isEvenTimestep==true)
{
- fP00dest = &DC[DIR_P00 *size_MatC];
- fM00dest = &DC[DIR_M00 *size_MatC];
- f0P0dest = &DC[DIR_0P0 *size_MatC];
- f0M0dest = &DC[DIR_0M0 *size_MatC];
- f00Pdest = &DC[DIR_00P *size_MatC];
- f00Mdest = &DC[DIR_00M *size_MatC];
- fPP0dest = &DC[DIR_PP0 *size_MatC];
- fMM0dest = &DC[DIR_MM0 *size_MatC];
- fPM0dest = &DC[DIR_PM0 *size_MatC];
- fMP0dest = &DC[DIR_MP0 *size_MatC];
- fP0Pdest = &DC[DIR_P0P *size_MatC];
- fM0Mdest = &DC[DIR_M0M *size_MatC];
- fP0Mdest = &DC[DIR_P0M *size_MatC];
- fM0Pdest = &DC[DIR_M0P *size_MatC];
- f0PPdest = &DC[DIR_0PP *size_MatC];
- f0MMdest = &DC[DIR_0MM *size_MatC];
- f0PMdest = &DC[DIR_0PM *size_MatC];
- f0MPdest = &DC[DIR_0MP *size_MatC];
- f000dest = &DC[DIR_000*size_MatC];
- fMMMdest = &DC[DIR_MMM *size_MatC];
- fMMPdest = &DC[DIR_MMP *size_MatC];
- fMPPdest = &DC[DIR_MPP *size_MatC];
- fMPMdest = &DC[DIR_MPM *size_MatC];
- fPPMdest = &DC[DIR_PPM *size_MatC];
- fPPPdest = &DC[DIR_PPP *size_MatC];
- fPMPdest = &DC[DIR_PMP *size_MatC];
- fPMMdest = &DC[DIR_PMM *size_MatC];
+ fP00dest = &DC[DIR_P00 * numberOfLBnodesCoarse];
+ fM00dest = &DC[DIR_M00 * numberOfLBnodesCoarse];
+ f0P0dest = &DC[DIR_0P0 * numberOfLBnodesCoarse];
+ f0M0dest = &DC[DIR_0M0 * numberOfLBnodesCoarse];
+ f00Pdest = &DC[DIR_00P * numberOfLBnodesCoarse];
+ f00Mdest = &DC[DIR_00M * numberOfLBnodesCoarse];
+ fPP0dest = &DC[DIR_PP0 * numberOfLBnodesCoarse];
+ fMM0dest = &DC[DIR_MM0 * numberOfLBnodesCoarse];
+ fPM0dest = &DC[DIR_PM0 * numberOfLBnodesCoarse];
+ fMP0dest = &DC[DIR_MP0 * numberOfLBnodesCoarse];
+ fP0Pdest = &DC[DIR_P0P * numberOfLBnodesCoarse];
+ fM0Mdest = &DC[DIR_M0M * numberOfLBnodesCoarse];
+ fP0Mdest = &DC[DIR_P0M * numberOfLBnodesCoarse];
+ fM0Pdest = &DC[DIR_M0P * numberOfLBnodesCoarse];
+ f0PPdest = &DC[DIR_0PP * numberOfLBnodesCoarse];
+ f0MMdest = &DC[DIR_0MM * numberOfLBnodesCoarse];
+ f0PMdest = &DC[DIR_0PM * numberOfLBnodesCoarse];
+ f0MPdest = &DC[DIR_0MP * numberOfLBnodesCoarse];
+ f000dest = &DC[DIR_000 * numberOfLBnodesCoarse];
+ fMMMdest = &DC[DIR_MMM * numberOfLBnodesCoarse];
+ fMMPdest = &DC[DIR_MMP * numberOfLBnodesCoarse];
+ fMPPdest = &DC[DIR_MPP * numberOfLBnodesCoarse];
+ fMPMdest = &DC[DIR_MPM * numberOfLBnodesCoarse];
+ fPPMdest = &DC[DIR_PPM * numberOfLBnodesCoarse];
+ fPPPdest = &DC[DIR_PPP * numberOfLBnodesCoarse];
+ fPMPdest = &DC[DIR_PMP * numberOfLBnodesCoarse];
+ fPMMdest = &DC[DIR_PMM * numberOfLBnodesCoarse];
}
else
{
- fP00dest = &DC[DIR_M00 *size_MatC];
- fM00dest = &DC[DIR_P00 *size_MatC];
- f0P0dest = &DC[DIR_0M0 *size_MatC];
- f0M0dest = &DC[DIR_0P0 *size_MatC];
- f00Pdest = &DC[DIR_00M *size_MatC];
- f00Mdest = &DC[DIR_00P *size_MatC];
- fPP0dest = &DC[DIR_MM0 *size_MatC];
- fMM0dest = &DC[DIR_PP0 *size_MatC];
- fPM0dest = &DC[DIR_MP0 *size_MatC];
- fMP0dest = &DC[DIR_PM0 *size_MatC];
- fP0Pdest = &DC[DIR_M0M *size_MatC];
- fM0Mdest = &DC[DIR_P0P *size_MatC];
- fP0Mdest = &DC[DIR_M0P *size_MatC];
- fM0Pdest = &DC[DIR_P0M *size_MatC];
- f0PPdest = &DC[DIR_0MM *size_MatC];
- f0MMdest = &DC[DIR_0PP *size_MatC];
- f0PMdest = &DC[DIR_0MP *size_MatC];
- f0MPdest = &DC[DIR_0PM *size_MatC];
- f000dest = &DC[DIR_000*size_MatC];
- fMMMdest = &DC[DIR_PPP *size_MatC];
- fMMPdest = &DC[DIR_PPM *size_MatC];
- fMPPdest = &DC[DIR_PMM *size_MatC];
- fMPMdest = &DC[DIR_PMP *size_MatC];
- fPPMdest = &DC[DIR_MMP *size_MatC];
- fPPPdest = &DC[DIR_MMM *size_MatC];
- fPMPdest = &DC[DIR_MPM *size_MatC];
- fPMMdest = &DC[DIR_MPP *size_MatC];
+ fP00dest = &DC[DIR_M00 * numberOfLBnodesCoarse];
+ fM00dest = &DC[DIR_P00 * numberOfLBnodesCoarse];
+ f0P0dest = &DC[DIR_0M0 * numberOfLBnodesCoarse];
+ f0M0dest = &DC[DIR_0P0 * numberOfLBnodesCoarse];
+ f00Pdest = &DC[DIR_00M * numberOfLBnodesCoarse];
+ f00Mdest = &DC[DIR_00P * numberOfLBnodesCoarse];
+ fPP0dest = &DC[DIR_MM0 * numberOfLBnodesCoarse];
+ fMM0dest = &DC[DIR_PP0 * numberOfLBnodesCoarse];
+ fPM0dest = &DC[DIR_MP0 * numberOfLBnodesCoarse];
+ fMP0dest = &DC[DIR_PM0 * numberOfLBnodesCoarse];
+ fP0Pdest = &DC[DIR_M0M * numberOfLBnodesCoarse];
+ fM0Mdest = &DC[DIR_P0P * numberOfLBnodesCoarse];
+ fP0Mdest = &DC[DIR_M0P * numberOfLBnodesCoarse];
+ fM0Pdest = &DC[DIR_P0M * numberOfLBnodesCoarse];
+ f0PPdest = &DC[DIR_0MM * numberOfLBnodesCoarse];
+ f0MMdest = &DC[DIR_0PP * numberOfLBnodesCoarse];
+ f0PMdest = &DC[DIR_0MP * numberOfLBnodesCoarse];
+ f0MPdest = &DC[DIR_0PM * numberOfLBnodesCoarse];
+ f000dest = &DC[DIR_000 * numberOfLBnodesCoarse];
+ fMMMdest = &DC[DIR_PPP * numberOfLBnodesCoarse];
+ fMMPdest = &DC[DIR_PPM * numberOfLBnodesCoarse];
+ fMPPdest = &DC[DIR_PMM * numberOfLBnodesCoarse];
+ fMPMdest = &DC[DIR_PMP * numberOfLBnodesCoarse];
+ fPPMdest = &DC[DIR_MMP * numberOfLBnodesCoarse];
+ fPPPdest = &DC[DIR_MMM * numberOfLBnodesCoarse];
+ fPMPdest = &DC[DIR_MPM * numberOfLBnodesCoarse];
+ fPMMdest = &DC[DIR_MPP * numberOfLBnodesCoarse];
}
Distributions6 G;
if (isEvenTimestep == true)
{
- G.g[DIR_P00] = &G6[DIR_P00 *size_MatC];
- G.g[DIR_M00] = &G6[DIR_M00 *size_MatC];
- G.g[DIR_0P0] = &G6[DIR_0P0 *size_MatC];
- G.g[DIR_0M0] = &G6[DIR_0M0 *size_MatC];
- G.g[DIR_00P] = &G6[DIR_00P *size_MatC];
- G.g[DIR_00M] = &G6[DIR_00M *size_MatC];
+ G.g[DIR_P00] = &G6[DIR_P00 * numberOfLBnodesCoarse];
+ G.g[DIR_M00] = &G6[DIR_M00 * numberOfLBnodesCoarse];
+ G.g[DIR_0P0] = &G6[DIR_0P0 * numberOfLBnodesCoarse];
+ G.g[DIR_0M0] = &G6[DIR_0M0 * numberOfLBnodesCoarse];
+ G.g[DIR_00P] = &G6[DIR_00P * numberOfLBnodesCoarse];
+ G.g[DIR_00M] = &G6[DIR_00M * numberOfLBnodesCoarse];
}
else
{
- G.g[DIR_M00] = &G6[DIR_P00 *size_MatC];
- G.g[DIR_P00] = &G6[DIR_M00 *size_MatC];
- G.g[DIR_0M0] = &G6[DIR_0P0 *size_MatC];
- G.g[DIR_0P0] = &G6[DIR_0M0 *size_MatC];
- G.g[DIR_00M] = &G6[DIR_00P *size_MatC];
- G.g[DIR_00P] = &G6[DIR_00M *size_MatC];
+ G.g[DIR_M00] = &G6[DIR_P00 * numberOfLBnodesCoarse];
+ G.g[DIR_P00] = &G6[DIR_M00 * numberOfLBnodesCoarse];
+ G.g[DIR_0M0] = &G6[DIR_0P0 * numberOfLBnodesCoarse];
+ G.g[DIR_0P0] = &G6[DIR_0M0 * numberOfLBnodesCoarse];
+ G.g[DIR_00M] = &G6[DIR_00P * numberOfLBnodesCoarse];
+ G.g[DIR_00P] = &G6[DIR_00M * numberOfLBnodesCoarse];
}
////////////////////////////////////////////////////////////////////////////////
diff --git a/src/gpu/VirtualFluids_GPU/GPU/GridScaling/scaleCF_compressible.cu b/src/gpu/VirtualFluids_GPU/GPU/GridScaling/scaleCF_compressible.cu
index f4160b89c047a7e6244a5579baae03d30b3c89cb..0724002cffa3a47820664851ffefd1c35dbe0235 100644
--- a/src/gpu/VirtualFluids_GPU/GPU/GridScaling/scaleCF_compressible.cu
+++ b/src/gpu/VirtualFluids_GPU/GPU/GridScaling/scaleCF_compressible.cu
@@ -32,12 +32,13 @@
//=======================================================================================
#include "DataTypes.h"
-#include "Kernel/Utilities/DistributionHelper.cuh"
-#include "Kernel/Utilities/ChimeraTransformation.h"
-#include "Kernel/Utilities/ScalingHelperFunctions.h"
+#include "LBM/GPUHelperFunctions/KernelUtilities.h"
+#include "LBM/GPUHelperFunctions/ChimeraTransformation.h"
+#include "LBM/GPUHelperFunctions/ScalingUtilities.h"
using namespace vf::lbm::constant;
using namespace vf::lbm::dir;
+using namespace vf::gpu;
//////////////////////////////////////////////////////////////////////////
//! \brief Calculate the interpolated distributions on the fine destination nodes
@@ -226,8 +227,8 @@ __global__ void scaleCF_compressible(
unsigned int* neighborXfine,
unsigned int* neighborYfine,
unsigned int* neighborZfine,
- unsigned int numberOfLBnodesCoarse,
- unsigned int numberOfLBnodesFine,
+ unsigned long long numberOfLBnodesCoarse,
+ unsigned long long numberOfLBnodesFine,
bool isEvenTimestep,
unsigned int* indicesCoarseMMM,
unsigned int* indicesFineMMM,
@@ -237,13 +238,13 @@ __global__ void scaleCF_compressible(
OffCF offsetCF)
{
////////////////////////////////////////////////////////////////////////////////
- //! - Get the thread index coordinates from threadId_100, blockId_100, blockDim and gridDim.
+ //! - Get the node index coordinates from threadId_100, blockId_100, blockDim and gridDim.
//!
- const unsigned k_thread = vf::gpu::getNodeIndex();
+ const unsigned nodeIndex = getNodeIndex();
//////////////////////////////////////////////////////////////////////////
//! - Return for non-interface node
- if (k_thread >= numberOfInterfaceNodes)
+ if (nodeIndex >= numberOfInterfaceNodes)
return;
//////////////////////////////////////////////////////////////////////////
@@ -252,8 +253,9 @@ __global__ void scaleCF_compressible(
//! href="https://doi.org/10.3390/computation5020019"><b>[ M. Geier et al. (2017),
//! DOI:10.3390/computation5020019 ]</b></a>
//!
- Distributions27 distFine = vf::gpu::getDistributionReferences27(distributionsFine, numberOfLBnodesFine, true);
- Distributions27 distCoarse = vf::gpu::getDistributionReferences27(distributionsCoarse, numberOfLBnodesCoarse, isEvenTimestep);
+ Distributions27 distFine, distCoarse;
+ getPointersToDistributions(distFine, distributionsFine, numberOfLBnodesFine, true);
+ getPointersToDistributions(distCoarse, distributionsCoarse, numberOfLBnodesCoarse, isEvenTimestep);
////////////////////////////////////////////////////////////////////////////////
//! - declare local variables for source nodes
@@ -289,7 +291,7 @@ __global__ void scaleCF_compressible(
// source node BSW = MMM
////////////////////////////////////////////////////////////////////////////////
// index of the base node and its neighbors
- unsigned int k_base_000 = indicesCoarseMMM[k_thread];
+ unsigned int k_base_000 = indicesCoarseMMM[nodeIndex];
unsigned int k_base_M00 = neighborXcoarse [k_base_000];
unsigned int k_base_0M0 = neighborYcoarse [k_base_000];
unsigned int k_base_00M = neighborZcoarse [k_base_000];
@@ -452,119 +454,240 @@ __global__ void scaleCF_compressible(
real c_000, c_100, c_010, c_001, c_200, c_020, c_002, c_110, c_101, c_011, c_111;
real d_000, d_100, d_010, d_001, d_110, d_101, d_011, d_111;
- a_000 = (-kxxMyyFromfcNEQ_PPM - kxxMyyFromfcNEQ_PPP + kxxMyyFromfcNEQ_MPM + kxxMyyFromfcNEQ_MPP -
- kxxMyyFromfcNEQ_PMM - kxxMyyFromfcNEQ_PMP + kxxMyyFromfcNEQ_MMM + kxxMyyFromfcNEQ_MMP -
- kxxMzzFromfcNEQ_PPM - kxxMzzFromfcNEQ_PPP + kxxMzzFromfcNEQ_MPM + kxxMzzFromfcNEQ_MPP -
- kxxMzzFromfcNEQ_PMM - kxxMzzFromfcNEQ_PMP + kxxMzzFromfcNEQ_MMM + kxxMzzFromfcNEQ_MMP -
- c2o1 * kxyFromfcNEQ_PPM - c2o1 * kxyFromfcNEQ_PPP - c2o1 * kxyFromfcNEQ_MPM - c2o1 * kxyFromfcNEQ_MPP +
- c2o1 * kxyFromfcNEQ_PMM + c2o1 * kxyFromfcNEQ_PMP + c2o1 * kxyFromfcNEQ_MMM + c2o1 * kxyFromfcNEQ_MMP +
- c2o1 * kxzFromfcNEQ_PPM - c2o1 * kxzFromfcNEQ_PPP + c2o1 * kxzFromfcNEQ_MPM - c2o1 * kxzFromfcNEQ_MPP +
- c2o1 * kxzFromfcNEQ_PMM - c2o1 * kxzFromfcNEQ_PMP + c2o1 * kxzFromfcNEQ_MMM - c2o1 * kxzFromfcNEQ_MMP +
- c8o1 * vx1_PPM + c8o1 * vx1_PPP + c8o1 * vx1_MPM + c8o1 * vx1_MPP + c8o1 * vx1_PMM + c8o1 * vx1_PMP +
- c8o1 * vx1_MMM + c8o1 * vx1_MMP + c2o1 * vx2_PPM + c2o1 * vx2_PPP - c2o1 * vx2_MPM - c2o1 * vx2_MPP -
- c2o1 * vx2_PMM - c2o1 * vx2_PMP + c2o1 * vx2_MMM + c2o1 * vx2_MMP - c2o1 * vx3_PPM + c2o1 * vx3_PPP +
- c2o1 * vx3_MPM - c2o1 * vx3_MPP - c2o1 * vx3_PMM + c2o1 * vx3_PMP + c2o1 * vx3_MMM - c2o1 * vx3_MMP) /
- c64o1;
- b_000 = (c2o1 * kxxMyyFromfcNEQ_PPM + c2o1 * kxxMyyFromfcNEQ_PPP + c2o1 * kxxMyyFromfcNEQ_MPM +
- c2o1 * kxxMyyFromfcNEQ_MPP - c2o1 * kxxMyyFromfcNEQ_PMM - c2o1 * kxxMyyFromfcNEQ_PMP -
- c2o1 * kxxMyyFromfcNEQ_MMM - c2o1 * kxxMyyFromfcNEQ_MMP - kxxMzzFromfcNEQ_PPM - kxxMzzFromfcNEQ_PPP -
- kxxMzzFromfcNEQ_MPM - kxxMzzFromfcNEQ_MPP + kxxMzzFromfcNEQ_PMM + kxxMzzFromfcNEQ_PMP +
- kxxMzzFromfcNEQ_MMM + kxxMzzFromfcNEQ_MMP - c2o1 * kxyFromfcNEQ_PPM - c2o1 * kxyFromfcNEQ_PPP +
- c2o1 * kxyFromfcNEQ_MPM + c2o1 * kxyFromfcNEQ_MPP - c2o1 * kxyFromfcNEQ_PMM - c2o1 * kxyFromfcNEQ_PMP +
- c2o1 * kxyFromfcNEQ_MMM + c2o1 * kxyFromfcNEQ_MMP + c2o1 * kyzFromfcNEQ_PPM - c2o1 * kyzFromfcNEQ_PPP +
- c2o1 * kyzFromfcNEQ_MPM - c2o1 * kyzFromfcNEQ_MPP + c2o1 * kyzFromfcNEQ_PMM - c2o1 * kyzFromfcNEQ_PMP +
- c2o1 * kyzFromfcNEQ_MMM - c2o1 * kyzFromfcNEQ_MMP + c2o1 * vx1_PPM + c2o1 * vx1_PPP - c2o1 * vx1_MPM -
- c2o1 * vx1_MPP - c2o1 * vx1_PMM - c2o1 * vx1_PMP + c2o1 * vx1_MMM + c2o1 * vx1_MMP + c8o1 * vx2_PPM +
- c8o1 * vx2_PPP + c8o1 * vx2_MPM + c8o1 * vx2_MPP + c8o1 * vx2_PMM + c8o1 * vx2_PMP + c8o1 * vx2_MMM +
- c8o1 * vx2_MMP - c2o1 * vx3_PPM + c2o1 * vx3_PPP - c2o1 * vx3_MPM + c2o1 * vx3_MPP + c2o1 * vx3_PMM -
- c2o1 * vx3_PMP + c2o1 * vx3_MMM - c2o1 * vx3_MMP) /
- c64o1;
- c_000 = (kxxMyyFromfcNEQ_PPM - kxxMyyFromfcNEQ_PPP + kxxMyyFromfcNEQ_MPM - kxxMyyFromfcNEQ_MPP +
- kxxMyyFromfcNEQ_PMM - kxxMyyFromfcNEQ_PMP + kxxMyyFromfcNEQ_MMM - kxxMyyFromfcNEQ_MMP -
- c2o1 * kxxMzzFromfcNEQ_PPM + c2o1 * kxxMzzFromfcNEQ_PPP - c2o1 * kxxMzzFromfcNEQ_MPM +
- c2o1 * kxxMzzFromfcNEQ_MPP - c2o1 * kxxMzzFromfcNEQ_PMM + c2o1 * kxxMzzFromfcNEQ_PMP -
- c2o1 * kxxMzzFromfcNEQ_MMM + c2o1 * kxxMzzFromfcNEQ_MMP - c2o1 * kxzFromfcNEQ_PPM -
- c2o1 * kxzFromfcNEQ_PPP + c2o1 * kxzFromfcNEQ_MPM + c2o1 * kxzFromfcNEQ_MPP - c2o1 * kxzFromfcNEQ_PMM -
- c2o1 * kxzFromfcNEQ_PMP + c2o1 * kxzFromfcNEQ_MMM + c2o1 * kxzFromfcNEQ_MMP - c2o1 * kyzFromfcNEQ_PPM -
- c2o1 * kyzFromfcNEQ_PPP - c2o1 * kyzFromfcNEQ_MPM - c2o1 * kyzFromfcNEQ_MPP + c2o1 * kyzFromfcNEQ_PMM +
- c2o1 * kyzFromfcNEQ_PMP + c2o1 * kyzFromfcNEQ_MMM + c2o1 * kyzFromfcNEQ_MMP - c2o1 * vx1_PPM +
- c2o1 * vx1_PPP + c2o1 * vx1_MPM - c2o1 * vx1_MPP - c2o1 * vx1_PMM + c2o1 * vx1_PMP + c2o1 * vx1_MMM -
- c2o1 * vx1_MMP - c2o1 * vx2_PPM + c2o1 * vx2_PPP - c2o1 * vx2_MPM + c2o1 * vx2_MPP + c2o1 * vx2_PMM -
- c2o1 * vx2_PMP + c2o1 * vx2_MMM - c2o1 * vx2_MMP + c8o1 * vx3_PPM + c8o1 * vx3_PPP + c8o1 * vx3_MPM +
- c8o1 * vx3_MPP + c8o1 * vx3_PMM + c8o1 * vx3_PMP + c8o1 * vx3_MMM + c8o1 * vx3_MMP) /
- c64o1;
- a_100 = (vx1_PPM + vx1_PPP - vx1_MPM - vx1_MPP + vx1_PMM + vx1_PMP - vx1_MMM - vx1_MMP) / c4o1;
- b_100 = (vx2_PPM + vx2_PPP - vx2_MPM - vx2_MPP + vx2_PMM + vx2_PMP - vx2_MMM - vx2_MMP) / c4o1;
- c_100 = (vx3_PPM + vx3_PPP - vx3_MPM - vx3_MPP + vx3_PMM + vx3_PMP - vx3_MMM - vx3_MMP) / c4o1;
- a_200 = (kxxMyyFromfcNEQ_PPM + kxxMyyFromfcNEQ_PPP - kxxMyyFromfcNEQ_MPM - kxxMyyFromfcNEQ_MPP +
- kxxMyyFromfcNEQ_PMM + kxxMyyFromfcNEQ_PMP - kxxMyyFromfcNEQ_MMM - kxxMyyFromfcNEQ_MMP +
- kxxMzzFromfcNEQ_PPM + kxxMzzFromfcNEQ_PPP - kxxMzzFromfcNEQ_MPM - kxxMzzFromfcNEQ_MPP +
- kxxMzzFromfcNEQ_PMM + kxxMzzFromfcNEQ_PMP - kxxMzzFromfcNEQ_MMM - kxxMzzFromfcNEQ_MMP + c2o1 * vx2_PPM +
- c2o1 * vx2_PPP - c2o1 * vx2_MPM - c2o1 * vx2_MPP - c2o1 * vx2_PMM - c2o1 * vx2_PMP + c2o1 * vx2_MMM +
- c2o1 * vx2_MMP - c2o1 * vx3_PPM + c2o1 * vx3_PPP + c2o1 * vx3_MPM - c2o1 * vx3_MPP - c2o1 * vx3_PMM +
- c2o1 * vx3_PMP + c2o1 * vx3_MMM - c2o1 * vx3_MMP) /
- c16o1;
- b_200 = (kxyFromfcNEQ_PPM + kxyFromfcNEQ_PPP - kxyFromfcNEQ_MPM - kxyFromfcNEQ_MPP + kxyFromfcNEQ_PMM +
- kxyFromfcNEQ_PMP - kxyFromfcNEQ_MMM - kxyFromfcNEQ_MMP - c2o1 * vx1_PPM - c2o1 * vx1_PPP +
- c2o1 * vx1_MPM + c2o1 * vx1_MPP + c2o1 * vx1_PMM + c2o1 * vx1_PMP - c2o1 * vx1_MMM - c2o1 * vx1_MMP) /
- c8o1;
- c_200 = (kxzFromfcNEQ_PPM + kxzFromfcNEQ_PPP - kxzFromfcNEQ_MPM - kxzFromfcNEQ_MPP + kxzFromfcNEQ_PMM +
- kxzFromfcNEQ_PMP - kxzFromfcNEQ_MMM - kxzFromfcNEQ_MMP + c2o1 * vx1_PPM - c2o1 * vx1_PPP -
- c2o1 * vx1_MPM + c2o1 * vx1_MPP + c2o1 * vx1_PMM - c2o1 * vx1_PMP - c2o1 * vx1_MMM + c2o1 * vx1_MMP) /
- c8o1;
- a_010 = (vx1_PPM + vx1_PPP + vx1_MPM + vx1_MPP - vx1_PMM - vx1_PMP - vx1_MMM - vx1_MMP) / c4o1;
- b_010 = (vx2_PPM + vx2_PPP + vx2_MPM + vx2_MPP - vx2_PMM - vx2_PMP - vx2_MMM - vx2_MMP) / c4o1;
- c_010 = (vx3_PPM + vx3_PPP + vx3_MPM + vx3_MPP - vx3_PMM - vx3_PMP - vx3_MMM - vx3_MMP) / c4o1;
- a_020 = (kxyFromfcNEQ_PPM + kxyFromfcNEQ_PPP + kxyFromfcNEQ_MPM + kxyFromfcNEQ_MPP - kxyFromfcNEQ_PMM -
- kxyFromfcNEQ_PMP - kxyFromfcNEQ_MMM - kxyFromfcNEQ_MMP - c2o1 * vx2_PPM - c2o1 * vx2_PPP +
- c2o1 * vx2_MPM + c2o1 * vx2_MPP + c2o1 * vx2_PMM + c2o1 * vx2_PMP - c2o1 * vx2_MMM - c2o1 * vx2_MMP) /
- c8o1;
- b_020 = (-c2o1 * kxxMyyFromfcNEQ_PPM - c2o1 * kxxMyyFromfcNEQ_PPP - c2o1 * kxxMyyFromfcNEQ_MPM -
- c2o1 * kxxMyyFromfcNEQ_MPP + c2o1 * kxxMyyFromfcNEQ_PMM + c2o1 * kxxMyyFromfcNEQ_PMP +
- c2o1 * kxxMyyFromfcNEQ_MMM + c2o1 * kxxMyyFromfcNEQ_MMP + kxxMzzFromfcNEQ_PPM + kxxMzzFromfcNEQ_PPP +
- kxxMzzFromfcNEQ_MPM + kxxMzzFromfcNEQ_MPP - kxxMzzFromfcNEQ_PMM - kxxMzzFromfcNEQ_PMP -
- kxxMzzFromfcNEQ_MMM - kxxMzzFromfcNEQ_MMP + c2o1 * vx1_PPM + c2o1 * vx1_PPP - c2o1 * vx1_MPM -
- c2o1 * vx1_MPP - c2o1 * vx1_PMM - c2o1 * vx1_PMP + c2o1 * vx1_MMM + c2o1 * vx1_MMP - c2o1 * vx3_PPM +
- c2o1 * vx3_PPP - c2o1 * vx3_MPM + c2o1 * vx3_MPP + c2o1 * vx3_PMM - c2o1 * vx3_PMP + c2o1 * vx3_MMM -
- c2o1 * vx3_MMP) /
- c16o1;
- c_020 = (kyzFromfcNEQ_PPM + kyzFromfcNEQ_PPP + kyzFromfcNEQ_MPM + kyzFromfcNEQ_MPP - kyzFromfcNEQ_PMM -
- kyzFromfcNEQ_PMP - kyzFromfcNEQ_MMM - kyzFromfcNEQ_MMP + c2o1 * vx2_PPM - c2o1 * vx2_PPP +
- c2o1 * vx2_MPM - c2o1 * vx2_MPP - c2o1 * vx2_PMM + c2o1 * vx2_PMP - c2o1 * vx2_MMM + c2o1 * vx2_MMP) /
- c8o1;
- a_001 = (-vx1_PPM + vx1_PPP - vx1_MPM + vx1_MPP - vx1_PMM + vx1_PMP - vx1_MMM + vx1_MMP) / c4o1;
- b_001 = (-vx2_PPM + vx2_PPP - vx2_MPM + vx2_MPP - vx2_PMM + vx2_PMP - vx2_MMM + vx2_MMP) / c4o1;
- c_001 = (-vx3_PPM + vx3_PPP - vx3_MPM + vx3_MPP - vx3_PMM + vx3_PMP - vx3_MMM + vx3_MMP) / c4o1;
- a_002 = (-kxzFromfcNEQ_PPM + kxzFromfcNEQ_PPP - kxzFromfcNEQ_MPM + kxzFromfcNEQ_MPP - kxzFromfcNEQ_PMM +
- kxzFromfcNEQ_PMP - kxzFromfcNEQ_MMM + kxzFromfcNEQ_MMP + c2o1 * vx3_PPM - c2o1 * vx3_PPP -
- c2o1 * vx3_MPM + c2o1 * vx3_MPP + c2o1 * vx3_PMM - c2o1 * vx3_PMP - c2o1 * vx3_MMM + c2o1 * vx3_MMP) /
- c8o1;
- b_002 = (-kyzFromfcNEQ_PPM + kyzFromfcNEQ_PPP - kyzFromfcNEQ_MPM + kyzFromfcNEQ_MPP - kyzFromfcNEQ_PMM +
- kyzFromfcNEQ_PMP - kyzFromfcNEQ_MMM + kyzFromfcNEQ_MMP + c2o1 * vx3_PPM - c2o1 * vx3_PPP +
- c2o1 * vx3_MPM - c2o1 * vx3_MPP - c2o1 * vx3_PMM + c2o1 * vx3_PMP - c2o1 * vx3_MMM + c2o1 * vx3_MMP) /
- c8o1;
- c_002 = (-kxxMyyFromfcNEQ_PPM + kxxMyyFromfcNEQ_PPP - kxxMyyFromfcNEQ_MPM + kxxMyyFromfcNEQ_MPP -
- kxxMyyFromfcNEQ_PMM + kxxMyyFromfcNEQ_PMP - kxxMyyFromfcNEQ_MMM + kxxMyyFromfcNEQ_MMP +
- c2o1 * kxxMzzFromfcNEQ_PPM - c2o1 * kxxMzzFromfcNEQ_PPP + c2o1 * kxxMzzFromfcNEQ_MPM -
- c2o1 * kxxMzzFromfcNEQ_MPP + c2o1 * kxxMzzFromfcNEQ_PMM - c2o1 * kxxMzzFromfcNEQ_PMP +
- c2o1 * kxxMzzFromfcNEQ_MMM - c2o1 * kxxMzzFromfcNEQ_MMP - c2o1 * vx1_PPM + c2o1 * vx1_PPP +
- c2o1 * vx1_MPM - c2o1 * vx1_MPP - c2o1 * vx1_PMM + c2o1 * vx1_PMP + c2o1 * vx1_MMM - c2o1 * vx1_MMP -
- c2o1 * vx2_PPM + c2o1 * vx2_PPP - c2o1 * vx2_MPM + c2o1 * vx2_MPP + c2o1 * vx2_PMM - c2o1 * vx2_PMP +
- c2o1 * vx2_MMM - c2o1 * vx2_MMP) /
- c16o1;
- a_110 = (vx1_PPM + vx1_PPP - vx1_MPM - vx1_MPP - vx1_PMM - vx1_PMP + vx1_MMM + vx1_MMP) / c2o1;
- b_110 = (vx2_PPM + vx2_PPP - vx2_MPM - vx2_MPP - vx2_PMM - vx2_PMP + vx2_MMM + vx2_MMP) / c2o1;
- c_110 = (vx3_PPM + vx3_PPP - vx3_MPM - vx3_MPP - vx3_PMM - vx3_PMP + vx3_MMM + vx3_MMP) / c2o1;
- a_101 = (-vx1_PPM + vx1_PPP + vx1_MPM - vx1_MPP - vx1_PMM + vx1_PMP + vx1_MMM - vx1_MMP) / c2o1;
- b_101 = (-vx2_PPM + vx2_PPP + vx2_MPM - vx2_MPP - vx2_PMM + vx2_PMP + vx2_MMM - vx2_MMP) / c2o1;
- c_101 = (-vx3_PPM + vx3_PPP + vx3_MPM - vx3_MPP - vx3_PMM + vx3_PMP + vx3_MMM - vx3_MMP) / c2o1;
- a_011 = (-vx1_PPM + vx1_PPP - vx1_MPM + vx1_MPP + vx1_PMM - vx1_PMP + vx1_MMM - vx1_MMP) / c2o1;
- b_011 = (-vx2_PPM + vx2_PPP - vx2_MPM + vx2_MPP + vx2_PMM - vx2_PMP + vx2_MMM - vx2_MMP) / c2o1;
- c_011 = (-vx3_PPM + vx3_PPP - vx3_MPM + vx3_MPP + vx3_PMM - vx3_PMP + vx3_MMM - vx3_MMP) / c2o1;
-
- a_111 = -vx1_PPM + vx1_PPP + vx1_MPM - vx1_MPP + vx1_PMM - vx1_PMP - vx1_MMM + vx1_MMP;
- b_111 = -vx2_PPM + vx2_PPP + vx2_MPM - vx2_MPP + vx2_PMM - vx2_PMP - vx2_MMM + vx2_MMP;
- c_111 = -vx3_PPM + vx3_PPP + vx3_MPM - vx3_MPP + vx3_PMM - vx3_PMP - vx3_MMM + vx3_MMP;
+ // a_000 = (-kxxMyyFromfcNEQ_PPM - kxxMyyFromfcNEQ_PPP + kxxMyyFromfcNEQ_MPM + kxxMyyFromfcNEQ_MPP -
+ // kxxMyyFromfcNEQ_PMM - kxxMyyFromfcNEQ_PMP + kxxMyyFromfcNEQ_MMM + kxxMyyFromfcNEQ_MMP -
+ // kxxMzzFromfcNEQ_PPM - kxxMzzFromfcNEQ_PPP + kxxMzzFromfcNEQ_MPM + kxxMzzFromfcNEQ_MPP -
+ // kxxMzzFromfcNEQ_PMM - kxxMzzFromfcNEQ_PMP + kxxMzzFromfcNEQ_MMM + kxxMzzFromfcNEQ_MMP -
+ // c2o1 * kxyFromfcNEQ_PPM - c2o1 * kxyFromfcNEQ_PPP - c2o1 * kxyFromfcNEQ_MPM - c2o1 * kxyFromfcNEQ_MPP +
+ // c2o1 * kxyFromfcNEQ_PMM + c2o1 * kxyFromfcNEQ_PMP + c2o1 * kxyFromfcNEQ_MMM + c2o1 * kxyFromfcNEQ_MMP +
+ // c2o1 * kxzFromfcNEQ_PPM - c2o1 * kxzFromfcNEQ_PPP + c2o1 * kxzFromfcNEQ_MPM - c2o1 * kxzFromfcNEQ_MPP +
+ // c2o1 * kxzFromfcNEQ_PMM - c2o1 * kxzFromfcNEQ_PMP + c2o1 * kxzFromfcNEQ_MMM - c2o1 * kxzFromfcNEQ_MMP +
+ // c8o1 * vx1_PPM + c8o1 * vx1_PPP + c8o1 * vx1_MPM + c8o1 * vx1_MPP + c8o1 * vx1_PMM + c8o1 * vx1_PMP +
+ // c8o1 * vx1_MMM + c8o1 * vx1_MMP + c2o1 * vx2_PPM + c2o1 * vx2_PPP - c2o1 * vx2_MPM - c2o1 * vx2_MPP -
+ // c2o1 * vx2_PMM - c2o1 * vx2_PMP + c2o1 * vx2_MMM + c2o1 * vx2_MMP - c2o1 * vx3_PPM + c2o1 * vx3_PPP +
+ // c2o1 * vx3_MPM - c2o1 * vx3_MPP - c2o1 * vx3_PMM + c2o1 * vx3_PMP + c2o1 * vx3_MMM - c2o1 * vx3_MMP) /
+ // c64o1;
+ a_000 =
+ c1o64 * (c2o1 * (((kxyFromfcNEQ_MMM - kxyFromfcNEQ_PPP) + (kxyFromfcNEQ_MMP - kxyFromfcNEQ_PPM)) +
+ ((kxyFromfcNEQ_PMM - kxyFromfcNEQ_MPP) + (kxyFromfcNEQ_PMP - kxyFromfcNEQ_MPM)) +
+ ((kxzFromfcNEQ_MMM - kxzFromfcNEQ_PPP) + (kxzFromfcNEQ_PPM - kxzFromfcNEQ_MMP)) +
+ ((kxzFromfcNEQ_PMM - kxzFromfcNEQ_MPP) + (kxzFromfcNEQ_MPM - kxzFromfcNEQ_PMP)) +
+ ((vx2_PPP + vx2_MMM) + (vx2_PPM + vx2_MMP)) - ((vx2_MPP + vx2_PMM) + (vx2_MPM + vx2_PMP)) +
+ ((vx3_PPP + vx3_MMM) - (vx3_PPM + vx3_MMP)) + ((vx3_PMP + vx3_MPM) - (vx3_MPP + vx3_PMM))) +
+ c8o1 * (((vx1_PPP + vx1_MMM) + (vx1_PPM + vx1_MMP)) + ((vx1_MPP + vx1_PMM) + (vx1_PMP + vx1_MPM))) +
+ ((kxxMyyFromfcNEQ_MMM - kxxMyyFromfcNEQ_PPP) + (kxxMyyFromfcNEQ_MMP - kxxMyyFromfcNEQ_PPM)) +
+ ((kxxMyyFromfcNEQ_MPP - kxxMyyFromfcNEQ_PMM) + (kxxMyyFromfcNEQ_MPM - kxxMyyFromfcNEQ_PMP)) +
+ ((kxxMzzFromfcNEQ_MMM - kxxMzzFromfcNEQ_PPP) + (kxxMzzFromfcNEQ_MMP - kxxMzzFromfcNEQ_PPM)) +
+ ((kxxMzzFromfcNEQ_MPP - kxxMzzFromfcNEQ_PMM) + (kxxMzzFromfcNEQ_MPM - kxxMzzFromfcNEQ_PMP)));
+
+ // b_000 = (c2o1 * kxxMyyFromfcNEQ_PPM + c2o1 * kxxMyyFromfcNEQ_PPP + c2o1 * kxxMyyFromfcNEQ_MPM +
+ // c2o1 * kxxMyyFromfcNEQ_MPP - c2o1 * kxxMyyFromfcNEQ_PMM - c2o1 * kxxMyyFromfcNEQ_PMP -
+ // c2o1 * kxxMyyFromfcNEQ_MMM - c2o1 * kxxMyyFromfcNEQ_MMP - kxxMzzFromfcNEQ_PPM - kxxMzzFromfcNEQ_PPP -
+ // kxxMzzFromfcNEQ_MPM - kxxMzzFromfcNEQ_MPP + kxxMzzFromfcNEQ_PMM + kxxMzzFromfcNEQ_PMP +
+ // kxxMzzFromfcNEQ_MMM + kxxMzzFromfcNEQ_MMP - c2o1 * kxyFromfcNEQ_PPM - c2o1 * kxyFromfcNEQ_PPP +
+ // c2o1 * kxyFromfcNEQ_MPM + c2o1 * kxyFromfcNEQ_MPP - c2o1 * kxyFromfcNEQ_PMM - c2o1 * kxyFromfcNEQ_PMP +
+ // c2o1 * kxyFromfcNEQ_MMM + c2o1 * kxyFromfcNEQ_MMP + c2o1 * kyzFromfcNEQ_PPM - c2o1 * kyzFromfcNEQ_PPP +
+ // c2o1 * kyzFromfcNEQ_MPM - c2o1 * kyzFromfcNEQ_MPP + c2o1 * kyzFromfcNEQ_PMM - c2o1 * kyzFromfcNEQ_PMP +
+ // c2o1 * kyzFromfcNEQ_MMM - c2o1 * kyzFromfcNEQ_MMP + c2o1 * vx1_PPM + c2o1 * vx1_PPP - c2o1 * vx1_MPM -
+ // c2o1 * vx1_MPP - c2o1 * vx1_PMM - c2o1 * vx1_PMP + c2o1 * vx1_MMM + c2o1 * vx1_MMP + c8o1 * vx2_PPM +
+ // c8o1 * vx2_PPP + c8o1 * vx2_MPM + c8o1 * vx2_MPP + c8o1 * vx2_PMM + c8o1 * vx2_PMP + c8o1 * vx2_MMM +
+ // c8o1 * vx2_MMP - c2o1 * vx3_PPM + c2o1 * vx3_PPP - c2o1 * vx3_MPM + c2o1 * vx3_MPP + c2o1 * vx3_PMM -
+ // c2o1 * vx3_PMP + c2o1 * vx3_MMM - c2o1 * vx3_MMP) /
+ // c64o1;
+ b_000 =
+ c1o64 * (c2o1 * (((kxxMyyFromfcNEQ_PPP - kxxMyyFromfcNEQ_MMM) + (kxxMyyFromfcNEQ_PPM - kxxMyyFromfcNEQ_MMP)) +
+ ((kxxMyyFromfcNEQ_MPP - kxxMyyFromfcNEQ_PMM) + (kxxMyyFromfcNEQ_MPM - kxxMyyFromfcNEQ_PMP)) +
+ ((kxyFromfcNEQ_MMM - kxyFromfcNEQ_PPP) + (kxyFromfcNEQ_MMP - kxyFromfcNEQ_PPM)) +
+ ((kxyFromfcNEQ_MPP - kxyFromfcNEQ_PMM) + (kxyFromfcNEQ_MPM - kxyFromfcNEQ_PMP)) +
+ ((kyzFromfcNEQ_MMM - kyzFromfcNEQ_PPP) + (kyzFromfcNEQ_PPM - kyzFromfcNEQ_MMP)) +
+ ((kyzFromfcNEQ_PMM - kyzFromfcNEQ_MPP) + (kyzFromfcNEQ_MPM - kyzFromfcNEQ_PMP)) +
+ ((vx1_PPP + vx1_MMM) + (vx1_PPM + vx1_MMP)) - ((vx1_MPM + vx1_MPP) + (vx1_PMM + vx1_PMP)) +
+ ((vx3_PPP + vx3_MMM) - (vx3_PPM + vx3_MMP)) + ((vx3_MPP + vx3_PMM) - (vx3_MPM + vx3_PMP))) +
+ c8o1 * (((vx2_PPP + vx2_MMM) + (vx2_PPM + vx2_MMP)) + ((vx2_MPP + vx2_PMM) + (vx2_MPM + vx2_PMP))) +
+ ((kxxMzzFromfcNEQ_MMM - kxxMzzFromfcNEQ_PPP) + (kxxMzzFromfcNEQ_MMP - kxxMzzFromfcNEQ_PPM)) +
+ ((kxxMzzFromfcNEQ_PMM - kxxMzzFromfcNEQ_MPP) + (kxxMzzFromfcNEQ_PMP - kxxMzzFromfcNEQ_MPM)));
+
+ // c_000 = (kxxMyyFromfcNEQ_PPM - kxxMyyFromfcNEQ_PPP + kxxMyyFromfcNEQ_MPM - kxxMyyFromfcNEQ_MPP +
+ // kxxMyyFromfcNEQ_PMM - kxxMyyFromfcNEQ_PMP + kxxMyyFromfcNEQ_MMM - kxxMyyFromfcNEQ_MMP -
+ // c2o1 * kxxMzzFromfcNEQ_PPM + c2o1 * kxxMzzFromfcNEQ_PPP - c2o1 * kxxMzzFromfcNEQ_MPM +
+ // c2o1 * kxxMzzFromfcNEQ_MPP - c2o1 * kxxMzzFromfcNEQ_PMM + c2o1 * kxxMzzFromfcNEQ_PMP -
+ // c2o1 * kxxMzzFromfcNEQ_MMM + c2o1 * kxxMzzFromfcNEQ_MMP - c2o1 * kxzFromfcNEQ_PPM -
+ // c2o1 * kxzFromfcNEQ_PPP + c2o1 * kxzFromfcNEQ_MPM + c2o1 * kxzFromfcNEQ_MPP - c2o1 * kxzFromfcNEQ_PMM -
+ // c2o1 * kxzFromfcNEQ_PMP + c2o1 * kxzFromfcNEQ_MMM + c2o1 * kxzFromfcNEQ_MMP - c2o1 * kyzFromfcNEQ_PPM -
+ // c2o1 * kyzFromfcNEQ_PPP - c2o1 * kyzFromfcNEQ_MPM - c2o1 * kyzFromfcNEQ_MPP + c2o1 * kyzFromfcNEQ_PMM +
+ // c2o1 * kyzFromfcNEQ_PMP + c2o1 * kyzFromfcNEQ_MMM + c2o1 * kyzFromfcNEQ_MMP - c2o1 * vx1_PPM +
+ // c2o1 * vx1_PPP + c2o1 * vx1_MPM - c2o1 * vx1_MPP - c2o1 * vx1_PMM + c2o1 * vx1_PMP + c2o1 * vx1_MMM -
+ // c2o1 * vx1_MMP - c2o1 * vx2_PPM + c2o1 * vx2_PPP - c2o1 * vx2_MPM + c2o1 * vx2_MPP + c2o1 * vx2_PMM -
+ // c2o1 * vx2_PMP + c2o1 * vx2_MMM - c2o1 * vx2_MMP + c8o1 * vx3_PPM + c8o1 * vx3_PPP + c8o1 * vx3_MPM +
+ // c8o1 * vx3_MPP + c8o1 * vx3_PMM + c8o1 * vx3_PMP + c8o1 * vx3_MMM + c8o1 * vx3_MMP) /
+ // c64o1;
+ c_000 =
+ c1o64 * (c2o1 * (((kxxMzzFromfcNEQ_PPP - kxxMzzFromfcNEQ_MMM) + (kxxMzzFromfcNEQ_MMP - kxxMzzFromfcNEQ_PPM)) +
+ ((kxxMzzFromfcNEQ_MPP - kxxMzzFromfcNEQ_PMM) + (kxxMzzFromfcNEQ_PMP - kxxMzzFromfcNEQ_MPM)) +
+ ((kxzFromfcNEQ_MMM - kxzFromfcNEQ_PPP) + (kxzFromfcNEQ_MMP - kxzFromfcNEQ_PPM)) +
+ ((kxzFromfcNEQ_MPP - kxzFromfcNEQ_PMM) + (kxzFromfcNEQ_MPM - kxzFromfcNEQ_PMP)) +
+ ((kyzFromfcNEQ_MMM - kyzFromfcNEQ_PPP) + (kyzFromfcNEQ_MMP - kyzFromfcNEQ_PPM)) +
+ ((kyzFromfcNEQ_PMM - kyzFromfcNEQ_MPP) + (kyzFromfcNEQ_PMP - kyzFromfcNEQ_MPM)) +
+ ((vx1_PPP + vx1_MMM) - (vx1_MMP + vx1_PPM)) + ((vx1_MPM + vx1_PMP) - (vx1_MPP + vx1_PMM)) +
+ ((vx2_PPP + vx2_MMM) - (vx2_MMP + vx2_PPM)) + ((vx2_MPP + vx2_PMM) - (vx2_MPM + vx2_PMP))) +
+ c8o1 * (((vx3_PPP + vx3_MMM) + (vx3_PPM + vx3_MMP)) + ((vx3_PMM + vx3_MPP) + (vx3_PMP + vx3_MPM))) +
+ ((kxxMyyFromfcNEQ_MMM - kxxMyyFromfcNEQ_PPP) + (kxxMyyFromfcNEQ_PPM - kxxMyyFromfcNEQ_MMP)) +
+ ((kxxMyyFromfcNEQ_PMM - kxxMyyFromfcNEQ_MPP) + (kxxMyyFromfcNEQ_MPM - kxxMyyFromfcNEQ_PMP)));
+
+ // a_100 = (vx1_PPM + vx1_PPP - vx1_MPM - vx1_MPP + vx1_PMM + vx1_PMP - vx1_MMM - vx1_MMP) / c4o1;
+ a_100 = c1o4 * (((vx1_PPP - vx1_MMM) + (vx1_PPM - vx1_MMP)) + ((vx1_PMM - vx1_MPP) + (vx1_PMP - vx1_MPM)));
+
+ // b_100 = (vx2_PPM + vx2_PPP - vx2_MPM - vx2_MPP + vx2_PMM + vx2_PMP - vx2_MMM - vx2_MMP) / c4o1;
+ b_100 = c1o4 * (((vx2_PPP - vx2_MMM) + (vx2_PPM - vx2_MMP)) + ((vx2_PMM - vx2_MPP) + (vx2_PMP - vx2_MPM)));
+
+ // c_100 = (vx3_PPM + vx3_PPP - vx3_MPM - vx3_MPP + vx3_PMM + vx3_PMP - vx3_MMM - vx3_MMP) / c4o1;
+ c_100 = c1o4 * (((vx3_PPP - vx3_MMM) + (vx3_PPM - vx3_MMP)) + ((vx3_PMM - vx3_MPP) + (vx3_PMP - vx3_MPM)));
+
+ // a_200 = (kxxMyyFromfcNEQ_PPM + kxxMyyFromfcNEQ_PPP - kxxMyyFromfcNEQ_MPM - kxxMyyFromfcNEQ_MPP +
+ // kxxMyyFromfcNEQ_PMM + kxxMyyFromfcNEQ_PMP - kxxMyyFromfcNEQ_MMM - kxxMyyFromfcNEQ_MMP +
+ // kxxMzzFromfcNEQ_PPM + kxxMzzFromfcNEQ_PPP - kxxMzzFromfcNEQ_MPM - kxxMzzFromfcNEQ_MPP +
+ // kxxMzzFromfcNEQ_PMM + kxxMzzFromfcNEQ_PMP - kxxMzzFromfcNEQ_MMM - kxxMzzFromfcNEQ_MMP + c2o1 * vx2_PPM +
+ // c2o1 * vx2_PPP - c2o1 * vx2_MPM - c2o1 * vx2_MPP - c2o1 * vx2_PMM - c2o1 * vx2_PMP + c2o1 * vx2_MMM +
+ // c2o1 * vx2_MMP - c2o1 * vx3_PPM + c2o1 * vx3_PPP + c2o1 * vx3_MPM - c2o1 * vx3_MPP - c2o1 * vx3_PMM +
+ // c2o1 * vx3_PMP + c2o1 * vx3_MMM - c2o1 * vx3_MMP) /
+ // c16o1;
+ a_200 =
+ c1o16 * (c2o1 * (((vx2_PPP + vx2_MMM) + (vx2_PPM - vx2_MPP)) + ((vx2_MMP - vx2_PMM) - (vx2_MPM + vx2_PMP)) +
+ ((vx3_PPP + vx3_MMM) - (vx3_PPM + vx3_MPP)) + ((vx3_MPM + vx3_PMP) - (vx3_MMP + vx3_PMM))) +
+ ((kxxMyyFromfcNEQ_PPP - kxxMyyFromfcNEQ_MMM) + (kxxMyyFromfcNEQ_PPM - kxxMyyFromfcNEQ_MMP)) +
+ ((kxxMyyFromfcNEQ_PMM - kxxMyyFromfcNEQ_MPP) + (kxxMyyFromfcNEQ_PMP - kxxMyyFromfcNEQ_MPM)) +
+ ((kxxMzzFromfcNEQ_PPP - kxxMzzFromfcNEQ_MMM) + (kxxMzzFromfcNEQ_PPM - kxxMzzFromfcNEQ_MMP)) +
+ ((kxxMzzFromfcNEQ_PMM - kxxMzzFromfcNEQ_MPP) + (kxxMzzFromfcNEQ_PMP - kxxMzzFromfcNEQ_MPM)));
+
+ // b_200 = (kxyFromfcNEQ_PPM + kxyFromfcNEQ_PPP - kxyFromfcNEQ_MPM - kxyFromfcNEQ_MPP + kxyFromfcNEQ_PMM +
+ // kxyFromfcNEQ_PMP - kxyFromfcNEQ_MMM - kxyFromfcNEQ_MMP - c2o1 * vx1_PPM - c2o1 * vx1_PPP +
+ // c2o1 * vx1_MPM + c2o1 * vx1_MPP + c2o1 * vx1_PMM + c2o1 * vx1_PMP - c2o1 * vx1_MMM - c2o1 * vx1_MMP) /
+ // c8o1;
+ b_200 =
+ c1o8 * (c2o1 * (-((vx1_PPP + vx1_MMM) + (vx1_PPM + vx1_MMP)) + ((vx1_MPP + vx1_PMM) + (vx1_MPM + vx1_PMP))) +
+ ((kxyFromfcNEQ_PPP - kxyFromfcNEQ_MMM) + (kxyFromfcNEQ_PPM - kxyFromfcNEQ_MMP)) +
+ ((kxyFromfcNEQ_PMM - kxyFromfcNEQ_MPP) + (kxyFromfcNEQ_PMP - kxyFromfcNEQ_MPM)));
+
+ // c_200 = (kxzFromfcNEQ_PPM + kxzFromfcNEQ_PPP - kxzFromfcNEQ_MPM - kxzFromfcNEQ_MPP + kxzFromfcNEQ_PMM +
+ // kxzFromfcNEQ_PMP - kxzFromfcNEQ_MMM - kxzFromfcNEQ_MMP + c2o1 * vx1_PPM - c2o1 * vx1_PPP - c2o1 *
+ // vx1_MPM + c2o1 * vx1_MPP + c2o1 * vx1_PMM - c2o1 * vx1_PMP - c2o1 * vx1_MMM + c2o1 * vx1_MMP) /
+ // c8o1;
+ c_200 = c1o8 * (c2o1 * (((vx1_PPM + vx1_MMP) - (vx1_PPP + vx1_MMM)) + ((vx1_MPP + vx1_PMM) - (vx1_MPM + vx1_PMP))) +
+ ((kxzFromfcNEQ_PPP - kxzFromfcNEQ_MMM) + (kxzFromfcNEQ_PPM - kxzFromfcNEQ_MMP)) +
+ ((kxzFromfcNEQ_PMM - kxzFromfcNEQ_MPP) + (kxzFromfcNEQ_PMP - kxzFromfcNEQ_MPM)));
+
+ // a_010 = (vx1_PPM + vx1_PPP + vx1_MPM + vx1_MPP - vx1_PMM - vx1_PMP - vx1_MMM - vx1_MMP) / c4o1;
+ a_010 = c1o4 * (((vx1_PPP - vx1_MMM) + (vx1_PPM - vx1_MMP)) + ((vx1_MPP - vx1_PMM) + (vx1_MPM - vx1_PMP)));
+
+ // b_010 = (vx2_PPM + vx2_PPP + vx2_MPM + vx2_MPP - vx2_PMM - vx2_PMP - vx2_MMM - vx2_MMP) / c4o1;
+ b_010 = c1o4 * (((vx2_PPP - vx2_MMM) + (vx2_PPM - vx2_MMP)) + ((vx2_MPP - vx2_PMM) + (vx2_MPM - vx2_PMP)));
+
+ // c_010 = (vx3_PPM + vx3_PPP + vx3_MPM + vx3_MPP - vx3_PMM - vx3_PMP - vx3_MMM - vx3_MMP) / c4o1;
+ c_010 = c1o4 * (((vx3_PPP - vx3_MMM) + (vx3_PPM - vx3_MMP)) + ((vx3_MPP - vx3_PMM) + (vx3_MPM - vx3_PMP)));
+
+ // a_020 = (kxyFromfcNEQ_PPM + kxyFromfcNEQ_PPP + kxyFromfcNEQ_MPM + kxyFromfcNEQ_MPP - kxyFromfcNEQ_PMM -
+ // kxyFromfcNEQ_PMP - kxyFromfcNEQ_MMM - kxyFromfcNEQ_MMP - c2o1 * vx2_PPM - c2o1 * vx2_PPP +
+ // c2o1 * vx2_MPM + c2o1 * vx2_MPP + c2o1 * vx2_PMM + c2o1 * vx2_PMP - c2o1 * vx2_MMM - c2o1 * vx2_MMP) /
+ // c8o1;
+ a_020 =
+ c1o8 * (c2o1 * (-((vx2_PPP + vx2_MMM) + (vx2_MMP + vx2_PPM)) + ((vx2_MPP + vx2_PMM) + (vx2_MPM + vx2_PMP))) +
+ ((kxyFromfcNEQ_PPP - kxyFromfcNEQ_MMM) + (kxyFromfcNEQ_PPM - kxyFromfcNEQ_MMP)) +
+ ((kxyFromfcNEQ_MPP - kxyFromfcNEQ_PMM) + (kxyFromfcNEQ_MPM - kxyFromfcNEQ_PMP)));
+
+ // b_020 = (-c2o1 * kxxMyyFromfcNEQ_PPM - c2o1 * kxxMyyFromfcNEQ_PPP - c2o1 * kxxMyyFromfcNEQ_MPM -
+ // c2o1 * kxxMyyFromfcNEQ_MPP + c2o1 * kxxMyyFromfcNEQ_PMM + c2o1 * kxxMyyFromfcNEQ_PMP +
+ // c2o1 * kxxMyyFromfcNEQ_MMM + c2o1 * kxxMyyFromfcNEQ_MMP + kxxMzzFromfcNEQ_PPM + kxxMzzFromfcNEQ_PPP +
+ // kxxMzzFromfcNEQ_MPM + kxxMzzFromfcNEQ_MPP - kxxMzzFromfcNEQ_PMM - kxxMzzFromfcNEQ_PMP -
+ // kxxMzzFromfcNEQ_MMM - kxxMzzFromfcNEQ_MMP + c2o1 * vx1_PPM + c2o1 * vx1_PPP - c2o1 * vx1_MPM -
+ // c2o1 * vx1_MPP - c2o1 * vx1_PMM - c2o1 * vx1_PMP + c2o1 * vx1_MMM + c2o1 * vx1_MMP - c2o1 * vx3_PPM +
+ // c2o1 * vx3_PPP - c2o1 * vx3_MPM + c2o1 * vx3_MPP + c2o1 * vx3_PMM - c2o1 * vx3_PMP + c2o1 * vx3_MMM -
+ // c2o1 * vx3_MMP) /
+ // c16o1;
+ b_020 =
+ c1o16 * (c2o1 * (((kxxMyyFromfcNEQ_MMM - kxxMyyFromfcNEQ_PPP) + (kxxMyyFromfcNEQ_MMP - kxxMyyFromfcNEQ_PPM)) +
+ ((kxxMyyFromfcNEQ_PMM - kxxMyyFromfcNEQ_MPP) + (kxxMyyFromfcNEQ_PMP - kxxMyyFromfcNEQ_MPM)) +
+ ((vx1_PPP + vx1_MMM) + (vx1_PPM + vx1_MMP)) - ((vx1_MPP + vx1_PMM) + (vx1_PMP + vx1_MPM)) +
+ ((vx3_PPP + vx3_MMM) - (vx3_PPM + vx3_MMP)) + ((vx3_MPP + vx3_PMM) - (vx3_MPM + vx3_PMP))) +
+ ((kxxMzzFromfcNEQ_PPP - kxxMzzFromfcNEQ_MMM) + (kxxMzzFromfcNEQ_PPM - kxxMzzFromfcNEQ_MMP)) +
+ ((kxxMzzFromfcNEQ_MPP - kxxMzzFromfcNEQ_PMM) + (kxxMzzFromfcNEQ_MPM - kxxMzzFromfcNEQ_PMP)));
+
+ // c_020 = (kyzFromfcNEQ_PPM + kyzFromfcNEQ_PPP + kyzFromfcNEQ_MPM + kyzFromfcNEQ_MPP - kyzFromfcNEQ_PMM -
+ // kyzFromfcNEQ_PMP - kyzFromfcNEQ_MMM - kyzFromfcNEQ_MMP + c2o1 * vx2_PPM - c2o1 * vx2_PPP + c2o1 *
+ // vx2_MPM - c2o1 * vx2_MPP - c2o1 * vx2_PMM + c2o1 * vx2_PMP - c2o1 * vx2_MMM + c2o1 * vx2_MMP) /
+ // c8o1;
+ c_020 = c1o8 * (c2o1 * (((vx2_MMP + vx2_PPM) - (vx2_PPP + vx2_MMM)) + ((vx2_PMP + vx2_MPM) - (vx2_MPP + vx2_PMM))) +
+ ((kyzFromfcNEQ_PPP - kyzFromfcNEQ_MMM) + (kyzFromfcNEQ_PPM - kyzFromfcNEQ_MMP)) +
+ ((kyzFromfcNEQ_MPP - kyzFromfcNEQ_PMM) + (kyzFromfcNEQ_MPM - kyzFromfcNEQ_PMP)));
+
+ // a_001 = (-vx1_PPM + vx1_PPP - vx1_MPM + vx1_MPP - vx1_PMM + vx1_PMP - vx1_MMM + vx1_MMP) / c4o1;
+ a_001 = c1o4 * (((vx1_PPP - vx1_MMM) + (vx1_MMP - vx1_PPM)) + ((vx1_MPP - vx1_PMM) + (vx1_PMP - vx1_MPM)));
+
+ // b_001 = (-vx2_PPM + vx2_PPP - vx2_MPM + vx2_MPP - vx2_PMM + vx2_PMP - vx2_MMM + vx2_MMP) / c4o1;
+ b_001 = c1o4 * (((vx2_PPP - vx2_MMM) + (vx2_MMP - vx2_PPM)) + ((vx2_MPP - vx2_PMM) + (vx2_PMP - vx2_MPM)));
+
+ // c_001 = (-vx3_PPM + vx3_PPP - vx3_MPM + vx3_MPP - vx3_PMM + vx3_PMP - vx3_MMM + vx3_MMP) / c4o1;
+ c_001 = c1o4 * (((vx3_PPP - vx3_MMM) + (vx3_MMP - vx3_PPM)) + ((vx3_MPP - vx3_PMM) + (vx3_PMP - vx3_MPM)));
+
+ // a_002 = (-kxzFromfcNEQ_PPM + kxzFromfcNEQ_PPP - kxzFromfcNEQ_MPM + kxzFromfcNEQ_MPP - kxzFromfcNEQ_PMM +
+ // kxzFromfcNEQ_PMP - kxzFromfcNEQ_MMM + kxzFromfcNEQ_MMP + c2o1 * vx3_PPM - c2o1 * vx3_PPP -
+ // c2o1 * vx3_MPM + c2o1 * vx3_MPP + c2o1 * vx3_PMM - c2o1 * vx3_PMP - c2o1 * vx3_MMM + c2o1 * vx3_MMP) /
+ // c8o1;
+ a_002 = c1o8 * (c2o1 * (((vx3_PPM + vx3_MMP) - (vx3_PPP + vx3_MMM)) + ((vx3_MPP + vx3_PMM) - (vx3_PMP + vx3_MPM))) +
+ ((kxzFromfcNEQ_PPP - kxzFromfcNEQ_MMM) + (kxzFromfcNEQ_MMP - kxzFromfcNEQ_PPM)) +
+ ((kxzFromfcNEQ_PMP - kxzFromfcNEQ_MPM) + (kxzFromfcNEQ_MPP - kxzFromfcNEQ_PMM)));
+
+ // b_002 = (-kyzFromfcNEQ_PPM + kyzFromfcNEQ_PPP - kyzFromfcNEQ_MPM + kyzFromfcNEQ_MPP - kyzFromfcNEQ_PMM +
+ // kyzFromfcNEQ_PMP - kyzFromfcNEQ_MMM + kyzFromfcNEQ_MMP + c2o1 * vx3_PPM - c2o1 * vx3_PPP + c2o1 *
+ // vx3_MPM - c2o1 * vx3_MPP - c2o1 * vx3_PMM + c2o1 * vx3_PMP - c2o1 * vx3_MMM + c2o1 * vx3_MMP) /
+ // c8o1;
+ b_002 = c1o8 * (c2o1 * (((vx3_PPM + vx3_MMP) - (vx3_PPP + vx3_MMM)) + ((vx3_MPM + vx3_PMP) - (vx3_PMM + vx3_MPP))) +
+ ((kyzFromfcNEQ_PPP - kyzFromfcNEQ_MMM) + (kyzFromfcNEQ_MMP - kyzFromfcNEQ_PPM)) +
+ ((kyzFromfcNEQ_PMP - kyzFromfcNEQ_MPM) + (kyzFromfcNEQ_MPP - kyzFromfcNEQ_PMM)));
+
+ // c_002 = (-kxxMyyFromfcNEQ_PPM + kxxMyyFromfcNEQ_PPP - kxxMyyFromfcNEQ_MPM + kxxMyyFromfcNEQ_MPP -
+ // kxxMyyFromfcNEQ_PMM + kxxMyyFromfcNEQ_PMP - kxxMyyFromfcNEQ_MMM + kxxMyyFromfcNEQ_MMP +
+ // c2o1 * kxxMzzFromfcNEQ_PPM - c2o1 * kxxMzzFromfcNEQ_PPP + c2o1 * kxxMzzFromfcNEQ_MPM -
+ // c2o1 * kxxMzzFromfcNEQ_MPP + c2o1 * kxxMzzFromfcNEQ_PMM - c2o1 * kxxMzzFromfcNEQ_PMP +
+ // c2o1 * kxxMzzFromfcNEQ_MMM - c2o1 * kxxMzzFromfcNEQ_MMP - c2o1 * vx1_PPM + c2o1 * vx1_PPP +
+ // c2o1 * vx1_MPM - c2o1 * vx1_MPP - c2o1 * vx1_PMM + c2o1 * vx1_PMP + c2o1 * vx1_MMM - c2o1 * vx1_MMP -
+ // c2o1 * vx2_PPM + c2o1 * vx2_PPP - c2o1 * vx2_MPM + c2o1 * vx2_MPP + c2o1 * vx2_PMM - c2o1 * vx2_PMP +
+ // c2o1 * vx2_MMM - c2o1 * vx2_MMP) /
+ // c16o1;
+ c_002 =
+ c1o16 * (c2o1 * (((kxxMzzFromfcNEQ_MMM - kxxMzzFromfcNEQ_PPP) + (kxxMzzFromfcNEQ_PPM - kxxMzzFromfcNEQ_MMP)) +
+ ((kxxMzzFromfcNEQ_MPM - kxxMzzFromfcNEQ_PMP) + (kxxMzzFromfcNEQ_PMM - kxxMzzFromfcNEQ_MPP)) +
+ ((vx1_PPP + vx1_MMM) - (vx1_MMP + vx1_PPM)) + ((vx1_MPM + vx1_PMP) - (vx1_PMM + vx1_MPP)) +
+ ((vx2_PPP + vx2_MMM) - (vx2_MMP + vx2_PPM)) + ((vx2_PMM + vx2_MPP) - (vx2_MPM + vx2_PMP))) +
+ ((kxxMyyFromfcNEQ_PPP - kxxMyyFromfcNEQ_MMM) + (kxxMyyFromfcNEQ_MMP - kxxMyyFromfcNEQ_PPM)) +
+ ((kxxMyyFromfcNEQ_PMP - kxxMyyFromfcNEQ_MPM) + (kxxMyyFromfcNEQ_MPP - kxxMyyFromfcNEQ_PMM)));
+
+ // a_110 = (vx1_PPM + vx1_PPP - vx1_MPM - vx1_MPP - vx1_PMM - vx1_PMP + vx1_MMM + vx1_MMP) / c2o1;
+ // b_110 = (vx2_PPM + vx2_PPP - vx2_MPM - vx2_MPP - vx2_PMM - vx2_PMP + vx2_MMM + vx2_MMP) / c2o1;
+ // c_110 = (vx3_PPM + vx3_PPP - vx3_MPM - vx3_MPP - vx3_PMM - vx3_PMP + vx3_MMM + vx3_MMP) / c2o1;
+ a_110 = c1o2 * (((vx1_PPP + vx1_MMM) + (vx1_MMP + vx1_PPM)) - ((vx1_MPM + vx1_PMP) + (vx1_PMM + vx1_MPP)));
+ b_110 = c1o2 * (((vx2_PPP + vx2_MMM) + (vx2_MMP + vx2_PPM)) - ((vx2_MPM + vx2_PMP) + (vx2_PMM + vx2_MPP)));
+ c_110 = c1o2 * (((vx3_PPP + vx3_MMM) + (vx3_MMP + vx3_PPM)) - ((vx3_MPM + vx3_PMP) + (vx3_PMM + vx3_MPP)));
+
+ // a_101 = (-vx1_PPM + vx1_PPP + vx1_MPM - vx1_MPP - vx1_PMM + vx1_PMP + vx1_MMM - vx1_MMP) / c2o1;
+ // b_101 = (-vx2_PPM + vx2_PPP + vx2_MPM - vx2_MPP - vx2_PMM + vx2_PMP + vx2_MMM - vx2_MMP) / c2o1;
+ // c_101 = (-vx3_PPM + vx3_PPP + vx3_MPM - vx3_MPP - vx3_PMM + vx3_PMP + vx3_MMM - vx3_MMP) / c2o1;
+ a_101 = c1o2 * (((vx1_PPP + vx1_MMM) - (vx1_MMP + vx1_PPM)) + ((vx1_MPM + vx1_PMP) - (vx1_PMM + vx1_MPP)));
+ b_101 = c1o2 * (((vx2_PPP + vx2_MMM) - (vx2_MMP + vx2_PPM)) + ((vx2_MPM + vx2_PMP) - (vx2_PMM + vx2_MPP)));
+ c_101 = c1o2 * (((vx3_PPP + vx3_MMM) - (vx3_MMP + vx3_PPM)) + ((vx3_MPM + vx3_PMP) - (vx3_PMM + vx3_MPP)));
+
+ // a_011 = (-vx1_PPM + vx1_PPP - vx1_MPM + vx1_MPP + vx1_PMM - vx1_PMP + vx1_MMM - vx1_MMP) / c2o1;
+ // b_011 = (-vx2_PPM + vx2_PPP - vx2_MPM + vx2_MPP + vx2_PMM - vx2_PMP + vx2_MMM - vx2_MMP) / c2o1;
+ // c_011 = (-vx3_PPM + vx3_PPP - vx3_MPM + vx3_MPP + vx3_PMM - vx3_PMP + vx3_MMM - vx3_MMP) / c2o1;
+ a_011 = c1o2 * (((vx1_PPP + vx1_MMM) - (vx1_MMP + vx1_PPM)) + ((vx1_PMM + vx1_MPP) - (vx1_MPM + vx1_PMP)));
+ b_011 = c1o2 * (((vx2_PPP + vx2_MMM) - (vx2_MMP + vx2_PPM)) + ((vx2_PMM + vx2_MPP) - (vx2_MPM + vx2_PMP)));
+ c_011 = c1o2 * (((vx3_PPP + vx3_MMM) - (vx3_MMP + vx3_PPM)) + ((vx3_PMM + vx3_MPP) - (vx3_MPM + vx3_PMP)));
+
+ // a_111 = -vx1_PPM + vx1_PPP + vx1_MPM - vx1_MPP + vx1_PMM - vx1_PMP - vx1_MMM + vx1_MMP;
+ // b_111 = -vx2_PPM + vx2_PPP + vx2_MPM - vx2_MPP + vx2_PMM - vx2_PMP - vx2_MMM + vx2_MMP;
+ // c_111 = -vx3_PPM + vx3_PPP + vx3_MPM - vx3_MPP + vx3_PMM - vx3_PMP - vx3_MMM + vx3_MMP;
+ a_111 = ((vx1_PPP - vx1_MMM) + (vx1_MMP - vx1_PPM)) + ((vx1_MPM - vx1_PMP) + (vx1_PMM - vx1_MPP));
+ b_111 = ((vx2_PPP - vx2_MMM) + (vx2_MMP - vx2_PPM)) + ((vx2_MPM - vx2_PMP) + (vx2_PMM - vx2_MPP));
+ c_111 = ((vx3_PPP - vx3_MMM) + (vx3_MMP - vx3_PPM)) + ((vx3_MPM - vx3_PMP) + (vx3_PMM - vx3_MPP));
//////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
@@ -618,9 +741,9 @@ __global__ void scaleCF_compressible(
////////////////////////////////////////////////////////////////////////////////
//! - Set the relative position of the offset cell {-1, 0, 1}
//!
- real xoff = offsetCF.xOffCF[k_thread];
- real yoff = offsetCF.yOffCF[k_thread];
- real zoff = offsetCF.zOffCF[k_thread];
+ real xoff = offsetCF.xOffCF[nodeIndex];
+ real yoff = offsetCF.yOffCF[nodeIndex];
+ real zoff = offsetCF.zOffCF[nodeIndex];
real xoff_sq = xoff * xoff;
real yoff_sq = yoff * yoff;
@@ -632,14 +755,29 @@ __global__ void scaleCF_compressible(
((xoff != c0o1) || (yoff != c0o1) || (zoff != c0o1))
? c0o1
: -c3o1 * (a_100 * a_100 + b_010 * b_010 + c_001 * c_001) - c6o1 * (b_100 * a_010 + c_100 * a_001 + c_010 * b_001);
- d_000 = ( drho_PPM + drho_PPP + drho_MPM + drho_MPP + drho_PMM + drho_PMP + drho_MMM + drho_MMP) * c1o8;
- d_100 = ( drho_PPM + drho_PPP - drho_MPM - drho_MPP + drho_PMM + drho_PMP - drho_MMM - drho_MMP) * c1o4;
- d_010 = ( drho_PPM + drho_PPP + drho_MPM + drho_MPP - drho_PMM - drho_PMP - drho_MMM - drho_MMP) * c1o4;
- d_001 = (-drho_PPM + drho_PPP - drho_MPM + drho_MPP - drho_PMM + drho_PMP - drho_MMM + drho_MMP) * c1o4;
- d_110 = ( drho_PPM + drho_PPP - drho_MPM - drho_MPP - drho_PMM - drho_PMP + drho_MMM + drho_MMP) * c1o2;
- d_101 = (-drho_PPM + drho_PPP + drho_MPM - drho_MPP - drho_PMM + drho_PMP + drho_MMM - drho_MMP) * c1o2;
- d_011 = (-drho_PPM + drho_PPP - drho_MPM + drho_MPP + drho_PMM - drho_PMP + drho_MMM - drho_MMP) * c1o2;
- d_111 = -drho_PPM + drho_PPP + drho_MPM - drho_MPP + drho_PMM - drho_PMP - drho_MMM + drho_MMP;
+ // d_000 = ( drho_PPM + drho_PPP + drho_MPM + drho_MPP + drho_PMM + drho_PMP + drho_MMM + drho_MMP) * c1o8;
+ d_000 = c1o8 * (((drho_PPP + drho_MMM) + (drho_PPM + drho_MMP)) + ((drho_PMM + drho_MPP) + (drho_PMP + drho_MPM)));
+
+ // d_100 = ( drho_PPM + drho_PPP - drho_MPM - drho_MPP + drho_PMM + drho_PMP - drho_MMM - drho_MMP) * c1o4;
+ d_100 = c1o4 * (((drho_PPP - drho_MMM) + (drho_PPM - drho_MMP)) + ((drho_PMM - drho_MPP) + (drho_PMP - drho_MPM)));
+
+ // d_010 = ( drho_PPM + drho_PPP + drho_MPM + drho_MPP - drho_PMM - drho_PMP - drho_MMM - drho_MMP) * c1o4;
+ d_010 = c1o4 * (((drho_PPP - drho_MMM) + (drho_PPM - drho_MMP)) + ((drho_MPP - drho_PMM) + (drho_MPM - drho_PMP)));
+
+ // d_001 = (-drho_PPM + drho_PPP - drho_MPM + drho_MPP - drho_PMM + drho_PMP - drho_MMM + drho_MMP) * c1o4;
+ d_001 = c1o4 * (((drho_PPP - drho_MMM) + (drho_MMP - drho_PPM)) + ((drho_MPP - drho_PMM) + (drho_PMP - drho_MPM)));
+
+ // d_110 = ( drho_PPM + drho_PPP - drho_MPM - drho_MPP - drho_PMM - drho_PMP + drho_MMM + drho_MMP) * c1o2;
+ d_110 = c1o2 * (((drho_PPP + drho_MMM) + (drho_PPM + drho_MMP)) - ((drho_PMM + drho_MPP) + (drho_PMP + drho_MPM)));
+
+ // d_101 = (-drho_PPM + drho_PPP + drho_MPM - drho_MPP - drho_PMM + drho_PMP + drho_MMM - drho_MMP) * c1o2;
+ d_101 = c1o2 * (((drho_PPP + drho_MMM) - (drho_PPM + drho_MMP)) + ((drho_PMP + drho_MPM) - (drho_PMM + drho_MPP)));
+
+ // d_011 = (-drho_PPM + drho_PPP - drho_MPM + drho_MPP + drho_PMM - drho_PMP + drho_MMM - drho_MMP) * c1o2;
+ d_011 = c1o2 * (((drho_PPP + drho_MMM) - (drho_PPM + drho_MMP)) + ((drho_PMM + drho_MPP) - (drho_PMP + drho_MPM)));
+
+ // d_111 = -drho_PPM + drho_PPP + drho_MPM - drho_MPP + drho_PMM - drho_PMP - drho_MMM + drho_MMP;
+ d_111 = (((drho_PPP - drho_MMM) + (drho_MMP - drho_PPM)) + ((drho_PMM - drho_MPP) + (drho_MPM - drho_PMP)));
//////////////////////////////////////////////////////////////////////////
//! - Extrapolation for refinement in to the wall (polynomial coefficients)
@@ -768,7 +906,7 @@ __global__ void scaleCF_compressible(
//////////////////////////////////////////////////////////////////////////
// index of the base node and its neighbors
- k_base_000 = indicesFineMMM[k_thread];
+ k_base_000 = indicesFineMMM[nodeIndex];
k_base_M00 = neighborXfine [k_base_000];
k_base_0M0 = neighborYfine [k_base_000];
k_base_00M = neighborZfine [k_base_000];
diff --git a/src/gpu/VirtualFluids_GPU/GPU/GridScaling/scaleFC_compressible.cu b/src/gpu/VirtualFluids_GPU/GPU/GridScaling/scaleFC_compressible.cu
index 3ab8b9d20279eff341ca42d20cee9fe7550a2039..e7d999d108e59bca98bf87b813f9479f1c601266 100644
--- a/src/gpu/VirtualFluids_GPU/GPU/GridScaling/scaleFC_compressible.cu
+++ b/src/gpu/VirtualFluids_GPU/GPU/GridScaling/scaleFC_compressible.cu
@@ -31,12 +31,13 @@
//! \author Martin Schoenherr, Anna Wellmann
//=======================================================================================
-#include "Kernel/Utilities/DistributionHelper.cuh"
-#include "Kernel/Utilities/ChimeraTransformation.h"
-#include "Kernel/Utilities/ScalingHelperFunctions.h"
+#include "LBM/GPUHelperFunctions/ChimeraTransformation.h"
+#include "LBM/GPUHelperFunctions/KernelUtilities.h"
+#include "LBM/GPUHelperFunctions/ScalingUtilities.h"
using namespace vf::lbm::constant;
using namespace vf::lbm::dir;
+using namespace vf::gpu;
//////////////////////////////////////////////////////////////////////////
//! \brief Interpolate from fine to coarse
@@ -54,8 +55,8 @@ __global__ void scaleFC_compressible(
unsigned int *neighborXfine,
unsigned int *neighborYfine,
unsigned int *neighborZfine,
- unsigned int numberOfLBnodesCoarse,
- unsigned int numberOfLBnodesFine,
+ unsigned long long numberOfLBnodesCoarse,
+ unsigned long long numberOfLBnodesFine,
bool isEvenTimestep,
unsigned int *indicesCoarse000,
unsigned int *indicesFineMMM,
@@ -65,13 +66,13 @@ __global__ void scaleFC_compressible(
OffFC offsetFC)
{
////////////////////////////////////////////////////////////////////////////////
- //! - Get the thread index coordinates from threadIdx, blockIdx, blockDim and gridDim.
+ //! - Get the node index coordinates from threadIdx, blockIdx, blockDim and gridDim.
//!
- const unsigned k_thread = vf::gpu::getNodeIndex();
+ const unsigned nodeIndex = getNodeIndex();
//////////////////////////////////////////////////////////////////////////
//! - Return for non-interface node
- if (k_thread >= numberOfInterfaceNodes)
+ if (nodeIndex >= numberOfInterfaceNodes)
return;
//////////////////////////////////////////////////////////////////////////
@@ -80,8 +81,9 @@ __global__ void scaleFC_compressible(
//! href="https://doi.org/10.3390/computation5020019"><b>[ M. Geier et al. (2017),
//! DOI:10.3390/computation5020019 ]</b></a>
//!
- Distributions27 distFine = vf::gpu::getDistributionReferences27(distributionsFine, numberOfLBnodesFine, true);
- Distributions27 distCoarse = vf::gpu::getDistributionReferences27(distributionsCoarse, numberOfLBnodesCoarse, isEvenTimestep);
+ Distributions27 distFine, distCoarse;
+ getPointersToDistributions(distFine, distributionsFine, numberOfLBnodesFine, true);
+ getPointersToDistributions(distCoarse, distributionsCoarse, numberOfLBnodesCoarse, isEvenTimestep);
////////////////////////////////////////////////////////////////////////////////
//! - declare local variables for source nodes
@@ -117,7 +119,7 @@ __global__ void scaleFC_compressible(
// source node BSW = MMM
//////////////////////////////////////////////////////////////////////////
// index of the base node and its neighbors
- unsigned int k_base_000 = indicesFineMMM[k_thread];
+ unsigned int k_base_000 = indicesFineMMM[nodeIndex];
unsigned int k_base_M00 = neighborXfine [k_base_000];
unsigned int k_base_0M0 = neighborYfine [k_base_000];
unsigned int k_base_00M = neighborZfine [k_base_000];
@@ -278,115 +280,120 @@ __global__ void scaleFC_compressible(
real c_000, c_100, c_010, c_001, c_200, c_020, c_002, c_110, c_101, c_011;
real d_000, d_100, d_010, d_001, d_110, d_101, d_011;
- a_000 = (-kxxMyyFromfcNEQ_PPM - kxxMyyFromfcNEQ_PPP + kxxMyyFromfcNEQ_MPM + kxxMyyFromfcNEQ_MPP -
- kxxMyyFromfcNEQ_PMM - kxxMyyFromfcNEQ_PMP + kxxMyyFromfcNEQ_MMM + kxxMyyFromfcNEQ_MMP -
- kxxMzzFromfcNEQ_PPM - kxxMzzFromfcNEQ_PPP + kxxMzzFromfcNEQ_MPM + kxxMzzFromfcNEQ_MPP -
- kxxMzzFromfcNEQ_PMM - kxxMzzFromfcNEQ_PMP + kxxMzzFromfcNEQ_MMM + kxxMzzFromfcNEQ_MMP -
- c2o1 * kxyFromfcNEQ_PPM - c2o1 * kxyFromfcNEQ_PPP - c2o1 * kxyFromfcNEQ_MPM - c2o1 * kxyFromfcNEQ_MPP +
- c2o1 * kxyFromfcNEQ_PMM + c2o1 * kxyFromfcNEQ_PMP + c2o1 * kxyFromfcNEQ_MMM + c2o1 * kxyFromfcNEQ_MMP +
- c2o1 * kxzFromfcNEQ_PPM - c2o1 * kxzFromfcNEQ_PPP + c2o1 * kxzFromfcNEQ_MPM - c2o1 * kxzFromfcNEQ_MPP +
- c2o1 * kxzFromfcNEQ_PMM - c2o1 * kxzFromfcNEQ_PMP + c2o1 * kxzFromfcNEQ_MMM - c2o1 * kxzFromfcNEQ_MMP +
- c8o1 * vx1_PPM + c8o1 * vx1_PPP + c8o1 * vx1_MPM + c8o1 * vx1_MPP + c8o1 * vx1_PMM + c8o1 * vx1_PMP +
- c8o1 * vx1_MMM + c8o1 * vx1_MMP + c2o1 * vx2_PPM + c2o1 * vx2_PPP - c2o1 * vx2_MPM - c2o1 * vx2_MPP -
- c2o1 * vx2_PMM - c2o1 * vx2_PMP + c2o1 * vx2_MMM + c2o1 * vx2_MMP - c2o1 * vx3_PPM + c2o1 * vx3_PPP +
- c2o1 * vx3_MPM - c2o1 * vx3_MPP - c2o1 * vx3_PMM + c2o1 * vx3_PMP + c2o1 * vx3_MMM - c2o1 * vx3_MMP) /
- c64o1;
- b_000 = (c2o1 * kxxMyyFromfcNEQ_PPM + c2o1 * kxxMyyFromfcNEQ_PPP + c2o1 * kxxMyyFromfcNEQ_MPM +
- c2o1 * kxxMyyFromfcNEQ_MPP - c2o1 * kxxMyyFromfcNEQ_PMM - c2o1 * kxxMyyFromfcNEQ_PMP -
- c2o1 * kxxMyyFromfcNEQ_MMM - c2o1 * kxxMyyFromfcNEQ_MMP - kxxMzzFromfcNEQ_PPM - kxxMzzFromfcNEQ_PPP -
- kxxMzzFromfcNEQ_MPM - kxxMzzFromfcNEQ_MPP + kxxMzzFromfcNEQ_PMM + kxxMzzFromfcNEQ_PMP +
- kxxMzzFromfcNEQ_MMM + kxxMzzFromfcNEQ_MMP - c2o1 * kxyFromfcNEQ_PPM - c2o1 * kxyFromfcNEQ_PPP +
- c2o1 * kxyFromfcNEQ_MPM + c2o1 * kxyFromfcNEQ_MPP - c2o1 * kxyFromfcNEQ_PMM - c2o1 * kxyFromfcNEQ_PMP +
- c2o1 * kxyFromfcNEQ_MMM + c2o1 * kxyFromfcNEQ_MMP + c2o1 * kyzFromfcNEQ_PPM - c2o1 * kyzFromfcNEQ_PPP +
- c2o1 * kyzFromfcNEQ_MPM - c2o1 * kyzFromfcNEQ_MPP + c2o1 * kyzFromfcNEQ_PMM - c2o1 * kyzFromfcNEQ_PMP +
- c2o1 * kyzFromfcNEQ_MMM - c2o1 * kyzFromfcNEQ_MMP + c2o1 * vx1_PPM + c2o1 * vx1_PPP - c2o1 * vx1_MPM -
- c2o1 * vx1_MPP - c2o1 * vx1_PMM - c2o1 * vx1_PMP + c2o1 * vx1_MMM + c2o1 * vx1_MMP + c8o1 * vx2_PPM +
- c8o1 * vx2_PPP + c8o1 * vx2_MPM + c8o1 * vx2_MPP + c8o1 * vx2_PMM + c8o1 * vx2_PMP + c8o1 * vx2_MMM +
- c8o1 * vx2_MMP - c2o1 * vx3_PPM + c2o1 * vx3_PPP - c2o1 * vx3_MPM + c2o1 * vx3_MPP + c2o1 * vx3_PMM -
- c2o1 * vx3_PMP + c2o1 * vx3_MMM - c2o1 * vx3_MMP) /
- c64o1;
- c_000 = (kxxMyyFromfcNEQ_PPM - kxxMyyFromfcNEQ_PPP + kxxMyyFromfcNEQ_MPM - kxxMyyFromfcNEQ_MPP +
- kxxMyyFromfcNEQ_PMM - kxxMyyFromfcNEQ_PMP + kxxMyyFromfcNEQ_MMM - kxxMyyFromfcNEQ_MMP -
- c2o1 * kxxMzzFromfcNEQ_PPM + c2o1 * kxxMzzFromfcNEQ_PPP - c2o1 * kxxMzzFromfcNEQ_MPM +
- c2o1 * kxxMzzFromfcNEQ_MPP - c2o1 * kxxMzzFromfcNEQ_PMM + c2o1 * kxxMzzFromfcNEQ_PMP -
- c2o1 * kxxMzzFromfcNEQ_MMM + c2o1 * kxxMzzFromfcNEQ_MMP - c2o1 * kxzFromfcNEQ_PPM -
- c2o1 * kxzFromfcNEQ_PPP + c2o1 * kxzFromfcNEQ_MPM + c2o1 * kxzFromfcNEQ_MPP - c2o1 * kxzFromfcNEQ_PMM -
- c2o1 * kxzFromfcNEQ_PMP + c2o1 * kxzFromfcNEQ_MMM + c2o1 * kxzFromfcNEQ_MMP - c2o1 * kyzFromfcNEQ_PPM -
- c2o1 * kyzFromfcNEQ_PPP - c2o1 * kyzFromfcNEQ_MPM - c2o1 * kyzFromfcNEQ_MPP + c2o1 * kyzFromfcNEQ_PMM +
- c2o1 * kyzFromfcNEQ_PMP + c2o1 * kyzFromfcNEQ_MMM + c2o1 * kyzFromfcNEQ_MMP - c2o1 * vx1_PPM +
- c2o1 * vx1_PPP + c2o1 * vx1_MPM - c2o1 * vx1_MPP - c2o1 * vx1_PMM + c2o1 * vx1_PMP + c2o1 * vx1_MMM -
- c2o1 * vx1_MMP - c2o1 * vx2_PPM + c2o1 * vx2_PPP - c2o1 * vx2_MPM + c2o1 * vx2_MPP + c2o1 * vx2_PMM -
- c2o1 * vx2_PMP + c2o1 * vx2_MMM - c2o1 * vx2_MMP + c8o1 * vx3_PPM + c8o1 * vx3_PPP + c8o1 * vx3_MPM +
- c8o1 * vx3_MPP + c8o1 * vx3_PMM + c8o1 * vx3_PMP + c8o1 * vx3_MMM + c8o1 * vx3_MMP) /
- c64o1;
- a_100 = (vx1_PPM + vx1_PPP - vx1_MPM - vx1_MPP + vx1_PMM + vx1_PMP - vx1_MMM - vx1_MMP) / c4o1;
- b_100 = (vx2_PPM + vx2_PPP - vx2_MPM - vx2_MPP + vx2_PMM + vx2_PMP - vx2_MMM - vx2_MMP) / c4o1;
- c_100 = (vx3_PPM + vx3_PPP - vx3_MPM - vx3_MPP + vx3_PMM + vx3_PMP - vx3_MMM - vx3_MMP) / c4o1;
- a_200 = (kxxMyyFromfcNEQ_PPM + kxxMyyFromfcNEQ_PPP - kxxMyyFromfcNEQ_MPM - kxxMyyFromfcNEQ_MPP +
- kxxMyyFromfcNEQ_PMM + kxxMyyFromfcNEQ_PMP - kxxMyyFromfcNEQ_MMM - kxxMyyFromfcNEQ_MMP +
- kxxMzzFromfcNEQ_PPM + kxxMzzFromfcNEQ_PPP - kxxMzzFromfcNEQ_MPM - kxxMzzFromfcNEQ_MPP +
- kxxMzzFromfcNEQ_PMM + kxxMzzFromfcNEQ_PMP - kxxMzzFromfcNEQ_MMM - kxxMzzFromfcNEQ_MMP + c2o1 * vx2_PPM +
- c2o1 * vx2_PPP - c2o1 * vx2_MPM - c2o1 * vx2_MPP - c2o1 * vx2_PMM - c2o1 * vx2_PMP + c2o1 * vx2_MMM +
- c2o1 * vx2_MMP - c2o1 * vx3_PPM + c2o1 * vx3_PPP + c2o1 * vx3_MPM - c2o1 * vx3_MPP - c2o1 * vx3_PMM +
- c2o1 * vx3_PMP + c2o1 * vx3_MMM - c2o1 * vx3_MMP) /
- c16o1;
- b_200 = (kxyFromfcNEQ_PPM + kxyFromfcNEQ_PPP - kxyFromfcNEQ_MPM - kxyFromfcNEQ_MPP + kxyFromfcNEQ_PMM +
- kxyFromfcNEQ_PMP - kxyFromfcNEQ_MMM - kxyFromfcNEQ_MMP - c2o1 * vx1_PPM - c2o1 * vx1_PPP +
- c2o1 * vx1_MPM + c2o1 * vx1_MPP + c2o1 * vx1_PMM + c2o1 * vx1_PMP - c2o1 * vx1_MMM - c2o1 * vx1_MMP) /
- c8o1;
- c_200 = (kxzFromfcNEQ_PPM + kxzFromfcNEQ_PPP - kxzFromfcNEQ_MPM - kxzFromfcNEQ_MPP + kxzFromfcNEQ_PMM +
- kxzFromfcNEQ_PMP - kxzFromfcNEQ_MMM - kxzFromfcNEQ_MMP + c2o1 * vx1_PPM - c2o1 * vx1_PPP -
- c2o1 * vx1_MPM + c2o1 * vx1_MPP + c2o1 * vx1_PMM - c2o1 * vx1_PMP - c2o1 * vx1_MMM + c2o1 * vx1_MMP) /
- c8o1;
- a_010 = (vx1_PPM + vx1_PPP + vx1_MPM + vx1_MPP - vx1_PMM - vx1_PMP - vx1_MMM - vx1_MMP) / c4o1;
- b_010 = (vx2_PPM + vx2_PPP + vx2_MPM + vx2_MPP - vx2_PMM - vx2_PMP - vx2_MMM - vx2_MMP) / c4o1;
- c_010 = (vx3_PPM + vx3_PPP + vx3_MPM + vx3_MPP - vx3_PMM - vx3_PMP - vx3_MMM - vx3_MMP) / c4o1;
- a_020 = (kxyFromfcNEQ_PPM + kxyFromfcNEQ_PPP + kxyFromfcNEQ_MPM + kxyFromfcNEQ_MPP - kxyFromfcNEQ_PMM -
- kxyFromfcNEQ_PMP - kxyFromfcNEQ_MMM - kxyFromfcNEQ_MMP - c2o1 * vx2_PPM - c2o1 * vx2_PPP +
- c2o1 * vx2_MPM + c2o1 * vx2_MPP + c2o1 * vx2_PMM + c2o1 * vx2_PMP - c2o1 * vx2_MMM - c2o1 * vx2_MMP) /
- c8o1;
- b_020 = (-c2o1 * kxxMyyFromfcNEQ_PPM - c2o1 * kxxMyyFromfcNEQ_PPP - c2o1 * kxxMyyFromfcNEQ_MPM -
- c2o1 * kxxMyyFromfcNEQ_MPP + c2o1 * kxxMyyFromfcNEQ_PMM + c2o1 * kxxMyyFromfcNEQ_PMP +
- c2o1 * kxxMyyFromfcNEQ_MMM + c2o1 * kxxMyyFromfcNEQ_MMP + kxxMzzFromfcNEQ_PPM + kxxMzzFromfcNEQ_PPP +
- kxxMzzFromfcNEQ_MPM + kxxMzzFromfcNEQ_MPP - kxxMzzFromfcNEQ_PMM - kxxMzzFromfcNEQ_PMP -
- kxxMzzFromfcNEQ_MMM - kxxMzzFromfcNEQ_MMP + c2o1 * vx1_PPM + c2o1 * vx1_PPP - c2o1 * vx1_MPM -
- c2o1 * vx1_MPP - c2o1 * vx1_PMM - c2o1 * vx1_PMP + c2o1 * vx1_MMM + c2o1 * vx1_MMP - c2o1 * vx3_PPM +
- c2o1 * vx3_PPP - c2o1 * vx3_MPM + c2o1 * vx3_MPP + c2o1 * vx3_PMM - c2o1 * vx3_PMP + c2o1 * vx3_MMM -
- c2o1 * vx3_MMP) /
- c16o1;
- c_020 = (kyzFromfcNEQ_PPM + kyzFromfcNEQ_PPP + kyzFromfcNEQ_MPM + kyzFromfcNEQ_MPP - kyzFromfcNEQ_PMM -
- kyzFromfcNEQ_PMP - kyzFromfcNEQ_MMM - kyzFromfcNEQ_MMP + c2o1 * vx2_PPM - c2o1 * vx2_PPP +
- c2o1 * vx2_MPM - c2o1 * vx2_MPP - c2o1 * vx2_PMM + c2o1 * vx2_PMP - c2o1 * vx2_MMM + c2o1 * vx2_MMP) /
- c8o1;
- a_001 = (-vx1_PPM + vx1_PPP - vx1_MPM + vx1_MPP - vx1_PMM + vx1_PMP - vx1_MMM + vx1_MMP) / c4o1;
- b_001 = (-vx2_PPM + vx2_PPP - vx2_MPM + vx2_MPP - vx2_PMM + vx2_PMP - vx2_MMM + vx2_MMP) / c4o1;
- c_001 = (-vx3_PPM + vx3_PPP - vx3_MPM + vx3_MPP - vx3_PMM + vx3_PMP - vx3_MMM + vx3_MMP) / c4o1;
- a_002 = (-kxzFromfcNEQ_PPM + kxzFromfcNEQ_PPP - kxzFromfcNEQ_MPM + kxzFromfcNEQ_MPP - kxzFromfcNEQ_PMM +
- kxzFromfcNEQ_PMP - kxzFromfcNEQ_MMM + kxzFromfcNEQ_MMP + c2o1 * vx3_PPM - c2o1 * vx3_PPP -
- c2o1 * vx3_MPM + c2o1 * vx3_MPP + c2o1 * vx3_PMM - c2o1 * vx3_PMP - c2o1 * vx3_MMM + c2o1 * vx3_MMP) /
- c8o1;
- b_002 = (-kyzFromfcNEQ_PPM + kyzFromfcNEQ_PPP - kyzFromfcNEQ_MPM + kyzFromfcNEQ_MPP - kyzFromfcNEQ_PMM +
- kyzFromfcNEQ_PMP - kyzFromfcNEQ_MMM + kyzFromfcNEQ_MMP + c2o1 * vx3_PPM - c2o1 * vx3_PPP +
- c2o1 * vx3_MPM - c2o1 * vx3_MPP - c2o1 * vx3_PMM + c2o1 * vx3_PMP - c2o1 * vx3_MMM + c2o1 * vx3_MMP) /
- c8o1;
- c_002 = (-kxxMyyFromfcNEQ_PPM + kxxMyyFromfcNEQ_PPP - kxxMyyFromfcNEQ_MPM + kxxMyyFromfcNEQ_MPP -
- kxxMyyFromfcNEQ_PMM + kxxMyyFromfcNEQ_PMP - kxxMyyFromfcNEQ_MMM + kxxMyyFromfcNEQ_MMP +
- c2o1 * kxxMzzFromfcNEQ_PPM - c2o1 * kxxMzzFromfcNEQ_PPP + c2o1 * kxxMzzFromfcNEQ_MPM -
- c2o1 * kxxMzzFromfcNEQ_MPP + c2o1 * kxxMzzFromfcNEQ_PMM - c2o1 * kxxMzzFromfcNEQ_PMP +
- c2o1 * kxxMzzFromfcNEQ_MMM - c2o1 * kxxMzzFromfcNEQ_MMP - c2o1 * vx1_PPM + c2o1 * vx1_PPP +
- c2o1 * vx1_MPM - c2o1 * vx1_MPP - c2o1 * vx1_PMM + c2o1 * vx1_PMP + c2o1 * vx1_MMM - c2o1 * vx1_MMP -
- c2o1 * vx2_PPM + c2o1 * vx2_PPP - c2o1 * vx2_MPM + c2o1 * vx2_MPP + c2o1 * vx2_PMM - c2o1 * vx2_PMP +
- c2o1 * vx2_MMM - c2o1 * vx2_MMP) /
- c16o1;
- a_110 = (vx1_PPM + vx1_PPP - vx1_MPM - vx1_MPP - vx1_PMM - vx1_PMP + vx1_MMM + vx1_MMP) / c2o1;
- b_110 = (vx2_PPM + vx2_PPP - vx2_MPM - vx2_MPP - vx2_PMM - vx2_PMP + vx2_MMM + vx2_MMP) / c2o1;
- c_110 = (vx3_PPM + vx3_PPP - vx3_MPM - vx3_MPP - vx3_PMM - vx3_PMP + vx3_MMM + vx3_MMP) / c2o1;
- a_101 = (-vx1_PPM + vx1_PPP + vx1_MPM - vx1_MPP - vx1_PMM + vx1_PMP + vx1_MMM - vx1_MMP) / c2o1;
- b_101 = (-vx2_PPM + vx2_PPP + vx2_MPM - vx2_MPP - vx2_PMM + vx2_PMP + vx2_MMM - vx2_MMP) / c2o1;
- c_101 = (-vx3_PPM + vx3_PPP + vx3_MPM - vx3_MPP - vx3_PMM + vx3_PMP + vx3_MMM - vx3_MMP) / c2o1;
- a_011 = (-vx1_PPM + vx1_PPP - vx1_MPM + vx1_MPP + vx1_PMM - vx1_PMP + vx1_MMM - vx1_MMP) / c2o1;
- b_011 = (-vx2_PPM + vx2_PPP - vx2_MPM + vx2_MPP + vx2_PMM - vx2_PMP + vx2_MMM - vx2_MMP) / c2o1;
- c_011 = (-vx3_PPM + vx3_PPP - vx3_MPM + vx3_MPP + vx3_PMM - vx3_PMP + vx3_MMM - vx3_MMP) / c2o1;
+ a_000 = c1o64 * (
+ c2o1 * (
+ ((kxyFromfcNEQ_MMM - kxyFromfcNEQ_PPP) + (kxyFromfcNEQ_MMP - kxyFromfcNEQ_PPM)) + ((kxyFromfcNEQ_PMM - kxyFromfcNEQ_MPP) + (kxyFromfcNEQ_PMP - kxyFromfcNEQ_MPM)) +
+ ((kxzFromfcNEQ_MMM - kxzFromfcNEQ_PPP) + (kxzFromfcNEQ_PPM - kxzFromfcNEQ_MMP)) + ((kxzFromfcNEQ_PMM - kxzFromfcNEQ_MPP) + (kxzFromfcNEQ_MPM - kxzFromfcNEQ_PMP)) +
+ ((vx2_PPP + vx2_MMM) + (vx2_PPM + vx2_MMP)) - ((vx2_MPP + vx2_PMM) + (vx2_MPM + vx2_PMP)) +
+ ((vx3_PPP + vx3_MMM) - (vx3_PPM + vx3_MMP)) + ((vx3_PMP + vx3_MPM) - (vx3_MPP + vx3_PMM))) +
+ c8o1 * (((vx1_PPP + vx1_MMM) + (vx1_PPM + vx1_MMP)) + ((vx1_MPP + vx1_PMM) + (vx1_PMP + vx1_MPM))) +
+ ((kxxMyyFromfcNEQ_MMM - kxxMyyFromfcNEQ_PPP) + (kxxMyyFromfcNEQ_MMP - kxxMyyFromfcNEQ_PPM)) +
+ ((kxxMyyFromfcNEQ_MPP - kxxMyyFromfcNEQ_PMM) + (kxxMyyFromfcNEQ_MPM - kxxMyyFromfcNEQ_PMP)) +
+ ((kxxMzzFromfcNEQ_MMM - kxxMzzFromfcNEQ_PPP) + (kxxMzzFromfcNEQ_MMP - kxxMzzFromfcNEQ_PPM)) +
+ ((kxxMzzFromfcNEQ_MPP - kxxMzzFromfcNEQ_PMM) + (kxxMzzFromfcNEQ_MPM - kxxMzzFromfcNEQ_PMP)));
+ b_000 = c1o64 * (
+ c2o1 * (
+ ((kxxMyyFromfcNEQ_PPP - kxxMyyFromfcNEQ_MMM) + (kxxMyyFromfcNEQ_PPM - kxxMyyFromfcNEQ_MMP)) +
+ ((kxxMyyFromfcNEQ_MPP - kxxMyyFromfcNEQ_PMM) + (kxxMyyFromfcNEQ_MPM - kxxMyyFromfcNEQ_PMP)) +
+ ((kxyFromfcNEQ_MMM - kxyFromfcNEQ_PPP) + (kxyFromfcNEQ_MMP - kxyFromfcNEQ_PPM)) +
+ ((kxyFromfcNEQ_MPP - kxyFromfcNEQ_PMM) + (kxyFromfcNEQ_MPM - kxyFromfcNEQ_PMP)) +
+ ((kyzFromfcNEQ_MMM - kyzFromfcNEQ_PPP) + (kyzFromfcNEQ_PPM - kyzFromfcNEQ_MMP)) +
+ ((kyzFromfcNEQ_PMM - kyzFromfcNEQ_MPP) + (kyzFromfcNEQ_MPM - kyzFromfcNEQ_PMP)) +
+ ((vx1_PPP + vx1_MMM) + (vx1_PPM + vx1_MMP)) - ((vx1_MPM + vx1_MPP) + (vx1_PMM + vx1_PMP)) +
+ ((vx3_PPP + vx3_MMM) - (vx3_PPM + vx3_MMP)) + ((vx3_MPP + vx3_PMM) - (vx3_MPM + vx3_PMP))) +
+ c8o1 * (((vx2_PPP + vx2_MMM) + (vx2_PPM + vx2_MMP)) + ((vx2_MPP + vx2_PMM) + (vx2_MPM + vx2_PMP))) +
+ ((kxxMzzFromfcNEQ_MMM - kxxMzzFromfcNEQ_PPP) + (kxxMzzFromfcNEQ_MMP - kxxMzzFromfcNEQ_PPM)) +
+ ((kxxMzzFromfcNEQ_PMM - kxxMzzFromfcNEQ_MPP) + (kxxMzzFromfcNEQ_PMP - kxxMzzFromfcNEQ_MPM)));
+ c_000 = c1o64 * (
+ c2o1 * (
+ ((kxxMzzFromfcNEQ_PPP - kxxMzzFromfcNEQ_MMM) + (kxxMzzFromfcNEQ_MMP - kxxMzzFromfcNEQ_PPM)) +
+ ((kxxMzzFromfcNEQ_MPP - kxxMzzFromfcNEQ_PMM) + (kxxMzzFromfcNEQ_PMP - kxxMzzFromfcNEQ_MPM)) +
+ ((kxzFromfcNEQ_MMM - kxzFromfcNEQ_PPP) + (kxzFromfcNEQ_MMP - kxzFromfcNEQ_PPM)) +
+ ((kxzFromfcNEQ_MPP - kxzFromfcNEQ_PMM) + (kxzFromfcNEQ_MPM - kxzFromfcNEQ_PMP)) +
+ ((kyzFromfcNEQ_MMM - kyzFromfcNEQ_PPP) + (kyzFromfcNEQ_MMP - kyzFromfcNEQ_PPM)) +
+ ((kyzFromfcNEQ_PMM - kyzFromfcNEQ_MPP) + (kyzFromfcNEQ_PMP - kyzFromfcNEQ_MPM)) +
+ ((vx1_PPP + vx1_MMM) - (vx1_MMP + vx1_PPM)) + ((vx1_MPM + vx1_PMP) - (vx1_MPP + vx1_PMM)) +
+ ((vx2_PPP + vx2_MMM) - (vx2_MMP + vx2_PPM)) + ((vx2_MPP + vx2_PMM) - (vx2_MPM + vx2_PMP))) +
+ c8o1 * (((vx3_PPP + vx3_MMM) + (vx3_PPM + vx3_MMP)) + ((vx3_PMM + vx3_MPP) + (vx3_PMP + vx3_MPM))) +
+ ((kxxMyyFromfcNEQ_MMM - kxxMyyFromfcNEQ_PPP) + (kxxMyyFromfcNEQ_PPM - kxxMyyFromfcNEQ_MMP)) +
+ ((kxxMyyFromfcNEQ_PMM - kxxMyyFromfcNEQ_MPP) + (kxxMyyFromfcNEQ_MPM - kxxMyyFromfcNEQ_PMP)));
+
+ a_100 = c1o4 * (((vx1_PPP - vx1_MMM) + (vx1_PPM - vx1_MMP)) + ((vx1_PMM - vx1_MPP) + (vx1_PMP - vx1_MPM)));
+ b_100 = c1o4 * (((vx2_PPP - vx2_MMM) + (vx2_PPM - vx2_MMP)) + ((vx2_PMM - vx2_MPP) + (vx2_PMP - vx2_MPM)));
+ c_100 = c1o4 * (((vx3_PPP - vx3_MMM) + (vx3_PPM - vx3_MMP)) + ((vx3_PMM - vx3_MPP) + (vx3_PMP - vx3_MPM)));
+
+ a_200 = c1o16 * (
+ c2o1 * (
+ ((vx2_PPP + vx2_MMM) + (vx2_PPM - vx2_MPP)) + ((vx2_MMP - vx2_PMM) - (vx2_MPM + vx2_PMP)) +
+ ((vx3_PPP + vx3_MMM) - (vx3_PPM + vx3_MPP)) + ((vx3_MPM + vx3_PMP) - (vx3_MMP + vx3_PMM))) +
+ ((kxxMyyFromfcNEQ_PPP - kxxMyyFromfcNEQ_MMM) + (kxxMyyFromfcNEQ_PPM - kxxMyyFromfcNEQ_MMP)) +
+ ((kxxMyyFromfcNEQ_PMM - kxxMyyFromfcNEQ_MPP) + (kxxMyyFromfcNEQ_PMP - kxxMyyFromfcNEQ_MPM)) +
+ ((kxxMzzFromfcNEQ_PPP - kxxMzzFromfcNEQ_MMM) + (kxxMzzFromfcNEQ_PPM - kxxMzzFromfcNEQ_MMP)) +
+ ((kxxMzzFromfcNEQ_PMM - kxxMzzFromfcNEQ_MPP) + (kxxMzzFromfcNEQ_PMP - kxxMzzFromfcNEQ_MPM)));
+ b_200 = c1o8 * (
+ c2o1 * (
+ -((vx1_PPP + vx1_MMM) + (vx1_PPM + vx1_MMP)) + ((vx1_MPP + vx1_PMM) + (vx1_MPM + vx1_PMP))) +
+ ((kxyFromfcNEQ_PPP - kxyFromfcNEQ_MMM) + (kxyFromfcNEQ_PPM - kxyFromfcNEQ_MMP)) +
+ ((kxyFromfcNEQ_PMM - kxyFromfcNEQ_MPP) + (kxyFromfcNEQ_PMP - kxyFromfcNEQ_MPM)));
+ c_200 = c1o8 * (
+ c2o1 * (
+ ((vx1_PPM + vx1_MMP) - (vx1_PPP + vx1_MMM)) + ((vx1_MPP + vx1_PMM) - (vx1_MPM + vx1_PMP))) +
+ ((kxzFromfcNEQ_PPP - kxzFromfcNEQ_MMM) + (kxzFromfcNEQ_PPM - kxzFromfcNEQ_MMP)) +
+ ((kxzFromfcNEQ_PMM - kxzFromfcNEQ_MPP) + (kxzFromfcNEQ_PMP - kxzFromfcNEQ_MPM)));
+
+ a_010 = c1o4 * (((vx1_PPP - vx1_MMM) + (vx1_PPM - vx1_MMP)) + ((vx1_MPP - vx1_PMM) + (vx1_MPM - vx1_PMP)));
+ b_010 = c1o4 * (((vx2_PPP - vx2_MMM) + (vx2_PPM - vx2_MMP)) + ((vx2_MPP - vx2_PMM) + (vx2_MPM - vx2_PMP)));
+ c_010 = c1o4 * (((vx3_PPP - vx3_MMM) + (vx3_PPM - vx3_MMP)) + ((vx3_MPP - vx3_PMM) + (vx3_MPM - vx3_PMP)));
+
+ a_020 = c1o8 * (
+ c2o1 * (-((vx2_PPP + vx2_MMM) + (vx2_MMP + vx2_PPM)) + ((vx2_MPP + vx2_PMM) + (vx2_MPM + vx2_PMP))) +
+ ((kxyFromfcNEQ_PPP - kxyFromfcNEQ_MMM) + (kxyFromfcNEQ_PPM - kxyFromfcNEQ_MMP)) +
+ ((kxyFromfcNEQ_MPP - kxyFromfcNEQ_PMM) + (kxyFromfcNEQ_MPM - kxyFromfcNEQ_PMP)));
+ b_020 = c1o16 * (
+ c2o1 * (
+ ((kxxMyyFromfcNEQ_MMM - kxxMyyFromfcNEQ_PPP) + (kxxMyyFromfcNEQ_MMP - kxxMyyFromfcNEQ_PPM)) +
+ ((kxxMyyFromfcNEQ_PMM - kxxMyyFromfcNEQ_MPP) + (kxxMyyFromfcNEQ_PMP - kxxMyyFromfcNEQ_MPM)) +
+ ((vx1_PPP + vx1_MMM) + (vx1_PPM + vx1_MMP)) - ((vx1_MPP + vx1_PMM) + (vx1_PMP + vx1_MPM)) +
+ ((vx3_PPP + vx3_MMM) - (vx3_PPM + vx3_MMP)) + ((vx3_MPP + vx3_PMM) - (vx3_MPM + vx3_PMP))) +
+ ((kxxMzzFromfcNEQ_PPP - kxxMzzFromfcNEQ_MMM) + (kxxMzzFromfcNEQ_PPM - kxxMzzFromfcNEQ_MMP)) +
+ ((kxxMzzFromfcNEQ_MPP - kxxMzzFromfcNEQ_PMM) + (kxxMzzFromfcNEQ_MPM - kxxMzzFromfcNEQ_PMP)));
+ c_020 = c1o8 * (
+ c2o1 * (((vx2_MMP + vx2_PPM) - (vx2_PPP + vx2_MMM)) + ((vx2_PMP + vx2_MPM) - (vx2_MPP + vx2_PMM))) +
+ ((kyzFromfcNEQ_PPP - kyzFromfcNEQ_MMM) + (kyzFromfcNEQ_PPM - kyzFromfcNEQ_MMP)) +
+ ((kyzFromfcNEQ_MPP - kyzFromfcNEQ_PMM) + (kyzFromfcNEQ_MPM - kyzFromfcNEQ_PMP)));
+
+ a_001 = c1o4 * (((vx1_PPP - vx1_MMM) + (vx1_MMP - vx1_PPM)) + ((vx1_MPP - vx1_PMM) + (vx1_PMP - vx1_MPM)));
+ b_001 = c1o4 * (((vx2_PPP - vx2_MMM) + (vx2_MMP - vx2_PPM)) + ((vx2_MPP - vx2_PMM) + (vx2_PMP - vx2_MPM)));
+ c_001 = c1o4 * (((vx3_PPP - vx3_MMM) + (vx3_MMP - vx3_PPM)) + ((vx3_MPP - vx3_PMM) + (vx3_PMP - vx3_MPM)));
+
+ a_002 = c1o8 * (
+ c2o1 * (((vx3_PPM + vx3_MMP) - (vx3_PPP + vx3_MMM)) + ((vx3_MPP + vx3_PMM) - (vx3_PMP + vx3_MPM))) +
+ ((kxzFromfcNEQ_PPP - kxzFromfcNEQ_MMM) + (kxzFromfcNEQ_MMP - kxzFromfcNEQ_PPM)) +
+ ((kxzFromfcNEQ_PMP - kxzFromfcNEQ_MPM) + (kxzFromfcNEQ_MPP - kxzFromfcNEQ_PMM)));
+ b_002 = c1o8 * (
+ c2o1 * (((vx3_PPM + vx3_MMP) - (vx3_PPP + vx3_MMM)) + ((vx3_MPM + vx3_PMP) - (vx3_PMM + vx3_MPP))) +
+ ((kyzFromfcNEQ_PPP - kyzFromfcNEQ_MMM) + (kyzFromfcNEQ_MMP - kyzFromfcNEQ_PPM)) +
+ ((kyzFromfcNEQ_PMP - kyzFromfcNEQ_MPM) + (kyzFromfcNEQ_MPP - kyzFromfcNEQ_PMM)));
+ c_002 = c1o16 * (
+ c2o1 * (
+ ((kxxMzzFromfcNEQ_MMM - kxxMzzFromfcNEQ_PPP) + (kxxMzzFromfcNEQ_PPM - kxxMzzFromfcNEQ_MMP)) +
+ ((kxxMzzFromfcNEQ_MPM - kxxMzzFromfcNEQ_PMP) + (kxxMzzFromfcNEQ_PMM - kxxMzzFromfcNEQ_MPP)) +
+ ((vx1_PPP + vx1_MMM) - (vx1_MMP + vx1_PPM)) + ((vx1_MPM + vx1_PMP) - (vx1_PMM + vx1_MPP)) +
+ ((vx2_PPP + vx2_MMM) - (vx2_MMP + vx2_PPM)) + ((vx2_PMM + vx2_MPP) - (vx2_MPM + vx2_PMP))) +
+ ((kxxMyyFromfcNEQ_PPP - kxxMyyFromfcNEQ_MMM) + (kxxMyyFromfcNEQ_MMP - kxxMyyFromfcNEQ_PPM)) +
+ ((kxxMyyFromfcNEQ_PMP - kxxMyyFromfcNEQ_MPM) + (kxxMyyFromfcNEQ_MPP - kxxMyyFromfcNEQ_PMM)));
+
+ a_110 = c1o2 * (((vx1_PPP + vx1_MMM) + (vx1_MMP + vx1_PPM)) - ((vx1_MPM + vx1_PMP) + (vx1_PMM + vx1_MPP)));
+ b_110 = c1o2 * (((vx2_PPP + vx2_MMM) + (vx2_MMP + vx2_PPM)) - ((vx2_MPM + vx2_PMP) + (vx2_PMM + vx2_MPP)));
+ c_110 = c1o2 * (((vx3_PPP + vx3_MMM) + (vx3_MMP + vx3_PPM)) - ((vx3_MPM + vx3_PMP) + (vx3_PMM + vx3_MPP)));
+
+ a_101 = c1o2 * (((vx1_PPP + vx1_MMM) - (vx1_MMP + vx1_PPM)) + ((vx1_MPM + vx1_PMP) - (vx1_PMM + vx1_MPP)));
+ b_101 = c1o2 * (((vx2_PPP + vx2_MMM) - (vx2_MMP + vx2_PPM)) + ((vx2_MPM + vx2_PMP) - (vx2_PMM + vx2_MPP)));
+ c_101 = c1o2 * (((vx3_PPP + vx3_MMM) - (vx3_MMP + vx3_PPM)) + ((vx3_MPM + vx3_PMP) - (vx3_PMM + vx3_MPP)));
+
+ a_011 = c1o2 * (((vx1_PPP + vx1_MMM) - (vx1_MMP + vx1_PPM)) + ((vx1_PMM + vx1_MPP) - (vx1_MPM + vx1_PMP)));
+ b_011 = c1o2 * (((vx2_PPP + vx2_MMM) - (vx2_MMP + vx2_PPM)) + ((vx2_PMM + vx2_MPP) - (vx2_MPM + vx2_PMP)));
+ c_011 = c1o2 * (((vx3_PPP + vx3_MMM) - (vx3_MMP + vx3_PPM)) + ((vx3_PMM + vx3_MPP) - (vx3_MPM + vx3_PMP)));
//////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
@@ -399,9 +406,9 @@ __global__ void scaleFC_compressible(
////////////////////////////////////////////////////////////////////////////////
//! - Set the relative position of the offset cell {-1, 0, 1}
//!
- real xoff = offsetFC.xOffFC[k_thread];
- real yoff = offsetFC.yOffFC[k_thread];
- real zoff = offsetFC.zOffFC[k_thread];
+ real xoff = offsetFC.xOffFC[nodeIndex];
+ real yoff = offsetFC.yOffFC[nodeIndex];
+ real zoff = offsetFC.zOffFC[nodeIndex];
real xoff_sq = xoff * xoff;
real yoff_sq = yoff * yoff;
@@ -412,15 +419,14 @@ __global__ void scaleFC_compressible(
//!
real LaplaceRho =
((xoff != c0o1) || (yoff != c0o1) || (zoff != c0o1))
- ? c0o1
- : -c3o1 * (a_100 * a_100 + b_010 * b_010 + c_001 * c_001) - c6o1 * (b_100 * a_010 + c_100 * a_001 + c_010 * b_001);
- d_000 = ( drho_PPM + drho_PPP + drho_MPM + drho_MPP + drho_PMM + drho_PMP + drho_MMM + drho_MMP - c2o1 * LaplaceRho) * c1o8;
- d_100 = ( drho_PPM + drho_PPP - drho_MPM - drho_MPP + drho_PMM + drho_PMP - drho_MMM - drho_MMP) * c1o4;
- d_010 = ( drho_PPM + drho_PPP + drho_MPM + drho_MPP - drho_PMM - drho_PMP - drho_MMM - drho_MMP) * c1o4;
- d_001 = (-drho_PPM + drho_PPP - drho_MPM + drho_MPP - drho_PMM + drho_PMP - drho_MMM + drho_MMP) * c1o4;
- d_110 = ( drho_PPM + drho_PPP - drho_MPM - drho_MPP - drho_PMM - drho_PMP + drho_MMM + drho_MMP) * c1o2;
- d_101 = (-drho_PPM + drho_PPP + drho_MPM - drho_MPP - drho_PMM + drho_PMP + drho_MMM - drho_MMP) * c1o2;
- d_011 = (-drho_PPM + drho_PPP - drho_MPM + drho_MPP + drho_PMM - drho_PMP + drho_MMM - drho_MMP) * c1o2;
+ ? c0o1 : -c3o1 * (a_100 * a_100 + b_010 * b_010 + c_001 * c_001) - c6o1 * (b_100 * a_010 + c_100 * a_001 + c_010 * b_001);
+ d_000 = c1o8 * ((((drho_PPP + drho_MMM) + (drho_PPM + drho_MMP)) + ((drho_PMM + drho_MPP) + (drho_PMP + drho_MPM))) - c2o1 * LaplaceRho);
+ d_100 = c1o4 * (((drho_PPP - drho_MMM) + (drho_PPM - drho_MMP)) + ((drho_PMM - drho_MPP) + (drho_PMP - drho_MPM)));
+ d_010 = c1o4 * (((drho_PPP - drho_MMM) + (drho_PPM - drho_MMP)) + ((drho_MPP - drho_PMM) + (drho_MPM - drho_PMP)));
+ d_001 = c1o4 * (((drho_PPP - drho_MMM) + (drho_MMP - drho_PPM)) + ((drho_MPP - drho_PMM) + (drho_PMP - drho_MPM)));
+ d_110 = c1o2 * (((drho_PPP + drho_MMM) + (drho_PPM + drho_MMP)) - ((drho_PMM + drho_MPP) + (drho_PMP + drho_MPM)));
+ d_101 = c1o2 * (((drho_PPP + drho_MMM) - (drho_PPM + drho_MMP)) + ((drho_PMP + drho_MPM) - (drho_PMM + drho_MPP)));
+ d_011 = c1o2 * (((drho_PPP + drho_MMM) - (drho_PPM + drho_MMP)) + ((drho_PMM + drho_MPP) - (drho_PMP + drho_MPM)));
//////////////////////////////////////////////////////////////////////////
@@ -639,7 +645,7 @@ __global__ void scaleFC_compressible(
////////////////////////////////////////////////////////////////////////////////////
// index of the destination node and its neighbors
- k_000 = indicesCoarse000[k_thread];
+ k_000 = indicesCoarse000[nodeIndex];
k_M00 = neighborXcoarse [k_000];
k_0M0 = neighborYcoarse [k_000];
k_00M = neighborZcoarse [k_000];
diff --git a/src/gpu/VirtualFluids_GPU/GPU/Init27.cu b/src/gpu/VirtualFluids_GPU/GPU/Init27.cu
index 6d497d2a1ab7ec305bec4f1ad1ed2e2d63c4dc27..23666fdcf6714d30b40b4750c52f129cc472761c 100644
--- a/src/gpu/VirtualFluids_GPU/GPU/Init27.cu
+++ b/src/gpu/VirtualFluids_GPU/GPU/Init27.cu
@@ -15,7 +15,7 @@ __global__ void LBInit27( int myid,
unsigned int* neighborY,
unsigned int* neighborZ,
real* vParabel,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
unsigned int grid_nx,
unsigned int grid_ny,
unsigned int grid_nz,
@@ -24,33 +24,33 @@ __global__ void LBInit27( int myid,
int maxlev)
{
Distributions27 D;
- D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_MPM * numberOfLBnodes];
////////////////////////////////////////////////////////////////////////////////
unsigned int k; // Zugriff auf arrays im device
//
@@ -142,32 +142,32 @@ __global__ void LBInit27( int myid,
real cu_sq=c3o2*(vx1*vx1+vx2*vx2+vx3*vx3);
(D.f[DIR_000])[kzero] = c8o27* (drho-cu_sq);
- (D.f[DIR_P00 ])[ke ] = c2o27* (drho+c3o1*( vx1 )+c9o2*( vx1 )*( vx1 )-cu_sq);
- (D.f[DIR_M00 ])[kw ] = c2o27* (drho+c3o1*(-vx1 )+c9o2*(-vx1 )*(-vx1 )-cu_sq);
- (D.f[DIR_0P0 ])[kn ] = c2o27* (drho+c3o1*( vx2 )+c9o2*( vx2 )*( vx2 )-cu_sq);
- (D.f[DIR_0M0 ])[ks ] = c2o27* (drho+c3o1*( -vx2 )+c9o2*( -vx2 )*( -vx2 )-cu_sq);
- (D.f[DIR_00P ])[kt ] = c2o27* (drho+c3o1*( vx3)+c9o2*( vx3)*( vx3)-cu_sq);
- (D.f[DIR_00M ])[kb ] = c2o27* (drho+c3o1*( -vx3)+c9o2*( -vx3)*( -vx3)-cu_sq);
- (D.f[DIR_PP0 ])[kne ] = c1o54* (drho+c3o1*( vx1+vx2 )+c9o2*( vx1+vx2 )*( vx1+vx2 )-cu_sq);
- (D.f[DIR_MM0 ])[ksw ] = c1o54* (drho+c3o1*(-vx1-vx2 )+c9o2*(-vx1-vx2 )*(-vx1-vx2 )-cu_sq);
- (D.f[DIR_PM0 ])[kse ] = c1o54* (drho+c3o1*( vx1-vx2 )+c9o2*( vx1-vx2 )*( vx1-vx2 )-cu_sq);
- (D.f[DIR_MP0 ])[knw ] = c1o54* (drho+c3o1*(-vx1+vx2 )+c9o2*(-vx1+vx2 )*(-vx1+vx2 )-cu_sq);
- (D.f[DIR_P0P ])[kte ] = c1o54* (drho+c3o1*( vx1 +vx3)+c9o2*( vx1 +vx3)*( vx1 +vx3)-cu_sq);
- (D.f[DIR_M0M ])[kbw ] = c1o54* (drho+c3o1*(-vx1 -vx3)+c9o2*(-vx1 -vx3)*(-vx1 -vx3)-cu_sq);
- (D.f[DIR_P0M ])[kbe ] = c1o54* (drho+c3o1*( vx1 -vx3)+c9o2*( vx1 -vx3)*( vx1 -vx3)-cu_sq);
- (D.f[DIR_M0P ])[ktw ] = c1o54* (drho+c3o1*(-vx1 +vx3)+c9o2*(-vx1 +vx3)*(-vx1 +vx3)-cu_sq);
- (D.f[DIR_0PP ])[ktn ] = c1o54* (drho+c3o1*( vx2+vx3)+c9o2*( vx2+vx3)*( vx2+vx3)-cu_sq);
- (D.f[DIR_0MM ])[kbs ] = c1o54* (drho+c3o1*( -vx2-vx3)+c9o2*( -vx2-vx3)*( -vx2-vx3)-cu_sq);
- (D.f[DIR_0PM ])[kbn ] = c1o54* (drho+c3o1*( vx2-vx3)+c9o2*( vx2-vx3)*( vx2-vx3)-cu_sq);
- (D.f[DIR_0MP ])[kts ] = c1o54* (drho+c3o1*( -vx2+vx3)+c9o2*( -vx2+vx3)*( -vx2+vx3)-cu_sq);
- (D.f[DIR_PPP ])[ktne ] = c1o216*(drho+c3o1*( vx1+vx2+vx3)+c9o2*( vx1+vx2+vx3)*( vx1+vx2+vx3)-cu_sq);
- (D.f[DIR_MMM ])[kbsw ] = c1o216*(drho+c3o1*(-vx1-vx2-vx3)+c9o2*(-vx1-vx2-vx3)*(-vx1-vx2-vx3)-cu_sq);
- (D.f[DIR_PPM ])[kbne ] = c1o216*(drho+c3o1*( vx1+vx2-vx3)+c9o2*( vx1+vx2-vx3)*( vx1+vx2-vx3)-cu_sq);
- (D.f[DIR_MMP ])[ktsw ] = c1o216*(drho+c3o1*(-vx1-vx2+vx3)+c9o2*(-vx1-vx2+vx3)*(-vx1-vx2+vx3)-cu_sq);
- (D.f[DIR_PMP ])[ktse ] = c1o216*(drho+c3o1*( vx1-vx2+vx3)+c9o2*( vx1-vx2+vx3)*( vx1-vx2+vx3)-cu_sq);
- (D.f[DIR_MPM ])[kbnw ] = c1o216*(drho+c3o1*(-vx1+vx2-vx3)+c9o2*(-vx1+vx2-vx3)*(-vx1+vx2-vx3)-cu_sq);
- (D.f[DIR_PMM ])[kbse ] = c1o216*(drho+c3o1*( vx1-vx2-vx3)+c9o2*( vx1-vx2-vx3)*( vx1-vx2-vx3)-cu_sq);
- (D.f[DIR_MPP ])[ktnw ] = c1o216*(drho+c3o1*(-vx1+vx2+vx3)+c9o2*(-vx1+vx2+vx3)*(-vx1+vx2+vx3)-cu_sq);
+ (D.f[DIR_P00])[ke ] = c2o27* (drho+c3o1*( vx1 )+c9o2*( vx1 )*( vx1 )-cu_sq);
+ (D.f[DIR_M00])[kw ] = c2o27* (drho+c3o1*(-vx1 )+c9o2*(-vx1 )*(-vx1 )-cu_sq);
+ (D.f[DIR_0P0])[kn ] = c2o27* (drho+c3o1*( vx2 )+c9o2*( vx2 )*( vx2 )-cu_sq);
+ (D.f[DIR_0M0])[ks ] = c2o27* (drho+c3o1*( -vx2 )+c9o2*( -vx2 )*( -vx2 )-cu_sq);
+ (D.f[DIR_00P])[kt ] = c2o27* (drho+c3o1*( vx3)+c9o2*( vx3)*( vx3)-cu_sq);
+ (D.f[DIR_00M])[kb ] = c2o27* (drho+c3o1*( -vx3)+c9o2*( -vx3)*( -vx3)-cu_sq);
+ (D.f[DIR_PP0])[kne ] = c1o54* (drho+c3o1*( vx1+vx2 )+c9o2*( vx1+vx2 )*( vx1+vx2 )-cu_sq);
+ (D.f[DIR_MM0])[ksw ] = c1o54* (drho+c3o1*(-vx1-vx2 )+c9o2*(-vx1-vx2 )*(-vx1-vx2 )-cu_sq);
+ (D.f[DIR_PM0])[kse ] = c1o54* (drho+c3o1*( vx1-vx2 )+c9o2*( vx1-vx2 )*( vx1-vx2 )-cu_sq);
+ (D.f[DIR_MP0])[knw ] = c1o54* (drho+c3o1*(-vx1+vx2 )+c9o2*(-vx1+vx2 )*(-vx1+vx2 )-cu_sq);
+ (D.f[DIR_P0P])[kte ] = c1o54* (drho+c3o1*( vx1 +vx3)+c9o2*( vx1 +vx3)*( vx1 +vx3)-cu_sq);
+ (D.f[DIR_M0M])[kbw ] = c1o54* (drho+c3o1*(-vx1 -vx3)+c9o2*(-vx1 -vx3)*(-vx1 -vx3)-cu_sq);
+ (D.f[DIR_P0M])[kbe ] = c1o54* (drho+c3o1*( vx1 -vx3)+c9o2*( vx1 -vx3)*( vx1 -vx3)-cu_sq);
+ (D.f[DIR_M0P])[ktw ] = c1o54* (drho+c3o1*(-vx1 +vx3)+c9o2*(-vx1 +vx3)*(-vx1 +vx3)-cu_sq);
+ (D.f[DIR_0PP])[ktn ] = c1o54* (drho+c3o1*( vx2+vx3)+c9o2*( vx2+vx3)*( vx2+vx3)-cu_sq);
+ (D.f[DIR_0MM])[kbs ] = c1o54* (drho+c3o1*( -vx2-vx3)+c9o2*( -vx2-vx3)*( -vx2-vx3)-cu_sq);
+ (D.f[DIR_0PM])[kbn ] = c1o54* (drho+c3o1*( vx2-vx3)+c9o2*( vx2-vx3)*( vx2-vx3)-cu_sq);
+ (D.f[DIR_0MP])[kts ] = c1o54* (drho+c3o1*( -vx2+vx3)+c9o2*( -vx2+vx3)*( -vx2+vx3)-cu_sq);
+ (D.f[DIR_PPP])[ktne ] = c1o216*(drho+c3o1*( vx1+vx2+vx3)+c9o2*( vx1+vx2+vx3)*( vx1+vx2+vx3)-cu_sq);
+ (D.f[DIR_MMM])[kbsw ] = c1o216*(drho+c3o1*(-vx1-vx2-vx3)+c9o2*(-vx1-vx2-vx3)*(-vx1-vx2-vx3)-cu_sq);
+ (D.f[DIR_PPM])[kbne ] = c1o216*(drho+c3o1*( vx1+vx2-vx3)+c9o2*( vx1+vx2-vx3)*( vx1+vx2-vx3)-cu_sq);
+ (D.f[DIR_MMP])[ktsw ] = c1o216*(drho+c3o1*(-vx1-vx2+vx3)+c9o2*(-vx1-vx2+vx3)*(-vx1-vx2+vx3)-cu_sq);
+ (D.f[DIR_PMP])[ktse ] = c1o216*(drho+c3o1*( vx1-vx2+vx3)+c9o2*( vx1-vx2+vx3)*( vx1-vx2+vx3)-cu_sq);
+ (D.f[DIR_MPM])[kbnw ] = c1o216*(drho+c3o1*(-vx1+vx2-vx3)+c9o2*(-vx1+vx2-vx3)*(-vx1+vx2-vx3)-cu_sq);
+ (D.f[DIR_PMM])[kbse ] = c1o216*(drho+c3o1*( vx1-vx2-vx3)+c9o2*( vx1-vx2-vx3)*( vx1-vx2-vx3)-cu_sq);
+ (D.f[DIR_MPP])[ktnw ] = c1o216*(drho+c3o1*(-vx1+vx2+vx3)+c9o2*(-vx1+vx2+vx3)*(-vx1+vx2+vx3)-cu_sq);
}
////////////////////////////////////////////////////////////////////////////////
@@ -191,7 +191,7 @@ __global__ void LBInitNonEqPartSP27( unsigned int* neighborX,
real* ux,
real* uy,
real* uz,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
real* DD,
real omega,
bool EvenOrOdd)
@@ -207,7 +207,7 @@ __global__ void LBInitNonEqPartSP27( unsigned int* neighborX,
const unsigned k = nx*(ny*z + y) + x;
//////////////////////////////////////////////////////////////////////////
- if(k<size_Mat)
+ if(k<numberOfLBnodes)
{
////////////////////////////////////////////////////////////////////////////////
unsigned int BC;
@@ -218,63 +218,63 @@ __global__ void LBInitNonEqPartSP27( unsigned int* neighborX,
Distributions27 D;
if (EvenOrOdd==true)
{
- D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
+ D.f[DIR_M00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_MPM * numberOfLBnodes];
}
//////////////////////////////////////////////////////////////////////////
real drho = rho[k];//0.0f;//
@@ -397,62 +397,62 @@ __global__ void LBInitNonEqPartSP27( unsigned int* neighborX,
real cu_sq=c3o2*(vx1*vx1+vx2*vx2+vx3*vx3);
(D.f[DIR_000])[kzero] = c8o27* (drho-cu_sq);
- (D.f[DIR_P00 ])[ke ] = c2o27* (drho+c3o1*( vx1 )+c9o2*( vx1 )*( vx1 )-cu_sq);
- (D.f[DIR_M00 ])[kw ] = c2o27* (drho+c3o1*(-vx1 )+c9o2*(-vx1 )*(-vx1 )-cu_sq);
- (D.f[DIR_0P0 ])[kn ] = c2o27* (drho+c3o1*( vx2 )+c9o2*( vx2 )*( vx2 )-cu_sq);
- (D.f[DIR_0M0 ])[ks ] = c2o27* (drho+c3o1*( -vx2 )+c9o2*( -vx2 )*( -vx2 )-cu_sq);
- (D.f[DIR_00P ])[kt ] = c2o27* (drho+c3o1*( vx3)+c9o2*( vx3)*( vx3)-cu_sq);
- (D.f[DIR_00M ])[kb ] = c2o27* (drho+c3o1*( -vx3)+c9o2*( -vx3)*( -vx3)-cu_sq);
- (D.f[DIR_PP0 ])[kne ] = c1o54* (drho+c3o1*( vx1+vx2 )+c9o2*( vx1+vx2 )*( vx1+vx2 )-cu_sq);
- (D.f[DIR_MM0 ])[ksw ] = c1o54* (drho+c3o1*(-vx1-vx2 )+c9o2*(-vx1-vx2 )*(-vx1-vx2 )-cu_sq);
- (D.f[DIR_PM0 ])[kse ] = c1o54* (drho+c3o1*( vx1-vx2 )+c9o2*( vx1-vx2 )*( vx1-vx2 )-cu_sq);
- (D.f[DIR_MP0 ])[knw ] = c1o54* (drho+c3o1*(-vx1+vx2 )+c9o2*(-vx1+vx2 )*(-vx1+vx2 )-cu_sq);
- (D.f[DIR_P0P ])[kte ] = c1o54* (drho+c3o1*( vx1 +vx3)+c9o2*( vx1 +vx3)*( vx1 +vx3)-cu_sq);
- (D.f[DIR_M0M ])[kbw ] = c1o54* (drho+c3o1*(-vx1 -vx3)+c9o2*(-vx1 -vx3)*(-vx1 -vx3)-cu_sq);
- (D.f[DIR_P0M ])[kbe ] = c1o54* (drho+c3o1*( vx1 -vx3)+c9o2*( vx1 -vx3)*( vx1 -vx3)-cu_sq);
- (D.f[DIR_M0P ])[ktw ] = c1o54* (drho+c3o1*(-vx1 +vx3)+c9o2*(-vx1 +vx3)*(-vx1 +vx3)-cu_sq);
- (D.f[DIR_0PP ])[ktn ] = c1o54* (drho+c3o1*( vx2+vx3)+c9o2*( vx2+vx3)*( vx2+vx3)-cu_sq);
- (D.f[DIR_0MM ])[kbs ] = c1o54* (drho+c3o1*( -vx2-vx3)+c9o2*( -vx2-vx3)*( -vx2-vx3)-cu_sq);
- (D.f[DIR_0PM ])[kbn ] = c1o54* (drho+c3o1*( vx2-vx3)+c9o2*( vx2-vx3)*( vx2-vx3)-cu_sq);
- (D.f[DIR_0MP ])[kts ] = c1o54* (drho+c3o1*( -vx2+vx3)+c9o2*( -vx2+vx3)*( -vx2+vx3)-cu_sq);
- (D.f[DIR_PPP ])[ktne ] = c1o216*(drho+c3o1*( vx1+vx2+vx3)+c9o2*( vx1+vx2+vx3)*( vx1+vx2+vx3)-cu_sq);
- (D.f[DIR_MMM ])[kbsw ] = c1o216*(drho+c3o1*(-vx1-vx2-vx3)+c9o2*(-vx1-vx2-vx3)*(-vx1-vx2-vx3)-cu_sq);
- (D.f[DIR_PPM ])[kbne ] = c1o216*(drho+c3o1*( vx1+vx2-vx3)+c9o2*( vx1+vx2-vx3)*( vx1+vx2-vx3)-cu_sq);
- (D.f[DIR_MMP ])[ktsw ] = c1o216*(drho+c3o1*(-vx1-vx2+vx3)+c9o2*(-vx1-vx2+vx3)*(-vx1-vx2+vx3)-cu_sq);
- (D.f[DIR_PMP ])[ktse ] = c1o216*(drho+c3o1*( vx1-vx2+vx3)+c9o2*( vx1-vx2+vx3)*( vx1-vx2+vx3)-cu_sq);
- (D.f[DIR_MPM ])[kbnw ] = c1o216*(drho+c3o1*(-vx1+vx2-vx3)+c9o2*(-vx1+vx2-vx3)*(-vx1+vx2-vx3)-cu_sq);
- (D.f[DIR_PMM ])[kbse ] = c1o216*(drho+c3o1*( vx1-vx2-vx3)+c9o2*( vx1-vx2-vx3)*( vx1-vx2-vx3)-cu_sq);
- (D.f[DIR_MPP ])[ktnw ] = c1o216*(drho+c3o1*(-vx1+vx2+vx3)+c9o2*(-vx1+vx2+vx3)*(-vx1+vx2+vx3)-cu_sq);
+ (D.f[DIR_P00])[ke ] = c2o27* (drho+c3o1*( vx1 )+c9o2*( vx1 )*( vx1 )-cu_sq);
+ (D.f[DIR_M00])[kw ] = c2o27* (drho+c3o1*(-vx1 )+c9o2*(-vx1 )*(-vx1 )-cu_sq);
+ (D.f[DIR_0P0])[kn ] = c2o27* (drho+c3o1*( vx2 )+c9o2*( vx2 )*( vx2 )-cu_sq);
+ (D.f[DIR_0M0])[ks ] = c2o27* (drho+c3o1*( -vx2 )+c9o2*( -vx2 )*( -vx2 )-cu_sq);
+ (D.f[DIR_00P])[kt ] = c2o27* (drho+c3o1*( vx3)+c9o2*( vx3)*( vx3)-cu_sq);
+ (D.f[DIR_00M])[kb ] = c2o27* (drho+c3o1*( -vx3)+c9o2*( -vx3)*( -vx3)-cu_sq);
+ (D.f[DIR_PP0])[kne ] = c1o54* (drho+c3o1*( vx1+vx2 )+c9o2*( vx1+vx2 )*( vx1+vx2 )-cu_sq);
+ (D.f[DIR_MM0])[ksw ] = c1o54* (drho+c3o1*(-vx1-vx2 )+c9o2*(-vx1-vx2 )*(-vx1-vx2 )-cu_sq);
+ (D.f[DIR_PM0])[kse ] = c1o54* (drho+c3o1*( vx1-vx2 )+c9o2*( vx1-vx2 )*( vx1-vx2 )-cu_sq);
+ (D.f[DIR_MP0])[knw ] = c1o54* (drho+c3o1*(-vx1+vx2 )+c9o2*(-vx1+vx2 )*(-vx1+vx2 )-cu_sq);
+ (D.f[DIR_P0P])[kte ] = c1o54* (drho+c3o1*( vx1 +vx3)+c9o2*( vx1 +vx3)*( vx1 +vx3)-cu_sq);
+ (D.f[DIR_M0M])[kbw ] = c1o54* (drho+c3o1*(-vx1 -vx3)+c9o2*(-vx1 -vx3)*(-vx1 -vx3)-cu_sq);
+ (D.f[DIR_P0M])[kbe ] = c1o54* (drho+c3o1*( vx1 -vx3)+c9o2*( vx1 -vx3)*( vx1 -vx3)-cu_sq);
+ (D.f[DIR_M0P])[ktw ] = c1o54* (drho+c3o1*(-vx1 +vx3)+c9o2*(-vx1 +vx3)*(-vx1 +vx3)-cu_sq);
+ (D.f[DIR_0PP])[ktn ] = c1o54* (drho+c3o1*( vx2+vx3)+c9o2*( vx2+vx3)*( vx2+vx3)-cu_sq);
+ (D.f[DIR_0MM])[kbs ] = c1o54* (drho+c3o1*( -vx2-vx3)+c9o2*( -vx2-vx3)*( -vx2-vx3)-cu_sq);
+ (D.f[DIR_0PM])[kbn ] = c1o54* (drho+c3o1*( vx2-vx3)+c9o2*( vx2-vx3)*( vx2-vx3)-cu_sq);
+ (D.f[DIR_0MP])[kts ] = c1o54* (drho+c3o1*( -vx2+vx3)+c9o2*( -vx2+vx3)*( -vx2+vx3)-cu_sq);
+ (D.f[DIR_PPP])[ktne ] = c1o216*(drho+c3o1*( vx1+vx2+vx3)+c9o2*( vx1+vx2+vx3)*( vx1+vx2+vx3)-cu_sq);
+ (D.f[DIR_MMM])[kbsw ] = c1o216*(drho+c3o1*(-vx1-vx2-vx3)+c9o2*(-vx1-vx2-vx3)*(-vx1-vx2-vx3)-cu_sq);
+ (D.f[DIR_PPM])[kbne ] = c1o216*(drho+c3o1*( vx1+vx2-vx3)+c9o2*( vx1+vx2-vx3)*( vx1+vx2-vx3)-cu_sq);
+ (D.f[DIR_MMP])[ktsw ] = c1o216*(drho+c3o1*(-vx1-vx2+vx3)+c9o2*(-vx1-vx2+vx3)*(-vx1-vx2+vx3)-cu_sq);
+ (D.f[DIR_PMP])[ktse ] = c1o216*(drho+c3o1*( vx1-vx2+vx3)+c9o2*( vx1-vx2+vx3)*( vx1-vx2+vx3)-cu_sq);
+ (D.f[DIR_MPM])[kbnw ] = c1o216*(drho+c3o1*(-vx1+vx2-vx3)+c9o2*(-vx1+vx2-vx3)*(-vx1+vx2-vx3)-cu_sq);
+ (D.f[DIR_PMM])[kbse ] = c1o216*(drho+c3o1*( vx1-vx2-vx3)+c9o2*( vx1-vx2-vx3)*( vx1-vx2-vx3)-cu_sq);
+ (D.f[DIR_MPP])[ktnw ] = c1o216*(drho+c3o1*(-vx1+vx2+vx3)+c9o2*(-vx1+vx2+vx3)*(-vx1+vx2+vx3)-cu_sq);
//////////////////////////////////////////////////////////////////////////
(D.f[DIR_000])[kzero] += f_ZERO;
- (D.f[DIR_P00 ])[ke ] += f_E ;
- (D.f[DIR_M00 ])[kw ] += f_E ;
- (D.f[DIR_0P0 ])[kn ] += f_N ;
- (D.f[DIR_0M0 ])[ks ] += f_N ;
- (D.f[DIR_00P ])[kt ] += f_T ;
- (D.f[DIR_00M ])[kb ] += f_T ;
- (D.f[DIR_PP0 ])[kne ] += f_NE ;
- (D.f[DIR_MM0 ])[ksw ] += f_NE ;
- (D.f[DIR_PM0 ])[kse ] += f_SE ;
- (D.f[DIR_MP0 ])[knw ] += f_SE ;
- (D.f[DIR_P0P ])[kte ] += f_TE ;
- (D.f[DIR_M0M ])[kbw ] += f_TE ;
- (D.f[DIR_P0M ])[kbe ] += f_BE ;
- (D.f[DIR_M0P ])[ktw ] += f_BE ;
- (D.f[DIR_0PP ])[ktn ] += f_TN ;
- (D.f[DIR_0MM ])[kbs ] += f_TN ;
- (D.f[DIR_0PM ])[kbn ] += f_BN ;
- (D.f[DIR_0MP ])[kts ] += f_BN ;
- (D.f[DIR_PPP ])[ktne ] += f_TNE ;
- (D.f[DIR_MMM ])[kbsw ] += f_TNE ;
- (D.f[DIR_PPM ])[kbne ] += f_TSW ;
- (D.f[DIR_MMP ])[ktsw ] += f_TSW ;
- (D.f[DIR_PMP ])[ktse ] += f_TSE ;
- (D.f[DIR_MPM ])[kbnw ] += f_TSE ;
- (D.f[DIR_PMM ])[kbse ] += f_TNW ;
- (D.f[DIR_MPP ])[ktnw ] += f_TNW ;
+ (D.f[DIR_P00])[ke ] += f_E ;
+ (D.f[DIR_M00])[kw ] += f_E ;
+ (D.f[DIR_0P0])[kn ] += f_N ;
+ (D.f[DIR_0M0])[ks ] += f_N ;
+ (D.f[DIR_00P])[kt ] += f_T ;
+ (D.f[DIR_00M])[kb ] += f_T ;
+ (D.f[DIR_PP0])[kne ] += f_NE ;
+ (D.f[DIR_MM0])[ksw ] += f_NE ;
+ (D.f[DIR_PM0])[kse ] += f_SE ;
+ (D.f[DIR_MP0])[knw ] += f_SE ;
+ (D.f[DIR_P0P])[kte ] += f_TE ;
+ (D.f[DIR_M0M])[kbw ] += f_TE ;
+ (D.f[DIR_P0M])[kbe ] += f_BE ;
+ (D.f[DIR_M0P])[ktw ] += f_BE ;
+ (D.f[DIR_0PP])[ktn ] += f_TN ;
+ (D.f[DIR_0MM])[kbs ] += f_TN ;
+ (D.f[DIR_0PM])[kbn ] += f_BN ;
+ (D.f[DIR_0MP])[kts ] += f_BN ;
+ (D.f[DIR_PPP])[ktne ] += f_TNE ;
+ (D.f[DIR_MMM])[kbsw ] += f_TNE ;
+ (D.f[DIR_PPM])[kbne ] += f_TSW ;
+ (D.f[DIR_MMP])[ktsw ] += f_TSW ;
+ (D.f[DIR_PMP])[ktse ] += f_TSE ;
+ (D.f[DIR_MPM])[kbnw ] += f_TSE ;
+ (D.f[DIR_PMM])[kbse ] += f_TNW ;
+ (D.f[DIR_MPP])[ktnw ] += f_TNW ;
//////////////////////////////////////////////////////////////////////////
}
@@ -460,7 +460,7 @@ __global__ void LBInitNonEqPartSP27( unsigned int* neighborX,
{
//////////////////////////////////////////////////////////////////////////
Distributions27 D;
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
//////////////////////////////////////////////////////////////////////////
(D.f[DIR_000])[k] = c96o1;
//////////////////////////////////////////////////////////////////////////
diff --git a/src/gpu/VirtualFluids_GPU/GPU/InitAdvectionDiffusion27.cu b/src/gpu/VirtualFluids_GPU/GPU/InitAdvectionDiffusion27.cu
index c091aa8b9a29017ddc0f6ea6584e805d7afc4859..7f67d1692f7e136a6537be6780fe8625adc33e22 100644
--- a/src/gpu/VirtualFluids_GPU/GPU/InitAdvectionDiffusion27.cu
+++ b/src/gpu/VirtualFluids_GPU/GPU/InitAdvectionDiffusion27.cu
@@ -47,7 +47,7 @@ __global__ void InitAD27(
real* velocityX,
real* velocityY,
real* velocityZ,
- uint size_Mat,
+ unsigned long long numberOfLBnodes,
real* distributionsAD,
bool isEvenTimestep)
{
@@ -68,7 +68,7 @@ __global__ void InitAD27(
//////////////////////////////////////////////////////////////////////////
// run for all indices in size_Mat and fluid nodes
- if ((k < size_Mat) && (typeOfGridNode[k] == GEO_FLUID))
+ if ((k < numberOfLBnodes) && (typeOfGridNode[k] == GEO_FLUID))
{
//////////////////////////////////////////////////////////////////////////
//! - Read distributions: style of reading and writing the distributions from/to stored arrays dependent on timestep is based on the esoteric twist algorithm \ref
@@ -77,63 +77,63 @@ __global__ void InitAD27(
Distributions27 distAD;
if (isEvenTimestep)
{
- distAD.f[DIR_P00 ] = &distributionsAD[DIR_P00 *size_Mat];
- distAD.f[DIR_M00 ] = &distributionsAD[DIR_M00 *size_Mat];
- distAD.f[DIR_0P0 ] = &distributionsAD[DIR_0P0 *size_Mat];
- distAD.f[DIR_0M0 ] = &distributionsAD[DIR_0M0 *size_Mat];
- distAD.f[DIR_00P ] = &distributionsAD[DIR_00P *size_Mat];
- distAD.f[DIR_00M ] = &distributionsAD[DIR_00M *size_Mat];
- distAD.f[DIR_PP0 ] = &distributionsAD[DIR_PP0 *size_Mat];
- distAD.f[DIR_MM0 ] = &distributionsAD[DIR_MM0 *size_Mat];
- distAD.f[DIR_PM0 ] = &distributionsAD[DIR_PM0 *size_Mat];
- distAD.f[DIR_MP0 ] = &distributionsAD[DIR_MP0 *size_Mat];
- distAD.f[DIR_P0P ] = &distributionsAD[DIR_P0P *size_Mat];
- distAD.f[DIR_M0M ] = &distributionsAD[DIR_M0M *size_Mat];
- distAD.f[DIR_P0M ] = &distributionsAD[DIR_P0M *size_Mat];
- distAD.f[DIR_M0P ] = &distributionsAD[DIR_M0P *size_Mat];
- distAD.f[DIR_0PP ] = &distributionsAD[DIR_0PP *size_Mat];
- distAD.f[DIR_0MM ] = &distributionsAD[DIR_0MM *size_Mat];
- distAD.f[DIR_0PM ] = &distributionsAD[DIR_0PM *size_Mat];
- distAD.f[DIR_0MP ] = &distributionsAD[DIR_0MP *size_Mat];
- distAD.f[DIR_000] = &distributionsAD[DIR_000*size_Mat];
- distAD.f[DIR_PPP ] = &distributionsAD[DIR_PPP *size_Mat];
- distAD.f[DIR_MMP ] = &distributionsAD[DIR_MMP *size_Mat];
- distAD.f[DIR_PMP ] = &distributionsAD[DIR_PMP *size_Mat];
- distAD.f[DIR_MPP ] = &distributionsAD[DIR_MPP *size_Mat];
- distAD.f[DIR_PPM ] = &distributionsAD[DIR_PPM *size_Mat];
- distAD.f[DIR_MMM ] = &distributionsAD[DIR_MMM *size_Mat];
- distAD.f[DIR_PMM ] = &distributionsAD[DIR_PMM *size_Mat];
- distAD.f[DIR_MPM ] = &distributionsAD[DIR_MPM *size_Mat];
+ distAD.f[DIR_P00] = &distributionsAD[DIR_P00 * numberOfLBnodes];
+ distAD.f[DIR_M00] = &distributionsAD[DIR_M00 * numberOfLBnodes];
+ distAD.f[DIR_0P0] = &distributionsAD[DIR_0P0 * numberOfLBnodes];
+ distAD.f[DIR_0M0] = &distributionsAD[DIR_0M0 * numberOfLBnodes];
+ distAD.f[DIR_00P] = &distributionsAD[DIR_00P * numberOfLBnodes];
+ distAD.f[DIR_00M] = &distributionsAD[DIR_00M * numberOfLBnodes];
+ distAD.f[DIR_PP0] = &distributionsAD[DIR_PP0 * numberOfLBnodes];
+ distAD.f[DIR_MM0] = &distributionsAD[DIR_MM0 * numberOfLBnodes];
+ distAD.f[DIR_PM0] = &distributionsAD[DIR_PM0 * numberOfLBnodes];
+ distAD.f[DIR_MP0] = &distributionsAD[DIR_MP0 * numberOfLBnodes];
+ distAD.f[DIR_P0P] = &distributionsAD[DIR_P0P * numberOfLBnodes];
+ distAD.f[DIR_M0M] = &distributionsAD[DIR_M0M * numberOfLBnodes];
+ distAD.f[DIR_P0M] = &distributionsAD[DIR_P0M * numberOfLBnodes];
+ distAD.f[DIR_M0P] = &distributionsAD[DIR_M0P * numberOfLBnodes];
+ distAD.f[DIR_0PP] = &distributionsAD[DIR_0PP * numberOfLBnodes];
+ distAD.f[DIR_0MM] = &distributionsAD[DIR_0MM * numberOfLBnodes];
+ distAD.f[DIR_0PM] = &distributionsAD[DIR_0PM * numberOfLBnodes];
+ distAD.f[DIR_0MP] = &distributionsAD[DIR_0MP * numberOfLBnodes];
+ distAD.f[DIR_000] = &distributionsAD[DIR_000 * numberOfLBnodes];
+ distAD.f[DIR_PPP] = &distributionsAD[DIR_PPP * numberOfLBnodes];
+ distAD.f[DIR_MMP] = &distributionsAD[DIR_MMP * numberOfLBnodes];
+ distAD.f[DIR_PMP] = &distributionsAD[DIR_PMP * numberOfLBnodes];
+ distAD.f[DIR_MPP] = &distributionsAD[DIR_MPP * numberOfLBnodes];
+ distAD.f[DIR_PPM] = &distributionsAD[DIR_PPM * numberOfLBnodes];
+ distAD.f[DIR_MMM] = &distributionsAD[DIR_MMM * numberOfLBnodes];
+ distAD.f[DIR_PMM] = &distributionsAD[DIR_PMM * numberOfLBnodes];
+ distAD.f[DIR_MPM] = &distributionsAD[DIR_MPM * numberOfLBnodes];
}
else
{
- distAD.f[DIR_M00 ] = &distributionsAD[DIR_P00 *size_Mat];
- distAD.f[DIR_P00 ] = &distributionsAD[DIR_M00 *size_Mat];
- distAD.f[DIR_0M0 ] = &distributionsAD[DIR_0P0 *size_Mat];
- distAD.f[DIR_0P0 ] = &distributionsAD[DIR_0M0 *size_Mat];
- distAD.f[DIR_00M ] = &distributionsAD[DIR_00P *size_Mat];
- distAD.f[DIR_00P ] = &distributionsAD[DIR_00M *size_Mat];
- distAD.f[DIR_MM0 ] = &distributionsAD[DIR_PP0 *size_Mat];
- distAD.f[DIR_PP0 ] = &distributionsAD[DIR_MM0 *size_Mat];
- distAD.f[DIR_MP0 ] = &distributionsAD[DIR_PM0 *size_Mat];
- distAD.f[DIR_PM0 ] = &distributionsAD[DIR_MP0 *size_Mat];
- distAD.f[DIR_M0M ] = &distributionsAD[DIR_P0P *size_Mat];
- distAD.f[DIR_P0P ] = &distributionsAD[DIR_M0M *size_Mat];
- distAD.f[DIR_M0P ] = &distributionsAD[DIR_P0M *size_Mat];
- distAD.f[DIR_P0M ] = &distributionsAD[DIR_M0P *size_Mat];
- distAD.f[DIR_0MM ] = &distributionsAD[DIR_0PP *size_Mat];
- distAD.f[DIR_0PP ] = &distributionsAD[DIR_0MM *size_Mat];
- distAD.f[DIR_0MP ] = &distributionsAD[DIR_0PM *size_Mat];
- distAD.f[DIR_0PM ] = &distributionsAD[DIR_0MP *size_Mat];
- distAD.f[DIR_000] = &distributionsAD[DIR_000*size_Mat];
- distAD.f[DIR_MMM ] = &distributionsAD[DIR_PPP *size_Mat];
- distAD.f[DIR_PPM ] = &distributionsAD[DIR_MMP *size_Mat];
- distAD.f[DIR_MPM ] = &distributionsAD[DIR_PMP *size_Mat];
- distAD.f[DIR_PMM ] = &distributionsAD[DIR_MPP *size_Mat];
- distAD.f[DIR_MMP ] = &distributionsAD[DIR_PPM *size_Mat];
- distAD.f[DIR_PPP ] = &distributionsAD[DIR_MMM *size_Mat];
- distAD.f[DIR_MPP ] = &distributionsAD[DIR_PMM *size_Mat];
- distAD.f[DIR_PMP ] = &distributionsAD[DIR_MPM *size_Mat];
+ distAD.f[DIR_M00] = &distributionsAD[DIR_P00 * numberOfLBnodes];
+ distAD.f[DIR_P00] = &distributionsAD[DIR_M00 * numberOfLBnodes];
+ distAD.f[DIR_0M0] = &distributionsAD[DIR_0P0 * numberOfLBnodes];
+ distAD.f[DIR_0P0] = &distributionsAD[DIR_0M0 * numberOfLBnodes];
+ distAD.f[DIR_00M] = &distributionsAD[DIR_00P * numberOfLBnodes];
+ distAD.f[DIR_00P] = &distributionsAD[DIR_00M * numberOfLBnodes];
+ distAD.f[DIR_MM0] = &distributionsAD[DIR_PP0 * numberOfLBnodes];
+ distAD.f[DIR_PP0] = &distributionsAD[DIR_MM0 * numberOfLBnodes];
+ distAD.f[DIR_MP0] = &distributionsAD[DIR_PM0 * numberOfLBnodes];
+ distAD.f[DIR_PM0] = &distributionsAD[DIR_MP0 * numberOfLBnodes];
+ distAD.f[DIR_M0M] = &distributionsAD[DIR_P0P * numberOfLBnodes];
+ distAD.f[DIR_P0P] = &distributionsAD[DIR_M0M * numberOfLBnodes];
+ distAD.f[DIR_M0P] = &distributionsAD[DIR_P0M * numberOfLBnodes];
+ distAD.f[DIR_P0M] = &distributionsAD[DIR_M0P * numberOfLBnodes];
+ distAD.f[DIR_0MM] = &distributionsAD[DIR_0PP * numberOfLBnodes];
+ distAD.f[DIR_0PP] = &distributionsAD[DIR_0MM * numberOfLBnodes];
+ distAD.f[DIR_0MP] = &distributionsAD[DIR_0PM * numberOfLBnodes];
+ distAD.f[DIR_0PM] = &distributionsAD[DIR_0MP * numberOfLBnodes];
+ distAD.f[DIR_000] = &distributionsAD[DIR_000 * numberOfLBnodes];
+ distAD.f[DIR_MMM] = &distributionsAD[DIR_PPP * numberOfLBnodes];
+ distAD.f[DIR_PPM] = &distributionsAD[DIR_MMP * numberOfLBnodes];
+ distAD.f[DIR_MPM] = &distributionsAD[DIR_PMP * numberOfLBnodes];
+ distAD.f[DIR_PMM] = &distributionsAD[DIR_MPP * numberOfLBnodes];
+ distAD.f[DIR_MMP] = &distributionsAD[DIR_PPM * numberOfLBnodes];
+ distAD.f[DIR_PPP] = &distributionsAD[DIR_MMM * numberOfLBnodes];
+ distAD.f[DIR_MPP] = &distributionsAD[DIR_PMM * numberOfLBnodes];
+ distAD.f[DIR_PMP] = &distributionsAD[DIR_MPM * numberOfLBnodes];
}
//////////////////////////////////////////////////////////////////////////
//! - Set local velocities and concetration
@@ -178,32 +178,32 @@ __global__ void InitAD27(
real cu_sq = c3o2*(vx1*vx1 + vx2*vx2 + vx3*vx3);
(distAD.f[DIR_000])[kzero] = c8o27 * conc * (c1o1 - cu_sq);
- (distAD.f[DIR_P00 ])[ke ] = c2o27 * conc * (c1o1 + c3o1 * ( vx1 ) + c9o2 * ( vx1 ) * ( vx1 ) - cu_sq);
- (distAD.f[DIR_M00 ])[kw ] = c2o27 * conc * (c1o1 + c3o1 * (-vx1 ) + c9o2 * (-vx1 ) * (-vx1 ) - cu_sq);
- (distAD.f[DIR_0P0 ])[kn ] = c2o27 * conc * (c1o1 + c3o1 * ( vx2 ) + c9o2 * ( vx2 ) * ( vx2 ) - cu_sq);
- (distAD.f[DIR_0M0 ])[ks ] = c2o27 * conc * (c1o1 + c3o1 * ( - vx2 ) + c9o2 * ( - vx2 ) * ( - vx2 ) - cu_sq);
- (distAD.f[DIR_00P ])[kt ] = c2o27 * conc * (c1o1 + c3o1 * ( vx3) + c9o2 * ( vx3) * ( vx3) - cu_sq);
- (distAD.f[DIR_00M ])[kb ] = c2o27 * conc * (c1o1 + c3o1 * ( - vx3) + c9o2 * ( - vx3) * ( - vx3) - cu_sq);
- (distAD.f[DIR_PP0 ])[kne ] = c1o54 * conc * (c1o1 + c3o1 * ( vx1 + vx2 ) + c9o2 * ( vx1 + vx2 ) * ( vx1 + vx2 ) - cu_sq);
- (distAD.f[DIR_MM0 ])[ksw ] = c1o54 * conc * (c1o1 + c3o1 * (-vx1 - vx2 ) + c9o2 * (-vx1 - vx2 ) * (-vx1 - vx2 ) - cu_sq);
- (distAD.f[DIR_PM0 ])[kse ] = c1o54 * conc * (c1o1 + c3o1 * ( vx1 - vx2 ) + c9o2 * ( vx1 - vx2 ) * ( vx1 - vx2 ) - cu_sq);
- (distAD.f[DIR_MP0 ])[knw ] = c1o54 * conc * (c1o1 + c3o1 * (-vx1 + vx2 ) + c9o2 * (-vx1 + vx2 ) * (-vx1 + vx2 ) - cu_sq);
- (distAD.f[DIR_P0P ])[kte ] = c1o54 * conc * (c1o1 + c3o1 * ( vx1 + vx3) + c9o2 * ( vx1 + vx3) * ( vx1 + vx3) - cu_sq);
- (distAD.f[DIR_M0M ])[kbw ] = c1o54 * conc * (c1o1 + c3o1 * (-vx1 - vx3) + c9o2 * (-vx1 - vx3) * (-vx1 - vx3) - cu_sq);
- (distAD.f[DIR_P0M ])[kbe ] = c1o54 * conc * (c1o1 + c3o1 * ( vx1 - vx3) + c9o2 * ( vx1 - vx3) * ( vx1 - vx3) - cu_sq);
- (distAD.f[DIR_M0P ])[ktw ] = c1o54 * conc * (c1o1 + c3o1 * (-vx1 + vx3) + c9o2 * (-vx1 + vx3) * (-vx1 + vx3) - cu_sq);
- (distAD.f[DIR_0PP ])[ktn ] = c1o54 * conc * (c1o1 + c3o1 * ( vx2 + vx3) + c9o2 * ( vx2 + vx3) * ( vx2 + vx3) - cu_sq);
- (distAD.f[DIR_0MM ])[kbs ] = c1o54 * conc * (c1o1 + c3o1 * ( - vx2 - vx3) + c9o2 * ( - vx2 - vx3) * ( - vx2 - vx3) - cu_sq);
- (distAD.f[DIR_0PM ])[kbn ] = c1o54 * conc * (c1o1 + c3o1 * ( vx2 - vx3) + c9o2 * ( vx2 - vx3) * ( vx2 - vx3) - cu_sq);
- (distAD.f[DIR_0MP ])[kts ] = c1o54 * conc * (c1o1 + c3o1 * ( - vx2 + vx3) + c9o2 * ( - vx2 + vx3) * ( - vx2 + vx3) - cu_sq);
- (distAD.f[DIR_PPP ])[ktne ] = c1o216 * conc * (c1o1 + c3o1 * ( vx1 + vx2 + vx3) + c9o2 * ( vx1 + vx2 + vx3) * ( vx1 + vx2 + vx3) - cu_sq);
- (distAD.f[DIR_MMM ])[kbsw ] = c1o216 * conc * (c1o1 + c3o1 * (-vx1 - vx2 - vx3) + c9o2 * (-vx1 - vx2 - vx3) * (-vx1 - vx2 - vx3) - cu_sq);
- (distAD.f[DIR_PPM ])[kbne ] = c1o216 * conc * (c1o1 + c3o1 * ( vx1 + vx2 - vx3) + c9o2 * ( vx1 + vx2 - vx3) * ( vx1 + vx2 - vx3) - cu_sq);
- (distAD.f[DIR_MMP ])[ktsw ] = c1o216 * conc * (c1o1 + c3o1 * (-vx1 - vx2 + vx3) + c9o2 * (-vx1 - vx2 + vx3) * (-vx1 - vx2 + vx3) - cu_sq);
- (distAD.f[DIR_PMP ])[ktse ] = c1o216 * conc * (c1o1 + c3o1 * ( vx1 - vx2 + vx3) + c9o2 * ( vx1 - vx2 + vx3) * ( vx1 - vx2 + vx3) - cu_sq);
- (distAD.f[DIR_MPM ])[kbnw ] = c1o216 * conc * (c1o1 + c3o1 * (-vx1 + vx2 - vx3) + c9o2 * (-vx1 + vx2 - vx3) * (-vx1 + vx2 - vx3) - cu_sq);
- (distAD.f[DIR_PMM ])[kbse ] = c1o216 * conc * (c1o1 + c3o1 * ( vx1 - vx2 - vx3) + c9o2 * ( vx1 - vx2 - vx3) * ( vx1 - vx2 - vx3) - cu_sq);
- (distAD.f[DIR_MPP ])[ktnw ] = c1o216 * conc * (c1o1 + c3o1 * (-vx1 + vx2 + vx3) + c9o2 * (-vx1 + vx2 + vx3) * (-vx1 + vx2 + vx3) - cu_sq);
+ (distAD.f[DIR_P00])[ke ] = c2o27 * conc * (c1o1 + c3o1 * ( vx1 ) + c9o2 * ( vx1 ) * ( vx1 ) - cu_sq);
+ (distAD.f[DIR_M00])[kw ] = c2o27 * conc * (c1o1 + c3o1 * (-vx1 ) + c9o2 * (-vx1 ) * (-vx1 ) - cu_sq);
+ (distAD.f[DIR_0P0])[kn ] = c2o27 * conc * (c1o1 + c3o1 * ( vx2 ) + c9o2 * ( vx2 ) * ( vx2 ) - cu_sq);
+ (distAD.f[DIR_0M0])[ks ] = c2o27 * conc * (c1o1 + c3o1 * ( - vx2 ) + c9o2 * ( - vx2 ) * ( - vx2 ) - cu_sq);
+ (distAD.f[DIR_00P])[kt ] = c2o27 * conc * (c1o1 + c3o1 * ( vx3) + c9o2 * ( vx3) * ( vx3) - cu_sq);
+ (distAD.f[DIR_00M])[kb ] = c2o27 * conc * (c1o1 + c3o1 * ( - vx3) + c9o2 * ( - vx3) * ( - vx3) - cu_sq);
+ (distAD.f[DIR_PP0])[kne ] = c1o54 * conc * (c1o1 + c3o1 * ( vx1 + vx2 ) + c9o2 * ( vx1 + vx2 ) * ( vx1 + vx2 ) - cu_sq);
+ (distAD.f[DIR_MM0])[ksw ] = c1o54 * conc * (c1o1 + c3o1 * (-vx1 - vx2 ) + c9o2 * (-vx1 - vx2 ) * (-vx1 - vx2 ) - cu_sq);
+ (distAD.f[DIR_PM0])[kse ] = c1o54 * conc * (c1o1 + c3o1 * ( vx1 - vx2 ) + c9o2 * ( vx1 - vx2 ) * ( vx1 - vx2 ) - cu_sq);
+ (distAD.f[DIR_MP0])[knw ] = c1o54 * conc * (c1o1 + c3o1 * (-vx1 + vx2 ) + c9o2 * (-vx1 + vx2 ) * (-vx1 + vx2 ) - cu_sq);
+ (distAD.f[DIR_P0P])[kte ] = c1o54 * conc * (c1o1 + c3o1 * ( vx1 + vx3) + c9o2 * ( vx1 + vx3) * ( vx1 + vx3) - cu_sq);
+ (distAD.f[DIR_M0M])[kbw ] = c1o54 * conc * (c1o1 + c3o1 * (-vx1 - vx3) + c9o2 * (-vx1 - vx3) * (-vx1 - vx3) - cu_sq);
+ (distAD.f[DIR_P0M])[kbe ] = c1o54 * conc * (c1o1 + c3o1 * ( vx1 - vx3) + c9o2 * ( vx1 - vx3) * ( vx1 - vx3) - cu_sq);
+ (distAD.f[DIR_M0P])[ktw ] = c1o54 * conc * (c1o1 + c3o1 * (-vx1 + vx3) + c9o2 * (-vx1 + vx3) * (-vx1 + vx3) - cu_sq);
+ (distAD.f[DIR_0PP])[ktn ] = c1o54 * conc * (c1o1 + c3o1 * ( vx2 + vx3) + c9o2 * ( vx2 + vx3) * ( vx2 + vx3) - cu_sq);
+ (distAD.f[DIR_0MM])[kbs ] = c1o54 * conc * (c1o1 + c3o1 * ( - vx2 - vx3) + c9o2 * ( - vx2 - vx3) * ( - vx2 - vx3) - cu_sq);
+ (distAD.f[DIR_0PM])[kbn ] = c1o54 * conc * (c1o1 + c3o1 * ( vx2 - vx3) + c9o2 * ( vx2 - vx3) * ( vx2 - vx3) - cu_sq);
+ (distAD.f[DIR_0MP])[kts ] = c1o54 * conc * (c1o1 + c3o1 * ( - vx2 + vx3) + c9o2 * ( - vx2 + vx3) * ( - vx2 + vx3) - cu_sq);
+ (distAD.f[DIR_PPP])[ktne ] = c1o216 * conc * (c1o1 + c3o1 * ( vx1 + vx2 + vx3) + c9o2 * ( vx1 + vx2 + vx3) * ( vx1 + vx2 + vx3) - cu_sq);
+ (distAD.f[DIR_MMM])[kbsw ] = c1o216 * conc * (c1o1 + c3o1 * (-vx1 - vx2 - vx3) + c9o2 * (-vx1 - vx2 - vx3) * (-vx1 - vx2 - vx3) - cu_sq);
+ (distAD.f[DIR_PPM])[kbne ] = c1o216 * conc * (c1o1 + c3o1 * ( vx1 + vx2 - vx3) + c9o2 * ( vx1 + vx2 - vx3) * ( vx1 + vx2 - vx3) - cu_sq);
+ (distAD.f[DIR_MMP])[ktsw ] = c1o216 * conc * (c1o1 + c3o1 * (-vx1 - vx2 + vx3) + c9o2 * (-vx1 - vx2 + vx3) * (-vx1 - vx2 + vx3) - cu_sq);
+ (distAD.f[DIR_PMP])[ktse ] = c1o216 * conc * (c1o1 + c3o1 * ( vx1 - vx2 + vx3) + c9o2 * ( vx1 - vx2 + vx3) * ( vx1 - vx2 + vx3) - cu_sq);
+ (distAD.f[DIR_MPM])[kbnw ] = c1o216 * conc * (c1o1 + c3o1 * (-vx1 + vx2 - vx3) + c9o2 * (-vx1 + vx2 - vx3) * (-vx1 + vx2 - vx3) - cu_sq);
+ (distAD.f[DIR_PMM])[kbse ] = c1o216 * conc * (c1o1 + c3o1 * ( vx1 - vx2 - vx3) + c9o2 * ( vx1 - vx2 - vx3) * ( vx1 - vx2 - vx3) - cu_sq);
+ (distAD.f[DIR_MPP])[ktnw ] = c1o216 * conc * (c1o1 + c3o1 * (-vx1 + vx2 + vx3) + c9o2 * (-vx1 + vx2 + vx3) * (-vx1 + vx2 + vx3) - cu_sq);
}
}
@@ -263,63 +263,63 @@ __global__ void InitAD27(
// Distributions27 D27;
// if (EvenOrOdd==true)
// {
-// D27.f[DIR_P00 ] = &DD27[DIR_P00 *size_Mat];
-// D27.f[DIR_M00 ] = &DD27[DIR_M00 *size_Mat];
-// D27.f[DIR_0P0 ] = &DD27[DIR_0P0 *size_Mat];
-// D27.f[DIR_0M0 ] = &DD27[DIR_0M0 *size_Mat];
-// D27.f[DIR_00P ] = &DD27[DIR_00P *size_Mat];
-// D27.f[DIR_00M ] = &DD27[DIR_00M *size_Mat];
-// D27.f[DIR_PP0 ] = &DD27[DIR_PP0 *size_Mat];
-// D27.f[DIR_MM0 ] = &DD27[DIR_MM0 *size_Mat];
-// D27.f[DIR_PM0 ] = &DD27[DIR_PM0 *size_Mat];
-// D27.f[DIR_MP0 ] = &DD27[DIR_MP0 *size_Mat];
-// D27.f[DIR_P0P ] = &DD27[DIR_P0P *size_Mat];
-// D27.f[DIR_M0M ] = &DD27[DIR_M0M *size_Mat];
-// D27.f[DIR_P0M ] = &DD27[DIR_P0M *size_Mat];
-// D27.f[DIR_M0P ] = &DD27[DIR_M0P *size_Mat];
-// D27.f[DIR_0PP ] = &DD27[DIR_0PP *size_Mat];
-// D27.f[DIR_0MM ] = &DD27[DIR_0MM *size_Mat];
-// D27.f[DIR_0PM ] = &DD27[DIR_0PM *size_Mat];
-// D27.f[DIR_0MP ] = &DD27[DIR_0MP *size_Mat];
-// D27.f[DIR_000] = &DD27[DIR_000*size_Mat];
-// D27.f[DIR_PPP ] = &DD27[DIR_PPP *size_Mat];
-// D27.f[DIR_MMP ] = &DD27[DIR_MMP *size_Mat];
-// D27.f[DIR_PMP ] = &DD27[DIR_PMP *size_Mat];
-// D27.f[DIR_MPP ] = &DD27[DIR_MPP *size_Mat];
-// D27.f[DIR_PPM ] = &DD27[DIR_PPM *size_Mat];
-// D27.f[DIR_MMM ] = &DD27[DIR_MMM *size_Mat];
-// D27.f[DIR_PMM ] = &DD27[DIR_PMM *size_Mat];
-// D27.f[DIR_MPM ] = &DD27[DIR_MPM *size_Mat];
+// D27.f[DIR_P00] = &DD27[DIR_P00 * size_Mat];
+// D27.f[DIR_M00] = &DD27[DIR_M00 * size_Mat];
+// D27.f[DIR_0P0] = &DD27[DIR_0P0 * size_Mat];
+// D27.f[DIR_0M0] = &DD27[DIR_0M0 * size_Mat];
+// D27.f[DIR_00P] = &DD27[DIR_00P * size_Mat];
+// D27.f[DIR_00M] = &DD27[DIR_00M * size_Mat];
+// D27.f[DIR_PP0] = &DD27[DIR_PP0 * size_Mat];
+// D27.f[DIR_MM0] = &DD27[DIR_MM0 * size_Mat];
+// D27.f[DIR_PM0] = &DD27[DIR_PM0 * size_Mat];
+// D27.f[DIR_MP0] = &DD27[DIR_MP0 * size_Mat];
+// D27.f[DIR_P0P] = &DD27[DIR_P0P * size_Mat];
+// D27.f[DIR_M0M] = &DD27[DIR_M0M * size_Mat];
+// D27.f[DIR_P0M] = &DD27[DIR_P0M * size_Mat];
+// D27.f[DIR_M0P] = &DD27[DIR_M0P * size_Mat];
+// D27.f[DIR_0PP] = &DD27[DIR_0PP * size_Mat];
+// D27.f[DIR_0MM] = &DD27[DIR_0MM * size_Mat];
+// D27.f[DIR_0PM] = &DD27[DIR_0PM * size_Mat];
+// D27.f[DIR_0MP] = &DD27[DIR_0MP * size_Mat];
+// D27.f[DIR_000] = &DD27[DIR_000 * size_Mat];
+// D27.f[DIR_PPP] = &DD27[DIR_PPP * size_Mat];
+// D27.f[DIR_MMP] = &DD27[DIR_MMP * size_Mat];
+// D27.f[DIR_PMP] = &DD27[DIR_PMP * size_Mat];
+// D27.f[DIR_MPP] = &DD27[DIR_MPP * size_Mat];
+// D27.f[DIR_PPM] = &DD27[DIR_PPM * size_Mat];
+// D27.f[DIR_MMM] = &DD27[DIR_MMM * size_Mat];
+// D27.f[DIR_PMM] = &DD27[DIR_PMM * size_Mat];
+// D27.f[DIR_MPM] = &DD27[DIR_MPM * size_Mat];
// }
// else
// {
-// D27.f[DIR_M00 ] = &DD27[DIR_P00 *size_Mat];
-// D27.f[DIR_P00 ] = &DD27[DIR_M00 *size_Mat];
-// D27.f[DIR_0M0 ] = &DD27[DIR_0P0 *size_Mat];
-// D27.f[DIR_0P0 ] = &DD27[DIR_0M0 *size_Mat];
-// D27.f[DIR_00M ] = &DD27[DIR_00P *size_Mat];
-// D27.f[DIR_00P ] = &DD27[DIR_00M *size_Mat];
-// D27.f[DIR_MM0 ] = &DD27[DIR_PP0 *size_Mat];
-// D27.f[DIR_PP0 ] = &DD27[DIR_MM0 *size_Mat];
-// D27.f[DIR_MP0 ] = &DD27[DIR_PM0 *size_Mat];
-// D27.f[DIR_PM0 ] = &DD27[DIR_MP0 *size_Mat];
-// D27.f[DIR_M0M ] = &DD27[DIR_P0P *size_Mat];
-// D27.f[DIR_P0P ] = &DD27[DIR_M0M *size_Mat];
-// D27.f[DIR_M0P ] = &DD27[DIR_P0M *size_Mat];
-// D27.f[DIR_P0M ] = &DD27[DIR_M0P *size_Mat];
-// D27.f[DIR_0MM ] = &DD27[DIR_0PP *size_Mat];
-// D27.f[DIR_0PP ] = &DD27[DIR_0MM *size_Mat];
-// D27.f[DIR_0MP ] = &DD27[DIR_0PM *size_Mat];
-// D27.f[DIR_0PM ] = &DD27[DIR_0MP *size_Mat];
-// D27.f[DIR_000] = &DD27[DIR_000*size_Mat];
-// D27.f[DIR_MMM ] = &DD27[DIR_PPP *size_Mat];
-// D27.f[DIR_PPM ] = &DD27[DIR_MMP *size_Mat];
-// D27.f[DIR_MPM ] = &DD27[DIR_PMP *size_Mat];
-// D27.f[DIR_PMM ] = &DD27[DIR_MPP *size_Mat];
-// D27.f[DIR_MMP ] = &DD27[DIR_PPM *size_Mat];
-// D27.f[DIR_PPP ] = &DD27[DIR_MMM *size_Mat];
-// D27.f[DIR_MPP ] = &DD27[DIR_PMM *size_Mat];
-// D27.f[DIR_PMP ] = &DD27[DIR_MPM *size_Mat];
+// D27.f[DIR_M00] = &DD27[DIR_P00 * size_Mat];
+// D27.f[DIR_P00] = &DD27[DIR_M00 * size_Mat];
+// D27.f[DIR_0M0] = &DD27[DIR_0P0 * size_Mat];
+// D27.f[DIR_0P0] = &DD27[DIR_0M0 * size_Mat];
+// D27.f[DIR_00M] = &DD27[DIR_00P * size_Mat];
+// D27.f[DIR_00P] = &DD27[DIR_00M * size_Mat];
+// D27.f[DIR_MM0] = &DD27[DIR_PP0 * size_Mat];
+// D27.f[DIR_PP0] = &DD27[DIR_MM0 * size_Mat];
+// D27.f[DIR_MP0] = &DD27[DIR_PM0 * size_Mat];
+// D27.f[DIR_PM0] = &DD27[DIR_MP0 * size_Mat];
+// D27.f[DIR_M0M] = &DD27[DIR_P0P * size_Mat];
+// D27.f[DIR_P0P] = &DD27[DIR_M0M * size_Mat];
+// D27.f[DIR_M0P] = &DD27[DIR_P0M * size_Mat];
+// D27.f[DIR_P0M] = &DD27[DIR_M0P * size_Mat];
+// D27.f[DIR_0MM] = &DD27[DIR_0PP * size_Mat];
+// D27.f[DIR_0PP] = &DD27[DIR_0MM * size_Mat];
+// D27.f[DIR_0MP] = &DD27[DIR_0PM * size_Mat];
+// D27.f[DIR_0PM] = &DD27[DIR_0MP * size_Mat];
+// D27.f[DIR_000] = &DD27[DIR_000 * size_Mat];
+// D27.f[DIR_MMM] = &DD27[DIR_PPP * size_Mat];
+// D27.f[DIR_PPM] = &DD27[DIR_MMP * size_Mat];
+// D27.f[DIR_MPM] = &DD27[DIR_PMP * size_Mat];
+// D27.f[DIR_PMM] = &DD27[DIR_MPP * size_Mat];
+// D27.f[DIR_MMP] = &DD27[DIR_PPM * size_Mat];
+// D27.f[DIR_PPP] = &DD27[DIR_MMM * size_Mat];
+// D27.f[DIR_MPP] = &DD27[DIR_PMM * size_Mat];
+// D27.f[DIR_PMP] = &DD27[DIR_MPM * size_Mat];
// }
// //////////////////////////////////////////////////////////////////////////
// real ConcD = Conc[k];
@@ -391,32 +391,32 @@ __global__ void InitAD27(
// real cu_sq=c3o2*(vx1*vx1+vx2*vx2+vx3*vx3);
// (D27.f[DIR_000])[kzero] = c8o27* ConcD*(c1o1-cu_sq);
-// (D27.f[DIR_P00 ])[ke ] = c2o27* ConcD*(c1o1+c3o1*( vx1 )+c9o2*( vx1 )*( vx1 )-cu_sq);
-// (D27.f[DIR_M00 ])[kw ] = c2o27* ConcD*(c1o1+c3o1*(-vx1 )+c9o2*(-vx1 )*(-vx1 )-cu_sq);
-// (D27.f[DIR_0P0 ])[kn ] = c2o27* ConcD*(c1o1+c3o1*( vx2 )+c9o2*( vx2 )*( vx2 )-cu_sq);
-// (D27.f[DIR_0M0 ])[ks ] = c2o27* ConcD*(c1o1+c3o1*( -vx2 )+c9o2*( -vx2 )*( -vx2 )-cu_sq);
-// (D27.f[DIR_00P ])[kt ] = c2o27* ConcD*(c1o1+c3o1*( vx3)+c9o2*( vx3)*( vx3)-cu_sq);
-// (D27.f[DIR_00M ])[kb ] = c2o27* ConcD*(c1o1+c3o1*( -vx3)+c9o2*( -vx3)*( -vx3)-cu_sq);
-// (D27.f[DIR_PP0 ])[kne ] = c1o54* ConcD*(c1o1+c3o1*( vx1+vx2 )+c9o2*( vx1+vx2 )*( vx1+vx2 )-cu_sq);
-// (D27.f[DIR_MM0 ])[ksw ] = c1o54* ConcD*(c1o1+c3o1*(-vx1-vx2 )+c9o2*(-vx1-vx2 )*(-vx1-vx2 )-cu_sq);
-// (D27.f[DIR_PM0 ])[kse ] = c1o54* ConcD*(c1o1+c3o1*( vx1-vx2 )+c9o2*( vx1-vx2 )*( vx1-vx2 )-cu_sq);
-// (D27.f[DIR_MP0 ])[knw ] = c1o54* ConcD*(c1o1+c3o1*(-vx1+vx2 )+c9o2*(-vx1+vx2 )*(-vx1+vx2 )-cu_sq);
-// (D27.f[DIR_P0P ])[kte ] = c1o54* ConcD*(c1o1+c3o1*( vx1 +vx3)+c9o2*( vx1 +vx3)*( vx1 +vx3)-cu_sq);
-// (D27.f[DIR_M0M ])[kbw ] = c1o54* ConcD*(c1o1+c3o1*(-vx1 -vx3)+c9o2*(-vx1 -vx3)*(-vx1 -vx3)-cu_sq);
-// (D27.f[DIR_P0M ])[kbe ] = c1o54* ConcD*(c1o1+c3o1*( vx1 -vx3)+c9o2*( vx1 -vx3)*( vx1 -vx3)-cu_sq);
-// (D27.f[DIR_M0P ])[ktw ] = c1o54* ConcD*(c1o1+c3o1*(-vx1 +vx3)+c9o2*(-vx1 +vx3)*(-vx1 +vx3)-cu_sq);
-// (D27.f[DIR_0PP ])[ktn ] = c1o54* ConcD*(c1o1+c3o1*( vx2+vx3)+c9o2*( vx2+vx3)*( vx2+vx3)-cu_sq);
-// (D27.f[DIR_0MM ])[kbs ] = c1o54* ConcD*(c1o1+c3o1*( -vx2-vx3)+c9o2*( -vx2-vx3)*( -vx2-vx3)-cu_sq);
-// (D27.f[DIR_0PM ])[kbn ] = c1o54* ConcD*(c1o1+c3o1*( vx2-vx3)+c9o2*( vx2-vx3)*( vx2-vx3)-cu_sq);
-// (D27.f[DIR_0MP ])[kts ] = c1o54* ConcD*(c1o1+c3o1*( -vx2+vx3)+c9o2*( -vx2+vx3)*( -vx2+vx3)-cu_sq);
-// (D27.f[DIR_PPP ])[ktne ] = c1o216*ConcD*(c1o1+c3o1*( vx1+vx2+vx3)+c9o2*( vx1+vx2+vx3)*( vx1+vx2+vx3)-cu_sq);
-// (D27.f[DIR_MMM ])[kbsw ] = c1o216*ConcD*(c1o1+c3o1*(-vx1-vx2-vx3)+c9o2*(-vx1-vx2-vx3)*(-vx1-vx2-vx3)-cu_sq);
-// (D27.f[DIR_PPM ])[kbne ] = c1o216*ConcD*(c1o1+c3o1*( vx1+vx2-vx3)+c9o2*( vx1+vx2-vx3)*( vx1+vx2-vx3)-cu_sq);
-// (D27.f[DIR_MMP ])[ktsw ] = c1o216*ConcD*(c1o1+c3o1*(-vx1-vx2+vx3)+c9o2*(-vx1-vx2+vx3)*(-vx1-vx2+vx3)-cu_sq);
-// (D27.f[DIR_PMP ])[ktse ] = c1o216*ConcD*(c1o1+c3o1*( vx1-vx2+vx3)+c9o2*( vx1-vx2+vx3)*( vx1-vx2+vx3)-cu_sq);
-// (D27.f[DIR_MPM ])[kbnw ] = c1o216*ConcD*(c1o1+c3o1*(-vx1+vx2-vx3)+c9o2*(-vx1+vx2-vx3)*(-vx1+vx2-vx3)-cu_sq);
-// (D27.f[DIR_PMM ])[kbse ] = c1o216*ConcD*(c1o1+c3o1*( vx1-vx2-vx3)+c9o2*( vx1-vx2-vx3)*( vx1-vx2-vx3)-cu_sq);
-// (D27.f[DIR_MPP ])[ktnw ] = c1o216*ConcD*(c1o1+c3o1*(-vx1+vx2+vx3)+c9o2*(-vx1+vx2+vx3)*(-vx1+vx2+vx3)-cu_sq);
+// (D27.f[DIR_P00])[ke ] = c2o27* ConcD*(c1o1+c3o1*( vx1 )+c9o2*( vx1 )*( vx1 )-cu_sq);
+// (D27.f[DIR_M00])[kw ] = c2o27* ConcD*(c1o1+c3o1*(-vx1 )+c9o2*(-vx1 )*(-vx1 )-cu_sq);
+// (D27.f[DIR_0P0])[kn ] = c2o27* ConcD*(c1o1+c3o1*( vx2 )+c9o2*( vx2 )*( vx2 )-cu_sq);
+// (D27.f[DIR_0M0])[ks ] = c2o27* ConcD*(c1o1+c3o1*( -vx2 )+c9o2*( -vx2 )*( -vx2 )-cu_sq);
+// (D27.f[DIR_00P])[kt ] = c2o27* ConcD*(c1o1+c3o1*( vx3)+c9o2*( vx3)*( vx3)-cu_sq);
+// (D27.f[DIR_00M])[kb ] = c2o27* ConcD*(c1o1+c3o1*( -vx3)+c9o2*( -vx3)*( -vx3)-cu_sq);
+// (D27.f[DIR_PP0])[kne ] = c1o54* ConcD*(c1o1+c3o1*( vx1+vx2 )+c9o2*( vx1+vx2 )*( vx1+vx2 )-cu_sq);
+// (D27.f[DIR_MM0])[ksw ] = c1o54* ConcD*(c1o1+c3o1*(-vx1-vx2 )+c9o2*(-vx1-vx2 )*(-vx1-vx2 )-cu_sq);
+// (D27.f[DIR_PM0])[kse ] = c1o54* ConcD*(c1o1+c3o1*( vx1-vx2 )+c9o2*( vx1-vx2 )*( vx1-vx2 )-cu_sq);
+// (D27.f[DIR_MP0])[knw ] = c1o54* ConcD*(c1o1+c3o1*(-vx1+vx2 )+c9o2*(-vx1+vx2 )*(-vx1+vx2 )-cu_sq);
+// (D27.f[DIR_P0P])[kte ] = c1o54* ConcD*(c1o1+c3o1*( vx1 +vx3)+c9o2*( vx1 +vx3)*( vx1 +vx3)-cu_sq);
+// (D27.f[DIR_M0M])[kbw ] = c1o54* ConcD*(c1o1+c3o1*(-vx1 -vx3)+c9o2*(-vx1 -vx3)*(-vx1 -vx3)-cu_sq);
+// (D27.f[DIR_P0M])[kbe ] = c1o54* ConcD*(c1o1+c3o1*( vx1 -vx3)+c9o2*( vx1 -vx3)*( vx1 -vx3)-cu_sq);
+// (D27.f[DIR_M0P])[ktw ] = c1o54* ConcD*(c1o1+c3o1*(-vx1 +vx3)+c9o2*(-vx1 +vx3)*(-vx1 +vx3)-cu_sq);
+// (D27.f[DIR_0PP])[ktn ] = c1o54* ConcD*(c1o1+c3o1*( vx2+vx3)+c9o2*( vx2+vx3)*( vx2+vx3)-cu_sq);
+// (D27.f[DIR_0MM])[kbs ] = c1o54* ConcD*(c1o1+c3o1*( -vx2-vx3)+c9o2*( -vx2-vx3)*( -vx2-vx3)-cu_sq);
+// (D27.f[DIR_0PM])[kbn ] = c1o54* ConcD*(c1o1+c3o1*( vx2-vx3)+c9o2*( vx2-vx3)*( vx2-vx3)-cu_sq);
+// (D27.f[DIR_0MP])[kts ] = c1o54* ConcD*(c1o1+c3o1*( -vx2+vx3)+c9o2*( -vx2+vx3)*( -vx2+vx3)-cu_sq);
+// (D27.f[DIR_PPP])[ktne ] = c1o216*ConcD*(c1o1+c3o1*( vx1+vx2+vx3)+c9o2*( vx1+vx2+vx3)*( vx1+vx2+vx3)-cu_sq);
+// (D27.f[DIR_MMM])[kbsw ] = c1o216*ConcD*(c1o1+c3o1*(-vx1-vx2-vx3)+c9o2*(-vx1-vx2-vx3)*(-vx1-vx2-vx3)-cu_sq);
+// (D27.f[DIR_PPM])[kbne ] = c1o216*ConcD*(c1o1+c3o1*( vx1+vx2-vx3)+c9o2*( vx1+vx2-vx3)*( vx1+vx2-vx3)-cu_sq);
+// (D27.f[DIR_MMP])[ktsw ] = c1o216*ConcD*(c1o1+c3o1*(-vx1-vx2+vx3)+c9o2*(-vx1-vx2+vx3)*(-vx1-vx2+vx3)-cu_sq);
+// (D27.f[DIR_PMP])[ktse ] = c1o216*ConcD*(c1o1+c3o1*( vx1-vx2+vx3)+c9o2*( vx1-vx2+vx3)*( vx1-vx2+vx3)-cu_sq);
+// (D27.f[DIR_MPM])[kbnw ] = c1o216*ConcD*(c1o1+c3o1*(-vx1+vx2-vx3)+c9o2*(-vx1+vx2-vx3)*(-vx1+vx2-vx3)-cu_sq);
+// (D27.f[DIR_PMM])[kbse ] = c1o216*ConcD*(c1o1+c3o1*( vx1-vx2-vx3)+c9o2*( vx1-vx2-vx3)*( vx1-vx2-vx3)-cu_sq);
+// (D27.f[DIR_MPP])[ktnw ] = c1o216*ConcD*(c1o1+c3o1*(-vx1+vx2+vx3)+c9o2*(-vx1+vx2+vx3)*(-vx1+vx2+vx3)-cu_sq);
// ////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
// }
// }
@@ -448,7 +448,7 @@ __global__ void InitAD7( unsigned int* neighborX,
real* ux,
real* uy,
real* uz,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
real* DD7,
bool EvenOrOdd)
{
@@ -463,7 +463,7 @@ __global__ void InitAD7( unsigned int* neighborX,
const unsigned k = nx*(ny*z + y) + x;
//////////////////////////////////////////////////////////////////////////
- if(k<size_Mat)
+ if(k<numberOfLBnodes)
{
////////////////////////////////////////////////////////////////////////////////
unsigned int BC;
@@ -474,23 +474,23 @@ __global__ void InitAD7( unsigned int* neighborX,
Distributions7 D7;
if (EvenOrOdd==true)
{
- D7.f[0] = &DD7[0*size_Mat];
- D7.f[1] = &DD7[1*size_Mat];
- D7.f[2] = &DD7[2*size_Mat];
- D7.f[3] = &DD7[3*size_Mat];
- D7.f[4] = &DD7[4*size_Mat];
- D7.f[5] = &DD7[5*size_Mat];
- D7.f[6] = &DD7[6*size_Mat];
+ D7.f[0] = &DD7[0*numberOfLBnodes];
+ D7.f[1] = &DD7[1*numberOfLBnodes];
+ D7.f[2] = &DD7[2*numberOfLBnodes];
+ D7.f[3] = &DD7[3*numberOfLBnodes];
+ D7.f[4] = &DD7[4*numberOfLBnodes];
+ D7.f[5] = &DD7[5*numberOfLBnodes];
+ D7.f[6] = &DD7[6*numberOfLBnodes];
}
else
{
- D7.f[0] = &DD7[0*size_Mat];
- D7.f[2] = &DD7[1*size_Mat];
- D7.f[1] = &DD7[2*size_Mat];
- D7.f[4] = &DD7[3*size_Mat];
- D7.f[3] = &DD7[4*size_Mat];
- D7.f[6] = &DD7[5*size_Mat];
- D7.f[5] = &DD7[6*size_Mat];
+ D7.f[0] = &DD7[0*numberOfLBnodes];
+ D7.f[2] = &DD7[1*numberOfLBnodes];
+ D7.f[1] = &DD7[2*numberOfLBnodes];
+ D7.f[4] = &DD7[3*numberOfLBnodes];
+ D7.f[3] = &DD7[4*numberOfLBnodes];
+ D7.f[6] = &DD7[5*numberOfLBnodes];
+ D7.f[5] = &DD7[6*numberOfLBnodes];
}
//////////////////////////////////////////////////////////////////////////
real ConcD = Conc[k];
diff --git a/src/gpu/VirtualFluids_GPU/GPU/KernelUtilities.h b/src/gpu/VirtualFluids_GPU/GPU/KernelUtilities.h
deleted file mode 100644
index 2f6a11aa17398b65858508c3f94b241c16551b37..0000000000000000000000000000000000000000
--- a/src/gpu/VirtualFluids_GPU/GPU/KernelUtilities.h
+++ /dev/null
@@ -1,177 +0,0 @@
-//=======================================================================================
-// ____ ____ __ ______ __________ __ __ __ __
-// \ \ | | | | | _ \ |___ ___| | | | | / \ | |
-// \ \ | | | | | |_) | | | | | | | / \ | |
-// \ \ | | | | | _ / | | | | | | / /\ \ | |
-// \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
-// \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
-// \ \ | | ________________________________________________________________
-// \ \ | | | ______________________________________________________________|
-// \ \| | | | __ __ __ __ ______ _______
-// \ | | |_____ | | | | | | | | | _ \ / _____)
-// \ | | _____| | | | | | | | | | | \ \ \_______
-// \ | | | | |_____ | \_/ | | | | |_/ / _____ |
-// \ _____| |__| |________| \_______/ |__| |______/ (_______/
-//
-// This file is part of VirtualFluids. VirtualFluids is free software: you can
-// redistribute it and/or modify it under the terms of the GNU General Public
-// License as published by the Free Software Foundation, either version 3 of
-// the License, or (at your option) any later version.
-//
-// VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
-// ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
-// FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
-// for more details.
-//
-// You should have received a copy of the GNU General Public License along
-// with VirtualFluids (see COPYING.txt). If not, see <http://www.gnu.org/licenses/>.
-//
-//! \file KernelUtilities.h
-//! \ingroup GPU
-//! \author Martin Schoenherr, Anna Wellmann
-//======================================================================================
-#ifndef KERNELUTILS_H
-#define KERNELUTILS_H
-
-#include "LBM/LB.h"
-#include "lbm/constants/D3Q27.h"
-#include "lbm/constants/NumericConstants.h"
-
-using namespace vf::lbm::constant;
-using namespace vf::lbm::dir;
-
-__inline__ __device__ void getPointersToDistributions(Distributions27 &dist, real *distributionArray, const uint numberOfLBnodes, const bool isEvenTimestep)
-{
- if (isEvenTimestep)
- {
- dist.f[DIR_P00 ] = &distributionArray[DIR_P00 *numberOfLBnodes];
- dist.f[DIR_M00 ] = &distributionArray[DIR_M00 *numberOfLBnodes];
- dist.f[DIR_0P0 ] = &distributionArray[DIR_0P0 *numberOfLBnodes];
- dist.f[DIR_0M0 ] = &distributionArray[DIR_0M0 *numberOfLBnodes];
- dist.f[DIR_00P ] = &distributionArray[DIR_00P *numberOfLBnodes];
- dist.f[DIR_00M ] = &distributionArray[DIR_00M *numberOfLBnodes];
- dist.f[DIR_PP0 ] = &distributionArray[DIR_PP0 *numberOfLBnodes];
- dist.f[DIR_MM0 ] = &distributionArray[DIR_MM0 *numberOfLBnodes];
- dist.f[DIR_PM0 ] = &distributionArray[DIR_PM0 *numberOfLBnodes];
- dist.f[DIR_MP0 ] = &distributionArray[DIR_MP0 *numberOfLBnodes];
- dist.f[DIR_P0P ] = &distributionArray[DIR_P0P *numberOfLBnodes];
- dist.f[DIR_M0M ] = &distributionArray[DIR_M0M *numberOfLBnodes];
- dist.f[DIR_P0M ] = &distributionArray[DIR_P0M *numberOfLBnodes];
- dist.f[DIR_M0P ] = &distributionArray[DIR_M0P *numberOfLBnodes];
- dist.f[DIR_0PP ] = &distributionArray[DIR_0PP *numberOfLBnodes];
- dist.f[DIR_0MM ] = &distributionArray[DIR_0MM *numberOfLBnodes];
- dist.f[DIR_0PM ] = &distributionArray[DIR_0PM *numberOfLBnodes];
- dist.f[DIR_0MP ] = &distributionArray[DIR_0MP *numberOfLBnodes];
- dist.f[DIR_000] = &distributionArray[DIR_000*numberOfLBnodes];
- dist.f[DIR_PPP ] = &distributionArray[DIR_PPP *numberOfLBnodes];
- dist.f[DIR_MMP ] = &distributionArray[DIR_MMP *numberOfLBnodes];
- dist.f[DIR_PMP ] = &distributionArray[DIR_PMP *numberOfLBnodes];
- dist.f[DIR_MPP ] = &distributionArray[DIR_MPP *numberOfLBnodes];
- dist.f[DIR_PPM ] = &distributionArray[DIR_PPM *numberOfLBnodes];
- dist.f[DIR_MMM ] = &distributionArray[DIR_MMM *numberOfLBnodes];
- dist.f[DIR_PMM ] = &distributionArray[DIR_PMM *numberOfLBnodes];
- dist.f[DIR_MPM ] = &distributionArray[DIR_MPM *numberOfLBnodes];
- }
- else
- {
- dist.f[DIR_M00 ] = &distributionArray[DIR_P00 *numberOfLBnodes];
- dist.f[DIR_P00 ] = &distributionArray[DIR_M00 *numberOfLBnodes];
- dist.f[DIR_0M0 ] = &distributionArray[DIR_0P0 *numberOfLBnodes];
- dist.f[DIR_0P0 ] = &distributionArray[DIR_0M0 *numberOfLBnodes];
- dist.f[DIR_00M ] = &distributionArray[DIR_00P *numberOfLBnodes];
- dist.f[DIR_00P ] = &distributionArray[DIR_00M *numberOfLBnodes];
- dist.f[DIR_MM0 ] = &distributionArray[DIR_PP0 *numberOfLBnodes];
- dist.f[DIR_PP0 ] = &distributionArray[DIR_MM0 *numberOfLBnodes];
- dist.f[DIR_MP0 ] = &distributionArray[DIR_PM0 *numberOfLBnodes];
- dist.f[DIR_PM0 ] = &distributionArray[DIR_MP0 *numberOfLBnodes];
- dist.f[DIR_M0M ] = &distributionArray[DIR_P0P *numberOfLBnodes];
- dist.f[DIR_P0P ] = &distributionArray[DIR_M0M *numberOfLBnodes];
- dist.f[DIR_M0P ] = &distributionArray[DIR_P0M *numberOfLBnodes];
- dist.f[DIR_P0M ] = &distributionArray[DIR_M0P *numberOfLBnodes];
- dist.f[DIR_0MM ] = &distributionArray[DIR_0PP *numberOfLBnodes];
- dist.f[DIR_0PP ] = &distributionArray[DIR_0MM *numberOfLBnodes];
- dist.f[DIR_0MP ] = &distributionArray[DIR_0PM *numberOfLBnodes];
- dist.f[DIR_0PM ] = &distributionArray[DIR_0MP *numberOfLBnodes];
- dist.f[DIR_000] = &distributionArray[DIR_000*numberOfLBnodes];
- dist.f[DIR_PPP ] = &distributionArray[DIR_MMM *numberOfLBnodes];
- dist.f[DIR_MMP ] = &distributionArray[DIR_PPM *numberOfLBnodes];
- dist.f[DIR_PMP ] = &distributionArray[DIR_MPM *numberOfLBnodes];
- dist.f[DIR_MPP ] = &distributionArray[DIR_PMM *numberOfLBnodes];
- dist.f[DIR_PPM ] = &distributionArray[DIR_MMP *numberOfLBnodes];
- dist.f[DIR_MMM ] = &distributionArray[DIR_PPP *numberOfLBnodes];
- dist.f[DIR_PMM ] = &distributionArray[DIR_MPP *numberOfLBnodes];
- dist.f[DIR_MPM ] = &distributionArray[DIR_PMP *numberOfLBnodes];
- }
-}
-
-__inline__ __device__ void getPointersToSubgridDistances(SubgridDistances27& subgridD, real* subgridDistances, const unsigned int numberOfSubgridIndices)
-{
- subgridD.q[DIR_P00 ] = &subgridDistances[DIR_P00 *numberOfSubgridIndices];
- subgridD.q[DIR_M00 ] = &subgridDistances[DIR_M00 *numberOfSubgridIndices];
- subgridD.q[DIR_0P0 ] = &subgridDistances[DIR_0P0 *numberOfSubgridIndices];
- subgridD.q[DIR_0M0 ] = &subgridDistances[DIR_0M0 *numberOfSubgridIndices];
- subgridD.q[DIR_00P ] = &subgridDistances[DIR_00P *numberOfSubgridIndices];
- subgridD.q[DIR_00M ] = &subgridDistances[DIR_00M *numberOfSubgridIndices];
- subgridD.q[DIR_PP0 ] = &subgridDistances[DIR_PP0 *numberOfSubgridIndices];
- subgridD.q[DIR_MM0 ] = &subgridDistances[DIR_MM0 *numberOfSubgridIndices];
- subgridD.q[DIR_PM0 ] = &subgridDistances[DIR_PM0 *numberOfSubgridIndices];
- subgridD.q[DIR_MP0 ] = &subgridDistances[DIR_MP0 *numberOfSubgridIndices];
- subgridD.q[DIR_P0P ] = &subgridDistances[DIR_P0P *numberOfSubgridIndices];
- subgridD.q[DIR_M0M ] = &subgridDistances[DIR_M0M *numberOfSubgridIndices];
- subgridD.q[DIR_P0M ] = &subgridDistances[DIR_P0M *numberOfSubgridIndices];
- subgridD.q[DIR_M0P ] = &subgridDistances[DIR_M0P *numberOfSubgridIndices];
- subgridD.q[DIR_0PP ] = &subgridDistances[DIR_0PP *numberOfSubgridIndices];
- subgridD.q[DIR_0MM ] = &subgridDistances[DIR_0MM *numberOfSubgridIndices];
- subgridD.q[DIR_0PM ] = &subgridDistances[DIR_0PM *numberOfSubgridIndices];
- subgridD.q[DIR_0MP ] = &subgridDistances[DIR_0MP *numberOfSubgridIndices];
- subgridD.q[DIR_000] = &subgridDistances[DIR_000 *numberOfSubgridIndices];
- subgridD.q[DIR_PPP ] = &subgridDistances[DIR_PPP *numberOfSubgridIndices];
- subgridD.q[DIR_MMP ] = &subgridDistances[DIR_MMP *numberOfSubgridIndices];
- subgridD.q[DIR_PMP ] = &subgridDistances[DIR_PMP *numberOfSubgridIndices];
- subgridD.q[DIR_MPP ] = &subgridDistances[DIR_MPP *numberOfSubgridIndices];
- subgridD.q[DIR_PPM ] = &subgridDistances[DIR_PPM *numberOfSubgridIndices];
- subgridD.q[DIR_MMM ] = &subgridDistances[DIR_MMM *numberOfSubgridIndices];
- subgridD.q[DIR_PMM ] = &subgridDistances[DIR_PMM *numberOfSubgridIndices];
- subgridD.q[DIR_MPM ] = &subgridDistances[DIR_MPM *numberOfSubgridIndices];
-}
-
-__inline__ __device__ real getEquilibriumForBC(const real& drho, const real& velocity, const real& cu_sq, const real weight)
-{
- return weight * (drho + c9o2 * velocity * velocity * (c1o1 + drho) - cu_sq);
-}
-
-__inline__ __device__ real getInterpolatedDistributionForVeloBC(const real& q, const real& f, const real& fInverse, const real& feq,
- const real& omega, const real& velocity, const real weight)
-{
-
- return (c1o1-q) / (c1o1+q) * (f - fInverse + (f + fInverse - c2o1 * feq * omega) / (c1o1 - omega)) * c1o2
- + (q * (f + fInverse) - c6o1 * weight * velocity) / (c1o1 + q);
-}
-
-__inline__ __device__ real getBounceBackDistributionForVeloBC( const real& f,
- const real& velocity, const real weight)
-{
-
- return f - (c6o1 * weight * velocity);
-}
-
-__inline__ __device__ real getInterpolatedDistributionForNoSlipBC(const real& q, const real& f, const real& fInverse, const real& feq,
- const real& omega)
-{
-
- return (c1o1-q) / (c1o1+q) * (f - fInverse + (f + fInverse - c2o1 * feq * omega) / (c1o1 - omega)) * c1o2
- + (q * (f + fInverse)) / (c1o1 + q);
-}
-
-
-__inline__ __device__ real getInterpolatedDistributionForVeloWithPressureBC(const real& q, const real& f, const real& fInverse, const real& feq,
- const real& omega, const real& drho, const real& velocity, const real weight)
-{
-
- return (c1o1-q) / (c1o1+q) * (f - fInverse + (f + fInverse - c2o1 * feq * omega) / (c1o1 - omega)) * c1o2
- + (q * (f + fInverse) - c6o1 * weight * velocity) / (c1o1 + q) - weight * drho;
-}
-
-
-
-#endif
diff --git a/src/gpu/VirtualFluids_GPU/GPU/KineticEnergyAnalyzer.cu b/src/gpu/VirtualFluids_GPU/GPU/KineticEnergyAnalyzer.cu
index 51368bbe09e6fc43a7a1ff6b8b15387417774964..b05cb9201ce30038bd6edf52e2e95a13c6f6d7d4 100644
--- a/src/gpu/VirtualFluids_GPU/GPU/KineticEnergyAnalyzer.cu
+++ b/src/gpu/VirtualFluids_GPU/GPU/KineticEnergyAnalyzer.cu
@@ -14,7 +14,7 @@
#include <iomanip>
-//#include "Core/Logger/Logger.h"
+#include "cuda/CudaGrid.h"
#include "Parameter/Parameter.h"
// includes, kernels
@@ -24,7 +24,7 @@
using namespace vf::lbm::constant;
using namespace vf::lbm::dir;
-__global__ void kineticEnergyKernel (real* vx, real* vy, real* vz, real* rho, uint* neighborX, uint* neighborY, uint* neighborZ, uint* neighborWSB, uint* geo, real* kineticEnergy, uint* isFluid, uint size_Mat);
+__global__ void kineticEnergyKernel (real* vx, real* vy, real* vz, real* rho, uint* neighborX, uint* neighborY, uint* neighborZ, uint* neighborWSB, uint* geo, real* kineticEnergy, uint* isFluid, unsigned long long numberOfLBnodes);
__host__ __device__ inline void kineticEnergyFunction(real* vx, real* vy, real* vz, real* rho, uint* neighborX, uint* neighborY, uint* neighborZ, uint* neighborWSB, uint* geo, real* kineticEnergy, uint* isFluid, uint index);
@@ -35,56 +35,42 @@ bool KineticEnergyAnalyzer::run(uint iter)
if( iter % this->analyzeIter != 0 ) return false;
int lev = 0;
- int size_Mat = this->para->getParD(lev)->numberOfNodes;
-
- thrust::device_vector<real> kineticEnergy(size_Mat, c0o1);
- thrust::device_vector<uint> isFluid (size_Mat, 0);
-
- unsigned int numberOfThreads = 128;
- int Grid = (size_Mat / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 grid(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- LBCalcMacCompSP27<<< grid, threads >>> (para->getParD(lev)->velocityX,
- para->getParD(lev)->velocityY,
- para->getParD(lev)->velocityZ,
- para->getParD(lev)->rho,
- para->getParD(lev)->pressure,
- para->getParD(lev)->typeOfGridNode,
- para->getParD(lev)->neighborX,
- para->getParD(lev)->neighborY,
- para->getParD(lev)->neighborZ,
- para->getParD(lev)->numberOfNodes,
- para->getParD(lev)->distributions.f[0],
- para->getParD(lev)->isEvenTimestep);
- getLastCudaError("LBCalcMacSP27 execution failed");
-
- kineticEnergyKernel <<< grid, threads >>> ( para->getParD(lev)->velocityX,
- para->getParD(lev)->velocityY,
- para->getParD(lev)->velocityZ,
- para->getParD(lev)->rho,
- para->getParD(lev)->neighborX,
- para->getParD(lev)->neighborY,
- para->getParD(lev)->neighborZ,
- para->getParD(lev)->neighborInverse,
- para->getParD(lev)->typeOfGridNode,
- kineticEnergy.data().get(),
- isFluid.data().get(),
- size_Mat);
- cudaDeviceSynchronize();
-
- getLastCudaError("kineticEnergyKernel execution failed");
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(para->getParD(lev)->numberofthreads, para->getParD(lev)->numberOfNodes);
+
+ thrust::device_vector<real> kineticEnergy( this->para->getParD(lev)->numberOfNodes, c0o1);
+ thrust::device_vector<uint> isFluid ( this->para->getParD(lev)->numberOfNodes, 0);
+
+ LBCalcMacCompSP27<<< grid.grid, grid.threads >>>(
+ para->getParD(lev)->velocityX,
+ para->getParD(lev)->velocityY,
+ para->getParD(lev)->velocityZ,
+ para->getParD(lev)->rho,
+ para->getParD(lev)->pressure,
+ para->getParD(lev)->typeOfGridNode,
+ para->getParD(lev)->neighborX,
+ para->getParD(lev)->neighborY,
+ para->getParD(lev)->neighborZ,
+ para->getParD(lev)->numberOfNodes,
+ para->getParD(lev)->distributions.f[0],
+ para->getParD(lev)->isEvenTimestep);
+ getLastCudaError("LBCalcMacCompSP27 execution failed");
+
+ kineticEnergyKernel<<< grid.grid, grid.threads >>>(
+ para->getParD(lev)->velocityX,
+ para->getParD(lev)->velocityY,
+ para->getParD(lev)->velocityZ,
+ para->getParD(lev)->rho,
+ para->getParD(lev)->neighborX,
+ para->getParD(lev)->neighborY,
+ para->getParD(lev)->neighborZ,
+ para->getParD(lev)->neighborInverse,
+ para->getParD(lev)->typeOfGridNode,
+ kineticEnergy.data().get(),
+ isFluid.data().get(),
+ para->getParD(lev)->numberOfNodes);
+ cudaDeviceSynchronize();
+
+ getLastCudaError("kineticEnergyKernel execution failed");
real EKin = thrust::reduce(kineticEnergy.begin(), kineticEnergy.end(), c0o1, thrust::plus<real>());
uint numberOfFluidNodes = thrust::reduce(isFluid.begin(), isFluid.end(), 0, thrust::plus<uint>());
@@ -99,7 +85,7 @@ bool KineticEnergyAnalyzer::run(uint iter)
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
-__global__ void kineticEnergyKernel(real* vx, real* vy, real* vz, real* rho, uint* neighborX, uint* neighborY, uint* neighborZ, uint* neighborWSB, uint* geo, real* kineticEnergy, uint* isFluid, uint size_Mat)
+__global__ void kineticEnergyKernel(real* vx, real* vy, real* vz, real* rho, uint* neighborX, uint* neighborY, uint* neighborZ, uint* neighborWSB, uint* geo, real* kineticEnergy, uint* isFluid, unsigned long long numberOfLBnodes)
{
//////////////////////////////////////////////////////////////////////////
const uint x = threadIdx.x; // Globaler x-Index
@@ -115,7 +101,7 @@ __global__ void kineticEnergyKernel(real* vx, real* vy, real* vz, real* rho, uin
//if( index % 34 == 0 || index % 34 == 33 ) return;
- if( index >= size_Mat) return;
+ if( index >= (uint)numberOfLBnodes) return;
unsigned int BC;
BC = geo[index];
diff --git a/src/gpu/VirtualFluids_GPU/GPU/LBMKernel.cu b/src/gpu/VirtualFluids_GPU/GPU/LBMKernel.cu
index 63fc5be0ebe5d4a26d4662ee8c0dddbc3098247a..4faea21102b6a68dd9a0aa30e9cecc7eba6051b0 100644
--- a/src/gpu/VirtualFluids_GPU/GPU/LBMKernel.cu
+++ b/src/gpu/VirtualFluids_GPU/GPU/LBMKernel.cu
@@ -18,2176 +18,1644 @@
#include "Parameter/Parameter.h"
//////////////////////////////////////////////////////////////////////////
-void KernelCas27( unsigned int grid_nx,
- unsigned int grid_ny,
- unsigned int grid_nz,
- real s9,
- unsigned int* bcMatD,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- real* DD,
- int size_Mat,
- bool EvenOrOdd)
-{
- dim3 threads ( grid_nx, 1, 1 );
- dim3 grid ( grid_ny, grid_nz ); // Gitter fuer Kollision und Propagation
-
- LB_Kernel_Casc27<<< grid, threads >>>( s9,
- bcMatD,
- neighborX,
- neighborY,
- neighborZ,
- DD,
- size_Mat,
- EvenOrOdd);
- getLastCudaError("LB_Kernel_Casc27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void KernelCasSP27( unsigned int numberOfThreads,
- real s9,
- unsigned int* bcMatD,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- real* DD,
- int size_Mat,
- bool EvenOrOdd)
-{
- int Grid = (size_Mat / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 grid(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- LB_Kernel_Casc_SP_27<<< grid, threads >>>(s9,
- bcMatD,
- neighborX,
- neighborY,
- neighborZ,
- DD,
- size_Mat,
- EvenOrOdd);
- getLastCudaError("LB_Kernel_Casc_SP_27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void KernelCasSPMS27( unsigned int numberOfThreads,
- real s9,
- unsigned int* bcMatD,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- real* DD,
- int size_Mat,
- bool EvenOrOdd)
-{
- int Grid = (size_Mat / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 grid(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- LB_Kernel_Casc_SP_MS_27<<< grid, threads >>>(s9,
- bcMatD,
- neighborX,
- neighborY,
- neighborZ,
- DD,
- size_Mat,
- EvenOrOdd);
- getLastCudaError("LB_Kernel_Casc_SP_MS_27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void KernelCasSPMSOHM27( unsigned int numberOfThreads,
- real s9,
- unsigned int* bcMatD,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- real* DD,
- int size_Mat,
- bool EvenOrOdd)
-{
- int Grid = (size_Mat / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 grid(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- LB_Kernel_Casc_SP_MS_OHM_27<<< grid, threads >>>( s9,
- bcMatD,
- neighborX,
- neighborY,
- neighborZ,
- DD,
- size_Mat,
- EvenOrOdd);
- getLastCudaError("LB_Kernel_Casc_SP_MS_OHM_27 execution failed");
+void KernelCas27(
+ unsigned int grid_nx,
+ unsigned int grid_ny,
+ unsigned int grid_nz,
+ real s9,
+ unsigned int* bcMatD,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ real* DD,
+ unsigned long long numberOfLBnodes,
+ bool EvenOrOdd)
+{
+ dim3 threads ( grid_nx, 1, 1 );
+ dim3 grid ( grid_ny, grid_nz ); // Gitter fuer Kollision und Propagation
+
+ LB_Kernel_Casc27<<< grid, threads >>>(
+ s9,
+ bcMatD,
+ neighborX,
+ neighborY,
+ neighborZ,
+ DD,
+ numberOfLBnodes,
+ EvenOrOdd);
+ getLastCudaError("LB_Kernel_Casc27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void KernelCasSP27(
+ unsigned int numberOfThreads,
+ real s9,
+ unsigned int* bcMatD,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ real* DD,
+ unsigned long long numberOfLBnodes,
+ bool EvenOrOdd)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, numberOfLBnodes);
+
+ LB_Kernel_Casc_SP_27<<< grid.grid, grid.threads >>>(
+ s9,
+ bcMatD,
+ neighborX,
+ neighborY,
+ neighborZ,
+ DD,
+ numberOfLBnodes,
+ EvenOrOdd);
+ getLastCudaError("LB_Kernel_Casc_SP_27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void KernelCasSPMS27(
+ unsigned int numberOfThreads,
+ real s9,
+ unsigned int* bcMatD,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ real* DD,
+ unsigned long long numberOfLBnodes,
+ bool EvenOrOdd)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, numberOfLBnodes);
+
+ LB_Kernel_Casc_SP_MS_27<<< grid.grid, grid.threads >>>(
+ s9,
+ bcMatD,
+ neighborX,
+ neighborY,
+ neighborZ,
+ DD,
+ numberOfLBnodes,
+ EvenOrOdd);
+ getLastCudaError("LB_Kernel_Casc_SP_MS_27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void KernelCasSPMSOHM27(
+ unsigned int numberOfThreads,
+ real s9,
+ unsigned int* bcMatD,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ real* DD,
+ unsigned long long numberOfLBnodes,
+ bool EvenOrOdd)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, numberOfLBnodes);
+
+ LB_Kernel_Casc_SP_MS_OHM_27<<< grid.grid, grid.threads >>>(
+ s9,
+ bcMatD,
+ neighborX,
+ neighborY,
+ neighborZ,
+ DD,
+ numberOfLBnodes,
+ EvenOrOdd);
+ getLastCudaError("LB_Kernel_Casc_SP_MS_OHM_27 execution failed");
}
//////////////////////////////////////////////////////////////////////////
void KernelKumCompSRTSP27(
- unsigned int numberOfThreads,
- real omega,
- unsigned int* bcMatD,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- real* DDStart,
- int size_Mat,
- int level,
- real* forces,
- bool EvenOrOdd)
-{
- int Grid = (size_Mat / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 grid(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- LB_Kernel_Kum_New_Comp_SRT_SP_27 <<< grid, threads >>>(
- omega,
- bcMatD,
- neighborX,
- neighborY,
- neighborZ,
- DDStart,
- size_Mat,
- level,
- forces,
- EvenOrOdd);
- getLastCudaError("LB_Kernel_Kum_New_Comp_SRT_SP_27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void KernelKum1hSP27( unsigned int numberOfThreads,
- real omega,
- real deltaPhi,
- real angularVelocity,
- unsigned int* bcMatD,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- real* coordX,
- real* coordY,
- real* coordZ,
- real* DDStart,
- int size_Mat,
- bool EvenOrOdd)
-{
- int Grid = (size_Mat / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 grid(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- LB_Kernel_Kum_1h_SP_27<<< grid, threads >>>(omega,
- deltaPhi,
- angularVelocity,
- bcMatD,
- neighborX,
- neighborY,
- neighborZ,
- coordX,
- coordY,
- coordZ,
- DDStart,
- size_Mat,
- EvenOrOdd);
- getLastCudaError("LB_Kernel_Kum_New_SP_27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void KernelCascadeSP27( unsigned int numberOfThreads,
- real s9,
- unsigned int* bcMatD,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- real* DD,
- int size_Mat,
- bool EvenOrOdd)
-{
- int Grid = (size_Mat / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 grid(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- LB_Kernel_Cascade_SP_27<<< grid, threads >>>(s9,
- bcMatD,
- neighborX,
- neighborY,
- neighborZ,
- DD,
- size_Mat,
- EvenOrOdd);
- getLastCudaError("LB_Kernel_Cascade_SP_27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void KernelKumNewSP27( unsigned int numberOfThreads,
- real s9,
- unsigned int* bcMatD,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- real* DD,
- int size_Mat,
- bool EvenOrOdd)
-{
- int Grid = (size_Mat / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 grid(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- LB_Kernel_Kum_New_SP_27<<< grid, threads >>>(s9,
- bcMatD,
- neighborX,
- neighborY,
- neighborZ,
- DD,
- size_Mat,
- EvenOrOdd);
- getLastCudaError("LB_Kernel_Kum_New_SP_27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void KernelKumNewCompSP27(unsigned int numberOfThreads,
- real s9,
- unsigned int* bcMatD,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- real* DD,
- int size_Mat,
- int size_Array,
- int level,
- real* forces,
- bool EvenOrOdd)
-{
- //int Grid = size_Array / numberOfThreads;
- //dim3 grid(Grid, 1, 1);
- //dim3 threads(numberOfThreads, 1, 1 );
-
- int Grid = (size_Mat / numberOfThreads) + 1;
- int Grid1, Grid2;
- if (Grid > 512)
- {
- Grid1 = 512;
- Grid2 = (Grid / Grid1) + 1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 grid(Grid1, Grid2, 1);
- dim3 threads(numberOfThreads, 1, 1);
-
- //LB_Kernel_Kum_New_Comp_SP_27<<< grid, threads >>>( s9,
- // bcMatD,
- // neighborX,
- // neighborY,
- // neighborZ,
- // DD,
- // size_Mat,
- // level,
- // forces,
- // EvenOrOdd);
- //getLastCudaError("LB_Kernel_Kum_New_Comp_SP_27 execution failed");
-}
-
-//////////////////////////////////////////////////////////////////////////
-void CumulantOnePreconditionedErrorDiffusionChimCompSP27(unsigned int numberOfThreads,
- real s9,
- unsigned int* bcMatD,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- real* DD,
- int size_Mat,
- int size_Array,
- int level,
- real* forces,
- bool EvenOrOdd)
-{
- //int Grid = size_Array / numberOfThreads;
- //dim3 grid(Grid, 1, 1);
- //dim3 threads(numberOfThreads, 1, 1 );
-
- int Grid = (size_Mat / numberOfThreads) + 1;
- int Grid1, Grid2;
- if (Grid > 512)
- {
- Grid1 = 512;
- Grid2 = (Grid / Grid1) + 1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 grid(Grid1, Grid2, 1);
- dim3 threads(numberOfThreads, 1, 1);
-
- Cumulant_One_preconditioned_errorDiffusion_chim_Comp_SP_27 <<< grid, threads >>>( s9,
- bcMatD,
- neighborX,
- neighborY,
- neighborZ,
- DD,
- size_Mat,
- level,
- forces,
- EvenOrOdd);
- getLastCudaError("Cumulant_One_preconditioned_chim_Comp_SP_27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void CumulantOnePreconditionedChimCompSP27( unsigned int numberOfThreads,
- real s9,
- unsigned int* bcMatD,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- real* DD,
- int size_Mat,
- int size_Array,
- int level,
- real* forces,
- bool EvenOrOdd)
-{
- //int Grid = size_Array / numberOfThreads;
- //dim3 grid(Grid, 1, 1);
- //dim3 threads(numberOfThreads, 1, 1 );
-
- int Grid = (size_Mat / numberOfThreads) + 1;
- int Grid1, Grid2;
- if (Grid > 512)
- {
- Grid1 = 512;
- Grid2 = (Grid / Grid1) + 1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 grid(Grid1, Grid2, 1);
- dim3 threads(numberOfThreads, 1, 1);
-
- Cumulant_One_preconditioned_chim_Comp_SP_27 <<< grid, threads >>>( s9,
- bcMatD,
- neighborX,
- neighborY,
- neighborZ,
- DD,
- size_Mat,
- level,
- forces,
- EvenOrOdd);
- getLastCudaError("Cumulant_One_preconditioned_chim_Comp_SP_27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void CumulantOneChimCompSP27(unsigned int numberOfThreads,
- real s9,
- unsigned int* bcMatD,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- real* DD,
- int size_Mat,
- int size_Array,
- int level,
- real* forces,
- bool EvenOrOdd)
-{
- //int Grid = size_Array / numberOfThreads;
- //dim3 grid(Grid, 1, 1);
- //dim3 threads(numberOfThreads, 1, 1 );
-
- int Grid = (size_Mat / numberOfThreads) + 1;
- int Grid1, Grid2;
- if (Grid > 512)
- {
- Grid1 = 512;
- Grid2 = (Grid / Grid1) + 1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 grid(Grid1, Grid2, 1);
- dim3 threads(numberOfThreads, 1, 1);
-
- Cumulant_One_chim_Comp_SP_27 <<< grid, threads >>>( s9,
- bcMatD,
- neighborX,
- neighborY,
- neighborZ,
- DD,
- size_Mat,
- level,
- forces,
- EvenOrOdd);
- getLastCudaError("Cumulant_One_chim_Comp_SP_27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void KernelKumIsoTestSP27(unsigned int numberOfThreads,
- real s9,
- unsigned int* bcMatD,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- real* DD,
- real* dxxUx,
- real* dyyUy,
- real* dzzUz,
- int size_Mat,
- bool EvenOrOdd)
-{
- int Grid = (size_Mat / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 grid(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- LB_Kernel_Kum_IsoTest_SP_27<<< grid, threads >>>(s9,
- bcMatD,
- neighborX,
- neighborY,
- neighborZ,
- DD,
- dxxUx,
- dyyUy,
- dzzUz,
- size_Mat,
- EvenOrOdd);
- getLastCudaError("LB_Kernel_Kum_IsoTest_SP_27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void KernelKumCompSP27( unsigned int numberOfThreads,
- real s9,
- unsigned int* bcMatD,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- real* DD,
- int size_Mat,
- bool EvenOrOdd)
-{
- int Grid = (size_Mat / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 grid(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- LB_Kernel_Kum_Comp_SP_27<<< grid, threads >>>(s9,
- bcMatD,
- neighborX,
- neighborY,
- neighborZ,
- DD,
- size_Mat,
- EvenOrOdd);
- getLastCudaError("LB_Kernel_Kum_Comp_SP_27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void KernelPMCumOneCompSP27(unsigned int numberOfThreads,
- real omega,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- real* DD,
- int size_Mat,
- int level,
- real* forces,
- real porosity,
- real darcy,
- real forchheimer,
- unsigned int sizeOfPorousMedia,
- unsigned int* nodeIdsPorousMedia,
- bool EvenOrOdd)
-{
- int Grid = (size_Mat / numberOfThreads) + 1;
- int Grid1, Grid2;
- if (Grid > 512)
- {
- Grid1 = 512;
- Grid2 = (Grid / Grid1) + 1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 grid(Grid1, Grid2, 1);
- dim3 threads(numberOfThreads, 1, 1);
-
- LB_Kernel_PM_Cum_One_Comp_SP_27 <<< grid, threads >>>(omega,
- neighborX,
- neighborY,
- neighborZ,
- DD,
- size_Mat,
- level,
- forces,
- porosity,
- darcy,
- forchheimer,
- sizeOfPorousMedia,
- nodeIdsPorousMedia,
- EvenOrOdd);
- getLastCudaError("LB_Kernel_PM_Cum_One_Comp_SP_27 execution failed");
+ unsigned int numberOfThreads,
+ real omega,
+ unsigned int* bcMatD,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ real* DDStart,
+ unsigned long long numberOfLBnodes,
+ int level,
+ real* forces,
+ bool EvenOrOdd)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, numberOfLBnodes);
+
+ LB_Kernel_Kum_New_Comp_SRT_SP_27 <<< grid.grid, grid.threads >>>(
+ omega,
+ bcMatD,
+ neighborX,
+ neighborY,
+ neighborZ,
+ DDStart,
+ numberOfLBnodes,
+ level,
+ forces,
+ EvenOrOdd);
+ getLastCudaError("LB_Kernel_Kum_New_Comp_SRT_SP_27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void KernelKum1hSP27(
+ unsigned int numberOfThreads,
+ real omega,
+ real deltaPhi,
+ real angularVelocity,
+ unsigned int* bcMatD,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ real* coordX,
+ real* coordY,
+ real* coordZ,
+ real* DDStart,
+ unsigned long long numberOfLBnodes,
+ bool EvenOrOdd)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, numberOfLBnodes);
+
+ LB_Kernel_Kum_1h_SP_27<<< grid.grid, grid.threads >>>(
+ omega,
+ deltaPhi,
+ angularVelocity,
+ bcMatD,
+ neighborX,
+ neighborY,
+ neighborZ,
+ coordX,
+ coordY,
+ coordZ,
+ DDStart,
+ numberOfLBnodes,
+ EvenOrOdd);
+ getLastCudaError("LB_Kernel_Kum_1h_SP_27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void KernelCascadeSP27(
+ unsigned int numberOfThreads,
+ real s9,
+ unsigned int* bcMatD,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ real* DD,
+ unsigned long long numberOfLBnodes,
+ bool EvenOrOdd)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, numberOfLBnodes);
+
+ LB_Kernel_Cascade_SP_27<<< grid.grid, grid.threads >>>(
+ s9,
+ bcMatD,
+ neighborX,
+ neighborY,
+ neighborZ,
+ DD,
+ numberOfLBnodes,
+ EvenOrOdd);
+ getLastCudaError("LB_Kernel_Cascade_SP_27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void KernelKumNewSP27(
+ unsigned int numberOfThreads,
+ real s9,
+ unsigned int* bcMatD,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ real* DD,
+ unsigned long long numberOfLBnodes,
+ bool EvenOrOdd)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, numberOfLBnodes);
+
+
+ LB_Kernel_Kum_New_SP_27<<< grid.grid, grid.threads >>>(
+ s9,
+ bcMatD,
+ neighborX,
+ neighborY,
+ neighborZ,
+ DD,
+ numberOfLBnodes,
+ EvenOrOdd);
+ getLastCudaError("LB_Kernel_Kum_New_SP_27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void KernelKumNewCompSP27(
+ unsigned int numberOfThreads,
+ real s9,
+ unsigned int* bcMatD,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ real* DD,
+ unsigned long long numberOfLBnodes,
+ int size_Array,
+ int level,
+ real* forces,
+ bool EvenOrOdd)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, numberOfLBnodes);
+
+ //LB_Kernel_Kum_New_Comp_SP_27<<< grid.grid, grid.threads >>>( s9,
+ // bcMatD,
+ // neighborX,
+ // neighborY,
+ // neighborZ,
+ // DD,
+ // numberOfLBnodes,
+ // level,
+ // forces,
+ // EvenOrOdd);
+ //getLastCudaError("LB_Kernel_Kum_New_Comp_SP_27 execution failed");
+}
+
+//////////////////////////////////////////////////////////////////////////
+void CumulantOnePreconditionedErrorDiffusionChimCompSP27(
+ unsigned int numberOfThreads,
+ real s9,
+ unsigned int* bcMatD,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ real* DD,
+ unsigned long long numberOfLBnodes,
+ int size_Array,
+ int level,
+ real* forces,
+ bool EvenOrOdd)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, numberOfLBnodes);
+
+ Cumulant_One_preconditioned_errorDiffusion_chim_Comp_SP_27 <<< grid.grid, grid.threads >>>(
+ s9,
+ bcMatD,
+ neighborX,
+ neighborY,
+ neighborZ,
+ DD,
+ numberOfLBnodes,
+ level,
+ forces,
+ EvenOrOdd);
+ getLastCudaError("Cumulant_One_preconditioned_chim_Comp_SP_27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void CumulantOnePreconditionedChimCompSP27(
+ unsigned int numberOfThreads,
+ real s9,
+ unsigned int* bcMatD,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ real* DD,
+ unsigned long long numberOfLBnodes,
+ int size_Array,
+ int level,
+ real* forces,
+ bool EvenOrOdd)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, numberOfLBnodes);
+
+ Cumulant_One_preconditioned_chim_Comp_SP_27 <<< grid.grid, grid.threads >>>(
+ s9,
+ bcMatD,
+ neighborX,
+ neighborY,
+ neighborZ,
+ DD,
+ numberOfLBnodes,
+ level,
+ forces,
+ EvenOrOdd);
+ getLastCudaError("Cumulant_One_preconditioned_chim_Comp_SP_27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void CumulantOneChimCompSP27(
+ unsigned int numberOfThreads,
+ real s9,
+ unsigned int* bcMatD,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ real* DD,
+ unsigned long long numberOfLBnodes,
+ int size_Array,
+ int level,
+ real* forces,
+ bool EvenOrOdd)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, numberOfLBnodes);
+
+ Cumulant_One_chim_Comp_SP_27 <<< grid.grid, grid.threads >>>(
+ s9,
+ bcMatD,
+ neighborX,
+ neighborY,
+ neighborZ,
+ DD,
+ numberOfLBnodes,
+ level,
+ forces,
+ EvenOrOdd);
+ getLastCudaError("Cumulant_One_chim_Comp_SP_27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void KernelKumIsoTestSP27(
+ unsigned int numberOfThreads,
+ real s9,
+ unsigned int* bcMatD,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ real* DD,
+ real* dxxUx,
+ real* dyyUy,
+ real* dzzUz,
+ unsigned long long numberOfLBnodes,
+ bool EvenOrOdd)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, numberOfLBnodes);
+
+ LB_Kernel_Kum_IsoTest_SP_27<<< grid.grid, grid.threads >>>(
+ s9,
+ bcMatD,
+ neighborX,
+ neighborY,
+ neighborZ,
+ DD,
+ dxxUx,
+ dyyUy,
+ dzzUz,
+ numberOfLBnodes,
+ EvenOrOdd);
+ getLastCudaError("LB_Kernel_Kum_IsoTest_SP_27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void KernelKumCompSP27(
+ unsigned int numberOfThreads,
+ real s9,
+ unsigned int* bcMatD,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ real* DD,
+ unsigned long long numberOfLBnodes,
+ bool EvenOrOdd)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, numberOfLBnodes);
+
+
+ LB_Kernel_Kum_Comp_SP_27<<< grid.grid, grid.threads >>>(
+ s9,
+ bcMatD,
+ neighborX,
+ neighborY,
+ neighborZ,
+ DD,
+ numberOfLBnodes,
+ EvenOrOdd);
+ getLastCudaError("LB_Kernel_Kum_Comp_SP_27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void KernelPMCumOneCompSP27(
+ unsigned int numberOfThreads,
+ real omega,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ real* DD,
+ unsigned long long numberOfLBnodes,
+ int level,
+ real* forces,
+ real porosity,
+ real darcy,
+ real forchheimer,
+ unsigned int sizeOfPorousMedia,
+ unsigned int* nodeIdsPorousMedia,
+ bool EvenOrOdd)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, numberOfLBnodes);
+
+ LB_Kernel_PM_Cum_One_Comp_SP_27 <<< grid.grid, grid.threads >>>(
+ omega,
+ neighborX,
+ neighborY,
+ neighborZ,
+ DD,
+ numberOfLBnodes,
+ level,
+ forces,
+ porosity,
+ darcy,
+ forchheimer,
+ sizeOfPorousMedia,
+ nodeIdsPorousMedia,
+ EvenOrOdd);
+ getLastCudaError("LB_Kernel_PM_Cum_One_Comp_SP_27 execution failed");
}
//////////////////////////////////////////////////////////////////////////
void KernelWaleBySoniMalavCumAA2016CompSP27(
- unsigned int numberOfThreads,
- real s9,
- unsigned int* bcMatD,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int* neighborWSB,
- real* veloX,
- real* veloY,
- real* veloZ,
- real* DD,
- real* turbulentViscosity,
- int size_Mat,
- int size_Array,
- int level,
- real* forces,
- bool EvenOrOdd)
-{
- //int Grid = size_Array / numberOfThreads;
- //dim3 grid(Grid, 1, 1);
- //dim3 threads(numberOfThreads, 1, 1 );
-
- int Grid = (size_Mat / numberOfThreads) + 1;
- int Grid1, Grid2;
- if (Grid > 512)
- {
- Grid1 = 512;
- Grid2 = (Grid / Grid1) + 1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 grid(Grid1, Grid2, 1);
- dim3 threads(numberOfThreads, 1, 1);
-
- LB_Kernel_WaleBySoniMalav_Cum_AA2016_Comp_SP_27 << < grid, threads >> >(
- s9,
- bcMatD,
- neighborX,
- neighborY,
- neighborZ,
- neighborWSB,
- veloX,
- veloY,
- veloZ,
- DD,
- turbulentViscosity,
- size_Mat,
- level,
- forces,
- EvenOrOdd);
- getLastCudaError("LB_Kernel_WaleBySoniMalav_Cum_AA2016_Comp_SP_27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void KernelADincomp7( unsigned int numberOfThreads,
- real diffusivity,
- unsigned int* bcMatD,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- real* DD,
- real* DD7,
- int size_Mat,
- bool EvenOrOdd)
-{
- int Grid = (size_Mat / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 grid(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- LB_Kernel_AD_Incomp_7<<< grid, threads >>>( diffusivity,
- bcMatD,
- neighborX,
- neighborY,
- neighborZ,
- DD,
- DD7,
- size_Mat,
- EvenOrOdd);
- getLastCudaError("LB_Kernel_AD_Incomp_7 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void KernelADincomp27( unsigned int numberOfThreads,
- real diffusivity,
- unsigned int* bcMatD,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- real* DD,
- real* DD27,
- int size_Mat,
- bool EvenOrOdd)
-{
- int Grid = (size_Mat / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 grid(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- LB_Kernel_AD_Incomp_27<<< grid, threads >>>( diffusivity,
- bcMatD,
- neighborX,
- neighborY,
- neighborZ,
- DD,
- DD27,
- size_Mat,
- EvenOrOdd);
- getLastCudaError("LB_Kernel_AD_Incomp_27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void Init27( int myid,
- int numprocs,
- real u0,
- unsigned int* geoD,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- real* vParab,
- unsigned int size_Mat,
- unsigned int grid_nx,
- unsigned int grid_ny,
- unsigned int grid_nz,
- real* DD,
- int level,
- int maxlevel)
-{
- dim3 threads ( grid_nx, 1, 1 );
- dim3 grid ( grid_ny, grid_nz ); // Gitter fuer Kollision und Propagation
-
- LBInit27<<< grid, threads >>> ( myid,
- numprocs,
- u0,
- geoD,
- neighborX,
- neighborY,
- neighborZ,
- vParab,
- size_Mat,
- grid_nx,
- grid_ny,
- grid_nz,
- DD,
- level,
- maxlevel);
- getLastCudaError("LBInit27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void InitNonEqPartSP27( unsigned int numberOfThreads,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int* neighborWSB,
- unsigned int* geoD,
- real* rho,
- real* ux,
- real* uy,
- real* uz,
- unsigned int size_Mat,
- real* DD,
- real omega,
- bool EvenOrOdd)
-{
- int Grid = (size_Mat / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 grid(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- LBInitNonEqPartSP27<<< grid, threads >>>( neighborX,
- neighborY,
- neighborZ,
- neighborWSB,
- geoD,
- rho,
- ux,
- uy,
- uz,
- size_Mat,
- DD,
- omega,
- EvenOrOdd);
- getLastCudaError("LBInitNonEqPartSP27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void InitThS7( unsigned int numberOfThreads,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int* geoD,
- real* Conc,
- real* ux,
- real* uy,
- real* uz,
- unsigned int size_Mat,
- real* DD7,
- bool EvenOrOdd)
-{
- int Grid = (size_Mat / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 grid(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- InitAD7<<< grid, threads >>>( neighborX,
- neighborY,
- neighborZ,
- geoD,
- Conc,
- ux,
- uy,
- uz,
- size_Mat,
- DD7,
- EvenOrOdd);
- getLastCudaError("InitAD7 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void InitADDev27( unsigned int numberOfThreads,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int* geoD,
- real* Conc,
- real* ux,
- real* uy,
- real* uz,
- unsigned int size_Mat,
- real* DD27,
- bool EvenOrOdd)
-{
- int Grid = (size_Mat / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 grid(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- InitAD27<<< grid, threads >>>(neighborX,
- neighborY,
- neighborZ,
- geoD,
- Conc,
- ux,
- uy,
- uz,
- size_Mat,
- DD27,
- EvenOrOdd);
- getLastCudaError("InitAD27 execution failed");
+ unsigned int numberOfThreads,
+ real s9,
+ unsigned int* bcMatD,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned int* neighborWSB,
+ real* veloX,
+ real* veloY,
+ real* veloZ,
+ real* DD,
+ real* turbulentViscosity,
+ unsigned long long numberOfLBnodes,
+ int size_Array,
+ int level,
+ real* forces,
+ bool EvenOrOdd)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, numberOfLBnodes);
+
+ LB_Kernel_WaleBySoniMalav_Cum_AA2016_Comp_SP_27 <<< grid.grid, grid.threads >>>(
+ s9,
+ bcMatD,
+ neighborX,
+ neighborY,
+ neighborZ,
+ neighborWSB,
+ veloX,
+ veloY,
+ veloZ,
+ DD,
+ turbulentViscosity,
+ numberOfLBnodes,
+ level,
+ forces,
+ EvenOrOdd);
+ getLastCudaError("LB_Kernel_WaleBySoniMalav_Cum_AA2016_Comp_SP_27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void KernelADincomp7(
+ unsigned int numberOfThreads,
+ real diffusivity,
+ unsigned int* bcMatD,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ real* DD,
+ real* DD7,
+ unsigned long long numberOfLBnodes,
+ bool EvenOrOdd)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, numberOfLBnodes);
+
+ LB_Kernel_AD_Incomp_7<<< grid.grid, grid.threads >>>(
+ diffusivity,
+ bcMatD,
+ neighborX,
+ neighborY,
+ neighborZ,
+ DD,
+ DD7,
+ numberOfLBnodes,
+ EvenOrOdd);
+ getLastCudaError("LB_Kernel_AD_Incomp_7 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void KernelADincomp27(
+ unsigned int numberOfThreads,
+ real diffusivity,
+ unsigned int* bcMatD,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ real* DD,
+ real* DD27,
+ unsigned long long numberOfLBnodes,
+ bool EvenOrOdd)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, numberOfLBnodes);
+
+ LB_Kernel_AD_Incomp_27<<< grid.grid, grid.threads >>>(
+ diffusivity,
+ bcMatD,
+ neighborX,
+ neighborY,
+ neighborZ,
+ DD,
+ DD27,
+ numberOfLBnodes,
+ EvenOrOdd);
+ getLastCudaError("LB_Kernel_AD_Incomp_27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void Init27(
+ int myid,
+ int numprocs,
+ real u0,
+ unsigned int* geoD,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ real* vParab,
+ unsigned long long numberOfLBnodes,
+ unsigned int grid_nx,
+ unsigned int grid_ny,
+ unsigned int grid_nz,
+ real* DD,
+ int level,
+ int maxlevel)
+{
+ dim3 threads ( grid_nx, 1, 1 );
+ dim3 grid ( grid_ny, grid_nz );
+
+ LBInit27<<< grid, threads >>> (
+ myid,
+ numprocs,
+ u0,
+ geoD,
+ neighborX,
+ neighborY,
+ neighborZ,
+ vParab,
+ numberOfLBnodes,
+ grid_nx,
+ grid_ny,
+ grid_nz,
+ DD,
+ level,
+ maxlevel);
+ getLastCudaError("LBInit27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void InitNonEqPartSP27(
+ unsigned int numberOfThreads,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned int* neighborWSB,
+ unsigned int* geoD,
+ real* rho,
+ real* ux,
+ real* uy,
+ real* uz,
+ unsigned long long numberOfLBnodes,
+ real* DD,
+ real omega,
+ bool EvenOrOdd)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, numberOfLBnodes);
+
+ LBInitNonEqPartSP27<<< grid.grid, grid.threads >>>(
+ neighborX,
+ neighborY,
+ neighborZ,
+ neighborWSB,
+ geoD,
+ rho,
+ ux,
+ uy,
+ uz,
+ numberOfLBnodes,
+ DD,
+ omega,
+ EvenOrOdd);
+ getLastCudaError("LBInitNonEqPartSP27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void InitThS7(
+ unsigned int numberOfThreads,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned int* geoD,
+ real* Conc,
+ real* ux,
+ real* uy,
+ real* uz,
+ unsigned long long numberOfLBnodes,
+ real* DD7,
+ bool EvenOrOdd)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, numberOfLBnodes);
+
+ InitAD7<<< grid.grid, grid.threads >>>(
+ neighborX,
+ neighborY,
+ neighborZ,
+ geoD,
+ Conc,
+ ux,
+ uy,
+ uz,
+ numberOfLBnodes,
+ DD7,
+ EvenOrOdd);
+ getLastCudaError("InitAD7 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void InitADDev27(
+ unsigned int numberOfThreads,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned int* geoD,
+ real* Conc,
+ real* ux,
+ real* uy,
+ real* uz,
+ unsigned long long numberOfLBnodes,
+ real* DD27,
+ bool EvenOrOdd)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, numberOfLBnodes);
+
+ InitAD27<<< grid.grid, grid.threads >>>(
+ neighborX,
+ neighborY,
+ neighborZ,
+ geoD,
+ Conc,
+ ux,
+ uy,
+ uz,
+ numberOfLBnodes,
+ DD27,
+ EvenOrOdd);
+ getLastCudaError("InitAD27 execution failed");
}
//////////////////////////////////////////////////////////////////////////
void PostProcessorF3_2018Fehlberg(
- unsigned int numberOfThreads,
- real omega,
- unsigned int* bcMatD,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- real* rhoOut,
- real* vxOut,
- real* vyOut,
- real* vzOut,
- real* DDStart,
- real* G6,
- int size_Mat,
- int level,
- real* forces,
- bool EvenOrOdd)
-{
- int Grid = (size_Mat / numberOfThreads) + 1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid / Grid1) + 1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 grid(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1);
-
- LB_PostProcessor_F3_2018_Fehlberg <<< grid, threads >>> ( omega,
- bcMatD,
- neighborX,
- neighborY,
- neighborZ,
- rhoOut,
- vxOut,
- vyOut,
- vzOut,
- DDStart,
- G6,
- size_Mat,
- level,
- forces,
- EvenOrOdd);
- getLastCudaError("LB_PostProcessor_F3_2018_Fehlberg execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void CalcMac27( real* vxD,
- real* vyD,
- real* vzD,
- real* rhoD,
- unsigned int* geoD,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- unsigned int grid_nx,
- unsigned int grid_ny,
- unsigned int grid_nz,
- real* DD,
- bool isEvenTimestep)
+ unsigned int numberOfThreads,
+ real omega,
+ unsigned int* bcMatD,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ real* rhoOut,
+ real* vxOut,
+ real* vyOut,
+ real* vzOut,
+ real* DDStart,
+ real* G6,
+ unsigned long long numberOfLBnodes,
+ int level,
+ real* forces,
+ bool EvenOrOdd)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, numberOfLBnodes);
+
+ LB_PostProcessor_F3_2018_Fehlberg <<< grid.grid, grid.threads >>> (
+ omega,
+ bcMatD,
+ neighborX,
+ neighborY,
+ neighborZ,
+ rhoOut,
+ vxOut,
+ vyOut,
+ vzOut,
+ DDStart,
+ G6,
+ numberOfLBnodes,
+ level,
+ forces,
+ EvenOrOdd);
+ getLastCudaError("LB_PostProcessor_F3_2018_Fehlberg execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void CalcMac27(
+ real* vxD,
+ real* vyD,
+ real* vzD,
+ real* rhoD,
+ unsigned int* geoD,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ unsigned int grid_nx,
+ unsigned int grid_ny,
+ unsigned int grid_nz,
+ real* DD,
+ bool isEvenTimestep)
{
dim3 threads ( grid_nx, 1, 1 );
dim3 grid ( grid_ny, grid_nz );
- LBCalcMac27<<< grid, threads >>> ( vxD,
- vyD,
- vzD,
- rhoD,
- geoD,
- neighborX,
- neighborY,
- neighborZ,
- size_Mat,
- DD,
- isEvenTimestep);
- getLastCudaError("LBCalcMac27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void CalcMacSP27( real* vxD,
- real* vyD,
- real* vzD,
- real* rhoD,
- real* pressD,
- unsigned int* geoD,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- unsigned int numberOfThreads,
- real* DD,
- bool isEvenTimestep)
-{
- int Grid = (size_Mat / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 grid(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- LBCalcMacSP27<<< grid, threads >>> ( vxD,
- vyD,
- vzD,
- rhoD,
- pressD,
- geoD,
- neighborX,
- neighborY,
- neighborZ,
- size_Mat,
- DD,
- isEvenTimestep);
- getLastCudaError("LBCalcMacSP27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void CalcMacCompSP27( real* vxD,
- real* vyD,
- real* vzD,
- real* rhoD,
- real* pressD,
- unsigned int* geoD,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- unsigned int numberOfThreads,
- real* DD,
- bool isEvenTimestep)
-{
- int Grid = (size_Mat / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 grid(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- LBCalcMacCompSP27<<< grid, threads >>> ( vxD,
- vyD,
- vzD,
- rhoD,
- pressD,
- geoD,
- neighborX,
- neighborY,
- neighborZ,
- size_Mat,
- DD,
- isEvenTimestep);
- getLastCudaError("LBCalcMacSP27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void CalcMacThS7( real* Conc,
- unsigned int* geoD,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- unsigned int numberOfThreads,
- real* DD7,
- bool isEvenTimestep)
-{
- int Grid = (size_Mat / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 grid(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- CalcConc7<<< grid, threads >>> (Conc,
- geoD,
- neighborX,
- neighborY,
- neighborZ,
- size_Mat,
- DD7,
- isEvenTimestep);
- getLastCudaError("CalcConc7 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void PlaneConcThS7(real* Conc,
- int* kPC,
- unsigned int numberOfPointskPC,
- unsigned int* geoD,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- unsigned int numberOfThreads,
- real* DD7,
- bool isEvenTimestep)
-{
- int Grid = (numberOfPointskPC / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 grid(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- GetPlaneConc7<<< grid, threads >>> ( Conc,
- kPC,
- numberOfPointskPC,
- geoD,
- neighborX,
- neighborY,
- neighborZ,
- size_Mat,
- DD7,
- isEvenTimestep);
- getLastCudaError("GetPlaneConc7 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void PlaneConcThS27(real* Conc,
- int* kPC,
- unsigned int numberOfPointskPC,
- unsigned int* geoD,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- unsigned int numberOfThreads,
- real* DD27,
- bool isEvenTimestep)
-{
- int Grid = (numberOfPointskPC / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 grid(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- GetPlaneConc27<<< grid, threads >>> ( Conc,
- kPC,
- numberOfPointskPC,
- geoD,
- neighborX,
- neighborY,
- neighborZ,
- size_Mat,
- DD27,
- isEvenTimestep);
- getLastCudaError("GetPlaneConc27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void CalcConcentration27( unsigned int numberOfThreads,
- real* Conc,
- unsigned int* geoD,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- real* DD27,
- bool isEvenTimestep)
-{
- int Grid = (size_Mat / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 grid(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- CalcConc27<<< grid, threads >>> ( Conc,
- geoD,
- neighborX,
- neighborY,
- neighborZ,
- size_Mat,
- DD27,
- isEvenTimestep);
- getLastCudaError("CalcConc27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void CalcMedSP27( real* vxD,
- real* vyD,
- real* vzD,
- real* rhoD,
- real* pressD,
- unsigned int* geoD,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- unsigned int numberOfThreads,
- real* DD,
- bool isEvenTimestep)
-{
- int Grid = (size_Mat / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 grid(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- LBCalcMedSP27<<< grid, threads >>> ( vxD,
- vyD,
- vzD,
- rhoD,
- pressD,
- geoD,
- neighborX,
- neighborY,
- neighborZ,
- size_Mat,
- DD,
- isEvenTimestep);
- getLastCudaError("LBCalcMedSP27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void CalcMedCompSP27( real* vxD,
- real* vyD,
- real* vzD,
- real* rhoD,
- real* pressD,
- unsigned int* geoD,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- unsigned int numberOfThreads,
- real* DD,
- bool isEvenTimestep)
-{
- int Grid = (size_Mat / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 grid(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- LBCalcMedCompSP27<<< grid, threads >>> ( vxD,
- vyD,
- vzD,
- rhoD,
- pressD,
- geoD,
- neighborX,
- neighborY,
- neighborZ,
- size_Mat,
- DD,
- isEvenTimestep);
- getLastCudaError("LBCalcMedSP27 execution failed");
+ LBCalcMac27<<< grid, threads >>> (
+ vxD,
+ vyD,
+ vzD,
+ rhoD,
+ geoD,
+ neighborX,
+ neighborY,
+ neighborZ,
+ numberOfLBnodes,
+ DD,
+ isEvenTimestep);
+ getLastCudaError("LBCalcMac27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void CalcMacSP27(
+ real* vxD,
+ real* vyD,
+ real* vzD,
+ real* rhoD,
+ real* pressD,
+ unsigned int* geoD,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ unsigned int numberOfThreads,
+ real* DD,
+ bool isEvenTimestep)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, numberOfLBnodes);
+
+ LBCalcMacSP27<<< grid.grid, grid.threads >>> (
+ vxD,
+ vyD,
+ vzD,
+ rhoD,
+ pressD,
+ geoD,
+ neighborX,
+ neighborY,
+ neighborZ,
+ numberOfLBnodes,
+ DD,
+ isEvenTimestep);
+ getLastCudaError("LBCalcMacSP27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void CalcMacCompSP27(
+ real* vxD,
+ real* vyD,
+ real* vzD,
+ real* rhoD,
+ real* pressD,
+ unsigned int* geoD,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ unsigned int numberOfThreads,
+ real* DD,
+ bool isEvenTimestep)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, numberOfLBnodes);
+
+ LBCalcMacCompSP27<<< grid.grid, grid.threads >>> (
+ vxD,
+ vyD,
+ vzD,
+ rhoD,
+ pressD,
+ geoD,
+ neighborX,
+ neighborY,
+ neighborZ,
+ numberOfLBnodes,
+ DD,
+ isEvenTimestep);
+ getLastCudaError("LBCalcMacCompSP27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void CalcMacThS7(
+ real* Conc,
+ unsigned int* geoD,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ unsigned int numberOfThreads,
+ real* DD7,
+ bool isEvenTimestep)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, numberOfLBnodes);
+
+ CalcConc7<<< grid.grid, grid.threads >>> (
+ Conc,
+ geoD,
+ neighborX,
+ neighborY,
+ neighborZ,
+ numberOfLBnodes,
+ DD7,
+ isEvenTimestep);
+ getLastCudaError("CalcConc7 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void PlaneConcThS7(
+ real* Conc,
+ int* kPC,
+ unsigned int numberOfPointskPC,
+ unsigned int* geoD,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ unsigned int numberOfThreads,
+ real* DD7,
+ bool isEvenTimestep)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, numberOfPointskPC);
+
+ GetPlaneConc7<<< grid.grid, grid.threads >>> (
+ Conc,
+ kPC,
+ numberOfPointskPC,
+ geoD,
+ neighborX,
+ neighborY,
+ neighborZ,
+ numberOfLBnodes,
+ DD7,
+ isEvenTimestep);
+ getLastCudaError("GetPlaneConc7 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void PlaneConcThS27(
+ real* Conc,
+ int* kPC,
+ unsigned int numberOfPointskPC,
+ unsigned int* geoD,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ unsigned int numberOfThreads,
+ real* DD27,
+ bool isEvenTimestep)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, numberOfPointskPC);
+
+ GetPlaneConc27<<< grid.grid, grid.threads >>> (
+ Conc,
+ kPC,
+ numberOfPointskPC,
+ geoD,
+ neighborX,
+ neighborY,
+ neighborZ,
+ numberOfLBnodes,
+ DD27,
+ isEvenTimestep);
+ getLastCudaError("GetPlaneConc27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void CalcConcentration27(
+ unsigned int numberOfThreads,
+ real* Conc,
+ unsigned int* geoD,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ real* DD27,
+ bool isEvenTimestep)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, numberOfLBnodes);
+
+ CalcConc27<<< grid.grid, grid.threads >>> (
+ Conc,
+ geoD,
+ neighborX,
+ neighborY,
+ neighborZ,
+ numberOfLBnodes,
+ DD27,
+ isEvenTimestep);
+ getLastCudaError("CalcConc27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void CalcMedSP27(
+ real* vxD,
+ real* vyD,
+ real* vzD,
+ real* rhoD,
+ real* pressD,
+ unsigned int* geoD,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ unsigned int numberOfThreads,
+ real* DD,
+ bool isEvenTimestep)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, numberOfLBnodes);
+
+ LBCalcMedSP27<<< grid.grid, grid.threads >>> (
+ vxD,
+ vyD,
+ vzD,
+ rhoD,
+ pressD,
+ geoD,
+ neighborX,
+ neighborY,
+ neighborZ,
+ numberOfLBnodes,
+ DD,
+ isEvenTimestep);
+ getLastCudaError("LBCalcMedSP27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void CalcMedCompSP27(
+ real* vxD,
+ real* vyD,
+ real* vzD,
+ real* rhoD,
+ real* pressD,
+ unsigned int* geoD,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ unsigned int numberOfThreads,
+ real* DD,
+ bool isEvenTimestep)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, numberOfLBnodes);
+
+ LBCalcMedCompSP27<<< grid.grid, grid.threads >>> (
+ vxD,
+ vyD,
+ vzD,
+ rhoD,
+ pressD,
+ geoD,
+ neighborX,
+ neighborY,
+ neighborZ,
+ numberOfLBnodes,
+ DD,
+ isEvenTimestep);
+ getLastCudaError("LBCalcMedCompSP27 execution failed");
}
//////////////////////////////////////////////////////////////////////////
void CalcMedCompAD27(
- real* vxD,
- real* vyD,
- real* vzD,
- real* rhoD,
- real* pressD,
- real* concD,
- unsigned int* geoD,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- unsigned int numberOfThreads,
- real* DD,
- real* DD_AD,
- bool isEvenTimestep)
-{
- int Grid = (size_Mat / numberOfThreads) + 1;
- int Grid1, Grid2;
- if (Grid > 512)
- {
- Grid1 = 512;
- Grid2 = (Grid / Grid1) + 1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 grid(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1);
-
- LBCalcMedCompAD27 <<< grid, threads >>> (
- vxD,
- vyD,
- vzD,
- rhoD,
- pressD,
- concD,
- geoD,
- neighborX,
- neighborY,
- neighborZ,
- size_Mat,
- DD,
- DD_AD,
- isEvenTimestep);
- getLastCudaError("LBCalcMedAD27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void CalcMacMedSP27( real* vxD,
- real* vyD,
- real* vzD,
- real* rhoD,
- real* pressD,
- unsigned int* geoD,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int tdiff,
- unsigned int size_Mat,
- unsigned int numberOfThreads,
- bool isEvenTimestep)
-{
- int Grid = (size_Mat / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 grid(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- LBCalcMacMedSP27<<< grid, threads >>> ( vxD,
- vyD,
- vzD,
- rhoD,
- pressD,
- geoD,
- neighborX,
- neighborY,
- neighborZ,
- tdiff,
- size_Mat,
- isEvenTimestep);
- getLastCudaError("LBCalcMacMedSP27 execution failed");
+ real* vxD,
+ real* vyD,
+ real* vzD,
+ real* rhoD,
+ real* pressD,
+ real* concD,
+ unsigned int* geoD,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ unsigned int numberOfThreads,
+ real* DD,
+ real* DD_AD,
+ bool isEvenTimestep)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, numberOfLBnodes);
+
+ LBCalcMedCompAD27 <<< grid.grid, grid.threads >>> (
+ vxD,
+ vyD,
+ vzD,
+ rhoD,
+ pressD,
+ concD,
+ geoD,
+ neighborX,
+ neighborY,
+ neighborZ,
+ numberOfLBnodes,
+ DD,
+ DD_AD,
+ isEvenTimestep);
+ getLastCudaError("LBCalcMedCompAD27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void CalcMacMedSP27(
+ real* vxD,
+ real* vyD,
+ real* vzD,
+ real* rhoD,
+ real* pressD,
+ unsigned int* geoD,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned int tdiff,
+ unsigned long long numberOfLBnodes,
+ unsigned int numberOfThreads,
+ bool isEvenTimestep)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, numberOfLBnodes);
+
+ LBCalcMacMedSP27<<< grid.grid, grid.threads >>> (
+ vxD,
+ vyD,
+ vzD,
+ rhoD,
+ pressD,
+ geoD,
+ neighborX,
+ neighborY,
+ neighborZ,
+ tdiff,
+ numberOfLBnodes,
+ isEvenTimestep);
+ getLastCudaError("LBCalcMacMedSP27 execution failed");
}
//////////////////////////////////////////////////////////////////////////
void ResetMedianValuesSP27(
- real* vxD,
- real* vyD,
- real* vzD,
- real* rhoD,
- real* pressD,
- unsigned int size_Mat,
- unsigned int numberOfThreads,
- bool isEvenTimestep)
-{
- int Grid = (size_Mat / numberOfThreads) + 1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid / Grid1) + 1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 grid(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1);
-
- LBResetMedianValuesSP27 << < grid, threads >> > (
- vxD,
- vyD,
- vzD,
- rhoD,
- pressD,
- size_Mat,
- isEvenTimestep);
- getLastCudaError("LBResetMedianValuesSP27 execution failed");
+ real* vxD,
+ real* vyD,
+ real* vzD,
+ real* rhoD,
+ real* pressD,
+ unsigned long long numberOfLBnodes,
+ unsigned int numberOfThreads,
+ bool isEvenTimestep)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, numberOfLBnodes);
+
+ LBResetMedianValuesSP27 <<< grid.grid, grid.threads >>> (
+ vxD,
+ vyD,
+ vzD,
+ rhoD,
+ pressD,
+ numberOfLBnodes,
+ isEvenTimestep);
+ getLastCudaError("LBResetMedianValuesSP27 execution failed");
}
//////////////////////////////////////////////////////////////////////////
void ResetMedianValuesAD27(
- real* vxD,
- real* vyD,
- real* vzD,
- real* rhoD,
- real* pressD,
- real* concD,
- unsigned int size_Mat,
- unsigned int numberOfThreads,
- bool isEvenTimestep)
-{
- int Grid = (size_Mat / numberOfThreads) + 1;
- int Grid1, Grid2;
- if (Grid > 512)
- {
- Grid1 = 512;
- Grid2 = (Grid / Grid1) + 1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 grid(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1);
-
- LBResetMedianValuesAD27 << < grid, threads >> > (
- vxD,
- vyD,
- vzD,
- rhoD,
- pressD,
- concD,
- size_Mat,
- isEvenTimestep);
- getLastCudaError("LBResetMedianValuesAD27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void Calc2ndMomentsIncompSP27(real* kxyFromfcNEQ,
- real* kyzFromfcNEQ,
- real* kxzFromfcNEQ,
- real* kxxMyyFromfcNEQ,
- real* kxxMzzFromfcNEQ,
- unsigned int* geoD,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- unsigned int numberOfThreads,
- real* DD,
- bool isEvenTimestep)
-{
- int Grid = (size_Mat / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 grid(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- LBCalc2ndMomentsIncompSP27<<< grid, threads >>> ( kxyFromfcNEQ,
- kyzFromfcNEQ,
- kxzFromfcNEQ,
- kxxMyyFromfcNEQ,
- kxxMzzFromfcNEQ,
- geoD,
- neighborX,
- neighborY,
- neighborZ,
- size_Mat,
- DD,
- isEvenTimestep);
- getLastCudaError("LBCalc2ndMomentsIncompSP27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void Calc2ndMomentsCompSP27( real* kxyFromfcNEQ,
- real* kyzFromfcNEQ,
- real* kxzFromfcNEQ,
- real* kxxMyyFromfcNEQ,
- real* kxxMzzFromfcNEQ,
- unsigned int* geoD,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- unsigned int numberOfThreads,
- real* DD,
- bool isEvenTimestep)
-{
- int Grid = (size_Mat / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 grid(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- LBCalc2ndMomentsCompSP27<<< grid, threads >>> (kxyFromfcNEQ,
- kyzFromfcNEQ,
- kxzFromfcNEQ,
- kxxMyyFromfcNEQ,
- kxxMzzFromfcNEQ,
- geoD,
- neighborX,
- neighborY,
- neighborZ,
- size_Mat,
- DD,
- isEvenTimestep);
- getLastCudaError("LBCalc2ndMomentsCompSP27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void Calc3rdMomentsIncompSP27(real* CUMbbb,
- real* CUMabc,
- real* CUMbac,
- real* CUMbca,
- real* CUMcba,
- real* CUMacb,
- real* CUMcab,
- unsigned int* geoD,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- unsigned int numberOfThreads,
- real* DD,
- bool isEvenTimestep)
-{
- int Grid = (size_Mat / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 grid(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- LBCalc3rdMomentsIncompSP27<<< grid, threads >>> ( CUMbbb,
- CUMabc,
- CUMbac,
- CUMbca,
- CUMcba,
- CUMacb,
- CUMcab,
- geoD,
- neighborX,
- neighborY,
- neighborZ,
- DD,
- size_Mat,
- isEvenTimestep);
- getLastCudaError("LBCalc3rdMomentsIncompSP27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void Calc3rdMomentsCompSP27( real* CUMbbb,
- real* CUMabc,
- real* CUMbac,
- real* CUMbca,
- real* CUMcba,
- real* CUMacb,
- real* CUMcab,
- unsigned int* geoD,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- unsigned int numberOfThreads,
- real* DD,
- bool isEvenTimestep)
-{
- int Grid = (size_Mat / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 grid(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- LBCalc3rdMomentsCompSP27<<< grid, threads >>> (CUMbbb,
- CUMabc,
- CUMbac,
- CUMbca,
- CUMcba,
- CUMacb,
- CUMcab,
- geoD,
- neighborX,
- neighborY,
- neighborZ,
- DD,
- size_Mat,
- isEvenTimestep);
- getLastCudaError("LBCalc3rdMomentsCompSP27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void CalcHigherMomentsIncompSP27(real* CUMcbb,
- real* CUMbcb,
- real* CUMbbc,
- real* CUMcca,
- real* CUMcac,
- real* CUMacc,
- real* CUMbcc,
- real* CUMcbc,
- real* CUMccb,
- real* CUMccc,
- unsigned int* geoD,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- unsigned int numberOfThreads,
- real* DD,
- bool isEvenTimestep)
-{
- int Grid = (size_Mat / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 grid(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- LBCalcHigherMomentsIncompSP27<<< grid, threads >>> (CUMcbb,
- CUMbcb,
- CUMbbc,
- CUMcca,
- CUMcac,
- CUMacc,
- CUMbcc,
- CUMcbc,
- CUMccb,
- CUMccc,
- geoD,
- neighborX,
- neighborY,
- neighborZ,
- DD,
- size_Mat,
- isEvenTimestep);
- getLastCudaError("LBCalcHigherMomentsIncompSP27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void CalcHigherMomentsCompSP27( real* CUMcbb,
- real* CUMbcb,
- real* CUMbbc,
- real* CUMcca,
- real* CUMcac,
- real* CUMacc,
- real* CUMbcc,
- real* CUMcbc,
- real* CUMccb,
- real* CUMccc,
- unsigned int* geoD,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- unsigned int numberOfThreads,
- real* DD,
- bool isEvenTimestep)
-{
- int Grid = (size_Mat / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 grid(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- LBCalcHigherMomentsCompSP27<<< grid, threads >>> ( CUMcbb,
- CUMbcb,
- CUMbbc,
- CUMcca,
- CUMcac,
- CUMacc,
- CUMbcc,
- CUMcbc,
- CUMccb,
- CUMccc,
- geoD,
- neighborX,
- neighborY,
- neighborZ,
- DD,
- size_Mat,
- isEvenTimestep);
- getLastCudaError("LBCalcHigherMomentsCompSP27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void LBCalcMeasurePoints27(real* vxMP,
- real* vyMP,
- real* vzMP,
- real* rhoMP,
- unsigned int* kMP,
- unsigned int numberOfPointskMP,
- unsigned int MPClockCycle,
- unsigned int t,
- unsigned int* geoD,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- real* DD,
- unsigned int numberOfThreads,
- bool isEvenTimestep)
-{
- int Grid = (numberOfPointskMP / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 grid(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- LBCalcMeasurePoints<<< grid, threads >>> (vxMP,
- vyMP,
- vzMP,
- rhoMP,
- kMP,
- numberOfPointskMP,
- MPClockCycle,
- t,
- geoD,
- neighborX,
- neighborY,
- neighborZ,
- size_Mat,
- DD,
- isEvenTimestep);
- getLastCudaError("LBCalcMeasurePoints execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void BcPress27( int nx,
- int ny,
- int tz,
- unsigned int grid_nx,
- unsigned int grid_ny,
- unsigned int* bcMatD,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- real* DD,
- unsigned int size_Mat,
- bool isEvenTimestep)
-{
- dim3 threads ( grid_nx, 1, 1 );
- dim3 grid ( grid_ny, 1 );
-
- LB_BC_Press_East27<<< grid, threads >>> ( nx,
- ny,
- tz,
- bcMatD,
- neighborX,
- neighborY,
- neighborZ,
- DD,
- size_Mat,
- isEvenTimestep);
- getLastCudaError("LB_BC_Press_East27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void BcVel27(int nx,
- int ny,
- int nz,
- int itz,
- unsigned int grid_nx,
- unsigned int grid_ny,
- unsigned int* bcMatD,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- real* DD,
- unsigned int size_Mat,
- bool isEvenTimestep,
- real u0x,
- real om)
-{
- dim3 threads ( grid_nx, 1, 1 );
- dim3 grid ( grid_ny, 1 );
-
- LB_BC_Vel_West_27<<< grid, threads >>> ( nx,
- ny,
- nz,
- itz,
- bcMatD,
- neighborX,
- neighborY,
- neighborZ,
- DD,
- size_Mat,
- isEvenTimestep,
- u0x,
- grid_nx,
- grid_ny,
- om);
- getLastCudaError("LB_BC_Vel_West_27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void QADPressDev7( unsigned int numberOfThreads,
- real* DD,
- real* DD7,
- real* temp,
- real* velo,
- real diffusivity,
- int* k_Q,
- real* QQ,
- unsigned int numberOfBCnodes,
- real om1,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep)
-{
- int Grid = (numberOfBCnodes / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 gridQ(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- QADPress7<<< gridQ, threads >>>( DD,
- DD7,
- temp,
- velo,
- diffusivity,
- k_Q,
- QQ,
- numberOfBCnodes,
- om1,
- neighborX,
- neighborY,
- neighborZ,
- size_Mat,
- isEvenTimestep);
- getLastCudaError("QADPress7 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void QADPressDev27(unsigned int numberOfThreads,
- real* DD,
- real* DD27,
- real* temp,
- real* velo,
- real diffusivity,
- int* k_Q,
- real* QQ,
- unsigned int numberOfBCnodes,
- real om1,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep)
-{
- int Grid = (numberOfBCnodes / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 gridQ(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- QADPress27<<< gridQ, threads >>>( DD,
- DD27,
- temp,
- velo,
- diffusivity,
- k_Q,
- QQ,
- numberOfBCnodes,
- om1,
- neighborX,
- neighborY,
- neighborZ,
- size_Mat,
- isEvenTimestep);
- getLastCudaError("QADPress27 execution failed");
+ real* vxD,
+ real* vyD,
+ real* vzD,
+ real* rhoD,
+ real* pressD,
+ real* concD,
+ unsigned long long numberOfLBnodes,
+ unsigned int numberOfThreads,
+ bool isEvenTimestep)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, numberOfLBnodes);
+
+ LBResetMedianValuesAD27 <<< grid.grid, grid.threads >>> (
+ vxD,
+ vyD,
+ vzD,
+ rhoD,
+ pressD,
+ concD,
+ numberOfLBnodes,
+ isEvenTimestep);
+ getLastCudaError("LBResetMedianValuesAD27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void Calc2ndMomentsIncompSP27(
+ real* kxyFromfcNEQ,
+ real* kyzFromfcNEQ,
+ real* kxzFromfcNEQ,
+ real* kxxMyyFromfcNEQ,
+ real* kxxMzzFromfcNEQ,
+ unsigned int* geoD,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ unsigned int numberOfThreads,
+ real* DD,
+ bool isEvenTimestep)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, numberOfLBnodes);
+
+ LBCalc2ndMomentsIncompSP27<<< grid.grid, grid.threads >>> (
+ kxyFromfcNEQ,
+ kyzFromfcNEQ,
+ kxzFromfcNEQ,
+ kxxMyyFromfcNEQ,
+ kxxMzzFromfcNEQ,
+ geoD,
+ neighborX,
+ neighborY,
+ neighborZ,
+ numberOfLBnodes,
+ DD,
+ isEvenTimestep);
+ getLastCudaError("LBCalc2ndMomentsIncompSP27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void Calc2ndMomentsCompSP27(
+ real* kxyFromfcNEQ,
+ real* kyzFromfcNEQ,
+ real* kxzFromfcNEQ,
+ real* kxxMyyFromfcNEQ,
+ real* kxxMzzFromfcNEQ,
+ unsigned int* geoD,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ unsigned int numberOfThreads,
+ real* DD,
+ bool isEvenTimestep)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, numberOfLBnodes);
+
+ LBCalc2ndMomentsCompSP27<<< grid.grid, grid.threads >>> (
+ kxyFromfcNEQ,
+ kyzFromfcNEQ,
+ kxzFromfcNEQ,
+ kxxMyyFromfcNEQ,
+ kxxMzzFromfcNEQ,
+ geoD,
+ neighborX,
+ neighborY,
+ neighborZ,
+ numberOfLBnodes,
+ DD,
+ isEvenTimestep);
+ getLastCudaError("LBCalc2ndMomentsCompSP27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void Calc3rdMomentsIncompSP27(
+ real* CUMbbb,
+ real* CUMabc,
+ real* CUMbac,
+ real* CUMbca,
+ real* CUMcba,
+ real* CUMacb,
+ real* CUMcab,
+ unsigned int* geoD,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ unsigned int numberOfThreads,
+ real* DD,
+ bool isEvenTimestep)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, numberOfLBnodes);
+
+ LBCalc3rdMomentsIncompSP27<<< grid.grid, grid.threads >>> (
+ CUMbbb,
+ CUMabc,
+ CUMbac,
+ CUMbca,
+ CUMcba,
+ CUMacb,
+ CUMcab,
+ geoD,
+ neighborX,
+ neighborY,
+ neighborZ,
+ DD,
+ numberOfLBnodes,
+ isEvenTimestep);
+ getLastCudaError("LBCalc3rdMomentsIncompSP27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void Calc3rdMomentsCompSP27(
+ real* CUMbbb,
+ real* CUMabc,
+ real* CUMbac,
+ real* CUMbca,
+ real* CUMcba,
+ real* CUMacb,
+ real* CUMcab,
+ unsigned int* geoD,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ unsigned int numberOfThreads,
+ real* DD,
+ bool isEvenTimestep)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, numberOfLBnodes);
+
+ LBCalc3rdMomentsCompSP27<<< grid.grid, grid.threads >>> (
+ CUMbbb,
+ CUMabc,
+ CUMbac,
+ CUMbca,
+ CUMcba,
+ CUMacb,
+ CUMcab,
+ geoD,
+ neighborX,
+ neighborY,
+ neighborZ,
+ DD,
+ numberOfLBnodes,
+ isEvenTimestep);
+ getLastCudaError("LBCalc3rdMomentsCompSP27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void CalcHigherMomentsIncompSP27(
+ real* CUMcbb,
+ real* CUMbcb,
+ real* CUMbbc,
+ real* CUMcca,
+ real* CUMcac,
+ real* CUMacc,
+ real* CUMbcc,
+ real* CUMcbc,
+ real* CUMccb,
+ real* CUMccc,
+ unsigned int* geoD,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ unsigned int numberOfThreads,
+ real* DD,
+ bool isEvenTimestep)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, numberOfLBnodes);
+
+ LBCalcHigherMomentsIncompSP27<<< grid.grid, grid.threads >>> (
+ CUMcbb,
+ CUMbcb,
+ CUMbbc,
+ CUMcca,
+ CUMcac,
+ CUMacc,
+ CUMbcc,
+ CUMcbc,
+ CUMccb,
+ CUMccc,
+ geoD,
+ neighborX,
+ neighborY,
+ neighborZ,
+ DD,
+ numberOfLBnodes,
+ isEvenTimestep);
+ getLastCudaError("LBCalcHigherMomentsIncompSP27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void CalcHigherMomentsCompSP27(
+ real* CUMcbb,
+ real* CUMbcb,
+ real* CUMbbc,
+ real* CUMcca,
+ real* CUMcac,
+ real* CUMacc,
+ real* CUMbcc,
+ real* CUMcbc,
+ real* CUMccb,
+ real* CUMccc,
+ unsigned int* geoD,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ unsigned int numberOfThreads,
+ real* DD,
+ bool isEvenTimestep)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, numberOfLBnodes);
+
+ LBCalcHigherMomentsCompSP27<<< grid.grid, grid.threads >>> (
+ CUMcbb,
+ CUMbcb,
+ CUMbbc,
+ CUMcca,
+ CUMcac,
+ CUMacc,
+ CUMbcc,
+ CUMcbc,
+ CUMccb,
+ CUMccc,
+ geoD,
+ neighborX,
+ neighborY,
+ neighborZ,
+ DD,
+ numberOfLBnodes,
+ isEvenTimestep);
+ getLastCudaError("LBCalcHigherMomentsCompSP27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void LBCalcMeasurePoints27(
+ real* vxMP,
+ real* vyMP,
+ real* vzMP,
+ real* rhoMP,
+ unsigned int* kMP,
+ unsigned int numberOfPointskMP,
+ unsigned int MPClockCycle,
+ unsigned int t,
+ unsigned int* geoD,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ real* DD,
+ unsigned int numberOfThreads,
+ bool isEvenTimestep)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, numberOfPointskMP);
+
+ LBCalcMeasurePoints<<< grid.grid, grid.threads >>> (
+ vxMP,
+ vyMP,
+ vzMP,
+ rhoMP,
+ kMP,
+ numberOfPointskMP,
+ MPClockCycle,
+ t,
+ geoD,
+ neighborX,
+ neighborY,
+ neighborZ,
+ numberOfLBnodes,
+ DD,
+ isEvenTimestep);
+ getLastCudaError("LBCalcMeasurePoints execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void BcPress27(
+ int nx,
+ int ny,
+ int tz,
+ unsigned int grid_nx,
+ unsigned int grid_ny,
+ unsigned int* bcMatD,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ real* DD,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep)
+{
+ dim3 threads ( grid_nx, 1, 1 );
+ dim3 grid ( grid_ny, 1 );
+
+ LB_BC_Press_East27<<< grid, threads >>> (
+ nx,
+ ny,
+ tz,
+ bcMatD,
+ neighborX,
+ neighborY,
+ neighborZ,
+ DD,
+ numberOfLBnodes,
+ isEvenTimestep);
+ getLastCudaError("LB_BC_Press_East27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void BcVel27(
+ int nx,
+ int ny,
+ int nz,
+ int itz,
+ unsigned int grid_nx,
+ unsigned int grid_ny,
+ unsigned int* bcMatD,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ real* DD,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep,
+ real u0x,
+ real om)
+{
+ dim3 threads ( grid_nx, 1, 1 );
+ dim3 grid ( grid_ny, 1 );
+
+ LB_BC_Vel_West_27<<< grid, threads >>> (
+ nx,
+ ny,
+ nz,
+ itz,
+ bcMatD,
+ neighborX,
+ neighborY,
+ neighborZ,
+ DD,
+ numberOfLBnodes,
+ isEvenTimestep,
+ u0x,
+ grid_nx,
+ grid_ny,
+ om);
+ getLastCudaError("LB_BC_Vel_West_27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void QADPressDev7(
+ unsigned int numberOfThreads,
+ real* DD,
+ real* DD7,
+ real* temp,
+ real* velo,
+ real diffusivity,
+ int* k_Q,
+ real* QQ,
+ unsigned int numberOfBCnodes,
+ real om1,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, numberOfBCnodes);
+
+ QADPress7<<< grid.grid, grid.threads >>>(
+ DD,
+ DD7,
+ temp,
+ velo,
+ diffusivity,
+ k_Q,
+ QQ,
+ numberOfBCnodes,
+ om1,
+ neighborX,
+ neighborY,
+ neighborZ,
+ numberOfLBnodes,
+ isEvenTimestep);
+ getLastCudaError("QADPress7 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void QADPressDev27(
+ unsigned int numberOfThreads,
+ real* DD,
+ real* DD27,
+ real* temp,
+ real* velo,
+ real diffusivity,
+ int* k_Q,
+ real* QQ,
+ unsigned int numberOfBCnodes,
+ real om1,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, numberOfBCnodes);
+
+ QADPress27<<< grid.grid, grid.threads >>>(
+ DD,
+ DD27,
+ temp,
+ velo,
+ diffusivity,
+ k_Q,
+ QQ,
+ numberOfBCnodes,
+ om1,
+ neighborX,
+ neighborY,
+ neighborZ,
+ numberOfLBnodes,
+ isEvenTimestep);
+ getLastCudaError("QADPress27 execution failed");
}
//////////////////////////////////////////////////////////////////////////
void QADPressNEQNeighborDev27(
- unsigned int numberOfThreads,
- real* DD,
- real* DD27,
- int* k_Q,
- int* k_N,
- int numberOfBCnodes,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep
- )
-{
-
- int Grid = (numberOfBCnodes / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 gridQ(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- QADPressNEQNeighbor27<<< gridQ, threads >>>(
- DD,
- DD27,
- k_Q,
- k_N,
- numberOfBCnodes,
- neighborX,
- neighborY,
- neighborZ,
- size_Mat,
- isEvenTimestep
- );
- getLastCudaError("QADPressNEQNeighbor27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void QADVelDev7(unsigned int numberOfThreads,
- real* DD,
- real* DD7,
- real* temp,
- real* velo,
- real diffusivity,
- int* k_Q,
- real* QQ,
- unsigned int numberOfBCnodes,
- real om1,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep)
-{
- int Grid = (numberOfBCnodes / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 gridQ(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- QADVel7<<< gridQ, threads >>> (
- DD,
- DD7,
- temp,
- velo,
- diffusivity,
- k_Q,
- QQ,
- numberOfBCnodes,
- om1,
- neighborX,
- neighborY,
- neighborZ,
- size_Mat,
- isEvenTimestep);
- getLastCudaError("QADVel7 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void QADVelDev27( unsigned int numberOfThreads,
- real* DD,
- real* DD27,
- real* temp,
- real* velo,
- real diffusivity,
- int* k_Q,
- real* QQ,
- unsigned int numberOfBCnodes,
- real om1,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep)
-{
- int Grid = (numberOfBCnodes / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 gridQ(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- QADVel27<<< gridQ, threads >>> ( DD,
- DD27,
- temp,
- velo,
- diffusivity,
- k_Q,
- QQ,
- numberOfBCnodes,
- om1,
- neighborX,
- neighborY,
- neighborZ,
- size_Mat,
- isEvenTimestep);
- getLastCudaError("QADVel27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void QADDev7(unsigned int numberOfThreads,
- real* DD,
- real* DD7,
- real* temp,
- real diffusivity,
- int* k_Q,
- real* QQ,
- unsigned int numberOfBCnodes,
- real om1,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep)
-{
- int Grid = (numberOfBCnodes / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 gridQ(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- QAD7<<< gridQ, threads >>> ( DD,
- DD7,
- temp,
- diffusivity,
- k_Q,
- QQ,
- numberOfBCnodes,
- om1,
- neighborX,
- neighborY,
- neighborZ,
- size_Mat,
- isEvenTimestep);
- getLastCudaError("QAD7 execution failed");
+ unsigned int numberOfThreads,
+ real* DD,
+ real* DD27,
+ int* k_Q,
+ int* k_N,
+ int numberOfBCnodes,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, numberOfBCnodes);
+
+ QADPressNEQNeighbor27<<< grid.grid, grid.threads >>>(
+ DD,
+ DD27,
+ k_Q,
+ k_N,
+ numberOfBCnodes,
+ neighborX,
+ neighborY,
+ neighborZ,
+ numberOfLBnodes,
+ isEvenTimestep);
+ getLastCudaError("QADPressNEQNeighbor27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void QADVelDev7(
+ unsigned int numberOfThreads,
+ real* DD,
+ real* DD7,
+ real* temp,
+ real* velo,
+ real diffusivity,
+ int* k_Q,
+ real* QQ,
+ unsigned int numberOfBCnodes,
+ real om1,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, numberOfBCnodes);
+
+ QADVel7<<< grid.grid, grid.threads >>> (
+ DD,
+ DD7,
+ temp,
+ velo,
+ diffusivity,
+ k_Q,
+ QQ,
+ numberOfBCnodes,
+ om1,
+ neighborX,
+ neighborY,
+ neighborZ,
+ numberOfLBnodes,
+ isEvenTimestep);
+ getLastCudaError("QADVel7 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void QADVelDev27(
+ unsigned int numberOfThreads,
+ real* DD,
+ real* DD27,
+ real* temp,
+ real* velo,
+ real diffusivity,
+ int* k_Q,
+ real* QQ,
+ unsigned int numberOfBCnodes,
+ real om1,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, numberOfBCnodes);
+
+ QADVel27<<< grid.grid, grid.threads >>> (
+ DD,
+ DD27,
+ temp,
+ velo,
+ diffusivity,
+ k_Q,
+ QQ,
+ numberOfBCnodes,
+ om1,
+ neighborX,
+ neighborY,
+ neighborZ,
+ numberOfLBnodes,
+ isEvenTimestep);
+ getLastCudaError("QADVel27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void QADDev7(
+ unsigned int numberOfThreads,
+ real* DD,
+ real* DD7,
+ real* temp,
+ real diffusivity,
+ int* k_Q,
+ real* QQ,
+ unsigned int numberOfBCnodes,
+ real om1,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, numberOfBCnodes);
+
+ QAD7<<< grid.grid, grid.threads >>> (
+ DD,
+ DD7,
+ temp,
+ diffusivity,
+ k_Q,
+ QQ,
+ numberOfBCnodes,
+ om1,
+ neighborX,
+ neighborY,
+ neighborZ,
+ numberOfLBnodes,
+ isEvenTimestep);
+ getLastCudaError("QAD7 execution failed");
}
@@ -2202,1700 +1670,1430 @@ void FactorizedCentralMomentsAdvectionDiffusionDeviceKernel(
uint* neighborZ,
real* distributions,
real* distributionsAD,
- int size_Mat,
+ unsigned long long numberOfLBnodes,
real* forces,
bool isEvenTimestep)
{
- int Grid = (size_Mat / numberOfThreads) + 1;
- dim3 grid(Grid, 1, 1);
- dim3 threads(numberOfThreads, 1, 1);
+ int Grid = (numberOfLBnodes / numberOfThreads) + 1;
+ dim3 grid(Grid, 1, 1);
+ dim3 threads(numberOfThreads, 1, 1);
- Factorized_Central_Moments_Advection_Diffusion_Device_Kernel <<< grid, threads >>> (
- omegaDiffusivity,
- typeOfGridNode,
- neighborX,
- neighborY,
- neighborZ,
- distributions,
- distributionsAD,
- size_Mat,
- forces,
- isEvenTimestep);
- getLastCudaError("Factorized_Central_Moments_Advection_Diffusion_Device_Kernel execution failed");
+ Factorized_Central_Moments_Advection_Diffusion_Device_Kernel <<< grid, threads >>> (
+ omegaDiffusivity,
+ typeOfGridNode,
+ neighborX,
+ neighborY,
+ neighborZ,
+ distributions,
+ distributionsAD,
+ numberOfLBnodes,
+ forces,
+ isEvenTimestep);
+ getLastCudaError("Factorized_Central_Moments_Advection_Diffusion_Device_Kernel execution failed");
}
//////////////////////////////////////////////////////////////////////////
void ADSlipVelDevComp(
- uint numberOfThreads,
- real * normalX,
- real * normalY,
- real * normalZ,
- real * distributions,
- real * distributionsAD,
- int* QindexArray,
- real * Qarrays,
- uint numberOfBCnodes,
- real omegaDiffusivity,
- uint * neighborX,
- uint * neighborY,
- uint * neighborZ,
- uint size_Mat,
- bool isEvenTimestep)
-{
- int Grid = (numberOfBCnodes / numberOfThreads) + 1;
- dim3 gridQ(Grid, 1, 1);
- dim3 threads(numberOfThreads, 1, 1);
-
- AD_SlipVelDeviceComp << < gridQ, threads >> > (
- normalX,
- normalY,
- normalZ,
- distributions,
- distributionsAD,
- QindexArray,
- Qarrays,
- numberOfBCnodes,
- omegaDiffusivity,
- neighborX,
- neighborY,
- neighborZ,
- size_Mat,
- isEvenTimestep);
- getLastCudaError("AD_SlipVelDeviceComp execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-
-void QADDirichletDev27( unsigned int numberOfThreads,
- real* DD,
- real* DD27,
- real* temp,
- real diffusivity,
- int* k_Q,
- real* QQ,
- unsigned int numberOfBCnodes,
- real om1,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep)
-{
- int Grid = (numberOfBCnodes / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 gridQ(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- QADDirichlet27<<< gridQ, threads >>> (
- DD,
- DD27,
- temp,
- diffusivity,
- k_Q,
- QQ,
- numberOfBCnodes,
- om1,
- neighborX,
- neighborY,
- neighborZ,
- size_Mat,
- isEvenTimestep);
- getLastCudaError("QADDirichletDev27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void QADBBDev27(unsigned int numberOfThreads,
- real* DD,
- real* DD27,
- real* temp,
- real diffusivity,
- int* k_Q,
- real* QQ,
- unsigned int numberOfBCnodes,
- real om1,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep)
-{
- int Grid = (numberOfBCnodes / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 gridQ(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- QADBB27<<< gridQ, threads >>> ( DD,
- DD27,
- temp,
- diffusivity,
- k_Q,
- QQ,
- numberOfBCnodes,
- om1,
- neighborX,
- neighborY,
- neighborZ,
- size_Mat,
- isEvenTimestep);
- getLastCudaError("QADBB27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void QNoSlipADincompDev7(unsigned int numberOfThreads,
- real* DD,
- real* DD7,
- real* temp,
- real diffusivity,
- int* k_Q,
- real* QQ,
- unsigned int numberOfBCnodes,
- real om1,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep)
-{
- int Grid = (numberOfBCnodes / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 gridQ(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- QNoSlipADincomp7<<< gridQ, threads >>> (
- DD,
- DD7,
- temp,
- diffusivity,
- k_Q,
- QQ,
- numberOfBCnodes,
- om1,
- neighborX,
- neighborY,
- neighborZ,
- size_Mat,
- isEvenTimestep);
- getLastCudaError("QNoSlipADincomp7 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void QNoSlipADincompDev27( unsigned int numberOfThreads,
- real* DD,
- real* DD27,
- real* temp,
- real diffusivity,
- int* k_Q,
- real* QQ,
- unsigned int numberOfBCnodes,
- real om1,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep)
-{
- int Grid = (numberOfBCnodes / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 gridQ(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- QNoSlipADincomp27<<< gridQ, threads >>> (
- DD,
- DD27,
- temp,
- diffusivity,
- k_Q,
- QQ,
- numberOfBCnodes,
- om1,
- neighborX,
- neighborY,
- neighborZ,
- size_Mat,
- isEvenTimestep);
- getLastCudaError("QNoSlipADincomp27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void QADVeloIncompDev7( unsigned int numberOfThreads,
- real* DD,
- real* DD7,
- real* temp,
- real* velo,
- real diffusivity,
- int* k_Q,
- real* QQ,
- unsigned int numberOfBCnodes,
- real om1,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep)
-{
- int Grid = (numberOfBCnodes / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 gridQ(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- QADVeloIncomp7<<< gridQ, threads >>> (
- DD,
- DD7,
- temp,
- velo,
- diffusivity,
- k_Q,
- QQ,
- numberOfBCnodes,
- om1,
- neighborX,
- neighborY,
- neighborZ,
- size_Mat,
- isEvenTimestep);
- getLastCudaError("QADVeloIncomp7 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void QADVeloIncompDev27( unsigned int numberOfThreads,
- real* DD,
- real* DD27,
- real* temp,
- real* velo,
- real diffusivity,
- int* k_Q,
- real* QQ,
- unsigned int numberOfBCnodes,
- real om1,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep)
-{
- int Grid = (numberOfBCnodes / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 gridQ(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- QADVeloIncomp27<<< gridQ, threads >>> (
- DD,
- DD27,
- temp,
- velo,
- diffusivity,
- k_Q,
- QQ,
- numberOfBCnodes,
- om1,
- neighborX,
- neighborY,
- neighborZ,
- size_Mat,
- isEvenTimestep);
- getLastCudaError("QADVeloIncomp27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void QADPressIncompDev7( unsigned int numberOfThreads,
- real* DD,
- real* DD7,
- real* temp,
- real* velo,
- real diffusivity,
- int* k_Q,
- real* QQ,
- unsigned int numberOfBCnodes,
- real om1,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep)
-{
- int Grid = (numberOfBCnodes / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 gridQ(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- QADPressIncomp7<<< gridQ, threads >>>(
- DD,
- DD7,
- temp,
- velo,
- diffusivity,
- k_Q,
- QQ,
- numberOfBCnodes,
- om1,
- neighborX,
- neighborY,
- neighborZ,
- size_Mat,
- isEvenTimestep);
- getLastCudaError("QADPressIncomp7 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void QADPressIncompDev27( unsigned int numberOfThreads,
- real* DD,
- real* DD27,
- real* temp,
- real* velo,
- real diffusivity,
- int* k_Q,
- real* QQ,
- unsigned int numberOfBCnodes,
- real om1,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep)
-{
- int Grid = (numberOfBCnodes / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 gridQ(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- QADPressIncomp27<<< gridQ, threads >>>(
- DD,
- DD27,
- temp,
- velo,
- diffusivity,
- k_Q,
- QQ,
- numberOfBCnodes,
- om1,
- neighborX,
- neighborY,
- neighborZ,
- size_Mat,
- isEvenTimestep);
- getLastCudaError("QADPressIncomp27 execution failed");
+ uint numberOfThreads,
+ real * normalX,
+ real * normalY,
+ real * normalZ,
+ real * distributions,
+ real * distributionsAD,
+ int* QindexArray,
+ real * Qarrays,
+ uint numberOfBCnodes,
+ real omegaDiffusivity,
+ uint * neighborX,
+ uint * neighborY,
+ uint * neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, numberOfBCnodes);
+
+ AD_SlipVelDeviceComp <<< grid.grid, grid.threads >>> (
+ normalX,
+ normalY,
+ normalZ,
+ distributions,
+ distributionsAD,
+ QindexArray,
+ Qarrays,
+ numberOfBCnodes,
+ omegaDiffusivity,
+ neighborX,
+ neighborY,
+ neighborZ,
+ numberOfLBnodes,
+ isEvenTimestep);
+ getLastCudaError("AD_SlipVelDeviceComp execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+
+void QADDirichletDev27(
+ unsigned int numberOfThreads,
+ real* DD,
+ real* DD27,
+ real* temp,
+ real diffusivity,
+ int* k_Q,
+ real* QQ,
+ unsigned int numberOfBCnodes,
+ real om1,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, numberOfBCnodes);
+
+ QADDirichlet27<<< grid.grid, grid.threads >>> (
+ DD,
+ DD27,
+ temp,
+ diffusivity,
+ k_Q,
+ QQ,
+ numberOfBCnodes,
+ om1,
+ neighborX,
+ neighborY,
+ neighborZ,
+ numberOfLBnodes,
+ isEvenTimestep);
+ getLastCudaError("QADDirichletDev27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void QADBBDev27(
+ unsigned int numberOfThreads,
+ real* DD,
+ real* DD27,
+ real* temp,
+ real diffusivity,
+ int* k_Q,
+ real* QQ,
+ unsigned int numberOfBCnodes,
+ real om1,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, numberOfBCnodes);
+
+ QADBB27<<< grid.grid, grid.threads >>> (
+ DD,
+ DD27,
+ temp,
+ diffusivity,
+ k_Q,
+ QQ,
+ numberOfBCnodes,
+ om1,
+ neighborX,
+ neighborY,
+ neighborZ,
+ numberOfLBnodes,
+ isEvenTimestep);
+ getLastCudaError("QADBB27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void QNoSlipADincompDev7(
+ unsigned int numberOfThreads,
+ real* DD,
+ real* DD7,
+ real* temp,
+ real diffusivity,
+ int* k_Q,
+ real* QQ,
+ unsigned int numberOfBCnodes,
+ real om1,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, numberOfBCnodes);
+
+ QNoSlipADincomp7<<< grid.grid, grid.threads >>> (
+ DD,
+ DD7,
+ temp,
+ diffusivity,
+ k_Q,
+ QQ,
+ numberOfBCnodes,
+ om1,
+ neighborX,
+ neighborY,
+ neighborZ,
+ numberOfLBnodes,
+ isEvenTimestep);
+ getLastCudaError("QNoSlipADincomp7 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void QNoSlipADincompDev27(
+ unsigned int numberOfThreads,
+ real* DD,
+ real* DD27,
+ real* temp,
+ real diffusivity,
+ int* k_Q,
+ real* QQ,
+ unsigned int numberOfBCnodes,
+ real om1,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, numberOfBCnodes);
+
+ QNoSlipADincomp27<<< grid.grid, grid.threads >>> (
+ DD,
+ DD27,
+ temp,
+ diffusivity,
+ k_Q,
+ QQ,
+ numberOfBCnodes,
+ om1,
+ neighborX,
+ neighborY,
+ neighborZ,
+ numberOfLBnodes,
+ isEvenTimestep);
+ getLastCudaError("QNoSlipADincomp27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void QADVeloIncompDev7(
+ unsigned int numberOfThreads,
+ real* DD,
+ real* DD7,
+ real* temp,
+ real* velo,
+ real diffusivity,
+ int* k_Q,
+ real* QQ,
+ unsigned int numberOfBCnodes,
+ real om1,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, numberOfBCnodes);
+
+ QADVeloIncomp7<<< grid.grid, grid.threads >>> (
+ DD,
+ DD7,
+ temp,
+ velo,
+ diffusivity,
+ k_Q,
+ QQ,
+ numberOfBCnodes,
+ om1,
+ neighborX,
+ neighborY,
+ neighborZ,
+ numberOfLBnodes,
+ isEvenTimestep);
+ getLastCudaError("QADVeloIncomp7 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void QADVeloIncompDev27(
+ unsigned int numberOfThreads,
+ real* DD,
+ real* DD27,
+ real* temp,
+ real* velo,
+ real diffusivity,
+ int* k_Q,
+ real* QQ,
+ unsigned int numberOfBCnodes,
+ real om1,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, numberOfBCnodes);
+
+ QADVeloIncomp27<<< grid.grid, grid.threads >>> (
+ DD,
+ DD27,
+ temp,
+ velo,
+ diffusivity,
+ k_Q,
+ QQ,
+ numberOfBCnodes,
+ om1,
+ neighborX,
+ neighborY,
+ neighborZ,
+ numberOfLBnodes,
+ isEvenTimestep);
+ getLastCudaError("QADVeloIncomp27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void QADPressIncompDev7(
+ unsigned int numberOfThreads,
+ real* DD,
+ real* DD7,
+ real* temp,
+ real* velo,
+ real diffusivity,
+ int* k_Q,
+ real* QQ,
+ unsigned int numberOfBCnodes,
+ real om1,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, numberOfBCnodes);
+
+ QADPressIncomp7<<< grid.grid, grid.threads >>>(
+ DD,
+ DD7,
+ temp,
+ velo,
+ diffusivity,
+ k_Q,
+ QQ,
+ numberOfBCnodes,
+ om1,
+ neighborX,
+ neighborY,
+ neighborZ,
+ numberOfLBnodes,
+ isEvenTimestep);
+ getLastCudaError("QADPressIncomp7 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void QADPressIncompDev27(
+ unsigned int numberOfThreads,
+ real* DD,
+ real* DD27,
+ real* temp,
+ real* velo,
+ real diffusivity,
+ int* k_Q,
+ real* QQ,
+ unsigned int numberOfBCnodes,
+ real om1,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, numberOfBCnodes);
+
+ QADPressIncomp27<<< grid.grid, grid.threads >>>(
+ DD,
+ DD27,
+ temp,
+ velo,
+ diffusivity,
+ k_Q,
+ QQ,
+ numberOfBCnodes,
+ om1,
+ neighborX,
+ neighborY,
+ neighborZ,
+ numberOfLBnodes,
+ isEvenTimestep);
+ getLastCudaError("QADPressIncomp27 execution failed");
}
//////////////////////////////////////////////////////////////////////////
void QDev27(LBMSimulationParameter* parameterDevice, QforBoundaryConditions* boundaryCondition)
{
- dim3 grid = vf::cuda::getCudaGrid( parameterDevice->numberofthreads, boundaryCondition->numberOfBCnodes);
- dim3 threads(parameterDevice->numberofthreads, 1, 1 );
-
- QDevice27<<< grid, threads >>> (
- parameterDevice->distributions.f[0],
- boundaryCondition->k,
- boundaryCondition->q27[0],
- boundaryCondition->numberOfBCnodes,
- parameterDevice->omega,
- parameterDevice->neighborX,
- parameterDevice->neighborY,
- parameterDevice->neighborZ,
- parameterDevice->numberOfNodes,
- parameterDevice->isEvenTimestep);
+ dim3 grid = vf::cuda::getCudaGrid( parameterDevice->numberofthreads, boundaryCondition->numberOfBCnodes);
+ dim3 threads(parameterDevice->numberofthreads, 1, 1 );
+ QDevice27<<< grid, threads >>> (
+ parameterDevice->distributions.f[0],
+ boundaryCondition->k,
+ boundaryCondition->q27[0],
+ boundaryCondition->numberOfBCnodes,
+ parameterDevice->omega,
+ parameterDevice->neighborX,
+ parameterDevice->neighborY,
+ parameterDevice->neighborZ,
+ parameterDevice->numberOfNodes,
+ parameterDevice->isEvenTimestep);
getLastCudaError("QDevice27 execution failed");
}
//////////////////////////////////////////////////////////////////////////
void QDevComp27(LBMSimulationParameter* parameterDevice, QforBoundaryConditions* boundaryCondition)
{
- dim3 grid = vf::cuda::getCudaGrid( parameterDevice->numberofthreads, boundaryCondition->numberOfBCnodes);
- dim3 threads(parameterDevice->numberofthreads, 1, 1 );
-
- QDeviceComp27<<< grid, threads >>> (
- parameterDevice->distributions.f[0],
- boundaryCondition->k,
- boundaryCondition->q27[0],
- boundaryCondition->numberOfBCnodes,
- parameterDevice->omega,
- parameterDevice->neighborX,
- parameterDevice->neighborY,
- parameterDevice->neighborZ,
- parameterDevice->numberOfNodes,
- parameterDevice->isEvenTimestep);
- getLastCudaError("QDeviceComp27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void QDevCompThinWalls27(unsigned int numberOfThreads,
- real* DD,
- int* k_Q,
- real* QQ,
- unsigned int numberOfBCnodes,
- real om1,
- unsigned int* geom,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int* neighborWSB,
- unsigned int size_Mat,
- bool isEvenTimestep)
-{
- int Grid = (numberOfBCnodes / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 gridQ(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- QDeviceCompThinWallsPartOne27 <<< gridQ, threads >>> (DD,
- k_Q,
- QQ,
- numberOfBCnodes,
- om1,
- neighborX,
- neighborY,
- neighborZ,
- size_Mat,
- isEvenTimestep);
- getLastCudaError("QDeviceCompThinWallsPartOne27 execution failed");
-
- QThinWallsPartTwo27 <<< gridQ, threads >>> ( DD,
- k_Q,
- QQ,
- numberOfBCnodes,
- geom,
- neighborX,
- neighborY,
- neighborZ,
- neighborWSB,
- size_Mat,
- isEvenTimestep);
- getLastCudaError("QThinWallsPartTwo27 execution failed");
+ dim3 grid = vf::cuda::getCudaGrid( parameterDevice->numberofthreads, boundaryCondition->numberOfBCnodes);
+ dim3 threads(parameterDevice->numberofthreads, 1, 1 );
+ QDeviceComp27<<< grid, threads >>> (
+ parameterDevice->distributions.f[0],
+ boundaryCondition->k,
+ boundaryCondition->q27[0],
+ boundaryCondition->numberOfBCnodes,
+ parameterDevice->omega,
+ parameterDevice->neighborX,
+ parameterDevice->neighborY,
+ parameterDevice->neighborZ,
+ parameterDevice->numberOfNodes,
+ parameterDevice->isEvenTimestep);
+ getLastCudaError("QDeviceComp27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void QDevCompThinWalls27(
+ unsigned int numberOfThreads,
+ real* DD,
+ int* k_Q,
+ real* QQ,
+ unsigned int numberOfBCnodes,
+ real om1,
+ unsigned int* geom,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned int* neighborWSB,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, numberOfBCnodes);
+
+ QDeviceCompThinWallsPartOne27 <<< grid.grid, grid.threads >>> (
+ DD,
+ k_Q,
+ QQ,
+ numberOfBCnodes,
+ om1,
+ neighborX,
+ neighborY,
+ neighborZ,
+ numberOfLBnodes,
+ isEvenTimestep);
+ getLastCudaError("QDeviceCompThinWallsPartOne27 execution failed");
+
+ QThinWallsPartTwo27 <<< grid.grid, grid.threads >>> (
+ DD,
+ k_Q,
+ QQ,
+ numberOfBCnodes,
+ geom,
+ neighborX,
+ neighborY,
+ neighborZ,
+ neighborWSB,
+ numberOfLBnodes,
+ isEvenTimestep);
+ getLastCudaError("QThinWallsPartTwo27 execution failed");
}
//////////////////////////////////////////////////////////////////////////
void QDev3rdMomentsComp27(LBMSimulationParameter* parameterDevice, QforBoundaryConditions* boundaryCondition)
{
- dim3 grid = vf::cuda::getCudaGrid( parameterDevice->numberofthreads, boundaryCondition->numberOfBCnodes);
- dim3 threads(parameterDevice->numberofthreads, 1, 1);
-
- QDevice3rdMomentsComp27<<< grid, threads >>> (
- parameterDevice->distributions.f[0],
- boundaryCondition->k,
- boundaryCondition->q27[0],
- boundaryCondition->numberOfBCnodes,
- parameterDevice->omega,
- parameterDevice->neighborX,
- parameterDevice->neighborY,
- parameterDevice->neighborZ,
- parameterDevice->numberOfNodes,
- parameterDevice->isEvenTimestep);
+ dim3 grid = vf::cuda::getCudaGrid( parameterDevice->numberofthreads, boundaryCondition->numberOfBCnodes);
+ dim3 threads(parameterDevice->numberofthreads, 1, 1);
+
+ QDevice3rdMomentsComp27<<< grid, threads >>> (
+ parameterDevice->distributions.f[0],
+ boundaryCondition->k,
+ boundaryCondition->q27[0],
+ boundaryCondition->numberOfBCnodes,
+ parameterDevice->omega,
+ parameterDevice->neighborX,
+ parameterDevice->neighborY,
+ parameterDevice->neighborZ,
+ parameterDevice->numberOfNodes,
+ parameterDevice->isEvenTimestep);
getLastCudaError("QDevice3rdMomentsComp27 execution failed");
}
//////////////////////////////////////////////////////////////////////////
-void QDevIncompHighNu27( unsigned int numberOfThreads,
- real* DD,
- int* k_Q,
- real* QQ,
- unsigned int numberOfBCnodes,
- real om1,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep)
-{
- int Grid = (numberOfBCnodes / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 gridQ(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- QDeviceIncompHighNu27<<< gridQ, threads >>> (
- DD,
- k_Q,
- QQ,
- numberOfBCnodes,
- om1,
- neighborX,
- neighborY,
- neighborZ,
- size_Mat,
- isEvenTimestep);
- getLastCudaError("QDeviceIncompHighNu27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void QDevCompHighNu27( unsigned int numberOfThreads,
- real* DD,
- int* k_Q,
- real* QQ,
- unsigned int numberOfBCnodes,
- real om1,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep)
-{
- int Grid = (numberOfBCnodes / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 gridQ(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- QDeviceCompHighNu27<<< gridQ, threads >>> (
- DD,
- k_Q,
- QQ,
- numberOfBCnodes,
- om1,
- neighborX,
- neighborY,
- neighborZ,
- size_Mat,
- isEvenTimestep);
- getLastCudaError("QDevice27 execution failed");
+void QDevIncompHighNu27(
+ unsigned int numberOfThreads,
+ real* DD,
+ int* k_Q,
+ real* QQ,
+ unsigned int numberOfBCnodes,
+ real om1,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, numberOfBCnodes);
+
+ QDeviceIncompHighNu27<<< grid.grid, grid.threads >>> (
+ DD,
+ k_Q,
+ QQ,
+ numberOfBCnodes,
+ om1,
+ neighborX,
+ neighborY,
+ neighborZ,
+ numberOfLBnodes,
+ isEvenTimestep);
+ getLastCudaError("QDeviceIncompHighNu27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void QDevCompHighNu27(
+ unsigned int numberOfThreads,
+ real* DD,
+ int* k_Q,
+ real* QQ,
+ unsigned int numberOfBCnodes,
+ real om1,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, numberOfBCnodes);
+
+ QDeviceCompHighNu27<<< grid.grid, grid.threads >>> (
+ DD,
+ k_Q,
+ QQ,
+ numberOfBCnodes,
+ om1,
+ neighborX,
+ neighborY,
+ neighborZ,
+ numberOfLBnodes,
+ isEvenTimestep);
+ getLastCudaError("QDeviceCompHighNu27 execution failed");
}
//////////////////////////////////////////////////////////////////////////
void QVelDevicePlainBB27(LBMSimulationParameter* parameterDevice, QforBoundaryConditions* boundaryCondition)
{
- dim3 grid = vf::cuda::getCudaGrid( parameterDevice->numberofthreads, boundaryCondition->numberOfBCnodes);
- dim3 threads(parameterDevice->numberofthreads, 1, 1 );
+ dim3 grid = vf::cuda::getCudaGrid( parameterDevice->numberofthreads, boundaryCondition->numberOfBCnodes);
+ dim3 threads(parameterDevice->numberofthreads, 1, 1 );
- QVelDevPlainBB27<<< grid, threads >>> (
- boundaryCondition->Vx,
- boundaryCondition->Vy,
- boundaryCondition->Vz,
- parameterDevice->distributions.f[0],
- boundaryCondition->k,
- boundaryCondition->q27[0],
- boundaryCondition->numberOfBCnodes,
- parameterDevice->neighborX,
- parameterDevice->neighborY,
- parameterDevice->neighborZ,
- parameterDevice->numberOfNodes,
- parameterDevice->isEvenTimestep);
- getLastCudaError("QVelDevicePlainBB27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void QVelDeviceCouette27(unsigned int numberOfThreads,
- real* vx,
- real* vy,
- real* vz,
- real* DD,
- int* k_Q,
- real* QQ,
- unsigned int numberOfBCnodes,
- real om1,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep)
-{
- int Grid = (numberOfBCnodes / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 gridQ(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- QVelDevCouette27<<< gridQ, threads >>> ( vx,
- vy,
- vz,
- DD,
- k_Q,
- QQ,
- numberOfBCnodes,
- om1,
- neighborX,
- neighborY,
- neighborZ,
- size_Mat,
- isEvenTimestep);
- getLastCudaError("QVelDevicePlainBB27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void QVelDevice1h27( unsigned int numberOfThreads,
- int nx,
- int ny,
- real* vx,
- real* vy,
- real* vz,
- real* DD,
- int* k_Q,
- real* QQ,
- unsigned int numberOfBCnodes,
- real om1,
- real Phi,
- real angularVelocity,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- real* coordX,
- real* coordY,
- real* coordZ,
- unsigned int size_Mat,
- bool isEvenTimestep)
-{
- int Grid = (numberOfBCnodes / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 gridQ(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- QVelDev1h27<<< gridQ, threads >>> (nx,
- ny,
- vx,
- vy,
- vz,
- DD,
- k_Q,
- QQ,
- numberOfBCnodes,
- om1,
- Phi,
- angularVelocity,
- neighborX,
- neighborY,
- neighborZ,
- coordX,
- coordY,
- coordZ,
- size_Mat,
- isEvenTimestep);
- getLastCudaError("QVelDevice27 execution failed");
+ QVelDevPlainBB27<<< grid, threads >>> (
+ boundaryCondition->Vx,
+ boundaryCondition->Vy,
+ boundaryCondition->Vz,
+ parameterDevice->distributions.f[0],
+ boundaryCondition->k,
+ boundaryCondition->q27[0],
+ boundaryCondition->numberOfBCnodes,
+ parameterDevice->neighborX,
+ parameterDevice->neighborY,
+ parameterDevice->neighborZ,
+ parameterDevice->numberOfNodes,
+ parameterDevice->isEvenTimestep);
+ getLastCudaError("QVelDevicePlainBB27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void QVelDeviceCouette27(
+ unsigned int numberOfThreads,
+ real* vx,
+ real* vy,
+ real* vz,
+ real* DD,
+ int* k_Q,
+ real* QQ,
+ unsigned int numberOfBCnodes,
+ real om1,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, numberOfBCnodes);
+
+ QVelDevCouette27<<< grid.grid, grid.threads >>> (
+ vx,
+ vy,
+ vz,
+ DD,
+ k_Q,
+ QQ,
+ numberOfBCnodes,
+ om1,
+ neighborX,
+ neighborY,
+ neighborZ,
+ numberOfLBnodes,
+ isEvenTimestep);
+ getLastCudaError("QVelDevCouette27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void QVelDevice1h27(
+ unsigned int numberOfThreads,
+ int nx,
+ int ny,
+ real* vx,
+ real* vy,
+ real* vz,
+ real* DD,
+ int* k_Q,
+ real* QQ,
+ unsigned int numberOfBCnodes,
+ real om1,
+ real Phi,
+ real angularVelocity,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ real* coordX,
+ real* coordY,
+ real* coordZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, numberOfBCnodes);
+
+ QVelDev1h27<<< grid.grid, grid.threads >>> (
+ nx,
+ ny,
+ vx,
+ vy,
+ vz,
+ DD,
+ k_Q,
+ QQ,
+ numberOfBCnodes,
+ om1,
+ Phi,
+ angularVelocity,
+ neighborX,
+ neighborY,
+ neighborZ,
+ coordX,
+ coordY,
+ coordZ,
+ numberOfLBnodes,
+ isEvenTimestep);
+ getLastCudaError("QVelDev1h27 execution failed");
}
//////////////////////////////////////////////////////////////////////////
void QVelDev27(LBMSimulationParameter* parameterDevice, QforBoundaryConditions* boundaryCondition)
{
- dim3 grid = vf::cuda::getCudaGrid( parameterDevice->numberofthreads, boundaryCondition->numberOfBCnodes);
- dim3 threads(parameterDevice->numberofthreads, 1, 1 );
+ dim3 grid = vf::cuda::getCudaGrid( parameterDevice->numberofthreads, boundaryCondition->numberOfBCnodes);
+ dim3 threads(parameterDevice->numberofthreads, 1, 1 );
- QVelDevice27<<< grid, threads >>> (
- parameterDevice->nx,
- parameterDevice->ny,
- boundaryCondition->Vx,
- boundaryCondition->Vy,
- boundaryCondition->Vz,
- parameterDevice->distributions.f[0],
- boundaryCondition->k,
- boundaryCondition->q27[0],
- boundaryCondition->numberOfBCnodes,
- parameterDevice->omega,
- parameterDevice->neighborX,
- parameterDevice->neighborY,
- parameterDevice->neighborZ,
- parameterDevice->numberOfNodes,
- parameterDevice->isEvenTimestep);
- getLastCudaError("QVelDevice27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void QVelDevCompPlusSlip27(unsigned int numberOfThreads,
- real* vx,
- real* vy,
- real* vz,
- real* DD,
- int* k_Q,
- real* QQ,
- unsigned int numberOfBCnodes,
- real om1,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep)
-{
- int Grid = (numberOfBCnodes / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 gridQ(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- QVelDeviceCompPlusSlip27<<< gridQ, threads >>> (
- vx,
- vy,
- vz,
- DD,
- k_Q,
- QQ,
- numberOfBCnodes,
- om1,
- neighborX,
- neighborY,
- neighborZ,
- size_Mat,
- isEvenTimestep);
- getLastCudaError("QVelDeviceCompPlusSlip27 execution failed");
+ QVelDevice27<<< grid, threads >>> (
+ parameterDevice->nx,
+ parameterDevice->ny,
+ boundaryCondition->Vx,
+ boundaryCondition->Vy,
+ boundaryCondition->Vz,
+ parameterDevice->distributions.f[0],
+ boundaryCondition->k,
+ boundaryCondition->q27[0],
+ boundaryCondition->numberOfBCnodes,
+ parameterDevice->omega,
+ parameterDevice->neighborX,
+ parameterDevice->neighborY,
+ parameterDevice->neighborZ,
+ parameterDevice->numberOfNodes,
+ parameterDevice->isEvenTimestep);
+ getLastCudaError("QVelDevice27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void QVelDevCompPlusSlip27(
+ unsigned int numberOfThreads,
+ real* vx,
+ real* vy,
+ real* vz,
+ real* DD,
+ int* k_Q,
+ real* QQ,
+ unsigned int numberOfBCnodes,
+ real om1,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, numberOfBCnodes);
+
+ QVelDeviceCompPlusSlip27<<< grid.grid, grid.threads >>> (
+ vx,
+ vy,
+ vz,
+ DD,
+ k_Q,
+ QQ,
+ numberOfBCnodes,
+ om1,
+ neighborX,
+ neighborY,
+ neighborZ,
+ numberOfLBnodes,
+ isEvenTimestep);
+ getLastCudaError("QVelDeviceCompPlusSlip27 execution failed");
}
//////////////////////////////////////////////////////////////////////////
void QVelDevComp27(LBMSimulationParameter* parameterDevice, QforBoundaryConditions* boundaryCondition)
{
- dim3 grid = vf::cuda::getCudaGrid(parameterDevice->numberofthreads, boundaryCondition->numberOfBCnodes);
- dim3 threads(parameterDevice->numberofthreads, 1, 1 );
+ dim3 grid = vf::cuda::getCudaGrid(parameterDevice->numberofthreads, boundaryCondition->numberOfBCnodes);
+ dim3 threads(parameterDevice->numberofthreads, 1, 1 );
- QVelDeviceComp27<<< grid, threads >>> (
- boundaryCondition->Vx,
- boundaryCondition->Vy,
- boundaryCondition->Vz,
- parameterDevice->distributions.f[0],
- boundaryCondition->k,
- boundaryCondition->q27[0],
- boundaryCondition->numberOfBCnodes,
- parameterDevice->omega,
- parameterDevice->neighborX,
- parameterDevice->neighborY,
- parameterDevice->neighborZ,
- parameterDevice->numberOfNodes,
- parameterDevice->isEvenTimestep);
+ QVelDeviceComp27<<< grid, threads >>> (
+ boundaryCondition->Vx,
+ boundaryCondition->Vy,
+ boundaryCondition->Vz,
+ parameterDevice->distributions.f[0],
+ boundaryCondition->k,
+ boundaryCondition->q27[0],
+ boundaryCondition->numberOfBCnodes,
+ parameterDevice->omega,
+ parameterDevice->neighborX,
+ parameterDevice->neighborY,
+ parameterDevice->neighborZ,
+ parameterDevice->numberOfNodes,
+ parameterDevice->isEvenTimestep);
getLastCudaError("QVelDeviceComp27 execution failed");
}
//////////////////////////////////////////////////////////////////////////
-void QVelDevCompThinWalls27(unsigned int numberOfThreads,
- real* vx,
- real* vy,
- real* vz,
- real* DD,
- int* k_Q,
- real* QQ,
- unsigned int numberOfBCnodes,
- real om1,
- unsigned int* geom,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int* neighborWSB,
- unsigned int size_Mat,
- bool isEvenTimestep)
-{
- int Grid = (numberOfBCnodes / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 gridQ(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- QVelDeviceCompThinWallsPartOne27<<< gridQ, threads >>> (vx,
- vy,
- vz,
- DD,
- k_Q,
- QQ,
- numberOfBCnodes,
- om1,
- neighborX,
- neighborY,
- neighborZ,
- size_Mat,
- isEvenTimestep);
- getLastCudaError("QVelDeviceCompThinWallsPartOne27 execution failed");
-
- QThinWallsPartTwo27 <<< gridQ, threads >>> (
- DD,
- k_Q,
- QQ,
- numberOfBCnodes,
- geom,
- neighborX,
- neighborY,
- neighborZ,
- neighborWSB,
- size_Mat,
- isEvenTimestep);
- getLastCudaError("QThinWallsPartTwo27 execution failed");
-}
-
-void QVelDevCompZeroPress27(LBMSimulationParameter* parameterDevice, QforBoundaryConditions* boundaryCondition)
+void QVelDevCompThinWalls27(
+ unsigned int numberOfThreads,
+ real* vx,
+ real* vy,
+ real* vz,
+ real* DD,
+ int* k_Q,
+ real* QQ,
+ unsigned int numberOfBCnodes,
+ real om1,
+ unsigned int* geom,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned int* neighborWSB,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, numberOfBCnodes);
+
+ QVelDeviceCompThinWallsPartOne27<<< grid.grid, grid.threads >>> (
+ vx,
+ vy,
+ vz,
+ DD,
+ k_Q,
+ QQ,
+ numberOfBCnodes,
+ om1,
+ neighborX,
+ neighborY,
+ neighborZ,
+ numberOfLBnodes,
+ isEvenTimestep);
+ getLastCudaError("QVelDeviceCompThinWallsPartOne27 execution failed");
+
+ QThinWallsPartTwo27 <<< grid.grid, grid.threads >>> (
+ DD,
+ k_Q,
+ QQ,
+ numberOfBCnodes,
+ geom,
+ neighborX,
+ neighborY,
+ neighborZ,
+ neighborWSB,
+ numberOfLBnodes,
+ isEvenTimestep);
+ getLastCudaError("QThinWallsPartTwo27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void QVelDevCompZeroPress27(LBMSimulationParameter *parameterDevice, QforBoundaryConditions *boundaryCondition)
{
dim3 grid = vf::cuda::getCudaGrid( parameterDevice->numberofthreads, boundaryCondition->numberOfBCnodes);
dim3 threads(parameterDevice->numberofthreads, 1, 1 );
- QVelDeviceCompZeroPress27<<< grid, threads >>> (
- boundaryCondition->Vx,
- boundaryCondition->Vy,
- boundaryCondition->Vz,
- parameterDevice->distributions.f[0],
- boundaryCondition->k,
- boundaryCondition->q27[0],
- boundaryCondition->numberOfBCnodes,
- parameterDevice->omega,
- parameterDevice->neighborX,
- parameterDevice->neighborY,
- parameterDevice->neighborZ,
- parameterDevice->numberOfNodes,
- parameterDevice->isEvenTimestep);
- getLastCudaError("QVelDeviceCompZeroPress27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void QVelDevIncompHighNu27(unsigned int numberOfThreads,
- real* vx,
- real* vy,
- real* vz,
- real* DD,
- int* k_Q,
- real* QQ,
- unsigned int numberOfBCnodes,
- real om1,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep)
-{
- int Grid = (numberOfBCnodes / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 gridQ(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- QVelDeviceIncompHighNu27<<< gridQ, threads >>> (
- vx,
- vy,
- vz,
- DD,
- k_Q,
- QQ,
- numberOfBCnodes,
- om1,
- neighborX,
- neighborY,
- neighborZ,
- size_Mat,
- isEvenTimestep);
- getLastCudaError("QVelDeviceIncompHighNu27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void QVelDevCompHighNu27( unsigned int numberOfThreads,
- real* vx,
- real* vy,
- real* vz,
- real* DD,
- int* k_Q,
- real* QQ,
- unsigned int numberOfBCnodes,
- real om1,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep)
-{
- int Grid = (numberOfBCnodes / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 gridQ(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- QVelDeviceCompHighNu27<<< gridQ, threads >>> (
- vx,
- vy,
- vz,
- DD,
- k_Q,
- QQ,
- numberOfBCnodes,
- om1,
- neighborX,
- neighborY,
- neighborZ,
- size_Mat,
- isEvenTimestep);
- getLastCudaError("QVelDeviceComp27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void QVeloDevEQ27(unsigned int numberOfThreads,
- real* VeloX,
- real* VeloY,
- real* VeloZ,
- real* DD,
- int* k_Q,
- int numberOfBCnodes,
- real om1,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep)
-{
- int Grid = (numberOfBCnodes / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 gridQ(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- QVeloDeviceEQ27<<< gridQ, threads >>> (VeloX,
- VeloY,
- VeloZ,
- DD,
- k_Q,
- numberOfBCnodes,
- om1,
- neighborX,
- neighborY,
- neighborZ,
- size_Mat,
- isEvenTimestep);
- getLastCudaError("QVeloDeviceEQ27 execution failed");
+ QVelDeviceCompZeroPress27<<< grid, threads >>> (
+ boundaryCondition->Vx,
+ boundaryCondition->Vy,
+ boundaryCondition->Vz,
+ parameterDevice->distributions.f[0],
+ boundaryCondition->k,
+ boundaryCondition->q27[0],
+ boundaryCondition->numberOfBCnodes,
+ parameterDevice->omega,
+ parameterDevice->neighborX,
+ parameterDevice->neighborY,
+ parameterDevice->neighborZ,
+ parameterDevice->numberOfNodes,
+ parameterDevice->isEvenTimestep);
+ getLastCudaError("QVelDeviceCompZeroPress27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void QVelDevIncompHighNu27(
+ unsigned int numberOfThreads,
+ real* vx,
+ real* vy,
+ real* vz,
+ real* DD,
+ int* k_Q,
+ real* QQ,
+ unsigned int numberOfBCnodes,
+ real om1,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, numberOfBCnodes);
+
+ QVelDeviceIncompHighNu27<<< grid.grid, grid.threads >>> (
+ vx,
+ vy,
+ vz,
+ DD,
+ k_Q,
+ QQ,
+ numberOfBCnodes,
+ om1,
+ neighborX,
+ neighborY,
+ neighborZ,
+ numberOfLBnodes,
+ isEvenTimestep);
+ getLastCudaError("QVelDeviceIncompHighNu27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void QVelDevCompHighNu27(
+ unsigned int numberOfThreads,
+ real* vx,
+ real* vy,
+ real* vz,
+ real* DD,
+ int* k_Q,
+ real* QQ,
+ unsigned int numberOfBCnodes,
+ real om1,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, numberOfBCnodes);
+
+ QVelDeviceCompHighNu27<<< grid.grid, grid.threads >>> (
+ vx,
+ vy,
+ vz,
+ DD,
+ k_Q,
+ QQ,
+ numberOfBCnodes,
+ om1,
+ neighborX,
+ neighborY,
+ neighborZ,
+ numberOfLBnodes,
+ isEvenTimestep);
+ getLastCudaError("QVelDeviceComp27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void QVeloDevEQ27(
+ unsigned int numberOfThreads,
+ real* VeloX,
+ real* VeloY,
+ real* VeloZ,
+ real* DD,
+ int* k_Q,
+ int numberOfBCnodes,
+ real om1,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, numberOfBCnodes);
+
+ QVeloDeviceEQ27<<< grid.grid, grid.threads >>> (
+ VeloX,
+ VeloY,
+ VeloZ,
+ DD,
+ k_Q,
+ numberOfBCnodes,
+ om1,
+ neighborX,
+ neighborY,
+ neighborZ,
+ numberOfLBnodes,
+ isEvenTimestep);
+ getLastCudaError("QVeloDeviceEQ27 execution failed");
}
//////////////////////////////////////////////////////////////////////////
void QVeloStreetDevEQ27(
- uint numberOfThreads,
- real* veloXfraction,
- real* veloYfraction,
- int* naschVelo,
- real* DD,
- int* naschIndex,
- int numberOfStreetNodes,
- real velocityRatio,
- uint* neighborX,
- uint* neighborY,
- uint* neighborZ,
- uint size_Mat,
- bool isEvenTimestep)
-{
- int Grid = (numberOfStreetNodes / numberOfThreads) + 1;
- int Grid1, Grid2;
- if (Grid > 512)
- {
- Grid1 = 512;
- Grid2 = (Grid / Grid1) + 1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 gridQ(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1);
-
- QVeloStreetDeviceEQ27 << < gridQ, threads >> > (
- veloXfraction,
- veloYfraction,
- naschVelo,
- DD,
- naschIndex,
- numberOfStreetNodes,
- velocityRatio,
- neighborX,
- neighborY,
- neighborZ,
- size_Mat,
- isEvenTimestep);
- getLastCudaError("QVeloStreetDeviceEQ27 execution failed");
+ uint numberOfThreads,
+ real* veloXfraction,
+ real* veloYfraction,
+ int* naschVelo,
+ real* DD,
+ int* naschIndex,
+ int numberOfStreetNodes,
+ real velocityRatio,
+ uint* neighborX,
+ uint* neighborY,
+ uint* neighborZ,
+ uint numberOfLBnodes,
+ bool isEvenTimestep)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, numberOfStreetNodes);
+
+ QVeloStreetDeviceEQ27 << < grid.grid, grid.threads >> > (
+ veloXfraction,
+ veloYfraction,
+ naschVelo,
+ DD,
+ naschIndex,
+ numberOfStreetNodes,
+ velocityRatio,
+ neighborX,
+ neighborY,
+ neighborZ,
+ numberOfLBnodes,
+ isEvenTimestep);
+ getLastCudaError("QVeloStreetDeviceEQ27 execution failed");
}
//////////////////////////////////////////////////////////////////////////
void QSlipDev27(LBMSimulationParameter* parameterDevice, QforBoundaryConditions* boundaryCondition)
{
- dim3 grid = vf::cuda::getCudaGrid( parameterDevice->numberofthreads, boundaryCondition->numberOfBCnodes);
- dim3 threads(parameterDevice->numberofthreads, 1, 1 );
+ dim3 grid = vf::cuda::getCudaGrid( parameterDevice->numberofthreads, boundaryCondition->numberOfBCnodes);
+ dim3 threads(parameterDevice->numberofthreads, 1, 1 );
- QSlipDevice27<<< grid, threads >>> (
- parameterDevice->distributions.f[0],
- boundaryCondition->k,
- boundaryCondition->q27[0],
- boundaryCondition->numberOfBCnodes,
- parameterDevice->omega,
- parameterDevice->neighborX,
- parameterDevice->neighborY,
- parameterDevice->neighborZ,
- parameterDevice->numberOfNodes,
- parameterDevice->isEvenTimestep);
- getLastCudaError("QSlipDevice27 execution failed");
+ QSlipDevice27<<< grid, threads >>> (
+ parameterDevice->distributions.f[0],
+ boundaryCondition->k,
+ boundaryCondition->q27[0],
+ boundaryCondition->numberOfBCnodes,
+ parameterDevice->omega,
+ parameterDevice->neighborX,
+ parameterDevice->neighborY,
+ parameterDevice->neighborZ,
+ parameterDevice->numberOfNodes,
+ parameterDevice->isEvenTimestep);
+ getLastCudaError("QSlipDevice27 execution failed");
}
//////////////////////////////////////////////////////////////////////////
void QSlipDevCompTurbulentViscosity27(LBMSimulationParameter* parameterDevice, QforBoundaryConditions* boundaryCondition)
{
- dim3 grid = vf::cuda::getCudaGrid( parameterDevice->numberofthreads, boundaryCondition->numberOfBCnodes);
- dim3 threads(parameterDevice->numberofthreads, 1, 1 );
+ dim3 grid = vf::cuda::getCudaGrid( parameterDevice->numberofthreads, boundaryCondition->numberOfBCnodes);
+ dim3 threads(parameterDevice->numberofthreads, 1, 1 );
- QSlipDeviceComp27TurbViscosity<<< grid, threads >>> (
- parameterDevice->distributions.f[0],
- boundaryCondition->k,
- boundaryCondition->q27[0],
- boundaryCondition->numberOfBCnodes,
- parameterDevice->omega,
- parameterDevice->neighborX,
- parameterDevice->neighborY,
- parameterDevice->neighborZ,
- parameterDevice->turbViscosity,
- parameterDevice->numberOfNodes,
- parameterDevice->isEvenTimestep);
- getLastCudaError("QSlipDeviceComp27TurbViscosity execution failed");
+ QSlipDeviceComp27TurbViscosity<<< grid, threads >>> (
+ parameterDevice->distributions.f[0],
+ boundaryCondition->k,
+ boundaryCondition->q27[0],
+ boundaryCondition->numberOfBCnodes,
+ parameterDevice->omega,
+ parameterDevice->neighborX,
+ parameterDevice->neighborY,
+ parameterDevice->neighborZ,
+ parameterDevice->turbViscosity,
+ parameterDevice->numberOfNodes,
+ parameterDevice->isEvenTimestep);
+ getLastCudaError("QSlipDeviceComp27TurbViscosity execution failed");
}
//////////////////////////////////////////////////////////////////////////
void QSlipPressureDevCompTurbulentViscosity27(LBMSimulationParameter* parameterDevice, QforBoundaryConditions* boundaryCondition)
{
- dim3 grid = vf::cuda::getCudaGrid( parameterDevice->numberofthreads, boundaryCondition->numberOfBCnodes);
- dim3 threads(parameterDevice->numberofthreads, 1, 1 );
+ dim3 grid = vf::cuda::getCudaGrid( parameterDevice->numberofthreads, boundaryCondition->numberOfBCnodes);
+ dim3 threads(parameterDevice->numberofthreads, 1, 1 );
- QSlipPressureDeviceComp27TurbViscosity<<< grid, threads >>> (
- parameterDevice->distributions.f[0],
- boundaryCondition->k,
- boundaryCondition->q27[0],
- boundaryCondition->numberOfBCnodes,
- parameterDevice->omega,
- parameterDevice->neighborX,
- parameterDevice->neighborY,
- parameterDevice->neighborZ,
- parameterDevice->turbViscosity,
- parameterDevice->numberOfNodes,
- parameterDevice->isEvenTimestep);
- getLastCudaError("QSlipDeviceComp27TurbViscosity execution failed");
+ QSlipPressureDeviceComp27TurbViscosity<<< grid, threads >>> (
+ parameterDevice->distributions.f[0],
+ boundaryCondition->k,
+ boundaryCondition->q27[0],
+ boundaryCondition->numberOfBCnodes,
+ parameterDevice->omega,
+ parameterDevice->neighborX,
+ parameterDevice->neighborY,
+ parameterDevice->neighborZ,
+ parameterDevice->turbViscosity,
+ parameterDevice->numberOfNodes,
+ parameterDevice->isEvenTimestep);
+ getLastCudaError("QSlipDeviceComp27TurbViscosity execution failed");
}
//////////////////////////////////////////////////////////////////////////
void QSlipDevComp27(LBMSimulationParameter* parameterDevice, QforBoundaryConditions* boundaryCondition)
{
- dim3 grid = vf::cuda::getCudaGrid( parameterDevice->numberofthreads, boundaryCondition->numberOfBCnodes);
- dim3 threads(parameterDevice->numberofthreads, 1, 1 );
+ dim3 grid = vf::cuda::getCudaGrid( parameterDevice->numberofthreads, boundaryCondition->numberOfBCnodes);
+ dim3 threads(parameterDevice->numberofthreads, 1, 1 );
- QSlipDeviceComp27<<< grid, threads >>> (
- parameterDevice->distributions.f[0],
- boundaryCondition->k,
- boundaryCondition->q27[0],
- boundaryCondition->numberOfBCnodes,
- parameterDevice->omega,
- parameterDevice->neighborX,
- parameterDevice->neighborY,
- parameterDevice->neighborZ,
- parameterDevice->numberOfNodes,
- parameterDevice->isEvenTimestep);
- getLastCudaError("QSlipDeviceComp27 execution failed");
+ QSlipDeviceComp27<<< grid, threads >>> (
+ parameterDevice->distributions.f[0],
+ boundaryCondition->k,
+ boundaryCondition->q27[0],
+ boundaryCondition->numberOfBCnodes,
+ parameterDevice->omega,
+ parameterDevice->neighborX,
+ parameterDevice->neighborY,
+ parameterDevice->neighborZ,
+ parameterDevice->numberOfNodes,
+ parameterDevice->isEvenTimestep);
+ getLastCudaError("QSlipDeviceComp27 execution failed");
}
//////////////////////////////////////////////////////////////////////////
void BBSlipDevComp27(LBMSimulationParameter* parameterDevice, QforBoundaryConditions* boundaryCondition)
{
- dim3 grid = vf::cuda::getCudaGrid( parameterDevice->numberofthreads, boundaryCondition->numberOfBCnodes);
- dim3 threads(parameterDevice->numberofthreads, 1, 1 );
+ dim3 grid = vf::cuda::getCudaGrid( parameterDevice->numberofthreads, boundaryCondition->numberOfBCnodes);
+ dim3 threads(parameterDevice->numberofthreads, 1, 1 );
- QSlipDeviceComp27<<< grid, threads >>> (
- parameterDevice->distributions.f[0],
- boundaryCondition->k,
- boundaryCondition->q27[0],
- boundaryCondition->numberOfBCnodes,
- parameterDevice->omega,
- parameterDevice->neighborX,
- parameterDevice->neighborY,
- parameterDevice->neighborZ,
- parameterDevice->numberOfNodes,
- parameterDevice->isEvenTimestep);
- getLastCudaError("BBSlipDeviceComp27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void QSlipGeomDevComp27(unsigned int numberOfThreads,
- real* DD,
- int* k_Q,
- real* QQ,
- unsigned int numberOfBCnodes,
- real om1,
- real* NormalX,
- real* NormalY,
- real* NormalZ,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep)
-{
- int Grid = (numberOfBCnodes / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 gridQ(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- QSlipGeomDeviceComp27<<< gridQ, threads >>> (DD,
- k_Q,
- QQ,
- numberOfBCnodes,
- om1,
- NormalX,
- NormalY,
- NormalZ,
- neighborX,
- neighborY,
- neighborZ,
- size_Mat,
- isEvenTimestep);
- getLastCudaError("QSlipGeomDeviceComp27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void QSlipNormDevComp27(unsigned int numberOfThreads,
- real* DD,
- int* k_Q,
- real* QQ,
- unsigned int numberOfBCnodes,
- real om1,
- real* NormalX,
- real* NormalY,
- real* NormalZ,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep)
-{
- int Grid = (numberOfBCnodes / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 gridQ(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- QSlipNormDeviceComp27<<< gridQ, threads >>> (DD,
- k_Q,
- QQ,
- numberOfBCnodes,
- om1,
- NormalX,
- NormalY,
- NormalZ,
- neighborX,
- neighborY,
- neighborZ,
- size_Mat,
- isEvenTimestep);
- getLastCudaError("QSlipGeomDeviceComp27 execution failed");
+ BBSlipDeviceComp27<<< grid, threads >>> (
+ parameterDevice->distributions.f[0],
+ boundaryCondition->k,
+ boundaryCondition->q27[0],
+ boundaryCondition->numberOfBCnodes,
+ parameterDevice->neighborX,
+ parameterDevice->neighborY,
+ parameterDevice->neighborZ,
+ parameterDevice->numberOfNodes,
+ parameterDevice->isEvenTimestep);
+ getLastCudaError("BBSlipDeviceComp27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void QSlipGeomDevComp27(
+ unsigned int numberOfThreads,
+ real* DD,
+ int* k_Q,
+ real* QQ,
+ unsigned int numberOfBCnodes,
+ real om1,
+ real* NormalX,
+ real* NormalY,
+ real* NormalZ,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep)
+{
+ vf::cuda::CudaGrid grid(numberOfThreads, numberOfBCnodes);
+
+ QSlipGeomDeviceComp27<<< grid.grid, grid.threads >>> (
+ DD,
+ k_Q,
+ QQ,
+ numberOfBCnodes,
+ om1,
+ NormalX,
+ NormalY,
+ NormalZ,
+ neighborX,
+ neighborY,
+ neighborZ,
+ numberOfLBnodes,
+ isEvenTimestep);
+ getLastCudaError("QSlipGeomDeviceComp27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void QSlipNormDevComp27(
+ unsigned int numberOfThreads,
+ real* DD,
+ int* k_Q,
+ real* QQ,
+ unsigned int numberOfBCnodes,
+ real om1,
+ real* NormalX,
+ real* NormalY,
+ real* NormalZ,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, numberOfBCnodes);
+
+ QSlipNormDeviceComp27<<< grid.grid, grid.threads >>> (
+ DD,
+ k_Q,
+ QQ,
+ numberOfBCnodes,
+ om1,
+ NormalX,
+ NormalY,
+ NormalZ,
+ neighborX,
+ neighborY,
+ neighborZ,
+ numberOfLBnodes,
+ isEvenTimestep);
+ getLastCudaError("QSlipNormDeviceComp27 execution failed");
}
//////////////////////////////////////////////////////////////////////////
void QStressDevComp27(Parameter *para, QforBoundaryConditions* boundaryCondition, const int level)
{
- dim3 grid = vf::cuda::getCudaGrid( para->getParD(level)->numberofthreads, boundaryCondition->numberOfBCnodes);
- dim3 threads(para->getParD(level)->numberofthreads, 1, 1 );
-
- QStressDeviceComp27<<< grid, threads >>> (
- para->getParD(level)->distributions.f[0],
- boundaryCondition->k,
- boundaryCondition->kN,
- boundaryCondition->q27[0],
- boundaryCondition->numberOfBCnodes,
- para->getParD(level)->omega,
- para->getParD(level)->turbViscosity,
- para->getParD(level)->velocityX,
- para->getParD(level)->velocityY,
- para->getParD(level)->velocityY,
- boundaryCondition->normalX,
- boundaryCondition->normalY,
- boundaryCondition->normalZ,
- boundaryCondition->Vx,
- boundaryCondition->Vy,
- boundaryCondition->Vz,
- boundaryCondition->Vx1,
- boundaryCondition->Vy1,
- boundaryCondition->Vz1,
- para->getParD(level)->wallModel.samplingOffset,
- para->getParD(level)->wallModel.z0,
- para->getHasWallModelMonitor(),
- para->getParD(level)->wallModel.u_star,
- para->getParD(level)->wallModel.Fx,
- para->getParD(level)->wallModel.Fy,
- para->getParD(level)->wallModel.Fz,
- para->getParD(level)->neighborX,
- para->getParD(level)->neighborY,
- para->getParD(level)->neighborZ,
- para->getParD(level)->numberOfNodes,
- para->getParD(level)->isEvenTimestep);
- getLastCudaError("QSlipDeviceComp27 execution failed");
+ dim3 grid = vf::cuda::getCudaGrid( para->getParD(level)->numberofthreads, boundaryCondition->numberOfBCnodes);
+ dim3 threads(para->getParD(level)->numberofthreads, 1, 1 );
+
+ QStressDeviceComp27<<< grid, threads >>> (
+ para->getParD(level)->distributions.f[0],
+ boundaryCondition->k,
+ boundaryCondition->kN,
+ boundaryCondition->q27[0],
+ boundaryCondition->numberOfBCnodes,
+ para->getParD(level)->omega,
+ para->getParD(level)->turbViscosity,
+ para->getParD(level)->velocityX,
+ para->getParD(level)->velocityY,
+ para->getParD(level)->velocityY,
+ boundaryCondition->normalX,
+ boundaryCondition->normalY,
+ boundaryCondition->normalZ,
+ boundaryCondition->Vx,
+ boundaryCondition->Vy,
+ boundaryCondition->Vz,
+ boundaryCondition->Vx1,
+ boundaryCondition->Vy1,
+ boundaryCondition->Vz1,
+ para->getParD(level)->wallModel.samplingOffset,
+ para->getParD(level)->wallModel.z0,
+ para->getHasWallModelMonitor(),
+ para->getParD(level)->wallModel.u_star,
+ para->getParD(level)->wallModel.Fx,
+ para->getParD(level)->wallModel.Fy,
+ para->getParD(level)->wallModel.Fz,
+ para->getParD(level)->neighborX,
+ para->getParD(level)->neighborY,
+ para->getParD(level)->neighborZ,
+ para->getParD(level)->numberOfNodes,
+ para->getParD(level)->isEvenTimestep);
+ getLastCudaError("QStressDeviceComp27 execution failed");
}
//////////////////////////////////////////////////////////////////////////
void BBStressDev27(Parameter *para, QforBoundaryConditions* boundaryCondition, const int level)
{
- dim3 grid = vf::cuda::getCudaGrid( para->getParD(level)->numberofthreads, boundaryCondition->numberOfBCnodes);
- dim3 threads(para->getParD(level)->numberofthreads, 1, 1 );
-
- BBStressDevice27<<< grid, threads >>> (
- para->getParD(level)->distributions.f[0],
- boundaryCondition->k,
- boundaryCondition->kN,
- boundaryCondition->q27[0],
- boundaryCondition->numberOfBCnodes,
- para->getParD(level)->velocityX,
- para->getParD(level)->velocityY,
- para->getParD(level)->velocityY,
- boundaryCondition->normalX,
- boundaryCondition->normalY,
- boundaryCondition->normalZ,
- boundaryCondition->Vx,
- boundaryCondition->Vy,
- boundaryCondition->Vz,
- boundaryCondition->Vx1,
- boundaryCondition->Vy1,
- boundaryCondition->Vz1,
- para->getParD(level)->wallModel.samplingOffset,
- para->getParD(level)->wallModel.z0,
- para->getHasWallModelMonitor(),
- para->getParD(level)->wallModel.u_star,
- para->getParD(level)->wallModel.Fx,
- para->getParD(level)->wallModel.Fy,
- para->getParD(level)->wallModel.Fz,
- para->getParD(level)->neighborX,
- para->getParD(level)->neighborY,
- para->getParD(level)->neighborZ,
- para->getParD(level)->numberOfNodes,
- para->getParD(level)->isEvenTimestep);
- getLastCudaError("BBStressDevice27 execution failed");
+ dim3 grid = vf::cuda::getCudaGrid( para->getParD(level)->numberofthreads, boundaryCondition->numberOfBCnodes);
+ dim3 threads(para->getParD(level)->numberofthreads, 1, 1 );
+
+ BBStressDevice27<<< grid, threads >>> (
+ para->getParD(level)->distributions.f[0],
+ boundaryCondition->k,
+ boundaryCondition->kN,
+ boundaryCondition->q27[0],
+ boundaryCondition->numberOfBCnodes,
+ para->getParD(level)->velocityX,
+ para->getParD(level)->velocityY,
+ para->getParD(level)->velocityY,
+ boundaryCondition->normalX,
+ boundaryCondition->normalY,
+ boundaryCondition->normalZ,
+ boundaryCondition->Vx,
+ boundaryCondition->Vy,
+ boundaryCondition->Vz,
+ boundaryCondition->Vx1,
+ boundaryCondition->Vy1,
+ boundaryCondition->Vz1,
+ para->getParD(level)->wallModel.samplingOffset,
+ para->getParD(level)->wallModel.z0,
+ para->getHasWallModelMonitor(),
+ para->getParD(level)->wallModel.u_star,
+ para->getParD(level)->wallModel.Fx,
+ para->getParD(level)->wallModel.Fy,
+ para->getParD(level)->wallModel.Fz,
+ para->getParD(level)->neighborX,
+ para->getParD(level)->neighborY,
+ para->getParD(level)->neighborZ,
+ para->getParD(level)->numberOfNodes,
+ para->getParD(level)->isEvenTimestep);
+ getLastCudaError("BBStressDevice27 execution failed");
}
//////////////////////////////////////////////////////////////////////////
void BBStressPressureDev27(Parameter *para, QforBoundaryConditions* boundaryCondition, const int level)
{
- dim3 grid = vf::cuda::getCudaGrid( para->getParD(level)->numberofthreads, boundaryCondition->numberOfBCnodes);
- dim3 threads(para->getParD(level)->numberofthreads, 1, 1 );
-
- BBStressPressureDevice27<<< grid, threads >>> (
- para->getParD(level)->distributions.f[0],
- boundaryCondition->k,
- boundaryCondition->kN,
- boundaryCondition->q27[0],
- boundaryCondition->numberOfBCnodes,
- para->getParD(level)->velocityX,
- para->getParD(level)->velocityY,
- para->getParD(level)->velocityY,
- boundaryCondition->normalX,
- boundaryCondition->normalY,
- boundaryCondition->normalZ,
- boundaryCondition->Vx,
- boundaryCondition->Vy,
- boundaryCondition->Vz,
- boundaryCondition->Vx1,
- boundaryCondition->Vy1,
- boundaryCondition->Vz1,
- para->getParD(level)->wallModel.samplingOffset,
- para->getParD(level)->wallModel.z0,
- para->getHasWallModelMonitor(),
- para->getParD(level)->wallModel.u_star,
- para->getParD(level)->wallModel.Fx,
- para->getParD(level)->wallModel.Fy,
- para->getParD(level)->wallModel.Fz,
- para->getParD(level)->neighborX,
- para->getParD(level)->neighborY,
- para->getParD(level)->neighborZ,
- para->getParD(level)->numberOfNodes,
- para->getParD(level)->isEvenTimestep);
- getLastCudaError("BBStressDevice27 execution failed");
+ dim3 grid = vf::cuda::getCudaGrid( para->getParD(level)->numberofthreads, boundaryCondition->numberOfBCnodes);
+ dim3 threads(para->getParD(level)->numberofthreads, 1, 1 );
+
+ BBStressPressureDevice27<<< grid, threads >>> (
+ para->getParD(level)->distributions.f[0],
+ boundaryCondition->k,
+ boundaryCondition->kN,
+ boundaryCondition->q27[0],
+ boundaryCondition->numberOfBCnodes,
+ para->getParD(level)->velocityX,
+ para->getParD(level)->velocityY,
+ para->getParD(level)->velocityY,
+ boundaryCondition->normalX,
+ boundaryCondition->normalY,
+ boundaryCondition->normalZ,
+ boundaryCondition->Vx,
+ boundaryCondition->Vy,
+ boundaryCondition->Vz,
+ boundaryCondition->Vx1,
+ boundaryCondition->Vy1,
+ boundaryCondition->Vz1,
+ para->getParD(level)->wallModel.samplingOffset,
+ para->getParD(level)->wallModel.z0,
+ para->getHasWallModelMonitor(),
+ para->getParD(level)->wallModel.u_star,
+ para->getParD(level)->wallModel.Fx,
+ para->getParD(level)->wallModel.Fy,
+ para->getParD(level)->wallModel.Fz,
+ para->getParD(level)->neighborX,
+ para->getParD(level)->neighborY,
+ para->getParD(level)->neighborZ,
+ para->getParD(level)->numberOfNodes,
+ para->getParD(level)->isEvenTimestep);
+ getLastCudaError("BBStressPressureDevice27 execution failed");
}
//////////////////////////////////////////////////////////////////////////
void QPressDev27(LBMSimulationParameter* parameterDevice, QforBoundaryConditions* boundaryCondition)
{
- dim3 grid = vf::cuda::getCudaGrid( parameterDevice->numberofthreads, boundaryCondition->numberOfBCnodes);
- dim3 threads(parameterDevice->numberofthreads, 1, 1 );
+ dim3 grid = vf::cuda::getCudaGrid( parameterDevice->numberofthreads, boundaryCondition->numberOfBCnodes);
+ dim3 threads(parameterDevice->numberofthreads, 1, 1 );
- QPressDevice27<<< grid, threads >>> (
- boundaryCondition->RhoBC,
- parameterDevice->distributions.f[0],
- boundaryCondition->k,
- boundaryCondition->q27[0],
- boundaryCondition->numberOfBCnodes,
- parameterDevice->omega,
- parameterDevice->neighborX,
- parameterDevice->neighborY,
- parameterDevice->neighborZ,
- parameterDevice->numberOfNodes,
- parameterDevice->isEvenTimestep);
- getLastCudaError("QPressDevice27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void QPressDevAntiBB27( unsigned int numberOfThreads,
- real* rhoBC,
- real* vx,
- real* vy,
- real* vz,
- real* DD,
- int* k_Q,
- real* QQ,
- int numberOfBCnodes,
- real om1,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep)
-{
- int Grid = (numberOfBCnodes / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 gridQ(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- QPressDeviceAntiBB27<<< gridQ, threads >>>( rhoBC,
- vx,
- vy,
- vz,
- DD,
- k_Q,
- QQ,
- numberOfBCnodes,
- om1,
- neighborX,
- neighborY,
- neighborZ,
- size_Mat,
- isEvenTimestep);
+ QPressDevice27<<< grid, threads >>> (
+ boundaryCondition->RhoBC,
+ parameterDevice->distributions.f[0],
+ boundaryCondition->k,
+ boundaryCondition->q27[0],
+ boundaryCondition->numberOfBCnodes,
+ parameterDevice->omega,
+ parameterDevice->neighborX,
+ parameterDevice->neighborY,
+ parameterDevice->neighborZ,
+ parameterDevice->numberOfNodes,
+ parameterDevice->isEvenTimestep);
+ getLastCudaError("QPressDevice27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void QPressDevAntiBB27(
+ unsigned int numberOfThreads,
+ real* rhoBC,
+ real* vx,
+ real* vy,
+ real* vz,
+ real* DD,
+ int* k_Q,
+ real* QQ,
+ int numberOfBCnodes,
+ real om1,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, numberOfBCnodes);
+
+ QPressDeviceAntiBB27<<< grid.grid, grid.threads >>>(
+ rhoBC,
+ vx,
+ vy,
+ vz,
+ DD,
+ k_Q,
+ QQ,
+ numberOfBCnodes,
+ om1,
+ neighborX,
+ neighborY,
+ neighborZ,
+ numberOfLBnodes,
+ isEvenTimestep);
getLastCudaError("QPressDeviceAntiBB27 execution failed");
}
//////////////////////////////////////////////////////////////////////////
-void QPressDevFixBackflow27( unsigned int numberOfThreads,
- real* rhoBC,
- real* DD,
- int* k_Q,
- unsigned int numberOfBCnodes,
- real om1,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep)
-{
- int Grid = (numberOfBCnodes / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 gridQ(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- QPressDeviceFixBackflow27<<< gridQ, threads >>> ( rhoBC,
- DD,
- k_Q,
- numberOfBCnodes,
- om1,
- neighborX,
- neighborY,
- neighborZ,
- size_Mat,
- isEvenTimestep);
- getLastCudaError("QPressDeviceFixBackflow27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void QPressDevDirDepBot27( unsigned int numberOfThreads,
- real* rhoBC,
- real* DD,
- int* k_Q,
- unsigned int numberOfBCnodes,
- real om1,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep)
-{
- int Grid = (numberOfBCnodes / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 gridQ(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- QPressDeviceDirDepBot27<<< gridQ, threads >>> ( rhoBC,
- DD,
- k_Q,
- numberOfBCnodes,
- om1,
- neighborX,
- neighborY,
- neighborZ,
- size_Mat,
- isEvenTimestep);
- getLastCudaError("QPressDeviceDirDepBot27 execution failed");
+void QPressDevFixBackflow27(
+ unsigned int numberOfThreads,
+ real* rhoBC,
+ real* DD,
+ int* k_Q,
+ unsigned int numberOfBCnodes,
+ real om1,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, numberOfBCnodes);
+
+ QPressDeviceFixBackflow27<<< grid.grid, grid.threads >>> (
+ rhoBC,
+ DD,
+ k_Q,
+ numberOfBCnodes,
+ om1,
+ neighborX,
+ neighborY,
+ neighborZ,
+ numberOfLBnodes,
+ isEvenTimestep);
+ getLastCudaError("QPressDeviceFixBackflow27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void QPressDevDirDepBot27(
+ unsigned int numberOfThreads,
+ real* rhoBC,
+ real* DD,
+ int* k_Q,
+ unsigned int numberOfBCnodes,
+ real om1,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, numberOfBCnodes);
+
+ QPressDeviceDirDepBot27<<< grid.grid, grid.threads >>> (
+ rhoBC,
+ DD,
+ k_Q,
+ numberOfBCnodes,
+ om1,
+ neighborX,
+ neighborY,
+ neighborZ,
+ numberOfLBnodes,
+ isEvenTimestep);
+ getLastCudaError("QPressDeviceDirDepBot27 execution failed");
}
//////////////////////////////////////////////////////////////////////////
void QPressNoRhoDev27(LBMSimulationParameter* parameterDevice, QforBoundaryConditions* boundaryCondition)
{
- dim3 grid = vf::cuda::getCudaGrid( parameterDevice->numberofthreads, boundaryCondition->numberOfBCnodes);
- dim3 threads(parameterDevice->numberofthreads, 1, 1 );
+ dim3 grid = vf::cuda::getCudaGrid( parameterDevice->numberofthreads, boundaryCondition->numberOfBCnodes);
+ dim3 threads(parameterDevice->numberofthreads, 1, 1 );
+
+ QPressNoRhoDevice27<<< grid, threads >>> (
+ boundaryCondition->RhoBC,
+ parameterDevice->distributions.f[0],
+ boundaryCondition->k,
+ boundaryCondition->kN,
+ boundaryCondition->numberOfBCnodes,
+ parameterDevice->omega,
+ parameterDevice->neighborX,
+ parameterDevice->neighborY,
+ parameterDevice->neighborZ,
+ parameterDevice->numberOfNodes,
+ parameterDevice->isEvenTimestep,
+ vf::lbm::dir::DIR_P00);
+ getLastCudaError("QPressNoRhoDevice27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void QPressZeroRhoOutflowDev27(LBMSimulationParameter* parameterDevice, QforBoundaryConditions* boundaryCondition)
+{
+ dim3 grid = vf::cuda::getCudaGrid( parameterDevice->numberofthreads, boundaryCondition->numberOfBCnodes);
+ dim3 threads(parameterDevice->numberofthreads, 1, 1 );
- QPressNoRhoDevice27<<< grid, threads >>> (
- boundaryCondition->RhoBC,
- parameterDevice->distributions.f[0],
- boundaryCondition->k,
- boundaryCondition->kN,
- boundaryCondition->numberOfBCnodes,
- parameterDevice->omega,
- parameterDevice->neighborX,
- parameterDevice->neighborY,
- parameterDevice->neighborZ,
- parameterDevice->numberOfNodes,
- parameterDevice->isEvenTimestep);
- getLastCudaError("QPressNoRhoDevice27 execution failed");
+ QPressZeroRhoOutflowDevice27<<< grid, threads >>> (
+ boundaryCondition->RhoBC,
+ parameterDevice->distributions.f[0],
+ boundaryCondition->k,
+ boundaryCondition->kN,
+ boundaryCondition->numberOfBCnodes,
+ parameterDevice->omega,
+ parameterDevice->neighborX,
+ parameterDevice->neighborY,
+ parameterDevice->neighborZ,
+ parameterDevice->numberOfNodes,
+ parameterDevice->isEvenTimestep,
+ vf::lbm::dir::DIR_P00,
+ parameterDevice->outflowPressureCorrectionFactor);
+ getLastCudaError("QPressZeroRhoOutflowDevice27 execution failed");
}
//////////////////////////////////////////////////////////////////////////
void QInflowScaleByPressDev27(LBMSimulationParameter* parameterDevice, QforBoundaryConditions* boundaryCondition)
{
- dim3 grid = vf::cuda::getCudaGrid( parameterDevice->numberofthreads, boundaryCondition->numberOfBCnodes);
- dim3 threads(parameterDevice->numberofthreads, 1, 1 );
+ dim3 grid = vf::cuda::getCudaGrid( parameterDevice->numberofthreads, boundaryCondition->numberOfBCnodes);
+ dim3 threads(parameterDevice->numberofthreads, 1, 1 );
- QInflowScaleByPressDevice27<<< grid, threads >>> (
- boundaryCondition->RhoBC,
- parameterDevice->distributions.f[0],
- boundaryCondition->k,
- boundaryCondition->kN,
- boundaryCondition->numberOfBCnodes,
- parameterDevice->omega,
- parameterDevice->neighborX,
- parameterDevice->neighborY,
- parameterDevice->neighborZ,
- parameterDevice->numberOfNodes,
- parameterDevice->isEvenTimestep);
- getLastCudaError("QInflowScaleByPressDevice27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void QPressDevOld27( unsigned int numberOfThreads,
- real* rhoBC,
- real* DD,
- int* k_Q,
- int* k_N,
- unsigned int numberOfBCnodes,
- real om1,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep)
-{
- int Grid = (numberOfBCnodes / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 gridQ(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- QPressDeviceOld27<<< gridQ, threads >>> ( rhoBC,
- DD,
- k_Q,
- k_N,
- numberOfBCnodes,
- om1,
- neighborX,
- neighborY,
- neighborZ,
- size_Mat,
- isEvenTimestep);
- getLastCudaError("QPressDeviceOld27 execution failed");
+ QInflowScaleByPressDevice27<<< grid, threads >>> (
+ boundaryCondition->RhoBC,
+ parameterDevice->distributions.f[0],
+ boundaryCondition->k,
+ boundaryCondition->kN,
+ boundaryCondition->numberOfBCnodes,
+ parameterDevice->omega,
+ parameterDevice->neighborX,
+ parameterDevice->neighborY,
+ parameterDevice->neighborZ,
+ parameterDevice->numberOfNodes,
+ parameterDevice->isEvenTimestep);
+ getLastCudaError("QInflowScaleByPressDevice27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void QPressDevOld27(
+ unsigned int numberOfThreads,
+ real* rhoBC,
+ real* DD,
+ int* k_Q,
+ int* k_N,
+ unsigned int numberOfBCnodes,
+ real om1,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, numberOfBCnodes);
+
+ QPressDeviceOld27<<< grid.grid, grid.threads >>> (
+ rhoBC,
+ DD,
+ k_Q,
+ k_N,
+ numberOfBCnodes,
+ om1,
+ neighborX,
+ neighborY,
+ neighborZ,
+ numberOfLBnodes,
+ isEvenTimestep);
+ getLastCudaError("QPressDeviceOld27 execution failed");
}
//////////////////////////////////////////////////////////////////////////
void QPressDevIncompNEQ27(LBMSimulationParameter* parameterDevice, QforBoundaryConditions* boundaryCondition)
{
- dim3 grid = vf::cuda::getCudaGrid( parameterDevice->numberofthreads, boundaryCondition->numberOfBCnodes);
- dim3 threads(parameterDevice->numberofthreads, 1, 1 );
+ dim3 grid = vf::cuda::getCudaGrid( parameterDevice->numberofthreads, boundaryCondition->numberOfBCnodes);
+ dim3 threads(parameterDevice->numberofthreads, 1, 1 );
- QPressDeviceIncompNEQ27<<< grid, threads >>> (
- boundaryCondition->RhoBC,
- parameterDevice->distributions.f[0],
- boundaryCondition->k,
- boundaryCondition->kN,
- boundaryCondition->numberOfBCnodes,
- parameterDevice->omega,
- parameterDevice->neighborX,
- parameterDevice->neighborY,
- parameterDevice->neighborZ,
- parameterDevice->numberOfNodes,
- parameterDevice->isEvenTimestep);
- getLastCudaError("QPressDeviceIncompNEQ27 execution failed");
+ QPressDeviceIncompNEQ27<<< grid, threads >>> (
+ boundaryCondition->RhoBC,
+ parameterDevice->distributions.f[0],
+ boundaryCondition->k,
+ boundaryCondition->kN,
+ boundaryCondition->numberOfBCnodes,
+ parameterDevice->omega,
+ parameterDevice->neighborX,
+ parameterDevice->neighborY,
+ parameterDevice->neighborZ,
+ parameterDevice->numberOfNodes,
+ parameterDevice->isEvenTimestep);
+ getLastCudaError("QPressDeviceIncompNEQ27 execution failed");
}
//////////////////////////////////////////////////////////////////////////
void QPressDevNEQ27(LBMSimulationParameter* parameterDevice, QforBoundaryConditions* boundaryCondition)
{
- dim3 grid = vf::cuda::getCudaGrid( parameterDevice->numberofthreads, boundaryCondition->numberOfBCnodes);
- dim3 threads(parameterDevice->numberofthreads, 1, 1 );
+ dim3 grid = vf::cuda::getCudaGrid( parameterDevice->numberofthreads, boundaryCondition->numberOfBCnodes);
+ dim3 threads(parameterDevice->numberofthreads, 1, 1 );
- QPressDeviceNEQ27<<< grid, threads >>> (
+ QPressDeviceNEQ27<<< grid, threads >>> (
boundaryCondition->RhoBC,
parameterDevice->distributions.f[0],
boundaryCondition->k,
@@ -3907,3241 +3105,2743 @@ void QPressDevNEQ27(LBMSimulationParameter* parameterDevice, QforBoundaryConditi
parameterDevice->neighborZ,
parameterDevice->numberOfNodes,
parameterDevice->isEvenTimestep);
- getLastCudaError("QPressDevNEQ27 execution failed");
+ getLastCudaError("QPressDevNEQ27 execution failed");
}
//////////////////////////////////////////////////////////////////////////
void QPressDevEQZ27(LBMSimulationParameter* parameterDevice, QforBoundaryConditions* boundaryCondition)
{
- dim3 grid = vf::cuda::getCudaGrid( parameterDevice->numberofthreads, boundaryCondition->numberOfBCnodes);
- dim3 threads(parameterDevice->numberofthreads, 1, 1 );
+ dim3 grid = vf::cuda::getCudaGrid( parameterDevice->numberofthreads, boundaryCondition->numberOfBCnodes);
+ dim3 threads(parameterDevice->numberofthreads, 1, 1 );
- QPressDeviceEQZ27<<< grid, threads >>> (
- boundaryCondition->RhoBC,
- parameterDevice->distributions.f[0],
- boundaryCondition->k,
- boundaryCondition->kN,
- parameterDevice->kDistTestRE.f[0],
- boundaryCondition->numberOfBCnodes,
- parameterDevice->omega,
- parameterDevice->neighborX,
- parameterDevice->neighborY,
- parameterDevice->neighborZ,
- parameterDevice->numberOfNodes,
- parameterDevice->isEvenTimestep);
- getLastCudaError("QPressDeviceEQZ27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void QPressDevZero27(unsigned int numberOfThreads,
- real* DD,
- int* k_Q,
- unsigned int numberOfBCnodes,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep)
-{
- int Grid = (numberOfBCnodes / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 gridQ(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- QPressDeviceZero27<<< gridQ, threads >>> (DD,
- k_Q,
- numberOfBCnodes,
- neighborX,
- neighborY,
- neighborZ,
- size_Mat,
- isEvenTimestep);
- getLastCudaError("QPressDeviceOld27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void QPressDevFake27( unsigned int numberOfThreads,
- real* rhoBC,
- real* DD,
- int* k_Q,
- int* k_N,
- unsigned int numberOfBCnodes,
- real om1,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep)
-{
- int Grid = (numberOfBCnodes / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 gridQ(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- QPressDeviceFake27<<< gridQ, threads >>> (rhoBC,
- DD,
- k_Q,
- k_N,
- numberOfBCnodes,
- om1,
- neighborX,
- neighborY,
- neighborZ,
- size_Mat,
- isEvenTimestep);
- getLastCudaError("QPressDeviceFake27 execution failed");
+ QPressDeviceEQZ27<<< grid, threads >>> (
+ boundaryCondition->RhoBC,
+ parameterDevice->distributions.f[0],
+ boundaryCondition->k,
+ boundaryCondition->kN,
+ parameterDevice->kDistTestRE.f[0],
+ boundaryCondition->numberOfBCnodes,
+ parameterDevice->omega,
+ parameterDevice->neighborX,
+ parameterDevice->neighborY,
+ parameterDevice->neighborZ,
+ parameterDevice->numberOfNodes,
+ parameterDevice->isEvenTimestep);
+ getLastCudaError("QPressDeviceEQZ27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void QPressDevZero27(
+ unsigned int numberOfThreads,
+ real* DD,
+ int* k_Q,
+ unsigned int numberOfBCnodes,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, numberOfBCnodes);
+
+ QPressDeviceZero27<<< grid.grid, grid.threads >>> (
+ DD,
+ k_Q,
+ numberOfBCnodes,
+ neighborX,
+ neighborY,
+ neighborZ,
+ numberOfLBnodes,
+ isEvenTimestep);
+ getLastCudaError("QPressDeviceOld27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void QPressDevFake27(
+ unsigned int numberOfThreads,
+ real* rhoBC,
+ real* DD,
+ int* k_Q,
+ int* k_N,
+ unsigned int numberOfBCnodes,
+ real om1,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, numberOfBCnodes);
+
+
+ QPressDeviceFake27<<< grid.grid, grid.threads >>> (
+ rhoBC,
+ DD,
+ k_Q,
+ k_N,
+ numberOfBCnodes,
+ om1,
+ neighborX,
+ neighborY,
+ neighborZ,
+ numberOfLBnodes,
+ isEvenTimestep);
+ getLastCudaError("QPressDeviceFake27 execution failed");
}
//////////////////////////////////////////////////////////////////////////
void BBDev27(LBMSimulationParameter* parameterDevice, QforBoundaryConditions* boundaryCondition)
{
- dim3 grid = vf::cuda::getCudaGrid( parameterDevice->numberofthreads, boundaryCondition->numberOfBCnodes);
- dim3 threads(parameterDevice->numberofthreads, 1, 1 );
+ dim3 grid = vf::cuda::getCudaGrid( parameterDevice->numberofthreads, boundaryCondition->numberOfBCnodes);
+ dim3 threads(parameterDevice->numberofthreads, 1, 1 );
- BBDevice27<<< grid, threads >>> (
- parameterDevice->distributions.f[0],
- boundaryCondition->k,
- boundaryCondition->q27[0],
- boundaryCondition->numberOfBCnodes,
- parameterDevice->neighborX,
- parameterDevice->neighborY,
- parameterDevice->neighborZ,
- parameterDevice->numberOfNodes,
- parameterDevice->isEvenTimestep);
- getLastCudaError("BBDevice27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void QPressDev27_IntBB( unsigned int numberOfThreads,
- real* rho,
- real* DD,
- int* k_Q,
- real* QQ,
- unsigned int numberOfBCnodes,
- real om1,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep)
-{
- int Grid = (numberOfBCnodes / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 gridQ(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- QPressDevice27_IntBB<<< gridQ, threads >>> (rho,
- DD,
- k_Q,
- QQ,
- numberOfBCnodes,
- om1,
- neighborX,
- neighborY,
- neighborZ,
- size_Mat,
- isEvenTimestep);
- getLastCudaError("QPressDevice27_IntBB execution failed");
+ BBDevice27<<< grid, threads >>> (
+ parameterDevice->distributions.f[0],
+ boundaryCondition->k,
+ boundaryCondition->q27[0],
+ boundaryCondition->numberOfBCnodes,
+ parameterDevice->neighborX,
+ parameterDevice->neighborY,
+ parameterDevice->neighborZ,
+ parameterDevice->numberOfNodes,
+ parameterDevice->isEvenTimestep);
+ getLastCudaError("BBDevice27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void QPressDev27_IntBB(
+ unsigned int numberOfThreads,
+ real* rho,
+ real* DD,
+ int* k_Q,
+ real* QQ,
+ unsigned int numberOfBCnodes,
+ real om1,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, numberOfBCnodes);
+
+ QPressDevice27_IntBB<<< grid.grid, grid.threads >>> (
+ rho,
+ DD,
+ k_Q,
+ QQ,
+ numberOfBCnodes,
+ om1,
+ neighborX,
+ neighborY,
+ neighborZ,
+ numberOfLBnodes,
+ isEvenTimestep);
+ getLastCudaError("QPressDevice27_IntBB execution failed");
}
// TODO: https://git.rz.tu-bs.de/irmb/VirtualFluids_dev/-/issues/29
//////////////////////////////////////////////////////////////////////////
-void PressSchlaffer27(unsigned int numberOfThreads,
- real* rhoBC,
- real* DD,
- real* vx0,
- real* vy0,
- real* vz0,
- real* deltaVz0,
- int* k_Q,
- int* k_N,
- int numberOfBCnodes,
- real om1,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep)
-{
- int Grid = (numberOfBCnodes / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 gridQ(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- PressSchlaff27<<< gridQ, threads >>>( rhoBC,
- DD,
- vx0,
- vy0,
- vz0,
- deltaVz0,
- k_Q,
- k_N,
- numberOfBCnodes,
- om1,
- neighborX,
- neighborY,
- neighborZ,
- size_Mat,
- isEvenTimestep);
- getLastCudaError("PressSchlaff27 execution failed");
+void PressSchlaffer27(
+ unsigned int numberOfThreads,
+ real* rhoBC,
+ real* DD,
+ real* vx0,
+ real* vy0,
+ real* vz0,
+ real* deltaVz0,
+ int* k_Q,
+ int* k_N,
+ int numberOfBCnodes,
+ real om1,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, numberOfBCnodes);
+
+ PressSchlaff27<<< grid.grid, grid.threads >>>(
+ rhoBC,
+ DD,
+ vx0,
+ vy0,
+ vz0,
+ deltaVz0,
+ k_Q,
+ k_N,
+ numberOfBCnodes,
+ om1,
+ neighborX,
+ neighborY,
+ neighborZ,
+ numberOfLBnodes,
+ isEvenTimestep);
+ getLastCudaError("PressSchlaff27 execution failed");
}
// TODO: https://git.rz.tu-bs.de/irmb/VirtualFluids_dev/-/issues/29
//////////////////////////////////////////////////////////////////////////
-void VelSchlaffer27( unsigned int numberOfThreads,
- int t,
- real* DD,
- real* vz0,
- real* deltaVz0,
- int* k_Q,
- int* k_N,
- int numberOfBCnodes,
- real om1,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep)
-{
- int Grid = (numberOfBCnodes / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 gridQ(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- VelSchlaff27<<< gridQ, threads >>>( t,
- DD,
- vz0,
- deltaVz0,
- k_Q,
- k_N,
- numberOfBCnodes,
- om1,
- neighborX,
- neighborY,
- neighborZ,
- size_Mat,
- isEvenTimestep);
- getLastCudaError("VelSchlaff27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void PropVelo( unsigned int numberOfThreads,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- real* rho,
- real* ux,
- real* uy,
- real* uz,
- int* k_Q,
- unsigned int size_Prop,
- unsigned int size_Mat,
- unsigned int* bcMatD,
- real* DD,
- bool EvenOrOdd)
-{
- int Grid = (size_Prop / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 grid(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- PropellerBC<<< grid, threads >>>(neighborX,
- neighborY,
- neighborZ,
- rho,
- ux,
- uy,
- uz,
- k_Q,
- size_Prop,
- size_Mat,
- bcMatD,
- DD,
- EvenOrOdd);
- getLastCudaError("PropellerBC execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void ScaleCF27( real* DC,
- real* DF,
- unsigned int* neighborCX,
- unsigned int* neighborCY,
- unsigned int* neighborCZ,
- unsigned int* neighborFX,
- unsigned int* neighborFY,
- unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
- bool isEvenTimestep,
- unsigned int* posCSWB,
- unsigned int* posFSWB,
- unsigned int kCF,
- real omCoarse,
- real omFine,
- real nu,
- unsigned int nxC,
- unsigned int nyC,
- unsigned int nxF,
- unsigned int nyF,
- unsigned int numberOfThreads)
-{
- int Grid = (kCF / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 gridINT_CF(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- scaleCF27<<< gridINT_CF, threads >>> ( DC,
- DF,
- neighborCX,
- neighborCY,
- neighborCZ,
- neighborFX,
- neighborFY,
- neighborFZ,
- size_MatC,
- size_MatF,
- isEvenTimestep,
- posCSWB,
- posFSWB,
- kCF,
- omCoarse,
- omFine,
- nu,
- nxC,
- nyC,
- nxF,
- nyF);
- getLastCudaError("scaleCF27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void ScaleCFEff27(real* DC,
- real* DF,
- unsigned int* neighborCX,
- unsigned int* neighborCY,
- unsigned int* neighborCZ,
- unsigned int* neighborFX,
- unsigned int* neighborFY,
- unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
- bool isEvenTimestep,
- unsigned int* posCSWB,
- unsigned int* posFSWB,
- unsigned int kCF,
- real omCoarse,
- real omFine,
- real nu,
- unsigned int nxC,
- unsigned int nyC,
- unsigned int nxF,
- unsigned int nyF,
- unsigned int numberOfThreads,
- OffCF offCF)
-{
- int Grid = (kCF / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 gridINT_CF(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- scaleCFEff27<<< gridINT_CF, threads >>> ( DC,
- DF,
- neighborCX,
- neighborCY,
- neighborCZ,
- neighborFX,
- neighborFY,
- neighborFZ,
- size_MatC,
- size_MatF,
- isEvenTimestep,
- posCSWB,
- posFSWB,
- kCF,
- omCoarse,
- omFine,
- nu,
- nxC,
- nyC,
- nxF,
- nyF,
- offCF);
- getLastCudaError("scaleCFEff27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void ScaleCFLast27(real* DC,
- real* DF,
- unsigned int* neighborCX,
- unsigned int* neighborCY,
- unsigned int* neighborCZ,
- unsigned int* neighborFX,
- unsigned int* neighborFY,
- unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
- bool isEvenTimestep,
- unsigned int* posCSWB,
- unsigned int* posFSWB,
- unsigned int kCF,
- real omCoarse,
- real omFine,
- real nu,
- unsigned int nxC,
- unsigned int nyC,
- unsigned int nxF,
- unsigned int nyF,
- unsigned int numberOfThreads,
- OffCF offCF)
-{
- int Grid = (kCF / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 gridINT_CF(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- scaleCFLast27<<< gridINT_CF, threads >>> (DC,
- DF,
- neighborCX,
- neighborCY,
- neighborCZ,
- neighborFX,
- neighborFY,
- neighborFZ,
- size_MatC,
- size_MatF,
- isEvenTimestep,
- posCSWB,
- posFSWB,
- kCF,
- omCoarse,
- omFine,
- nu,
- nxC,
- nyC,
- nxF,
- nyF,
- offCF);
- getLastCudaError("scaleCFLast27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void ScaleCFpress27( real* DC,
- real* DF,
- unsigned int* neighborCX,
- unsigned int* neighborCY,
- unsigned int* neighborCZ,
- unsigned int* neighborFX,
- unsigned int* neighborFY,
- unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
- bool isEvenTimestep,
- unsigned int* posCSWB,
- unsigned int* posFSWB,
- unsigned int kCF,
- real omCoarse,
- real omFine,
- real nu,
- unsigned int nxC,
- unsigned int nyC,
- unsigned int nxF,
- unsigned int nyF,
- unsigned int numberOfThreads,
- OffCF offCF)
-{
- int Grid = (kCF / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 gridINT_CF(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- scaleCFpress27<<< gridINT_CF, threads >>>(DC,
- DF,
- neighborCX,
- neighborCY,
- neighborCZ,
- neighborFX,
- neighborFY,
- neighborFZ,
- size_MatC,
- size_MatF,
- isEvenTimestep,
- posCSWB,
- posFSWB,
- kCF,
- omCoarse,
- omFine,
- nu,
- nxC,
- nyC,
- nxF,
- nyF,
- offCF);
- getLastCudaError("scaleCFpress27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void ScaleCF_Fix_27( real* DC,
- real* DF,
- unsigned int* neighborCX,
- unsigned int* neighborCY,
- unsigned int* neighborCZ,
- unsigned int* neighborFX,
- unsigned int* neighborFY,
- unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
- bool isEvenTimestep,
- unsigned int* posCSWB,
- unsigned int* posFSWB,
- unsigned int kCF,
- real omCoarse,
- real omFine,
- real nu,
- unsigned int nxC,
- unsigned int nyC,
- unsigned int nxF,
- unsigned int nyF,
- unsigned int numberOfThreads,
- OffCF offCF)
-{
- int Grid = (kCF / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 gridINT_CF(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- scaleCF_Fix_27<<< gridINT_CF, threads >>>(DC,
- DF,
- neighborCX,
- neighborCY,
- neighborCZ,
- neighborFX,
- neighborFY,
- neighborFZ,
- size_MatC,
- size_MatF,
- isEvenTimestep,
- posCSWB,
- posFSWB,
- kCF,
- omCoarse,
- omFine,
- nu,
- nxC,
- nyC,
- nxF,
- nyF,
- offCF);
- getLastCudaError("scaleCF_Fix_27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void ScaleCF_Fix_comp_27( real* DC,
- real* DF,
- unsigned int* neighborCX,
- unsigned int* neighborCY,
- unsigned int* neighborCZ,
- unsigned int* neighborFX,
- unsigned int* neighborFY,
- unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
- bool isEvenTimestep,
- unsigned int* posCSWB,
- unsigned int* posFSWB,
- unsigned int kCF,
- real omCoarse,
- real omFine,
- real nu,
- unsigned int nxC,
- unsigned int nyC,
- unsigned int nxF,
- unsigned int nyF,
- unsigned int numberOfThreads,
- OffCF offCF)
-{
- int Grid = (kCF / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 gridINT_CF(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- scaleCF_Fix_comp_27<<< gridINT_CF, threads >>>( DC,
- DF,
- neighborCX,
- neighborCY,
- neighborCZ,
- neighborFX,
- neighborFY,
- neighborFZ,
- size_MatC,
- size_MatF,
- isEvenTimestep,
- posCSWB,
- posFSWB,
- kCF,
- omCoarse,
- omFine,
- nu,
- nxC,
- nyC,
- nxF,
- nyF,
- offCF);
- getLastCudaError("scaleCF_Fix_27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void ScaleCF_0817_comp_27(real* DC,
- real* DF,
- unsigned int* neighborCX,
- unsigned int* neighborCY,
- unsigned int* neighborCZ,
- unsigned int* neighborFX,
- unsigned int* neighborFY,
- unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
- bool isEvenTimestep,
- unsigned int* posCSWB,
- unsigned int* posFSWB,
- unsigned int kCF,
- real omCoarse,
- real omFine,
- real nu,
- unsigned int nxC,
- unsigned int nyC,
- unsigned int nxF,
- unsigned int nyF,
- unsigned int numberOfThreads,
- OffCF offCF,
- CUstream_st *stream)
-{
- int Grid = (kCF / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 gridINT_CF(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- scaleCF_0817_comp_27<<< gridINT_CF, threads, 0, stream >>>( DC,
- DF,
- neighborCX,
- neighborCY,
- neighborCZ,
- neighborFX,
- neighborFY,
- neighborFZ,
- size_MatC,
- size_MatF,
- isEvenTimestep,
- posCSWB,
- posFSWB,
- kCF,
- omCoarse,
- omFine,
- nu,
- nxC,
- nyC,
- nxF,
- nyF,
- offCF);
- getLastCudaError("scaleCF_0817_27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void ScaleCF_comp_D3Q27F3_2018(real* DC,
- real* DF,
- real* G6,
- unsigned int* neighborCX,
- unsigned int* neighborCY,
- unsigned int* neighborCZ,
- unsigned int* neighborFX,
- unsigned int* neighborFY,
- unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
- bool isEvenTimestep,
- unsigned int* posCSWB,
- unsigned int* posFSWB,
- unsigned int kCF,
- real omCoarse,
- real omFine,
- real nu,
- unsigned int nxC,
- unsigned int nyC,
- unsigned int nxF,
- unsigned int nyF,
- unsigned int numberOfThreads,
- OffCF offCF)
-{
- int Grid = (kCF / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 gridINT_CF(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- scaleCF_comp_D3Q27F3_2018 <<< gridINT_CF, threads >>>(DC,
- DF,
- G6,
- neighborCX,
- neighborCY,
- neighborCZ,
- neighborFX,
- neighborFY,
- neighborFZ,
- size_MatC,
- size_MatF,
- isEvenTimestep,
- posCSWB,
- posFSWB,
- kCF,
- omCoarse,
- omFine,
- nu,
- nxC,
- nyC,
- nxF,
- nyF,
- offCF);
- getLastCudaError("scaleCF_comp_D3Q27F3_2018 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void ScaleCF_comp_D3Q27F3(real* DC,
- real* DF,
- real* G6,
- unsigned int* neighborCX,
- unsigned int* neighborCY,
- unsigned int* neighborCZ,
- unsigned int* neighborFX,
- unsigned int* neighborFY,
- unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
- bool isEvenTimestep,
- unsigned int* posCSWB,
- unsigned int* posFSWB,
- unsigned int kCF,
- real omCoarse,
- real omFine,
- real nu,
- unsigned int nxC,
- unsigned int nyC,
- unsigned int nxF,
- unsigned int nyF,
- unsigned int numberOfThreads,
- OffCF offCF,
- CUstream_st *stream)
-{
- int Grid = (kCF / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 gridINT_CF(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- scaleCF_comp_D3Q27F3 <<< gridINT_CF, threads, 0, stream >>>( DC,
- DF,
- G6,
- neighborCX,
- neighborCY,
- neighborCZ,
- neighborFX,
- neighborFY,
- neighborFZ,
- size_MatC,
- size_MatF,
- isEvenTimestep,
- posCSWB,
- posFSWB,
- kCF,
- omCoarse,
- omFine,
- nu,
- nxC,
- nyC,
- nxF,
- nyF,
- offCF);
- getLastCudaError("scaleCF_comp_D3Q27F3 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void ScaleCF_staggered_time_comp_27( real* DC,
- real* DF,
- unsigned int* neighborCX,
- unsigned int* neighborCY,
- unsigned int* neighborCZ,
- unsigned int* neighborFX,
- unsigned int* neighborFY,
- unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
- bool isEvenTimestep,
- unsigned int* posCSWB,
- unsigned int* posFSWB,
- unsigned int kCF,
- real omCoarse,
- real omFine,
- real nu,
- unsigned int nxC,
- unsigned int nyC,
- unsigned int nxF,
- unsigned int nyF,
- unsigned int numberOfThreads,
- OffCF offCF)
-{
- int Grid = (kCF / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 gridINT_CF(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- scaleCF_staggered_time_comp_27<<< gridINT_CF, threads >>>( DC,
- DF,
- neighborCX,
- neighborCY,
- neighborCZ,
- neighborFX,
- neighborFY,
- neighborFZ,
- size_MatC,
- size_MatF,
- isEvenTimestep,
- posCSWB,
- posFSWB,
- kCF,
- omCoarse,
- omFine,
- nu,
- nxC,
- nyC,
- nxF,
- nyF,
- offCF);
- getLastCudaError("scaleCF_Fix_27 execution failed");
+void VelSchlaffer27(
+ unsigned int numberOfThreads,
+ int t,
+ real* DD,
+ real* vz0,
+ real* deltaVz0,
+ int* k_Q,
+ int* k_N,
+ int numberOfBCnodes,
+ real om1,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, numberOfBCnodes);
+
+ VelSchlaff27<<< grid.grid, grid.threads >>>(
+ t,
+ DD,
+ vz0,
+ deltaVz0,
+ k_Q,
+ k_N,
+ numberOfBCnodes,
+ om1,
+ neighborX,
+ neighborY,
+ neighborZ,
+ numberOfLBnodes,
+ isEvenTimestep);
+ getLastCudaError("VelSchlaff27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void QPrecursorDevCompZeroPress(LBMSimulationParameter* parameterDevice,
+ QforPrecursorBoundaryConditions* boundaryCondition,
+ real timeRatio,
+ real velocityRatio)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(parameterDevice->numberofthreads, boundaryCondition->numberOfBCnodes);
+
+ QPrecursorDeviceCompZeroPress<<< grid.grid, grid.threads >>>(
+ boundaryCondition->k,
+ boundaryCondition->numberOfBCnodes,
+ boundaryCondition->numberOfPrecursorNodes,
+ boundaryCondition->sizeQ,
+ parameterDevice->omega,
+ parameterDevice->distributions.f[0],
+ boundaryCondition->q27[0],
+ parameterDevice->neighborX,
+ parameterDevice->neighborY,
+ parameterDevice->neighborZ,
+ boundaryCondition->planeNeighbor0PP,
+ boundaryCondition->planeNeighbor0PM,
+ boundaryCondition->planeNeighbor0MP,
+ boundaryCondition->planeNeighbor0MM,
+ boundaryCondition->weights0PP,
+ boundaryCondition->weights0PM,
+ boundaryCondition->weights0MP,
+ boundaryCondition->weights0MM,
+ boundaryCondition->last,
+ boundaryCondition->current,
+ boundaryCondition->velocityX,
+ boundaryCondition->velocityY,
+ boundaryCondition->velocityZ,
+ timeRatio,
+ velocityRatio,
+ parameterDevice->numberOfNodes,
+ parameterDevice->isEvenTimestep);
+ getLastCudaError("QPrecursorDeviceCompZeroPress execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void PrecursorDevEQ27( LBMSimulationParameter* parameterDevice,
+ QforPrecursorBoundaryConditions* boundaryCondition,
+ real timeRatio,
+ real velocityRatio)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(parameterDevice->numberofthreads, boundaryCondition->numberOfBCnodes);
+
+ PrecursorDeviceEQ27<<< grid.grid, grid.threads >>>(
+ boundaryCondition->k,
+ boundaryCondition->numberOfBCnodes,
+ boundaryCondition->numberOfPrecursorNodes,
+ parameterDevice->omega,
+ parameterDevice->distributions.f[0],
+ parameterDevice->neighborX,
+ parameterDevice->neighborX,
+ parameterDevice->neighborX,
+ boundaryCondition->planeNeighbor0PP,
+ boundaryCondition->planeNeighbor0PM,
+ boundaryCondition->planeNeighbor0MP,
+ boundaryCondition->planeNeighbor0MM,
+ boundaryCondition->weights0PP,
+ boundaryCondition->weights0PM,
+ boundaryCondition->weights0MP,
+ boundaryCondition->weights0MM,
+ boundaryCondition->last,
+ boundaryCondition->current,
+ boundaryCondition->velocityX,
+ boundaryCondition->velocityY,
+ boundaryCondition->velocityZ,
+ timeRatio,
+ velocityRatio,
+ parameterDevice->numberOfNodes,
+ parameterDevice->isEvenTimestep);
+ getLastCudaError("PrecursorDeviceEQ27 execution failed");
+
+}
+//////////////////////////////////////////////////////////////////////////
+void PrecursorDevDistributions( LBMSimulationParameter* parameterDevice,
+ QforPrecursorBoundaryConditions* boundaryCondition,
+ real timeRatio,
+ real velocityRatio)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(parameterDevice->numberofthreads, boundaryCondition->numberOfBCnodes);
+
+ PrecursorDeviceDistributions<<< grid.grid, grid.threads >>>(
+ boundaryCondition->k,
+ boundaryCondition->numberOfBCnodes,
+ boundaryCondition->numberOfPrecursorNodes,
+ parameterDevice->distributions.f[0],
+ parameterDevice->neighborX,
+ parameterDevice->neighborY,
+ parameterDevice->neighborZ,
+ boundaryCondition->planeNeighbor0PP,
+ boundaryCondition->planeNeighbor0PM,
+ boundaryCondition->planeNeighbor0MP,
+ boundaryCondition->planeNeighbor0MM,
+ boundaryCondition->weights0PP,
+ boundaryCondition->weights0PM,
+ boundaryCondition->weights0MP,
+ boundaryCondition->weights0MM,
+ boundaryCondition->last,
+ boundaryCondition->current,
+ timeRatio,
+ parameterDevice->numberOfNodes,
+ parameterDevice->isEvenTimestep);
+ getLastCudaError("PrecursorDeviceDistributions execution failed");
+
+}
+
+//////////////////////////////////////////////////////////////////////////
+void QPrecursorDevDistributions( LBMSimulationParameter* parameterDevice,
+ QforPrecursorBoundaryConditions* boundaryCondition,
+ real timeRatio,
+ real velocityRatio)
+{
+
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(parameterDevice->numberofthreads, boundaryCondition->numberOfBCnodes);
+
+ QPrecursorDeviceDistributions<<< grid.grid, grid.threads >>>(
+ boundaryCondition->k,
+ boundaryCondition->q27[0],
+ boundaryCondition->sizeQ,
+ boundaryCondition->numberOfBCnodes,
+ boundaryCondition->numberOfPrecursorNodes,
+ parameterDevice->distributions.f[0],
+ parameterDevice->neighborX,
+ parameterDevice->neighborY,
+ parameterDevice->neighborZ,
+ boundaryCondition->planeNeighbor0PP,
+ boundaryCondition->planeNeighbor0PM,
+ boundaryCondition->planeNeighbor0MP,
+ boundaryCondition->planeNeighbor0MM,
+ boundaryCondition->weights0PP,
+ boundaryCondition->weights0PM,
+ boundaryCondition->weights0MP,
+ boundaryCondition->weights0MM,
+ boundaryCondition->last,
+ boundaryCondition->current,
+ timeRatio,
+ parameterDevice->numberOfNodes,
+ parameterDevice->isEvenTimestep);
+ getLastCudaError("QPrecursorDeviceCompZeroPress execution failed");
+
+}
+//////////////////////////////////////////////////////////////////////////
+extern "C" void PropVelo(
+ unsigned int numberOfThreads,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ real* rho,
+ real* ux,
+ real* uy,
+ real* uz,
+ int* k_Q,
+ unsigned int size_Prop,
+ unsigned long long numberOfLBnodes,
+ unsigned int* bcMatD,
+ real* DD,
+ bool EvenOrOdd)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, size_Prop);
+
+ PropellerBC<<< grid.grid, grid.threads >>>(
+ neighborX,
+ neighborY,
+ neighborZ,
+ rho,
+ ux,
+ uy,
+ uz,
+ k_Q,
+ size_Prop,
+ numberOfLBnodes,
+ bcMatD,
+ DD,
+ EvenOrOdd);
+ getLastCudaError("PropellerBC execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void ScaleCF27(
+ real* DC,
+ real* DF,
+ unsigned int* neighborCX,
+ unsigned int* neighborCY,
+ unsigned int* neighborCZ,
+ unsigned int* neighborFX,
+ unsigned int* neighborFY,
+ unsigned int* neighborFZ,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
+ bool isEvenTimestep,
+ unsigned int* posCSWB,
+ unsigned int* posFSWB,
+ unsigned int kCF,
+ real omCoarse,
+ real omFine,
+ real nu,
+ unsigned int nxC,
+ unsigned int nyC,
+ unsigned int nxF,
+ unsigned int nyF,
+ unsigned int numberOfThreads)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, kCF);
+
+ scaleCF27<<< grid.grid, grid.threads >>> (
+ DC,
+ DF,
+ neighborCX,
+ neighborCY,
+ neighborCZ,
+ neighborFX,
+ neighborFY,
+ neighborFZ,
+ numberOfLBnodesC,
+ numberOfLBnodesF,
+ isEvenTimestep,
+ posCSWB,
+ posFSWB,
+ kCF,
+ omCoarse,
+ omFine,
+ nu,
+ nxC,
+ nyC,
+ nxF,
+ nyF);
+ getLastCudaError("scaleCF27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void ScaleCFEff27(
+ real* DC,
+ real* DF,
+ unsigned int* neighborCX,
+ unsigned int* neighborCY,
+ unsigned int* neighborCZ,
+ unsigned int* neighborFX,
+ unsigned int* neighborFY,
+ unsigned int* neighborFZ,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
+ bool isEvenTimestep,
+ unsigned int* posCSWB,
+ unsigned int* posFSWB,
+ unsigned int kCF,
+ real omCoarse,
+ real omFine,
+ real nu,
+ unsigned int nxC,
+ unsigned int nyC,
+ unsigned int nxF,
+ unsigned int nyF,
+ unsigned int numberOfThreads,
+ OffCF offCF)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, kCF);
+
+ scaleCFEff27<<< grid.grid, grid.threads >>> (
+ DC,
+ DF,
+ neighborCX,
+ neighborCY,
+ neighborCZ,
+ neighborFX,
+ neighborFY,
+ neighborFZ,
+ numberOfLBnodesC,
+ numberOfLBnodesF,
+ isEvenTimestep,
+ posCSWB,
+ posFSWB,
+ kCF,
+ omCoarse,
+ omFine,
+ nu,
+ nxC,
+ nyC,
+ nxF,
+ nyF,
+ offCF);
+ getLastCudaError("scaleCFEff27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void ScaleCFLast27(
+ real* DC,
+ real* DF,
+ unsigned int* neighborCX,
+ unsigned int* neighborCY,
+ unsigned int* neighborCZ,
+ unsigned int* neighborFX,
+ unsigned int* neighborFY,
+ unsigned int* neighborFZ,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
+ bool isEvenTimestep,
+ unsigned int* posCSWB,
+ unsigned int* posFSWB,
+ unsigned int kCF,
+ real omCoarse,
+ real omFine,
+ real nu,
+ unsigned int nxC,
+ unsigned int nyC,
+ unsigned int nxF,
+ unsigned int nyF,
+ unsigned int numberOfThreads,
+ OffCF offCF)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, kCF);
+
+ scaleCFLast27<<< grid.grid, grid.threads >>> (
+ DC,
+ DF,
+ neighborCX,
+ neighborCY,
+ neighborCZ,
+ neighborFX,
+ neighborFY,
+ neighborFZ,
+ numberOfLBnodesC,
+ numberOfLBnodesF,
+ isEvenTimestep,
+ posCSWB,
+ posFSWB,
+ kCF,
+ omCoarse,
+ omFine,
+ nu,
+ nxC,
+ nyC,
+ nxF,
+ nyF,
+ offCF);
+ getLastCudaError("scaleCFLast27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void ScaleCFpress27(
+ real* DC,
+ real* DF,
+ unsigned int* neighborCX,
+ unsigned int* neighborCY,
+ unsigned int* neighborCZ,
+ unsigned int* neighborFX,
+ unsigned int* neighborFY,
+ unsigned int* neighborFZ,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
+ bool isEvenTimestep,
+ unsigned int* posCSWB,
+ unsigned int* posFSWB,
+ unsigned int kCF,
+ real omCoarse,
+ real omFine,
+ real nu,
+ unsigned int nxC,
+ unsigned int nyC,
+ unsigned int nxF,
+ unsigned int nyF,
+ unsigned int numberOfThreads,
+ OffCF offCF)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, kCF);
+
+ scaleCFpress27<<< grid.grid, grid.threads >>>(
+ DC,
+ DF,
+ neighborCX,
+ neighborCY,
+ neighborCZ,
+ neighborFX,
+ neighborFY,
+ neighborFZ,
+ numberOfLBnodesC,
+ numberOfLBnodesF,
+ isEvenTimestep,
+ posCSWB,
+ posFSWB,
+ kCF,
+ omCoarse,
+ omFine,
+ nu,
+ nxC,
+ nyC,
+ nxF,
+ nyF,
+ offCF);
+ getLastCudaError("scaleCFpress27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void ScaleCF_Fix_27(
+ real* DC,
+ real* DF,
+ unsigned int* neighborCX,
+ unsigned int* neighborCY,
+ unsigned int* neighborCZ,
+ unsigned int* neighborFX,
+ unsigned int* neighborFY,
+ unsigned int* neighborFZ,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
+ bool isEvenTimestep,
+ unsigned int* posCSWB,
+ unsigned int* posFSWB,
+ unsigned int kCF,
+ real omCoarse,
+ real omFine,
+ real nu,
+ unsigned int nxC,
+ unsigned int nyC,
+ unsigned int nxF,
+ unsigned int nyF,
+ unsigned int numberOfThreads,
+ OffCF offCF)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, kCF);
+
+ scaleCF_Fix_27<<< grid.grid, grid.threads >>>(
+ DC,
+ DF,
+ neighborCX,
+ neighborCY,
+ neighborCZ,
+ neighborFX,
+ neighborFY,
+ neighborFZ,
+ numberOfLBnodesC,
+ numberOfLBnodesF,
+ isEvenTimestep,
+ posCSWB,
+ posFSWB,
+ kCF,
+ omCoarse,
+ omFine,
+ nu,
+ nxC,
+ nyC,
+ nxF,
+ nyF,
+ offCF);
+ getLastCudaError("scaleCF_Fix_27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void ScaleCF_Fix_comp_27(
+ real* DC,
+ real* DF,
+ unsigned int* neighborCX,
+ unsigned int* neighborCY,
+ unsigned int* neighborCZ,
+ unsigned int* neighborFX,
+ unsigned int* neighborFY,
+ unsigned int* neighborFZ,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
+ bool isEvenTimestep,
+ unsigned int* posCSWB,
+ unsigned int* posFSWB,
+ unsigned int kCF,
+ real omCoarse,
+ real omFine,
+ real nu,
+ unsigned int nxC,
+ unsigned int nyC,
+ unsigned int nxF,
+ unsigned int nyF,
+ unsigned int numberOfThreads,
+ OffCF offCF)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, kCF);
+
+ scaleCF_Fix_comp_27<<< grid.grid, grid.threads >>>(
+ DC,
+ DF,
+ neighborCX,
+ neighborCY,
+ neighborCZ,
+ neighborFX,
+ neighborFY,
+ neighborFZ,
+ numberOfLBnodesC,
+ numberOfLBnodesF,
+ isEvenTimestep,
+ posCSWB,
+ posFSWB,
+ kCF,
+ omCoarse,
+ omFine,
+ nu,
+ nxC,
+ nyC,
+ nxF,
+ nyF,
+ offCF);
+ getLastCudaError("scaleCF_Fix_comp_27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void ScaleCF_0817_comp_27(
+ real* DC,
+ real* DF,
+ unsigned int* neighborCX,
+ unsigned int* neighborCY,
+ unsigned int* neighborCZ,
+ unsigned int* neighborFX,
+ unsigned int* neighborFY,
+ unsigned int* neighborFZ,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
+ bool isEvenTimestep,
+ unsigned int* posCSWB,
+ unsigned int* posFSWB,
+ unsigned int kCF,
+ real omCoarse,
+ real omFine,
+ real nu,
+ unsigned int nxC,
+ unsigned int nyC,
+ unsigned int nxF,
+ unsigned int nyF,
+ unsigned int numberOfThreads,
+ OffCF offCF,
+ CUstream_st *stream)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, kCF);
+
+ scaleCF_0817_comp_27<<< grid.grid, grid.threads, 0, stream >>>(
+ DC,
+ DF,
+ neighborCX,
+ neighborCY,
+ neighborCZ,
+ neighborFX,
+ neighborFY,
+ neighborFZ,
+ numberOfLBnodesC,
+ numberOfLBnodesF,
+ isEvenTimestep,
+ posCSWB,
+ posFSWB,
+ kCF,
+ omCoarse,
+ omFine,
+ nu,
+ nxC,
+ nyC,
+ nxF,
+ nyF,
+ offCF);
+ getLastCudaError("scaleCF_0817_27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void ScaleCF_comp_D3Q27F3_2018(
+ real* DC,
+ real* DF,
+ real* G6,
+ unsigned int* neighborCX,
+ unsigned int* neighborCY,
+ unsigned int* neighborCZ,
+ unsigned int* neighborFX,
+ unsigned int* neighborFY,
+ unsigned int* neighborFZ,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
+ bool isEvenTimestep,
+ unsigned int* posCSWB,
+ unsigned int* posFSWB,
+ unsigned int kCF,
+ real omCoarse,
+ real omFine,
+ real nu,
+ unsigned int nxC,
+ unsigned int nyC,
+ unsigned int nxF,
+ unsigned int nyF,
+ unsigned int numberOfThreads,
+ OffCF offCF)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, kCF);
+
+ scaleCF_comp_D3Q27F3_2018 <<< grid.grid, grid.threads >>>(
+ DC,
+ DF,
+ G6,
+ neighborCX,
+ neighborCY,
+ neighborCZ,
+ neighborFX,
+ neighborFY,
+ neighborFZ,
+ numberOfLBnodesC,
+ numberOfLBnodesF,
+ isEvenTimestep,
+ posCSWB,
+ posFSWB,
+ kCF,
+ omCoarse,
+ omFine,
+ nu,
+ nxC,
+ nyC,
+ nxF,
+ nyF,
+ offCF);
+ getLastCudaError("scaleCF_comp_D3Q27F3_2018 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void ScaleCF_comp_D3Q27F3(
+ real* DC,
+ real* DF,
+ real* G6,
+ unsigned int* neighborCX,
+ unsigned int* neighborCY,
+ unsigned int* neighborCZ,
+ unsigned int* neighborFX,
+ unsigned int* neighborFY,
+ unsigned int* neighborFZ,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
+ bool isEvenTimestep,
+ unsigned int* posCSWB,
+ unsigned int* posFSWB,
+ unsigned int kCF,
+ real omCoarse,
+ real omFine,
+ real nu,
+ unsigned int nxC,
+ unsigned int nyC,
+ unsigned int nxF,
+ unsigned int nyF,
+ unsigned int numberOfThreads,
+ OffCF offCF,
+ CUstream_st *stream)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, kCF);
+
+ scaleCF_comp_D3Q27F3 <<< grid.grid, grid.threads, 0, stream >>>(
+ DC,
+ DF,
+ G6,
+ neighborCX,
+ neighborCY,
+ neighborCZ,
+ neighborFX,
+ neighborFY,
+ neighborFZ,
+ numberOfLBnodesC,
+ numberOfLBnodesF,
+ isEvenTimestep,
+ posCSWB,
+ posFSWB,
+ kCF,
+ omCoarse,
+ omFine,
+ nu,
+ nxC,
+ nyC,
+ nxF,
+ nyF,
+ offCF);
+ getLastCudaError("scaleCF_comp_D3Q27F3 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void ScaleCF_staggered_time_comp_27(
+ real* DC,
+ real* DF,
+ unsigned int* neighborCX,
+ unsigned int* neighborCY,
+ unsigned int* neighborCZ,
+ unsigned int* neighborFX,
+ unsigned int* neighborFY,
+ unsigned int* neighborFZ,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
+ bool isEvenTimestep,
+ unsigned int* posCSWB,
+ unsigned int* posFSWB,
+ unsigned int kCF,
+ real omCoarse,
+ real omFine,
+ real nu,
+ unsigned int nxC,
+ unsigned int nyC,
+ unsigned int nxF,
+ unsigned int nyF,
+ unsigned int numberOfThreads,
+ OffCF offCF)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, kCF);
+
+ scaleCF_staggered_time_comp_27<<< grid.grid, grid.threads >>>(
+ DC,
+ DF,
+ neighborCX,
+ neighborCY,
+ neighborCZ,
+ neighborFX,
+ neighborFY,
+ neighborFZ,
+ numberOfLBnodesC,
+ numberOfLBnodesF,
+ isEvenTimestep,
+ posCSWB,
+ posFSWB,
+ kCF,
+ omCoarse,
+ omFine,
+ nu,
+ nxC,
+ nyC,
+ nxF,
+ nyF,
+ offCF);
+ getLastCudaError("scaleCF_staggered_time_comp_27 execution failed");
}
//////////////////////////////////////////////////////////////////////////
void ScaleCF_RhoSq_comp_27(LBMSimulationParameter * parameterDeviceC, LBMSimulationParameter* parameterDeviceF, ICellCF * icellCF, OffCF& offsetCF, CUstream_st *stream)
{
- dim3 grid = vf::cuda::getCudaGrid(parameterDeviceC->numberofthreads, icellCF->kCF);
- dim3 threads(parameterDeviceC->numberofthreads, 1, 1 );
-
- scaleCF_RhoSq_comp_27<<<grid, threads, 0, stream>>>(
- parameterDeviceC->distributions.f[0],
- parameterDeviceF->distributions.f[0],
- parameterDeviceC->neighborX,
- parameterDeviceC->neighborY,
- parameterDeviceC->neighborZ,
- parameterDeviceF->neighborX,
- parameterDeviceF->neighborY,
- parameterDeviceF->neighborZ,
- parameterDeviceC->numberOfNodes,
- parameterDeviceF->numberOfNodes,
- parameterDeviceC->isEvenTimestep,
- icellCF->ICellCFC,
- icellCF->ICellCFF,
- icellCF->kCF,
- parameterDeviceC->omega,
- parameterDeviceF->omega,
- parameterDeviceC->vis,
- parameterDeviceC->nx,
- parameterDeviceC->ny,
- parameterDeviceF->nx,
- parameterDeviceF->ny,
- offsetCF);
- getLastCudaError("scaleCF_RhoSq_27 execution failed");
+ dim3 grid = vf::cuda::getCudaGrid(parameterDeviceC->numberofthreads, icellCF->kCF);
+ dim3 threads(parameterDeviceC->numberofthreads, 1, 1 );
+
+ scaleCF_RhoSq_comp_27<<<grid, threads, 0, stream>>>(
+ parameterDeviceC->distributions.f[0],
+ parameterDeviceF->distributions.f[0],
+ parameterDeviceC->neighborX,
+ parameterDeviceC->neighborY,
+ parameterDeviceC->neighborZ,
+ parameterDeviceF->neighborX,
+ parameterDeviceF->neighborY,
+ parameterDeviceF->neighborZ,
+ parameterDeviceC->numberOfNodes,
+ parameterDeviceF->numberOfNodes,
+ parameterDeviceC->isEvenTimestep,
+ icellCF->ICellCFC,
+ icellCF->ICellCFF,
+ icellCF->kCF,
+ parameterDeviceC->omega,
+ parameterDeviceF->omega,
+ parameterDeviceC->vis,
+ parameterDeviceC->nx,
+ parameterDeviceC->ny,
+ parameterDeviceF->nx,
+ parameterDeviceF->ny,
+ offsetCF);
+ getLastCudaError("scaleCF_RhoSq_27 execution failed");
}
void ScaleCF_compressible(LBMSimulationParameter * parameterDeviceC, LBMSimulationParameter* parameterDeviceF, ICellCF * icellCF, OffCF& offsetCF, CUstream_st *stream)
{
- dim3 grid = vf::cuda::getCudaGrid(parameterDeviceC->numberofthreads, icellCF->kCF);
- dim3 threads(parameterDeviceC->numberofthreads, 1, 1 );
-
- scaleCF_compressible<<<grid, threads, 0, stream>>>(
- parameterDeviceC->distributions.f[0],
- parameterDeviceF->distributions.f[0],
- parameterDeviceC->neighborX,
- parameterDeviceC->neighborY,
- parameterDeviceC->neighborZ,
- parameterDeviceF->neighborX,
- parameterDeviceF->neighborY,
- parameterDeviceF->neighborZ,
- parameterDeviceC->numberOfNodes,
- parameterDeviceF->numberOfNodes,
- parameterDeviceC->isEvenTimestep,
- icellCF->ICellCFC,
- icellCF->ICellCFF,
- icellCF->kCF,
- parameterDeviceC->omega,
- parameterDeviceF->omega,
- offsetCF);
- getLastCudaError("scaleCF_compressible execution failed");
-}
-
-//////////////////////////////////////////////////////////////////////////
-void ScaleCF_RhoSq_3rdMom_comp_27(real* DC,
- real* DF,
- unsigned int* neighborCX,
- unsigned int* neighborCY,
- unsigned int* neighborCZ,
- unsigned int* neighborFX,
- unsigned int* neighborFY,
- unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
- bool isEvenTimestep,
- unsigned int* posCSWB,
- unsigned int* posFSWB,
- unsigned int kCF,
- real omCoarse,
- real omFine,
- real nu,
- unsigned int nxC,
- unsigned int nyC,
- unsigned int nxF,
- unsigned int nyF,
- unsigned int numberOfThreads,
- OffCF offCF,
- CUstream_st *stream)
-{
- int Grid = (kCF / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 gridINT_CF(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- scaleCF_RhoSq_3rdMom_comp_27<<< gridINT_CF, threads, 0, stream >>>( DC,
- DF,
- neighborCX,
- neighborCY,
- neighborCZ,
- neighborFX,
- neighborFY,
- neighborFZ,
- size_MatC,
- size_MatF,
- isEvenTimestep,
- posCSWB,
- posFSWB,
- kCF,
- omCoarse,
- omFine,
- nu,
- nxC,
- nyC,
- nxF,
- nyF,
- offCF);
- getLastCudaError("scaleCF_RhoSq_3rdMom_comp_27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void ScaleCF_AA2016_comp_27(real* DC,
- real* DF,
- unsigned int* neighborCX,
- unsigned int* neighborCY,
- unsigned int* neighborCZ,
- unsigned int* neighborFX,
- unsigned int* neighborFY,
- unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
- bool isEvenTimestep,
- unsigned int* posCSWB,
- unsigned int* posFSWB,
- unsigned int kCF,
- real omCoarse,
- real omFine,
- real nu,
- unsigned int nxC,
- unsigned int nyC,
- unsigned int nxF,
- unsigned int nyF,
- unsigned int numberOfThreads,
- OffCF offCF,
- CUstream_st *stream)
-{
- int Grid = (kCF / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 gridINT_CF(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- scaleCF_AA2016_comp_27<<< gridINT_CF, threads, 0, stream >>>(DC,
- DF,
- neighborCX,
- neighborCY,
- neighborCZ,
- neighborFX,
- neighborFY,
- neighborFZ,
- size_MatC,
- size_MatF,
- isEvenTimestep,
- posCSWB,
- posFSWB,
- kCF,
- omCoarse,
- omFine,
- nu,
- nxC,
- nyC,
- nxF,
- nyF,
- offCF);
- getLastCudaError("scaleCF_AA2016_comp_27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void ScaleCF_NSPress_27( real* DC,
- real* DF,
- unsigned int* neighborCX,
- unsigned int* neighborCY,
- unsigned int* neighborCZ,
- unsigned int* neighborFX,
- unsigned int* neighborFY,
- unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
- bool isEvenTimestep,
- unsigned int* posCSWB,
- unsigned int* posFSWB,
- unsigned int kCF,
- real omCoarse,
- real omFine,
- real nu,
- unsigned int nxC,
- unsigned int nyC,
- unsigned int nxF,
- unsigned int nyF,
- unsigned int numberOfThreads,
- OffCF offCF)
-{
- int Grid = (kCF / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 gridINT_CF(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- scaleCF_NSPress_27<<< gridINT_CF, threads >>>(DC,
- DF,
- neighborCX,
- neighborCY,
- neighborCZ,
- neighborFX,
- neighborFY,
- neighborFZ,
- size_MatC,
- size_MatF,
- isEvenTimestep,
- posCSWB,
- posFSWB,
- kCF,
- omCoarse,
- omFine,
- nu,
- nxC,
- nyC,
- nxF,
- nyF,
- offCF);
- getLastCudaError("scaleCF_Fix_27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void ScaleCFThSMG7( real* DC,
- real* DF,
- real* DD7C,
- real* DD7F,
- unsigned int* neighborCX,
- unsigned int* neighborCY,
- unsigned int* neighborCZ,
- unsigned int* neighborFX,
- unsigned int* neighborFY,
- unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
- bool isEvenTimestep,
- unsigned int* posCSWB,
- unsigned int* posFSWB,
- unsigned int kCF,
- real nu,
- real diffusivity_fine,
- unsigned int numberOfThreads,
- OffCF offCF)
-{
- int Grid = (kCF / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 gridINT_CF(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- scaleCFThSMG7<<< gridINT_CF, threads >>> (DC,
- DF,
- DD7C,
- DD7F,
- neighborCX,
- neighborCY,
- neighborCZ,
- neighborFX,
- neighborFY,
- neighborFZ,
- size_MatC,
- size_MatF,
- isEvenTimestep,
- posCSWB,
- posFSWB,
- kCF,
- nu,
- diffusivity_fine,
- offCF);
- getLastCudaError("scaleCFThSMG7 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void ScaleCFThS7( real* DC,
- real* DF,
- real* DD7C,
- real* DD7F,
- unsigned int* neighborCX,
- unsigned int* neighborCY,
- unsigned int* neighborCZ,
- unsigned int* neighborFX,
- unsigned int* neighborFY,
- unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
- bool isEvenTimestep,
- unsigned int* posCSWB,
- unsigned int* posFSWB,
- unsigned int kCF,
- real nu,
- real diffusivity_fine,
- unsigned int numberOfThreads)
-{
- int Grid = (kCF / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 gridINT_CF(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- scaleCFThS7<<< gridINT_CF, threads >>> ( DC,
- DF,
- DD7C,
- DD7F,
- neighborCX,
- neighborCY,
- neighborCZ,
- neighborFX,
- neighborFY,
- neighborFZ,
- size_MatC,
- size_MatF,
- isEvenTimestep,
- posCSWB,
- posFSWB,
- kCF,
- nu,
- diffusivity_fine);
- getLastCudaError("scaleCFThS7 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void ScaleCFThS27( real* DC,
- real* DF,
- real* DD27C,
- real* DD27F,
- unsigned int* neighborCX,
- unsigned int* neighborCY,
- unsigned int* neighborCZ,
- unsigned int* neighborFX,
- unsigned int* neighborFY,
- unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
- bool isEvenTimestep,
- unsigned int* posCSWB,
- unsigned int* posFSWB,
- unsigned int kCF,
- real nu,
- real diffusivity_fine,
- unsigned int numberOfThreads,
- OffCF offCF)
-{
- int Grid = (kCF / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 gridINT_CF(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- scaleCFThS27<<< gridINT_CF, threads >>> ( DC,
- DF,
- DD27C,
- DD27F,
- neighborCX,
- neighborCY,
- neighborCZ,
- neighborFX,
- neighborFY,
- neighborFZ,
- size_MatC,
- size_MatF,
- isEvenTimestep,
- posCSWB,
- posFSWB,
- kCF,
- nu,
- diffusivity_fine,
- offCF);
- getLastCudaError("scaleCFThS27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void ScaleFC27( real* DC,
- real* DF,
- unsigned int* neighborCX,
- unsigned int* neighborCY,
- unsigned int* neighborCZ,
- unsigned int* neighborFX,
- unsigned int* neighborFY,
- unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
- bool isEvenTimestep,
- unsigned int* posC,
- unsigned int* posFSWB,
- unsigned int kFC,
- real omCoarse,
- real omFine,
- real nu,
- unsigned int nxC,
- unsigned int nyC,
- unsigned int nxF,
- unsigned int nyF,
- unsigned int numberOfThreads)
-{
- int Grid = (kFC / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 gridINT_FC(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- scaleFC27<<< gridINT_FC, threads >>> ( DC,
- DF,
- neighborCX,
- neighborCY,
- neighborCZ,
- neighborFX,
- neighborFY,
- neighborFZ,
- size_MatC,
- size_MatF,
- isEvenTimestep,
- posC,
- posFSWB,
- kFC,
- omCoarse,
- omFine,
- nu,
- nxC,
- nyC,
- nxF,
- nyF);
- getLastCudaError("scaleFC27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void ScaleFCEff27(real* DC,
- real* DF,
- unsigned int* neighborCX,
- unsigned int* neighborCY,
- unsigned int* neighborCZ,
- unsigned int* neighborFX,
- unsigned int* neighborFY,
- unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
- bool isEvenTimestep,
- unsigned int* posC,
- unsigned int* posFSWB,
- unsigned int kFC,
- real omCoarse,
- real omFine,
- real nu,
- unsigned int nxC,
- unsigned int nyC,
- unsigned int nxF,
- unsigned int nyF,
- unsigned int numberOfThreads,
- OffFC offFC)
-{
- int Grid = (kFC / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 gridINT_FC(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- scaleFCEff27<<< gridINT_FC, threads >>> ( DC,
- DF,
- neighborCX,
- neighborCY,
- neighborCZ,
- neighborFX,
- neighborFY,
- neighborFZ,
- size_MatC,
- size_MatF,
- isEvenTimestep,
- posC,
- posFSWB,
- kFC,
- omCoarse,
- omFine,
- nu,
- nxC,
- nyC,
- nxF,
- nyF,
- offFC);
- getLastCudaError("scaleFCEff27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void ScaleFCLast27(real* DC,
- real* DF,
- unsigned int* neighborCX,
- unsigned int* neighborCY,
- unsigned int* neighborCZ,
- unsigned int* neighborFX,
- unsigned int* neighborFY,
- unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
- bool isEvenTimestep,
- unsigned int* posC,
- unsigned int* posFSWB,
- unsigned int kFC,
- real omCoarse,
- real omFine,
- real nu,
- unsigned int nxC,
- unsigned int nyC,
- unsigned int nxF,
- unsigned int nyF,
- unsigned int numberOfThreads,
- OffFC offFC)
-{
- int Grid = (kFC / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 gridINT_FC(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- scaleFCLast27<<< gridINT_FC, threads >>> (DC,
- DF,
- neighborCX,
- neighborCY,
- neighborCZ,
- neighborFX,
- neighborFY,
- neighborFZ,
- size_MatC,
- size_MatF,
- isEvenTimestep,
- posC,
- posFSWB,
- kFC,
- omCoarse,
- omFine,
- nu,
- nxC,
- nyC,
- nxF,
- nyF,
- offFC);
- getLastCudaError("Kernel execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void ScaleFCpress27(real* DC,
- real* DF,
- unsigned int* neighborCX,
- unsigned int* neighborCY,
- unsigned int* neighborCZ,
- unsigned int* neighborFX,
- unsigned int* neighborFY,
- unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
- bool isEvenTimestep,
- unsigned int* posC,
- unsigned int* posFSWB,
- unsigned int kFC,
- real omCoarse,
- real omFine,
- real nu,
- unsigned int nxC,
- unsigned int nyC,
- unsigned int nxF,
- unsigned int nyF,
- unsigned int numberOfThreads,
- OffFC offFC)
-{
- int Grid = (kFC / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 gridINT_FC(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- scaleFCpress27<<< gridINT_FC, threads >>> ( DC,
- DF,
- neighborCX,
- neighborCY,
- neighborCZ,
- neighborFX,
- neighborFY,
- neighborFZ,
- size_MatC,
- size_MatF,
- isEvenTimestep,
- posC,
- posFSWB,
- kFC,
- omCoarse,
- omFine,
- nu,
- nxC,
- nyC,
- nxF,
- nyF,
- offFC);
- getLastCudaError("scaleFCpress27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void ScaleFC_Fix_27(real* DC,
- real* DF,
- unsigned int* neighborCX,
- unsigned int* neighborCY,
- unsigned int* neighborCZ,
- unsigned int* neighborFX,
- unsigned int* neighborFY,
- unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
- bool isEvenTimestep,
- unsigned int* posC,
- unsigned int* posFSWB,
- unsigned int kFC,
- real omCoarse,
- real omFine,
- real nu,
- unsigned int nxC,
- unsigned int nyC,
- unsigned int nxF,
- unsigned int nyF,
- unsigned int numberOfThreads,
- OffFC offFC)
-{
- int Grid = (kFC / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 gridINT_FC(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- scaleFC_Fix_27<<< gridINT_FC, threads >>> ( DC,
- DF,
- neighborCX,
- neighborCY,
- neighborCZ,
- neighborFX,
- neighborFY,
- neighborFZ,
- size_MatC,
- size_MatF,
- isEvenTimestep,
- posC,
- posFSWB,
- kFC,
- omCoarse,
- omFine,
- nu,
- nxC,
- nyC,
- nxF,
- nyF,
- offFC);
- getLastCudaError("scaleFC_Fix_27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void ScaleFC_Fix_comp_27( real* DC,
- real* DF,
- unsigned int* neighborCX,
- unsigned int* neighborCY,
- unsigned int* neighborCZ,
- unsigned int* neighborFX,
- unsigned int* neighborFY,
- unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
- bool isEvenTimestep,
- unsigned int* posC,
- unsigned int* posFSWB,
- unsigned int kFC,
- real omCoarse,
- real omFine,
- real nu,
- unsigned int nxC,
- unsigned int nyC,
- unsigned int nxF,
- unsigned int nyF,
- unsigned int numberOfThreads,
- OffFC offFC)
-{
- int Grid = (kFC / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 gridINT_FC(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- scaleFC_Fix_comp_27<<< gridINT_FC, threads >>> ( DC,
- DF,
- neighborCX,
- neighborCY,
- neighborCZ,
- neighborFX,
- neighborFY,
- neighborFZ,
- size_MatC,
- size_MatF,
- isEvenTimestep,
- posC,
- posFSWB,
- kFC,
- omCoarse,
- omFine,
- nu,
- nxC,
- nyC,
- nxF,
- nyF,
- offFC);
- getLastCudaError("scaleFC_Fix_27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void ScaleFC_0817_comp_27( real* DC,
- real* DF,
- unsigned int* neighborCX,
- unsigned int* neighborCY,
- unsigned int* neighborCZ,
- unsigned int* neighborFX,
- unsigned int* neighborFY,
- unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
- bool isEvenTimestep,
- unsigned int* posC,
- unsigned int* posFSWB,
- unsigned int kFC,
- real omCoarse,
- real omFine,
- real nu,
- unsigned int nxC,
- unsigned int nyC,
- unsigned int nxF,
- unsigned int nyF,
- unsigned int numberOfThreads,
- OffFC offFC,
- CUstream_st *stream)
-{
- int Grid = (kFC / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 gridINT_FC(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- scaleFC_0817_comp_27<<< gridINT_FC, threads, 0, stream >>> (DC,
- DF,
- neighborCX,
- neighborCY,
- neighborCZ,
- neighborFX,
- neighborFY,
- neighborFZ,
- size_MatC,
- size_MatF,
- isEvenTimestep,
- posC,
- posFSWB,
- kFC,
- omCoarse,
- omFine,
- nu,
- nxC,
- nyC,
- nxF,
- nyF,
- offFC);
- getLastCudaError("scaleFC_0817_27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void ScaleFC_comp_D3Q27F3_2018( real* DC,
- real* DF,
- real* G6,
- unsigned int* neighborCX,
- unsigned int* neighborCY,
- unsigned int* neighborCZ,
- unsigned int* neighborFX,
- unsigned int* neighborFY,
- unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
- bool isEvenTimestep,
- unsigned int* posC,
- unsigned int* posFSWB,
- unsigned int kFC,
- real omCoarse,
- real omFine,
- real nu,
- unsigned int nxC,
- unsigned int nyC,
- unsigned int nxF,
- unsigned int nyF,
- unsigned int numberOfThreads,
- OffFC offFC)
-{
- int Grid = (kFC / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 gridINT_FC(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- scaleFC_comp_D3Q27F3_2018 <<< gridINT_FC, threads >>> (DC,
- DF,
- G6,
- neighborCX,
- neighborCY,
- neighborCZ,
- neighborFX,
- neighborFY,
- neighborFZ,
- size_MatC,
- size_MatF,
- isEvenTimestep,
- posC,
- posFSWB,
- kFC,
- omCoarse,
- omFine,
- nu,
- nxC,
- nyC,
- nxF,
- nyF,
- offFC);
- getLastCudaError("scaleFC_comp_D3Q27F3_2018 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void ScaleFC_comp_D3Q27F3( real* DC,
- real* DF,
- real* G6,
- unsigned int* neighborCX,
- unsigned int* neighborCY,
- unsigned int* neighborCZ,
- unsigned int* neighborFX,
- unsigned int* neighborFY,
- unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
- bool isEvenTimestep,
- unsigned int* posC,
- unsigned int* posFSWB,
- unsigned int kFC,
- real omCoarse,
- real omFine,
- real nu,
- unsigned int nxC,
- unsigned int nyC,
- unsigned int nxF,
- unsigned int nyF,
- unsigned int numberOfThreads,
- OffFC offFC,
- CUstream_st *stream)
-{
- int Grid = (kFC / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 gridINT_FC(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- scaleFC_comp_D3Q27F3 <<< gridINT_FC, threads, 0, stream >>> (DC,
- DF,
- G6,
- neighborCX,
- neighborCY,
- neighborCZ,
- neighborFX,
- neighborFY,
- neighborFZ,
- size_MatC,
- size_MatF,
- isEvenTimestep,
- posC,
- posFSWB,
- kFC,
- omCoarse,
- omFine,
- nu,
- nxC,
- nyC,
- nxF,
- nyF,
- offFC);
- getLastCudaError("scaleFC_0817_27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void ScaleFC_staggered_time_comp_27( real* DC,
- real* DF,
- unsigned int* neighborCX,
- unsigned int* neighborCY,
- unsigned int* neighborCZ,
- unsigned int* neighborFX,
- unsigned int* neighborFY,
- unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
- bool isEvenTimestep,
- unsigned int* posC,
- unsigned int* posFSWB,
- unsigned int kFC,
- real omCoarse,
- real omFine,
- real nu,
- unsigned int nxC,
- unsigned int nyC,
- unsigned int nxF,
- unsigned int nyF,
- unsigned int numberOfThreads,
- OffFC offFC)
-{
- int Grid = (kFC / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 gridINT_FC(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- scaleFC_staggered_time_comp_27<<< gridINT_FC, threads >>> ( DC,
- DF,
- neighborCX,
- neighborCY,
- neighborCZ,
- neighborFX,
- neighborFY,
- neighborFZ,
- size_MatC,
- size_MatF,
- isEvenTimestep,
- posC,
- posFSWB,
- kFC,
- omCoarse,
- omFine,
- nu,
- nxC,
- nyC,
- nxF,
- nyF,
- offFC);
- getLastCudaError("scaleFC_Fix_27 execution failed");
+ dim3 grid = vf::cuda::getCudaGrid(parameterDeviceC->numberofthreads, icellCF->kCF);
+ dim3 threads(parameterDeviceC->numberofthreads, 1, 1 );
+
+ scaleCF_compressible<<<grid, threads, 0, stream>>>(
+ parameterDeviceC->distributions.f[0],
+ parameterDeviceF->distributions.f[0],
+ parameterDeviceC->neighborX,
+ parameterDeviceC->neighborY,
+ parameterDeviceC->neighborZ,
+ parameterDeviceF->neighborX,
+ parameterDeviceF->neighborY,
+ parameterDeviceF->neighborZ,
+ parameterDeviceC->numberOfNodes,
+ parameterDeviceF->numberOfNodes,
+ parameterDeviceC->isEvenTimestep,
+ icellCF->ICellCFC,
+ icellCF->ICellCFF,
+ icellCF->kCF,
+ parameterDeviceC->omega,
+ parameterDeviceF->omega,
+ offsetCF);
+ getLastCudaError("scaleCF_compressible execution failed");
+}
+
+//////////////////////////////////////////////////////////////////////////
+void ScaleCF_RhoSq_3rdMom_comp_27(
+ real* DC,
+ real* DF,
+ unsigned int* neighborCX,
+ unsigned int* neighborCY,
+ unsigned int* neighborCZ,
+ unsigned int* neighborFX,
+ unsigned int* neighborFY,
+ unsigned int* neighborFZ,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
+ bool isEvenTimestep,
+ unsigned int* posCSWB,
+ unsigned int* posFSWB,
+ unsigned int kCF,
+ real omCoarse,
+ real omFine,
+ real nu,
+ unsigned int nxC,
+ unsigned int nyC,
+ unsigned int nxF,
+ unsigned int nyF,
+ unsigned int numberOfThreads,
+ OffCF offCF,
+ CUstream_st *stream)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, kCF);
+
+ scaleCF_RhoSq_3rdMom_comp_27<<< grid.grid, grid.threads, 0, stream >>>(
+ DC,
+ DF,
+ neighborCX,
+ neighborCY,
+ neighborCZ,
+ neighborFX,
+ neighborFY,
+ neighborFZ,
+ numberOfLBnodesC,
+ numberOfLBnodesF,
+ isEvenTimestep,
+ posCSWB,
+ posFSWB,
+ kCF,
+ omCoarse,
+ omFine,
+ nu,
+ nxC,
+ nyC,
+ nxF,
+ nyF,
+ offCF);
+ getLastCudaError("scaleCF_RhoSq_3rdMom_comp_27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void ScaleCF_AA2016_comp_27(
+ real* DC,
+ real* DF,
+ unsigned int* neighborCX,
+ unsigned int* neighborCY,
+ unsigned int* neighborCZ,
+ unsigned int* neighborFX,
+ unsigned int* neighborFY,
+ unsigned int* neighborFZ,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
+ bool isEvenTimestep,
+ unsigned int* posCSWB,
+ unsigned int* posFSWB,
+ unsigned int kCF,
+ real omCoarse,
+ real omFine,
+ real nu,
+ unsigned int nxC,
+ unsigned int nyC,
+ unsigned int nxF,
+ unsigned int nyF,
+ unsigned int numberOfThreads,
+ OffCF offCF,
+ CUstream_st *stream)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, kCF);
+
+ scaleCF_AA2016_comp_27<<< grid.grid, grid.threads, 0, stream >>>(
+ DC,
+ DF,
+ neighborCX,
+ neighborCY,
+ neighborCZ,
+ neighborFX,
+ neighborFY,
+ neighborFZ,
+ numberOfLBnodesC,
+ numberOfLBnodesF,
+ isEvenTimestep,
+ posCSWB,
+ posFSWB,
+ kCF,
+ omCoarse,
+ omFine,
+ nu,
+ nxC,
+ nyC,
+ nxF,
+ nyF,
+ offCF);
+ getLastCudaError("scaleCF_AA2016_comp_27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void ScaleCF_NSPress_27(
+ real* DC,
+ real* DF,
+ unsigned int* neighborCX,
+ unsigned int* neighborCY,
+ unsigned int* neighborCZ,
+ unsigned int* neighborFX,
+ unsigned int* neighborFY,
+ unsigned int* neighborFZ,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
+ bool isEvenTimestep,
+ unsigned int* posCSWB,
+ unsigned int* posFSWB,
+ unsigned int kCF,
+ real omCoarse,
+ real omFine,
+ real nu,
+ unsigned int nxC,
+ unsigned int nyC,
+ unsigned int nxF,
+ unsigned int nyF,
+ unsigned int numberOfThreads,
+ OffCF offCF)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, kCF);
+
+ scaleCF_NSPress_27<<< grid.grid, grid.threads >>>(
+ DC,
+ DF,
+ neighborCX,
+ neighborCY,
+ neighborCZ,
+ neighborFX,
+ neighborFY,
+ neighborFZ,
+ numberOfLBnodesC,
+ numberOfLBnodesF,
+ isEvenTimestep,
+ posCSWB,
+ posFSWB,
+ kCF,
+ omCoarse,
+ omFine,
+ nu,
+ nxC,
+ nyC,
+ nxF,
+ nyF,
+ offCF);
+ getLastCudaError("scaleCF_NSPress_27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void ScaleCFThSMG7(
+ real* DC,
+ real* DF,
+ real* DD7C,
+ real* DD7F,
+ unsigned int* neighborCX,
+ unsigned int* neighborCY,
+ unsigned int* neighborCZ,
+ unsigned int* neighborFX,
+ unsigned int* neighborFY,
+ unsigned int* neighborFZ,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
+ bool isEvenTimestep,
+ unsigned int* posCSWB,
+ unsigned int* posFSWB,
+ unsigned int kCF,
+ real nu,
+ real diffusivity_fine,
+ unsigned int numberOfThreads,
+ OffCF offCF)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, kCF);
+
+ scaleCFThSMG7<<< grid.grid, grid.threads >>> (
+ DC,
+ DF,
+ DD7C,
+ DD7F,
+ neighborCX,
+ neighborCY,
+ neighborCZ,
+ neighborFX,
+ neighborFY,
+ neighborFZ,
+ numberOfLBnodesC,
+ numberOfLBnodesF,
+ isEvenTimestep,
+ posCSWB,
+ posFSWB,
+ kCF,
+ nu,
+ diffusivity_fine,
+ offCF);
+ getLastCudaError("scaleCFThSMG7 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void ScaleCFThS7(
+ real* DC,
+ real* DF,
+ real* DD7C,
+ real* DD7F,
+ unsigned int* neighborCX,
+ unsigned int* neighborCY,
+ unsigned int* neighborCZ,
+ unsigned int* neighborFX,
+ unsigned int* neighborFY,
+ unsigned int* neighborFZ,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
+ bool isEvenTimestep,
+ unsigned int* posCSWB,
+ unsigned int* posFSWB,
+ unsigned int kCF,
+ real nu,
+ real diffusivity_fine,
+ unsigned int numberOfThreads)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, kCF);
+
+ scaleCFThS7<<< grid.grid, grid.threads >>> (
+ DC,
+ DF,
+ DD7C,
+ DD7F,
+ neighborCX,
+ neighborCY,
+ neighborCZ,
+ neighborFX,
+ neighborFY,
+ neighborFZ,
+ numberOfLBnodesC,
+ numberOfLBnodesF,
+ isEvenTimestep,
+ posCSWB,
+ posFSWB,
+ kCF,
+ nu,
+ diffusivity_fine);
+ getLastCudaError("scaleCFThS7 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void ScaleCFThS27(
+ real* DC,
+ real* DF,
+ real* DD27C,
+ real* DD27F,
+ unsigned int* neighborCX,
+ unsigned int* neighborCY,
+ unsigned int* neighborCZ,
+ unsigned int* neighborFX,
+ unsigned int* neighborFY,
+ unsigned int* neighborFZ,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
+ bool isEvenTimestep,
+ unsigned int* posCSWB,
+ unsigned int* posFSWB,
+ unsigned int kCF,
+ real nu,
+ real diffusivity_fine,
+ unsigned int numberOfThreads,
+ OffCF offCF)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, kCF);
+
+ scaleCFThS27<<< grid.grid, grid.threads >>> (
+ DC,
+ DF,
+ DD27C,
+ DD27F,
+ neighborCX,
+ neighborCY,
+ neighborCZ,
+ neighborFX,
+ neighborFY,
+ neighborFZ,
+ numberOfLBnodesC,
+ numberOfLBnodesF,
+ isEvenTimestep,
+ posCSWB,
+ posFSWB,
+ kCF,
+ nu,
+ diffusivity_fine,
+ offCF);
+ getLastCudaError("scaleCFThS27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void ScaleFC27(
+ real* DC,
+ real* DF,
+ unsigned int* neighborCX,
+ unsigned int* neighborCY,
+ unsigned int* neighborCZ,
+ unsigned int* neighborFX,
+ unsigned int* neighborFY,
+ unsigned int* neighborFZ,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
+ bool isEvenTimestep,
+ unsigned int* posC,
+ unsigned int* posFSWB,
+ unsigned int kFC,
+ real omCoarse,
+ real omFine,
+ real nu,
+ unsigned int nxC,
+ unsigned int nyC,
+ unsigned int nxF,
+ unsigned int nyF,
+ unsigned int numberOfThreads)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, kFC);
+
+ scaleFC27<<< grid.grid, grid.threads >>> (
+ DC,
+ DF,
+ neighborCX,
+ neighborCY,
+ neighborCZ,
+ neighborFX,
+ neighborFY,
+ neighborFZ,
+ numberOfLBnodesC,
+ numberOfLBnodesF,
+ isEvenTimestep,
+ posC,
+ posFSWB,
+ kFC,
+ omCoarse,
+ omFine,
+ nu,
+ nxC,
+ nyC,
+ nxF,
+ nyF);
+ getLastCudaError("scaleFC27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void ScaleFCEff27(
+ real* DC,
+ real* DF,
+ unsigned int* neighborCX,
+ unsigned int* neighborCY,
+ unsigned int* neighborCZ,
+ unsigned int* neighborFX,
+ unsigned int* neighborFY,
+ unsigned int* neighborFZ,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
+ bool isEvenTimestep,
+ unsigned int* posC,
+ unsigned int* posFSWB,
+ unsigned int kFC,
+ real omCoarse,
+ real omFine,
+ real nu,
+ unsigned int nxC,
+ unsigned int nyC,
+ unsigned int nxF,
+ unsigned int nyF,
+ unsigned int numberOfThreads,
+ OffFC offFC)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, kFC);
+
+ scaleFCEff27<<< grid.grid, grid.threads >>> (
+ DC,
+ DF,
+ neighborCX,
+ neighborCY,
+ neighborCZ,
+ neighborFX,
+ neighborFY,
+ neighborFZ,
+ numberOfLBnodesC,
+ numberOfLBnodesF,
+ isEvenTimestep,
+ posC,
+ posFSWB,
+ kFC,
+ omCoarse,
+ omFine,
+ nu,
+ nxC,
+ nyC,
+ nxF,
+ nyF,
+ offFC);
+ getLastCudaError("scaleFCEff27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void ScaleFCLast27(
+ real* DC,
+ real* DF,
+ unsigned int* neighborCX,
+ unsigned int* neighborCY,
+ unsigned int* neighborCZ,
+ unsigned int* neighborFX,
+ unsigned int* neighborFY,
+ unsigned int* neighborFZ,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
+ bool isEvenTimestep,
+ unsigned int* posC,
+ unsigned int* posFSWB,
+ unsigned int kFC,
+ real omCoarse,
+ real omFine,
+ real nu,
+ unsigned int nxC,
+ unsigned int nyC,
+ unsigned int nxF,
+ unsigned int nyF,
+ unsigned int numberOfThreads,
+ OffFC offFC)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, kFC);
+
+ scaleFCLast27<<< grid.grid, grid.threads >>> (
+ DC,
+ DF,
+ neighborCX,
+ neighborCY,
+ neighborCZ,
+ neighborFX,
+ neighborFY,
+ neighborFZ,
+ numberOfLBnodesC,
+ numberOfLBnodesF,
+ isEvenTimestep,
+ posC,
+ posFSWB,
+ kFC,
+ omCoarse,
+ omFine,
+ nu,
+ nxC,
+ nyC,
+ nxF,
+ nyF,
+ offFC);
+ getLastCudaError("Kernel execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void ScaleFCpress27(
+ real* DC,
+ real* DF,
+ unsigned int* neighborCX,
+ unsigned int* neighborCY,
+ unsigned int* neighborCZ,
+ unsigned int* neighborFX,
+ unsigned int* neighborFY,
+ unsigned int* neighborFZ,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
+ bool isEvenTimestep,
+ unsigned int* posC,
+ unsigned int* posFSWB,
+ unsigned int kFC,
+ real omCoarse,
+ real omFine,
+ real nu,
+ unsigned int nxC,
+ unsigned int nyC,
+ unsigned int nxF,
+ unsigned int nyF,
+ unsigned int numberOfThreads,
+ OffFC offFC)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, kFC);
+
+ scaleFCpress27<<< grid.grid, grid.threads >>> (
+ DC,
+ DF,
+ neighborCX,
+ neighborCY,
+ neighborCZ,
+ neighborFX,
+ neighborFY,
+ neighborFZ,
+ numberOfLBnodesC,
+ numberOfLBnodesF,
+ isEvenTimestep,
+ posC,
+ posFSWB,
+ kFC,
+ omCoarse,
+ omFine,
+ nu,
+ nxC,
+ nyC,
+ nxF,
+ nyF,
+ offFC);
+ getLastCudaError("scaleFCpress27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void ScaleFC_Fix_27(
+ real* DC,
+ real* DF,
+ unsigned int* neighborCX,
+ unsigned int* neighborCY,
+ unsigned int* neighborCZ,
+ unsigned int* neighborFX,
+ unsigned int* neighborFY,
+ unsigned int* neighborFZ,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
+ bool isEvenTimestep,
+ unsigned int* posC,
+ unsigned int* posFSWB,
+ unsigned int kFC,
+ real omCoarse,
+ real omFine,
+ real nu,
+ unsigned int nxC,
+ unsigned int nyC,
+ unsigned int nxF,
+ unsigned int nyF,
+ unsigned int numberOfThreads,
+ OffFC offFC)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, kFC);
+
+ scaleFC_Fix_27<<< grid.grid, grid.threads >>> (
+ DC,
+ DF,
+ neighborCX,
+ neighborCY,
+ neighborCZ,
+ neighborFX,
+ neighborFY,
+ neighborFZ,
+ numberOfLBnodesC,
+ numberOfLBnodesF,
+ isEvenTimestep,
+ posC,
+ posFSWB,
+ kFC,
+ omCoarse,
+ omFine,
+ nu,
+ nxC,
+ nyC,
+ nxF,
+ nyF,
+ offFC);
+ getLastCudaError("scaleFC_Fix_27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void ScaleFC_Fix_comp_27(
+ real* DC,
+ real* DF,
+ unsigned int* neighborCX,
+ unsigned int* neighborCY,
+ unsigned int* neighborCZ,
+ unsigned int* neighborFX,
+ unsigned int* neighborFY,
+ unsigned int* neighborFZ,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
+ bool isEvenTimestep,
+ unsigned int* posC,
+ unsigned int* posFSWB,
+ unsigned int kFC,
+ real omCoarse,
+ real omFine,
+ real nu,
+ unsigned int nxC,
+ unsigned int nyC,
+ unsigned int nxF,
+ unsigned int nyF,
+ unsigned int numberOfThreads,
+ OffFC offFC)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, kFC);
+
+ scaleFC_Fix_comp_27<<< grid.grid, grid.threads >>> (
+ DC,
+ DF,
+ neighborCX,
+ neighborCY,
+ neighborCZ,
+ neighborFX,
+ neighborFY,
+ neighborFZ,
+ numberOfLBnodesC,
+ numberOfLBnodesF,
+ isEvenTimestep,
+ posC,
+ posFSWB,
+ kFC,
+ omCoarse,
+ omFine,
+ nu,
+ nxC,
+ nyC,
+ nxF,
+ nyF,
+ offFC);
+ getLastCudaError("scaleFC_Fix_comp_27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void ScaleFC_0817_comp_27(
+ real* DC,
+ real* DF,
+ unsigned int* neighborCX,
+ unsigned int* neighborCY,
+ unsigned int* neighborCZ,
+ unsigned int* neighborFX,
+ unsigned int* neighborFY,
+ unsigned int* neighborFZ,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
+ bool isEvenTimestep,
+ unsigned int* posC,
+ unsigned int* posFSWB,
+ unsigned int kFC,
+ real omCoarse,
+ real omFine,
+ real nu,
+ unsigned int nxC,
+ unsigned int nyC,
+ unsigned int nxF,
+ unsigned int nyF,
+ unsigned int numberOfThreads,
+ OffFC offFC,
+ CUstream_st *stream)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, kFC);
+
+ scaleFC_0817_comp_27<<< grid.grid, grid.threads, 0, stream >>> (
+ DC,
+ DF,
+ neighborCX,
+ neighborCY,
+ neighborCZ,
+ neighborFX,
+ neighborFY,
+ neighborFZ,
+ numberOfLBnodesC,
+ numberOfLBnodesF,
+ isEvenTimestep,
+ posC,
+ posFSWB,
+ kFC,
+ omCoarse,
+ omFine,
+ nu,
+ nxC,
+ nyC,
+ nxF,
+ nyF,
+ offFC);
+ getLastCudaError("scaleFC_0817_comp_27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void ScaleFC_comp_D3Q27F3_2018(
+ real* DC,
+ real* DF,
+ real* G6,
+ unsigned int* neighborCX,
+ unsigned int* neighborCY,
+ unsigned int* neighborCZ,
+ unsigned int* neighborFX,
+ unsigned int* neighborFY,
+ unsigned int* neighborFZ,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
+ bool isEvenTimestep,
+ unsigned int* posC,
+ unsigned int* posFSWB,
+ unsigned int kFC,
+ real omCoarse,
+ real omFine,
+ real nu,
+ unsigned int nxC,
+ unsigned int nyC,
+ unsigned int nxF,
+ unsigned int nyF,
+ unsigned int numberOfThreads,
+ OffFC offFC)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, kFC);
+
+ scaleFC_comp_D3Q27F3_2018 <<< grid.grid, grid.threads >>> (
+ DC,
+ DF,
+ G6,
+ neighborCX,
+ neighborCY,
+ neighborCZ,
+ neighborFX,
+ neighborFY,
+ neighborFZ,
+ numberOfLBnodesC,
+ numberOfLBnodesF,
+ isEvenTimestep,
+ posC,
+ posFSWB,
+ kFC,
+ omCoarse,
+ omFine,
+ nu,
+ nxC,
+ nyC,
+ nxF,
+ nyF,
+ offFC);
+ getLastCudaError("scaleFC_comp_D3Q27F3_2018 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void ScaleFC_comp_D3Q27F3(
+ real* DC,
+ real* DF,
+ real* G6,
+ unsigned int* neighborCX,
+ unsigned int* neighborCY,
+ unsigned int* neighborCZ,
+ unsigned int* neighborFX,
+ unsigned int* neighborFY,
+ unsigned int* neighborFZ,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
+ bool isEvenTimestep,
+ unsigned int* posC,
+ unsigned int* posFSWB,
+ unsigned int kFC,
+ real omCoarse,
+ real omFine,
+ real nu,
+ unsigned int nxC,
+ unsigned int nyC,
+ unsigned int nxF,
+ unsigned int nyF,
+ unsigned int numberOfThreads,
+ OffFC offFC,
+ CUstream_st *stream)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, kFC);
+
+ scaleFC_comp_D3Q27F3 <<< grid.grid, grid.threads, 0, stream >>> (
+ DC,
+ DF,
+ G6,
+ neighborCX,
+ neighborCY,
+ neighborCZ,
+ neighborFX,
+ neighborFY,
+ neighborFZ,
+ numberOfLBnodesC,
+ numberOfLBnodesF,
+ isEvenTimestep,
+ posC,
+ posFSWB,
+ kFC,
+ omCoarse,
+ omFine,
+ nu,
+ nxC,
+ nyC,
+ nxF,
+ nyF,
+ offFC);
+ getLastCudaError("scaleFC_comp_D3Q27F3 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void ScaleFC_staggered_time_comp_27(
+ real* DC,
+ real* DF,
+ unsigned int* neighborCX,
+ unsigned int* neighborCY,
+ unsigned int* neighborCZ,
+ unsigned int* neighborFX,
+ unsigned int* neighborFY,
+ unsigned int* neighborFZ,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
+ bool isEvenTimestep,
+ unsigned int* posC,
+ unsigned int* posFSWB,
+ unsigned int kFC,
+ real omCoarse,
+ real omFine,
+ real nu,
+ unsigned int nxC,
+ unsigned int nyC,
+ unsigned int nxF,
+ unsigned int nyF,
+ unsigned int numberOfThreads,
+ OffFC offFC)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, kFC);
+
+ scaleFC_staggered_time_comp_27<<< grid.grid, grid.threads >>> (
+ DC,
+ DF,
+ neighborCX,
+ neighborCY,
+ neighborCZ,
+ neighborFX,
+ neighborFY,
+ neighborFZ,
+ numberOfLBnodesC,
+ numberOfLBnodesF,
+ isEvenTimestep,
+ posC,
+ posFSWB,
+ kFC,
+ omCoarse,
+ omFine,
+ nu,
+ nxC,
+ nyC,
+ nxF,
+ nyF,
+ offFC);
+ getLastCudaError("scaleFC_staggered_time_comp_27 execution failed");
}
//////////////////////////////////////////////////////////////////////////
void ScaleFC_RhoSq_comp_27(LBMSimulationParameter * parameterDeviceC, LBMSimulationParameter* parameterDeviceF, ICellFC * icellFC, OffFC &offsetFC, CUstream_st *stream)
{
- dim3 grid = vf::cuda::getCudaGrid(parameterDeviceC->numberofthreads, icellFC->kFC);
- dim3 threads(parameterDeviceC->numberofthreads, 1, 1 );
-
- scaleFC_RhoSq_comp_27<<<grid, threads, 0, stream>>>(
- parameterDeviceC->distributions.f[0],
- parameterDeviceF->distributions.f[0],
- parameterDeviceC->neighborX,
- parameterDeviceC->neighborY,
- parameterDeviceC->neighborZ,
- parameterDeviceF->neighborX,
- parameterDeviceF->neighborY,
- parameterDeviceF->neighborZ,
- parameterDeviceC->numberOfNodes,
- parameterDeviceF->numberOfNodes,
- parameterDeviceC->isEvenTimestep,
- icellFC->ICellFCC,
- icellFC->ICellFCF,
- icellFC->kFC,
- parameterDeviceC->omega,
- parameterDeviceF->omega,
- parameterDeviceC->vis,
- parameterDeviceC->nx,
- parameterDeviceC->ny,
- parameterDeviceF->nx,
- parameterDeviceF->ny,
- offsetFC);
- getLastCudaError("scaleFC_RhoSq_27 execution failed");
+ dim3 grid = vf::cuda::getCudaGrid(parameterDeviceC->numberofthreads, icellFC->kFC);
+ dim3 threads(parameterDeviceC->numberofthreads, 1, 1 );
+
+ scaleFC_RhoSq_comp_27<<<grid, threads, 0, stream>>>(
+ parameterDeviceC->distributions.f[0],
+ parameterDeviceF->distributions.f[0],
+ parameterDeviceC->neighborX,
+ parameterDeviceC->neighborY,
+ parameterDeviceC->neighborZ,
+ parameterDeviceF->neighborX,
+ parameterDeviceF->neighborY,
+ parameterDeviceF->neighborZ,
+ parameterDeviceC->numberOfNodes,
+ parameterDeviceF->numberOfNodes,
+ parameterDeviceC->isEvenTimestep,
+ icellFC->ICellFCC,
+ icellFC->ICellFCF,
+ icellFC->kFC,
+ parameterDeviceC->omega,
+ parameterDeviceF->omega,
+ parameterDeviceC->vis,
+ parameterDeviceC->nx,
+ parameterDeviceC->ny,
+ parameterDeviceF->nx,
+ parameterDeviceF->ny,
+ offsetFC);
+ getLastCudaError("scaleFC_RhoSq_comp_27 execution failed");
}
//////////////////////////////////////////////////////////////////////////
void ScaleFC_compressible(LBMSimulationParameter * parameterDeviceC, LBMSimulationParameter* parameterDeviceF, ICellFC * icellFC, OffFC &offsetFC, CUstream_st *stream)
{
- dim3 grid = vf::cuda::getCudaGrid(parameterDeviceC->numberofthreads, icellFC->kFC);
- dim3 threads(parameterDeviceC->numberofthreads, 1, 1 );
-
- scaleFC_compressible<<<grid, threads, 0, stream>>>(
- parameterDeviceC->distributions.f[0],
- parameterDeviceF->distributions.f[0],
- parameterDeviceC->neighborX,
- parameterDeviceC->neighborY,
- parameterDeviceC->neighborZ,
- parameterDeviceF->neighborX,
- parameterDeviceF->neighborY,
- parameterDeviceF->neighborZ,
- parameterDeviceC->numberOfNodes,
- parameterDeviceF->numberOfNodes,
- parameterDeviceC->isEvenTimestep,
- icellFC->ICellFCC,
- icellFC->ICellFCF,
- icellFC->kFC,
- parameterDeviceC->omega,
- parameterDeviceF->omega,
- offsetFC);
- getLastCudaError("scaleFC_compressible execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void ScaleFC_RhoSq_3rdMom_comp_27( real* DC,
- real* DF,
- unsigned int* neighborCX,
- unsigned int* neighborCY,
- unsigned int* neighborCZ,
- unsigned int* neighborFX,
- unsigned int* neighborFY,
- unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
- bool isEvenTimestep,
- unsigned int* posC,
- unsigned int* posFSWB,
- unsigned int kFC,
- real omCoarse,
- real omFine,
- real nu,
- unsigned int nxC,
- unsigned int nyC,
- unsigned int nxF,
- unsigned int nyF,
- unsigned int numberOfThreads,
- OffFC offFC,
- CUstream_st *stream)
-{
- int Grid = (kFC / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 gridINT_FC(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- scaleFC_RhoSq_3rdMom_comp_27<<< gridINT_FC, threads, 0, stream >>>(DC,
- DF,
- neighborCX,
- neighborCY,
- neighborCZ,
- neighborFX,
- neighborFY,
- neighborFZ,
- size_MatC,
- size_MatF,
- isEvenTimestep,
- posC,
- posFSWB,
- kFC,
- omCoarse,
- omFine,
- nu,
- nxC,
- nyC,
- nxF,
- nyF,
- offFC);
- getLastCudaError("scaleFC_RhoSq_3rdMom_comp_27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void ScaleFC_AA2016_comp_27( real* DC,
- real* DF,
- unsigned int* neighborCX,
- unsigned int* neighborCY,
- unsigned int* neighborCZ,
- unsigned int* neighborFX,
- unsigned int* neighborFY,
- unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
- bool isEvenTimestep,
- unsigned int* posC,
- unsigned int* posFSWB,
- unsigned int kFC,
- real omCoarse,
- real omFine,
- real nu,
- unsigned int nxC,
- unsigned int nyC,
- unsigned int nxF,
- unsigned int nyF,
- unsigned int numberOfThreads,
- OffFC offFC,
- CUstream_st *stream)
-{
- int Grid = (kFC / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 gridINT_FC(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- scaleFC_AA2016_comp_27<<< gridINT_FC, threads, 0, stream >>>(DC,
- DF,
- neighborCX,
- neighborCY,
- neighborCZ,
- neighborFX,
- neighborFY,
- neighborFZ,
- size_MatC,
- size_MatF,
- isEvenTimestep,
- posC,
- posFSWB,
- kFC,
- omCoarse,
- omFine,
- nu,
- nxC,
- nyC,
- nxF,
- nyF,
- offFC);
- getLastCudaError("scaleFC_AA2016_comp_27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void ScaleFC_NSPress_27(real* DC,
- real* DF,
- unsigned int* neighborCX,
- unsigned int* neighborCY,
- unsigned int* neighborCZ,
- unsigned int* neighborFX,
- unsigned int* neighborFY,
- unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
- bool isEvenTimestep,
- unsigned int* posC,
- unsigned int* posFSWB,
- unsigned int kFC,
- real omCoarse,
- real omFine,
- real nu,
- unsigned int nxC,
- unsigned int nyC,
- unsigned int nxF,
- unsigned int nyF,
- unsigned int numberOfThreads,
- OffFC offFC)
-{
- int Grid = (kFC / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 gridINT_FC(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- scaleFC_NSPress_27<<< gridINT_FC, threads >>> ( DC,
- DF,
- neighborCX,
- neighborCY,
- neighborCZ,
- neighborFX,
- neighborFY,
- neighborFZ,
- size_MatC,
- size_MatF,
- isEvenTimestep,
- posC,
- posFSWB,
- kFC,
- omCoarse,
- omFine,
- nu,
- nxC,
- nyC,
- nxF,
- nyF,
- offFC);
- getLastCudaError("scaleFC_Fix_27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void ScaleFCThSMG7(real* DC,
- real* DF,
- real* DD7C,
- real* DD7F,
- unsigned int* neighborCX,
- unsigned int* neighborCY,
- unsigned int* neighborCZ,
- unsigned int* neighborFX,
- unsigned int* neighborFY,
- unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
- bool isEvenTimestep,
- unsigned int* posC,
- unsigned int* posFSWB,
- unsigned int kFC,
- real nu,
- real diffusivity_coarse,
- unsigned int numberOfThreads,
- OffFC offFC)
-{
- int Grid = (kFC / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 gridINT_FC(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- scaleFCThSMG7<<< gridINT_FC, threads >>>( DC,
- DF,
- DD7C,
- DD7F,
- neighborCX,
- neighborCY,
- neighborCZ,
- neighborFX,
- neighborFY,
- neighborFZ,
- size_MatC,
- size_MatF,
- isEvenTimestep,
- posC,
- posFSWB,
- kFC,
- nu,
- diffusivity_coarse,
- offFC);
- getLastCudaError("scaleFCThSMG7 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void ScaleFCThS7( real* DC,
- real* DF,
- real* DD7C,
- real* DD7F,
- unsigned int* neighborCX,
- unsigned int* neighborCY,
- unsigned int* neighborCZ,
- unsigned int* neighborFX,
- unsigned int* neighborFY,
- unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
- bool isEvenTimestep,
- unsigned int* posC,
- unsigned int* posFSWB,
- unsigned int kFC,
- real nu,
- real diffusivity_coarse,
- unsigned int numberOfThreads)
-{
- int Grid = (kFC / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 gridINT_FC(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- scaleFCThS7<<< gridINT_FC, threads >>>(DC,
- DF,
- DD7C,
- DD7F,
- neighborCX,
- neighborCY,
- neighborCZ,
- neighborFX,
- neighborFY,
- neighborFZ,
- size_MatC,
- size_MatF,
- isEvenTimestep,
- posC,
- posFSWB,
- kFC,
- nu,
- diffusivity_coarse);
- getLastCudaError("scaleFCThS7 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void ScaleFCThS27( real* DC,
- real* DF,
- real* DD27C,
- real* DD27F,
- unsigned int* neighborCX,
- unsigned int* neighborCY,
- unsigned int* neighborCZ,
- unsigned int* neighborFX,
- unsigned int* neighborFY,
- unsigned int* neighborFZ,
- unsigned int size_MatC,
- unsigned int size_MatF,
- bool isEvenTimestep,
- unsigned int* posC,
- unsigned int* posFSWB,
- unsigned int kFC,
- real nu,
- real diffusivity_coarse,
- unsigned int numberOfThreads,
- OffFC offFC)
-{
- int Grid = (kFC / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 gridINT_FC(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- scaleFCThS27<<< gridINT_FC, threads >>>( DC,
- DF,
- DD27C,
- DD27F,
- neighborCX,
- neighborCY,
- neighborCZ,
- neighborFX,
- neighborFY,
- neighborFZ,
- size_MatC,
- size_MatF,
- isEvenTimestep,
- posC,
- posFSWB,
- kFC,
- nu,
- diffusivity_coarse,
- offFC);
- getLastCudaError("scaleFCThS27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void DragLiftPostD27(real* DD,
- int* k_Q,
- real* QQ,
- int numberOfBCnodes,
- double *DragX,
- double *DragY,
- double *DragZ,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep,
- unsigned int numberOfThreads)
-{
- int Grid = (numberOfBCnodes / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 grid(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- DragLiftPost27<<< grid, threads >>>(DD,
- k_Q,
- QQ,
- numberOfBCnodes,
- DragX,
- DragY,
- DragZ,
- neighborX,
- neighborY,
- neighborZ,
- size_Mat,
- isEvenTimestep);
- getLastCudaError("DragLift27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void DragLiftPreD27( real* DD,
- int* k_Q,
- real* QQ,
- int numberOfBCnodes,
- double *DragX,
- double *DragY,
- double *DragZ,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep,
- unsigned int numberOfThreads)
-{
- int Grid = (numberOfBCnodes / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 grid(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- DragLiftPre27<<< grid, threads >>>( DD,
- k_Q,
- QQ,
- numberOfBCnodes,
- DragX,
- DragY,
- DragZ,
- neighborX,
- neighborY,
- neighborZ,
- size_Mat,
- isEvenTimestep);
- getLastCudaError("DragLift27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void CalcCPtop27(real* DD,
- int* cpIndex,
- int nonCp,
- double *cpPress,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep,
- unsigned int numberOfThreads)
-{
- int Grid = (nonCp / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 grid(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- CalcCP27<<< grid, threads >>>(DD,
- cpIndex,
- nonCp,
- cpPress,
- neighborX,
- neighborY,
- neighborZ,
- size_Mat,
- isEvenTimestep);
- getLastCudaError("CalcCP27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void CalcCPbottom27( real* DD,
- int* cpIndex,
- int nonCp,
- double *cpPress,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep,
- unsigned int numberOfThreads)
-{
- int Grid = (nonCp / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 grid(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- CalcCP27<<< grid, threads >>>(DD,
- cpIndex,
- nonCp,
- cpPress,
- neighborX,
- neighborY,
- neighborZ,
- size_Mat,
- isEvenTimestep);
- getLastCudaError("CalcCP27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void GetSendFsPreDev27(real* DD,
- real* bufferFs,
- int* sendIndex,
- int buffmax,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep,
- unsigned int numberOfThreads,
- cudaStream_t stream)
-{
- int Grid = (buffmax / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 grid(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- getSendFsPre27<<< grid, threads, 0, stream >>>(DD,
- bufferFs,
- sendIndex,
- buffmax,
- neighborX,
- neighborY,
- neighborZ,
- size_Mat,
- isEvenTimestep);
- getLastCudaError("getSendFsPre27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void GetSendFsPostDev27(real* DD,
- real* bufferFs,
- int* sendIndex,
- int buffmax,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep,
- unsigned int numberOfThreads,
- cudaStream_t stream)
-{
- int Grid = (buffmax / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 grid(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- getSendFsPost27<<< grid, threads, 0, stream >>>(DD,
- bufferFs,
- sendIndex,
- buffmax,
- neighborX,
- neighborY,
- neighborZ,
- size_Mat,
- isEvenTimestep);
- getLastCudaError("getSendFsPost27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void SetRecvFsPreDev27(real* DD,
- real* bufferFs,
- int* recvIndex,
- int buffmax,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep,
- unsigned int numberOfThreads,
- cudaStream_t stream)
-{
- int Grid = (buffmax / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 grid(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- setRecvFsPre27<<< grid, threads, 0, stream >>>(DD,
- bufferFs,
- recvIndex,
- buffmax,
- neighborX,
- neighborY,
- neighborZ,
- size_Mat,
- isEvenTimestep);
- getLastCudaError("setRecvFsPre27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void SetRecvFsPostDev27(real* DD,
- real* bufferFs,
- int* recvIndex,
- int buffmax,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep,
- unsigned int numberOfThreads,
- cudaStream_t stream)
-{
- int Grid = (buffmax / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 grid(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- setRecvFsPost27<<< grid, threads, 0, stream >>>(DD,
- bufferFs,
- recvIndex,
- buffmax,
- neighborX,
- neighborY,
- neighborZ,
- size_Mat,
- isEvenTimestep);
- getLastCudaError("setRecvFsPost27 execution failed");
+ dim3 grid = vf::cuda::getCudaGrid(parameterDeviceC->numberofthreads, icellFC->kFC);
+ dim3 threads(parameterDeviceC->numberofthreads, 1, 1 );
+
+ scaleFC_compressible<<<grid, threads, 0, stream>>>(
+ parameterDeviceC->distributions.f[0],
+ parameterDeviceF->distributions.f[0],
+ parameterDeviceC->neighborX,
+ parameterDeviceC->neighborY,
+ parameterDeviceC->neighborZ,
+ parameterDeviceF->neighborX,
+ parameterDeviceF->neighborY,
+ parameterDeviceF->neighborZ,
+ parameterDeviceC->numberOfNodes,
+ parameterDeviceF->numberOfNodes,
+ parameterDeviceC->isEvenTimestep,
+ icellFC->ICellFCC,
+ icellFC->ICellFCF,
+ icellFC->kFC,
+ parameterDeviceC->omega,
+ parameterDeviceF->omega,
+ offsetFC);
+ getLastCudaError("scaleFC_compressible execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void ScaleFC_RhoSq_3rdMom_comp_27(
+ real* DC,
+ real* DF,
+ unsigned int* neighborCX,
+ unsigned int* neighborCY,
+ unsigned int* neighborCZ,
+ unsigned int* neighborFX,
+ unsigned int* neighborFY,
+ unsigned int* neighborFZ,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
+ bool isEvenTimestep,
+ unsigned int* posC,
+ unsigned int* posFSWB,
+ unsigned int kFC,
+ real omCoarse,
+ real omFine,
+ real nu,
+ unsigned int nxC,
+ unsigned int nyC,
+ unsigned int nxF,
+ unsigned int nyF,
+ unsigned int numberOfThreads,
+ OffFC offFC,
+ CUstream_st *stream)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, kFC);
+
+ scaleFC_RhoSq_3rdMom_comp_27<<< grid.grid, grid.threads, 0, stream >>>(
+ DC,
+ DF,
+ neighborCX,
+ neighborCY,
+ neighborCZ,
+ neighborFX,
+ neighborFY,
+ neighborFZ,
+ numberOfLBnodesC,
+ numberOfLBnodesF,
+ isEvenTimestep,
+ posC,
+ posFSWB,
+ kFC,
+ omCoarse,
+ omFine,
+ nu,
+ nxC,
+ nyC,
+ nxF,
+ nyF,
+ offFC);
+ getLastCudaError("scaleFC_RhoSq_3rdMom_comp_27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void ScaleFC_AA2016_comp_27(
+ real* DC,
+ real* DF,
+ unsigned int* neighborCX,
+ unsigned int* neighborCY,
+ unsigned int* neighborCZ,
+ unsigned int* neighborFX,
+ unsigned int* neighborFY,
+ unsigned int* neighborFZ,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
+ bool isEvenTimestep,
+ unsigned int* posC,
+ unsigned int* posFSWB,
+ unsigned int kFC,
+ real omCoarse,
+ real omFine,
+ real nu,
+ unsigned int nxC,
+ unsigned int nyC,
+ unsigned int nxF,
+ unsigned int nyF,
+ unsigned int numberOfThreads,
+ OffFC offFC,
+ CUstream_st *stream)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, kFC);
+
+ scaleFC_AA2016_comp_27<<< grid.grid, grid.threads, 0, stream >>>(
+ DC,
+ DF,
+ neighborCX,
+ neighborCY,
+ neighborCZ,
+ neighborFX,
+ neighborFY,
+ neighborFZ,
+ numberOfLBnodesC,
+ numberOfLBnodesF,
+ isEvenTimestep,
+ posC,
+ posFSWB,
+ kFC,
+ omCoarse,
+ omFine,
+ nu,
+ nxC,
+ nyC,
+ nxF,
+ nyF,
+ offFC);
+ getLastCudaError("scaleFC_AA2016_comp_27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void ScaleFC_NSPress_27(
+ real* DC,
+ real* DF,
+ unsigned int* neighborCX,
+ unsigned int* neighborCY,
+ unsigned int* neighborCZ,
+ unsigned int* neighborFX,
+ unsigned int* neighborFY,
+ unsigned int* neighborFZ,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
+ bool isEvenTimestep,
+ unsigned int* posC,
+ unsigned int* posFSWB,
+ unsigned int kFC,
+ real omCoarse,
+ real omFine,
+ real nu,
+ unsigned int nxC,
+ unsigned int nyC,
+ unsigned int nxF,
+ unsigned int nyF,
+ unsigned int numberOfThreads,
+ OffFC offFC)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, kFC);
+
+ scaleFC_NSPress_27<<< grid.grid, grid.threads >>> (
+ DC,
+ DF,
+ neighborCX,
+ neighborCY,
+ neighborCZ,
+ neighborFX,
+ neighborFY,
+ neighborFZ,
+ numberOfLBnodesC,
+ numberOfLBnodesF,
+ isEvenTimestep,
+ posC,
+ posFSWB,
+ kFC,
+ omCoarse,
+ omFine,
+ nu,
+ nxC,
+ nyC,
+ nxF,
+ nyF,
+ offFC);
+ getLastCudaError("scaleFC_NSPress_27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void ScaleFCThSMG7(
+ real* DC,
+ real* DF,
+ real* DD7C,
+ real* DD7F,
+ unsigned int* neighborCX,
+ unsigned int* neighborCY,
+ unsigned int* neighborCZ,
+ unsigned int* neighborFX,
+ unsigned int* neighborFY,
+ unsigned int* neighborFZ,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
+ bool isEvenTimestep,
+ unsigned int* posC,
+ unsigned int* posFSWB,
+ unsigned int kFC,
+ real nu,
+ real diffusivity_coarse,
+ unsigned int numberOfThreads,
+ OffFC offFC)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, kFC);
+
+ scaleFCThSMG7<<< grid.grid, grid.threads >>>(
+ DC,
+ DF,
+ DD7C,
+ DD7F,
+ neighborCX,
+ neighborCY,
+ neighborCZ,
+ neighborFX,
+ neighborFY,
+ neighborFZ,
+ numberOfLBnodesC,
+ numberOfLBnodesF,
+ isEvenTimestep,
+ posC,
+ posFSWB,
+ kFC,
+ nu,
+ diffusivity_coarse,
+ offFC);
+ getLastCudaError("scaleFCThSMG7 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void ScaleFCThS7(
+ real* DC,
+ real* DF,
+ real* DD7C,
+ real* DD7F,
+ unsigned int* neighborCX,
+ unsigned int* neighborCY,
+ unsigned int* neighborCZ,
+ unsigned int* neighborFX,
+ unsigned int* neighborFY,
+ unsigned int* neighborFZ,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
+ bool isEvenTimestep,
+ unsigned int* posC,
+ unsigned int* posFSWB,
+ unsigned int kFC,
+ real nu,
+ real diffusivity_coarse,
+ unsigned int numberOfThreads)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, kFC);
+
+ scaleFCThS7<<< grid.grid, grid.threads >>>(
+ DC,
+ DF,
+ DD7C,
+ DD7F,
+ neighborCX,
+ neighborCY,
+ neighborCZ,
+ neighborFX,
+ neighborFY,
+ neighborFZ,
+ numberOfLBnodesC,
+ numberOfLBnodesF,
+ isEvenTimestep,
+ posC,
+ posFSWB,
+ kFC,
+ nu,
+ diffusivity_coarse);
+ getLastCudaError("scaleFCThS7 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void ScaleFCThS27(
+ real* DC,
+ real* DF,
+ real* DD27C,
+ real* DD27F,
+ unsigned int* neighborCX,
+ unsigned int* neighborCY,
+ unsigned int* neighborCZ,
+ unsigned int* neighborFX,
+ unsigned int* neighborFY,
+ unsigned int* neighborFZ,
+ unsigned long long numberOfLBnodesC,
+ unsigned long long numberOfLBnodesF,
+ bool isEvenTimestep,
+ unsigned int* posC,
+ unsigned int* posFSWB,
+ unsigned int kFC,
+ real nu,
+ real diffusivity_coarse,
+ unsigned int numberOfThreads,
+ OffFC offFC)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, kFC);
+
+ scaleFCThS27<<< grid.grid, grid.threads >>>(
+ DC,
+ DF,
+ DD27C,
+ DD27F,
+ neighborCX,
+ neighborCY,
+ neighborCZ,
+ neighborFX,
+ neighborFY,
+ neighborFZ,
+ numberOfLBnodesC,
+ numberOfLBnodesF,
+ isEvenTimestep,
+ posC,
+ posFSWB,
+ kFC,
+ nu,
+ diffusivity_coarse,
+ offFC);
+ getLastCudaError("scaleFCThS27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void DragLiftPostD27(
+ real* DD,
+ int* k_Q,
+ real* QQ,
+ int numberOfBCnodes,
+ double *DragX,
+ double *DragY,
+ double *DragZ,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep,
+ unsigned int numberOfThreads)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, numberOfBCnodes);
+
+ DragLiftPost27<<< grid.grid, grid.threads >>>(
+ DD,
+ k_Q,
+ QQ,
+ numberOfBCnodes,
+ DragX,
+ DragY,
+ DragZ,
+ neighborX,
+ neighborY,
+ neighborZ,
+ numberOfLBnodes,
+ isEvenTimestep);
+ getLastCudaError("DragLiftPost27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void DragLiftPreD27(
+ real* DD,
+ int* k_Q,
+ real* QQ,
+ int numberOfBCnodes,
+ double *DragX,
+ double *DragY,
+ double *DragZ,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep,
+ unsigned int numberOfThreads)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, numberOfBCnodes);
+
+ DragLiftPre27<<< grid.grid, grid.threads >>>(
+ DD,
+ k_Q,
+ QQ,
+ numberOfBCnodes,
+ DragX,
+ DragY,
+ DragZ,
+ neighborX,
+ neighborY,
+ neighborZ,
+ numberOfLBnodes,
+ isEvenTimestep);
+ getLastCudaError("DragLiftPre27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void CalcCPtop27(
+ real* DD,
+ int* cpIndex,
+ int nonCp,
+ double *cpPress,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep,
+ unsigned int numberOfThreads)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, nonCp);
+
+ CalcCP27<<< grid.grid, grid.threads >>>(
+ DD,
+ cpIndex,
+ nonCp,
+ cpPress,
+ neighborX,
+ neighborY,
+ neighborZ,
+ numberOfLBnodes,
+ isEvenTimestep);
+ getLastCudaError("CalcCP27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void CalcCPbottom27(
+ real* DD,
+ int* cpIndex,
+ int nonCp,
+ double *cpPress,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep,
+ unsigned int numberOfThreads)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, nonCp);
+
+ CalcCP27<<< grid.grid, grid.threads >>>(
+ DD,
+ cpIndex,
+ nonCp,
+ cpPress,
+ neighborX,
+ neighborY,
+ neighborZ,
+ numberOfLBnodes,
+ isEvenTimestep);
+ getLastCudaError("CalcCP27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void GetSendFsPreDev27(
+ real* DD,
+ real* bufferFs,
+ int* sendIndex,
+ int buffmax,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep,
+ unsigned int numberOfThreads,
+ cudaStream_t stream)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, buffmax);
+
+ getSendFsPre27<<< grid.grid, grid.threads, 0, stream >>>(
+ DD,
+ bufferFs,
+ sendIndex,
+ buffmax,
+ neighborX,
+ neighborY,
+ neighborZ,
+ numberOfLBnodes,
+ isEvenTimestep);
+ getLastCudaError("getSendFsPre27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void GetSendFsPostDev27(
+ real* DD,
+ real* bufferFs,
+ int* sendIndex,
+ int buffmax,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep,
+ unsigned int numberOfThreads,
+ cudaStream_t stream)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, buffmax);
+
+ getSendFsPost27<<< grid.grid, grid.threads, 0, stream >>>(
+ DD,
+ bufferFs,
+ sendIndex,
+ buffmax,
+ neighborX,
+ neighborY,
+ neighborZ,
+ numberOfLBnodes,
+ isEvenTimestep);
+ getLastCudaError("getSendFsPost27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void SetRecvFsPreDev27(
+ real* DD,
+ real* bufferFs,
+ int* recvIndex,
+ int buffmax,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep,
+ unsigned int numberOfThreads,
+ cudaStream_t stream)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, buffmax);
+
+ setRecvFsPre27<<< grid.grid, grid.threads, 0, stream >>>(
+ DD,
+ bufferFs,
+ recvIndex,
+ buffmax,
+ neighborX,
+ neighborY,
+ neighborZ,
+ numberOfLBnodes,
+ isEvenTimestep);
+ getLastCudaError("setRecvFsPre27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void SetRecvFsPostDev27(
+ real* DD,
+ real* bufferFs,
+ int* recvIndex,
+ int buffmax,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep,
+ unsigned int numberOfThreads,
+ cudaStream_t stream)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, buffmax);
+
+ setRecvFsPost27<<< grid.grid, grid.threads, 0, stream >>>(
+ DD,
+ bufferFs,
+ recvIndex,
+ buffmax,
+ neighborX,
+ neighborY,
+ neighborZ,
+ numberOfLBnodes,
+ isEvenTimestep);
+ getLastCudaError("setRecvFsPost27 execution failed");
}
//////////////////////////////////////////////////////////////////////////
void getSendGsDevF3(
- real* G6,
- real* bufferGs,
- int* sendIndex,
- int buffmax,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep,
- unsigned int numberOfThreads)
-{
- int Grid = (buffmax / numberOfThreads) + 1;
- int Grid1, Grid2;
- if (Grid > 512)
- {
- Grid1 = 512;
- Grid2 = (Grid / Grid1) + 1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 grid(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1);
-
- getSendGsF3 <<< grid, threads >>> (
- G6,
- bufferGs,
- sendIndex,
- buffmax,
- neighborX,
- neighborY,
- neighborZ,
- size_Mat,
- isEvenTimestep);
- getLastCudaError("getSendGsF3 execution failed");
+ real* G6,
+ real* bufferGs,
+ int* sendIndex,
+ int buffmax,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep,
+ unsigned int numberOfThreads)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, buffmax);
+
+ getSendGsF3 <<< grid.grid, grid.threads >>> (
+ G6,
+ bufferGs,
+ sendIndex,
+ buffmax,
+ neighborX,
+ neighborY,
+ neighborZ,
+ numberOfLBnodes,
+ isEvenTimestep);
+ getLastCudaError("getSendGsF3 execution failed");
}
//////////////////////////////////////////////////////////////////////////
void setRecvGsDevF3(
- real* G6,
- real* bufferGs,
- int* recvIndex,
- int buffmax,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep,
- unsigned int numberOfThreads)
-{
- int Grid = (buffmax / numberOfThreads) + 1;
- int Grid1, Grid2;
- if (Grid > 512)
- {
- Grid1 = 512;
- Grid2 = (Grid / Grid1) + 1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 grid(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1);
-
- setRecvGsF3 <<< grid, threads >>> (
- G6,
- bufferGs,
- recvIndex,
- buffmax,
- neighborX,
- neighborY,
- neighborZ,
- size_Mat,
- isEvenTimestep);
- getLastCudaError("setRecvGsF3 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void WallFuncDev27(unsigned int numberOfThreads,
- real* vx,
- real* vy,
- real* vz,
- real* DD,
- int* k_Q,
- real* QQ,
- unsigned int numberOfBCnodes,
- real om1,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep)
-{
- int Grid = (numberOfBCnodes / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 gridQ(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- WallFunction27<<< gridQ, threads >>> (
- vx,
- vy,
- vz,
- DD,
- k_Q,
- QQ,
- numberOfBCnodes,
- om1,
- neighborX,
- neighborY,
- neighborZ,
- size_Mat,
- isEvenTimestep);
- getLastCudaError("WallFunction27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void SetOutputWallVelocitySP27(unsigned int numberOfThreads,
- real* vxD,
- real* vyD,
- real* vzD,
- real* vxWall,
- real* vyWall,
- real* vzWall,
- int numberOfWallNodes,
- int* kWallNodes,
- real* rhoD,
- real* pressD,
- unsigned int* geoD,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- real* DD,
- bool isEvenTimestep)
-{
- int Grid = (numberOfWallNodes / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 gridQ(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- LBSetOutputWallVelocitySP27<<< gridQ, threads >>> ( vxD,
- vyD,
- vzD,
- vxWall,
- vyWall,
- vzWall,
- numberOfWallNodes,
- kWallNodes,
- rhoD,
- pressD,
- geoD,
- neighborX,
- neighborY,
- neighborZ,
- size_Mat,
- DD,
- isEvenTimestep);
- getLastCudaError("LBSetOutputWallVelocitySP27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void GetVelotoForce27(unsigned int numberOfThreads,
- real* DD,
- int* bcIndex,
- int nonAtBC,
- real* Vx,
- real* Vy,
- real* Vz,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep)
-{
- int Grid = (nonAtBC / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 gridQ(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- GetVeloforForcing27<<< gridQ, threads >>> (DD,
- bcIndex,
- nonAtBC,
- Vx,
- Vy,
- Vz,
- neighborX,
- neighborY,
- neighborZ,
- size_Mat,
- isEvenTimestep);
- getLastCudaError("GetVeloforForcing27 execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void InitParticlesDevice(real* coordX,
- real* coordY,
- real* coordZ,
- real* coordParticleXlocal,
- real* coordParticleYlocal,
- real* coordParticleZlocal,
- real* coordParticleXglobal,
- real* coordParticleYglobal,
- real* coordParticleZglobal,
- real* veloParticleX,
- real* veloParticleY,
- real* veloParticleZ,
- real* randArray,
- unsigned int* particleID,
- unsigned int* cellBaseID,
- unsigned int* bcMatD,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int* neighborWSB,
- int level,
- unsigned int numberOfParticles,
- unsigned int size_Mat,
- unsigned int numberOfThreads)
-{
- int Grid = (numberOfParticles / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 gridQ(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- InitParticles<<< gridQ, threads >>> (coordX,
- coordY,
- coordZ,
- coordParticleXlocal,
- coordParticleYlocal,
- coordParticleZlocal,
- coordParticleXglobal,
- coordParticleYglobal,
- coordParticleZglobal,
- veloParticleX,
- veloParticleY,
- veloParticleZ,
- randArray,
- particleID,
- cellBaseID,
- bcMatD,
- neighborX,
- neighborY,
- neighborZ,
- neighborWSB,
- level,
- numberOfParticles,
- size_Mat);
- getLastCudaError("InitParticles execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void MoveParticlesDevice(real* coordX,
- real* coordY,
- real* coordZ,
- real* coordParticleXlocal,
- real* coordParticleYlocal,
- real* coordParticleZlocal,
- real* coordParticleXglobal,
- real* coordParticleYglobal,
- real* coordParticleZglobal,
- real* veloParticleX,
- real* veloParticleY,
- real* veloParticleZ,
- real* DD,
- real omega,
- unsigned int* particleID,
- unsigned int* cellBaseID,
- unsigned int* bcMatD,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int* neighborWSB,
- int level,
- unsigned int timestep,
- unsigned int numberOfTimesteps,
- unsigned int numberOfParticles,
- unsigned int size_Mat,
- unsigned int numberOfThreads,
- bool isEvenTimestep)
-{
- int Grid = (numberOfParticles / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 gridQ(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- MoveParticles<<< gridQ, threads >>> (coordX,
- coordY,
- coordZ,
- coordParticleXlocal,
- coordParticleYlocal,
- coordParticleZlocal,
- coordParticleXglobal,
- coordParticleYglobal,
- coordParticleZglobal,
- veloParticleX,
- veloParticleY,
- veloParticleZ,
- DD,
- omega,
- particleID,
- cellBaseID,
- bcMatD,
- neighborX,
- neighborY,
- neighborZ,
- neighborWSB,
- level,
- timestep,
- numberOfTimesteps,
- numberOfParticles,
- size_Mat,
- isEvenTimestep);
- getLastCudaError("MoveParticles execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void initRandomDevice(curandState* state,
- unsigned int size_Mat,
- unsigned int numberOfThreads)
-{
- int Grid = (size_Mat / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 gridQ(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- initRandom<<< gridQ, threads >>> (state);
- getLastCudaError("initRandom execution failed");
-}
-//////////////////////////////////////////////////////////////////////////
-void generateRandomValuesDevice( curandState* state,
- unsigned int size_Mat,
- real* randArray,
- unsigned int numberOfThreads)
-{
- int Grid = (size_Mat / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 gridQ(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- generateRandomValues<<< gridQ, threads >>> (state,randArray);
- getLastCudaError("generateRandomValues execution failed");
+ real* G6,
+ real* bufferGs,
+ int* recvIndex,
+ int buffmax,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep,
+ unsigned int numberOfThreads)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, buffmax);
+
+ setRecvGsF3 <<< grid.grid, grid.threads >>> (
+ G6,
+ bufferGs,
+ recvIndex,
+ buffmax,
+ neighborX,
+ neighborY,
+ neighborZ,
+ numberOfLBnodes,
+ isEvenTimestep);
+ getLastCudaError("setRecvGsF3 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void WallFuncDev27(
+ unsigned int numberOfThreads,
+ real* vx,
+ real* vy,
+ real* vz,
+ real* DD,
+ int* k_Q,
+ real* QQ,
+ unsigned int numberOfBCnodes,
+ real om1,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, numberOfBCnodes);
+
+ WallFunction27<<< grid.grid, grid.threads >>> (
+ vx,
+ vy,
+ vz,
+ DD,
+ k_Q,
+ QQ,
+ numberOfBCnodes,
+ om1,
+ neighborX,
+ neighborY,
+ neighborZ,
+ numberOfLBnodes,
+ isEvenTimestep);
+ getLastCudaError("WallFunction27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void SetOutputWallVelocitySP27(
+ unsigned int numberOfThreads,
+ real* vxD,
+ real* vyD,
+ real* vzD,
+ real* vxWall,
+ real* vyWall,
+ real* vzWall,
+ int numberOfWallNodes,
+ int* kWallNodes,
+ real* rhoD,
+ real* pressD,
+ unsigned int* geoD,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ real* DD,
+ bool isEvenTimestep)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, numberOfWallNodes);
+
+ LBSetOutputWallVelocitySP27<<< grid.grid, grid.threads >>> (
+ vxD,
+ vyD,
+ vzD,
+ vxWall,
+ vyWall,
+ vzWall,
+ numberOfWallNodes,
+ kWallNodes,
+ rhoD,
+ pressD,
+ geoD,
+ neighborX,
+ neighborY,
+ neighborZ,
+ numberOfLBnodes,
+ DD,
+ isEvenTimestep);
+ getLastCudaError("LBSetOutputWallVelocitySP27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void GetVelotoForce27(
+ unsigned int numberOfThreads,
+ real* DD,
+ int* bcIndex,
+ int nonAtBC,
+ real* Vx,
+ real* Vy,
+ real* Vz,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, nonAtBC);
+
+ GetVeloforForcing27<<< grid.grid, grid.threads >>> (
+ DD,
+ bcIndex,
+ nonAtBC,
+ Vx,
+ Vy,
+ Vz,
+ neighborX,
+ neighborY,
+ neighborZ,
+ numberOfLBnodes,
+ isEvenTimestep);
+ getLastCudaError("GetVeloforForcing27 execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void InitParticlesDevice(
+ real* coordX,
+ real* coordY,
+ real* coordZ,
+ real* coordParticleXlocal,
+ real* coordParticleYlocal,
+ real* coordParticleZlocal,
+ real* coordParticleXglobal,
+ real* coordParticleYglobal,
+ real* coordParticleZglobal,
+ real* veloParticleX,
+ real* veloParticleY,
+ real* veloParticleZ,
+ real* randArray,
+ unsigned int* particleID,
+ unsigned int* cellBaseID,
+ unsigned int* bcMatD,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned int* neighborWSB,
+ int level,
+ unsigned int numberOfParticles,
+ unsigned long long numberOfLBnodes,
+ unsigned int numberOfThreads)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, numberOfParticles);
+
+ InitParticles<<< grid.grid, grid.threads >>> (
+ coordX,
+ coordY,
+ coordZ,
+ coordParticleXlocal,
+ coordParticleYlocal,
+ coordParticleZlocal,
+ coordParticleXglobal,
+ coordParticleYglobal,
+ coordParticleZglobal,
+ veloParticleX,
+ veloParticleY,
+ veloParticleZ,
+ randArray,
+ particleID,
+ cellBaseID,
+ bcMatD,
+ neighborX,
+ neighborY,
+ neighborZ,
+ neighborWSB,
+ level,
+ numberOfParticles,
+ numberOfLBnodes);
+ getLastCudaError("InitParticles execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void MoveParticlesDevice(
+ real* coordX,
+ real* coordY,
+ real* coordZ,
+ real* coordParticleXlocal,
+ real* coordParticleYlocal,
+ real* coordParticleZlocal,
+ real* coordParticleXglobal,
+ real* coordParticleYglobal,
+ real* coordParticleZglobal,
+ real* veloParticleX,
+ real* veloParticleY,
+ real* veloParticleZ,
+ real* DD,
+ real omega,
+ unsigned int* particleID,
+ unsigned int* cellBaseID,
+ unsigned int* bcMatD,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned int* neighborWSB,
+ int level,
+ unsigned int timestep,
+ unsigned int numberOfTimesteps,
+ unsigned int numberOfParticles,
+ unsigned long long numberOfLBnodes,
+ unsigned int numberOfThreads,
+ bool isEvenTimestep)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, numberOfParticles);
+
+ MoveParticles<<< grid.grid, grid.threads >>> (
+ coordX,
+ coordY,
+ coordZ,
+ coordParticleXlocal,
+ coordParticleYlocal,
+ coordParticleZlocal,
+ coordParticleXglobal,
+ coordParticleYglobal,
+ coordParticleZglobal,
+ veloParticleX,
+ veloParticleY,
+ veloParticleZ,
+ DD,
+ omega,
+ particleID,
+ cellBaseID,
+ bcMatD,
+ neighborX,
+ neighborY,
+ neighborZ,
+ neighborWSB,
+ level,
+ timestep,
+ numberOfTimesteps,
+ numberOfParticles,
+ numberOfLBnodes,
+ isEvenTimestep);
+ getLastCudaError("MoveParticles execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void initRandomDevice(
+ curandState* state,
+ unsigned long long numberOfLBnodes,
+ unsigned int numberOfThreads)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, numberOfLBnodes);
+ initRandom<<< grid.grid, grid.threads >>> (state);
+ getLastCudaError("initRandom execution failed");
+}
+//////////////////////////////////////////////////////////////////////////
+void generateRandomValuesDevice(
+ curandState* state,
+ unsigned long long numberOfLBnodes,
+ real* randArray,
+ unsigned int numberOfThreads)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, numberOfLBnodes);
+ generateRandomValues<<< grid.grid, grid.threads >>> (state,randArray);
+ getLastCudaError("generateRandomValues execution failed");
}
//////////////////////////////////////////////////////////////////////////
void CalcTurbulenceIntensityDevice(
- real* vxx,
- real* vyy,
- real* vzz,
- real* vxy,
- real* vxz,
- real* vyz,
- real* vx_mean,
- real* vy_mean,
- real* vz_mean,
- real* DD,
- uint* typeOfGridNode,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep,
- uint numberOfThreads)
-{
- int Grid = (size_Mat / numberOfThreads)+1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid/Grid1)+1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 gridQ(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1 );
-
- CalcTurbulenceIntensity<<<gridQ, threads>>>(
- vxx,
- vyy,
- vzz,
- vxy,
- vxz,
- vyz,
- vx_mean,
- vy_mean,
- vz_mean,
- DD,
- typeOfGridNode,
- neighborX,
- neighborY,
- neighborZ,
- size_Mat,
- isEvenTimestep);
-
- getLastCudaError("CalcTurbulenceIntensity execution failed");
+ real* vxx,
+ real* vyy,
+ real* vzz,
+ real* vxy,
+ real* vxz,
+ real* vyz,
+ real* vx_mean,
+ real* vy_mean,
+ real* vz_mean,
+ real* DD,
+ uint* typeOfGridNode,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep,
+ uint numberOfThreads)
+{
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(numberOfThreads, numberOfLBnodes);
+ CalcTurbulenceIntensity<<<grid.grid, grid.threads>>>(
+ vxx,
+ vyy,
+ vzz,
+ vxy,
+ vxz,
+ vyz,
+ vx_mean,
+ vy_mean,
+ vz_mean,
+ DD,
+ typeOfGridNode,
+ neighborX,
+ neighborY,
+ neighborZ,
+ numberOfLBnodes,
+ isEvenTimestep);
+ getLastCudaError("CalcTurbulenceIntensity execution failed");
}
-
-
-
-
-
-
-
-
-
-
-
-
-
diff --git a/src/gpu/VirtualFluids_GPU/GPU/NoSlipBCs27.cu b/src/gpu/VirtualFluids_GPU/GPU/NoSlipBCs27.cu
index 314687c4b29a32962b386d7c083f72b754388e5b..79dedee58afb7b11c4c3ede9911f54df65cf859f 100644
--- a/src/gpu/VirtualFluids_GPU/GPU/NoSlipBCs27.cu
+++ b/src/gpu/VirtualFluids_GPU/GPU/NoSlipBCs27.cu
@@ -1,92 +1,117 @@
-// _ ___ __ __________ _ __ ______________ __
-// | | / (_)____/ /___ ______ _/ / ____/ /_ __(_)___/ /____ / ___/ __ / / / /
-// | | / / / ___/ __/ / / / __ `/ / /_ / / / / / / __ / ___/ / /___/ /_/ / / / /
-// | |/ / / / / /_/ /_/ / /_/ / / __/ / / /_/ / / /_/ (__ ) / /_) / ____/ /__/ /
-// |___/_/_/ \__/\__,_/\__,_/_/_/ /_/\__,_/_/\__,_/____/ \____/_/ \_____/
+//=======================================================================================
+// ____ ____ __ ______ __________ __ __ __ __
+// \ \ | | | | | _ \ |___ ___| | | | | / \ | |
+// \ \ | | | | | |_) | | | | | | | / \ | |
+// \ \ | | | | | _ / | | | | | | / /\ \ | |
+// \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
+// \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
+// \ \ | | ________________________________________________________________
+// \ \ | | | ______________________________________________________________|
+// \ \| | | | __ __ __ __ ______ _______
+// \ | | |_____ | | | | | | | | | _ \ / _____)
+// \ | | _____| | | | | | | | | | | \ \ \_______
+// \ | | | | |_____ | \_/ | | | | |_/ / _____ |
+// \ _____| |__| |________| \_______/ |__| |______/ (_______/
//
-//////////////////////////////////////////////////////////////////////////
-/* Device code */
+// This file is part of VirtualFluids. VirtualFluids is free software: you can
+// redistribute it and/or modify it under the terms of the GNU General Public
+// License as published by the Free Software Foundation, either version 3 of
+// the License, or (at your option) any later version.
+//
+// VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
+// ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+// FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
+// for more details.
+//
+// You should have received a copy of the GNU General Public License along
+// with VirtualFluids (see COPYING.txt). If not, see <http://www.gnu.org/licenses/>.
+//
+//! \file NoSlipBCs27.cu
+//! \ingroup GPU
+//! \author Martin Schoenherr, Anna Wellmann
+//======================================================================================
#include "LBM/LB.h"
#include "lbm/constants/D3Q27.h"
#include <lbm/constants/NumericConstants.h>
-#include "KernelUtilities.h"
+#include "LBM/GPUHelperFunctions/KernelUtilities.h"
using namespace vf::lbm::constant;
using namespace vf::lbm::dir;
+using namespace vf::gpu;
//////////////////////////////////////////////////////////////////////////////
__global__ void QDevice3rdMomentsComp27(
- real* distributions,
- int* subgridDistanceIndices,
- real* subgridDistances,
- unsigned int numberOfBCnodes,
- real omega,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int numberOfLBnodes,
- bool isEvenTimestep)
+ real* distributions,
+ int* subgridDistanceIndices,
+ real* subgridDistances,
+ unsigned int numberOfBCnodes,
+ real omega,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep)
{
Distributions27 D;
if (isEvenTimestep==true)
{
- D.f[DIR_P00 ] = &distributions[DIR_P00 *numberOfLBnodes];
- D.f[DIR_M00 ] = &distributions[DIR_M00 *numberOfLBnodes];
- D.f[DIR_0P0 ] = &distributions[DIR_0P0 *numberOfLBnodes];
- D.f[DIR_0M0 ] = &distributions[DIR_0M0 *numberOfLBnodes];
- D.f[DIR_00P ] = &distributions[DIR_00P *numberOfLBnodes];
- D.f[DIR_00M ] = &distributions[DIR_00M *numberOfLBnodes];
- D.f[DIR_PP0 ] = &distributions[DIR_PP0 *numberOfLBnodes];
- D.f[DIR_MM0 ] = &distributions[DIR_MM0 *numberOfLBnodes];
- D.f[DIR_PM0 ] = &distributions[DIR_PM0 *numberOfLBnodes];
- D.f[DIR_MP0 ] = &distributions[DIR_MP0 *numberOfLBnodes];
- D.f[DIR_P0P ] = &distributions[DIR_P0P *numberOfLBnodes];
- D.f[DIR_M0M ] = &distributions[DIR_M0M *numberOfLBnodes];
- D.f[DIR_P0M ] = &distributions[DIR_P0M *numberOfLBnodes];
- D.f[DIR_M0P ] = &distributions[DIR_M0P *numberOfLBnodes];
- D.f[DIR_0PP ] = &distributions[DIR_0PP *numberOfLBnodes];
- D.f[DIR_0MM ] = &distributions[DIR_0MM *numberOfLBnodes];
- D.f[DIR_0PM ] = &distributions[DIR_0PM *numberOfLBnodes];
- D.f[DIR_0MP ] = &distributions[DIR_0MP *numberOfLBnodes];
- D.f[DIR_000] = &distributions[DIR_000*numberOfLBnodes];
- D.f[DIR_PPP ] = &distributions[DIR_PPP *numberOfLBnodes];
- D.f[DIR_MMP ] = &distributions[DIR_MMP *numberOfLBnodes];
- D.f[DIR_PMP ] = &distributions[DIR_PMP *numberOfLBnodes];
- D.f[DIR_MPP ] = &distributions[DIR_MPP *numberOfLBnodes];
- D.f[DIR_PPM ] = &distributions[DIR_PPM *numberOfLBnodes];
- D.f[DIR_MMM ] = &distributions[DIR_MMM *numberOfLBnodes];
- D.f[DIR_PMM ] = &distributions[DIR_PMM *numberOfLBnodes];
- D.f[DIR_MPM ] = &distributions[DIR_MPM *numberOfLBnodes];
+ D.f[DIR_P00] = &distributions[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &distributions[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &distributions[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &distributions[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &distributions[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &distributions[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &distributions[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &distributions[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &distributions[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &distributions[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &distributions[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &distributions[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &distributions[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &distributions[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &distributions[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &distributions[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &distributions[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &distributions[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &distributions[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &distributions[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &distributions[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &distributions[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &distributions[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &distributions[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &distributions[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &distributions[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &distributions[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00 ] = &distributions[DIR_P00 *numberOfLBnodes];
- D.f[DIR_P00 ] = &distributions[DIR_M00 *numberOfLBnodes];
- D.f[DIR_0M0 ] = &distributions[DIR_0P0 *numberOfLBnodes];
- D.f[DIR_0P0 ] = &distributions[DIR_0M0 *numberOfLBnodes];
- D.f[DIR_00M ] = &distributions[DIR_00P *numberOfLBnodes];
- D.f[DIR_00P ] = &distributions[DIR_00M *numberOfLBnodes];
- D.f[DIR_MM0 ] = &distributions[DIR_PP0 *numberOfLBnodes];
- D.f[DIR_PP0 ] = &distributions[DIR_MM0 *numberOfLBnodes];
- D.f[DIR_MP0 ] = &distributions[DIR_PM0 *numberOfLBnodes];
- D.f[DIR_PM0 ] = &distributions[DIR_MP0 *numberOfLBnodes];
- D.f[DIR_M0M ] = &distributions[DIR_P0P *numberOfLBnodes];
- D.f[DIR_P0P ] = &distributions[DIR_M0M *numberOfLBnodes];
- D.f[DIR_M0P ] = &distributions[DIR_P0M *numberOfLBnodes];
- D.f[DIR_P0M ] = &distributions[DIR_M0P *numberOfLBnodes];
- D.f[DIR_0MM ] = &distributions[DIR_0PP *numberOfLBnodes];
- D.f[DIR_0PP ] = &distributions[DIR_0MM *numberOfLBnodes];
- D.f[DIR_0MP ] = &distributions[DIR_0PM *numberOfLBnodes];
- D.f[DIR_0PM ] = &distributions[DIR_0MP *numberOfLBnodes];
- D.f[DIR_000] = &distributions[DIR_000*numberOfLBnodes];
- D.f[DIR_PPP ] = &distributions[DIR_MMM *numberOfLBnodes];
- D.f[DIR_MMP ] = &distributions[DIR_PPM *numberOfLBnodes];
- D.f[DIR_PMP ] = &distributions[DIR_MPM *numberOfLBnodes];
- D.f[DIR_MPP ] = &distributions[DIR_PMM *numberOfLBnodes];
- D.f[DIR_PPM ] = &distributions[DIR_MMP *numberOfLBnodes];
- D.f[DIR_MMM ] = &distributions[DIR_PPP *numberOfLBnodes];
- D.f[DIR_PMM ] = &distributions[DIR_MPP *numberOfLBnodes];
- D.f[DIR_MPM ] = &distributions[DIR_PMP *numberOfLBnodes];
+ D.f[DIR_M00] = &distributions[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &distributions[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &distributions[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &distributions[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &distributions[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &distributions[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &distributions[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &distributions[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &distributions[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &distributions[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &distributions[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &distributions[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &distributions[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &distributions[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &distributions[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &distributions[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &distributions[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &distributions[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &distributions[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &distributions[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MMP] = &distributions[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PMP] = &distributions[DIR_MPM * numberOfLBnodes];
+ D.f[DIR_MPP] = &distributions[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PPM] = &distributions[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MMM] = &distributions[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PMM] = &distributions[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MPM] = &distributions[DIR_PMP * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////
const unsigned x = threadIdx.x; // Globaler x-Index
@@ -107,24 +132,24 @@ __global__ void QDevice3rdMomentsComp27(
*q_dirBE, *q_dirTW, *q_dirTN, *q_dirBS, *q_dirBN, *q_dirTS,
*q_dirTNE, *q_dirTSW, *q_dirTSE, *q_dirTNW, *q_dirBNE, *q_dirBSW,
*q_dirBSE, *q_dirBNW;
- q_dirE = &subgridDistances[DIR_P00 * numberOfBCnodes];
- q_dirW = &subgridDistances[DIR_M00 * numberOfBCnodes];
- q_dirN = &subgridDistances[DIR_0P0 * numberOfBCnodes];
- q_dirS = &subgridDistances[DIR_0M0 * numberOfBCnodes];
- q_dirT = &subgridDistances[DIR_00P * numberOfBCnodes];
- q_dirB = &subgridDistances[DIR_00M * numberOfBCnodes];
- q_dirNE = &subgridDistances[DIR_PP0 * numberOfBCnodes];
- q_dirSW = &subgridDistances[DIR_MM0 * numberOfBCnodes];
- q_dirSE = &subgridDistances[DIR_PM0 * numberOfBCnodes];
- q_dirNW = &subgridDistances[DIR_MP0 * numberOfBCnodes];
- q_dirTE = &subgridDistances[DIR_P0P * numberOfBCnodes];
- q_dirBW = &subgridDistances[DIR_M0M * numberOfBCnodes];
- q_dirBE = &subgridDistances[DIR_P0M * numberOfBCnodes];
- q_dirTW = &subgridDistances[DIR_M0P * numberOfBCnodes];
- q_dirTN = &subgridDistances[DIR_0PP * numberOfBCnodes];
- q_dirBS = &subgridDistances[DIR_0MM * numberOfBCnodes];
- q_dirBN = &subgridDistances[DIR_0PM * numberOfBCnodes];
- q_dirTS = &subgridDistances[DIR_0MP * numberOfBCnodes];
+ q_dirE = &subgridDistances[DIR_P00 * numberOfBCnodes];
+ q_dirW = &subgridDistances[DIR_M00 * numberOfBCnodes];
+ q_dirN = &subgridDistances[DIR_0P0 * numberOfBCnodes];
+ q_dirS = &subgridDistances[DIR_0M0 * numberOfBCnodes];
+ q_dirT = &subgridDistances[DIR_00P * numberOfBCnodes];
+ q_dirB = &subgridDistances[DIR_00M * numberOfBCnodes];
+ q_dirNE = &subgridDistances[DIR_PP0 * numberOfBCnodes];
+ q_dirSW = &subgridDistances[DIR_MM0 * numberOfBCnodes];
+ q_dirSE = &subgridDistances[DIR_PM0 * numberOfBCnodes];
+ q_dirNW = &subgridDistances[DIR_MP0 * numberOfBCnodes];
+ q_dirTE = &subgridDistances[DIR_P0P * numberOfBCnodes];
+ q_dirBW = &subgridDistances[DIR_M0M * numberOfBCnodes];
+ q_dirBE = &subgridDistances[DIR_P0M * numberOfBCnodes];
+ q_dirTW = &subgridDistances[DIR_M0P * numberOfBCnodes];
+ q_dirTN = &subgridDistances[DIR_0PP * numberOfBCnodes];
+ q_dirBS = &subgridDistances[DIR_0MM * numberOfBCnodes];
+ q_dirBN = &subgridDistances[DIR_0PM * numberOfBCnodes];
+ q_dirTS = &subgridDistances[DIR_0MP * numberOfBCnodes];
q_dirTNE = &subgridDistances[DIR_PPP * numberOfBCnodes];
q_dirTSW = &subgridDistances[DIR_MMP * numberOfBCnodes];
q_dirTSE = &subgridDistances[DIR_PMP * numberOfBCnodes];
@@ -167,32 +192,32 @@ __global__ void QDevice3rdMomentsComp27(
real f_E, f_W, f_N, f_S, f_T, f_B, f_NE, f_SW, f_SE, f_NW, f_TE, f_BW, f_BE,
f_TW, f_TN, f_BS, f_BN, f_TS, f_TNE, f_TSW, f_TSE, f_TNW, f_BNE, f_BSW, f_BSE, f_BNW;
- f_W = (D.f[DIR_P00 ])[ke ];
- f_E = (D.f[DIR_M00 ])[kw ];
- f_S = (D.f[DIR_0P0 ])[kn ];
- f_N = (D.f[DIR_0M0 ])[ks ];
- f_B = (D.f[DIR_00P ])[kt ];
- f_T = (D.f[DIR_00M ])[kb ];
- f_SW = (D.f[DIR_PP0 ])[kne ];
- f_NE = (D.f[DIR_MM0 ])[ksw ];
- f_NW = (D.f[DIR_PM0 ])[kse ];
- f_SE = (D.f[DIR_MP0 ])[knw ];
- f_BW = (D.f[DIR_P0P ])[kte ];
- f_TE = (D.f[DIR_M0M ])[kbw ];
- f_TW = (D.f[DIR_P0M ])[kbe ];
- f_BE = (D.f[DIR_M0P ])[ktw ];
- f_BS = (D.f[DIR_0PP ])[ktn ];
- f_TN = (D.f[DIR_0MM ])[kbs ];
- f_TS = (D.f[DIR_0PM ])[kbn ];
- f_BN = (D.f[DIR_0MP ])[kts ];
- f_BSW = (D.f[DIR_PPP ])[ktne ];
- f_BNE = (D.f[DIR_MMP ])[ktsw ];
- f_BNW = (D.f[DIR_PMP ])[ktse ];
- f_BSE = (D.f[DIR_MPP ])[ktnw ];
- f_TSW = (D.f[DIR_PPM ])[kbne ];
- f_TNE = (D.f[DIR_MMM ])[kbsw ];
- f_TNW = (D.f[DIR_PMM ])[kbse ];
- f_TSE = (D.f[DIR_MPM ])[kbnw ];
+ f_W = (D.f[DIR_P00])[ke ];
+ f_E = (D.f[DIR_M00])[kw ];
+ f_S = (D.f[DIR_0P0])[kn ];
+ f_N = (D.f[DIR_0M0])[ks ];
+ f_B = (D.f[DIR_00P])[kt ];
+ f_T = (D.f[DIR_00M])[kb ];
+ f_SW = (D.f[DIR_PP0])[kne ];
+ f_NE = (D.f[DIR_MM0])[ksw ];
+ f_NW = (D.f[DIR_PM0])[kse ];
+ f_SE = (D.f[DIR_MP0])[knw ];
+ f_BW = (D.f[DIR_P0P])[kte ];
+ f_TE = (D.f[DIR_M0M])[kbw ];
+ f_TW = (D.f[DIR_P0M])[kbe ];
+ f_BE = (D.f[DIR_M0P])[ktw ];
+ f_BS = (D.f[DIR_0PP])[ktn ];
+ f_TN = (D.f[DIR_0MM])[kbs ];
+ f_TS = (D.f[DIR_0PM])[kbn ];
+ f_BN = (D.f[DIR_0MP])[kts ];
+ f_BSW = (D.f[DIR_PPP])[ktne ];
+ f_BNE = (D.f[DIR_MMP])[ktsw ];
+ f_BNW = (D.f[DIR_PMP])[ktse ];
+ f_BSE = (D.f[DIR_MPP])[ktnw ];
+ f_TSW = (D.f[DIR_PPM])[kbne ];
+ f_TNE = (D.f[DIR_MMM])[kbsw ];
+ f_TNW = (D.f[DIR_PMM])[kbse ];
+ f_TSE = (D.f[DIR_MPM])[kbnw ];
////////////////////////////////////////////////////////////////////////////////
real vx1, vx2, vx3, drho, feq, q, m3;
drho = f_TSE + f_TNW + f_TNE + f_TSW + f_BSE + f_BNW + f_BNE + f_BSW +
@@ -217,63 +242,63 @@ __global__ void QDevice3rdMomentsComp27(
//////////////////////////////////////////////////////////////////////////
if (isEvenTimestep==false)
{
- D.f[DIR_P00 ] = &distributions[DIR_P00 *numberOfLBnodes];
- D.f[DIR_M00 ] = &distributions[DIR_M00 *numberOfLBnodes];
- D.f[DIR_0P0 ] = &distributions[DIR_0P0 *numberOfLBnodes];
- D.f[DIR_0M0 ] = &distributions[DIR_0M0 *numberOfLBnodes];
- D.f[DIR_00P ] = &distributions[DIR_00P *numberOfLBnodes];
- D.f[DIR_00M ] = &distributions[DIR_00M *numberOfLBnodes];
- D.f[DIR_PP0 ] = &distributions[DIR_PP0 *numberOfLBnodes];
- D.f[DIR_MM0 ] = &distributions[DIR_MM0 *numberOfLBnodes];
- D.f[DIR_PM0 ] = &distributions[DIR_PM0 *numberOfLBnodes];
- D.f[DIR_MP0 ] = &distributions[DIR_MP0 *numberOfLBnodes];
- D.f[DIR_P0P ] = &distributions[DIR_P0P *numberOfLBnodes];
- D.f[DIR_M0M ] = &distributions[DIR_M0M *numberOfLBnodes];
- D.f[DIR_P0M ] = &distributions[DIR_P0M *numberOfLBnodes];
- D.f[DIR_M0P ] = &distributions[DIR_M0P *numberOfLBnodes];
- D.f[DIR_0PP ] = &distributions[DIR_0PP *numberOfLBnodes];
- D.f[DIR_0MM ] = &distributions[DIR_0MM *numberOfLBnodes];
- D.f[DIR_0PM ] = &distributions[DIR_0PM *numberOfLBnodes];
- D.f[DIR_0MP ] = &distributions[DIR_0MP *numberOfLBnodes];
- D.f[DIR_000] = &distributions[DIR_000*numberOfLBnodes];
- D.f[DIR_PPP ] = &distributions[DIR_PPP *numberOfLBnodes];
- D.f[DIR_MMP ] = &distributions[DIR_MMP *numberOfLBnodes];
- D.f[DIR_PMP ] = &distributions[DIR_PMP *numberOfLBnodes];
- D.f[DIR_MPP ] = &distributions[DIR_MPP *numberOfLBnodes];
- D.f[DIR_PPM ] = &distributions[DIR_PPM *numberOfLBnodes];
- D.f[DIR_MMM ] = &distributions[DIR_MMM *numberOfLBnodes];
- D.f[DIR_PMM ] = &distributions[DIR_PMM *numberOfLBnodes];
- D.f[DIR_MPM ] = &distributions[DIR_MPM *numberOfLBnodes];
+ D.f[DIR_P00] = &distributions[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &distributions[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &distributions[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &distributions[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &distributions[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &distributions[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &distributions[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &distributions[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &distributions[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &distributions[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &distributions[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &distributions[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &distributions[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &distributions[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &distributions[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &distributions[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &distributions[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &distributions[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &distributions[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &distributions[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &distributions[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &distributions[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &distributions[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &distributions[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &distributions[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &distributions[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &distributions[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00 ] = &distributions[DIR_P00 *numberOfLBnodes];
- D.f[DIR_P00 ] = &distributions[DIR_M00 *numberOfLBnodes];
- D.f[DIR_0M0 ] = &distributions[DIR_0P0 *numberOfLBnodes];
- D.f[DIR_0P0 ] = &distributions[DIR_0M0 *numberOfLBnodes];
- D.f[DIR_00M ] = &distributions[DIR_00P *numberOfLBnodes];
- D.f[DIR_00P ] = &distributions[DIR_00M *numberOfLBnodes];
- D.f[DIR_MM0 ] = &distributions[DIR_PP0 *numberOfLBnodes];
- D.f[DIR_PP0 ] = &distributions[DIR_MM0 *numberOfLBnodes];
- D.f[DIR_MP0 ] = &distributions[DIR_PM0 *numberOfLBnodes];
- D.f[DIR_PM0 ] = &distributions[DIR_MP0 *numberOfLBnodes];
- D.f[DIR_M0M ] = &distributions[DIR_P0P *numberOfLBnodes];
- D.f[DIR_P0P ] = &distributions[DIR_M0M *numberOfLBnodes];
- D.f[DIR_M0P ] = &distributions[DIR_P0M *numberOfLBnodes];
- D.f[DIR_P0M ] = &distributions[DIR_M0P *numberOfLBnodes];
- D.f[DIR_0MM ] = &distributions[DIR_0PP *numberOfLBnodes];
- D.f[DIR_0PP ] = &distributions[DIR_0MM *numberOfLBnodes];
- D.f[DIR_0MP ] = &distributions[DIR_0PM *numberOfLBnodes];
- D.f[DIR_0PM ] = &distributions[DIR_0MP *numberOfLBnodes];
- D.f[DIR_000] = &distributions[DIR_000*numberOfLBnodes];
- D.f[DIR_PPP ] = &distributions[DIR_MMM *numberOfLBnodes];
- D.f[DIR_MMP ] = &distributions[DIR_PPM *numberOfLBnodes];
- D.f[DIR_PMP ] = &distributions[DIR_MPM *numberOfLBnodes];
- D.f[DIR_MPP ] = &distributions[DIR_PMM *numberOfLBnodes];
- D.f[DIR_PPM ] = &distributions[DIR_MMP *numberOfLBnodes];
- D.f[DIR_MMM ] = &distributions[DIR_PPP *numberOfLBnodes];
- D.f[DIR_PMM ] = &distributions[DIR_MPP *numberOfLBnodes];
- D.f[DIR_MPM ] = &distributions[DIR_PMP *numberOfLBnodes];
+ D.f[DIR_M00] = &distributions[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &distributions[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &distributions[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &distributions[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &distributions[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &distributions[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &distributions[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &distributions[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &distributions[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &distributions[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &distributions[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &distributions[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &distributions[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &distributions[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &distributions[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &distributions[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &distributions[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &distributions[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &distributions[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &distributions[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MMP] = &distributions[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PMP] = &distributions[DIR_MPM * numberOfLBnodes];
+ D.f[DIR_MPP] = &distributions[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PPM] = &distributions[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MMM] = &distributions[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PMM] = &distributions[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MPM] = &distributions[DIR_PMP * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
//Test
@@ -559,77 +584,78 @@ __global__ void QDevice3rdMomentsComp27(
//////////////////////////////////////////////////////////////////////////////
-__global__ void QDeviceIncompHighNu27(real* DD,
- int* k_Q,
- real* QQ,
- unsigned int numberOfBCnodes,
- real om1,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int numberOfLBnodes,
- bool isEvenTimestep)
+__global__ void QDeviceIncompHighNu27(
+ real* DD,
+ int* k_Q,
+ real* QQ,
+ unsigned int numberOfBCnodes,
+ real om1,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep)
{
Distributions27 D;
if (isEvenTimestep==true)
{
- D.f[DIR_P00 ] = &DD[DIR_P00 *numberOfLBnodes];
- D.f[DIR_M00 ] = &DD[DIR_M00 *numberOfLBnodes];
- D.f[DIR_0P0 ] = &DD[DIR_0P0 *numberOfLBnodes];
- D.f[DIR_0M0 ] = &DD[DIR_0M0 *numberOfLBnodes];
- D.f[DIR_00P ] = &DD[DIR_00P *numberOfLBnodes];
- D.f[DIR_00M ] = &DD[DIR_00M *numberOfLBnodes];
- D.f[DIR_PP0 ] = &DD[DIR_PP0 *numberOfLBnodes];
- D.f[DIR_MM0 ] = &DD[DIR_MM0 *numberOfLBnodes];
- D.f[DIR_PM0 ] = &DD[DIR_PM0 *numberOfLBnodes];
- D.f[DIR_MP0 ] = &DD[DIR_MP0 *numberOfLBnodes];
- D.f[DIR_P0P ] = &DD[DIR_P0P *numberOfLBnodes];
- D.f[DIR_M0M ] = &DD[DIR_M0M *numberOfLBnodes];
- D.f[DIR_P0M ] = &DD[DIR_P0M *numberOfLBnodes];
- D.f[DIR_M0P ] = &DD[DIR_M0P *numberOfLBnodes];
- D.f[DIR_0PP ] = &DD[DIR_0PP *numberOfLBnodes];
- D.f[DIR_0MM ] = &DD[DIR_0MM *numberOfLBnodes];
- D.f[DIR_0PM ] = &DD[DIR_0PM *numberOfLBnodes];
- D.f[DIR_0MP ] = &DD[DIR_0MP *numberOfLBnodes];
- D.f[DIR_000] = &DD[DIR_000*numberOfLBnodes];
- D.f[DIR_PPP ] = &DD[DIR_PPP *numberOfLBnodes];
- D.f[DIR_MMP ] = &DD[DIR_MMP *numberOfLBnodes];
- D.f[DIR_PMP ] = &DD[DIR_PMP *numberOfLBnodes];
- D.f[DIR_MPP ] = &DD[DIR_MPP *numberOfLBnodes];
- D.f[DIR_PPM ] = &DD[DIR_PPM *numberOfLBnodes];
- D.f[DIR_MMM ] = &DD[DIR_MMM *numberOfLBnodes];
- D.f[DIR_PMM ] = &DD[DIR_PMM *numberOfLBnodes];
- D.f[DIR_MPM ] = &DD[DIR_MPM *numberOfLBnodes];
+ D.f[DIR_P00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00 ] = &DD[DIR_P00 *numberOfLBnodes];
- D.f[DIR_P00 ] = &DD[DIR_M00 *numberOfLBnodes];
- D.f[DIR_0M0 ] = &DD[DIR_0P0 *numberOfLBnodes];
- D.f[DIR_0P0 ] = &DD[DIR_0M0 *numberOfLBnodes];
- D.f[DIR_00M ] = &DD[DIR_00P *numberOfLBnodes];
- D.f[DIR_00P ] = &DD[DIR_00M *numberOfLBnodes];
- D.f[DIR_MM0 ] = &DD[DIR_PP0 *numberOfLBnodes];
- D.f[DIR_PP0 ] = &DD[DIR_MM0 *numberOfLBnodes];
- D.f[DIR_MP0 ] = &DD[DIR_PM0 *numberOfLBnodes];
- D.f[DIR_PM0 ] = &DD[DIR_MP0 *numberOfLBnodes];
- D.f[DIR_M0M ] = &DD[DIR_P0P *numberOfLBnodes];
- D.f[DIR_P0P ] = &DD[DIR_M0M *numberOfLBnodes];
- D.f[DIR_M0P ] = &DD[DIR_P0M *numberOfLBnodes];
- D.f[DIR_P0M ] = &DD[DIR_M0P *numberOfLBnodes];
- D.f[DIR_0MM ] = &DD[DIR_0PP *numberOfLBnodes];
- D.f[DIR_0PP ] = &DD[DIR_0MM *numberOfLBnodes];
- D.f[DIR_0MP ] = &DD[DIR_0PM *numberOfLBnodes];
- D.f[DIR_0PM ] = &DD[DIR_0MP *numberOfLBnodes];
- D.f[DIR_000] = &DD[DIR_000*numberOfLBnodes];
- D.f[DIR_PPP ] = &DD[DIR_MMM *numberOfLBnodes];
- D.f[DIR_MMP ] = &DD[DIR_PPM *numberOfLBnodes];
- D.f[DIR_PMP ] = &DD[DIR_MPM *numberOfLBnodes];
- D.f[DIR_MPP ] = &DD[DIR_PMM *numberOfLBnodes];
- D.f[DIR_PPM ] = &DD[DIR_MMP *numberOfLBnodes];
- D.f[DIR_MMM ] = &DD[DIR_PPP *numberOfLBnodes];
- D.f[DIR_PMM ] = &DD[DIR_MPP *numberOfLBnodes];
- D.f[DIR_MPM ] = &DD[DIR_PMP *numberOfLBnodes];
+ D.f[DIR_M00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_MPM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_PMP * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////
const unsigned x = threadIdx.x; // Globaler x-Index
@@ -650,24 +676,24 @@ __global__ void QDeviceIncompHighNu27(real* DD,
*q_dirBE, *q_dirTW, *q_dirTN, *q_dirBS, *q_dirBN, *q_dirTS,
*q_dirTNE, *q_dirTSW, *q_dirTSE, *q_dirTNW, *q_dirBNE, *q_dirBSW,
*q_dirBSE, *q_dirBNW;
- q_dirE = &QQ[DIR_P00 * numberOfBCnodes];
- q_dirW = &QQ[DIR_M00 * numberOfBCnodes];
- q_dirN = &QQ[DIR_0P0 * numberOfBCnodes];
- q_dirS = &QQ[DIR_0M0 * numberOfBCnodes];
- q_dirT = &QQ[DIR_00P * numberOfBCnodes];
- q_dirB = &QQ[DIR_00M * numberOfBCnodes];
- q_dirNE = &QQ[DIR_PP0 * numberOfBCnodes];
- q_dirSW = &QQ[DIR_MM0 * numberOfBCnodes];
- q_dirSE = &QQ[DIR_PM0 * numberOfBCnodes];
- q_dirNW = &QQ[DIR_MP0 * numberOfBCnodes];
- q_dirTE = &QQ[DIR_P0P * numberOfBCnodes];
- q_dirBW = &QQ[DIR_M0M * numberOfBCnodes];
- q_dirBE = &QQ[DIR_P0M * numberOfBCnodes];
- q_dirTW = &QQ[DIR_M0P * numberOfBCnodes];
- q_dirTN = &QQ[DIR_0PP * numberOfBCnodes];
- q_dirBS = &QQ[DIR_0MM * numberOfBCnodes];
- q_dirBN = &QQ[DIR_0PM * numberOfBCnodes];
- q_dirTS = &QQ[DIR_0MP * numberOfBCnodes];
+ q_dirE = &QQ[DIR_P00 * numberOfBCnodes];
+ q_dirW = &QQ[DIR_M00 * numberOfBCnodes];
+ q_dirN = &QQ[DIR_0P0 * numberOfBCnodes];
+ q_dirS = &QQ[DIR_0M0 * numberOfBCnodes];
+ q_dirT = &QQ[DIR_00P * numberOfBCnodes];
+ q_dirB = &QQ[DIR_00M * numberOfBCnodes];
+ q_dirNE = &QQ[DIR_PP0 * numberOfBCnodes];
+ q_dirSW = &QQ[DIR_MM0 * numberOfBCnodes];
+ q_dirSE = &QQ[DIR_PM0 * numberOfBCnodes];
+ q_dirNW = &QQ[DIR_MP0 * numberOfBCnodes];
+ q_dirTE = &QQ[DIR_P0P * numberOfBCnodes];
+ q_dirBW = &QQ[DIR_M0M * numberOfBCnodes];
+ q_dirBE = &QQ[DIR_P0M * numberOfBCnodes];
+ q_dirTW = &QQ[DIR_M0P * numberOfBCnodes];
+ q_dirTN = &QQ[DIR_0PP * numberOfBCnodes];
+ q_dirBS = &QQ[DIR_0MM * numberOfBCnodes];
+ q_dirBN = &QQ[DIR_0PM * numberOfBCnodes];
+ q_dirTS = &QQ[DIR_0MP * numberOfBCnodes];
q_dirTNE = &QQ[DIR_PPP * numberOfBCnodes];
q_dirTSW = &QQ[DIR_MMP * numberOfBCnodes];
q_dirTSE = &QQ[DIR_PMP * numberOfBCnodes];
@@ -710,32 +736,32 @@ __global__ void QDeviceIncompHighNu27(real* DD,
real f_E, f_W, f_N, f_S, f_T, f_B, f_NE, f_SW, f_SE, f_NW, f_TE, f_BW, f_BE,
f_TW, f_TN, f_BS, f_BN, f_TS, f_TNE, f_TSW, f_TSE, f_TNW, f_BNE, f_BSW, f_BSE, f_BNW;
- f_E = (D.f[DIR_P00 ])[ke ];
- f_W = (D.f[DIR_M00 ])[kw ];
- f_N = (D.f[DIR_0P0 ])[kn ];
- f_S = (D.f[DIR_0M0 ])[ks ];
- f_T = (D.f[DIR_00P ])[kt ];
- f_B = (D.f[DIR_00M ])[kb ];
- f_NE = (D.f[DIR_PP0 ])[kne ];
- f_SW = (D.f[DIR_MM0 ])[ksw ];
- f_SE = (D.f[DIR_PM0 ])[kse ];
- f_NW = (D.f[DIR_MP0 ])[knw ];
- f_TE = (D.f[DIR_P0P ])[kte ];
- f_BW = (D.f[DIR_M0M ])[kbw ];
- f_BE = (D.f[DIR_P0M ])[kbe ];
- f_TW = (D.f[DIR_M0P ])[ktw ];
- f_TN = (D.f[DIR_0PP ])[ktn ];
- f_BS = (D.f[DIR_0MM ])[kbs ];
- f_BN = (D.f[DIR_0PM ])[kbn ];
- f_TS = (D.f[DIR_0MP ])[kts ];
- f_TNE = (D.f[DIR_PPP ])[ktne ];
- f_TSW = (D.f[DIR_MMP ])[ktsw ];
- f_TSE = (D.f[DIR_PMP ])[ktse ];
- f_TNW = (D.f[DIR_MPP ])[ktnw ];
- f_BNE = (D.f[DIR_PPM ])[kbne ];
- f_BSW = (D.f[DIR_MMM ])[kbsw ];
- f_BSE = (D.f[DIR_PMM ])[kbse ];
- f_BNW = (D.f[DIR_MPM ])[kbnw ];
+ f_E = (D.f[DIR_P00])[ke ];
+ f_W = (D.f[DIR_M00])[kw ];
+ f_N = (D.f[DIR_0P0])[kn ];
+ f_S = (D.f[DIR_0M0])[ks ];
+ f_T = (D.f[DIR_00P])[kt ];
+ f_B = (D.f[DIR_00M])[kb ];
+ f_NE = (D.f[DIR_PP0])[kne ];
+ f_SW = (D.f[DIR_MM0])[ksw ];
+ f_SE = (D.f[DIR_PM0])[kse ];
+ f_NW = (D.f[DIR_MP0])[knw ];
+ f_TE = (D.f[DIR_P0P])[kte ];
+ f_BW = (D.f[DIR_M0M])[kbw ];
+ f_BE = (D.f[DIR_P0M])[kbe ];
+ f_TW = (D.f[DIR_M0P])[ktw ];
+ f_TN = (D.f[DIR_0PP])[ktn ];
+ f_BS = (D.f[DIR_0MM])[kbs ];
+ f_BN = (D.f[DIR_0PM])[kbn ];
+ f_TS = (D.f[DIR_0MP])[kts ];
+ f_TNE = (D.f[DIR_PPP])[ktne ];
+ f_TSW = (D.f[DIR_MMP])[ktsw ];
+ f_TSE = (D.f[DIR_PMP])[ktse ];
+ f_TNW = (D.f[DIR_MPP])[ktnw ];
+ f_BNE = (D.f[DIR_PPM])[kbne ];
+ f_BSW = (D.f[DIR_MMM])[kbsw ];
+ f_BSE = (D.f[DIR_PMM])[kbse ];
+ f_BNW = (D.f[DIR_MPM])[kbnw ];
////////////////////////////////////////////////////////////////////////////////
real vx1, vx2, vx3, drho, feq, q;
drho = f_TSE + f_TNW + f_TNE + f_TSW + f_BSE + f_BNW + f_BNE + f_BSW +
@@ -760,63 +786,63 @@ __global__ void QDeviceIncompHighNu27(real* DD,
//////////////////////////////////////////////////////////////////////////
if (isEvenTimestep==false)
{
- D.f[DIR_P00 ] = &DD[DIR_P00 *numberOfLBnodes];
- D.f[DIR_M00 ] = &DD[DIR_M00 *numberOfLBnodes];
- D.f[DIR_0P0 ] = &DD[DIR_0P0 *numberOfLBnodes];
- D.f[DIR_0M0 ] = &DD[DIR_0M0 *numberOfLBnodes];
- D.f[DIR_00P ] = &DD[DIR_00P *numberOfLBnodes];
- D.f[DIR_00M ] = &DD[DIR_00M *numberOfLBnodes];
- D.f[DIR_PP0 ] = &DD[DIR_PP0 *numberOfLBnodes];
- D.f[DIR_MM0 ] = &DD[DIR_MM0 *numberOfLBnodes];
- D.f[DIR_PM0 ] = &DD[DIR_PM0 *numberOfLBnodes];
- D.f[DIR_MP0 ] = &DD[DIR_MP0 *numberOfLBnodes];
- D.f[DIR_P0P ] = &DD[DIR_P0P *numberOfLBnodes];
- D.f[DIR_M0M ] = &DD[DIR_M0M *numberOfLBnodes];
- D.f[DIR_P0M ] = &DD[DIR_P0M *numberOfLBnodes];
- D.f[DIR_M0P ] = &DD[DIR_M0P *numberOfLBnodes];
- D.f[DIR_0PP ] = &DD[DIR_0PP *numberOfLBnodes];
- D.f[DIR_0MM ] = &DD[DIR_0MM *numberOfLBnodes];
- D.f[DIR_0PM ] = &DD[DIR_0PM *numberOfLBnodes];
- D.f[DIR_0MP ] = &DD[DIR_0MP *numberOfLBnodes];
- D.f[DIR_000] = &DD[DIR_000*numberOfLBnodes];
- D.f[DIR_PPP ] = &DD[DIR_PPP *numberOfLBnodes];
- D.f[DIR_MMP ] = &DD[DIR_MMP *numberOfLBnodes];
- D.f[DIR_PMP ] = &DD[DIR_PMP *numberOfLBnodes];
- D.f[DIR_MPP ] = &DD[DIR_MPP *numberOfLBnodes];
- D.f[DIR_PPM ] = &DD[DIR_PPM *numberOfLBnodes];
- D.f[DIR_MMM ] = &DD[DIR_MMM *numberOfLBnodes];
- D.f[DIR_PMM ] = &DD[DIR_PMM *numberOfLBnodes];
- D.f[DIR_MPM ] = &DD[DIR_MPM *numberOfLBnodes];
+ D.f[DIR_P00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00 ] = &DD[DIR_P00 *numberOfLBnodes];
- D.f[DIR_P00 ] = &DD[DIR_M00 *numberOfLBnodes];
- D.f[DIR_0M0 ] = &DD[DIR_0P0 *numberOfLBnodes];
- D.f[DIR_0P0 ] = &DD[DIR_0M0 *numberOfLBnodes];
- D.f[DIR_00M ] = &DD[DIR_00P *numberOfLBnodes];
- D.f[DIR_00P ] = &DD[DIR_00M *numberOfLBnodes];
- D.f[DIR_MM0 ] = &DD[DIR_PP0 *numberOfLBnodes];
- D.f[DIR_PP0 ] = &DD[DIR_MM0 *numberOfLBnodes];
- D.f[DIR_MP0 ] = &DD[DIR_PM0 *numberOfLBnodes];
- D.f[DIR_PM0 ] = &DD[DIR_MP0 *numberOfLBnodes];
- D.f[DIR_M0M ] = &DD[DIR_P0P *numberOfLBnodes];
- D.f[DIR_P0P ] = &DD[DIR_M0M *numberOfLBnodes];
- D.f[DIR_M0P ] = &DD[DIR_P0M *numberOfLBnodes];
- D.f[DIR_P0M ] = &DD[DIR_M0P *numberOfLBnodes];
- D.f[DIR_0MM ] = &DD[DIR_0PP *numberOfLBnodes];
- D.f[DIR_0PP ] = &DD[DIR_0MM *numberOfLBnodes];
- D.f[DIR_0MP ] = &DD[DIR_0PM *numberOfLBnodes];
- D.f[DIR_0PM ] = &DD[DIR_0MP *numberOfLBnodes];
- D.f[DIR_000] = &DD[DIR_000*numberOfLBnodes];
- D.f[DIR_PPP ] = &DD[DIR_MMM *numberOfLBnodes];
- D.f[DIR_MMP ] = &DD[DIR_PPM *numberOfLBnodes];
- D.f[DIR_PMP ] = &DD[DIR_MPM *numberOfLBnodes];
- D.f[DIR_MPP ] = &DD[DIR_PMM *numberOfLBnodes];
- D.f[DIR_PPM ] = &DD[DIR_MMP *numberOfLBnodes];
- D.f[DIR_MMM ] = &DD[DIR_PPP *numberOfLBnodes];
- D.f[DIR_PMM ] = &DD[DIR_MPP *numberOfLBnodes];
- D.f[DIR_MPM ] = &DD[DIR_PMP *numberOfLBnodes];
+ D.f[DIR_M00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_MPM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_PMP * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
//Test
@@ -1055,77 +1081,77 @@ __global__ void QDeviceIncompHighNu27(real* DD,
//////////////////////////////////////////////////////////////////////////////
__global__ void QDeviceCompHighNu27(
- real* DD,
- int* k_Q,
- real* QQ,
- unsigned int numberOfBCnodes,
- real om1,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep)
+ real* DD,
+ int* k_Q,
+ real* QQ,
+ unsigned int numberOfBCnodes,
+ real om1,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep)
{
Distributions27 D;
if (isEvenTimestep==true)
{
- D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
+ D.f[DIR_M00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_MPM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_PMP * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////
const unsigned x = threadIdx.x; // Globaler x-Index
@@ -1146,24 +1172,24 @@ __global__ void QDeviceCompHighNu27(
*q_dirBE, *q_dirTW, *q_dirTN, *q_dirBS, *q_dirBN, *q_dirTS,
*q_dirTNE, *q_dirTSW, *q_dirTSE, *q_dirTNW, *q_dirBNE, *q_dirBSW,
*q_dirBSE, *q_dirBNW;
- q_dirE = &QQ[DIR_P00 * numberOfBCnodes];
- q_dirW = &QQ[DIR_M00 * numberOfBCnodes];
- q_dirN = &QQ[DIR_0P0 * numberOfBCnodes];
- q_dirS = &QQ[DIR_0M0 * numberOfBCnodes];
- q_dirT = &QQ[DIR_00P * numberOfBCnodes];
- q_dirB = &QQ[DIR_00M * numberOfBCnodes];
- q_dirNE = &QQ[DIR_PP0 * numberOfBCnodes];
- q_dirSW = &QQ[DIR_MM0 * numberOfBCnodes];
- q_dirSE = &QQ[DIR_PM0 * numberOfBCnodes];
- q_dirNW = &QQ[DIR_MP0 * numberOfBCnodes];
- q_dirTE = &QQ[DIR_P0P * numberOfBCnodes];
- q_dirBW = &QQ[DIR_M0M * numberOfBCnodes];
- q_dirBE = &QQ[DIR_P0M * numberOfBCnodes];
- q_dirTW = &QQ[DIR_M0P * numberOfBCnodes];
- q_dirTN = &QQ[DIR_0PP * numberOfBCnodes];
- q_dirBS = &QQ[DIR_0MM * numberOfBCnodes];
- q_dirBN = &QQ[DIR_0PM * numberOfBCnodes];
- q_dirTS = &QQ[DIR_0MP * numberOfBCnodes];
+ q_dirE = &QQ[DIR_P00 * numberOfBCnodes];
+ q_dirW = &QQ[DIR_M00 * numberOfBCnodes];
+ q_dirN = &QQ[DIR_0P0 * numberOfBCnodes];
+ q_dirS = &QQ[DIR_0M0 * numberOfBCnodes];
+ q_dirT = &QQ[DIR_00P * numberOfBCnodes];
+ q_dirB = &QQ[DIR_00M * numberOfBCnodes];
+ q_dirNE = &QQ[DIR_PP0 * numberOfBCnodes];
+ q_dirSW = &QQ[DIR_MM0 * numberOfBCnodes];
+ q_dirSE = &QQ[DIR_PM0 * numberOfBCnodes];
+ q_dirNW = &QQ[DIR_MP0 * numberOfBCnodes];
+ q_dirTE = &QQ[DIR_P0P * numberOfBCnodes];
+ q_dirBW = &QQ[DIR_M0M * numberOfBCnodes];
+ q_dirBE = &QQ[DIR_P0M * numberOfBCnodes];
+ q_dirTW = &QQ[DIR_M0P * numberOfBCnodes];
+ q_dirTN = &QQ[DIR_0PP * numberOfBCnodes];
+ q_dirBS = &QQ[DIR_0MM * numberOfBCnodes];
+ q_dirBN = &QQ[DIR_0PM * numberOfBCnodes];
+ q_dirTS = &QQ[DIR_0MP * numberOfBCnodes];
q_dirTNE = &QQ[DIR_PPP * numberOfBCnodes];
q_dirTSW = &QQ[DIR_MMP * numberOfBCnodes];
q_dirTSE = &QQ[DIR_PMP * numberOfBCnodes];
@@ -1206,58 +1232,58 @@ __global__ void QDeviceCompHighNu27(
real f_E, f_W, f_N, f_S, f_T, f_B, f_NE, f_SW, f_SE, f_NW, f_TE, f_BW, f_BE,
f_TW, f_TN, f_BS, f_BN, f_TS, f_TNE, f_TSW, f_TSE, f_TNW, f_BNE, f_BSW, f_BSE, f_BNW;
- f_E = (D.f[DIR_P00 ])[ke ];
- f_W = (D.f[DIR_M00 ])[kw ];
- f_N = (D.f[DIR_0P0 ])[kn ];
- f_S = (D.f[DIR_0M0 ])[ks ];
- f_T = (D.f[DIR_00P ])[kt ];
- f_B = (D.f[DIR_00M ])[kb ];
- f_NE = (D.f[DIR_PP0 ])[kne ];
- f_SW = (D.f[DIR_MM0 ])[ksw ];
- f_SE = (D.f[DIR_PM0 ])[kse ];
- f_NW = (D.f[DIR_MP0 ])[knw ];
- f_TE = (D.f[DIR_P0P ])[kte ];
- f_BW = (D.f[DIR_M0M ])[kbw ];
- f_BE = (D.f[DIR_P0M ])[kbe ];
- f_TW = (D.f[DIR_M0P ])[ktw ];
- f_TN = (D.f[DIR_0PP ])[ktn ];
- f_BS = (D.f[DIR_0MM ])[kbs ];
- f_BN = (D.f[DIR_0PM ])[kbn ];
- f_TS = (D.f[DIR_0MP ])[kts ];
- f_TNE = (D.f[DIR_PPP ])[ktne ];
- f_TSW = (D.f[DIR_MMP ])[ktsw ];
- f_TSE = (D.f[DIR_PMP ])[ktse ];
- f_TNW = (D.f[DIR_MPP ])[ktnw ];
- f_BNE = (D.f[DIR_PPM ])[kbne ];
- f_BSW = (D.f[DIR_MMM ])[kbsw ];
- f_BSE = (D.f[DIR_PMM ])[kbse ];
- f_BNW = (D.f[DIR_MPM ])[kbnw ];
- //f_W = (D.f[DIR_P00 ])[ke ];
- //f_E = (D.f[DIR_M00 ])[kw ];
- //f_S = (D.f[DIR_0P0 ])[kn ];
- //f_N = (D.f[DIR_0M0 ])[ks ];
- //f_B = (D.f[DIR_00P ])[kt ];
- //f_T = (D.f[DIR_00M ])[kb ];
- //f_SW = (D.f[DIR_PP0 ])[kne ];
- //f_NE = (D.f[DIR_MM0 ])[ksw ];
- //f_NW = (D.f[DIR_PM0 ])[kse ];
- //f_SE = (D.f[DIR_MP0 ])[knw ];
- //f_BW = (D.f[DIR_P0P ])[kte ];
- //f_TE = (D.f[DIR_M0M ])[kbw ];
- //f_TW = (D.f[DIR_P0M ])[kbe ];
- //f_BE = (D.f[DIR_M0P ])[ktw ];
- //f_BS = (D.f[DIR_0PP ])[ktn ];
- //f_TN = (D.f[DIR_0MM ])[kbs ];
- //f_TS = (D.f[DIR_0PM ])[kbn ];
- //f_BN = (D.f[DIR_0MP ])[kts ];
- //f_BSW = (D.f[DIR_PPP ])[ktne ];
- //f_BNE = (D.f[DIR_MMP ])[ktsw ];
- //f_BNW = (D.f[DIR_PMP ])[ktse ];
- //f_BSE = (D.f[DIR_MPP ])[ktnw ];
- //f_TSW = (D.f[DIR_PPM ])[kbne ];
- //f_TNE = (D.f[DIR_MMM ])[kbsw ];
- //f_TNW = (D.f[DIR_PMM ])[kbse ];
- //f_TSE = (D.f[DIR_MPM ])[kbnw ];
+ f_E = (D.f[DIR_P00])[ke ];
+ f_W = (D.f[DIR_M00])[kw ];
+ f_N = (D.f[DIR_0P0])[kn ];
+ f_S = (D.f[DIR_0M0])[ks ];
+ f_T = (D.f[DIR_00P])[kt ];
+ f_B = (D.f[DIR_00M])[kb ];
+ f_NE = (D.f[DIR_PP0])[kne ];
+ f_SW = (D.f[DIR_MM0])[ksw ];
+ f_SE = (D.f[DIR_PM0])[kse ];
+ f_NW = (D.f[DIR_MP0])[knw ];
+ f_TE = (D.f[DIR_P0P])[kte ];
+ f_BW = (D.f[DIR_M0M])[kbw ];
+ f_BE = (D.f[DIR_P0M])[kbe ];
+ f_TW = (D.f[DIR_M0P])[ktw ];
+ f_TN = (D.f[DIR_0PP])[ktn ];
+ f_BS = (D.f[DIR_0MM])[kbs ];
+ f_BN = (D.f[DIR_0PM])[kbn ];
+ f_TS = (D.f[DIR_0MP])[kts ];
+ f_TNE = (D.f[DIR_PPP])[ktne ];
+ f_TSW = (D.f[DIR_MMP])[ktsw ];
+ f_TSE = (D.f[DIR_PMP])[ktse ];
+ f_TNW = (D.f[DIR_MPP])[ktnw ];
+ f_BNE = (D.f[DIR_PPM])[kbne ];
+ f_BSW = (D.f[DIR_MMM])[kbsw ];
+ f_BSE = (D.f[DIR_PMM])[kbse ];
+ f_BNW = (D.f[DIR_MPM])[kbnw ];
+ //f_W = (D.f[DIR_P00])[ke ];
+ //f_E = (D.f[DIR_M00])[kw ];
+ //f_S = (D.f[DIR_0P0])[kn ];
+ //f_N = (D.f[DIR_0M0])[ks ];
+ //f_B = (D.f[DIR_00P])[kt ];
+ //f_T = (D.f[DIR_00M])[kb ];
+ //f_SW = (D.f[DIR_PP0])[kne ];
+ //f_NE = (D.f[DIR_MM0])[ksw ];
+ //f_NW = (D.f[DIR_PM0])[kse ];
+ //f_SE = (D.f[DIR_MP0])[knw ];
+ //f_BW = (D.f[DIR_P0P])[kte ];
+ //f_TE = (D.f[DIR_M0M])[kbw ];
+ //f_TW = (D.f[DIR_P0M])[kbe ];
+ //f_BE = (D.f[DIR_M0P])[ktw ];
+ //f_BS = (D.f[DIR_0PP])[ktn ];
+ //f_TN = (D.f[DIR_0MM])[kbs ];
+ //f_TS = (D.f[DIR_0PM])[kbn ];
+ //f_BN = (D.f[DIR_0MP])[kts ];
+ //f_BSW = (D.f[DIR_PPP])[ktne ];
+ //f_BNE = (D.f[DIR_MMP])[ktsw ];
+ //f_BNW = (D.f[DIR_PMP])[ktse ];
+ //f_BSE = (D.f[DIR_MPP])[ktnw ];
+ //f_TSW = (D.f[DIR_PPM])[kbne ];
+ //f_TNE = (D.f[DIR_MMM])[kbsw ];
+ //f_TNW = (D.f[DIR_PMM])[kbse ];
+ //f_TSE = (D.f[DIR_MPM])[kbnw ];
////////////////////////////////////////////////////////////////////////////////
real vx1, vx2, vx3, drho, feq, q;
drho = f_TSE + f_TNW + f_TNE + f_TSW + f_BSE + f_BNW + f_BNE + f_BSW +
@@ -1282,63 +1308,63 @@ __global__ void QDeviceCompHighNu27(
//////////////////////////////////////////////////////////////////////////
if (isEvenTimestep==false)
{
- D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
+ D.f[DIR_M00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_MPM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_PMP * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
//Test
@@ -1629,16 +1655,16 @@ __global__ void QDeviceCompHighNu27(
//////////////////////////////////////////////////////////////////////////////
__global__ void QDeviceComp27(
- real* distributions,
- int* subgridDistanceIndices,
- real* subgridDistances,
- unsigned int numberOfBCnodes,
- real omega,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int numberOfLBnodes,
- bool isEvenTimestep)
+ real* distributions,
+ int* subgridDistanceIndices,
+ real* subgridDistances,
+ unsigned int numberOfBCnodes,
+ real omega,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep)
{
//////////////////////////////////////////////////////////////////////////
//! The no-slip boundary condition is executed in the following steps
@@ -1646,16 +1672,9 @@ __global__ void QDeviceComp27(
////////////////////////////////////////////////////////////////////////////////
//! - Get node index coordinates from threadIdx, blockIdx, blockDim and gridDim.
//!
- const unsigned x = threadIdx.x; // global x-index
- const unsigned y = blockIdx.x; // global y-index
- const unsigned z = blockIdx.y; // global z-index
-
- const unsigned nx = blockDim.x;
- const unsigned ny = gridDim.x;
+ const unsigned nodeIndex = getNodeIndex();
- const unsigned k = nx*(ny*z + y) + x;
-
- if(k < numberOfBCnodes)
+ if(nodeIndex < numberOfBCnodes)
{
//////////////////////////////////////////////////////////////////////////
//! - Read distributions: style of reading and writing the distributions from/to stored arrays dependent on timestep is based on the esoteric twist algorithm \ref
@@ -1673,7 +1692,7 @@ __global__ void QDeviceComp27(
////////////////////////////////////////////////////////////////////////////////
//! - Set neighbor indices (necessary for indirect addressing)
//!
- unsigned int indexOfBCnode = subgridDistanceIndices[k];
+ unsigned int indexOfBCnode = subgridDistanceIndices[nodeIndex];
unsigned int kzero= indexOfBCnode;
unsigned int ke = indexOfBCnode;
unsigned int kw = neighborX[indexOfBCnode];
@@ -1705,32 +1724,32 @@ __global__ void QDeviceComp27(
////////////////////////////////////////////////////////////////////////////////
//! - Set local distributions
//!
- real f_W = (dist.f[DIR_P00 ])[ke ];
- real f_E = (dist.f[DIR_M00 ])[kw ];
- real f_S = (dist.f[DIR_0P0 ])[kn ];
- real f_N = (dist.f[DIR_0M0 ])[ks ];
- real f_B = (dist.f[DIR_00P ])[kt ];
- real f_T = (dist.f[DIR_00M ])[kb ];
- real f_SW = (dist.f[DIR_PP0 ])[kne ];
- real f_NE = (dist.f[DIR_MM0 ])[ksw ];
- real f_NW = (dist.f[DIR_PM0 ])[kse ];
- real f_SE = (dist.f[DIR_MP0 ])[knw ];
- real f_BW = (dist.f[DIR_P0P ])[kte ];
- real f_TE = (dist.f[DIR_M0M ])[kbw ];
- real f_TW = (dist.f[DIR_P0M ])[kbe ];
- real f_BE = (dist.f[DIR_M0P ])[ktw ];
- real f_BS = (dist.f[DIR_0PP ])[ktn ];
- real f_TN = (dist.f[DIR_0MM ])[kbs ];
- real f_TS = (dist.f[DIR_0PM ])[kbn ];
- real f_BN = (dist.f[DIR_0MP ])[kts ];
- real f_BSW = (dist.f[DIR_PPP ])[ktne ];
- real f_BNE = (dist.f[DIR_MMP ])[ktsw ];
- real f_BNW = (dist.f[DIR_PMP ])[ktse ];
- real f_BSE = (dist.f[DIR_MPP ])[ktnw ];
- real f_TSW = (dist.f[DIR_PPM ])[kbne ];
- real f_TNE = (dist.f[DIR_MMM ])[kbsw ];
- real f_TNW = (dist.f[DIR_PMM ])[kbse ];
- real f_TSE = (dist.f[DIR_MPM ])[kbnw ];
+ real f_W = (dist.f[DIR_P00])[ke ];
+ real f_E = (dist.f[DIR_M00])[kw ];
+ real f_S = (dist.f[DIR_0P0])[kn ];
+ real f_N = (dist.f[DIR_0M0])[ks ];
+ real f_B = (dist.f[DIR_00P])[kt ];
+ real f_T = (dist.f[DIR_00M])[kb ];
+ real f_SW = (dist.f[DIR_PP0])[kne ];
+ real f_NE = (dist.f[DIR_MM0])[ksw ];
+ real f_NW = (dist.f[DIR_PM0])[kse ];
+ real f_SE = (dist.f[DIR_MP0])[knw ];
+ real f_BW = (dist.f[DIR_P0P])[kte ];
+ real f_TE = (dist.f[DIR_M0M])[kbw ];
+ real f_TW = (dist.f[DIR_P0M])[kbe ];
+ real f_BE = (dist.f[DIR_M0P])[ktw ];
+ real f_BS = (dist.f[DIR_0PP])[ktn ];
+ real f_TN = (dist.f[DIR_0MM])[kbs ];
+ real f_TS = (dist.f[DIR_0PM])[kbn ];
+ real f_BN = (dist.f[DIR_0MP])[kts ];
+ real f_BSW = (dist.f[DIR_PPP])[ktne ];
+ real f_BNE = (dist.f[DIR_MMP])[ktsw ];
+ real f_BNW = (dist.f[DIR_PMP])[ktse ];
+ real f_BSE = (dist.f[DIR_MPP])[ktnw ];
+ real f_TSW = (dist.f[DIR_PPM])[kbne ];
+ real f_TNE = (dist.f[DIR_MMM])[kbsw ];
+ real f_TNW = (dist.f[DIR_PMM])[kbse ];
+ real f_TSE = (dist.f[DIR_MPM])[kbnw ];
////////////////////////////////////////////////////////////////////////////////
//! - Calculate macroscopic quantities
@@ -1761,7 +1780,7 @@ __global__ void QDeviceComp27(
////////////////////////////////////////////////////////////////////////////////
//! - Update distributions with subgrid distance (q) between zero and one
real feq, q, velocityLB;
- q = (subgridD.q[DIR_P00])[k];
+ q = (subgridD.q[DIR_P00])[nodeIndex];
if (q>=c0o1 && q<=c1o1) // only update distribution for q between zero and one
{
velocityLB = vx1;
@@ -1769,7 +1788,7 @@ __global__ void QDeviceComp27(
(dist.f[DIR_M00])[kw] = getInterpolatedDistributionForNoSlipBC(q, f_E, f_W, feq, omega);
}
- q = (subgridD.q[DIR_M00])[k];
+ q = (subgridD.q[DIR_M00])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = -vx1;
@@ -1777,7 +1796,7 @@ __global__ void QDeviceComp27(
(dist.f[DIR_P00])[ke] = getInterpolatedDistributionForNoSlipBC(q, f_W, f_E, feq, omega);
}
- q = (subgridD.q[DIR_0P0])[k];
+ q = (subgridD.q[DIR_0P0])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = vx2;
@@ -1785,7 +1804,7 @@ __global__ void QDeviceComp27(
(dist.f[DIR_0M0])[ks] = getInterpolatedDistributionForNoSlipBC(q, f_N, f_S, feq, omega);
}
- q = (subgridD.q[DIR_0M0])[k];
+ q = (subgridD.q[DIR_0M0])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = -vx2;
@@ -1793,7 +1812,7 @@ __global__ void QDeviceComp27(
(dist.f[DIR_0P0])[kn] = getInterpolatedDistributionForNoSlipBC(q, f_S, f_N, feq, omega);
}
- q = (subgridD.q[DIR_00P])[k];
+ q = (subgridD.q[DIR_00P])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = vx3;
@@ -1801,7 +1820,7 @@ __global__ void QDeviceComp27(
(dist.f[DIR_00M])[kb] = getInterpolatedDistributionForNoSlipBC(q, f_T, f_B, feq, omega);
}
- q = (subgridD.q[DIR_00M])[k];
+ q = (subgridD.q[DIR_00M])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = -vx3;
@@ -1809,7 +1828,7 @@ __global__ void QDeviceComp27(
(dist.f[DIR_00P])[kt] = getInterpolatedDistributionForNoSlipBC(q, f_B, f_T, feq, omega);
}
- q = (subgridD.q[DIR_PP0])[k];
+ q = (subgridD.q[DIR_PP0])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = vx1 + vx2;
@@ -1817,7 +1836,7 @@ __global__ void QDeviceComp27(
(dist.f[DIR_MM0])[ksw] = getInterpolatedDistributionForNoSlipBC(q, f_NE, f_SW, feq, omega);
}
- q = (subgridD.q[DIR_MM0])[k];
+ q = (subgridD.q[DIR_MM0])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = -vx1 - vx2;
@@ -1825,7 +1844,7 @@ __global__ void QDeviceComp27(
(dist.f[DIR_PP0])[kne] = getInterpolatedDistributionForNoSlipBC(q, f_SW, f_NE, feq, omega);
}
- q = (subgridD.q[DIR_PM0])[k];
+ q = (subgridD.q[DIR_PM0])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = vx1 - vx2;
@@ -1833,7 +1852,7 @@ __global__ void QDeviceComp27(
(dist.f[DIR_MP0])[knw] = getInterpolatedDistributionForNoSlipBC(q, f_SE, f_NW, feq, omega);
}
- q = (subgridD.q[DIR_MP0])[k];
+ q = (subgridD.q[DIR_MP0])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = -vx1 + vx2;
@@ -1841,7 +1860,7 @@ __global__ void QDeviceComp27(
(dist.f[DIR_PM0])[kse] = getInterpolatedDistributionForNoSlipBC(q, f_NW, f_SE, feq, omega);
}
- q = (subgridD.q[DIR_P0P])[k];
+ q = (subgridD.q[DIR_P0P])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = vx1 + vx3;
@@ -1849,7 +1868,7 @@ __global__ void QDeviceComp27(
(dist.f[DIR_M0M])[kbw] = getInterpolatedDistributionForNoSlipBC(q, f_TE, f_BW, feq, omega);
}
- q = (subgridD.q[DIR_M0M])[k];
+ q = (subgridD.q[DIR_M0M])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = -vx1 - vx3;
@@ -1857,7 +1876,7 @@ __global__ void QDeviceComp27(
(dist.f[DIR_P0P])[kte] = getInterpolatedDistributionForNoSlipBC(q, f_BW, f_TE, feq, omega);
}
- q = (subgridD.q[DIR_P0M])[k];
+ q = (subgridD.q[DIR_P0M])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = vx1 - vx3;
@@ -1865,7 +1884,7 @@ __global__ void QDeviceComp27(
(dist.f[DIR_M0P])[ktw] = getInterpolatedDistributionForNoSlipBC(q, f_BE, f_TW, feq, omega);
}
- q = (subgridD.q[DIR_M0P])[k];
+ q = (subgridD.q[DIR_M0P])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = -vx1 + vx3;
@@ -1873,7 +1892,7 @@ __global__ void QDeviceComp27(
(dist.f[DIR_P0M])[kbe] = getInterpolatedDistributionForNoSlipBC(q, f_TW, f_BE, feq, omega);
}
- q = (subgridD.q[DIR_0PP])[k];
+ q = (subgridD.q[DIR_0PP])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = vx2 + vx3;
@@ -1881,7 +1900,7 @@ __global__ void QDeviceComp27(
(dist.f[DIR_0MM])[kbs] = getInterpolatedDistributionForNoSlipBC(q, f_TN, f_BS, feq, omega);
}
- q = (subgridD.q[DIR_0MM])[k];
+ q = (subgridD.q[DIR_0MM])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = -vx2 - vx3;
@@ -1889,7 +1908,7 @@ __global__ void QDeviceComp27(
(dist.f[DIR_0PP])[ktn] = getInterpolatedDistributionForNoSlipBC(q, f_BS, f_TN, feq, omega);
}
- q = (subgridD.q[DIR_0PM])[k];
+ q = (subgridD.q[DIR_0PM])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = vx2 - vx3;
@@ -1897,7 +1916,7 @@ __global__ void QDeviceComp27(
(dist.f[DIR_0MP])[kts] = getInterpolatedDistributionForNoSlipBC(q, f_BN, f_TS, feq, omega);
}
- q = (subgridD.q[DIR_0MP])[k];
+ q = (subgridD.q[DIR_0MP])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = -vx2 + vx3;
@@ -1905,7 +1924,7 @@ __global__ void QDeviceComp27(
(dist.f[DIR_0PM])[kbn] = getInterpolatedDistributionForNoSlipBC(q, f_TS, f_BN, feq, omega);
}
- q = (subgridD.q[DIR_PPP])[k];
+ q = (subgridD.q[DIR_PPP])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = vx1 + vx2 + vx3;
@@ -1913,7 +1932,7 @@ __global__ void QDeviceComp27(
(dist.f[DIR_MMM])[kbsw] = getInterpolatedDistributionForNoSlipBC(q, f_TNE, f_BSW, feq, omega);
}
- q = (subgridD.q[DIR_MMM])[k];
+ q = (subgridD.q[DIR_MMM])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = -vx1 - vx2 - vx3;
@@ -1921,7 +1940,7 @@ __global__ void QDeviceComp27(
(dist.f[DIR_PPP])[ktne] = getInterpolatedDistributionForNoSlipBC(q, f_BSW, f_TNE, feq, omega);
}
- q = (subgridD.q[DIR_PPM])[k];
+ q = (subgridD.q[DIR_PPM])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = vx1 + vx2 - vx3;
@@ -1929,7 +1948,7 @@ __global__ void QDeviceComp27(
(dist.f[DIR_MMP])[ktsw] = getInterpolatedDistributionForNoSlipBC(q, f_BNE, f_TSW, feq, omega);
}
- q = (subgridD.q[DIR_MMP])[k];
+ q = (subgridD.q[DIR_MMP])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = -vx1 - vx2 + vx3;
@@ -1937,7 +1956,7 @@ __global__ void QDeviceComp27(
(dist.f[DIR_PPM])[kbne] = getInterpolatedDistributionForNoSlipBC(q, f_TSW, f_BNE, feq, omega);
}
- q = (subgridD.q[DIR_PMP])[k];
+ q = (subgridD.q[DIR_PMP])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = vx1 - vx2 + vx3;
@@ -1945,7 +1964,7 @@ __global__ void QDeviceComp27(
(dist.f[DIR_MPM])[kbnw] = getInterpolatedDistributionForNoSlipBC(q, f_TSE, f_BNW, feq, omega);
}
- q = (subgridD.q[DIR_MPM])[k];
+ q = (subgridD.q[DIR_MPM])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = -vx1 + vx2 - vx3;
@@ -1953,7 +1972,7 @@ __global__ void QDeviceComp27(
(dist.f[DIR_PMP])[ktse] = getInterpolatedDistributionForNoSlipBC(q, f_BNW, f_TSE, feq, omega);
}
- q = (subgridD.q[DIR_PMM])[k];
+ q = (subgridD.q[DIR_PMM])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = vx1 - vx2 - vx3;
@@ -1961,7 +1980,7 @@ __global__ void QDeviceComp27(
(dist.f[DIR_MPP])[ktnw] = getInterpolatedDistributionForNoSlipBC(q, f_BSE, f_TNW, feq, omega);
}
- q = (subgridD.q[DIR_MPP])[k];
+ q = (subgridD.q[DIR_MPP])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = -vx1 + vx2 + vx3;
@@ -2011,16 +2030,17 @@ __global__ void QDeviceComp27(
//////////////////////////////////////////////////////////////////////////////
-__global__ void QDevice27(real* distributions,
- int* subgridDistanceIndices,
- real* subgridDistances,
- unsigned int numberOfBCnodes,
- real omega,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int numberOfLBnodes,
- bool isEvenTimestep)
+__global__ void QDevice27(
+ real* distributions,
+ int* subgridDistanceIndices,
+ real* subgridDistances,
+ unsigned int numberOfBCnodes,
+ real omega,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep)
{
//////////////////////////////////////////////////////////////////////////
//! The no-slip boundary condition is executed in the following steps
@@ -2028,19 +2048,12 @@ __global__ void QDevice27(real* distributions,
////////////////////////////////////////////////////////////////////////////////
//! - Get node index coordinates from threadIdx, blockIdx, blockDim and gridDim.
//!
- const unsigned x = threadIdx.x; // global x-index
- const unsigned y = blockIdx.x; // global y-index
- const unsigned z = blockIdx.y; // global z-index
-
- const unsigned nx = blockDim.x;
- const unsigned ny = gridDim.x;
-
- const unsigned k = nx*(ny*z + y) + x;
+ const unsigned nodeIndex = getNodeIndex();
//////////////////////////////////////////////////////////////////////////
//! - Run for all indices in size of boundary condition (numberOfBCnodes)
//!
- if(k < numberOfBCnodes)
+ if(nodeIndex < numberOfBCnodes)
{
//////////////////////////////////////////////////////////////////////////
@@ -2059,7 +2072,7 @@ __global__ void QDevice27(real* distributions,
////////////////////////////////////////////////////////////////////////////////
//! - Set neighbor indices (necessary for indirect addressing)
//!
- unsigned int indexOfBCnode = subgridDistanceIndices[k];
+ unsigned int indexOfBCnode = subgridDistanceIndices[nodeIndex];
unsigned int kzero= indexOfBCnode;
unsigned int ke = indexOfBCnode;
unsigned int kw = neighborX[indexOfBCnode];
@@ -2091,32 +2104,32 @@ __global__ void QDevice27(real* distributions,
////////////////////////////////////////////////////////////////////////////////
//! - Set local distributions
//!
- real f_W = (dist.f[DIR_P00 ])[ke ];
- real f_E = (dist.f[DIR_M00 ])[kw ];
- real f_S = (dist.f[DIR_0P0 ])[kn ];
- real f_N = (dist.f[DIR_0M0 ])[ks ];
- real f_B = (dist.f[DIR_00P ])[kt ];
- real f_T = (dist.f[DIR_00M ])[kb ];
- real f_SW = (dist.f[DIR_PP0 ])[kne ];
- real f_NE = (dist.f[DIR_MM0 ])[ksw ];
- real f_NW = (dist.f[DIR_PM0 ])[kse ];
- real f_SE = (dist.f[DIR_MP0 ])[knw ];
- real f_BW = (dist.f[DIR_P0P ])[kte ];
- real f_TE = (dist.f[DIR_M0M ])[kbw ];
- real f_TW = (dist.f[DIR_P0M ])[kbe ];
- real f_BE = (dist.f[DIR_M0P ])[ktw ];
- real f_BS = (dist.f[DIR_0PP ])[ktn ];
- real f_TN = (dist.f[DIR_0MM ])[kbs ];
- real f_TS = (dist.f[DIR_0PM ])[kbn ];
- real f_BN = (dist.f[DIR_0MP ])[kts ];
- real f_BSW = (dist.f[DIR_PPP ])[ktne ];
- real f_BNE = (dist.f[DIR_MMP ])[ktsw ];
- real f_BNW = (dist.f[DIR_PMP ])[ktse ];
- real f_BSE = (dist.f[DIR_MPP ])[ktnw ];
- real f_TSW = (dist.f[DIR_PPM ])[kbne ];
- real f_TNE = (dist.f[DIR_MMM ])[kbsw ];
- real f_TNW = (dist.f[DIR_PMM ])[kbse ];
- real f_TSE = (dist.f[DIR_MPM ])[kbnw ];
+ real f_W = (dist.f[DIR_P00])[ke ];
+ real f_E = (dist.f[DIR_M00])[kw ];
+ real f_S = (dist.f[DIR_0P0])[kn ];
+ real f_N = (dist.f[DIR_0M0])[ks ];
+ real f_B = (dist.f[DIR_00P])[kt ];
+ real f_T = (dist.f[DIR_00M])[kb ];
+ real f_SW = (dist.f[DIR_PP0])[kne ];
+ real f_NE = (dist.f[DIR_MM0])[ksw ];
+ real f_NW = (dist.f[DIR_PM0])[kse ];
+ real f_SE = (dist.f[DIR_MP0])[knw ];
+ real f_BW = (dist.f[DIR_P0P])[kte ];
+ real f_TE = (dist.f[DIR_M0M])[kbw ];
+ real f_TW = (dist.f[DIR_P0M])[kbe ];
+ real f_BE = (dist.f[DIR_M0P])[ktw ];
+ real f_BS = (dist.f[DIR_0PP])[ktn ];
+ real f_TN = (dist.f[DIR_0MM])[kbs ];
+ real f_TS = (dist.f[DIR_0PM])[kbn ];
+ real f_BN = (dist.f[DIR_0MP])[kts ];
+ real f_BSW = (dist.f[DIR_PPP])[ktne ];
+ real f_BNE = (dist.f[DIR_MMP])[ktsw ];
+ real f_BNW = (dist.f[DIR_PMP])[ktse ];
+ real f_BSE = (dist.f[DIR_MPP])[ktnw ];
+ real f_TSW = (dist.f[DIR_PPM])[kbne ];
+ real f_TNE = (dist.f[DIR_MMM])[kbsw ];
+ real f_TNW = (dist.f[DIR_PMM])[kbse ];
+ real f_TSE = (dist.f[DIR_MPM])[kbnw ];
////////////////////////////////////////////////////////////////////////////////
//! - Calculate macroscopic quantities
@@ -2148,7 +2161,7 @@ __global__ void QDevice27(real* distributions,
//! - Update distributions with subgrid distance (q) between zero and one
//!
real feq, q, velocityLB;
- q = (subgridD.q[DIR_P00])[k];
+ q = (subgridD.q[DIR_P00])[nodeIndex];
if (q>=c0o1 && q<=c1o1) // only update distribution for q between zero and one
{
velocityLB = vx1;
@@ -2156,7 +2169,7 @@ __global__ void QDevice27(real* distributions,
(dist.f[DIR_M00])[kw] = getInterpolatedDistributionForNoSlipBC(q, f_E, f_W, feq, omega);
}
- q = (subgridD.q[DIR_M00])[k];
+ q = (subgridD.q[DIR_M00])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = -vx1;
@@ -2164,7 +2177,7 @@ __global__ void QDevice27(real* distributions,
(dist.f[DIR_P00])[ke] = getInterpolatedDistributionForNoSlipBC(q, f_W, f_E, feq, omega);
}
- q = (subgridD.q[DIR_0P0])[k];
+ q = (subgridD.q[DIR_0P0])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = vx2;
@@ -2172,7 +2185,7 @@ __global__ void QDevice27(real* distributions,
(dist.f[DIR_0M0])[ks] = getInterpolatedDistributionForNoSlipBC(q, f_N, f_S, feq, omega);
}
- q = (subgridD.q[DIR_0M0])[k];
+ q = (subgridD.q[DIR_0M0])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = -vx2;
@@ -2180,7 +2193,7 @@ __global__ void QDevice27(real* distributions,
(dist.f[DIR_0P0])[kn] = getInterpolatedDistributionForNoSlipBC(q, f_S, f_N, feq, omega);
}
- q = (subgridD.q[DIR_00P])[k];
+ q = (subgridD.q[DIR_00P])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = vx3;
@@ -2188,7 +2201,7 @@ __global__ void QDevice27(real* distributions,
(dist.f[DIR_00M])[kb] = getInterpolatedDistributionForNoSlipBC(q, f_T, f_B, feq, omega);
}
- q = (subgridD.q[DIR_00M])[k];
+ q = (subgridD.q[DIR_00M])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = -vx3;
@@ -2196,7 +2209,7 @@ __global__ void QDevice27(real* distributions,
(dist.f[DIR_00P])[kt] = getInterpolatedDistributionForNoSlipBC(q, f_B, f_T, feq, omega);
}
- q = (subgridD.q[DIR_PP0])[k];
+ q = (subgridD.q[DIR_PP0])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = vx1 + vx2;
@@ -2204,7 +2217,7 @@ __global__ void QDevice27(real* distributions,
(dist.f[DIR_MM0])[ksw] = getInterpolatedDistributionForNoSlipBC(q, f_NE, f_SW, feq, omega);
}
- q = (subgridD.q[DIR_MM0])[k];
+ q = (subgridD.q[DIR_MM0])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = -vx1 - vx2;
@@ -2212,7 +2225,7 @@ __global__ void QDevice27(real* distributions,
(dist.f[DIR_PP0])[kne] = getInterpolatedDistributionForNoSlipBC(q, f_SW, f_NE, feq, omega);
}
- q = (subgridD.q[DIR_PM0])[k];
+ q = (subgridD.q[DIR_PM0])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = vx1 - vx2;
@@ -2220,7 +2233,7 @@ __global__ void QDevice27(real* distributions,
(dist.f[DIR_MP0])[knw] = getInterpolatedDistributionForNoSlipBC(q, f_SE, f_NW, feq, omega);
}
- q = (subgridD.q[DIR_MP0])[k];
+ q = (subgridD.q[DIR_MP0])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = -vx1 + vx2;
@@ -2228,7 +2241,7 @@ __global__ void QDevice27(real* distributions,
(dist.f[DIR_PM0])[kse] = getInterpolatedDistributionForNoSlipBC(q, f_NW, f_SE, feq, omega);
}
- q = (subgridD.q[DIR_P0P])[k];
+ q = (subgridD.q[DIR_P0P])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = vx1 + vx3;
@@ -2236,7 +2249,7 @@ __global__ void QDevice27(real* distributions,
(dist.f[DIR_M0M])[kbw] = getInterpolatedDistributionForNoSlipBC(q, f_TE, f_BW, feq, omega);
}
- q = (subgridD.q[DIR_M0M])[k];
+ q = (subgridD.q[DIR_M0M])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = -vx1 - vx3;
@@ -2244,7 +2257,7 @@ __global__ void QDevice27(real* distributions,
(dist.f[DIR_P0P])[kte] = getInterpolatedDistributionForNoSlipBC(q, f_BW, f_TE, feq, omega);
}
- q = (subgridD.q[DIR_P0M])[k];
+ q = (subgridD.q[DIR_P0M])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = vx1 - vx3;
@@ -2252,7 +2265,7 @@ __global__ void QDevice27(real* distributions,
(dist.f[DIR_M0P])[ktw] = getInterpolatedDistributionForNoSlipBC(q, f_BE, f_TW, feq, omega);
}
- q = (subgridD.q[DIR_M0P])[k];
+ q = (subgridD.q[DIR_M0P])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = -vx1 + vx3;
@@ -2260,7 +2273,7 @@ __global__ void QDevice27(real* distributions,
(dist.f[DIR_P0M])[kbe] = getInterpolatedDistributionForNoSlipBC(q, f_TW, f_BE, feq, omega);
}
- q = (subgridD.q[DIR_0PP])[k];
+ q = (subgridD.q[DIR_0PP])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = vx2 + vx3;
@@ -2268,7 +2281,7 @@ __global__ void QDevice27(real* distributions,
(dist.f[DIR_0MM])[kbs] = getInterpolatedDistributionForNoSlipBC(q, f_TN, f_BS, feq, omega);
}
- q = (subgridD.q[DIR_0MM])[k];
+ q = (subgridD.q[DIR_0MM])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = -vx2 - vx3;
@@ -2276,7 +2289,7 @@ __global__ void QDevice27(real* distributions,
(dist.f[DIR_0PP])[ktn] = getInterpolatedDistributionForNoSlipBC(q, f_BS, f_TN, feq, omega);
}
- q = (subgridD.q[DIR_0PM])[k];
+ q = (subgridD.q[DIR_0PM])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = vx2 - vx3;
@@ -2284,7 +2297,7 @@ __global__ void QDevice27(real* distributions,
(dist.f[DIR_0MP])[kts] = getInterpolatedDistributionForNoSlipBC(q, f_BN, f_TS, feq, omega);
}
- q = (subgridD.q[DIR_0MP])[k];
+ q = (subgridD.q[DIR_0MP])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = -vx2 + vx3;
@@ -2292,7 +2305,7 @@ __global__ void QDevice27(real* distributions,
(dist.f[DIR_0PM])[kbn] = getInterpolatedDistributionForNoSlipBC(q, f_TS, f_BN, feq, omega);
}
- q = (subgridD.q[DIR_PPP])[k];
+ q = (subgridD.q[DIR_PPP])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = vx1 + vx2 + vx3;
@@ -2300,7 +2313,7 @@ __global__ void QDevice27(real* distributions,
(dist.f[DIR_MMM])[kbsw] = getInterpolatedDistributionForNoSlipBC(q, f_TNE, f_BSW, feq, omega);
}
- q = (subgridD.q[DIR_MMM])[k];
+ q = (subgridD.q[DIR_MMM])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = -vx1 - vx2 - vx3;
@@ -2308,7 +2321,7 @@ __global__ void QDevice27(real* distributions,
(dist.f[DIR_PPP])[ktne] = getInterpolatedDistributionForNoSlipBC(q, f_BSW, f_TNE, feq, omega);
}
- q = (subgridD.q[DIR_PPM])[k];
+ q = (subgridD.q[DIR_PPM])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = vx1 + vx2 - vx3;
@@ -2316,7 +2329,7 @@ __global__ void QDevice27(real* distributions,
(dist.f[DIR_MMP])[ktsw] = getInterpolatedDistributionForNoSlipBC(q, f_BNE, f_TSW, feq, omega);
}
- q = (subgridD.q[DIR_MMP])[k];
+ q = (subgridD.q[DIR_MMP])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = -vx1 - vx2 + vx3;
@@ -2324,7 +2337,7 @@ __global__ void QDevice27(real* distributions,
(dist.f[DIR_PPM])[kbne] = getInterpolatedDistributionForNoSlipBC(q, f_TSW, f_BNE, feq, omega);
}
- q = (subgridD.q[DIR_PMP])[k];
+ q = (subgridD.q[DIR_PMP])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = vx1 - vx2 + vx3;
@@ -2332,7 +2345,7 @@ __global__ void QDevice27(real* distributions,
(dist.f[DIR_MPM])[kbnw] = getInterpolatedDistributionForNoSlipBC(q, f_TSE, f_BNW, feq, omega);
}
- q = (subgridD.q[DIR_MPM])[k];
+ q = (subgridD.q[DIR_MPM])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = -vx1 + vx2 - vx3;
@@ -2340,7 +2353,7 @@ __global__ void QDevice27(real* distributions,
(dist.f[DIR_PMP])[ktse] = getInterpolatedDistributionForNoSlipBC(q, f_BNW, f_TSE, feq, omega);
}
- q = (subgridD.q[DIR_PMM])[k];
+ q = (subgridD.q[DIR_PMM])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = vx1 - vx2 - vx3;
@@ -2348,7 +2361,7 @@ __global__ void QDevice27(real* distributions,
(dist.f[DIR_MPP])[ktnw] = getInterpolatedDistributionForNoSlipBC(q, f_BSE, f_TNW, feq, omega);
}
- q = (subgridD.q[DIR_MPP])[k];
+ q = (subgridD.q[DIR_MPP])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = -vx1 + vx2 + vx3;
@@ -2398,15 +2411,16 @@ __global__ void QDevice27(real* distributions,
//////////////////////////////////////////////////////////////////////////////
-__global__ void BBDevice27(real* distributions,
- int* subgridDistanceIndices,
- real* subgridDistances,
- unsigned int numberOfBCnodes,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int numberOfLBnodes,
- bool isEvenTimestep)
+__global__ void BBDevice27(
+ real* distributions,
+ int* subgridDistanceIndices,
+ real* subgridDistances,
+ unsigned int numberOfBCnodes,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep)
{
//////////////////////////////////////////////////////////////////////////
//! The no-slip boundary condition is executed in the following steps
@@ -2414,18 +2428,11 @@ __global__ void BBDevice27(real* distributions,
////////////////////////////////////////////////////////////////////////////////
//! - Get node index coordinates from threadIdx, blockIdx, blockDim and gridDim.
//!
- const unsigned x = threadIdx.x; // global x-index
- const unsigned y = blockIdx.x; // global y-index
- const unsigned z = blockIdx.y; // global z-index
-
- const unsigned nx = blockDim.x;
- const unsigned ny = gridDim.x;
-
- const unsigned k = nx*(ny*z + y) + x;
+ const unsigned nodeIndex = getNodeIndex();
//////////////////////////////////////////////////////////////////////////
// run for all indices in size of boundary condition (numberOfBCnodes)
- if(k < numberOfBCnodes)
+ if(nodeIndex < numberOfBCnodes)
{
//////////////////////////////////////////////////////////////////////////
//! - Read distributions: style of reading and writing the distributions from/to stored arrays dependent on timestep is based on the esoteric twist algorithm \ref
@@ -2443,7 +2450,7 @@ __global__ void BBDevice27(real* distributions,
////////////////////////////////////////////////////////////////////////////////
//! - Set neighbor indices (necessary for indirect addressing)
//!
- unsigned int indexOfBCnode = subgridDistanceIndices[k];
+ unsigned int indexOfBCnode = subgridDistanceIndices[nodeIndex];
unsigned int ke = indexOfBCnode;
unsigned int kw = neighborX[indexOfBCnode];
unsigned int kn = indexOfBCnode;
@@ -2474,32 +2481,32 @@ __global__ void BBDevice27(real* distributions,
////////////////////////////////////////////////////////////////////////////////
//! - Set local distributions
//!
- real f_W = (dist.f[DIR_P00 ])[ke ];
- real f_E = (dist.f[DIR_M00 ])[kw ];
- real f_S = (dist.f[DIR_0P0 ])[kn ];
- real f_N = (dist.f[DIR_0M0 ])[ks ];
- real f_B = (dist.f[DIR_00P ])[kt ];
- real f_T = (dist.f[DIR_00M ])[kb ];
- real f_SW = (dist.f[DIR_PP0 ])[kne ];
- real f_NE = (dist.f[DIR_MM0 ])[ksw ];
- real f_NW = (dist.f[DIR_PM0 ])[kse ];
- real f_SE = (dist.f[DIR_MP0 ])[knw ];
- real f_BW = (dist.f[DIR_P0P ])[kte ];
- real f_TE = (dist.f[DIR_M0M ])[kbw ];
- real f_TW = (dist.f[DIR_P0M ])[kbe ];
- real f_BE = (dist.f[DIR_M0P ])[ktw ];
- real f_BS = (dist.f[DIR_0PP ])[ktn ];
- real f_TN = (dist.f[DIR_0MM ])[kbs ];
- real f_TS = (dist.f[DIR_0PM ])[kbn ];
- real f_BN = (dist.f[DIR_0MP ])[kts ];
- real f_BSW = (dist.f[DIR_PPP ])[ktne ];
- real f_BNE = (dist.f[DIR_MMP ])[ktsw ];
- real f_BNW = (dist.f[DIR_PMP ])[ktse ];
- real f_BSE = (dist.f[DIR_MPP ])[ktnw ];
- real f_TSW = (dist.f[DIR_PPM ])[kbne ];
- real f_TNE = (dist.f[DIR_MMM ])[kbsw ];
- real f_TNW = (dist.f[DIR_PMM ])[kbse ];
- real f_TSE = (dist.f[DIR_MPM ])[kbnw ];
+ real f_W = (dist.f[DIR_P00])[ke ];
+ real f_E = (dist.f[DIR_M00])[kw ];
+ real f_S = (dist.f[DIR_0P0])[kn ];
+ real f_N = (dist.f[DIR_0M0])[ks ];
+ real f_B = (dist.f[DIR_00P])[kt ];
+ real f_T = (dist.f[DIR_00M])[kb ];
+ real f_SW = (dist.f[DIR_PP0])[kne ];
+ real f_NE = (dist.f[DIR_MM0])[ksw ];
+ real f_NW = (dist.f[DIR_PM0])[kse ];
+ real f_SE = (dist.f[DIR_MP0])[knw ];
+ real f_BW = (dist.f[DIR_P0P])[kte ];
+ real f_TE = (dist.f[DIR_M0M])[kbw ];
+ real f_TW = (dist.f[DIR_P0M])[kbe ];
+ real f_BE = (dist.f[DIR_M0P])[ktw ];
+ real f_BS = (dist.f[DIR_0PP])[ktn ];
+ real f_TN = (dist.f[DIR_0MM])[kbs ];
+ real f_TS = (dist.f[DIR_0PM])[kbn ];
+ real f_BN = (dist.f[DIR_0MP])[kts ];
+ real f_BSW = (dist.f[DIR_PPP])[ktne ];
+ real f_BNE = (dist.f[DIR_MMP])[ktsw ];
+ real f_BNW = (dist.f[DIR_PMP])[ktse ];
+ real f_BSE = (dist.f[DIR_MPP])[ktnw ];
+ real f_TSW = (dist.f[DIR_PPM])[kbne ];
+ real f_TNE = (dist.f[DIR_MMM])[kbsw ];
+ real f_TNW = (dist.f[DIR_PMM])[kbse ];
+ real f_TSE = (dist.f[DIR_MPM])[kbnw ];
////////////////////////////////////////////////////////////////////////////////
//! - change the pointer to write the results in the correct array
@@ -2509,32 +2516,32 @@ __global__ void BBDevice27(real* distributions,
////////////////////////////////////////////////////////////////////////////////
//! - rewrite distributions if there is a sub-grid distance (q) in same direction
real q;
- q = (subgridD.q[DIR_P00 ])[k]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_M00 ])[kw ]=f_E ;
- q = (subgridD.q[DIR_M00 ])[k]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_P00 ])[ke ]=f_W ;
- q = (subgridD.q[DIR_0P0 ])[k]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_0M0 ])[ks ]=f_N ;
- q = (subgridD.q[DIR_0M0 ])[k]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_0P0 ])[kn ]=f_S ;
- q = (subgridD.q[DIR_00P ])[k]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_00M ])[kb ]=f_T ;
- q = (subgridD.q[DIR_00M ])[k]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_00P ])[kt ]=f_B ;
- q = (subgridD.q[DIR_PP0 ])[k]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_MM0 ])[ksw ]=f_NE ;
- q = (subgridD.q[DIR_MM0 ])[k]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_PP0 ])[kne ]=f_SW ;
- q = (subgridD.q[DIR_PM0 ])[k]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_MP0 ])[knw ]=f_SE ;
- q = (subgridD.q[DIR_MP0 ])[k]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_PM0 ])[kse ]=f_NW ;
- q = (subgridD.q[DIR_P0P ])[k]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_M0M ])[kbw ]=f_TE ;
- q = (subgridD.q[DIR_M0M ])[k]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_P0P ])[kte ]=f_BW ;
- q = (subgridD.q[DIR_P0M ])[k]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_M0P ])[ktw ]=f_BE ;
- q = (subgridD.q[DIR_M0P ])[k]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_P0M ])[kbe ]=f_TW ;
- q = (subgridD.q[DIR_0PP ])[k]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_0MM ])[kbs ]=f_TN ;
- q = (subgridD.q[DIR_0MM ])[k]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_0PP ])[ktn ]=f_BS ;
- q = (subgridD.q[DIR_0PM ])[k]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_0MP ])[kts ]=f_BN ;
- q = (subgridD.q[DIR_0MP ])[k]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_0PM ])[kbn ]=f_TS ;
- q = (subgridD.q[DIR_PPP])[k]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_MMM])[kbsw]=f_TNE;
- q = (subgridD.q[DIR_MMM])[k]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_PPP])[ktne]=f_BSW;
- q = (subgridD.q[DIR_PPM])[k]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_MMP])[ktsw]=f_BNE;
- q = (subgridD.q[DIR_MMP])[k]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_PPM])[kbne]=f_TSW;
- q = (subgridD.q[DIR_PMP])[k]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_MPM])[kbnw]=f_TSE;
- q = (subgridD.q[DIR_MPM])[k]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_PMP])[ktse]=f_BNW;
- q = (subgridD.q[DIR_PMM])[k]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_MPP])[ktnw]=f_BSE;
- q = (subgridD.q[DIR_MPP])[k]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_PMM])[kbse]=f_TNW;
+ q = (subgridD.q[DIR_P00])[nodeIndex]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_M00])[kw ]=f_E ;
+ q = (subgridD.q[DIR_M00])[nodeIndex]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_P00])[ke ]=f_W ;
+ q = (subgridD.q[DIR_0P0])[nodeIndex]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_0M0])[ks ]=f_N ;
+ q = (subgridD.q[DIR_0M0])[nodeIndex]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_0P0])[kn ]=f_S ;
+ q = (subgridD.q[DIR_00P])[nodeIndex]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_00M])[kb ]=f_T ;
+ q = (subgridD.q[DIR_00M])[nodeIndex]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_00P])[kt ]=f_B ;
+ q = (subgridD.q[DIR_PP0])[nodeIndex]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_MM0])[ksw ]=f_NE ;
+ q = (subgridD.q[DIR_MM0])[nodeIndex]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_PP0])[kne ]=f_SW ;
+ q = (subgridD.q[DIR_PM0])[nodeIndex]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_MP0])[knw ]=f_SE ;
+ q = (subgridD.q[DIR_MP0])[nodeIndex]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_PM0])[kse ]=f_NW ;
+ q = (subgridD.q[DIR_P0P])[nodeIndex]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_M0M])[kbw ]=f_TE ;
+ q = (subgridD.q[DIR_M0M])[nodeIndex]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_P0P])[kte ]=f_BW ;
+ q = (subgridD.q[DIR_P0M])[nodeIndex]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_M0P])[ktw ]=f_BE ;
+ q = (subgridD.q[DIR_M0P])[nodeIndex]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_P0M])[kbe ]=f_TW ;
+ q = (subgridD.q[DIR_0PP])[nodeIndex]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_0MM])[kbs ]=f_TN ;
+ q = (subgridD.q[DIR_0MM])[nodeIndex]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_0PP])[ktn ]=f_BS ;
+ q = (subgridD.q[DIR_0PM])[nodeIndex]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_0MP])[kts ]=f_BN ;
+ q = (subgridD.q[DIR_0MP])[nodeIndex]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_0PM])[kbn ]=f_TS ;
+ q = (subgridD.q[DIR_PPP])[nodeIndex]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_MMM])[kbsw]=f_TNE;
+ q = (subgridD.q[DIR_MMM])[nodeIndex]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_PPP])[ktne]=f_BSW;
+ q = (subgridD.q[DIR_PPM])[nodeIndex]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_MMP])[ktsw]=f_BNE;
+ q = (subgridD.q[DIR_MMP])[nodeIndex]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_PPM])[kbne]=f_TSW;
+ q = (subgridD.q[DIR_PMP])[nodeIndex]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_MPM])[kbnw]=f_TSE;
+ q = (subgridD.q[DIR_MPM])[nodeIndex]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_PMP])[ktse]=f_BNW;
+ q = (subgridD.q[DIR_PMM])[nodeIndex]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_MPP])[ktnw]=f_BSE;
+ q = (subgridD.q[DIR_MPP])[nodeIndex]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_PMM])[kbse]=f_TNW;
}
}
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
diff --git a/src/gpu/VirtualFluids_GPU/GPU/Particles.cu b/src/gpu/VirtualFluids_GPU/GPU/Particles.cu
index 3a3ab784e6a7901c41d402629172c3c6154ffde9..22d9df4a3b4ae706dcf9b76d93940122015248f1 100644
--- a/src/gpu/VirtualFluids_GPU/GPU/Particles.cu
+++ b/src/gpu/VirtualFluids_GPU/GPU/Particles.cu
@@ -29,7 +29,7 @@ __global__ void InitParticles( real* coordX,
unsigned int* neighborWSB,
int level,
unsigned int numberOfParticles,
- unsigned int size_Mat)
+ unsigned long long numberOfLBnodes)
{
////////////////////////////////////////////////////////////////////////////////
const unsigned x = threadIdx.x; // Globaler x-Index
@@ -72,12 +72,12 @@ __global__ void InitParticles( real* coordX,
////////////////////////////////////////////////////////////////////////////////
//find random node of the fluid domain
- unsigned int cbID = (unsigned int)(randArray[k]*size_Mat);
- for(int i = 0; i < size_Mat;i++)
+ unsigned int cbID = (unsigned int)(randArray[k]*numberOfLBnodes);
+ for(int i = 0; i < numberOfLBnodes;i++)
{
//if (coordX[cbID] < 15 && coordX[cbID] > 5 && coordY[cbID] < 15 && coordY[cbID] > 5 && coordZ[cbID] < 15 && coordZ[cbID] > 5) break;
if (coordX[cbID] < 5 && coordX[cbID] > 2) break;
- cbID = (unsigned int)(randArray[k]*(size_Mat - i));
+ cbID = (unsigned int)(randArray[k]*(numberOfLBnodes - i));
}
real coordinateX;
@@ -183,7 +183,7 @@ __global__ void MoveParticles( real* coordX,
unsigned int timestep,
unsigned int numberOfTimesteps,
unsigned int numberOfParticles,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep)
{
////////////////////////////////////////////////////////////////////////////////
@@ -248,63 +248,63 @@ __global__ void MoveParticles( real* coordX,
{
if (isEvenTimestep==true)
{
- feC = &DD[DIR_P00 *size_Mat];
- fwC = &DD[DIR_M00 *size_Mat];
- fnC = &DD[DIR_0P0 *size_Mat];
- fsC = &DD[DIR_0M0 *size_Mat];
- ftC = &DD[DIR_00P *size_Mat];
- fbC = &DD[DIR_00M *size_Mat];
- fneC = &DD[DIR_PP0 *size_Mat];
- fswC = &DD[DIR_MM0 *size_Mat];
- fseC = &DD[DIR_PM0 *size_Mat];
- fnwC = &DD[DIR_MP0 *size_Mat];
- fteC = &DD[DIR_P0P *size_Mat];
- fbwC = &DD[DIR_M0M *size_Mat];
- fbeC = &DD[DIR_P0M *size_Mat];
- ftwC = &DD[DIR_M0P *size_Mat];
- ftnC = &DD[DIR_0PP *size_Mat];
- fbsC = &DD[DIR_0MM *size_Mat];
- fbnC = &DD[DIR_0PM *size_Mat];
- ftsC = &DD[DIR_0MP *size_Mat];
- fzeroC = &DD[DIR_000*size_Mat];
- ftneC = &DD[DIR_PPP *size_Mat];
- ftswC = &DD[DIR_MMP *size_Mat];
- ftseC = &DD[DIR_PMP *size_Mat];
- ftnwC = &DD[DIR_MPP *size_Mat];
- fbneC = &DD[DIR_PPM *size_Mat];
- fbswC = &DD[DIR_MMM *size_Mat];
- fbseC = &DD[DIR_PMM *size_Mat];
- fbnwC = &DD[DIR_MPM *size_Mat];
+ feC = &DD[DIR_P00 * numberOfLBnodes];
+ fwC = &DD[DIR_M00 * numberOfLBnodes];
+ fnC = &DD[DIR_0P0 * numberOfLBnodes];
+ fsC = &DD[DIR_0M0 * numberOfLBnodes];
+ ftC = &DD[DIR_00P * numberOfLBnodes];
+ fbC = &DD[DIR_00M * numberOfLBnodes];
+ fneC = &DD[DIR_PP0 * numberOfLBnodes];
+ fswC = &DD[DIR_MM0 * numberOfLBnodes];
+ fseC = &DD[DIR_PM0 * numberOfLBnodes];
+ fnwC = &DD[DIR_MP0 * numberOfLBnodes];
+ fteC = &DD[DIR_P0P * numberOfLBnodes];
+ fbwC = &DD[DIR_M0M * numberOfLBnodes];
+ fbeC = &DD[DIR_P0M * numberOfLBnodes];
+ ftwC = &DD[DIR_M0P * numberOfLBnodes];
+ ftnC = &DD[DIR_0PP * numberOfLBnodes];
+ fbsC = &DD[DIR_0MM * numberOfLBnodes];
+ fbnC = &DD[DIR_0PM * numberOfLBnodes];
+ ftsC = &DD[DIR_0MP * numberOfLBnodes];
+ fzeroC = &DD[DIR_000 * numberOfLBnodes];
+ ftneC = &DD[DIR_PPP * numberOfLBnodes];
+ ftswC = &DD[DIR_MMP * numberOfLBnodes];
+ ftseC = &DD[DIR_PMP * numberOfLBnodes];
+ ftnwC = &DD[DIR_MPP * numberOfLBnodes];
+ fbneC = &DD[DIR_PPM * numberOfLBnodes];
+ fbswC = &DD[DIR_MMM * numberOfLBnodes];
+ fbseC = &DD[DIR_PMM * numberOfLBnodes];
+ fbnwC = &DD[DIR_MPM * numberOfLBnodes];
}
else
{
- fwC = &DD[DIR_P00 *size_Mat];
- feC = &DD[DIR_M00 *size_Mat];
- fsC = &DD[DIR_0P0 *size_Mat];
- fnC = &DD[DIR_0M0 *size_Mat];
- fbC = &DD[DIR_00P *size_Mat];
- ftC = &DD[DIR_00M *size_Mat];
- fswC = &DD[DIR_PP0 *size_Mat];
- fneC = &DD[DIR_MM0 *size_Mat];
- fnwC = &DD[DIR_PM0 *size_Mat];
- fseC = &DD[DIR_MP0 *size_Mat];
- fbwC = &DD[DIR_P0P *size_Mat];
- fteC = &DD[DIR_M0M *size_Mat];
- ftwC = &DD[DIR_P0M *size_Mat];
- fbeC = &DD[DIR_M0P *size_Mat];
- fbsC = &DD[DIR_0PP *size_Mat];
- ftnC = &DD[DIR_0MM *size_Mat];
- ftsC = &DD[DIR_0PM *size_Mat];
- fbnC = &DD[DIR_0MP *size_Mat];
- fzeroC = &DD[DIR_000*size_Mat];
- fbswC = &DD[DIR_PPP *size_Mat];
- fbneC = &DD[DIR_MMP *size_Mat];
- fbnwC = &DD[DIR_PMP *size_Mat];
- fbseC = &DD[DIR_MPP *size_Mat];
- ftswC = &DD[DIR_PPM *size_Mat];
- ftneC = &DD[DIR_MMM *size_Mat];
- ftnwC = &DD[DIR_PMM *size_Mat];
- ftseC = &DD[DIR_MPM *size_Mat];
+ fwC = &DD[DIR_P00 * numberOfLBnodes];
+ feC = &DD[DIR_M00 * numberOfLBnodes];
+ fsC = &DD[DIR_0P0 * numberOfLBnodes];
+ fnC = &DD[DIR_0M0 * numberOfLBnodes];
+ fbC = &DD[DIR_00P * numberOfLBnodes];
+ ftC = &DD[DIR_00M * numberOfLBnodes];
+ fswC = &DD[DIR_PP0 * numberOfLBnodes];
+ fneC = &DD[DIR_MM0 * numberOfLBnodes];
+ fnwC = &DD[DIR_PM0 * numberOfLBnodes];
+ fseC = &DD[DIR_MP0 * numberOfLBnodes];
+ fbwC = &DD[DIR_P0P * numberOfLBnodes];
+ fteC = &DD[DIR_M0M * numberOfLBnodes];
+ ftwC = &DD[DIR_P0M * numberOfLBnodes];
+ fbeC = &DD[DIR_M0P * numberOfLBnodes];
+ fbsC = &DD[DIR_0PP * numberOfLBnodes];
+ ftnC = &DD[DIR_0MM * numberOfLBnodes];
+ ftsC = &DD[DIR_0PM * numberOfLBnodes];
+ fbnC = &DD[DIR_0MP * numberOfLBnodes];
+ fzeroC = &DD[DIR_000 * numberOfLBnodes];
+ fbswC = &DD[DIR_PPP * numberOfLBnodes];
+ fbneC = &DD[DIR_MMP * numberOfLBnodes];
+ fbnwC = &DD[DIR_PMP * numberOfLBnodes];
+ fbseC = &DD[DIR_MPP * numberOfLBnodes];
+ ftswC = &DD[DIR_PPM * numberOfLBnodes];
+ ftneC = &DD[DIR_MMM * numberOfLBnodes];
+ ftnwC = &DD[DIR_PMM * numberOfLBnodes];
+ ftseC = &DD[DIR_MPM * numberOfLBnodes];
}
//////////////////////////////////////////////////////////////////////////
@@ -1055,7 +1055,7 @@ __global__ void MoveParticlesWithoutBCs( real* coordX,
unsigned int timestep,
unsigned int numberOfTimesteps,
unsigned int numberOfParticles,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep)
{
////////////////////////////////////////////////////////////////////////////////
@@ -1114,63 +1114,63 @@ __global__ void MoveParticlesWithoutBCs( real* coordX,
{
if (isEvenTimestep==true)
{
- feC = &DD[DIR_P00 *size_Mat];
- fwC = &DD[DIR_M00 *size_Mat];
- fnC = &DD[DIR_0P0 *size_Mat];
- fsC = &DD[DIR_0M0 *size_Mat];
- ftC = &DD[DIR_00P *size_Mat];
- fbC = &DD[DIR_00M *size_Mat];
- fneC = &DD[DIR_PP0 *size_Mat];
- fswC = &DD[DIR_MM0 *size_Mat];
- fseC = &DD[DIR_PM0 *size_Mat];
- fnwC = &DD[DIR_MP0 *size_Mat];
- fteC = &DD[DIR_P0P *size_Mat];
- fbwC = &DD[DIR_M0M *size_Mat];
- fbeC = &DD[DIR_P0M *size_Mat];
- ftwC = &DD[DIR_M0P *size_Mat];
- ftnC = &DD[DIR_0PP *size_Mat];
- fbsC = &DD[DIR_0MM *size_Mat];
- fbnC = &DD[DIR_0PM *size_Mat];
- ftsC = &DD[DIR_0MP *size_Mat];
- fzeroC = &DD[DIR_000*size_Mat];
- ftneC = &DD[DIR_PPP *size_Mat];
- ftswC = &DD[DIR_MMP *size_Mat];
- ftseC = &DD[DIR_PMP *size_Mat];
- ftnwC = &DD[DIR_MPP *size_Mat];
- fbneC = &DD[DIR_PPM *size_Mat];
- fbswC = &DD[DIR_MMM *size_Mat];
- fbseC = &DD[DIR_PMM *size_Mat];
- fbnwC = &DD[DIR_MPM *size_Mat];
+ feC = &DD[DIR_P00 * numberOfLBnodes];
+ fwC = &DD[DIR_M00 * numberOfLBnodes];
+ fnC = &DD[DIR_0P0 * numberOfLBnodes];
+ fsC = &DD[DIR_0M0 * numberOfLBnodes];
+ ftC = &DD[DIR_00P * numberOfLBnodes];
+ fbC = &DD[DIR_00M * numberOfLBnodes];
+ fneC = &DD[DIR_PP0 * numberOfLBnodes];
+ fswC = &DD[DIR_MM0 * numberOfLBnodes];
+ fseC = &DD[DIR_PM0 * numberOfLBnodes];
+ fnwC = &DD[DIR_MP0 * numberOfLBnodes];
+ fteC = &DD[DIR_P0P * numberOfLBnodes];
+ fbwC = &DD[DIR_M0M * numberOfLBnodes];
+ fbeC = &DD[DIR_P0M * numberOfLBnodes];
+ ftwC = &DD[DIR_M0P * numberOfLBnodes];
+ ftnC = &DD[DIR_0PP * numberOfLBnodes];
+ fbsC = &DD[DIR_0MM * numberOfLBnodes];
+ fbnC = &DD[DIR_0PM * numberOfLBnodes];
+ ftsC = &DD[DIR_0MP * numberOfLBnodes];
+ fzeroC = &DD[DIR_000 * numberOfLBnodes];
+ ftneC = &DD[DIR_PPP * numberOfLBnodes];
+ ftswC = &DD[DIR_MMP * numberOfLBnodes];
+ ftseC = &DD[DIR_PMP * numberOfLBnodes];
+ ftnwC = &DD[DIR_MPP * numberOfLBnodes];
+ fbneC = &DD[DIR_PPM * numberOfLBnodes];
+ fbswC = &DD[DIR_MMM * numberOfLBnodes];
+ fbseC = &DD[DIR_PMM * numberOfLBnodes];
+ fbnwC = &DD[DIR_MPM * numberOfLBnodes];
}
else
{
- fwC = &DD[DIR_P00 *size_Mat];
- feC = &DD[DIR_M00 *size_Mat];
- fsC = &DD[DIR_0P0 *size_Mat];
- fnC = &DD[DIR_0M0 *size_Mat];
- fbC = &DD[DIR_00P *size_Mat];
- ftC = &DD[DIR_00M *size_Mat];
- fswC = &DD[DIR_PP0 *size_Mat];
- fneC = &DD[DIR_MM0 *size_Mat];
- fnwC = &DD[DIR_PM0 *size_Mat];
- fseC = &DD[DIR_MP0 *size_Mat];
- fbwC = &DD[DIR_P0P *size_Mat];
- fteC = &DD[DIR_M0M *size_Mat];
- ftwC = &DD[DIR_P0M *size_Mat];
- fbeC = &DD[DIR_M0P *size_Mat];
- fbsC = &DD[DIR_0PP *size_Mat];
- ftnC = &DD[DIR_0MM *size_Mat];
- ftsC = &DD[DIR_0PM *size_Mat];
- fbnC = &DD[DIR_0MP *size_Mat];
- fzeroC = &DD[DIR_000*size_Mat];
- fbswC = &DD[DIR_PPP *size_Mat];
- fbneC = &DD[DIR_MMP *size_Mat];
- fbnwC = &DD[DIR_PMP *size_Mat];
- fbseC = &DD[DIR_MPP *size_Mat];
- ftswC = &DD[DIR_PPM *size_Mat];
- ftneC = &DD[DIR_MMM *size_Mat];
- ftnwC = &DD[DIR_PMM *size_Mat];
- ftseC = &DD[DIR_MPM *size_Mat];
+ fwC = &DD[DIR_P00 * numberOfLBnodes];
+ feC = &DD[DIR_M00 * numberOfLBnodes];
+ fsC = &DD[DIR_0P0 * numberOfLBnodes];
+ fnC = &DD[DIR_0M0 * numberOfLBnodes];
+ fbC = &DD[DIR_00P * numberOfLBnodes];
+ ftC = &DD[DIR_00M * numberOfLBnodes];
+ fswC = &DD[DIR_PP0 * numberOfLBnodes];
+ fneC = &DD[DIR_MM0 * numberOfLBnodes];
+ fnwC = &DD[DIR_PM0 * numberOfLBnodes];
+ fseC = &DD[DIR_MP0 * numberOfLBnodes];
+ fbwC = &DD[DIR_P0P * numberOfLBnodes];
+ fteC = &DD[DIR_M0M * numberOfLBnodes];
+ ftwC = &DD[DIR_P0M * numberOfLBnodes];
+ fbeC = &DD[DIR_M0P * numberOfLBnodes];
+ fbsC = &DD[DIR_0PP * numberOfLBnodes];
+ ftnC = &DD[DIR_0MM * numberOfLBnodes];
+ ftsC = &DD[DIR_0PM * numberOfLBnodes];
+ fbnC = &DD[DIR_0MP * numberOfLBnodes];
+ fzeroC = &DD[DIR_000 * numberOfLBnodes];
+ fbswC = &DD[DIR_PPP * numberOfLBnodes];
+ fbneC = &DD[DIR_MMP * numberOfLBnodes];
+ fbnwC = &DD[DIR_PMP * numberOfLBnodes];
+ fbseC = &DD[DIR_MPP * numberOfLBnodes];
+ ftswC = &DD[DIR_PPM * numberOfLBnodes];
+ ftneC = &DD[DIR_MMM * numberOfLBnodes];
+ ftnwC = &DD[DIR_PMM * numberOfLBnodes];
+ ftseC = &DD[DIR_MPM * numberOfLBnodes];
}
//////////////////////////////////////////////////////////////////////////
@@ -1928,7 +1928,7 @@ __global__ void ParticleNoSlipDeviceComp27(real* coordX,
real* NormalX,
real* NormalY,
real* NormalZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep)
{
@@ -1937,63 +1937,63 @@ __global__ void ParticleNoSlipDeviceComp27(real* coordX,
//Distributions27 D;
//if (isEvenTimestep==true)
//{
- // D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- // D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- // D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- // D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- // D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- // D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- // D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- // D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- // D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- // D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- // D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- // D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- // D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- // D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- // D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- // D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- // D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- // D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- // D.f[DIR_000] = &DD[DIR_000*size_Mat];
- // D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- // D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- // D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- // D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- // D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- // D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- // D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- // D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
+ // D.f[DIR_P00] = &DD[DIR_P00 * size_Mat];
+ // D.f[DIR_M00] = &DD[DIR_M00 * size_Mat];
+ // D.f[DIR_0P0] = &DD[DIR_0P0 * size_Mat];
+ // D.f[DIR_0M0] = &DD[DIR_0M0 * size_Mat];
+ // D.f[DIR_00P] = &DD[DIR_00P * size_Mat];
+ // D.f[DIR_00M] = &DD[DIR_00M * size_Mat];
+ // D.f[DIR_PP0] = &DD[DIR_PP0 * size_Mat];
+ // D.f[DIR_MM0] = &DD[DIR_MM0 * size_Mat];
+ // D.f[DIR_PM0] = &DD[DIR_PM0 * size_Mat];
+ // D.f[DIR_MP0] = &DD[DIR_MP0 * size_Mat];
+ // D.f[DIR_P0P] = &DD[DIR_P0P * size_Mat];
+ // D.f[DIR_M0M] = &DD[DIR_M0M * size_Mat];
+ // D.f[DIR_P0M] = &DD[DIR_P0M * size_Mat];
+ // D.f[DIR_M0P] = &DD[DIR_M0P * size_Mat];
+ // D.f[DIR_0PP] = &DD[DIR_0PP * size_Mat];
+ // D.f[DIR_0MM] = &DD[DIR_0MM * size_Mat];
+ // D.f[DIR_0PM] = &DD[DIR_0PM * size_Mat];
+ // D.f[DIR_0MP] = &DD[DIR_0MP * size_Mat];
+ // D.f[DIR_000] = &DD[DIR_000 * size_Mat];
+ // D.f[DIR_PPP] = &DD[DIR_PPP * size_Mat];
+ // D.f[DIR_MMP] = &DD[DIR_MMP * size_Mat];
+ // D.f[DIR_PMP] = &DD[DIR_PMP * size_Mat];
+ // D.f[DIR_MPP] = &DD[DIR_MPP * size_Mat];
+ // D.f[DIR_PPM] = &DD[DIR_PPM * size_Mat];
+ // D.f[DIR_MMM] = &DD[DIR_MMM * size_Mat];
+ // D.f[DIR_PMM] = &DD[DIR_PMM * size_Mat];
+ // D.f[DIR_MPM] = &DD[DIR_MPM * size_Mat];
//}
//else
//{
- // D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
- // D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
- // D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
- // D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
- // D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
- // D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
- // D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
- // D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
- // D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
- // D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
- // D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
- // D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
- // D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
- // D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
- // D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
- // D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
- // D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
- // D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
- // D.f[DIR_000] = &DD[DIR_000*size_Mat];
- // D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
- // D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
- // D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
- // D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
- // D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
- // D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
- // D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
- // D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
+ // D.f[DIR_M00] = &DD[DIR_P00 * size_Mat];
+ // D.f[DIR_P00] = &DD[DIR_M00 * size_Mat];
+ // D.f[DIR_0M0] = &DD[DIR_0P0 * size_Mat];
+ // D.f[DIR_0P0] = &DD[DIR_0M0 * size_Mat];
+ // D.f[DIR_00M] = &DD[DIR_00P * size_Mat];
+ // D.f[DIR_00P] = &DD[DIR_00M * size_Mat];
+ // D.f[DIR_MM0] = &DD[DIR_PP0 * size_Mat];
+ // D.f[DIR_PP0] = &DD[DIR_MM0 * size_Mat];
+ // D.f[DIR_MP0] = &DD[DIR_PM0 * size_Mat];
+ // D.f[DIR_PM0] = &DD[DIR_MP0 * size_Mat];
+ // D.f[DIR_M0M] = &DD[DIR_P0P * size_Mat];
+ // D.f[DIR_P0P] = &DD[DIR_M0M * size_Mat];
+ // D.f[DIR_M0P] = &DD[DIR_P0M * size_Mat];
+ // D.f[DIR_P0M] = &DD[DIR_M0P * size_Mat];
+ // D.f[DIR_0MM] = &DD[DIR_0PP * size_Mat];
+ // D.f[DIR_0PP] = &DD[DIR_0MM * size_Mat];
+ // D.f[DIR_0MP] = &DD[DIR_0PM * size_Mat];
+ // D.f[DIR_0PM] = &DD[DIR_0MP * size_Mat];
+ // D.f[DIR_000] = &DD[DIR_000 * size_Mat];
+ // D.f[DIR_PPP] = &DD[DIR_MMM * size_Mat];
+ // D.f[DIR_MMP] = &DD[DIR_PPM * size_Mat];
+ // D.f[DIR_PMP] = &DD[DIR_MPM * size_Mat];
+ // D.f[DIR_MPP] = &DD[DIR_PMM * size_Mat];
+ // D.f[DIR_PPM] = &DD[DIR_MMP * size_Mat];
+ // D.f[DIR_MMM] = &DD[DIR_PPP * size_Mat];
+ // D.f[DIR_PMM] = &DD[DIR_MPP * size_Mat];
+ // D.f[DIR_MPM] = &DD[DIR_PMP * size_Mat];
//}
//////////////////////////////////////////////////////////////////////////////////
//const unsigned x = threadIdx.x; // Globaler x-Index
@@ -2015,24 +2015,24 @@ __global__ void ParticleNoSlipDeviceComp27(real* coordX,
// // *q_dirBE, *q_dirTW, *q_dirTN, *q_dirBS, *q_dirBN, *q_dirTS,
// // *q_dirTNE, *q_dirTSW, *q_dirTSE, *q_dirTNW, *q_dirBNE, *q_dirBSW,
// // *q_dirBSE, *q_dirBNW;
- // // q_dirE = &QQ[DIR_P00 * numberOfBCnodes];
- // q_dirW = &QQ[DIR_M00 * numberOfBCnodes];
- // // q_dirN = &QQ[DIR_0P0 * numberOfBCnodes];
- // q_dirS = &QQ[DIR_0M0 * numberOfBCnodes];
- // // q_dirT = &QQ[DIR_00P * numberOfBCnodes];
- // q_dirB = &QQ[DIR_00M * numberOfBCnodes];
- // // q_dirNE = &QQ[DIR_PP0 * numberOfBCnodes];
- // // q_dirSW = &QQ[DIR_MM0 * numberOfBCnodes];
- // // q_dirSE = &QQ[DIR_PM0 * numberOfBCnodes];
- // // q_dirNW = &QQ[DIR_MP0 * numberOfBCnodes];
- // // q_dirTE = &QQ[DIR_P0P * numberOfBCnodes];
- // // q_dirBW = &QQ[DIR_M0M * numberOfBCnodes];
- // // q_dirBE = &QQ[DIR_P0M * numberOfBCnodes];
- // // q_dirTW = &QQ[DIR_M0P * numberOfBCnodes];
- // // q_dirTN = &QQ[DIR_0PP * numberOfBCnodes];
- // // q_dirBS = &QQ[DIR_0MM * numberOfBCnodes];
- // // q_dirBN = &QQ[DIR_0PM * numberOfBCnodes];
- // // q_dirTS = &QQ[DIR_0MP * numberOfBCnodes];
+ // // q_dirE = &QQ[DIR_P00 * numberOfBCnodes];
+ // q_dirW = &QQ[DIR_M00 * numberOfBCnodes];
+ // // q_dirN = &QQ[DIR_0P0 * numberOfBCnodes];
+ // q_dirS = &QQ[DIR_0M0 * numberOfBCnodes];
+ // // q_dirT = &QQ[DIR_00P * numberOfBCnodes];
+ // q_dirB = &QQ[DIR_00M * numberOfBCnodes];
+ // // q_dirNE = &QQ[DIR_PP0 * numberOfBCnodes];
+ // // q_dirSW = &QQ[DIR_MM0 * numberOfBCnodes];
+ // // q_dirSE = &QQ[DIR_PM0 * numberOfBCnodes];
+ // // q_dirNW = &QQ[DIR_MP0 * numberOfBCnodes];
+ // // q_dirTE = &QQ[DIR_P0P * numberOfBCnodes];
+ // // q_dirBW = &QQ[DIR_M0M * numberOfBCnodes];
+ // // q_dirBE = &QQ[DIR_P0M * numberOfBCnodes];
+ // // q_dirTW = &QQ[DIR_M0P * numberOfBCnodes];
+ // // q_dirTN = &QQ[DIR_0PP * numberOfBCnodes];
+ // // q_dirBS = &QQ[DIR_0MM * numberOfBCnodes];
+ // // q_dirBN = &QQ[DIR_0PM * numberOfBCnodes];
+ // // q_dirTS = &QQ[DIR_0MP * numberOfBCnodes];
// // q_dirTNE = &QQ[DIR_PPP * numberOfBCnodes];
// // q_dirTSW = &QQ[DIR_MMP * numberOfBCnodes];
// // q_dirTSE = &QQ[DIR_PMP * numberOfBCnodes];
@@ -2047,24 +2047,24 @@ __global__ void ParticleNoSlipDeviceComp27(real* coordX,
// // *nx_dirBE, *nx_dirTW, *nx_dirTN, *nx_dirBS, *nx_dirBN, *nx_dirTS,
// // *nx_dirTNE, *nx_dirTSW, *nx_dirTSE, *nx_dirTNW, *nx_dirBNE, *nx_dirBSW,
// // *nx_dirBSE, *nx_dirBNW;
- // // nx_dirE = &NormalX[DIR_P00 * numberOfBCnodes];
- // // nx_dirW = &NormalX[DIR_M00 * numberOfBCnodes];
- // // nx_dirN = &NormalX[DIR_0P0 * numberOfBCnodes];
- // // nx_dirS = &NormalX[DIR_0M0 * numberOfBCnodes];
- // // nx_dirT = &NormalX[DIR_00P * numberOfBCnodes];
- // // nx_dirB = &NormalX[DIR_00M * numberOfBCnodes];
- // // nx_dirNE = &NormalX[DIR_PP0 * numberOfBCnodes];
- // // nx_dirSW = &NormalX[DIR_MM0 * numberOfBCnodes];
- // // nx_dirSE = &NormalX[DIR_PM0 * numberOfBCnodes];
- // // nx_dirNW = &NormalX[DIR_MP0 * numberOfBCnodes];
- // // nx_dirTE = &NormalX[DIR_P0P * numberOfBCnodes];
- // // nx_dirBW = &NormalX[DIR_M0M * numberOfBCnodes];
- // // nx_dirBE = &NormalX[DIR_P0M * numberOfBCnodes];
- // // nx_dirTW = &NormalX[DIR_M0P * numberOfBCnodes];
- // // nx_dirTN = &NormalX[DIR_0PP * numberOfBCnodes];
- // // nx_dirBS = &NormalX[DIR_0MM * numberOfBCnodes];
- // // nx_dirBN = &NormalX[DIR_0PM * numberOfBCnodes];
- // // nx_dirTS = &NormalX[DIR_0MP * numberOfBCnodes];
+ // // nx_dirE = &NormalX[DIR_P00 * numberOfBCnodes];
+ // // nx_dirW = &NormalX[DIR_M00 * numberOfBCnodes];
+ // // nx_dirN = &NormalX[DIR_0P0 * numberOfBCnodes];
+ // // nx_dirS = &NormalX[DIR_0M0 * numberOfBCnodes];
+ // // nx_dirT = &NormalX[DIR_00P * numberOfBCnodes];
+ // // nx_dirB = &NormalX[DIR_00M * numberOfBCnodes];
+ // // nx_dirNE = &NormalX[DIR_PP0 * numberOfBCnodes];
+ // // nx_dirSW = &NormalX[DIR_MM0 * numberOfBCnodes];
+ // // nx_dirSE = &NormalX[DIR_PM0 * numberOfBCnodes];
+ // // nx_dirNW = &NormalX[DIR_MP0 * numberOfBCnodes];
+ // // nx_dirTE = &NormalX[DIR_P0P * numberOfBCnodes];
+ // // nx_dirBW = &NormalX[DIR_M0M * numberOfBCnodes];
+ // // nx_dirBE = &NormalX[DIR_P0M * numberOfBCnodes];
+ // // nx_dirTW = &NormalX[DIR_M0P * numberOfBCnodes];
+ // // nx_dirTN = &NormalX[DIR_0PP * numberOfBCnodes];
+ // // nx_dirBS = &NormalX[DIR_0MM * numberOfBCnodes];
+ // // nx_dirBN = &NormalX[DIR_0PM * numberOfBCnodes];
+ // // nx_dirTS = &NormalX[DIR_0MP * numberOfBCnodes];
// // nx_dirTNE = &NormalX[DIR_PPP * numberOfBCnodes];
// // nx_dirTSW = &NormalX[DIR_MMP * numberOfBCnodes];
// // nx_dirTSE = &NormalX[DIR_PMP * numberOfBCnodes];
@@ -2079,24 +2079,24 @@ __global__ void ParticleNoSlipDeviceComp27(real* coordX,
// // *ny_dirBE, *ny_dirTW, *ny_dirTN, *ny_dirBS, *ny_dirBN, *ny_dirTS,
// // *ny_dirTNE, *ny_dirTSW, *ny_dirTSE, *ny_dirTNW, *ny_dirBNE, *ny_dirBSW,
// // *ny_dirBSE, *ny_dirBNW;
- // // ny_dirE = &NormalY[DIR_P00 * numberOfBCnodes];
- // // ny_dirW = &NormalY[DIR_M00 * numberOfBCnodes];
- // // ny_dirN = &NormalY[DIR_0P0 * numberOfBCnodes];
- // // ny_dirS = &NormalY[DIR_0M0 * numberOfBCnodes];
- // // ny_dirT = &NormalY[DIR_00P * numberOfBCnodes];
- // // ny_dirB = &NormalY[DIR_00M * numberOfBCnodes];
- // // ny_dirNE = &NormalY[DIR_PP0 * numberOfBCnodes];
- // // ny_dirSW = &NormalY[DIR_MM0 * numberOfBCnodes];
- // // ny_dirSE = &NormalY[DIR_PM0 * numberOfBCnodes];
- // // ny_dirNW = &NormalY[DIR_MP0 * numberOfBCnodes];
- // // ny_dirTE = &NormalY[DIR_P0P * numberOfBCnodes];
- // // ny_dirBW = &NormalY[DIR_M0M * numberOfBCnodes];
- // // ny_dirBE = &NormalY[DIR_P0M * numberOfBCnodes];
- // // ny_dirTW = &NormalY[DIR_M0P * numberOfBCnodes];
- // // ny_dirTN = &NormalY[DIR_0PP * numberOfBCnodes];
- // // ny_dirBS = &NormalY[DIR_0MM * numberOfBCnodes];
- // // ny_dirBN = &NormalY[DIR_0PM * numberOfBCnodes];
- // // ny_dirTS = &NormalY[DIR_0MP * numberOfBCnodes];
+ // // ny_dirE = &NormalY[DIR_P00 * numberOfBCnodes];
+ // // ny_dirW = &NormalY[DIR_M00 * numberOfBCnodes];
+ // // ny_dirN = &NormalY[DIR_0P0 * numberOfBCnodes];
+ // // ny_dirS = &NormalY[DIR_0M0 * numberOfBCnodes];
+ // // ny_dirT = &NormalY[DIR_00P * numberOfBCnodes];
+ // // ny_dirB = &NormalY[DIR_00M * numberOfBCnodes];
+ // // ny_dirNE = &NormalY[DIR_PP0 * numberOfBCnodes];
+ // // ny_dirSW = &NormalY[DIR_MM0 * numberOfBCnodes];
+ // // ny_dirSE = &NormalY[DIR_PM0 * numberOfBCnodes];
+ // // ny_dirNW = &NormalY[DIR_MP0 * numberOfBCnodes];
+ // // ny_dirTE = &NormalY[DIR_P0P * numberOfBCnodes];
+ // // ny_dirBW = &NormalY[DIR_M0M * numberOfBCnodes];
+ // // ny_dirBE = &NormalY[DIR_P0M * numberOfBCnodes];
+ // // ny_dirTW = &NormalY[DIR_M0P * numberOfBCnodes];
+ // // ny_dirTN = &NormalY[DIR_0PP * numberOfBCnodes];
+ // // ny_dirBS = &NormalY[DIR_0MM * numberOfBCnodes];
+ // // ny_dirBN = &NormalY[DIR_0PM * numberOfBCnodes];
+ // // ny_dirTS = &NormalY[DIR_0MP * numberOfBCnodes];
// // ny_dirTNE = &NormalY[DIR_PPP * numberOfBCnodes];
// // ny_dirTSW = &NormalY[DIR_MMP * numberOfBCnodes];
// // ny_dirTSE = &NormalY[DIR_PMP * numberOfBCnodes];
@@ -2111,24 +2111,24 @@ __global__ void ParticleNoSlipDeviceComp27(real* coordX,
// // *nz_dirBE, *nz_dirTW, *nz_dirTN, *nz_dirBS, *nz_dirBN, *nz_dirTS,
// // *nz_dirTNE, *nz_dirTSW, *nz_dirTSE, *nz_dirTNW, *nz_dirBNE, *nz_dirBSW,
// // *nz_dirBSE, *nz_dirBNW;
- // // nz_dirE = &NormalZ[DIR_P00 * numberOfBCnodes];
- // // nz_dirW = &NormalZ[DIR_M00 * numberOfBCnodes];
- // // nz_dirN = &NormalZ[DIR_0P0 * numberOfBCnodes];
- // // nz_dirS = &NormalZ[DIR_0M0 * numberOfBCnodes];
- // // nz_dirT = &NormalZ[DIR_00P * numberOfBCnodes];
- // // nz_dirB = &NormalZ[DIR_00M * numberOfBCnodes];
- // // nz_dirNE = &NormalZ[DIR_PP0 * numberOfBCnodes];
- // // nz_dirSW = &NormalZ[DIR_MM0 * numberOfBCnodes];
- // // nz_dirSE = &NormalZ[DIR_PM0 * numberOfBCnodes];
- // // nz_dirNW = &NormalZ[DIR_MP0 * numberOfBCnodes];
- // // nz_dirTE = &NormalZ[DIR_P0P * numberOfBCnodes];
- // // nz_dirBW = &NormalZ[DIR_M0M * numberOfBCnodes];
- // // nz_dirBE = &NormalZ[DIR_P0M * numberOfBCnodes];
- // // nz_dirTW = &NormalZ[DIR_M0P * numberOfBCnodes];
- // // nz_dirTN = &NormalZ[DIR_0PP * numberOfBCnodes];
- // // nz_dirBS = &NormalZ[DIR_0MM * numberOfBCnodes];
- // // nz_dirBN = &NormalZ[DIR_0PM * numberOfBCnodes];
- // // nz_dirTS = &NormalZ[DIR_0MP * numberOfBCnodes];
+ // // nz_dirE = &NormalZ[DIR_P00 * numberOfBCnodes];
+ // // nz_dirW = &NormalZ[DIR_M00 * numberOfBCnodes];
+ // // nz_dirN = &NormalZ[DIR_0P0 * numberOfBCnodes];
+ // // nz_dirS = &NormalZ[DIR_0M0 * numberOfBCnodes];
+ // // nz_dirT = &NormalZ[DIR_00P * numberOfBCnodes];
+ // // nz_dirB = &NormalZ[DIR_00M * numberOfBCnodes];
+ // // nz_dirNE = &NormalZ[DIR_PP0 * numberOfBCnodes];
+ // // nz_dirSW = &NormalZ[DIR_MM0 * numberOfBCnodes];
+ // // nz_dirSE = &NormalZ[DIR_PM0 * numberOfBCnodes];
+ // // nz_dirNW = &NormalZ[DIR_MP0 * numberOfBCnodes];
+ // // nz_dirTE = &NormalZ[DIR_P0P * numberOfBCnodes];
+ // // nz_dirBW = &NormalZ[DIR_M0M * numberOfBCnodes];
+ // // nz_dirBE = &NormalZ[DIR_P0M * numberOfBCnodes];
+ // // nz_dirTW = &NormalZ[DIR_M0P * numberOfBCnodes];
+ // // nz_dirTN = &NormalZ[DIR_0PP * numberOfBCnodes];
+ // // nz_dirBS = &NormalZ[DIR_0MM * numberOfBCnodes];
+ // // nz_dirBN = &NormalZ[DIR_0PM * numberOfBCnodes];
+ // // nz_dirTS = &NormalZ[DIR_0MP * numberOfBCnodes];
// // nz_dirTNE = &NormalZ[DIR_PPP * numberOfBCnodes];
// // nz_dirTSW = &NormalZ[DIR_MMP * numberOfBCnodes];
// // nz_dirTSE = &NormalZ[DIR_PMP * numberOfBCnodes];
@@ -2190,32 +2190,32 @@ __global__ void ParticleNoSlipDeviceComp27(real* coordX,
// unsigned int ktne = KQK;
// unsigned int kbsw = neighborZ[ksw];
// ////////////////////////////////////////////////////////////////////////////////
- // real f_W = (D.f[DIR_P00 ])[ke ];
- // real f_E = (D.f[DIR_M00 ])[kw ];
- // real f_S = (D.f[DIR_0P0 ])[kn ];
- // real f_N = (D.f[DIR_0M0 ])[ks ];
- // real f_B = (D.f[DIR_00P ])[kt ];
- // real f_T = (D.f[DIR_00M ])[kb ];
- // real f_SW = (D.f[DIR_PP0 ])[kne ];
- // real f_NE = (D.f[DIR_MM0 ])[ksw ];
- // real f_NW = (D.f[DIR_PM0 ])[kse ];
- // real f_SE = (D.f[DIR_MP0 ])[knw ];
- // real f_BW = (D.f[DIR_P0P ])[kte ];
- // real f_TE = (D.f[DIR_M0M ])[kbw ];
- // real f_TW = (D.f[DIR_P0M ])[kbe ];
- // real f_BE = (D.f[DIR_M0P ])[ktw ];
- // real f_BS = (D.f[DIR_0PP ])[ktn ];
- // real f_TN = (D.f[DIR_0MM ])[kbs ];
- // real f_TS = (D.f[DIR_0PM ])[kbn ];
- // real f_BN = (D.f[DIR_0MP ])[kts ];
- // real f_BSW = (D.f[DIR_PPP ])[ktne ];
- // real f_BNE = (D.f[DIR_MMP ])[ktsw ];
- // real f_BNW = (D.f[DIR_PMP ])[ktse ];
- // real f_BSE = (D.f[DIR_MPP ])[ktnw ];
- // real f_TSW = (D.f[DIR_PPM ])[kbne ];
- // real f_TNE = (D.f[DIR_MMM ])[kbsw ];
- // real f_TNW = (D.f[DIR_PMM ])[kbse ];
- // real f_TSE = (D.f[DIR_MPM ])[kbnw ];
+ // real f_W = (D.f[DIR_P00])[ke ];
+ // real f_E = (D.f[DIR_M00])[kw ];
+ // real f_S = (D.f[DIR_0P0])[kn ];
+ // real f_N = (D.f[DIR_0M0])[ks ];
+ // real f_B = (D.f[DIR_00P])[kt ];
+ // real f_T = (D.f[DIR_00M])[kb ];
+ // real f_SW = (D.f[DIR_PP0])[kne ];
+ // real f_NE = (D.f[DIR_MM0])[ksw ];
+ // real f_NW = (D.f[DIR_PM0])[kse ];
+ // real f_SE = (D.f[DIR_MP0])[knw ];
+ // real f_BW = (D.f[DIR_P0P])[kte ];
+ // real f_TE = (D.f[DIR_M0M])[kbw ];
+ // real f_TW = (D.f[DIR_P0M])[kbe ];
+ // real f_BE = (D.f[DIR_M0P])[ktw ];
+ // real f_BS = (D.f[DIR_0PP])[ktn ];
+ // real f_TN = (D.f[DIR_0MM])[kbs ];
+ // real f_TS = (D.f[DIR_0PM])[kbn ];
+ // real f_BN = (D.f[DIR_0MP])[kts ];
+ // real f_BSW = (D.f[DIR_PPP])[ktne ];
+ // real f_BNE = (D.f[DIR_MMP])[ktsw ];
+ // real f_BNW = (D.f[DIR_PMP])[ktse ];
+ // real f_BSE = (D.f[DIR_MPP])[ktnw ];
+ // real f_TSW = (D.f[DIR_PPM])[kbne ];
+ // real f_TNE = (D.f[DIR_MMM])[kbsw ];
+ // real f_TNW = (D.f[DIR_PMM])[kbse ];
+ // real f_TSE = (D.f[DIR_MPM])[kbnw ];
// ////////////////////////////////////////////////////////////////////////////////
// // real feq, q;
// real vx1, vx2, vx3, drho;
@@ -2241,63 +2241,63 @@ __global__ void ParticleNoSlipDeviceComp27(real* coordX,
// //////////////////////////////////////////////////////////////////////////
// if (isEvenTimestep==false)
// {
- // D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- // D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- // D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- // D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- // D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- // D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- // D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- // D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- // D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- // D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- // D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- // D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- // D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- // D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- // D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- // D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- // D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- // D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- // D.f[DIR_000] = &DD[DIR_000*size_Mat];
- // D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- // D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- // D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- // D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- // D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- // D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- // D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- // D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
+ // D.f[DIR_P00] = &DD[DIR_P00 * size_Mat];
+ // D.f[DIR_M00] = &DD[DIR_M00 * size_Mat];
+ // D.f[DIR_0P0] = &DD[DIR_0P0 * size_Mat];
+ // D.f[DIR_0M0] = &DD[DIR_0M0 * size_Mat];
+ // D.f[DIR_00P] = &DD[DIR_00P * size_Mat];
+ // D.f[DIR_00M] = &DD[DIR_00M * size_Mat];
+ // D.f[DIR_PP0] = &DD[DIR_PP0 * size_Mat];
+ // D.f[DIR_MM0] = &DD[DIR_MM0 * size_Mat];
+ // D.f[DIR_PM0] = &DD[DIR_PM0 * size_Mat];
+ // D.f[DIR_MP0] = &DD[DIR_MP0 * size_Mat];
+ // D.f[DIR_P0P] = &DD[DIR_P0P * size_Mat];
+ // D.f[DIR_M0M] = &DD[DIR_M0M * size_Mat];
+ // D.f[DIR_P0M] = &DD[DIR_P0M * size_Mat];
+ // D.f[DIR_M0P] = &DD[DIR_M0P * size_Mat];
+ // D.f[DIR_0PP] = &DD[DIR_0PP * size_Mat];
+ // D.f[DIR_0MM] = &DD[DIR_0MM * size_Mat];
+ // D.f[DIR_0PM] = &DD[DIR_0PM * size_Mat];
+ // D.f[DIR_0MP] = &DD[DIR_0MP * size_Mat];
+ // D.f[DIR_000] = &DD[DIR_000 * size_Mat];
+ // D.f[DIR_PPP] = &DD[DIR_PPP * size_Mat];
+ // D.f[DIR_MMP] = &DD[DIR_MMP * size_Mat];
+ // D.f[DIR_PMP] = &DD[DIR_PMP * size_Mat];
+ // D.f[DIR_MPP] = &DD[DIR_MPP * size_Mat];
+ // D.f[DIR_PPM] = &DD[DIR_PPM * size_Mat];
+ // D.f[DIR_MMM] = &DD[DIR_MMM * size_Mat];
+ // D.f[DIR_PMM] = &DD[DIR_PMM * size_Mat];
+ // D.f[DIR_MPM] = &DD[DIR_MPM * size_Mat];
// }
// else
// {
- // D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
- // D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
- // D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
- // D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
- // D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
- // D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
- // D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
- // D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
- // D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
- // D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
- // D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
- // D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
- // D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
- // D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
- // D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
- // D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
- // D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
- // D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
- // D.f[DIR_000] = &DD[DIR_000*size_Mat];
- // D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
- // D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
- // D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
- // D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
- // D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
- // D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
- // D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
- // D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
+ // D.f[DIR_M00] = &DD[DIR_P00 * size_Mat];
+ // D.f[DIR_P00] = &DD[DIR_M00 * size_Mat];
+ // D.f[DIR_0M0] = &DD[DIR_0P0 * size_Mat];
+ // D.f[DIR_0P0] = &DD[DIR_0M0 * size_Mat];
+ // D.f[DIR_00M] = &DD[DIR_00P * size_Mat];
+ // D.f[DIR_00P] = &DD[DIR_00M * size_Mat];
+ // D.f[DIR_MM0] = &DD[DIR_PP0 * size_Mat];
+ // D.f[DIR_PP0] = &DD[DIR_MM0 * size_Mat];
+ // D.f[DIR_MP0] = &DD[DIR_PM0 * size_Mat];
+ // D.f[DIR_PM0] = &DD[DIR_MP0 * size_Mat];
+ // D.f[DIR_M0M] = &DD[DIR_P0P * size_Mat];
+ // D.f[DIR_P0P] = &DD[DIR_M0M * size_Mat];
+ // D.f[DIR_M0P] = &DD[DIR_P0M * size_Mat];
+ // D.f[DIR_P0M] = &DD[DIR_M0P * size_Mat];
+ // D.f[DIR_0MM] = &DD[DIR_0PP * size_Mat];
+ // D.f[DIR_0PP] = &DD[DIR_0MM * size_Mat];
+ // D.f[DIR_0MP] = &DD[DIR_0PM * size_Mat];
+ // D.f[DIR_0PM] = &DD[DIR_0MP * size_Mat];
+ // D.f[DIR_000] = &DD[DIR_000 * size_Mat];
+ // D.f[DIR_PPP] = &DD[DIR_MMM * size_Mat];
+ // D.f[DIR_MMP] = &DD[DIR_PPM * size_Mat];
+ // D.f[DIR_PMP] = &DD[DIR_MPM * size_Mat];
+ // D.f[DIR_MPP] = &DD[DIR_PMM * size_Mat];
+ // D.f[DIR_PPM] = &DD[DIR_MMP * size_Mat];
+ // D.f[DIR_MMM] = &DD[DIR_PPP * size_Mat];
+ // D.f[DIR_PMM] = &DD[DIR_MPP * size_Mat];
+ // D.f[DIR_MPM] = &DD[DIR_PMP * size_Mat];
// }
//}
}
diff --git a/src/gpu/VirtualFluids_GPU/GPU/PrecursorBCs27.cu b/src/gpu/VirtualFluids_GPU/GPU/PrecursorBCs27.cu
new file mode 100644
index 0000000000000000000000000000000000000000..177eb41587896dd7993b06f98a1506abfc4f3f5f
--- /dev/null
+++ b/src/gpu/VirtualFluids_GPU/GPU/PrecursorBCs27.cu
@@ -0,0 +1,1157 @@
+//=======================================================================================
+// ____ ____ __ ______ __________ __ __ __ __
+// \ \ | | | | | _ \ |___ ___| | | | | / \ | |
+// \ \ | | | | | |_) | | | | | | | / \ | |
+// \ \ | | | | | _ / | | | | | | / /\ \ | |
+// \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
+// \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
+// \ \ | | ________________________________________________________________
+// \ \ | | | ______________________________________________________________|
+// \ \| | | | __ __ __ __ ______ _______
+// \ | | |_____ | | | | | | | | | _ \ / _____)
+// \ | | _____| | | | | | | | | | | \ \ \_______
+// \ | | | | |_____ | \_/ | | | | |_/ / _____ |
+// \ _____| |__| |________| \_______/ |__| |______/ (_______/
+//
+// This file is part of VirtualFluids. VirtualFluids is free software: you can
+// redistribute it and/or modify it under the terms of the GNU General Public
+// License as published by the Free Software Foundation, either version 3 of
+// the License, or (at your option) any later version.
+//
+// VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
+// ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+// FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
+// for more details.
+//
+// You should have received a copy of the GNU General Public License along
+// with VirtualFluids (see COPYING.txt). If not, see <http://www.gnu.org/licenses/>.
+//
+//! \file PrecursorBCs27.cu
+//! \ingroup GPU
+//! \author Henry Korb, Henrik Asmuth
+//======================================================================================
+#include "LBM/LB.h"
+#include <lbm/constants/NumericConstants.h>
+#include <lbm/constants/D3Q27.h>
+#include <lbm/MacroscopicQuantities.h>
+
+#include "LBM/GPUHelperFunctions/KernelUtilities.h"
+
+using namespace vf::lbm::constant;
+using namespace vf::lbm::dir;
+using namespace vf::gpu;
+
+__global__ void QPrecursorDeviceCompZeroPress(
+ int* subgridDistanceIndices,
+ int numberOfBCnodes,
+ int numberOfPrecursorNodes,
+ int sizeQ,
+ real omega,
+ real* distributions,
+ real* subgridDistances,
+ uint* neighborX,
+ uint* neighborY,
+ uint* neighborZ,
+ uint* neighbors0PP,
+ uint* neighbors0PM,
+ uint* neighbors0MP,
+ uint* neighbors0MM,
+ real* weights0PP,
+ real* weights0PM,
+ real* weights0MP,
+ real* weights0MM,
+ real* vLast,
+ real* vCurrent,
+ real velocityX,
+ real velocityY,
+ real velocityZ,
+ real timeRatio,
+ real velocityRatio,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep)
+{
+ ////////////////////////////////////////////////////////////////////////////////
+ //! - Get node index coordinates from threadIdx, blockIdx, blockDim and gridDim.
+ //!
+ const unsigned nodeIndex = getNodeIndex();
+
+ if(nodeIndex>=numberOfBCnodes) return;
+
+ ////////////////////////////////////////////////////////////////////////////////
+ // interpolation of velocity
+ real vxLastInterpd, vyLastInterpd, vzLastInterpd;
+ real vxNextInterpd, vyNextInterpd, vzNextInterpd;
+
+ uint kNeighbor0PP = neighbors0PP[nodeIndex];
+ real d0PP = weights0PP[nodeIndex];
+
+ real* vxLast = vLast;
+ real* vyLast = &vLast[numberOfPrecursorNodes];
+ real* vzLast = &vLast[2*numberOfPrecursorNodes];
+
+ real* vxCurrent = vCurrent;
+ real* vyCurrent = &vCurrent[numberOfPrecursorNodes];
+ real* vzCurrent = &vCurrent[2*numberOfPrecursorNodes];
+
+ if(d0PP < 1e6)
+ {
+ uint kNeighbor0PM = neighbors0PM[nodeIndex];
+ uint kNeighbor0MP = neighbors0MP[nodeIndex];
+ uint kNeighbor0MM = neighbors0MM[nodeIndex];
+
+ real d0PM = weights0PM[nodeIndex];
+ real d0MP = weights0MP[nodeIndex];
+ real d0MM = weights0MM[nodeIndex];
+
+ real invWeightSum = 1.f/(d0PP+d0PM+d0MP+d0MM);
+
+ vxLastInterpd = (vxLast[kNeighbor0PP]*d0PP + vxLast[kNeighbor0PM]*d0PM + vxLast[kNeighbor0MP]*d0MP + vxLast[kNeighbor0MM]*d0MM)*invWeightSum;
+ vyLastInterpd = (vyLast[kNeighbor0PP]*d0PP + vyLast[kNeighbor0PM]*d0PM + vyLast[kNeighbor0MP]*d0MP + vyLast[kNeighbor0MM]*d0MM)*invWeightSum;
+ vzLastInterpd = (vzLast[kNeighbor0PP]*d0PP + vzLast[kNeighbor0PM]*d0PM + vzLast[kNeighbor0MP]*d0MP + vzLast[kNeighbor0MM]*d0MM)*invWeightSum;
+
+ vxNextInterpd = (vxCurrent[kNeighbor0PP]*d0PP + vxCurrent[kNeighbor0PM]*d0PM + vxCurrent[kNeighbor0MP]*d0MP + vxCurrent[kNeighbor0MM]*d0MM)*invWeightSum;
+ vyNextInterpd = (vyCurrent[kNeighbor0PP]*d0PP + vyCurrent[kNeighbor0PM]*d0PM + vyCurrent[kNeighbor0MP]*d0MP + vyCurrent[kNeighbor0MM]*d0MM)*invWeightSum;
+ vzNextInterpd = (vzCurrent[kNeighbor0PP]*d0PP + vzCurrent[kNeighbor0PM]*d0PM + vzCurrent[kNeighbor0MP]*d0MP + vzCurrent[kNeighbor0MM]*d0MM)*invWeightSum;
+ }
+ else
+ {
+ vxLastInterpd = vxLast[kNeighbor0PP];
+ vyLastInterpd = vyLast[kNeighbor0PP];
+ vzLastInterpd = vzLast[kNeighbor0PP];
+
+ vxNextInterpd = vxCurrent[kNeighbor0PP];
+ vyNextInterpd = vyCurrent[kNeighbor0PP];
+ vzNextInterpd = vzCurrent[kNeighbor0PP];
+ }
+
+ // if(k==16300)s printf("%f %f %f\n", vxLastInterpd, vyLastInterpd, vzLastInterpd);
+ real VeloX = (velocityX + (1.f-timeRatio)*vxLastInterpd + timeRatio*vxNextInterpd)/velocityRatio;
+ real VeloY = (velocityY + (1.f-timeRatio)*vyLastInterpd + timeRatio*vyNextInterpd)/velocityRatio;
+ real VeloZ = (velocityZ + (1.f-timeRatio)*vzLastInterpd + timeRatio*vzNextInterpd)/velocityRatio;
+ // From here on just a copy of QVelDeviceCompZeroPress
+ ////////////////////////////////////////////////////////////////////////////////
+
+ //////////////////////////////////////////////////////////////////////////
+ //! - Read distributions: style of reading and writing the distributions from/to stored arrays dependent on timestep
+ //! is based on the esoteric twist algorithm \ref <a href="https://doi.org/10.3390/computation5020019"><b>[ M. Geier
+ //! et al. (2017), DOI:10.3390/computation5020019 ]</b></a>
+ //!
+ Distributions27 dist;
+ getPointersToDistributions(dist, distributions, numberOfLBnodes, isEvenTimestep);
+
+ unsigned int KQK = subgridDistanceIndices[nodeIndex];
+ unsigned int k000= KQK;
+ unsigned int kP00 = KQK;
+ unsigned int kM00 = neighborX[KQK];
+ unsigned int k0P0 = KQK;
+ unsigned int k0M0 = neighborY[KQK];
+ unsigned int k00P = KQK;
+ unsigned int k00M = neighborZ[KQK];
+ unsigned int kMM0 = neighborY[kM00];
+ unsigned int kPP0 = KQK;
+ unsigned int kPM0 = k0M0;
+ unsigned int kMP0 = kM00;
+ unsigned int kM0M = neighborZ[kM00];
+ unsigned int kP0P = KQK;
+ unsigned int kP0M = k00M;
+ unsigned int kM0P = kM00;
+ unsigned int k0PP = KQK;
+ unsigned int k0MM = neighborZ[k0M0];
+ unsigned int k0PM = k00M;
+ unsigned int k0MP = k0M0;
+ unsigned int kPMP = k0M0;
+ unsigned int kMPM = kM0M;
+ unsigned int kMPP = kM00;
+ unsigned int kPMM = k0MM;
+ unsigned int kMMP = kMM0;
+ unsigned int kPPM = k00M;
+ unsigned int kPPP = KQK;
+ unsigned int kMMM = neighborZ[kMM0];
+
+ ////////////////////////////////////////////////////////////////////////////////
+ //! - Set local distributions
+ //!
+ real f_M00 = (dist.f[DIR_P00])[kP00];
+ real f_P00 = (dist.f[DIR_M00])[kM00];
+ real f_0M0 = (dist.f[DIR_0P0])[k0P0];
+ real f_0P0 = (dist.f[DIR_0M0])[k0M0];
+ real f_00M = (dist.f[DIR_00P])[k00P];
+ real f_00P = (dist.f[DIR_00M])[k00M];
+ real f_MM0 = (dist.f[DIR_PP0])[kPP0];
+ real f_PP0 = (dist.f[DIR_MM0])[kMM0];
+ real f_MP0 = (dist.f[DIR_PM0])[kPM0];
+ real f_PM0 = (dist.f[DIR_MP0])[kMP0];
+ real f_M0M = (dist.f[DIR_P0P])[kP0P];
+ real f_P0P = (dist.f[DIR_M0M])[kM0M];
+ real f_M0P = (dist.f[DIR_P0M])[kP0M];
+ real f_P0M = (dist.f[DIR_M0P])[kM0P];
+ real f_0MM = (dist.f[DIR_0PP])[k0PP];
+ real f_0PP = (dist.f[DIR_0MM])[k0MM];
+ real f_0MP = (dist.f[DIR_0PM])[k0PM];
+ real f_0PM = (dist.f[DIR_0MP])[k0MP];
+ real f_MMM = (dist.f[DIR_PPP])[kPPP];
+ real f_PPM = (dist.f[DIR_MMP])[kMMP];
+ real f_MPM = (dist.f[DIR_PMP])[kPMP];
+ real f_PMM = (dist.f[DIR_MPP])[kMPP];
+ real f_MMP = (dist.f[DIR_PPM])[kPPM];
+ real f_PPP = (dist.f[DIR_MMM])[kMMM];
+ real f_MPP = (dist.f[DIR_PMM])[kPMM];
+ real f_PMP = (dist.f[DIR_MPM])[kMPM];
+
+ SubgridDistances27 subgridD;
+ getPointersToSubgridDistances(subgridD, subgridDistances, numberOfBCnodes);
+
+ ////////////////////////////////////////////////////////////////////////////////
+ real drho = f_PMP + f_MPP + f_PPP + f_MMP + f_PMM + f_MPM + f_PPM + f_MMM +
+ f_0PM + f_0PP + f_0MP + f_0MM + f_P0M + f_M0P + f_P0P + f_M0M + f_PM0 + f_MP0 + f_PP0 + f_MM0 +
+ f_00P + f_00M + f_0P0 + f_0M0 + f_P00 + f_M00 + ((dist.f[DIR_000])[k000]);
+
+ real vx1 = (((f_PMP - f_MPM) - (f_MPP - f_PMM)) + ((f_PPP - f_MMM) - (f_MMP - f_PPM)) +
+ ((f_P0M - f_M0P) + (f_P0P - f_M0M)) + ((f_PM0 - f_MP0) + (f_PP0 - f_MM0)) +
+ (f_P00 - f_M00)) / (c1o1 + drho);
+
+
+ real vx2 = ((-(f_PMP - f_MPM) + (f_MPP - f_PMM)) + ((f_PPP - f_MMM) - (f_MMP - f_PPM)) +
+ ((f_0PM - f_0MP) + (f_0PP - f_0MM)) + (-(f_PM0 - f_MP0) + (f_PP0 - f_MM0)) +
+ (f_0P0 - f_0M0)) / (c1o1 + drho);
+
+ real vx3 = (((f_PMP - f_MPM) + (f_MPP - f_PMM)) + ((f_PPP - f_MMM) + (f_MMP - f_PPM)) +
+ (-(f_0PM - f_0MP) + (f_0PP - f_0MM)) + ((f_P0P - f_M0M) - (f_P0M - f_M0P)) +
+ (f_00P - f_00M)) / (c1o1 + drho);
+
+
+ // if(k==16383 || k==0) printf("k %d kQ %d drho = %f u %f v %f w %f\n",k, KQK, drho, vx1, vx2, vx3);
+ real cu_sq=c3o2*(vx1*vx1+vx2*vx2+vx3*vx3) * (c1o1 + drho);
+ //////////////////////////////////////////////////////////////////////////
+
+
+ ////////////////////////////////////////////////////////////////////////////////
+ //! - Update distributions with subgrid distance (q) between zero and one
+ real feq, q, velocityLB, velocityBC;
+ q = (subgridD.q[DIR_P00])[nodeIndex];
+ if (q>=c0o1 && q<=c1o1) // only update distribution for q between zero and one
+ {
+ velocityLB = vx1;
+ feq = getEquilibriumForBC(drho, velocityLB, cu_sq, c2o27);
+ velocityBC = VeloX;
+ (dist.f[DIR_M00])[kM00] = getInterpolatedDistributionForVeloWithPressureBC(q, f_P00, f_M00, feq, omega, drho, velocityBC, c2o27);
+ }
+
+ q = (subgridD.q[DIR_M00])[nodeIndex];
+ if (q>=c0o1 && q<=c1o1)
+ {
+ velocityLB = -vx1;
+ feq = getEquilibriumForBC(drho, velocityLB, cu_sq, c2o27);
+ velocityBC = -VeloX;
+ (dist.f[DIR_P00])[kP00] = getInterpolatedDistributionForVeloWithPressureBC(q, f_M00, f_P00, feq, omega, drho, velocityBC, c2o27);
+ }
+
+ q = (subgridD.q[DIR_0P0])[nodeIndex];
+ if (q>=c0o1 && q<=c1o1)
+ {
+ velocityLB = vx2;
+ feq = getEquilibriumForBC(drho, velocityLB, cu_sq, c2o27);
+ velocityBC = VeloY;
+ (dist.f[DIR_0M0])[DIR_0M0] = getInterpolatedDistributionForVeloWithPressureBC(q, f_0P0, f_0M0, feq, omega, drho, velocityBC, c2o27);
+ }
+
+ q = (subgridD.q[DIR_0M0])[nodeIndex];
+ if (q>=c0o1 && q<=c1o1)
+ {
+ velocityLB = -vx2;
+ feq = getEquilibriumForBC(drho, velocityLB, cu_sq, c2o27);
+ velocityBC = -VeloY;
+ (dist.f[DIR_0P0])[k0P0] = getInterpolatedDistributionForVeloWithPressureBC(q, f_0M0, f_0P0, feq, omega, drho, velocityBC, c2o27);
+ }
+
+ q = (subgridD.q[DIR_00P])[nodeIndex];
+ if (q>=c0o1 && q<=c1o1)
+ {
+ velocityLB = vx3;
+ feq = getEquilibriumForBC(drho, velocityLB, cu_sq, c2o27);
+ velocityBC = VeloZ;
+ (dist.f[DIR_00M])[k00M] = getInterpolatedDistributionForVeloWithPressureBC(q, f_00P, f_00M, feq, omega, drho, velocityBC, c2o27);
+ }
+
+ q = (subgridD.q[DIR_00M])[nodeIndex];
+ if (q>=c0o1 && q<=c1o1)
+ {
+ velocityLB = -vx3;
+ feq = getEquilibriumForBC(drho, velocityLB, cu_sq, c2o27);
+ velocityBC = -VeloZ;
+ (dist.f[DIR_00P])[k00P] = getInterpolatedDistributionForVeloWithPressureBC(q, f_00M, f_00P, feq, omega, drho, velocityBC, c2o27);
+ }
+
+ q = (subgridD.q[DIR_PP0])[nodeIndex];
+ if (q>=c0o1 && q<=c1o1)
+ {
+ velocityLB = vx1 + vx2;
+ feq = getEquilibriumForBC(drho, velocityLB, cu_sq, c1o54);
+ velocityBC = VeloX + VeloY;
+ (dist.f[DIR_MM0])[kMM0] = getInterpolatedDistributionForVeloWithPressureBC(q, f_PP0, f_MM0, feq, omega, drho, velocityBC, c1o54);
+ }
+
+ q = (subgridD.q[DIR_MM0])[nodeIndex];
+ if (q>=c0o1 && q<=c1o1)
+ {
+ velocityLB = -vx1 - vx2;
+ feq = getEquilibriumForBC(drho, velocityLB, cu_sq, c1o54);
+ velocityBC = -VeloX - VeloY;
+ (dist.f[DIR_PP0])[kPP0] = getInterpolatedDistributionForVeloWithPressureBC(q, f_MM0, f_PP0, feq, omega, drho, velocityBC, c1o54);
+ }
+
+ q = (subgridD.q[DIR_PM0])[nodeIndex];
+ if (q>=c0o1 && q<=c1o1)
+ {
+ velocityLB = vx1 - vx2;
+ feq = getEquilibriumForBC(drho, velocityLB, cu_sq, c1o54);
+ velocityBC = VeloX - VeloY;
+ (dist.f[DIR_MP0])[kMP0] = getInterpolatedDistributionForVeloWithPressureBC(q, f_PM0, f_MP0, feq, omega, drho, velocityBC, c1o54);
+ }
+
+ q = (subgridD.q[DIR_MP0])[nodeIndex];
+ if (q>=c0o1 && q<=c1o1)
+ {
+ velocityLB = -vx1 + vx2;
+ feq = getEquilibriumForBC(drho, velocityLB, cu_sq, c1o54);
+ velocityBC = -VeloX + VeloY;
+ (dist.f[DIR_PM0])[kPM0] = getInterpolatedDistributionForVeloWithPressureBC(q, f_MP0, f_PM0, feq, omega, drho, velocityBC, c1o54);
+ }
+
+ q = (subgridD.q[DIR_P0P])[nodeIndex];
+ if (q>=c0o1 && q<=c1o1)
+ {
+ velocityLB = vx1 + vx3;
+ feq = getEquilibriumForBC(drho, velocityLB, cu_sq, c1o54);
+ velocityBC = VeloX + VeloZ;
+ (dist.f[DIR_M0M])[kM0M] = getInterpolatedDistributionForVeloWithPressureBC(q, f_P0P, f_M0M, feq, omega, drho, velocityBC, c1o54);
+ }
+
+ q = (subgridD.q[DIR_M0M])[nodeIndex];
+ if (q>=c0o1 && q<=c1o1)
+ {
+ velocityLB = -vx1 - vx3;
+ feq = getEquilibriumForBC(drho, velocityLB, cu_sq, c1o54);
+ velocityBC = -VeloX - VeloZ;
+ (dist.f[DIR_P0P])[kP0P] = getInterpolatedDistributionForVeloWithPressureBC(q, f_M0M, f_P0P, feq, omega, drho, velocityBC, c1o54);
+ }
+
+ q = (subgridD.q[DIR_P0M])[nodeIndex];
+ if (q>=c0o1 && q<=c1o1)
+ {
+ velocityLB = vx1 - vx3;
+ feq = getEquilibriumForBC(drho, velocityLB, cu_sq, c1o54);
+ velocityBC = VeloX - VeloZ;
+ (dist.f[DIR_M0P])[kM0P] = getInterpolatedDistributionForVeloWithPressureBC(q, f_P0M, f_M0P, feq, omega, drho, velocityBC, c1o54);
+ }
+
+ q = (subgridD.q[DIR_M0P])[nodeIndex];
+ if (q>=c0o1 && q<=c1o1)
+ {
+ velocityLB = -vx1 + vx3;
+ feq = getEquilibriumForBC(drho, velocityLB, cu_sq, c1o54);
+ velocityBC = -VeloX + VeloZ;
+ (dist.f[DIR_P0M])[kP0M] = getInterpolatedDistributionForVeloWithPressureBC(q, f_M0P, f_P0M, feq, omega, drho, velocityBC, c1o54);
+ }
+
+ q = (subgridD.q[DIR_0PP])[nodeIndex];
+ if (q>=c0o1 && q<=c1o1)
+ {
+ velocityLB = vx2 + vx3;
+ feq = getEquilibriumForBC(drho, velocityLB, cu_sq, c1o54);
+ velocityBC = VeloY + VeloZ;
+ (dist.f[DIR_0MM])[k0MM] = getInterpolatedDistributionForVeloWithPressureBC(q, f_0PP, f_0MM, feq, omega, drho, velocityBC, c1o54);
+ }
+
+ q = (subgridD.q[DIR_0MM])[nodeIndex];
+ if (q>=c0o1 && q<=c1o1)
+ {
+ velocityLB = -vx2 - vx3;
+ feq = getEquilibriumForBC(drho, velocityLB, cu_sq, c1o54);
+ velocityBC = -VeloY - VeloZ;
+ (dist.f[DIR_0PP])[k0PP] = getInterpolatedDistributionForVeloWithPressureBC(q, f_0MM, f_0PP, feq, omega, drho, velocityBC, c1o54);
+ }
+
+ q = (subgridD.q[DIR_0PM])[nodeIndex];
+ if (q>=c0o1 && q<=c1o1)
+ {
+ velocityLB = vx2 - vx3;
+ feq = getEquilibriumForBC(drho, velocityLB, cu_sq, c1o54);
+ velocityBC = VeloY - VeloZ;
+ (dist.f[DIR_0MP])[k0MP] = getInterpolatedDistributionForVeloWithPressureBC(q, f_0PM, f_0PP, feq, omega, drho, velocityBC, c1o54);
+ }
+
+ q = (subgridD.q[DIR_0MP])[nodeIndex];
+ if (q>=c0o1 && q<=c1o1)
+ {
+ velocityLB = -vx2 + vx3;
+ feq = getEquilibriumForBC(drho, velocityLB, cu_sq, c1o54);
+ velocityBC = -VeloY + VeloZ;
+ (dist.f[DIR_0PM])[k0PM] = getInterpolatedDistributionForVeloWithPressureBC(q, f_0PP, f_0PM, feq, omega, drho, velocityBC, c1o54);
+ }
+
+ q = (subgridD.q[DIR_PPP])[nodeIndex];
+ if (q>=c0o1 && q<=c1o1)
+ {
+ velocityLB = vx1 + vx2 + vx3;
+ feq = getEquilibriumForBC(drho, velocityLB, cu_sq, c1o216);
+ velocityBC = VeloX + VeloY + VeloZ;
+ (dist.f[DIR_MMM])[kMMM] = getInterpolatedDistributionForVeloWithPressureBC(q, f_PPP, f_MMM, feq, omega, drho, velocityBC, c1o216);
+ }
+
+ q = (subgridD.q[DIR_MMM])[nodeIndex];
+ if (q>=c0o1 && q<=c1o1)
+ {
+ velocityLB = -vx1 - vx2 - vx3;
+ feq = getEquilibriumForBC(drho, velocityLB, cu_sq, c1o216);
+ velocityBC = -VeloX - VeloY - VeloZ;
+ (dist.f[DIR_PPP])[kPPP] = getInterpolatedDistributionForVeloWithPressureBC(q, f_MMM, f_PPP, feq, omega, drho, velocityBC, c1o216);
+ }
+
+ q = (subgridD.q[DIR_PPM])[nodeIndex];
+ if (q>=c0o1 && q<=c1o1)
+ {
+ velocityLB = vx1 + vx2 - vx3;
+ feq = getEquilibriumForBC(drho, velocityLB, cu_sq, c1o216);
+ velocityBC = VeloX + VeloY - VeloZ;
+ (dist.f[DIR_MMP])[kMMP] = getInterpolatedDistributionForVeloWithPressureBC(q, f_PPM, f_MMP, feq, omega, drho, velocityBC, c1o216);
+ }
+
+ q = (subgridD.q[DIR_MMP])[nodeIndex];
+ if (q>=c0o1 && q<=c1o1)
+ {
+ velocityLB = -vx1 - vx2 + vx3;
+ feq = getEquilibriumForBC(drho, velocityLB, cu_sq, c1o216);
+ velocityBC = -VeloX - VeloY + VeloZ;
+ (dist.f[DIR_PPM])[kPPM] = getInterpolatedDistributionForVeloWithPressureBC(q, f_MMP, f_PPM, feq, omega, drho, velocityBC, c1o216);
+ }
+
+ q = (subgridD.q[DIR_PMP])[nodeIndex];
+ if (q>=c0o1 && q<=c1o1)
+ {
+ velocityLB = vx1 - vx2 + vx3;
+ feq = getEquilibriumForBC(drho, velocityLB, cu_sq, c1o216);
+ velocityBC = VeloX - VeloY + VeloZ;
+ (dist.f[DIR_MPM])[kMPM] = getInterpolatedDistributionForVeloWithPressureBC(q, f_PMP, f_MPM, feq, omega, drho, velocityBC, c1o216);
+ }
+
+ q = (subgridD.q[DIR_MPM])[nodeIndex];
+ if (q>=c0o1 && q<=c1o1)
+ {
+ velocityLB = -vx1 + vx2 - vx3;
+ feq = getEquilibriumForBC(drho, velocityLB, cu_sq, c1o216);
+ velocityBC = -VeloX + VeloY - VeloZ;
+ (dist.f[DIR_PMP])[kPMP] = getInterpolatedDistributionForVeloWithPressureBC(q, f_MPM, f_PMP, feq, omega, drho, velocityBC, c1o216);
+ }
+
+ q = (subgridD.q[DIR_PMM])[nodeIndex];
+ if (q>=c0o1 && q<=c1o1)
+ {
+ velocityLB = vx1 - vx2 - vx3;
+ feq = getEquilibriumForBC(drho, velocityLB, cu_sq, c1o216);
+ velocityBC = VeloX - VeloY - VeloZ;
+ (dist.f[DIR_MPP])[kMPP] = getInterpolatedDistributionForVeloWithPressureBC(q, f_PMM, f_MPP, feq, omega, drho, velocityBC, c1o216);
+ }
+
+ q = (subgridD.q[DIR_MPP])[nodeIndex];
+ if (q>=c0o1 && q<=c1o1)
+ {
+ velocityLB = -vx1 + vx2 + vx3;
+ feq = getEquilibriumForBC(drho, velocityLB, cu_sq, c1o216);
+ velocityBC = -VeloX + VeloY + VeloZ;
+ (dist.f[DIR_PMM])[kPMM] = getInterpolatedDistributionForVeloWithPressureBC(q, f_MPP, f_PMM, feq, omega, drho, velocityBC, c1o216);
+ }
+}
+////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
+__global__ void PrecursorDeviceEQ27(
+ int *subgridDistanceIndices,
+ int numberOfBCnodes,
+ int numberOfPrecursorNodes,
+ real omega,
+ real* distributions,
+ uint* neighborX,
+ uint* neighborY,
+ uint* neighborZ,
+ uint* neighbors0PP,
+ uint* neighbors0PM,
+ uint* neighbors0MP,
+ uint* neighbors0MM,
+ real* weights0PP,
+ real* weights0PM,
+ real* weights0MP,
+ real* weights0MM,
+ real* vLast,
+ real* vCurrent,
+ real velocityX,
+ real velocityY,
+ real velocityZ,
+ real timeRatio,
+ real velocityRatio,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep)
+{
+ ////////////////////////////////////////////////////////////////////////////////
+ //! - Get node index coordinates from threadIdx, blockIdx, blockDim and gridDim.
+ //!
+ const unsigned nodeIndex = getNodeIndex();
+
+ if(nodeIndex>=numberOfBCnodes) return;
+
+ ////////////////////////////////////////////////////////////////////////////////
+ // interpolation of velocity
+ real vxLastInterpd, vyLastInterpd, vzLastInterpd;
+ real vxNextInterpd, vyNextInterpd, vzNextInterpd;
+
+ uint kNeighbor0PP = neighbors0PP[nodeIndex];
+ real d0PP = weights0PP[nodeIndex];
+
+ real* vxLast = vLast;
+ real* vyLast = &vLast[numberOfPrecursorNodes];
+ real* vzLast = &vLast[2*numberOfPrecursorNodes];
+
+ real* vxCurrent = vCurrent;
+ real* vyCurrent = &vCurrent[numberOfPrecursorNodes];
+ real* vzCurrent = &vCurrent[2*numberOfPrecursorNodes];
+
+ if(d0PP < 1e6)
+ {
+ uint kNeighbor0PM = neighbors0PM[nodeIndex];
+ uint kNeighbor0MP = neighbors0MP[nodeIndex];
+ uint kNeighbor0MM = neighbors0MM[nodeIndex];
+
+ real d0PM = weights0PM[nodeIndex];
+ real d0MP = weights0MP[nodeIndex];
+ real d0MM = weights0MM[nodeIndex];
+
+ real invWeightSum = 1.f/(d0PP+d0PM+d0MP+d0MM);
+
+ vxLastInterpd = (vxLast[kNeighbor0PP]*d0PP + vxLast[kNeighbor0PM]*d0PM + vxLast[kNeighbor0MP]*d0MP + vxLast[kNeighbor0MM]*d0MM)*invWeightSum;
+ vyLastInterpd = (vyLast[kNeighbor0PP]*d0PP + vyLast[kNeighbor0PM]*d0PM + vyLast[kNeighbor0MP]*d0MP + vyLast[kNeighbor0MM]*d0MM)*invWeightSum;
+ vzLastInterpd = (vzLast[kNeighbor0PP]*d0PP + vzLast[kNeighbor0PM]*d0PM + vzLast[kNeighbor0MP]*d0MP + vzLast[kNeighbor0MM]*d0MM)*invWeightSum;
+
+ vxNextInterpd = (vxCurrent[kNeighbor0PP]*d0PP + vxCurrent[kNeighbor0PM]*d0PM + vxCurrent[kNeighbor0MP]*d0MP + vxCurrent[kNeighbor0MM]*d0MM)*invWeightSum;
+ vyNextInterpd = (vyCurrent[kNeighbor0PP]*d0PP + vyCurrent[kNeighbor0PM]*d0PM + vyCurrent[kNeighbor0MP]*d0MP + vyCurrent[kNeighbor0MM]*d0MM)*invWeightSum;
+ vzNextInterpd = (vzCurrent[kNeighbor0PP]*d0PP + vzCurrent[kNeighbor0PM]*d0PM + vzCurrent[kNeighbor0MP]*d0MP + vzCurrent[kNeighbor0MM]*d0MM)*invWeightSum;
+ }
+ else
+ {
+ vxLastInterpd = vxLast[kNeighbor0PP];
+ vyLastInterpd = vyLast[kNeighbor0PP];
+ vzLastInterpd = vzLast[kNeighbor0PP];
+
+ vxNextInterpd = vxCurrent[kNeighbor0PP];
+ vyNextInterpd = vyCurrent[kNeighbor0PP];
+ vzNextInterpd = vzCurrent[kNeighbor0PP];
+ }
+
+ // if(k==16300) printf("%f %f %f\n", vxLastInterpd, vyLastInterpd, vzLastInterpd);
+ real VeloX = (velocityX + (1.f-timeRatio)*vxLastInterpd + timeRatio*vxNextInterpd)/velocityRatio;
+ real VeloY = (velocityY + (1.f-timeRatio)*vyLastInterpd + timeRatio*vyNextInterpd)/velocityRatio;
+ real VeloZ = (velocityZ + (1.f-timeRatio)*vzLastInterpd + timeRatio*vzNextInterpd)/velocityRatio;
+ // From here on just a copy of QVelDeviceCompZeroPress
+ ////////////////////////////////////////////////////////////////////////////////
+
+ //////////////////////////////////////////////////////////////////////////
+ //! - Read distributions: style of reading and writing the distributions from/to stored arrays dependent on timestep
+ //! is based on the esoteric twist algorithm \ref <a href="https://doi.org/10.3390/computation5020019"><b>[ M. Geier
+ //! et al. (2017), DOI:10.3390/computation5020019 ]</b></a>
+ //!
+ Distributions27 dist;
+ getPointersToDistributions(dist, distributions, numberOfLBnodes, !isEvenTimestep);
+
+ unsigned int KQK = subgridDistanceIndices[nodeIndex]; //QK
+ unsigned int k000 = KQK; //000
+ unsigned int kP00 = KQK; //P00
+ unsigned int kM00 = neighborX[KQK]; //M00
+ unsigned int k0P0 = KQK; //n
+ unsigned int k0M0 = neighborY[KQK]; //s
+ unsigned int k00P = KQK; //t
+ unsigned int k00M = neighborZ[KQK]; //b
+ unsigned int kMM0 = neighborY[kM00]; //sw
+ unsigned int kPP0 = KQK; //ne
+ unsigned int kPM0 = k0M0; //se
+ unsigned int kMP0 = kM00; //nw
+ unsigned int kM0M = neighborZ[kM00]; //bw
+ unsigned int kP0P = KQK; //te
+ unsigned int kP0M = k00M; //be
+ unsigned int k0PP = KQK; //tn
+ unsigned int k0MM = neighborZ[k0M0]; //bs
+ unsigned int kM0P = kM00; //tw
+ unsigned int k0PM = k00M; //bn
+ unsigned int k0MP = k0M0; //ts
+ unsigned int kPMP = k0M0; //tse
+ unsigned int kMPM = kM0M; //bnw
+ unsigned int kMPP = kM00; //tnw
+ unsigned int kPMM = k0MM; //bse
+ unsigned int kMMP = kMM0; //tsw
+ unsigned int kPPM = k00M; //bne
+ unsigned int kPPP = KQK; //tne
+ unsigned int kMMM = neighborZ[kMM0]; //bsw
+
+ //////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
+ // based on BGK Plus Comp
+ //////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
+ real f_M00 = (dist.f[DIR_P00])[kP00];
+ real f_P00 = (dist.f[DIR_M00])[kM00];
+ real f_0M0 = (dist.f[DIR_0P0])[k0P0];
+ real f_0P0 = (dist.f[DIR_0M0])[k0M0];
+ real f_00M = (dist.f[DIR_00P])[k00P];
+ real f_00P = (dist.f[DIR_00M])[k00M];
+ real f_MM0 = (dist.f[DIR_PP0])[kPP0];
+ real f_PP0 = (dist.f[DIR_MM0])[kMM0];
+ real f_MP0 = (dist.f[DIR_PM0])[kPM0];
+ real f_PM0 = (dist.f[DIR_MP0])[kMP0];
+ real f_M0M = (dist.f[DIR_P0P])[kP0P];
+ real f_P0P = (dist.f[DIR_M0M])[kM0M];
+ real f_M0P = (dist.f[DIR_P0M])[kP0M];
+ real f_P0M = (dist.f[DIR_M0P])[kM0P];
+ real f_0MM = (dist.f[DIR_0PP])[k0PP];
+ real f_0PP = (dist.f[DIR_0MM])[k0MM];
+ real f_0PM = (dist.f[DIR_0MP])[k0MP];
+ real f_0MP = (dist.f[DIR_0PM])[k0PM];
+ real f_000 = (dist.f[DIR_000])[k000];
+ real f_MMM = (dist.f[DIR_PPP])[kPPP];
+ real f_PPM = (dist.f[DIR_MMP])[kMMP];
+ real f_MPM = (dist.f[DIR_PMP])[kPMP];
+ real f_PMM = (dist.f[DIR_MPP])[kMPP];
+ real f_MMP = (dist.f[DIR_PPM])[kPPM];
+ real f_PPP = (dist.f[DIR_MMM])[kMMM];
+ real f_MPP = (dist.f[DIR_PMM])[kPMM];
+ real f_PMP = (dist.f[DIR_MPM])[kMPM];
+
+ ////////////////////////////////////////////////////////////////////////////////
+ //! - Set macroscopic quantities
+ //!
+ real drho = c0o1;
+
+ real vx1 = VeloX;
+
+ real vx2 = VeloY;
+
+ real vx3 = VeloZ;
+
+ real cusq = c3o2 * (vx1 * vx1 + vx2 * vx2 + vx3 * vx3);
+
+ ////////////////////////////////////////////////////////////////////////////////
+ f_000 = c8o27* (drho-(drho+c1o1)*cusq);
+ f_P00 = c2o27* (drho+(drho+c1o1)*(c3o1*( vx1 )+c9o2*( vx1 )*( vx1 )-cusq));
+ f_M00 = c2o27* (drho+(drho+c1o1)*(c3o1*(-vx1 )+c9o2*(-vx1 )*(-vx1 )-cusq));
+ f_0P0 = c2o27* (drho+(drho+c1o1)*(c3o1*( vx2 )+c9o2*( vx2 )*( vx2 )-cusq));
+ f_0M0 = c2o27* (drho+(drho+c1o1)*(c3o1*( -vx2 )+c9o2*( -vx2 )*( -vx2 )-cusq));
+ f_00P = c2o27* (drho+(drho+c1o1)*(c3o1*( vx3)+c9o2*( vx3)*( vx3)-cusq));
+ f_00M = c2o27* (drho+(drho+c1o1)*(c3o1*( -vx3)+c9o2*( -vx3)*( -vx3)-cusq));
+ f_PP0 = c1o54* (drho+(drho+c1o1)*(c3o1*( vx1+vx2 )+c9o2*( vx1+vx2 )*( vx1+vx2 )-cusq));
+ f_MM0 = c1o54* (drho+(drho+c1o1)*(c3o1*(-vx1-vx2 )+c9o2*(-vx1-vx2 )*(-vx1-vx2 )-cusq));
+ f_PM0 = c1o54* (drho+(drho+c1o1)*(c3o1*( vx1-vx2 )+c9o2*( vx1-vx2 )*( vx1-vx2 )-cusq));
+ f_MP0 = c1o54* (drho+(drho+c1o1)*(c3o1*(-vx1+vx2 )+c9o2*(-vx1+vx2 )*(-vx1+vx2 )-cusq));
+ f_P0P = c1o54* (drho+(drho+c1o1)*(c3o1*( vx1 +vx3)+c9o2*( vx1 +vx3)*( vx1 +vx3)-cusq));
+ f_M0M = c1o54* (drho+(drho+c1o1)*(c3o1*(-vx1 -vx3)+c9o2*(-vx1 -vx3)*(-vx1 -vx3)-cusq));
+ f_P0M = c1o54* (drho+(drho+c1o1)*(c3o1*( vx1 -vx3)+c9o2*( vx1 -vx3)*( vx1 -vx3)-cusq));
+ f_M0P = c1o54* (drho+(drho+c1o1)*(c3o1*(-vx1 +vx3)+c9o2*(-vx1 +vx3)*(-vx1 +vx3)-cusq));
+ f_0PP = c1o54* (drho+(drho+c1o1)*(c3o1*( vx2+vx3)+c9o2*( vx2+vx3)*( vx2+vx3)-cusq));
+ f_0MM = c1o54* (drho+(drho+c1o1)*(c3o1*( -vx2-vx3)+c9o2*( -vx2-vx3)*( -vx2-vx3)-cusq));
+ f_0PM = c1o54* (drho+(drho+c1o1)*(c3o1*( vx2-vx3)+c9o2*( vx2-vx3)*( vx2-vx3)-cusq));
+ f_0MP = c1o54* (drho+(drho+c1o1)*(c3o1*( -vx2+vx3)+c9o2*( -vx2+vx3)*( -vx2+vx3)-cusq));
+ f_PPP = c1o216*(drho+(drho+c1o1)*(c3o1*( vx1+vx2+vx3)+c9o2*( vx1+vx2+vx3)*( vx1+vx2+vx3)-cusq));
+ f_MMM = c1o216*(drho+(drho+c1o1)*(c3o1*(-vx1-vx2-vx3)+c9o2*(-vx1-vx2-vx3)*(-vx1-vx2-vx3)-cusq));
+ f_PPM = c1o216*(drho+(drho+c1o1)*(c3o1*( vx1+vx2-vx3)+c9o2*( vx1+vx2-vx3)*( vx1+vx2-vx3)-cusq));
+ f_MMP = c1o216*(drho+(drho+c1o1)*(c3o1*(-vx1-vx2+vx3)+c9o2*(-vx1-vx2+vx3)*(-vx1-vx2+vx3)-cusq));
+ f_PMP = c1o216*(drho+(drho+c1o1)*(c3o1*( vx1-vx2+vx3)+c9o2*( vx1-vx2+vx3)*( vx1-vx2+vx3)-cusq));
+ f_MPM = c1o216*(drho+(drho+c1o1)*(c3o1*(-vx1+vx2-vx3)+c9o2*(-vx1+vx2-vx3)*(-vx1+vx2-vx3)-cusq));
+ f_PMM = c1o216*(drho+(drho+c1o1)*(c3o1*( vx1-vx2-vx3)+c9o2*( vx1-vx2-vx3)*( vx1-vx2-vx3)-cusq));
+ f_MPP = c1o216*(drho+(drho+c1o1)*(c3o1*(-vx1+vx2+vx3)+c9o2*(-vx1+vx2+vx3)*(-vx1+vx2+vx3)-cusq));
+
+ ////////////////////////////////////////////////////////////////////////////////
+ //! write the new distributions to the bc nodes
+ //!
+ (dist.f[DIR_P00])[kP00] = f_M00;
+ (dist.f[DIR_PP0])[kPP0] = f_MM0;
+ (dist.f[DIR_P0M])[kP0M] = f_M0P;
+ (dist.f[DIR_PM0])[kPM0] = f_MP0;
+ (dist.f[DIR_PMP])[kPMP] = f_MPM;
+ (dist.f[DIR_P0P])[kP0P] = f_M0M;
+ (dist.f[DIR_PPM])[kPPM] = f_MMP;
+ (dist.f[DIR_PPP])[kPPP] = f_MMM;
+ (dist.f[DIR_PMM])[kPMM] = f_MPP;
+
+ (dist.f[DIR_M00])[kM00] = f_P00;
+ (dist.f[DIR_MM0])[kMM0] = f_PP0;
+ (dist.f[DIR_M0M])[kM0M] = f_P0P;
+ (dist.f[DIR_MP0])[kMP0] = f_PM0;
+ (dist.f[DIR_M0P])[kM0P] = f_P0M;
+ (dist.f[DIR_MMM])[kMMM] = f_PPP;
+ (dist.f[DIR_MMP])[kMMP] = f_PPM;
+ (dist.f[DIR_MPP])[kMPP] = f_PMM;
+ (dist.f[DIR_MPM])[kMPM] = f_PMP;
+
+ (dist.f[DIR_0P0])[k0P0] = f_0M0;
+ (dist.f[DIR_0M0])[k0M0] = f_0P0;
+ (dist.f[DIR_00P])[k00P] = f_00M;
+ (dist.f[DIR_00M])[k00M] = f_00P;
+ (dist.f[DIR_0PP])[k0PP] = f_0MM;
+ (dist.f[DIR_0MM])[k0MM] = f_0PP;
+ (dist.f[DIR_0PM])[k0PM] = f_0MP;
+ (dist.f[DIR_0MP])[k0MP] = f_0PM;
+ (dist.f[DIR_000])[k000] = f_000;
+}
+////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
+__global__ void PrecursorDeviceDistributions(
+ int *subgridDistanceIndices,
+ int numberOfBCnodes,
+ int numberOfPrecursorNodes,
+ real* distributions,
+ uint* neighborX,
+ uint* neighborY,
+ uint* neighborZ,
+ uint* neighbors0PP,
+ uint* neighbors0PM,
+ uint* neighbors0MP,
+ uint* neighbors0MM,
+ real* weights0PP,
+ real* weights0PM,
+ real* weights0MP,
+ real* weights0MM,
+ real* fsLast,
+ real* fsNext,
+ real timeRatio,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep)
+{
+ ////////////////////////////////////////////////////////////////////////////////
+ //! - Get node index coordinates from threadIdx, blockIdx, blockDim and gridDim.
+ //!
+ const unsigned nodeIndex = getNodeIndex();
+
+ if(nodeIndex>=numberOfBCnodes) return;
+
+ uint kNeighbor0PP = neighbors0PP[nodeIndex];
+ real d0PP = weights0PP[nodeIndex];
+
+ real f0LastInterp, f1LastInterp, f2LastInterp, f3LastInterp, f4LastInterp, f5LastInterp, f6LastInterp, f7LastInterp, f8LastInterp;
+ real f0NextInterp, f1NextInterp, f2NextInterp, f3NextInterp, f4NextInterp, f5NextInterp, f6NextInterp, f7NextInterp, f8NextInterp;
+
+ real* f0Last = fsLast;
+ real* f1Last = &fsLast[ numberOfPrecursorNodes];
+ real* f2Last = &fsLast[2*numberOfPrecursorNodes];
+ real* f3Last = &fsLast[3*numberOfPrecursorNodes];
+ real* f4Last = &fsLast[4*numberOfPrecursorNodes];
+ real* f5Last = &fsLast[5*numberOfPrecursorNodes];
+ real* f6Last = &fsLast[6*numberOfPrecursorNodes];
+ real* f7Last = &fsLast[7*numberOfPrecursorNodes];
+ real* f8Last = &fsLast[8*numberOfPrecursorNodes];
+
+ real* f0Next = fsNext;
+ real* f1Next = &fsNext[ numberOfPrecursorNodes];
+ real* f2Next = &fsNext[2*numberOfPrecursorNodes];
+ real* f3Next = &fsNext[3*numberOfPrecursorNodes];
+ real* f4Next = &fsNext[4*numberOfPrecursorNodes];
+ real* f5Next = &fsNext[5*numberOfPrecursorNodes];
+ real* f6Next = &fsNext[6*numberOfPrecursorNodes];
+ real* f7Next = &fsNext[7*numberOfPrecursorNodes];
+ real* f8Next = &fsNext[8*numberOfPrecursorNodes];
+
+
+ if(d0PP<1e6)
+ {
+ uint kNeighbor0PM = neighbors0PM[nodeIndex];
+ uint kNeighbor0MP = neighbors0MP[nodeIndex];
+ uint kNeighbor0MM = neighbors0MM[nodeIndex];
+
+ real d0PM = weights0PM[nodeIndex];
+ real d0MP = weights0MP[nodeIndex];
+ real d0MM = weights0MM[nodeIndex];
+
+ real invWeightSum = 1.f/(d0PP+d0PM+d0MP+d0MM);
+
+ f0LastInterp = (f0Last[kNeighbor0PP]*d0PP + f0Last[kNeighbor0PM]*d0PM + f0Last[kNeighbor0MP]*d0MP + f0Last[kNeighbor0MM]*d0MM)*invWeightSum;
+ f0NextInterp = (f0Next[kNeighbor0PP]*d0PP + f0Next[kNeighbor0PM]*d0PM + f0Next[kNeighbor0MP]*d0MP + f0Next[kNeighbor0MM]*d0MM)*invWeightSum;
+
+ f1LastInterp = (f1Last[kNeighbor0PP]*d0PP + f1Last[kNeighbor0PM]*d0PM + f1Last[kNeighbor0MP]*d0MP + f1Last[kNeighbor0MM]*d0MM)*invWeightSum;
+ f1NextInterp = (f1Next[kNeighbor0PP]*d0PP + f1Next[kNeighbor0PM]*d0PM + f1Next[kNeighbor0MP]*d0MP + f1Next[kNeighbor0MM]*d0MM)*invWeightSum;
+
+ f2LastInterp = (f2Last[kNeighbor0PP]*d0PP + f2Last[kNeighbor0PM]*d0PM + f2Last[kNeighbor0MP]*d0MP + f2Last[kNeighbor0MM]*d0MM)*invWeightSum;
+ f2NextInterp = (f2Next[kNeighbor0PP]*d0PP + f2Next[kNeighbor0PM]*d0PM + f2Next[kNeighbor0MP]*d0MP + f2Next[kNeighbor0MM]*d0MM)*invWeightSum;
+
+ f3LastInterp = (f3Last[kNeighbor0PP]*d0PP + f3Last[kNeighbor0PM]*d0PM + f3Last[kNeighbor0MP]*d0MP + f3Last[kNeighbor0MM]*d0MM)*invWeightSum;
+ f3NextInterp = (f3Next[kNeighbor0PP]*d0PP + f3Next[kNeighbor0PM]*d0PM + f3Next[kNeighbor0MP]*d0MP + f3Next[kNeighbor0MM]*d0MM)*invWeightSum;
+
+ f4LastInterp = (f4Last[kNeighbor0PP]*d0PP + f4Last[kNeighbor0PM]*d0PM + f4Last[kNeighbor0MP]*d0MP + f4Last[kNeighbor0MM]*d0MM)*invWeightSum;
+ f4NextInterp = (f4Next[kNeighbor0PP]*d0PP + f4Next[kNeighbor0PM]*d0PM + f4Next[kNeighbor0MP]*d0MP + f4Next[kNeighbor0MM]*d0MM)*invWeightSum;
+
+ f5LastInterp = (f5Last[kNeighbor0PP]*d0PP + f5Last[kNeighbor0PM]*d0PM + f5Last[kNeighbor0MP]*d0MP + f5Last[kNeighbor0MM]*d0MM)*invWeightSum;
+ f5NextInterp = (f5Next[kNeighbor0PP]*d0PP + f5Next[kNeighbor0PM]*d0PM + f5Next[kNeighbor0MP]*d0MP + f5Next[kNeighbor0MM]*d0MM)*invWeightSum;
+
+ f6LastInterp = (f6Last[kNeighbor0PP]*d0PP + f6Last[kNeighbor0PM]*d0PM + f6Last[kNeighbor0MP]*d0MP + f6Last[kNeighbor0MM]*d0MM)*invWeightSum;
+ f6NextInterp = (f6Next[kNeighbor0PP]*d0PP + f6Next[kNeighbor0PM]*d0PM + f6Next[kNeighbor0MP]*d0MP + f6Next[kNeighbor0MM]*d0MM)*invWeightSum;
+
+ f7LastInterp = (f7Last[kNeighbor0PP]*d0PP + f7Last[kNeighbor0PM]*d0PM + f7Last[kNeighbor0MP]*d0MP + f7Last[kNeighbor0MM]*d0MM)*invWeightSum;
+ f7NextInterp = (f7Next[kNeighbor0PP]*d0PP + f7Next[kNeighbor0PM]*d0PM + f7Next[kNeighbor0MP]*d0MP + f7Next[kNeighbor0MM]*d0MM)*invWeightSum;
+
+ f8LastInterp = (f8Last[kNeighbor0PP]*d0PP + f8Last[kNeighbor0PM]*d0PM + f8Last[kNeighbor0MP]*d0MP + f8Last[kNeighbor0MM]*d0MM)*invWeightSum;
+ f8NextInterp = (f8Next[kNeighbor0PP]*d0PP + f8Next[kNeighbor0PM]*d0PM + f8Next[kNeighbor0MP]*d0MP + f8Next[kNeighbor0MM]*d0MM)*invWeightSum;
+
+ } else {
+ f0LastInterp = f0Last[kNeighbor0PP];
+ f1LastInterp = f1Last[kNeighbor0PP];
+ f2LastInterp = f2Last[kNeighbor0PP];
+ f3LastInterp = f3Last[kNeighbor0PP];
+ f4LastInterp = f4Last[kNeighbor0PP];
+ f5LastInterp = f5Last[kNeighbor0PP];
+ f6LastInterp = f6Last[kNeighbor0PP];
+ f7LastInterp = f7Last[kNeighbor0PP];
+ f8LastInterp = f8Last[kNeighbor0PP];
+
+ f0NextInterp = f0Next[kNeighbor0PP];
+ f1NextInterp = f1Next[kNeighbor0PP];
+ f2NextInterp = f2Next[kNeighbor0PP];
+ f3NextInterp = f3Next[kNeighbor0PP];
+ f4NextInterp = f4Next[kNeighbor0PP];
+ f5NextInterp = f5Next[kNeighbor0PP];
+ f6NextInterp = f6Next[kNeighbor0PP];
+ f7NextInterp = f7Next[kNeighbor0PP];
+ f8NextInterp = f8Next[kNeighbor0PP];
+ }
+ //////////////////////////////////////////////////////////////////////////
+ //! - Read distributions: style of reading and writing the distributions from/to stored arrays dependent on timestep
+ //! is based on the esoteric twist algorithm \ref <a href="https://doi.org/10.3390/computation5020019"><b>[ M. Geier
+ //! et al. (2017), DOI:10.3390/computation5020019 ]</b></a>
+ //!
+ Distributions27 dist;
+ getPointersToDistributions(dist, distributions, numberOfLBnodes, !isEvenTimestep);
+
+ unsigned int KQK = subgridDistanceIndices[nodeIndex];
+ // unsigned int k000= KQK;
+ unsigned int kP00 = KQK;
+ // unsigned int kM00 = neighborX[KQK];
+ // unsigned int k0P0 = KQK;
+ unsigned int k0M0 = neighborY[KQK];
+ // unsigned int k00P = KQK;
+ unsigned int k00M = neighborZ[KQK];
+ // unsigned int kMM0 = neighborY[kM00];
+ unsigned int kPP0 = KQK;
+ unsigned int kPM0 = k0M0;
+ // unsigned int kMP0 = kM00;
+ // unsigned int kM0M = neighborZ[kM00];
+ unsigned int kP0P = KQK;
+ unsigned int kP0M = k00M;
+ // unsigned int kM0P = kM00;
+ unsigned int k0MM = neighborZ[k0M0];
+ // unsigned int k0PM = k00M;
+ // unsigned int k0MP = k0M0;
+ unsigned int kPMP = k0M0;
+ // unsigned int kMPM = kM0M;
+ // unsigned int kMPP = kM00;
+ unsigned int kPMM = k0MM;
+ // unsigned int kMMP = kMM0;
+ unsigned int kPPM = k00M;
+ unsigned int kPPP = KQK;
+ // unsigned int kMMM = neighborZ[kMM0];
+
+ dist.f[DIR_P00][kP00] = f0LastInterp*(1.f-timeRatio) + f0NextInterp*timeRatio;
+ dist.f[DIR_PP0][kPP0] = f1LastInterp*(1.f-timeRatio) + f1NextInterp*timeRatio;
+ dist.f[DIR_PM0][kPM0] = f2LastInterp*(1.f-timeRatio) + f2NextInterp*timeRatio;
+ dist.f[DIR_P0P][kP0P] = f3LastInterp*(1.f-timeRatio) + f3NextInterp*timeRatio;
+ dist.f[DIR_P0M][kP0M] = f4LastInterp*(1.f-timeRatio) + f4NextInterp*timeRatio;
+ dist.f[DIR_PPP][kPPP] = f5LastInterp*(1.f-timeRatio) + f5NextInterp*timeRatio;
+ dist.f[DIR_PMP][kPMP] = f6LastInterp*(1.f-timeRatio) + f6NextInterp*timeRatio;
+ dist.f[DIR_PPM][kPPM] = f7LastInterp*(1.f-timeRatio) + f7NextInterp*timeRatio;
+ dist.f[DIR_PMM][kPMM] = f8LastInterp*(1.f-timeRatio) + f8NextInterp*timeRatio;
+}
+////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
+// NOTE: Has not been tested after bug fix!
+__global__ void QPrecursorDeviceDistributions(
+ int* subgridDistanceIndices,
+ real* subgridDistances,
+ int sizeQ,
+ int numberOfBCnodes,
+ int numberOfPrecursorNodes,
+ real* distributions,
+ uint* neighborX,
+ uint* neighborY,
+ uint* neighborZ,
+ uint* neighbors0PP,
+ uint* neighbors0PM,
+ uint* neighbors0MP,
+ uint* neighbors0MM,
+ real* weights0PP,
+ real* weights0PM,
+ real* weights0MP,
+ real* weights0MM,
+ real* fsLast,
+ real* fsNext,
+ real timeRatio,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep)
+{
+ ////////////////////////////////////////////////////////////////////////////////
+ //! - Get node index coordinates from threadIdx, blockIdx, blockDim and gridDim.
+ //!
+ const unsigned nodeIndex = getNodeIndex();
+
+ if(nodeIndex>=numberOfBCnodes) return;
+
+ uint kNeighbor0PP = neighbors0PP[nodeIndex];
+ real d0PP = weights0PP[nodeIndex];
+
+ real f0LastInterp, f1LastInterp, f2LastInterp, f3LastInterp, f4LastInterp, f5LastInterp, f6LastInterp, f7LastInterp, f8LastInterp;
+ real f0NextInterp, f1NextInterp, f2NextInterp, f3NextInterp, f4NextInterp, f5NextInterp, f6NextInterp, f7NextInterp, f8NextInterp;
+
+ real* f0Last = fsLast;
+ real* f1Last = &fsLast[ numberOfPrecursorNodes];
+ real* f2Last = &fsLast[2*numberOfPrecursorNodes];
+ real* f3Last = &fsLast[3*numberOfPrecursorNodes];
+ real* f4Last = &fsLast[4*numberOfPrecursorNodes];
+ real* f5Last = &fsLast[5*numberOfPrecursorNodes];
+ real* f6Last = &fsLast[6*numberOfPrecursorNodes];
+ real* f7Last = &fsLast[7*numberOfPrecursorNodes];
+ real* f8Last = &fsLast[8*numberOfPrecursorNodes];
+
+ real* f0Next = fsNext;
+ real* f1Next = &fsNext[ numberOfPrecursorNodes];
+ real* f2Next = &fsNext[2*numberOfPrecursorNodes];
+ real* f3Next = &fsNext[3*numberOfPrecursorNodes];
+ real* f4Next = &fsNext[4*numberOfPrecursorNodes];
+ real* f5Next = &fsNext[5*numberOfPrecursorNodes];
+ real* f6Next = &fsNext[6*numberOfPrecursorNodes];
+ real* f7Next = &fsNext[7*numberOfPrecursorNodes];
+ real* f8Next = &fsNext[8*numberOfPrecursorNodes];
+
+
+ if(d0PP<1e6)
+ {
+ uint kNeighbor0PM = neighbors0PM[nodeIndex];
+ uint kNeighbor0MP = neighbors0MP[nodeIndex];
+ uint kNeighbor0MM = neighbors0MM[nodeIndex];
+
+ real d0PM = weights0PM[nodeIndex];
+ real d0MP = weights0MP[nodeIndex];
+ real d0MM = weights0MM[nodeIndex];
+
+ real invWeightSum = 1.f/(d0PP+d0PM+d0MP+d0MM);
+
+ f0LastInterp = (f0Last[kNeighbor0PP]*d0PP + f0Last[kNeighbor0PM]*d0PM + f0Last[kNeighbor0MP]*d0MP + f0Last[kNeighbor0MM]*d0MM)*invWeightSum;
+ f0NextInterp = (f0Next[kNeighbor0PP]*d0PP + f0Next[kNeighbor0PM]*d0PM + f0Next[kNeighbor0MP]*d0MP + f0Next[kNeighbor0MM]*d0MM)*invWeightSum;
+
+ f1LastInterp = (f1Last[kNeighbor0PP]*d0PP + f1Last[kNeighbor0PM]*d0PM + f1Last[kNeighbor0MP]*d0MP + f1Last[kNeighbor0MM]*d0MM)*invWeightSum;
+ f1NextInterp = (f1Next[kNeighbor0PP]*d0PP + f1Next[kNeighbor0PM]*d0PM + f1Next[kNeighbor0MP]*d0MP + f1Next[kNeighbor0MM]*d0MM)*invWeightSum;
+
+ f2LastInterp = (f2Last[kNeighbor0PP]*d0PP + f2Last[kNeighbor0PM]*d0PM + f2Last[kNeighbor0MP]*d0MP + f2Last[kNeighbor0MM]*d0MM)*invWeightSum;
+ f2NextInterp = (f2Next[kNeighbor0PP]*d0PP + f2Next[kNeighbor0PM]*d0PM + f2Next[kNeighbor0MP]*d0MP + f2Next[kNeighbor0MM]*d0MM)*invWeightSum;
+
+ f3LastInterp = (f3Last[kNeighbor0PP]*d0PP + f3Last[kNeighbor0PM]*d0PM + f3Last[kNeighbor0MP]*d0MP + f3Last[kNeighbor0MM]*d0MM)*invWeightSum;
+ f3NextInterp = (f3Next[kNeighbor0PP]*d0PP + f3Next[kNeighbor0PM]*d0PM + f3Next[kNeighbor0MP]*d0MP + f3Next[kNeighbor0MM]*d0MM)*invWeightSum;
+
+ f4LastInterp = (f4Last[kNeighbor0PP]*d0PP + f4Last[kNeighbor0PM]*d0PM + f4Last[kNeighbor0MP]*d0MP + f4Last[kNeighbor0MM]*d0MM)*invWeightSum;
+ f4NextInterp = (f4Next[kNeighbor0PP]*d0PP + f4Next[kNeighbor0PM]*d0PM + f4Next[kNeighbor0MP]*d0MP + f4Next[kNeighbor0MM]*d0MM)*invWeightSum;
+
+ f5LastInterp = (f5Last[kNeighbor0PP]*d0PP + f5Last[kNeighbor0PM]*d0PM + f5Last[kNeighbor0MP]*d0MP + f5Last[kNeighbor0MM]*d0MM)*invWeightSum;
+ f5NextInterp = (f5Next[kNeighbor0PP]*d0PP + f5Next[kNeighbor0PM]*d0PM + f5Next[kNeighbor0MP]*d0MP + f5Next[kNeighbor0MM]*d0MM)*invWeightSum;
+
+ f6LastInterp = (f6Last[kNeighbor0PP]*d0PP + f6Last[kNeighbor0PM]*d0PM + f6Last[kNeighbor0MP]*d0MP + f6Last[kNeighbor0MM]*d0MM)*invWeightSum;
+ f6NextInterp = (f6Next[kNeighbor0PP]*d0PP + f6Next[kNeighbor0PM]*d0PM + f6Next[kNeighbor0MP]*d0MP + f6Next[kNeighbor0MM]*d0MM)*invWeightSum;
+
+ f7LastInterp = (f7Last[kNeighbor0PP]*d0PP + f7Last[kNeighbor0PM]*d0PM + f7Last[kNeighbor0MP]*d0MP + f7Last[kNeighbor0MM]*d0MM)*invWeightSum;
+ f7NextInterp = (f7Next[kNeighbor0PP]*d0PP + f7Next[kNeighbor0PM]*d0PM + f7Next[kNeighbor0MP]*d0MP + f7Next[kNeighbor0MM]*d0MM)*invWeightSum;
+
+ f8LastInterp = (f8Last[kNeighbor0PP]*d0PP + f8Last[kNeighbor0PM]*d0PM + f8Last[kNeighbor0MP]*d0MP + f8Last[kNeighbor0MM]*d0MM)*invWeightSum;
+ f8NextInterp = (f8Next[kNeighbor0PP]*d0PP + f8Next[kNeighbor0PM]*d0PM + f8Next[kNeighbor0MP]*d0MP + f8Next[kNeighbor0MM]*d0MM)*invWeightSum;
+
+ } else {
+ f0LastInterp = f0Last[kNeighbor0PP];
+ f1LastInterp = f1Last[kNeighbor0PP];
+ f2LastInterp = f2Last[kNeighbor0PP];
+ f3LastInterp = f3Last[kNeighbor0PP];
+ f4LastInterp = f4Last[kNeighbor0PP];
+ f5LastInterp = f5Last[kNeighbor0PP];
+ f6LastInterp = f6Last[kNeighbor0PP];
+ f7LastInterp = f7Last[kNeighbor0PP];
+ f8LastInterp = f8Last[kNeighbor0PP];
+
+ f0NextInterp = f0Next[kNeighbor0PP];
+ f1NextInterp = f1Next[kNeighbor0PP];
+ f2NextInterp = f2Next[kNeighbor0PP];
+ f3NextInterp = f3Next[kNeighbor0PP];
+ f4NextInterp = f4Next[kNeighbor0PP];
+ f5NextInterp = f5Next[kNeighbor0PP];
+ f6NextInterp = f6Next[kNeighbor0PP];
+ f7NextInterp = f7Next[kNeighbor0PP];
+ f8NextInterp = f8Next[kNeighbor0PP];
+ }
+ //////////////////////////////////////////////////////////////////////////
+ //! - Read distributions: style of reading and writing the distributions from/to stored arrays dependent on timestep
+ //! is based on the esoteric twist algorithm \ref <a href="https://doi.org/10.3390/computation5020019"><b>[ M. Geier
+ //! et al. (2017), DOI:10.3390/computation5020019 ]</b></a>
+ //!
+ Distributions27 dist;
+ getPointersToDistributions(dist, distributions, numberOfLBnodes, !isEvenTimestep);
+
+ unsigned int KQK = subgridDistanceIndices[nodeIndex];
+ // unsigned int k000= KQK;
+ unsigned int kP00 = KQK;
+ // unsigned int kM00 = neighborX[KQK];
+ // unsigned int k0P0 = KQK;
+ unsigned int k0M0 = neighborY[KQK];
+ // unsigned int k00P = KQK;
+ unsigned int k00M = neighborZ[KQK];
+ // unsigned int kMM0 = neighborY[kM00];
+ unsigned int kPP0 = KQK;
+ unsigned int kPM0 = k0M0;
+ // unsigned int kMP0 = kM00;
+ // unsigned int kM0M = neighborZ[kM00];
+ unsigned int kP0P = KQK;
+ unsigned int kP0M = k00M;
+ // unsigned int kM0P = kM00;
+ unsigned int k0MM = neighborZ[k0M0];
+ // unsigned int k0PM = k00M;
+ // unsigned int k0MP = k0M0;
+ unsigned int kPMP = k0M0;
+ // unsigned int kMPM = kM0M;
+ // unsigned int kMPP = kM00;
+ unsigned int kPMM = k0MM;
+ // unsigned int kMMP = kMM0;
+ unsigned int kPPM = k00M;
+ unsigned int kPPP = KQK;
+ // unsigned int kMMM = neighborZ[kMM0];
+ SubgridDistances27 qs;
+ getPointersToSubgridDistances(qs, subgridDistances, sizeQ);
+
+ real q;
+ q = qs.q[DIR_P00][nodeIndex]; if(q>= c0o1 && q <= c1o1) dist.f[DIR_P00][kP00] = f0LastInterp*(1.f-timeRatio) + f0NextInterp*timeRatio;
+ q = qs.q[DIR_PP0][nodeIndex]; if(q>= c0o1 && q <= c1o1) dist.f[DIR_PP0][kPP0] = f1LastInterp*(1.f-timeRatio) + f1NextInterp*timeRatio;
+ q = qs.q[DIR_PM0][nodeIndex]; if(q>= c0o1 && q <= c1o1) dist.f[DIR_PM0][kPM0] = f2LastInterp*(1.f-timeRatio) + f2NextInterp*timeRatio;
+ q = qs.q[DIR_P0P][nodeIndex]; if(q>= c0o1 && q <= c1o1) dist.f[DIR_P0P][kP0P] = f3LastInterp*(1.f-timeRatio) + f3NextInterp*timeRatio;
+ q = qs.q[DIR_P0M][nodeIndex]; if(q>= c0o1 && q <= c1o1) dist.f[DIR_P0M][kP0M] = f4LastInterp*(1.f-timeRatio) + f4NextInterp*timeRatio;
+ q = qs.q[DIR_PPP][nodeIndex]; if(q>= c0o1 && q <= c1o1) dist.f[DIR_PPP][kPPP] = f5LastInterp*(1.f-timeRatio) + f5NextInterp*timeRatio;
+ q = qs.q[DIR_PMP][nodeIndex]; if(q>= c0o1 && q <= c1o1) dist.f[DIR_PMP][kPMP] = f6LastInterp*(1.f-timeRatio) + f6NextInterp*timeRatio;
+ q = qs.q[DIR_PPM][nodeIndex]; if(q>= c0o1 && q <= c1o1) dist.f[DIR_PPM][kPPM] = f7LastInterp*(1.f-timeRatio) + f7NextInterp*timeRatio;
+ q = qs.q[DIR_PMM][nodeIndex]; if(q>= c0o1 && q <= c1o1) dist.f[DIR_PMM][kPMM] = f8LastInterp*(1.f-timeRatio) + f8NextInterp*timeRatio;
+
+}
+////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
diff --git a/src/gpu/VirtualFluids_GPU/GPU/PressBCs27.cu b/src/gpu/VirtualFluids_GPU/GPU/PressBCs27.cu
index ccb2ce79c63515e59e4f9ae75016f44ced71a170..02cfd2bce3723162b645cef568c87ca3b1dd2720 100644
--- a/src/gpu/VirtualFluids_GPU/GPU/PressBCs27.cu
+++ b/src/gpu/VirtualFluids_GPU/GPU/PressBCs27.cu
@@ -1,29 +1,63 @@
-/* Device code */
-#include "LBM/LB.h"
+//=======================================================================================
+// ____ ____ __ ______ __________ __ __ __ __
+// \ \ | | | | | _ \ |___ ___| | | | | / \ | |
+// \ \ | | | | | |_) | | | | | | | / \ | |
+// \ \ | | | | | _ / | | | | | | / /\ \ | |
+// \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
+// \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
+// \ \ | | ________________________________________________________________
+// \ \ | | | ______________________________________________________________|
+// \ \| | | | __ __ __ __ ______ _______
+// \ | | |_____ | | | | | | | | | _ \ / _____)
+// \ | | _____| | | | | | | | | | | \ \ \_______
+// \ | | | | |_____ | \_/ | | | | |_/ / _____ |
+// \ _____| |__| |________| \_______/ |__| |______/ (_______/
+//
+// This file is part of VirtualFluids. VirtualFluids is free software: you can
+// redistribute it and/or modify it under the terms of the GNU General Public
+// License as published by the Free Software Foundation, either version 3 of
+// the License, or (at your option) any later version.
+//
+// VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
+// ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+// FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
+// for more details.
+//
+// You should have received a copy of the GNU General Public License along
+// with VirtualFluids (see COPYING.txt). If not, see <http://www.gnu.org/licenses/>.
+//
+//! \file PressBCs27.cu
+//! \ingroup GPU
+//! \author Martin Schoenherr, Anna Wellmann
+//======================================================================================
+#include "LBM/LB.h"
#include "lbm/constants/D3Q27.h"
#include "lbm/constants/NumericConstants.h"
-#include "KernelUtilities.h"
+#include "lbm/MacroscopicQuantities.h"
+#include "LBM/GPUHelperFunctions/KernelUtilities.h"
using namespace vf::lbm::constant;
using namespace vf::lbm::dir;
+using namespace vf::gpu;
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
-__global__ void QInflowScaleByPressDevice27( real* rhoBC,
- real* DD,
- int* k_Q,
- int* k_N,
- int numberOfBCnodes,
- real om1,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep)
+__global__ void QInflowScaleByPressDevice27(
+ real* rhoBC,
+ real* DD,
+ int* k_Q,
+ int* k_N,
+ int numberOfBCnodes,
+ real om1,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep)
{
////////////////////////////////////////////////////////////////////////////////
- const unsigned x = threadIdx.x; // Globaler x-Index
- const unsigned y = blockIdx.x; // Globaler y-Index
- const unsigned z = blockIdx.y; // Globaler z-Index
+ const unsigned x = threadIdx.x; // Globaler x-Index
+ const unsigned y = blockIdx.x; // Globaler y-Index
+ const unsigned z = blockIdx.y; // Globaler z-Index
const unsigned nx = blockDim.x;
const unsigned ny = gridDim.x;
@@ -97,141 +131,141 @@ __global__ void QInflowScaleByPressDevice27( real* rhoBC,
Distributions27 D;
if (isEvenTimestep==true)
{
- D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
- }
+ D.f[DIR_P00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_MPM * numberOfLBnodes];
+ }
else
{
- D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
+ D.f[DIR_M00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_MPM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_PMP * numberOfLBnodes];
}
//////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
- real f1_E = (D.f[DIR_P00 ])[k1e ];
- real f1_W = (D.f[DIR_M00 ])[k1w ];
- real f1_N = (D.f[DIR_0P0 ])[k1n ];
- real f1_S = (D.f[DIR_0M0 ])[k1s ];
- real f1_T = (D.f[DIR_00P ])[k1t ];
- real f1_B = (D.f[DIR_00M ])[k1b ];
- real f1_NE = (D.f[DIR_PP0 ])[k1ne ];
- real f1_SW = (D.f[DIR_MM0 ])[k1sw ];
- real f1_SE = (D.f[DIR_PM0 ])[k1se ];
- real f1_NW = (D.f[DIR_MP0 ])[k1nw ];
- real f1_TE = (D.f[DIR_P0P ])[k1te ];
- real f1_BW = (D.f[DIR_M0M ])[k1bw ];
- real f1_BE = (D.f[DIR_P0M ])[k1be ];
- real f1_TW = (D.f[DIR_M0P ])[k1tw ];
- real f1_TN = (D.f[DIR_0PP ])[k1tn ];
- real f1_BS = (D.f[DIR_0MM ])[k1bs ];
- real f1_BN = (D.f[DIR_0PM ])[k1bn ];
- real f1_TS = (D.f[DIR_0MP ])[k1ts ];
+ real f1_E = (D.f[DIR_P00])[k1e ];
+ real f1_W = (D.f[DIR_M00])[k1w ];
+ real f1_N = (D.f[DIR_0P0])[k1n ];
+ real f1_S = (D.f[DIR_0M0])[k1s ];
+ real f1_T = (D.f[DIR_00P])[k1t ];
+ real f1_B = (D.f[DIR_00M])[k1b ];
+ real f1_NE = (D.f[DIR_PP0])[k1ne ];
+ real f1_SW = (D.f[DIR_MM0])[k1sw ];
+ real f1_SE = (D.f[DIR_PM0])[k1se ];
+ real f1_NW = (D.f[DIR_MP0])[k1nw ];
+ real f1_TE = (D.f[DIR_P0P])[k1te ];
+ real f1_BW = (D.f[DIR_M0M])[k1bw ];
+ real f1_BE = (D.f[DIR_P0M])[k1be ];
+ real f1_TW = (D.f[DIR_M0P])[k1tw ];
+ real f1_TN = (D.f[DIR_0PP])[k1tn ];
+ real f1_BS = (D.f[DIR_0MM])[k1bs ];
+ real f1_BN = (D.f[DIR_0PM])[k1bn ];
+ real f1_TS = (D.f[DIR_0MP])[k1ts ];
//real f1_ZERO = (D.f[DIR_000])[k1zero];
- real f1_TNE = (D.f[DIR_PPP ])[k1tne ];
- real f1_TSW = (D.f[DIR_MMP ])[k1tsw ];
- real f1_TSE = (D.f[DIR_PMP ])[k1tse ];
- real f1_TNW = (D.f[DIR_MPP ])[k1tnw ];
- real f1_BNE = (D.f[DIR_PPM ])[k1bne ];
- real f1_BSW = (D.f[DIR_MMM ])[k1bsw ];
- real f1_BSE = (D.f[DIR_PMM ])[k1bse ];
- real f1_BNW = (D.f[DIR_MPM ])[k1bnw ];
+ real f1_TNE = (D.f[DIR_PPP])[k1tne ];
+ real f1_TSW = (D.f[DIR_MMP])[k1tsw ];
+ real f1_TSE = (D.f[DIR_PMP])[k1tse ];
+ real f1_TNW = (D.f[DIR_MPP])[k1tnw ];
+ real f1_BNE = (D.f[DIR_PPM])[k1bne ];
+ real f1_BSW = (D.f[DIR_MMM])[k1bsw ];
+ real f1_BSE = (D.f[DIR_PMM])[k1bse ];
+ real f1_BNW = (D.f[DIR_MPM])[k1bnw ];
//////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
- real f_E = (D.f[DIR_P00 ])[ke ];
- real f_W = (D.f[DIR_M00 ])[kw ];
- real f_N = (D.f[DIR_0P0 ])[kn ];
- real f_S = (D.f[DIR_0M0 ])[ks ];
- real f_T = (D.f[DIR_00P ])[kt ];
- real f_B = (D.f[DIR_00M ])[kb ];
- real f_NE = (D.f[DIR_PP0 ])[kne ];
- real f_SW = (D.f[DIR_MM0 ])[ksw ];
- real f_SE = (D.f[DIR_PM0 ])[kse ];
- real f_NW = (D.f[DIR_MP0 ])[knw ];
- real f_TE = (D.f[DIR_P0P ])[kte ];
- real f_BW = (D.f[DIR_M0M ])[kbw ];
- real f_BE = (D.f[DIR_P0M ])[kbe ];
- real f_TW = (D.f[DIR_M0P ])[ktw ];
- real f_TN = (D.f[DIR_0PP ])[ktn ];
- real f_BS = (D.f[DIR_0MM ])[kbs ];
- real f_BN = (D.f[DIR_0PM ])[kbn ];
- real f_TS = (D.f[DIR_0MP ])[kts ];
+ real f_E = (D.f[DIR_P00])[ke ];
+ real f_W = (D.f[DIR_M00])[kw ];
+ real f_N = (D.f[DIR_0P0])[kn ];
+ real f_S = (D.f[DIR_0M0])[ks ];
+ real f_T = (D.f[DIR_00P])[kt ];
+ real f_B = (D.f[DIR_00M])[kb ];
+ real f_NE = (D.f[DIR_PP0])[kne ];
+ real f_SW = (D.f[DIR_MM0])[ksw ];
+ real f_SE = (D.f[DIR_PM0])[kse ];
+ real f_NW = (D.f[DIR_MP0])[knw ];
+ real f_TE = (D.f[DIR_P0P])[kte ];
+ real f_BW = (D.f[DIR_M0M])[kbw ];
+ real f_BE = (D.f[DIR_P0M])[kbe ];
+ real f_TW = (D.f[DIR_M0P])[ktw ];
+ real f_TN = (D.f[DIR_0PP])[ktn ];
+ real f_BS = (D.f[DIR_0MM])[kbs ];
+ real f_BN = (D.f[DIR_0PM])[kbn ];
+ real f_TS = (D.f[DIR_0MP])[kts ];
//real f_ZERO = (D.f[DIR_000])[kzero];
- real f_TNE = (D.f[DIR_PPP ])[ktne ];
- real f_TSW = (D.f[DIR_MMP ])[ktsw ];
- real f_TSE = (D.f[DIR_PMP ])[ktse ];
- real f_TNW = (D.f[DIR_MPP ])[ktnw ];
- real f_BNE = (D.f[DIR_PPM ])[kbne ];
- real f_BSW = (D.f[DIR_MMM ])[kbsw ];
- real f_BSE = (D.f[DIR_PMM ])[kbse ];
- real f_BNW = (D.f[DIR_MPM ])[kbnw ];
+ real f_TNE = (D.f[DIR_PPP])[ktne ];
+ real f_TSW = (D.f[DIR_MMP])[ktsw ];
+ real f_TSE = (D.f[DIR_PMP])[ktse ];
+ real f_TNW = (D.f[DIR_MPP])[ktnw ];
+ real f_BNE = (D.f[DIR_PPM])[kbne ];
+ real f_BSW = (D.f[DIR_MMM])[kbsw ];
+ real f_BSE = (D.f[DIR_PMM])[kbse ];
+ real f_BNW = (D.f[DIR_MPM])[kbnw ];
//////////////////////////////////////////////////////////////////////////
// real vx1, vx2, vx3;
real drho, drho1;
//////////////////////////////////////////////////////////////////////////
- //Dichte
+ //Dichte
drho1 = f1_TSE + f1_TNW + f1_TNE + f1_TSW + f1_BSE + f1_BNW + f1_BNE + f1_BSW +
- f1_BN + f1_TS + f1_TN + f1_BS + f1_BE + f1_TW + f1_TE + f1_BW + f1_SE + f1_NW + f1_NE + f1_SW +
- f1_T + f1_B + f1_N + f1_S + f1_E + f1_W + ((D.f[DIR_000])[k1zero]);
+ f1_BN + f1_TS + f1_TN + f1_BS + f1_BE + f1_TW + f1_TE + f1_BW + f1_SE + f1_NW + f1_NE + f1_SW +
+ f1_T + f1_B + f1_N + f1_S + f1_E + f1_W + ((D.f[DIR_000])[k1zero]);
drho = f_TSE + f_TNW + f_TNE + f_TSW + f_BSE + f_BNW + f_BNE + f_BSW +
- f_BN + f_TS + f_TN + f_BS + f_BE + f_TW + f_TE + f_BW + f_SE + f_NW + f_NE + f_SW +
- f_T + f_B + f_N + f_S + f_E + f_W + ((D.f[DIR_000])[kzero]);
+ f_BN + f_TS + f_TN + f_BS + f_BE + f_TW + f_TE + f_BW + f_SE + f_NW + f_NE + f_SW +
+ f_T + f_B + f_N + f_S + f_E + f_W + ((D.f[DIR_000])[kzero]);
//////////////////////////////////////////////////////////////////////////
- //Schallgeschwindigkeit
- real cs = c1o1 / sqrtf(c3o1);
+ //Schallgeschwindigkeit
+ real cs = c1o1 / sqrtf(c3o1);
//////////////////////////////////////////////////////////////////////////
- real rhoInterpol = drho1 * cs + (c1o1 - cs) * drho;
- //real diffRho = (rhoBC[k] + one) / (rhoInterpol + one);
- real diffRhoToAdd = rhoBC[k] - rhoInterpol;
- /////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
- //no velocity
- //////////////////////////////////////////
+ real rhoInterpol = drho1 * cs + (c1o1 - cs) * drho;
+ //real diffRho = (rhoBC[k] + one) / (rhoInterpol + one);
+ real diffRhoToAdd = rhoBC[k] - rhoInterpol;
+ /////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
+ //no velocity
+ //////////////////////////////////////////
f_E = f1_E * cs + (c1o1 - cs) * f_E ;
f_W = f1_W * cs + (c1o1 - cs) * f_W ;
f_N = f1_N * cs + (c1o1 - cs) * f_N ;
@@ -258,16 +292,16 @@ __global__ void QInflowScaleByPressDevice27( real* rhoBC,
f_BSW = f1_BSW * cs + (c1o1 - cs) * f_BSW ;
f_BSE = f1_BSE * cs + (c1o1 - cs) * f_BSE ;
f_BNW = f1_BNW * cs + (c1o1 - cs) * f_BNW ;
- /////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
- //scale by press
- //////////////////////////////////////////
- //f_E = (f_E + c2over27 ) * diffRho - c2over27 ;
+ /////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
+ //scale by press
+ //////////////////////////////////////////
+ //f_E = (f_E + c2over27 ) * diffRho - c2over27 ;
// f_W = (f_W + c2over27 ) * diffRho - c2over27 ;
// f_N = (f_N + c2over27 ) * diffRho - c2over27 ;
// f_S = (f_S + c2over27 ) * diffRho - c2over27 ;
// f_T = (f_T + c2over27 ) * diffRho - c2over27 ;
// f_B = (f_B + c2over27 ) * diffRho - c2over27 ;
- //f_NE = (f_NE + c1over54 ) * diffRho - c1over54 ;
+ //f_NE = (f_NE + c1over54 ) * diffRho - c1over54 ;
// f_SW = (f_SW + c1over54 ) * diffRho - c1over54 ;
// f_SE = (f_SE + c1over54 ) * diffRho - c1over54 ;
// f_NW = (f_NW + c1over54 ) * diffRho - c1over54 ;
@@ -287,16 +321,16 @@ __global__ void QInflowScaleByPressDevice27( real* rhoBC,
// f_BSW = (f_BSW + c1over216) * diffRho - c1over216;
// f_BSE = (f_BSE + c1over216) * diffRho - c1over216;
// f_BNW = (f_BNW + c1over216) * diffRho - c1over216;
- /////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
- // add press
- //////////////////////////////////////////
- f_E = (f_E + c2o27 * diffRhoToAdd);
+ /////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
+ // add press
+ //////////////////////////////////////////
+ f_E = (f_E + c2o27 * diffRhoToAdd);
f_W = (f_W + c2o27 * diffRhoToAdd);
f_N = (f_N + c2o27 * diffRhoToAdd);
f_S = (f_S + c2o27 * diffRhoToAdd);
f_T = (f_T + c2o27 * diffRhoToAdd);
f_B = (f_B + c2o27 * diffRhoToAdd);
- f_NE = (f_NE + c1o54 * diffRhoToAdd);
+ f_NE = (f_NE + c1o54 * diffRhoToAdd);
f_SW = (f_SW + c1o54 * diffRhoToAdd);
f_SE = (f_SE + c1o54 * diffRhoToAdd);
f_NW = (f_NW + c1o54 * diffRhoToAdd);
@@ -316,111 +350,111 @@ __global__ void QInflowScaleByPressDevice27( real* rhoBC,
f_BSW = (f_BSW + c1o216 * diffRhoToAdd);
f_BSE = (f_BSE + c1o216 * diffRhoToAdd);
f_BNW = (f_BNW + c1o216 * diffRhoToAdd);
- /////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
+ /////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
- //////////////////////////////////////////////////////////////////////////
+ //////////////////////////////////////////////////////////////////////////
if (isEvenTimestep==false)
{
- D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
- }
+ D.f[DIR_P00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_MPM * numberOfLBnodes];
+ }
else
{
- D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
+ D.f[DIR_M00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_MPM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_PMP * numberOfLBnodes];
}
//////////////////////////////////////////////////////////////////////////
//__syncthreads();
- // -X
- //(D.f[DIR_P00 ])[ke ] = f_E ;
- //(D.f[DIR_PM0 ])[kse ] = f_SE ;
- //(D.f[DIR_PP0 ])[kne ] = f_NE ;
- //(D.f[DIR_P0M ])[kbe ] = f_BE ;
- //(D.f[DIR_P0P ])[kte ] = f_TE ;
- //(D.f[DIR_PMP ])[ktse ] = f_TSE ;
- //(D.f[DIR_PPP ])[ktne ] = f_TNE ;
- //(D.f[DIR_PMM ])[kbse ] = f_BSE ;
- //(D.f[DIR_PPM ])[kbne ] = f_BNE ;
- // X
- (D.f[DIR_M00 ])[kw ] = f_W ;
- (D.f[DIR_MM0 ])[ksw ] = f_SW ;
- (D.f[DIR_MP0 ])[knw ] = f_NW ;
- (D.f[DIR_M0M ])[kbw ] = f_BW ;
- (D.f[DIR_M0P ])[ktw ] = f_TW ;
- (D.f[DIR_MMP ])[ktsw ] = f_TSW ;
- (D.f[DIR_MPP ])[ktnw ] = f_TNW ;
- (D.f[DIR_MMM ])[kbsw ] = f_BSW ;
- (D.f[DIR_MPM ])[kbnw ] = f_BNW ;
- // Y
- //(D.f[DIR_0M0 ])[ks ] = f_S ;
- //(D.f[DIR_PM0 ])[kse ] = f_SE ;
- //(D.f[DIR_MM0 ])[ksw ] = f_SW ;
- //(D.f[DIR_0MP ])[kts ] = f_TS ;
- //(D.f[DIR_0MM ])[kbs ] = f_BS ;
- //(D.f[DIR_PMP ])[ktse ] = f_TSE ;
- //(D.f[DIR_MMP ])[ktsw ] = f_TSW ;
- //(D.f[DIR_PMM ])[kbse ] = f_BSE ;
- //(D.f[DIR_MMM ])[kbsw ] = f_BSW ;
- // Z
- //(D.f[DIR_00M ])[kb ] = f_B ;
- //(D.f[DIR_P0M ])[kbe ] = f_BE ;
- //(D.f[DIR_M0M ])[kbw ] = f_BW ;
- //(D.f[DIR_0PM ])[kbn ] = f_BN ;
- //(D.f[DIR_0MM ])[kbs ] = f_BS ;
- //(D.f[DIR_PPM ])[kbne ] = f_BNE ;
- //(D.f[DIR_MPM ])[kbnw ] = f_BNW ;
- //(D.f[DIR_PMM ])[kbse ] = f_BSE ;
- //(D.f[DIR_MMM ])[kbsw ] = f_BSW ;
+ // -X
+ //(D.f[DIR_P00])[ke ] = f_E ;
+ //(D.f[DIR_PM0])[kse ] = f_SE ;
+ //(D.f[DIR_PP0])[kne ] = f_NE ;
+ //(D.f[DIR_P0M])[kbe ] = f_BE ;
+ //(D.f[DIR_P0P])[kte ] = f_TE ;
+ //(D.f[DIR_PMP])[ktse ] = f_TSE ;
+ //(D.f[DIR_PPP])[ktne ] = f_TNE ;
+ //(D.f[DIR_PMM])[kbse ] = f_BSE ;
+ //(D.f[DIR_PPM])[kbne ] = f_BNE ;
+ // X
+ (D.f[DIR_M00])[kw ] = f_W ;
+ (D.f[DIR_MM0])[ksw ] = f_SW ;
+ (D.f[DIR_MP0])[knw ] = f_NW ;
+ (D.f[DIR_M0M])[kbw ] = f_BW ;
+ (D.f[DIR_M0P])[ktw ] = f_TW ;
+ (D.f[DIR_MMP])[ktsw ] = f_TSW ;
+ (D.f[DIR_MPP])[ktnw ] = f_TNW ;
+ (D.f[DIR_MMM])[kbsw ] = f_BSW ;
+ (D.f[DIR_MPM])[kbnw ] = f_BNW ;
+ // Y
+ //(D.f[DIR_0M0])[ks ] = f_S ;
+ //(D.f[DIR_PM0])[kse ] = f_SE ;
+ //(D.f[DIR_MM0])[ksw ] = f_SW ;
+ //(D.f[DIR_0MP])[kts ] = f_TS ;
+ //(D.f[DIR_0MM])[kbs ] = f_BS ;
+ //(D.f[DIR_PMP])[ktse ] = f_TSE ;
+ //(D.f[DIR_MMP])[ktsw ] = f_TSW ;
+ //(D.f[DIR_PMM])[kbse ] = f_BSE ;
+ //(D.f[DIR_MMM])[kbsw ] = f_BSW ;
+ // Z
+ //(D.f[DIR_00M])[kb ] = f_B ;
+ //(D.f[DIR_P0M])[kbe ] = f_BE ;
+ //(D.f[DIR_M0M])[kbw ] = f_BW ;
+ //(D.f[DIR_0PM])[kbn ] = f_BN ;
+ //(D.f[DIR_0MM])[kbs ] = f_BS ;
+ //(D.f[DIR_PPM])[kbne ] = f_BNE ;
+ //(D.f[DIR_MPM])[kbnw ] = f_BNW ;
+ //(D.f[DIR_PMM])[kbse ] = f_BSE ;
+ //(D.f[DIR_MMM])[kbsw ] = f_BSW ;
//////////////////////////////////////////////////////////////////////////
}
}
@@ -465,22 +499,23 @@ __global__ void QInflowScaleByPressDevice27( real* rhoBC,
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
-__global__ void QPressDeviceIncompNEQ27( real* rhoBC,
- real* DD,
- int* k_Q,
- int* k_N,
- int numberOfBCnodes,
- real om1,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep)
+__global__ void QPressDeviceIncompNEQ27(
+ real* rhoBC,
+ real* DD,
+ int* k_Q,
+ int* k_N,
+ int numberOfBCnodes,
+ real om1,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep)
{
////////////////////////////////////////////////////////////////////////////////
- const unsigned x = threadIdx.x; // Globaler x-Index
- const unsigned y = blockIdx.x; // Globaler y-Index
- const unsigned z = blockIdx.y; // Globaler z-Index
+ const unsigned x = threadIdx.x; // Globaler x-Index
+ const unsigned y = blockIdx.x; // Globaler y-Index
+ const unsigned z = blockIdx.y; // Globaler z-Index
const unsigned nx = blockDim.x;
const unsigned ny = gridDim.x;
@@ -554,112 +589,112 @@ __global__ void QPressDeviceIncompNEQ27( real* rhoBC,
Distributions27 D;
if (isEvenTimestep==true) //// ACHTUNG PREColl !!!!!!!!!!!!!!
{
- D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
- }
+ D.f[DIR_P00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_MPM * numberOfLBnodes];
+ }
else
{
- D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
+ D.f[DIR_M00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_MPM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_PMP * numberOfLBnodes];
}
//////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
real f1_E,f1_W,f1_N,f1_S,f1_T,f1_B,f1_NE,f1_SW,f1_SE,f1_NW,f1_TE,f1_BW,f1_BE,f1_TW,f1_TN,f1_BS,f1_BN,f1_TS,f1_ZERO,
f1_TNE,f1_TSW,f1_TSE,f1_TNW,f1_BNE,f1_BSW,f1_BSE,f1_BNW;
- f1_W = (D.f[DIR_P00 ])[k1e ];
- f1_E = (D.f[DIR_M00 ])[k1w ];
- f1_S = (D.f[DIR_0P0 ])[k1n ];
- f1_N = (D.f[DIR_0M0 ])[k1s ];
- f1_B = (D.f[DIR_00P ])[k1t ];
- f1_T = (D.f[DIR_00M ])[k1b ];
- f1_SW = (D.f[DIR_PP0 ])[k1ne ];
- f1_NE = (D.f[DIR_MM0 ])[k1sw ];
- f1_NW = (D.f[DIR_PM0 ])[k1se ];
- f1_SE = (D.f[DIR_MP0 ])[k1nw ];
- f1_BW = (D.f[DIR_P0P ])[k1te ];
- f1_TE = (D.f[DIR_M0M ])[k1bw ];
- f1_TW = (D.f[DIR_P0M ])[k1be ];
- f1_BE = (D.f[DIR_M0P ])[k1tw ];
- f1_BS = (D.f[DIR_0PP ])[k1tn ];
- f1_TN = (D.f[DIR_0MM ])[k1bs ];
- f1_TS = (D.f[DIR_0PM ])[k1bn ];
- f1_BN = (D.f[DIR_0MP ])[k1ts ];
+ f1_W = (D.f[DIR_P00])[k1e ];
+ f1_E = (D.f[DIR_M00])[k1w ];
+ f1_S = (D.f[DIR_0P0])[k1n ];
+ f1_N = (D.f[DIR_0M0])[k1s ];
+ f1_B = (D.f[DIR_00P])[k1t ];
+ f1_T = (D.f[DIR_00M])[k1b ];
+ f1_SW = (D.f[DIR_PP0])[k1ne ];
+ f1_NE = (D.f[DIR_MM0])[k1sw ];
+ f1_NW = (D.f[DIR_PM0])[k1se ];
+ f1_SE = (D.f[DIR_MP0])[k1nw ];
+ f1_BW = (D.f[DIR_P0P])[k1te ];
+ f1_TE = (D.f[DIR_M0M])[k1bw ];
+ f1_TW = (D.f[DIR_P0M])[k1be ];
+ f1_BE = (D.f[DIR_M0P])[k1tw ];
+ f1_BS = (D.f[DIR_0PP])[k1tn ];
+ f1_TN = (D.f[DIR_0MM])[k1bs ];
+ f1_TS = (D.f[DIR_0PM])[k1bn ];
+ f1_BN = (D.f[DIR_0MP])[k1ts ];
f1_ZERO = (D.f[DIR_000])[k1zero];
- f1_BSW = (D.f[DIR_PPP ])[k1tne ];
- f1_BNE = (D.f[DIR_MMP ])[k1tsw ];
- f1_BNW = (D.f[DIR_PMP ])[k1tse ];
- f1_BSE = (D.f[DIR_MPP ])[k1tnw ];
- f1_TSW = (D.f[DIR_PPM ])[k1bne ];
- f1_TNE = (D.f[DIR_MMM ])[k1bsw ];
- f1_TNW = (D.f[DIR_PMM ])[k1bse ];
- f1_TSE = (D.f[DIR_MPM ])[k1bnw ];
+ f1_BSW = (D.f[DIR_PPP])[k1tne ];
+ f1_BNE = (D.f[DIR_MMP])[k1tsw ];
+ f1_BNW = (D.f[DIR_PMP])[k1tse ];
+ f1_BSE = (D.f[DIR_MPP])[k1tnw ];
+ f1_TSW = (D.f[DIR_PPM])[k1bne ];
+ f1_TNE = (D.f[DIR_MMM])[k1bsw ];
+ f1_TNW = (D.f[DIR_PMM])[k1bse ];
+ f1_TSE = (D.f[DIR_MPM])[k1bnw ];
//////////////////////////////////////////////////////////////////////////
real drho1 = f1_ZERO+f1_E+f1_W+f1_N+f1_S+f1_T+f1_B+f1_NE+f1_SW+f1_SE+f1_NW+f1_TE+f1_BW+f1_BE+f1_TW+f1_TN+f1_BS+f1_BN+f1_TS+
f1_TNE+f1_TSW+f1_TSE+f1_TNW+f1_BNE+f1_BSW+f1_BSE+f1_BNW;
real vx1 = ((f1_TSE - f1_BNW) - (f1_TNW - f1_BSE)) + ((f1_TNE - f1_BSW) - (f1_TSW - f1_BNE)) +
- ((f1_BE - f1_TW) + (f1_TE - f1_BW)) + ((f1_SE - f1_NW) + (f1_NE - f1_SW)) +
- (f1_E - f1_W);
+ ((f1_BE - f1_TW) + (f1_TE - f1_BW)) + ((f1_SE - f1_NW) + (f1_NE - f1_SW)) +
+ (f1_E - f1_W);
real vx2 = (-(f1_TSE - f1_BNW) + (f1_TNW - f1_BSE)) + ((f1_TNE - f1_BSW) - (f1_TSW - f1_BNE)) +
- ((f1_BN - f1_TS) + (f1_TN - f1_BS)) + (-(f1_SE - f1_NW) + (f1_NE - f1_SW)) +
- (f1_N - f1_S);
+ ((f1_BN - f1_TS) + (f1_TN - f1_BS)) + (-(f1_SE - f1_NW) + (f1_NE - f1_SW)) +
+ (f1_N - f1_S);
real vx3 = ((f1_TSE - f1_BNW) + (f1_TNW - f1_BSE)) + ((f1_TNE - f1_BSW) + (f1_TSW - f1_BNE)) +
- (-(f1_BN - f1_TS) + (f1_TN - f1_BS)) + ((f1_TE - f1_BW) - (f1_BE - f1_TW)) +
- (f1_T - f1_B);
+ (-(f1_BN - f1_TS) + (f1_TN - f1_BS)) + ((f1_TE - f1_BW) - (f1_BE - f1_TW)) +
+ (f1_T - f1_B);
real cusq=c3o2*(vx1*vx1+vx2*vx2+vx3*vx3);
@@ -690,15 +725,15 @@ __global__ void QPressDeviceIncompNEQ27( real* rhoBC,
f1_BNW -= c1o216*(drho1+(drho1+c1o1)*(c3o1*(-vx1+vx2-vx3)+c9o2*(-vx1+vx2-vx3)*(-vx1+vx2-vx3)-cusq));
f1_BSE -= c1o216*(drho1+(drho1+c1o1)*(c3o1*( vx1-vx2-vx3)+c9o2*( vx1-vx2-vx3)*( vx1-vx2-vx3)-cusq));
f1_TNW -= c1o216*(drho1+(drho1+c1o1)*(c3o1*(-vx1+vx2+vx3)+c9o2*(-vx1+vx2+vx3)*(-vx1+vx2+vx3)-cusq));
-
- drho1 = rhoBC[k];
- //if(vx1 < zero){
- // vx1 *= 0.9;
- //}
- //if(vx2 < zero){
- // vx2 *= c1o10;//0.9;
- //}
+ drho1 = rhoBC[k];
+
+ //if(vx1 < zero){
+ // vx1 *= 0.9;
+ //}
+ //if(vx2 < zero){
+ // vx2 *= c1o10;//0.9;
+ //}
f1_ZERO += c8o27* (drho1-(drho1+c1o1)*cusq);
f1_E += c2o27* (drho1+(drho1+c1o1)*(c3o1*( vx1 )+c9o2*( vx1 )*( vx1 )-cusq));
@@ -728,39 +763,39 @@ __global__ void QPressDeviceIncompNEQ27( real* rhoBC,
f1_BSE += c1o216*(drho1+(drho1+c1o1)*(c3o1*( vx1-vx2-vx3)+c9o2*( vx1-vx2-vx3)*( vx1-vx2-vx3)-cusq));
f1_TNW += c1o216*(drho1+(drho1+c1o1)*(c3o1*(-vx1+vx2+vx3)+c9o2*(-vx1+vx2+vx3)*(-vx1+vx2+vx3)-cusq));
- //drho1 = (drho1 + rhoBC[k])/2.f;
- //drho1 = drho1 - rhoBC[k];
+ //drho1 = (drho1 + rhoBC[k])/2.f;
+ //drho1 = drho1 - rhoBC[k];
//////////////////////////////////////////////////////////////////////////
__syncthreads();
- (D.f[DIR_P00 ])[ke ] = f1_W ;
- (D.f[DIR_M00 ])[kw ] = f1_E ;
- (D.f[DIR_0P0 ])[kn ] = f1_S ;
- (D.f[DIR_0M0 ])[ks ] = f1_N ;
- (D.f[DIR_00P ])[kt ] = f1_B ;
- (D.f[DIR_00M ])[kb ] = f1_T ;
- (D.f[DIR_PP0 ])[kne ] = f1_SW ;
- (D.f[DIR_MM0 ])[ksw ] = f1_NE ;
- (D.f[DIR_PM0 ])[kse ] = f1_NW ;
- (D.f[DIR_MP0 ])[knw ] = f1_SE ;
- (D.f[DIR_P0P ])[kte ] = f1_BW ;
- (D.f[DIR_M0M ])[kbw ] = f1_TE ;
- (D.f[DIR_P0M ])[kbe ] = f1_TW ;
- (D.f[DIR_M0P ])[ktw ] = f1_BE ;
- (D.f[DIR_0PP ])[ktn ] = f1_BS ;
- (D.f[DIR_0MM ])[kbs ] = f1_TN ;
- (D.f[DIR_0PM ])[kbn ] = f1_TS ;
- (D.f[DIR_0MP ])[kts ] = f1_BN ;
- (D.f[DIR_000])[kzero] = f1_ZERO;
- (D.f[DIR_PPP ])[ktne ] = f1_BSW ;
- (D.f[DIR_MMP ])[ktsw ] = f1_BNE ;
- (D.f[DIR_PMP ])[ktse ] = f1_BNW ;
- (D.f[DIR_MPP ])[ktnw ] = f1_BSE ;
- (D.f[DIR_PPM ])[kbne ] = f1_TSW ;
- (D.f[DIR_MMM ])[kbsw ] = f1_TNE ;
- (D.f[DIR_PMM ])[kbse ] = f1_TNW ;
- (D.f[DIR_MPM ])[kbnw ] = f1_TSE ;
+ (D.f[DIR_P00])[ke ] = f1_W ;
+ (D.f[DIR_M00])[kw ] = f1_E ;
+ (D.f[DIR_0P0])[kn ] = f1_S ;
+ (D.f[DIR_0M0])[ks ] = f1_N ;
+ (D.f[DIR_00P])[kt ] = f1_B ;
+ (D.f[DIR_00M])[kb ] = f1_T ;
+ (D.f[DIR_PP0])[kne ] = f1_SW ;
+ (D.f[DIR_MM0])[ksw ] = f1_NE ;
+ (D.f[DIR_PM0])[kse ] = f1_NW ;
+ (D.f[DIR_MP0])[knw ] = f1_SE ;
+ (D.f[DIR_P0P])[kte ] = f1_BW ;
+ (D.f[DIR_M0M])[kbw ] = f1_TE ;
+ (D.f[DIR_P0M])[kbe ] = f1_TW ;
+ (D.f[DIR_M0P])[ktw ] = f1_BE ;
+ (D.f[DIR_0PP])[ktn ] = f1_BS ;
+ (D.f[DIR_0MM])[kbs ] = f1_TN ;
+ (D.f[DIR_0PM])[kbn ] = f1_TS ;
+ (D.f[DIR_0MP])[kts ] = f1_BN ;
+ (D.f[DIR_000])[kzero] = f1_ZERO;
+ (D.f[DIR_PPP])[ktne ] = f1_BSW ;
+ (D.f[DIR_MMP])[ktsw ] = f1_BNE ;
+ (D.f[DIR_PMP])[ktse ] = f1_BNW ;
+ (D.f[DIR_MPP])[ktnw ] = f1_BSE ;
+ (D.f[DIR_PPM])[kbne ] = f1_TSW ;
+ (D.f[DIR_MMM])[kbsw ] = f1_TNE ;
+ (D.f[DIR_PMM])[kbse ] = f1_TNW ;
+ (D.f[DIR_MPM])[kbnw ] = f1_TSE ;
}
}
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
@@ -804,54 +839,49 @@ __global__ void QPressDeviceIncompNEQ27( real* rhoBC,
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
-__global__ void QPressDeviceNEQ27(real* rhoBC,
- real* distribution,
- int* bcNodeIndices,
- int* bcNeighborIndices,
- int numberOfBCnodes,
- real omega1,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int numberOfLBnodes,
- bool isEvenTimestep)
+__global__ void QPressDeviceNEQ27(
+ real* rhoBC,
+ real* distributions,
+ int* bcNodeIndices,
+ int* bcNeighborIndices,
+ int numberOfBCnodes,
+ real omega1,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep)
{
- //////////////////////////////////////////////////////////////////////////
- //! The pressure boundary condition is executed in the following steps
- //!
- ////////////////////////////////////////////////////////////////////////////////
- //! - Get node index coordinates from threadIdx, blockIdx, blockDim and gridDim.
- //!
- const unsigned x = threadIdx.x; // global x-index
- const unsigned y = blockIdx.x; // global y-index
- const unsigned z = blockIdx.y; // global z-index
-
- const unsigned nx = blockDim.x;
- const unsigned ny = gridDim.x;
+ ////////////////////////////////////////////////////////////////////////////////
+ //! The pressure boundary condition is executed in the following steps
+ //!
- const unsigned k = nx*(ny*z + y) + x;
+ ////////////////////////////////////////////////////////////////////////////////
+ //! - Get node index coordinates from threadIdx, blockIdx, blockDim and gridDim.
+ //!
+ const unsigned nodeIndex = getNodeIndex();
- //////////////////////////////////////////////////////////////////////////
+ ////////////////////////////////////////////////////////////////////////////////
//! - Run for all indices in size of boundary condition (numberOfBCnodes)
//!
- if(k < numberOfBCnodes)
+ if(nodeIndex < numberOfBCnodes)
{
//////////////////////////////////////////////////////////////////////////
//! - Read distributions: style of reading and writing the distributions from/to stored arrays dependent on timestep is based on the esoteric twist algorithm \ref
//! <a href="https://doi.org/10.3390/computation5020019"><b>[ M. Geier et al. (2017), DOI:10.3390/computation5020019 ]</b></a>
//!
Distributions27 dist;
- getPointersToDistributions(dist, distribution, numberOfLBnodes, isEvenTimestep);
+ getPointersToDistributions(dist, distributions, numberOfLBnodes, isEvenTimestep);
////////////////////////////////////////////////////////////////////////////////
//! - Set local pressure
//!
- real rhoBClocal = rhoBC[k];
+ real rhoBClocal = rhoBC[nodeIndex];
////////////////////////////////////////////////////////////////////////////////
//! - Set neighbor indices (necessary for indirect addressing)
//!
- unsigned int KQK = bcNodeIndices[k];
+ unsigned int KQK = bcNodeIndices[nodeIndex];
unsigned int kzero= KQK;
unsigned int ke = KQK;
unsigned int kw = neighborX[KQK];
@@ -882,7 +912,7 @@ __global__ void QPressDeviceNEQ27(real* rhoBC,
////////////////////////////////////////////////////////////////////////////////
//! - Set neighbor indices (necessary for indirect addressing) for neighboring node
//!
- unsigned int K1QK = bcNeighborIndices[k];
+ unsigned int K1QK = bcNeighborIndices[nodeIndex];
unsigned int k1zero= K1QK;
unsigned int k1e = K1QK;
unsigned int k1w = neighborX[K1QK];
@@ -914,52 +944,52 @@ __global__ void QPressDeviceNEQ27(real* rhoBC,
////////////////////////////////////////////////////////////////////////////////
//! - Set local distributions for neighboring node
//!
- real f1_W = (dist.f[DIR_P00 ])[k1e ];
- real f1_E = (dist.f[DIR_M00 ])[k1w ];
- real f1_S = (dist.f[DIR_0P0 ])[k1n ];
- real f1_N = (dist.f[DIR_0M0 ])[k1s ];
- real f1_B = (dist.f[DIR_00P ])[k1t ];
- real f1_T = (dist.f[DIR_00M ])[k1b ];
- real f1_SW = (dist.f[DIR_PP0 ])[k1ne ];
- real f1_NE = (dist.f[DIR_MM0 ])[k1sw ];
- real f1_NW = (dist.f[DIR_PM0 ])[k1se ];
- real f1_SE = (dist.f[DIR_MP0 ])[k1nw ];
- real f1_BW = (dist.f[DIR_P0P ])[k1te ];
- real f1_TE = (dist.f[DIR_M0M ])[k1bw ];
- real f1_TW = (dist.f[DIR_P0M ])[k1be ];
- real f1_BE = (dist.f[DIR_M0P ])[k1tw ];
- real f1_BS = (dist.f[DIR_0PP ])[k1tn ];
- real f1_TN = (dist.f[DIR_0MM ])[k1bs ];
- real f1_TS = (dist.f[DIR_0PM ])[k1bn ];
- real f1_BN = (dist.f[DIR_0MP ])[k1ts ];
+ real f1_W = (dist.f[DIR_P00])[k1e ];
+ real f1_E = (dist.f[DIR_M00])[k1w ];
+ real f1_S = (dist.f[DIR_0P0])[k1n ];
+ real f1_N = (dist.f[DIR_0M0])[k1s ];
+ real f1_B = (dist.f[DIR_00P])[k1t ];
+ real f1_T = (dist.f[DIR_00M])[k1b ];
+ real f1_SW = (dist.f[DIR_PP0])[k1ne ];
+ real f1_NE = (dist.f[DIR_MM0])[k1sw ];
+ real f1_NW = (dist.f[DIR_PM0])[k1se ];
+ real f1_SE = (dist.f[DIR_MP0])[k1nw ];
+ real f1_BW = (dist.f[DIR_P0P])[k1te ];
+ real f1_TE = (dist.f[DIR_M0M])[k1bw ];
+ real f1_TW = (dist.f[DIR_P0M])[k1be ];
+ real f1_BE = (dist.f[DIR_M0P])[k1tw ];
+ real f1_BS = (dist.f[DIR_0PP])[k1tn ];
+ real f1_TN = (dist.f[DIR_0MM])[k1bs ];
+ real f1_TS = (dist.f[DIR_0PM])[k1bn ];
+ real f1_BN = (dist.f[DIR_0MP])[k1ts ];
real f1_ZERO = (dist.f[DIR_000])[k1zero];
- real f1_BSW = (dist.f[DIR_PPP ])[k1tne ];
- real f1_BNE = (dist.f[DIR_MMP ])[k1tsw ];
- real f1_BNW = (dist.f[DIR_PMP ])[k1tse ];
- real f1_BSE = (dist.f[DIR_MPP ])[k1tnw ];
- real f1_TSW = (dist.f[DIR_PPM ])[k1bne ];
- real f1_TNE = (dist.f[DIR_MMM ])[k1bsw ];
- real f1_TNW = (dist.f[DIR_PMM ])[k1bse ];
- real f1_TSE = (dist.f[DIR_MPM ])[k1bnw ];
+ real f1_BSW = (dist.f[DIR_PPP])[k1tne ];
+ real f1_BNE = (dist.f[DIR_MMP])[k1tsw ];
+ real f1_BNW = (dist.f[DIR_PMP])[k1tse ];
+ real f1_BSE = (dist.f[DIR_MPP])[k1tnw ];
+ real f1_TSW = (dist.f[DIR_PPM])[k1bne ];
+ real f1_TNE = (dist.f[DIR_MMM])[k1bsw ];
+ real f1_TNW = (dist.f[DIR_PMM])[k1bse ];
+ real f1_TSE = (dist.f[DIR_MPM])[k1bnw ];
////////////////////////////////////////////////////////////////////////////////
//! - Calculate macroscopic quantities (for neighboring node)
//!
real drho1 = f1_TSE + f1_TNW + f1_TNE + f1_TSW + f1_BSE + f1_BNW + f1_BNE + f1_BSW +
- f1_BN + f1_TS + f1_TN + f1_BS + f1_BE + f1_TW + f1_TE + f1_BW + f1_SE + f1_NW + f1_NE + f1_SW +
- f1_T + f1_B + f1_N + f1_S + f1_E + f1_W + ((dist.f[DIR_000])[kzero]);
+ f1_BN + f1_TS + f1_TN + f1_BS + f1_BE + f1_TW + f1_TE + f1_BW + f1_SE + f1_NW + f1_NE + f1_SW +
+ f1_T + f1_B + f1_N + f1_S + f1_E + f1_W + ((dist.f[DIR_000])[kzero]);
real vx1 = (((f1_TSE - f1_BNW) - (f1_TNW - f1_BSE)) + ((f1_TNE - f1_BSW) - (f1_TSW - f1_BNE)) +
((f1_BE - f1_TW) + (f1_TE - f1_BW)) + ((f1_SE - f1_NW) + (f1_NE - f1_SW)) +
- (f1_E - f1_W)) / (c1o1 + drho1);
+ (f1_E - f1_W)) / (c1o1 + drho1);
real vx2 = ((-(f1_TSE - f1_BNW) + (f1_TNW - f1_BSE)) + ((f1_TNE - f1_BSW) - (f1_TSW - f1_BNE)) +
((f1_BN - f1_TS) + (f1_TN - f1_BS)) + (-(f1_SE - f1_NW) + (f1_NE - f1_SW)) +
- (f1_N - f1_S)) / (c1o1 + drho1);
+ (f1_N - f1_S)) / (c1o1 + drho1);
real vx3 = (((f1_TSE - f1_BNW) + (f1_TNW - f1_BSE)) + ((f1_TNE - f1_BSW) + (f1_TSW - f1_BNE)) +
(-(f1_BN - f1_TS) + (f1_TN - f1_BS)) + ((f1_TE - f1_BW) - (f1_BE - f1_TW)) +
- (f1_T - f1_B)) / (c1o1 + drho1);
+ (f1_T - f1_B)) / (c1o1 + drho1);
real cusq = c3o2 * (vx1 * vx1 + vx2 * vx2 + vx3 * vx3);
@@ -1037,33 +1067,33 @@ __global__ void QPressDeviceNEQ27(real* rhoBC,
////////////////////////////////////////////////////////////////////////////////
//! write the new distributions to the bc nodes
//!
- (dist.f[DIR_P00 ])[ke ] = f1_W ;
- (dist.f[DIR_M00 ])[kw ] = f1_E ;
- (dist.f[DIR_0P0 ])[kn ] = f1_S ;
- (dist.f[DIR_0M0 ])[ks ] = f1_N ;
- (dist.f[DIR_00P ])[kt ] = f1_B ;
- (dist.f[DIR_00M ])[kb ] = f1_T ;
- (dist.f[DIR_PP0 ])[kne ] = f1_SW ;
- (dist.f[DIR_MM0 ])[ksw ] = f1_NE ;
- (dist.f[DIR_PM0 ])[kse ] = f1_NW ;
- (dist.f[DIR_MP0 ])[knw ] = f1_SE ;
- (dist.f[DIR_P0P ])[kte ] = f1_BW ;
- (dist.f[DIR_M0M ])[kbw ] = f1_TE ;
- (dist.f[DIR_P0M ])[kbe ] = f1_TW ;
- (dist.f[DIR_M0P ])[ktw ] = f1_BE ;
- (dist.f[DIR_0PP ])[ktn ] = f1_BS ;
- (dist.f[DIR_0MM ])[kbs ] = f1_TN ;
- (dist.f[DIR_0PM ])[kbn ] = f1_TS ;
- (dist.f[DIR_0MP ])[kts ] = f1_BN ;
+ (dist.f[DIR_P00])[ke ] = f1_W ;
+ (dist.f[DIR_M00])[kw ] = f1_E ;
+ (dist.f[DIR_0P0])[kn ] = f1_S ;
+ (dist.f[DIR_0M0])[ks ] = f1_N ;
+ (dist.f[DIR_00P])[kt ] = f1_B ;
+ (dist.f[DIR_00M])[kb ] = f1_T ;
+ (dist.f[DIR_PP0])[kne ] = f1_SW ;
+ (dist.f[DIR_MM0])[ksw ] = f1_NE ;
+ (dist.f[DIR_PM0])[kse ] = f1_NW ;
+ (dist.f[DIR_MP0])[knw ] = f1_SE ;
+ (dist.f[DIR_P0P])[kte ] = f1_BW ;
+ (dist.f[DIR_M0M])[kbw ] = f1_TE ;
+ (dist.f[DIR_P0M])[kbe ] = f1_TW ;
+ (dist.f[DIR_M0P])[ktw ] = f1_BE ;
+ (dist.f[DIR_0PP])[ktn ] = f1_BS ;
+ (dist.f[DIR_0MM])[kbs ] = f1_TN ;
+ (dist.f[DIR_0PM])[kbn ] = f1_TS ;
+ (dist.f[DIR_0MP])[kts ] = f1_BN ;
(dist.f[DIR_000])[kzero] = f1_ZERO;
- (dist.f[DIR_PPP ])[ktne ] = f1_BSW ;
- (dist.f[DIR_MMP ])[ktsw ] = f1_BNE ;
- (dist.f[DIR_PMP ])[ktse ] = f1_BNW ;
- (dist.f[DIR_MPP ])[ktnw ] = f1_BSE ;
- (dist.f[DIR_PPM ])[kbne ] = f1_TSW ;
- (dist.f[DIR_MMM ])[kbsw ] = f1_TNE ;
- (dist.f[DIR_PMM ])[kbse ] = f1_TNW ;
- (dist.f[DIR_MPM ])[kbnw ] = f1_TSE ;
+ (dist.f[DIR_PPP])[ktne ] = f1_BSW ;
+ (dist.f[DIR_MMP])[ktsw ] = f1_BNE ;
+ (dist.f[DIR_PMP])[ktse ] = f1_BNW ;
+ (dist.f[DIR_MPP])[ktnw ] = f1_BSE ;
+ (dist.f[DIR_PPM])[kbne ] = f1_TSW ;
+ (dist.f[DIR_MMM])[kbsw ] = f1_TNE ;
+ (dist.f[DIR_PMM])[kbse ] = f1_TNW ;
+ (dist.f[DIR_MPM])[kbnw ] = f1_TSE ;
}
}
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
@@ -1107,16 +1137,17 @@ __global__ void QPressDeviceNEQ27(real* rhoBC,
////////////////////////////////////////////////////////////////////////////////
-__global__ void LB_BC_Press_East27( int nx,
- int ny,
- int tz,
- unsigned int* bcMatD,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- real* DD,
- unsigned int size_Mat,
- bool isEvenTimestep)
+__global__ void LB_BC_Press_East27(
+ int nx,
+ int ny,
+ int tz,
+ unsigned int* bcMatD,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ real* DD,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep)
{
//thread-index
int ty = blockIdx.x;
@@ -1124,9 +1155,9 @@ __global__ void LB_BC_Press_East27( int nx,
int k, k1, nxny; // Zugriff auf arrays im device
- int x = tx + STARTOFFX; // Globaler x-Index
- int y = ty + STARTOFFY; // Globaler y-Index
- int z = tz + STARTOFFZ; // Globaler z-Index
+ int x = tx + STARTOFFX; // Globaler x-Index
+ int y = ty + STARTOFFY; // Globaler y-Index
+ int z = tz + STARTOFFZ; // Globaler z-Index
k = nx*(ny*z + y) + x;
nxny = nx*ny;
@@ -1137,63 +1168,63 @@ __global__ void LB_BC_Press_East27( int nx,
Distributions27 D;
if (isEvenTimestep==true)
{
- D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
- }
+ D.f[DIR_P00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_MPM * numberOfLBnodes];
+ }
else
{
- D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
+ D.f[DIR_M00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_MPM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_PMP * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////
////////////////////////////////////////////////////////////////////////////////
@@ -1312,69 +1343,69 @@ __global__ void LB_BC_Press_East27( int nx,
real f1_E,f1_W,f1_N,f1_S,f1_T,f1_B,f1_NE,f1_SW,f1_SE,f1_NW,f1_TE,f1_BW,f1_BE,f1_TW,f1_TN,f1_BS,f1_BN,f1_TS,f1_ZERO,
f1_TNE,f1_TSW,f1_TSE,f1_TNW,f1_BNE,f1_BSW,f1_BSE,f1_BNW;
- f1_W = (D.f[DIR_P00 ])[k1e ];
- f1_E = (D.f[DIR_M00 ])[k1w ];
- f1_S = (D.f[DIR_0P0 ])[k1n ];
- f1_N = (D.f[DIR_0M0 ])[k1s ];
- f1_B = (D.f[DIR_00P ])[k1t ];
- f1_T = (D.f[DIR_00M ])[k1b ];
- f1_SW = (D.f[DIR_PP0 ])[k1ne ];
- f1_NE = (D.f[DIR_MM0 ])[k1sw ];
- f1_NW = (D.f[DIR_PM0 ])[k1se ];
- f1_SE = (D.f[DIR_MP0 ])[k1nw ];
- f1_BW = (D.f[DIR_P0P ])[k1te ];
- f1_TE = (D.f[DIR_M0M ])[k1bw ];
- f1_TW = (D.f[DIR_P0M ])[k1be ];
- f1_BE = (D.f[DIR_M0P ])[k1tw ];
- f1_BS = (D.f[DIR_0PP ])[k1tn ];
- f1_TN = (D.f[DIR_0MM ])[k1bs ];
- f1_TS = (D.f[DIR_0PM ])[k1bn ];
- f1_BN = (D.f[DIR_0MP ])[k1ts ];
+ f1_W = (D.f[DIR_P00])[k1e ];
+ f1_E = (D.f[DIR_M00])[k1w ];
+ f1_S = (D.f[DIR_0P0])[k1n ];
+ f1_N = (D.f[DIR_0M0])[k1s ];
+ f1_B = (D.f[DIR_00P])[k1t ];
+ f1_T = (D.f[DIR_00M])[k1b ];
+ f1_SW = (D.f[DIR_PP0])[k1ne ];
+ f1_NE = (D.f[DIR_MM0])[k1sw ];
+ f1_NW = (D.f[DIR_PM0])[k1se ];
+ f1_SE = (D.f[DIR_MP0])[k1nw ];
+ f1_BW = (D.f[DIR_P0P])[k1te ];
+ f1_TE = (D.f[DIR_M0M])[k1bw ];
+ f1_TW = (D.f[DIR_P0M])[k1be ];
+ f1_BE = (D.f[DIR_M0P])[k1tw ];
+ f1_BS = (D.f[DIR_0PP])[k1tn ];
+ f1_TN = (D.f[DIR_0MM])[k1bs ];
+ f1_TS = (D.f[DIR_0PM])[k1bn ];
+ f1_BN = (D.f[DIR_0MP])[k1ts ];
f1_ZERO = (D.f[DIR_000])[k1zero];
- f1_BSW = (D.f[DIR_PPP ])[k1tne ];
- f1_BNE = (D.f[DIR_MMP ])[k1tsw ];
- f1_BNW = (D.f[DIR_PMP ])[k1tse ];
- f1_BSE = (D.f[DIR_MPP ])[k1tnw ];
- f1_TSW = (D.f[DIR_PPM ])[k1bne ];
- f1_TNE = (D.f[DIR_MMM ])[k1bsw ];
- f1_TNW = (D.f[DIR_PMM ])[k1bse ];
- f1_TSE = (D.f[DIR_MPM ])[k1bnw ];
+ f1_BSW = (D.f[DIR_PPP])[k1tne ];
+ f1_BNE = (D.f[DIR_MMP])[k1tsw ];
+ f1_BNW = (D.f[DIR_PMP])[k1tse ];
+ f1_BSE = (D.f[DIR_MPP])[k1tnw ];
+ f1_TSW = (D.f[DIR_PPM])[k1bne ];
+ f1_TNE = (D.f[DIR_MMM])[k1bsw ];
+ f1_TNW = (D.f[DIR_PMM])[k1bse ];
+ f1_TSE = (D.f[DIR_MPM])[k1bnw ];
real drho1 = f1_ZERO+f1_E+f1_W+f1_N+f1_S+f1_T+f1_B+f1_NE+f1_SW+f1_SE+f1_NW+f1_TE+f1_BW+f1_BE+f1_TW+f1_TN+f1_BS+f1_BN+f1_TS+
f1_TNE+f1_TSW+f1_TSE+f1_TNW+f1_BNE+f1_BSW+f1_BSE+f1_BNW;
__syncthreads();
- (D.f[DIR_P00 ])[ke ] = f1_W -c2o27*drho1;
- (D.f[DIR_M00 ])[kw ] = f1_E -c2o27*drho1;
- (D.f[DIR_0P0 ])[kn ] = f1_S -c2o27*drho1;
- (D.f[DIR_0M0 ])[ks ] = f1_N -c2o27*drho1;
- (D.f[DIR_00P ])[kt ] = f1_B -c2o27*drho1;
- (D.f[DIR_00M ])[kb ] = f1_T -c2o27*drho1;
- (D.f[DIR_PP0 ])[kne ] = f1_SW -c1o54*drho1;
- (D.f[DIR_MM0 ])[ksw ] = f1_NE -c1o54*drho1;
- (D.f[DIR_PM0 ])[kse ] = f1_NW -c1o54*drho1;
- (D.f[DIR_MP0 ])[knw ] = f1_SE -c1o54*drho1;
- (D.f[DIR_P0P ])[kte ] = f1_BW -c1o54*drho1;
- (D.f[DIR_M0M ])[kbw ] = f1_TE -c1o54*drho1;
- (D.f[DIR_P0M ])[kbe ] = f1_TW -c1o54*drho1;
- (D.f[DIR_M0P ])[ktw ] = f1_BE -c1o54*drho1;
- (D.f[DIR_0PP ])[ktn ] = f1_BS -c1o54*drho1;
- (D.f[DIR_0MM ])[kbs ] = f1_TN -c1o54*drho1;
- (D.f[DIR_0PM ])[kbn ] = f1_TS -c1o54*drho1;
- (D.f[DIR_0MP ])[kts ] = f1_BN -c1o54*drho1;
+ (D.f[DIR_P00])[ke ] = f1_W -c2o27*drho1;
+ (D.f[DIR_M00])[kw ] = f1_E -c2o27*drho1;
+ (D.f[DIR_0P0])[kn ] = f1_S -c2o27*drho1;
+ (D.f[DIR_0M0])[ks ] = f1_N -c2o27*drho1;
+ (D.f[DIR_00P])[kt ] = f1_B -c2o27*drho1;
+ (D.f[DIR_00M])[kb ] = f1_T -c2o27*drho1;
+ (D.f[DIR_PP0])[kne ] = f1_SW -c1o54*drho1;
+ (D.f[DIR_MM0])[ksw ] = f1_NE -c1o54*drho1;
+ (D.f[DIR_PM0])[kse ] = f1_NW -c1o54*drho1;
+ (D.f[DIR_MP0])[knw ] = f1_SE -c1o54*drho1;
+ (D.f[DIR_P0P])[kte ] = f1_BW -c1o54*drho1;
+ (D.f[DIR_M0M])[kbw ] = f1_TE -c1o54*drho1;
+ (D.f[DIR_P0M])[kbe ] = f1_TW -c1o54*drho1;
+ (D.f[DIR_M0P])[ktw ] = f1_BE -c1o54*drho1;
+ (D.f[DIR_0PP])[ktn ] = f1_BS -c1o54*drho1;
+ (D.f[DIR_0MM])[kbs ] = f1_TN -c1o54*drho1;
+ (D.f[DIR_0PM])[kbn ] = f1_TS -c1o54*drho1;
+ (D.f[DIR_0MP])[kts ] = f1_BN -c1o54*drho1;
(D.f[DIR_000])[kzero] = f1_ZERO-c8o27*drho1;
- (D.f[DIR_PPP ])[ktne ] = f1_BSW -c1o216*drho1;
- (D.f[DIR_MMP ])[ktsw ] = f1_BNE -c1o216*drho1;
- (D.f[DIR_PMP ])[ktse ] = f1_BNW -c1o216*drho1;
- (D.f[DIR_MPP ])[ktnw ] = f1_BSE -c1o216*drho1;
- (D.f[DIR_PPM ])[kbne ] = f1_TSW -c1o216*drho1;
- (D.f[DIR_MMM ])[kbsw ] = f1_TNE -c1o216*drho1;
- (D.f[DIR_PMM ])[kbse ] = f1_TNW -c1o216*drho1;
- (D.f[DIR_MPM ])[kbnw ] = f1_TSE -c1o216*drho1;
+ (D.f[DIR_PPP])[ktne ] = f1_BSW -c1o216*drho1;
+ (D.f[DIR_MMP])[ktsw ] = f1_BNE -c1o216*drho1;
+ (D.f[DIR_PMP])[ktse ] = f1_BNW -c1o216*drho1;
+ (D.f[DIR_MPP])[ktnw ] = f1_BSE -c1o216*drho1;
+ (D.f[DIR_PPM])[kbne ] = f1_TSW -c1o216*drho1;
+ (D.f[DIR_MMM])[kbsw ] = f1_TNE -c1o216*drho1;
+ (D.f[DIR_PMM])[kbse ] = f1_TNW -c1o216*drho1;
+ (D.f[DIR_MPM])[kbnw ] = f1_TSE -c1o216*drho1;
}
__syncthreads();
-}
+}
//////////////////////////////////////////////////////////////////////////////
@@ -1416,83 +1447,84 @@ __global__ void LB_BC_Press_East27( int nx,
//////////////////////////////////////////////////////////////////////////////
-__global__ void QPressDevice27(real* rhoBC,
- real* DD,
- int* k_Q,
- real* QQ,
- unsigned int numberOfBCnodes,
- real om1,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep)
+__global__ void QPressDevice27(
+ real* rhoBC,
+ real* DD,
+ int* k_Q,
+ real* QQ,
+ unsigned int numberOfBCnodes,
+ real om1,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep)
{
Distributions27 D;
if (isEvenTimestep==true)
{
- D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
- }
+ D.f[DIR_P00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_MPM * numberOfLBnodes];
+ }
else
{
- D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
+ D.f[DIR_M00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_MPM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_PMP * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////
- const unsigned x = threadIdx.x; // Globaler x-Index
- const unsigned y = blockIdx.x; // Globaler y-Index
- const unsigned z = blockIdx.y; // Globaler z-Index
+ const unsigned x = threadIdx.x; // Globaler x-Index
+ const unsigned y = blockIdx.x; // Globaler y-Index
+ const unsigned z = blockIdx.y; // Globaler z-Index
const unsigned nx = blockDim.x;
const unsigned ny = gridDim.x;
@@ -1502,29 +1534,29 @@ __global__ void QPressDevice27(real* rhoBC,
if(k<numberOfBCnodes)
{
- real *q_dirE, *q_dirW, *q_dirN, *q_dirS, *q_dirT, *q_dirB,
+ real *q_dirE, *q_dirW, *q_dirN, *q_dirS, *q_dirT, *q_dirB,
*q_dirNE, *q_dirSW, *q_dirSE, *q_dirNW, *q_dirTE, *q_dirBW,
*q_dirBE, *q_dirTW, *q_dirTN, *q_dirBS, *q_dirBN, *q_dirTS,
*q_dirTNE, *q_dirTSW, *q_dirTSE, *q_dirTNW, *q_dirBNE, *q_dirBSW,
- *q_dirBSE, *q_dirBNW;
- q_dirE = &QQ[DIR_P00 * numberOfBCnodes];
- q_dirW = &QQ[DIR_M00 * numberOfBCnodes];
- q_dirN = &QQ[DIR_0P0 * numberOfBCnodes];
- q_dirS = &QQ[DIR_0M0 * numberOfBCnodes];
- q_dirT = &QQ[DIR_00P * numberOfBCnodes];
- q_dirB = &QQ[DIR_00M * numberOfBCnodes];
- q_dirNE = &QQ[DIR_PP0 * numberOfBCnodes];
- q_dirSW = &QQ[DIR_MM0 * numberOfBCnodes];
- q_dirSE = &QQ[DIR_PM0 * numberOfBCnodes];
- q_dirNW = &QQ[DIR_MP0 * numberOfBCnodes];
- q_dirTE = &QQ[DIR_P0P * numberOfBCnodes];
- q_dirBW = &QQ[DIR_M0M * numberOfBCnodes];
- q_dirBE = &QQ[DIR_P0M * numberOfBCnodes];
- q_dirTW = &QQ[DIR_M0P * numberOfBCnodes];
- q_dirTN = &QQ[DIR_0PP * numberOfBCnodes];
- q_dirBS = &QQ[DIR_0MM * numberOfBCnodes];
- q_dirBN = &QQ[DIR_0PM * numberOfBCnodes];
- q_dirTS = &QQ[DIR_0MP * numberOfBCnodes];
+ *q_dirBSE, *q_dirBNW;
+ q_dirE = &QQ[DIR_P00 * numberOfBCnodes];
+ q_dirW = &QQ[DIR_M00 * numberOfBCnodes];
+ q_dirN = &QQ[DIR_0P0 * numberOfBCnodes];
+ q_dirS = &QQ[DIR_0M0 * numberOfBCnodes];
+ q_dirT = &QQ[DIR_00P * numberOfBCnodes];
+ q_dirB = &QQ[DIR_00M * numberOfBCnodes];
+ q_dirNE = &QQ[DIR_PP0 * numberOfBCnodes];
+ q_dirSW = &QQ[DIR_MM0 * numberOfBCnodes];
+ q_dirSE = &QQ[DIR_PM0 * numberOfBCnodes];
+ q_dirNW = &QQ[DIR_MP0 * numberOfBCnodes];
+ q_dirTE = &QQ[DIR_P0P * numberOfBCnodes];
+ q_dirBW = &QQ[DIR_M0M * numberOfBCnodes];
+ q_dirBE = &QQ[DIR_P0M * numberOfBCnodes];
+ q_dirTW = &QQ[DIR_M0P * numberOfBCnodes];
+ q_dirTN = &QQ[DIR_0PP * numberOfBCnodes];
+ q_dirBS = &QQ[DIR_0MM * numberOfBCnodes];
+ q_dirBN = &QQ[DIR_0PM * numberOfBCnodes];
+ q_dirTS = &QQ[DIR_0MP * numberOfBCnodes];
q_dirTNE = &QQ[DIR_PPP * numberOfBCnodes];
q_dirTSW = &QQ[DIR_MMP * numberOfBCnodes];
q_dirTSE = &QQ[DIR_PMP * numberOfBCnodes];
@@ -1567,46 +1599,46 @@ __global__ void QPressDevice27(real* rhoBC,
real f_E, f_W, f_N, f_S, f_T, f_B, f_NE, f_SW, f_SE, f_NW, f_TE, f_BW, f_BE,
f_TW, f_TN, f_BS, f_BN, f_TS, f_TNE, f_TSW, f_TSE, f_TNW, f_BNE, f_BSW, f_BSE, f_BNW;
- f_W = (D.f[DIR_P00 ])[ke ];
- f_E = (D.f[DIR_M00 ])[kw ];
- f_S = (D.f[DIR_0P0 ])[kn ];
- f_N = (D.f[DIR_0M0 ])[ks ];
- f_B = (D.f[DIR_00P ])[kt ];
- f_T = (D.f[DIR_00M ])[kb ];
- f_SW = (D.f[DIR_PP0 ])[kne ];
- f_NE = (D.f[DIR_MM0 ])[ksw ];
- f_NW = (D.f[DIR_PM0 ])[kse ];
- f_SE = (D.f[DIR_MP0 ])[knw ];
- f_BW = (D.f[DIR_P0P ])[kte ];
- f_TE = (D.f[DIR_M0M ])[kbw ];
- f_TW = (D.f[DIR_P0M ])[kbe ];
- f_BE = (D.f[DIR_M0P ])[ktw ];
- f_BS = (D.f[DIR_0PP ])[ktn ];
- f_TN = (D.f[DIR_0MM ])[kbs ];
- f_TS = (D.f[DIR_0PM ])[kbn ];
- f_BN = (D.f[DIR_0MP ])[kts ];
- f_BSW = (D.f[DIR_PPP ])[ktne ];
- f_BNE = (D.f[DIR_MMP ])[ktsw ];
- f_BNW = (D.f[DIR_PMP ])[ktse ];
- f_BSE = (D.f[DIR_MPP ])[ktnw ];
- f_TSW = (D.f[DIR_PPM ])[kbne ];
- f_TNE = (D.f[DIR_MMM ])[kbsw ];
- f_TNW = (D.f[DIR_PMM ])[kbse ];
- f_TSE = (D.f[DIR_MPM ])[kbnw ];
+ f_W = (D.f[DIR_P00])[ke ];
+ f_E = (D.f[DIR_M00])[kw ];
+ f_S = (D.f[DIR_0P0])[kn ];
+ f_N = (D.f[DIR_0M0])[ks ];
+ f_B = (D.f[DIR_00P])[kt ];
+ f_T = (D.f[DIR_00M])[kb ];
+ f_SW = (D.f[DIR_PP0])[kne ];
+ f_NE = (D.f[DIR_MM0])[ksw ];
+ f_NW = (D.f[DIR_PM0])[kse ];
+ f_SE = (D.f[DIR_MP0])[knw ];
+ f_BW = (D.f[DIR_P0P])[kte ];
+ f_TE = (D.f[DIR_M0M])[kbw ];
+ f_TW = (D.f[DIR_P0M])[kbe ];
+ f_BE = (D.f[DIR_M0P])[ktw ];
+ f_BS = (D.f[DIR_0PP])[ktn ];
+ f_TN = (D.f[DIR_0MM])[kbs ];
+ f_TS = (D.f[DIR_0PM])[kbn ];
+ f_BN = (D.f[DIR_0MP])[kts ];
+ f_BSW = (D.f[DIR_PPP])[ktne ];
+ f_BNE = (D.f[DIR_MMP])[ktsw ];
+ f_BNW = (D.f[DIR_PMP])[ktse ];
+ f_BSE = (D.f[DIR_MPP])[ktnw ];
+ f_TSW = (D.f[DIR_PPM])[kbne ];
+ f_TNE = (D.f[DIR_MMM])[kbsw ];
+ f_TNW = (D.f[DIR_PMM])[kbse ];
+ f_TSE = (D.f[DIR_MPM])[kbnw ];
////////////////////////////////////////////////////////////////////////////////
real q, vx1, vx2, vx3, drho;
vx1 = ((f_TSE - f_BNW) - (f_TNW - f_BSE)) + ((f_TNE - f_BSW) - (f_TSW - f_BNE)) +
((f_BE - f_TW) + (f_TE - f_BW)) + ((f_SE - f_NW) + (f_NE - f_SW)) +
- (f_E - f_W);
+ (f_E - f_W);
vx2 = (-(f_TSE - f_BNW) + (f_TNW - f_BSE)) + ((f_TNE - f_BSW) - (f_TSW - f_BNE)) +
((f_BN - f_TS) + (f_TN - f_BS)) + (-(f_SE - f_NW) + (f_NE - f_SW)) +
- (f_N - f_S);
+ (f_N - f_S);
vx3 = ((f_TSE - f_BNW) + (f_TNW - f_BSE)) + ((f_TNE - f_BSW) + (f_TSW - f_BNE)) +
(-(f_BN - f_TS) + (f_TN - f_BS)) + ((f_TE - f_BW) - (f_BE - f_TW)) +
- (f_T - f_B);
+ (f_T - f_B);
real cu_sq=c3o2*(vx1*vx1+vx2*vx2+vx3*vx3);
//////////////////////////////////////////////////////////////////////////
@@ -1616,245 +1648,245 @@ __global__ void QPressDevice27(real* rhoBC,
////////////////////////////////////////////////////////////////////////////////
if (isEvenTimestep==false)
{
- D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
- }
+ D.f[DIR_P00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_MPM * numberOfLBnodes];
+ }
else
{
- D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
+ D.f[DIR_M00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_MPM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_PMP * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
q = q_dirE[k];
if (q>=c0o1 && q<=c1o1)
{
- (D.f[DIR_M00])[kw]=c2o27* (drho+c3o1*(-vx1 )+c9o2*(-vx1 )*(-vx1 )-cu_sq);
- //(D.f[DIR_P00])[ke]=c2over27* (drho+three*( vx1 )+c9over2*( vx1 )*( vx1 )-cu_sq);
+ (D.f[DIR_M00])[kw]=c2o27* (drho+c3o1*(-vx1 )+c9o2*(-vx1 )*(-vx1 )-cu_sq);
+ //(D.f[DIR_P00])[ke]=c2over27* (drho+three*( vx1 )+c9over2*( vx1 )*( vx1 )-cu_sq);
}
q = q_dirW[k];
if (q>=c0o1 && q<=c1o1)
{
- (D.f[DIR_P00])[ke]=c2o27* (drho+c3o1*( vx1 )+c9o2*( vx1 )*( vx1 )-cu_sq);
- //(D.f[DIR_M00])[kw]=c2over27* (drho+three*(-vx1 )+c9over2*(-vx1 )*(-vx1 )-cu_sq);
+ (D.f[DIR_P00])[ke]=c2o27* (drho+c3o1*( vx1 )+c9o2*( vx1 )*( vx1 )-cu_sq);
+ //(D.f[DIR_M00])[kw]=c2over27* (drho+three*(-vx1 )+c9over2*(-vx1 )*(-vx1 )-cu_sq);
}
q = q_dirN[k];
if (q>=c0o1 && q<=c1o1)
{
- (D.f[DIR_0M0])[ks]=c2o27* (drho+c3o1*( -vx2 )+c9o2*( -vx2 )*( -vx2 )-cu_sq);
- //(D.f[DIR_0P0])[kn]=c2over27* (drho+three*( vx2 )+c9over2*( vx2 )*( vx2 )-cu_sq);
+ (D.f[DIR_0M0])[ks]=c2o27* (drho+c3o1*( -vx2 )+c9o2*( -vx2 )*( -vx2 )-cu_sq);
+ //(D.f[DIR_0P0])[kn]=c2over27* (drho+three*( vx2 )+c9over2*( vx2 )*( vx2 )-cu_sq);
}
q = q_dirS[k];
if (q>=c0o1 && q<=c1o1)
{
- (D.f[DIR_0P0])[kn]=c2o27* (drho+c3o1*( vx2 )+c9o2*( vx2 )*( vx2 )-cu_sq);
- //(D.f[DIR_0M0])[ks]=c2over27* (drho+three*( -vx2 )+c9over2*( -vx2 )*( -vx2 )-cu_sq);
+ (D.f[DIR_0P0])[kn]=c2o27* (drho+c3o1*( vx2 )+c9o2*( vx2 )*( vx2 )-cu_sq);
+ //(D.f[DIR_0M0])[ks]=c2over27* (drho+three*( -vx2 )+c9over2*( -vx2 )*( -vx2 )-cu_sq);
}
q = q_dirT[k];
if (q>=c0o1 && q<=c1o1)
{
- (D.f[DIR_00M])[kb]=c2o27* (drho+c3o1*( -vx3)+c9o2*( -vx3)*( -vx3)-cu_sq);
- //(D.f[DIR_00P])[kt]=c2over27* (drho+three*( vx3)+c9over2*( vx3)*( vx3)-cu_sq);
+ (D.f[DIR_00M])[kb]=c2o27* (drho+c3o1*( -vx3)+c9o2*( -vx3)*( -vx3)-cu_sq);
+ //(D.f[DIR_00P])[kt]=c2over27* (drho+three*( vx3)+c9over2*( vx3)*( vx3)-cu_sq);
}
q = q_dirB[k];
if (q>=c0o1 && q<=c1o1)
{
- (D.f[DIR_00P])[kt]=c2o27* (drho+c3o1*( vx3)+c9o2*( vx3)*( vx3)-cu_sq);
- //(D.f[DIR_00M])[kb]=c2over27* (drho+three*( -vx3)+c9over2*( -vx3)*( -vx3)-cu_sq);
+ (D.f[DIR_00P])[kt]=c2o27* (drho+c3o1*( vx3)+c9o2*( vx3)*( vx3)-cu_sq);
+ //(D.f[DIR_00M])[kb]=c2over27* (drho+three*( -vx3)+c9over2*( -vx3)*( -vx3)-cu_sq);
}
q = q_dirNE[k];
if (q>=c0o1 && q<=c1o1)
{
- (D.f[DIR_MM0])[ksw]=c1o54* (drho+c3o1*(-vx1-vx2 )+c9o2*(-vx1-vx2 )*(-vx1-vx2 )-cu_sq);
- //(D.f[DIR_PP0])[kne]=c1over54* (drho+three*( vx1+vx2 )+c9over2*( vx1+vx2 )*( vx1+vx2 )-cu_sq);
+ (D.f[DIR_MM0])[ksw]=c1o54* (drho+c3o1*(-vx1-vx2 )+c9o2*(-vx1-vx2 )*(-vx1-vx2 )-cu_sq);
+ //(D.f[DIR_PP0])[kne]=c1over54* (drho+three*( vx1+vx2 )+c9over2*( vx1+vx2 )*( vx1+vx2 )-cu_sq);
}
q = q_dirSW[k];
if (q>=c0o1 && q<=c1o1)
{
- (D.f[DIR_PP0])[kne]=c1o54* (drho+c3o1*( vx1+vx2 )+c9o2*( vx1+vx2 )*( vx1+vx2 )-cu_sq);
- //(D.f[DIR_MM0])[ksw]=c1over54* (drho+three*(-vx1-vx2 )+c9over2*(-vx1-vx2 )*(-vx1-vx2 )-cu_sq);
+ (D.f[DIR_PP0])[kne]=c1o54* (drho+c3o1*( vx1+vx2 )+c9o2*( vx1+vx2 )*( vx1+vx2 )-cu_sq);
+ //(D.f[DIR_MM0])[ksw]=c1over54* (drho+three*(-vx1-vx2 )+c9over2*(-vx1-vx2 )*(-vx1-vx2 )-cu_sq);
}
q = q_dirSE[k];
if (q>=c0o1 && q<=c1o1)
{
- (D.f[DIR_MP0])[knw]=c1o54* (drho+c3o1*(-vx1+vx2 )+c9o2*(-vx1+vx2 )*(-vx1+vx2 )-cu_sq);
- //(D.f[DIR_PM0])[kse]=c1over54* (drho+three*( vx1-vx2 )+c9over2*( vx1-vx2 )*( vx1-vx2 )-cu_sq);
+ (D.f[DIR_MP0])[knw]=c1o54* (drho+c3o1*(-vx1+vx2 )+c9o2*(-vx1+vx2 )*(-vx1+vx2 )-cu_sq);
+ //(D.f[DIR_PM0])[kse]=c1over54* (drho+three*( vx1-vx2 )+c9over2*( vx1-vx2 )*( vx1-vx2 )-cu_sq);
}
q = q_dirNW[k];
if (q>=c0o1 && q<=c1o1)
{
- (D.f[DIR_PM0])[kse]=c1o54* (drho+c3o1*( vx1-vx2 )+c9o2*( vx1-vx2 )*( vx1-vx2 )-cu_sq);
- //(D.f[DIR_MP0])[knw]=c1over54* (drho+three*(-vx1+vx2 )+c9over2*(-vx1+vx2 )*(-vx1+vx2 )-cu_sq);
+ (D.f[DIR_PM0])[kse]=c1o54* (drho+c3o1*( vx1-vx2 )+c9o2*( vx1-vx2 )*( vx1-vx2 )-cu_sq);
+ //(D.f[DIR_MP0])[knw]=c1over54* (drho+three*(-vx1+vx2 )+c9over2*(-vx1+vx2 )*(-vx1+vx2 )-cu_sq);
}
q = q_dirTE[k];
if (q>=c0o1 && q<=c1o1)
{
- (D.f[DIR_M0M])[kbw]=c1o54* (drho+c3o1*(-vx1 -vx3)+c9o2*(-vx1 -vx3)*(-vx1 -vx3)-cu_sq);
- //(D.f[DIR_P0P])[kte]=c1over54* (drho+three*( vx1 +vx3)+c9over2*( vx1 +vx3)*( vx1 +vx3)-cu_sq);
+ (D.f[DIR_M0M])[kbw]=c1o54* (drho+c3o1*(-vx1 -vx3)+c9o2*(-vx1 -vx3)*(-vx1 -vx3)-cu_sq);
+ //(D.f[DIR_P0P])[kte]=c1over54* (drho+three*( vx1 +vx3)+c9over2*( vx1 +vx3)*( vx1 +vx3)-cu_sq);
}
q = q_dirBW[k];
if (q>=c0o1 && q<=c1o1)
{
- (D.f[DIR_P0P])[kte]=c1o54* (drho+c3o1*( vx1 +vx3)+c9o2*( vx1 +vx3)*( vx1 +vx3)-cu_sq);
- //(D.f[DIR_M0M])[kbw]=c1over54* (drho+three*(-vx1 -vx3)+c9over2*(-vx1 -vx3)*(-vx1 -vx3)-cu_sq);
+ (D.f[DIR_P0P])[kte]=c1o54* (drho+c3o1*( vx1 +vx3)+c9o2*( vx1 +vx3)*( vx1 +vx3)-cu_sq);
+ //(D.f[DIR_M0M])[kbw]=c1over54* (drho+three*(-vx1 -vx3)+c9over2*(-vx1 -vx3)*(-vx1 -vx3)-cu_sq);
}
q = q_dirBE[k];
if (q>=c0o1 && q<=c1o1)
{
- (D.f[DIR_M0P])[ktw]=c1o54* (drho+c3o1*(-vx1 +vx3)+c9o2*(-vx1 +vx3)*(-vx1 +vx3)-cu_sq);
- //(D.f[DIR_P0M])[kbe]=c1over54* (drho+three*( vx1 -vx3)+c9over2*( vx1 -vx3)*( vx1 -vx3)-cu_sq);
+ (D.f[DIR_M0P])[ktw]=c1o54* (drho+c3o1*(-vx1 +vx3)+c9o2*(-vx1 +vx3)*(-vx1 +vx3)-cu_sq);
+ //(D.f[DIR_P0M])[kbe]=c1over54* (drho+three*( vx1 -vx3)+c9over2*( vx1 -vx3)*( vx1 -vx3)-cu_sq);
}
q = q_dirTW[k];
if (q>=c0o1 && q<=c1o1)
{
- (D.f[DIR_P0M])[kbe]=c1o54* (drho+c3o1*( vx1 -vx3)+c9o2*( vx1 -vx3)*( vx1 -vx3)-cu_sq);
- //(D.f[DIR_M0P])[ktw]=c1over54* (drho+three*(-vx1 +vx3)+c9over2*(-vx1 +vx3)*(-vx1 +vx3)-cu_sq);
+ (D.f[DIR_P0M])[kbe]=c1o54* (drho+c3o1*( vx1 -vx3)+c9o2*( vx1 -vx3)*( vx1 -vx3)-cu_sq);
+ //(D.f[DIR_M0P])[ktw]=c1over54* (drho+three*(-vx1 +vx3)+c9over2*(-vx1 +vx3)*(-vx1 +vx3)-cu_sq);
}
q = q_dirTN[k];
if (q>=c0o1 && q<=c1o1)
{
- (D.f[DIR_0MM])[kbs]=c1o54* (drho+c3o1*( -vx2-vx3)+c9o2*( -vx2-vx3)*( -vx2-vx3)-cu_sq);
- //(D.f[DIR_0PP])[ktn]=c1over54* (drho+three*( vx2+vx3)+c9over2*( vx2+vx3)*( vx2+vx3)-cu_sq);
+ (D.f[DIR_0MM])[kbs]=c1o54* (drho+c3o1*( -vx2-vx3)+c9o2*( -vx2-vx3)*( -vx2-vx3)-cu_sq);
+ //(D.f[DIR_0PP])[ktn]=c1over54* (drho+three*( vx2+vx3)+c9over2*( vx2+vx3)*( vx2+vx3)-cu_sq);
}
q = q_dirBS[k];
if (q>=c0o1 && q<=c1o1)
{
- (D.f[DIR_0PP])[ktn]=c1o54* (drho+c3o1*( vx2+vx3)+c9o2*( vx2+vx3)*( vx2+vx3)-cu_sq);
- //(D.f[DIR_0MM])[kbs]=c1over54* (drho+three*( -vx2-vx3)+c9over2*( -vx2-vx3)*( -vx2-vx3)-cu_sq);
+ (D.f[DIR_0PP])[ktn]=c1o54* (drho+c3o1*( vx2+vx3)+c9o2*( vx2+vx3)*( vx2+vx3)-cu_sq);
+ //(D.f[DIR_0MM])[kbs]=c1over54* (drho+three*( -vx2-vx3)+c9over2*( -vx2-vx3)*( -vx2-vx3)-cu_sq);
}
q = q_dirBN[k];
if (q>=c0o1 && q<=c1o1)
{
- (D.f[DIR_0MP])[kts]=c1o54* (drho+c3o1*( -vx2+vx3)+c9o2*( -vx2+vx3)*( -vx2+vx3)-cu_sq);
- //(D.f[DIR_0PM])[kbn]=c1over54* (drho+three*( vx2-vx3)+c9over2*( vx2-vx3)*( vx2-vx3)-cu_sq);
+ (D.f[DIR_0MP])[kts]=c1o54* (drho+c3o1*( -vx2+vx3)+c9o2*( -vx2+vx3)*( -vx2+vx3)-cu_sq);
+ //(D.f[DIR_0PM])[kbn]=c1over54* (drho+three*( vx2-vx3)+c9over2*( vx2-vx3)*( vx2-vx3)-cu_sq);
}
q = q_dirTS[k];
if (q>=c0o1 && q<=c1o1)
{
- (D.f[DIR_0PM])[kbn]=c1o54* (drho+c3o1*( vx2-vx3)+c9o2*( vx2-vx3)*( vx2-vx3)-cu_sq);
- //(D.f[DIR_0MP])[kts]=c1over54* (drho+three*( -vx2+vx3)+c9over2*( -vx2+vx3)*( -vx2+vx3)-cu_sq);
+ (D.f[DIR_0PM])[kbn]=c1o54* (drho+c3o1*( vx2-vx3)+c9o2*( vx2-vx3)*( vx2-vx3)-cu_sq);
+ //(D.f[DIR_0MP])[kts]=c1over54* (drho+three*( -vx2+vx3)+c9over2*( -vx2+vx3)*( -vx2+vx3)-cu_sq);
}
q = q_dirTNE[k];
if (q>=c0o1 && q<=c1o1)
{
- (D.f[DIR_MMM])[kbsw]=c1o216*(drho+c3o1*(-vx1-vx2-vx3)+c9o2*(-vx1-vx2-vx3)*(-vx1-vx2-vx3)-cu_sq);
- //(D.f[DIR_PPP])[ktne]=c1over216*(drho+three*( vx1+vx2+vx3)+c9over2*( vx1+vx2+vx3)*( vx1+vx2+vx3)-cu_sq);
+ (D.f[DIR_MMM])[kbsw]=c1o216*(drho+c3o1*(-vx1-vx2-vx3)+c9o2*(-vx1-vx2-vx3)*(-vx1-vx2-vx3)-cu_sq);
+ //(D.f[DIR_PPP])[ktne]=c1over216*(drho+three*( vx1+vx2+vx3)+c9over2*( vx1+vx2+vx3)*( vx1+vx2+vx3)-cu_sq);
}
q = q_dirBSW[k];
if (q>=c0o1 && q<=c1o1)
{
- (D.f[DIR_PPP])[ktne]=c1o216*(drho+c3o1*( vx1+vx2+vx3)+c9o2*( vx1+vx2+vx3)*( vx1+vx2+vx3)-cu_sq);
- //(D.f[DIR_MMM])[kbsw]=c1over216*(drho+three*(-vx1-vx2-vx3)+c9over2*(-vx1-vx2-vx3)*(-vx1-vx2-vx3)-cu_sq);
+ (D.f[DIR_PPP])[ktne]=c1o216*(drho+c3o1*( vx1+vx2+vx3)+c9o2*( vx1+vx2+vx3)*( vx1+vx2+vx3)-cu_sq);
+ //(D.f[DIR_MMM])[kbsw]=c1over216*(drho+three*(-vx1-vx2-vx3)+c9over2*(-vx1-vx2-vx3)*(-vx1-vx2-vx3)-cu_sq);
}
q = q_dirBNE[k];
if (q>=c0o1 && q<=c1o1)
{
- (D.f[DIR_MMP])[ktsw]=c1o216*(drho+c3o1*(-vx1-vx2+vx3)+c9o2*(-vx1-vx2+vx3)*(-vx1-vx2+vx3)-cu_sq);
- //(D.f[DIR_PPM])[kbne]=c1over216*(drho+three*( vx1+vx2-vx3)+c9over2*( vx1+vx2-vx3)*( vx1+vx2-vx3)-cu_sq);
+ (D.f[DIR_MMP])[ktsw]=c1o216*(drho+c3o1*(-vx1-vx2+vx3)+c9o2*(-vx1-vx2+vx3)*(-vx1-vx2+vx3)-cu_sq);
+ //(D.f[DIR_PPM])[kbne]=c1over216*(drho+three*( vx1+vx2-vx3)+c9over2*( vx1+vx2-vx3)*( vx1+vx2-vx3)-cu_sq);
}
q = q_dirTSW[k];
if (q>=c0o1 && q<=c1o1)
{
- (D.f[DIR_PPM])[kbne]=c1o216*(drho+c3o1*( vx1+vx2-vx3)+c9o2*( vx1+vx2-vx3)*( vx1+vx2-vx3)-cu_sq);
- //(D.f[DIR_MMP])[ktsw]=c1over216*(drho+three*(-vx1-vx2+vx3)+c9over2*(-vx1-vx2+vx3)*(-vx1-vx2+vx3)-cu_sq);
+ (D.f[DIR_PPM])[kbne]=c1o216*(drho+c3o1*( vx1+vx2-vx3)+c9o2*( vx1+vx2-vx3)*( vx1+vx2-vx3)-cu_sq);
+ //(D.f[DIR_MMP])[ktsw]=c1over216*(drho+three*(-vx1-vx2+vx3)+c9over2*(-vx1-vx2+vx3)*(-vx1-vx2+vx3)-cu_sq);
}
q = q_dirTSE[k];
if (q>=c0o1 && q<=c1o1)
{
- (D.f[DIR_MPM])[kbnw]=c1o216*(drho+c3o1*(-vx1+vx2-vx3)+c9o2*(-vx1+vx2-vx3)*(-vx1+vx2-vx3)-cu_sq);
- //(D.f[DIR_PMP])[ktse]=c1over216*(drho+three*( vx1-vx2+vx3)+c9over2*( vx1-vx2+vx3)*( vx1-vx2+vx3)-cu_sq);
+ (D.f[DIR_MPM])[kbnw]=c1o216*(drho+c3o1*(-vx1+vx2-vx3)+c9o2*(-vx1+vx2-vx3)*(-vx1+vx2-vx3)-cu_sq);
+ //(D.f[DIR_PMP])[ktse]=c1over216*(drho+three*( vx1-vx2+vx3)+c9over2*( vx1-vx2+vx3)*( vx1-vx2+vx3)-cu_sq);
}
q = q_dirBNW[k];
if (q>=c0o1 && q<=c1o1)
{
- (D.f[DIR_PMP])[ktse]=c1o216*(drho+c3o1*( vx1-vx2+vx3)+c9o2*( vx1-vx2+vx3)*( vx1-vx2+vx3)-cu_sq);
- //(D.f[DIR_MPM])[kbnw]=c1over216*(drho+three*(-vx1+vx2-vx3)+c9over2*(-vx1+vx2-vx3)*(-vx1+vx2-vx3)-cu_sq);
+ (D.f[DIR_PMP])[ktse]=c1o216*(drho+c3o1*( vx1-vx2+vx3)+c9o2*( vx1-vx2+vx3)*( vx1-vx2+vx3)-cu_sq);
+ //(D.f[DIR_MPM])[kbnw]=c1over216*(drho+three*(-vx1+vx2-vx3)+c9over2*(-vx1+vx2-vx3)*(-vx1+vx2-vx3)-cu_sq);
}
q = q_dirBSE[k];
if (q>=c0o1 && q<=c1o1)
{
- (D.f[DIR_MPP])[ktnw]=c1o216*(drho+c3o1*(-vx1+vx2+vx3)+c9o2*(-vx1+vx2+vx3)*(-vx1+vx2+vx3)-cu_sq);
- //(D.f[DIR_PMM])[kbse]=c1over216*(drho+three*( vx1-vx2-vx3)+c9over2*( vx1-vx2-vx3)*( vx1-vx2-vx3)-cu_sq);
+ (D.f[DIR_MPP])[ktnw]=c1o216*(drho+c3o1*(-vx1+vx2+vx3)+c9o2*(-vx1+vx2+vx3)*(-vx1+vx2+vx3)-cu_sq);
+ //(D.f[DIR_PMM])[kbse]=c1over216*(drho+three*( vx1-vx2-vx3)+c9over2*( vx1-vx2-vx3)*( vx1-vx2-vx3)-cu_sq);
}
q = q_dirTNW[k];
if (q>=c0o1 && q<=c1o1)
{
- (D.f[DIR_PMM])[kbse]=c1o216*(drho+c3o1*( vx1-vx2-vx3)+c9o2*( vx1-vx2-vx3)*( vx1-vx2-vx3)-cu_sq);
- //(D.f[DIR_MPP])[ktnw]=c1over216*(drho+three*(-vx1+vx2+vx3)+c9over2*(-vx1+vx2+vx3)*(-vx1+vx2+vx3)-cu_sq);
+ (D.f[DIR_PMM])[kbse]=c1o216*(drho+c3o1*( vx1-vx2-vx3)+c9o2*( vx1-vx2-vx3)*( vx1-vx2-vx3)-cu_sq);
+ //(D.f[DIR_MPP])[ktnw]=c1over216*(drho+three*(-vx1+vx2+vx3)+c9over2*(-vx1+vx2+vx3)*(-vx1+vx2+vx3)-cu_sq);
}
}
}
@@ -1899,86 +1931,87 @@ __global__ void QPressDevice27(real* rhoBC,
//////////////////////////////////////////////////////////////////////////////
-__global__ void QPressDeviceAntiBB27( real* rhoBC,
- real* vx,
- real* vy,
- real* vz,
- real* DD,
- int* k_Q,
- real* QQ,
- int numberOfBCnodes,
- real om1,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep)
+__global__ void QPressDeviceAntiBB27(
+ real* rhoBC,
+ real* vx,
+ real* vy,
+ real* vz,
+ real* DD,
+ int* k_Q,
+ real* QQ,
+ int numberOfBCnodes,
+ real om1,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep)
{
Distributions27 D;
if (isEvenTimestep==true)
{
- D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
- }
+ D.f[DIR_P00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_MPM * numberOfLBnodes];
+ }
else
{
- D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
+ D.f[DIR_M00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_MPM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_PMP * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////
- const unsigned x = threadIdx.x; // Globaler x-Index
- const unsigned y = blockIdx.x; // Globaler y-Index
- const unsigned z = blockIdx.y; // Globaler z-Index
+ const unsigned x = threadIdx.x; // Globaler x-Index
+ const unsigned y = blockIdx.x; // Globaler y-Index
+ const unsigned z = blockIdx.y; // Globaler z-Index
const unsigned nx = blockDim.x;
const unsigned ny = gridDim.x;
@@ -1988,37 +2021,37 @@ __global__ void QPressDeviceAntiBB27( real* rhoBC,
if(k<numberOfBCnodes)
{
- real *q_dirE, *q_dirW, *q_dirN, *q_dirS, *q_dirT, *q_dirB,
+ real *q_dirE, *q_dirW, *q_dirN, *q_dirS, *q_dirT, *q_dirB,
*q_dirNE, *q_dirSW, *q_dirSE, *q_dirNW, *q_dirTE, *q_dirBW,
*q_dirBE, *q_dirTW, *q_dirTN, *q_dirBS, *q_dirBN, *q_dirTS,
*q_dirTNE, *q_dirTSW, *q_dirTSE, *q_dirTNW, *q_dirBNE, *q_dirBSW,
- *q_dirBSE, *q_dirBNW;
- q_dirE = &QQ[DIR_P00 *numberOfBCnodes];
- q_dirW = &QQ[DIR_M00 *numberOfBCnodes];
- q_dirN = &QQ[DIR_0P0 *numberOfBCnodes];
- q_dirS = &QQ[DIR_0M0 *numberOfBCnodes];
- q_dirT = &QQ[DIR_00P *numberOfBCnodes];
- q_dirB = &QQ[DIR_00M *numberOfBCnodes];
- q_dirNE = &QQ[DIR_PP0 *numberOfBCnodes];
- q_dirSW = &QQ[DIR_MM0 *numberOfBCnodes];
- q_dirSE = &QQ[DIR_PM0 *numberOfBCnodes];
- q_dirNW = &QQ[DIR_MP0 *numberOfBCnodes];
- q_dirTE = &QQ[DIR_P0P *numberOfBCnodes];
- q_dirBW = &QQ[DIR_M0M *numberOfBCnodes];
- q_dirBE = &QQ[DIR_P0M *numberOfBCnodes];
- q_dirTW = &QQ[DIR_M0P *numberOfBCnodes];
- q_dirTN = &QQ[DIR_0PP *numberOfBCnodes];
- q_dirBS = &QQ[DIR_0MM *numberOfBCnodes];
- q_dirBN = &QQ[DIR_0PM *numberOfBCnodes];
- q_dirTS = &QQ[DIR_0MP *numberOfBCnodes];
- q_dirTNE = &QQ[DIR_PPP *numberOfBCnodes];
- q_dirTSW = &QQ[DIR_MMP *numberOfBCnodes];
- q_dirTSE = &QQ[DIR_PMP *numberOfBCnodes];
- q_dirTNW = &QQ[DIR_MPP *numberOfBCnodes];
- q_dirBNE = &QQ[DIR_PPM *numberOfBCnodes];
- q_dirBSW = &QQ[DIR_MMM *numberOfBCnodes];
- q_dirBSE = &QQ[DIR_PMM *numberOfBCnodes];
- q_dirBNW = &QQ[DIR_MPM *numberOfBCnodes];
+ *q_dirBSE, *q_dirBNW;
+ q_dirE = &QQ[DIR_P00 * numberOfBCnodes];
+ q_dirW = &QQ[DIR_M00 * numberOfBCnodes];
+ q_dirN = &QQ[DIR_0P0 * numberOfBCnodes];
+ q_dirS = &QQ[DIR_0M0 * numberOfBCnodes];
+ q_dirT = &QQ[DIR_00P * numberOfBCnodes];
+ q_dirB = &QQ[DIR_00M * numberOfBCnodes];
+ q_dirNE = &QQ[DIR_PP0 * numberOfBCnodes];
+ q_dirSW = &QQ[DIR_MM0 * numberOfBCnodes];
+ q_dirSE = &QQ[DIR_PM0 * numberOfBCnodes];
+ q_dirNW = &QQ[DIR_MP0 * numberOfBCnodes];
+ q_dirTE = &QQ[DIR_P0P * numberOfBCnodes];
+ q_dirBW = &QQ[DIR_M0M * numberOfBCnodes];
+ q_dirBE = &QQ[DIR_P0M * numberOfBCnodes];
+ q_dirTW = &QQ[DIR_M0P * numberOfBCnodes];
+ q_dirTN = &QQ[DIR_0PP * numberOfBCnodes];
+ q_dirBS = &QQ[DIR_0MM * numberOfBCnodes];
+ q_dirBN = &QQ[DIR_0PM * numberOfBCnodes];
+ q_dirTS = &QQ[DIR_0MP * numberOfBCnodes];
+ q_dirTNE = &QQ[DIR_PPP * numberOfBCnodes];
+ q_dirTSW = &QQ[DIR_MMP * numberOfBCnodes];
+ q_dirTSE = &QQ[DIR_PMP * numberOfBCnodes];
+ q_dirTNW = &QQ[DIR_MPP * numberOfBCnodes];
+ q_dirBNE = &QQ[DIR_PPM * numberOfBCnodes];
+ q_dirBSW = &QQ[DIR_MMM * numberOfBCnodes];
+ q_dirBSE = &QQ[DIR_PMM * numberOfBCnodes];
+ q_dirBNW = &QQ[DIR_MPM * numberOfBCnodes];
////////////////////////////////////////////////////////////////////////////////
//index
unsigned int KQK = k_Q[k];
@@ -2053,123 +2086,123 @@ __global__ void QPressDeviceAntiBB27( real* rhoBC,
real f_E, f_W, f_N, f_S, f_T, f_B, f_NE, f_SW, f_SE, f_NW, f_TE, f_BW, f_BE,
f_TW, f_TN, f_BS, f_BN, f_TS, f_TNE, f_TSW, f_TSE, f_TNW, f_BNE, f_BSW, f_BSE, f_BNW, f_ZERO;
- f_W = (D.f[DIR_P00 ])[ke ];
- f_E = (D.f[DIR_M00 ])[kw ];
- f_S = (D.f[DIR_0P0 ])[kn ];
- f_N = (D.f[DIR_0M0 ])[ks ];
- f_B = (D.f[DIR_00P ])[kt ];
- f_T = (D.f[DIR_00M ])[kb ];
- f_SW = (D.f[DIR_PP0 ])[kne ];
- f_NE = (D.f[DIR_MM0 ])[ksw ];
- f_NW = (D.f[DIR_PM0 ])[kse ];
- f_SE = (D.f[DIR_MP0 ])[knw ];
- f_BW = (D.f[DIR_P0P ])[kte ];
- f_TE = (D.f[DIR_M0M ])[kbw ];
- f_TW = (D.f[DIR_P0M ])[kbe ];
- f_BE = (D.f[DIR_M0P ])[ktw ];
- f_BS = (D.f[DIR_0PP ])[ktn ];
- f_TN = (D.f[DIR_0MM ])[kbs ];
- f_TS = (D.f[DIR_0PM ])[kbn ];
- f_BN = (D.f[DIR_0MP ])[kts ];
- f_BSW = (D.f[DIR_PPP ])[ktne ];
- f_BNE = (D.f[DIR_MMP ])[ktsw ];
- f_BNW = (D.f[DIR_PMP ])[ktse ];
- f_BSE = (D.f[DIR_MPP ])[ktnw ];
- f_TSW = (D.f[DIR_PPM ])[kbne ];
- f_TNE = (D.f[DIR_MMM ])[kbsw ];
- f_TNW = (D.f[DIR_PMM ])[kbse ];
- f_TSE = (D.f[DIR_MPM ])[kbnw ];
+ f_W = (D.f[DIR_P00])[ke ];
+ f_E = (D.f[DIR_M00])[kw ];
+ f_S = (D.f[DIR_0P0])[kn ];
+ f_N = (D.f[DIR_0M0])[ks ];
+ f_B = (D.f[DIR_00P])[kt ];
+ f_T = (D.f[DIR_00M])[kb ];
+ f_SW = (D.f[DIR_PP0])[kne ];
+ f_NE = (D.f[DIR_MM0])[ksw ];
+ f_NW = (D.f[DIR_PM0])[kse ];
+ f_SE = (D.f[DIR_MP0])[knw ];
+ f_BW = (D.f[DIR_P0P])[kte ];
+ f_TE = (D.f[DIR_M0M])[kbw ];
+ f_TW = (D.f[DIR_P0M])[kbe ];
+ f_BE = (D.f[DIR_M0P])[ktw ];
+ f_BS = (D.f[DIR_0PP])[ktn ];
+ f_TN = (D.f[DIR_0MM])[kbs ];
+ f_TS = (D.f[DIR_0PM])[kbn ];
+ f_BN = (D.f[DIR_0MP])[kts ];
+ f_BSW = (D.f[DIR_PPP])[ktne ];
+ f_BNE = (D.f[DIR_MMP])[ktsw ];
+ f_BNW = (D.f[DIR_PMP])[ktse ];
+ f_BSE = (D.f[DIR_MPP])[ktnw ];
+ f_TSW = (D.f[DIR_PPM])[kbne ];
+ f_TNE = (D.f[DIR_MMM])[kbsw ];
+ f_TNW = (D.f[DIR_PMM])[kbse ];
+ f_TSE = (D.f[DIR_MPM])[kbnw ];
f_ZERO = (D.f[DIR_000])[kzero];
////////////////////////////////////////////////////////////////////////////////
//real vx1, vx2, vx3, drho;
//vx1 = ((f_TSE - f_BNW) - (f_TNW - f_BSE)) + ((f_TNE - f_BSW) - (f_TSW - f_BNE)) +
// ((f_BE - f_TW) + (f_TE - f_BW)) + ((f_SE - f_NW) + (f_NE - f_SW)) +
- // (f_E - f_W);
+ // (f_E - f_W);
//vx2 = (-(f_TSE - f_BNW) + (f_TNW - f_BSE)) + ((f_TNE - f_BSW) - (f_TSW - f_BNE)) +
// ((f_BN - f_TS) + (f_TN - f_BS)) + (-(f_SE - f_NW) + (f_NE - f_SW)) +
- // (f_N - f_S);
+ // (f_N - f_S);
//vx3 = ((f_TSE - f_BNW) + (f_TNW - f_BSE)) + ((f_TNE - f_BSW) + (f_TSW - f_BNE)) +
// (-(f_BN - f_TS) + (f_TN - f_BS)) + ((f_TE - f_BW) - (f_BE - f_TW)) +
- // (f_T - f_B);
+ // (f_T - f_B);
//real cu_sq=c3o2*(vx1*vx1+vx2*vx2+vx3*vx3);
//////////////////////////////////////////////////////////////////////////
real drho = f_ZERO+f_E+f_W+f_N+f_S+f_T+f_B+f_NE+f_SW+f_SE+f_NW+f_TE+f_BW+f_BE+f_TW+f_TN+f_BS+f_BN+f_TS+
- f_TNE+f_TSW+f_TSE+f_TNW+f_BNE+f_BSW+f_BSE+f_BNW;
+ f_TNE+f_TSW+f_TSE+f_TNW+f_BNE+f_BSW+f_BSE+f_BNW;
drho = drho - rhoBC[k];
- drho *= 0.01f;
+ drho *= 0.01f;
////////////////////////////////////////////////////////////////////////////////
- real q;
+ real q;
//deltaRho = (rhoBC[k] + one) / (deltaRho + one);
////////////////////////////////////////////////////////////////////////////////
if (isEvenTimestep==false)
{
- D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
- }
+ D.f[DIR_P00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_MPM * numberOfLBnodes];
+ }
else
{
- D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
+ D.f[DIR_M00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_MPM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_PMP * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
q = q_dirE[k];
if (q>=c0o1 && q<=c1o1)
{
- (D.f[DIR_M00])[kw]=f_W-c2o27*drho;
+ (D.f[DIR_M00])[kw]=f_W-c2o27*drho;
}
q = q_dirW[k];
@@ -2181,19 +2214,19 @@ __global__ void QPressDeviceAntiBB27( real* rhoBC,
q = q_dirN[k];
if (q>=c0o1 && q<=c1o1)
{
- (D.f[DIR_0M0])[ks]=f_S-c2o27*drho;
+ (D.f[DIR_0M0])[ks]=f_S-c2o27*drho;
}
q = q_dirS[k];
if (q>=c0o1 && q<=c1o1)
{
- (D.f[DIR_0P0])[kn]=f_N-c2o27*drho;
+ (D.f[DIR_0P0])[kn]=f_N-c2o27*drho;
}
q = q_dirT[k];
if (q>=c0o1 && q<=c1o1)
{
- (D.f[DIR_00M])[kb]=f_B-c2o27*drho;
+ (D.f[DIR_00M])[kb]=f_B-c2o27*drho;
}
q = q_dirB[k];
@@ -2229,13 +2262,13 @@ __global__ void QPressDeviceAntiBB27( real* rhoBC,
q = q_dirTE[k];
if (q>=c0o1 && q<=c1o1)
{
- (D.f[DIR_M0M])[kbw]=f_BW-c1o54*drho;
+ (D.f[DIR_M0M])[kbw]=f_BW-c1o54*drho;
}
q = q_dirBW[k];
if (q>=c0o1 && q<=c1o1)
{
- (D.f[DIR_P0P])[kte]=f_TE-c1o54*drho;
+ (D.f[DIR_P0P])[kte]=f_TE-c1o54*drho;
}
q = q_dirBE[k];
@@ -2364,21 +2397,22 @@ __global__ void QPressDeviceAntiBB27( real* rhoBC,
//////////////////////////////////////////////////////////////////////////////
-__global__ void QPressDeviceFixBackflow27( real* rhoBC,
- real* DD,
- int* k_Q,
- int numberOfBCnodes,
- real om1,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep)
+__global__ void QPressDeviceFixBackflow27(
+ real* rhoBC,
+ real* DD,
+ int* k_Q,
+ int numberOfBCnodes,
+ real om1,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep)
{
////////////////////////////////////////////////////////////////////////////////
- const unsigned x = threadIdx.x; // Globaler x-Index
- const unsigned y = blockIdx.x; // Globaler y-Index
- const unsigned z = blockIdx.y; // Globaler z-Index
+ const unsigned x = threadIdx.x; // Globaler x-Index
+ const unsigned y = blockIdx.x; // Globaler y-Index
+ const unsigned z = blockIdx.y; // Globaler z-Index
const unsigned nx = blockDim.x;
const unsigned ny = gridDim.x;
@@ -2426,63 +2460,63 @@ __global__ void QPressDeviceFixBackflow27( real* rhoBC,
Distributions27 D;
if (isEvenTimestep==false)
{
- D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
- }
+ D.f[DIR_P00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_MPM * numberOfLBnodes];
+ }
else
{
- D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
+ D.f[DIR_M00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_MPM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_PMP * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
(D.f[DIR_M00])[kw] = c2o27 * deltaRho;
@@ -2555,21 +2589,22 @@ __global__ void QPressDeviceFixBackflow27( real* rhoBC,
//////////////////////////////////////////////////////////////////////////////
-__global__ void QPressDeviceDirDepBot27( real* rhoBC,
- real* DD,
- int* k_Q,
- int numberOfBCnodes,
- real om1,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep)
+__global__ void QPressDeviceDirDepBot27(
+ real* rhoBC,
+ real* DD,
+ int* k_Q,
+ int numberOfBCnodes,
+ real om1,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep)
{
////////////////////////////////////////////////////////////////////////////////
- const unsigned x = threadIdx.x; // Globaler x-Index
- const unsigned y = blockIdx.x; // Globaler y-Index
- const unsigned z = blockIdx.y; // Globaler z-Index
+ const unsigned x = threadIdx.x; // Globaler x-Index
+ const unsigned y = blockIdx.x; // Globaler y-Index
+ const unsigned z = blockIdx.y; // Globaler z-Index
const unsigned nx = blockDim.x;
const unsigned ny = gridDim.x;
@@ -2617,86 +2652,86 @@ __global__ void QPressDeviceDirDepBot27( real* rhoBC,
Distributions27 D;
if (isEvenTimestep==false)
{
- D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
- }
+ D.f[DIR_P00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_MPM * numberOfLBnodes];
+ }
else
{
- D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
+ D.f[DIR_M00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_MPM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_PMP * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
real f_E,f_W,f_N,f_S,f_T,f_NE,f_SW,f_SE,f_NW,f_TE,f_TW,f_TN,f_TS,f_ZERO,f_TNE,f_TSW,f_TSE,f_TNW;//,
//f_B,f_BW,f_BE,f_BS,f_BN,f_BSW,f_BNE,f_BNW,f_BSE;
- f_E = (D.f[DIR_P00 ])[ke ];
- f_W = (D.f[DIR_M00 ])[kw ];
- f_N = (D.f[DIR_0P0 ])[kn ];
- f_S = (D.f[DIR_0M0 ])[ks ];
- f_T = (D.f[DIR_00P ])[kt ];
- f_NE = (D.f[DIR_PP0 ])[kne ];
- f_SW = (D.f[DIR_MM0 ])[ksw ];
- f_SE = (D.f[DIR_PM0 ])[kse ];
- f_NW = (D.f[DIR_MP0 ])[knw ];
- f_TE = (D.f[DIR_P0P ])[kte ];
- f_TW = (D.f[DIR_M0P ])[ktw ];
- f_TN = (D.f[DIR_0PP ])[ktn ];
- f_TS = (D.f[DIR_0MP ])[kts ];
+ f_E = (D.f[DIR_P00])[ke ];
+ f_W = (D.f[DIR_M00])[kw ];
+ f_N = (D.f[DIR_0P0])[kn ];
+ f_S = (D.f[DIR_0M0])[ks ];
+ f_T = (D.f[DIR_00P])[kt ];
+ f_NE = (D.f[DIR_PP0])[kne ];
+ f_SW = (D.f[DIR_MM0])[ksw ];
+ f_SE = (D.f[DIR_PM0])[kse ];
+ f_NW = (D.f[DIR_MP0])[knw ];
+ f_TE = (D.f[DIR_P0P])[kte ];
+ f_TW = (D.f[DIR_M0P])[ktw ];
+ f_TN = (D.f[DIR_0PP])[ktn ];
+ f_TS = (D.f[DIR_0MP])[kts ];
f_ZERO = (D.f[DIR_000])[kzero];
- f_TNE = (D.f[DIR_PPP ])[ktne ];
- f_TSW = (D.f[DIR_MMP ])[ktsw ];
- f_TSE = (D.f[DIR_PMP ])[ktse ];
- f_TNW = (D.f[DIR_MPP ])[ktnw ];
+ f_TNE = (D.f[DIR_PPP])[ktne ];
+ f_TSW = (D.f[DIR_MMP])[ktsw ];
+ f_TSE = (D.f[DIR_PMP])[ktse ];
+ f_TNW = (D.f[DIR_MPP])[ktnw ];
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
//f_B = (four*rho- four*f_SW- eight*f_TSW-four*f_W- eight*f_TW- four*f_NW- eight*f_TNW-four*f_S- eight*f_TS-four*f_ZERO+ f_T-four*f_N- eight*f_TN- four*f_SE- eight*f_TSE-four*f_E- eight*f_TE- four*f_NE- eight*f_TNE)/nine;
@@ -2793,496 +2828,474 @@ __global__ void QPressDeviceDirDepBot27( real* rhoBC,
-
-
+__host__ __device__ real computeOutflowDistribution(const real* const &f, const real* const &f1, const int dir, const real cs)
+{
+ return f1[dir] * cs + (c1o1 - cs) * f[dir];
+}
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
-__global__ void QPressNoRhoDevice27( real* rhoBC,
- real* DD,
- int* k_Q,
- int* k_N,
- int numberOfBCnodes,
- real om1,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep)
+__global__ void QPressNoRhoDevice27(
+ real* rhoBC,
+ real* distributions,
+ int* k_Q,
+ int* k_N,
+ int numberOfBCnodes,
+ real om1,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep,
+ int direction)
{
////////////////////////////////////////////////////////////////////////////////
- const unsigned x = threadIdx.x; // Globaler x-Index
- const unsigned y = blockIdx.x; // Globaler y-Index
- const unsigned z = blockIdx.y; // Globaler z-Index
+ //! - Get the node index coordinates from threadIdx, blockIdx, blockDim and gridDim.
+ //!
+ const unsigned nodeIndex = getNodeIndex();
- const unsigned nx = blockDim.x;
- const unsigned ny = gridDim.x;
+ //////////////////////////////////////////////////////////////////////////
- const unsigned k = nx*(ny*z + y) + x;
+ if(nodeIndex >= numberOfBCnodes) return;
+
+ ////////////////////////////////////////////////////////////////////////////////
+ //index
+ unsigned int KQK = k_Q[nodeIndex];
+ // unsigned int kzero= KQK;
+ unsigned int ke = KQK;
+ unsigned int kw = neighborX[KQK];
+ unsigned int kn = KQK;
+ unsigned int ks = neighborY[KQK];
+ unsigned int kt = KQK;
+ unsigned int kb = neighborZ[KQK];
+ unsigned int ksw = neighborY[kw];
+ unsigned int kne = KQK;
+ unsigned int kse = ks;
+ unsigned int knw = kw;
+ unsigned int kbw = neighborZ[kw];
+ unsigned int kte = KQK;
+ unsigned int kbe = kb;
+ unsigned int ktw = kw;
+ unsigned int kbs = neighborZ[ks];
+ unsigned int ktn = KQK;
+ unsigned int kbn = kb;
+ unsigned int kts = ks;
+ unsigned int ktse = ks;
+ unsigned int kbnw = kbw;
+ unsigned int ktnw = kw;
+ unsigned int kbse = kbs;
+ unsigned int ktsw = ksw;
+ unsigned int kbne = kb;
+ unsigned int ktne = KQK;
+ unsigned int kbsw = neighborZ[ksw];
+ ////////////////////////////////////////////////////////////////////////////////
+ //index1
+ unsigned int K1QK = k_N[nodeIndex];
+ //unsigned int k1zero= K1QK;
+ unsigned int k1e = K1QK;
+ unsigned int k1w = neighborX[K1QK];
+ unsigned int k1n = K1QK;
+ unsigned int k1s = neighborY[K1QK];
+ unsigned int k1t = K1QK;
+ unsigned int k1b = neighborZ[K1QK];
+ unsigned int k1sw = neighborY[k1w];
+ unsigned int k1ne = K1QK;
+ unsigned int k1se = k1s;
+ unsigned int k1nw = k1w;
+ unsigned int k1bw = neighborZ[k1w];
+ unsigned int k1te = K1QK;
+ unsigned int k1be = k1b;
+ unsigned int k1tw = k1w;
+ unsigned int k1bs = neighborZ[k1s];
+ unsigned int k1tn = K1QK;
+ unsigned int k1bn = k1b;
+ unsigned int k1ts = k1s;
+ unsigned int k1tse = k1s;
+ unsigned int k1bnw = k1bw;
+ unsigned int k1tnw = k1w;
+ unsigned int k1bse = k1bs;
+ unsigned int k1tsw = k1sw;
+ unsigned int k1bne = k1b;
+ unsigned int k1tne = K1QK;
+ unsigned int k1bsw = neighborZ[k1sw];
+ ////////////////////////////////////////////////////////////////////////////////
+ Distributions27 dist;
+ getPointersToDistributions(dist, distributions, numberOfLBnodes, isEvenTimestep);
+ real f[27], f1[27];
+ //////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
+ f1[DIR_P00] = (dist.f[DIR_P00])[k1e ];
+ f1[DIR_M00] = (dist.f[DIR_M00])[k1w ];
+ f1[DIR_0P0] = (dist.f[DIR_0P0])[k1n ];
+ f1[DIR_0M0] = (dist.f[DIR_0M0])[k1s ];
+ f1[DIR_00P] = (dist.f[DIR_00P])[k1t ];
+ f1[DIR_00M] = (dist.f[DIR_00M])[k1b ];
+ f1[DIR_PP0] = (dist.f[DIR_PP0])[k1ne ];
+ f1[DIR_MM0] = (dist.f[DIR_MM0])[k1sw ];
+ f1[DIR_PM0] = (dist.f[DIR_PM0])[k1se ];
+ f1[DIR_MP0] = (dist.f[DIR_MP0])[k1nw ];
+ f1[DIR_P0P] = (dist.f[DIR_P0P])[k1te ];
+ f1[DIR_M0M] = (dist.f[DIR_M0M])[k1bw ];
+ f1[DIR_P0M] = (dist.f[DIR_P0M])[k1be ];
+ f1[DIR_M0P] = (dist.f[DIR_M0P])[k1tw ];
+ f1[DIR_0PP] = (dist.f[DIR_0PP])[k1tn ];
+ f1[DIR_0MM] = (dist.f[DIR_0MM])[k1bs ];
+ f1[DIR_0PM] = (dist.f[DIR_0PM])[k1bn ];
+ f1[DIR_0MP] = (dist.f[DIR_0MP])[k1ts ];
+ // f1[DIR_000] = (dist.f[DIR_000])[k1zero];
+ f1[DIR_PPP] = (dist.f[DIR_PPP])[k1tne ];
+ f1[DIR_MMP] = (dist.f[DIR_MMP])[k1tsw ];
+ f1[DIR_PMP] = (dist.f[DIR_PMP])[k1tse ];
+ f1[DIR_MPP] = (dist.f[DIR_MPP])[k1tnw ];
+ f1[DIR_PPM] = (dist.f[DIR_PPM])[k1bne ];
+ f1[DIR_MMM] = (dist.f[DIR_MMM])[k1bsw ];
+ f1[DIR_PMM] = (dist.f[DIR_PMM])[k1bse ];
+ f1[DIR_MPM] = (dist.f[DIR_MPM])[k1bnw ];
+ //////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
+ f[DIR_P00] = (dist.f[DIR_P00])[ke ];
+ f[DIR_M00] = (dist.f[DIR_M00])[kw ];
+ f[DIR_0P0] = (dist.f[DIR_0P0])[kn ];
+ f[DIR_0M0] = (dist.f[DIR_0M0])[ks ];
+ f[DIR_00P] = (dist.f[DIR_00P])[kt ];
+ f[DIR_00M] = (dist.f[DIR_00M])[kb ];
+ f[DIR_PP0] = (dist.f[DIR_PP0])[kne ];
+ f[DIR_MM0] = (dist.f[DIR_MM0])[ksw ];
+ f[DIR_PM0] = (dist.f[DIR_PM0])[kse ];
+ f[DIR_MP0] = (dist.f[DIR_MP0])[knw ];
+ f[DIR_P0P] = (dist.f[DIR_P0P])[kte ];
+ f[DIR_M0M] = (dist.f[DIR_M0M])[kbw ];
+ f[DIR_P0M] = (dist.f[DIR_P0M])[kbe ];
+ f[DIR_M0P] = (dist.f[DIR_M0P])[ktw ];
+ f[DIR_0PP] = (dist.f[DIR_0PP])[ktn ];
+ f[DIR_0MM] = (dist.f[DIR_0MM])[kbs ];
+ f[DIR_0PM] = (dist.f[DIR_0PM])[kbn ];
+ f[DIR_0MP] = (dist.f[DIR_0MP])[kts ];
+ // f[DIR_000] = (dist.f[DIR_000])[kzero];
+ f[DIR_PPP] = (dist.f[DIR_PPP])[ktne ];
+ f[DIR_MMP] = (dist.f[DIR_MMP])[ktsw ];
+ f[DIR_PMP] = (dist.f[DIR_PMP])[ktse ];
+ f[DIR_MPP] = (dist.f[DIR_MPP])[ktnw ];
+ f[DIR_PPM] = (dist.f[DIR_PPM])[kbne ];
+ f[DIR_MMM] = (dist.f[DIR_MMM])[kbsw ];
+ f[DIR_PMM] = (dist.f[DIR_PMM])[kbse ];
+ f[DIR_MPM] = (dist.f[DIR_MPM])[kbnw ];
//////////////////////////////////////////////////////////////////////////
- if(k<numberOfBCnodes)
+
+ real cs = c1o1 / sqrtf(c3o1);
+
+ //////////////////////////////////////////////////////////////////////////
+ getPointersToDistributions(dist, distributions, numberOfLBnodes, !isEvenTimestep);
+ switch(direction)
{
- ////////////////////////////////////////////////////////////////////////////////
- //index
- unsigned int KQK = k_Q[k];
- //unsigned int kzero= KQK;
- unsigned int ke = KQK;
- unsigned int kw = neighborX[KQK];
- unsigned int kn = KQK;
- unsigned int ks = neighborY[KQK];
- unsigned int kt = KQK;
- unsigned int kb = neighborZ[KQK];
- unsigned int ksw = neighborY[kw];
- unsigned int kne = KQK;
- unsigned int kse = ks;
- unsigned int knw = kw;
- unsigned int kbw = neighborZ[kw];
- unsigned int kte = KQK;
- unsigned int kbe = kb;
- unsigned int ktw = kw;
- unsigned int kbs = neighborZ[ks];
- unsigned int ktn = KQK;
- unsigned int kbn = kb;
- unsigned int kts = ks;
- unsigned int ktse = ks;
- unsigned int kbnw = kbw;
- unsigned int ktnw = kw;
- unsigned int kbse = kbs;
- unsigned int ktsw = ksw;
- unsigned int kbne = kb;
- unsigned int ktne = KQK;
- unsigned int kbsw = neighborZ[ksw];
- ////////////////////////////////////////////////////////////////////////////////
- //index1
- unsigned int K1QK = k_N[k];
- //unsigned int k1zero= K1QK;
- unsigned int k1e = K1QK;
- unsigned int k1w = neighborX[K1QK];
- unsigned int k1n = K1QK;
- unsigned int k1s = neighborY[K1QK];
- unsigned int k1t = K1QK;
- unsigned int k1b = neighborZ[K1QK];
- unsigned int k1sw = neighborY[k1w];
- unsigned int k1ne = K1QK;
- unsigned int k1se = k1s;
- unsigned int k1nw = k1w;
- unsigned int k1bw = neighborZ[k1w];
- unsigned int k1te = K1QK;
- unsigned int k1be = k1b;
- unsigned int k1tw = k1w;
- unsigned int k1bs = neighborZ[k1s];
- unsigned int k1tn = K1QK;
- unsigned int k1bn = k1b;
- unsigned int k1ts = k1s;
- unsigned int k1tse = k1s;
- unsigned int k1bnw = k1bw;
- unsigned int k1tnw = k1w;
- unsigned int k1bse = k1bs;
- unsigned int k1tsw = k1sw;
- unsigned int k1bne = k1b;
- unsigned int k1tne = K1QK;
- unsigned int k1bsw = neighborZ[k1sw];
- ////////////////////////////////////////////////////////////////////////////////
- Distributions27 D;
- if (isEvenTimestep==true)
- {
- D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
- }
- else
- {
- D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
- }
- //////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
- real f1_E = (D.f[DIR_P00 ])[k1e ];
- real f1_W = (D.f[DIR_M00 ])[k1w ];
- real f1_N = (D.f[DIR_0P0 ])[k1n ];
- real f1_S = (D.f[DIR_0M0 ])[k1s ];
- real f1_T = (D.f[DIR_00P ])[k1t ];
- real f1_B = (D.f[DIR_00M ])[k1b ];
- real f1_NE = (D.f[DIR_PP0 ])[k1ne ];
- real f1_SW = (D.f[DIR_MM0 ])[k1sw ];
- real f1_SE = (D.f[DIR_PM0 ])[k1se ];
- real f1_NW = (D.f[DIR_MP0 ])[k1nw ];
- real f1_TE = (D.f[DIR_P0P ])[k1te ];
- real f1_BW = (D.f[DIR_M0M ])[k1bw ];
- real f1_BE = (D.f[DIR_P0M ])[k1be ];
- real f1_TW = (D.f[DIR_M0P ])[k1tw ];
- real f1_TN = (D.f[DIR_0PP ])[k1tn ];
- real f1_BS = (D.f[DIR_0MM ])[k1bs ];
- real f1_BN = (D.f[DIR_0PM ])[k1bn ];
- real f1_TS = (D.f[DIR_0MP ])[k1ts ];
- //real f1_ZERO = (D.f[DIR_000])[k1zero];
- real f1_TNE = (D.f[DIR_PPP ])[k1tne ];
- real f1_TSW = (D.f[DIR_MMP ])[k1tsw ];
- real f1_TSE = (D.f[DIR_PMP ])[k1tse ];
- real f1_TNW = (D.f[DIR_MPP ])[k1tnw ];
- real f1_BNE = (D.f[DIR_PPM ])[k1bne ];
- real f1_BSW = (D.f[DIR_MMM ])[k1bsw ];
- real f1_BSE = (D.f[DIR_PMM ])[k1bse ];
- real f1_BNW = (D.f[DIR_MPM ])[k1bnw ];
- //////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
- real f_E = (D.f[DIR_P00 ])[ke ];
- real f_W = (D.f[DIR_M00 ])[kw ];
- real f_N = (D.f[DIR_0P0 ])[kn ];
- real f_S = (D.f[DIR_0M0 ])[ks ];
- real f_T = (D.f[DIR_00P ])[kt ];
- real f_B = (D.f[DIR_00M ])[kb ];
- real f_NE = (D.f[DIR_PP0 ])[kne ];
- real f_SW = (D.f[DIR_MM0 ])[ksw ];
- real f_SE = (D.f[DIR_PM0 ])[kse ];
- real f_NW = (D.f[DIR_MP0 ])[knw ];
- real f_TE = (D.f[DIR_P0P ])[kte ];
- real f_BW = (D.f[DIR_M0M ])[kbw ];
- real f_BE = (D.f[DIR_P0M ])[kbe ];
- real f_TW = (D.f[DIR_M0P ])[ktw ];
- real f_TN = (D.f[DIR_0PP ])[ktn ];
- real f_BS = (D.f[DIR_0MM ])[kbs ];
- real f_BN = (D.f[DIR_0PM ])[kbn ];
- real f_TS = (D.f[DIR_0MP ])[kts ];
- //real f_ZERO = (D.f[DIR_000])[kzero];
- real f_TNE = (D.f[DIR_PPP ])[ktne ];
- real f_TSW = (D.f[DIR_MMP ])[ktsw ];
- real f_TSE = (D.f[DIR_PMP ])[ktse ];
- real f_TNW = (D.f[DIR_MPP ])[ktnw ];
- real f_BNE = (D.f[DIR_PPM ])[kbne ];
- real f_BSW = (D.f[DIR_MMM ])[kbsw ];
- real f_BSE = (D.f[DIR_PMM ])[kbse ];
- real f_BNW = (D.f[DIR_MPM ])[kbnw ];
- //////////////////////////////////////////////////////////////////////////
+ case MZZ:
+ (dist.f[DIR_P00])[ke ] = computeOutflowDistribution(f, f1, DIR_P00, cs);
+ (dist.f[DIR_PM0])[kse ] = computeOutflowDistribution(f, f1, DIR_PM0, cs);
+ (dist.f[DIR_PP0])[kne ] = computeOutflowDistribution(f, f1, DIR_PP0, cs);
+ (dist.f[DIR_P0M])[kbe ] = computeOutflowDistribution(f, f1, DIR_P0M, cs);
+ (dist.f[DIR_P0P])[kte ] = computeOutflowDistribution(f, f1, DIR_P0P, cs);
+ (dist.f[DIR_PMP])[ktse ] = computeOutflowDistribution(f, f1, DIR_PMP, cs);
+ (dist.f[DIR_PPP])[ktne ] = computeOutflowDistribution(f, f1, DIR_PPP, cs);
+ (dist.f[DIR_PMM])[kbse ] = computeOutflowDistribution(f, f1, DIR_PMM, cs);
+ (dist.f[DIR_PPM])[kbne ] = computeOutflowDistribution(f, f1, DIR_PPM, cs);
+ break;
+
+ case PZZ:
+ (dist.f[DIR_M00])[kw ] = computeOutflowDistribution(f, f1, DIR_M00, cs);
+ (dist.f[DIR_MM0])[ksw ] = computeOutflowDistribution(f, f1, DIR_MM0, cs);
+ (dist.f[DIR_MP0])[knw ] = computeOutflowDistribution(f, f1, DIR_MP0, cs);
+ (dist.f[DIR_M0M])[kbw ] = computeOutflowDistribution(f, f1, DIR_M0M, cs);
+ (dist.f[DIR_M0P])[ktw ] = computeOutflowDistribution(f, f1, DIR_M0P, cs);
+ (dist.f[DIR_MMP])[ktsw ] = computeOutflowDistribution(f, f1, DIR_MMP, cs);
+ (dist.f[DIR_MPP])[ktnw ] = computeOutflowDistribution(f, f1, DIR_MPP, cs);
+ (dist.f[DIR_MMM])[kbsw ] = computeOutflowDistribution(f, f1, DIR_MMM, cs);
+ (dist.f[DIR_MPM])[kbnw ] = computeOutflowDistribution(f, f1, DIR_MPM, cs);
+ break;
+
+ case ZMZ:
+ (dist.f[DIR_0P0])[kn ] = computeOutflowDistribution(f, f1, DIR_0P0, cs);
+ (dist.f[DIR_PP0])[kne ] = computeOutflowDistribution(f, f1, DIR_PP0, cs);
+ (dist.f[DIR_MP0])[knw ] = computeOutflowDistribution(f, f1, DIR_MP0, cs);
+ (dist.f[DIR_0PP])[ktn ] = computeOutflowDistribution(f, f1, DIR_0PP, cs);
+ (dist.f[DIR_0PM])[kbn ] = computeOutflowDistribution(f, f1, DIR_0PM, cs);
+ (dist.f[DIR_PPP])[ktne ] = computeOutflowDistribution(f, f1, DIR_PPP, cs);
+ (dist.f[DIR_MPP])[ktnw ] = computeOutflowDistribution(f, f1, DIR_MPP, cs);
+ (dist.f[DIR_PPM])[kbne ] = computeOutflowDistribution(f, f1, DIR_PPM, cs);
+ (dist.f[DIR_MPM])[kbnw ] = computeOutflowDistribution(f, f1, DIR_MPM, cs);
+ break;
+
+ case ZPZ:
+ (dist.f[DIR_0M0])[ks ] = computeOutflowDistribution(f, f1, DIR_0M0, cs);
+ (dist.f[DIR_PM0])[kse ] = computeOutflowDistribution(f, f1, DIR_PM0, cs);
+ (dist.f[DIR_MM0])[ksw ] = computeOutflowDistribution(f, f1, DIR_MM0, cs);
+ (dist.f[DIR_0MP])[kts ] = computeOutflowDistribution(f, f1, DIR_0MP, cs);
+ (dist.f[DIR_0MM])[kbs ] = computeOutflowDistribution(f, f1, DIR_0MM, cs);
+ (dist.f[DIR_PMP])[ktse ] = computeOutflowDistribution(f, f1, DIR_PMP, cs);
+ (dist.f[DIR_MMP])[ktsw ] = computeOutflowDistribution(f, f1, DIR_MMP, cs);
+ (dist.f[DIR_PMM])[kbse ] = computeOutflowDistribution(f, f1, DIR_PMM, cs);
+ (dist.f[DIR_MMM])[kbsw ] = computeOutflowDistribution(f, f1, DIR_MMM, cs);
+ break;
+
+ case ZZM:
+ (dist.f[DIR_00P])[kt ] = computeOutflowDistribution(f, f1, DIR_00P, cs);
+ (dist.f[DIR_P0P])[kte ] = computeOutflowDistribution(f, f1, DIR_P0P, cs);
+ (dist.f[DIR_M0P])[ktw ] = computeOutflowDistribution(f, f1, DIR_M0P, cs);
+ (dist.f[DIR_0PP])[ktn ] = computeOutflowDistribution(f, f1, DIR_0PP, cs);
+ (dist.f[DIR_0MP])[kts ] = computeOutflowDistribution(f, f1, DIR_0MP, cs);
+ (dist.f[DIR_PPP])[ktne ] = computeOutflowDistribution(f, f1, DIR_PPP, cs);
+ (dist.f[DIR_MPP])[ktnw ] = computeOutflowDistribution(f, f1, DIR_MPP, cs);
+ (dist.f[DIR_PMP])[ktse ] = computeOutflowDistribution(f, f1, DIR_PMP, cs);
+ (dist.f[DIR_MMP])[ktsw ] = computeOutflowDistribution(f, f1, DIR_MMP, cs);
+ break;
+
+ case ZZP:
+ (dist.f[DIR_00M])[kb ] = computeOutflowDistribution(f, f1, DIR_00M, cs);
+ (dist.f[DIR_P0M])[kbe ] = computeOutflowDistribution(f, f1, DIR_P0M, cs);
+ (dist.f[DIR_M0M])[kbw ] = computeOutflowDistribution(f, f1, DIR_M0M, cs);
+ (dist.f[DIR_0PM])[kbn ] = computeOutflowDistribution(f, f1, DIR_0PM, cs);
+ (dist.f[DIR_0MM])[kbs ] = computeOutflowDistribution(f, f1, DIR_0MM, cs);
+ (dist.f[DIR_PPM])[kbne ] = computeOutflowDistribution(f, f1, DIR_PPM, cs);
+ (dist.f[DIR_MPM])[kbnw ] = computeOutflowDistribution(f, f1, DIR_MPM, cs);
+ (dist.f[DIR_PMM])[kbse ] = computeOutflowDistribution(f, f1, DIR_PMM, cs);
+ (dist.f[DIR_MMM])[kbsw ] = computeOutflowDistribution(f, f1, DIR_MMM, cs);
+ break;
+ default:
+ break;
+ }
+}
+////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
- //real vx1, vx2, vx3, drho;
- //real vx1, vx2, vx3, drho, drho1;
- //////////////////////////////////////////////////////////////////////////
- //Dichte
- // drho1 = f1_TSE + f1_TNW + f1_TNE + f1_TSW + f1_BSE + f1_BNW + f1_BNE + f1_BSW +
- // f1_BN + f1_TS + f1_TN + f1_BS + f1_BE + f1_TW + f1_TE + f1_BW + f1_SE + f1_NW + f1_NE + f1_SW +
- // f1_T + f1_B + f1_N + f1_S + f1_E + f1_W + ((D.f[DIR_000])[k1zero]);
- // drho = f_TSE + f_TNW + f_TNE + f_TSW + f_BSE + f_BNW + f_BNE + f_BSW +
- // f_BN + f_TS + f_TN + f_BS + f_BE + f_TW + f_TE + f_BW + f_SE + f_NW + f_NE + f_SW +
- // f_T + f_B + f_N + f_S + f_E + f_W + ((D.f[DIR_000])[kzero]);
-
- //////////////////////////////////////////////////////////////////////////
- //Ux
- //vx1 = (((f_TSE - f_BNW) - (f_TNW - f_BSE)) + ((f_TNE - f_BSW) - (f_TSW - f_BNE)) +
- // ((f_BE - f_TW) + (f_TE - f_BW)) + ((f_SE - f_NW) + (f_NE - f_SW)) +
- // (f_E - f_W)) /(one + drho);
- // vx2 = ((-(f_TSE - f_BNW) + (f_TNW - f_BSE)) + ((f_TNE - f_BSW) - (f_TSW - f_BNE)) +
- // ((f_BN - f_TS) + (f_TN - f_BS)) + (-(f_SE - f_NW) + (f_NE - f_SW)) +
- // (f_N - f_S)) /(one + drho);
- // vx3 = (((f_TSE - f_BNW) + (f_TNW - f_BSE)) + ((f_TNE - f_BSW) + (f_TSW - f_BNE)) +
- // (-(f_BN - f_TS) + (f_TN - f_BS)) + ((f_TE - f_BW) - (f_BE - f_TW)) +
- // (f_T - f_B)) /(one + drho);
- //real cu_sq=c3o2*(vx1*vx1+vx2*vx2+vx3*vx3);
- // //////////////////////////////////////////////////////////////////////////
- ////real omega = om1;
- // real cusq = c3o2*(vx1*vx1+vx2*vx2+vx3*vx3);
- // //////////////////////////////////////////////////////////////////////////
- ////T�st MK
- ////if(vx1 < zero) vx1 = zero;
- // //////////////////////////////////////////////////////////////////////////
- // real fZERO = c8over27* (drho1-(one + drho1)*(cusq)) ;
- // real fE = c2over27* (drho1+(one + drho1)*(three*( vx1 )+c9over2*( vx1 )*( vx1 )-cusq));
- // real fW = c2over27* (drho1+(one + drho1)*(three*(-vx1 )+c9over2*(-vx1 )*(-vx1 )-cusq));
- // real fN = c2over27* (drho1+(one + drho1)*(three*( vx2 )+c9over2*( vx2 )*( vx2 )-cusq));
- // real fS = c2over27* (drho1+(one + drho1)*(three*( -vx2 )+c9over2*( -vx2 )*( -vx2 )-cusq));
- // real fT = c2over27* (drho1+(one + drho1)*(three*( vx3)+c9over2*( vx3)*( vx3)-cusq));
- // real fB = c2over27* (drho1+(one + drho1)*(three*( -vx3)+c9over2*( -vx3)*( -vx3)-cusq));
- // real fNE = c1over54* (drho1+(one + drho1)*(three*( vx1+vx2 )+c9over2*( vx1+vx2 )*( vx1+vx2 )-cusq));
- // real fSW = c1over54* (drho1+(one + drho1)*(three*(-vx1-vx2 )+c9over2*(-vx1-vx2 )*(-vx1-vx2 )-cusq));
- // real fSE = c1over54* (drho1+(one + drho1)*(three*( vx1-vx2 )+c9over2*( vx1-vx2 )*( vx1-vx2 )-cusq));
- // real fNW = c1over54* (drho1+(one + drho1)*(three*(-vx1+vx2 )+c9over2*(-vx1+vx2 )*(-vx1+vx2 )-cusq));
- // real fTE = c1over54* (drho1+(one + drho1)*(three*( vx1 +vx3)+c9over2*( vx1 +vx3)*( vx1 +vx3)-cusq));
- // real fBW = c1over54* (drho1+(one + drho1)*(three*(-vx1 -vx3)+c9over2*(-vx1 -vx3)*(-vx1 -vx3)-cusq));
- // real fBE = c1over54* (drho1+(one + drho1)*(three*( vx1 -vx3)+c9over2*( vx1 -vx3)*( vx1 -vx3)-cusq));
- // real fTW = c1over54* (drho1+(one + drho1)*(three*(-vx1 +vx3)+c9over2*(-vx1 +vx3)*(-vx1 +vx3)-cusq));
- // real fTN = c1over54* (drho1+(one + drho1)*(three*( vx2+vx3)+c9over2*( vx2+vx3)*( vx2+vx3)-cusq));
- // real fBS = c1over54* (drho1+(one + drho1)*(three*( -vx2-vx3)+c9over2*( -vx2-vx3)*( -vx2-vx3)-cusq));
- // real fBN = c1over54* (drho1+(one + drho1)*(three*( vx2-vx3)+c9over2*( vx2-vx3)*( vx2-vx3)-cusq));
- // real fTS = c1over54* (drho1+(one + drho1)*(three*( -vx2+vx3)+c9over2*( -vx2+vx3)*( -vx2+vx3)-cusq));
- // real fTNE = c1over216* (drho1+(one + drho1)*(three*( vx1+vx2+vx3)+c9over2*( vx1+vx2+vx3)*( vx1+vx2+vx3)-cusq));
- // real fBSW = c1over216* (drho1+(one + drho1)*(three*(-vx1-vx2-vx3)+c9over2*(-vx1-vx2-vx3)*(-vx1-vx2-vx3)-cusq));
- // real fBNE = c1over216* (drho1+(one + drho1)*(three*( vx1+vx2-vx3)+c9over2*( vx1+vx2-vx3)*( vx1+vx2-vx3)-cusq));
- // real fTSW = c1over216* (drho1+(one + drho1)*(three*(-vx1-vx2+vx3)+c9over2*(-vx1-vx2+vx3)*(-vx1-vx2+vx3)-cusq));
- // real fTSE = c1over216* (drho1+(one + drho1)*(three*( vx1-vx2+vx3)+c9over2*( vx1-vx2+vx3)*( vx1-vx2+vx3)-cusq));
- // real fBNW = c1over216* (drho1+(one + drho1)*(three*(-vx1+vx2-vx3)+c9over2*(-vx1+vx2-vx3)*(-vx1+vx2-vx3)-cusq));
- // real fBSE = c1over216* (drho1+(one + drho1)*(three*( vx1-vx2-vx3)+c9over2*( vx1-vx2-vx3)*( vx1-vx2-vx3)-cusq));
- // real fTNW = c1over216* (drho1+(one + drho1)*(three*(-vx1+vx2+vx3)+c9over2*(-vx1+vx2+vx3)*(-vx1+vx2+vx3)-cusq));
-
- real cs = c1o1 / sqrtf(c3o1);
- /////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
- //no velocity
- //////////////////////////////////////////
- f_E = f1_E * cs + (c1o1 - cs) * f_E ;
- f_W = f1_W * cs + (c1o1 - cs) * f_W ;
- f_N = f1_N * cs + (c1o1 - cs) * f_N ;
- f_S = f1_S * cs + (c1o1 - cs) * f_S ;
- f_T = f1_T * cs + (c1o1 - cs) * f_T ;
- f_B = f1_B * cs + (c1o1 - cs) * f_B ;
- f_NE = f1_NE * cs + (c1o1 - cs) * f_NE ;
- f_SW = f1_SW * cs + (c1o1 - cs) * f_SW ;
- f_SE = f1_SE * cs + (c1o1 - cs) * f_SE ;
- f_NW = f1_NW * cs + (c1o1 - cs) * f_NW ;
- f_TE = f1_TE * cs + (c1o1 - cs) * f_TE ;
- f_BW = f1_BW * cs + (c1o1 - cs) * f_BW ;
- f_BE = f1_BE * cs + (c1o1 - cs) * f_BE ;
- f_TW = f1_TW * cs + (c1o1 - cs) * f_TW ;
- f_TN = f1_TN * cs + (c1o1 - cs) * f_TN ;
- f_BS = f1_BS * cs + (c1o1 - cs) * f_BS ;
- f_BN = f1_BN * cs + (c1o1 - cs) * f_BN ;
- f_TS = f1_TS * cs + (c1o1 - cs) * f_TS ;
- f_TNE = f1_TNE * cs + (c1o1 - cs) * f_TNE ;
- f_TSW = f1_TSW * cs + (c1o1 - cs) * f_TSW ;
- f_TSE = f1_TSE * cs + (c1o1 - cs) * f_TSE ;
- f_TNW = f1_TNW * cs + (c1o1 - cs) * f_TNW ;
- f_BNE = f1_BNE * cs + (c1o1 - cs) * f_BNE ;
- f_BSW = f1_BSW * cs + (c1o1 - cs) * f_BSW ;
- f_BSE = f1_BSE * cs + (c1o1 - cs) * f_BSE ;
- f_BNW = f1_BNW * cs + (c1o1 - cs) * f_BNW ;
- /////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
-
- /////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
- //with velocity
- //if(true){//vx1 >= zero){
- // real csMvx = one / sqrtf(three) - vx1;
- // //real csMvy = one / sqrtf(three) - vx2;
- // ///////////////////////////////////////////
- // // X
- // f_W = f1_W * csMvx + (one - csMvx) * f_W ;//- c2over27 * ((drho + drho1)*c1o2-((drho + drho1)*c1o2 )*three*vx1);
- // f_NW = f1_NW * csMvx + (one - csMvx) * f_NW ;//- c1over54 * ((drho + drho1)*c1o2-((drho + drho1)*c1o2 )*three*vx1);
- // f_SW = f1_SW * csMvx + (one - csMvx) * f_SW ;//- c1over54 * ((drho + drho1)*c1o2-((drho + drho1)*c1o2 )*three*vx1);
- // f_TW = f1_TW * csMvx + (one - csMvx) * f_TW ;//- c1over54 * ((drho + drho1)*c1o2-((drho + drho1)*c1o2 )*three*vx1);
- // f_BW = f1_BW * csMvx + (one - csMvx) * f_BW ;//- c1over54 * ((drho + drho1)*c1o2-((drho + drho1)*c1o2 )*three*vx1);
- // f_TNW = f1_TNW * csMvx + (one - csMvx) * f_TNW ;//- c1over216 * ((drho + drho1)*c1o2-((drho + drho1)*c1o2 )*three*vx1);
- // f_TSW = f1_TSW * csMvx + (one - csMvx) * f_TSW ;//- c1over216 * ((drho + drho1)*c1o2-((drho + drho1)*c1o2 )*three*vx1);
- // f_BNW = f1_BNW * csMvx + (one - csMvx) * f_BNW ;//- c1over216 * ((drho + drho1)*c1o2-((drho + drho1)*c1o2 )*three*vx1);
- // f_BSW = f1_BSW * csMvx + (one - csMvx) * f_BSW ;//- c1over216 * ((drho + drho1)*c1o2-((drho + drho1)*c1o2 )*three*vx1);
- // ///////////////////////////////////////////
- // // Y
- // //f_S = f1_S * csMvy + (one - csMvy) * f_S ;//- c2over27 * ((drho + drho1)*c1o2-((drho + drho1)*c1o2 )*three*vx2);
- // //f_SE = f1_SE * csMvy + (one - csMvy) * f_SE ;//- c1over54 * ((drho + drho1)*c1o2-((drho + drho1)*c1o2 )*three*vx2);
- // //f_SW = f1_SW * csMvy + (one - csMvy) * f_SW ;//- c1over54 * ((drho + drho1)*c1o2-((drho + drho1)*c1o2 )*three*vx2);
- // //f_TS = f1_TS * csMvy + (one - csMvy) * f_TS ;//- c1over54 * ((drho + drho1)*c1o2-((drho + drho1)*c1o2 )*three*vx2);
- // //f_BS = f1_BS * csMvy + (one - csMvy) * f_BS ;//- c1over54 * ((drho + drho1)*c1o2-((drho + drho1)*c1o2 )*three*vx2);
- // //f_TSE = f1_TSE * csMvy + (one - csMvy) * f_TSE ;//- c1over216 * ((drho + drho1)*c1o2-((drho + drho1)*c1o2 )*three*vx2);
- // //f_TSW = f1_TSW * csMvy + (one - csMvy) * f_TSW ;//- c1over216 * ((drho + drho1)*c1o2-((drho + drho1)*c1o2 )*three*vx2);
- // //f_BSE = f1_BSE * csMvy + (one - csMvy) * f_BSE ;//- c1over216 * ((drho + drho1)*c1o2-((drho + drho1)*c1o2 )*three*vx2);
- // //f_BSW = f1_BSW * csMvy + (one - csMvy) * f_BSW ;//- c1over216 * ((drho + drho1)*c1o2-((drho + drho1)*c1o2 )*three*vx2);
- // //f_S = f1_S * csMvy + (one - csMvy) * f_S;
- // //f_SE = f1_SE * csMvy + (one - csMvy) * f_SE;
- // //f_SW = f1_SW * csMvy + (one - csMvy) * f_SW;
- // //f_TS = f1_TS * csMvy + (one - csMvy) * f_TS;
- // //f_BS = f1_BS * csMvy + (one - csMvy) * f_BS;
- // //f_TSE = f1_TSE * csMvy + (one - csMvy) * f_TSE;
- // //f_TSW = f1_TSW * csMvy + (one - csMvy) * f_TSW;
- // //f_BSE = f1_BSE * csMvy + (one - csMvy) * f_BSE;
- // //f_BSW = f1_BSW * csMvy + (one - csMvy) * f_BSW;
- // //////////////////////////////////////////////////////////////////////////
- //}
- //else
- //{
- // ///////////////////////////////////////////
- // // X
- // vx1 = vx1 * 0.9;
- // f_W = f_E - six * c2over27 * ( vx1 );
- // f_NW = f_SE - six * c1over54 * ( vx1-vx2 );
- // f_SW = f_NE - six * c1over54 * ( vx1+vx2 );
- // f_TW = f_BE - six * c1over54 * ( vx1 -vx3);
- // f_BW = f_TE - six * c1over54 * ( vx1 +vx3);
- // f_TNW = f_BSE - six * c1over216 * ( vx1-vx2-vx3);
- // f_TSW = f_BNE - six * c1over216 * ( vx1+vx2-vx3);
- // f_BNW = f_TSE - six * c1over216 * ( vx1-vx2+vx3);
- // f_BSW = f_TNE - six * c1over216 * ( vx1+vx2+vx3);
- // ///////////////////////////////////////////
- // // Y
- // //vx2 = vx2 * 0.9;
- // //f_S = f_N - six * c2over27 * ( vx2 );
- // //f_SE = f_NW - six * c1over54 * (-vx1+vx2 );
- // //f_SW = f_NE - six * c1over54 * ( vx1+vx2 );
- // //f_TS = f_BN - six * c1over54 * ( vx2-vx3);
- // //f_BS = f_TN - six * c1over54 * ( vx2+vx3);
- // //f_TSE = f_BNW - six * c1over216 * (-vx1+vx2-vx3);
- // //f_TSW = f_BNE - six * c1over216 * ( vx1+vx2-vx3);
- // //f_BSE = f_TNW - six * c1over216 * (-vx1+vx2+vx3);
- // //f_BSW = f_TNE - six * c1over216 * ( vx1+vx2+vx3);
- // ///////////////////////////////////////////
- //}
- /////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
-
- //////////////////////////////////////////////////////////////////////////
- if (isEvenTimestep==false)
- {
- D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
- }
- else
- {
- D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
- }
- //////////////////////////////////////////////////////////////////////////
- //__syncthreads();
- // -X
- //(D.f[DIR_P00 ])[ke ] = f_E ;
- //(D.f[DIR_PM0 ])[kse ] = f_SE ;
- //(D.f[DIR_PP0 ])[kne ] = f_NE ;
- //(D.f[DIR_P0M ])[kbe ] = f_BE ;
- //(D.f[DIR_P0P ])[kte ] = f_TE ;
- //(D.f[DIR_PMP ])[ktse ] = f_TSE ;
- //(D.f[DIR_PPP ])[ktne ] = f_TNE ;
- //(D.f[DIR_PMM ])[kbse ] = f_BSE ;
- //(D.f[DIR_PPM ])[kbne ] = f_BNE ;
- // X
- (D.f[DIR_M00 ])[kw ] = f_W ;
- (D.f[DIR_MM0 ])[ksw ] = f_SW ;
- (D.f[DIR_MP0 ])[knw ] = f_NW ;
- (D.f[DIR_M0M ])[kbw ] = f_BW ;
- (D.f[DIR_M0P ])[ktw ] = f_TW ;
- (D.f[DIR_MMP ])[ktsw ] = f_TSW ;
- (D.f[DIR_MPP ])[ktnw ] = f_TNW ;
- (D.f[DIR_MMM ])[kbsw ] = f_BSW ;
- (D.f[DIR_MPM ])[kbnw ] = f_BNW ;
- // Y
- //(D.f[DIR_0M0 ])[ks ] = f_S ;
- //(D.f[DIR_PM0 ])[kse ] = f_SE ;
- //(D.f[DIR_MM0 ])[ksw ] = f_SW ;
- //(D.f[DIR_0MP ])[kts ] = f_TS ;
- //(D.f[DIR_0MM ])[kbs ] = f_BS ;
- //(D.f[DIR_PMP ])[ktse ] = f_TSE ;
- //(D.f[DIR_MMP ])[ktsw ] = f_TSW ;
- //(D.f[DIR_PMM ])[kbse ] = f_BSE ;
- //(D.f[DIR_MMM ])[kbsw ] = f_BSW ;
- // Z
- //(D.f[DIR_00M ])[kb ] = f_B ;
- //(D.f[DIR_P0M ])[kbe ] = f_BE ;
- //(D.f[DIR_M0M ])[kbw ] = f_BW ;
- //(D.f[DIR_0PM ])[kbn ] = f_BN ;
- //(D.f[DIR_0MM ])[kbs ] = f_BS ;
- //(D.f[DIR_PPM ])[kbne ] = f_BNE ;
- //(D.f[DIR_MPM ])[kbnw ] = f_BNW ;
- //(D.f[DIR_PMM ])[kbse ] = f_BSE ;
- //(D.f[DIR_MMM ])[kbsw ] = f_BSW ;
- //////////////////////////////////////////////////////////////////////////
- }
-}
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
+__host__ __device__ real computeOutflowDistribution(const real* const &f, const real* const &f1, const int dir, const real rhoCorrection, const real cs, const real weight)
+{
+ return f1[dir ] * cs + (c1o1 - cs) * f[dir ] - weight *rhoCorrection;
+}
+
+__global__ void QPressZeroRhoOutflowDevice27(
+ real* rhoBC,
+ real* distributions,
+ int* k_Q,
+ int* k_N,
+ int numberOfBCnodes,
+ real om1,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep,
+ int direction,
+ real densityCorrectionFactor)
+{
+ ////////////////////////////////////////////////////////////////////////////////
+ //! - Get the node index coordinates from threadIdx, blockIdx, blockDim and gridDim.
+ //!
+ const unsigned nodeIndex = getNodeIndex();
+ //////////////////////////////////////////////////////////////////////////
+ if( nodeIndex >= numberOfBCnodes ) return;
+ ////////////////////////////////////////////////////////////////////////////////
+ //index
+ uint k_000 = k_Q[nodeIndex];
+ uint k_M00 = neighborX[k_000];
+ uint k_0M0 = neighborY[k_000];
+ uint k_00M = neighborZ[k_000];
+ uint k_MM0 = neighborY[k_M00];
+ uint k_M0M = neighborZ[k_M00];
+ uint k_0MM = neighborZ[k_0M0];
+ uint k_MMM = neighborZ[k_MM0];
+ ////////////////////////////////////////////////////////////////////////////////
+ //index of neighbor
+ uint kN_000 = k_N[nodeIndex];
+ uint kN_M00 = neighborX[k_000];
+ uint kN_0M0 = neighborY[k_000];
+ uint kN_00M = neighborZ[k_000];
+ uint kN_MM0 = neighborY[k_M00];
+ uint kN_M0M = neighborZ[k_M00];
+ uint kN_0MM = neighborZ[k_0M0];
+ uint kN_MMM = neighborZ[k_MM0];
+ ////////////////////////////////////////////////////////////////////////////////
+ Distributions27 dist;
+ getPointersToDistributions(dist, distributions, numberOfLBnodes, isEvenTimestep);
+ real f[27], fN[27];
+ //////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
+ f[DIR_000] = (dist.f[DIR_000])[k_000];
+ f[DIR_P00] = (dist.f[DIR_P00])[k_000];
+ f[DIR_M00] = (dist.f[DIR_M00])[k_M00];
+ f[DIR_0P0] = (dist.f[DIR_0P0])[k_000];
+ f[DIR_0M0] = (dist.f[DIR_0M0])[k_0M0];
+ f[DIR_00P] = (dist.f[DIR_00P])[k_000];
+ f[DIR_00M] = (dist.f[DIR_00M])[k_00M];
+ f[DIR_PP0] = (dist.f[DIR_PP0])[k_000];
+ f[DIR_MM0] = (dist.f[DIR_MM0])[k_MM0];
+ f[DIR_PM0] = (dist.f[DIR_PM0])[k_0M0];
+ f[DIR_MP0] = (dist.f[DIR_MP0])[k_M00];
+ f[DIR_P0P] = (dist.f[DIR_P0P])[k_000];
+ f[DIR_M0M] = (dist.f[DIR_M0M])[k_M0M];
+ f[DIR_P0M] = (dist.f[DIR_P0M])[k_00M];
+ f[DIR_M0P] = (dist.f[DIR_M0P])[k_M00];
+ f[DIR_0PP] = (dist.f[DIR_0PP])[k_000];
+ f[DIR_0MM] = (dist.f[DIR_0MM])[k_0MM];
+ f[DIR_0PM] = (dist.f[DIR_0PM])[k_00M];
+ f[DIR_0MP] = (dist.f[DIR_0MP])[k_0M0];
+ f[DIR_PPP] = (dist.f[DIR_PPP])[k_000];
+ f[DIR_MPP] = (dist.f[DIR_MPP])[k_M00];
+ f[DIR_PMP] = (dist.f[DIR_PMP])[k_0M0];
+ f[DIR_MMP] = (dist.f[DIR_MMP])[k_MM0];
+ f[DIR_PPM] = (dist.f[DIR_PPM])[k_00M];
+ f[DIR_MPM] = (dist.f[DIR_MPM])[k_M0M];
+ f[DIR_PMM] = (dist.f[DIR_PMM])[k_0MM];
+ f[DIR_MMM] = (dist.f[DIR_MMM])[k_MMM];
+ //////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
+ fN[DIR_000] = (dist.f[DIR_000])[kN_000];
+ fN[DIR_P00] = (dist.f[DIR_P00])[kN_000];
+ fN[DIR_M00] = (dist.f[DIR_M00])[kN_M00];
+ fN[DIR_0P0] = (dist.f[DIR_0P0])[kN_000];
+ fN[DIR_0M0] = (dist.f[DIR_0M0])[kN_0M0];
+ fN[DIR_00P] = (dist.f[DIR_00P])[kN_000];
+ fN[DIR_00M] = (dist.f[DIR_00M])[kN_00M];
+ fN[DIR_PP0] = (dist.f[DIR_PP0])[kN_000];
+ fN[DIR_MM0] = (dist.f[DIR_MM0])[kN_MM0];
+ fN[DIR_PM0] = (dist.f[DIR_PM0])[kN_0M0];
+ fN[DIR_MP0] = (dist.f[DIR_MP0])[kN_M00];
+ fN[DIR_P0P] = (dist.f[DIR_P0P])[kN_000];
+ fN[DIR_M0M] = (dist.f[DIR_M0M])[kN_M0M];
+ fN[DIR_P0M] = (dist.f[DIR_P0M])[kN_00M];
+ fN[DIR_M0P] = (dist.f[DIR_M0P])[kN_M00];
+ fN[DIR_0PP] = (dist.f[DIR_0PP])[kN_000];
+ fN[DIR_0MM] = (dist.f[DIR_0MM])[kN_0MM];
+ fN[DIR_0PM] = (dist.f[DIR_0PM])[kN_00M];
+ fN[DIR_0MP] = (dist.f[DIR_0MP])[kN_0M0];
+ fN[DIR_PPP] = (dist.f[DIR_PPP])[kN_000];
+ fN[DIR_MPP] = (dist.f[DIR_MPP])[kN_M00];
+ fN[DIR_PMP] = (dist.f[DIR_PMP])[kN_0M0];
+ fN[DIR_MMP] = (dist.f[DIR_MMP])[kN_MM0];
+ fN[DIR_PPM] = (dist.f[DIR_PPM])[kN_00M];
+ fN[DIR_MPM] = (dist.f[DIR_MPM])[kN_M0M];
+ fN[DIR_PMM] = (dist.f[DIR_PMM])[kN_0MM];
+ fN[DIR_MMM] = (dist.f[DIR_MMM])[kN_MMM];
+ //////////////////////////////////////////////////////////////////////////
+ real drho = vf::lbm::getDensity(f);
+ real rhoCorrection = densityCorrectionFactor*drho;
+ real cs = c1o1 / sqrtf(c3o1);
+ getPointersToDistributions(dist, distributions, numberOfLBnodes, !isEvenTimestep);
+ switch(direction)
+ {
+ case MZZ:
+ (dist.f[DIR_P00])[k_000] = computeOutflowDistribution(f, fN, DIR_P00 , rhoCorrection, cs, c2o27);
+ (dist.f[DIR_PM0])[k_0M0] = computeOutflowDistribution(f, fN, DIR_PM0, rhoCorrection, cs, c1o54);
+ (dist.f[DIR_PP0])[k_000] = computeOutflowDistribution(f, fN, DIR_PP0, rhoCorrection, cs, c1o54);
+ (dist.f[DIR_P0M])[k_00M] = computeOutflowDistribution(f, fN, DIR_P0M, rhoCorrection, cs, c1o54);
+ (dist.f[DIR_P0P])[k_000] = computeOutflowDistribution(f, fN, DIR_P0P, rhoCorrection, cs, c1o54);
+ (dist.f[DIR_PMP])[k_0M0] = computeOutflowDistribution(f, fN, DIR_PMP, rhoCorrection, cs, c1o216);
+ (dist.f[DIR_PPP])[k_000] = computeOutflowDistribution(f, fN, DIR_PPP, rhoCorrection, cs, c1o216);
+ (dist.f[DIR_PMM])[k_0MM] = computeOutflowDistribution(f, fN, DIR_PMM, rhoCorrection, cs, c1o216);
+ (dist.f[DIR_PPM])[k_00M] = computeOutflowDistribution(f, fN, DIR_PPM, rhoCorrection, cs, c1o216);
+ break;
+
+ case PZZ:
+ (dist.f[DIR_M00])[k_M00] = computeOutflowDistribution(f, fN, DIR_M00, rhoCorrection, cs, c2o27);
+ (dist.f[DIR_MM0])[k_MM0] = computeOutflowDistribution(f, fN, DIR_MM0, rhoCorrection, cs, c1o54);
+ (dist.f[DIR_MP0])[k_M00] = computeOutflowDistribution(f, fN, DIR_MP0, rhoCorrection, cs, c1o54);
+ (dist.f[DIR_M0M])[k_M0M] = computeOutflowDistribution(f, fN, DIR_M0M, rhoCorrection, cs, c1o54);
+ (dist.f[DIR_M0P])[k_M00] = computeOutflowDistribution(f, fN, DIR_M0P, rhoCorrection, cs, c1o54);
+ (dist.f[DIR_MMP])[k_MM0] = computeOutflowDistribution(f, fN, DIR_MMP, rhoCorrection, cs, c1o216);
+ (dist.f[DIR_MPP])[k_M00] = computeOutflowDistribution(f, fN, DIR_MPP, rhoCorrection, cs, c1o216);
+ (dist.f[DIR_MMM])[k_MMM] = computeOutflowDistribution(f, fN, DIR_MMM, rhoCorrection, cs, c1o216);
+ (dist.f[DIR_MPM])[k_M0M] = computeOutflowDistribution(f, fN, DIR_MPM, rhoCorrection, cs, c1o216);
+ break;
+
+ case ZMZ:
+ (dist.f[DIR_0P0])[k_000] = computeOutflowDistribution(f, fN, DIR_0P0, rhoCorrection, cs, c2o27);
+ (dist.f[DIR_PP0])[k_000] = computeOutflowDistribution(f, fN, DIR_PP0, rhoCorrection, cs, c1o54);
+ (dist.f[DIR_MP0])[k_M00] = computeOutflowDistribution(f, fN, DIR_MP0, rhoCorrection, cs, c1o54);
+ (dist.f[DIR_0PP])[k_000] = computeOutflowDistribution(f, fN, DIR_0PP, rhoCorrection, cs, c1o54);
+ (dist.f[DIR_0PM])[k_00M] = computeOutflowDistribution(f, fN, DIR_0PM, rhoCorrection, cs, c1o54);
+ (dist.f[DIR_PPP])[k_000] = computeOutflowDistribution(f, fN, DIR_PPP, rhoCorrection, cs, c1o216);
+ (dist.f[DIR_MPP])[k_M00] = computeOutflowDistribution(f, fN, DIR_MPP, rhoCorrection, cs, c1o216);
+ (dist.f[DIR_PPM])[k_00M] = computeOutflowDistribution(f, fN, DIR_PPM, rhoCorrection, cs, c1o216);
+ (dist.f[DIR_MPM])[k_M0M] = computeOutflowDistribution(f, fN, DIR_MPM, rhoCorrection, cs, c1o216);
+ break;
+
+ case ZPZ:
+ (dist.f[DIR_0M0])[k_0M0] =computeOutflowDistribution(f, fN, DIR_0M0, rhoCorrection, cs, c2o27);
+ (dist.f[DIR_PM0])[k_0M0] =computeOutflowDistribution(f, fN, DIR_PM0, rhoCorrection, cs, c1o54);
+ (dist.f[DIR_MM0])[k_MM0] =computeOutflowDistribution(f, fN, DIR_MM0, rhoCorrection, cs, c1o54);
+ (dist.f[DIR_0MP])[k_0M0] =computeOutflowDistribution(f, fN, DIR_0MP, rhoCorrection, cs, c1o54);
+ (dist.f[DIR_0MM])[k_0MM] =computeOutflowDistribution(f, fN, DIR_0MM, rhoCorrection, cs, c1o54);
+ (dist.f[DIR_PMP])[k_0M0] =computeOutflowDistribution(f, fN, DIR_PMP, rhoCorrection, cs, c1o216);
+ (dist.f[DIR_MMP])[k_MM0] =computeOutflowDistribution(f, fN, DIR_MMP, rhoCorrection, cs, c1o216);
+ (dist.f[DIR_PMM])[k_0MM] =computeOutflowDistribution(f, fN, DIR_PMM, rhoCorrection, cs, c1o216);
+ (dist.f[DIR_MMM])[k_MMM] =computeOutflowDistribution(f, fN, DIR_MMM, rhoCorrection, cs, c1o216);
+ break;
+
+ case ZZM:
+ (dist.f[DIR_00P])[k_000] = computeOutflowDistribution(f, fN, DIR_00P, rhoCorrection, cs, c2o27);
+ (dist.f[DIR_P0P])[k_000] = computeOutflowDistribution(f, fN, DIR_P0P, rhoCorrection, cs, c1o54);
+ (dist.f[DIR_M0P])[k_M00] = computeOutflowDistribution(f, fN, DIR_M0P, rhoCorrection, cs, c1o54);
+ (dist.f[DIR_0PP])[k_000] = computeOutflowDistribution(f, fN, DIR_0PP, rhoCorrection, cs, c1o54);
+ (dist.f[DIR_0MP])[k_0M0] = computeOutflowDistribution(f, fN, DIR_0MP, rhoCorrection, cs, c1o54);
+ (dist.f[DIR_PPP])[k_000] = computeOutflowDistribution(f, fN, DIR_PPP, rhoCorrection, cs, c1o216);
+ (dist.f[DIR_MPP])[k_M00] = computeOutflowDistribution(f, fN, DIR_MPP, rhoCorrection, cs, c1o216);
+ (dist.f[DIR_PMP])[k_0M0] = computeOutflowDistribution(f, fN, DIR_PMP, rhoCorrection, cs, c1o216);
+ (dist.f[DIR_MMP])[k_MM0] = computeOutflowDistribution(f, fN, DIR_MMP, rhoCorrection, cs, c1o216);
+ break;
+
+ case ZZP:
+ (dist.f[DIR_00M])[k_00M] = computeOutflowDistribution(f, fN, DIR_00M, rhoCorrection, cs, c2o27);
+ (dist.f[DIR_P0M])[k_00M] = computeOutflowDistribution(f, fN, DIR_P0M, rhoCorrection, cs, c1o54);
+ (dist.f[DIR_M0M])[k_M0M] = computeOutflowDistribution(f, fN, DIR_M0M, rhoCorrection, cs, c1o54);
+ (dist.f[DIR_0PM])[k_00M] = computeOutflowDistribution(f, fN, DIR_0PM, rhoCorrection, cs, c1o54);
+ (dist.f[DIR_0MM])[k_0MM] = computeOutflowDistribution(f, fN, DIR_0MM, rhoCorrection, cs, c1o54);
+ (dist.f[DIR_PPM])[k_00M] = computeOutflowDistribution(f, fN, DIR_PPM, rhoCorrection, cs, c1o216);
+ (dist.f[DIR_MPM])[k_M0M] = computeOutflowDistribution(f, fN, DIR_MPM, rhoCorrection, cs, c1o216);
+ (dist.f[DIR_PMM])[k_0MM] = computeOutflowDistribution(f, fN, DIR_PMM, rhoCorrection, cs, c1o216);
+ (dist.f[DIR_MMM])[k_MMM] = computeOutflowDistribution(f, fN, DIR_MMM, rhoCorrection, cs, c1o216);
+ break;
+ default:
+ break;
+ }
+}
+////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
@@ -3314,22 +3327,23 @@ __global__ void QPressNoRhoDevice27( real* rhoBC,
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
-__global__ void QPressDeviceOld27(real* rhoBC,
- real* DD,
- int* k_Q,
- int* k_N,
- int numberOfBCnodes,
- real om1,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep)
+__global__ void QPressDeviceOld27(
+ real* rhoBC,
+ real* DD,
+ int* k_Q,
+ int* k_N,
+ int numberOfBCnodes,
+ real om1,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep)
{
////////////////////////////////////////////////////////////////////////////////
- const unsigned x = threadIdx.x; // Globaler x-Index
- const unsigned y = blockIdx.x; // Globaler y-Index
- const unsigned z = blockIdx.y; // Globaler z-Index
+ const unsigned x = threadIdx.x; // Globaler x-Index
+ const unsigned y = blockIdx.x; // Globaler y-Index
+ const unsigned z = blockIdx.y; // Globaler z-Index
const unsigned nx = blockDim.x;
const unsigned ny = gridDim.x;
@@ -3403,133 +3417,133 @@ __global__ void QPressDeviceOld27(real* rhoBC,
Distributions27 D;
if (isEvenTimestep==false)
{
- D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
- }
+ D.f[DIR_P00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_MPM * numberOfLBnodes];
+ }
else
{
- D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
+ D.f[DIR_M00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_MPM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_PMP * numberOfLBnodes];
}
//////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
real f1_E,f1_W,f1_N,f1_S,f1_T,f1_B,f1_NE,f1_SW,f1_SE,f1_NW,f1_TE,f1_BW,f1_BE,f1_TW,f1_TN,f1_BS,f1_BN,f1_TS,f1_ZERO,
f1_TNE,f1_TSW,f1_TSE,f1_TNW,f1_BNE,f1_BSW,f1_BSE,f1_BNW;
- f1_W = (D.f[DIR_P00 ])[k1e ];
- f1_E = (D.f[DIR_M00 ])[k1w ];
- f1_S = (D.f[DIR_0P0 ])[k1n ];
- f1_N = (D.f[DIR_0M0 ])[k1s ];
- f1_B = (D.f[DIR_00P ])[k1t ];
- f1_T = (D.f[DIR_00M ])[k1b ];
- f1_SW = (D.f[DIR_PP0 ])[k1ne ];
- f1_NE = (D.f[DIR_MM0 ])[k1sw ];
- f1_NW = (D.f[DIR_PM0 ])[k1se ];
- f1_SE = (D.f[DIR_MP0 ])[k1nw ];
- f1_BW = (D.f[DIR_P0P ])[k1te ];
- f1_TE = (D.f[DIR_M0M ])[k1bw ];
- f1_TW = (D.f[DIR_P0M ])[k1be ];
- f1_BE = (D.f[DIR_M0P ])[k1tw ];
- f1_BS = (D.f[DIR_0PP ])[k1tn ];
- f1_TN = (D.f[DIR_0MM ])[k1bs ];
- f1_TS = (D.f[DIR_0PM ])[k1bn ];
- f1_BN = (D.f[DIR_0MP ])[k1ts ];
+ f1_W = (D.f[DIR_P00])[k1e ];
+ f1_E = (D.f[DIR_M00])[k1w ];
+ f1_S = (D.f[DIR_0P0])[k1n ];
+ f1_N = (D.f[DIR_0M0])[k1s ];
+ f1_B = (D.f[DIR_00P])[k1t ];
+ f1_T = (D.f[DIR_00M])[k1b ];
+ f1_SW = (D.f[DIR_PP0])[k1ne ];
+ f1_NE = (D.f[DIR_MM0])[k1sw ];
+ f1_NW = (D.f[DIR_PM0])[k1se ];
+ f1_SE = (D.f[DIR_MP0])[k1nw ];
+ f1_BW = (D.f[DIR_P0P])[k1te ];
+ f1_TE = (D.f[DIR_M0M])[k1bw ];
+ f1_TW = (D.f[DIR_P0M])[k1be ];
+ f1_BE = (D.f[DIR_M0P])[k1tw ];
+ f1_BS = (D.f[DIR_0PP])[k1tn ];
+ f1_TN = (D.f[DIR_0MM])[k1bs ];
+ f1_TS = (D.f[DIR_0PM])[k1bn ];
+ f1_BN = (D.f[DIR_0MP])[k1ts ];
f1_ZERO = (D.f[DIR_000])[k1zero];
- f1_BSW = (D.f[DIR_PPP ])[k1tne ];
- f1_BNE = (D.f[DIR_MMP ])[k1tsw ];
- f1_BNW = (D.f[DIR_PMP ])[k1tse ];
- f1_BSE = (D.f[DIR_MPP ])[k1tnw ];
- f1_TSW = (D.f[DIR_PPM ])[k1bne ];
- f1_TNE = (D.f[DIR_MMM ])[k1bsw ];
- f1_TNW = (D.f[DIR_PMM ])[k1bse ];
- f1_TSE = (D.f[DIR_MPM ])[k1bnw ];
+ f1_BSW = (D.f[DIR_PPP])[k1tne ];
+ f1_BNE = (D.f[DIR_MMP])[k1tsw ];
+ f1_BNW = (D.f[DIR_PMP])[k1tse ];
+ f1_BSE = (D.f[DIR_MPP])[k1tnw ];
+ f1_TSW = (D.f[DIR_PPM])[k1bne ];
+ f1_TNE = (D.f[DIR_MMM])[k1bsw ];
+ f1_TNW = (D.f[DIR_PMM])[k1bse ];
+ f1_TSE = (D.f[DIR_MPM])[k1bnw ];
//////////////////////////////////////////////////////////////////////////
real drho1 = f1_ZERO+f1_E+f1_W+f1_N+f1_S+f1_T+f1_B+f1_NE+f1_SW+f1_SE+f1_NW+f1_TE+f1_BW+f1_BE+f1_TW+f1_TN+f1_BS+f1_BN+f1_TS+
f1_TNE+f1_TSW+f1_TSE+f1_TNW+f1_BNE+f1_BSW+f1_BSE+f1_BNW;
- //drho1 = (drho1 + rhoBC[k])/2.f;
- drho1 = drho1 - rhoBC[k];
+ //drho1 = (drho1 + rhoBC[k])/2.f;
+ drho1 = drho1 - rhoBC[k];
//////////////////////////////////////////////////////////////////////////
__syncthreads();
- (D.f[DIR_P00 ])[ke ] = f1_W -c2o27*drho1; // c1o100; // zero; //
- (D.f[DIR_M00 ])[kw ] = f1_E -c2o27*drho1; // c1o100; // zero; //
- (D.f[DIR_0P0 ])[kn ] = f1_S -c2o27*drho1; // c1o100; // zero; //
- (D.f[DIR_0M0 ])[ks ] = f1_N -c2o27*drho1; // c1o100; // zero; //
- (D.f[DIR_00P ])[kt ] = f1_B -c2o27*drho1; // c1o100; // zero; //
- (D.f[DIR_00M ])[kb ] = f1_T -c2o27*drho1; // c1o100; // zero; //
- (D.f[DIR_PP0 ])[kne ] = f1_SW -c1o54*drho1; // c1o100; // zero; //
- (D.f[DIR_MM0 ])[ksw ] = f1_NE -c1o54*drho1; // c1o100; // zero; //
- (D.f[DIR_PM0 ])[kse ] = f1_NW -c1o54*drho1; // c1o100; // zero; //
- (D.f[DIR_MP0 ])[knw ] = f1_SE -c1o54*drho1; // c1o100; // zero; //
- (D.f[DIR_P0P ])[kte ] = f1_BW -c1o54*drho1; // c1o100; // zero; //
- (D.f[DIR_M0M ])[kbw ] = f1_TE -c1o54*drho1; // c1o100; // zero; //
- (D.f[DIR_P0M ])[kbe ] = f1_TW -c1o54*drho1; // c1o100; // zero; //
- (D.f[DIR_M0P ])[ktw ] = f1_BE -c1o54*drho1; // c1o100; // zero; //
- (D.f[DIR_0PP ])[ktn ] = f1_BS -c1o54*drho1; // c1o100; // zero; //
- (D.f[DIR_0MM ])[kbs ] = f1_TN -c1o54*drho1; // c1o100; // zero; //
- (D.f[DIR_0PM ])[kbn ] = f1_TS -c1o54*drho1; // c1o100; // zero; //
- (D.f[DIR_0MP ])[kts ] = f1_BN -c1o54*drho1; // c1o100; // zero; //
+ (D.f[DIR_P00])[ke ] = f1_W -c2o27*drho1; // c1o100; // zero; //
+ (D.f[DIR_M00])[kw ] = f1_E -c2o27*drho1; // c1o100; // zero; //
+ (D.f[DIR_0P0])[kn ] = f1_S -c2o27*drho1; // c1o100; // zero; //
+ (D.f[DIR_0M0])[ks ] = f1_N -c2o27*drho1; // c1o100; // zero; //
+ (D.f[DIR_00P])[kt ] = f1_B -c2o27*drho1; // c1o100; // zero; //
+ (D.f[DIR_00M])[kb ] = f1_T -c2o27*drho1; // c1o100; // zero; //
+ (D.f[DIR_PP0])[kne ] = f1_SW -c1o54*drho1; // c1o100; // zero; //
+ (D.f[DIR_MM0])[ksw ] = f1_NE -c1o54*drho1; // c1o100; // zero; //
+ (D.f[DIR_PM0])[kse ] = f1_NW -c1o54*drho1; // c1o100; // zero; //
+ (D.f[DIR_MP0])[knw ] = f1_SE -c1o54*drho1; // c1o100; // zero; //
+ (D.f[DIR_P0P])[kte ] = f1_BW -c1o54*drho1; // c1o100; // zero; //
+ (D.f[DIR_M0M])[kbw ] = f1_TE -c1o54*drho1; // c1o100; // zero; //
+ (D.f[DIR_P0M])[kbe ] = f1_TW -c1o54*drho1; // c1o100; // zero; //
+ (D.f[DIR_M0P])[ktw ] = f1_BE -c1o54*drho1; // c1o100; // zero; //
+ (D.f[DIR_0PP])[ktn ] = f1_BS -c1o54*drho1; // c1o100; // zero; //
+ (D.f[DIR_0MM])[kbs ] = f1_TN -c1o54*drho1; // c1o100; // zero; //
+ (D.f[DIR_0PM])[kbn ] = f1_TS -c1o54*drho1; // c1o100; // zero; //
+ (D.f[DIR_0MP])[kts ] = f1_BN -c1o54*drho1; // c1o100; // zero; //
(D.f[DIR_000])[kzero] = f1_ZERO-c8o27*drho1; // c1o100; // zero; //
- (D.f[DIR_PPP ])[ktne ] = f1_BSW -c1o216*drho1; // c1o100; // zero; //
- (D.f[DIR_MMP ])[ktsw ] = f1_BNE -c1o216*drho1; // c1o100; // zero; //
- (D.f[DIR_PMP ])[ktse ] = f1_BNW -c1o216*drho1; // c1o100; // zero; //
- (D.f[DIR_MPP ])[ktnw ] = f1_BSE -c1o216*drho1; // c1o100; // zero; //
- (D.f[DIR_PPM ])[kbne ] = f1_TSW -c1o216*drho1; // c1o100; // zero; //
- (D.f[DIR_MMM ])[kbsw ] = f1_TNE -c1o216*drho1; // c1o100; // zero; //
- (D.f[DIR_PMM ])[kbse ] = f1_TNW -c1o216*drho1; // c1o100; // zero; //
- (D.f[DIR_MPM ])[kbnw ] = f1_TSE -c1o216*drho1; // c1o100; // zero; //
+ (D.f[DIR_PPP])[ktne ] = f1_BSW -c1o216*drho1; // c1o100; // zero; //
+ (D.f[DIR_MMP])[ktsw ] = f1_BNE -c1o216*drho1; // c1o100; // zero; //
+ (D.f[DIR_PMP])[ktse ] = f1_BNW -c1o216*drho1; // c1o100; // zero; //
+ (D.f[DIR_MPP])[ktnw ] = f1_BSE -c1o216*drho1; // c1o100; // zero; //
+ (D.f[DIR_PPM])[kbne ] = f1_TSW -c1o216*drho1; // c1o100; // zero; //
+ (D.f[DIR_MMM])[kbsw ] = f1_TNE -c1o216*drho1; // c1o100; // zero; //
+ (D.f[DIR_PMM])[kbse ] = f1_TNW -c1o216*drho1; // c1o100; // zero; //
+ (D.f[DIR_MPM])[kbnw ] = f1_TSE -c1o216*drho1; // c1o100; // zero; //
}
}
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
@@ -3573,23 +3587,24 @@ __global__ void QPressDeviceOld27(real* rhoBC,
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
-__global__ void QPressDeviceEQZ27(real* rhoBC,
- real* DD,
- int* k_Q,
- int* k_N,
- real* kTestRE,
- int numberOfBCnodes,
- real om1,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep)
+__global__ void QPressDeviceEQZ27(
+ real* rhoBC,
+ real* DD,
+ int* k_Q,
+ int* k_N,
+ real* kTestRE,
+ int numberOfBCnodes,
+ real om1,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep)
{
////////////////////////////////////////////////////////////////////////////////
- const unsigned x = threadIdx.x; // Globaler x-Index
- const unsigned y = blockIdx.x; // Globaler y-Index
- const unsigned z = blockIdx.y; // Globaler z-Index
+ const unsigned x = threadIdx.x; // Globaler x-Index
+ const unsigned y = blockIdx.x; // Globaler y-Index
+ const unsigned z = blockIdx.y; // Globaler z-Index
const unsigned nx = blockDim.x;
const unsigned ny = gridDim.x;
@@ -3663,168 +3678,168 @@ __global__ void QPressDeviceEQZ27(real* rhoBC,
Distributions27 D;
if (isEvenTimestep==true)
{
- D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
- }
+ D.f[DIR_P00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_MPM * numberOfLBnodes];
+ }
else
{
- D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
+ D.f[DIR_M00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_MPM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_PMP * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////
// Distributions27 kDistTest;
- // kDistTest.f[DIR_P00 ] = &kTestRE[DIR_P00 *numberOfBCnodes];
- // kDistTest.f[DIR_M00 ] = &kTestRE[DIR_M00 *numberOfBCnodes];
- // kDistTest.f[DIR_0P0 ] = &kTestRE[DIR_0P0 *numberOfBCnodes];
- // kDistTest.f[DIR_0M0 ] = &kTestRE[DIR_0M0 *numberOfBCnodes];
- // kDistTest.f[DIR_00P ] = &kTestRE[DIR_00P *numberOfBCnodes];
- // kDistTest.f[DIR_00M ] = &kTestRE[DIR_00M *numberOfBCnodes];
- // kDistTest.f[DIR_PP0 ] = &kTestRE[DIR_PP0 *numberOfBCnodes];
- // kDistTest.f[DIR_MM0 ] = &kTestRE[DIR_MM0 *numberOfBCnodes];
- // kDistTest.f[DIR_PM0 ] = &kTestRE[DIR_PM0 *numberOfBCnodes];
- // kDistTest.f[DIR_MP0 ] = &kTestRE[DIR_MP0 *numberOfBCnodes];
- // kDistTest.f[DIR_P0P ] = &kTestRE[DIR_P0P *numberOfBCnodes];
- // kDistTest.f[DIR_M0M ] = &kTestRE[DIR_M0M *numberOfBCnodes];
- // kDistTest.f[DIR_P0M ] = &kTestRE[DIR_P0M *numberOfBCnodes];
- // kDistTest.f[DIR_M0P ] = &kTestRE[DIR_M0P *numberOfBCnodes];
- // kDistTest.f[DIR_0PP ] = &kTestRE[DIR_0PP *numberOfBCnodes];
- // kDistTest.f[DIR_0MM ] = &kTestRE[DIR_0MM *numberOfBCnodes];
- // kDistTest.f[DIR_0PM ] = &kTestRE[DIR_0PM *numberOfBCnodes];
- // kDistTest.f[DIR_0MP ] = &kTestRE[DIR_0MP *numberOfBCnodes];
- // kDistTest.f[DIR_000] = &kTestRE[DIR_000*numberOfBCnodes];
- // kDistTest.f[DIR_PPP ] = &kTestRE[DIR_PPP *numberOfBCnodes];
- // kDistTest.f[DIR_MMP ] = &kTestRE[DIR_MMP *numberOfBCnodes];
- // kDistTest.f[DIR_PMP ] = &kTestRE[DIR_PMP *numberOfBCnodes];
- // kDistTest.f[DIR_MPP ] = &kTestRE[DIR_MPP *numberOfBCnodes];
- // kDistTest.f[DIR_PPM ] = &kTestRE[DIR_PPM *numberOfBCnodes];
- // kDistTest.f[DIR_MMM ] = &kTestRE[DIR_MMM *numberOfBCnodes];
- // kDistTest.f[DIR_PMM ] = &kTestRE[DIR_PMM *numberOfBCnodes];
- // kDistTest.f[DIR_MPM ] = &kTestRE[DIR_MPM *numberOfBCnodes];
+ // kDistTest.f[DIR_P00] = &kTestRE[DIR_P00 * numberOfBCnodes];
+ // kDistTest.f[DIR_M00] = &kTestRE[DIR_M00 * numberOfBCnodes];
+ // kDistTest.f[DIR_0P0] = &kTestRE[DIR_0P0 * numberOfBCnodes];
+ // kDistTest.f[DIR_0M0] = &kTestRE[DIR_0M0 * numberOfBCnodes];
+ // kDistTest.f[DIR_00P] = &kTestRE[DIR_00P * numberOfBCnodes];
+ // kDistTest.f[DIR_00M] = &kTestRE[DIR_00M * numberOfBCnodes];
+ // kDistTest.f[DIR_PP0] = &kTestRE[DIR_PP0 * numberOfBCnodes];
+ // kDistTest.f[DIR_MM0] = &kTestRE[DIR_MM0 * numberOfBCnodes];
+ // kDistTest.f[DIR_PM0] = &kTestRE[DIR_PM0 * numberOfBCnodes];
+ // kDistTest.f[DIR_MP0] = &kTestRE[DIR_MP0 * numberOfBCnodes];
+ // kDistTest.f[DIR_P0P] = &kTestRE[DIR_P0P * numberOfBCnodes];
+ // kDistTest.f[DIR_M0M] = &kTestRE[DIR_M0M * numberOfBCnodes];
+ // kDistTest.f[DIR_P0M] = &kTestRE[DIR_P0M * numberOfBCnodes];
+ // kDistTest.f[DIR_M0P] = &kTestRE[DIR_M0P * numberOfBCnodes];
+ // kDistTest.f[DIR_0PP] = &kTestRE[DIR_0PP * numberOfBCnodes];
+ // kDistTest.f[DIR_0MM] = &kTestRE[DIR_0MM * numberOfBCnodes];
+ // kDistTest.f[DIR_0PM] = &kTestRE[DIR_0PM * numberOfBCnodes];
+ // kDistTest.f[DIR_0MP] = &kTestRE[DIR_0MP * numberOfBCnodes];
+ // kDistTest.f[DIR_000] = &kTestRE[DIR_000 * numberOfBCnodes];
+ // kDistTest.f[DIR_PPP] = &kTestRE[DIR_PPP * numberOfBCnodes];
+ // kDistTest.f[DIR_MMP] = &kTestRE[DIR_MMP * numberOfBCnodes];
+ // kDistTest.f[DIR_PMP] = &kTestRE[DIR_PMP * numberOfBCnodes];
+ // kDistTest.f[DIR_MPP] = &kTestRE[DIR_MPP * numberOfBCnodes];
+ // kDistTest.f[DIR_PPM] = &kTestRE[DIR_PPM * numberOfBCnodes];
+ // kDistTest.f[DIR_MMM] = &kTestRE[DIR_MMM * numberOfBCnodes];
+ // kDistTest.f[DIR_PMM] = &kTestRE[DIR_PMM * numberOfBCnodes];
+ // kDistTest.f[DIR_MPM] = &kTestRE[DIR_MPM * numberOfBCnodes];
// ////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
// //real f1_E,f1_W,f1_N,f1_S,f1_T,f1_B,f1_NE,f1_SW,f1_SE,f1_NW,f1_TE,f1_BW,f1_BE,f1_TW,f1_TN,f1_BS,f1_BN,f1_TS,f1_ZERO,f1_TNE,f1_TSW,f1_TSE,f1_TNW,f1_BNE,f1_BSW,f1_BSE,f1_BNW;
- // //f1_W = (D.f[DIR_P00 ])[k1e ];
- // //f1_E = (D.f[DIR_M00 ])[k1w ];
- // //f1_S = (D.f[DIR_0P0 ])[k1n ];
- // //f1_N = (D.f[DIR_0M0 ])[k1s ];
- // //f1_B = (D.f[DIR_00P ])[k1t ];
- // //f1_T = (D.f[DIR_00M ])[k1b ];
- // //f1_SW = (D.f[DIR_PP0 ])[k1ne ];
- // //f1_NE = (D.f[DIR_MM0 ])[k1sw ];
- // //f1_NW = (D.f[DIR_PM0 ])[k1se ];
- // //f1_SE = (D.f[DIR_MP0 ])[k1nw ];
- // //f1_BW = (D.f[DIR_P0P ])[k1te ];
- // //f1_TE = (D.f[DIR_M0M ])[k1bw ];
- // //f1_TW = (D.f[DIR_P0M ])[k1be ];
- // //f1_BE = (D.f[DIR_M0P ])[k1tw ];
- // //f1_BS = (D.f[DIR_0PP ])[k1tn ];
- // //f1_TN = (D.f[DIR_0MM ])[k1bs ];
- // //f1_TS = (D.f[DIR_0PM ])[k1bn ];
- // //f1_BN = (D.f[DIR_0MP ])[k1ts ];
+ // //f1_W = (D.f[DIR_P00])[k1e ];
+ // //f1_E = (D.f[DIR_M00])[k1w ];
+ // //f1_S = (D.f[DIR_0P0])[k1n ];
+ // //f1_N = (D.f[DIR_0M0])[k1s ];
+ // //f1_B = (D.f[DIR_00P])[k1t ];
+ // //f1_T = (D.f[DIR_00M])[k1b ];
+ // //f1_SW = (D.f[DIR_PP0])[k1ne ];
+ // //f1_NE = (D.f[DIR_MM0])[k1sw ];
+ // //f1_NW = (D.f[DIR_PM0])[k1se ];
+ // //f1_SE = (D.f[DIR_MP0])[k1nw ];
+ // //f1_BW = (D.f[DIR_P0P])[k1te ];
+ // //f1_TE = (D.f[DIR_M0M])[k1bw ];
+ // //f1_TW = (D.f[DIR_P0M])[k1be ];
+ // //f1_BE = (D.f[DIR_M0P])[k1tw ];
+ // //f1_BS = (D.f[DIR_0PP])[k1tn ];
+ // //f1_TN = (D.f[DIR_0MM])[k1bs ];
+ // //f1_TS = (D.f[DIR_0PM])[k1bn ];
+ // //f1_BN = (D.f[DIR_0MP])[k1ts ];
// //f1_ZERO = (D.f[DIR_000])[k1zero];
- // //f1_BSW = (D.f[DIR_PPP ])[k1tne ];
- // //f1_BNE = (D.f[DIR_MMP ])[k1tsw ];
- // //f1_BNW = (D.f[DIR_PMP ])[k1tse ];
- // //f1_BSE = (D.f[DIR_MPP ])[k1tnw ];
- // //f1_TSW = (D.f[DIR_PPM ])[k1bne ];
- // //f1_TNE = (D.f[DIR_MMM ])[k1bsw ];
- // //f1_TNW = (D.f[DIR_PMM ])[k1bse ];
- // //f1_TSE = (D.f[DIR_MPM ])[k1bnw ];
+ // //f1_BSW = (D.f[DIR_PPP])[k1tne ];
+ // //f1_BNE = (D.f[DIR_MMP])[k1tsw ];
+ // //f1_BNW = (D.f[DIR_PMP])[k1tse ];
+ // //f1_BSE = (D.f[DIR_MPP])[k1tnw ];
+ // //f1_TSW = (D.f[DIR_PPM])[k1bne ];
+ // //f1_TNE = (D.f[DIR_MMM])[k1bsw ];
+ // //f1_TNW = (D.f[DIR_PMM])[k1bse ];
+ // //f1_TSE = (D.f[DIR_MPM])[k1bnw ];
// ////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
// //////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
// real f1_E,f1_W,f1_N,f1_S,f1_T,f1_B,f1_NE,f1_SW,f1_SE,f1_NW,f1_TE,f1_BW,f1_BE,f1_TW,f1_TN,f1_BS,f1_BN,f1_TS,f1_ZERO,f1_TNE,f1_TSW,f1_TSE,f1_TNW,f1_BNE,f1_BSW,f1_BSE,f1_BNW;
- // f1_E = (D.f[DIR_P00 ])[k1e ];
- // f1_W = (D.f[DIR_M00 ])[k1w ];
- // f1_N = (D.f[DIR_0P0 ])[k1n ];
- // f1_S = (D.f[DIR_0M0 ])[k1s ];
- // f1_T = (D.f[DIR_00P ])[k1t ];
- // f1_B = (D.f[DIR_00M ])[k1b ];
- // f1_NE = (D.f[DIR_PP0 ])[k1ne ];
- // f1_SW = (D.f[DIR_MM0 ])[k1sw ];
- // f1_SE = (D.f[DIR_PM0 ])[k1se ];
- // f1_NW = (D.f[DIR_MP0 ])[k1nw ];
- // f1_TE = (D.f[DIR_P0P ])[k1te ];
- // f1_BW = (D.f[DIR_M0M ])[k1bw ];
- // f1_BE = (D.f[DIR_P0M ])[k1be ];
- // f1_TW = (D.f[DIR_M0P ])[k1tw ];
- // f1_TN = (D.f[DIR_0PP ])[k1tn ];
- // f1_BS = (D.f[DIR_0MM ])[k1bs ];
- // f1_BN = (D.f[DIR_0PM ])[k1bn ];
- // f1_TS = (D.f[DIR_0MP ])[k1ts ];
+ // f1_E = (D.f[DIR_P00])[k1e ];
+ // f1_W = (D.f[DIR_M00])[k1w ];
+ // f1_N = (D.f[DIR_0P0])[k1n ];
+ // f1_S = (D.f[DIR_0M0])[k1s ];
+ // f1_T = (D.f[DIR_00P])[k1t ];
+ // f1_B = (D.f[DIR_00M])[k1b ];
+ // f1_NE = (D.f[DIR_PP0])[k1ne ];
+ // f1_SW = (D.f[DIR_MM0])[k1sw ];
+ // f1_SE = (D.f[DIR_PM0])[k1se ];
+ // f1_NW = (D.f[DIR_MP0])[k1nw ];
+ // f1_TE = (D.f[DIR_P0P])[k1te ];
+ // f1_BW = (D.f[DIR_M0M])[k1bw ];
+ // f1_BE = (D.f[DIR_P0M])[k1be ];
+ // f1_TW = (D.f[DIR_M0P])[k1tw ];
+ // f1_TN = (D.f[DIR_0PP])[k1tn ];
+ // f1_BS = (D.f[DIR_0MM])[k1bs ];
+ // f1_BN = (D.f[DIR_0PM])[k1bn ];
+ // f1_TS = (D.f[DIR_0MP])[k1ts ];
// f1_ZERO = (D.f[DIR_000])[k1zero];
- // f1_TNE = (D.f[DIR_PPP ])[k1tne ];
- // f1_TSW = (D.f[DIR_MMP ])[k1tsw ];
- // f1_TSE = (D.f[DIR_PMP ])[k1tse ];
- // f1_TNW = (D.f[DIR_MPP ])[k1tnw ];
- // f1_BNE = (D.f[DIR_PPM ])[k1bne ];
- // f1_BSW = (D.f[DIR_MMM ])[k1bsw ];
- // f1_BSE = (D.f[DIR_PMM ])[k1bse ];
- // f1_BNW = (D.f[DIR_MPM ])[k1bnw ];
+ // f1_TNE = (D.f[DIR_PPP])[k1tne ];
+ // f1_TSW = (D.f[DIR_MMP])[k1tsw ];
+ // f1_TSE = (D.f[DIR_PMP])[k1tse ];
+ // f1_TNW = (D.f[DIR_MPP])[k1tnw ];
+ // f1_BNE = (D.f[DIR_PPM])[k1bne ];
+ // f1_BSW = (D.f[DIR_MMM])[k1bsw ];
+ // f1_BSE = (D.f[DIR_PMM])[k1bse ];
+ // f1_BNW = (D.f[DIR_MPM])[k1bnw ];
// //////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
// //////////////////////////////////////////////////////////////////////////
// real drho1 = f1_ZERO+f1_E+f1_W+f1_N+f1_S+f1_T+f1_B+f1_NE+f1_SW+f1_SE+f1_NW+f1_TE+f1_BW+f1_BE+f1_TW+f1_TN+f1_BS+f1_BN+f1_TS+ f1_TNE+f1_TSW+f1_TSE+f1_TNW+f1_BNE+f1_BSW+f1_BSE+f1_BNW;
- //real vx1 = (((f1_TNE-f1_BSW)+(f1_BSE-f1_TNW)+(f1_BNE-f1_TSW)+(f1_TSE-f1_BNW)) + (((f1_NE-f1_SW)+(f1_TE-f1_BW))+((f1_SE-f1_NW)+(f1_BE-f1_TW))) + (f1_E-f1_W)) / (one + drho1);
- //real vx2 = (((f1_TNE-f1_BSW)+(f1_TNW-f1_BSE)+(f1_BNE-f1_TSW)+(f1_BNW-f1_TSE)) + (((f1_NE-f1_SW)+(f1_TN-f1_BS))+((f1_BN-f1_TS)+(f1_NW-f1_SE))) + (f1_N-f1_S)) / (one + drho1);
- //real vx3 = (((f1_TNE-f1_BSW)+(f1_TNW-f1_BSE)+(f1_TSW-f1_BNE)+(f1_TSE-f1_BNW)) + (((f1_TE-f1_BW)+(f1_TN-f1_BS))+((f1_TW-f1_BE)+(f1_TS-f1_BN))) + (f1_T-f1_B)) / (one + drho1);
+ //real vx1 = (((f1_TNE-f1_BSW)+(f1_BSE-f1_TNW)+(f1_BNE-f1_TSW)+(f1_TSE-f1_BNW)) + (((f1_NE-f1_SW)+(f1_TE-f1_BW))+((f1_SE-f1_NW)+(f1_BE-f1_TW))) + (f1_E-f1_W)) / (one + drho1);
+ //real vx2 = (((f1_TNE-f1_BSW)+(f1_TNW-f1_BSE)+(f1_BNE-f1_TSW)+(f1_BNW-f1_TSE)) + (((f1_NE-f1_SW)+(f1_TN-f1_BS))+((f1_BN-f1_TS)+(f1_NW-f1_SE))) + (f1_N-f1_S)) / (one + drho1);
+ //real vx3 = (((f1_TNE-f1_BSW)+(f1_TNW-f1_BSE)+(f1_TSW-f1_BNE)+(f1_TSE-f1_BNW)) + (((f1_TE-f1_BW)+(f1_TN-f1_BS))+((f1_TW-f1_BE)+(f1_TS-f1_BN))) + (f1_T-f1_B)) / (one + drho1);
// //////////////////////////////////////////////////////////////////////////
- ////real omega = om1;
+ ////real omega = om1;
// real cusq = c3o2*(vx1*vx1+vx2*vx2+vx3*vx3);
// //////////////////////////////////////////////////////////////////////////
- ////T�st MK
- ////if(vx1 < zero) vx1 = zero;
+ ////T�st MK
+ ////if(vx1 < zero) vx1 = zero;
// //////////////////////////////////////////////////////////////////////////
- ////becomes higher with neighbor source and lower with local source
+ ////becomes higher with neighbor source and lower with local source
// //real fZERO = c8over27* (rhoBC[k]-(one + rhoBC[k])*(cusq)) ;
// //real fE = c2over27* (rhoBC[k]+(one + rhoBC[k])*(three*( vx1 )+c9over2*( vx1 )*( vx1 )-cusq));
// //real fW = c2over27* (rhoBC[k]+(one + rhoBC[k])*(three*(-vx1 )+c9over2*(-vx1 )*(-vx1 )-cusq));
@@ -3853,7 +3868,7 @@ __global__ void QPressDeviceEQZ27(real* rhoBC,
// //real fBSE = c1over216* (rhoBC[k]+(one + rhoBC[k])*(three*( vx1-vx2-vx3)+c9over2*( vx1-vx2-vx3)*( vx1-vx2-vx3)-cusq));
// //real fTNW = c1over216* (rhoBC[k]+(one + rhoBC[k])*(three*(-vx1+vx2+vx3)+c9over2*(-vx1+vx2+vx3)*(-vx1+vx2+vx3)-cusq));
// //////////////////////////////////////////////////////////////////////////
- //// based on VirtualFluids (kucher + fard)
+ //// based on VirtualFluids (kucher + fard)
// real fZERO = c8over27 * rhoBC[k] * (one - cusq);
// real fE = c2over27 * rhoBC[k] * (one + three * ( vx1 ) + c9over2 * ( vx1 ) * ( vx1 ) - cusq);
// real fW = c2over27 * rhoBC[k] * (one + three * (-vx1 ) + c9over2 * (-vx1 ) * (-vx1 ) - cusq);
@@ -3882,7 +3897,7 @@ __global__ void QPressDeviceEQZ27(real* rhoBC,
// real fBSE = c1over216 * rhoBC[k] * (one + three * ( vx1-vx2-vx3) + c9over2 * ( vx1-vx2-vx3) * ( vx1-vx2-vx3) - cusq);
// real fTNW = c1over216 * rhoBC[k] * (one + three * (-vx1+vx2+vx3) + c9over2 * (-vx1+vx2+vx3) * (-vx1+vx2+vx3) - cusq);
//// ////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
- //////test
+ //////test
//// real fZERO = c8over27 * ((drho1 + rhoBC[k]) / two) * (one - cusq);
//// real fE = c2over27 * ((drho1 + rhoBC[k]) / two) * (one + three * ( vx1 ) + c9over2 * ( vx1 ) * ( vx1 ) - cusq);
//// real fW = c2over27 * ((drho1 + rhoBC[k]) / two) * (one + three * (-vx1 ) + c9over2 * (-vx1 ) * (-vx1 ) - cusq);
@@ -3911,190 +3926,190 @@ __global__ void QPressDeviceEQZ27(real* rhoBC,
//// real fBSE = c1over216 * ((drho1 + rhoBC[k]) / two) * (one + three * ( vx1-vx2-vx3) + c9over2 * ( vx1-vx2-vx3) * ( vx1-vx2-vx3) - cusq);
//// real fTNW = c1over216 * ((drho1 + rhoBC[k]) / two) * (one + three * (-vx1+vx2+vx3) + c9over2 * (-vx1+vx2+vx3) * (-vx1+vx2+vx3) - cusq);
- //////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
+ //////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
// based on BGK Plus Comp
- //////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
- //double mfabb = (D.f[DIR_P00 ])[k1e ];
- //double mfcbb = (D.f[DIR_M00 ])[k1w ];
- //double mfbab = (D.f[DIR_0P0 ])[k1n ];
- //double mfbcb = (D.f[DIR_0M0 ])[k1s ];
- //double mfbba = (D.f[DIR_00P ])[k1t ];
- //double mfbbc = (D.f[DIR_00M ])[k1b ];
- //double mfaab = (D.f[DIR_PP0 ])[k1ne ];
- //double mfccb = (D.f[DIR_MM0 ])[k1sw ];
- //double mfacb = (D.f[DIR_PM0 ])[k1se ];
- //double mfcab = (D.f[DIR_MP0 ])[k1nw ];
- //double mfaba = (D.f[DIR_P0P ])[k1te ];
- //double mfcbc = (D.f[DIR_M0M ])[k1bw ];
- //double mfabc = (D.f[DIR_P0M ])[k1be ];
- //double mfcba = (D.f[DIR_M0P ])[k1tw ];
- //double mfbaa = (D.f[DIR_0PP ])[k1tn ];
- //double mfbcc = (D.f[DIR_0MM ])[k1bs ];
- //double mfbac = (D.f[DIR_0PM ])[k1bn ];
- //double mfbca = (D.f[DIR_0MP ])[k1ts ];
- //double mfbbb = (D.f[DIR_000])[k1zero];
- //double mfaaa = (D.f[DIR_PPP ])[k1tne ];
- //double mfcca = (D.f[DIR_MMP ])[k1tsw ];
- //double mfaca = (D.f[DIR_PMP ])[k1tse ];
- //double mfcaa = (D.f[DIR_MPP ])[k1tnw ];
- //double mfaac = (D.f[DIR_PPM ])[k1bne ];
- //double mfccc = (D.f[DIR_MMM ])[k1bsw ];
- //double mfacc = (D.f[DIR_PMM ])[k1bse ];
- //double mfcac = (D.f[DIR_MPM ])[k1bnw ];
- real mfabb = (D.f[DIR_P00 ])[k1e ];
- real mfcbb = (D.f[DIR_M00 ])[k1w ];
- real mfbab = (D.f[DIR_0P0 ])[k1n ];
- real mfbcb = (D.f[DIR_0M0 ])[k1s ];
- real mfbba = (D.f[DIR_00P ])[k1t ];
- real mfbbc = (D.f[DIR_00M ])[k1b ];
- real mfaab = (D.f[DIR_PP0 ])[k1ne ];
- real mfccb = (D.f[DIR_MM0 ])[k1sw ];
- real mfacb = (D.f[DIR_PM0 ])[k1se ];
- real mfcab = (D.f[DIR_MP0 ])[k1nw ];
- real mfaba = (D.f[DIR_P0P ])[k1te ];
- real mfcbc = (D.f[DIR_M0M ])[k1bw ];
- real mfabc = (D.f[DIR_P0M ])[k1be ];
- real mfcba = (D.f[DIR_M0P ])[k1tw ];
- real mfbaa = (D.f[DIR_0PP ])[k1tn ];
- real mfbcc = (D.f[DIR_0MM ])[k1bs ];
- real mfbac = (D.f[DIR_0PM ])[k1bn ];
- real mfbca = (D.f[DIR_0MP ])[k1ts ];
- real mfbbb = (D.f[DIR_000])[k1zero];
- real mfaaa = (D.f[DIR_PPP ])[k1tne ];
- real mfcca = (D.f[DIR_MMP ])[k1tsw ];
- real mfaca = (D.f[DIR_PMP ])[k1tse ];
- real mfcaa = (D.f[DIR_MPP ])[k1tnw ];
- real mfaac = (D.f[DIR_PPM ])[k1bne ];
- real mfccc = (D.f[DIR_MMM ])[k1bsw ];
- real mfacc = (D.f[DIR_PMM ])[k1bse ];
- real mfcac = (D.f[DIR_MPM ])[k1bnw ];
-
- //real mfcbb = (D.f[DIR_P00 ])[ke ];
- //real mfabb = (D.f[DIR_M00 ])[kw ];
- //real mfbcb = (D.f[DIR_0P0 ])[kn ];
- //real mfbab = (D.f[DIR_0M0 ])[ks ];
- //real mfbbc = (D.f[DIR_00P ])[kt ];
- //real mfbba = (D.f[DIR_00M ])[kb ];
- //real mfccb = (D.f[DIR_PP0 ])[kne ];
- //real mfaab = (D.f[DIR_MM0 ])[ksw ];
- //real mfcab = (D.f[DIR_PM0 ])[kse ];
- //real mfacb = (D.f[DIR_MP0 ])[knw ];
- //real mfcbc = (D.f[DIR_P0P ])[kte ];
- //real mfaba = (D.f[DIR_M0M ])[kbw ];
- //real mfcba = (D.f[DIR_P0M ])[kbe ];
- //real mfabc = (D.f[DIR_M0P ])[ktw ];
- //real mfbcc = (D.f[DIR_0PP ])[ktn ];
- //real mfbaa = (D.f[DIR_0MM ])[kbs ];
- //real mfbca = (D.f[DIR_0PM ])[kbn ];
- //real mfbac = (D.f[DIR_0MP ])[kts ];
- //real mfbbb = (D.f[DIR_000])[kzero];
- //real mfccc = (D.f[DIR_PPP ])[ktne ];
- //real mfaac = (D.f[DIR_MMP ])[ktsw ];
- //real mfcac = (D.f[DIR_PMP ])[ktse ];
- //real mfacc = (D.f[DIR_MPP ])[ktnw ];
- //real mfcca = (D.f[DIR_PPM ])[kbne ];
- //real mfaaa = (D.f[DIR_MMM ])[kbsw ];
- //real mfcaa = (D.f[DIR_PMM ])[kbse ];
- //real mfaca = (D.f[DIR_MPM ])[kbnw ];
- ////////////////////////////////////////////////////////////////////////////////////
- //real rho = (((((mfccc+mfaaa) + (mfaca+mfcac)) + ((mfacc+mfcaa) + (mfaac+mfcca))) +
- // (((mfbac+mfbca) + (mfbaa+mfbcc)) + ((mfabc+mfcba) + (mfaba+mfcbc)) + ((mfacb+mfcab) + (mfaab+mfccb))) +
- // ((mfabb+mfcbb) + (mfbab+mfbcb)) + (mfbba+mfbbc)) + mfbbb) + one;//!!!!Achtung + one
- ////////////////////////////////////////////////////////////////////////////////////
- real rho = rhoBC[k];
- ////////////////////////////////////////////////////////////////////////////////////
- real OoRho = c1o1 / (rho * 1.5f);
- ////////////////////////////////////////////////////////////////////////////////////
- real vvx = ((((mfccc-mfaaa) + (mfcac-mfaca)) + ((mfcaa-mfacc) + (mfcca-mfaac))) +
- (((mfcba-mfabc) + (mfcbc-mfaba)) + ((mfcab-mfacb) + (mfccb-mfaab))) +
- (mfcbb-mfabb)) * OoRho;
- real vvy =((((mfccc-mfaaa) + (mfaca-mfcac)) + ((mfacc-mfcaa) + (mfcca-mfaac))) +
- (((mfbca-mfbac) + (mfbcc-mfbaa)) + ((mfacb-mfcab) + (mfccb-mfaab))) +
- (mfbcb-mfbab)) * OoRho;
- real vvz =((((mfccc-mfaaa) + (mfcac-mfaca)) + ((mfacc-mfcaa) + (mfaac-mfcca))) +
- (((mfbac-mfbca) + (mfbcc-mfbaa)) + ((mfabc-mfcba) + (mfcbc-mfaba))) +
- (mfbbc-mfbba)) * OoRho;
- /////////////////////////
- //Test Values
- //double vvx = 0.016;
- //double vvy = zero;
- //double vvz = zero;
- ////////////////////////////////////////////////////////////////////////////////////////
- ////round off error test
- //if(vvx!=zero){
- // (kDistTest.f[DIR_P00 ])[k] = mfabb;
- // (kDistTest.f[DIR_M00 ])[k] = mfcbb;
- // (kDistTest.f[DIR_0P0 ])[k] = mfbab;
- // (kDistTest.f[DIR_0M0 ])[k] = mfbcb;
- // (kDistTest.f[DIR_00P ])[k] = mfbba;
- // (kDistTest.f[DIR_00M ])[k] = mfbbc;
- // (kDistTest.f[DIR_PP0 ])[k] = mfaab;
- // (kDistTest.f[DIR_MM0 ])[k] = mfccb;
- // (kDistTest.f[DIR_PM0 ])[k] = mfacb;
- // (kDistTest.f[DIR_MP0 ])[k] = mfcab;
- // (kDistTest.f[DIR_P0P ])[k] = mfaba;
- // (kDistTest.f[DIR_M0M ])[k] = mfcbc;
- // (kDistTest.f[DIR_P0M ])[k] = mfabc;
- // (kDistTest.f[DIR_M0P ])[k] = mfcba;
- // (kDistTest.f[DIR_0PP ])[k] = mfbaa;
- // (kDistTest.f[DIR_0MM ])[k] = mfbcc;
- // (kDistTest.f[DIR_0PM ])[k] = mfbac;
- // (kDistTest.f[DIR_0MP ])[k] = mfbca;
- // (kDistTest.f[DIR_000])[k] = KQK;
- // (kDistTest.f[DIR_PPP ])[k] = mfaaa;
- // (kDistTest.f[DIR_MMP ])[k] = mfcca;
- // (kDistTest.f[DIR_PMP ])[k] = mfaca;
- // (kDistTest.f[DIR_MPP ])[k] = mfcaa;
- // (kDistTest.f[DIR_PPM ])[k] = mfaac;
- // (kDistTest.f[DIR_MMM ])[k] = mfccc;
- // (kDistTest.f[DIR_PMM ])[k] = mfacc;
- // (kDistTest.f[DIR_MPM ])[k] = mfcac;
- //}else{
- // (kDistTest.f[DIR_P00 ])[k] = zero;
- // (kDistTest.f[DIR_M00 ])[k] = zero;
- // (kDistTest.f[DIR_0P0 ])[k] = zero;
- // (kDistTest.f[DIR_0M0 ])[k] = zero;
- // (kDistTest.f[DIR_00P ])[k] = zero;
- // (kDistTest.f[DIR_00M ])[k] = zero;
- // (kDistTest.f[DIR_PP0 ])[k] = zero;
- // (kDistTest.f[DIR_MM0 ])[k] = zero;
- // (kDistTest.f[DIR_PM0 ])[k] = zero;
- // (kDistTest.f[DIR_MP0 ])[k] = zero;
- // (kDistTest.f[DIR_P0P ])[k] = zero;
- // (kDistTest.f[DIR_M0M ])[k] = zero;
- // (kDistTest.f[DIR_P0M ])[k] = zero;
- // (kDistTest.f[DIR_M0P ])[k] = zero;
- // (kDistTest.f[DIR_0PP ])[k] = zero;
- // (kDistTest.f[DIR_0MM ])[k] = zero;
- // (kDistTest.f[DIR_0PM ])[k] = zero;
- // (kDistTest.f[DIR_0MP ])[k] = zero;
- // (kDistTest.f[DIR_000])[k] = zero;
- // (kDistTest.f[DIR_PPP ])[k] = zero;
- // (kDistTest.f[DIR_MMP ])[k] = zero;
- // (kDistTest.f[DIR_PMP ])[k] = zero;
- // (kDistTest.f[DIR_MPP ])[k] = zero;
- // (kDistTest.f[DIR_PPM ])[k] = zero;
- // (kDistTest.f[DIR_MMM ])[k] = zero;
- // (kDistTest.f[DIR_PMM ])[k] = zero;
- // (kDistTest.f[DIR_MPM ])[k] = zero;
- //}
-
- //////////////////////////////////////////////////////////////////////////////////////
- //// first bad fix for negative x velocity
- ////if(vvx > zero) vvx = zero;
- //////////////////////////////////////////////////////////////////////////////////////
- ////// second bad fix for negative x velocity
- ////if(vvx > zero){
- //// vvx = -vvx;
- //// vvy = -vvy;
- //// vvz = -vvz;
- ////}
- ////////////////////////////////////////////////////////////////////////////////////
- double vx2 = vvx * vvx;
- double vy2 = vvy * vvy;
- double vz2 = vvz * vvz;
- //////////////////////////////////////////////////////////////////////////////////
- //original
+ //////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
+ //double mfabb = (D.f[DIR_P00])[k1e ];
+ //double mfcbb = (D.f[DIR_M00])[k1w ];
+ //double mfbab = (D.f[DIR_0P0])[k1n ];
+ //double mfbcb = (D.f[DIR_0M0])[k1s ];
+ //double mfbba = (D.f[DIR_00P])[k1t ];
+ //double mfbbc = (D.f[DIR_00M])[k1b ];
+ //double mfaab = (D.f[DIR_PP0])[k1ne ];
+ //double mfccb = (D.f[DIR_MM0])[k1sw ];
+ //double mfacb = (D.f[DIR_PM0])[k1se ];
+ //double mfcab = (D.f[DIR_MP0])[k1nw ];
+ //double mfaba = (D.f[DIR_P0P])[k1te ];
+ //double mfcbc = (D.f[DIR_M0M])[k1bw ];
+ //double mfabc = (D.f[DIR_P0M])[k1be ];
+ //double mfcba = (D.f[DIR_M0P])[k1tw ];
+ //double mfbaa = (D.f[DIR_0PP])[k1tn ];
+ //double mfbcc = (D.f[DIR_0MM])[k1bs ];
+ //double mfbac = (D.f[DIR_0PM])[k1bn ];
+ //double mfbca = (D.f[DIR_0MP])[k1ts ];
+ //double mfbbb = (D.f[DIR_000])[k1zero];
+ //double mfaaa = (D.f[DIR_PPP])[k1tne ];
+ //double mfcca = (D.f[DIR_MMP])[k1tsw ];
+ //double mfaca = (D.f[DIR_PMP])[k1tse ];
+ //double mfcaa = (D.f[DIR_MPP])[k1tnw ];
+ //double mfaac = (D.f[DIR_PPM])[k1bne ];
+ //double mfccc = (D.f[DIR_MMM])[k1bsw ];
+ //double mfacc = (D.f[DIR_PMM])[k1bse ];
+ //double mfcac = (D.f[DIR_MPM])[k1bnw ];
+ real mfabb = (D.f[DIR_P00])[k1e ];
+ real mfcbb = (D.f[DIR_M00])[k1w ];
+ real mfbab = (D.f[DIR_0P0])[k1n ];
+ real mfbcb = (D.f[DIR_0M0])[k1s ];
+ real mfbba = (D.f[DIR_00P])[k1t ];
+ real mfbbc = (D.f[DIR_00M])[k1b ];
+ real mfaab = (D.f[DIR_PP0])[k1ne ];
+ real mfccb = (D.f[DIR_MM0])[k1sw ];
+ real mfacb = (D.f[DIR_PM0])[k1se ];
+ real mfcab = (D.f[DIR_MP0])[k1nw ];
+ real mfaba = (D.f[DIR_P0P])[k1te ];
+ real mfcbc = (D.f[DIR_M0M])[k1bw ];
+ real mfabc = (D.f[DIR_P0M])[k1be ];
+ real mfcba = (D.f[DIR_M0P])[k1tw ];
+ real mfbaa = (D.f[DIR_0PP])[k1tn ];
+ real mfbcc = (D.f[DIR_0MM])[k1bs ];
+ real mfbac = (D.f[DIR_0PM])[k1bn ];
+ real mfbca = (D.f[DIR_0MP])[k1ts ];
+ real mfbbb = (D.f[DIR_000])[k1zero];
+ real mfaaa = (D.f[DIR_PPP])[k1tne ];
+ real mfcca = (D.f[DIR_MMP])[k1tsw ];
+ real mfaca = (D.f[DIR_PMP])[k1tse ];
+ real mfcaa = (D.f[DIR_MPP])[k1tnw ];
+ real mfaac = (D.f[DIR_PPM])[k1bne ];
+ real mfccc = (D.f[DIR_MMM])[k1bsw ];
+ real mfacc = (D.f[DIR_PMM])[k1bse ];
+ real mfcac = (D.f[DIR_MPM])[k1bnw ];
+
+ //real mfcbb = (D.f[DIR_P00])[ke ];
+ //real mfabb = (D.f[DIR_M00])[kw ];
+ //real mfbcb = (D.f[DIR_0P0])[kn ];
+ //real mfbab = (D.f[DIR_0M0])[ks ];
+ //real mfbbc = (D.f[DIR_00P])[kt ];
+ //real mfbba = (D.f[DIR_00M])[kb ];
+ //real mfccb = (D.f[DIR_PP0])[kne ];
+ //real mfaab = (D.f[DIR_MM0])[ksw ];
+ //real mfcab = (D.f[DIR_PM0])[kse ];
+ //real mfacb = (D.f[DIR_MP0])[knw ];
+ //real mfcbc = (D.f[DIR_P0P])[kte ];
+ //real mfaba = (D.f[DIR_M0M])[kbw ];
+ //real mfcba = (D.f[DIR_P0M])[kbe ];
+ //real mfabc = (D.f[DIR_M0P])[ktw ];
+ //real mfbcc = (D.f[DIR_0PP])[ktn ];
+ //real mfbaa = (D.f[DIR_0MM])[kbs ];
+ //real mfbca = (D.f[DIR_0PM])[kbn ];
+ //real mfbac = (D.f[DIR_0MP])[kts ];
+ //real mfbbb = (D.f[DIR_000])[kzero];
+ //real mfccc = (D.f[DIR_PPP])[ktne ];
+ //real mfaac = (D.f[DIR_MMP])[ktsw ];
+ //real mfcac = (D.f[DIR_PMP])[ktse ];
+ //real mfacc = (D.f[DIR_MPP])[ktnw ];
+ //real mfcca = (D.f[DIR_PPM])[kbne ];
+ //real mfaaa = (D.f[DIR_MMM])[kbsw ];
+ //real mfcaa = (D.f[DIR_PMM])[kbse ];
+ //real mfaca = (D.f[DIR_MPM])[kbnw ];
+ ////////////////////////////////////////////////////////////////////////////////////
+ //real rho = (((((mfccc+mfaaa) + (mfaca+mfcac)) + ((mfacc+mfcaa) + (mfaac+mfcca))) +
+ // (((mfbac+mfbca) + (mfbaa+mfbcc)) + ((mfabc+mfcba) + (mfaba+mfcbc)) + ((mfacb+mfcab) + (mfaab+mfccb))) +
+ // ((mfabb+mfcbb) + (mfbab+mfbcb)) + (mfbba+mfbbc)) + mfbbb) + one;//!!!!Achtung + one
+ ////////////////////////////////////////////////////////////////////////////////////
+ real rho = rhoBC[k];
+ ////////////////////////////////////////////////////////////////////////////////////
+ real OoRho = c1o1 / (rho * 1.5f);
+ ////////////////////////////////////////////////////////////////////////////////////
+ real vvx = ((((mfccc-mfaaa) + (mfcac-mfaca)) + ((mfcaa-mfacc) + (mfcca-mfaac))) +
+ (((mfcba-mfabc) + (mfcbc-mfaba)) + ((mfcab-mfacb) + (mfccb-mfaab))) +
+ (mfcbb-mfabb)) * OoRho;
+ real vvy =((((mfccc-mfaaa) + (mfaca-mfcac)) + ((mfacc-mfcaa) + (mfcca-mfaac))) +
+ (((mfbca-mfbac) + (mfbcc-mfbaa)) + ((mfacb-mfcab) + (mfccb-mfaab))) +
+ (mfbcb-mfbab)) * OoRho;
+ real vvz =((((mfccc-mfaaa) + (mfcac-mfaca)) + ((mfacc-mfcaa) + (mfaac-mfcca))) +
+ (((mfbac-mfbca) + (mfbcc-mfbaa)) + ((mfabc-mfcba) + (mfcbc-mfaba))) +
+ (mfbbc-mfbba)) * OoRho;
+ /////////////////////////
+ //Test Values
+ //double vvx = 0.016;
+ //double vvy = zero;
+ //double vvz = zero;
+ ////////////////////////////////////////////////////////////////////////////////////////
+ ////round off error test
+ //if(vvx!=zero){
+ // (kDistTest.f[DIR_P00])[k] = mfabb;
+ // (kDistTest.f[DIR_M00])[k] = mfcbb;
+ // (kDistTest.f[DIR_0P0])[k] = mfbab;
+ // (kDistTest.f[DIR_0M0])[k] = mfbcb;
+ // (kDistTest.f[DIR_00P])[k] = mfbba;
+ // (kDistTest.f[DIR_00M])[k] = mfbbc;
+ // (kDistTest.f[DIR_PP0])[k] = mfaab;
+ // (kDistTest.f[DIR_MM0])[k] = mfccb;
+ // (kDistTest.f[DIR_PM0])[k] = mfacb;
+ // (kDistTest.f[DIR_MP0])[k] = mfcab;
+ // (kDistTest.f[DIR_P0P])[k] = mfaba;
+ // (kDistTest.f[DIR_M0M])[k] = mfcbc;
+ // (kDistTest.f[DIR_P0M])[k] = mfabc;
+ // (kDistTest.f[DIR_M0P])[k] = mfcba;
+ // (kDistTest.f[DIR_0PP])[k] = mfbaa;
+ // (kDistTest.f[DIR_0MM])[k] = mfbcc;
+ // (kDistTest.f[DIR_0PM])[k] = mfbac;
+ // (kDistTest.f[DIR_0MP])[k] = mfbca;
+ // (kDistTest.f[DIR_000])[k] = KQK;
+ // (kDistTest.f[DIR_PPP])[k] = mfaaa;
+ // (kDistTest.f[DIR_MMP])[k] = mfcca;
+ // (kDistTest.f[DIR_PMP])[k] = mfaca;
+ // (kDistTest.f[DIR_MPP])[k] = mfcaa;
+ // (kDistTest.f[DIR_PPM])[k] = mfaac;
+ // (kDistTest.f[DIR_MMM])[k] = mfccc;
+ // (kDistTest.f[DIR_PMM])[k] = mfacc;
+ // (kDistTest.f[DIR_MPM])[k] = mfcac;
+ //}else{
+ // (kDistTest.f[DIR_P00])[k] = zero;
+ // (kDistTest.f[DIR_M00])[k] = zero;
+ // (kDistTest.f[DIR_0P0])[k] = zero;
+ // (kDistTest.f[DIR_0M0])[k] = zero;
+ // (kDistTest.f[DIR_00P])[k] = zero;
+ // (kDistTest.f[DIR_00M])[k] = zero;
+ // (kDistTest.f[DIR_PP0])[k] = zero;
+ // (kDistTest.f[DIR_MM0])[k] = zero;
+ // (kDistTest.f[DIR_PM0])[k] = zero;
+ // (kDistTest.f[DIR_MP0])[k] = zero;
+ // (kDistTest.f[DIR_P0P])[k] = zero;
+ // (kDistTest.f[DIR_M0M])[k] = zero;
+ // (kDistTest.f[DIR_P0M])[k] = zero;
+ // (kDistTest.f[DIR_M0P])[k] = zero;
+ // (kDistTest.f[DIR_0PP])[k] = zero;
+ // (kDistTest.f[DIR_0MM])[k] = zero;
+ // (kDistTest.f[DIR_0PM])[k] = zero;
+ // (kDistTest.f[DIR_0MP])[k] = zero;
+ // (kDistTest.f[DIR_000])[k] = zero;
+ // (kDistTest.f[DIR_PPP])[k] = zero;
+ // (kDistTest.f[DIR_MMP])[k] = zero;
+ // (kDistTest.f[DIR_PMP])[k] = zero;
+ // (kDistTest.f[DIR_MPP])[k] = zero;
+ // (kDistTest.f[DIR_PPM])[k] = zero;
+ // (kDistTest.f[DIR_MMM])[k] = zero;
+ // (kDistTest.f[DIR_PMM])[k] = zero;
+ // (kDistTest.f[DIR_MPM])[k] = zero;
+ //}
+
+ //////////////////////////////////////////////////////////////////////////////////////
+ //// first bad fix for negative x velocity
+ ////if(vvx > zero) vvx = zero;
+ //////////////////////////////////////////////////////////////////////////////////////
+ ////// second bad fix for negative x velocity
+ ////if(vvx > zero){
+ //// vvx = -vvx;
+ //// vvy = -vvy;
+ //// vvz = -vvz;
+ ////}
+ ////////////////////////////////////////////////////////////////////////////////////
+ double vx2 = vvx * vvx;
+ double vy2 = vvy * vvy;
+ double vz2 = vvz * vvz;
+ //////////////////////////////////////////////////////////////////////////////////
+ //original
real XXb = -c2o3 + vx2;
real XXc = -c1o2 * (XXb + c1o1 + vvx);
real XXa = XXc + vvx;
@@ -4104,213 +4119,213 @@ __global__ void QPressDeviceEQZ27(real* rhoBC,
real ZZb = -c2o3 + vz2;
real ZZc = -c1o2 * (ZZb + c1o1 + vvz);
real ZZa = ZZc + vvz;
- //////////////////////////////////////////////////////////////////////////////////
- //unkonditioniert
- mfcbb = -(rhoBC[k] + c1o1) * XXc * YYb * ZZb - c2o27;
- mfabb = -(rhoBC[k] + c1o1) * XXa * YYb * ZZb - c2o27;
- mfbcb = -(rhoBC[k] + c1o1) * XXb * YYc * ZZb - c2o27;
- mfbab = -(rhoBC[k] + c1o1) * XXb * YYa * ZZb - c2o27;
- mfbbc = -(rhoBC[k] + c1o1) * XXb * YYb * ZZc - c2o27;
- mfbba = -(rhoBC[k] + c1o1) * XXb * YYb * ZZa - c2o27;
- mfccb = -(rhoBC[k] + c1o1) * XXc * YYc * ZZb - c1o54;
- mfaab = -(rhoBC[k] + c1o1) * XXa * YYa * ZZb - c1o54;
- mfcab = -(rhoBC[k] + c1o1) * XXc * YYa * ZZb - c1o54;
- mfacb = -(rhoBC[k] + c1o1) * XXa * YYc * ZZb - c1o54;
- mfcbc = -(rhoBC[k] + c1o1) * XXc * YYb * ZZc - c1o54;
- mfaba = -(rhoBC[k] + c1o1) * XXa * YYb * ZZa - c1o54;
- mfcba = -(rhoBC[k] + c1o1) * XXc * YYb * ZZa - c1o54;
- mfabc = -(rhoBC[k] + c1o1) * XXa * YYb * ZZc - c1o54;
- mfbcc = -(rhoBC[k] + c1o1) * XXb * YYc * ZZc - c1o54;
- mfbaa = -(rhoBC[k] + c1o1) * XXb * YYa * ZZa - c1o54;
- mfbca = -(rhoBC[k] + c1o1) * XXb * YYc * ZZa - c1o54;
- mfbac = -(rhoBC[k] + c1o1) * XXb * YYa * ZZc - c1o54;
- mfbbb = -(rhoBC[k] + c1o1) * XXb * YYb * ZZb - c8o27;
- mfccc = -(rhoBC[k] + c1o1) * XXc * YYc * ZZc - c1o216;
- mfaac = -(rhoBC[k] + c1o1) * XXa * YYa * ZZc - c1o216;
- mfcac = -(rhoBC[k] + c1o1) * XXc * YYa * ZZc - c1o216;
- mfacc = -(rhoBC[k] + c1o1) * XXa * YYc * ZZc - c1o216;
- mfcca = -(rhoBC[k] + c1o1) * XXc * YYc * ZZa - c1o216;
- mfaaa = -(rhoBC[k] + c1o1) * XXa * YYa * ZZa - c1o216;
- mfcaa = -(rhoBC[k] + c1o1) * XXc * YYa * ZZa - c1o216;
- mfaca = -(rhoBC[k] + c1o1) * XXa * YYc * ZZa - c1o216;
- //////////////////////////////////////////////////////////
- ////konditioniert
- //double OneOver216RhoPlusOne = c1over216*(rhoBC[k]+one);
- //double OnoOver216Rho = c1over216*rhoBC[k];
- //mfcbb = OnoOver216Rho*sixteen + OneOver216RhoPlusOne*twelve*(-(two*vy2) - two*vz2 + three*vy2*vz2 + vvx*(-two + three*vy2)*(-two + three*vz2) + vx2*(-two + three*vy2)*(-two + three*vz2));
- //mfabb = OnoOver216Rho*sixteen - OneOver216RhoPlusOne*twelve*(two*vy2 + two*vz2 - three*vy2*vz2 + vvx*(-two + three*vy2)*(-two + three*vz2) + vx2*(-four + six*vy2 + six*vz2 - nine*vy2*vz2));
- //mfbcb = four*(-(four*OneOver216RhoPlusOne) + four*OnoOver216Rho + OneOver216RhoPlusOne*(-two + three*vx2)*(one + three*vvy + three*vy2)*(-two + three*vz2));
- //mfbab = four*(four*OnoOver216Rho - OneOver216RhoPlusOne*three*(vvy*(-two + three*vx2)*(-two + three*vz2) - one*vx2*(one + three*vy2)*(-two + three*vz2) + two*(-(two*vy2) + vz2 + three*vy2*vz2)));
- //mfbbc = four*(-(four*OneOver216RhoPlusOne) + four*OnoOver216Rho + OneOver216RhoPlusOne*(-two + three*vx2)*(-two + three*vy2)*(one + three*vvz + three*vz2));
- //mfbba = four*(four*OnoOver216Rho - OneOver216RhoPlusOne*three*(vvz*(-two + three*vx2)*(-two + three*vy2) - one*vx2*(-two + three*vy2)*(one + three*vz2) + two*(vy2 - two*vz2 + three*vy2*vz2)));
- //mfccb = -(two*(-(OnoOver216Rho*two) + OneOver216RhoPlusOne*three*(-(two*vx2) - two*vy2 - six*vx2*vy2 + vz2 + three*vx2*vz2 + three*vy2*vz2 + nine*vx2*vy2*vz2 + vvy*(one + three*vx2)*(-two + three*vz2) + vvx*(one + three*vvy + three*vy2)*(-two + three*vz2))));
- //mfaab = -(two*(-(OnoOver216Rho*two) + OneOver216RhoPlusOne*three*(-(two*vx2) - two*vy2 - six*vx2*vy2 + vz2 + three*vx2*vz2 + three*vy2*vz2 + nine*vx2*vy2*vz2 - one*vvy*(one + three*vx2)*(-two + three*vz2) + vvx*(-one + three*vvy - three*vy2)*(-two + three*vz2))));
- //mfcab = two*(OnoOver216Rho*two + OneOver216RhoPlusOne*three*(two*vx2 + two*vy2 + six*vx2*vy2 - one*vz2 - three*vx2*vz2 - three*vy2*vz2 - nine*vx2*vy2*vz2 + vvy*(one + three*vx2)*(-two + three*vz2) + vvx*(-one + three*vvy - three*vy2)*(-two + three*vz2)));
- //mfacb = two*(OnoOver216Rho*two + OneOver216RhoPlusOne*three*(two*vx2 + two*vy2 + six*vx2*vy2 - one*vz2 - three*vx2*vz2 - three*vy2*vz2 - nine*vx2*vy2*vz2 - one*vvy*(one + three*vx2)*(-two + three*vz2) + vvx*(one + three*vvy + three*vy2)*(-two + three*vz2)));
- //mfcbc = -(two*(-(OnoOver216Rho*two) + OneOver216RhoPlusOne*three*(-(two*vx2) + vy2 + three*vx2*vy2 + vvz*(one + three*vx2)*(-two + three*vy2) - two*vz2 - six*vx2*vz2 + three*vy2*vz2 + nine*vx2*vy2*vz2 + vvx*(-two + three*vy2)*(one + three*vvz + three*vz2))));
- //mfaba = -(two*(-(OnoOver216Rho*two) + OneOver216RhoPlusOne*three*(-(two*vx2) + vy2 + three*vx2*vy2 - one*vvz*(one + three*vx2)*(-two + three*vy2) - two*vz2 - six*vx2*vz2 + three*vy2*vz2 + nine*vx2*vy2*vz2 + vvx*(-two + three*vy2)*(-one + three*vvz - three*vz2))));
- //mfcba = two*(OnoOver216Rho*two + OneOver216RhoPlusOne*three*(two*vx2 - one*vy2 - three*vx2*vy2 + vvz*(one + three*vx2)*(-two + three*vy2) + two*vz2 + six*vx2*vz2 - three*vy2*vz2 - nine*vx2*vy2*vz2 + vvx*(-two + three*vy2)*(-one + three*vvz - three*vz2)));
- //mfabc = two*(OnoOver216Rho*two + OneOver216RhoPlusOne*three*(two*vx2 - one*vy2 - three*vx2*vy2 - one*vvz*(one + three*vx2)*(-two + three*vy2) + two*vz2 + six*vx2*vz2 - three*vy2*vz2 - nine*vx2*vy2*vz2 + vvx*(-two + three*vy2)*(one + three*vvz + three*vz2)));
- //mfbcc = -(two*(-(OnoOver216Rho*two) + OneOver216RhoPlusOne*three*(vx2 - two*vy2 + three*vx2*vy2 + vvz*(-two + three*vx2)*(one + three*vy2) - two*vz2 + three*vx2*vz2 - six*vy2*vz2 + nine*vx2*vy2*vz2 + vvy*(-two + three*vx2)*(one + three*vvz + three*vz2))));
- //mfbaa = -(two*(-(OnoOver216Rho*two) + OneOver216RhoPlusOne*three*(vx2 - two*vy2 + three*vx2*vy2 - one*vvz*(-two + three*vx2)*(one + three*vy2) - two*vz2 + three*vx2*vz2 - six*vy2*vz2 + nine*vx2*vy2*vz2 + vvy*(-two + three*vx2)*(-one + three*vvz - three*vz2))));
- //mfbca = two*(OnoOver216Rho*two + OneOver216RhoPlusOne*three*(-(one*vx2) + two*vy2 - three*vx2*vy2 + vvz*(-two + three*vx2)*(one + three*vy2) + two*vz2 - three*vx2*vz2 + six*vy2*vz2 - nine*vx2*vy2*vz2 + vvy*(-two + three*vx2)*(-one + three*vvz - three*vz2)));
- //mfbac = two*(OnoOver216Rho*two + OneOver216RhoPlusOne*three*(-(one*vx2) + two*vy2 - three*vx2*vy2 - one*vvz*(-two + three*vx2)*(one + three*vy2) + two*vz2 - three*vx2*vz2 + six*vy2*vz2 - nine*vx2*vy2*vz2 + vvy*(-two + three*vx2)*(one + three*vvz + three*vz2)));
- //mfbbb = eight*(eight*OnoOver216Rho + OneOver216RhoPlusOne*three*(four*vy2 + four*vz2 - six*vy2*vz2 + vx2*(-two + three*vy2)*(-two + three*vz2)));
- //mfccc = OnoOver216Rho + OneOver216RhoPlusOne*three*(vvz + vx2 + three*vvz*vx2 + vy2 + three*vvz*vy2 + three*vx2*vy2 + nine*vvz*vx2*vy2 + vz2 + three*vx2*vz2 + three*vy2*vz2 + nine*vx2*vy2*vz2 + vvy*(one + three*vx2)*(one + three*vvz + three*vz2) + vvx*(one + three*vvy + three*vy2)*(one + three*vvz + three*vz2));
- //mfaac = OnoOver216Rho + OneOver216RhoPlusOne*three*(vvz + vx2 + three*vvz*vx2 + vy2 + three*vvz*vy2 + three*vx2*vy2 + nine*vvz*vx2*vy2 + vz2 + three*vx2*vz2 + three*vy2*vz2 + nine*vx2*vy2*vz2 - one*vvy*(one + three*vx2)*(one + three*vvz + three*vz2) + vvx*(-one + three*vvy - three*vy2)*(one + three*vvz + three*vz2));
- //mfcac = OnoOver216Rho + OneOver216RhoPlusOne*three*(vvz + vx2 + three*vvz*vx2 + vy2 + three*vvz*vy2 + three*vx2*vy2 + nine*vvz*vx2*vy2 + vz2 + three*vx2*vz2 + three*vy2*vz2 + nine*vx2*vy2*vz2 - one*vvy*(one + three*vx2)*(one + three*vvz + three*vz2) - one*vvx*(-one + three*vvy - three*vy2)*(one + three*vvz + three*vz2));
- //mfacc = OnoOver216Rho + OneOver216RhoPlusOne*three*(vvz + vx2 + three*vvz*vx2 + vy2 + three*vvz*vy2 + three*vx2*vy2 + nine*vvz*vx2*vy2 + vz2 + three*vx2*vz2 + three*vy2*vz2 + nine*vx2*vy2*vz2 + vvy*(one + three*vx2)*(one + three*vvz + three*vz2) - one*vvx*(one + three*vvy + three*vy2)*(one + three*vvz + three*vz2));
- //mfcca = OnoOver216Rho + OneOver216RhoPlusOne*three*(-(one*vvz) + vx2 - three*vvz*vx2 + vy2 - three*vvz*vy2 + three*vx2*vy2 - nine*vvz*vx2*vy2 + vz2 + three*vx2*vz2 + three*vy2*vz2 + nine*vx2*vy2*vz2 - one*vvy*(one + three*vx2)*(-one + three*vvz - three*vz2) - one*vvx*(one + three*vvy + three*vy2)*(-one + three*vvz - three*vz2));
- //mfaaa = OnoOver216Rho - OneOver216RhoPlusOne*three*(vvz - one*vx2 + three*vvz*vx2 - one*vy2 + three*vvz*vy2 - three*vx2*vy2 + nine*vvz*vx2*vy2 - one*vz2 - three*vx2*vz2 - three*vy2*vz2 - nine*vx2*vy2*vz2 - one*vvy*(one + three*vx2)*(-one + three*vvz - three*vz2) + vvx*(-one + three*vvy - three*vy2)*(-one + three*vvz - three*vz2));
- //mfcaa = OnoOver216Rho + OneOver216RhoPlusOne*three*(-(one*vvz) + vx2 - three*vvz*vx2 + vy2 - three*vvz*vy2 + three*vx2*vy2 - nine*vvz*vx2*vy2 + vz2 + three*vx2*vz2 + three*vy2*vz2 + nine*vx2*vy2*vz2 + vvy*(one + three*vx2)*(-one + three*vvz - three*vz2) + vvx*(-one + three*vvy - three*vy2)*(-one + three*vvz - three*vz2));
- //mfaca = OnoOver216Rho + OneOver216RhoPlusOne*three*(-(one*vvz) + vx2 - three*vvz*vx2 + vy2 - three*vvz*vy2 + three*vx2*vy2 - nine*vvz*vx2*vy2 + vz2 + three*vx2*vz2 + three*vy2*vz2 + nine*vx2*vy2*vz2 - one*vvy*(one + three*vx2)*(-one + three*vvz - three*vz2) + vvx*(one + three*vvy + three*vy2)*(-one + three*vvz - three*vz2));
+ //////////////////////////////////////////////////////////////////////////////////
+ //unkonditioniert
+ mfcbb = -(rhoBC[k] + c1o1) * XXc * YYb * ZZb - c2o27;
+ mfabb = -(rhoBC[k] + c1o1) * XXa * YYb * ZZb - c2o27;
+ mfbcb = -(rhoBC[k] + c1o1) * XXb * YYc * ZZb - c2o27;
+ mfbab = -(rhoBC[k] + c1o1) * XXb * YYa * ZZb - c2o27;
+ mfbbc = -(rhoBC[k] + c1o1) * XXb * YYb * ZZc - c2o27;
+ mfbba = -(rhoBC[k] + c1o1) * XXb * YYb * ZZa - c2o27;
+ mfccb = -(rhoBC[k] + c1o1) * XXc * YYc * ZZb - c1o54;
+ mfaab = -(rhoBC[k] + c1o1) * XXa * YYa * ZZb - c1o54;
+ mfcab = -(rhoBC[k] + c1o1) * XXc * YYa * ZZb - c1o54;
+ mfacb = -(rhoBC[k] + c1o1) * XXa * YYc * ZZb - c1o54;
+ mfcbc = -(rhoBC[k] + c1o1) * XXc * YYb * ZZc - c1o54;
+ mfaba = -(rhoBC[k] + c1o1) * XXa * YYb * ZZa - c1o54;
+ mfcba = -(rhoBC[k] + c1o1) * XXc * YYb * ZZa - c1o54;
+ mfabc = -(rhoBC[k] + c1o1) * XXa * YYb * ZZc - c1o54;
+ mfbcc = -(rhoBC[k] + c1o1) * XXb * YYc * ZZc - c1o54;
+ mfbaa = -(rhoBC[k] + c1o1) * XXb * YYa * ZZa - c1o54;
+ mfbca = -(rhoBC[k] + c1o1) * XXb * YYc * ZZa - c1o54;
+ mfbac = -(rhoBC[k] + c1o1) * XXb * YYa * ZZc - c1o54;
+ mfbbb = -(rhoBC[k] + c1o1) * XXb * YYb * ZZb - c8o27;
+ mfccc = -(rhoBC[k] + c1o1) * XXc * YYc * ZZc - c1o216;
+ mfaac = -(rhoBC[k] + c1o1) * XXa * YYa * ZZc - c1o216;
+ mfcac = -(rhoBC[k] + c1o1) * XXc * YYa * ZZc - c1o216;
+ mfacc = -(rhoBC[k] + c1o1) * XXa * YYc * ZZc - c1o216;
+ mfcca = -(rhoBC[k] + c1o1) * XXc * YYc * ZZa - c1o216;
+ mfaaa = -(rhoBC[k] + c1o1) * XXa * YYa * ZZa - c1o216;
+ mfcaa = -(rhoBC[k] + c1o1) * XXc * YYa * ZZa - c1o216;
+ mfaca = -(rhoBC[k] + c1o1) * XXa * YYc * ZZa - c1o216;
+ //////////////////////////////////////////////////////////
+ ////konditioniert
+ //double OneOver216RhoPlusOne = c1over216*(rhoBC[k]+one);
+ //double OnoOver216Rho = c1over216*rhoBC[k];
+ //mfcbb = OnoOver216Rho*sixteen + OneOver216RhoPlusOne*twelve*(-(two*vy2) - two*vz2 + three*vy2*vz2 + vvx*(-two + three*vy2)*(-two + three*vz2) + vx2*(-two + three*vy2)*(-two + three*vz2));
+ //mfabb = OnoOver216Rho*sixteen - OneOver216RhoPlusOne*twelve*(two*vy2 + two*vz2 - three*vy2*vz2 + vvx*(-two + three*vy2)*(-two + three*vz2) + vx2*(-four + six*vy2 + six*vz2 - nine*vy2*vz2));
+ //mfbcb = four*(-(four*OneOver216RhoPlusOne) + four*OnoOver216Rho + OneOver216RhoPlusOne*(-two + three*vx2)*(one + three*vvy + three*vy2)*(-two + three*vz2));
+ //mfbab = four*(four*OnoOver216Rho - OneOver216RhoPlusOne*three*(vvy*(-two + three*vx2)*(-two + three*vz2) - one*vx2*(one + three*vy2)*(-two + three*vz2) + two*(-(two*vy2) + vz2 + three*vy2*vz2)));
+ //mfbbc = four*(-(four*OneOver216RhoPlusOne) + four*OnoOver216Rho + OneOver216RhoPlusOne*(-two + three*vx2)*(-two + three*vy2)*(one + three*vvz + three*vz2));
+ //mfbba = four*(four*OnoOver216Rho - OneOver216RhoPlusOne*three*(vvz*(-two + three*vx2)*(-two + three*vy2) - one*vx2*(-two + three*vy2)*(one + three*vz2) + two*(vy2 - two*vz2 + three*vy2*vz2)));
+ //mfccb = -(two*(-(OnoOver216Rho*two) + OneOver216RhoPlusOne*three*(-(two*vx2) - two*vy2 - six*vx2*vy2 + vz2 + three*vx2*vz2 + three*vy2*vz2 + nine*vx2*vy2*vz2 + vvy*(one + three*vx2)*(-two + three*vz2) + vvx*(one + three*vvy + three*vy2)*(-two + three*vz2))));
+ //mfaab = -(two*(-(OnoOver216Rho*two) + OneOver216RhoPlusOne*three*(-(two*vx2) - two*vy2 - six*vx2*vy2 + vz2 + three*vx2*vz2 + three*vy2*vz2 + nine*vx2*vy2*vz2 - one*vvy*(one + three*vx2)*(-two + three*vz2) + vvx*(-one + three*vvy - three*vy2)*(-two + three*vz2))));
+ //mfcab = two*(OnoOver216Rho*two + OneOver216RhoPlusOne*three*(two*vx2 + two*vy2 + six*vx2*vy2 - one*vz2 - three*vx2*vz2 - three*vy2*vz2 - nine*vx2*vy2*vz2 + vvy*(one + three*vx2)*(-two + three*vz2) + vvx*(-one + three*vvy - three*vy2)*(-two + three*vz2)));
+ //mfacb = two*(OnoOver216Rho*two + OneOver216RhoPlusOne*three*(two*vx2 + two*vy2 + six*vx2*vy2 - one*vz2 - three*vx2*vz2 - three*vy2*vz2 - nine*vx2*vy2*vz2 - one*vvy*(one + three*vx2)*(-two + three*vz2) + vvx*(one + three*vvy + three*vy2)*(-two + three*vz2)));
+ //mfcbc = -(two*(-(OnoOver216Rho*two) + OneOver216RhoPlusOne*three*(-(two*vx2) + vy2 + three*vx2*vy2 + vvz*(one + three*vx2)*(-two + three*vy2) - two*vz2 - six*vx2*vz2 + three*vy2*vz2 + nine*vx2*vy2*vz2 + vvx*(-two + three*vy2)*(one + three*vvz + three*vz2))));
+ //mfaba = -(two*(-(OnoOver216Rho*two) + OneOver216RhoPlusOne*three*(-(two*vx2) + vy2 + three*vx2*vy2 - one*vvz*(one + three*vx2)*(-two + three*vy2) - two*vz2 - six*vx2*vz2 + three*vy2*vz2 + nine*vx2*vy2*vz2 + vvx*(-two + three*vy2)*(-one + three*vvz - three*vz2))));
+ //mfcba = two*(OnoOver216Rho*two + OneOver216RhoPlusOne*three*(two*vx2 - one*vy2 - three*vx2*vy2 + vvz*(one + three*vx2)*(-two + three*vy2) + two*vz2 + six*vx2*vz2 - three*vy2*vz2 - nine*vx2*vy2*vz2 + vvx*(-two + three*vy2)*(-one + three*vvz - three*vz2)));
+ //mfabc = two*(OnoOver216Rho*two + OneOver216RhoPlusOne*three*(two*vx2 - one*vy2 - three*vx2*vy2 - one*vvz*(one + three*vx2)*(-two + three*vy2) + two*vz2 + six*vx2*vz2 - three*vy2*vz2 - nine*vx2*vy2*vz2 + vvx*(-two + three*vy2)*(one + three*vvz + three*vz2)));
+ //mfbcc = -(two*(-(OnoOver216Rho*two) + OneOver216RhoPlusOne*three*(vx2 - two*vy2 + three*vx2*vy2 + vvz*(-two + three*vx2)*(one + three*vy2) - two*vz2 + three*vx2*vz2 - six*vy2*vz2 + nine*vx2*vy2*vz2 + vvy*(-two + three*vx2)*(one + three*vvz + three*vz2))));
+ //mfbaa = -(two*(-(OnoOver216Rho*two) + OneOver216RhoPlusOne*three*(vx2 - two*vy2 + three*vx2*vy2 - one*vvz*(-two + three*vx2)*(one + three*vy2) - two*vz2 + three*vx2*vz2 - six*vy2*vz2 + nine*vx2*vy2*vz2 + vvy*(-two + three*vx2)*(-one + three*vvz - three*vz2))));
+ //mfbca = two*(OnoOver216Rho*two + OneOver216RhoPlusOne*three*(-(one*vx2) + two*vy2 - three*vx2*vy2 + vvz*(-two + three*vx2)*(one + three*vy2) + two*vz2 - three*vx2*vz2 + six*vy2*vz2 - nine*vx2*vy2*vz2 + vvy*(-two + three*vx2)*(-one + three*vvz - three*vz2)));
+ //mfbac = two*(OnoOver216Rho*two + OneOver216RhoPlusOne*three*(-(one*vx2) + two*vy2 - three*vx2*vy2 - one*vvz*(-two + three*vx2)*(one + three*vy2) + two*vz2 - three*vx2*vz2 + six*vy2*vz2 - nine*vx2*vy2*vz2 + vvy*(-two + three*vx2)*(one + three*vvz + three*vz2)));
+ //mfbbb = eight*(eight*OnoOver216Rho + OneOver216RhoPlusOne*three*(four*vy2 + four*vz2 - six*vy2*vz2 + vx2*(-two + three*vy2)*(-two + three*vz2)));
+ //mfccc = OnoOver216Rho + OneOver216RhoPlusOne*three*(vvz + vx2 + three*vvz*vx2 + vy2 + three*vvz*vy2 + three*vx2*vy2 + nine*vvz*vx2*vy2 + vz2 + three*vx2*vz2 + three*vy2*vz2 + nine*vx2*vy2*vz2 + vvy*(one + three*vx2)*(one + three*vvz + three*vz2) + vvx*(one + three*vvy + three*vy2)*(one + three*vvz + three*vz2));
+ //mfaac = OnoOver216Rho + OneOver216RhoPlusOne*three*(vvz + vx2 + three*vvz*vx2 + vy2 + three*vvz*vy2 + three*vx2*vy2 + nine*vvz*vx2*vy2 + vz2 + three*vx2*vz2 + three*vy2*vz2 + nine*vx2*vy2*vz2 - one*vvy*(one + three*vx2)*(one + three*vvz + three*vz2) + vvx*(-one + three*vvy - three*vy2)*(one + three*vvz + three*vz2));
+ //mfcac = OnoOver216Rho + OneOver216RhoPlusOne*three*(vvz + vx2 + three*vvz*vx2 + vy2 + three*vvz*vy2 + three*vx2*vy2 + nine*vvz*vx2*vy2 + vz2 + three*vx2*vz2 + three*vy2*vz2 + nine*vx2*vy2*vz2 - one*vvy*(one + three*vx2)*(one + three*vvz + three*vz2) - one*vvx*(-one + three*vvy - three*vy2)*(one + three*vvz + three*vz2));
+ //mfacc = OnoOver216Rho + OneOver216RhoPlusOne*three*(vvz + vx2 + three*vvz*vx2 + vy2 + three*vvz*vy2 + three*vx2*vy2 + nine*vvz*vx2*vy2 + vz2 + three*vx2*vz2 + three*vy2*vz2 + nine*vx2*vy2*vz2 + vvy*(one + three*vx2)*(one + three*vvz + three*vz2) - one*vvx*(one + three*vvy + three*vy2)*(one + three*vvz + three*vz2));
+ //mfcca = OnoOver216Rho + OneOver216RhoPlusOne*three*(-(one*vvz) + vx2 - three*vvz*vx2 + vy2 - three*vvz*vy2 + three*vx2*vy2 - nine*vvz*vx2*vy2 + vz2 + three*vx2*vz2 + three*vy2*vz2 + nine*vx2*vy2*vz2 - one*vvy*(one + three*vx2)*(-one + three*vvz - three*vz2) - one*vvx*(one + three*vvy + three*vy2)*(-one + three*vvz - three*vz2));
+ //mfaaa = OnoOver216Rho - OneOver216RhoPlusOne*three*(vvz - one*vx2 + three*vvz*vx2 - one*vy2 + three*vvz*vy2 - three*vx2*vy2 + nine*vvz*vx2*vy2 - one*vz2 - three*vx2*vz2 - three*vy2*vz2 - nine*vx2*vy2*vz2 - one*vvy*(one + three*vx2)*(-one + three*vvz - three*vz2) + vvx*(-one + three*vvy - three*vy2)*(-one + three*vvz - three*vz2));
+ //mfcaa = OnoOver216Rho + OneOver216RhoPlusOne*three*(-(one*vvz) + vx2 - three*vvz*vx2 + vy2 - three*vvz*vy2 + three*vx2*vy2 - nine*vvz*vx2*vy2 + vz2 + three*vx2*vz2 + three*vy2*vz2 + nine*vx2*vy2*vz2 + vvy*(one + three*vx2)*(-one + three*vvz - three*vz2) + vvx*(-one + three*vvy - three*vy2)*(-one + three*vvz - three*vz2));
+ //mfaca = OnoOver216Rho + OneOver216RhoPlusOne*three*(-(one*vvz) + vx2 - three*vvz*vx2 + vy2 - three*vvz*vy2 + three*vx2*vy2 - nine*vvz*vx2*vy2 + vz2 + three*vx2*vz2 + three*vy2*vz2 + nine*vx2*vy2*vz2 - one*vvy*(one + three*vx2)*(-one + three*vvz - three*vz2) + vvx*(one + three*vvy + three*vy2)*(-one + three*vvz - three*vz2));
//////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
//if (isEvenTimestep==true)
//{
- // D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- // D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- // D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- // D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- // D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- // D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- // D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- // D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- // D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- // D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- // D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- // D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- // D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- // D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- // D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- // D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- // D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- // D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- // D.f[DIR_000] = &DD[DIR_000*size_Mat];
- // D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- // D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- // D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- // D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- // D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- // D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- // D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- // D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
- //}
+ // D.f[DIR_P00] = &DD[DIR_P00 * size_Mat];
+ // D.f[DIR_M00] = &DD[DIR_M00 * size_Mat];
+ // D.f[DIR_0P0] = &DD[DIR_0P0 * size_Mat];
+ // D.f[DIR_0M0] = &DD[DIR_0M0 * size_Mat];
+ // D.f[DIR_00P] = &DD[DIR_00P * size_Mat];
+ // D.f[DIR_00M] = &DD[DIR_00M * size_Mat];
+ // D.f[DIR_PP0] = &DD[DIR_PP0 * size_Mat];
+ // D.f[DIR_MM0] = &DD[DIR_MM0 * size_Mat];
+ // D.f[DIR_PM0] = &DD[DIR_PM0 * size_Mat];
+ // D.f[DIR_MP0] = &DD[DIR_MP0 * size_Mat];
+ // D.f[DIR_P0P] = &DD[DIR_P0P * size_Mat];
+ // D.f[DIR_M0M] = &DD[DIR_M0M * size_Mat];
+ // D.f[DIR_P0M] = &DD[DIR_P0M * size_Mat];
+ // D.f[DIR_M0P] = &DD[DIR_M0P * size_Mat];
+ // D.f[DIR_0PP] = &DD[DIR_0PP * size_Mat];
+ // D.f[DIR_0MM] = &DD[DIR_0MM * size_Mat];
+ // D.f[DIR_0PM] = &DD[DIR_0PM * size_Mat];
+ // D.f[DIR_0MP] = &DD[DIR_0MP * size_Mat];
+ // D.f[DIR_000] = &DD[DIR_000 * size_Mat];
+ // D.f[DIR_PPP] = &DD[DIR_PPP * size_Mat];
+ // D.f[DIR_MMP] = &DD[DIR_MMP * size_Mat];
+ // D.f[DIR_PMP] = &DD[DIR_PMP * size_Mat];
+ // D.f[DIR_MPP] = &DD[DIR_MPP * size_Mat];
+ // D.f[DIR_PPM] = &DD[DIR_PPM * size_Mat];
+ // D.f[DIR_MMM] = &DD[DIR_MMM * size_Mat];
+ // D.f[DIR_PMM] = &DD[DIR_PMM * size_Mat];
+ // D.f[DIR_MPM] = &DD[DIR_MPM * size_Mat];
+ //}
//else
//{
- // D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
- // D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
- // D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
- // D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
- // D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
- // D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
- // D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
- // D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
- // D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
- // D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
- // D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
- // D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
- // D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
- // D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
- // D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
- // D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
- // D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
- // D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
- // D.f[DIR_000] = &DD[DIR_000*size_Mat];
- // D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
- // D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
- // D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
- // D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
- // D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
- // D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
- // D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
- // D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
+ // D.f[DIR_M00] = &DD[DIR_P00 * size_Mat];
+ // D.f[DIR_P00] = &DD[DIR_M00 * size_Mat];
+ // D.f[DIR_0M0] = &DD[DIR_0P0 * size_Mat];
+ // D.f[DIR_0P0] = &DD[DIR_0M0 * size_Mat];
+ // D.f[DIR_00M] = &DD[DIR_00P * size_Mat];
+ // D.f[DIR_00P] = &DD[DIR_00M * size_Mat];
+ // D.f[DIR_MM0] = &DD[DIR_PP0 * size_Mat];
+ // D.f[DIR_PP0] = &DD[DIR_MM0 * size_Mat];
+ // D.f[DIR_MP0] = &DD[DIR_PM0 * size_Mat];
+ // D.f[DIR_PM0] = &DD[DIR_MP0 * size_Mat];
+ // D.f[DIR_M0M] = &DD[DIR_P0P * size_Mat];
+ // D.f[DIR_P0P] = &DD[DIR_M0M * size_Mat];
+ // D.f[DIR_M0P] = &DD[DIR_P0M * size_Mat];
+ // D.f[DIR_P0M] = &DD[DIR_M0P * size_Mat];
+ // D.f[DIR_0MM] = &DD[DIR_0PP * size_Mat];
+ // D.f[DIR_0PP] = &DD[DIR_0MM * size_Mat];
+ // D.f[DIR_0MP] = &DD[DIR_0PM * size_Mat];
+ // D.f[DIR_0PM] = &DD[DIR_0MP * size_Mat];
+ // D.f[DIR_000] = &DD[DIR_000 * size_Mat];
+ // D.f[DIR_PPP] = &DD[DIR_MMM * size_Mat];
+ // D.f[DIR_MMP] = &DD[DIR_PPM * size_Mat];
+ // D.f[DIR_PMP] = &DD[DIR_MPM * size_Mat];
+ // D.f[DIR_MPP] = &DD[DIR_PMM * size_Mat];
+ // D.f[DIR_PPM] = &DD[DIR_MMP * size_Mat];
+ // D.f[DIR_MMM] = &DD[DIR_PPP * size_Mat];
+ // D.f[DIR_PMM] = &DD[DIR_MPP * size_Mat];
+ // D.f[DIR_MPM] = &DD[DIR_PMP * size_Mat];
//}
//////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
//__syncthreads();
- (D.f[DIR_P00 ])[ke ] = mfabb;//mfcbb;
- (D.f[DIR_M00 ])[kw ] = mfcbb;//mfabb;
- (D.f[DIR_0P0 ])[kn ] = mfbab;//mfbcb;
- (D.f[DIR_0M0 ])[ks ] = mfbcb;//mfbab;
- (D.f[DIR_00P ])[kt ] = mfbba;//mfbbc;
- (D.f[DIR_00M ])[kb ] = mfbbc;//mfbba;
- (D.f[DIR_PP0 ])[kne ] = mfaab;//mfccb;
- (D.f[DIR_MM0 ])[ksw ] = mfccb;//mfaab;
- (D.f[DIR_PM0 ])[kse ] = mfacb;//mfcab;
- (D.f[DIR_MP0 ])[knw ] = mfcab;//mfacb;
- (D.f[DIR_P0P ])[kte ] = mfaba;//mfcbc;
- (D.f[DIR_M0M ])[kbw ] = mfcbc;//mfaba;
- (D.f[DIR_P0M ])[kbe ] = mfabc;//mfcba;
- (D.f[DIR_M0P ])[ktw ] = mfcba;//mfabc;
- (D.f[DIR_0PP ])[ktn ] = mfbaa;//mfbcc;
- (D.f[DIR_0MM ])[kbs ] = mfbcc;//mfbaa;
- (D.f[DIR_0PM ])[kbn ] = mfbac;//mfbca;
- (D.f[DIR_0MP ])[kts ] = mfbca;//mfbac;
- (D.f[DIR_000])[kzero] = mfbbb;//mfbbb;
- (D.f[DIR_PPP ])[ktne ] = mfaaa;//mfccc;
- (D.f[DIR_MMP ])[ktsw ] = mfcca;//mfaac;
- (D.f[DIR_PMP ])[ktse ] = mfaca;//mfcac;
- (D.f[DIR_MPP ])[ktnw ] = mfcaa;//mfacc;
- (D.f[DIR_PPM ])[kbne ] = mfaac;//mfcca;
- (D.f[DIR_MMM ])[kbsw ] = mfccc;//mfaaa;
- (D.f[DIR_PMM ])[kbse ] = mfacc;//mfcaa;
- (D.f[DIR_MPM ])[kbnw ] = mfcac;//mfaca;
- //(D.f[DIR_P00 ])[ke ] = mfcbb;
- //(D.f[DIR_M00 ])[kw ] = mfabb;
- //(D.f[DIR_0P0 ])[kn ] = mfbcb;
- //(D.f[DIR_0M0 ])[ks ] = mfbab;
- //(D.f[DIR_00P ])[kt ] = mfbbc;
- //(D.f[DIR_00M ])[kb ] = mfbba;
- //(D.f[DIR_PP0 ])[kne ] = mfccb;
- //(D.f[DIR_MM0 ])[ksw ] = mfaab;
- //(D.f[DIR_PM0 ])[kse ] = mfcab;
- //(D.f[DIR_MP0 ])[knw ] = mfacb;
- //(D.f[DIR_P0P ])[kte ] = mfcbc;
- //(D.f[DIR_M0M ])[kbw ] = mfaba;
- //(D.f[DIR_P0M ])[kbe ] = mfcba;
- //(D.f[DIR_M0P ])[ktw ] = mfabc;
- //(D.f[DIR_0PP ])[ktn ] = mfbcc;
- //(D.f[DIR_0MM ])[kbs ] = mfbaa;
- //(D.f[DIR_0PM ])[kbn ] = mfbca;
- //(D.f[DIR_0MP ])[kts ] = mfbac;
- //(D.f[DIR_000])[kzero] = mfbbb;
- //(D.f[DIR_PPP ])[ktne ] = mfccc;
- //(D.f[DIR_MMP ])[ktsw ] = mfaac;
- //(D.f[DIR_PMP ])[ktse ] = mfcac;
- //(D.f[DIR_MPP ])[ktnw ] = mfacc;
- //(D.f[DIR_PPM ])[kbne ] = mfcca;
- //(D.f[DIR_MMM ])[kbsw ] = mfaaa;
- //(D.f[DIR_PMM ])[kbse ] = mfcaa;
- //(D.f[DIR_MPM ])[kbnw ] = mfaca;
-
- //(D.f[DIR_P00 ])[ke ] = fE ; //f1_E ; //fW; //fE ;
- //(D.f[DIR_M00 ])[kw ] = fW ; //f1_W ; //fE; //fW ;
- //(D.f[DIR_0P0 ])[kn ] = fN ; //f1_N ; //fS; //fN ;
- //(D.f[DIR_0M0 ])[ks ] = fS ; //f1_S ; //fN; //fS ;
- //(D.f[DIR_00P ])[kt ] = fT ; //f1_T ; //fB; //fT ;
- //(D.f[DIR_00M ])[kb ] = fB ; //f1_B ; //fT; //fB ;
- //(D.f[DIR_PP0 ])[kne ] = fNE; //f1_NE; //fSW; //fNE;
- //(D.f[DIR_MM0 ])[ksw ] = fSW; //f1_SW; //fNE; //fSW;
- //(D.f[DIR_PM0 ])[kse ] = fSE; //f1_SE; //fNW; //fSE;
- //(D.f[DIR_MP0 ])[knw ] = fNW; //f1_NW; //fSE; //fNW;
- //(D.f[DIR_P0P ])[kte ] = fTE; //f1_TE; //fBW; //fTE;
- //(D.f[DIR_M0M ])[kbw ] = fBW; //f1_BW; //fTE; //fBW;
- //(D.f[DIR_P0M ])[kbe ] = fBE; //f1_BE; //fTW; //fBE;
- //(D.f[DIR_M0P ])[ktw ] = fTW; //f1_TW; //fBE; //fTW;
- //(D.f[DIR_0PP ])[ktn ] = fTN; //f1_TN; //fBS; //fTN;
- //(D.f[DIR_0MM ])[kbs ] = fBS; //f1_BS; //fTN; //fBS;
- //(D.f[DIR_0PM ])[kbn ] = fBN; //f1_BN; //fTS; //fBN;
- //(D.f[DIR_0MP ])[kts ] = fTS; //f1_TS; //fBN; //fTS;
+ (D.f[DIR_P00])[ke ] = mfabb;//mfcbb;
+ (D.f[DIR_M00])[kw ] = mfcbb;//mfabb;
+ (D.f[DIR_0P0])[kn ] = mfbab;//mfbcb;
+ (D.f[DIR_0M0])[ks ] = mfbcb;//mfbab;
+ (D.f[DIR_00P])[kt ] = mfbba;//mfbbc;
+ (D.f[DIR_00M])[kb ] = mfbbc;//mfbba;
+ (D.f[DIR_PP0])[kne ] = mfaab;//mfccb;
+ (D.f[DIR_MM0])[ksw ] = mfccb;//mfaab;
+ (D.f[DIR_PM0])[kse ] = mfacb;//mfcab;
+ (D.f[DIR_MP0])[knw ] = mfcab;//mfacb;
+ (D.f[DIR_P0P])[kte ] = mfaba;//mfcbc;
+ (D.f[DIR_M0M])[kbw ] = mfcbc;//mfaba;
+ (D.f[DIR_P0M])[kbe ] = mfabc;//mfcba;
+ (D.f[DIR_M0P])[ktw ] = mfcba;//mfabc;
+ (D.f[DIR_0PP])[ktn ] = mfbaa;//mfbcc;
+ (D.f[DIR_0MM])[kbs ] = mfbcc;//mfbaa;
+ (D.f[DIR_0PM])[kbn ] = mfbac;//mfbca;
+ (D.f[DIR_0MP])[kts ] = mfbca;//mfbac;
+ (D.f[DIR_000])[kzero] = mfbbb;//mfbbb;
+ (D.f[DIR_PPP])[ktne ] = mfaaa;//mfccc;
+ (D.f[DIR_MMP])[ktsw ] = mfcca;//mfaac;
+ (D.f[DIR_PMP])[ktse ] = mfaca;//mfcac;
+ (D.f[DIR_MPP])[ktnw ] = mfcaa;//mfacc;
+ (D.f[DIR_PPM])[kbne ] = mfaac;//mfcca;
+ (D.f[DIR_MMM])[kbsw ] = mfccc;//mfaaa;
+ (D.f[DIR_PMM])[kbse ] = mfacc;//mfcaa;
+ (D.f[DIR_MPM])[kbnw ] = mfcac;//mfaca;
+ //(D.f[DIR_P00])[ke ] = mfcbb;
+ //(D.f[DIR_M00])[kw ] = mfabb;
+ //(D.f[DIR_0P0])[kn ] = mfbcb;
+ //(D.f[DIR_0M0])[ks ] = mfbab;
+ //(D.f[DIR_00P])[kt ] = mfbbc;
+ //(D.f[DIR_00M])[kb ] = mfbba;
+ //(D.f[DIR_PP0])[kne ] = mfccb;
+ //(D.f[DIR_MM0])[ksw ] = mfaab;
+ //(D.f[DIR_PM0])[kse ] = mfcab;
+ //(D.f[DIR_MP0])[knw ] = mfacb;
+ //(D.f[DIR_P0P])[kte ] = mfcbc;
+ //(D.f[DIR_M0M])[kbw ] = mfaba;
+ //(D.f[DIR_P0M])[kbe ] = mfcba;
+ //(D.f[DIR_M0P])[ktw ] = mfabc;
+ //(D.f[DIR_0PP])[ktn ] = mfbcc;
+ //(D.f[DIR_0MM])[kbs ] = mfbaa;
+ //(D.f[DIR_0PM])[kbn ] = mfbca;
+ //(D.f[DIR_0MP])[kts ] = mfbac;
+ //(D.f[DIR_000])[kzero] = mfbbb;
+ //(D.f[DIR_PPP])[ktne ] = mfccc;
+ //(D.f[DIR_MMP])[ktsw ] = mfaac;
+ //(D.f[DIR_PMP])[ktse ] = mfcac;
+ //(D.f[DIR_MPP])[ktnw ] = mfacc;
+ //(D.f[DIR_PPM])[kbne ] = mfcca;
+ //(D.f[DIR_MMM])[kbsw ] = mfaaa;
+ //(D.f[DIR_PMM])[kbse ] = mfcaa;
+ //(D.f[DIR_MPM])[kbnw ] = mfaca;
+
+ //(D.f[DIR_P00])[ke ] = fE ; //f1_E ; //fW; //fE ;
+ //(D.f[DIR_M00])[kw ] = fW ; //f1_W ; //fE; //fW ;
+ //(D.f[DIR_0P0])[kn ] = fN ; //f1_N ; //fS; //fN ;
+ //(D.f[DIR_0M0])[ks ] = fS ; //f1_S ; //fN; //fS ;
+ //(D.f[DIR_00P])[kt ] = fT ; //f1_T ; //fB; //fT ;
+ //(D.f[DIR_00M])[kb ] = fB ; //f1_B ; //fT; //fB ;
+ //(D.f[DIR_PP0])[kne ] = fNE; //f1_NE; //fSW; //fNE;
+ //(D.f[DIR_MM0])[ksw ] = fSW; //f1_SW; //fNE; //fSW;
+ //(D.f[DIR_PM0])[kse ] = fSE; //f1_SE; //fNW; //fSE;
+ //(D.f[DIR_MP0])[knw ] = fNW; //f1_NW; //fSE; //fNW;
+ //(D.f[DIR_P0P])[kte ] = fTE; //f1_TE; //fBW; //fTE;
+ //(D.f[DIR_M0M])[kbw ] = fBW; //f1_BW; //fTE; //fBW;
+ //(D.f[DIR_P0M])[kbe ] = fBE; //f1_BE; //fTW; //fBE;
+ //(D.f[DIR_M0P])[ktw ] = fTW; //f1_TW; //fBE; //fTW;
+ //(D.f[DIR_0PP])[ktn ] = fTN; //f1_TN; //fBS; //fTN;
+ //(D.f[DIR_0MM])[kbs ] = fBS; //f1_BS; //fTN; //fBS;
+ //(D.f[DIR_0PM])[kbn ] = fBN; //f1_BN; //fTS; //fBN;
+ //(D.f[DIR_0MP])[kts ] = fTS; //f1_TS; //fBN; //fTS;
//(D.f[DIR_000])[kzero] = fZERO;//f1_ZERO; //fZERO; //fZERO;
- //(D.f[DIR_PPP ])[ktne ] = fTNE; //f1_TNE; //fBSW; //fTNE;
- //(D.f[DIR_MMM ])[kbsw ] = fBSW; //f1_BSW; //fTNE; //fBSW;
- //(D.f[DIR_PPM ])[kbne ] = fBNE; //f1_BNE; //fTSW; //fBNE;
- //(D.f[DIR_MMP ])[ktsw ] = fTSW; //f1_TSW; //fBNE; //fTSW;
- //(D.f[DIR_PMP ])[ktse ] = fTSE; //f1_TSE; //fBNW; //fTSE;
- //(D.f[DIR_MPM ])[kbnw ] = fBNW; //f1_BNW; //fTSE; //fBNW;
- //(D.f[DIR_PMM ])[kbse ] = fBSE; //f1_BSE; //fTNW; //fBSE;
- //(D.f[DIR_MPP ])[ktnw ] = fTNW; //f1_TNW; //fBSE; //fTNW;
+ //(D.f[DIR_PPP])[ktne ] = fTNE; //f1_TNE; //fBSW; //fTNE;
+ //(D.f[DIR_MMM])[kbsw ] = fBSW; //f1_BSW; //fTNE; //fBSW;
+ //(D.f[DIR_PPM])[kbne ] = fBNE; //f1_BNE; //fTSW; //fBNE;
+ //(D.f[DIR_MMP])[ktsw ] = fTSW; //f1_TSW; //fBNE; //fTSW;
+ //(D.f[DIR_PMP])[ktse ] = fTSE; //f1_TSE; //fBNW; //fTSE;
+ //(D.f[DIR_MPM])[kbnw ] = fBNW; //f1_BNW; //fTSE; //fBNW;
+ //(D.f[DIR_PMM])[kbse ] = fBSE; //f1_BSE; //fTNW; //fBSE;
+ //(D.f[DIR_MPP])[ktnw ] = fTNW; //f1_TNW; //fBSE; //fTNW;
}
}
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
@@ -4354,19 +4369,20 @@ __global__ void QPressDeviceEQZ27(real* rhoBC,
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
-__global__ void QPressDeviceZero27( real* DD,
- int* k_Q,
- unsigned int numberOfBCnodes,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep)
+__global__ void QPressDeviceZero27(
+ real* DD,
+ int* k_Q,
+ unsigned int numberOfBCnodes,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep)
{
////////////////////////////////////////////////////////////////////////////////
- const unsigned x = threadIdx.x; // Globaler x-Index
- const unsigned y = blockIdx.x; // Globaler y-Index
- const unsigned z = blockIdx.y; // Globaler z-Index
+ const unsigned x = threadIdx.x; // Globaler x-Index
+ const unsigned y = blockIdx.x; // Globaler y-Index
+ const unsigned z = blockIdx.y; // Globaler z-Index
const unsigned nx = blockDim.x;
const unsigned ny = gridDim.x;
@@ -4410,94 +4426,94 @@ __global__ void QPressDeviceZero27( real* DD,
Distributions27 D;
if (isEvenTimestep==false)
{
- D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
- }
+ D.f[DIR_P00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_MPM * numberOfLBnodes];
+ }
else
{
- D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
+ D.f[DIR_M00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_MPM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_PMP * numberOfLBnodes];
}
//////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
//__syncthreads();
- //////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
- (D.f[DIR_P00 ])[ke ] =c0o1;
- (D.f[DIR_M00 ])[kw ] =c0o1;
- (D.f[DIR_0P0 ])[kn ] =c0o1;
- (D.f[DIR_0M0 ])[ks ] =c0o1;
- (D.f[DIR_00P ])[kt ] =c0o1;
- (D.f[DIR_00M ])[kb ] =c0o1;
- (D.f[DIR_PP0 ])[kne ] =c0o1;
- (D.f[DIR_MM0 ])[ksw ] =c0o1;
- (D.f[DIR_PM0 ])[kse ] =c0o1;
- (D.f[DIR_MP0 ])[knw ] =c0o1;
- (D.f[DIR_P0P ])[kte ] =c0o1;
- (D.f[DIR_M0M ])[kbw ] =c0o1;
- (D.f[DIR_P0M ])[kbe ] =c0o1;
- (D.f[DIR_M0P ])[ktw ] =c0o1;
- (D.f[DIR_0PP ])[ktn ] =c0o1;
- (D.f[DIR_0MM ])[kbs ] =c0o1;
- (D.f[DIR_0PM ])[kbn ] =c0o1;
- (D.f[DIR_0MP ])[kts ] =c0o1;
+ //////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
+ (D.f[DIR_P00])[ke ] =c0o1;
+ (D.f[DIR_M00])[kw ] =c0o1;
+ (D.f[DIR_0P0])[kn ] =c0o1;
+ (D.f[DIR_0M0])[ks ] =c0o1;
+ (D.f[DIR_00P])[kt ] =c0o1;
+ (D.f[DIR_00M])[kb ] =c0o1;
+ (D.f[DIR_PP0])[kne ] =c0o1;
+ (D.f[DIR_MM0])[ksw ] =c0o1;
+ (D.f[DIR_PM0])[kse ] =c0o1;
+ (D.f[DIR_MP0])[knw ] =c0o1;
+ (D.f[DIR_P0P])[kte ] =c0o1;
+ (D.f[DIR_M0M])[kbw ] =c0o1;
+ (D.f[DIR_P0M])[kbe ] =c0o1;
+ (D.f[DIR_M0P])[ktw ] =c0o1;
+ (D.f[DIR_0PP])[ktn ] =c0o1;
+ (D.f[DIR_0MM])[kbs ] =c0o1;
+ (D.f[DIR_0PM])[kbn ] =c0o1;
+ (D.f[DIR_0MP])[kts ] =c0o1;
(D.f[DIR_000])[kzero] =c0o1;
- (D.f[DIR_PPP ])[ktne ] =c0o1;
- (D.f[DIR_MMP ])[ktsw ] =c0o1;
- (D.f[DIR_PMP ])[ktse ] =c0o1;
- (D.f[DIR_MPP ])[ktnw ] =c0o1;
- (D.f[DIR_PPM ])[kbne ] =c0o1;
- (D.f[DIR_MMM ])[kbsw ] =c0o1;
- (D.f[DIR_PMM ])[kbse ] =c0o1;
- (D.f[DIR_MPM ])[kbnw ] =c0o1;
+ (D.f[DIR_PPP])[ktne ] =c0o1;
+ (D.f[DIR_MMP])[ktsw ] =c0o1;
+ (D.f[DIR_PMP])[ktse ] =c0o1;
+ (D.f[DIR_MPP])[ktnw ] =c0o1;
+ (D.f[DIR_PPM])[kbne ] =c0o1;
+ (D.f[DIR_MMM])[kbsw ] =c0o1;
+ (D.f[DIR_PMM])[kbse ] =c0o1;
+ (D.f[DIR_MPM])[kbnw ] =c0o1;
}
}
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
@@ -4541,22 +4557,23 @@ __global__ void QPressDeviceZero27( real* DD,
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
-__global__ void QPressDeviceFake27( real* rhoBC,
- real* DD,
- int* k_Q,
- int* k_N,
- int numberOfBCnodes,
- real om1,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep)
+__global__ void QPressDeviceFake27(
+ real* rhoBC,
+ real* DD,
+ int* k_Q,
+ int* k_N,
+ int numberOfBCnodes,
+ real om1,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep)
{
////////////////////////////////////////////////////////////////////////////////
- const unsigned x = threadIdx.x; // Globaler x-Index
- const unsigned y = blockIdx.x; // Globaler y-Index
- const unsigned z = blockIdx.y; // Globaler z-Index
+ const unsigned x = threadIdx.x; // Globaler x-Index
+ const unsigned y = blockIdx.x; // Globaler y-Index
+ const unsigned z = blockIdx.y; // Globaler z-Index
const unsigned nx = blockDim.x;
const unsigned ny = gridDim.x;
@@ -4630,148 +4647,148 @@ __global__ void QPressDeviceFake27( real* rhoBC,
Distributions27 D;
if (isEvenTimestep==false)
{
- D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
- }
+ D.f[DIR_P00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_MPM * numberOfLBnodes];
+ }
else
{
- D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
+ D.f[DIR_M00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_MPM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_PMP * numberOfLBnodes];
}
//////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
real f1_E,f1_W,f1_N,f1_S,f1_T,f1_B,f1_NE,f1_SW,f1_SE,f1_NW,f1_TE,f1_BW,f1_BE,f1_TW,f1_TN,f1_BS,f1_BN,f1_TS,f1_ZERO,
f1_TNE,f1_TSW,f1_TSE,f1_TNW,f1_BNE,f1_BSW,f1_BSE,f1_BNW;
- f1_W = (D.f[DIR_P00 ])[k1e ];
- f1_E = (D.f[DIR_M00 ])[k1w ];
- f1_S = (D.f[DIR_0P0 ])[k1n ];
- f1_N = (D.f[DIR_0M0 ])[k1s ];
- f1_B = (D.f[DIR_00P ])[k1t ];
- f1_T = (D.f[DIR_00M ])[k1b ];
- f1_SW = (D.f[DIR_PP0 ])[k1ne ];
- f1_NE = (D.f[DIR_MM0 ])[k1sw ];
- f1_NW = (D.f[DIR_PM0 ])[k1se ];
- f1_SE = (D.f[DIR_MP0 ])[k1nw ];
- f1_BW = (D.f[DIR_P0P ])[k1te ];
- f1_TE = (D.f[DIR_M0M ])[k1bw ];
- f1_TW = (D.f[DIR_P0M ])[k1be ];
- f1_BE = (D.f[DIR_M0P ])[k1tw ];
- f1_BS = (D.f[DIR_0PP ])[k1tn ];
- f1_TN = (D.f[DIR_0MM ])[k1bs ];
- f1_TS = (D.f[DIR_0PM ])[k1bn ];
- f1_BN = (D.f[DIR_0MP ])[k1ts ];
+ f1_W = (D.f[DIR_P00])[k1e ];
+ f1_E = (D.f[DIR_M00])[k1w ];
+ f1_S = (D.f[DIR_0P0])[k1n ];
+ f1_N = (D.f[DIR_0M0])[k1s ];
+ f1_B = (D.f[DIR_00P])[k1t ];
+ f1_T = (D.f[DIR_00M])[k1b ];
+ f1_SW = (D.f[DIR_PP0])[k1ne ];
+ f1_NE = (D.f[DIR_MM0])[k1sw ];
+ f1_NW = (D.f[DIR_PM0])[k1se ];
+ f1_SE = (D.f[DIR_MP0])[k1nw ];
+ f1_BW = (D.f[DIR_P0P])[k1te ];
+ f1_TE = (D.f[DIR_M0M])[k1bw ];
+ f1_TW = (D.f[DIR_P0M])[k1be ];
+ f1_BE = (D.f[DIR_M0P])[k1tw ];
+ f1_BS = (D.f[DIR_0PP])[k1tn ];
+ f1_TN = (D.f[DIR_0MM])[k1bs ];
+ f1_TS = (D.f[DIR_0PM])[k1bn ];
+ f1_BN = (D.f[DIR_0MP])[k1ts ];
f1_ZERO = (D.f[DIR_000])[k1zero];
- f1_BSW = (D.f[DIR_PPP ])[k1tne ];
- f1_BNE = (D.f[DIR_MMP ])[k1tsw ];
- f1_BNW = (D.f[DIR_PMP ])[k1tse ];
- f1_BSE = (D.f[DIR_MPP ])[k1tnw ];
- f1_TSW = (D.f[DIR_PPM ])[k1bne ];
- f1_TNE = (D.f[DIR_MMM ])[k1bsw ];
- f1_TNW = (D.f[DIR_PMM ])[k1bse ];
- f1_TSE = (D.f[DIR_MPM ])[k1bnw ];
+ f1_BSW = (D.f[DIR_PPP])[k1tne ];
+ f1_BNE = (D.f[DIR_MMP])[k1tsw ];
+ f1_BNW = (D.f[DIR_PMP])[k1tse ];
+ f1_BSE = (D.f[DIR_MPP])[k1tnw ];
+ f1_TSW = (D.f[DIR_PPM])[k1bne ];
+ f1_TNE = (D.f[DIR_MMM])[k1bsw ];
+ f1_TNW = (D.f[DIR_PMM])[k1bse ];
+ f1_TSE = (D.f[DIR_MPM])[k1bnw ];
////////////////////////////////////////////////////////////////////////////////
real vx1, vx2, vx3;
vx1 = ((f1_TSE - f1_BNW) - (f1_TNW - f1_BSE)) + ((f1_TNE - f1_BSW) - (f1_TSW - f1_BNE)) +
((f1_BE - f1_TW) + (f1_TE - f1_BW)) + ((f1_SE - f1_NW) + (f1_NE - f1_SW)) +
- (f1_E - f1_W);
+ (f1_E - f1_W);
vx2 = (-(f1_TSE - f1_BNW) + (f1_TNW - f1_BSE)) + ((f1_TNE - f1_BSW) - (f1_TSW - f1_BNE)) +
((f1_BN - f1_TS) + (f1_TN - f1_BS)) + (-(f1_SE - f1_NW) + (f1_NE - f1_SW)) +
- (f1_N - f1_S);
+ (f1_N - f1_S);
vx3 = ((f1_TSE - f1_BNW) + (f1_TNW - f1_BSE)) + ((f1_TNE - f1_BSW) + (f1_TSW - f1_BNE)) +
(-(f1_BN - f1_TS) + (f1_TN - f1_BS)) + ((f1_TE - f1_BW) - (f1_BE - f1_TW)) +
- (f1_T - f1_B);
+ (f1_T - f1_B);
real cu_sq=c3o2*(vx1*vx1+vx2*vx2+vx3*vx3);
//////////////////////////////////////////////////////////////////////////
real drho1 = f1_ZERO+f1_E+f1_W+f1_N+f1_S+f1_T+f1_B+f1_NE+f1_SW+f1_SE+f1_NW+f1_TE+f1_BW+f1_BE+f1_TW+f1_TN+f1_BS+f1_BN+f1_TS+
f1_TNE+f1_TSW+f1_TSE+f1_TNW+f1_BNE+f1_BSW+f1_BSE+f1_BNW;
- //drho1 = (drho1 + rhoBC[k])/2.f;
- drho1 = drho1 - rhoBC[k];
+ //drho1 = (drho1 + rhoBC[k])/2.f;
+ drho1 = drho1 - rhoBC[k];
__syncthreads();
- (D.f[DIR_P00 ])[ke ] = c2o27* (rhoBC[k]+c3o1*(-vx1 )+c9o2*(-vx1 )*(-vx1 )-cu_sq);
- (D.f[DIR_M00 ])[kw ] = c2o27* (rhoBC[k]+c3o1*( vx1 )+c9o2*( vx1 )*( vx1 )-cu_sq);
- (D.f[DIR_0P0 ])[kn ] = c2o27* (rhoBC[k]+c3o1*( -vx2 )+c9o2*( -vx2 )*( -vx2 )-cu_sq);
- (D.f[DIR_0M0 ])[ks ] = c2o27* (rhoBC[k]+c3o1*( vx2 )+c9o2*( vx2 )*( vx2 )-cu_sq);
- (D.f[DIR_00P ])[kt ] = c2o27* (rhoBC[k]+c3o1*( -vx3)+c9o2*( -vx3)*( -vx3)-cu_sq);
- (D.f[DIR_00M ])[kb ] = c2o27* (rhoBC[k]+c3o1*( vx3)+c9o2*( vx3)*( vx3)-cu_sq);
- (D.f[DIR_PP0 ])[kne ] = f1_SW -c1o54*drho1; // c1o100; // zero; //
- (D.f[DIR_MM0 ])[ksw ] = f1_NE -c1o54*drho1; // c1o100; // zero; //
- (D.f[DIR_PM0 ])[kse ] = f1_NW -c1o54*drho1; // c1o100; // zero; //
- (D.f[DIR_MP0 ])[knw ] = f1_SE -c1o54*drho1; // c1o100; // zero; //
- (D.f[DIR_P0P ])[kte ] = f1_BW -c1o54*drho1; // c1o100; // zero; //
- (D.f[DIR_M0M ])[kbw ] = f1_TE -c1o54*drho1; // c1o100; // zero; //
- (D.f[DIR_P0M ])[kbe ] = f1_TW -c1o54*drho1; // c1o100; // zero; //
- (D.f[DIR_M0P ])[ktw ] = f1_BE -c1o54*drho1; // c1o100; // zero; //
- (D.f[DIR_0PP ])[ktn ] = f1_BS -c1o54*drho1; // c1o100; // zero; //
- (D.f[DIR_0MM ])[kbs ] = f1_TN -c1o54*drho1; // c1o100; // zero; //
- (D.f[DIR_0PM ])[kbn ] = f1_TS -c1o54*drho1; // c1o100; // zero; //
- (D.f[DIR_0MP ])[kts ] = f1_BN -c1o54*drho1; // c1o100; // zero; //
+ (D.f[DIR_P00])[ke ] = c2o27* (rhoBC[k]+c3o1*(-vx1 )+c9o2*(-vx1 )*(-vx1 )-cu_sq);
+ (D.f[DIR_M00])[kw ] = c2o27* (rhoBC[k]+c3o1*( vx1 )+c9o2*( vx1 )*( vx1 )-cu_sq);
+ (D.f[DIR_0P0])[kn ] = c2o27* (rhoBC[k]+c3o1*( -vx2 )+c9o2*( -vx2 )*( -vx2 )-cu_sq);
+ (D.f[DIR_0M0])[ks ] = c2o27* (rhoBC[k]+c3o1*( vx2 )+c9o2*( vx2 )*( vx2 )-cu_sq);
+ (D.f[DIR_00P])[kt ] = c2o27* (rhoBC[k]+c3o1*( -vx3)+c9o2*( -vx3)*( -vx3)-cu_sq);
+ (D.f[DIR_00M])[kb ] = c2o27* (rhoBC[k]+c3o1*( vx3)+c9o2*( vx3)*( vx3)-cu_sq);
+ (D.f[DIR_PP0])[kne ] = f1_SW -c1o54*drho1; // c1o100; // zero; //
+ (D.f[DIR_MM0])[ksw ] = f1_NE -c1o54*drho1; // c1o100; // zero; //
+ (D.f[DIR_PM0])[kse ] = f1_NW -c1o54*drho1; // c1o100; // zero; //
+ (D.f[DIR_MP0])[knw ] = f1_SE -c1o54*drho1; // c1o100; // zero; //
+ (D.f[DIR_P0P])[kte ] = f1_BW -c1o54*drho1; // c1o100; // zero; //
+ (D.f[DIR_M0M])[kbw ] = f1_TE -c1o54*drho1; // c1o100; // zero; //
+ (D.f[DIR_P0M])[kbe ] = f1_TW -c1o54*drho1; // c1o100; // zero; //
+ (D.f[DIR_M0P])[ktw ] = f1_BE -c1o54*drho1; // c1o100; // zero; //
+ (D.f[DIR_0PP])[ktn ] = f1_BS -c1o54*drho1; // c1o100; // zero; //
+ (D.f[DIR_0MM])[kbs ] = f1_TN -c1o54*drho1; // c1o100; // zero; //
+ (D.f[DIR_0PM])[kbn ] = f1_TS -c1o54*drho1; // c1o100; // zero; //
+ (D.f[DIR_0MP])[kts ] = f1_BN -c1o54*drho1; // c1o100; // zero; //
(D.f[DIR_000])[kzero] = f1_ZERO-c8o27*drho1; // c1o100; // zero; //
- (D.f[DIR_PPP ])[ktne ] = f1_BSW -c1o216*drho1; // c1o100; // zero; //
- (D.f[DIR_MMP ])[ktsw ] = f1_BNE -c1o216*drho1; // c1o100; // zero; //
- (D.f[DIR_PMP ])[ktse ] = f1_BNW -c1o216*drho1; // c1o100; // zero; //
- (D.f[DIR_MPP ])[ktnw ] = f1_BSE -c1o216*drho1; // c1o100; // zero; //
- (D.f[DIR_PPM ])[kbne ] = f1_TSW -c1o216*drho1; // c1o100; // zero; //
- (D.f[DIR_MMM ])[kbsw ] = f1_TNE -c1o216*drho1; // c1o100; // zero; //
- (D.f[DIR_PMM ])[kbse ] = f1_TNW -c1o216*drho1; // c1o100; // zero; //
- (D.f[DIR_MPM ])[kbnw ] = f1_TSE -c1o216*drho1; // c1o100; // zero; //
+ (D.f[DIR_PPP])[ktne ] = f1_BSW -c1o216*drho1; // c1o100; // zero; //
+ (D.f[DIR_MMP])[ktsw ] = f1_BNE -c1o216*drho1; // c1o100; // zero; //
+ (D.f[DIR_PMP])[ktse ] = f1_BNW -c1o216*drho1; // c1o100; // zero; //
+ (D.f[DIR_MPP])[ktnw ] = f1_BSE -c1o216*drho1; // c1o100; // zero; //
+ (D.f[DIR_PPM])[kbne ] = f1_TSW -c1o216*drho1; // c1o100; // zero; //
+ (D.f[DIR_MMM])[kbsw ] = f1_TNE -c1o216*drho1; // c1o100; // zero; //
+ (D.f[DIR_PMM])[kbse ] = f1_TNW -c1o216*drho1; // c1o100; // zero; //
+ (D.f[DIR_MPM])[kbnw ] = f1_TSE -c1o216*drho1; // c1o100; // zero; //
}
}
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
@@ -4815,461 +4832,462 @@ __global__ void QPressDeviceFake27( real* rhoBC,
//////////////////////////////////////////////////////////////////////////
-__global__ void QPressDevice27_IntBB(real* rho,
- real* DD,
- int* k_Q,
- real* QQ,
- unsigned int numberOfBCnodes,
- real om1,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep)
+__global__ void QPressDevice27_IntBB(
+ real* rho,
+ real* DD,
+ int* k_Q,
+ real* QQ,
+ unsigned int numberOfBCnodes,
+ real om1,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep)
{
- Distributions27 D;
- if (isEvenTimestep==true)
- {
- D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
- }
- else
- {
- D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
- }
- ////////////////////////////////////////////////////////////////////////////////
- const unsigned x = threadIdx.x; // Globaler x-Index
- const unsigned y = blockIdx.x; // Globaler y-Index
- const unsigned z = blockIdx.y; // Globaler z-Index
-
- const unsigned nx = blockDim.x;
- const unsigned ny = gridDim.x;
-
- const unsigned k = nx*(ny*z + y) + x;
- //////////////////////////////////////////////////////////////////////////
-
- if(k < numberOfBCnodes)
- {
- ////////////////////////////////////////////////////////////////////////////////
- //real VeloX = vx[k];
- //real VeloY = vy[k];
- //real VeloZ = vz[k]; //(16.0*(u0*2.0)*bbx*bby*(grid_nx-bbx)*(grid_ny-bby))/(grid_nx*grid_nx*grid_ny*grid_ny)
- ////////////////////////////////////////////////////////////////////////////////
- real *q_dirE, *q_dirW, *q_dirN, *q_dirS, *q_dirT, *q_dirB,
- *q_dirNE, *q_dirSW, *q_dirSE, *q_dirNW, *q_dirTE, *q_dirBW,
- *q_dirBE, *q_dirTW, *q_dirTN, *q_dirBS, *q_dirBN, *q_dirTS,
- *q_dirTNE, *q_dirTSW, *q_dirTSE, *q_dirTNW, *q_dirBNE, *q_dirBSW,
- *q_dirBSE, *q_dirBNW;
- q_dirE = &QQ[DIR_P00 * numberOfBCnodes];
- q_dirW = &QQ[DIR_M00 * numberOfBCnodes];
- q_dirN = &QQ[DIR_0P0 * numberOfBCnodes];
- q_dirS = &QQ[DIR_0M0 * numberOfBCnodes];
- q_dirT = &QQ[DIR_00P * numberOfBCnodes];
- q_dirB = &QQ[DIR_00M * numberOfBCnodes];
- q_dirNE = &QQ[DIR_PP0 * numberOfBCnodes];
- q_dirSW = &QQ[DIR_MM0 * numberOfBCnodes];
- q_dirSE = &QQ[DIR_PM0 * numberOfBCnodes];
- q_dirNW = &QQ[DIR_MP0 * numberOfBCnodes];
- q_dirTE = &QQ[DIR_P0P * numberOfBCnodes];
- q_dirBW = &QQ[DIR_M0M * numberOfBCnodes];
- q_dirBE = &QQ[DIR_P0M * numberOfBCnodes];
- q_dirTW = &QQ[DIR_M0P * numberOfBCnodes];
- q_dirTN = &QQ[DIR_0PP * numberOfBCnodes];
- q_dirBS = &QQ[DIR_0MM * numberOfBCnodes];
- q_dirBN = &QQ[DIR_0PM * numberOfBCnodes];
- q_dirTS = &QQ[DIR_0MP * numberOfBCnodes];
- q_dirTNE = &QQ[DIR_PPP * numberOfBCnodes];
- q_dirTSW = &QQ[DIR_MMP * numberOfBCnodes];
- q_dirTSE = &QQ[DIR_PMP * numberOfBCnodes];
- q_dirTNW = &QQ[DIR_MPP * numberOfBCnodes];
- q_dirBNE = &QQ[DIR_PPM * numberOfBCnodes];
- q_dirBSW = &QQ[DIR_MMM * numberOfBCnodes];
- q_dirBSE = &QQ[DIR_PMM * numberOfBCnodes];
- q_dirBNW = &QQ[DIR_MPM * numberOfBCnodes];
- ////////////////////////////////////////////////////////////////////////////////
- //index
- unsigned int KQK = k_Q[k];
- unsigned int kzero= KQK;
- unsigned int ke = KQK;
- unsigned int kw = neighborX[KQK];
- unsigned int kn = KQK;
- unsigned int ks = neighborY[KQK];
- unsigned int kt = KQK;
- unsigned int kb = neighborZ[KQK];
- unsigned int ksw = neighborY[kw];
- unsigned int kne = KQK;
- unsigned int kse = ks;
- unsigned int knw = kw;
- unsigned int kbw = neighborZ[kw];
- unsigned int kte = KQK;
- unsigned int kbe = kb;
- unsigned int ktw = kw;
- unsigned int kbs = neighborZ[ks];
- unsigned int ktn = KQK;
- unsigned int kbn = kb;
- unsigned int kts = ks;
- unsigned int ktse = ks;
- unsigned int kbnw = kbw;
- unsigned int ktnw = kw;
- unsigned int kbse = kbs;
- unsigned int ktsw = ksw;
- unsigned int kbne = kb;
- unsigned int ktne = KQK;
- unsigned int kbsw = neighborZ[ksw];
- ////////////////////////////////////////////////////////////////////////////////
- real f_E, f_W, f_N, f_S, f_T, f_B, f_NE, f_SW, f_SE, f_NW, f_TE, f_BW, f_BE,
- f_TW, f_TN, f_BS, f_BN, f_TS, f_TNE, f_TSW, f_TSE, f_TNW, f_BNE, f_BSW, f_BSE, f_BNW;
-
- f_W = (D.f[DIR_P00 ])[ke ];
- f_E = (D.f[DIR_M00 ])[kw ];
- f_S = (D.f[DIR_0P0 ])[kn ];
- f_N = (D.f[DIR_0M0 ])[ks ];
- f_B = (D.f[DIR_00P ])[kt ];
- f_T = (D.f[DIR_00M ])[kb ];
- f_SW = (D.f[DIR_PP0 ])[kne ];
- f_NE = (D.f[DIR_MM0 ])[ksw ];
- f_NW = (D.f[DIR_PM0 ])[kse ];
- f_SE = (D.f[DIR_MP0 ])[knw ];
- f_BW = (D.f[DIR_P0P ])[kte ];
- f_TE = (D.f[DIR_M0M ])[kbw ];
- f_TW = (D.f[DIR_P0M ])[kbe ];
- f_BE = (D.f[DIR_M0P ])[ktw ];
- f_BS = (D.f[DIR_0PP ])[ktn ];
- f_TN = (D.f[DIR_0MM ])[kbs ];
- f_TS = (D.f[DIR_0PM ])[kbn ];
- f_BN = (D.f[DIR_0MP ])[kts ];
- f_BSW = (D.f[DIR_PPP ])[ktne ];
- f_BNE = (D.f[DIR_MMP ])[ktsw ];
- f_BNW = (D.f[DIR_PMP ])[ktse ];
- f_BSE = (D.f[DIR_MPP ])[ktnw ];
- f_TSW = (D.f[DIR_PPM ])[kbne ];
- f_TNE = (D.f[DIR_MMM ])[kbsw ];
- f_TNW = (D.f[DIR_PMM ])[kbse ];
- f_TSE = (D.f[DIR_MPM ])[kbnw ];
- ////////////////////////////////////////////////////////////////////////////////
- real vx1, vx2, vx3, drho, feq, q;
- drho = f_TSE + f_TNW + f_TNE + f_TSW + f_BSE + f_BNW + f_BNE + f_BSW +
- f_BN + f_TS + f_TN + f_BS + f_BE + f_TW + f_TE + f_BW + f_SE + f_NW + f_NE + f_SW +
- f_T + f_B + f_N + f_S + f_E + f_W + ((D.f[DIR_000])[kzero]);
-
- vx1 = (((f_TSE - f_BNW) - (f_TNW - f_BSE)) + ((f_TNE - f_BSW) - (f_TSW - f_BNE)) +
- ((f_BE - f_TW) + (f_TE - f_BW)) + ((f_SE - f_NW) + (f_NE - f_SW)) +
- (f_E - f_W))/(c1o1+drho);
-
-
- vx2 = ((-(f_TSE - f_BNW) + (f_TNW - f_BSE)) + ((f_TNE - f_BSW) - (f_TSW - f_BNE)) +
- ((f_BN - f_TS) + (f_TN - f_BS)) + (-(f_SE - f_NW) + (f_NE - f_SW)) +
- (f_N - f_S))/(c1o1+drho);
-
- vx3 = (((f_TSE - f_BNW) + (f_TNW - f_BSE)) + ((f_TNE - f_BSW) + (f_TSW - f_BNE)) +
- (-(f_BN - f_TS) + (f_TN - f_BS)) + ((f_TE - f_BW) - (f_BE - f_TW)) +
- (f_T - f_B))/(c1o1+drho);
-
- real cu_sq=c3o2*(vx1*vx1+vx2*vx2+vx3*vx3);
-
- //////////////////////////////////////////////////////////////////////////
- if (isEvenTimestep==false)
- {
- D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
- }
- else
- {
- D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
- }
- ////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
- //Test
- //(D.f[DIR_000])[k]=c1o10;
- real rhoDiff = drho - rho[k];
- real VeloX = vx1;
- real VeloY = vx2;
- real VeloZ = vx3;
- ////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
-
- q = q_dirE[k];
- if (q>=c0o1 && q<=c1o1)
- {
- feq=c2o27* (drho+c9o2*( vx1 )*( vx1 )-cu_sq);
- (D.f[DIR_M00])[kw]=(c1o1-q)/(c1o1+q)*(f_E-f_W+(f_E+f_W-c2o1*feq*om1)/(c1o1-om1))*c1o2+(q*(f_E+f_W)-c2o27*(rhoDiff + c6o1*( VeloX )))/(c1o1+q);
- }
-
- q = q_dirW[k];
- if (q>=c0o1 && q<=c1o1)
- {
- feq=c2o27* (drho+c9o2*(-vx1 )*(-vx1 )-cu_sq);
- (D.f[DIR_P00])[ke]=(c1o1-q)/(c1o1+q)*(f_W-f_E+(f_W+f_E-c2o1*feq*om1)/(c1o1-om1))*c1o2+(q*(f_W+f_E)-c2o27*(rhoDiff + c6o1*(-VeloX )))/(c1o1+q);
- }
-
- q = q_dirN[k];
- if (q>=c0o1 && q<=c1o1)
- {
- feq=c2o27* (drho+c9o2*( vx2 )*( vx2 )-cu_sq);
- (D.f[DIR_0M0])[ks]=(c1o1-q)/(c1o1+q)*(f_N-f_S+(f_N+f_S-c2o1*feq*om1)/(c1o1-om1))*c1o2+(q*(f_N+f_S)-c2o27*(rhoDiff + c6o1*( VeloY )))/(c1o1+q);
- }
-
- q = q_dirS[k];
- if (q>=c0o1 && q<=c1o1)
- {
- feq=c2o27* (drho+c9o2*( -vx2 )*( -vx2 )-cu_sq);
- (D.f[DIR_0P0])[kn]=(c1o1-q)/(c1o1+q)*(f_S-f_N+(f_S+f_N-c2o1*feq*om1)/(c1o1-om1))*c1o2+(q*(f_S+f_N)-c2o27*(rhoDiff + c6o1*(-VeloY )))/(c1o1+q);
- }
-
- q = q_dirT[k];
- if (q>=c0o1 && q<=c1o1)
- {
- feq=c2o27* (drho+c9o2*( vx3)*( vx3)-cu_sq);
- (D.f[DIR_00M])[kb]=(c1o1-q)/(c1o1+q)*(f_T-f_B+(f_T+f_B-c2o1*feq*om1)/(c1o1-om1))*c1o2+(q*(f_T+f_B)-c2o27*(rhoDiff + c6o1*( VeloZ )))/(c1o1+q);
- }
-
- q = q_dirB[k];
- if (q>=c0o1 && q<=c1o1)
- {
- feq=c2o27* (drho+c9o2*( -vx3)*( -vx3)-cu_sq);
- (D.f[DIR_00P])[kt]=(c1o1-q)/(c1o1+q)*(f_B-f_T+(f_B+f_T-c2o1*feq*om1)/(c1o1-om1))*c1o2+(q*(f_B+f_T)-c2o27*(rhoDiff + c6o1*(-VeloZ )))/(c1o1+q);
- }
-
- q = q_dirNE[k];
- if (q>=c0o1 && q<=c1o1)
- {
- feq=c1o54* (drho+c9o2*( vx1+vx2 )*( vx1+vx2 )-cu_sq);
- (D.f[DIR_MM0])[ksw]=(c1o1-q)/(c1o1+q)*(f_NE-f_SW+(f_NE+f_SW-c2o1*feq*om1)/(c1o1-om1))*c1o2+(q*(f_NE+f_SW)-c1o54*(rhoDiff + c6o1*(VeloX+VeloY)))/(c1o1+q);
- }
-
- q = q_dirSW[k];
- if (q>=c0o1 && q<=c1o1)
- {
- feq=c1o54* (drho+c9o2*(-vx1-vx2 )*(-vx1-vx2 )-cu_sq);
- (D.f[DIR_PP0])[kne]=(c1o1-q)/(c1o1+q)*(f_SW-f_NE+(f_SW+f_NE-c2o1*feq*om1)/(c1o1-om1))*c1o2+(q*(f_SW+f_NE)-c1o54*(rhoDiff + c6o1*(-VeloX-VeloY)))/(c1o1+q);
- }
-
- q = q_dirSE[k];
- if (q>=c0o1 && q<=c1o1)
- {
- feq=c1o54* (drho+c9o2*( vx1-vx2 )*( vx1-vx2 )-cu_sq);
- (D.f[DIR_MP0])[knw]=(c1o1-q)/(c1o1+q)*(f_SE-f_NW+(f_SE+f_NW-c2o1*feq*om1)/(c1o1-om1))*c1o2+(q*(f_SE+f_NW)-c1o54*(rhoDiff + c6o1*( VeloX-VeloY)))/(c1o1+q);
- }
-
- q = q_dirNW[k];
- if (q>=c0o1 && q<=c1o1)
- {
- feq=c1o54* (drho+c9o2*(-vx1+vx2 )*(-vx1+vx2 )-cu_sq);
- (D.f[DIR_PM0])[kse]=(c1o1-q)/(c1o1+q)*(f_NW-f_SE+(f_NW+f_SE-c2o1*feq*om1)/(c1o1-om1))*c1o2+(q*(f_NW+f_SE)-c1o54*(rhoDiff + c6o1*(-VeloX+VeloY)))/(c1o1+q);
- }
-
- q = q_dirTE[k];
- if (q>=c0o1 && q<=c1o1)
- {
- feq=c1o54* (drho+c9o2*( vx1 +vx3)*( vx1 +vx3)-cu_sq);
- (D.f[DIR_M0M])[kbw]=(c1o1-q)/(c1o1+q)*(f_TE-f_BW+(f_TE+f_BW-c2o1*feq*om1)/(c1o1-om1))*c1o2+(q*(f_TE+f_BW)-c1o54*(rhoDiff + c6o1*( VeloX+VeloZ)))/(c1o1+q);
- }
-
- q = q_dirBW[k];
- if (q>=c0o1 && q<=c1o1)
- {
- feq=c1o54* (drho+c9o2*(-vx1 -vx3)*(-vx1 -vx3)-cu_sq);
- (D.f[DIR_P0P])[kte]=(c1o1-q)/(c1o1+q)*(f_BW-f_TE+(f_BW+f_TE-c2o1*feq*om1)/(c1o1-om1))*c1o2+(q*(f_BW+f_TE)-c1o54*(rhoDiff + c6o1*(-VeloX-VeloZ)))/(c1o1+q);
- }
-
- q = q_dirBE[k];
- if (q>=c0o1 && q<=c1o1)
- {
- feq=c1o54* (drho+c9o2*( vx1 -vx3)*( vx1 -vx3)-cu_sq);
- (D.f[DIR_M0P])[ktw]=(c1o1-q)/(c1o1+q)*(f_BE-f_TW+(f_BE+f_TW-c2o1*feq*om1)/(c1o1-om1))*c1o2+(q*(f_BE+f_TW)-c1o54*(rhoDiff + c6o1*( VeloX-VeloZ)))/(c1o1+q);
- }
-
- q = q_dirTW[k];
- if (q>=c0o1 && q<=c1o1)
- {
- feq=c1o54* (drho+c9o2*(-vx1 +vx3)*(-vx1 +vx3)-cu_sq);
- (D.f[DIR_P0M])[kbe]=(c1o1-q)/(c1o1+q)*(f_TW-f_BE+(f_TW+f_BE-c2o1*feq*om1)/(c1o1-om1))*c1o2+(q*(f_TW+f_BE)-c1o54*(rhoDiff + c6o1*(-VeloX+VeloZ)))/(c1o1+q);
- }
-
- q = q_dirTN[k];
- if (q>=c0o1 && q<=c1o1)
- {
- feq=c1o54* (drho+c9o2*( vx2+vx3)*( vx2+vx3)-cu_sq);
- (D.f[DIR_0MM])[kbs]=(c1o1-q)/(c1o1+q)*(f_TN-f_BS+(f_TN+f_BS-c2o1*feq*om1)/(c1o1-om1))*c1o2+(q*(f_TN+f_BS)-c1o54*(rhoDiff + c6o1*( VeloY+VeloZ)))/(c1o1+q);
- }
-
- q = q_dirBS[k];
- if (q>=c0o1 && q<=c1o1)
- {
- feq=c1o54* (drho+c9o2*( -vx2-vx3)*( -vx2-vx3)-cu_sq);
- (D.f[DIR_0PP])[ktn]=(c1o1-q)/(c1o1+q)*(f_BS-f_TN+(f_BS+f_TN-c2o1*feq*om1)/(c1o1-om1))*c1o2+(q*(f_BS+f_TN)-c1o54*(rhoDiff + c6o1*( -VeloY-VeloZ)))/(c1o1+q);
- }
-
- q = q_dirBN[k];
- if (q>=c0o1 && q<=c1o1)
- {
- feq=c1o54* (drho+c9o2*( vx2-vx3)*( vx2-vx3)-cu_sq);
- (D.f[DIR_0MP])[kts]=(c1o1-q)/(c1o1+q)*(f_BN-f_TS+(f_BN+f_TS-c2o1*feq*om1)/(c1o1-om1))*c1o2+(q*(f_BN+f_TS)-c1o54*(rhoDiff + c6o1*( VeloY-VeloZ)))/(c1o1+q);
- }
-
- q = q_dirTS[k];
- if (q>=c0o1 && q<=c1o1)
- {
- feq=c1o54* (drho+c9o2*( -vx2+vx3)*( -vx2+vx3)-cu_sq);
- (D.f[DIR_0PM])[kbn]=(c1o1-q)/(c1o1+q)*(f_TS-f_BN+(f_TS+f_BN-c2o1*feq*om1)/(c1o1-om1))*c1o2+(q*(f_TS+f_BN)-c1o54*(rhoDiff + c6o1*( -VeloY+VeloZ)))/(c1o1+q);
- }
-
- q = q_dirTNE[k];
- if (q>=c0o1 && q<=c1o1)
- {
- feq=c1o216*(drho+c9o2*( vx1+vx2+vx3)*( vx1+vx2+vx3)-cu_sq);
- (D.f[DIR_MMM])[kbsw]=(c1o1-q)/(c1o1+q)*(f_TNE-f_BSW+(f_TNE+f_BSW-c2o1*feq*om1)/(c1o1-om1))*c1o2+(q*(f_TNE+f_BSW)-c1o216*(rhoDiff + c6o1*( VeloX+VeloY+VeloZ)))/(c1o1+q);
- }
-
- q = q_dirBSW[k];
- if (q>=c0o1 && q<=c1o1)
- {
- feq=c1o216*(drho+c9o2*(-vx1-vx2-vx3)*(-vx1-vx2-vx3)-cu_sq);
- (D.f[DIR_PPP])[ktne]=(c1o1-q)/(c1o1+q)*(f_BSW-f_TNE+(f_BSW+f_TNE-c2o1*feq*om1)/(c1o1-om1))*c1o2+(q*(f_BSW+f_TNE)-c1o216*(rhoDiff + c6o1*(-VeloX-VeloY-VeloZ)))/(c1o1+q);
- }
-
- q = q_dirBNE[k];
- if (q>=c0o1 && q<=c1o1)
- {
- feq=c1o216*(drho+c9o2*( vx1+vx2-vx3)*( vx1+vx2-vx3)-cu_sq);
- (D.f[DIR_MMP])[ktsw]=(c1o1-q)/(c1o1+q)*(f_BNE-f_TSW+(f_BNE+f_TSW-c2o1*feq*om1)/(c1o1-om1))*c1o2+(q*(f_BNE+f_TSW)-c1o216*(rhoDiff + c6o1*( VeloX+VeloY-VeloZ)))/(c1o1+q);
- }
-
- q = q_dirTSW[k];
- if (q>=c0o1 && q<=c1o1)
- {
- feq=c1o216*(drho+c9o2*(-vx1-vx2+vx3)*(-vx1-vx2+vx3)-cu_sq);
- (D.f[DIR_PPM])[kbne]=(c1o1-q)/(c1o1+q)*(f_TSW-f_BNE+(f_TSW+f_BNE-c2o1*feq*om1)/(c1o1-om1))*c1o2+(q*(f_TSW+f_BNE)-c1o216*(rhoDiff + c6o1*(-VeloX-VeloY+VeloZ)))/(c1o1+q);
- }
-
- q = q_dirTSE[k];
- if (q>=c0o1 && q<=c1o1)
- {
- feq=c1o216*(drho+c9o2*( vx1-vx2+vx3)*( vx1-vx2+vx3)-cu_sq);
- (D.f[DIR_MPM])[kbnw]=(c1o1-q)/(c1o1+q)*(f_TSE-f_BNW+(f_TSE+f_BNW-c2o1*feq*om1)/(c1o1-om1))*c1o2+(q*(f_TSE+f_BNW)-c1o216*(rhoDiff + c6o1*( VeloX-VeloY+VeloZ)))/(c1o1+q);
- }
-
- q = q_dirBNW[k];
- if (q>=c0o1 && q<=c1o1)
- {
- feq=c1o216*(drho+c9o2*(-vx1+vx2-vx3)*(-vx1+vx2-vx3)-cu_sq);
- (D.f[DIR_PMP])[ktse]=(c1o1-q)/(c1o1+q)*(f_BNW-f_TSE+(f_BNW+f_TSE-c2o1*feq*om1)/(c1o1-om1))*c1o2+(q*(f_BNW+f_TSE)-c1o216*(rhoDiff + c6o1*(-VeloX+VeloY-VeloZ)))/(c1o1+q);
- }
-
- q = q_dirBSE[k];
- if (q>=c0o1 && q<=c1o1)
- {
- feq=c1o216*(drho+c9o2*( vx1-vx2-vx3)*( vx1-vx2-vx3)-cu_sq);
- (D.f[DIR_MPP])[ktnw]=(c1o1-q)/(c1o1+q)*(f_BSE-f_TNW+(f_BSE+f_TNW-c2o1*feq*om1)/(c1o1-om1))*c1o2+(q*(f_BSE+f_TNW)-c1o216*(rhoDiff + c6o1*( VeloX-VeloY-VeloZ)))/(c1o1+q);
- }
-
- q = q_dirTNW[k];
- if (q>=c0o1 && q<=c1o1)
- {
- feq=c1o216*(drho+c9o2*(-vx1+vx2+vx3)*(-vx1+vx2+vx3)-cu_sq);
- (D.f[DIR_PMM])[kbse]=(c1o1-q)/(c1o1+q)*(f_TNW-f_BSE+(f_TNW+f_BSE-c2o1*feq*om1)/(c1o1-om1))*c1o2+(q*(f_TNW+f_BSE)-c1o216*(rhoDiff + c6o1*(-VeloX+VeloY+VeloZ)))/(c1o1+q);
- }
- }
+ Distributions27 D;
+ if (isEvenTimestep==true)
+ {
+ D.f[DIR_P00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_MPM * numberOfLBnodes];
+ }
+ else
+ {
+ D.f[DIR_M00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_MPM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_PMP * numberOfLBnodes];
+ }
+ ////////////////////////////////////////////////////////////////////////////////
+ const unsigned x = threadIdx.x; // Globaler x-Index
+ const unsigned y = blockIdx.x; // Globaler y-Index
+ const unsigned z = blockIdx.y; // Globaler z-Index
+
+ const unsigned nx = blockDim.x;
+ const unsigned ny = gridDim.x;
+
+ const unsigned k = nx*(ny*z + y) + x;
+ //////////////////////////////////////////////////////////////////////////
+
+ if(k < numberOfBCnodes)
+ {
+ ////////////////////////////////////////////////////////////////////////////////
+ //real VeloX = vx[k];
+ //real VeloY = vy[k];
+ //real VeloZ = vz[k]; //(16.0*(u0*2.0)*bbx*bby*(grid_nx-bbx)*(grid_ny-bby))/(grid_nx*grid_nx*grid_ny*grid_ny)
+ ////////////////////////////////////////////////////////////////////////////////
+ real *q_dirE, *q_dirW, *q_dirN, *q_dirS, *q_dirT, *q_dirB,
+ *q_dirNE, *q_dirSW, *q_dirSE, *q_dirNW, *q_dirTE, *q_dirBW,
+ *q_dirBE, *q_dirTW, *q_dirTN, *q_dirBS, *q_dirBN, *q_dirTS,
+ *q_dirTNE, *q_dirTSW, *q_dirTSE, *q_dirTNW, *q_dirBNE, *q_dirBSW,
+ *q_dirBSE, *q_dirBNW;
+ q_dirE = &QQ[DIR_P00 * numberOfBCnodes];
+ q_dirW = &QQ[DIR_M00 * numberOfBCnodes];
+ q_dirN = &QQ[DIR_0P0 * numberOfBCnodes];
+ q_dirS = &QQ[DIR_0M0 * numberOfBCnodes];
+ q_dirT = &QQ[DIR_00P * numberOfBCnodes];
+ q_dirB = &QQ[DIR_00M * numberOfBCnodes];
+ q_dirNE = &QQ[DIR_PP0 * numberOfBCnodes];
+ q_dirSW = &QQ[DIR_MM0 * numberOfBCnodes];
+ q_dirSE = &QQ[DIR_PM0 * numberOfBCnodes];
+ q_dirNW = &QQ[DIR_MP0 * numberOfBCnodes];
+ q_dirTE = &QQ[DIR_P0P * numberOfBCnodes];
+ q_dirBW = &QQ[DIR_M0M * numberOfBCnodes];
+ q_dirBE = &QQ[DIR_P0M * numberOfBCnodes];
+ q_dirTW = &QQ[DIR_M0P * numberOfBCnodes];
+ q_dirTN = &QQ[DIR_0PP * numberOfBCnodes];
+ q_dirBS = &QQ[DIR_0MM * numberOfBCnodes];
+ q_dirBN = &QQ[DIR_0PM * numberOfBCnodes];
+ q_dirTS = &QQ[DIR_0MP * numberOfBCnodes];
+ q_dirTNE = &QQ[DIR_PPP * numberOfBCnodes];
+ q_dirTSW = &QQ[DIR_MMP * numberOfBCnodes];
+ q_dirTSE = &QQ[DIR_PMP * numberOfBCnodes];
+ q_dirTNW = &QQ[DIR_MPP * numberOfBCnodes];
+ q_dirBNE = &QQ[DIR_PPM * numberOfBCnodes];
+ q_dirBSW = &QQ[DIR_MMM * numberOfBCnodes];
+ q_dirBSE = &QQ[DIR_PMM * numberOfBCnodes];
+ q_dirBNW = &QQ[DIR_MPM * numberOfBCnodes];
+ ////////////////////////////////////////////////////////////////////////////////
+ //index
+ unsigned int KQK = k_Q[k];
+ unsigned int kzero= KQK;
+ unsigned int ke = KQK;
+ unsigned int kw = neighborX[KQK];
+ unsigned int kn = KQK;
+ unsigned int ks = neighborY[KQK];
+ unsigned int kt = KQK;
+ unsigned int kb = neighborZ[KQK];
+ unsigned int ksw = neighborY[kw];
+ unsigned int kne = KQK;
+ unsigned int kse = ks;
+ unsigned int knw = kw;
+ unsigned int kbw = neighborZ[kw];
+ unsigned int kte = KQK;
+ unsigned int kbe = kb;
+ unsigned int ktw = kw;
+ unsigned int kbs = neighborZ[ks];
+ unsigned int ktn = KQK;
+ unsigned int kbn = kb;
+ unsigned int kts = ks;
+ unsigned int ktse = ks;
+ unsigned int kbnw = kbw;
+ unsigned int ktnw = kw;
+ unsigned int kbse = kbs;
+ unsigned int ktsw = ksw;
+ unsigned int kbne = kb;
+ unsigned int ktne = KQK;
+ unsigned int kbsw = neighborZ[ksw];
+ ////////////////////////////////////////////////////////////////////////////////
+ real f_E, f_W, f_N, f_S, f_T, f_B, f_NE, f_SW, f_SE, f_NW, f_TE, f_BW, f_BE,
+ f_TW, f_TN, f_BS, f_BN, f_TS, f_TNE, f_TSW, f_TSE, f_TNW, f_BNE, f_BSW, f_BSE, f_BNW;
+
+ f_W = (D.f[DIR_P00])[ke ];
+ f_E = (D.f[DIR_M00])[kw ];
+ f_S = (D.f[DIR_0P0])[kn ];
+ f_N = (D.f[DIR_0M0])[ks ];
+ f_B = (D.f[DIR_00P])[kt ];
+ f_T = (D.f[DIR_00M])[kb ];
+ f_SW = (D.f[DIR_PP0])[kne ];
+ f_NE = (D.f[DIR_MM0])[ksw ];
+ f_NW = (D.f[DIR_PM0])[kse ];
+ f_SE = (D.f[DIR_MP0])[knw ];
+ f_BW = (D.f[DIR_P0P])[kte ];
+ f_TE = (D.f[DIR_M0M])[kbw ];
+ f_TW = (D.f[DIR_P0M])[kbe ];
+ f_BE = (D.f[DIR_M0P])[ktw ];
+ f_BS = (D.f[DIR_0PP])[ktn ];
+ f_TN = (D.f[DIR_0MM])[kbs ];
+ f_TS = (D.f[DIR_0PM])[kbn ];
+ f_BN = (D.f[DIR_0MP])[kts ];
+ f_BSW = (D.f[DIR_PPP])[ktne ];
+ f_BNE = (D.f[DIR_MMP])[ktsw ];
+ f_BNW = (D.f[DIR_PMP])[ktse ];
+ f_BSE = (D.f[DIR_MPP])[ktnw ];
+ f_TSW = (D.f[DIR_PPM])[kbne ];
+ f_TNE = (D.f[DIR_MMM])[kbsw ];
+ f_TNW = (D.f[DIR_PMM])[kbse ];
+ f_TSE = (D.f[DIR_MPM])[kbnw ];
+ ////////////////////////////////////////////////////////////////////////////////
+ real vx1, vx2, vx3, drho, feq, q;
+ drho = f_TSE + f_TNW + f_TNE + f_TSW + f_BSE + f_BNW + f_BNE + f_BSW +
+ f_BN + f_TS + f_TN + f_BS + f_BE + f_TW + f_TE + f_BW + f_SE + f_NW + f_NE + f_SW +
+ f_T + f_B + f_N + f_S + f_E + f_W + ((D.f[DIR_000])[kzero]);
+
+ vx1 = (((f_TSE - f_BNW) - (f_TNW - f_BSE)) + ((f_TNE - f_BSW) - (f_TSW - f_BNE)) +
+ ((f_BE - f_TW) + (f_TE - f_BW)) + ((f_SE - f_NW) + (f_NE - f_SW)) +
+ (f_E - f_W))/(c1o1+drho);
+
+
+ vx2 = ((-(f_TSE - f_BNW) + (f_TNW - f_BSE)) + ((f_TNE - f_BSW) - (f_TSW - f_BNE)) +
+ ((f_BN - f_TS) + (f_TN - f_BS)) + (-(f_SE - f_NW) + (f_NE - f_SW)) +
+ (f_N - f_S))/(c1o1+drho);
+
+ vx3 = (((f_TSE - f_BNW) + (f_TNW - f_BSE)) + ((f_TNE - f_BSW) + (f_TSW - f_BNE)) +
+ (-(f_BN - f_TS) + (f_TN - f_BS)) + ((f_TE - f_BW) - (f_BE - f_TW)) +
+ (f_T - f_B))/(c1o1+drho);
+
+ real cu_sq=c3o2*(vx1*vx1+vx2*vx2+vx3*vx3);
+
+ //////////////////////////////////////////////////////////////////////////
+ if (isEvenTimestep==false)
+ {
+ D.f[DIR_P00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_MPM * numberOfLBnodes];
+ }
+ else
+ {
+ D.f[DIR_M00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_MPM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_PMP * numberOfLBnodes];
+ }
+ ////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
+ //Test
+ //(D.f[DIR_000])[k]=c1o10;
+ real rhoDiff = drho - rho[k];
+ real VeloX = vx1;
+ real VeloY = vx2;
+ real VeloZ = vx3;
+ ////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
+
+ q = q_dirE[k];
+ if (q>=c0o1 && q<=c1o1)
+ {
+ feq=c2o27* (drho+c9o2*( vx1 )*( vx1 )-cu_sq);
+ (D.f[DIR_M00])[kw]=(c1o1-q)/(c1o1+q)*(f_E-f_W+(f_E+f_W-c2o1*feq*om1)/(c1o1-om1))*c1o2+(q*(f_E+f_W)-c2o27*(rhoDiff + c6o1*( VeloX )))/(c1o1+q);
+ }
+
+ q = q_dirW[k];
+ if (q>=c0o1 && q<=c1o1)
+ {
+ feq=c2o27* (drho+c9o2*(-vx1 )*(-vx1 )-cu_sq);
+ (D.f[DIR_P00])[ke]=(c1o1-q)/(c1o1+q)*(f_W-f_E+(f_W+f_E-c2o1*feq*om1)/(c1o1-om1))*c1o2+(q*(f_W+f_E)-c2o27*(rhoDiff + c6o1*(-VeloX )))/(c1o1+q);
+ }
+
+ q = q_dirN[k];
+ if (q>=c0o1 && q<=c1o1)
+ {
+ feq=c2o27* (drho+c9o2*( vx2 )*( vx2 )-cu_sq);
+ (D.f[DIR_0M0])[ks]=(c1o1-q)/(c1o1+q)*(f_N-f_S+(f_N+f_S-c2o1*feq*om1)/(c1o1-om1))*c1o2+(q*(f_N+f_S)-c2o27*(rhoDiff + c6o1*( VeloY )))/(c1o1+q);
+ }
+
+ q = q_dirS[k];
+ if (q>=c0o1 && q<=c1o1)
+ {
+ feq=c2o27* (drho+c9o2*( -vx2 )*( -vx2 )-cu_sq);
+ (D.f[DIR_0P0])[kn]=(c1o1-q)/(c1o1+q)*(f_S-f_N+(f_S+f_N-c2o1*feq*om1)/(c1o1-om1))*c1o2+(q*(f_S+f_N)-c2o27*(rhoDiff + c6o1*(-VeloY )))/(c1o1+q);
+ }
+
+ q = q_dirT[k];
+ if (q>=c0o1 && q<=c1o1)
+ {
+ feq=c2o27* (drho+c9o2*( vx3)*( vx3)-cu_sq);
+ (D.f[DIR_00M])[kb]=(c1o1-q)/(c1o1+q)*(f_T-f_B+(f_T+f_B-c2o1*feq*om1)/(c1o1-om1))*c1o2+(q*(f_T+f_B)-c2o27*(rhoDiff + c6o1*( VeloZ )))/(c1o1+q);
+ }
+
+ q = q_dirB[k];
+ if (q>=c0o1 && q<=c1o1)
+ {
+ feq=c2o27* (drho+c9o2*( -vx3)*( -vx3)-cu_sq);
+ (D.f[DIR_00P])[kt]=(c1o1-q)/(c1o1+q)*(f_B-f_T+(f_B+f_T-c2o1*feq*om1)/(c1o1-om1))*c1o2+(q*(f_B+f_T)-c2o27*(rhoDiff + c6o1*(-VeloZ )))/(c1o1+q);
+ }
+
+ q = q_dirNE[k];
+ if (q>=c0o1 && q<=c1o1)
+ {
+ feq=c1o54* (drho+c9o2*( vx1+vx2 )*( vx1+vx2 )-cu_sq);
+ (D.f[DIR_MM0])[ksw]=(c1o1-q)/(c1o1+q)*(f_NE-f_SW+(f_NE+f_SW-c2o1*feq*om1)/(c1o1-om1))*c1o2+(q*(f_NE+f_SW)-c1o54*(rhoDiff + c6o1*(VeloX+VeloY)))/(c1o1+q);
+ }
+
+ q = q_dirSW[k];
+ if (q>=c0o1 && q<=c1o1)
+ {
+ feq=c1o54* (drho+c9o2*(-vx1-vx2 )*(-vx1-vx2 )-cu_sq);
+ (D.f[DIR_PP0])[kne]=(c1o1-q)/(c1o1+q)*(f_SW-f_NE+(f_SW+f_NE-c2o1*feq*om1)/(c1o1-om1))*c1o2+(q*(f_SW+f_NE)-c1o54*(rhoDiff + c6o1*(-VeloX-VeloY)))/(c1o1+q);
+ }
+
+ q = q_dirSE[k];
+ if (q>=c0o1 && q<=c1o1)
+ {
+ feq=c1o54* (drho+c9o2*( vx1-vx2 )*( vx1-vx2 )-cu_sq);
+ (D.f[DIR_MP0])[knw]=(c1o1-q)/(c1o1+q)*(f_SE-f_NW+(f_SE+f_NW-c2o1*feq*om1)/(c1o1-om1))*c1o2+(q*(f_SE+f_NW)-c1o54*(rhoDiff + c6o1*( VeloX-VeloY)))/(c1o1+q);
+ }
+
+ q = q_dirNW[k];
+ if (q>=c0o1 && q<=c1o1)
+ {
+ feq=c1o54* (drho+c9o2*(-vx1+vx2 )*(-vx1+vx2 )-cu_sq);
+ (D.f[DIR_PM0])[kse]=(c1o1-q)/(c1o1+q)*(f_NW-f_SE+(f_NW+f_SE-c2o1*feq*om1)/(c1o1-om1))*c1o2+(q*(f_NW+f_SE)-c1o54*(rhoDiff + c6o1*(-VeloX+VeloY)))/(c1o1+q);
+ }
+
+ q = q_dirTE[k];
+ if (q>=c0o1 && q<=c1o1)
+ {
+ feq=c1o54* (drho+c9o2*( vx1 +vx3)*( vx1 +vx3)-cu_sq);
+ (D.f[DIR_M0M])[kbw]=(c1o1-q)/(c1o1+q)*(f_TE-f_BW+(f_TE+f_BW-c2o1*feq*om1)/(c1o1-om1))*c1o2+(q*(f_TE+f_BW)-c1o54*(rhoDiff + c6o1*( VeloX+VeloZ)))/(c1o1+q);
+ }
+
+ q = q_dirBW[k];
+ if (q>=c0o1 && q<=c1o1)
+ {
+ feq=c1o54* (drho+c9o2*(-vx1 -vx3)*(-vx1 -vx3)-cu_sq);
+ (D.f[DIR_P0P])[kte]=(c1o1-q)/(c1o1+q)*(f_BW-f_TE+(f_BW+f_TE-c2o1*feq*om1)/(c1o1-om1))*c1o2+(q*(f_BW+f_TE)-c1o54*(rhoDiff + c6o1*(-VeloX-VeloZ)))/(c1o1+q);
+ }
+
+ q = q_dirBE[k];
+ if (q>=c0o1 && q<=c1o1)
+ {
+ feq=c1o54* (drho+c9o2*( vx1 -vx3)*( vx1 -vx3)-cu_sq);
+ (D.f[DIR_M0P])[ktw]=(c1o1-q)/(c1o1+q)*(f_BE-f_TW+(f_BE+f_TW-c2o1*feq*om1)/(c1o1-om1))*c1o2+(q*(f_BE+f_TW)-c1o54*(rhoDiff + c6o1*( VeloX-VeloZ)))/(c1o1+q);
+ }
+
+ q = q_dirTW[k];
+ if (q>=c0o1 && q<=c1o1)
+ {
+ feq=c1o54* (drho+c9o2*(-vx1 +vx3)*(-vx1 +vx3)-cu_sq);
+ (D.f[DIR_P0M])[kbe]=(c1o1-q)/(c1o1+q)*(f_TW-f_BE+(f_TW+f_BE-c2o1*feq*om1)/(c1o1-om1))*c1o2+(q*(f_TW+f_BE)-c1o54*(rhoDiff + c6o1*(-VeloX+VeloZ)))/(c1o1+q);
+ }
+
+ q = q_dirTN[k];
+ if (q>=c0o1 && q<=c1o1)
+ {
+ feq=c1o54* (drho+c9o2*( vx2+vx3)*( vx2+vx3)-cu_sq);
+ (D.f[DIR_0MM])[kbs]=(c1o1-q)/(c1o1+q)*(f_TN-f_BS+(f_TN+f_BS-c2o1*feq*om1)/(c1o1-om1))*c1o2+(q*(f_TN+f_BS)-c1o54*(rhoDiff + c6o1*( VeloY+VeloZ)))/(c1o1+q);
+ }
+
+ q = q_dirBS[k];
+ if (q>=c0o1 && q<=c1o1)
+ {
+ feq=c1o54* (drho+c9o2*( -vx2-vx3)*( -vx2-vx3)-cu_sq);
+ (D.f[DIR_0PP])[ktn]=(c1o1-q)/(c1o1+q)*(f_BS-f_TN+(f_BS+f_TN-c2o1*feq*om1)/(c1o1-om1))*c1o2+(q*(f_BS+f_TN)-c1o54*(rhoDiff + c6o1*( -VeloY-VeloZ)))/(c1o1+q);
+ }
+
+ q = q_dirBN[k];
+ if (q>=c0o1 && q<=c1o1)
+ {
+ feq=c1o54* (drho+c9o2*( vx2-vx3)*( vx2-vx3)-cu_sq);
+ (D.f[DIR_0MP])[kts]=(c1o1-q)/(c1o1+q)*(f_BN-f_TS+(f_BN+f_TS-c2o1*feq*om1)/(c1o1-om1))*c1o2+(q*(f_BN+f_TS)-c1o54*(rhoDiff + c6o1*( VeloY-VeloZ)))/(c1o1+q);
+ }
+
+ q = q_dirTS[k];
+ if (q>=c0o1 && q<=c1o1)
+ {
+ feq=c1o54* (drho+c9o2*( -vx2+vx3)*( -vx2+vx3)-cu_sq);
+ (D.f[DIR_0PM])[kbn]=(c1o1-q)/(c1o1+q)*(f_TS-f_BN+(f_TS+f_BN-c2o1*feq*om1)/(c1o1-om1))*c1o2+(q*(f_TS+f_BN)-c1o54*(rhoDiff + c6o1*( -VeloY+VeloZ)))/(c1o1+q);
+ }
+
+ q = q_dirTNE[k];
+ if (q>=c0o1 && q<=c1o1)
+ {
+ feq=c1o216*(drho+c9o2*( vx1+vx2+vx3)*( vx1+vx2+vx3)-cu_sq);
+ (D.f[DIR_MMM])[kbsw]=(c1o1-q)/(c1o1+q)*(f_TNE-f_BSW+(f_TNE+f_BSW-c2o1*feq*om1)/(c1o1-om1))*c1o2+(q*(f_TNE+f_BSW)-c1o216*(rhoDiff + c6o1*( VeloX+VeloY+VeloZ)))/(c1o1+q);
+ }
+
+ q = q_dirBSW[k];
+ if (q>=c0o1 && q<=c1o1)
+ {
+ feq=c1o216*(drho+c9o2*(-vx1-vx2-vx3)*(-vx1-vx2-vx3)-cu_sq);
+ (D.f[DIR_PPP])[ktne]=(c1o1-q)/(c1o1+q)*(f_BSW-f_TNE+(f_BSW+f_TNE-c2o1*feq*om1)/(c1o1-om1))*c1o2+(q*(f_BSW+f_TNE)-c1o216*(rhoDiff + c6o1*(-VeloX-VeloY-VeloZ)))/(c1o1+q);
+ }
+
+ q = q_dirBNE[k];
+ if (q>=c0o1 && q<=c1o1)
+ {
+ feq=c1o216*(drho+c9o2*( vx1+vx2-vx3)*( vx1+vx2-vx3)-cu_sq);
+ (D.f[DIR_MMP])[ktsw]=(c1o1-q)/(c1o1+q)*(f_BNE-f_TSW+(f_BNE+f_TSW-c2o1*feq*om1)/(c1o1-om1))*c1o2+(q*(f_BNE+f_TSW)-c1o216*(rhoDiff + c6o1*( VeloX+VeloY-VeloZ)))/(c1o1+q);
+ }
+
+ q = q_dirTSW[k];
+ if (q>=c0o1 && q<=c1o1)
+ {
+ feq=c1o216*(drho+c9o2*(-vx1-vx2+vx3)*(-vx1-vx2+vx3)-cu_sq);
+ (D.f[DIR_PPM])[kbne]=(c1o1-q)/(c1o1+q)*(f_TSW-f_BNE+(f_TSW+f_BNE-c2o1*feq*om1)/(c1o1-om1))*c1o2+(q*(f_TSW+f_BNE)-c1o216*(rhoDiff + c6o1*(-VeloX-VeloY+VeloZ)))/(c1o1+q);
+ }
+
+ q = q_dirTSE[k];
+ if (q>=c0o1 && q<=c1o1)
+ {
+ feq=c1o216*(drho+c9o2*( vx1-vx2+vx3)*( vx1-vx2+vx3)-cu_sq);
+ (D.f[DIR_MPM])[kbnw]=(c1o1-q)/(c1o1+q)*(f_TSE-f_BNW+(f_TSE+f_BNW-c2o1*feq*om1)/(c1o1-om1))*c1o2+(q*(f_TSE+f_BNW)-c1o216*(rhoDiff + c6o1*( VeloX-VeloY+VeloZ)))/(c1o1+q);
+ }
+
+ q = q_dirBNW[k];
+ if (q>=c0o1 && q<=c1o1)
+ {
+ feq=c1o216*(drho+c9o2*(-vx1+vx2-vx3)*(-vx1+vx2-vx3)-cu_sq);
+ (D.f[DIR_PMP])[ktse]=(c1o1-q)/(c1o1+q)*(f_BNW-f_TSE+(f_BNW+f_TSE-c2o1*feq*om1)/(c1o1-om1))*c1o2+(q*(f_BNW+f_TSE)-c1o216*(rhoDiff + c6o1*(-VeloX+VeloY-VeloZ)))/(c1o1+q);
+ }
+
+ q = q_dirBSE[k];
+ if (q>=c0o1 && q<=c1o1)
+ {
+ feq=c1o216*(drho+c9o2*( vx1-vx2-vx3)*( vx1-vx2-vx3)-cu_sq);
+ (D.f[DIR_MPP])[ktnw]=(c1o1-q)/(c1o1+q)*(f_BSE-f_TNW+(f_BSE+f_TNW-c2o1*feq*om1)/(c1o1-om1))*c1o2+(q*(f_BSE+f_TNW)-c1o216*(rhoDiff + c6o1*( VeloX-VeloY-VeloZ)))/(c1o1+q);
+ }
+
+ q = q_dirTNW[k];
+ if (q>=c0o1 && q<=c1o1)
+ {
+ feq=c1o216*(drho+c9o2*(-vx1+vx2+vx3)*(-vx1+vx2+vx3)-cu_sq);
+ (D.f[DIR_PMM])[kbse]=(c1o1-q)/(c1o1+q)*(f_TNW-f_BSE+(f_TNW+f_BSE-c2o1*feq*om1)/(c1o1-om1))*c1o2+(q*(f_TNW+f_BSE)-c1o216*(rhoDiff + c6o1*(-VeloX+VeloY+VeloZ)))/(c1o1+q);
+ }
+ }
}
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
diff --git a/src/gpu/VirtualFluids_GPU/GPU/SchlafferBCs27.cu b/src/gpu/VirtualFluids_GPU/GPU/SchlafferBCs27.cu
index 8675780d26e63656b04fdfc1f9836b1eba8d1b87..5d4572e234fdcad072e9b666c911f3250c32346a 100644
--- a/src/gpu/VirtualFluids_GPU/GPU/SchlafferBCs27.cu
+++ b/src/gpu/VirtualFluids_GPU/GPU/SchlafferBCs27.cu
@@ -21,7 +21,7 @@ __global__ void PressSchlaff27(real* rhoBC,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep)
{
////////////////////////////////////////////////////////////////////////////////
@@ -71,94 +71,94 @@ __global__ void PressSchlaff27(real* rhoBC,
Distributions27 D;
if (isEvenTimestep==true)
{
- D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
+ D.f[DIR_M00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_MPM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_PMP * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
real f1_E,f1_W,f1_N,f1_S,f1_T,f1_B,f1_NE,f1_SW,f1_SE,f1_NW,f1_TE,f1_BW,f1_BE,f1_TW,f1_TN,f1_BS,f1_BN,f1_TS,f1_ZERO,f1_TNE,f1_TSW,f1_TSE,f1_TNW,f1_BNE,f1_BSW,f1_BSE,f1_BNW;
- f1_E = (D.f[DIR_P00 ])[ke ];
- f1_W = (D.f[DIR_M00 ])[kw ];
- f1_N = (D.f[DIR_0P0 ])[kn ];
- f1_S = (D.f[DIR_0M0 ])[ks ];
- f1_T = (D.f[DIR_00P ])[kt ];
- f1_B = (D.f[DIR_00M ])[kb ];
- f1_NE = (D.f[DIR_PP0 ])[kne ];
- f1_SW = (D.f[DIR_MM0 ])[ksw ];
- f1_SE = (D.f[DIR_PM0 ])[kse ];
- f1_NW = (D.f[DIR_MP0 ])[knw ];
- f1_TE = (D.f[DIR_P0P ])[kte ];
- f1_BW = (D.f[DIR_M0M ])[kbw ];
- f1_BE = (D.f[DIR_P0M ])[kbe ];
- f1_TW = (D.f[DIR_M0P ])[ktw ];
- f1_TN = (D.f[DIR_0PP ])[ktn ];
- f1_BS = (D.f[DIR_0MM ])[kbs ];
- f1_BN = (D.f[DIR_0PM ])[kbn ];
- f1_TS = (D.f[DIR_0MP ])[kts ];
+ f1_E = (D.f[DIR_P00])[ke ];
+ f1_W = (D.f[DIR_M00])[kw ];
+ f1_N = (D.f[DIR_0P0])[kn ];
+ f1_S = (D.f[DIR_0M0])[ks ];
+ f1_T = (D.f[DIR_00P])[kt ];
+ f1_B = (D.f[DIR_00M])[kb ];
+ f1_NE = (D.f[DIR_PP0])[kne ];
+ f1_SW = (D.f[DIR_MM0])[ksw ];
+ f1_SE = (D.f[DIR_PM0])[kse ];
+ f1_NW = (D.f[DIR_MP0])[knw ];
+ f1_TE = (D.f[DIR_P0P])[kte ];
+ f1_BW = (D.f[DIR_M0M])[kbw ];
+ f1_BE = (D.f[DIR_P0M])[kbe ];
+ f1_TW = (D.f[DIR_M0P])[ktw ];
+ f1_TN = (D.f[DIR_0PP])[ktn ];
+ f1_BS = (D.f[DIR_0MM])[kbs ];
+ f1_BN = (D.f[DIR_0PM])[kbn ];
+ f1_TS = (D.f[DIR_0MP])[kts ];
f1_ZERO = (D.f[DIR_000])[kzero];
- f1_TNE = (D.f[DIR_PPP ])[ktne ];
- f1_TSW = (D.f[DIR_MMP ])[ktsw ];
- f1_TSE = (D.f[DIR_PMP ])[ktse ];
- f1_TNW = (D.f[DIR_MPP ])[ktnw ];
- f1_BNE = (D.f[DIR_PPM ])[kbne ];
- f1_BSW = (D.f[DIR_MMM ])[kbsw ];
- f1_BSE = (D.f[DIR_PMM ])[kbse ];
- f1_BNW = (D.f[DIR_MPM ])[kbnw ];
+ f1_TNE = (D.f[DIR_PPP])[ktne ];
+ f1_TSW = (D.f[DIR_MMP])[ktsw ];
+ f1_TSE = (D.f[DIR_PMP])[ktse ];
+ f1_TNW = (D.f[DIR_MPP])[ktnw ];
+ f1_BNE = (D.f[DIR_PPM])[kbne ];
+ f1_BSW = (D.f[DIR_MMM])[kbsw ];
+ f1_BSE = (D.f[DIR_PMM])[kbse ];
+ f1_BNW = (D.f[DIR_MPM])[kbnw ];
//////////////////////////////////////////////////////////////////////////
real cs = c1o1/sqrt(c3o1);
real csp1 = cs + c1o1;
@@ -222,15 +222,15 @@ __global__ void PressSchlaff27(real* rhoBC,
deltaVz0[k] = tempDeltaV;
- (D.f[DIR_00M ])[kb ] = f1_B ;
- (D.f[DIR_M0M ])[kbw ] = f1_BW ;
- (D.f[DIR_P0M ])[kbe ] = f1_BE ;
- (D.f[DIR_0MM ])[kbs ] = f1_BS ;
- (D.f[DIR_0PM ])[kbn ] = f1_BN ;
- (D.f[DIR_PPM ])[kbne ] = f1_BNE ;
- (D.f[DIR_MMM ])[kbsw ] = f1_BSW ;
- (D.f[DIR_PMM ])[kbse ] = f1_BSE ;
- (D.f[DIR_MPM ])[kbnw ] = f1_BNW ;
+ (D.f[DIR_00M])[kb ] = f1_B ;
+ (D.f[DIR_M0M])[kbw ] = f1_BW ;
+ (D.f[DIR_P0M])[kbe ] = f1_BE ;
+ (D.f[DIR_0MM])[kbs ] = f1_BS ;
+ (D.f[DIR_0PM])[kbn ] = f1_BN ;
+ (D.f[DIR_PPM])[kbne ] = f1_BNE ;
+ (D.f[DIR_MMM])[kbsw ] = f1_BSW ;
+ (D.f[DIR_PMM])[kbse ] = f1_BSE ;
+ (D.f[DIR_MPM])[kbnw ] = f1_BNW ;
}
}
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
@@ -285,7 +285,7 @@ __global__ void VelSchlaff27( int t,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep)
{
////////////////////////////////////////////////////////////////////////////////
@@ -335,122 +335,122 @@ __global__ void VelSchlaff27( int t,
Distributions27 D;
if (isEvenTimestep==true)
{
- D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
+ D.f[DIR_M00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_MPM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_PMP * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
real f1_E,f1_W,f1_N,f1_S,f1_T,f1_B,f1_NE,f1_SW,f1_SE,f1_NW,f1_TE,f1_BW,f1_BE,f1_TW,f1_TN,f1_BS,f1_BN,f1_TS,f1_ZERO,
f1_TNE,f1_TSW,f1_TSE,f1_TNW,f1_BNE,f1_BSW,f1_BSE,f1_BNW;
- f1_E = (D.f[DIR_P00 ])[ke ];
- f1_W = (D.f[DIR_M00 ])[kw ];
- f1_N = (D.f[DIR_0P0 ])[kn ];
- f1_S = (D.f[DIR_0M0 ])[ks ];
- f1_T = (D.f[DIR_00P ])[kt ];
- f1_B = (D.f[DIR_00M ])[kb ];
- f1_NE = (D.f[DIR_PP0 ])[kne ];
- f1_SW = (D.f[DIR_MM0 ])[ksw ];
- f1_SE = (D.f[DIR_PM0 ])[kse ];
- f1_NW = (D.f[DIR_MP0 ])[knw ];
- f1_TE = (D.f[DIR_P0P ])[kte ];
- f1_BW = (D.f[DIR_M0M ])[kbw ];
- f1_BE = (D.f[DIR_P0M ])[kbe ];
- f1_TW = (D.f[DIR_M0P ])[ktw ];
- f1_TN = (D.f[DIR_0PP ])[ktn ];
- f1_BS = (D.f[DIR_0MM ])[kbs ];
- f1_BN = (D.f[DIR_0PM ])[kbn ];
- f1_TS = (D.f[DIR_0MP ])[kts ];
+ f1_E = (D.f[DIR_P00])[ke ];
+ f1_W = (D.f[DIR_M00])[kw ];
+ f1_N = (D.f[DIR_0P0])[kn ];
+ f1_S = (D.f[DIR_0M0])[ks ];
+ f1_T = (D.f[DIR_00P])[kt ];
+ f1_B = (D.f[DIR_00M])[kb ];
+ f1_NE = (D.f[DIR_PP0])[kne ];
+ f1_SW = (D.f[DIR_MM0])[ksw ];
+ f1_SE = (D.f[DIR_PM0])[kse ];
+ f1_NW = (D.f[DIR_MP0])[knw ];
+ f1_TE = (D.f[DIR_P0P])[kte ];
+ f1_BW = (D.f[DIR_M0M])[kbw ];
+ f1_BE = (D.f[DIR_P0M])[kbe ];
+ f1_TW = (D.f[DIR_M0P])[ktw ];
+ f1_TN = (D.f[DIR_0PP])[ktn ];
+ f1_BS = (D.f[DIR_0MM])[kbs ];
+ f1_BN = (D.f[DIR_0PM])[kbn ];
+ f1_TS = (D.f[DIR_0MP])[kts ];
f1_ZERO = (D.f[DIR_000])[kzero];
- f1_TNE = (D.f[DIR_PPP ])[ktne ];
- f1_TSW = (D.f[DIR_MMP ])[ktsw ];
- f1_TSE = (D.f[DIR_PMP ])[ktse ];
- f1_TNW = (D.f[DIR_MPP ])[ktnw ];
- f1_BNE = (D.f[DIR_PPM ])[kbne ];
- f1_BSW = (D.f[DIR_MMM ])[kbsw ];
- f1_BSE = (D.f[DIR_PMM ])[kbse ];
- f1_BNW = (D.f[DIR_MPM ])[kbnw ];
- //f1_W = (D.f[DIR_P00 ])[ke ];
- //f1_E = (D.f[DIR_M00 ])[kw ];
- //f1_S = (D.f[DIR_0P0 ])[kn ];
- //f1_N = (D.f[DIR_0M0 ])[ks ];
- //f1_B = (D.f[DIR_00P ])[kt ];
- //f1_T = (D.f[DIR_00M ])[kb ];
- //f1_SW = (D.f[DIR_PP0 ])[kne ];
- //f1_NE = (D.f[DIR_MM0 ])[ksw ];
- //f1_NW = (D.f[DIR_PM0 ])[kse ];
- //f1_SE = (D.f[DIR_MP0 ])[knw ];
- //f1_BW = (D.f[DIR_P0P ])[kte ];
- //f1_TE = (D.f[DIR_M0M ])[kbw ];
- //f1_TW = (D.f[DIR_P0M ])[kbe ];
- //f1_BE = (D.f[DIR_M0P ])[ktw ];
- //f1_BS = (D.f[DIR_0PP ])[ktn ];
- //f1_TN = (D.f[DIR_0MM ])[kbs ];
- //f1_TS = (D.f[DIR_0PM ])[kbn ];
- //f1_BN = (D.f[DIR_0MP ])[kts ];
+ f1_TNE = (D.f[DIR_PPP])[ktne ];
+ f1_TSW = (D.f[DIR_MMP])[ktsw ];
+ f1_TSE = (D.f[DIR_PMP])[ktse ];
+ f1_TNW = (D.f[DIR_MPP])[ktnw ];
+ f1_BNE = (D.f[DIR_PPM])[kbne ];
+ f1_BSW = (D.f[DIR_MMM])[kbsw ];
+ f1_BSE = (D.f[DIR_PMM])[kbse ];
+ f1_BNW = (D.f[DIR_MPM])[kbnw ];
+ //f1_W = (D.f[DIR_P00])[ke ];
+ //f1_E = (D.f[DIR_M00])[kw ];
+ //f1_S = (D.f[DIR_0P0])[kn ];
+ //f1_N = (D.f[DIR_0M0])[ks ];
+ //f1_B = (D.f[DIR_00P])[kt ];
+ //f1_T = (D.f[DIR_00M])[kb ];
+ //f1_SW = (D.f[DIR_PP0])[kne ];
+ //f1_NE = (D.f[DIR_MM0])[ksw ];
+ //f1_NW = (D.f[DIR_PM0])[kse ];
+ //f1_SE = (D.f[DIR_MP0])[knw ];
+ //f1_BW = (D.f[DIR_P0P])[kte ];
+ //f1_TE = (D.f[DIR_M0M])[kbw ];
+ //f1_TW = (D.f[DIR_P0M])[kbe ];
+ //f1_BE = (D.f[DIR_M0P])[ktw ];
+ //f1_BS = (D.f[DIR_0PP])[ktn ];
+ //f1_TN = (D.f[DIR_0MM])[kbs ];
+ //f1_TS = (D.f[DIR_0PM])[kbn ];
+ //f1_BN = (D.f[DIR_0MP])[kts ];
//f1_ZERO = (D.f[DIR_000])[kzero];
- //f1_BSW = (D.f[DIR_PPP ])[ktne ];
- //f1_BNE = (D.f[DIR_MMP ])[ktsw ];
- //f1_BNW = (D.f[DIR_PMP ])[ktse ];
- //f1_BSE = (D.f[DIR_MPP ])[ktnw ];
- //f1_TSW = (D.f[DIR_PPM ])[kbne ];
- //f1_TNE = (D.f[DIR_MMM ])[kbsw ];
- //f1_TNW = (D.f[DIR_PMM ])[kbse ];
- //f1_TSE = (D.f[DIR_MPM ])[kbnw ];
+ //f1_BSW = (D.f[DIR_PPP])[ktne ];
+ //f1_BNE = (D.f[DIR_MMP])[ktsw ];
+ //f1_BNW = (D.f[DIR_PMP])[ktse ];
+ //f1_BSE = (D.f[DIR_MPP])[ktnw ];
+ //f1_TSW = (D.f[DIR_PPM])[kbne ];
+ //f1_TNE = (D.f[DIR_MMM])[kbsw ];
+ //f1_TNW = (D.f[DIR_PMM])[kbse ];
+ //f1_TSE = (D.f[DIR_MPM])[kbnw ];
//////////////////////////////////////////////////////////////////////////
real cs = c1o1/sqrt(c3o1);
real csp1 = cs + c1o1;
@@ -522,64 +522,64 @@ __global__ void VelSchlaff27( int t,
f1_TNW = f1_BSE - c1o36 * (VX - VY - VZ);
deltaVz0[k] = tempDeltaV;
- (D.f[DIR_00P ])[kt ] = f1_T ;
- (D.f[DIR_P0P ])[kte ] = f1_TE ;
- (D.f[DIR_M0P ])[ktw ] = f1_TW ;
- (D.f[DIR_0PP ])[ktn ] = f1_TN ;
- (D.f[DIR_0MP ])[kts ] = f1_TS ;
- (D.f[DIR_PPP ])[ktne ] = f1_TNE;
- (D.f[DIR_MMP ])[ktsw ] = f1_TSW;
- (D.f[DIR_PMP ])[ktse ] = f1_TSE;
- (D.f[DIR_MPP ])[ktnw ] = f1_TNW;
-
- //(D.f[DIR_00M ])[kb ] = f1_B ;
- //(D.f[DIR_M0M ])[kbw ] = f1_BW ;
- //(D.f[DIR_P0M ])[kbe ] = f1_BE ;
- //(D.f[DIR_0MM ])[kbs ] = f1_BS ;
- //(D.f[DIR_0PM ])[kbn ] = f1_BN ;
- //(D.f[DIR_PPM ])[kbne ] = f1_BNE ;
- //(D.f[DIR_MMM ])[kbsw ] = f1_BSW ;
- //(D.f[DIR_PMM ])[kbse ] = f1_BSE ;
- //(D.f[DIR_MPM ])[kbnw ] = f1_BNW ;
-
-
- //(D.f[DIR_00P ])[kt ] = f1_B ;
- //(D.f[DIR_P0P ])[kte ] = f1_BW ;
- //(D.f[DIR_M0P ])[ktw ] = f1_BE ;
- //(D.f[DIR_0PP ])[ktn ] = f1_BS ;
- //(D.f[DIR_0MP ])[kts ] = f1_BN ;
- //(D.f[DIR_PPP ])[ktne ] = f1_BSW;
- //(D.f[DIR_MMP ])[ktsw ] = f1_BNE;
- //(D.f[DIR_PMP ])[ktse ] = f1_BNW;
- //(D.f[DIR_MPP ])[ktnw ] = f1_BSE;
-
- //(D.f[DIR_P00 ])[ke ] = f1_W -c2over27*drho1;
- //(D.f[DIR_M00 ])[kw ] = f1_E -c2over27*drho1;
- //(D.f[DIR_0P0 ])[kn ] = f1_S -c2over27*drho1;
- //(D.f[DIR_0M0 ])[ks ] = f1_N -c2over27*drho1;
- //(D.f[DIR_00P ])[kt ] = f1_B -c2over27*drho1;
- //(D.f[DIR_00M ])[kb ] = f1_T -c2over27*drho1;
- //(D.f[DIR_PP0 ])[kne ] = f1_SW -c1over54*drho1;
- //(D.f[DIR_MM0 ])[ksw ] = f1_NE -c1over54*drho1;
- //(D.f[DIR_PM0 ])[kse ] = f1_NW -c1over54*drho1;
- //(D.f[DIR_MP0 ])[knw ] = f1_SE -c1over54*drho1;
- //(D.f[DIR_P0P ])[kte ] = f1_BW -c1over54*drho1;
- //(D.f[DIR_M0M ])[kbw ] = f1_TE -c1over54*drho1;
- //(D.f[DIR_P0M ])[kbe ] = f1_TW -c1over54*drho1;
- //(D.f[DIR_M0P ])[ktw ] = f1_BE -c1over54*drho1;
- //(D.f[DIR_0PP ])[ktn ] = f1_BS -c1over54*drho1;
- //(D.f[DIR_0MM ])[kbs ] = f1_TN -c1over54*drho1;
- //(D.f[DIR_0PM ])[kbn ] = f1_TS -c1over54*drho1;
- //(D.f[DIR_0MP ])[kts ] = f1_BN -c1over54*drho1;
+ (D.f[DIR_00P])[kt ] = f1_T ;
+ (D.f[DIR_P0P])[kte ] = f1_TE ;
+ (D.f[DIR_M0P])[ktw ] = f1_TW ;
+ (D.f[DIR_0PP])[ktn ] = f1_TN ;
+ (D.f[DIR_0MP])[kts ] = f1_TS ;
+ (D.f[DIR_PPP])[ktne ] = f1_TNE;
+ (D.f[DIR_MMP])[ktsw ] = f1_TSW;
+ (D.f[DIR_PMP])[ktse ] = f1_TSE;
+ (D.f[DIR_MPP])[ktnw ] = f1_TNW;
+
+ //(D.f[DIR_00M])[kb ] = f1_B ;
+ //(D.f[DIR_M0M])[kbw ] = f1_BW ;
+ //(D.f[DIR_P0M])[kbe ] = f1_BE ;
+ //(D.f[DIR_0MM])[kbs ] = f1_BS ;
+ //(D.f[DIR_0PM])[kbn ] = f1_BN ;
+ //(D.f[DIR_PPM])[kbne ] = f1_BNE ;
+ //(D.f[DIR_MMM])[kbsw ] = f1_BSW ;
+ //(D.f[DIR_PMM])[kbse ] = f1_BSE ;
+ //(D.f[DIR_MPM])[kbnw ] = f1_BNW ;
+
+
+ //(D.f[DIR_00P])[kt ] = f1_B ;
+ //(D.f[DIR_P0P])[kte ] = f1_BW ;
+ //(D.f[DIR_M0P])[ktw ] = f1_BE ;
+ //(D.f[DIR_0PP])[ktn ] = f1_BS ;
+ //(D.f[DIR_0MP])[kts ] = f1_BN ;
+ //(D.f[DIR_PPP])[ktne ] = f1_BSW;
+ //(D.f[DIR_MMP])[ktsw ] = f1_BNE;
+ //(D.f[DIR_PMP])[ktse ] = f1_BNW;
+ //(D.f[DIR_MPP])[ktnw ] = f1_BSE;
+
+ //(D.f[DIR_P00])[ke ] = f1_W -c2over27*drho1;
+ //(D.f[DIR_M00])[kw ] = f1_E -c2over27*drho1;
+ //(D.f[DIR_0P0])[kn ] = f1_S -c2over27*drho1;
+ //(D.f[DIR_0M0])[ks ] = f1_N -c2over27*drho1;
+ //(D.f[DIR_00P])[kt ] = f1_B -c2over27*drho1;
+ //(D.f[DIR_00M])[kb ] = f1_T -c2over27*drho1;
+ //(D.f[DIR_PP0])[kne ] = f1_SW -c1over54*drho1;
+ //(D.f[DIR_MM0])[ksw ] = f1_NE -c1over54*drho1;
+ //(D.f[DIR_PM0])[kse ] = f1_NW -c1over54*drho1;
+ //(D.f[DIR_MP0])[knw ] = f1_SE -c1over54*drho1;
+ //(D.f[DIR_P0P])[kte ] = f1_BW -c1over54*drho1;
+ //(D.f[DIR_M0M])[kbw ] = f1_TE -c1over54*drho1;
+ //(D.f[DIR_P0M])[kbe ] = f1_TW -c1over54*drho1;
+ //(D.f[DIR_M0P])[ktw ] = f1_BE -c1over54*drho1;
+ //(D.f[DIR_0PP])[ktn ] = f1_BS -c1over54*drho1;
+ //(D.f[DIR_0MM])[kbs ] = f1_TN -c1over54*drho1;
+ //(D.f[DIR_0PM])[kbn ] = f1_TS -c1over54*drho1;
+ //(D.f[DIR_0MP])[kts ] = f1_BN -c1over54*drho1;
//(D.f[DIR_000])[kzero] = f1_ZERO-c8over27*drho1;
- //(D.f[DIR_PPP ])[ktne ] = f1_BSW -c1over216*drho1;
- //(D.f[DIR_MMP ])[ktsw ] = f1_BNE -c1over216*drho1;
- //(D.f[DIR_PMP ])[ktse ] = f1_BNW -c1over216*drho1;
- //(D.f[DIR_MPP ])[ktnw ] = f1_BSE -c1over216*drho1;
- //(D.f[DIR_PPM ])[kbne ] = f1_TSW -c1over216*drho1;
- //(D.f[DIR_MMM ])[kbsw ] = f1_TNE -c1over216*drho1;
- //(D.f[DIR_PMM ])[kbse ] = f1_TNW -c1over216*drho1;
- //(D.f[DIR_MPM ])[kbnw ] = f1_TSE -c1over216*drho1;
+ //(D.f[DIR_PPP])[ktne ] = f1_BSW -c1over216*drho1;
+ //(D.f[DIR_MMP])[ktsw ] = f1_BNE -c1over216*drho1;
+ //(D.f[DIR_PMP])[ktse ] = f1_BNW -c1over216*drho1;
+ //(D.f[DIR_MPP])[ktnw ] = f1_BSE -c1over216*drho1;
+ //(D.f[DIR_PPM])[kbne ] = f1_TSW -c1over216*drho1;
+ //(D.f[DIR_MMM])[kbsw ] = f1_TNE -c1over216*drho1;
+ //(D.f[DIR_PMM])[kbse ] = f1_TNW -c1over216*drho1;
+ //(D.f[DIR_MPM])[kbnw ] = f1_TSE -c1over216*drho1;
}
}
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
diff --git a/src/gpu/VirtualFluids_GPU/GPU/SetForcing27.cu b/src/gpu/VirtualFluids_GPU/GPU/SetForcing27.cu
index 8dbf2c670a549f9a6afe581510205c31246b50cb..07fc5853eb7042d5567c38a03cb27418142bf642 100644
--- a/src/gpu/VirtualFluids_GPU/GPU/SetForcing27.cu
+++ b/src/gpu/VirtualFluids_GPU/GPU/SetForcing27.cu
@@ -16,69 +16,69 @@ __global__ void GetVeloforForcing27( real* DD,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep)
{
Distributions27 D;
if (isEvenTimestep==false)
{
- D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
+ D.f[DIR_M00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_MPM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_PMP * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////
const unsigned x = threadIdx.x; // Globaler x-Index
@@ -123,33 +123,33 @@ __global__ void GetVeloforForcing27( real* DD,
unsigned int ktne = KQK;
unsigned int kbsw = neighborZ[ksw];
////////////////////////////////////////////////////////////////////////////////
- real mfcbb = (D.f[DIR_P00 ])[ke ];
- real mfabb = (D.f[DIR_M00 ])[kw ];
- real mfbcb = (D.f[DIR_0P0 ])[kn ];
- real mfbab = (D.f[DIR_0M0 ])[ks ];
- real mfbbc = (D.f[DIR_00P ])[kt ];
- real mfbba = (D.f[DIR_00M ])[kb ];
- real mfccb = (D.f[DIR_PP0 ])[kne ];
- real mfaab = (D.f[DIR_MM0 ])[ksw ];
- real mfcab = (D.f[DIR_PM0 ])[kse ];
- real mfacb = (D.f[DIR_MP0 ])[knw ];
- real mfcbc = (D.f[DIR_P0P ])[kte ];
- real mfaba = (D.f[DIR_M0M ])[kbw ];
- real mfcba = (D.f[DIR_P0M ])[kbe ];
- real mfabc = (D.f[DIR_M0P ])[ktw ];
- real mfbcc = (D.f[DIR_0PP ])[ktn ];
- real mfbaa = (D.f[DIR_0MM ])[kbs ];
- real mfbca = (D.f[DIR_0PM ])[kbn ];
- real mfbac = (D.f[DIR_0MP ])[kts ];
+ real mfcbb = (D.f[DIR_P00])[ke ];
+ real mfabb = (D.f[DIR_M00])[kw ];
+ real mfbcb = (D.f[DIR_0P0])[kn ];
+ real mfbab = (D.f[DIR_0M0])[ks ];
+ real mfbbc = (D.f[DIR_00P])[kt ];
+ real mfbba = (D.f[DIR_00M])[kb ];
+ real mfccb = (D.f[DIR_PP0])[kne ];
+ real mfaab = (D.f[DIR_MM0])[ksw ];
+ real mfcab = (D.f[DIR_PM0])[kse ];
+ real mfacb = (D.f[DIR_MP0])[knw ];
+ real mfcbc = (D.f[DIR_P0P])[kte ];
+ real mfaba = (D.f[DIR_M0M])[kbw ];
+ real mfcba = (D.f[DIR_P0M])[kbe ];
+ real mfabc = (D.f[DIR_M0P])[ktw ];
+ real mfbcc = (D.f[DIR_0PP])[ktn ];
+ real mfbaa = (D.f[DIR_0MM])[kbs ];
+ real mfbca = (D.f[DIR_0PM])[kbn ];
+ real mfbac = (D.f[DIR_0MP])[kts ];
real mfbbb = (D.f[DIR_000])[kzero];
- real mfccc = (D.f[DIR_PPP ])[ktne ];
- real mfaac = (D.f[DIR_MMP ])[ktsw ];
- real mfcac = (D.f[DIR_PMP ])[ktse ];
- real mfacc = (D.f[DIR_MPP ])[ktnw ];
- real mfcca = (D.f[DIR_PPM ])[kbne ];
- real mfaaa = (D.f[DIR_MMM ])[kbsw ];
- real mfcaa = (D.f[DIR_PMM ])[kbse ];
- real mfaca = (D.f[DIR_MPM ])[kbnw ];
+ real mfccc = (D.f[DIR_PPP])[ktne ];
+ real mfaac = (D.f[DIR_MMP])[ktsw ];
+ real mfcac = (D.f[DIR_PMP])[ktse ];
+ real mfacc = (D.f[DIR_MPP])[ktnw ];
+ real mfcca = (D.f[DIR_PPM])[kbne ];
+ real mfaaa = (D.f[DIR_MMM])[kbsw ];
+ real mfcaa = (D.f[DIR_PMM])[kbse ];
+ real mfaca = (D.f[DIR_MPM])[kbnw ];
////////////////////////////////////////////////////////////////////////////////////
real rho = (mfccc+mfaaa + mfaca+mfcac + mfacc+mfcaa + mfaac+mfcca +
mfbac+mfbca + mfbaa+mfbcc + mfabc+mfcba + mfaba+mfcbc + mfacb+mfcab + mfaab+mfccb +
diff --git a/src/gpu/VirtualFluids_GPU/GPU/SlipBCs27.cu b/src/gpu/VirtualFluids_GPU/GPU/SlipBCs27.cu
index 0079c927373e90c1e408d2c57ace0595bcfdff15..cc8ca53d15ac02686b850a70ab181bb47285a7d1 100644
--- a/src/gpu/VirtualFluids_GPU/GPU/SlipBCs27.cu
+++ b/src/gpu/VirtualFluids_GPU/GPU/SlipBCs27.cu
@@ -1,84 +1,117 @@
-/* Device code */
+//=======================================================================================
+// ____ ____ __ ______ __________ __ __ __ __
+// \ \ | | | | | _ \ |___ ___| | | | | / \ | |
+// \ \ | | | | | |_) | | | | | | | / \ | |
+// \ \ | | | | | _ / | | | | | | / /\ \ | |
+// \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
+// \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
+// \ \ | | ________________________________________________________________
+// \ \ | | | ______________________________________________________________|
+// \ \| | | | __ __ __ __ ______ _______
+// \ | | |_____ | | | | | | | | | _ \ / _____)
+// \ | | _____| | | | | | | | | | | \ \ \_______
+// \ | | | | |_____ | \_/ | | | | |_/ / _____ |
+// \ _____| |__| |________| \_______/ |__| |______/ (_______/
+//
+// This file is part of VirtualFluids. VirtualFluids is free software: you can
+// redistribute it and/or modify it under the terms of the GNU General Public
+// License as published by the Free Software Foundation, either version 3 of
+// the License, or (at your option) any later version.
+//
+// VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
+// ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+// FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
+// for more details.
+//
+// You should have received a copy of the GNU General Public License along
+// with VirtualFluids (see COPYING.txt). If not, see <http://www.gnu.org/licenses/>.
+//
+//! \file SlipBCs27.cu
+//! \ingroup GPU
+//! \author Martin Schoenherr, Anna Wellmann
+//======================================================================================
#include "LBM/LB.h"
#include "lbm/constants/D3Q27.h"
#include "lbm/constants/NumericConstants.h"
-#include "KernelUtilities.h"
+#include "LBM/GPUHelperFunctions/KernelUtilities.h"
using namespace vf::lbm::constant;
using namespace vf::lbm::dir;
+using namespace vf::gpu;
//////////////////////////////////////////////////////////////////////////////
-__global__ void QSlipDevice27(real* DD,
- int* k_Q,
- real* QQ,
- unsigned int numberOfBCnodes,
- real om1,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep)
+__global__ void QSlipDevice27(
+ real* DD,
+ int* k_Q,
+ real* QQ,
+ unsigned int numberOfBCnodes,
+ real om1,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep)
{
Distributions27 D;
if (isEvenTimestep==true)
{
- D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
+ D.f[DIR_M00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_MPM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_PMP * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////
const unsigned x = threadIdx.x; // Globaler x-Index
@@ -99,24 +132,24 @@ __global__ void QSlipDevice27(real* DD,
*q_dirBE, *q_dirTW, *q_dirTN, *q_dirBS, *q_dirBN, *q_dirTS,
*q_dirTNE, *q_dirTSW, *q_dirTSE, *q_dirTNW, *q_dirBNE, *q_dirBSW,
*q_dirBSE, *q_dirBNW;
- q_dirE = &QQ[DIR_P00 * numberOfBCnodes];
- q_dirW = &QQ[DIR_M00 * numberOfBCnodes];
- q_dirN = &QQ[DIR_0P0 * numberOfBCnodes];
- q_dirS = &QQ[DIR_0M0 * numberOfBCnodes];
- q_dirT = &QQ[DIR_00P * numberOfBCnodes];
- q_dirB = &QQ[DIR_00M * numberOfBCnodes];
- q_dirNE = &QQ[DIR_PP0 * numberOfBCnodes];
- q_dirSW = &QQ[DIR_MM0 * numberOfBCnodes];
- q_dirSE = &QQ[DIR_PM0 * numberOfBCnodes];
- q_dirNW = &QQ[DIR_MP0 * numberOfBCnodes];
- q_dirTE = &QQ[DIR_P0P * numberOfBCnodes];
- q_dirBW = &QQ[DIR_M0M * numberOfBCnodes];
- q_dirBE = &QQ[DIR_P0M * numberOfBCnodes];
- q_dirTW = &QQ[DIR_M0P * numberOfBCnodes];
- q_dirTN = &QQ[DIR_0PP * numberOfBCnodes];
- q_dirBS = &QQ[DIR_0MM * numberOfBCnodes];
- q_dirBN = &QQ[DIR_0PM * numberOfBCnodes];
- q_dirTS = &QQ[DIR_0MP * numberOfBCnodes];
+ q_dirE = &QQ[DIR_P00 * numberOfBCnodes];
+ q_dirW = &QQ[DIR_M00 * numberOfBCnodes];
+ q_dirN = &QQ[DIR_0P0 * numberOfBCnodes];
+ q_dirS = &QQ[DIR_0M0 * numberOfBCnodes];
+ q_dirT = &QQ[DIR_00P * numberOfBCnodes];
+ q_dirB = &QQ[DIR_00M * numberOfBCnodes];
+ q_dirNE = &QQ[DIR_PP0 * numberOfBCnodes];
+ q_dirSW = &QQ[DIR_MM0 * numberOfBCnodes];
+ q_dirSE = &QQ[DIR_PM0 * numberOfBCnodes];
+ q_dirNW = &QQ[DIR_MP0 * numberOfBCnodes];
+ q_dirTE = &QQ[DIR_P0P * numberOfBCnodes];
+ q_dirBW = &QQ[DIR_M0M * numberOfBCnodes];
+ q_dirBE = &QQ[DIR_P0M * numberOfBCnodes];
+ q_dirTW = &QQ[DIR_M0P * numberOfBCnodes];
+ q_dirTN = &QQ[DIR_0PP * numberOfBCnodes];
+ q_dirBS = &QQ[DIR_0MM * numberOfBCnodes];
+ q_dirBN = &QQ[DIR_0PM * numberOfBCnodes];
+ q_dirTS = &QQ[DIR_0MP * numberOfBCnodes];
q_dirTNE = &QQ[DIR_PPP * numberOfBCnodes];
q_dirTSW = &QQ[DIR_MMP * numberOfBCnodes];
q_dirTSE = &QQ[DIR_PMP * numberOfBCnodes];
@@ -156,32 +189,32 @@ __global__ void QSlipDevice27(real* DD,
unsigned int ktne = KQK;
unsigned int kbsw = neighborZ[ksw];
////////////////////////////////////////////////////////////////////////////////
- real f_W = (D.f[DIR_P00 ])[ke ];
- real f_E = (D.f[DIR_M00 ])[kw ];
- real f_S = (D.f[DIR_0P0 ])[kn ];
- real f_N = (D.f[DIR_0M0 ])[ks ];
- real f_B = (D.f[DIR_00P ])[kt ];
- real f_T = (D.f[DIR_00M ])[kb ];
- real f_SW = (D.f[DIR_PP0 ])[kne ];
- real f_NE = (D.f[DIR_MM0 ])[ksw ];
- real f_NW = (D.f[DIR_PM0 ])[kse ];
- real f_SE = (D.f[DIR_MP0 ])[knw ];
- real f_BW = (D.f[DIR_P0P ])[kte ];
- real f_TE = (D.f[DIR_M0M ])[kbw ];
- real f_TW = (D.f[DIR_P0M ])[kbe ];
- real f_BE = (D.f[DIR_M0P ])[ktw ];
- real f_BS = (D.f[DIR_0PP ])[ktn ];
- real f_TN = (D.f[DIR_0MM ])[kbs ];
- real f_TS = (D.f[DIR_0PM ])[kbn ];
- real f_BN = (D.f[DIR_0MP ])[kts ];
- real f_BSW = (D.f[DIR_PPP ])[ktne ];
- real f_BNE = (D.f[DIR_MMP ])[ktsw ];
- real f_BNW = (D.f[DIR_PMP ])[ktse ];
- real f_BSE = (D.f[DIR_MPP ])[ktnw ];
- real f_TSW = (D.f[DIR_PPM ])[kbne ];
- real f_TNE = (D.f[DIR_MMM ])[kbsw ];
- real f_TNW = (D.f[DIR_PMM ])[kbse ];
- real f_TSE = (D.f[DIR_MPM ])[kbnw ];
+ real f_W = (D.f[DIR_P00])[ke ];
+ real f_E = (D.f[DIR_M00])[kw ];
+ real f_S = (D.f[DIR_0P0])[kn ];
+ real f_N = (D.f[DIR_0M0])[ks ];
+ real f_B = (D.f[DIR_00P])[kt ];
+ real f_T = (D.f[DIR_00M])[kb ];
+ real f_SW = (D.f[DIR_PP0])[kne ];
+ real f_NE = (D.f[DIR_MM0])[ksw ];
+ real f_NW = (D.f[DIR_PM0])[kse ];
+ real f_SE = (D.f[DIR_MP0])[knw ];
+ real f_BW = (D.f[DIR_P0P])[kte ];
+ real f_TE = (D.f[DIR_M0M])[kbw ];
+ real f_TW = (D.f[DIR_P0M])[kbe ];
+ real f_BE = (D.f[DIR_M0P])[ktw ];
+ real f_BS = (D.f[DIR_0PP])[ktn ];
+ real f_TN = (D.f[DIR_0MM])[kbs ];
+ real f_TS = (D.f[DIR_0PM])[kbn ];
+ real f_BN = (D.f[DIR_0MP])[kts ];
+ real f_BSW = (D.f[DIR_PPP])[ktne ];
+ real f_BNE = (D.f[DIR_MMP])[ktsw ];
+ real f_BNW = (D.f[DIR_PMP])[ktse ];
+ real f_BSE = (D.f[DIR_MPP])[ktnw ];
+ real f_TSW = (D.f[DIR_PPM])[kbne ];
+ real f_TNE = (D.f[DIR_MMM])[kbsw ];
+ real f_TNW = (D.f[DIR_PMM])[kbse ];
+ real f_TSE = (D.f[DIR_MPM])[kbnw ];
////////////////////////////////////////////////////////////////////////////////
real vx1, vx2, vx3, drho, feq, q;
drho = f_TSE + f_TNW + f_TNE + f_TSW + f_BSE + f_BNW + f_BNE + f_BSW +
@@ -206,63 +239,63 @@ __global__ void QSlipDevice27(real* DD,
//////////////////////////////////////////////////////////////////////////
if (isEvenTimestep==false)
{
- D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
+ D.f[DIR_M00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_MPM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_PMP * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
//Test
@@ -659,32 +692,26 @@ __global__ void QSlipDevice27(real* DD,
//////////////////////////////////////////////////////////////////////////////
__global__ void QSlipDeviceComp27(
- real* distributions,
- int* subgridDistanceIndices,
- real* subgridDistances,
- unsigned int numberOfBCnodes,
- real omega,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int numberOfLBnodes,
- bool isEvenTimestep)
+ real* distributions,
+ int* subgridDistanceIndices,
+ real* subgridDistances,
+ unsigned int numberOfBCnodes,
+ real omega,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep)
{
//! The slip boundary condition is executed in the following steps
//!
+
////////////////////////////////////////////////////////////////////////////////
//! - Get node index coordinates from threadIdx, blockIdx, blockDim and gridDim.
//!
- const unsigned x = threadIdx.x; // global x-index
- const unsigned y = blockIdx.x; // global y-index
- const unsigned z = blockIdx.y; // global z-index
-
- const unsigned nx = blockDim.x;
- const unsigned ny = gridDim.x;
+ const unsigned nodeIndex = getNodeIndex();
- const unsigned k = nx*(ny*z + y) + x;
-
- if(k < numberOfBCnodes)
+ if(nodeIndex < numberOfBCnodes)
{
//////////////////////////////////////////////////////////////////////////
//! - Read distributions: style of reading and writing the distributions from/to stored arrays dependent on timestep is based on the esoteric twist algorithm \ref
@@ -702,7 +729,7 @@ __global__ void QSlipDeviceComp27(
////////////////////////////////////////////////////////////////////////////////
//! - Set neighbor indices (necessary for indirect addressing)
//!
- unsigned int indexOfBCnode = subgridDistanceIndices[k];
+ unsigned int indexOfBCnode = subgridDistanceIndices[nodeIndex];
unsigned int kzero= indexOfBCnode;
unsigned int ke = indexOfBCnode;
unsigned int kw = neighborX[indexOfBCnode];
@@ -734,32 +761,32 @@ __global__ void QSlipDeviceComp27(
////////////////////////////////////////////////////////////////////////////////
//! - Set local distributions
//!
- real f_W = (dist.f[DIR_P00 ])[ke ];
- real f_E = (dist.f[DIR_M00 ])[kw ];
- real f_S = (dist.f[DIR_0P0 ])[kn ];
- real f_N = (dist.f[DIR_0M0 ])[ks ];
- real f_B = (dist.f[DIR_00P ])[kt ];
- real f_T = (dist.f[DIR_00M ])[kb ];
- real f_SW = (dist.f[DIR_PP0 ])[kne ];
- real f_NE = (dist.f[DIR_MM0 ])[ksw ];
- real f_NW = (dist.f[DIR_PM0 ])[kse ];
- real f_SE = (dist.f[DIR_MP0 ])[knw ];
- real f_BW = (dist.f[DIR_P0P ])[kte ];
- real f_TE = (dist.f[DIR_M0M ])[kbw ];
- real f_TW = (dist.f[DIR_P0M ])[kbe ];
- real f_BE = (dist.f[DIR_M0P ])[ktw ];
- real f_BS = (dist.f[DIR_0PP ])[ktn ];
- real f_TN = (dist.f[DIR_0MM ])[kbs ];
- real f_TS = (dist.f[DIR_0PM ])[kbn ];
- real f_BN = (dist.f[DIR_0MP ])[kts ];
- real f_BSW = (dist.f[DIR_PPP ])[ktne ];
- real f_BNE = (dist.f[DIR_MMP ])[ktsw ];
- real f_BNW = (dist.f[DIR_PMP ])[ktse ];
- real f_BSE = (dist.f[DIR_MPP ])[ktnw ];
- real f_TSW = (dist.f[DIR_PPM ])[kbne ];
- real f_TNE = (dist.f[DIR_MMM ])[kbsw ];
- real f_TNW = (dist.f[DIR_PMM ])[kbse ];
- real f_TSE = (dist.f[DIR_MPM ])[kbnw ];
+ real f_W = (dist.f[DIR_P00])[ke ];
+ real f_E = (dist.f[DIR_M00])[kw ];
+ real f_S = (dist.f[DIR_0P0])[kn ];
+ real f_N = (dist.f[DIR_0M0])[ks ];
+ real f_B = (dist.f[DIR_00P])[kt ];
+ real f_T = (dist.f[DIR_00M])[kb ];
+ real f_SW = (dist.f[DIR_PP0])[kne ];
+ real f_NE = (dist.f[DIR_MM0])[ksw ];
+ real f_NW = (dist.f[DIR_PM0])[kse ];
+ real f_SE = (dist.f[DIR_MP0])[knw ];
+ real f_BW = (dist.f[DIR_P0P])[kte ];
+ real f_TE = (dist.f[DIR_M0M])[kbw ];
+ real f_TW = (dist.f[DIR_P0M])[kbe ];
+ real f_BE = (dist.f[DIR_M0P])[ktw ];
+ real f_BS = (dist.f[DIR_0PP])[ktn ];
+ real f_TN = (dist.f[DIR_0MM])[kbs ];
+ real f_TS = (dist.f[DIR_0PM])[kbn ];
+ real f_BN = (dist.f[DIR_0MP])[kts ];
+ real f_BSW = (dist.f[DIR_PPP])[ktne ];
+ real f_BNE = (dist.f[DIR_MMP])[ktsw ];
+ real f_BNW = (dist.f[DIR_PMP])[ktse ];
+ real f_BSE = (dist.f[DIR_MPP])[ktnw ];
+ real f_TSW = (dist.f[DIR_PPM])[kbne ];
+ real f_TNE = (dist.f[DIR_MMM])[kbsw ];
+ real f_TNW = (dist.f[DIR_PMM])[kbse ];
+ real f_TSE = (dist.f[DIR_MPM])[kbnw ];
////////////////////////////////////////////////////////////////////////////////
//! - Calculate macroscopic quantities
@@ -804,7 +831,7 @@ __global__ void QSlipDeviceComp27(
bool y = false;
bool z = false;
- q = (subgridD.q[DIR_P00])[k];
+ q = (subgridD.q[DIR_P00])[nodeIndex];
if (q>=c0o1 && q<=c1o1) // only update distribution for q between zero and one
{
VeloX = c0o1;
@@ -816,7 +843,7 @@ __global__ void QSlipDeviceComp27(
(dist.f[DIR_M00])[kw] = getInterpolatedDistributionForVeloBC(q, f_E, f_W, feq, omega, velocityBC, c2o27);
}
- q = (subgridD.q[DIR_M00])[k];
+ q = (subgridD.q[DIR_M00])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloX = c0o1;
@@ -828,7 +855,7 @@ __global__ void QSlipDeviceComp27(
(dist.f[DIR_P00])[ke] = getInterpolatedDistributionForVeloBC(q, f_W, f_E, feq, omega, velocityBC, c2o27);
}
- q = (subgridD.q[DIR_0P0])[k];
+ q = (subgridD.q[DIR_0P0])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloY = c0o1;
@@ -840,7 +867,7 @@ __global__ void QSlipDeviceComp27(
(dist.f[DIR_0M0])[ks] = getInterpolatedDistributionForVeloBC(q, f_N, f_S, feq, omega, velocityBC, c2o27);
}
- q = (subgridD.q[DIR_0M0])[k];
+ q = (subgridD.q[DIR_0M0])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloY = c0o1;
@@ -852,7 +879,7 @@ __global__ void QSlipDeviceComp27(
(dist.f[DIR_0P0])[kn] = getInterpolatedDistributionForVeloBC(q, f_S, f_N, feq, omega, velocityBC, c2o27);
}
- q = (subgridD.q[DIR_00P])[k];
+ q = (subgridD.q[DIR_00P])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloZ = c0o1;
@@ -864,7 +891,7 @@ __global__ void QSlipDeviceComp27(
(dist.f[DIR_00M])[kb] = getInterpolatedDistributionForVeloBC(q, f_T, f_B, feq, omega, velocityBC, c2o27);
}
- q = (subgridD.q[DIR_00M])[k];
+ q = (subgridD.q[DIR_00M])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloZ = c0o1;
@@ -876,7 +903,7 @@ __global__ void QSlipDeviceComp27(
(dist.f[DIR_00P])[kt] = getInterpolatedDistributionForVeloBC(q, f_B, f_T, feq, omega, velocityBC, c2o27);
}
- q = (subgridD.q[DIR_PP0])[k];
+ q = (subgridD.q[DIR_PP0])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloX = slipLength*vx1;
@@ -890,7 +917,7 @@ __global__ void QSlipDeviceComp27(
(dist.f[DIR_MM0])[ksw] = getInterpolatedDistributionForVeloBC(q, f_NE, f_SW, feq, omega, velocityBC, c1o54);
}
- q = (subgridD.q[DIR_MM0])[k];
+ q = (subgridD.q[DIR_MM0])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloX = slipLength*vx1;
@@ -904,7 +931,7 @@ __global__ void QSlipDeviceComp27(
(dist.f[DIR_PP0])[kne] = getInterpolatedDistributionForVeloBC(q, f_SW, f_NE, feq, omega, velocityBC, c1o54);
}
- q = (subgridD.q[DIR_PM0])[k];
+ q = (subgridD.q[DIR_PM0])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloX = slipLength*vx1;
@@ -918,7 +945,7 @@ __global__ void QSlipDeviceComp27(
(dist.f[DIR_MP0])[knw] = getInterpolatedDistributionForVeloBC(q, f_SE, f_NW, feq, omega, velocityBC, c1o54);
}
- q = (subgridD.q[DIR_MP0])[k];
+ q = (subgridD.q[DIR_MP0])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloX = slipLength*vx1;
@@ -932,7 +959,7 @@ __global__ void QSlipDeviceComp27(
(dist.f[DIR_PM0])[kse] = getInterpolatedDistributionForVeloBC(q, f_NW, f_SE, feq, omega, velocityBC, c1o54);
}
- q = (subgridD.q[DIR_P0P])[k];
+ q = (subgridD.q[DIR_P0P])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloX = slipLength*vx1;
@@ -946,7 +973,7 @@ __global__ void QSlipDeviceComp27(
(dist.f[DIR_M0M])[kbw] = getInterpolatedDistributionForVeloBC(q, f_TE, f_BW, feq, omega, velocityBC, c1o54);
}
- q = (subgridD.q[DIR_M0M])[k];
+ q = (subgridD.q[DIR_M0M])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloX = slipLength*vx1;
@@ -955,12 +982,12 @@ __global__ void QSlipDeviceComp27(
if (z == true) VeloZ = c0o1;
velocityLB = -vx1 - vx3;
- feq = getEquilibriumForBC(drho, velocityLB, cu_sq, c1o54);
+ feq = getEquilibriumForBC(drho, velocityLB, cu_sq, c1o54);
velocityBC = -VeloX - VeloZ;
(dist.f[DIR_P0P])[kte] = getInterpolatedDistributionForVeloBC(q, f_BW, f_TE, feq, omega, velocityBC, c1o54);
}
- q = (subgridD.q[DIR_P0M])[k];
+ q = (subgridD.q[DIR_P0M])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloX = slipLength*vx1;
@@ -974,7 +1001,7 @@ __global__ void QSlipDeviceComp27(
(dist.f[DIR_M0P])[ktw] = getInterpolatedDistributionForVeloBC(q, f_BE, f_TW, feq, omega, velocityBC, c1o54);
}
- q = (subgridD.q[DIR_M0P])[k];
+ q = (subgridD.q[DIR_M0P])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloX = slipLength*vx1;
@@ -988,7 +1015,7 @@ __global__ void QSlipDeviceComp27(
(dist.f[DIR_P0M])[kbe] = getInterpolatedDistributionForVeloBC(q, f_TW, f_BE, feq, omega, velocityBC, c1o54);
}
- q = (subgridD.q[DIR_0PP])[k];
+ q = (subgridD.q[DIR_0PP])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloY = slipLength*vx2;
@@ -1002,7 +1029,7 @@ __global__ void QSlipDeviceComp27(
(dist.f[DIR_0MM])[kbs] = getInterpolatedDistributionForVeloBC(q, f_TN, f_BS, feq, omega, velocityBC, c1o54);
}
- q = (subgridD.q[DIR_0MM])[k];
+ q = (subgridD.q[DIR_0MM])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloY = slipLength*vx2;
@@ -1017,7 +1044,7 @@ __global__ void QSlipDeviceComp27(
}
- q = (subgridD.q[DIR_0PM])[k];
+ q = (subgridD.q[DIR_0PM])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloY = slipLength*vx2;
@@ -1031,7 +1058,7 @@ __global__ void QSlipDeviceComp27(
(dist.f[DIR_0MP])[kts] = getInterpolatedDistributionForVeloBC(q, f_BN, f_TS, feq, omega, velocityBC, c1o54);
}
- q = (subgridD.q[DIR_0MP])[k];
+ q = (subgridD.q[DIR_0MP])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloY = slipLength*vx2;
@@ -1045,7 +1072,7 @@ __global__ void QSlipDeviceComp27(
(dist.f[DIR_0PM])[kbn] = getInterpolatedDistributionForVeloBC(q, f_TS, f_BN, feq, omega, velocityBC, c1o54);
}
- q = (subgridD.q[DIR_PPP])[k];
+ q = (subgridD.q[DIR_PPP])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloX = slipLength*vx1;
@@ -1060,7 +1087,7 @@ __global__ void QSlipDeviceComp27(
(dist.f[DIR_MMM])[kbsw] = getInterpolatedDistributionForVeloBC(q, f_TNE, f_BSW, feq, omega, velocityBC, c1o216);
}
- q = (subgridD.q[DIR_MMM])[k];
+ q = (subgridD.q[DIR_MMM])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloX = slipLength*vx1;
@@ -1076,7 +1103,7 @@ __global__ void QSlipDeviceComp27(
}
- q = (subgridD.q[DIR_PPM])[k];
+ q = (subgridD.q[DIR_PPM])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloX = slipLength*vx1;
@@ -1091,7 +1118,7 @@ __global__ void QSlipDeviceComp27(
(dist.f[DIR_MMP])[ktsw] = getInterpolatedDistributionForVeloBC(q, f_BNE, f_TSW, feq, omega, velocityBC, c1o216);
}
- q = (subgridD.q[DIR_MMP])[k];
+ q = (subgridD.q[DIR_MMP])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloX = slipLength*vx1;
@@ -1106,7 +1133,7 @@ __global__ void QSlipDeviceComp27(
(dist.f[DIR_PPM])[kbne] = getInterpolatedDistributionForVeloBC(q, f_TSW, f_BNE, feq, omega, velocityBC, c1o216);
}
- q = (subgridD.q[DIR_PMP])[k];
+ q = (subgridD.q[DIR_PMP])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloX = slipLength*vx1;
@@ -1121,7 +1148,7 @@ __global__ void QSlipDeviceComp27(
(dist.f[DIR_MPM])[kbnw] = getInterpolatedDistributionForVeloBC(q, f_TSE, f_BNW, feq, omega, velocityBC, c1o216);
}
- q = (subgridD.q[DIR_MPM])[k];
+ q = (subgridD.q[DIR_MPM])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloX = slipLength*vx1;
@@ -1136,7 +1163,7 @@ __global__ void QSlipDeviceComp27(
(dist.f[DIR_PMP])[ktse] = getInterpolatedDistributionForVeloBC(q, f_BNW, f_TSE, feq, omega, velocityBC, c1o216);
}
- q = (subgridD.q[DIR_PMM])[k];
+ q = (subgridD.q[DIR_PMM])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloX = slipLength*vx1;
@@ -1151,7 +1178,7 @@ __global__ void QSlipDeviceComp27(
(dist.f[DIR_MPP])[ktnw] = getInterpolatedDistributionForVeloBC(q, f_BSE, f_TNW, feq, omega, velocityBC, c1o216);
}
- q = (subgridD.q[DIR_MPP])[k];
+ q = (subgridD.q[DIR_MPP])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloX = slipLength*vx1;
@@ -1169,34 +1196,53 @@ __global__ void QSlipDeviceComp27(
}
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
//////////////////////////////////////////////////////////////////////////////
__global__ void BBSlipDeviceComp27(
- real* distributions,
- int* subgridDistanceIndices,
- real* subgridDistances,
- unsigned int numberOfBCnodes,
- real omega,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int numberOfLBnodes,
- bool isEvenTimestep)
+ real* distributions,
+ int* subgridDistanceIndices,
+ real* subgridDistances,
+ unsigned int numberOfBCnodes,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep)
{
//! The slip boundary condition is executed in the following steps
//!
+
////////////////////////////////////////////////////////////////////////////////
//! - Get node index coordinates from threadIdx, blockIdx, blockDim and gridDim.
//!
- const unsigned x = threadIdx.x; // global x-index
- const unsigned y = blockIdx.x; // global y-index
- const unsigned z = blockIdx.y; // global z-index
-
- const unsigned nx = blockDim.x;
- const unsigned ny = gridDim.x;
-
- const unsigned k = nx*(ny*z + y) + x;
+ const unsigned nodeIndex = getNodeIndex();
- if(k < numberOfBCnodes)
+ if(nodeIndex < numberOfBCnodes)
{
//////////////////////////////////////////////////////////////////////////
//! - Read distributions: style of reading and writing the distributions from/to stored arrays dependent on timestep is based on the esoteric twist algorithm \ref
@@ -1214,7 +1260,7 @@ __global__ void BBSlipDeviceComp27(
////////////////////////////////////////////////////////////////////////////////
//! - Set neighbor indices (necessary for indirect addressing)
//!
- unsigned int indexOfBCnode = subgridDistanceIndices[k];
+ unsigned int indexOfBCnode = subgridDistanceIndices[nodeIndex];
unsigned int kzero= indexOfBCnode;
unsigned int ke = indexOfBCnode;
unsigned int kw = neighborX[indexOfBCnode];
@@ -1246,32 +1292,32 @@ __global__ void BBSlipDeviceComp27(
////////////////////////////////////////////////////////////////////////////////
//! - Set local distributions
//!
- real f_W = (dist.f[DIR_P00 ])[ke ];
- real f_E = (dist.f[DIR_M00 ])[kw ];
- real f_S = (dist.f[DIR_0P0 ])[kn ];
- real f_N = (dist.f[DIR_0M0 ])[ks ];
- real f_B = (dist.f[DIR_00P ])[kt ];
- real f_T = (dist.f[DIR_00M ])[kb ];
- real f_SW = (dist.f[DIR_PP0 ])[kne ];
- real f_NE = (dist.f[DIR_MM0 ])[ksw ];
- real f_NW = (dist.f[DIR_PM0 ])[kse ];
- real f_SE = (dist.f[DIR_MP0 ])[knw ];
- real f_BW = (dist.f[DIR_P0P ])[kte ];
- real f_TE = (dist.f[DIR_M0M ])[kbw ];
- real f_TW = (dist.f[DIR_P0M ])[kbe ];
- real f_BE = (dist.f[DIR_M0P ])[ktw ];
- real f_BS = (dist.f[DIR_0PP ])[ktn ];
- real f_TN = (dist.f[DIR_0MM ])[kbs ];
- real f_TS = (dist.f[DIR_0PM ])[kbn ];
- real f_BN = (dist.f[DIR_0MP ])[kts ];
- real f_BSW = (dist.f[DIR_PPP ])[ktne ];
- real f_BNE = (dist.f[DIR_MMP ])[ktsw ];
- real f_BNW = (dist.f[DIR_PMP ])[ktse ];
- real f_BSE = (dist.f[DIR_MPP ])[ktnw ];
- real f_TSW = (dist.f[DIR_PPM ])[kbne ];
- real f_TNE = (dist.f[DIR_MMM ])[kbsw ];
- real f_TNW = (dist.f[DIR_PMM ])[kbse ];
- real f_TSE = (dist.f[DIR_MPM ])[kbnw ];
+ real f_W = (dist.f[DIR_P00])[ke ];
+ real f_E = (dist.f[DIR_M00])[kw ];
+ real f_S = (dist.f[DIR_0P0])[kn ];
+ real f_N = (dist.f[DIR_0M0])[ks ];
+ real f_B = (dist.f[DIR_00P])[kt ];
+ real f_T = (dist.f[DIR_00M])[kb ];
+ real f_SW = (dist.f[DIR_PP0])[kne ];
+ real f_NE = (dist.f[DIR_MM0])[ksw ];
+ real f_NW = (dist.f[DIR_PM0])[kse ];
+ real f_SE = (dist.f[DIR_MP0])[knw ];
+ real f_BW = (dist.f[DIR_P0P])[kte ];
+ real f_TE = (dist.f[DIR_M0M])[kbw ];
+ real f_TW = (dist.f[DIR_P0M])[kbe ];
+ real f_BE = (dist.f[DIR_M0P])[ktw ];
+ real f_BS = (dist.f[DIR_0PP])[ktn ];
+ real f_TN = (dist.f[DIR_0MM])[kbs ];
+ real f_TS = (dist.f[DIR_0PM])[kbn ];
+ real f_BN = (dist.f[DIR_0MP])[kts ];
+ real f_BSW = (dist.f[DIR_PPP])[ktne ];
+ real f_BNE = (dist.f[DIR_MMP])[ktsw ];
+ real f_BNW = (dist.f[DIR_PMP])[ktse ];
+ real f_BSE = (dist.f[DIR_MPP])[ktnw ];
+ real f_TSW = (dist.f[DIR_PPM])[kbne ];
+ real f_TNE = (dist.f[DIR_MMM])[kbsw ];
+ real f_TNW = (dist.f[DIR_PMM])[kbse ];
+ real f_TSE = (dist.f[DIR_MPM])[kbnw ];
////////////////////////////////////////////////////////////////////////////////
//! - Calculate macroscopic quantities
@@ -1316,7 +1362,7 @@ __global__ void BBSlipDeviceComp27(
bool y = false;
bool z = false;
- q = (subgridD.q[DIR_P00])[k];
+ q = (subgridD.q[DIR_P00])[nodeIndex];
if (q>=c0o1 && q<=c1o1) // only update distribution for q between zero and one
{
VeloX = c0o1;
@@ -1326,7 +1372,7 @@ __global__ void BBSlipDeviceComp27(
(dist.f[DIR_M00])[kw] = getBounceBackDistributionForVeloBC(f_W, velocityBC, c2o27);
}
- q = (subgridD.q[DIR_M00])[k];
+ q = (subgridD.q[DIR_M00])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloX = c0o1;
@@ -1336,7 +1382,7 @@ __global__ void BBSlipDeviceComp27(
(dist.f[DIR_P00])[ke] = getBounceBackDistributionForVeloBC(f_E, velocityBC, c2o27);
}
- q = (subgridD.q[DIR_0P0])[k];
+ q = (subgridD.q[DIR_0P0])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloY = c0o1;
@@ -1346,7 +1392,7 @@ __global__ void BBSlipDeviceComp27(
(dist.f[DIR_0M0])[ks] = getBounceBackDistributionForVeloBC(f_S, velocityBC, c2o27);
}
- q = (subgridD.q[DIR_0M0])[k];
+ q = (subgridD.q[DIR_0M0])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloY = c0o1;
@@ -1356,7 +1402,7 @@ __global__ void BBSlipDeviceComp27(
(dist.f[DIR_0P0])[kn] = getBounceBackDistributionForVeloBC(f_N, velocityBC, c2o27);
}
- q = (subgridD.q[DIR_00P])[k];
+ q = (subgridD.q[DIR_00P])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloZ = c0o1;
@@ -1366,7 +1412,7 @@ __global__ void BBSlipDeviceComp27(
(dist.f[DIR_00M])[kb] = getBounceBackDistributionForVeloBC(f_B, velocityBC, c2o27);
}
- q = (subgridD.q[DIR_00M])[k];
+ q = (subgridD.q[DIR_00M])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloZ = c0o1;
@@ -1376,7 +1422,7 @@ __global__ void BBSlipDeviceComp27(
(dist.f[DIR_00P])[kt] = getBounceBackDistributionForVeloBC(f_T, velocityBC, c2o27);
}
- q = (subgridD.q[DIR_PP0])[k];
+ q = (subgridD.q[DIR_PP0])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloX = slipLength*vx1;
@@ -1388,7 +1434,7 @@ __global__ void BBSlipDeviceComp27(
(dist.f[DIR_MM0])[ksw] = getBounceBackDistributionForVeloBC(f_SW, velocityBC, c1o54);
}
- q = (subgridD.q[DIR_MM0])[k];
+ q = (subgridD.q[DIR_MM0])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloX = slipLength*vx1;
@@ -1400,7 +1446,7 @@ __global__ void BBSlipDeviceComp27(
(dist.f[DIR_PP0])[kne] = getBounceBackDistributionForVeloBC(f_NE, velocityBC, c1o54);
}
- q = (subgridD.q[DIR_PM0])[k];
+ q = (subgridD.q[DIR_PM0])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloX = slipLength*vx1;
@@ -1412,7 +1458,7 @@ __global__ void BBSlipDeviceComp27(
(dist.f[DIR_MP0])[knw] = getBounceBackDistributionForVeloBC(f_NW, velocityBC, c1o54);
}
- q = (subgridD.q[DIR_MP0])[k];
+ q = (subgridD.q[DIR_MP0])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloX = slipLength*vx1;
@@ -1424,7 +1470,7 @@ __global__ void BBSlipDeviceComp27(
(dist.f[DIR_PM0])[kse] = getBounceBackDistributionForVeloBC(f_SE, velocityBC, c1o54);
}
- q = (subgridD.q[DIR_P0P])[k];
+ q = (subgridD.q[DIR_P0P])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloX = slipLength*vx1;
@@ -1436,7 +1482,7 @@ __global__ void BBSlipDeviceComp27(
(dist.f[DIR_M0M])[kbw] = getBounceBackDistributionForVeloBC(f_BW, velocityBC, c1o54);
}
- q = (subgridD.q[DIR_M0M])[k];
+ q = (subgridD.q[DIR_M0M])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloX = slipLength*vx1;
@@ -1444,11 +1490,11 @@ __global__ void BBSlipDeviceComp27(
if (x == true) VeloX = c0o1;
if (z == true) VeloZ = c0o1;
- velocityBC = -VeloX - VeloZ;
- (dist.f[DIR_P0P])[kte] = getBounceBackDistributionForVeloBC(f_TE, velocityBC, c1o54);
+ velocityBC = -VeloX - VeloZ;
+ (dist.f[DIR_P0P])[kte] = getBounceBackDistributionForVeloBC(f_TE, velocityBC, c1o54);
}
- q = (subgridD.q[DIR_P0M])[k];
+ q = (subgridD.q[DIR_P0M])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloX = slipLength*vx1;
@@ -1460,7 +1506,7 @@ __global__ void BBSlipDeviceComp27(
(dist.f[DIR_M0P])[ktw] = getBounceBackDistributionForVeloBC(f_TW, velocityBC, c1o54);
}
- q = (subgridD.q[DIR_M0P])[k];
+ q = (subgridD.q[DIR_M0P])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloX = slipLength*vx1;
@@ -1472,7 +1518,7 @@ __global__ void BBSlipDeviceComp27(
(dist.f[DIR_P0M])[kbe] = getBounceBackDistributionForVeloBC(f_BE, velocityBC, c1o54);
}
- q = (subgridD.q[DIR_0PP])[k];
+ q = (subgridD.q[DIR_0PP])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloY = slipLength*vx2;
@@ -1484,7 +1530,7 @@ __global__ void BBSlipDeviceComp27(
(dist.f[DIR_0MM])[kbs] = getBounceBackDistributionForVeloBC(f_BS, velocityBC, c1o54);
}
- q = (subgridD.q[DIR_0MM])[k];
+ q = (subgridD.q[DIR_0MM])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloY = slipLength*vx2;
@@ -1497,7 +1543,7 @@ __global__ void BBSlipDeviceComp27(
}
- q = (subgridD.q[DIR_0PM])[k];
+ q = (subgridD.q[DIR_0PM])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloY = slipLength*vx2;
@@ -1509,7 +1555,7 @@ __global__ void BBSlipDeviceComp27(
(dist.f[DIR_0MP])[kts] = getBounceBackDistributionForVeloBC(f_TS, velocityBC, c1o54);
}
- q = (subgridD.q[DIR_0MP])[k];
+ q = (subgridD.q[DIR_0MP])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloY = slipLength*vx2;
@@ -1521,7 +1567,7 @@ __global__ void BBSlipDeviceComp27(
(dist.f[DIR_0PM])[kbn] = getBounceBackDistributionForVeloBC(f_BN, velocityBC, c1o54);
}
- q = (subgridD.q[DIR_PPP])[k];
+ q = (subgridD.q[DIR_PPP])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloX = slipLength*vx1;
@@ -1535,7 +1581,7 @@ __global__ void BBSlipDeviceComp27(
(dist.f[DIR_MMM])[kbsw] = getBounceBackDistributionForVeloBC(f_TNE, velocityBC, c1o216);
}
- q = (subgridD.q[DIR_MMM])[k];
+ q = (subgridD.q[DIR_MMM])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloX = slipLength*vx1;
@@ -1550,7 +1596,7 @@ __global__ void BBSlipDeviceComp27(
}
- q = (subgridD.q[DIR_PPM])[k];
+ q = (subgridD.q[DIR_PPM])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloX = slipLength*vx1;
@@ -1564,7 +1610,7 @@ __global__ void BBSlipDeviceComp27(
(dist.f[DIR_MMP])[ktsw] = getBounceBackDistributionForVeloBC(f_TSW, velocityBC, c1o216);
}
- q = (subgridD.q[DIR_MMP])[k];
+ q = (subgridD.q[DIR_MMP])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloX = slipLength*vx1;
@@ -1578,7 +1624,7 @@ __global__ void BBSlipDeviceComp27(
(dist.f[DIR_PPM])[kbne] = getBounceBackDistributionForVeloBC(f_BNE, velocityBC, c1o216);
}
- q = (subgridD.q[DIR_PMP])[k];
+ q = (subgridD.q[DIR_PMP])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloX = slipLength*vx1;
@@ -1592,7 +1638,7 @@ __global__ void BBSlipDeviceComp27(
(dist.f[DIR_MPM])[kbnw] = getBounceBackDistributionForVeloBC(f_BNW, velocityBC, c1o216);
}
- q = (subgridD.q[DIR_MPM])[k];
+ q = (subgridD.q[DIR_MPM])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloX = slipLength*vx1;
@@ -1606,7 +1652,7 @@ __global__ void BBSlipDeviceComp27(
(dist.f[DIR_PMP])[ktse] = getBounceBackDistributionForVeloBC(f_TSE, velocityBC, c1o216);
}
- q = (subgridD.q[DIR_PMM])[k];
+ q = (subgridD.q[DIR_PMM])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloX = slipLength*vx1;
@@ -1620,7 +1666,7 @@ __global__ void BBSlipDeviceComp27(
(dist.f[DIR_MPP])[ktnw] = getBounceBackDistributionForVeloBC(f_TNW, velocityBC, c1o216);
}
- q = (subgridD.q[DIR_MPP])[k];
+ q = (subgridD.q[DIR_MPP])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloX = slipLength*vx1;
@@ -1638,35 +1684,55 @@ __global__ void BBSlipDeviceComp27(
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
////////////////////////////////////////////////////////////////////////////
__global__ void QSlipDeviceComp27TurbViscosity(
- real* distributions,
- int* subgridDistanceIndices,
- real* subgridDistances,
- unsigned int numberOfBCnodes,
- real omega,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- real* turbViscosity,
- unsigned int numberOfLBnodes,
- bool isEvenTimestep)
+ real* distributions,
+ int* subgridDistanceIndices,
+ real* subgridDistances,
+ unsigned int numberOfBCnodes,
+ real omega,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ real* turbViscosity,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep)
{
//! The slip boundary condition is executed in the following steps
//!
+
////////////////////////////////////////////////////////////////////////////////
//! - Get node index coordinates from threadIdx, blockIdx, blockDim and gridDim.
//!
- const unsigned x = threadIdx.x; // global x-index
- const unsigned y = blockIdx.x; // global y-index
- const unsigned z = blockIdx.y; // global z-index
-
- const unsigned nx = blockDim.x;
- const unsigned ny = gridDim.x;
+ const unsigned nodeIndex = getNodeIndex();
- const unsigned k = nx*(ny*z + y) + x;
-
- if(k < numberOfBCnodes)
+ if(nodeIndex < numberOfBCnodes)
{
//////////////////////////////////////////////////////////////////////////
//! - Read distributions: style of reading and writing the distributions from/to stored arrays dependent on timestep is based on the esoteric twist algorithm \ref
@@ -1684,7 +1750,7 @@ __global__ void QSlipDeviceComp27TurbViscosity(
////////////////////////////////////////////////////////////////////////////////
//! - Set neighbor indices (necessary for indirect addressing)
//!
- unsigned int indexOfBCnode = subgridDistanceIndices[k];
+ unsigned int indexOfBCnode = subgridDistanceIndices[nodeIndex];
unsigned int kzero= indexOfBCnode;
unsigned int ke = indexOfBCnode;
unsigned int kw = neighborX[indexOfBCnode];
@@ -1716,32 +1782,32 @@ __global__ void QSlipDeviceComp27TurbViscosity(
////////////////////////////////////////////////////////////////////////////////
//! - Set local distributions
//!
- real f_W = (dist.f[DIR_P00 ])[ke ];
- real f_E = (dist.f[DIR_M00 ])[kw ];
- real f_S = (dist.f[DIR_0P0 ])[kn ];
- real f_N = (dist.f[DIR_0M0 ])[ks ];
- real f_B = (dist.f[DIR_00P ])[kt ];
- real f_T = (dist.f[DIR_00M ])[kb ];
- real f_SW = (dist.f[DIR_PP0 ])[kne ];
- real f_NE = (dist.f[DIR_MM0 ])[ksw ];
- real f_NW = (dist.f[DIR_PM0 ])[kse ];
- real f_SE = (dist.f[DIR_MP0 ])[knw ];
- real f_BW = (dist.f[DIR_P0P ])[kte ];
- real f_TE = (dist.f[DIR_M0M ])[kbw ];
- real f_TW = (dist.f[DIR_P0M ])[kbe ];
- real f_BE = (dist.f[DIR_M0P ])[ktw ];
- real f_BS = (dist.f[DIR_0PP ])[ktn ];
- real f_TN = (dist.f[DIR_0MM ])[kbs ];
- real f_TS = (dist.f[DIR_0PM ])[kbn ];
- real f_BN = (dist.f[DIR_0MP ])[kts ];
- real f_BSW = (dist.f[DIR_PPP ])[ktne ];
- real f_BNE = (dist.f[DIR_MMP ])[ktsw ];
- real f_BNW = (dist.f[DIR_PMP ])[ktse ];
- real f_BSE = (dist.f[DIR_MPP ])[ktnw ];
- real f_TSW = (dist.f[DIR_PPM ])[kbne ];
- real f_TNE = (dist.f[DIR_MMM ])[kbsw ];
- real f_TNW = (dist.f[DIR_PMM ])[kbse ];
- real f_TSE = (dist.f[DIR_MPM ])[kbnw ];
+ real f_W = (dist.f[DIR_P00])[ke ];
+ real f_E = (dist.f[DIR_M00])[kw ];
+ real f_S = (dist.f[DIR_0P0])[kn ];
+ real f_N = (dist.f[DIR_0M0])[ks ];
+ real f_B = (dist.f[DIR_00P])[kt ];
+ real f_T = (dist.f[DIR_00M])[kb ];
+ real f_SW = (dist.f[DIR_PP0])[kne ];
+ real f_NE = (dist.f[DIR_MM0])[ksw ];
+ real f_NW = (dist.f[DIR_PM0])[kse ];
+ real f_SE = (dist.f[DIR_MP0])[knw ];
+ real f_BW = (dist.f[DIR_P0P])[kte ];
+ real f_TE = (dist.f[DIR_M0M])[kbw ];
+ real f_TW = (dist.f[DIR_P0M])[kbe ];
+ real f_BE = (dist.f[DIR_M0P])[ktw ];
+ real f_BS = (dist.f[DIR_0PP])[ktn ];
+ real f_TN = (dist.f[DIR_0MM])[kbs ];
+ real f_TS = (dist.f[DIR_0PM])[kbn ];
+ real f_BN = (dist.f[DIR_0MP])[kts ];
+ real f_BSW = (dist.f[DIR_PPP])[ktne ];
+ real f_BNE = (dist.f[DIR_MMP])[ktsw ];
+ real f_BNW = (dist.f[DIR_PMP])[ktse ];
+ real f_BSE = (dist.f[DIR_MPP])[ktnw ];
+ real f_TSW = (dist.f[DIR_PPM])[kbne ];
+ real f_TNE = (dist.f[DIR_MMM])[kbsw ];
+ real f_TNW = (dist.f[DIR_PMM])[kbse ];
+ real f_TSE = (dist.f[DIR_MPM])[kbnw ];
////////////////////////////////////////////////////////////////////////////////
//! - Calculate macroscopic quantities
@@ -1791,7 +1857,7 @@ __global__ void QSlipDeviceComp27TurbViscosity(
bool y = false;
bool z = false;
- q = (subgridD.q[DIR_P00])[k];
+ q = (subgridD.q[DIR_P00])[nodeIndex];
if (q>=c0o1 && q<=c1o1) // only update distribution for q between zero and one
{
VeloX = c0o1;
@@ -1803,7 +1869,7 @@ __global__ void QSlipDeviceComp27TurbViscosity(
(dist.f[DIR_M00])[kw] = getInterpolatedDistributionForVeloBC(q, f_E, f_W, feq, om_turb, velocityBC, c2o27);
}
- q = (subgridD.q[DIR_M00])[k];
+ q = (subgridD.q[DIR_M00])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloX = c0o1;
@@ -1815,7 +1881,7 @@ __global__ void QSlipDeviceComp27TurbViscosity(
(dist.f[DIR_P00])[ke] = getInterpolatedDistributionForVeloBC(q, f_W, f_E, feq, om_turb, velocityBC, c2o27);
}
- q = (subgridD.q[DIR_0P0])[k];
+ q = (subgridD.q[DIR_0P0])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloY = c0o1;
@@ -1827,7 +1893,7 @@ __global__ void QSlipDeviceComp27TurbViscosity(
(dist.f[DIR_0M0])[ks] = getInterpolatedDistributionForVeloBC(q, f_N, f_S, feq, om_turb, velocityBC, c2o27);
}
- q = (subgridD.q[DIR_0M0])[k];
+ q = (subgridD.q[DIR_0M0])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloY = c0o1;
@@ -1839,7 +1905,7 @@ __global__ void QSlipDeviceComp27TurbViscosity(
(dist.f[DIR_0P0])[kn] = getInterpolatedDistributionForVeloBC(q, f_S, f_N, feq, om_turb, velocityBC, c2o27);
}
- q = (subgridD.q[DIR_00P])[k];
+ q = (subgridD.q[DIR_00P])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloZ = c0o1;
@@ -1851,7 +1917,7 @@ __global__ void QSlipDeviceComp27TurbViscosity(
(dist.f[DIR_00M])[kb] = getInterpolatedDistributionForVeloBC(q, f_T, f_B, feq, om_turb, velocityBC, c2o27);
}
- q = (subgridD.q[DIR_00M])[k];
+ q = (subgridD.q[DIR_00M])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloZ = c0o1;
@@ -1863,7 +1929,7 @@ __global__ void QSlipDeviceComp27TurbViscosity(
(dist.f[DIR_00P])[kt] = getInterpolatedDistributionForVeloBC(q, f_B, f_T, feq, om_turb, velocityBC, c2o27);
}
- q = (subgridD.q[DIR_PP0])[k];
+ q = (subgridD.q[DIR_PP0])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloX = slipLength*vx1;
@@ -1877,7 +1943,7 @@ __global__ void QSlipDeviceComp27TurbViscosity(
(dist.f[DIR_MM0])[ksw] = getInterpolatedDistributionForVeloBC(q, f_NE, f_SW, feq, om_turb, velocityBC, c1o54);
}
- q = (subgridD.q[DIR_MM0])[k];
+ q = (subgridD.q[DIR_MM0])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloX = slipLength*vx1;
@@ -1891,7 +1957,7 @@ __global__ void QSlipDeviceComp27TurbViscosity(
(dist.f[DIR_PP0])[kne] = getInterpolatedDistributionForVeloBC(q, f_SW, f_NE, feq, om_turb, velocityBC, c1o54);
}
- q = (subgridD.q[DIR_PM0])[k];
+ q = (subgridD.q[DIR_PM0])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloX = slipLength*vx1;
@@ -1905,7 +1971,7 @@ __global__ void QSlipDeviceComp27TurbViscosity(
(dist.f[DIR_MP0])[knw] = getInterpolatedDistributionForVeloBC(q, f_SE, f_NW, feq, om_turb, velocityBC, c1o54);
}
- q = (subgridD.q[DIR_MP0])[k];
+ q = (subgridD.q[DIR_MP0])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloX = slipLength*vx1;
@@ -1919,7 +1985,7 @@ __global__ void QSlipDeviceComp27TurbViscosity(
(dist.f[DIR_PM0])[kse] = getInterpolatedDistributionForVeloBC(q, f_NW, f_SE, feq, om_turb, velocityBC, c1o54);
}
- q = (subgridD.q[DIR_P0P])[k];
+ q = (subgridD.q[DIR_P0P])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloX = slipLength*vx1;
@@ -1933,7 +1999,7 @@ __global__ void QSlipDeviceComp27TurbViscosity(
(dist.f[DIR_M0M])[kbw] = getInterpolatedDistributionForVeloBC(q, f_TE, f_BW, feq, om_turb, velocityBC, c1o54);
}
- q = (subgridD.q[DIR_M0M])[k];
+ q = (subgridD.q[DIR_M0M])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloX = slipLength*vx1;
@@ -1941,13 +2007,13 @@ __global__ void QSlipDeviceComp27TurbViscosity(
if (x == true) VeloX = c0o1;
if (z == true) VeloZ = c0o1;
- velocityLB = -vx1 - vx3;
- feq = getEquilibriumForBC(drho, velocityLB, cu_sq, c1o54);
- velocityBC = -VeloX - VeloZ;
- (dist.f[DIR_P0P])[kte] = getInterpolatedDistributionForVeloBC(q, f_BW, f_TE, feq, om_turb, velocityBC, c1o54);
+ velocityLB = -vx1 - vx3;
+ feq = getEquilibriumForBC(drho, velocityLB, cu_sq, c1o54);
+ velocityBC = -VeloX - VeloZ;
+ (dist.f[DIR_P0P])[kte] = getInterpolatedDistributionForVeloBC(q, f_BW, f_TE, feq, om_turb, velocityBC, c1o54);
}
- q = (subgridD.q[DIR_P0M])[k];
+ q = (subgridD.q[DIR_P0M])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloX = slipLength*vx1;
@@ -1961,7 +2027,7 @@ __global__ void QSlipDeviceComp27TurbViscosity(
(dist.f[DIR_M0P])[ktw] = getInterpolatedDistributionForVeloBC(q, f_BE, f_TW, feq, om_turb, velocityBC, c1o54);
}
- q = (subgridD.q[DIR_M0P])[k];
+ q = (subgridD.q[DIR_M0P])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloX = slipLength*vx1;
@@ -1975,7 +2041,7 @@ __global__ void QSlipDeviceComp27TurbViscosity(
(dist.f[DIR_P0M])[kbe] = getInterpolatedDistributionForVeloBC(q, f_TW, f_BE, feq, om_turb, velocityBC, c1o54);
}
- q = (subgridD.q[DIR_0PP])[k];
+ q = (subgridD.q[DIR_0PP])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloY = slipLength*vx2;
@@ -1989,7 +2055,7 @@ __global__ void QSlipDeviceComp27TurbViscosity(
(dist.f[DIR_0MM])[kbs] = getInterpolatedDistributionForVeloBC(q, f_TN, f_BS, feq, om_turb, velocityBC, c1o54);
}
- q = (subgridD.q[DIR_0MM])[k];
+ q = (subgridD.q[DIR_0MM])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloY = slipLength*vx2;
@@ -2004,7 +2070,7 @@ __global__ void QSlipDeviceComp27TurbViscosity(
}
- q = (subgridD.q[DIR_0PM])[k];
+ q = (subgridD.q[DIR_0PM])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloY = slipLength*vx2;
@@ -2018,7 +2084,7 @@ __global__ void QSlipDeviceComp27TurbViscosity(
(dist.f[DIR_0MP])[kts] = getInterpolatedDistributionForVeloBC(q, f_BN, f_TS, feq, om_turb, velocityBC, c1o54);
}
- q = (subgridD.q[DIR_0MP])[k];
+ q = (subgridD.q[DIR_0MP])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloY = slipLength*vx2;
@@ -2032,7 +2098,7 @@ __global__ void QSlipDeviceComp27TurbViscosity(
(dist.f[DIR_0PM])[kbn] = getInterpolatedDistributionForVeloBC(q, f_TS, f_BN, feq, om_turb, velocityBC, c1o54);
}
- q = (subgridD.q[DIR_PPP])[k];
+ q = (subgridD.q[DIR_PPP])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloX = slipLength*vx1;
@@ -2047,7 +2113,7 @@ __global__ void QSlipDeviceComp27TurbViscosity(
(dist.f[DIR_MMM])[kbsw] = getInterpolatedDistributionForVeloBC(q, f_TNE, f_BSW, feq, om_turb, velocityBC, c1o216);
}
- q = (subgridD.q[DIR_MMM])[k];
+ q = (subgridD.q[DIR_MMM])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloX = slipLength*vx1;
@@ -2063,7 +2129,7 @@ __global__ void QSlipDeviceComp27TurbViscosity(
}
- q = (subgridD.q[DIR_PPM])[k];
+ q = (subgridD.q[DIR_PPM])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloX = slipLength*vx1;
@@ -2078,7 +2144,7 @@ __global__ void QSlipDeviceComp27TurbViscosity(
(dist.f[DIR_MMP])[ktsw] = getInterpolatedDistributionForVeloBC(q, f_BNE, f_TSW, feq, om_turb, velocityBC, c1o216);
}
- q = (subgridD.q[DIR_MMP])[k];
+ q = (subgridD.q[DIR_MMP])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloX = slipLength*vx1;
@@ -2093,7 +2159,7 @@ __global__ void QSlipDeviceComp27TurbViscosity(
(dist.f[DIR_PPM])[kbne] = getInterpolatedDistributionForVeloBC(q, f_TSW, f_BNE, feq, om_turb, velocityBC, c1o216);
}
- q = (subgridD.q[DIR_PMP])[k];
+ q = (subgridD.q[DIR_PMP])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloX = slipLength*vx1;
@@ -2108,7 +2174,7 @@ __global__ void QSlipDeviceComp27TurbViscosity(
(dist.f[DIR_MPM])[kbnw] = getInterpolatedDistributionForVeloBC(q, f_TSE, f_BNW, feq, om_turb, velocityBC, c1o216);
}
- q = (subgridD.q[DIR_MPM])[k];
+ q = (subgridD.q[DIR_MPM])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloX = slipLength*vx1;
@@ -2123,7 +2189,7 @@ __global__ void QSlipDeviceComp27TurbViscosity(
(dist.f[DIR_PMP])[ktse] = getInterpolatedDistributionForVeloBC(q, f_BNW, f_TSE, feq, om_turb, velocityBC, c1o216);
}
- q = (subgridD.q[DIR_PMM])[k];
+ q = (subgridD.q[DIR_PMM])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloX = slipLength*vx1;
@@ -2138,7 +2204,7 @@ __global__ void QSlipDeviceComp27TurbViscosity(
(dist.f[DIR_MPP])[ktnw] = getInterpolatedDistributionForVeloBC(q, f_BSE, f_TNW, feq, om_turb, velocityBC, c1o216);
}
- q = (subgridD.q[DIR_MPP])[k];
+ q = (subgridD.q[DIR_MPP])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloX = slipLength*vx1;
@@ -2154,37 +2220,59 @@ __global__ void QSlipDeviceComp27TurbViscosity(
}
}
}
+////////////////////////////////////////////////////////////////////////////
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
////////////////////////////////////////////////////////////////////////////
__global__ void QSlipPressureDeviceComp27TurbViscosity(
- real* distributions,
- int* subgridDistanceIndices,
- real* subgridDistances,
- unsigned int numberOfBCnodes,
- real omega,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- real* turbViscosity,
- unsigned int numberOfLBnodes,
- bool isEvenTimestep)
+ real* distributions,
+ int* subgridDistanceIndices,
+ real* subgridDistances,
+ unsigned int numberOfBCnodes,
+ real omega,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ real* turbViscosity,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep)
{
//! The slip boundary condition is executed in the following steps
//!
////////////////////////////////////////////////////////////////////////////////
//! - Get node index coordinates from threadIdx, blockIdx, blockDim and gridDim.
//!
- const unsigned x = threadIdx.x; // global x-index
- const unsigned y = blockIdx.x; // global y-index
- const unsigned z = blockIdx.y; // global z-index
-
- const unsigned nx = blockDim.x;
- const unsigned ny = gridDim.x;
-
- const unsigned k = nx*(ny*z + y) + x;
+ const unsigned nodeIndex = getNodeIndex();
- if(k < numberOfBCnodes)
+ if(nodeIndex < numberOfBCnodes)
{
//////////////////////////////////////////////////////////////////////////
//! - Read distributions: style of reading and writing the distributions from/to stored arrays dependent on timestep is based on the esoteric twist algorithm \ref
@@ -2202,7 +2290,7 @@ __global__ void QSlipPressureDeviceComp27TurbViscosity(
////////////////////////////////////////////////////////////////////////////////
//! - Set neighbor indices (necessary for indirect addressing)
//!
- unsigned int indexOfBCnode = subgridDistanceIndices[k];
+ unsigned int indexOfBCnode = subgridDistanceIndices[nodeIndex];
unsigned int kzero= indexOfBCnode;
unsigned int ke = indexOfBCnode;
unsigned int kw = neighborX[indexOfBCnode];
@@ -2234,32 +2322,32 @@ __global__ void QSlipPressureDeviceComp27TurbViscosity(
////////////////////////////////////////////////////////////////////////////////
//! - Set local distributions
//!
- real f_W = (dist.f[DIR_P00 ])[ke ];
- real f_E = (dist.f[DIR_M00 ])[kw ];
- real f_S = (dist.f[DIR_0P0 ])[kn ];
- real f_N = (dist.f[DIR_0M0 ])[ks ];
- real f_B = (dist.f[DIR_00P ])[kt ];
- real f_T = (dist.f[DIR_00M ])[kb ];
- real f_SW = (dist.f[DIR_PP0 ])[kne ];
- real f_NE = (dist.f[DIR_MM0 ])[ksw ];
- real f_NW = (dist.f[DIR_PM0 ])[kse ];
- real f_SE = (dist.f[DIR_MP0 ])[knw ];
- real f_BW = (dist.f[DIR_P0P ])[kte ];
- real f_TE = (dist.f[DIR_M0M ])[kbw ];
- real f_TW = (dist.f[DIR_P0M ])[kbe ];
- real f_BE = (dist.f[DIR_M0P ])[ktw ];
- real f_BS = (dist.f[DIR_0PP ])[ktn ];
- real f_TN = (dist.f[DIR_0MM ])[kbs ];
- real f_TS = (dist.f[DIR_0PM ])[kbn ];
- real f_BN = (dist.f[DIR_0MP ])[kts ];
- real f_BSW = (dist.f[DIR_PPP ])[ktne ];
- real f_BNE = (dist.f[DIR_MMP ])[ktsw ];
- real f_BNW = (dist.f[DIR_PMP ])[ktse ];
- real f_BSE = (dist.f[DIR_MPP ])[ktnw ];
- real f_TSW = (dist.f[DIR_PPM ])[kbne ];
- real f_TNE = (dist.f[DIR_MMM ])[kbsw ];
- real f_TNW = (dist.f[DIR_PMM ])[kbse ];
- real f_TSE = (dist.f[DIR_MPM ])[kbnw ];
+ real f_W = (dist.f[DIR_P00])[ke ];
+ real f_E = (dist.f[DIR_M00])[kw ];
+ real f_S = (dist.f[DIR_0P0])[kn ];
+ real f_N = (dist.f[DIR_0M0])[ks ];
+ real f_B = (dist.f[DIR_00P])[kt ];
+ real f_T = (dist.f[DIR_00M])[kb ];
+ real f_SW = (dist.f[DIR_PP0])[kne ];
+ real f_NE = (dist.f[DIR_MM0])[ksw ];
+ real f_NW = (dist.f[DIR_PM0])[kse ];
+ real f_SE = (dist.f[DIR_MP0])[knw ];
+ real f_BW = (dist.f[DIR_P0P])[kte ];
+ real f_TE = (dist.f[DIR_M0M])[kbw ];
+ real f_TW = (dist.f[DIR_P0M])[kbe ];
+ real f_BE = (dist.f[DIR_M0P])[ktw ];
+ real f_BS = (dist.f[DIR_0PP])[ktn ];
+ real f_TN = (dist.f[DIR_0MM])[kbs ];
+ real f_TS = (dist.f[DIR_0PM])[kbn ];
+ real f_BN = (dist.f[DIR_0MP])[kts ];
+ real f_BSW = (dist.f[DIR_PPP])[ktne ];
+ real f_BNE = (dist.f[DIR_MMP])[ktsw ];
+ real f_BNW = (dist.f[DIR_PMP])[ktse ];
+ real f_BSE = (dist.f[DIR_MPP])[ktnw ];
+ real f_TSW = (dist.f[DIR_PPM])[kbne ];
+ real f_TNE = (dist.f[DIR_MMM])[kbsw ];
+ real f_TNW = (dist.f[DIR_PMM])[kbse ];
+ real f_TSE = (dist.f[DIR_MPM])[kbnw ];
////////////////////////////////////////////////////////////////////////////////
//! - Calculate macroscopic quantities
@@ -2309,7 +2397,7 @@ __global__ void QSlipPressureDeviceComp27TurbViscosity(
bool y = false;
bool z = false;
- q = (subgridD.q[DIR_P00])[k];
+ q = (subgridD.q[DIR_P00])[nodeIndex];
if (q>=c0o1 && q<=c1o1) // only update distribution for q between zero and one
{
VeloX = c0o1;
@@ -2321,7 +2409,7 @@ __global__ void QSlipPressureDeviceComp27TurbViscosity(
(dist.f[DIR_M00])[kw] = getInterpolatedDistributionForVeloWithPressureBC(q, f_E, f_W, feq, om_turb, drho, velocityBC, c2o27);
}
- q = (subgridD.q[DIR_M00])[k];
+ q = (subgridD.q[DIR_M00])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloX = c0o1;
@@ -2333,7 +2421,7 @@ __global__ void QSlipPressureDeviceComp27TurbViscosity(
(dist.f[DIR_P00])[ke] = getInterpolatedDistributionForVeloWithPressureBC(q, f_W, f_E, feq, om_turb, drho, velocityBC, c2o27);
}
- q = (subgridD.q[DIR_0P0])[k];
+ q = (subgridD.q[DIR_0P0])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloY = c0o1;
@@ -2345,7 +2433,7 @@ __global__ void QSlipPressureDeviceComp27TurbViscosity(
(dist.f[DIR_0M0])[ks] = getInterpolatedDistributionForVeloWithPressureBC(q, f_N, f_S, feq, om_turb, drho, velocityBC, c2o27);
}
- q = (subgridD.q[DIR_0M0])[k];
+ q = (subgridD.q[DIR_0M0])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloY = c0o1;
@@ -2357,7 +2445,7 @@ __global__ void QSlipPressureDeviceComp27TurbViscosity(
(dist.f[DIR_0P0])[kn] = getInterpolatedDistributionForVeloWithPressureBC(q, f_S, f_N, feq, om_turb, drho, velocityBC, c2o27);
}
- q = (subgridD.q[DIR_00P])[k];
+ q = (subgridD.q[DIR_00P])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloZ = c0o1;
@@ -2369,7 +2457,7 @@ __global__ void QSlipPressureDeviceComp27TurbViscosity(
(dist.f[DIR_00M])[kb] = getInterpolatedDistributionForVeloWithPressureBC(q, f_T, f_B, feq, om_turb, drho, velocityBC, c2o27);
}
- q = (subgridD.q[DIR_00M])[k];
+ q = (subgridD.q[DIR_00M])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloZ = c0o1;
@@ -2381,7 +2469,7 @@ __global__ void QSlipPressureDeviceComp27TurbViscosity(
(dist.f[DIR_00P])[kt] = getInterpolatedDistributionForVeloWithPressureBC(q, f_B, f_T, feq, om_turb, drho, velocityBC, c2o27);
}
- q = (subgridD.q[DIR_PP0])[k];
+ q = (subgridD.q[DIR_PP0])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloX = slipLength*vx1;
@@ -2395,7 +2483,7 @@ __global__ void QSlipPressureDeviceComp27TurbViscosity(
(dist.f[DIR_MM0])[ksw] = getInterpolatedDistributionForVeloWithPressureBC(q, f_NE, f_SW, feq, om_turb, drho, velocityBC, c1o54);
}
- q = (subgridD.q[DIR_MM0])[k];
+ q = (subgridD.q[DIR_MM0])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloX = slipLength*vx1;
@@ -2409,7 +2497,7 @@ __global__ void QSlipPressureDeviceComp27TurbViscosity(
(dist.f[DIR_PP0])[kne] = getInterpolatedDistributionForVeloWithPressureBC(q, f_SW, f_NE, feq, om_turb, drho, velocityBC, c1o54);
}
- q = (subgridD.q[DIR_PM0])[k];
+ q = (subgridD.q[DIR_PM0])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloX = slipLength*vx1;
@@ -2423,7 +2511,7 @@ __global__ void QSlipPressureDeviceComp27TurbViscosity(
(dist.f[DIR_MP0])[knw] = getInterpolatedDistributionForVeloWithPressureBC(q, f_SE, f_NW, feq, om_turb, drho, velocityBC, c1o54);
}
- q = (subgridD.q[DIR_MP0])[k];
+ q = (subgridD.q[DIR_MP0])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloX = slipLength*vx1;
@@ -2437,7 +2525,7 @@ __global__ void QSlipPressureDeviceComp27TurbViscosity(
(dist.f[DIR_PM0])[kse] = getInterpolatedDistributionForVeloWithPressureBC(q, f_NW, f_SE, feq, om_turb, drho, velocityBC, c1o54);
}
- q = (subgridD.q[DIR_P0P])[k];
+ q = (subgridD.q[DIR_P0P])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloX = slipLength*vx1;
@@ -2451,7 +2539,7 @@ __global__ void QSlipPressureDeviceComp27TurbViscosity(
(dist.f[DIR_M0M])[kbw] = getInterpolatedDistributionForVeloWithPressureBC(q, f_TE, f_BW, feq, om_turb, drho, velocityBC, c1o54);
}
- q = (subgridD.q[DIR_M0M])[k];
+ q = (subgridD.q[DIR_M0M])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloX = slipLength*vx1;
@@ -2459,13 +2547,13 @@ __global__ void QSlipPressureDeviceComp27TurbViscosity(
if (x == true) VeloX = c0o1;
if (z == true) VeloZ = c0o1;
- velocityLB = -vx1 - vx3;
- feq = getEquilibriumForBC(drho, velocityLB, cu_sq, c1o54);
- velocityBC = -VeloX - VeloZ;
- (dist.f[DIR_P0P])[kte] = getInterpolatedDistributionForVeloWithPressureBC(q, f_BW, f_TE, feq, om_turb, drho, velocityBC, c1o54);
+ velocityLB = -vx1 - vx3;
+ feq = getEquilibriumForBC(drho, velocityLB, cu_sq, c1o54);
+ velocityBC = -VeloX - VeloZ;
+ (dist.f[DIR_P0P])[kte] = getInterpolatedDistributionForVeloWithPressureBC(q, f_BW, f_TE, feq, om_turb, drho, velocityBC, c1o54);
}
- q = (subgridD.q[DIR_P0M])[k];
+ q = (subgridD.q[DIR_P0M])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloX = slipLength*vx1;
@@ -2479,7 +2567,7 @@ __global__ void QSlipPressureDeviceComp27TurbViscosity(
(dist.f[DIR_M0P])[ktw] = getInterpolatedDistributionForVeloWithPressureBC(q, f_BE, f_TW, feq, om_turb, drho, velocityBC, c1o54);
}
- q = (subgridD.q[DIR_M0P])[k];
+ q = (subgridD.q[DIR_M0P])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloX = slipLength*vx1;
@@ -2493,7 +2581,7 @@ __global__ void QSlipPressureDeviceComp27TurbViscosity(
(dist.f[DIR_P0M])[kbe] = getInterpolatedDistributionForVeloWithPressureBC(q, f_TW, f_BE, feq, om_turb, drho, velocityBC, c1o54);
}
- q = (subgridD.q[DIR_0PP])[k];
+ q = (subgridD.q[DIR_0PP])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloY = slipLength*vx2;
@@ -2507,7 +2595,7 @@ __global__ void QSlipPressureDeviceComp27TurbViscosity(
(dist.f[DIR_0MM])[kbs] = getInterpolatedDistributionForVeloWithPressureBC(q, f_TN, f_BS, feq, om_turb, drho, velocityBC, c1o54);
}
- q = (subgridD.q[DIR_0MM])[k];
+ q = (subgridD.q[DIR_0MM])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloY = slipLength*vx2;
@@ -2522,7 +2610,7 @@ __global__ void QSlipPressureDeviceComp27TurbViscosity(
}
- q = (subgridD.q[DIR_0PM])[k];
+ q = (subgridD.q[DIR_0PM])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloY = slipLength*vx2;
@@ -2536,7 +2624,7 @@ __global__ void QSlipPressureDeviceComp27TurbViscosity(
(dist.f[DIR_0MP])[kts] = getInterpolatedDistributionForVeloWithPressureBC(q, f_BN, f_TS, feq, om_turb, drho, velocityBC, c1o54);
}
- q = (subgridD.q[DIR_0MP])[k];
+ q = (subgridD.q[DIR_0MP])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloY = slipLength*vx2;
@@ -2550,7 +2638,7 @@ __global__ void QSlipPressureDeviceComp27TurbViscosity(
(dist.f[DIR_0PM])[kbn] = getInterpolatedDistributionForVeloWithPressureBC(q, f_TS, f_BN, feq, om_turb, drho, velocityBC, c1o54);
}
- q = (subgridD.q[DIR_PPP])[k];
+ q = (subgridD.q[DIR_PPP])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloX = slipLength*vx1;
@@ -2565,7 +2653,7 @@ __global__ void QSlipPressureDeviceComp27TurbViscosity(
(dist.f[DIR_MMM])[kbsw] = getInterpolatedDistributionForVeloWithPressureBC(q, f_TNE, f_BSW, feq, om_turb, drho, velocityBC, c1o216);
}
- q = (subgridD.q[DIR_MMM])[k];
+ q = (subgridD.q[DIR_MMM])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloX = slipLength*vx1;
@@ -2581,7 +2669,7 @@ __global__ void QSlipPressureDeviceComp27TurbViscosity(
}
- q = (subgridD.q[DIR_PPM])[k];
+ q = (subgridD.q[DIR_PPM])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloX = slipLength*vx1;
@@ -2596,7 +2684,7 @@ __global__ void QSlipPressureDeviceComp27TurbViscosity(
(dist.f[DIR_MMP])[ktsw] = getInterpolatedDistributionForVeloWithPressureBC(q, f_BNE, f_TSW, feq, om_turb, drho, velocityBC, c1o216);
}
- q = (subgridD.q[DIR_MMP])[k];
+ q = (subgridD.q[DIR_MMP])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloX = slipLength*vx1;
@@ -2611,7 +2699,7 @@ __global__ void QSlipPressureDeviceComp27TurbViscosity(
(dist.f[DIR_PPM])[kbne] = getInterpolatedDistributionForVeloWithPressureBC(q, f_TSW, f_BNE, feq, om_turb, drho, velocityBC, c1o216);
}
- q = (subgridD.q[DIR_PMP])[k];
+ q = (subgridD.q[DIR_PMP])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloX = slipLength*vx1;
@@ -2626,7 +2714,7 @@ __global__ void QSlipPressureDeviceComp27TurbViscosity(
(dist.f[DIR_MPM])[kbnw] = getInterpolatedDistributionForVeloWithPressureBC(q, f_TSE, f_BNW, feq, om_turb, drho, velocityBC, c1o216);
}
- q = (subgridD.q[DIR_MPM])[k];
+ q = (subgridD.q[DIR_MPM])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloX = slipLength*vx1;
@@ -2641,7 +2729,7 @@ __global__ void QSlipPressureDeviceComp27TurbViscosity(
(dist.f[DIR_PMP])[ktse] = getInterpolatedDistributionForVeloWithPressureBC(q, f_BNW, f_TSE, feq, om_turb, drho, velocityBC, c1o216);
}
- q = (subgridD.q[DIR_PMM])[k];
+ q = (subgridD.q[DIR_PMM])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloX = slipLength*vx1;
@@ -2656,7 +2744,7 @@ __global__ void QSlipPressureDeviceComp27TurbViscosity(
(dist.f[DIR_MPP])[ktnw] = getInterpolatedDistributionForVeloWithPressureBC(q, f_BSE, f_TNW, feq, om_turb, drho, velocityBC, c1o216);
}
- q = (subgridD.q[DIR_MPP])[k];
+ q = (subgridD.q[DIR_MPP])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
VeloX = slipLength*vx1;
@@ -2688,63 +2776,63 @@ __global__ void QSlipPressureDeviceComp27TurbViscosity(
// Distributions27 D;
// if (isEvenTimestep==true)
// {
-// D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
-// D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
-// D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
-// D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
-// D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
-// D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
-// D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
-// D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
-// D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
-// D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
-// D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
-// D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
-// D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
-// D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
-// D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
-// D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
-// D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
-// D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
-// D.f[DIR_000] = &DD[DIR_000*size_Mat];
-// D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
-// D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
-// D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
-// D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
-// D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
-// D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
-// D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
-// D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
+// D.f[DIR_P00] = &DD[DIR_P00 * size_Mat];
+// D.f[DIR_M00] = &DD[DIR_M00 * size_Mat];
+// D.f[DIR_0P0] = &DD[DIR_0P0 * size_Mat];
+// D.f[DIR_0M0] = &DD[DIR_0M0 * size_Mat];
+// D.f[DIR_00P] = &DD[DIR_00P * size_Mat];
+// D.f[DIR_00M] = &DD[DIR_00M * size_Mat];
+// D.f[DIR_PP0] = &DD[DIR_PP0 * size_Mat];
+// D.f[DIR_MM0] = &DD[DIR_MM0 * size_Mat];
+// D.f[DIR_PM0] = &DD[DIR_PM0 * size_Mat];
+// D.f[DIR_MP0] = &DD[DIR_MP0 * size_Mat];
+// D.f[DIR_P0P] = &DD[DIR_P0P * size_Mat];
+// D.f[DIR_M0M] = &DD[DIR_M0M * size_Mat];
+// D.f[DIR_P0M] = &DD[DIR_P0M * size_Mat];
+// D.f[DIR_M0P] = &DD[DIR_M0P * size_Mat];
+// D.f[DIR_0PP] = &DD[DIR_0PP * size_Mat];
+// D.f[DIR_0MM] = &DD[DIR_0MM * size_Mat];
+// D.f[DIR_0PM] = &DD[DIR_0PM * size_Mat];
+// D.f[DIR_0MP] = &DD[DIR_0MP * size_Mat];
+// D.f[DIR_000] = &DD[DIR_000 * size_Mat];
+// D.f[DIR_PPP] = &DD[DIR_PPP * size_Mat];
+// D.f[DIR_MMP] = &DD[DIR_MMP * size_Mat];
+// D.f[DIR_PMP] = &DD[DIR_PMP * size_Mat];
+// D.f[DIR_MPP] = &DD[DIR_MPP * size_Mat];
+// D.f[DIR_PPM] = &DD[DIR_PPM * size_Mat];
+// D.f[DIR_MMM] = &DD[DIR_MMM * size_Mat];
+// D.f[DIR_PMM] = &DD[DIR_PMM * size_Mat];
+// D.f[DIR_MPM] = &DD[DIR_MPM * size_Mat];
// }
// else
// {
-// D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
-// D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
-// D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
-// D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
-// D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
-// D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
-// D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
-// D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
-// D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
-// D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
-// D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
-// D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
-// D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
-// D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
-// D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
-// D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
-// D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
-// D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
-// D.f[DIR_000] = &DD[DIR_000*size_Mat];
-// D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
-// D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
-// D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
-// D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
-// D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
-// D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
-// D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
-// D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
+// D.f[DIR_M00] = &DD[DIR_P00 * size_Mat];
+// D.f[DIR_P00] = &DD[DIR_M00 * size_Mat];
+// D.f[DIR_0M0] = &DD[DIR_0P0 * size_Mat];
+// D.f[DIR_0P0] = &DD[DIR_0M0 * size_Mat];
+// D.f[DIR_00M] = &DD[DIR_00P * size_Mat];
+// D.f[DIR_00P] = &DD[DIR_00M * size_Mat];
+// D.f[DIR_MM0] = &DD[DIR_PP0 * size_Mat];
+// D.f[DIR_PP0] = &DD[DIR_MM0 * size_Mat];
+// D.f[DIR_MP0] = &DD[DIR_PM0 * size_Mat];
+// D.f[DIR_PM0] = &DD[DIR_MP0 * size_Mat];
+// D.f[DIR_M0M] = &DD[DIR_P0P * size_Mat];
+// D.f[DIR_P0P] = &DD[DIR_M0M * size_Mat];
+// D.f[DIR_M0P] = &DD[DIR_P0M * size_Mat];
+// D.f[DIR_P0M] = &DD[DIR_M0P * size_Mat];
+// D.f[DIR_0MM] = &DD[DIR_0PP * size_Mat];
+// D.f[DIR_0PP] = &DD[DIR_0MM * size_Mat];
+// D.f[DIR_0MP] = &DD[DIR_0PM * size_Mat];
+// D.f[DIR_0PM] = &DD[DIR_0MP * size_Mat];
+// D.f[DIR_000] = &DD[DIR_000 * size_Mat];
+// D.f[DIR_PPP] = &DD[DIR_MMM * size_Mat];
+// D.f[DIR_MMP] = &DD[DIR_PPM * size_Mat];
+// D.f[DIR_PMP] = &DD[DIR_MPM * size_Mat];
+// D.f[DIR_MPP] = &DD[DIR_PMM * size_Mat];
+// D.f[DIR_PPM] = &DD[DIR_MMP * size_Mat];
+// D.f[DIR_MMM] = &DD[DIR_PPP * size_Mat];
+// D.f[DIR_PMM] = &DD[DIR_MPP * size_Mat];
+// D.f[DIR_MPM] = &DD[DIR_PMP * size_Mat];
// }
// ////////////////////////////////////////////////////////////////////////////////
// const unsigned x = threadIdx.x; // Globaler x-Index
@@ -2765,24 +2853,24 @@ __global__ void QSlipPressureDeviceComp27TurbViscosity(
// *q_dirBE, *q_dirTW, *q_dirTN, *q_dirBS, *q_dirBN, *q_dirTS,
// *q_dirTNE, *q_dirTSW, *q_dirTSE, *q_dirTNW, *q_dirBNE, *q_dirBSW,
// *q_dirBSE, *q_dirBNW;
-// q_dirE = &QQ[DIR_P00 * numberOfBCnodes];
-// q_dirW = &QQ[DIR_M00 * numberOfBCnodes];
-// q_dirN = &QQ[DIR_0P0 * numberOfBCnodes];
-// q_dirS = &QQ[DIR_0M0 * numberOfBCnodes];
-// q_dirT = &QQ[DIR_00P * numberOfBCnodes];
-// q_dirB = &QQ[DIR_00M * numberOfBCnodes];
-// q_dirNE = &QQ[DIR_PP0 * numberOfBCnodes];
-// q_dirSW = &QQ[DIR_MM0 * numberOfBCnodes];
-// q_dirSE = &QQ[DIR_PM0 * numberOfBCnodes];
-// q_dirNW = &QQ[DIR_MP0 * numberOfBCnodes];
-// q_dirTE = &QQ[DIR_P0P * numberOfBCnodes];
-// q_dirBW = &QQ[DIR_M0M * numberOfBCnodes];
-// q_dirBE = &QQ[DIR_P0M * numberOfBCnodes];
-// q_dirTW = &QQ[DIR_M0P * numberOfBCnodes];
-// q_dirTN = &QQ[DIR_0PP * numberOfBCnodes];
-// q_dirBS = &QQ[DIR_0MM * numberOfBCnodes];
-// q_dirBN = &QQ[DIR_0PM * numberOfBCnodes];
-// q_dirTS = &QQ[DIR_0MP * numberOfBCnodes];
+// q_dirE = &QQ[DIR_P00 * numberOfBCnodes];
+// q_dirW = &QQ[DIR_M00 * numberOfBCnodes];
+// q_dirN = &QQ[DIR_0P0 * numberOfBCnodes];
+// q_dirS = &QQ[DIR_0M0 * numberOfBCnodes];
+// q_dirT = &QQ[DIR_00P * numberOfBCnodes];
+// q_dirB = &QQ[DIR_00M * numberOfBCnodes];
+// q_dirNE = &QQ[DIR_PP0 * numberOfBCnodes];
+// q_dirSW = &QQ[DIR_MM0 * numberOfBCnodes];
+// q_dirSE = &QQ[DIR_PM0 * numberOfBCnodes];
+// q_dirNW = &QQ[DIR_MP0 * numberOfBCnodes];
+// q_dirTE = &QQ[DIR_P0P * numberOfBCnodes];
+// q_dirBW = &QQ[DIR_M0M * numberOfBCnodes];
+// q_dirBE = &QQ[DIR_P0M * numberOfBCnodes];
+// q_dirTW = &QQ[DIR_M0P * numberOfBCnodes];
+// q_dirTN = &QQ[DIR_0PP * numberOfBCnodes];
+// q_dirBS = &QQ[DIR_0MM * numberOfBCnodes];
+// q_dirBN = &QQ[DIR_0PM * numberOfBCnodes];
+// q_dirTS = &QQ[DIR_0MP * numberOfBCnodes];
// q_dirTNE = &QQ[DIR_PPP * numberOfBCnodes];
// q_dirTSW = &QQ[DIR_MMP * numberOfBCnodes];
// q_dirTSE = &QQ[DIR_PMP * numberOfBCnodes];
@@ -2823,32 +2911,32 @@ __global__ void QSlipPressureDeviceComp27TurbViscosity(
// unsigned int kbsw = neighborZ[ksw];
// ////////////////////////////////////////////////////////////////////////////////
-// real f_W = (D.f[DIR_P00 ])[ke ];
-// real f_E = (D.f[DIR_M00 ])[kw ];
-// real f_S = (D.f[DIR_0P0 ])[kn ];
-// real f_N = (D.f[DIR_0M0 ])[ks ];
-// real f_B = (D.f[DIR_00P ])[kt ];
-// real f_T = (D.f[DIR_00M ])[kb ];
-// real f_SW = (D.f[DIR_PP0 ])[kne ];
-// real f_NE = (D.f[DIR_MM0 ])[ksw ];
-// real f_NW = (D.f[DIR_PM0 ])[kse ];
-// real f_SE = (D.f[DIR_MP0 ])[knw ];
-// real f_BW = (D.f[DIR_P0P ])[kte ];
-// real f_TE = (D.f[DIR_M0M ])[kbw ];
-// real f_TW = (D.f[DIR_P0M ])[kbe ];
-// real f_BE = (D.f[DIR_M0P ])[ktw ];
-// real f_BS = (D.f[DIR_0PP ])[ktn ];
-// real f_TN = (D.f[DIR_0MM ])[kbs ];
-// real f_TS = (D.f[DIR_0PM ])[kbn ];
-// real f_BN = (D.f[DIR_0MP ])[kts ];
-// real f_BSW = (D.f[DIR_PPP ])[ktne ];
-// real f_BNE = (D.f[DIR_MMP ])[ktsw ];
-// real f_BNW = (D.f[DIR_PMP ])[ktse ];
-// real f_BSE = (D.f[DIR_MPP ])[ktnw ];
-// real f_TSW = (D.f[DIR_PPM ])[kbne ];
-// real f_TNE = (D.f[DIR_MMM ])[kbsw ];
-// real f_TNW = (D.f[DIR_PMM ])[kbse ];
-// real f_TSE = (D.f[DIR_MPM ])[kbnw ];
+// real f_W = (D.f[DIR_P00])[ke ];
+// real f_E = (D.f[DIR_M00])[kw ];
+// real f_S = (D.f[DIR_0P0])[kn ];
+// real f_N = (D.f[DIR_0M0])[ks ];
+// real f_B = (D.f[DIR_00P])[kt ];
+// real f_T = (D.f[DIR_00M])[kb ];
+// real f_SW = (D.f[DIR_PP0])[kne ];
+// real f_NE = (D.f[DIR_MM0])[ksw ];
+// real f_NW = (D.f[DIR_PM0])[kse ];
+// real f_SE = (D.f[DIR_MP0])[knw ];
+// real f_BW = (D.f[DIR_P0P])[kte ];
+// real f_TE = (D.f[DIR_M0M])[kbw ];
+// real f_TW = (D.f[DIR_P0M])[kbe ];
+// real f_BE = (D.f[DIR_M0P])[ktw ];
+// real f_BS = (D.f[DIR_0PP])[ktn ];
+// real f_TN = (D.f[DIR_0MM])[kbs ];
+// real f_TS = (D.f[DIR_0PM])[kbn ];
+// real f_BN = (D.f[DIR_0MP])[kts ];
+// real f_BSW = (D.f[DIR_PPP])[ktne ];
+// real f_BNE = (D.f[DIR_MMP])[ktsw ];
+// real f_BNW = (D.f[DIR_PMP])[ktse ];
+// real f_BSE = (D.f[DIR_MPP])[ktnw ];
+// real f_TSW = (D.f[DIR_PPM])[kbne ];
+// real f_TNE = (D.f[DIR_MMM])[kbsw ];
+// real f_TNW = (D.f[DIR_PMM])[kbse ];
+// real f_TSE = (D.f[DIR_MPM])[kbnw ];
// ////////////////////////////////////////////////////////////////////////////////
// real vx1, vx2, vx3, drho, feq, q;
// drho = f_TSE + f_TNW + f_TNE + f_TSW + f_BSE + f_BNW + f_BNE + f_BSW +
@@ -2873,63 +2961,63 @@ __global__ void QSlipPressureDeviceComp27TurbViscosity(
// //////////////////////////////////////////////////////////////////////////
// if (isEvenTimestep==false)
// {
-// D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
-// D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
-// D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
-// D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
-// D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
-// D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
-// D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
-// D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
-// D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
-// D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
-// D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
-// D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
-// D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
-// D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
-// D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
-// D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
-// D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
-// D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
-// D.f[DIR_000] = &DD[DIR_000*size_Mat];
-// D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
-// D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
-// D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
-// D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
-// D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
-// D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
-// D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
-// D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
+// D.f[DIR_P00] = &DD[DIR_P00 * size_Mat];
+// D.f[DIR_M00] = &DD[DIR_M00 * size_Mat];
+// D.f[DIR_0P0] = &DD[DIR_0P0 * size_Mat];
+// D.f[DIR_0M0] = &DD[DIR_0M0 * size_Mat];
+// D.f[DIR_00P] = &DD[DIR_00P * size_Mat];
+// D.f[DIR_00M] = &DD[DIR_00M * size_Mat];
+// D.f[DIR_PP0] = &DD[DIR_PP0 * size_Mat];
+// D.f[DIR_MM0] = &DD[DIR_MM0 * size_Mat];
+// D.f[DIR_PM0] = &DD[DIR_PM0 * size_Mat];
+// D.f[DIR_MP0] = &DD[DIR_MP0 * size_Mat];
+// D.f[DIR_P0P] = &DD[DIR_P0P * size_Mat];
+// D.f[DIR_M0M] = &DD[DIR_M0M * size_Mat];
+// D.f[DIR_P0M] = &DD[DIR_P0M * size_Mat];
+// D.f[DIR_M0P] = &DD[DIR_M0P * size_Mat];
+// D.f[DIR_0PP] = &DD[DIR_0PP * size_Mat];
+// D.f[DIR_0MM] = &DD[DIR_0MM * size_Mat];
+// D.f[DIR_0PM] = &DD[DIR_0PM * size_Mat];
+// D.f[DIR_0MP] = &DD[DIR_0MP * size_Mat];
+// D.f[DIR_000] = &DD[DIR_000 * size_Mat];
+// D.f[DIR_PPP] = &DD[DIR_PPP * size_Mat];
+// D.f[DIR_MMP] = &DD[DIR_MMP * size_Mat];
+// D.f[DIR_PMP] = &DD[DIR_PMP * size_Mat];
+// D.f[DIR_MPP] = &DD[DIR_MPP * size_Mat];
+// D.f[DIR_PPM] = &DD[DIR_PPM * size_Mat];
+// D.f[DIR_MMM] = &DD[DIR_MMM * size_Mat];
+// D.f[DIR_PMM] = &DD[DIR_PMM * size_Mat];
+// D.f[DIR_MPM] = &DD[DIR_MPM * size_Mat];
// }
// else
// {
-// D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
-// D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
-// D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
-// D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
-// D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
-// D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
-// D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
-// D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
-// D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
-// D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
-// D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
-// D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
-// D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
-// D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
-// D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
-// D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
-// D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
-// D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
-// D.f[DIR_000] = &DD[DIR_000*size_Mat];
-// D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
-// D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
-// D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
-// D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
-// D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
-// D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
-// D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
-// D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
+// D.f[DIR_M00] = &DD[DIR_P00 * size_Mat];
+// D.f[DIR_P00] = &DD[DIR_M00 * size_Mat];
+// D.f[DIR_0M0] = &DD[DIR_0P0 * size_Mat];
+// D.f[DIR_0P0] = &DD[DIR_0M0 * size_Mat];
+// D.f[DIR_00M] = &DD[DIR_00P * size_Mat];
+// D.f[DIR_00P] = &DD[DIR_00M * size_Mat];
+// D.f[DIR_MM0] = &DD[DIR_PP0 * size_Mat];
+// D.f[DIR_PP0] = &DD[DIR_MM0 * size_Mat];
+// D.f[DIR_MP0] = &DD[DIR_PM0 * size_Mat];
+// D.f[DIR_PM0] = &DD[DIR_MP0 * size_Mat];
+// D.f[DIR_M0M] = &DD[DIR_P0P * size_Mat];
+// D.f[DIR_P0P] = &DD[DIR_M0M * size_Mat];
+// D.f[DIR_M0P] = &DD[DIR_P0M * size_Mat];
+// D.f[DIR_P0M] = &DD[DIR_M0P * size_Mat];
+// D.f[DIR_0MM] = &DD[DIR_0PP * size_Mat];
+// D.f[DIR_0PP] = &DD[DIR_0MM * size_Mat];
+// D.f[DIR_0MP] = &DD[DIR_0PM * size_Mat];
+// D.f[DIR_0PM] = &DD[DIR_0MP * size_Mat];
+// D.f[DIR_000] = &DD[DIR_000 * size_Mat];
+// D.f[DIR_PPP] = &DD[DIR_MMM * size_Mat];
+// D.f[DIR_MMP] = &DD[DIR_PPM * size_Mat];
+// D.f[DIR_PMP] = &DD[DIR_MPM * size_Mat];
+// D.f[DIR_MPP] = &DD[DIR_PMM * size_Mat];
+// D.f[DIR_PPM] = &DD[DIR_MMP * size_Mat];
+// D.f[DIR_MMM] = &DD[DIR_PPP * size_Mat];
+// D.f[DIR_PMM] = &DD[DIR_MPP * size_Mat];
+// D.f[DIR_MPM] = &DD[DIR_PMP * size_Mat];
// }
// ////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
// //Test
@@ -3378,80 +3466,81 @@ __global__ void QSlipPressureDeviceComp27TurbViscosity(
//////////////////////////////////////////////////////////////////////////////
-__global__ void QSlipGeomDeviceComp27(real* DD,
- int* k_Q,
- real* QQ,
- unsigned int numberOfBCnodes,
- real om1,
- real* NormalX,
- real* NormalY,
- real* NormalZ,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep)
+__global__ void QSlipGeomDeviceComp27(
+ real* DD,
+ int* k_Q,
+ real* QQ,
+ unsigned int numberOfBCnodes,
+ real om1,
+ real* NormalX,
+ real* NormalY,
+ real* NormalZ,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep)
{
Distributions27 D;
if (isEvenTimestep==true)
{
- D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
+ D.f[DIR_M00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_MPM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_PMP * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////
const unsigned x = threadIdx.x; // Globaler x-Index
@@ -3472,24 +3561,24 @@ __global__ void QSlipGeomDeviceComp27(real* DD,
*q_dirBE, *q_dirTW, *q_dirTN, *q_dirBS, *q_dirBN, *q_dirTS,
*q_dirTNE, *q_dirTSW, *q_dirTSE, *q_dirTNW, *q_dirBNE, *q_dirBSW,
*q_dirBSE, *q_dirBNW;
- q_dirE = &QQ[DIR_P00 * numberOfBCnodes];
- q_dirW = &QQ[DIR_M00 * numberOfBCnodes];
- q_dirN = &QQ[DIR_0P0 * numberOfBCnodes];
- q_dirS = &QQ[DIR_0M0 * numberOfBCnodes];
- q_dirT = &QQ[DIR_00P * numberOfBCnodes];
- q_dirB = &QQ[DIR_00M * numberOfBCnodes];
- q_dirNE = &QQ[DIR_PP0 * numberOfBCnodes];
- q_dirSW = &QQ[DIR_MM0 * numberOfBCnodes];
- q_dirSE = &QQ[DIR_PM0 * numberOfBCnodes];
- q_dirNW = &QQ[DIR_MP0 * numberOfBCnodes];
- q_dirTE = &QQ[DIR_P0P * numberOfBCnodes];
- q_dirBW = &QQ[DIR_M0M * numberOfBCnodes];
- q_dirBE = &QQ[DIR_P0M * numberOfBCnodes];
- q_dirTW = &QQ[DIR_M0P * numberOfBCnodes];
- q_dirTN = &QQ[DIR_0PP * numberOfBCnodes];
- q_dirBS = &QQ[DIR_0MM * numberOfBCnodes];
- q_dirBN = &QQ[DIR_0PM * numberOfBCnodes];
- q_dirTS = &QQ[DIR_0MP * numberOfBCnodes];
+ q_dirE = &QQ[DIR_P00 * numberOfBCnodes];
+ q_dirW = &QQ[DIR_M00 * numberOfBCnodes];
+ q_dirN = &QQ[DIR_0P0 * numberOfBCnodes];
+ q_dirS = &QQ[DIR_0M0 * numberOfBCnodes];
+ q_dirT = &QQ[DIR_00P * numberOfBCnodes];
+ q_dirB = &QQ[DIR_00M * numberOfBCnodes];
+ q_dirNE = &QQ[DIR_PP0 * numberOfBCnodes];
+ q_dirSW = &QQ[DIR_MM0 * numberOfBCnodes];
+ q_dirSE = &QQ[DIR_PM0 * numberOfBCnodes];
+ q_dirNW = &QQ[DIR_MP0 * numberOfBCnodes];
+ q_dirTE = &QQ[DIR_P0P * numberOfBCnodes];
+ q_dirBW = &QQ[DIR_M0M * numberOfBCnodes];
+ q_dirBE = &QQ[DIR_P0M * numberOfBCnodes];
+ q_dirTW = &QQ[DIR_M0P * numberOfBCnodes];
+ q_dirTN = &QQ[DIR_0PP * numberOfBCnodes];
+ q_dirBS = &QQ[DIR_0MM * numberOfBCnodes];
+ q_dirBN = &QQ[DIR_0PM * numberOfBCnodes];
+ q_dirTS = &QQ[DIR_0MP * numberOfBCnodes];
q_dirTNE = &QQ[DIR_PPP * numberOfBCnodes];
q_dirTSW = &QQ[DIR_MMP * numberOfBCnodes];
q_dirTSE = &QQ[DIR_PMP * numberOfBCnodes];
@@ -3504,24 +3593,24 @@ __global__ void QSlipGeomDeviceComp27(real* DD,
*nx_dirBE, *nx_dirTW, *nx_dirTN, *nx_dirBS, *nx_dirBN, *nx_dirTS,
*nx_dirTNE, *nx_dirTSW, *nx_dirTSE, *nx_dirTNW, *nx_dirBNE, *nx_dirBSW,
*nx_dirBSE, *nx_dirBNW;
- nx_dirE = &NormalX[DIR_P00 * numberOfBCnodes];
- nx_dirW = &NormalX[DIR_M00 * numberOfBCnodes];
- nx_dirN = &NormalX[DIR_0P0 * numberOfBCnodes];
- nx_dirS = &NormalX[DIR_0M0 * numberOfBCnodes];
- nx_dirT = &NormalX[DIR_00P * numberOfBCnodes];
- nx_dirB = &NormalX[DIR_00M * numberOfBCnodes];
- nx_dirNE = &NormalX[DIR_PP0 * numberOfBCnodes];
- nx_dirSW = &NormalX[DIR_MM0 * numberOfBCnodes];
- nx_dirSE = &NormalX[DIR_PM0 * numberOfBCnodes];
- nx_dirNW = &NormalX[DIR_MP0 * numberOfBCnodes];
- nx_dirTE = &NormalX[DIR_P0P * numberOfBCnodes];
- nx_dirBW = &NormalX[DIR_M0M * numberOfBCnodes];
- nx_dirBE = &NormalX[DIR_P0M * numberOfBCnodes];
- nx_dirTW = &NormalX[DIR_M0P * numberOfBCnodes];
- nx_dirTN = &NormalX[DIR_0PP * numberOfBCnodes];
- nx_dirBS = &NormalX[DIR_0MM * numberOfBCnodes];
- nx_dirBN = &NormalX[DIR_0PM * numberOfBCnodes];
- nx_dirTS = &NormalX[DIR_0MP * numberOfBCnodes];
+ nx_dirE = &NormalX[DIR_P00 * numberOfBCnodes];
+ nx_dirW = &NormalX[DIR_M00 * numberOfBCnodes];
+ nx_dirN = &NormalX[DIR_0P0 * numberOfBCnodes];
+ nx_dirS = &NormalX[DIR_0M0 * numberOfBCnodes];
+ nx_dirT = &NormalX[DIR_00P * numberOfBCnodes];
+ nx_dirB = &NormalX[DIR_00M * numberOfBCnodes];
+ nx_dirNE = &NormalX[DIR_PP0 * numberOfBCnodes];
+ nx_dirSW = &NormalX[DIR_MM0 * numberOfBCnodes];
+ nx_dirSE = &NormalX[DIR_PM0 * numberOfBCnodes];
+ nx_dirNW = &NormalX[DIR_MP0 * numberOfBCnodes];
+ nx_dirTE = &NormalX[DIR_P0P * numberOfBCnodes];
+ nx_dirBW = &NormalX[DIR_M0M * numberOfBCnodes];
+ nx_dirBE = &NormalX[DIR_P0M * numberOfBCnodes];
+ nx_dirTW = &NormalX[DIR_M0P * numberOfBCnodes];
+ nx_dirTN = &NormalX[DIR_0PP * numberOfBCnodes];
+ nx_dirBS = &NormalX[DIR_0MM * numberOfBCnodes];
+ nx_dirBN = &NormalX[DIR_0PM * numberOfBCnodes];
+ nx_dirTS = &NormalX[DIR_0MP * numberOfBCnodes];
nx_dirTNE = &NormalX[DIR_PPP * numberOfBCnodes];
nx_dirTSW = &NormalX[DIR_MMP * numberOfBCnodes];
nx_dirTSE = &NormalX[DIR_PMP * numberOfBCnodes];
@@ -3536,24 +3625,24 @@ __global__ void QSlipGeomDeviceComp27(real* DD,
*ny_dirBE, *ny_dirTW, *ny_dirTN, *ny_dirBS, *ny_dirBN, *ny_dirTS,
*ny_dirTNE, *ny_dirTSW, *ny_dirTSE, *ny_dirTNW, *ny_dirBNE, *ny_dirBSW,
*ny_dirBSE, *ny_dirBNW;
- ny_dirE = &NormalY[DIR_P00 * numberOfBCnodes];
- ny_dirW = &NormalY[DIR_M00 * numberOfBCnodes];
- ny_dirN = &NormalY[DIR_0P0 * numberOfBCnodes];
- ny_dirS = &NormalY[DIR_0M0 * numberOfBCnodes];
- ny_dirT = &NormalY[DIR_00P * numberOfBCnodes];
- ny_dirB = &NormalY[DIR_00M * numberOfBCnodes];
- ny_dirNE = &NormalY[DIR_PP0 * numberOfBCnodes];
- ny_dirSW = &NormalY[DIR_MM0 * numberOfBCnodes];
- ny_dirSE = &NormalY[DIR_PM0 * numberOfBCnodes];
- ny_dirNW = &NormalY[DIR_MP0 * numberOfBCnodes];
- ny_dirTE = &NormalY[DIR_P0P * numberOfBCnodes];
- ny_dirBW = &NormalY[DIR_M0M * numberOfBCnodes];
- ny_dirBE = &NormalY[DIR_P0M * numberOfBCnodes];
- ny_dirTW = &NormalY[DIR_M0P * numberOfBCnodes];
- ny_dirTN = &NormalY[DIR_0PP * numberOfBCnodes];
- ny_dirBS = &NormalY[DIR_0MM * numberOfBCnodes];
- ny_dirBN = &NormalY[DIR_0PM * numberOfBCnodes];
- ny_dirTS = &NormalY[DIR_0MP * numberOfBCnodes];
+ ny_dirE = &NormalY[DIR_P00 * numberOfBCnodes];
+ ny_dirW = &NormalY[DIR_M00 * numberOfBCnodes];
+ ny_dirN = &NormalY[DIR_0P0 * numberOfBCnodes];
+ ny_dirS = &NormalY[DIR_0M0 * numberOfBCnodes];
+ ny_dirT = &NormalY[DIR_00P * numberOfBCnodes];
+ ny_dirB = &NormalY[DIR_00M * numberOfBCnodes];
+ ny_dirNE = &NormalY[DIR_PP0 * numberOfBCnodes];
+ ny_dirSW = &NormalY[DIR_MM0 * numberOfBCnodes];
+ ny_dirSE = &NormalY[DIR_PM0 * numberOfBCnodes];
+ ny_dirNW = &NormalY[DIR_MP0 * numberOfBCnodes];
+ ny_dirTE = &NormalY[DIR_P0P * numberOfBCnodes];
+ ny_dirBW = &NormalY[DIR_M0M * numberOfBCnodes];
+ ny_dirBE = &NormalY[DIR_P0M * numberOfBCnodes];
+ ny_dirTW = &NormalY[DIR_M0P * numberOfBCnodes];
+ ny_dirTN = &NormalY[DIR_0PP * numberOfBCnodes];
+ ny_dirBS = &NormalY[DIR_0MM * numberOfBCnodes];
+ ny_dirBN = &NormalY[DIR_0PM * numberOfBCnodes];
+ ny_dirTS = &NormalY[DIR_0MP * numberOfBCnodes];
ny_dirTNE = &NormalY[DIR_PPP * numberOfBCnodes];
ny_dirTSW = &NormalY[DIR_MMP * numberOfBCnodes];
ny_dirTSE = &NormalY[DIR_PMP * numberOfBCnodes];
@@ -3568,24 +3657,24 @@ __global__ void QSlipGeomDeviceComp27(real* DD,
*nz_dirBE, *nz_dirTW, *nz_dirTN, *nz_dirBS, *nz_dirBN, *nz_dirTS,
*nz_dirTNE, *nz_dirTSW, *nz_dirTSE, *nz_dirTNW, *nz_dirBNE, *nz_dirBSW,
*nz_dirBSE, *nz_dirBNW;
- nz_dirE = &NormalZ[DIR_P00 * numberOfBCnodes];
- nz_dirW = &NormalZ[DIR_M00 * numberOfBCnodes];
- nz_dirN = &NormalZ[DIR_0P0 * numberOfBCnodes];
- nz_dirS = &NormalZ[DIR_0M0 * numberOfBCnodes];
- nz_dirT = &NormalZ[DIR_00P * numberOfBCnodes];
- nz_dirB = &NormalZ[DIR_00M * numberOfBCnodes];
- nz_dirNE = &NormalZ[DIR_PP0 * numberOfBCnodes];
- nz_dirSW = &NormalZ[DIR_MM0 * numberOfBCnodes];
- nz_dirSE = &NormalZ[DIR_PM0 * numberOfBCnodes];
- nz_dirNW = &NormalZ[DIR_MP0 * numberOfBCnodes];
- nz_dirTE = &NormalZ[DIR_P0P * numberOfBCnodes];
- nz_dirBW = &NormalZ[DIR_M0M * numberOfBCnodes];
- nz_dirBE = &NormalZ[DIR_P0M * numberOfBCnodes];
- nz_dirTW = &NormalZ[DIR_M0P * numberOfBCnodes];
- nz_dirTN = &NormalZ[DIR_0PP * numberOfBCnodes];
- nz_dirBS = &NormalZ[DIR_0MM * numberOfBCnodes];
- nz_dirBN = &NormalZ[DIR_0PM * numberOfBCnodes];
- nz_dirTS = &NormalZ[DIR_0MP * numberOfBCnodes];
+ nz_dirE = &NormalZ[DIR_P00 * numberOfBCnodes];
+ nz_dirW = &NormalZ[DIR_M00 * numberOfBCnodes];
+ nz_dirN = &NormalZ[DIR_0P0 * numberOfBCnodes];
+ nz_dirS = &NormalZ[DIR_0M0 * numberOfBCnodes];
+ nz_dirT = &NormalZ[DIR_00P * numberOfBCnodes];
+ nz_dirB = &NormalZ[DIR_00M * numberOfBCnodes];
+ nz_dirNE = &NormalZ[DIR_PP0 * numberOfBCnodes];
+ nz_dirSW = &NormalZ[DIR_MM0 * numberOfBCnodes];
+ nz_dirSE = &NormalZ[DIR_PM0 * numberOfBCnodes];
+ nz_dirNW = &NormalZ[DIR_MP0 * numberOfBCnodes];
+ nz_dirTE = &NormalZ[DIR_P0P * numberOfBCnodes];
+ nz_dirBW = &NormalZ[DIR_M0M * numberOfBCnodes];
+ nz_dirBE = &NormalZ[DIR_P0M * numberOfBCnodes];
+ nz_dirTW = &NormalZ[DIR_M0P * numberOfBCnodes];
+ nz_dirTN = &NormalZ[DIR_0PP * numberOfBCnodes];
+ nz_dirBS = &NormalZ[DIR_0MM * numberOfBCnodes];
+ nz_dirBN = &NormalZ[DIR_0PM * numberOfBCnodes];
+ nz_dirTS = &NormalZ[DIR_0MP * numberOfBCnodes];
nz_dirTNE = &NormalZ[DIR_PPP * numberOfBCnodes];
nz_dirTSW = &NormalZ[DIR_MMP * numberOfBCnodes];
nz_dirTSE = &NormalZ[DIR_PMP * numberOfBCnodes];
@@ -3625,32 +3714,32 @@ __global__ void QSlipGeomDeviceComp27(real* DD,
unsigned int ktne = KQK;
unsigned int kbsw = neighborZ[ksw];
////////////////////////////////////////////////////////////////////////////////
- real f_W = (D.f[DIR_P00 ])[ke ];
- real f_E = (D.f[DIR_M00 ])[kw ];
- real f_S = (D.f[DIR_0P0 ])[kn ];
- real f_N = (D.f[DIR_0M0 ])[ks ];
- real f_B = (D.f[DIR_00P ])[kt ];
- real f_T = (D.f[DIR_00M ])[kb ];
- real f_SW = (D.f[DIR_PP0 ])[kne ];
- real f_NE = (D.f[DIR_MM0 ])[ksw ];
- real f_NW = (D.f[DIR_PM0 ])[kse ];
- real f_SE = (D.f[DIR_MP0 ])[knw ];
- real f_BW = (D.f[DIR_P0P ])[kte ];
- real f_TE = (D.f[DIR_M0M ])[kbw ];
- real f_TW = (D.f[DIR_P0M ])[kbe ];
- real f_BE = (D.f[DIR_M0P ])[ktw ];
- real f_BS = (D.f[DIR_0PP ])[ktn ];
- real f_TN = (D.f[DIR_0MM ])[kbs ];
- real f_TS = (D.f[DIR_0PM ])[kbn ];
- real f_BN = (D.f[DIR_0MP ])[kts ];
- real f_BSW = (D.f[DIR_PPP ])[ktne ];
- real f_BNE = (D.f[DIR_MMP ])[ktsw ];
- real f_BNW = (D.f[DIR_PMP ])[ktse ];
- real f_BSE = (D.f[DIR_MPP ])[ktnw ];
- real f_TSW = (D.f[DIR_PPM ])[kbne ];
- real f_TNE = (D.f[DIR_MMM ])[kbsw ];
- real f_TNW = (D.f[DIR_PMM ])[kbse ];
- real f_TSE = (D.f[DIR_MPM ])[kbnw ];
+ real f_W = (D.f[DIR_P00])[ke ];
+ real f_E = (D.f[DIR_M00])[kw ];
+ real f_S = (D.f[DIR_0P0])[kn ];
+ real f_N = (D.f[DIR_0M0])[ks ];
+ real f_B = (D.f[DIR_00P])[kt ];
+ real f_T = (D.f[DIR_00M])[kb ];
+ real f_SW = (D.f[DIR_PP0])[kne ];
+ real f_NE = (D.f[DIR_MM0])[ksw ];
+ real f_NW = (D.f[DIR_PM0])[kse ];
+ real f_SE = (D.f[DIR_MP0])[knw ];
+ real f_BW = (D.f[DIR_P0P])[kte ];
+ real f_TE = (D.f[DIR_M0M])[kbw ];
+ real f_TW = (D.f[DIR_P0M])[kbe ];
+ real f_BE = (D.f[DIR_M0P])[ktw ];
+ real f_BS = (D.f[DIR_0PP])[ktn ];
+ real f_TN = (D.f[DIR_0MM])[kbs ];
+ real f_TS = (D.f[DIR_0PM])[kbn ];
+ real f_BN = (D.f[DIR_0MP])[kts ];
+ real f_BSW = (D.f[DIR_PPP])[ktne ];
+ real f_BNE = (D.f[DIR_MMP])[ktsw ];
+ real f_BNW = (D.f[DIR_PMP])[ktse ];
+ real f_BSE = (D.f[DIR_MPP])[ktnw ];
+ real f_TSW = (D.f[DIR_PPM])[kbne ];
+ real f_TNE = (D.f[DIR_MMM])[kbsw ];
+ real f_TNW = (D.f[DIR_PMM])[kbse ];
+ real f_TSE = (D.f[DIR_MPM])[kbnw ];
////////////////////////////////////////////////////////////////////////////////
real vx1, vx2, vx3, drho, feq, q;
drho = f_TSE + f_TNW + f_TNE + f_TSW + f_BSE + f_BNW + f_BNE + f_BSW +
@@ -3675,63 +3764,63 @@ __global__ void QSlipGeomDeviceComp27(real* DD,
//////////////////////////////////////////////////////////////////////////
if (isEvenTimestep==false)
{
- D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
+ D.f[DIR_M00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_MPM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_PMP * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
real VeloX = vx1;
@@ -4264,80 +4353,81 @@ __global__ void QSlipGeomDeviceComp27(real* DD,
//////////////////////////////////////////////////////////////////////////////
-__global__ void QSlipNormDeviceComp27(real* DD,
- int* k_Q,
- real* QQ,
- unsigned int numberOfBCnodes,
- real om1,
- real* NormalX,
- real* NormalY,
- real* NormalZ,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep)
+__global__ void QSlipNormDeviceComp27(
+ real* DD,
+ int* k_Q,
+ real* QQ,
+ unsigned int numberOfBCnodes,
+ real om1,
+ real* NormalX,
+ real* NormalY,
+ real* NormalZ,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep)
{
Distributions27 D;
if (isEvenTimestep==true)
{
- D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
+ D.f[DIR_M00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_MPM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_PMP * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////
const unsigned x = threadIdx.x; // Globaler x-Index
@@ -4358,24 +4448,24 @@ __global__ void QSlipNormDeviceComp27(real* DD,
*q_dirBE, *q_dirTW, *q_dirTN, *q_dirBS, *q_dirBN, *q_dirTS,
*q_dirTNE, *q_dirTSW, *q_dirTSE, *q_dirTNW, *q_dirBNE, *q_dirBSW,
*q_dirBSE, *q_dirBNW;
- q_dirE = &QQ[DIR_P00 * numberOfBCnodes];
- q_dirW = &QQ[DIR_M00 * numberOfBCnodes];
- q_dirN = &QQ[DIR_0P0 * numberOfBCnodes];
- q_dirS = &QQ[DIR_0M0 * numberOfBCnodes];
- q_dirT = &QQ[DIR_00P * numberOfBCnodes];
- q_dirB = &QQ[DIR_00M * numberOfBCnodes];
- q_dirNE = &QQ[DIR_PP0 * numberOfBCnodes];
- q_dirSW = &QQ[DIR_MM0 * numberOfBCnodes];
- q_dirSE = &QQ[DIR_PM0 * numberOfBCnodes];
- q_dirNW = &QQ[DIR_MP0 * numberOfBCnodes];
- q_dirTE = &QQ[DIR_P0P * numberOfBCnodes];
- q_dirBW = &QQ[DIR_M0M * numberOfBCnodes];
- q_dirBE = &QQ[DIR_P0M * numberOfBCnodes];
- q_dirTW = &QQ[DIR_M0P * numberOfBCnodes];
- q_dirTN = &QQ[DIR_0PP * numberOfBCnodes];
- q_dirBS = &QQ[DIR_0MM * numberOfBCnodes];
- q_dirBN = &QQ[DIR_0PM * numberOfBCnodes];
- q_dirTS = &QQ[DIR_0MP * numberOfBCnodes];
+ q_dirE = &QQ[DIR_P00 * numberOfBCnodes];
+ q_dirW = &QQ[DIR_M00 * numberOfBCnodes];
+ q_dirN = &QQ[DIR_0P0 * numberOfBCnodes];
+ q_dirS = &QQ[DIR_0M0 * numberOfBCnodes];
+ q_dirT = &QQ[DIR_00P * numberOfBCnodes];
+ q_dirB = &QQ[DIR_00M * numberOfBCnodes];
+ q_dirNE = &QQ[DIR_PP0 * numberOfBCnodes];
+ q_dirSW = &QQ[DIR_MM0 * numberOfBCnodes];
+ q_dirSE = &QQ[DIR_PM0 * numberOfBCnodes];
+ q_dirNW = &QQ[DIR_MP0 * numberOfBCnodes];
+ q_dirTE = &QQ[DIR_P0P * numberOfBCnodes];
+ q_dirBW = &QQ[DIR_M0M * numberOfBCnodes];
+ q_dirBE = &QQ[DIR_P0M * numberOfBCnodes];
+ q_dirTW = &QQ[DIR_M0P * numberOfBCnodes];
+ q_dirTN = &QQ[DIR_0PP * numberOfBCnodes];
+ q_dirBS = &QQ[DIR_0MM * numberOfBCnodes];
+ q_dirBN = &QQ[DIR_0PM * numberOfBCnodes];
+ q_dirTS = &QQ[DIR_0MP * numberOfBCnodes];
q_dirTNE = &QQ[DIR_PPP * numberOfBCnodes];
q_dirTSW = &QQ[DIR_MMP * numberOfBCnodes];
q_dirTSE = &QQ[DIR_PMP * numberOfBCnodes];
@@ -4390,24 +4480,24 @@ __global__ void QSlipNormDeviceComp27(real* DD,
*nx_dirBE, *nx_dirTW, *nx_dirTN, *nx_dirBS, *nx_dirBN, *nx_dirTS,
*nx_dirTNE, *nx_dirTSW, *nx_dirTSE, *nx_dirTNW, *nx_dirBNE, *nx_dirBSW,
*nx_dirBSE, *nx_dirBNW;
- nx_dirE = &NormalX[DIR_P00 * numberOfBCnodes];
- nx_dirW = &NormalX[DIR_M00 * numberOfBCnodes];
- nx_dirN = &NormalX[DIR_0P0 * numberOfBCnodes];
- nx_dirS = &NormalX[DIR_0M0 * numberOfBCnodes];
- nx_dirT = &NormalX[DIR_00P * numberOfBCnodes];
- nx_dirB = &NormalX[DIR_00M * numberOfBCnodes];
- nx_dirNE = &NormalX[DIR_PP0 * numberOfBCnodes];
- nx_dirSW = &NormalX[DIR_MM0 * numberOfBCnodes];
- nx_dirSE = &NormalX[DIR_PM0 * numberOfBCnodes];
- nx_dirNW = &NormalX[DIR_MP0 * numberOfBCnodes];
- nx_dirTE = &NormalX[DIR_P0P * numberOfBCnodes];
- nx_dirBW = &NormalX[DIR_M0M * numberOfBCnodes];
- nx_dirBE = &NormalX[DIR_P0M * numberOfBCnodes];
- nx_dirTW = &NormalX[DIR_M0P * numberOfBCnodes];
- nx_dirTN = &NormalX[DIR_0PP * numberOfBCnodes];
- nx_dirBS = &NormalX[DIR_0MM * numberOfBCnodes];
- nx_dirBN = &NormalX[DIR_0PM * numberOfBCnodes];
- nx_dirTS = &NormalX[DIR_0MP * numberOfBCnodes];
+ nx_dirE = &NormalX[DIR_P00 * numberOfBCnodes];
+ nx_dirW = &NormalX[DIR_M00 * numberOfBCnodes];
+ nx_dirN = &NormalX[DIR_0P0 * numberOfBCnodes];
+ nx_dirS = &NormalX[DIR_0M0 * numberOfBCnodes];
+ nx_dirT = &NormalX[DIR_00P * numberOfBCnodes];
+ nx_dirB = &NormalX[DIR_00M * numberOfBCnodes];
+ nx_dirNE = &NormalX[DIR_PP0 * numberOfBCnodes];
+ nx_dirSW = &NormalX[DIR_MM0 * numberOfBCnodes];
+ nx_dirSE = &NormalX[DIR_PM0 * numberOfBCnodes];
+ nx_dirNW = &NormalX[DIR_MP0 * numberOfBCnodes];
+ nx_dirTE = &NormalX[DIR_P0P * numberOfBCnodes];
+ nx_dirBW = &NormalX[DIR_M0M * numberOfBCnodes];
+ nx_dirBE = &NormalX[DIR_P0M * numberOfBCnodes];
+ nx_dirTW = &NormalX[DIR_M0P * numberOfBCnodes];
+ nx_dirTN = &NormalX[DIR_0PP * numberOfBCnodes];
+ nx_dirBS = &NormalX[DIR_0MM * numberOfBCnodes];
+ nx_dirBN = &NormalX[DIR_0PM * numberOfBCnodes];
+ nx_dirTS = &NormalX[DIR_0MP * numberOfBCnodes];
nx_dirTNE = &NormalX[DIR_PPP * numberOfBCnodes];
nx_dirTSW = &NormalX[DIR_MMP * numberOfBCnodes];
nx_dirTSE = &NormalX[DIR_PMP * numberOfBCnodes];
@@ -4422,24 +4512,24 @@ __global__ void QSlipNormDeviceComp27(real* DD,
*ny_dirBE, *ny_dirTW, *ny_dirTN, *ny_dirBS, *ny_dirBN, *ny_dirTS,
*ny_dirTNE, *ny_dirTSW, *ny_dirTSE, *ny_dirTNW, *ny_dirBNE, *ny_dirBSW,
*ny_dirBSE, *ny_dirBNW;
- ny_dirE = &NormalY[DIR_P00 * numberOfBCnodes];
- ny_dirW = &NormalY[DIR_M00 * numberOfBCnodes];
- ny_dirN = &NormalY[DIR_0P0 * numberOfBCnodes];
- ny_dirS = &NormalY[DIR_0M0 * numberOfBCnodes];
- ny_dirT = &NormalY[DIR_00P * numberOfBCnodes];
- ny_dirB = &NormalY[DIR_00M * numberOfBCnodes];
- ny_dirNE = &NormalY[DIR_PP0 * numberOfBCnodes];
- ny_dirSW = &NormalY[DIR_MM0 * numberOfBCnodes];
- ny_dirSE = &NormalY[DIR_PM0 * numberOfBCnodes];
- ny_dirNW = &NormalY[DIR_MP0 * numberOfBCnodes];
- ny_dirTE = &NormalY[DIR_P0P * numberOfBCnodes];
- ny_dirBW = &NormalY[DIR_M0M * numberOfBCnodes];
- ny_dirBE = &NormalY[DIR_P0M * numberOfBCnodes];
- ny_dirTW = &NormalY[DIR_M0P * numberOfBCnodes];
- ny_dirTN = &NormalY[DIR_0PP * numberOfBCnodes];
- ny_dirBS = &NormalY[DIR_0MM * numberOfBCnodes];
- ny_dirBN = &NormalY[DIR_0PM * numberOfBCnodes];
- ny_dirTS = &NormalY[DIR_0MP * numberOfBCnodes];
+ ny_dirE = &NormalY[DIR_P00 * numberOfBCnodes];
+ ny_dirW = &NormalY[DIR_M00 * numberOfBCnodes];
+ ny_dirN = &NormalY[DIR_0P0 * numberOfBCnodes];
+ ny_dirS = &NormalY[DIR_0M0 * numberOfBCnodes];
+ ny_dirT = &NormalY[DIR_00P * numberOfBCnodes];
+ ny_dirB = &NormalY[DIR_00M * numberOfBCnodes];
+ ny_dirNE = &NormalY[DIR_PP0 * numberOfBCnodes];
+ ny_dirSW = &NormalY[DIR_MM0 * numberOfBCnodes];
+ ny_dirSE = &NormalY[DIR_PM0 * numberOfBCnodes];
+ ny_dirNW = &NormalY[DIR_MP0 * numberOfBCnodes];
+ ny_dirTE = &NormalY[DIR_P0P * numberOfBCnodes];
+ ny_dirBW = &NormalY[DIR_M0M * numberOfBCnodes];
+ ny_dirBE = &NormalY[DIR_P0M * numberOfBCnodes];
+ ny_dirTW = &NormalY[DIR_M0P * numberOfBCnodes];
+ ny_dirTN = &NormalY[DIR_0PP * numberOfBCnodes];
+ ny_dirBS = &NormalY[DIR_0MM * numberOfBCnodes];
+ ny_dirBN = &NormalY[DIR_0PM * numberOfBCnodes];
+ ny_dirTS = &NormalY[DIR_0MP * numberOfBCnodes];
ny_dirTNE = &NormalY[DIR_PPP * numberOfBCnodes];
ny_dirTSW = &NormalY[DIR_MMP * numberOfBCnodes];
ny_dirTSE = &NormalY[DIR_PMP * numberOfBCnodes];
@@ -4454,24 +4544,24 @@ __global__ void QSlipNormDeviceComp27(real* DD,
*nz_dirBE, *nz_dirTW, *nz_dirTN, *nz_dirBS, *nz_dirBN, *nz_dirTS,
*nz_dirTNE, *nz_dirTSW, *nz_dirTSE, *nz_dirTNW, *nz_dirBNE, *nz_dirBSW,
*nz_dirBSE, *nz_dirBNW;
- nz_dirE = &NormalZ[DIR_P00 * numberOfBCnodes];
- nz_dirW = &NormalZ[DIR_M00 * numberOfBCnodes];
- nz_dirN = &NormalZ[DIR_0P0 * numberOfBCnodes];
- nz_dirS = &NormalZ[DIR_0M0 * numberOfBCnodes];
- nz_dirT = &NormalZ[DIR_00P * numberOfBCnodes];
- nz_dirB = &NormalZ[DIR_00M * numberOfBCnodes];
- nz_dirNE = &NormalZ[DIR_PP0 * numberOfBCnodes];
- nz_dirSW = &NormalZ[DIR_MM0 * numberOfBCnodes];
- nz_dirSE = &NormalZ[DIR_PM0 * numberOfBCnodes];
- nz_dirNW = &NormalZ[DIR_MP0 * numberOfBCnodes];
- nz_dirTE = &NormalZ[DIR_P0P * numberOfBCnodes];
- nz_dirBW = &NormalZ[DIR_M0M * numberOfBCnodes];
- nz_dirBE = &NormalZ[DIR_P0M * numberOfBCnodes];
- nz_dirTW = &NormalZ[DIR_M0P * numberOfBCnodes];
- nz_dirTN = &NormalZ[DIR_0PP * numberOfBCnodes];
- nz_dirBS = &NormalZ[DIR_0MM * numberOfBCnodes];
- nz_dirBN = &NormalZ[DIR_0PM * numberOfBCnodes];
- nz_dirTS = &NormalZ[DIR_0MP * numberOfBCnodes];
+ nz_dirE = &NormalZ[DIR_P00 * numberOfBCnodes];
+ nz_dirW = &NormalZ[DIR_M00 * numberOfBCnodes];
+ nz_dirN = &NormalZ[DIR_0P0 * numberOfBCnodes];
+ nz_dirS = &NormalZ[DIR_0M0 * numberOfBCnodes];
+ nz_dirT = &NormalZ[DIR_00P * numberOfBCnodes];
+ nz_dirB = &NormalZ[DIR_00M * numberOfBCnodes];
+ nz_dirNE = &NormalZ[DIR_PP0 * numberOfBCnodes];
+ nz_dirSW = &NormalZ[DIR_MM0 * numberOfBCnodes];
+ nz_dirSE = &NormalZ[DIR_PM0 * numberOfBCnodes];
+ nz_dirNW = &NormalZ[DIR_MP0 * numberOfBCnodes];
+ nz_dirTE = &NormalZ[DIR_P0P * numberOfBCnodes];
+ nz_dirBW = &NormalZ[DIR_M0M * numberOfBCnodes];
+ nz_dirBE = &NormalZ[DIR_P0M * numberOfBCnodes];
+ nz_dirTW = &NormalZ[DIR_M0P * numberOfBCnodes];
+ nz_dirTN = &NormalZ[DIR_0PP * numberOfBCnodes];
+ nz_dirBS = &NormalZ[DIR_0MM * numberOfBCnodes];
+ nz_dirBN = &NormalZ[DIR_0PM * numberOfBCnodes];
+ nz_dirTS = &NormalZ[DIR_0MP * numberOfBCnodes];
nz_dirTNE = &NormalZ[DIR_PPP * numberOfBCnodes];
nz_dirTSW = &NormalZ[DIR_MMP * numberOfBCnodes];
nz_dirTSE = &NormalZ[DIR_PMP * numberOfBCnodes];
@@ -4511,32 +4601,32 @@ __global__ void QSlipNormDeviceComp27(real* DD,
unsigned int ktne = KQK;
unsigned int kbsw = neighborZ[ksw];
////////////////////////////////////////////////////////////////////////////////
- real f_W = (D.f[DIR_P00 ])[ke ];
- real f_E = (D.f[DIR_M00 ])[kw ];
- real f_S = (D.f[DIR_0P0 ])[kn ];
- real f_N = (D.f[DIR_0M0 ])[ks ];
- real f_B = (D.f[DIR_00P ])[kt ];
- real f_T = (D.f[DIR_00M ])[kb ];
- real f_SW = (D.f[DIR_PP0 ])[kne ];
- real f_NE = (D.f[DIR_MM0 ])[ksw ];
- real f_NW = (D.f[DIR_PM0 ])[kse ];
- real f_SE = (D.f[DIR_MP0 ])[knw ];
- real f_BW = (D.f[DIR_P0P ])[kte ];
- real f_TE = (D.f[DIR_M0M ])[kbw ];
- real f_TW = (D.f[DIR_P0M ])[kbe ];
- real f_BE = (D.f[DIR_M0P ])[ktw ];
- real f_BS = (D.f[DIR_0PP ])[ktn ];
- real f_TN = (D.f[DIR_0MM ])[kbs ];
- real f_TS = (D.f[DIR_0PM ])[kbn ];
- real f_BN = (D.f[DIR_0MP ])[kts ];
- real f_BSW = (D.f[DIR_PPP ])[ktne ];
- real f_BNE = (D.f[DIR_MMP ])[ktsw ];
- real f_BNW = (D.f[DIR_PMP ])[ktse ];
- real f_BSE = (D.f[DIR_MPP ])[ktnw ];
- real f_TSW = (D.f[DIR_PPM ])[kbne ];
- real f_TNE = (D.f[DIR_MMM ])[kbsw ];
- real f_TNW = (D.f[DIR_PMM ])[kbse ];
- real f_TSE = (D.f[DIR_MPM ])[kbnw ];
+ real f_W = (D.f[DIR_P00])[ke ];
+ real f_E = (D.f[DIR_M00])[kw ];
+ real f_S = (D.f[DIR_0P0])[kn ];
+ real f_N = (D.f[DIR_0M0])[ks ];
+ real f_B = (D.f[DIR_00P])[kt ];
+ real f_T = (D.f[DIR_00M])[kb ];
+ real f_SW = (D.f[DIR_PP0])[kne ];
+ real f_NE = (D.f[DIR_MM0])[ksw ];
+ real f_NW = (D.f[DIR_PM0])[kse ];
+ real f_SE = (D.f[DIR_MP0])[knw ];
+ real f_BW = (D.f[DIR_P0P])[kte ];
+ real f_TE = (D.f[DIR_M0M])[kbw ];
+ real f_TW = (D.f[DIR_P0M])[kbe ];
+ real f_BE = (D.f[DIR_M0P])[ktw ];
+ real f_BS = (D.f[DIR_0PP])[ktn ];
+ real f_TN = (D.f[DIR_0MM])[kbs ];
+ real f_TS = (D.f[DIR_0PM])[kbn ];
+ real f_BN = (D.f[DIR_0MP])[kts ];
+ real f_BSW = (D.f[DIR_PPP])[ktne ];
+ real f_BNE = (D.f[DIR_MMP])[ktsw ];
+ real f_BNW = (D.f[DIR_PMP])[ktse ];
+ real f_BSE = (D.f[DIR_MPP])[ktnw ];
+ real f_TSW = (D.f[DIR_PPM])[kbne ];
+ real f_TNE = (D.f[DIR_MMM])[kbsw ];
+ real f_TNW = (D.f[DIR_PMM])[kbse ];
+ real f_TSE = (D.f[DIR_MPM])[kbnw ];
////////////////////////////////////////////////////////////////////////////////
real vx1, vx2, vx3, drho, feq, q;
drho = f_TSE + f_TNW + f_TNE + f_TSW + f_BSE + f_BNW + f_BNE + f_BSW +
@@ -4561,63 +4651,63 @@ __global__ void QSlipNormDeviceComp27(real* DD,
//////////////////////////////////////////////////////////////////////////
if (isEvenTimestep==false)
{
- D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
+ D.f[DIR_M00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_MPM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_PMP * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
real VeloX = vx1;
diff --git a/src/gpu/VirtualFluids_GPU/GPU/StressBCs27.cu b/src/gpu/VirtualFluids_GPU/GPU/StressBCs27.cu
index 74e2faa38638228aa5d499aa74226405ab109f7d..3208299e93940dabe52faa7d0b3c684c45596660 100644
--- a/src/gpu/VirtualFluids_GPU/GPU/StressBCs27.cu
+++ b/src/gpu/VirtualFluids_GPU/GPU/StressBCs27.cu
@@ -43,28 +43,30 @@
#include "LBM/LB.h"
#include "lbm/constants/D3Q27.h"
#include <lbm/constants/NumericConstants.h>
-#include "KernelUtilities.h"
+#include "LBM/GPUHelperFunctions/KernelUtilities.h"
using namespace vf::lbm::constant;
using namespace vf::lbm::dir;
+using namespace vf::gpu;
//////////////////////////////////////////////////////////////////////////////
-__host__ __device__ __forceinline__ void iMEM(uint k, uint kN,
- real* _wallNormalX, real* _wallNormalY, real* _wallNormalZ,
- real* vx, real* vy, real* vz,
- real* vx_el, real* vy_el, real* vz_el, //!>mean (temporally filtered) velocities at exchange location
- real* vx_w_mean, real* vy_w_mean, real* vz_w_mean, //!>mean (temporally filtered) velocities at wall-adjactent node
- real vx_w_inst, real vy_w_inst, real vz_w_inst, //!>instantaneous velocities at wall-adjactent node
- real rho,
- int* samplingOffset,
- real q,
- real forceFactor, //!>e.g., 1.0 for simple-bounce back, or (1+q) for interpolated single-node bounce-back as in Geier et al (2015)
- real eps, //!>filter constant in temporal averaging
- real* z0, //!>aerodynamic roughness length
- bool hasWallModelMonitor,
- real* u_star_monitor,
- real wallMomentumX, real wallMomentumY, real wallMomentumZ,
- real& wallVelocityX, real& wallVelocityY, real&wallVelocityZ)
+__host__ __device__ __forceinline__ void iMEM(
+ uint k, uint kN,
+ real* _wallNormalX, real* _wallNormalY, real* _wallNormalZ,
+ real* vx, real* vy, real* vz,
+ real* vx_el, real* vy_el, real* vz_el, //!>mean (temporally filtered) velocities at exchange location
+ real* vx_w_mean, real* vy_w_mean, real* vz_w_mean, //!>mean (temporally filtered) velocities at wall-adjactent node
+ real vx_w_inst, real vy_w_inst, real vz_w_inst, //!>instantaneous velocities at wall-adjactent node
+ real rho,
+ int* samplingOffset,
+ real q,
+ real forceFactor, //!>e.g., 1.0 for simple-bounce back, or (1+q) for interpolated single-node bounce-back as in Geier et al (2015)
+ real eps, //!>filter constant in temporal averaging
+ real* z0, //!>aerodynamic roughness length
+ bool hasWallModelMonitor,
+ real* u_star_monitor,
+ real wallMomentumX, real wallMomentumY, real wallMomentumZ,
+ real& wallVelocityX, real& wallVelocityY, real&wallVelocityZ)
{
real wallNormalX = _wallNormalX[k];
real wallNormalY = _wallNormalY[k];
@@ -136,99 +138,100 @@ __host__ __device__ __forceinline__ void iMEM(uint k, uint kN,
}
//////////////////////////////////////////////////////////////////////////////
-__global__ void QStressDeviceComp27(real* DD,
- int* k_Q,
- int* k_N,
- real* QQ,
- unsigned int numberOfBCnodes,
- real om1,
- real* turbViscosity,
- real* vx,
- real* vy,
- real* vz,
- real* normalX,
- real* normalY,
- real* normalZ,
- real* vx_el,
- real* vy_el,
- real* vz_el,
- real* vx_w_mean,
- real* vy_w_mean,
- real* vz_w_mean,
- int* samplingOffset,
- real* z0,
- bool hasWallModelMonitor,
- real* u_star_monitor,
- real* Fx_monitor,
- real* Fy_monitor,
- real* Fz_monitor,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep)
+__global__ void QStressDeviceComp27(
+ real* DD,
+ int* k_Q,
+ int* k_N,
+ real* QQ,
+ unsigned int numberOfBCnodes,
+ real om1,
+ real* turbViscosity,
+ real* vx,
+ real* vy,
+ real* vz,
+ real* normalX,
+ real* normalY,
+ real* normalZ,
+ real* vx_el,
+ real* vy_el,
+ real* vz_el,
+ real* vx_w_mean,
+ real* vy_w_mean,
+ real* vz_w_mean,
+ int* samplingOffset,
+ real* z0,
+ bool hasWallModelMonitor,
+ real* u_star_monitor,
+ real* Fx_monitor,
+ real* Fy_monitor,
+ real* Fz_monitor,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep)
{
Distributions27 D;
if (isEvenTimestep==true)//get right array of post coll f's
{
- D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
+ D.f[DIR_M00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_MPM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_PMP * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////
const unsigned x = threadIdx.x; // Globaler x-Index
@@ -249,24 +252,24 @@ __global__ void QStressDeviceComp27(real* DD,
*q_dirBE, *q_dirTW, *q_dirTN, *q_dirBS, *q_dirBN, *q_dirTS,
*q_dirTNE, *q_dirTSW, *q_dirTSE, *q_dirTNW, *q_dirBNE, *q_dirBSW,
*q_dirBSE, *q_dirBNW;
- q_dirE = &QQ[DIR_P00 * numberOfBCnodes];
- q_dirW = &QQ[DIR_M00 * numberOfBCnodes];
- q_dirN = &QQ[DIR_0P0 * numberOfBCnodes];
- q_dirS = &QQ[DIR_0M0 * numberOfBCnodes];
- q_dirT = &QQ[DIR_00P * numberOfBCnodes];
- q_dirB = &QQ[DIR_00M * numberOfBCnodes];
- q_dirNE = &QQ[DIR_PP0 * numberOfBCnodes];
- q_dirSW = &QQ[DIR_MM0 * numberOfBCnodes];
- q_dirSE = &QQ[DIR_PM0 * numberOfBCnodes];
- q_dirNW = &QQ[DIR_MP0 * numberOfBCnodes];
- q_dirTE = &QQ[DIR_P0P * numberOfBCnodes];
- q_dirBW = &QQ[DIR_M0M * numberOfBCnodes];
- q_dirBE = &QQ[DIR_P0M * numberOfBCnodes];
- q_dirTW = &QQ[DIR_M0P * numberOfBCnodes];
- q_dirTN = &QQ[DIR_0PP * numberOfBCnodes];
- q_dirBS = &QQ[DIR_0MM * numberOfBCnodes];
- q_dirBN = &QQ[DIR_0PM * numberOfBCnodes];
- q_dirTS = &QQ[DIR_0MP * numberOfBCnodes];
+ q_dirE = &QQ[DIR_P00 * numberOfBCnodes];
+ q_dirW = &QQ[DIR_M00 * numberOfBCnodes];
+ q_dirN = &QQ[DIR_0P0 * numberOfBCnodes];
+ q_dirS = &QQ[DIR_0M0 * numberOfBCnodes];
+ q_dirT = &QQ[DIR_00P * numberOfBCnodes];
+ q_dirB = &QQ[DIR_00M * numberOfBCnodes];
+ q_dirNE = &QQ[DIR_PP0 * numberOfBCnodes];
+ q_dirSW = &QQ[DIR_MM0 * numberOfBCnodes];
+ q_dirSE = &QQ[DIR_PM0 * numberOfBCnodes];
+ q_dirNW = &QQ[DIR_MP0 * numberOfBCnodes];
+ q_dirTE = &QQ[DIR_P0P * numberOfBCnodes];
+ q_dirBW = &QQ[DIR_M0M * numberOfBCnodes];
+ q_dirBE = &QQ[DIR_P0M * numberOfBCnodes];
+ q_dirTW = &QQ[DIR_M0P * numberOfBCnodes];
+ q_dirTN = &QQ[DIR_0PP * numberOfBCnodes];
+ q_dirBS = &QQ[DIR_0MM * numberOfBCnodes];
+ q_dirBN = &QQ[DIR_0PM * numberOfBCnodes];
+ q_dirTS = &QQ[DIR_0MP * numberOfBCnodes];
q_dirTNE = &QQ[DIR_PPP * numberOfBCnodes];
q_dirTSW = &QQ[DIR_MMP * numberOfBCnodes];
q_dirTSE = &QQ[DIR_PMP * numberOfBCnodes];
@@ -309,32 +312,32 @@ __global__ void QStressDeviceComp27(real* DD,
real f_E, f_W, f_N, f_S, f_T, f_B, f_NE, f_SW, f_SE, f_NW, f_TE, f_BW, f_BE,
f_TW, f_TN, f_BS, f_BN, f_TS, f_TNE, f_TSW, f_TSE, f_TNW, f_BNE, f_BSW, f_BSE, f_BNW;
- f_W = (D.f[DIR_P00 ])[ke ]; //post-coll f's
- f_E = (D.f[DIR_M00 ])[kw ];
- f_S = (D.f[DIR_0P0 ])[kn ];
- f_N = (D.f[DIR_0M0 ])[ks ];
- f_B = (D.f[DIR_00P ])[kt ];
- f_T = (D.f[DIR_00M ])[kb ];
- f_SW = (D.f[DIR_PP0 ])[kne ];
- f_NE = (D.f[DIR_MM0 ])[ksw ];
- f_NW = (D.f[DIR_PM0 ])[kse ];
- f_SE = (D.f[DIR_MP0 ])[knw ];
- f_BW = (D.f[DIR_P0P ])[kte ];
- f_TE = (D.f[DIR_M0M ])[kbw ];
- f_TW = (D.f[DIR_P0M ])[kbe ];
- f_BE = (D.f[DIR_M0P ])[ktw ];
- f_BS = (D.f[DIR_0PP ])[ktn ];
- f_TN = (D.f[DIR_0MM ])[kbs ];
- f_TS = (D.f[DIR_0PM ])[kbn ];
- f_BN = (D.f[DIR_0MP ])[kts ];
- f_BSW = (D.f[DIR_PPP ])[ktne ];
- f_BNE = (D.f[DIR_MMP ])[ktsw ];
- f_BNW = (D.f[DIR_PMP ])[ktse ];
- f_BSE = (D.f[DIR_MPP ])[ktnw ];
- f_TSW = (D.f[DIR_PPM ])[kbne ];
- f_TNE = (D.f[DIR_MMM ])[kbsw ];
- f_TNW = (D.f[DIR_PMM ])[kbse ];
- f_TSE = (D.f[DIR_MPM ])[kbnw ];
+ f_W = (D.f[DIR_P00])[ke ]; //post-coll f's
+ f_E = (D.f[DIR_M00])[kw ];
+ f_S = (D.f[DIR_0P0])[kn ];
+ f_N = (D.f[DIR_0M0])[ks ];
+ f_B = (D.f[DIR_00P])[kt ];
+ f_T = (D.f[DIR_00M])[kb ];
+ f_SW = (D.f[DIR_PP0])[kne ];
+ f_NE = (D.f[DIR_MM0])[ksw ];
+ f_NW = (D.f[DIR_PM0])[kse ];
+ f_SE = (D.f[DIR_MP0])[knw ];
+ f_BW = (D.f[DIR_P0P])[kte ];
+ f_TE = (D.f[DIR_M0M])[kbw ];
+ f_TW = (D.f[DIR_P0M])[kbe ];
+ f_BE = (D.f[DIR_M0P])[ktw ];
+ f_BS = (D.f[DIR_0PP])[ktn ];
+ f_TN = (D.f[DIR_0MM])[kbs ];
+ f_TS = (D.f[DIR_0PM])[kbn ];
+ f_BN = (D.f[DIR_0MP])[kts ];
+ f_BSW = (D.f[DIR_PPP])[ktne ];
+ f_BNE = (D.f[DIR_MMP])[ktsw ];
+ f_BNW = (D.f[DIR_PMP])[ktse ];
+ f_BSE = (D.f[DIR_MPP])[ktnw ];
+ f_TSW = (D.f[DIR_PPM])[kbne ];
+ f_TNE = (D.f[DIR_MMM])[kbsw ];
+ f_TNW = (D.f[DIR_PMM])[kbse ];
+ f_TSE = (D.f[DIR_MPM])[kbnw ];
////////////////////////////////////////////////////////////////////////////////
real vx1, vx2, vx3, drho, feq, q;
@@ -361,63 +364,63 @@ __global__ void QStressDeviceComp27(real* DD,
//////////////////////////////////////////////////////////////////////////
if (isEvenTimestep==false) //get adress where incoming f's should be written to
{
- D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
+ D.f[DIR_M00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_MPM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_PMP * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
//Compute incoming f's with zero wall velocity
@@ -968,69 +971,69 @@ __global__ void BBStressDevice27( real* DD,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep)
{
Distributions27 D;
if (isEvenTimestep==true)
{
- D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
+ D.f[DIR_M00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_MPM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_PMP * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////
const unsigned x = threadIdx.x; // Globaler x-Index
@@ -1051,24 +1054,24 @@ __global__ void BBStressDevice27( real* DD,
*q_dirBE, *q_dirTW, *q_dirTN, *q_dirBS, *q_dirBN, *q_dirTS,
*q_dirTNE, *q_dirTSW, *q_dirTSE, *q_dirTNW, *q_dirBNE, *q_dirBSW,
*q_dirBSE, *q_dirBNW;
- q_dirE = &QQ[DIR_P00 * numberOfBCnodes];
- q_dirW = &QQ[DIR_M00 * numberOfBCnodes];
- q_dirN = &QQ[DIR_0P0 * numberOfBCnodes];
- q_dirS = &QQ[DIR_0M0 * numberOfBCnodes];
- q_dirT = &QQ[DIR_00P * numberOfBCnodes];
- q_dirB = &QQ[DIR_00M * numberOfBCnodes];
- q_dirNE = &QQ[DIR_PP0 * numberOfBCnodes];
- q_dirSW = &QQ[DIR_MM0 * numberOfBCnodes];
- q_dirSE = &QQ[DIR_PM0 * numberOfBCnodes];
- q_dirNW = &QQ[DIR_MP0 * numberOfBCnodes];
- q_dirTE = &QQ[DIR_P0P * numberOfBCnodes];
- q_dirBW = &QQ[DIR_M0M * numberOfBCnodes];
- q_dirBE = &QQ[DIR_P0M * numberOfBCnodes];
- q_dirTW = &QQ[DIR_M0P * numberOfBCnodes];
- q_dirTN = &QQ[DIR_0PP * numberOfBCnodes];
- q_dirBS = &QQ[DIR_0MM * numberOfBCnodes];
- q_dirBN = &QQ[DIR_0PM * numberOfBCnodes];
- q_dirTS = &QQ[DIR_0MP * numberOfBCnodes];
+ q_dirE = &QQ[DIR_P00 * numberOfBCnodes];
+ q_dirW = &QQ[DIR_M00 * numberOfBCnodes];
+ q_dirN = &QQ[DIR_0P0 * numberOfBCnodes];
+ q_dirS = &QQ[DIR_0M0 * numberOfBCnodes];
+ q_dirT = &QQ[DIR_00P * numberOfBCnodes];
+ q_dirB = &QQ[DIR_00M * numberOfBCnodes];
+ q_dirNE = &QQ[DIR_PP0 * numberOfBCnodes];
+ q_dirSW = &QQ[DIR_MM0 * numberOfBCnodes];
+ q_dirSE = &QQ[DIR_PM0 * numberOfBCnodes];
+ q_dirNW = &QQ[DIR_MP0 * numberOfBCnodes];
+ q_dirTE = &QQ[DIR_P0P * numberOfBCnodes];
+ q_dirBW = &QQ[DIR_M0M * numberOfBCnodes];
+ q_dirBE = &QQ[DIR_P0M * numberOfBCnodes];
+ q_dirTW = &QQ[DIR_M0P * numberOfBCnodes];
+ q_dirTN = &QQ[DIR_0PP * numberOfBCnodes];
+ q_dirBS = &QQ[DIR_0MM * numberOfBCnodes];
+ q_dirBN = &QQ[DIR_0PM * numberOfBCnodes];
+ q_dirTS = &QQ[DIR_0MP * numberOfBCnodes];
q_dirTNE = &QQ[DIR_PPP * numberOfBCnodes];
q_dirTSW = &QQ[DIR_MMP * numberOfBCnodes];
q_dirTSE = &QQ[DIR_PMP * numberOfBCnodes];
@@ -1112,32 +1115,32 @@ __global__ void BBStressDevice27( real* DD,
real f_E, f_W, f_N, f_S, f_T, f_B, f_NE, f_SW, f_SE, f_NW, f_TE, f_BW, f_BE,
f_TW, f_TN, f_BS, f_BN, f_TS, f_TNE, f_TSW, f_TSE, f_TNW, f_BNE, f_BSW, f_BSE, f_BNW;
- f_W = (D.f[DIR_P00 ])[ke ];
- f_E = (D.f[DIR_M00 ])[kw ];
- f_S = (D.f[DIR_0P0 ])[kn ];
- f_N = (D.f[DIR_0M0 ])[ks ];
- f_B = (D.f[DIR_00P ])[kt ];
- f_T = (D.f[DIR_00M ])[kb ];
- f_SW = (D.f[DIR_PP0 ])[kne ];
- f_NE = (D.f[DIR_MM0 ])[ksw ];
- f_NW = (D.f[DIR_PM0 ])[kse ];
- f_SE = (D.f[DIR_MP0 ])[knw ];
- f_BW = (D.f[DIR_P0P ])[kte ];
- f_TE = (D.f[DIR_M0M ])[kbw ];
- f_TW = (D.f[DIR_P0M ])[kbe ];
- f_BE = (D.f[DIR_M0P ])[ktw ];
- f_BS = (D.f[DIR_0PP ])[ktn ];
- f_TN = (D.f[DIR_0MM ])[kbs ];
- f_TS = (D.f[DIR_0PM ])[kbn ];
- f_BN = (D.f[DIR_0MP ])[kts ];
- f_BSW = (D.f[DIR_PPP ])[ktne ];
- f_BNE = (D.f[DIR_MMP ])[ktsw ];
- f_BNW = (D.f[DIR_PMP ])[ktse ];
- f_BSE = (D.f[DIR_MPP ])[ktnw ];
- f_TSW = (D.f[DIR_PPM ])[kbne ];
- f_TNE = (D.f[DIR_MMM ])[kbsw ];
- f_TNW = (D.f[DIR_PMM ])[kbse ];
- f_TSE = (D.f[DIR_MPM ])[kbnw ];
+ f_W = (D.f[DIR_P00])[ke ];
+ f_E = (D.f[DIR_M00])[kw ];
+ f_S = (D.f[DIR_0P0])[kn ];
+ f_N = (D.f[DIR_0M0])[ks ];
+ f_B = (D.f[DIR_00P])[kt ];
+ f_T = (D.f[DIR_00M])[kb ];
+ f_SW = (D.f[DIR_PP0])[kne ];
+ f_NE = (D.f[DIR_MM0])[ksw ];
+ f_NW = (D.f[DIR_PM0])[kse ];
+ f_SE = (D.f[DIR_MP0])[knw ];
+ f_BW = (D.f[DIR_P0P])[kte ];
+ f_TE = (D.f[DIR_M0M])[kbw ];
+ f_TW = (D.f[DIR_P0M])[kbe ];
+ f_BE = (D.f[DIR_M0P])[ktw ];
+ f_BS = (D.f[DIR_0PP])[ktn ];
+ f_TN = (D.f[DIR_0MM])[kbs ];
+ f_TS = (D.f[DIR_0PM])[kbn ];
+ f_BN = (D.f[DIR_0MP])[kts ];
+ f_BSW = (D.f[DIR_PPP])[ktne ];
+ f_BNE = (D.f[DIR_MMP])[ktsw ];
+ f_BNW = (D.f[DIR_PMP])[ktse ];
+ f_BSE = (D.f[DIR_MPP])[ktnw ];
+ f_TSW = (D.f[DIR_PPM])[kbne ];
+ f_TNE = (D.f[DIR_MMM])[kbsw ];
+ f_TNW = (D.f[DIR_PMM])[kbse ];
+ f_TSE = (D.f[DIR_MPM])[kbnw ];
////////////////////////////////////////////////////////////////////////////////
real vx1, vx2, vx3, drho;
@@ -1161,63 +1164,63 @@ __global__ void BBStressDevice27( real* DD,
//////////////////////////////////////////////////////////////////////////
if (isEvenTimestep==false)
{
- D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
+ D.f[DIR_M00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_MPM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_PMP * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
real f_E_in, f_W_in, f_N_in, f_S_in, f_T_in, f_B_in, f_NE_in, f_SW_in, f_SE_in, f_NW_in, f_TE_in, f_BW_in, f_BE_in,
@@ -1715,69 +1718,69 @@ __global__ void BBStressPressureDevice27( real* DD,
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep)
{
Distributions27 D;
if (isEvenTimestep==true)
{
- D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
+ D.f[DIR_M00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_MPM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_PMP * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////
const unsigned x = threadIdx.x; // Globaler x-Index
@@ -1798,24 +1801,24 @@ __global__ void BBStressPressureDevice27( real* DD,
*q_dirBE, *q_dirTW, *q_dirTN, *q_dirBS, *q_dirBN, *q_dirTS,
*q_dirTNE, *q_dirTSW, *q_dirTSE, *q_dirTNW, *q_dirBNE, *q_dirBSW,
*q_dirBSE, *q_dirBNW;
- q_dirE = &QQ[DIR_P00 * numberOfBCnodes];
- q_dirW = &QQ[DIR_M00 * numberOfBCnodes];
- q_dirN = &QQ[DIR_0P0 * numberOfBCnodes];
- q_dirS = &QQ[DIR_0M0 * numberOfBCnodes];
- q_dirT = &QQ[DIR_00P * numberOfBCnodes];
- q_dirB = &QQ[DIR_00M * numberOfBCnodes];
- q_dirNE = &QQ[DIR_PP0 * numberOfBCnodes];
- q_dirSW = &QQ[DIR_MM0 * numberOfBCnodes];
- q_dirSE = &QQ[DIR_PM0 * numberOfBCnodes];
- q_dirNW = &QQ[DIR_MP0 * numberOfBCnodes];
- q_dirTE = &QQ[DIR_P0P * numberOfBCnodes];
- q_dirBW = &QQ[DIR_M0M * numberOfBCnodes];
- q_dirBE = &QQ[DIR_P0M * numberOfBCnodes];
- q_dirTW = &QQ[DIR_M0P * numberOfBCnodes];
- q_dirTN = &QQ[DIR_0PP * numberOfBCnodes];
- q_dirBS = &QQ[DIR_0MM * numberOfBCnodes];
- q_dirBN = &QQ[DIR_0PM * numberOfBCnodes];
- q_dirTS = &QQ[DIR_0MP * numberOfBCnodes];
+ q_dirE = &QQ[DIR_P00 * numberOfBCnodes];
+ q_dirW = &QQ[DIR_M00 * numberOfBCnodes];
+ q_dirN = &QQ[DIR_0P0 * numberOfBCnodes];
+ q_dirS = &QQ[DIR_0M0 * numberOfBCnodes];
+ q_dirT = &QQ[DIR_00P * numberOfBCnodes];
+ q_dirB = &QQ[DIR_00M * numberOfBCnodes];
+ q_dirNE = &QQ[DIR_PP0 * numberOfBCnodes];
+ q_dirSW = &QQ[DIR_MM0 * numberOfBCnodes];
+ q_dirSE = &QQ[DIR_PM0 * numberOfBCnodes];
+ q_dirNW = &QQ[DIR_MP0 * numberOfBCnodes];
+ q_dirTE = &QQ[DIR_P0P * numberOfBCnodes];
+ q_dirBW = &QQ[DIR_M0M * numberOfBCnodes];
+ q_dirBE = &QQ[DIR_P0M * numberOfBCnodes];
+ q_dirTW = &QQ[DIR_M0P * numberOfBCnodes];
+ q_dirTN = &QQ[DIR_0PP * numberOfBCnodes];
+ q_dirBS = &QQ[DIR_0MM * numberOfBCnodes];
+ q_dirBN = &QQ[DIR_0PM * numberOfBCnodes];
+ q_dirTS = &QQ[DIR_0MP * numberOfBCnodes];
q_dirTNE = &QQ[DIR_PPP * numberOfBCnodes];
q_dirTSW = &QQ[DIR_MMP * numberOfBCnodes];
q_dirTSE = &QQ[DIR_PMP * numberOfBCnodes];
@@ -1859,32 +1862,32 @@ __global__ void BBStressPressureDevice27( real* DD,
real f_E, f_W, f_N, f_S, f_T, f_B, f_NE, f_SW, f_SE, f_NW, f_TE, f_BW, f_BE,
f_TW, f_TN, f_BS, f_BN, f_TS, f_TNE, f_TSW, f_TSE, f_TNW, f_BNE, f_BSW, f_BSE, f_BNW;
- f_W = (D.f[DIR_P00 ])[ke ];
- f_E = (D.f[DIR_M00 ])[kw ];
- f_S = (D.f[DIR_0P0 ])[kn ];
- f_N = (D.f[DIR_0M0 ])[ks ];
- f_B = (D.f[DIR_00P ])[kt ];
- f_T = (D.f[DIR_00M ])[kb ];
- f_SW = (D.f[DIR_PP0 ])[kne ];
- f_NE = (D.f[DIR_MM0 ])[ksw ];
- f_NW = (D.f[DIR_PM0 ])[kse ];
- f_SE = (D.f[DIR_MP0 ])[knw ];
- f_BW = (D.f[DIR_P0P ])[kte ];
- f_TE = (D.f[DIR_M0M ])[kbw ];
- f_TW = (D.f[DIR_P0M ])[kbe ];
- f_BE = (D.f[DIR_M0P ])[ktw ];
- f_BS = (D.f[DIR_0PP ])[ktn ];
- f_TN = (D.f[DIR_0MM ])[kbs ];
- f_TS = (D.f[DIR_0PM ])[kbn ];
- f_BN = (D.f[DIR_0MP ])[kts ];
- f_BSW = (D.f[DIR_PPP ])[ktne ];
- f_BNE = (D.f[DIR_MMP ])[ktsw ];
- f_BNW = (D.f[DIR_PMP ])[ktse ];
- f_BSE = (D.f[DIR_MPP ])[ktnw ];
- f_TSW = (D.f[DIR_PPM ])[kbne ];
- f_TNE = (D.f[DIR_MMM ])[kbsw ];
- f_TNW = (D.f[DIR_PMM ])[kbse ];
- f_TSE = (D.f[DIR_MPM ])[kbnw ];
+ f_W = (D.f[DIR_P00])[ke ];
+ f_E = (D.f[DIR_M00])[kw ];
+ f_S = (D.f[DIR_0P0])[kn ];
+ f_N = (D.f[DIR_0M0])[ks ];
+ f_B = (D.f[DIR_00P])[kt ];
+ f_T = (D.f[DIR_00M])[kb ];
+ f_SW = (D.f[DIR_PP0])[kne ];
+ f_NE = (D.f[DIR_MM0])[ksw ];
+ f_NW = (D.f[DIR_PM0])[kse ];
+ f_SE = (D.f[DIR_MP0])[knw ];
+ f_BW = (D.f[DIR_P0P])[kte ];
+ f_TE = (D.f[DIR_M0M])[kbw ];
+ f_TW = (D.f[DIR_P0M])[kbe ];
+ f_BE = (D.f[DIR_M0P])[ktw ];
+ f_BS = (D.f[DIR_0PP])[ktn ];
+ f_TN = (D.f[DIR_0MM])[kbs ];
+ f_TS = (D.f[DIR_0PM])[kbn ];
+ f_BN = (D.f[DIR_0MP])[kts ];
+ f_BSW = (D.f[DIR_PPP])[ktne ];
+ f_BNE = (D.f[DIR_MMP])[ktsw ];
+ f_BNW = (D.f[DIR_PMP])[ktse ];
+ f_BSE = (D.f[DIR_MPP])[ktnw ];
+ f_TSW = (D.f[DIR_PPM])[kbne ];
+ f_TNE = (D.f[DIR_MMM])[kbsw ];
+ f_TNW = (D.f[DIR_PMM])[kbse ];
+ f_TSE = (D.f[DIR_MPM])[kbnw ];
////////////////////////////////////////////////////////////////////////////////
real vx1, vx2, vx3, drho;
@@ -1908,63 +1911,63 @@ __global__ void BBStressPressureDevice27( real* DD,
//////////////////////////////////////////////////////////////////////////
if (isEvenTimestep==false)
{
- D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
+ D.f[DIR_M00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_MPM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_PMP * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
real f_E_in, f_W_in, f_N_in, f_S_in, f_T_in, f_B_in, f_NE_in, f_SW_in, f_SE_in, f_NW_in, f_TE_in, f_BW_in, f_BE_in,
diff --git a/src/gpu/VirtualFluids_GPU/GPU/ThinWallBCs27.cu b/src/gpu/VirtualFluids_GPU/GPU/ThinWallBCs27.cu
index 55f810628f370976289d1492e9916d5d3fa0dbb8..b96d961c9b92ae5d041beeb23482d7144e7a8acb 100644
--- a/src/gpu/VirtualFluids_GPU/GPU/ThinWallBCs27.cu
+++ b/src/gpu/VirtualFluids_GPU/GPU/ThinWallBCs27.cu
@@ -27,69 +27,69 @@ __global__ void QVelDeviceCompThinWallsPartOne27(
uint* neighborX,
uint* neighborY,
uint* neighborZ,
- uint size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep)
{
Distributions27 D;
if (isEvenTimestep==true)
{
- D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
+ D.f[DIR_M00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_MPM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_PMP * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////
const unsigned x = threadIdx.x; // Globaler x-Index
@@ -114,24 +114,24 @@ __global__ void QVelDeviceCompThinWallsPartOne27(
*q_dirBE, *q_dirTW, *q_dirTN, *q_dirBS, *q_dirBN, *q_dirTS,
*q_dirTNE, *q_dirTSW, *q_dirTSE, *q_dirTNW, *q_dirBNE, *q_dirBSW,
*q_dirBSE, *q_dirBNW;
- q_dirE = &QQ[DIR_P00 * numberOfBCnodes];
- q_dirW = &QQ[DIR_M00 * numberOfBCnodes];
- q_dirN = &QQ[DIR_0P0 * numberOfBCnodes];
- q_dirS = &QQ[DIR_0M0 * numberOfBCnodes];
- q_dirT = &QQ[DIR_00P * numberOfBCnodes];
- q_dirB = &QQ[DIR_00M * numberOfBCnodes];
- q_dirNE = &QQ[DIR_PP0 * numberOfBCnodes];
- q_dirSW = &QQ[DIR_MM0 * numberOfBCnodes];
- q_dirSE = &QQ[DIR_PM0 * numberOfBCnodes];
- q_dirNW = &QQ[DIR_MP0 * numberOfBCnodes];
- q_dirTE = &QQ[DIR_P0P * numberOfBCnodes];
- q_dirBW = &QQ[DIR_M0M * numberOfBCnodes];
- q_dirBE = &QQ[DIR_P0M * numberOfBCnodes];
- q_dirTW = &QQ[DIR_M0P * numberOfBCnodes];
- q_dirTN = &QQ[DIR_0PP * numberOfBCnodes];
- q_dirBS = &QQ[DIR_0MM * numberOfBCnodes];
- q_dirBN = &QQ[DIR_0PM * numberOfBCnodes];
- q_dirTS = &QQ[DIR_0MP * numberOfBCnodes];
+ q_dirE = &QQ[DIR_P00 * numberOfBCnodes];
+ q_dirW = &QQ[DIR_M00 * numberOfBCnodes];
+ q_dirN = &QQ[DIR_0P0 * numberOfBCnodes];
+ q_dirS = &QQ[DIR_0M0 * numberOfBCnodes];
+ q_dirT = &QQ[DIR_00P * numberOfBCnodes];
+ q_dirB = &QQ[DIR_00M * numberOfBCnodes];
+ q_dirNE = &QQ[DIR_PP0 * numberOfBCnodes];
+ q_dirSW = &QQ[DIR_MM0 * numberOfBCnodes];
+ q_dirSE = &QQ[DIR_PM0 * numberOfBCnodes];
+ q_dirNW = &QQ[DIR_MP0 * numberOfBCnodes];
+ q_dirTE = &QQ[DIR_P0P * numberOfBCnodes];
+ q_dirBW = &QQ[DIR_M0M * numberOfBCnodes];
+ q_dirBE = &QQ[DIR_P0M * numberOfBCnodes];
+ q_dirTW = &QQ[DIR_M0P * numberOfBCnodes];
+ q_dirTN = &QQ[DIR_0PP * numberOfBCnodes];
+ q_dirBS = &QQ[DIR_0MM * numberOfBCnodes];
+ q_dirBN = &QQ[DIR_0PM * numberOfBCnodes];
+ q_dirTS = &QQ[DIR_0MP * numberOfBCnodes];
q_dirTNE = &QQ[DIR_PPP * numberOfBCnodes];
q_dirTSW = &QQ[DIR_MMP * numberOfBCnodes];
q_dirTSE = &QQ[DIR_PMP * numberOfBCnodes];
@@ -174,32 +174,32 @@ __global__ void QVelDeviceCompThinWallsPartOne27(
real f_E, f_W, f_N, f_S, f_T, f_B, f_NE, f_SW, f_SE, f_NW, f_TE, f_BW, f_BE,
f_TW, f_TN, f_BS, f_BN, f_TS, f_TNE, f_TSW, f_TSE, f_TNW, f_BNE, f_BSW, f_BSE, f_BNW;
- f_W = (D.f[DIR_P00 ])[ke ];
- f_E = (D.f[DIR_M00 ])[kw ];
- f_S = (D.f[DIR_0P0 ])[kn ];
- f_N = (D.f[DIR_0M0 ])[ks ];
- f_B = (D.f[DIR_00P ])[kt ];
- f_T = (D.f[DIR_00M ])[kb ];
- f_SW = (D.f[DIR_PP0 ])[kne ];
- f_NE = (D.f[DIR_MM0 ])[ksw ];
- f_NW = (D.f[DIR_PM0 ])[kse ];
- f_SE = (D.f[DIR_MP0 ])[knw ];
- f_BW = (D.f[DIR_P0P ])[kte ];
- f_TE = (D.f[DIR_M0M ])[kbw ];
- f_TW = (D.f[DIR_P0M ])[kbe ];
- f_BE = (D.f[DIR_M0P ])[ktw ];
- f_BS = (D.f[DIR_0PP ])[ktn ];
- f_TN = (D.f[DIR_0MM ])[kbs ];
- f_TS = (D.f[DIR_0PM ])[kbn ];
- f_BN = (D.f[DIR_0MP ])[kts ];
- f_BSW = (D.f[DIR_PPP ])[ktne ];
- f_BNE = (D.f[DIR_MMP ])[ktsw ];
- f_BNW = (D.f[DIR_PMP ])[ktse ];
- f_BSE = (D.f[DIR_MPP ])[ktnw ];
- f_TSW = (D.f[DIR_PPM ])[kbne ];
- f_TNE = (D.f[DIR_MMM ])[kbsw ];
- f_TNW = (D.f[DIR_PMM ])[kbse ];
- f_TSE = (D.f[DIR_MPM ])[kbnw ];
+ f_W = (D.f[DIR_P00])[ke ];
+ f_E = (D.f[DIR_M00])[kw ];
+ f_S = (D.f[DIR_0P0])[kn ];
+ f_N = (D.f[DIR_0M0])[ks ];
+ f_B = (D.f[DIR_00P])[kt ];
+ f_T = (D.f[DIR_00M])[kb ];
+ f_SW = (D.f[DIR_PP0])[kne ];
+ f_NE = (D.f[DIR_MM0])[ksw ];
+ f_NW = (D.f[DIR_PM0])[kse ];
+ f_SE = (D.f[DIR_MP0])[knw ];
+ f_BW = (D.f[DIR_P0P])[kte ];
+ f_TE = (D.f[DIR_M0M])[kbw ];
+ f_TW = (D.f[DIR_P0M])[kbe ];
+ f_BE = (D.f[DIR_M0P])[ktw ];
+ f_BS = (D.f[DIR_0PP])[ktn ];
+ f_TN = (D.f[DIR_0MM])[kbs ];
+ f_TS = (D.f[DIR_0PM])[kbn ];
+ f_BN = (D.f[DIR_0MP])[kts ];
+ f_BSW = (D.f[DIR_PPP])[ktne ];
+ f_BNE = (D.f[DIR_MMP])[ktsw ];
+ f_BNW = (D.f[DIR_PMP])[ktse ];
+ f_BSE = (D.f[DIR_MPP])[ktnw ];
+ f_TSW = (D.f[DIR_PPM])[kbne ];
+ f_TNE = (D.f[DIR_MMM])[kbsw ];
+ f_TNW = (D.f[DIR_PMM])[kbse ];
+ f_TSE = (D.f[DIR_MPM])[kbnw ];
////////////////////////////////////////////////////////////////////////////////
real vx1, vx2, vx3, drho, feq, q;
drho = f_TSE + f_TNW + f_TNE + f_TSW + f_BSE + f_BNW + f_BNE + f_BSW +
@@ -456,69 +456,69 @@ __global__ void QDeviceCompThinWallsPartOne27(
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep)
{
Distributions27 D;
if (isEvenTimestep == true)
{
- D.f[DIR_P00] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_M00] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0P0] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0M0] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00P] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00M] = &DD[DIR_00M *size_Mat];
- D.f[DIR_PP0] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_MM0] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_PM0] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_MP0] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_P0P] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_M0M] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_P0M] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_M0P] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0PP] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0MM] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0PM] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0MP] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_MMP] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_PMP] = &DD[DIR_PMP *size_Mat];
- D.f[DIR_MPP] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_PPM] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_MMM] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_PMM] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_MPM] = &DD[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_P00] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0M0] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0P0] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00M] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00P] = &DD[DIR_00M *size_Mat];
- D.f[DIR_MM0] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_PP0] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_MP0] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_PM0] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_M0M] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_P0P] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_M0P] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_P0M] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0MM] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0PP] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0MP] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0PM] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_MMP] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_PMP] = &DD[DIR_MPM *size_Mat];
- D.f[DIR_MPP] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_PPM] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_MMM] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_PMM] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_MPM] = &DD[DIR_PMP *size_Mat];
+ D.f[DIR_M00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_MPM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_PMP * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////
const unsigned x = threadIdx.x; // Globaler x-Index
@@ -539,24 +539,24 @@ __global__ void QDeviceCompThinWallsPartOne27(
*q_dirBE, *q_dirTW, *q_dirTN, *q_dirBS, *q_dirBN, *q_dirTS,
*q_dirTNE, *q_dirTSW, *q_dirTSE, *q_dirTNW, *q_dirBNE, *q_dirBSW,
*q_dirBSE, *q_dirBNW;
- q_dirE = &QQ[DIR_P00 * numberOfBCnodes];
- q_dirW = &QQ[DIR_M00 * numberOfBCnodes];
- q_dirN = &QQ[DIR_0P0 * numberOfBCnodes];
- q_dirS = &QQ[DIR_0M0 * numberOfBCnodes];
- q_dirT = &QQ[DIR_00P * numberOfBCnodes];
- q_dirB = &QQ[DIR_00M * numberOfBCnodes];
- q_dirNE = &QQ[DIR_PP0 * numberOfBCnodes];
- q_dirSW = &QQ[DIR_MM0 * numberOfBCnodes];
- q_dirSE = &QQ[DIR_PM0 * numberOfBCnodes];
- q_dirNW = &QQ[DIR_MP0 * numberOfBCnodes];
- q_dirTE = &QQ[DIR_P0P * numberOfBCnodes];
- q_dirBW = &QQ[DIR_M0M * numberOfBCnodes];
- q_dirBE = &QQ[DIR_P0M * numberOfBCnodes];
- q_dirTW = &QQ[DIR_M0P * numberOfBCnodes];
- q_dirTN = &QQ[DIR_0PP * numberOfBCnodes];
- q_dirBS = &QQ[DIR_0MM * numberOfBCnodes];
- q_dirBN = &QQ[DIR_0PM * numberOfBCnodes];
- q_dirTS = &QQ[DIR_0MP * numberOfBCnodes];
+ q_dirE = &QQ[DIR_P00 * numberOfBCnodes];
+ q_dirW = &QQ[DIR_M00 * numberOfBCnodes];
+ q_dirN = &QQ[DIR_0P0 * numberOfBCnodes];
+ q_dirS = &QQ[DIR_0M0 * numberOfBCnodes];
+ q_dirT = &QQ[DIR_00P * numberOfBCnodes];
+ q_dirB = &QQ[DIR_00M * numberOfBCnodes];
+ q_dirNE = &QQ[DIR_PP0 * numberOfBCnodes];
+ q_dirSW = &QQ[DIR_MM0 * numberOfBCnodes];
+ q_dirSE = &QQ[DIR_PM0 * numberOfBCnodes];
+ q_dirNW = &QQ[DIR_MP0 * numberOfBCnodes];
+ q_dirTE = &QQ[DIR_P0P * numberOfBCnodes];
+ q_dirBW = &QQ[DIR_M0M * numberOfBCnodes];
+ q_dirBE = &QQ[DIR_P0M * numberOfBCnodes];
+ q_dirTW = &QQ[DIR_M0P * numberOfBCnodes];
+ q_dirTN = &QQ[DIR_0PP * numberOfBCnodes];
+ q_dirBS = &QQ[DIR_0MM * numberOfBCnodes];
+ q_dirBN = &QQ[DIR_0PM * numberOfBCnodes];
+ q_dirTS = &QQ[DIR_0MP * numberOfBCnodes];
q_dirTNE = &QQ[DIR_PPP * numberOfBCnodes];
q_dirTSW = &QQ[DIR_MMP * numberOfBCnodes];
q_dirTSE = &QQ[DIR_PMP * numberOfBCnodes];
@@ -882,7 +882,7 @@ __global__ void QThinWallsPartTwo27(
uint* neighborY,
uint* neighborZ,
uint* neighborWSB,
- uint size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep)
{
////////////////////////////////////////////////////////////////////////////////
@@ -904,24 +904,24 @@ __global__ void QThinWallsPartTwo27(
*q_dirBE, *q_dirTW, *q_dirTN, *q_dirBS, *q_dirBN, *q_dirTS,
*q_dirTNE, *q_dirTSW, *q_dirTSE, *q_dirTNW, *q_dirBNE, *q_dirBSW,
*q_dirBSE, *q_dirBNW;
- q_dirE = &QQ[DIR_P00 * numberOfBCnodes];
- q_dirW = &QQ[DIR_M00 * numberOfBCnodes];
- q_dirN = &QQ[DIR_0P0 * numberOfBCnodes];
- q_dirS = &QQ[DIR_0M0 * numberOfBCnodes];
- q_dirT = &QQ[DIR_00P * numberOfBCnodes];
- q_dirB = &QQ[DIR_00M * numberOfBCnodes];
- q_dirNE = &QQ[DIR_PP0 * numberOfBCnodes];
- q_dirSW = &QQ[DIR_MM0 * numberOfBCnodes];
- q_dirSE = &QQ[DIR_PM0 * numberOfBCnodes];
- q_dirNW = &QQ[DIR_MP0 * numberOfBCnodes];
- q_dirTE = &QQ[DIR_P0P * numberOfBCnodes];
- q_dirBW = &QQ[DIR_M0M * numberOfBCnodes];
- q_dirBE = &QQ[DIR_P0M * numberOfBCnodes];
- q_dirTW = &QQ[DIR_M0P * numberOfBCnodes];
- q_dirTN = &QQ[DIR_0PP * numberOfBCnodes];
- q_dirBS = &QQ[DIR_0MM * numberOfBCnodes];
- q_dirBN = &QQ[DIR_0PM * numberOfBCnodes];
- q_dirTS = &QQ[DIR_0MP * numberOfBCnodes];
+ q_dirE = &QQ[DIR_P00 * numberOfBCnodes];
+ q_dirW = &QQ[DIR_M00 * numberOfBCnodes];
+ q_dirN = &QQ[DIR_0P0 * numberOfBCnodes];
+ q_dirS = &QQ[DIR_0M0 * numberOfBCnodes];
+ q_dirT = &QQ[DIR_00P * numberOfBCnodes];
+ q_dirB = &QQ[DIR_00M * numberOfBCnodes];
+ q_dirNE = &QQ[DIR_PP0 * numberOfBCnodes];
+ q_dirSW = &QQ[DIR_MM0 * numberOfBCnodes];
+ q_dirSE = &QQ[DIR_PM0 * numberOfBCnodes];
+ q_dirNW = &QQ[DIR_MP0 * numberOfBCnodes];
+ q_dirTE = &QQ[DIR_P0P * numberOfBCnodes];
+ q_dirBW = &QQ[DIR_M0M * numberOfBCnodes];
+ q_dirBE = &QQ[DIR_P0M * numberOfBCnodes];
+ q_dirTW = &QQ[DIR_M0P * numberOfBCnodes];
+ q_dirTN = &QQ[DIR_0PP * numberOfBCnodes];
+ q_dirBS = &QQ[DIR_0MM * numberOfBCnodes];
+ q_dirBN = &QQ[DIR_0PM * numberOfBCnodes];
+ q_dirTS = &QQ[DIR_0MP * numberOfBCnodes];
q_dirTNE = &QQ[DIR_PPP * numberOfBCnodes];
q_dirTSW = &QQ[DIR_MMP * numberOfBCnodes];
q_dirTSE = &QQ[DIR_PMP * numberOfBCnodes];
@@ -980,123 +980,123 @@ __global__ void QThinWallsPartTwo27(
Distributions27 D, DN;
if (isEvenTimestep == true)
{
- D.f[DIR_P00] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_M00] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0P0] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0M0] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00P] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00M] = &DD[DIR_00M *size_Mat];
- D.f[DIR_PP0] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_MM0] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_PM0] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_MP0] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_P0P] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_M0M] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_P0M] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_M0P] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0PP] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0MM] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0PM] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0MP] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_MMP] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_PMP] = &DD[DIR_PMP *size_Mat];
- D.f[DIR_MPP] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_PPM] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_MMM] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_PMM] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_MPM] = &DD[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_P00] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0M0] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0P0] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00M] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00P] = &DD[DIR_00M *size_Mat];
- D.f[DIR_MM0] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_PP0] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_MP0] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_PM0] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_M0M] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_P0P] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_M0P] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_P0M] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0MM] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0PP] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0MP] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0PM] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_MMP] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_PMP] = &DD[DIR_MPM *size_Mat];
- D.f[DIR_MPP] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_PPM] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_MMM] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_PMM] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_MPM] = &DD[DIR_PMP *size_Mat];
+ D.f[DIR_M00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_MPM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_PMP * numberOfLBnodes];
}
if (isEvenTimestep==false)
{
- DN.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- DN.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- DN.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- DN.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- DN.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- DN.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- DN.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- DN.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- DN.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- DN.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- DN.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- DN.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- DN.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- DN.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- DN.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- DN.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- DN.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- DN.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- DN.f[DIR_000] = &DD[DIR_000*size_Mat];
- DN.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- DN.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- DN.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- DN.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- DN.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- DN.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- DN.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- DN.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
+ DN.f[DIR_P00] = &DD[DIR_P00 * numberOfLBnodes];
+ DN.f[DIR_M00] = &DD[DIR_M00 * numberOfLBnodes];
+ DN.f[DIR_0P0] = &DD[DIR_0P0 * numberOfLBnodes];
+ DN.f[DIR_0M0] = &DD[DIR_0M0 * numberOfLBnodes];
+ DN.f[DIR_00P] = &DD[DIR_00P * numberOfLBnodes];
+ DN.f[DIR_00M] = &DD[DIR_00M * numberOfLBnodes];
+ DN.f[DIR_PP0] = &DD[DIR_PP0 * numberOfLBnodes];
+ DN.f[DIR_MM0] = &DD[DIR_MM0 * numberOfLBnodes];
+ DN.f[DIR_PM0] = &DD[DIR_PM0 * numberOfLBnodes];
+ DN.f[DIR_MP0] = &DD[DIR_MP0 * numberOfLBnodes];
+ DN.f[DIR_P0P] = &DD[DIR_P0P * numberOfLBnodes];
+ DN.f[DIR_M0M] = &DD[DIR_M0M * numberOfLBnodes];
+ DN.f[DIR_P0M] = &DD[DIR_P0M * numberOfLBnodes];
+ DN.f[DIR_M0P] = &DD[DIR_M0P * numberOfLBnodes];
+ DN.f[DIR_0PP] = &DD[DIR_0PP * numberOfLBnodes];
+ DN.f[DIR_0MM] = &DD[DIR_0MM * numberOfLBnodes];
+ DN.f[DIR_0PM] = &DD[DIR_0PM * numberOfLBnodes];
+ DN.f[DIR_0MP] = &DD[DIR_0MP * numberOfLBnodes];
+ DN.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ DN.f[DIR_PPP] = &DD[DIR_PPP * numberOfLBnodes];
+ DN.f[DIR_MMP] = &DD[DIR_MMP * numberOfLBnodes];
+ DN.f[DIR_PMP] = &DD[DIR_PMP * numberOfLBnodes];
+ DN.f[DIR_MPP] = &DD[DIR_MPP * numberOfLBnodes];
+ DN.f[DIR_PPM] = &DD[DIR_PPM * numberOfLBnodes];
+ DN.f[DIR_MMM] = &DD[DIR_MMM * numberOfLBnodes];
+ DN.f[DIR_PMM] = &DD[DIR_PMM * numberOfLBnodes];
+ DN.f[DIR_MPM] = &DD[DIR_MPM * numberOfLBnodes];
}
else
{
- DN.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
- DN.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
- DN.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
- DN.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
- DN.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
- DN.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
- DN.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
- DN.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
- DN.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
- DN.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
- DN.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
- DN.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
- DN.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
- DN.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
- DN.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
- DN.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
- DN.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
- DN.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
- DN.f[DIR_000] = &DD[DIR_000*size_Mat];
- DN.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
- DN.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
- DN.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
- DN.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
- DN.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
- DN.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
- DN.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
- DN.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
+ DN.f[DIR_M00] = &DD[DIR_P00 * numberOfLBnodes];
+ DN.f[DIR_P00] = &DD[DIR_M00 * numberOfLBnodes];
+ DN.f[DIR_0M0] = &DD[DIR_0P0 * numberOfLBnodes];
+ DN.f[DIR_0P0] = &DD[DIR_0M0 * numberOfLBnodes];
+ DN.f[DIR_00M] = &DD[DIR_00P * numberOfLBnodes];
+ DN.f[DIR_00P] = &DD[DIR_00M * numberOfLBnodes];
+ DN.f[DIR_MM0] = &DD[DIR_PP0 * numberOfLBnodes];
+ DN.f[DIR_PP0] = &DD[DIR_MM0 * numberOfLBnodes];
+ DN.f[DIR_MP0] = &DD[DIR_PM0 * numberOfLBnodes];
+ DN.f[DIR_PM0] = &DD[DIR_MP0 * numberOfLBnodes];
+ DN.f[DIR_M0M] = &DD[DIR_P0P * numberOfLBnodes];
+ DN.f[DIR_P0P] = &DD[DIR_M0M * numberOfLBnodes];
+ DN.f[DIR_M0P] = &DD[DIR_P0M * numberOfLBnodes];
+ DN.f[DIR_P0M] = &DD[DIR_M0P * numberOfLBnodes];
+ DN.f[DIR_0MM] = &DD[DIR_0PP * numberOfLBnodes];
+ DN.f[DIR_0PP] = &DD[DIR_0MM * numberOfLBnodes];
+ DN.f[DIR_0MP] = &DD[DIR_0PM * numberOfLBnodes];
+ DN.f[DIR_0PM] = &DD[DIR_0MP * numberOfLBnodes];
+ DN.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ DN.f[DIR_PPP] = &DD[DIR_MMM * numberOfLBnodes];
+ DN.f[DIR_MMP] = &DD[DIR_PPM * numberOfLBnodes];
+ DN.f[DIR_PMP] = &DD[DIR_MPM * numberOfLBnodes];
+ DN.f[DIR_MPP] = &DD[DIR_PMM * numberOfLBnodes];
+ DN.f[DIR_PPM] = &DD[DIR_MMP * numberOfLBnodes];
+ DN.f[DIR_MMM] = &DD[DIR_PPP * numberOfLBnodes];
+ DN.f[DIR_PMM] = &DD[DIR_MPP * numberOfLBnodes];
+ DN.f[DIR_MPM] = &DD[DIR_PMP * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
//directions allways exchange
@@ -1106,24 +1106,24 @@ __global__ void QThinWallsPartTwo27(
//( 1 1 1) ( 1 0 0) ( 0 1 0) ( 0 0 1) ( 1 1 0) ( 1 0 1) ( 0 1 1) (-1 -1 1) (-1 1 -1) ( 1 -1 -1) (-1 1 0) (-1 0 1) ( 0 -1 1)
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
real q, tmp;
- q = q_dirE[k]; if (q>=c0o1 && q<=c1o1){ if (geom[kw ] < GEO_FLUID){tmp = (DN.f[DIR_M00 ])[kw ]; (DN.f[DIR_M00 ])[kw ]=(D.f[DIR_M00 ])[kw ]; (D.f[DIR_M00 ])[kw ]=tmp;}}
- q = q_dirW[k]; if (q>=c0o1 && q<=c1o1){ {tmp = (DN.f[DIR_P00 ])[ke ]; (DN.f[DIR_P00 ])[ke ]=(D.f[DIR_P00 ])[ke ]; (D.f[DIR_P00 ])[ke ]=tmp;}}
- q = q_dirN[k]; if (q>=c0o1 && q<=c1o1){ if (geom[ks ] < GEO_FLUID){tmp = (DN.f[DIR_0M0 ])[ks ]; (DN.f[DIR_0M0 ])[ks ]=(D.f[DIR_0M0 ])[ks ]; (D.f[DIR_0M0 ])[ks ]=tmp;}}
- q = q_dirS[k]; if (q>=c0o1 && q<=c1o1){ {tmp = (DN.f[DIR_0P0 ])[kn ]; (DN.f[DIR_0P0 ])[kn ]=(D.f[DIR_0P0 ])[kn ]; (D.f[DIR_0P0 ])[kn ]=tmp;}}
- q = q_dirT[k]; if (q>=c0o1 && q<=c1o1){ if (geom[kb ] < GEO_FLUID){tmp = (DN.f[DIR_00M ])[kb ]; (DN.f[DIR_00M ])[kb ]=(D.f[DIR_00M ])[kb ]; (D.f[DIR_00M ])[kb ]=tmp;}}
- q = q_dirB[k]; if (q>=c0o1 && q<=c1o1){ {tmp = (DN.f[DIR_00P ])[kt ]; (DN.f[DIR_00P ])[kt ]=(D.f[DIR_00P ])[kt ]; (D.f[DIR_00P ])[kt ]=tmp;}}
- q = q_dirNE[k]; if (q>=c0o1 && q<=c1o1){ if (geom[ksw ] < GEO_FLUID){tmp = (DN.f[DIR_MM0 ])[ksw ]; (DN.f[DIR_MM0 ])[ksw ]=(D.f[DIR_MM0 ])[ksw ]; (D.f[DIR_MM0 ])[ksw ]=tmp;}}
- q = q_dirSW[k]; if (q>=c0o1 && q<=c1o1){ {tmp = (DN.f[DIR_PP0 ])[kne ]; (DN.f[DIR_PP0 ])[kne ]=(D.f[DIR_PP0 ])[kne ]; (D.f[DIR_PP0 ])[kne ]=tmp;}}
- q = q_dirSE[k]; if (q>=c0o1 && q<=c1o1){ {tmp = (DN.f[DIR_MP0 ])[knw ]; (DN.f[DIR_MP0 ])[knw ]=(D.f[DIR_MP0 ])[knw ]; (D.f[DIR_MP0 ])[knw ]=tmp;}}
- q = q_dirNW[k]; if (q>=c0o1 && q<=c1o1){ if (geom[kmp0] < GEO_FLUID){tmp = (DN.f[DIR_PM0 ])[kse ]; (DN.f[DIR_PM0 ])[kse ]=(D.f[DIR_PM0 ])[kse ]; (D.f[DIR_PM0 ])[kse ]=tmp;}}
- q = q_dirTE[k]; if (q>=c0o1 && q<=c1o1){ if (geom[kbw ] < GEO_FLUID){tmp = (DN.f[DIR_M0M ])[kbw ]; (DN.f[DIR_M0M ])[kbw ]=(D.f[DIR_M0M ])[kbw ]; (D.f[DIR_M0M ])[kbw ]=tmp;}}
- q = q_dirBW[k]; if (q>=c0o1 && q<=c1o1){ {tmp = (DN.f[DIR_P0P ])[kte ]; (DN.f[DIR_P0P ])[kte ]=(D.f[DIR_P0P ])[kte ]; (D.f[DIR_P0P ])[kte ]=tmp;}}
- q = q_dirBE[k]; if (q>=c0o1 && q<=c1o1){ {tmp = (DN.f[DIR_M0P ])[ktw ]; (DN.f[DIR_M0P ])[ktw ]=(D.f[DIR_M0P ])[ktw ]; (D.f[DIR_M0P ])[ktw ]=tmp;}}
- q = q_dirTW[k]; if (q>=c0o1 && q<=c1o1){ if (geom[km0p] < GEO_FLUID){tmp = (DN.f[DIR_P0M ])[kbe ]; (DN.f[DIR_P0M ])[kbe ]=(D.f[DIR_P0M ])[kbe ]; (D.f[DIR_P0M ])[kbe ]=tmp;}}
- q = q_dirTN[k]; if (q>=c0o1 && q<=c1o1){ if (geom[kbs ] < GEO_FLUID){tmp = (DN.f[DIR_0MM ])[kbs ]; (DN.f[DIR_0MM ])[kbs ]=(D.f[DIR_0MM ])[kbs ]; (D.f[DIR_0MM ])[kbs ]=tmp;}}
- q = q_dirBS[k]; if (q>=c0o1 && q<=c1o1){ {tmp = (DN.f[DIR_0PP ])[ktn ]; (DN.f[DIR_0PP ])[ktn ]=(D.f[DIR_0PP ])[ktn ]; (D.f[DIR_0PP ])[ktn ]=tmp;}}
- q = q_dirBN[k]; if (q>=c0o1 && q<=c1o1){ {tmp = (DN.f[DIR_0MP ])[kts ]; (DN.f[DIR_0MP ])[kts ]=(D.f[DIR_0MP ])[kts ]; (D.f[DIR_0MP ])[kts ]=tmp;}}
- q = q_dirTS[k]; if (q>=c0o1 && q<=c1o1){ if (geom[k0mp] < GEO_FLUID){tmp = (DN.f[DIR_0PM ])[kbn ]; (DN.f[DIR_0PM ])[kbn ]=(D.f[DIR_0PM ])[kbn ]; (D.f[DIR_0PM ])[kbn ]=tmp;}}
+ q = q_dirE[k]; if (q>=c0o1 && q<=c1o1){ if (geom[kw ] < GEO_FLUID){tmp = (DN.f[DIR_M00])[kw ]; (DN.f[DIR_M00])[kw ]=(D.f[DIR_M00])[kw ]; (D.f[DIR_M00])[kw ]=tmp;}}
+ q = q_dirW[k]; if (q>=c0o1 && q<=c1o1){ {tmp = (DN.f[DIR_P00])[ke ]; (DN.f[DIR_P00])[ke ]=(D.f[DIR_P00])[ke ]; (D.f[DIR_P00])[ke ]=tmp;}}
+ q = q_dirN[k]; if (q>=c0o1 && q<=c1o1){ if (geom[ks ] < GEO_FLUID){tmp = (DN.f[DIR_0M0])[ks ]; (DN.f[DIR_0M0])[ks ]=(D.f[DIR_0M0])[ks ]; (D.f[DIR_0M0])[ks ]=tmp;}}
+ q = q_dirS[k]; if (q>=c0o1 && q<=c1o1){ {tmp = (DN.f[DIR_0P0])[kn ]; (DN.f[DIR_0P0])[kn ]=(D.f[DIR_0P0])[kn ]; (D.f[DIR_0P0])[kn ]=tmp;}}
+ q = q_dirT[k]; if (q>=c0o1 && q<=c1o1){ if (geom[kb ] < GEO_FLUID){tmp = (DN.f[DIR_00M])[kb ]; (DN.f[DIR_00M])[kb ]=(D.f[DIR_00M])[kb ]; (D.f[DIR_00M])[kb ]=tmp;}}
+ q = q_dirB[k]; if (q>=c0o1 && q<=c1o1){ {tmp = (DN.f[DIR_00P])[kt ]; (DN.f[DIR_00P])[kt ]=(D.f[DIR_00P])[kt ]; (D.f[DIR_00P])[kt ]=tmp;}}
+ q = q_dirNE[k]; if (q>=c0o1 && q<=c1o1){ if (geom[ksw ] < GEO_FLUID){tmp = (DN.f[DIR_MM0])[ksw ]; (DN.f[DIR_MM0])[ksw ]=(D.f[DIR_MM0])[ksw ]; (D.f[DIR_MM0])[ksw ]=tmp;}}
+ q = q_dirSW[k]; if (q>=c0o1 && q<=c1o1){ {tmp = (DN.f[DIR_PP0])[kne ]; (DN.f[DIR_PP0])[kne ]=(D.f[DIR_PP0])[kne ]; (D.f[DIR_PP0])[kne ]=tmp;}}
+ q = q_dirSE[k]; if (q>=c0o1 && q<=c1o1){ {tmp = (DN.f[DIR_MP0])[knw ]; (DN.f[DIR_MP0])[knw ]=(D.f[DIR_MP0])[knw ]; (D.f[DIR_MP0])[knw ]=tmp;}}
+ q = q_dirNW[k]; if (q>=c0o1 && q<=c1o1){ if (geom[kmp0] < GEO_FLUID){tmp = (DN.f[DIR_PM0])[kse ]; (DN.f[DIR_PM0])[kse ]=(D.f[DIR_PM0])[kse ]; (D.f[DIR_PM0])[kse ]=tmp;}}
+ q = q_dirTE[k]; if (q>=c0o1 && q<=c1o1){ if (geom[kbw ] < GEO_FLUID){tmp = (DN.f[DIR_M0M])[kbw ]; (DN.f[DIR_M0M])[kbw ]=(D.f[DIR_M0M])[kbw ]; (D.f[DIR_M0M])[kbw ]=tmp;}}
+ q = q_dirBW[k]; if (q>=c0o1 && q<=c1o1){ {tmp = (DN.f[DIR_P0P])[kte ]; (DN.f[DIR_P0P])[kte ]=(D.f[DIR_P0P])[kte ]; (D.f[DIR_P0P])[kte ]=tmp;}}
+ q = q_dirBE[k]; if (q>=c0o1 && q<=c1o1){ {tmp = (DN.f[DIR_M0P])[ktw ]; (DN.f[DIR_M0P])[ktw ]=(D.f[DIR_M0P])[ktw ]; (D.f[DIR_M0P])[ktw ]=tmp;}}
+ q = q_dirTW[k]; if (q>=c0o1 && q<=c1o1){ if (geom[km0p] < GEO_FLUID){tmp = (DN.f[DIR_P0M])[kbe ]; (DN.f[DIR_P0M])[kbe ]=(D.f[DIR_P0M])[kbe ]; (D.f[DIR_P0M])[kbe ]=tmp;}}
+ q = q_dirTN[k]; if (q>=c0o1 && q<=c1o1){ if (geom[kbs ] < GEO_FLUID){tmp = (DN.f[DIR_0MM])[kbs ]; (DN.f[DIR_0MM])[kbs ]=(D.f[DIR_0MM])[kbs ]; (D.f[DIR_0MM])[kbs ]=tmp;}}
+ q = q_dirBS[k]; if (q>=c0o1 && q<=c1o1){ {tmp = (DN.f[DIR_0PP])[ktn ]; (DN.f[DIR_0PP])[ktn ]=(D.f[DIR_0PP])[ktn ]; (D.f[DIR_0PP])[ktn ]=tmp;}}
+ q = q_dirBN[k]; if (q>=c0o1 && q<=c1o1){ {tmp = (DN.f[DIR_0MP])[kts ]; (DN.f[DIR_0MP])[kts ]=(D.f[DIR_0MP])[kts ]; (D.f[DIR_0MP])[kts ]=tmp;}}
+ q = q_dirTS[k]; if (q>=c0o1 && q<=c1o1){ if (geom[k0mp] < GEO_FLUID){tmp = (DN.f[DIR_0PM])[kbn ]; (DN.f[DIR_0PM])[kbn ]=(D.f[DIR_0PM])[kbn ]; (D.f[DIR_0PM])[kbn ]=tmp;}}
q = q_dirTNE[k]; if (q>=c0o1 && q<=c1o1){ if (geom[kbsw] < GEO_FLUID){tmp = (DN.f[DIR_MMM])[kbsw]; (DN.f[DIR_MMM])[kbsw]=(D.f[DIR_MMM])[kbsw]; (D.f[DIR_MMM])[kbsw]=tmp;}}
q = q_dirBSW[k]; if (q>=c0o1 && q<=c1o1){ {tmp = (DN.f[DIR_PPP])[ktne]; (DN.f[DIR_PPP])[ktne]=(D.f[DIR_PPP])[ktne]; (D.f[DIR_PPP])[ktne]=tmp;}}
q = q_dirBNE[k]; if (q>=c0o1 && q<=c1o1){ {tmp = (DN.f[DIR_MMP])[ktsw]; (DN.f[DIR_MMP])[ktsw]=(D.f[DIR_MMP])[ktsw]; (D.f[DIR_MMP])[ktsw]=tmp;}}
diff --git a/src/gpu/VirtualFluids_GPU/GPU/TurbulenceIntensity.cu b/src/gpu/VirtualFluids_GPU/GPU/TurbulenceIntensity.cu
index f8cf8ab13c39d55477bf006cd27f7943dcb5b53a..3f440454ef272b13c24fe2a2882d67d32d32a841 100644
--- a/src/gpu/VirtualFluids_GPU/GPU/TurbulenceIntensity.cu
+++ b/src/gpu/VirtualFluids_GPU/GPU/TurbulenceIntensity.cu
@@ -9,14 +9,16 @@
/* Device code */
#include "LBM/LB.h"
#include "lbm/constants/D3Q27.h"
-#include <lbm/constants/NumericConstants.h>
+#include "lbm/constants/NumericConstants.h"
#include "lbm/MacroscopicQuantities.h"
#include "../Kernel/Utilities/DistributionHelper.cuh"
+#include "LBM/GPUHelperFunctions/KernelUtilities.h"
using namespace vf::lbm::constant;
using namespace vf::lbm::dir;
+using namespace vf::gpu;
//////////////////////////////////////////////////////////////////////////////
__global__ void CalcTurbulenceIntensity(
@@ -34,19 +36,21 @@ __global__ void CalcTurbulenceIntensity(
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep)
{
- const unsigned k = vf::gpu::getNodeIndex();
+ ////////////////////////////////////////////////////////////////////////////////
+ //! - Get node index coordinates from threadIdx, blockIdx, blockDim and gridDim.
+ //!
+ const unsigned nodeIndex = getNodeIndex();
- if (k >= size_Mat)
+ if (nodeIndex >= numberOfLBnodes)
return;
- if (!vf::gpu::isValidFluidNode(typeOfGridNode[k]))
+ if (!isValidFluidNode(typeOfGridNode[nodeIndex]))
return;
- vf::gpu::DistributionWrapper distr_wrapper(distributions, size_Mat, isEvenTimestep, k, neighborX, neighborY,
- neighborZ);
+ DistributionWrapper distr_wrapper(distributions, numberOfLBnodes, isEvenTimestep, nodeIndex, neighborX, neighborY, neighborZ);
const auto &distribution = distr_wrapper.distribution;
// analogue to LBCalcMacCompSP27
@@ -58,16 +62,16 @@ __global__ void CalcTurbulenceIntensity(
// compute subtotals:
// fluctuations
- vxx[k] = vxx[k] + vx * vx;
- vyy[k] = vyy[k] + vy * vy;
- vzz[k] = vzz[k] + vz * vz;
- vxy[k] = vxy[k] + vx * vy;
- vxz[k] = vxz[k] + vx * vz;
- vyz[k] = vyz[k] + vy * vz;
+ vxx[nodeIndex] = vxx[nodeIndex] + vx * vx;
+ vyy[nodeIndex] = vyy[nodeIndex] + vy * vy;
+ vzz[nodeIndex] = vzz[nodeIndex] + vz * vz;
+ vxy[nodeIndex] = vxy[nodeIndex] + vx * vy;
+ vxz[nodeIndex] = vxz[nodeIndex] + vx * vz;
+ vyz[nodeIndex] = vyz[nodeIndex] + vy * vz;
// velocity (for mean velocity)
- vx_mean[k] = vx_mean[k] + vx;
- vy_mean[k] = vy_mean[k] + vy;
- vz_mean[k] = vz_mean[k] + vz;
+ vx_mean[nodeIndex] = vx_mean[nodeIndex] + vx;
+ vy_mean[nodeIndex] = vy_mean[nodeIndex] + vy;
+ vz_mean[nodeIndex] = vz_mean[nodeIndex] + vz;
}
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
diff --git a/src/gpu/VirtualFluids_GPU/GPU/TurbulentViscosityInlines.cuh b/src/gpu/VirtualFluids_GPU/GPU/TurbulentViscosityInlines.cuh
index eb301515527a9e8a3056676b0d4dffe8197c7dbe..58856f624fa1dfd2488c3061721e9dac53a67d07 100644
--- a/src/gpu/VirtualFluids_GPU/GPU/TurbulentViscosityInlines.cuh
+++ b/src/gpu/VirtualFluids_GPU/GPU/TurbulentViscosityInlines.cuh
@@ -53,8 +53,8 @@ __inline__ __device__ real calcTurbulentViscosityQR(real C, real dxux, real dyuy
//! Second invariant of the strain-rate tensor
real Q = c1o2*( dxux*dxux + dyuy*dyuy + dzuz*dzuz ) + c1o4*( Dxy*Dxy + Dxz*Dxz + Dyz*Dyz);
//! Third invariant of the strain-rate tensor (determinant)
- real R = - dxux*dyuy*dzuz - c1o4*( Dxy*Dxz*Dyz + dxux*Dyz*Dyz + dyuy*Dxz*Dxz + dzuz*Dxy*Dxy );
-
+ // real R = - dxux*dyuy*dzuz - c1o4*( Dxy*Dxz*Dyz + dxux*Dyz*Dyz + dyuy*Dxz*Dxz + dzuz*Dxy*Dxy );
+ real R = - dxux*dyuy*dzuz + c1o4*( -Dxy*Dxz*Dyz + dxux*Dyz*Dyz + dyuy*Dxz*Dxz + dzuz*Dxy*Dxy );
return C * max(R, c0o1) / Q;
}
diff --git a/src/gpu/VirtualFluids_GPU/GPU/TurbulentViscosityKernels.cu b/src/gpu/VirtualFluids_GPU/GPU/TurbulentViscosityKernels.cu
index 3719ca3712e6f63a77f62bf314af7d19eea01f4c..7147629c448b8b730e4ae8c4eff8a0a400863de9 100644
--- a/src/gpu/VirtualFluids_GPU/GPU/TurbulentViscosityKernels.cu
+++ b/src/gpu/VirtualFluids_GPU/GPU/TurbulentViscosityKernels.cu
@@ -38,6 +38,7 @@
#include <cuda_runtime.h>
#include <helper_cuda.h>
#include "LBM/LB.h"
+#include "LBM/GPUHelperFunctions/KernelUtilities.h"
using namespace vf::lbm::constant;
@@ -52,34 +53,31 @@ __host__ __device__ __forceinline__ void calcDerivatives(const uint& k, uint& kM
dvz = ((fluidP ? vz[kP] : vz[k])-(fluidM ? vz[kM] : vz[k]))*div;
}
-__global__ void calcAMD(real* vx,
- real* vy,
- real* vz,
- real* turbulentViscosity,
- uint* neighborX,
- uint* neighborY,
- uint* neighborZ,
- uint* neighborWSB,
- uint* typeOfGridNode,
- uint size_Mat,
- real SGSConstant)
+__global__ void calcAMD(
+ real* vx,
+ real* vy,
+ real* vz,
+ real* turbulentViscosity,
+ uint* neighborX,
+ uint* neighborY,
+ uint* neighborZ,
+ uint* neighborWSB,
+ uint* typeOfGridNode,
+ unsigned long long numberOfLBnodes,
+ real SGSConstant)
{
+ ////////////////////////////////////////////////////////////////////////////////
+ //! - Get node index coordinates from threadIdx, blockIdx, blockDim and gridDim.
+ //!
+ const unsigned nodeIndex = vf::gpu::getNodeIndex();
- const uint x = threadIdx.x;
- const uint y = blockIdx.x;
- const uint z = blockIdx.y;
+ if(nodeIndex >= numberOfLBnodes) return;
+ if(typeOfGridNode[nodeIndex] != GEO_FLUID) return;
- const uint nx = blockDim.x;
- const uint ny = gridDim.x;
-
- const uint k = nx*(ny*z + y) + x;
- if(k >= size_Mat) return;
- if(typeOfGridNode[k] != GEO_FLUID) return;
-
- uint kPx = neighborX[k];
- uint kPy = neighborY[k];
- uint kPz = neighborZ[k];
- uint kMxyz = neighborWSB[k];
+ uint kPx = neighborX[nodeIndex];
+ uint kPy = neighborY[nodeIndex];
+ uint kPz = neighborZ[nodeIndex];
+ uint kMxyz = neighborWSB[nodeIndex];
uint kMx = neighborZ[neighborY[kMxyz]];
uint kMy = neighborZ[neighborX[kMxyz]];
uint kMz = neighborY[neighborX[kMxyz]];
@@ -88,9 +86,9 @@ __global__ void calcAMD(real* vx,
dvydx, dvydy, dvydz,
dvzdx, dvzdy, dvzdz;
- calcDerivatives(k, kMx, kPx, typeOfGridNode, vx, vy, vz, dvxdx, dvydx, dvzdx);
- calcDerivatives(k, kMy, kPy, typeOfGridNode, vx, vy, vz, dvxdy, dvydy, dvzdy);
- calcDerivatives(k, kMz, kPz, typeOfGridNode, vx, vy, vz, dvxdz, dvydz, dvzdz);
+ calcDerivatives(nodeIndex, kMx, kPx, typeOfGridNode, vx, vy, vz, dvxdx, dvydx, dvzdx);
+ calcDerivatives(nodeIndex, kMy, kPy, typeOfGridNode, vx, vy, vz, dvxdy, dvydy, dvzdy);
+ calcDerivatives(nodeIndex, kMz, kPz, typeOfGridNode, vx, vy, vz, dvxdz, dvydz, dvzdz);
real denominator = dvxdx*dvxdx + dvydx*dvydx + dvzdx*dvzdx +
dvxdy*dvxdy + dvydy*dvydy + dvzdy*dvzdy +
@@ -102,7 +100,7 @@ __global__ void calcAMD(real* vx,
(dvxdx*dvzdx + dvxdy*dvzdy + dvxdz*dvzdz) * (dvxdz+dvzdx) +
(dvydx*dvzdx + dvydy*dvzdy + dvydz*dvzdz) * (dvydz+dvzdy);
- turbulentViscosity[k] = max(c0o1,-SGSConstant*enumerator)/denominator;
+ turbulentViscosity[nodeIndex] = denominator != c0o1 ? max(c0o1,-SGSConstant*enumerator)/denominator : c0o1;
}
void calcTurbulentViscosityAMD(Parameter* para, int level)
diff --git a/src/gpu/VirtualFluids_GPU/GPU/VelocityBCs27.cu b/src/gpu/VirtualFluids_GPU/GPU/VelocityBCs27.cu
index 05c85e8b546aeaa964b1dbb61cbf01dd9b82ca1a..ccf9d1771ec0e1895e5cb79fae63675429b02c73 100644
--- a/src/gpu/VirtualFluids_GPU/GPU/VelocityBCs27.cu
+++ b/src/gpu/VirtualFluids_GPU/GPU/VelocityBCs27.cu
@@ -1,96 +1,120 @@
-// _ ___ __ __________ _ __ ______________ __
-// | | / (_)____/ /___ ______ _/ / ____/ /_ __(_)___/ /____ / ___/ __ / / / /
-// | | / / / ___/ __/ / / / __ `/ / /_ / / / / / / __ / ___/ / /___/ /_/ / / / /
-// | |/ / / / / /_/ /_/ / /_/ / / __/ / / /_/ / / /_/ (__ ) / /_) / ____/ /__/ /
-// |___/_/_/ \__/\__,_/\__,_/_/_/ /_/\__,_/_/\__,_/____/ \____/_/ \_____/
+//=======================================================================================
+// ____ ____ __ ______ __________ __ __ __ __
+// \ \ | | | | | _ \ |___ ___| | | | | / \ | |
+// \ \ | | | | | |_) | | | | | | | / \ | |
+// \ \ | | | | | _ / | | | | | | / /\ \ | |
+// \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
+// \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
+// \ \ | | ________________________________________________________________
+// \ \ | | | ______________________________________________________________|
+// \ \| | | | __ __ __ __ ______ _______
+// \ | | |_____ | | | | | | | | | _ \ / _____)
+// \ | | _____| | | | | | | | | | | \ \ \_______
+// \ | | | | |_____ | \_/ | | | | |_/ / _____ |
+// \ _____| |__| |________| \_______/ |__| |______/ (_______/
//
-//////////////////////////////////////////////////////////////////////////
-
-/* Device code */
+// This file is part of VirtualFluids. VirtualFluids is free software: you can
+// redistribute it and/or modify it under the terms of the GNU General Public
+// License as published by the Free Software Foundation, either version 3 of
+// the License, or (at your option) any later version.
+//
+// VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
+// ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+// FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
+// for more details.
+//
+// You should have received a copy of the GNU General Public License along
+// with VirtualFluids (see COPYING.txt). If not, see <http://www.gnu.org/licenses/>.
+//
+//! \file VelocityBCs27.cu
+//! \ingroup GPU
+//! \author Martin Schoenherr, Anna Wellmann
+//======================================================================================
#include "LBM/LB.h"
#include "lbm/constants/D3Q27.h"
#include "lbm/constants/NumericConstants.h"
-#include "KernelUtilities.h"
+#include "LBM/GPUHelperFunctions/KernelUtilities.h"
using namespace vf::lbm::constant;
using namespace vf::lbm::dir;
+using namespace vf::gpu;
//////////////////////////////////////////////////////////////////////////////
__global__ void QVelDeviceCompPlusSlip27(
- real* vx,
- real* vy,
- real* vz,
- real* DD,
- int* k_Q,
- real* QQ,
- unsigned int numberOfBCnodes,
- real om1,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep)
+ real* vx,
+ real* vy,
+ real* vz,
+ real* DD,
+ int* k_Q,
+ real* QQ,
+ unsigned int numberOfBCnodes,
+ real om1,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep)
{
Distributions27 D;
if (isEvenTimestep==true)
{
- D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
+ D.f[DIR_M00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_MPM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_PMP * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////
const unsigned x = threadIdx.x; // Globaler x-Index
@@ -115,24 +139,24 @@ __global__ void QVelDeviceCompPlusSlip27(
*q_dirBE, *q_dirTW, *q_dirTN, *q_dirBS, *q_dirBN, *q_dirTS,
*q_dirTNE, *q_dirTSW, *q_dirTSE, *q_dirTNW, *q_dirBNE, *q_dirBSW,
*q_dirBSE, *q_dirBNW;
- q_dirE = &QQ[DIR_P00 * numberOfBCnodes];
- q_dirW = &QQ[DIR_M00 * numberOfBCnodes];
- q_dirN = &QQ[DIR_0P0 * numberOfBCnodes];
- q_dirS = &QQ[DIR_0M0 * numberOfBCnodes];
- q_dirT = &QQ[DIR_00P * numberOfBCnodes];
- q_dirB = &QQ[DIR_00M * numberOfBCnodes];
- q_dirNE = &QQ[DIR_PP0 * numberOfBCnodes];
- q_dirSW = &QQ[DIR_MM0 * numberOfBCnodes];
- q_dirSE = &QQ[DIR_PM0 * numberOfBCnodes];
- q_dirNW = &QQ[DIR_MP0 * numberOfBCnodes];
- q_dirTE = &QQ[DIR_P0P * numberOfBCnodes];
- q_dirBW = &QQ[DIR_M0M * numberOfBCnodes];
- q_dirBE = &QQ[DIR_P0M * numberOfBCnodes];
- q_dirTW = &QQ[DIR_M0P * numberOfBCnodes];
- q_dirTN = &QQ[DIR_0PP * numberOfBCnodes];
- q_dirBS = &QQ[DIR_0MM * numberOfBCnodes];
- q_dirBN = &QQ[DIR_0PM * numberOfBCnodes];
- q_dirTS = &QQ[DIR_0MP * numberOfBCnodes];
+ q_dirE = &QQ[DIR_P00 * numberOfBCnodes];
+ q_dirW = &QQ[DIR_M00 * numberOfBCnodes];
+ q_dirN = &QQ[DIR_0P0 * numberOfBCnodes];
+ q_dirS = &QQ[DIR_0M0 * numberOfBCnodes];
+ q_dirT = &QQ[DIR_00P * numberOfBCnodes];
+ q_dirB = &QQ[DIR_00M * numberOfBCnodes];
+ q_dirNE = &QQ[DIR_PP0 * numberOfBCnodes];
+ q_dirSW = &QQ[DIR_MM0 * numberOfBCnodes];
+ q_dirSE = &QQ[DIR_PM0 * numberOfBCnodes];
+ q_dirNW = &QQ[DIR_MP0 * numberOfBCnodes];
+ q_dirTE = &QQ[DIR_P0P * numberOfBCnodes];
+ q_dirBW = &QQ[DIR_M0M * numberOfBCnodes];
+ q_dirBE = &QQ[DIR_P0M * numberOfBCnodes];
+ q_dirTW = &QQ[DIR_M0P * numberOfBCnodes];
+ q_dirTN = &QQ[DIR_0PP * numberOfBCnodes];
+ q_dirBS = &QQ[DIR_0MM * numberOfBCnodes];
+ q_dirBN = &QQ[DIR_0PM * numberOfBCnodes];
+ q_dirTS = &QQ[DIR_0MP * numberOfBCnodes];
q_dirTNE = &QQ[DIR_PPP * numberOfBCnodes];
q_dirTSW = &QQ[DIR_MMP * numberOfBCnodes];
q_dirTSE = &QQ[DIR_PMP * numberOfBCnodes];
@@ -225,63 +249,63 @@ __global__ void QVelDeviceCompPlusSlip27(
//////////////////////////////////////////////////////////////////////////
if (isEvenTimestep == false)
{
- D.f[DIR_P00] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_M00] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0P0] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0M0] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00P] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00M] = &DD[DIR_00M *size_Mat];
- D.f[DIR_PP0] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_MM0] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_PM0] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_MP0] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_P0P] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_M0M] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_P0M] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_M0P] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0PP] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0MM] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0PM] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0MP] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_MMP] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_PMP] = &DD[DIR_PMP *size_Mat];
- D.f[DIR_MPP] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_PPM] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_MMM] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_PMM] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_MPM] = &DD[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_P00] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0M0] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0P0] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00M] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00P] = &DD[DIR_00M *size_Mat];
- D.f[DIR_MM0] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_PP0] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_MP0] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_PM0] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_M0M] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_P0P] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_M0P] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_P0M] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0MM] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0PP] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0MP] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0PM] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_MMP] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_PMP] = &DD[DIR_MPM *size_Mat];
- D.f[DIR_MPP] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_PPM] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_MMM] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_PMM] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_MPM] = &DD[DIR_PMP *size_Mat];
+ D.f[DIR_M00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_MPM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_PMP * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
//Test
@@ -553,18 +577,19 @@ __global__ void QVelDeviceCompPlusSlip27(
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
-__global__ void QVeloDeviceEQ27(real* VeloX,
- real* VeloY,
- real* VeloZ,
- real* DD,
- int* k_Q,
- int numberOfBCnodes,
- real om1,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep)
+__global__ void QVeloDeviceEQ27(
+ real* VeloX,
+ real* VeloY,
+ real* VeloZ,
+ real* DD,
+ int* k_Q,
+ int numberOfBCnodes,
+ real om1,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep)
{
////////////////////////////////////////////////////////////////////////////////
const unsigned x = threadIdx.x; // Globaler x-Index
@@ -613,95 +638,95 @@ __global__ void QVeloDeviceEQ27(real* VeloX,
Distributions27 D;
if (isEvenTimestep==true)
{
- D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
+ D.f[DIR_M00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_MPM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_PMP * numberOfLBnodes];
}
//////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
// based on BGK Plus Comp
//////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
- real mfcbb = (D.f[DIR_P00 ])[ke ];
- real mfabb = (D.f[DIR_M00 ])[kw ];
- real mfbcb = (D.f[DIR_0P0 ])[kn ];
- real mfbab = (D.f[DIR_0M0 ])[ks ];
- real mfbbc = (D.f[DIR_00P ])[kt ];
- real mfbba = (D.f[DIR_00M ])[kb ];
- real mfccb = (D.f[DIR_PP0 ])[kne ];
- real mfaab = (D.f[DIR_MM0 ])[ksw ];
- real mfcab = (D.f[DIR_PM0 ])[kse ];
- real mfacb = (D.f[DIR_MP0 ])[knw ];
- real mfcbc = (D.f[DIR_P0P ])[kte ];
- real mfaba = (D.f[DIR_M0M ])[kbw ];
- real mfcba = (D.f[DIR_P0M ])[kbe ];
- real mfabc = (D.f[DIR_M0P ])[ktw ];
- real mfbcc = (D.f[DIR_0PP ])[ktn ];
- real mfbaa = (D.f[DIR_0MM ])[kbs ];
- real mfbca = (D.f[DIR_0PM ])[kbn ];
- real mfbac = (D.f[DIR_0MP ])[kts ];
+ real mfcbb = (D.f[DIR_P00])[ke ];
+ real mfabb = (D.f[DIR_M00])[kw ];
+ real mfbcb = (D.f[DIR_0P0])[kn ];
+ real mfbab = (D.f[DIR_0M0])[ks ];
+ real mfbbc = (D.f[DIR_00P])[kt ];
+ real mfbba = (D.f[DIR_00M])[kb ];
+ real mfccb = (D.f[DIR_PP0])[kne ];
+ real mfaab = (D.f[DIR_MM0])[ksw ];
+ real mfcab = (D.f[DIR_PM0])[kse ];
+ real mfacb = (D.f[DIR_MP0])[knw ];
+ real mfcbc = (D.f[DIR_P0P])[kte ];
+ real mfaba = (D.f[DIR_M0M])[kbw ];
+ real mfcba = (D.f[DIR_P0M])[kbe ];
+ real mfabc = (D.f[DIR_M0P])[ktw ];
+ real mfbcc = (D.f[DIR_0PP])[ktn ];
+ real mfbaa = (D.f[DIR_0MM])[kbs ];
+ real mfbca = (D.f[DIR_0PM])[kbn ];
+ real mfbac = (D.f[DIR_0MP])[kts ];
real mfbbb = (D.f[DIR_000])[kzero];
- real mfccc = (D.f[DIR_PPP ])[ktne ];
- real mfaac = (D.f[DIR_MMP ])[ktsw ];
- real mfcac = (D.f[DIR_PMP ])[ktse ];
- real mfacc = (D.f[DIR_MPP ])[ktnw ];
- real mfcca = (D.f[DIR_PPM ])[kbne ];
- real mfaaa = (D.f[DIR_MMM ])[kbsw ];
- real mfcaa = (D.f[DIR_PMM ])[kbse ];
- real mfaca = (D.f[DIR_MPM ])[kbnw ];
+ real mfccc = (D.f[DIR_PPP])[ktne ];
+ real mfaac = (D.f[DIR_MMP])[ktsw ];
+ real mfcac = (D.f[DIR_PMP])[ktse ];
+ real mfacc = (D.f[DIR_MPP])[ktnw ];
+ real mfcca = (D.f[DIR_PPM])[kbne ];
+ real mfaaa = (D.f[DIR_MMM])[kbsw ];
+ real mfcaa = (D.f[DIR_PMM])[kbse ];
+ real mfaca = (D.f[DIR_MPM])[kbnw ];
////////////////////////////////////////////////////////////////////////////////////
real rho = (mfccc+mfaaa + mfaca+mfcac + mfacc+mfcaa + mfaac+mfcca +
mfbac+mfbca + mfbaa+mfbcc + mfabc+mfcba + mfaba+mfcbc + mfacb+mfcab + mfaab+mfccb +
@@ -763,33 +788,33 @@ __global__ void QVeloDeviceEQ27(real* VeloX,
mfcaa = -rho * XXc * YYa * ZZa - c1o216;
mfaca = -rho * XXa * YYc * ZZa - c1o216;
//////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
- (D.f[DIR_P00 ])[ke ] = mfabb;//mfcbb;
- (D.f[DIR_M00 ])[kw ] = mfcbb;//mfabb;
- (D.f[DIR_0P0 ])[kn ] = mfbab;//mfbcb;
- (D.f[DIR_0M0 ])[ks ] = mfbcb;//mfbab;
- (D.f[DIR_00P ])[kt ] = mfbba;//mfbbc;
- (D.f[DIR_00M ])[kb ] = mfbbc;//mfbba;
- (D.f[DIR_PP0 ])[kne ] = mfaab;//mfccb;
- (D.f[DIR_MM0 ])[ksw ] = mfccb;//mfaab;
- (D.f[DIR_PM0 ])[kse ] = mfacb;//mfcab;
- (D.f[DIR_MP0 ])[knw ] = mfcab;//mfacb;
- (D.f[DIR_P0P ])[kte ] = mfaba;//mfcbc;
- (D.f[DIR_M0M ])[kbw ] = mfcbc;//mfaba;
- (D.f[DIR_P0M ])[kbe ] = mfabc;//mfcba;
- (D.f[DIR_M0P ])[ktw ] = mfcba;//mfabc;
- (D.f[DIR_0PP ])[ktn ] = mfbaa;//mfbcc;
- (D.f[DIR_0MM ])[kbs ] = mfbcc;//mfbaa;
- (D.f[DIR_0PM ])[kbn ] = mfbac;//mfbca;
- (D.f[DIR_0MP ])[kts ] = mfbca;//mfbac;
+ (D.f[DIR_P00])[ke ] = mfabb;//mfcbb;
+ (D.f[DIR_M00])[kw ] = mfcbb;//mfabb;
+ (D.f[DIR_0P0])[kn ] = mfbab;//mfbcb;
+ (D.f[DIR_0M0])[ks ] = mfbcb;//mfbab;
+ (D.f[DIR_00P])[kt ] = mfbba;//mfbbc;
+ (D.f[DIR_00M])[kb ] = mfbbc;//mfbba;
+ (D.f[DIR_PP0])[kne ] = mfaab;//mfccb;
+ (D.f[DIR_MM0])[ksw ] = mfccb;//mfaab;
+ (D.f[DIR_PM0])[kse ] = mfacb;//mfcab;
+ (D.f[DIR_MP0])[knw ] = mfcab;//mfacb;
+ (D.f[DIR_P0P])[kte ] = mfaba;//mfcbc;
+ (D.f[DIR_M0M])[kbw ] = mfcbc;//mfaba;
+ (D.f[DIR_P0M])[kbe ] = mfabc;//mfcba;
+ (D.f[DIR_M0P])[ktw ] = mfcba;//mfabc;
+ (D.f[DIR_0PP])[ktn ] = mfbaa;//mfbcc;
+ (D.f[DIR_0MM])[kbs ] = mfbcc;//mfbaa;
+ (D.f[DIR_0PM])[kbn ] = mfbac;//mfbca;
+ (D.f[DIR_0MP])[kts ] = mfbca;//mfbac;
(D.f[DIR_000])[kzero] = mfbbb;//mfbbb;
- (D.f[DIR_PPP ])[ktne ] = mfaaa;//mfccc;
- (D.f[DIR_MMP ])[ktsw ] = mfcca;//mfaac;
- (D.f[DIR_PMP ])[ktse ] = mfaca;//mfcac;
- (D.f[DIR_MPP ])[ktnw ] = mfcaa;//mfacc;
- (D.f[DIR_PPM ])[kbne ] = mfaac;//mfcca;
- (D.f[DIR_MMM ])[kbsw ] = mfccc;//mfaaa;
- (D.f[DIR_PMM ])[kbse ] = mfacc;//mfcaa;
- (D.f[DIR_MPM ])[kbnw ] = mfcac;//mfaca;
+ (D.f[DIR_PPP])[ktne ] = mfaaa;//mfccc;
+ (D.f[DIR_MMP])[ktsw ] = mfcca;//mfaac;
+ (D.f[DIR_PMP])[ktse ] = mfaca;//mfcac;
+ (D.f[DIR_MPP])[ktnw ] = mfcaa;//mfacc;
+ (D.f[DIR_PPM])[kbne ] = mfaac;//mfcca;
+ (D.f[DIR_MMM])[kbsw ] = mfccc;//mfaaa;
+ (D.f[DIR_PMM])[kbse ] = mfacc;//mfcaa;
+ (D.f[DIR_MPM])[kbnw ] = mfcac;//mfaca;
}
}
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
@@ -834,18 +859,18 @@ __global__ void QVeloDeviceEQ27(real* VeloX,
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
__global__ void QVeloStreetDeviceEQ27(
- real* veloXfraction,
- real* veloYfraction,
- int* naschVelo,
- real* DD,
- int* naschIndex,
- int numberOfStreetNodes,
- real velocityRatio,
- uint* neighborX,
- uint* neighborY,
- uint* neighborZ,
- uint size_Mat,
- bool isEvenTimestep)
+ real* veloXfraction,
+ real* veloYfraction,
+ int* naschVelo,
+ real* DD,
+ int* naschIndex,
+ int numberOfStreetNodes,
+ real velocityRatio,
+ uint* neighborX,
+ uint* neighborY,
+ uint* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep)
{
////////////////////////////////////////////////////////////////////////////////
const unsigned x = threadIdx.x; // Globaler x-Index
@@ -894,95 +919,95 @@ __global__ void QVeloStreetDeviceEQ27(
Distributions27 D;
if (isEvenTimestep == true)
{
- D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
+ D.f[DIR_M00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_MPM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_PMP * numberOfLBnodes];
}
//////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
// based on BGK Plus Comp
//////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
- real mfcbb = (D.f[DIR_P00 ])[ke ];
- real mfabb = (D.f[DIR_M00 ])[kw ];
- real mfbcb = (D.f[DIR_0P0 ])[kn ];
- real mfbab = (D.f[DIR_0M0 ])[ks ];
- real mfbbc = (D.f[DIR_00P ])[kt ];
- real mfbba = (D.f[DIR_00M ])[kb ];
- real mfccb = (D.f[DIR_PP0 ])[kne ];
- real mfaab = (D.f[DIR_MM0 ])[ksw ];
- real mfcab = (D.f[DIR_PM0 ])[kse ];
- real mfacb = (D.f[DIR_MP0 ])[knw ];
- real mfcbc = (D.f[DIR_P0P ])[kte ];
- real mfaba = (D.f[DIR_M0M ])[kbw ];
- real mfcba = (D.f[DIR_P0M ])[kbe ];
- real mfabc = (D.f[DIR_M0P ])[ktw ];
- real mfbcc = (D.f[DIR_0PP ])[ktn ];
- real mfbaa = (D.f[DIR_0MM ])[kbs ];
- real mfbca = (D.f[DIR_0PM ])[kbn ];
- real mfbac = (D.f[DIR_0MP ])[kts ];
+ real mfcbb = (D.f[DIR_P00])[ke ];
+ real mfabb = (D.f[DIR_M00])[kw ];
+ real mfbcb = (D.f[DIR_0P0])[kn ];
+ real mfbab = (D.f[DIR_0M0])[ks ];
+ real mfbbc = (D.f[DIR_00P])[kt ];
+ real mfbba = (D.f[DIR_00M])[kb ];
+ real mfccb = (D.f[DIR_PP0])[kne ];
+ real mfaab = (D.f[DIR_MM0])[ksw ];
+ real mfcab = (D.f[DIR_PM0])[kse ];
+ real mfacb = (D.f[DIR_MP0])[knw ];
+ real mfcbc = (D.f[DIR_P0P])[kte ];
+ real mfaba = (D.f[DIR_M0M])[kbw ];
+ real mfcba = (D.f[DIR_P0M])[kbe ];
+ real mfabc = (D.f[DIR_M0P])[ktw ];
+ real mfbcc = (D.f[DIR_0PP])[ktn ];
+ real mfbaa = (D.f[DIR_0MM])[kbs ];
+ real mfbca = (D.f[DIR_0PM])[kbn ];
+ real mfbac = (D.f[DIR_0MP])[kts ];
real mfbbb = (D.f[DIR_000])[kzero];
- real mfccc = (D.f[DIR_PPP ])[ktne ];
- real mfaac = (D.f[DIR_MMP ])[ktsw ];
- real mfcac = (D.f[DIR_PMP ])[ktse ];
- real mfacc = (D.f[DIR_MPP ])[ktnw ];
- real mfcca = (D.f[DIR_PPM ])[kbne ];
- real mfaaa = (D.f[DIR_MMM ])[kbsw ];
- real mfcaa = (D.f[DIR_PMM ])[kbse ];
- real mfaca = (D.f[DIR_MPM ])[kbnw ];
+ real mfccc = (D.f[DIR_PPP])[ktne ];
+ real mfaac = (D.f[DIR_MMP])[ktsw ];
+ real mfcac = (D.f[DIR_PMP])[ktse ];
+ real mfacc = (D.f[DIR_MPP])[ktnw ];
+ real mfcca = (D.f[DIR_PPM])[kbne ];
+ real mfaaa = (D.f[DIR_MMM])[kbsw ];
+ real mfcaa = (D.f[DIR_PMM])[kbse ];
+ real mfaca = (D.f[DIR_MPM])[kbnw ];
////////////////////////////////////////////////////////////////////////////////////
real rho = (mfccc + mfaaa + mfaca + mfcac + mfacc + mfcaa + mfaac + mfcca +
mfbac + mfbca + mfbaa + mfbcc + mfabc + mfcba + mfaba + mfcbc + mfacb + mfcab + mfaab + mfccb +
@@ -1049,33 +1074,33 @@ __global__ void QVeloStreetDeviceEQ27(
mfcaa = -rho * XXc * YYa * ZZa - c1o216;
mfaca = -rho * XXa * YYc * ZZa - c1o216;
//////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
- (D.f[DIR_P00 ])[ke ] = mfabb;//mfcbb;
- (D.f[DIR_M00 ])[kw ] = mfcbb;//mfabb;
- (D.f[DIR_0P0 ])[kn ] = mfbab;//mfbcb;
- (D.f[DIR_0M0 ])[ks ] = mfbcb;//mfbab;
- (D.f[DIR_00P ])[kt ] = mfbba;//mfbbc;
- (D.f[DIR_00M ])[kb ] = mfbbc;//mfbba;
- (D.f[DIR_PP0 ])[kne ] = mfaab;//mfccb;
- (D.f[DIR_MM0 ])[ksw ] = mfccb;//mfaab;
- (D.f[DIR_PM0 ])[kse ] = mfacb;//mfcab;
- (D.f[DIR_MP0 ])[knw ] = mfcab;//mfacb;
- (D.f[DIR_P0P ])[kte ] = mfaba;//mfcbc;
- (D.f[DIR_M0M ])[kbw ] = mfcbc;//mfaba;
- (D.f[DIR_P0M ])[kbe ] = mfabc;//mfcba;
- (D.f[DIR_M0P ])[ktw ] = mfcba;//mfabc;
- (D.f[DIR_0PP ])[ktn ] = mfbaa;//mfbcc;
- (D.f[DIR_0MM ])[kbs ] = mfbcc;//mfbaa;
- (D.f[DIR_0PM ])[kbn ] = mfbac;//mfbca;
- (D.f[DIR_0MP ])[kts ] = mfbca;//mfbac;
+ (D.f[DIR_P00])[ke ] = mfabb;//mfcbb;
+ (D.f[DIR_M00])[kw ] = mfcbb;//mfabb;
+ (D.f[DIR_0P0])[kn ] = mfbab;//mfbcb;
+ (D.f[DIR_0M0])[ks ] = mfbcb;//mfbab;
+ (D.f[DIR_00P])[kt ] = mfbba;//mfbbc;
+ (D.f[DIR_00M])[kb ] = mfbbc;//mfbba;
+ (D.f[DIR_PP0])[kne ] = mfaab;//mfccb;
+ (D.f[DIR_MM0])[ksw ] = mfccb;//mfaab;
+ (D.f[DIR_PM0])[kse ] = mfacb;//mfcab;
+ (D.f[DIR_MP0])[knw ] = mfcab;//mfacb;
+ (D.f[DIR_P0P])[kte ] = mfaba;//mfcbc;
+ (D.f[DIR_M0M])[kbw ] = mfcbc;//mfaba;
+ (D.f[DIR_P0M])[kbe ] = mfabc;//mfcba;
+ (D.f[DIR_M0P])[ktw ] = mfcba;//mfabc;
+ (D.f[DIR_0PP])[ktn ] = mfbaa;//mfbcc;
+ (D.f[DIR_0MM])[kbs ] = mfbcc;//mfbaa;
+ (D.f[DIR_0PM])[kbn ] = mfbac;//mfbca;
+ (D.f[DIR_0MP])[kts ] = mfbca;//mfbac;
(D.f[DIR_000])[kzero] = mfbbb;//mfbbb;
- (D.f[DIR_PPP ])[ktne ] = mfaaa;//mfccc;
- (D.f[DIR_MMP ])[ktsw ] = mfcca;//mfaac;
- (D.f[DIR_PMP ])[ktse ] = mfaca;//mfcac;
- (D.f[DIR_MPP ])[ktnw ] = mfcaa;//mfacc;
- (D.f[DIR_PPM ])[kbne ] = mfaac;//mfcca;
- (D.f[DIR_MMM ])[kbsw ] = mfccc;//mfaaa;
- (D.f[DIR_PMM ])[kbse ] = mfacc;//mfcaa;
- (D.f[DIR_MPM ])[kbnw ] = mfcac;//mfaca;
+ (D.f[DIR_PPP])[ktne ] = mfaaa;//mfccc;
+ (D.f[DIR_MMP])[ktsw ] = mfcca;//mfaac;
+ (D.f[DIR_PMP])[ktse ] = mfaca;//mfcac;
+ (D.f[DIR_MPP])[ktnw ] = mfcaa;//mfacc;
+ (D.f[DIR_PPM])[kbne ] = mfaac;//mfcca;
+ (D.f[DIR_MMM])[kbsw ] = mfccc;//mfaaa;
+ (D.f[DIR_PMM])[kbse ] = mfacc;//mfcaa;
+ (D.f[DIR_MPM])[kbnw ] = mfcac;//mfaca;
}
}
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
@@ -1120,80 +1145,80 @@ __global__ void QVeloStreetDeviceEQ27(
//////////////////////////////////////////////////////////////////////////////
__global__ void QVelDeviceIncompHighNu27(
- real* vx,
- real* vy,
- real* vz,
- real* DD,
- int* k_Q,
- real* QQ,
- unsigned int numberOfBCnodes,
- real om1,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep)
+ real* vx,
+ real* vy,
+ real* vz,
+ real* DD,
+ int* k_Q,
+ real* QQ,
+ unsigned int numberOfBCnodes,
+ real om1,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep)
{
Distributions27 D;
if (isEvenTimestep==true)
{
- D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
+ D.f[DIR_M00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_MPM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_PMP * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////
const unsigned x = threadIdx.x; // Globaler x-Index
@@ -1218,24 +1243,24 @@ __global__ void QVelDeviceIncompHighNu27(
*q_dirBE, *q_dirTW, *q_dirTN, *q_dirBS, *q_dirBN, *q_dirTS,
*q_dirTNE, *q_dirTSW, *q_dirTSE, *q_dirTNW, *q_dirBNE, *q_dirBSW,
*q_dirBSE, *q_dirBNW;
- q_dirE = &QQ[DIR_P00 * numberOfBCnodes];
- q_dirW = &QQ[DIR_M00 * numberOfBCnodes];
- q_dirN = &QQ[DIR_0P0 * numberOfBCnodes];
- q_dirS = &QQ[DIR_0M0 * numberOfBCnodes];
- q_dirT = &QQ[DIR_00P * numberOfBCnodes];
- q_dirB = &QQ[DIR_00M * numberOfBCnodes];
- q_dirNE = &QQ[DIR_PP0 * numberOfBCnodes];
- q_dirSW = &QQ[DIR_MM0 * numberOfBCnodes];
- q_dirSE = &QQ[DIR_PM0 * numberOfBCnodes];
- q_dirNW = &QQ[DIR_MP0 * numberOfBCnodes];
- q_dirTE = &QQ[DIR_P0P * numberOfBCnodes];
- q_dirBW = &QQ[DIR_M0M * numberOfBCnodes];
- q_dirBE = &QQ[DIR_P0M * numberOfBCnodes];
- q_dirTW = &QQ[DIR_M0P * numberOfBCnodes];
- q_dirTN = &QQ[DIR_0PP * numberOfBCnodes];
- q_dirBS = &QQ[DIR_0MM * numberOfBCnodes];
- q_dirBN = &QQ[DIR_0PM * numberOfBCnodes];
- q_dirTS = &QQ[DIR_0MP * numberOfBCnodes];
+ q_dirE = &QQ[DIR_P00 * numberOfBCnodes];
+ q_dirW = &QQ[DIR_M00 * numberOfBCnodes];
+ q_dirN = &QQ[DIR_0P0 * numberOfBCnodes];
+ q_dirS = &QQ[DIR_0M0 * numberOfBCnodes];
+ q_dirT = &QQ[DIR_00P * numberOfBCnodes];
+ q_dirB = &QQ[DIR_00M * numberOfBCnodes];
+ q_dirNE = &QQ[DIR_PP0 * numberOfBCnodes];
+ q_dirSW = &QQ[DIR_MM0 * numberOfBCnodes];
+ q_dirSE = &QQ[DIR_PM0 * numberOfBCnodes];
+ q_dirNW = &QQ[DIR_MP0 * numberOfBCnodes];
+ q_dirTE = &QQ[DIR_P0P * numberOfBCnodes];
+ q_dirBW = &QQ[DIR_M0M * numberOfBCnodes];
+ q_dirBE = &QQ[DIR_P0M * numberOfBCnodes];
+ q_dirTW = &QQ[DIR_M0P * numberOfBCnodes];
+ q_dirTN = &QQ[DIR_0PP * numberOfBCnodes];
+ q_dirBS = &QQ[DIR_0MM * numberOfBCnodes];
+ q_dirBN = &QQ[DIR_0PM * numberOfBCnodes];
+ q_dirTS = &QQ[DIR_0MP * numberOfBCnodes];
q_dirTNE = &QQ[DIR_PPP * numberOfBCnodes];
q_dirTSW = &QQ[DIR_MMP * numberOfBCnodes];
q_dirTSE = &QQ[DIR_PMP * numberOfBCnodes];
@@ -1278,32 +1303,32 @@ __global__ void QVelDeviceIncompHighNu27(
real f_E, f_W, f_N, f_S, f_T, f_B, f_NE, f_SW, f_SE, f_NW, f_TE, f_BW, f_BE,
f_TW, f_TN, f_BS, f_BN, f_TS, f_TNE, f_TSW, f_TSE, f_TNW, f_BNE, f_BSW, f_BSE, f_BNW;
- f_E = (D.f[DIR_P00 ])[ke ];
- f_W = (D.f[DIR_M00 ])[kw ];
- f_N = (D.f[DIR_0P0 ])[kn ];
- f_S = (D.f[DIR_0M0 ])[ks ];
- f_T = (D.f[DIR_00P ])[kt ];
- f_B = (D.f[DIR_00M ])[kb ];
- f_NE = (D.f[DIR_PP0 ])[kne ];
- f_SW = (D.f[DIR_MM0 ])[ksw ];
- f_SE = (D.f[DIR_PM0 ])[kse ];
- f_NW = (D.f[DIR_MP0 ])[knw ];
- f_TE = (D.f[DIR_P0P ])[kte ];
- f_BW = (D.f[DIR_M0M ])[kbw ];
- f_BE = (D.f[DIR_P0M ])[kbe ];
- f_TW = (D.f[DIR_M0P ])[ktw ];
- f_TN = (D.f[DIR_0PP ])[ktn ];
- f_BS = (D.f[DIR_0MM ])[kbs ];
- f_BN = (D.f[DIR_0PM ])[kbn ];
- f_TS = (D.f[DIR_0MP ])[kts ];
- f_TNE = (D.f[DIR_PPP ])[ktne ];
- f_TSW = (D.f[DIR_MMP ])[ktsw ];
- f_TSE = (D.f[DIR_PMP ])[ktse ];
- f_TNW = (D.f[DIR_MPP ])[ktnw ];
- f_BNE = (D.f[DIR_PPM ])[kbne ];
- f_BSW = (D.f[DIR_MMM ])[kbsw ];
- f_BSE = (D.f[DIR_PMM ])[kbse ];
- f_BNW = (D.f[DIR_MPM ])[kbnw ];
+ f_E = (D.f[DIR_P00])[ke ];
+ f_W = (D.f[DIR_M00])[kw ];
+ f_N = (D.f[DIR_0P0])[kn ];
+ f_S = (D.f[DIR_0M0])[ks ];
+ f_T = (D.f[DIR_00P])[kt ];
+ f_B = (D.f[DIR_00M])[kb ];
+ f_NE = (D.f[DIR_PP0])[kne ];
+ f_SW = (D.f[DIR_MM0])[ksw ];
+ f_SE = (D.f[DIR_PM0])[kse ];
+ f_NW = (D.f[DIR_MP0])[knw ];
+ f_TE = (D.f[DIR_P0P])[kte ];
+ f_BW = (D.f[DIR_M0M])[kbw ];
+ f_BE = (D.f[DIR_P0M])[kbe ];
+ f_TW = (D.f[DIR_M0P])[ktw ];
+ f_TN = (D.f[DIR_0PP])[ktn ];
+ f_BS = (D.f[DIR_0MM])[kbs ];
+ f_BN = (D.f[DIR_0PM])[kbn ];
+ f_TS = (D.f[DIR_0MP])[kts ];
+ f_TNE = (D.f[DIR_PPP])[ktne ];
+ f_TSW = (D.f[DIR_MMP])[ktsw ];
+ f_TSE = (D.f[DIR_PMP])[ktse ];
+ f_TNW = (D.f[DIR_MPP])[ktnw ];
+ f_BNE = (D.f[DIR_PPM])[kbne ];
+ f_BSW = (D.f[DIR_MMM])[kbsw ];
+ f_BSE = (D.f[DIR_PMM])[kbse ];
+ f_BNW = (D.f[DIR_MPM])[kbnw ];
////////////////////////////////////////////////////////////////////////////////
real vx1, vx2, vx3, drho, feq, q;
drho = f_TSE + f_TNW + f_TNE + f_TSW + f_BSE + f_BNW + f_BNE + f_BSW +
@@ -1328,63 +1353,63 @@ __global__ void QVelDeviceIncompHighNu27(
//////////////////////////////////////////////////////////////////////////
if (isEvenTimestep==false)
{
- D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
+ D.f[DIR_M00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_MPM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_PMP * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
//Test
@@ -1618,80 +1643,80 @@ __global__ void QVelDeviceIncompHighNu27(
//////////////////////////////////////////////////////////////////////////////
__global__ void QVelDeviceCompHighNu27(
- real* vx,
- real* vy,
- real* vz,
- real* DD,
- int* k_Q,
- real* QQ,
- unsigned int numberOfBCnodes,
- real om1,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep)
+ real* vx,
+ real* vy,
+ real* vz,
+ real* DD,
+ int* k_Q,
+ real* QQ,
+ unsigned int numberOfBCnodes,
+ real om1,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep)
{
Distributions27 D;
if (isEvenTimestep==true)
{
- D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
+ D.f[DIR_M00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_MPM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_PMP * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////
const unsigned x = threadIdx.x; // Globaler x-Index
@@ -1716,24 +1741,24 @@ __global__ void QVelDeviceCompHighNu27(
*q_dirBE, *q_dirTW, *q_dirTN, *q_dirBS, *q_dirBN, *q_dirTS,
*q_dirTNE, *q_dirTSW, *q_dirTSE, *q_dirTNW, *q_dirBNE, *q_dirBSW,
*q_dirBSE, *q_dirBNW;
- q_dirE = &QQ[DIR_P00 * numberOfBCnodes];
- q_dirW = &QQ[DIR_M00 * numberOfBCnodes];
- q_dirN = &QQ[DIR_0P0 * numberOfBCnodes];
- q_dirS = &QQ[DIR_0M0 * numberOfBCnodes];
- q_dirT = &QQ[DIR_00P * numberOfBCnodes];
- q_dirB = &QQ[DIR_00M * numberOfBCnodes];
- q_dirNE = &QQ[DIR_PP0 * numberOfBCnodes];
- q_dirSW = &QQ[DIR_MM0 * numberOfBCnodes];
- q_dirSE = &QQ[DIR_PM0 * numberOfBCnodes];
- q_dirNW = &QQ[DIR_MP0 * numberOfBCnodes];
- q_dirTE = &QQ[DIR_P0P * numberOfBCnodes];
- q_dirBW = &QQ[DIR_M0M * numberOfBCnodes];
- q_dirBE = &QQ[DIR_P0M * numberOfBCnodes];
- q_dirTW = &QQ[DIR_M0P * numberOfBCnodes];
- q_dirTN = &QQ[DIR_0PP * numberOfBCnodes];
- q_dirBS = &QQ[DIR_0MM * numberOfBCnodes];
- q_dirBN = &QQ[DIR_0PM * numberOfBCnodes];
- q_dirTS = &QQ[DIR_0MP * numberOfBCnodes];
+ q_dirE = &QQ[DIR_P00 * numberOfBCnodes];
+ q_dirW = &QQ[DIR_M00 * numberOfBCnodes];
+ q_dirN = &QQ[DIR_0P0 * numberOfBCnodes];
+ q_dirS = &QQ[DIR_0M0 * numberOfBCnodes];
+ q_dirT = &QQ[DIR_00P * numberOfBCnodes];
+ q_dirB = &QQ[DIR_00M * numberOfBCnodes];
+ q_dirNE = &QQ[DIR_PP0 * numberOfBCnodes];
+ q_dirSW = &QQ[DIR_MM0 * numberOfBCnodes];
+ q_dirSE = &QQ[DIR_PM0 * numberOfBCnodes];
+ q_dirNW = &QQ[DIR_MP0 * numberOfBCnodes];
+ q_dirTE = &QQ[DIR_P0P * numberOfBCnodes];
+ q_dirBW = &QQ[DIR_M0M * numberOfBCnodes];
+ q_dirBE = &QQ[DIR_P0M * numberOfBCnodes];
+ q_dirTW = &QQ[DIR_M0P * numberOfBCnodes];
+ q_dirTN = &QQ[DIR_0PP * numberOfBCnodes];
+ q_dirBS = &QQ[DIR_0MM * numberOfBCnodes];
+ q_dirBN = &QQ[DIR_0PM * numberOfBCnodes];
+ q_dirTS = &QQ[DIR_0MP * numberOfBCnodes];
q_dirTNE = &QQ[DIR_PPP * numberOfBCnodes];
q_dirTSW = &QQ[DIR_MMP * numberOfBCnodes];
q_dirTSE = &QQ[DIR_PMP * numberOfBCnodes];
@@ -1776,58 +1801,58 @@ __global__ void QVelDeviceCompHighNu27(
real f_E, f_W, f_N, f_S, f_T, f_B, f_NE, f_SW, f_SE, f_NW, f_TE, f_BW, f_BE,
f_TW, f_TN, f_BS, f_BN, f_TS, f_TNE, f_TSW, f_TSE, f_TNW, f_BNE, f_BSW, f_BSE, f_BNW;
- f_E = (D.f[DIR_P00 ])[ke ];
- f_W = (D.f[DIR_M00 ])[kw ];
- f_N = (D.f[DIR_0P0 ])[kn ];
- f_S = (D.f[DIR_0M0 ])[ks ];
- f_T = (D.f[DIR_00P ])[kt ];
- f_B = (D.f[DIR_00M ])[kb ];
- f_NE = (D.f[DIR_PP0 ])[kne ];
- f_SW = (D.f[DIR_MM0 ])[ksw ];
- f_SE = (D.f[DIR_PM0 ])[kse ];
- f_NW = (D.f[DIR_MP0 ])[knw ];
- f_TE = (D.f[DIR_P0P ])[kte ];
- f_BW = (D.f[DIR_M0M ])[kbw ];
- f_BE = (D.f[DIR_P0M ])[kbe ];
- f_TW = (D.f[DIR_M0P ])[ktw ];
- f_TN = (D.f[DIR_0PP ])[ktn ];
- f_BS = (D.f[DIR_0MM ])[kbs ];
- f_BN = (D.f[DIR_0PM ])[kbn ];
- f_TS = (D.f[DIR_0MP ])[kts ];
- f_TNE = (D.f[DIR_PPP ])[ktne ];
- f_TSW = (D.f[DIR_MMP ])[ktsw ];
- f_TSE = (D.f[DIR_PMP ])[ktse ];
- f_TNW = (D.f[DIR_MPP ])[ktnw ];
- f_BNE = (D.f[DIR_PPM ])[kbne ];
- f_BSW = (D.f[DIR_MMM ])[kbsw ];
- f_BSE = (D.f[DIR_PMM ])[kbse ];
- f_BNW = (D.f[DIR_MPM ])[kbnw ];
- //f_W = (D.f[DIR_P00 ])[ke ];
- //f_E = (D.f[DIR_M00 ])[kw ];
- //f_S = (D.f[DIR_0P0 ])[kn ];
- //f_N = (D.f[DIR_0M0 ])[ks ];
- //f_B = (D.f[DIR_00P ])[kt ];
- //f_T = (D.f[DIR_00M ])[kb ];
- //f_SW = (D.f[DIR_PP0 ])[kne ];
- //f_NE = (D.f[DIR_MM0 ])[ksw ];
- //f_NW = (D.f[DIR_PM0 ])[kse ];
- //f_SE = (D.f[DIR_MP0 ])[knw ];
- //f_BW = (D.f[DIR_P0P ])[kte ];
- //f_TE = (D.f[DIR_M0M ])[kbw ];
- //f_TW = (D.f[DIR_P0M ])[kbe ];
- //f_BE = (D.f[DIR_M0P ])[ktw ];
- //f_BS = (D.f[DIR_0PP ])[ktn ];
- //f_TN = (D.f[DIR_0MM ])[kbs ];
- //f_TS = (D.f[DIR_0PM ])[kbn ];
- //f_BN = (D.f[DIR_0MP ])[kts ];
- //f_BSW = (D.f[DIR_PPP ])[ktne ];
- //f_BNE = (D.f[DIR_MMP ])[ktsw ];
- //f_BNW = (D.f[DIR_PMP ])[ktse ];
- //f_BSE = (D.f[DIR_MPP ])[ktnw ];
- //f_TSW = (D.f[DIR_PPM ])[kbne ];
- //f_TNE = (D.f[DIR_MMM ])[kbsw ];
- //f_TNW = (D.f[DIR_PMM ])[kbse ];
- //f_TSE = (D.f[DIR_MPM ])[kbnw ];
+ f_E = (D.f[DIR_P00])[ke ];
+ f_W = (D.f[DIR_M00])[kw ];
+ f_N = (D.f[DIR_0P0])[kn ];
+ f_S = (D.f[DIR_0M0])[ks ];
+ f_T = (D.f[DIR_00P])[kt ];
+ f_B = (D.f[DIR_00M])[kb ];
+ f_NE = (D.f[DIR_PP0])[kne ];
+ f_SW = (D.f[DIR_MM0])[ksw ];
+ f_SE = (D.f[DIR_PM0])[kse ];
+ f_NW = (D.f[DIR_MP0])[knw ];
+ f_TE = (D.f[DIR_P0P])[kte ];
+ f_BW = (D.f[DIR_M0M])[kbw ];
+ f_BE = (D.f[DIR_P0M])[kbe ];
+ f_TW = (D.f[DIR_M0P])[ktw ];
+ f_TN = (D.f[DIR_0PP])[ktn ];
+ f_BS = (D.f[DIR_0MM])[kbs ];
+ f_BN = (D.f[DIR_0PM])[kbn ];
+ f_TS = (D.f[DIR_0MP])[kts ];
+ f_TNE = (D.f[DIR_PPP])[ktne ];
+ f_TSW = (D.f[DIR_MMP])[ktsw ];
+ f_TSE = (D.f[DIR_PMP])[ktse ];
+ f_TNW = (D.f[DIR_MPP])[ktnw ];
+ f_BNE = (D.f[DIR_PPM])[kbne ];
+ f_BSW = (D.f[DIR_MMM])[kbsw ];
+ f_BSE = (D.f[DIR_PMM])[kbse ];
+ f_BNW = (D.f[DIR_MPM])[kbnw ];
+ //f_W = (D.f[DIR_P00])[ke ];
+ //f_E = (D.f[DIR_M00])[kw ];
+ //f_S = (D.f[DIR_0P0])[kn ];
+ //f_N = (D.f[DIR_0M0])[ks ];
+ //f_B = (D.f[DIR_00P])[kt ];
+ //f_T = (D.f[DIR_00M])[kb ];
+ //f_SW = (D.f[DIR_PP0])[kne ];
+ //f_NE = (D.f[DIR_MM0])[ksw ];
+ //f_NW = (D.f[DIR_PM0])[kse ];
+ //f_SE = (D.f[DIR_MP0])[knw ];
+ //f_BW = (D.f[DIR_P0P])[kte ];
+ //f_TE = (D.f[DIR_M0M])[kbw ];
+ //f_TW = (D.f[DIR_P0M])[kbe ];
+ //f_BE = (D.f[DIR_M0P])[ktw ];
+ //f_BS = (D.f[DIR_0PP])[ktn ];
+ //f_TN = (D.f[DIR_0MM])[kbs ];
+ //f_TS = (D.f[DIR_0PM])[kbn ];
+ //f_BN = (D.f[DIR_0MP])[kts ];
+ //f_BSW = (D.f[DIR_PPP])[ktne ];
+ //f_BNE = (D.f[DIR_MMP])[ktsw ];
+ //f_BNW = (D.f[DIR_PMP])[ktse ];
+ //f_BSE = (D.f[DIR_MPP])[ktnw ];
+ //f_TSW = (D.f[DIR_PPM])[kbne ];
+ //f_TNE = (D.f[DIR_MMM])[kbsw ];
+ //f_TNW = (D.f[DIR_PMM])[kbse ];
+ //f_TSE = (D.f[DIR_MPM])[kbnw ];
////////////////////////////////////////////////////////////////////////////////
real vx1, vx2, vx3, drho, feq, q;
drho = f_TSE + f_TNW + f_TNE + f_TSW + f_BSE + f_BNW + f_BNE + f_BSW +
@@ -1852,63 +1877,63 @@ __global__ void QVelDeviceCompHighNu27(
//////////////////////////////////////////////////////////////////////////
if (isEvenTimestep==false)
{
- D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
+ D.f[DIR_M00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_MPM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_PMP * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
//Test
@@ -2194,39 +2219,32 @@ __global__ void QVelDeviceCompHighNu27(
//////////////////////////////////////////////////////////////////////////////
__global__ void QVelDeviceCompZeroPress27(
- real* velocityX,
- real* velocityY,
- real* velocityZ,
- real* distribution,
- int* subgridDistanceIndices,
- real* subgridDistances,
- unsigned int numberOfBCnodes,
- real omega,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int numberOfLBnodes,
- bool isEvenTimestep)
+ real* velocityX,
+ real* velocityY,
+ real* velocityZ,
+ real* distribution,
+ int* subgridDistanceIndices,
+ real* subgridDistances,
+ unsigned int numberOfBCnodes,
+ real omega,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep)
{
//////////////////////////////////////////////////////////////////////////
- //! The velocity boundary condition is executed in the following steps
- //!
- ////////////////////////////////////////////////////////////////////////////////
- //! - Get node index coordinates from threadIdx, blockIdx, blockDim and gridDim.
- //!
- const unsigned x = threadIdx.x; // global x-index
- const unsigned y = blockIdx.x; // global y-index
- const unsigned z = blockIdx.y; // global z-index
-
- const unsigned nx = blockDim.x;
- const unsigned ny = gridDim.x;
-
- const unsigned k = nx*(ny*z + y) + x;
+ //! The velocity boundary condition is executed in the following steps
+ //!
+ ////////////////////////////////////////////////////////////////////////////////
+ //! - Get node index coordinates from threadIdx, blockIdx, blockDim and gridDim.
+ //!
+ const unsigned nodeIndex = getNodeIndex();
//////////////////////////////////////////////////////////////////////////
//! - Run for all indices in size of boundary condition (numberOfBCnodes)
//!
- if(k < numberOfBCnodes)
+ if(nodeIndex < numberOfBCnodes)
{
//////////////////////////////////////////////////////////////////////////
@@ -2239,9 +2257,9 @@ __global__ void QVelDeviceCompZeroPress27(
////////////////////////////////////////////////////////////////////////////////
//! - Set local velocities
//!
- real VeloX = velocityX[k];
- real VeloY = velocityY[k];
- real VeloZ = velocityZ[k];
+ real VeloX = velocityX[nodeIndex];
+ real VeloY = velocityY[nodeIndex];
+ real VeloZ = velocityZ[nodeIndex];
////////////////////////////////////////////////////////////////////////////////
@@ -2253,7 +2271,7 @@ __global__ void QVelDeviceCompZeroPress27(
////////////////////////////////////////////////////////////////////////////////
//! - Set neighbor indices (necessary for indirect addressing)
//!
- unsigned int KQK = subgridDistanceIndices[k];
+ unsigned int KQK = subgridDistanceIndices[nodeIndex];
unsigned int kzero= KQK;
unsigned int ke = KQK;
unsigned int kw = neighborX[KQK];
@@ -2285,32 +2303,32 @@ __global__ void QVelDeviceCompZeroPress27(
////////////////////////////////////////////////////////////////////////////////
//! - Set local distributions
//!
- real f_W = (dist.f[DIR_P00 ])[ke ];
- real f_E = (dist.f[DIR_M00 ])[kw ];
- real f_S = (dist.f[DIR_0P0 ])[kn ];
- real f_N = (dist.f[DIR_0M0 ])[ks ];
- real f_B = (dist.f[DIR_00P ])[kt ];
- real f_T = (dist.f[DIR_00M ])[kb ];
- real f_SW = (dist.f[DIR_PP0 ])[kne ];
- real f_NE = (dist.f[DIR_MM0 ])[ksw ];
- real f_NW = (dist.f[DIR_PM0 ])[kse ];
- real f_SE = (dist.f[DIR_MP0 ])[knw ];
- real f_BW = (dist.f[DIR_P0P ])[kte ];
- real f_TE = (dist.f[DIR_M0M ])[kbw ];
- real f_TW = (dist.f[DIR_P0M ])[kbe ];
- real f_BE = (dist.f[DIR_M0P ])[ktw ];
- real f_BS = (dist.f[DIR_0PP ])[ktn ];
- real f_TN = (dist.f[DIR_0MM ])[kbs ];
- real f_TS = (dist.f[DIR_0PM ])[kbn ];
- real f_BN = (dist.f[DIR_0MP ])[kts ];
- real f_BSW = (dist.f[DIR_PPP ])[ktne ];
- real f_BNE = (dist.f[DIR_MMP ])[ktsw ];
- real f_BNW = (dist.f[DIR_PMP ])[ktse ];
- real f_BSE = (dist.f[DIR_MPP ])[ktnw ];
- real f_TSW = (dist.f[DIR_PPM ])[kbne ];
- real f_TNE = (dist.f[DIR_MMM ])[kbsw ];
- real f_TNW = (dist.f[DIR_PMM ])[kbse ];
- real f_TSE = (dist.f[DIR_MPM ])[kbnw ];
+ real f_W = (dist.f[DIR_P00])[ke ];
+ real f_E = (dist.f[DIR_M00])[kw ];
+ real f_S = (dist.f[DIR_0P0])[kn ];
+ real f_N = (dist.f[DIR_0M0])[ks ];
+ real f_B = (dist.f[DIR_00P])[kt ];
+ real f_T = (dist.f[DIR_00M])[kb ];
+ real f_SW = (dist.f[DIR_PP0])[kne ];
+ real f_NE = (dist.f[DIR_MM0])[ksw ];
+ real f_NW = (dist.f[DIR_PM0])[kse ];
+ real f_SE = (dist.f[DIR_MP0])[knw ];
+ real f_BW = (dist.f[DIR_P0P])[kte ];
+ real f_TE = (dist.f[DIR_M0M])[kbw ];
+ real f_TW = (dist.f[DIR_P0M])[kbe ];
+ real f_BE = (dist.f[DIR_M0P])[ktw ];
+ real f_BS = (dist.f[DIR_0PP])[ktn ];
+ real f_TN = (dist.f[DIR_0MM])[kbs ];
+ real f_TS = (dist.f[DIR_0PM])[kbn ];
+ real f_BN = (dist.f[DIR_0MP])[kts ];
+ real f_BSW = (dist.f[DIR_PPP])[ktne ];
+ real f_BNE = (dist.f[DIR_MMP])[ktsw ];
+ real f_BNW = (dist.f[DIR_PMP])[ktse ];
+ real f_BSE = (dist.f[DIR_MPP])[ktnw ];
+ real f_TSW = (dist.f[DIR_PPM])[kbne ];
+ real f_TNE = (dist.f[DIR_MMM])[kbsw ];
+ real f_TNW = (dist.f[DIR_PMM])[kbse ];
+ real f_TSE = (dist.f[DIR_MPM])[kbnw ];
////////////////////////////////////////////////////////////////////////////////
//! - Calculate macroscopic quantities
@@ -2342,7 +2360,7 @@ __global__ void QVelDeviceCompZeroPress27(
////////////////////////////////////////////////////////////////////////////////
//! - Update distributions with subgrid distance (q) between zero and one
real feq, q, velocityLB, velocityBC;
- q = (subgridD.q[DIR_P00])[k];
+ q = (subgridD.q[DIR_P00])[nodeIndex];
if (q>=c0o1 && q<=c1o1) // only update distribution for q between zero and one
{
velocityLB = vx1;
@@ -2351,7 +2369,7 @@ __global__ void QVelDeviceCompZeroPress27(
(dist.f[DIR_M00])[kw] = getInterpolatedDistributionForVeloWithPressureBC(q, f_E, f_W, feq, omega, drho, velocityBC, c2o27);
}
- q = (subgridD.q[DIR_M00])[k];
+ q = (subgridD.q[DIR_M00])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = -vx1;
@@ -2360,7 +2378,7 @@ __global__ void QVelDeviceCompZeroPress27(
(dist.f[DIR_P00])[ke] = getInterpolatedDistributionForVeloWithPressureBC(q, f_W, f_E, feq, omega, drho, velocityBC, c2o27);
}
- q = (subgridD.q[DIR_0P0])[k];
+ q = (subgridD.q[DIR_0P0])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = vx2;
@@ -2369,7 +2387,7 @@ __global__ void QVelDeviceCompZeroPress27(
(dist.f[DIR_0M0])[ks] = getInterpolatedDistributionForVeloWithPressureBC(q, f_N, f_S, feq, omega, drho, velocityBC, c2o27);
}
- q = (subgridD.q[DIR_0M0])[k];
+ q = (subgridD.q[DIR_0M0])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = -vx2;
@@ -2378,7 +2396,7 @@ __global__ void QVelDeviceCompZeroPress27(
(dist.f[DIR_0P0])[kn] = getInterpolatedDistributionForVeloWithPressureBC(q, f_S, f_N, feq, omega, drho, velocityBC, c2o27);
}
- q = (subgridD.q[DIR_00P])[k];
+ q = (subgridD.q[DIR_00P])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = vx3;
@@ -2387,7 +2405,7 @@ __global__ void QVelDeviceCompZeroPress27(
(dist.f[DIR_00M])[kb] = getInterpolatedDistributionForVeloWithPressureBC(q, f_T, f_B, feq, omega, drho, velocityBC, c2o27);
}
- q = (subgridD.q[DIR_00M])[k];
+ q = (subgridD.q[DIR_00M])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = -vx3;
@@ -2396,7 +2414,7 @@ __global__ void QVelDeviceCompZeroPress27(
(dist.f[DIR_00P])[kt] = getInterpolatedDistributionForVeloWithPressureBC(q, f_B, f_T, feq, omega, drho, velocityBC, c2o27);
}
- q = (subgridD.q[DIR_PP0])[k];
+ q = (subgridD.q[DIR_PP0])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = vx1 + vx2;
@@ -2405,7 +2423,7 @@ __global__ void QVelDeviceCompZeroPress27(
(dist.f[DIR_MM0])[ksw] = getInterpolatedDistributionForVeloWithPressureBC(q, f_NE, f_SW, feq, omega, drho, velocityBC, c1o54);
}
- q = (subgridD.q[DIR_MM0])[k];
+ q = (subgridD.q[DIR_MM0])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = -vx1 - vx2;
@@ -2414,7 +2432,7 @@ __global__ void QVelDeviceCompZeroPress27(
(dist.f[DIR_PP0])[kne] = getInterpolatedDistributionForVeloWithPressureBC(q, f_SW, f_NE, feq, omega, drho, velocityBC, c1o54);
}
- q = (subgridD.q[DIR_PM0])[k];
+ q = (subgridD.q[DIR_PM0])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = vx1 - vx2;
@@ -2423,7 +2441,7 @@ __global__ void QVelDeviceCompZeroPress27(
(dist.f[DIR_MP0])[knw] = getInterpolatedDistributionForVeloWithPressureBC(q, f_SE, f_NW, feq, omega, drho, velocityBC, c1o54);
}
- q = (subgridD.q[DIR_MP0])[k];
+ q = (subgridD.q[DIR_MP0])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = -vx1 + vx2;
@@ -2432,7 +2450,7 @@ __global__ void QVelDeviceCompZeroPress27(
(dist.f[DIR_PM0])[kse] = getInterpolatedDistributionForVeloWithPressureBC(q, f_NW, f_SE, feq, omega, drho, velocityBC, c1o54);
}
- q = (subgridD.q[DIR_P0P])[k];
+ q = (subgridD.q[DIR_P0P])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = vx1 + vx3;
@@ -2441,7 +2459,7 @@ __global__ void QVelDeviceCompZeroPress27(
(dist.f[DIR_M0M])[kbw] = getInterpolatedDistributionForVeloWithPressureBC(q, f_TE, f_BW, feq, omega, drho, velocityBC, c1o54);
}
- q = (subgridD.q[DIR_M0M])[k];
+ q = (subgridD.q[DIR_M0M])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = -vx1 - vx3;
@@ -2450,7 +2468,7 @@ __global__ void QVelDeviceCompZeroPress27(
(dist.f[DIR_P0P])[kte] = getInterpolatedDistributionForVeloWithPressureBC(q, f_BW, f_TE, feq, omega, drho, velocityBC, c1o54);
}
- q = (subgridD.q[DIR_P0M])[k];
+ q = (subgridD.q[DIR_P0M])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = vx1 - vx3;
@@ -2459,7 +2477,7 @@ __global__ void QVelDeviceCompZeroPress27(
(dist.f[DIR_M0P])[ktw] = getInterpolatedDistributionForVeloWithPressureBC(q, f_BE, f_TW, feq, omega, drho, velocityBC, c1o54);
}
- q = (subgridD.q[DIR_M0P])[k];
+ q = (subgridD.q[DIR_M0P])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = -vx1 + vx3;
@@ -2468,7 +2486,7 @@ __global__ void QVelDeviceCompZeroPress27(
(dist.f[DIR_P0M])[kbe] = getInterpolatedDistributionForVeloWithPressureBC(q, f_TW, f_BE, feq, omega, drho, velocityBC, c1o54);
}
- q = (subgridD.q[DIR_0PP])[k];
+ q = (subgridD.q[DIR_0PP])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = vx2 + vx3;
@@ -2477,7 +2495,7 @@ __global__ void QVelDeviceCompZeroPress27(
(dist.f[DIR_0MM])[kbs] = getInterpolatedDistributionForVeloWithPressureBC(q, f_TN, f_BS, feq, omega, drho, velocityBC, c1o54);
}
- q = (subgridD.q[DIR_0MM])[k];
+ q = (subgridD.q[DIR_0MM])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = -vx2 - vx3;
@@ -2486,7 +2504,7 @@ __global__ void QVelDeviceCompZeroPress27(
(dist.f[DIR_0PP])[ktn] = getInterpolatedDistributionForVeloWithPressureBC(q, f_BS, f_TN, feq, omega, drho, velocityBC, c1o54);
}
- q = (subgridD.q[DIR_0PM])[k];
+ q = (subgridD.q[DIR_0PM])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = vx2 - vx3;
@@ -2495,7 +2513,7 @@ __global__ void QVelDeviceCompZeroPress27(
(dist.f[DIR_0MP])[kts] = getInterpolatedDistributionForVeloWithPressureBC(q, f_BN, f_TS, feq, omega, drho, velocityBC, c1o54);
}
- q = (subgridD.q[DIR_0MP])[k];
+ q = (subgridD.q[DIR_0MP])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = -vx2 + vx3;
@@ -2504,7 +2522,7 @@ __global__ void QVelDeviceCompZeroPress27(
(dist.f[DIR_0PM])[kbn] = getInterpolatedDistributionForVeloWithPressureBC(q, f_TS, f_BN, feq, omega, drho, velocityBC, c1o54);
}
- q = (subgridD.q[DIR_PPP])[k];
+ q = (subgridD.q[DIR_PPP])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = vx1 + vx2 + vx3;
@@ -2513,7 +2531,7 @@ __global__ void QVelDeviceCompZeroPress27(
(dist.f[DIR_MMM])[kbsw] = getInterpolatedDistributionForVeloWithPressureBC(q, f_TNE, f_BSW, feq, omega, drho, velocityBC, c1o216);
}
- q = (subgridD.q[DIR_MMM])[k];
+ q = (subgridD.q[DIR_MMM])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = -vx1 - vx2 - vx3;
@@ -2522,7 +2540,7 @@ __global__ void QVelDeviceCompZeroPress27(
(dist.f[DIR_PPP])[ktne] = getInterpolatedDistributionForVeloWithPressureBC(q, f_BSW, f_TNE, feq, omega, drho, velocityBC, c1o216);
}
- q = (subgridD.q[DIR_PPM])[k];
+ q = (subgridD.q[DIR_PPM])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = vx1 + vx2 - vx3;
@@ -2531,7 +2549,7 @@ __global__ void QVelDeviceCompZeroPress27(
(dist.f[DIR_MMP])[ktsw] = getInterpolatedDistributionForVeloWithPressureBC(q, f_BNE, f_TSW, feq, omega, drho, velocityBC, c1o216);
}
- q = (subgridD.q[DIR_MMP])[k];
+ q = (subgridD.q[DIR_MMP])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = -vx1 - vx2 + vx3;
@@ -2540,7 +2558,7 @@ __global__ void QVelDeviceCompZeroPress27(
(dist.f[DIR_PPM])[kbne] = getInterpolatedDistributionForVeloWithPressureBC(q, f_TSW, f_BNE, feq, omega, drho, velocityBC, c1o216);
}
- q = (subgridD.q[DIR_PMP])[k];
+ q = (subgridD.q[DIR_PMP])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = vx1 - vx2 + vx3;
@@ -2549,7 +2567,7 @@ __global__ void QVelDeviceCompZeroPress27(
(dist.f[DIR_MPM])[kbnw] = getInterpolatedDistributionForVeloWithPressureBC(q, f_TSE, f_BNW, feq, omega, drho, velocityBC, c1o216);
}
- q = (subgridD.q[DIR_MPM])[k];
+ q = (subgridD.q[DIR_MPM])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = -vx1 + vx2 - vx3;
@@ -2558,7 +2576,7 @@ __global__ void QVelDeviceCompZeroPress27(
(dist.f[DIR_PMP])[ktse] = getInterpolatedDistributionForVeloWithPressureBC(q, f_BNW, f_TSE, feq, omega, drho, velocityBC, c1o216);
}
- q = (subgridD.q[DIR_PMM])[k];
+ q = (subgridD.q[DIR_PMM])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = vx1 - vx2 - vx3;
@@ -2567,7 +2585,7 @@ __global__ void QVelDeviceCompZeroPress27(
(dist.f[DIR_MPP])[ktnw] = getInterpolatedDistributionForVeloWithPressureBC(q, f_BSE, f_TNW, feq, omega, drho, velocityBC, c1o216);
}
- q = (subgridD.q[DIR_MPP])[k];
+ q = (subgridD.q[DIR_MPP])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = -vx1 + vx2 + vx3;
@@ -2619,87 +2637,88 @@ __global__ void QVelDeviceCompZeroPress27(
//////////////////////////////////////////////////////////////////////////////
-__global__ void QVelDeviceCompZeroPress1h27( int inx,
- int iny,
- real* vx,
- real* vy,
- real* vz,
- real* DD,
- int* k_Q,
- real* QQ,
- unsigned int numberOfBCnodes,
- real om1,
- real Phi,
- real angularVelocity,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- real* coordX,
- real* coordY,
- real* coordZ,
- unsigned int size_Mat,
- bool isEvenTimestep)
+__global__ void QVelDeviceCompZeroPress1h27(
+ int inx,
+ int iny,
+ real* vx,
+ real* vy,
+ real* vz,
+ real* DD,
+ int* k_Q,
+ real* QQ,
+ unsigned int numberOfBCnodes,
+ real om1,
+ real Phi,
+ real angularVelocity,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ real* coordX,
+ real* coordY,
+ real* coordZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep)
{
Distributions27 D;
if (isEvenTimestep==true)
{
- D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
+ D.f[DIR_M00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_MPM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_PMP * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////
const unsigned x = threadIdx.x; // Globaler x-Index
@@ -2738,24 +2757,24 @@ __global__ void QVelDeviceCompZeroPress1h27( int inx,
*q_dirBE, *q_dirTW, *q_dirTN, *q_dirBS, *q_dirBN, *q_dirTS,
*q_dirTNE, *q_dirTSW, *q_dirTSE, *q_dirTNW, *q_dirBNE, *q_dirBSW,
*q_dirBSE, *q_dirBNW;
- q_dirE = &QQ[DIR_P00 * numberOfBCnodes];
- q_dirW = &QQ[DIR_M00 * numberOfBCnodes];
- q_dirN = &QQ[DIR_0P0 * numberOfBCnodes];
- q_dirS = &QQ[DIR_0M0 * numberOfBCnodes];
- q_dirT = &QQ[DIR_00P * numberOfBCnodes];
- q_dirB = &QQ[DIR_00M * numberOfBCnodes];
- q_dirNE = &QQ[DIR_PP0 * numberOfBCnodes];
- q_dirSW = &QQ[DIR_MM0 * numberOfBCnodes];
- q_dirSE = &QQ[DIR_PM0 * numberOfBCnodes];
- q_dirNW = &QQ[DIR_MP0 * numberOfBCnodes];
- q_dirTE = &QQ[DIR_P0P * numberOfBCnodes];
- q_dirBW = &QQ[DIR_M0M * numberOfBCnodes];
- q_dirBE = &QQ[DIR_P0M * numberOfBCnodes];
- q_dirTW = &QQ[DIR_M0P * numberOfBCnodes];
- q_dirTN = &QQ[DIR_0PP * numberOfBCnodes];
- q_dirBS = &QQ[DIR_0MM * numberOfBCnodes];
- q_dirBN = &QQ[DIR_0PM * numberOfBCnodes];
- q_dirTS = &QQ[DIR_0MP * numberOfBCnodes];
+ q_dirE = &QQ[DIR_P00 * numberOfBCnodes];
+ q_dirW = &QQ[DIR_M00 * numberOfBCnodes];
+ q_dirN = &QQ[DIR_0P0 * numberOfBCnodes];
+ q_dirS = &QQ[DIR_0M0 * numberOfBCnodes];
+ q_dirT = &QQ[DIR_00P * numberOfBCnodes];
+ q_dirB = &QQ[DIR_00M * numberOfBCnodes];
+ q_dirNE = &QQ[DIR_PP0 * numberOfBCnodes];
+ q_dirSW = &QQ[DIR_MM0 * numberOfBCnodes];
+ q_dirSE = &QQ[DIR_PM0 * numberOfBCnodes];
+ q_dirNW = &QQ[DIR_MP0 * numberOfBCnodes];
+ q_dirTE = &QQ[DIR_P0P * numberOfBCnodes];
+ q_dirBW = &QQ[DIR_M0M * numberOfBCnodes];
+ q_dirBE = &QQ[DIR_P0M * numberOfBCnodes];
+ q_dirTW = &QQ[DIR_M0P * numberOfBCnodes];
+ q_dirTN = &QQ[DIR_0PP * numberOfBCnodes];
+ q_dirBS = &QQ[DIR_0MM * numberOfBCnodes];
+ q_dirBN = &QQ[DIR_0PM * numberOfBCnodes];
+ q_dirTS = &QQ[DIR_0MP * numberOfBCnodes];
q_dirTNE = &QQ[DIR_PPP * numberOfBCnodes];
q_dirTSW = &QQ[DIR_MMP * numberOfBCnodes];
q_dirTSE = &QQ[DIR_PMP * numberOfBCnodes];
@@ -2797,63 +2816,63 @@ __global__ void QVelDeviceCompZeroPress1h27( int inx,
//////////////////////////////////////////////////////////////////////////
if (isEvenTimestep==false)
{
- D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
+ D.f[DIR_M00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_MPM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_PMP * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
real vx1, vx2, vx3, drho, feq, q, cu_sq;
@@ -3090,21 +3109,22 @@ __global__ void QVelDeviceCompZeroPress1h27( int inx,
//////////////////////////////////////////////////////////////////////////////
-__global__ void LB_BC_Vel_West_27( int nx,
- int ny,
- int nz,
- int itz,
- unsigned int* bcMatD,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- real* DD,
- unsigned int size_Mat,
- bool isEvenTimestep,
- real u0x,
- unsigned int grid_nx,
- unsigned int grid_ny,
- real om)
+__global__ void LB_BC_Vel_West_27(
+ int nx,
+ int ny,
+ int nz,
+ int itz,
+ unsigned int* bcMatD,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ real* DD,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep,
+ real u0x,
+ unsigned int grid_nx,
+ unsigned int grid_ny,
+ real om)
{
//thread-index
unsigned int ity = blockIdx.x;
@@ -3125,63 +3145,63 @@ __global__ void LB_BC_Vel_West_27( int nx,
Distributions27 D;
if (isEvenTimestep==true)
{
- D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
+ D.f[DIR_M00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_MPM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_PMP * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////
@@ -3300,33 +3320,33 @@ __global__ void LB_BC_Vel_West_27( int nx,
real f1_E,f1_W,f1_N,f1_S,f1_T,f1_B,f1_NE,f1_SW,f1_SE,f1_NW,f1_TE,f1_BW,f1_BE,f1_TW,f1_TN,f1_BS,f1_BN,f1_TS,f1_ZERO,
f1_TNE,f1_TSW,f1_TSE,f1_TNW,f1_BNE,f1_BSW,f1_BSE,f1_BNW;
- f1_W = (D.f[DIR_P00 ])[k1e ];
- f1_E = (D.f[DIR_M00 ])[k1w ];
- f1_S = (D.f[DIR_0P0 ])[k1n ];
- f1_N = (D.f[DIR_0M0 ])[k1s ];
- f1_B = (D.f[DIR_00P ])[k1t ];
- f1_T = (D.f[DIR_00M ])[k1b ];
- f1_SW = (D.f[DIR_PP0 ])[k1ne ];
- f1_NE = (D.f[DIR_MM0 ])[k1sw ];
- f1_NW = (D.f[DIR_PM0 ])[k1se ];
- f1_SE = (D.f[DIR_MP0 ])[k1nw ];
- f1_BW = (D.f[DIR_P0P ])[k1te ];
- f1_TE = (D.f[DIR_M0M ])[k1bw ];
- f1_TW = (D.f[DIR_P0M ])[k1be ];
- f1_BE = (D.f[DIR_M0P ])[k1tw ];
- f1_BS = (D.f[DIR_0PP ])[k1tn ];
- f1_TN = (D.f[DIR_0MM ])[k1bs ];
- f1_TS = (D.f[DIR_0PM ])[k1bn ];
- f1_BN = (D.f[DIR_0MP ])[k1ts ];
+ f1_W = (D.f[DIR_P00])[k1e ];
+ f1_E = (D.f[DIR_M00])[k1w ];
+ f1_S = (D.f[DIR_0P0])[k1n ];
+ f1_N = (D.f[DIR_0M0])[k1s ];
+ f1_B = (D.f[DIR_00P])[k1t ];
+ f1_T = (D.f[DIR_00M])[k1b ];
+ f1_SW = (D.f[DIR_PP0])[k1ne ];
+ f1_NE = (D.f[DIR_MM0])[k1sw ];
+ f1_NW = (D.f[DIR_PM0])[k1se ];
+ f1_SE = (D.f[DIR_MP0])[k1nw ];
+ f1_BW = (D.f[DIR_P0P])[k1te ];
+ f1_TE = (D.f[DIR_M0M])[k1bw ];
+ f1_TW = (D.f[DIR_P0M])[k1be ];
+ f1_BE = (D.f[DIR_M0P])[k1tw ];
+ f1_BS = (D.f[DIR_0PP])[k1tn ];
+ f1_TN = (D.f[DIR_0MM])[k1bs ];
+ f1_TS = (D.f[DIR_0PM])[k1bn ];
+ f1_BN = (D.f[DIR_0MP])[k1ts ];
f1_ZERO = (D.f[DIR_000])[k1zero];
- f1_BSW = (D.f[DIR_PPP ])[k1tne ];
- f1_BNE = (D.f[DIR_MMP ])[k1tsw ];
- f1_BNW = (D.f[DIR_PMP ])[k1tse ];
- f1_BSE = (D.f[DIR_MPP ])[k1tnw ];
- f1_TSW = (D.f[DIR_PPM ])[k1bne ];
- f1_TNE = (D.f[DIR_MMM ])[k1bsw ];
- f1_TNW = (D.f[DIR_PMM ])[k1bse ];
- f1_TSE = (D.f[DIR_MPM ])[k1bnw ];
+ f1_BSW = (D.f[DIR_PPP])[k1tne ];
+ f1_BNE = (D.f[DIR_MMP])[k1tsw ];
+ f1_BNW = (D.f[DIR_PMP])[k1tse ];
+ f1_BSE = (D.f[DIR_MPP])[k1tnw ];
+ f1_TSW = (D.f[DIR_PPM])[k1bne ];
+ f1_TNE = (D.f[DIR_MMM])[k1bsw ];
+ f1_TNW = (D.f[DIR_PMM])[k1bse ];
+ f1_TSE = (D.f[DIR_MPM])[k1bnw ];
real drho1 = f1_ZERO+f1_E+f1_W+f1_N+f1_S+f1_T+f1_B+f1_NE+f1_SW+f1_SE+f1_NW+f1_TE+f1_BW+f1_BE+f1_TW+f1_TN+f1_BS+f1_BN+f1_TS+
f1_TNE+f1_TSW+f1_TSE+f1_TNW+f1_BNE+f1_BSW+f1_BSE+f1_BNW;
@@ -3343,32 +3363,32 @@ __global__ void LB_BC_Vel_West_27( int nx,
real cu_sq=c3o2*(vx1*vx1+vx2*vx2+vx3*vx3);
(D.f[DIR_000])[kzero] = c8o27* (drho-cu_sq);
- (D.f[DIR_P00 ])[ke ] = c2o27* (drho+c3o1*( vx1 )+c9o2*( vx1 )*( vx1 )-cu_sq);
- (D.f[DIR_M00 ])[kw ] = c2o27* (drho+c3o1*(-vx1 )+c9o2*(-vx1 )*(-vx1 )-cu_sq);
- (D.f[DIR_0P0 ])[kn ] = c2o27* (drho+c3o1*( vx2 )+c9o2*( vx2 )*( vx2 )-cu_sq);
- (D.f[DIR_0M0 ])[ks ] = c2o27* (drho+c3o1*( -vx2 )+c9o2*( -vx2 )*( -vx2 )-cu_sq);
- (D.f[DIR_00P ])[kt ] = c2o27* (drho+c3o1*( vx3)+c9o2*( vx3)*( vx3)-cu_sq);
- (D.f[DIR_00M ])[kb ] = c2o27* (drho+c3o1*( -vx3)+c9o2*( -vx3)*( -vx3)-cu_sq);
- (D.f[DIR_PP0 ])[kne ] = c1o54* (drho+c3o1*( vx1+vx2 )+c9o2*( vx1+vx2 )*( vx1+vx2 )-cu_sq);
- (D.f[DIR_MM0 ])[ksw ] = c1o54* (drho+c3o1*(-vx1-vx2 )+c9o2*(-vx1-vx2 )*(-vx1-vx2 )-cu_sq);
- (D.f[DIR_PM0 ])[kse ] = c1o54* (drho+c3o1*( vx1-vx2 )+c9o2*( vx1-vx2 )*( vx1-vx2 )-cu_sq);
- (D.f[DIR_MP0 ])[knw ] = c1o54* (drho+c3o1*(-vx1+vx2 )+c9o2*(-vx1+vx2 )*(-vx1+vx2 )-cu_sq);
- (D.f[DIR_P0P ])[kte ] = c1o54* (drho+c3o1*( vx1 +vx3)+c9o2*( vx1 +vx3)*( vx1 +vx3)-cu_sq);
- (D.f[DIR_M0M ])[kbw ] = c1o54* (drho+c3o1*(-vx1 -vx3)+c9o2*(-vx1 -vx3)*(-vx1 -vx3)-cu_sq);
- (D.f[DIR_P0M ])[kbe ] = c1o54* (drho+c3o1*( vx1 -vx3)+c9o2*( vx1 -vx3)*( vx1 -vx3)-cu_sq);
- (D.f[DIR_M0P ])[ktw ] = c1o54* (drho+c3o1*(-vx1 +vx3)+c9o2*(-vx1 +vx3)*(-vx1 +vx3)-cu_sq);
- (D.f[DIR_0PP ])[ktn ] = c1o54* (drho+c3o1*( vx2+vx3)+c9o2*( vx2+vx3)*( vx2+vx3)-cu_sq);
- (D.f[DIR_0MM ])[kbs ] = c1o54* (drho+c3o1*( -vx2-vx3)+c9o2*( -vx2-vx3)*( -vx2-vx3)-cu_sq);
- (D.f[DIR_0PM ])[kbn ] = c1o54* (drho+c3o1*( vx2-vx3)+c9o2*( vx2-vx3)*( vx2-vx3)-cu_sq);
- (D.f[DIR_0MP ])[kts ] = c1o54* (drho+c3o1*( -vx2+vx3)+c9o2*( -vx2+vx3)*( -vx2+vx3)-cu_sq);
- (D.f[DIR_PPP ])[ktne ] = c1o216*(drho+c3o1*( vx1+vx2+vx3)+c9o2*( vx1+vx2+vx3)*( vx1+vx2+vx3)-cu_sq);
- (D.f[DIR_MMM ])[kbsw ] = c1o216*(drho+c3o1*(-vx1-vx2-vx3)+c9o2*(-vx1-vx2-vx3)*(-vx1-vx2-vx3)-cu_sq);
- (D.f[DIR_PPM ])[kbne ] = c1o216*(drho+c3o1*( vx1+vx2-vx3)+c9o2*( vx1+vx2-vx3)*( vx1+vx2-vx3)-cu_sq);
- (D.f[DIR_MMP ])[ktsw ] = c1o216*(drho+c3o1*(-vx1-vx2+vx3)+c9o2*(-vx1-vx2+vx3)*(-vx1-vx2+vx3)-cu_sq);
- (D.f[DIR_PMP ])[ktse ] = c1o216*(drho+c3o1*( vx1-vx2+vx3)+c9o2*( vx1-vx2+vx3)*( vx1-vx2+vx3)-cu_sq);
- (D.f[DIR_MPM ])[kbnw ] = c1o216*(drho+c3o1*(-vx1+vx2-vx3)+c9o2*(-vx1+vx2-vx3)*(-vx1+vx2-vx3)-cu_sq);
- (D.f[DIR_PMM ])[kbse ] = c1o216*(drho+c3o1*( vx1-vx2-vx3)+c9o2*( vx1-vx2-vx3)*( vx1-vx2-vx3)-cu_sq);
- (D.f[DIR_MPP ])[ktnw ] = c1o216*(drho+c3o1*(-vx1+vx2+vx3)+c9o2*(-vx1+vx2+vx3)*(-vx1+vx2+vx3)-cu_sq);
+ (D.f[DIR_P00])[ke ] = c2o27* (drho+c3o1*( vx1 )+c9o2*( vx1 )*( vx1 )-cu_sq);
+ (D.f[DIR_M00])[kw ] = c2o27* (drho+c3o1*(-vx1 )+c9o2*(-vx1 )*(-vx1 )-cu_sq);
+ (D.f[DIR_0P0])[kn ] = c2o27* (drho+c3o1*( vx2 )+c9o2*( vx2 )*( vx2 )-cu_sq);
+ (D.f[DIR_0M0])[ks ] = c2o27* (drho+c3o1*( -vx2 )+c9o2*( -vx2 )*( -vx2 )-cu_sq);
+ (D.f[DIR_00P])[kt ] = c2o27* (drho+c3o1*( vx3)+c9o2*( vx3)*( vx3)-cu_sq);
+ (D.f[DIR_00M])[kb ] = c2o27* (drho+c3o1*( -vx3)+c9o2*( -vx3)*( -vx3)-cu_sq);
+ (D.f[DIR_PP0])[kne ] = c1o54* (drho+c3o1*( vx1+vx2 )+c9o2*( vx1+vx2 )*( vx1+vx2 )-cu_sq);
+ (D.f[DIR_MM0])[ksw ] = c1o54* (drho+c3o1*(-vx1-vx2 )+c9o2*(-vx1-vx2 )*(-vx1-vx2 )-cu_sq);
+ (D.f[DIR_PM0])[kse ] = c1o54* (drho+c3o1*( vx1-vx2 )+c9o2*( vx1-vx2 )*( vx1-vx2 )-cu_sq);
+ (D.f[DIR_MP0])[knw ] = c1o54* (drho+c3o1*(-vx1+vx2 )+c9o2*(-vx1+vx2 )*(-vx1+vx2 )-cu_sq);
+ (D.f[DIR_P0P])[kte ] = c1o54* (drho+c3o1*( vx1 +vx3)+c9o2*( vx1 +vx3)*( vx1 +vx3)-cu_sq);
+ (D.f[DIR_M0M])[kbw ] = c1o54* (drho+c3o1*(-vx1 -vx3)+c9o2*(-vx1 -vx3)*(-vx1 -vx3)-cu_sq);
+ (D.f[DIR_P0M])[kbe ] = c1o54* (drho+c3o1*( vx1 -vx3)+c9o2*( vx1 -vx3)*( vx1 -vx3)-cu_sq);
+ (D.f[DIR_M0P])[ktw ] = c1o54* (drho+c3o1*(-vx1 +vx3)+c9o2*(-vx1 +vx3)*(-vx1 +vx3)-cu_sq);
+ (D.f[DIR_0PP])[ktn ] = c1o54* (drho+c3o1*( vx2+vx3)+c9o2*( vx2+vx3)*( vx2+vx3)-cu_sq);
+ (D.f[DIR_0MM])[kbs ] = c1o54* (drho+c3o1*( -vx2-vx3)+c9o2*( -vx2-vx3)*( -vx2-vx3)-cu_sq);
+ (D.f[DIR_0PM])[kbn ] = c1o54* (drho+c3o1*( vx2-vx3)+c9o2*( vx2-vx3)*( vx2-vx3)-cu_sq);
+ (D.f[DIR_0MP])[kts ] = c1o54* (drho+c3o1*( -vx2+vx3)+c9o2*( -vx2+vx3)*( -vx2+vx3)-cu_sq);
+ (D.f[DIR_PPP])[ktne ] = c1o216*(drho+c3o1*( vx1+vx2+vx3)+c9o2*( vx1+vx2+vx3)*( vx1+vx2+vx3)-cu_sq);
+ (D.f[DIR_MMM])[kbsw ] = c1o216*(drho+c3o1*(-vx1-vx2-vx3)+c9o2*(-vx1-vx2-vx3)*(-vx1-vx2-vx3)-cu_sq);
+ (D.f[DIR_PPM])[kbne ] = c1o216*(drho+c3o1*( vx1+vx2-vx3)+c9o2*( vx1+vx2-vx3)*( vx1+vx2-vx3)-cu_sq);
+ (D.f[DIR_MMP])[ktsw ] = c1o216*(drho+c3o1*(-vx1-vx2+vx3)+c9o2*(-vx1-vx2+vx3)*(-vx1-vx2+vx3)-cu_sq);
+ (D.f[DIR_PMP])[ktse ] = c1o216*(drho+c3o1*( vx1-vx2+vx3)+c9o2*( vx1-vx2+vx3)*( vx1-vx2+vx3)-cu_sq);
+ (D.f[DIR_MPM])[kbnw ] = c1o216*(drho+c3o1*(-vx1+vx2-vx3)+c9o2*(-vx1+vx2-vx3)*(-vx1+vx2-vx3)-cu_sq);
+ (D.f[DIR_PMM])[kbse ] = c1o216*(drho+c3o1*( vx1-vx2-vx3)+c9o2*( vx1-vx2-vx3)*( vx1-vx2-vx3)-cu_sq);
+ (D.f[DIR_MPP])[ktnw ] = c1o216*(drho+c3o1*(-vx1+vx2+vx3)+c9o2*(-vx1+vx2+vx3)*(-vx1+vx2+vx3)-cu_sq);
}
__syncthreads();
}
@@ -3414,18 +3434,18 @@ __global__ void LB_BC_Vel_West_27( int nx,
//////////////////////////////////////////////////////////////////////////////
__global__ void QVelDevPlainBB27(
- real* velocityX,
- real* velocityY,
- real* velocityZ,
- real* distributions,
- int* subgridDistanceIndices,
- real* subgridDistances,
- uint numberOfBCnodes,
- uint* neighborX,
- uint* neighborY,
- uint* neighborZ,
- uint numberOfLBnodes,
- bool isEvenTimestep)
+ real* velocityX,
+ real* velocityY,
+ real* velocityZ,
+ real* distributions,
+ int* subgridDistanceIndices,
+ real* subgridDistances,
+ uint numberOfBCnodes,
+ uint* neighborX,
+ uint* neighborY,
+ uint* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep)
{
//////////////////////////////////////////////////////////////////////////
//! The velocity boundary condition is executed in the following steps
@@ -3433,18 +3453,11 @@ __global__ void QVelDevPlainBB27(
////////////////////////////////////////////////////////////////////////////////
//! - Get node index coordinates from threadIdx, blockIdx, blockDim and gridDim.
//!
- const unsigned x = threadIdx.x; // global x-index
- const unsigned y = blockIdx.x; // global y-index
- const unsigned z = blockIdx.y; // global z-index
-
- const unsigned nx = blockDim.x;
- const unsigned ny = gridDim.x;
-
- const unsigned k = nx*(ny*z + y) + x;
+ const unsigned nodeIndex = getNodeIndex();
//////////////////////////////////////////////////////////////////////////
// run for all indices in size of boundary condition (numberOfBCnodes)
- if(k < numberOfBCnodes)
+ if(nodeIndex < numberOfBCnodes)
{
//////////////////////////////////////////////////////////////////////////
//! - Read distributions: style of reading and writing the distributions from/to stored arrays dependent on timestep is based on the esoteric twist algorithm \ref
@@ -3456,9 +3469,9 @@ __global__ void QVelDevPlainBB27(
////////////////////////////////////////////////////////////////////////////////
//! - Set local velocities
//!
- real VeloX = velocityX[k];
- real VeloY = velocityY[k];
- real VeloZ = velocityZ[k];
+ real VeloX = velocityX[nodeIndex];
+ real VeloY = velocityY[nodeIndex];
+ real VeloZ = velocityZ[nodeIndex];
////////////////////////////////////////////////////////////////////////////////
//! - Set local subgrid distances (q's)
@@ -3469,7 +3482,7 @@ __global__ void QVelDevPlainBB27(
////////////////////////////////////////////////////////////////////////////////
//! - Set neighbor indices (necessary for indirect addressing)
//!
- uint indexOfBCnode = subgridDistanceIndices[k];
+ uint indexOfBCnode = subgridDistanceIndices[nodeIndex];
uint ke = indexOfBCnode;
uint kw = neighborX[indexOfBCnode];
uint kn = indexOfBCnode;
@@ -3500,32 +3513,32 @@ __global__ void QVelDevPlainBB27(
////////////////////////////////////////////////////////////////////////////////
//! - Set local distributions
//!
- real f_W = (dist.f[DIR_P00 ])[ke ];
- real f_E = (dist.f[DIR_M00 ])[kw ];
- real f_S = (dist.f[DIR_0P0 ])[kn ];
- real f_N = (dist.f[DIR_0M0 ])[ks ];
- real f_B = (dist.f[DIR_00P ])[kt ];
- real f_T = (dist.f[DIR_00M ])[kb ];
- real f_SW = (dist.f[DIR_PP0 ])[kne ];
- real f_NE = (dist.f[DIR_MM0 ])[ksw ];
- real f_NW = (dist.f[DIR_PM0 ])[kse ];
- real f_SE = (dist.f[DIR_MP0 ])[knw ];
- real f_BW = (dist.f[DIR_P0P ])[kte ];
- real f_TE = (dist.f[DIR_M0M ])[kbw ];
- real f_TW = (dist.f[DIR_P0M ])[kbe ];
- real f_BE = (dist.f[DIR_M0P ])[ktw ];
- real f_BS = (dist.f[DIR_0PP ])[ktn ];
- real f_TN = (dist.f[DIR_0MM ])[kbs ];
- real f_TS = (dist.f[DIR_0PM ])[kbn ];
- real f_BN = (dist.f[DIR_0MP ])[kts ];
- real f_BSW = (dist.f[DIR_PPP ])[ktne ];
- real f_BNE = (dist.f[DIR_MMP ])[ktsw ];
- real f_BNW = (dist.f[DIR_PMP ])[ktse ];
- real f_BSE = (dist.f[DIR_MPP ])[ktnw ];
- real f_TSW = (dist.f[DIR_PPM ])[kbne ];
- real f_TNE = (dist.f[DIR_MMM ])[kbsw ];
- real f_TNW = (dist.f[DIR_PMM ])[kbse ];
- real f_TSE = (dist.f[DIR_MPM ])[kbnw ];
+ real f_W = (dist.f[DIR_P00])[ke ];
+ real f_E = (dist.f[DIR_M00])[kw ];
+ real f_S = (dist.f[DIR_0P0])[kn ];
+ real f_N = (dist.f[DIR_0M0])[ks ];
+ real f_B = (dist.f[DIR_00P])[kt ];
+ real f_T = (dist.f[DIR_00M])[kb ];
+ real f_SW = (dist.f[DIR_PP0])[kne ];
+ real f_NE = (dist.f[DIR_MM0])[ksw ];
+ real f_NW = (dist.f[DIR_PM0])[kse ];
+ real f_SE = (dist.f[DIR_MP0])[knw ];
+ real f_BW = (dist.f[DIR_P0P])[kte ];
+ real f_TE = (dist.f[DIR_M0M])[kbw ];
+ real f_TW = (dist.f[DIR_P0M])[kbe ];
+ real f_BE = (dist.f[DIR_M0P])[ktw ];
+ real f_BS = (dist.f[DIR_0PP])[ktn ];
+ real f_TN = (dist.f[DIR_0MM])[kbs ];
+ real f_TS = (dist.f[DIR_0PM])[kbn ];
+ real f_BN = (dist.f[DIR_0MP])[kts ];
+ real f_BSW = (dist.f[DIR_PPP])[ktne ];
+ real f_BNE = (dist.f[DIR_MMP])[ktsw ];
+ real f_BNW = (dist.f[DIR_PMP])[ktse ];
+ real f_BSE = (dist.f[DIR_MPP])[ktnw ];
+ real f_TSW = (dist.f[DIR_PPM])[kbne ];
+ real f_TNE = (dist.f[DIR_MMM])[kbsw ];
+ real f_TNW = (dist.f[DIR_PMM])[kbse ];
+ real f_TSE = (dist.f[DIR_MPM])[kbnw ];
////////////////////////////////////////////////////////////////////////////////
//! - change the pointer to write the results in the correct array
@@ -3535,32 +3548,32 @@ __global__ void QVelDevPlainBB27(
////////////////////////////////////////////////////////////////////////////////
//! - rewrite distributions if there is a sub-grid distance (q) in same direction
real q;
- q = (subgridD.q[DIR_P00 ])[k]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_M00 ])[kw ]=f_E + c4o9 * (-VeloX);
- q = (subgridD.q[DIR_M00 ])[k]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_P00 ])[ke ]=f_W + c4o9 * ( VeloX);
- q = (subgridD.q[DIR_0P0 ])[k]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_0M0 ])[ks ]=f_N + c4o9 * (-VeloY);
- q = (subgridD.q[DIR_0M0 ])[k]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_0P0 ])[kn ]=f_S + c4o9 * ( VeloY);
- q = (subgridD.q[DIR_00P ])[k]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_00M ])[kb ]=f_T + c4o9 * (-VeloZ);
- q = (subgridD.q[DIR_00M ])[k]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_00P ])[kt ]=f_B + c4o9 * ( VeloZ);
- q = (subgridD.q[DIR_PP0 ])[k]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_MM0 ])[ksw ]=f_NE + c1o9 * (-VeloX - VeloY);
- q = (subgridD.q[DIR_MM0 ])[k]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_PP0 ])[kne ]=f_SW + c1o9 * ( VeloX + VeloY);
- q = (subgridD.q[DIR_PM0 ])[k]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_MP0 ])[knw ]=f_SE + c1o9 * (-VeloX + VeloY);
- q = (subgridD.q[DIR_MP0 ])[k]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_PM0 ])[kse ]=f_NW + c1o9 * ( VeloX - VeloY);
- q = (subgridD.q[DIR_P0P ])[k]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_M0M ])[kbw ]=f_TE + c1o9 * (-VeloX - VeloZ);
- q = (subgridD.q[DIR_M0M ])[k]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_P0P ])[kte ]=f_BW + c1o9 * ( VeloX + VeloZ);
- q = (subgridD.q[DIR_P0M ])[k]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_M0P ])[ktw ]=f_BE + c1o9 * (-VeloX + VeloZ);
- q = (subgridD.q[DIR_M0P ])[k]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_P0M ])[kbe ]=f_TW + c1o9 * ( VeloX - VeloZ);
- q = (subgridD.q[DIR_0PP ])[k]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_0MM ])[kbs ]=f_TN + c1o9 * (-VeloY - VeloZ);
- q = (subgridD.q[DIR_0MM ])[k]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_0PP ])[ktn ]=f_BS + c1o9 * ( VeloY + VeloZ);
- q = (subgridD.q[DIR_0PM ])[k]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_0MP ])[kts ]=f_BN + c1o9 * (-VeloY + VeloZ);
- q = (subgridD.q[DIR_0MP ])[k]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_0PM ])[kbn ]=f_TS + c1o9 * ( VeloY - VeloZ);
- q = (subgridD.q[DIR_PPP])[k]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_MMM])[kbsw]=f_TNE + c1o36 * (-VeloX - VeloY - VeloZ);
- q = (subgridD.q[DIR_MMM])[k]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_PPP])[ktne]=f_BSW + c1o36 * ( VeloX + VeloY + VeloZ);
- q = (subgridD.q[DIR_PPM])[k]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_MMP])[ktsw]=f_BNE + c1o36 * (-VeloX - VeloY + VeloZ);
- q = (subgridD.q[DIR_MMP])[k]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_PPM])[kbne]=f_TSW + c1o36 * ( VeloX + VeloY - VeloZ);
- q = (subgridD.q[DIR_PMP])[k]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_MPM])[kbnw]=f_TSE + c1o36 * (-VeloX + VeloY - VeloZ);
- q = (subgridD.q[DIR_MPM])[k]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_PMP])[ktse]=f_BNW + c1o36 * ( VeloX - VeloY + VeloZ);
- q = (subgridD.q[DIR_PMM])[k]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_MPP])[ktnw]=f_BSE + c1o36 * (-VeloX + VeloY + VeloZ);
- q = (subgridD.q[DIR_MPP])[k]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_PMM])[kbse]=f_TNW + c1o36 * ( VeloX - VeloY - VeloZ);
+ q = (subgridD.q[DIR_P00])[nodeIndex]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_M00])[kw ]=f_E + c4o9 * (-VeloX);
+ q = (subgridD.q[DIR_M00])[nodeIndex]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_P00])[ke ]=f_W + c4o9 * ( VeloX);
+ q = (subgridD.q[DIR_0P0])[nodeIndex]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_0M0])[ks ]=f_N + c4o9 * (-VeloY);
+ q = (subgridD.q[DIR_0M0])[nodeIndex]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_0P0])[kn ]=f_S + c4o9 * ( VeloY);
+ q = (subgridD.q[DIR_00P])[nodeIndex]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_00M])[kb ]=f_T + c4o9 * (-VeloZ);
+ q = (subgridD.q[DIR_00M])[nodeIndex]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_00P])[kt ]=f_B + c4o9 * ( VeloZ);
+ q = (subgridD.q[DIR_PP0])[nodeIndex]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_MM0])[ksw ]=f_NE + c1o9 * (-VeloX - VeloY);
+ q = (subgridD.q[DIR_MM0])[nodeIndex]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_PP0])[kne ]=f_SW + c1o9 * ( VeloX + VeloY);
+ q = (subgridD.q[DIR_PM0])[nodeIndex]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_MP0])[knw ]=f_SE + c1o9 * (-VeloX + VeloY);
+ q = (subgridD.q[DIR_MP0])[nodeIndex]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_PM0])[kse ]=f_NW + c1o9 * ( VeloX - VeloY);
+ q = (subgridD.q[DIR_P0P])[nodeIndex]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_M0M])[kbw ]=f_TE + c1o9 * (-VeloX - VeloZ);
+ q = (subgridD.q[DIR_M0M])[nodeIndex]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_P0P])[kte ]=f_BW + c1o9 * ( VeloX + VeloZ);
+ q = (subgridD.q[DIR_P0M])[nodeIndex]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_M0P])[ktw ]=f_BE + c1o9 * (-VeloX + VeloZ);
+ q = (subgridD.q[DIR_M0P])[nodeIndex]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_P0M])[kbe ]=f_TW + c1o9 * ( VeloX - VeloZ);
+ q = (subgridD.q[DIR_0PP])[nodeIndex]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_0MM])[kbs ]=f_TN + c1o9 * (-VeloY - VeloZ);
+ q = (subgridD.q[DIR_0MM])[nodeIndex]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_0PP])[ktn ]=f_BS + c1o9 * ( VeloY + VeloZ);
+ q = (subgridD.q[DIR_0PM])[nodeIndex]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_0MP])[kts ]=f_BN + c1o9 * (-VeloY + VeloZ);
+ q = (subgridD.q[DIR_0MP])[nodeIndex]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_0PM])[kbn ]=f_TS + c1o9 * ( VeloY - VeloZ);
+ q = (subgridD.q[DIR_PPP])[nodeIndex]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_MMM])[kbsw]=f_TNE + c1o36 * (-VeloX - VeloY - VeloZ);
+ q = (subgridD.q[DIR_MMM])[nodeIndex]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_PPP])[ktne]=f_BSW + c1o36 * ( VeloX + VeloY + VeloZ);
+ q = (subgridD.q[DIR_PPM])[nodeIndex]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_MMP])[ktsw]=f_BNE + c1o36 * (-VeloX - VeloY + VeloZ);
+ q = (subgridD.q[DIR_MMP])[nodeIndex]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_PPM])[kbne]=f_TSW + c1o36 * ( VeloX + VeloY - VeloZ);
+ q = (subgridD.q[DIR_PMP])[nodeIndex]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_MPM])[kbnw]=f_TSE + c1o36 * (-VeloX + VeloY - VeloZ);
+ q = (subgridD.q[DIR_MPM])[nodeIndex]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_PMP])[ktse]=f_BNW + c1o36 * ( VeloX - VeloY + VeloZ);
+ q = (subgridD.q[DIR_PMM])[nodeIndex]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_MPP])[ktnw]=f_BSE + c1o36 * (-VeloX + VeloY + VeloZ);
+ q = (subgridD.q[DIR_MPP])[nodeIndex]; if (q>=c0o1 && q<=c1o1) (dist.f[DIR_PMM])[kbse]=f_TNW + c1o36 * ( VeloX - VeloY - VeloZ);
}
}
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
@@ -3604,80 +3617,81 @@ __global__ void QVelDevPlainBB27(
//////////////////////////////////////////////////////////////////////////////
-__global__ void QVelDevCouette27(real* vx,
- real* vy,
- real* vz,
- real* DD,
- int* k_Q,
- real* QQ,
- unsigned int numberOfBCnodes,
- real om1,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep)
+__global__ void QVelDevCouette27(
+ real* vx,
+ real* vy,
+ real* vz,
+ real* DD,
+ int* k_Q,
+ real* QQ,
+ unsigned int numberOfBCnodes,
+ real om1,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep)
{
Distributions27 D;
if (isEvenTimestep==true)
{
- D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
+ D.f[DIR_M00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_MPM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_PMP * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////
const unsigned x = threadIdx.x; // Globaler x-Index
@@ -3702,24 +3716,24 @@ __global__ void QVelDevCouette27(real* vx,
*q_dirBE, *q_dirTW, *q_dirTN, *q_dirBS, *q_dirBN, *q_dirTS,
*q_dirTNE, *q_dirTSW, *q_dirTSE, *q_dirTNW, *q_dirBNE, *q_dirBSW,
*q_dirBSE, *q_dirBNW;
- q_dirE = &QQ[DIR_P00 * numberOfBCnodes];
- q_dirW = &QQ[DIR_M00 * numberOfBCnodes];
- q_dirN = &QQ[DIR_0P0 * numberOfBCnodes];
- q_dirS = &QQ[DIR_0M0 * numberOfBCnodes];
- q_dirT = &QQ[DIR_00P * numberOfBCnodes];
- q_dirB = &QQ[DIR_00M * numberOfBCnodes];
- q_dirNE = &QQ[DIR_PP0 * numberOfBCnodes];
- q_dirSW = &QQ[DIR_MM0 * numberOfBCnodes];
- q_dirSE = &QQ[DIR_PM0 * numberOfBCnodes];
- q_dirNW = &QQ[DIR_MP0 * numberOfBCnodes];
- q_dirTE = &QQ[DIR_P0P * numberOfBCnodes];
- q_dirBW = &QQ[DIR_M0M * numberOfBCnodes];
- q_dirBE = &QQ[DIR_P0M * numberOfBCnodes];
- q_dirTW = &QQ[DIR_M0P * numberOfBCnodes];
- q_dirTN = &QQ[DIR_0PP * numberOfBCnodes];
- q_dirBS = &QQ[DIR_0MM * numberOfBCnodes];
- q_dirBN = &QQ[DIR_0PM * numberOfBCnodes];
- q_dirTS = &QQ[DIR_0MP * numberOfBCnodes];
+ q_dirE = &QQ[DIR_P00 * numberOfBCnodes];
+ q_dirW = &QQ[DIR_M00 * numberOfBCnodes];
+ q_dirN = &QQ[DIR_0P0 * numberOfBCnodes];
+ q_dirS = &QQ[DIR_0M0 * numberOfBCnodes];
+ q_dirT = &QQ[DIR_00P * numberOfBCnodes];
+ q_dirB = &QQ[DIR_00M * numberOfBCnodes];
+ q_dirNE = &QQ[DIR_PP0 * numberOfBCnodes];
+ q_dirSW = &QQ[DIR_MM0 * numberOfBCnodes];
+ q_dirSE = &QQ[DIR_PM0 * numberOfBCnodes];
+ q_dirNW = &QQ[DIR_MP0 * numberOfBCnodes];
+ q_dirTE = &QQ[DIR_P0P * numberOfBCnodes];
+ q_dirBW = &QQ[DIR_M0M * numberOfBCnodes];
+ q_dirBE = &QQ[DIR_P0M * numberOfBCnodes];
+ q_dirTW = &QQ[DIR_M0P * numberOfBCnodes];
+ q_dirTN = &QQ[DIR_0PP * numberOfBCnodes];
+ q_dirBS = &QQ[DIR_0MM * numberOfBCnodes];
+ q_dirBN = &QQ[DIR_0PM * numberOfBCnodes];
+ q_dirTS = &QQ[DIR_0MP * numberOfBCnodes];
q_dirTNE = &QQ[DIR_PPP * numberOfBCnodes];
q_dirTSW = &QQ[DIR_MMP * numberOfBCnodes];
q_dirTSE = &QQ[DIR_PMP * numberOfBCnodes];
@@ -3761,94 +3775,94 @@ __global__ void QVelDevCouette27(real* vx,
////////////////////////////////////////////////////////////////////////////////
////////////////////////////////////////////////////////////////////////////////
- real f_W = (D.f[DIR_P00 ])[ke ];
- real f_E = (D.f[DIR_M00 ])[kw ];
- real f_S = (D.f[DIR_0P0 ])[kn ];
- real f_N = (D.f[DIR_0M0 ])[ks ];
- real f_B = (D.f[DIR_00P ])[kt ];
- real f_T = (D.f[DIR_00M ])[kb ];
- real f_SW = (D.f[DIR_PP0 ])[kne ];
- real f_NE = (D.f[DIR_MM0 ])[ksw ];
- real f_NW = (D.f[DIR_PM0 ])[kse ];
- real f_SE = (D.f[DIR_MP0 ])[knw ];
- real f_BW = (D.f[DIR_P0P ])[kte ];
- real f_TE = (D.f[DIR_M0M ])[kbw ];
- real f_TW = (D.f[DIR_P0M ])[kbe ];
- real f_BE = (D.f[DIR_M0P ])[ktw ];
- real f_BS = (D.f[DIR_0PP ])[ktn ];
- real f_TN = (D.f[DIR_0MM ])[kbs ];
- real f_TS = (D.f[DIR_0PM ])[kbn ];
- real f_BN = (D.f[DIR_0MP ])[kts ];
- real f_BSW = (D.f[DIR_PPP ])[ktne ];
- real f_BNE = (D.f[DIR_MMP ])[ktsw ];
- real f_BNW = (D.f[DIR_PMP ])[ktse ];
- real f_BSE = (D.f[DIR_MPP ])[ktnw ];
- real f_TSW = (D.f[DIR_PPM ])[kbne ];
- real f_TNE = (D.f[DIR_MMM ])[kbsw ];
- real f_TNW = (D.f[DIR_PMM ])[kbse ];
- real f_TSE = (D.f[DIR_MPM ])[kbnw ];
+ real f_W = (D.f[DIR_P00])[ke ];
+ real f_E = (D.f[DIR_M00])[kw ];
+ real f_S = (D.f[DIR_0P0])[kn ];
+ real f_N = (D.f[DIR_0M0])[ks ];
+ real f_B = (D.f[DIR_00P])[kt ];
+ real f_T = (D.f[DIR_00M])[kb ];
+ real f_SW = (D.f[DIR_PP0])[kne ];
+ real f_NE = (D.f[DIR_MM0])[ksw ];
+ real f_NW = (D.f[DIR_PM0])[kse ];
+ real f_SE = (D.f[DIR_MP0])[knw ];
+ real f_BW = (D.f[DIR_P0P])[kte ];
+ real f_TE = (D.f[DIR_M0M])[kbw ];
+ real f_TW = (D.f[DIR_P0M])[kbe ];
+ real f_BE = (D.f[DIR_M0P])[ktw ];
+ real f_BS = (D.f[DIR_0PP])[ktn ];
+ real f_TN = (D.f[DIR_0MM])[kbs ];
+ real f_TS = (D.f[DIR_0PM])[kbn ];
+ real f_BN = (D.f[DIR_0MP])[kts ];
+ real f_BSW = (D.f[DIR_PPP])[ktne ];
+ real f_BNE = (D.f[DIR_MMP])[ktsw ];
+ real f_BNW = (D.f[DIR_PMP])[ktse ];
+ real f_BSE = (D.f[DIR_MPP])[ktnw ];
+ real f_TSW = (D.f[DIR_PPM])[kbne ];
+ real f_TNE = (D.f[DIR_MMM])[kbsw ];
+ real f_TNW = (D.f[DIR_PMM])[kbse ];
+ real f_TSE = (D.f[DIR_MPM])[kbnw ];
////////////////////////////////////////////////////////////////////////////////
////////////////////////////////////////////////////////////////////////////////
if (isEvenTimestep==false)
{
- D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
+ D.f[DIR_M00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_MPM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_PMP * numberOfLBnodes];
}
//////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
/////// FlowDirection Y !!!!!!!!!! ///////////////////////////////////
@@ -3868,24 +3882,24 @@ __global__ void QVelDevCouette27(real* vx,
//////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
//set distributions
real q;
- q = q_dirE[k]; if (q>=c0o1 && q<=c1o1) (D.f[DIR_M00 ])[kw ]=f_E + ms*c2o27 * VeloX;
- q = q_dirW[k]; if (q>=c0o1 && q<=c1o1) (D.f[DIR_P00 ])[ke ]=f_W - ms*c2o27 * VeloX;
- q = q_dirN[k]; if (q>=c0o1 && q<=c1o1) (D.f[DIR_0M0 ])[ks ]=f_N + ms*c2o27 * VeloY;
- q = q_dirS[k]; if (q>=c0o1 && q<=c1o1) (D.f[DIR_0P0 ])[kn ]=f_S - ms*c2o27 * VeloY;
- q = q_dirT[k]; if (q>=c0o1 && q<=c1o1) (D.f[DIR_00M ])[kb ]=f_T + ms*c2o27 * VeloZ - c3o2*c2o27*((c2o1*VeloY-vx2)*(c2o1*VeloY-vx2)-vx2*vx2)*on;
- q = q_dirB[k]; if (q>=c0o1 && q<=c1o1) (D.f[DIR_00P ])[kt ]=f_B - ms*c2o27 * VeloZ;
- q = q_dirNE[k]; if (q>=c0o1 && q<=c1o1) (D.f[DIR_MM0 ])[ksw ]=f_NE + ms*c1o54 * VeloX + ms*c1o54 * VeloY;
- q = q_dirSW[k]; if (q>=c0o1 && q<=c1o1) (D.f[DIR_PP0 ])[kne ]=f_SW - ms*c1o54 * VeloX - ms*c1o54 * VeloY;
- q = q_dirSE[k]; if (q>=c0o1 && q<=c1o1) (D.f[DIR_MP0 ])[knw ]=f_SE + ms*c1o54 * VeloX - ms*c1o54 * VeloY;
- q = q_dirNW[k]; if (q>=c0o1 && q<=c1o1) (D.f[DIR_PM0 ])[kse ]=f_NW - ms*c1o54 * VeloX + ms*c1o54 * VeloY;
- q = q_dirTE[k]; if (q>=c0o1 && q<=c1o1) (D.f[DIR_M0M ])[kbw ]=f_TE + ms*c1o54 * VeloX + ms*c1o54 * VeloZ - c3o2*c1o54*((c2o1*VeloY-vx2)*(c2o1*VeloY-vx2)-vx2*vx2)*on-c1o12*kxxMyyFromfcNEQ;
- q = q_dirBW[k]; if (q>=c0o1 && q<=c1o1) (D.f[DIR_P0P ])[kte ]=f_BW - ms*c1o54 * VeloX - ms*c1o54 * VeloZ;
- q = q_dirBE[k]; if (q>=c0o1 && q<=c1o1) (D.f[DIR_M0P ])[ktw ]=f_BE + ms*c1o54 * VeloX - ms*c1o54 * VeloZ;
- q = q_dirTW[k]; if (q>=c0o1 && q<=c1o1) (D.f[DIR_P0M ])[kbe ]=f_TW - ms*c1o54 * VeloX + ms*c1o54 * VeloZ - c3o2*c1o54*((c2o1*VeloY-vx2)*(c2o1*VeloY-vx2)-vx2*vx2)*on-c1o12*kxxMyyFromfcNEQ;
- q = q_dirTN[k]; if (q>=c0o1 && q<=c1o1) (D.f[DIR_0MM ])[kbs ]=f_TN + ms*c1o54 * VeloY + ms*c1o54 * VeloZ + c3o1*c1o54*((c2o1*VeloY-vx2)*(c2o1*VeloY-vx2)-vx2*vx2)*on+c1o12*kxxMyyFromfcNEQ;
- q = q_dirBS[k]; if (q>=c0o1 && q<=c1o1) (D.f[DIR_0PP ])[ktn ]=f_BS - ms*c1o54 * VeloY - ms*c1o54 * VeloZ;
- q = q_dirBN[k]; if (q>=c0o1 && q<=c1o1) (D.f[DIR_0MP ])[kts ]=f_BN + ms*c1o54 * VeloY - ms*c1o54 * VeloZ;
- q = q_dirTS[k]; if (q>=c0o1 && q<=c1o1) (D.f[DIR_0PM ])[kbn ]=f_TS - ms*c1o54 * VeloY + ms*c1o54 * VeloZ + c3o1*c1o54*((c2o1*VeloY-vx2)*(c2o1*VeloY-vx2)-vx2*vx2)*on+c1o12*kxxMyyFromfcNEQ;
+ q = q_dirE[k]; if (q>=c0o1 && q<=c1o1) (D.f[DIR_M00])[kw ]=f_E + ms*c2o27 * VeloX;
+ q = q_dirW[k]; if (q>=c0o1 && q<=c1o1) (D.f[DIR_P00])[ke ]=f_W - ms*c2o27 * VeloX;
+ q = q_dirN[k]; if (q>=c0o1 && q<=c1o1) (D.f[DIR_0M0])[ks ]=f_N + ms*c2o27 * VeloY;
+ q = q_dirS[k]; if (q>=c0o1 && q<=c1o1) (D.f[DIR_0P0])[kn ]=f_S - ms*c2o27 * VeloY;
+ q = q_dirT[k]; if (q>=c0o1 && q<=c1o1) (D.f[DIR_00M])[kb ]=f_T + ms*c2o27 * VeloZ - c3o2*c2o27*((c2o1*VeloY-vx2)*(c2o1*VeloY-vx2)-vx2*vx2)*on;
+ q = q_dirB[k]; if (q>=c0o1 && q<=c1o1) (D.f[DIR_00P])[kt ]=f_B - ms*c2o27 * VeloZ;
+ q = q_dirNE[k]; if (q>=c0o1 && q<=c1o1) (D.f[DIR_MM0])[ksw ]=f_NE + ms*c1o54 * VeloX + ms*c1o54 * VeloY;
+ q = q_dirSW[k]; if (q>=c0o1 && q<=c1o1) (D.f[DIR_PP0])[kne ]=f_SW - ms*c1o54 * VeloX - ms*c1o54 * VeloY;
+ q = q_dirSE[k]; if (q>=c0o1 && q<=c1o1) (D.f[DIR_MP0])[knw ]=f_SE + ms*c1o54 * VeloX - ms*c1o54 * VeloY;
+ q = q_dirNW[k]; if (q>=c0o1 && q<=c1o1) (D.f[DIR_PM0])[kse ]=f_NW - ms*c1o54 * VeloX + ms*c1o54 * VeloY;
+ q = q_dirTE[k]; if (q>=c0o1 && q<=c1o1) (D.f[DIR_M0M])[kbw ]=f_TE + ms*c1o54 * VeloX + ms*c1o54 * VeloZ - c3o2*c1o54*((c2o1*VeloY-vx2)*(c2o1*VeloY-vx2)-vx2*vx2)*on-c1o12*kxxMyyFromfcNEQ;
+ q = q_dirBW[k]; if (q>=c0o1 && q<=c1o1) (D.f[DIR_P0P])[kte ]=f_BW - ms*c1o54 * VeloX - ms*c1o54 * VeloZ;
+ q = q_dirBE[k]; if (q>=c0o1 && q<=c1o1) (D.f[DIR_M0P])[ktw ]=f_BE + ms*c1o54 * VeloX - ms*c1o54 * VeloZ;
+ q = q_dirTW[k]; if (q>=c0o1 && q<=c1o1) (D.f[DIR_P0M])[kbe ]=f_TW - ms*c1o54 * VeloX + ms*c1o54 * VeloZ - c3o2*c1o54*((c2o1*VeloY-vx2)*(c2o1*VeloY-vx2)-vx2*vx2)*on-c1o12*kxxMyyFromfcNEQ;
+ q = q_dirTN[k]; if (q>=c0o1 && q<=c1o1) (D.f[DIR_0MM])[kbs ]=f_TN + ms*c1o54 * VeloY + ms*c1o54 * VeloZ + c3o1*c1o54*((c2o1*VeloY-vx2)*(c2o1*VeloY-vx2)-vx2*vx2)*on+c1o12*kxxMyyFromfcNEQ;
+ q = q_dirBS[k]; if (q>=c0o1 && q<=c1o1) (D.f[DIR_0PP])[ktn ]=f_BS - ms*c1o54 * VeloY - ms*c1o54 * VeloZ;
+ q = q_dirBN[k]; if (q>=c0o1 && q<=c1o1) (D.f[DIR_0MP])[kts ]=f_BN + ms*c1o54 * VeloY - ms*c1o54 * VeloZ;
+ q = q_dirTS[k]; if (q>=c0o1 && q<=c1o1) (D.f[DIR_0PM])[kbn ]=f_TS - ms*c1o54 * VeloY + ms*c1o54 * VeloZ + c3o1*c1o54*((c2o1*VeloY-vx2)*(c2o1*VeloY-vx2)-vx2*vx2)*on+c1o12*kxxMyyFromfcNEQ;
q = q_dirTNE[k]; if (q>=c0o1 && q<=c1o1) (D.f[DIR_MMM])[kbsw]=f_TNE + ms*c1o216 * VeloX + ms*c1o216 * VeloY + ms*c1o216 * VeloZ + c3o1*c1o216*((c2o1*VeloY-vx2)*(c2o1*VeloY-vx2)-vx2*vx2)*on;
q = q_dirBSW[k]; if (q>=c0o1 && q<=c1o1) (D.f[DIR_PPP])[ktne]=f_BSW - ms*c1o216 * VeloX - ms*c1o216 * VeloY - ms*c1o216 * VeloZ;
q = q_dirBNE[k]; if (q>=c0o1 && q<=c1o1) (D.f[DIR_MMP])[ktsw]=f_BNE + ms*c1o216 * VeloX + ms*c1o216 * VeloY - ms*c1o216 * VeloZ;
@@ -3894,24 +3908,24 @@ __global__ void QVelDevCouette27(real* vx,
q = q_dirBNW[k]; if (q>=c0o1 && q<=c1o1) (D.f[DIR_PMP])[ktse]=f_BNW - ms*c1o216 * VeloX + ms*c1o216 * VeloY - ms*c1o216 * VeloZ;
q = q_dirBSE[k]; if (q>=c0o1 && q<=c1o1) (D.f[DIR_MPP])[ktnw]=f_BSE + ms*c1o216 * VeloX - ms*c1o216 * VeloY - ms*c1o216 * VeloZ;
q = q_dirTNW[k]; if (q>=c0o1 && q<=c1o1) (D.f[DIR_PMM])[kbse]=f_TNW - ms*c1o216 * VeloX + ms*c1o216 * VeloY + ms*c1o216 * VeloZ + c3o1*c1o216*((c2o1*VeloY-vx2)*(c2o1*VeloY-vx2)-vx2*vx2)*on;
- //q = q_dirE[k]; if (q>=zero && q<=one) (D.f[DIR_M00 ])[kw ]=f_E + ms*c2over27 * VeloX;
- // q = q_dirW[k]; if (q>=zero && q<=one) (D.f[DIR_P00 ])[ke ]=f_W - ms*c2over27 * VeloX;
- // q = q_dirN[k]; if (q>=zero && q<=one) (D.f[DIR_0M0 ])[ks ]=f_N + ms*c2over27 * VeloY;
- // q = q_dirS[k]; if (q>=zero && q<=one) (D.f[DIR_0P0 ])[kn ]=f_S - ms*c2over27 * VeloY;
- //q = q_dirT[k]; if (q>=zero && q<=one) (D.f[DIR_00M ])[kb ]=f_T + ms*c2over27 * VeloZ - c1o9*((two*VeloY-vx2)*(two*VeloY-vx2)-vx2*vx2)*on;
- // q = q_dirB[k]; if (q>=zero && q<=one) (D.f[DIR_00P ])[kt ]=f_B - ms*c2over27 * VeloZ;
- // q = q_dirNE[k]; if (q>=zero && q<=one) (D.f[DIR_MM0 ])[ksw ]=f_NE + ms*c1over54 * VeloX + ms*c1over54 * VeloY;
- //q = q_dirSW[k]; if (q>=zero && q<=one) (D.f[DIR_PP0 ])[kne ]=f_SW - ms*c1over54 * VeloX - ms*c1over54 * VeloY;
- //q = q_dirSE[k]; if (q>=zero && q<=one) (D.f[DIR_MP0 ])[knw ]=f_SE + ms*c1over54 * VeloX - ms*c1over54 * VeloY;
- //q = q_dirNW[k]; if (q>=zero && q<=one) (D.f[DIR_PM0 ])[kse ]=f_NW - ms*c1over54 * VeloX + ms*c1over54 * VeloY;
- //q = q_dirTE[k]; if (q>=zero && q<=one) (D.f[DIR_M0M ])[kbw ]=f_TE + ms*c1over54 * VeloX + ms*c1over54 * VeloZ - c1o36*((two*VeloY-vx2)*(two*VeloY-vx2)-vx2*vx2)*on;
- //q = q_dirBW[k]; if (q>=zero && q<=one) (D.f[DIR_P0P ])[kte ]=f_BW - ms*c1over54 * VeloX - ms*c1over54 * VeloZ;
- //q = q_dirBE[k]; if (q>=zero && q<=one) (D.f[DIR_M0P ])[ktw ]=f_BE + ms*c1over54 * VeloX - ms*c1over54 * VeloZ;
- //q = q_dirTW[k]; if (q>=zero && q<=one) (D.f[DIR_P0M ])[kbe ]=f_TW - ms*c1over54 * VeloX + ms*c1over54 * VeloZ - c1o36*((two*VeloY-vx2)*(two*VeloY-vx2)-vx2*vx2)*on;
- //q = q_dirTN[k]; if (q>=zero && q<=one) (D.f[DIR_0MM ])[kbs ]=f_TN + ms*c1over54 * VeloY + ms*c1over54 * VeloZ + c1o2*c1o9*((two*VeloY-vx2)*(two*VeloY-vx2)-vx2*vx2)*on;
- //q = q_dirBS[k]; if (q>=zero && q<=one) (D.f[DIR_0PP ])[ktn ]=f_BS - ms*c1over54 * VeloY - ms*c1over54 * VeloZ;
- //q = q_dirBN[k]; if (q>=zero && q<=one) (D.f[DIR_0MP ])[kts ]=f_BN + ms*c1over54 * VeloY - ms*c1over54 * VeloZ;
- //q = q_dirTS[k]; if (q>=zero && q<=one) (D.f[DIR_0PM ])[kbn ]=f_TS - ms*c1over54 * VeloY + ms*c1over54 * VeloZ + c1o2*c1o9*((two*VeloY-vx2)*(two*VeloY-vx2)-vx2*vx2)*on;
+ //q = q_dirE[k]; if (q>=zero && q<=one) (D.f[DIR_M00])[kw ]=f_E + ms*c2over27 * VeloX;
+ // q = q_dirW[k]; if (q>=zero && q<=one) (D.f[DIR_P00])[ke ]=f_W - ms*c2over27 * VeloX;
+ // q = q_dirN[k]; if (q>=zero && q<=one) (D.f[DIR_0M0])[ks ]=f_N + ms*c2over27 * VeloY;
+ // q = q_dirS[k]; if (q>=zero && q<=one) (D.f[DIR_0P0])[kn ]=f_S - ms*c2over27 * VeloY;
+ //q = q_dirT[k]; if (q>=zero && q<=one) (D.f[DIR_00M])[kb ]=f_T + ms*c2over27 * VeloZ - c1o9*((two*VeloY-vx2)*(two*VeloY-vx2)-vx2*vx2)*on;
+ // q = q_dirB[k]; if (q>=zero && q<=one) (D.f[DIR_00P])[kt ]=f_B - ms*c2over27 * VeloZ;
+ // q = q_dirNE[k]; if (q>=zero && q<=one) (D.f[DIR_MM0])[ksw ]=f_NE + ms*c1over54 * VeloX + ms*c1over54 * VeloY;
+ //q = q_dirSW[k]; if (q>=zero && q<=one) (D.f[DIR_PP0])[kne ]=f_SW - ms*c1over54 * VeloX - ms*c1over54 * VeloY;
+ //q = q_dirSE[k]; if (q>=zero && q<=one) (D.f[DIR_MP0])[knw ]=f_SE + ms*c1over54 * VeloX - ms*c1over54 * VeloY;
+ //q = q_dirNW[k]; if (q>=zero && q<=one) (D.f[DIR_PM0])[kse ]=f_NW - ms*c1over54 * VeloX + ms*c1over54 * VeloY;
+ //q = q_dirTE[k]; if (q>=zero && q<=one) (D.f[DIR_M0M])[kbw ]=f_TE + ms*c1over54 * VeloX + ms*c1over54 * VeloZ - c1o36*((two*VeloY-vx2)*(two*VeloY-vx2)-vx2*vx2)*on;
+ //q = q_dirBW[k]; if (q>=zero && q<=one) (D.f[DIR_P0P])[kte ]=f_BW - ms*c1over54 * VeloX - ms*c1over54 * VeloZ;
+ //q = q_dirBE[k]; if (q>=zero && q<=one) (D.f[DIR_M0P])[ktw ]=f_BE + ms*c1over54 * VeloX - ms*c1over54 * VeloZ;
+ //q = q_dirTW[k]; if (q>=zero && q<=one) (D.f[DIR_P0M])[kbe ]=f_TW - ms*c1over54 * VeloX + ms*c1over54 * VeloZ - c1o36*((two*VeloY-vx2)*(two*VeloY-vx2)-vx2*vx2)*on;
+ //q = q_dirTN[k]; if (q>=zero && q<=one) (D.f[DIR_0MM])[kbs ]=f_TN + ms*c1over54 * VeloY + ms*c1over54 * VeloZ + c1o2*c1o9*((two*VeloY-vx2)*(two*VeloY-vx2)-vx2*vx2)*on;
+ //q = q_dirBS[k]; if (q>=zero && q<=one) (D.f[DIR_0PP])[ktn ]=f_BS - ms*c1over54 * VeloY - ms*c1over54 * VeloZ;
+ //q = q_dirBN[k]; if (q>=zero && q<=one) (D.f[DIR_0MP])[kts ]=f_BN + ms*c1over54 * VeloY - ms*c1over54 * VeloZ;
+ //q = q_dirTS[k]; if (q>=zero && q<=one) (D.f[DIR_0PM])[kbn ]=f_TS - ms*c1over54 * VeloY + ms*c1over54 * VeloZ + c1o2*c1o9*((two*VeloY-vx2)*(two*VeloY-vx2)-vx2*vx2)*on;
// q = q_dirTNE[k]; if (q>=zero && q<=one) (D.f[DIR_MMM])[kbsw]=f_TNE + ms*c1over216 * VeloX + ms*c1over216 * VeloY + ms*c1over216 * VeloZ + c1o2*c1o36*((two*VeloY-vx2)*(two*VeloY-vx2)-vx2*vx2)*on;
// q = q_dirBSW[k]; if (q>=zero && q<=one) (D.f[DIR_PPP])[ktne]=f_BSW - ms*c1over216 * VeloX - ms*c1over216 * VeloY - ms*c1over216 * VeloZ;
// q = q_dirBNE[k]; if (q>=zero && q<=one) (D.f[DIR_MMP])[ktsw]=f_BNE + ms*c1over216 * VeloX + ms*c1over216 * VeloY - ms*c1over216 * VeloZ;
@@ -3964,87 +3978,88 @@ __global__ void QVelDevCouette27(real* vx,
//////////////////////////////////////////////////////////////////////////////
-__global__ void QVelDev1h27( int inx,
- int iny,
- real* vx,
- real* vy,
- real* vz,
- real* DD,
- int* k_Q,
- real* QQ,
- unsigned int numberOfBCnodes,
- real om1,
- real Phi,
- real angularVelocity,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- real* coordX,
- real* coordY,
- real* coordZ,
- unsigned int size_Mat,
- bool isEvenTimestep)
+__global__ void QVelDev1h27(
+ int inx,
+ int iny,
+ real* vx,
+ real* vy,
+ real* vz,
+ real* DD,
+ int* k_Q,
+ real* QQ,
+ unsigned int numberOfBCnodes,
+ real om1,
+ real Phi,
+ real angularVelocity,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ real* coordX,
+ real* coordY,
+ real* coordZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep)
{
Distributions27 D;
if (isEvenTimestep==true)
{
- D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
+ D.f[DIR_M00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_MPM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_PMP * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////
const unsigned x = threadIdx.x; // Globaler x-Index
@@ -4079,24 +4094,24 @@ __global__ void QVelDev1h27( int inx,
*q_dirBE, *q_dirTW, *q_dirTN, *q_dirBS, *q_dirBN, *q_dirTS,
*q_dirTNE, *q_dirTSW, *q_dirTSE, *q_dirTNW, *q_dirBNE, *q_dirBSW,
*q_dirBSE, *q_dirBNW;
- q_dirE = &QQ[DIR_P00 * numberOfBCnodes];
- q_dirW = &QQ[DIR_M00 * numberOfBCnodes];
- q_dirN = &QQ[DIR_0P0 * numberOfBCnodes];
- q_dirS = &QQ[DIR_0M0 * numberOfBCnodes];
- q_dirT = &QQ[DIR_00P * numberOfBCnodes];
- q_dirB = &QQ[DIR_00M * numberOfBCnodes];
- q_dirNE = &QQ[DIR_PP0 * numberOfBCnodes];
- q_dirSW = &QQ[DIR_MM0 * numberOfBCnodes];
- q_dirSE = &QQ[DIR_PM0 * numberOfBCnodes];
- q_dirNW = &QQ[DIR_MP0 * numberOfBCnodes];
- q_dirTE = &QQ[DIR_P0P * numberOfBCnodes];
- q_dirBW = &QQ[DIR_M0M * numberOfBCnodes];
- q_dirBE = &QQ[DIR_P0M * numberOfBCnodes];
- q_dirTW = &QQ[DIR_M0P * numberOfBCnodes];
- q_dirTN = &QQ[DIR_0PP * numberOfBCnodes];
- q_dirBS = &QQ[DIR_0MM * numberOfBCnodes];
- q_dirBN = &QQ[DIR_0PM * numberOfBCnodes];
- q_dirTS = &QQ[DIR_0MP * numberOfBCnodes];
+ q_dirE = &QQ[DIR_P00 * numberOfBCnodes];
+ q_dirW = &QQ[DIR_M00 * numberOfBCnodes];
+ q_dirN = &QQ[DIR_0P0 * numberOfBCnodes];
+ q_dirS = &QQ[DIR_0M0 * numberOfBCnodes];
+ q_dirT = &QQ[DIR_00P * numberOfBCnodes];
+ q_dirB = &QQ[DIR_00M * numberOfBCnodes];
+ q_dirNE = &QQ[DIR_PP0 * numberOfBCnodes];
+ q_dirSW = &QQ[DIR_MM0 * numberOfBCnodes];
+ q_dirSE = &QQ[DIR_PM0 * numberOfBCnodes];
+ q_dirNW = &QQ[DIR_MP0 * numberOfBCnodes];
+ q_dirTE = &QQ[DIR_P0P * numberOfBCnodes];
+ q_dirBW = &QQ[DIR_M0M * numberOfBCnodes];
+ q_dirBE = &QQ[DIR_P0M * numberOfBCnodes];
+ q_dirTW = &QQ[DIR_M0P * numberOfBCnodes];
+ q_dirTN = &QQ[DIR_0PP * numberOfBCnodes];
+ q_dirBS = &QQ[DIR_0MM * numberOfBCnodes];
+ q_dirBN = &QQ[DIR_0PM * numberOfBCnodes];
+ q_dirTS = &QQ[DIR_0MP * numberOfBCnodes];
q_dirTNE = &QQ[DIR_PPP * numberOfBCnodes];
q_dirTSW = &QQ[DIR_MMP * numberOfBCnodes];
q_dirTSE = &QQ[DIR_PMP * numberOfBCnodes];
@@ -4167,32 +4182,32 @@ __global__ void QVelDev1h27( int inx,
//real f_E, f_W, f_N, f_S, f_T, f_B, f_NE, f_SW, f_SE, f_NW, f_TE, f_BW, f_BE,
// f_TW, f_TN, f_BS, f_BN, f_TS, f_TNE, f_TSW, f_TSE, f_TNW, f_BNE, f_BSW, f_BSE, f_BNW;
- //f_W = (D.f[DIR_P00 ])[ke ];
- //f_E = (D.f[DIR_M00 ])[kw ];
- //f_S = (D.f[DIR_0P0 ])[kn ];
- //f_N = (D.f[DIR_0M0 ])[ks ];
- //f_B = (D.f[DIR_00P ])[kt ];
- //f_T = (D.f[DIR_00M ])[kb ];
- //f_SW = (D.f[DIR_PP0 ])[kne ];
- //f_NE = (D.f[DIR_MM0 ])[ksw ];
- //f_NW = (D.f[DIR_PM0 ])[kse ];
- //f_SE = (D.f[DIR_MP0 ])[knw ];
- //f_BW = (D.f[DIR_P0P ])[kte ];
- //f_TE = (D.f[DIR_M0M ])[kbw ];
- //f_TW = (D.f[DIR_P0M ])[kbe ];
- //f_BE = (D.f[DIR_M0P ])[ktw ];
- //f_BS = (D.f[DIR_0PP ])[ktn ];
- //f_TN = (D.f[DIR_0MM ])[kbs ];
- //f_TS = (D.f[DIR_0PM ])[kbn ];
- //f_BN = (D.f[DIR_0MP ])[kts ];
- //f_BSW = (D.f[DIR_PPP ])[ktne ];
- //f_BNE = (D.f[DIR_MMP ])[ktsw ];
- //f_BNW = (D.f[DIR_PMP ])[ktse ];
- //f_BSE = (D.f[DIR_MPP ])[ktnw ];
- //f_TSW = (D.f[DIR_PPM ])[kbne ];
- //f_TNE = (D.f[DIR_MMM ])[kbsw ];
- //f_TNW = (D.f[DIR_PMM ])[kbse ];
- //f_TSE = (D.f[DIR_MPM ])[kbnw ];
+ //f_W = (D.f[DIR_P00])[ke ];
+ //f_E = (D.f[DIR_M00])[kw ];
+ //f_S = (D.f[DIR_0P0])[kn ];
+ //f_N = (D.f[DIR_0M0])[ks ];
+ //f_B = (D.f[DIR_00P])[kt ];
+ //f_T = (D.f[DIR_00M])[kb ];
+ //f_SW = (D.f[DIR_PP0])[kne ];
+ //f_NE = (D.f[DIR_MM0])[ksw ];
+ //f_NW = (D.f[DIR_PM0])[kse ];
+ //f_SE = (D.f[DIR_MP0])[knw ];
+ //f_BW = (D.f[DIR_P0P])[kte ];
+ //f_TE = (D.f[DIR_M0M])[kbw ];
+ //f_TW = (D.f[DIR_P0M])[kbe ];
+ //f_BE = (D.f[DIR_M0P])[ktw ];
+ //f_BS = (D.f[DIR_0PP])[ktn ];
+ //f_TN = (D.f[DIR_0MM])[kbs ];
+ //f_TS = (D.f[DIR_0PM])[kbn ];
+ //f_BN = (D.f[DIR_0MP])[kts ];
+ //f_BSW = (D.f[DIR_PPP])[ktne ];
+ //f_BNE = (D.f[DIR_MMP])[ktsw ];
+ //f_BNW = (D.f[DIR_PMP])[ktse ];
+ //f_BSE = (D.f[DIR_MPP])[ktnw ];
+ //f_TSW = (D.f[DIR_PPM])[kbne ];
+ //f_TNE = (D.f[DIR_MMM])[kbsw ];
+ //f_TNW = (D.f[DIR_PMM])[kbse ];
+ //f_TSE = (D.f[DIR_MPM])[kbnw ];
////////////////////////////////////////////////////////////////////////////////
real /*vx1, vx2,*/ vx3, drho, feq, q, cu_sq;
//drho = f_TSE + f_TNW + f_TNE + f_TSW + f_BSE + f_BNW + f_BNE + f_BSW +
@@ -4217,63 +4232,63 @@ __global__ void QVelDev1h27( int inx,
//////////////////////////////////////////////////////////////////////////
if (isEvenTimestep==false)
{
- D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
+ D.f[DIR_M00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_MPM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_PMP * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
//Test
@@ -4748,39 +4763,32 @@ __global__ void QVelDev1h27( int inx,
//////////////////////////////////////////////////////////////////////////////
__global__ void QVelDeviceComp27(
- real* velocityX,
- real* velocityY,
- real* velocityZ,
- real* distributions,
- int* subgridDistanceIndices,
- real* subgridDistances,
- unsigned int numberOfBCnodes,
- real omega,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int numberOfLBnodes,
- bool isEvenTimestep)
+ real* velocityX,
+ real* velocityY,
+ real* velocityZ,
+ real* distributions,
+ int* subgridDistanceIndices,
+ real* subgridDistances,
+ unsigned int numberOfBCnodes,
+ real omega,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep)
{
//////////////////////////////////////////////////////////////////////////
//! The velocity boundary condition is executed in the following steps
//!
- ////////////////////////////////////////////////////////////////////////////////
- //! - Get node index coordinates from threadIdx, blockIdx, blockDim and gridDim.
- //!
- const unsigned x = threadIdx.x; // global x-index
- const unsigned y = blockIdx.x; // global y-index
- const unsigned z = blockIdx.y; // global z-index
-
- const unsigned nx = blockDim.x;
- const unsigned ny = gridDim.x;
-
- const unsigned k = nx*(ny*z + y) + x;
+ ////////////////////////////////////////////////////////////////////////////////
+ //! - Get node index coordinates from threadIdx, blockIdx, blockDim and gridDim.
+ //!
+ const unsigned nodeIndex = getNodeIndex();
//////////////////////////////////////////////////////////////////////////
//! - Run for all indices in size of boundary condition (numberOfBCnodes)
//!
- if(k < numberOfBCnodes)
+ if(nodeIndex < numberOfBCnodes)
{
//////////////////////////////////////////////////////////////////////////
//! - Read distributions: style of reading and writing the distributions from/to stored arrays dependent on timestep is based on the esoteric twist algorithm \ref
@@ -4792,9 +4800,9 @@ __global__ void QVelDeviceComp27(
////////////////////////////////////////////////////////////////////////////////
//! - Set local velocities
//!
- real VeloX = velocityX[k];
- real VeloY = velocityY[k];
- real VeloZ = velocityZ[k];
+ real VeloX = velocityX[nodeIndex];
+ real VeloY = velocityY[nodeIndex];
+ real VeloZ = velocityZ[nodeIndex];
////////////////////////////////////////////////////////////////////////////////
//! - Set local subgrid distances (q's)
@@ -4805,7 +4813,7 @@ __global__ void QVelDeviceComp27(
////////////////////////////////////////////////////////////////////////////////
//! - Set neighbor indices (necessary for indirect addressing)
//!
- unsigned int indexOfBCnode = subgridDistanceIndices[k];
+ unsigned int indexOfBCnode = subgridDistanceIndices[nodeIndex];
unsigned int kzero= indexOfBCnode;
unsigned int ke = indexOfBCnode;
unsigned int kw = neighborX[indexOfBCnode];
@@ -4837,32 +4845,32 @@ __global__ void QVelDeviceComp27(
////////////////////////////////////////////////////////////////////////////////
//! - Set local distributions
//!
- real f_W = (dist.f[DIR_P00 ])[ke ];
- real f_E = (dist.f[DIR_M00 ])[kw ];
- real f_S = (dist.f[DIR_0P0 ])[kn ];
- real f_N = (dist.f[DIR_0M0 ])[ks ];
- real f_B = (dist.f[DIR_00P ])[kt ];
- real f_T = (dist.f[DIR_00M ])[kb ];
- real f_SW = (dist.f[DIR_PP0 ])[kne ];
- real f_NE = (dist.f[DIR_MM0 ])[ksw ];
- real f_NW = (dist.f[DIR_PM0 ])[kse ];
- real f_SE = (dist.f[DIR_MP0 ])[knw ];
- real f_BW = (dist.f[DIR_P0P ])[kte ];
- real f_TE = (dist.f[DIR_M0M ])[kbw ];
- real f_TW = (dist.f[DIR_P0M ])[kbe ];
- real f_BE = (dist.f[DIR_M0P ])[ktw ];
- real f_BS = (dist.f[DIR_0PP ])[ktn ];
- real f_TN = (dist.f[DIR_0MM ])[kbs ];
- real f_TS = (dist.f[DIR_0PM ])[kbn ];
- real f_BN = (dist.f[DIR_0MP ])[kts ];
- real f_BSW = (dist.f[DIR_PPP ])[ktne ];
- real f_BNE = (dist.f[DIR_MMP ])[ktsw ];
- real f_BNW = (dist.f[DIR_PMP ])[ktse ];
- real f_BSE = (dist.f[DIR_MPP ])[ktnw ];
- real f_TSW = (dist.f[DIR_PPM ])[kbne ];
- real f_TNE = (dist.f[DIR_MMM ])[kbsw ];
- real f_TNW = (dist.f[DIR_PMM ])[kbse ];
- real f_TSE = (dist.f[DIR_MPM ])[kbnw ];
+ real f_W = (dist.f[DIR_P00])[ke ];
+ real f_E = (dist.f[DIR_M00])[kw ];
+ real f_S = (dist.f[DIR_0P0])[kn ];
+ real f_N = (dist.f[DIR_0M0])[ks ];
+ real f_B = (dist.f[DIR_00P])[kt ];
+ real f_T = (dist.f[DIR_00M])[kb ];
+ real f_SW = (dist.f[DIR_PP0])[kne ];
+ real f_NE = (dist.f[DIR_MM0])[ksw ];
+ real f_NW = (dist.f[DIR_PM0])[kse ];
+ real f_SE = (dist.f[DIR_MP0])[knw ];
+ real f_BW = (dist.f[DIR_P0P])[kte ];
+ real f_TE = (dist.f[DIR_M0M])[kbw ];
+ real f_TW = (dist.f[DIR_P0M])[kbe ];
+ real f_BE = (dist.f[DIR_M0P])[ktw ];
+ real f_BS = (dist.f[DIR_0PP])[ktn ];
+ real f_TN = (dist.f[DIR_0MM])[kbs ];
+ real f_TS = (dist.f[DIR_0PM])[kbn ];
+ real f_BN = (dist.f[DIR_0MP])[kts ];
+ real f_BSW = (dist.f[DIR_PPP])[ktne ];
+ real f_BNE = (dist.f[DIR_MMP])[ktsw ];
+ real f_BNW = (dist.f[DIR_PMP])[ktse ];
+ real f_BSE = (dist.f[DIR_MPP])[ktnw ];
+ real f_TSW = (dist.f[DIR_PPM])[kbne ];
+ real f_TNE = (dist.f[DIR_MMM])[kbsw ];
+ real f_TNW = (dist.f[DIR_PMM])[kbse ];
+ real f_TSE = (dist.f[DIR_MPM])[kbnw ];
////////////////////////////////////////////////////////////////////////////////
//! - Calculate macroscopic quantities
@@ -4894,7 +4902,7 @@ __global__ void QVelDeviceComp27(
//! - Update distributions with subgrid distance (q) between zero and one
//!
real feq, q, velocityLB, velocityBC;
- q = (subgridD.q[DIR_P00])[k];
+ q = (subgridD.q[DIR_P00])[nodeIndex];
if (q>=c0o1 && q<=c1o1) // only update distribution for q between zero and one
{
velocityLB = vx1;
@@ -4903,7 +4911,7 @@ __global__ void QVelDeviceComp27(
(dist.f[DIR_M00])[kw] = getInterpolatedDistributionForVeloBC(q, f_E, f_W, feq, omega, velocityBC, c2o27);
}
- q = (subgridD.q[DIR_M00])[k];
+ q = (subgridD.q[DIR_M00])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = -vx1;
@@ -4912,7 +4920,7 @@ __global__ void QVelDeviceComp27(
(dist.f[DIR_P00])[ke] = getInterpolatedDistributionForVeloBC(q, f_W, f_E, feq, omega, velocityBC, c2o27);
}
- q = (subgridD.q[DIR_0P0])[k];
+ q = (subgridD.q[DIR_0P0])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = vx2;
@@ -4921,7 +4929,7 @@ __global__ void QVelDeviceComp27(
(dist.f[DIR_0M0])[ks] = getInterpolatedDistributionForVeloBC(q, f_N, f_S, feq, omega, velocityBC, c2o27);
}
- q = (subgridD.q[DIR_0M0])[k];
+ q = (subgridD.q[DIR_0M0])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = -vx2;
@@ -4930,7 +4938,7 @@ __global__ void QVelDeviceComp27(
(dist.f[DIR_0P0])[kn] = getInterpolatedDistributionForVeloBC(q, f_S, f_N, feq, omega, velocityBC, c2o27);
}
- q = (subgridD.q[DIR_00P])[k];
+ q = (subgridD.q[DIR_00P])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = vx3;
@@ -4939,7 +4947,7 @@ __global__ void QVelDeviceComp27(
(dist.f[DIR_00M])[kb] = getInterpolatedDistributionForVeloBC(q, f_T, f_B, feq, omega, velocityBC, c2o27);
}
- q = (subgridD.q[DIR_00M])[k];
+ q = (subgridD.q[DIR_00M])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = -vx3;
@@ -4948,7 +4956,7 @@ __global__ void QVelDeviceComp27(
(dist.f[DIR_00P])[kt] = getInterpolatedDistributionForVeloBC(q, f_B, f_T, feq, omega, velocityBC, c2o27);
}
- q = (subgridD.q[DIR_PP0])[k];
+ q = (subgridD.q[DIR_PP0])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = vx1 + vx2;
@@ -4957,7 +4965,7 @@ __global__ void QVelDeviceComp27(
(dist.f[DIR_MM0])[ksw] = getInterpolatedDistributionForVeloBC(q, f_NE, f_SW, feq, omega, velocityBC, c1o54);
}
- q = (subgridD.q[DIR_MM0])[k];
+ q = (subgridD.q[DIR_MM0])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = -vx1 - vx2;
@@ -4966,7 +4974,7 @@ __global__ void QVelDeviceComp27(
(dist.f[DIR_PP0])[kne] = getInterpolatedDistributionForVeloBC(q, f_SW, f_NE, feq, omega, velocityBC, c1o54);
}
- q = (subgridD.q[DIR_PM0])[k];
+ q = (subgridD.q[DIR_PM0])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = vx1 - vx2;
@@ -4975,7 +4983,7 @@ __global__ void QVelDeviceComp27(
(dist.f[DIR_MP0])[knw] = getInterpolatedDistributionForVeloBC(q, f_SE, f_NW, feq, omega, velocityBC, c1o54);
}
- q = (subgridD.q[DIR_MP0])[k];
+ q = (subgridD.q[DIR_MP0])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = -vx1 + vx2;
@@ -4984,7 +4992,7 @@ __global__ void QVelDeviceComp27(
(dist.f[DIR_PM0])[kse] = getInterpolatedDistributionForVeloBC(q, f_NW, f_SE, feq, omega, velocityBC, c1o54);
}
- q = (subgridD.q[DIR_P0P])[k];
+ q = (subgridD.q[DIR_P0P])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = vx1 + vx3;
@@ -4993,7 +5001,7 @@ __global__ void QVelDeviceComp27(
(dist.f[DIR_M0M])[kbw] = getInterpolatedDistributionForVeloBC(q, f_TE, f_BW, feq, omega, velocityBC, c1o54);
}
- q = (subgridD.q[DIR_M0M])[k];
+ q = (subgridD.q[DIR_M0M])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = -vx1 - vx3;
@@ -5002,7 +5010,7 @@ __global__ void QVelDeviceComp27(
(dist.f[DIR_P0P])[kte] = getInterpolatedDistributionForVeloBC(q, f_BW, f_TE, feq, omega, velocityBC, c1o54);
}
- q = (subgridD.q[DIR_P0M])[k];
+ q = (subgridD.q[DIR_P0M])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = vx1 - vx3;
@@ -5011,7 +5019,7 @@ __global__ void QVelDeviceComp27(
(dist.f[DIR_M0P])[ktw] = getInterpolatedDistributionForVeloBC(q, f_BE, f_TW, feq, omega, velocityBC, c1o54);
}
- q = (subgridD.q[DIR_M0P])[k];
+ q = (subgridD.q[DIR_M0P])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = -vx1 + vx3;
@@ -5020,7 +5028,7 @@ __global__ void QVelDeviceComp27(
(dist.f[DIR_P0M])[kbe] = getInterpolatedDistributionForVeloBC(q, f_TW, f_BE, feq, omega, velocityBC, c1o54);
}
- q = (subgridD.q[DIR_0PP])[k];
+ q = (subgridD.q[DIR_0PP])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = vx2 + vx3;
@@ -5029,7 +5037,7 @@ __global__ void QVelDeviceComp27(
(dist.f[DIR_0MM])[kbs] = getInterpolatedDistributionForVeloBC(q, f_TN, f_BS, feq, omega, velocityBC, c1o54);
}
- q = (subgridD.q[DIR_0MM])[k];
+ q = (subgridD.q[DIR_0MM])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = -vx2 - vx3;
@@ -5038,7 +5046,7 @@ __global__ void QVelDeviceComp27(
(dist.f[DIR_0PP])[ktn] = getInterpolatedDistributionForVeloBC(q, f_BS, f_TN, feq, omega, velocityBC, c1o54);
}
- q = (subgridD.q[DIR_0PM])[k];
+ q = (subgridD.q[DIR_0PM])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = vx2 - vx3;
@@ -5047,7 +5055,7 @@ __global__ void QVelDeviceComp27(
(dist.f[DIR_0MP])[kts] = getInterpolatedDistributionForVeloBC(q, f_BN, f_TS, feq, omega, velocityBC, c1o54);
}
- q = (subgridD.q[DIR_0MP])[k];
+ q = (subgridD.q[DIR_0MP])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = -vx2 + vx3;
@@ -5056,7 +5064,7 @@ __global__ void QVelDeviceComp27(
(dist.f[DIR_0PM])[kbn] = getInterpolatedDistributionForVeloBC(q, f_TS, f_BN, feq, omega, velocityBC, c1o54);
}
- q = (subgridD.q[DIR_PPP])[k];
+ q = (subgridD.q[DIR_PPP])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = vx1 + vx2 + vx3;
@@ -5065,7 +5073,7 @@ __global__ void QVelDeviceComp27(
(dist.f[DIR_MMM])[kbsw] = getInterpolatedDistributionForVeloBC(q, f_TNE, f_BSW, feq, omega, velocityBC, c1o216);
}
- q = (subgridD.q[DIR_MMM])[k];
+ q = (subgridD.q[DIR_MMM])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = -vx1 - vx2 - vx3;
@@ -5074,7 +5082,7 @@ __global__ void QVelDeviceComp27(
(dist.f[DIR_PPP])[ktne] = getInterpolatedDistributionForVeloBC(q, f_BSW, f_TNE, feq, omega, velocityBC, c1o216);
}
- q = (subgridD.q[DIR_PPM])[k];
+ q = (subgridD.q[DIR_PPM])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = vx1 + vx2 - vx3;
@@ -5083,7 +5091,7 @@ __global__ void QVelDeviceComp27(
(dist.f[DIR_MMP])[ktsw] = getInterpolatedDistributionForVeloBC(q, f_BNE, f_TSW, feq, omega, velocityBC, c1o216);
}
- q = (subgridD.q[DIR_MMP])[k];
+ q = (subgridD.q[DIR_MMP])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = -vx1 - vx2 + vx3;
@@ -5092,7 +5100,7 @@ __global__ void QVelDeviceComp27(
(dist.f[DIR_PPM])[kbne] = getInterpolatedDistributionForVeloBC(q, f_TSW, f_BNE, feq, omega, velocityBC, c1o216);
}
- q = (subgridD.q[DIR_PMP])[k];
+ q = (subgridD.q[DIR_PMP])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = vx1 - vx2 + vx3;
@@ -5101,7 +5109,7 @@ __global__ void QVelDeviceComp27(
(dist.f[DIR_MPM])[kbnw] = getInterpolatedDistributionForVeloBC(q, f_TSE, f_BNW, feq, omega, velocityBC, c1o216);
}
- q = (subgridD.q[DIR_MPM])[k];
+ q = (subgridD.q[DIR_MPM])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = -vx1 + vx2 - vx3;
@@ -5110,7 +5118,7 @@ __global__ void QVelDeviceComp27(
(dist.f[DIR_PMP])[ktse] = getInterpolatedDistributionForVeloBC(q, f_BNW, f_TSE, feq, omega, velocityBC, c1o216);
}
- q = (subgridD.q[DIR_PMM])[k];
+ q = (subgridD.q[DIR_PMM])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = vx1 - vx2 - vx3;
@@ -5119,7 +5127,7 @@ __global__ void QVelDeviceComp27(
(dist.f[DIR_MPP])[ktnw] = getInterpolatedDistributionForVeloBC(q, f_BSE, f_TNW, feq, omega, velocityBC, c1o216);
}
- q = (subgridD.q[DIR_MPP])[k];
+ q = (subgridD.q[DIR_MPP])[nodeIndex];
if (q>=c0o1 && q<=c1o1)
{
velocityLB = -vx1 + vx2 + vx3;
@@ -5170,82 +5178,83 @@ __global__ void QVelDeviceComp27(
//////////////////////////////////////////////////////////////////////////////
-__global__ void QVelDevice27(int inx,
- int iny,
- real* vx,
- real* vy,
- real* vz,
- real* DD,
- int* k_Q,
- real* QQ,
- unsigned int numberOfBCnodes,
- real om1,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- unsigned int size_Mat,
- bool isEvenTimestep)
+__global__ void QVelDevice27(
+ int inx,
+ int iny,
+ real* vx,
+ real* vy,
+ real* vz,
+ real* DD,
+ int* k_Q,
+ real* QQ,
+ unsigned int numberOfBCnodes,
+ real om1,
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ unsigned long long numberOfLBnodes,
+ bool isEvenTimestep)
{
Distributions27 D;
if (isEvenTimestep==true)
{
- D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
+ D.f[DIR_M00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_MPM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_PMP * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////
const unsigned x = threadIdx.x; // Globaler x-Index
@@ -5270,24 +5279,24 @@ __global__ void QVelDevice27(int inx,
*q_dirBE, *q_dirTW, *q_dirTN, *q_dirBS, *q_dirBN, *q_dirTS,
*q_dirTNE, *q_dirTSW, *q_dirTSE, *q_dirTNW, *q_dirBNE, *q_dirBSW,
*q_dirBSE, *q_dirBNW;
- q_dirE = &QQ[DIR_P00 * numberOfBCnodes];
- q_dirW = &QQ[DIR_M00 * numberOfBCnodes];
- q_dirN = &QQ[DIR_0P0 * numberOfBCnodes];
- q_dirS = &QQ[DIR_0M0 * numberOfBCnodes];
- q_dirT = &QQ[DIR_00P * numberOfBCnodes];
- q_dirB = &QQ[DIR_00M * numberOfBCnodes];
- q_dirNE = &QQ[DIR_PP0 * numberOfBCnodes];
- q_dirSW = &QQ[DIR_MM0 * numberOfBCnodes];
- q_dirSE = &QQ[DIR_PM0 * numberOfBCnodes];
- q_dirNW = &QQ[DIR_MP0 * numberOfBCnodes];
- q_dirTE = &QQ[DIR_P0P * numberOfBCnodes];
- q_dirBW = &QQ[DIR_M0M * numberOfBCnodes];
- q_dirBE = &QQ[DIR_P0M * numberOfBCnodes];
- q_dirTW = &QQ[DIR_M0P * numberOfBCnodes];
- q_dirTN = &QQ[DIR_0PP * numberOfBCnodes];
- q_dirBS = &QQ[DIR_0MM * numberOfBCnodes];
- q_dirBN = &QQ[DIR_0PM * numberOfBCnodes];
- q_dirTS = &QQ[DIR_0MP * numberOfBCnodes];
+ q_dirE = &QQ[DIR_P00 * numberOfBCnodes];
+ q_dirW = &QQ[DIR_M00 * numberOfBCnodes];
+ q_dirN = &QQ[DIR_0P0 * numberOfBCnodes];
+ q_dirS = &QQ[DIR_0M0 * numberOfBCnodes];
+ q_dirT = &QQ[DIR_00P * numberOfBCnodes];
+ q_dirB = &QQ[DIR_00M * numberOfBCnodes];
+ q_dirNE = &QQ[DIR_PP0 * numberOfBCnodes];
+ q_dirSW = &QQ[DIR_MM0 * numberOfBCnodes];
+ q_dirSE = &QQ[DIR_PM0 * numberOfBCnodes];
+ q_dirNW = &QQ[DIR_MP0 * numberOfBCnodes];
+ q_dirTE = &QQ[DIR_P0P * numberOfBCnodes];
+ q_dirBW = &QQ[DIR_M0M * numberOfBCnodes];
+ q_dirBE = &QQ[DIR_P0M * numberOfBCnodes];
+ q_dirTW = &QQ[DIR_M0P * numberOfBCnodes];
+ q_dirTN = &QQ[DIR_0PP * numberOfBCnodes];
+ q_dirBS = &QQ[DIR_0MM * numberOfBCnodes];
+ q_dirBN = &QQ[DIR_0PM * numberOfBCnodes];
+ q_dirTS = &QQ[DIR_0MP * numberOfBCnodes];
q_dirTNE = &QQ[DIR_PPP * numberOfBCnodes];
q_dirTSW = &QQ[DIR_MMP * numberOfBCnodes];
q_dirTSE = &QQ[DIR_PMP * numberOfBCnodes];
@@ -5358,32 +5367,32 @@ __global__ void QVelDevice27(int inx,
real f_E, f_W, f_N, f_S, f_T, f_B, f_NE, f_SW, f_SE, f_NW, f_TE, f_BW, f_BE,
f_TW, f_TN, f_BS, f_BN, f_TS, f_TNE, f_TSW, f_TSE, f_TNW, f_BNE, f_BSW, f_BSE, f_BNW;
- f_W = (D.f[DIR_P00 ])[ke ];
- f_E = (D.f[DIR_M00 ])[kw ];
- f_S = (D.f[DIR_0P0 ])[kn ];
- f_N = (D.f[DIR_0M0 ])[ks ];
- f_B = (D.f[DIR_00P ])[kt ];
- f_T = (D.f[DIR_00M ])[kb ];
- f_SW = (D.f[DIR_PP0 ])[kne ];
- f_NE = (D.f[DIR_MM0 ])[ksw ];
- f_NW = (D.f[DIR_PM0 ])[kse ];
- f_SE = (D.f[DIR_MP0 ])[knw ];
- f_BW = (D.f[DIR_P0P ])[kte ];
- f_TE = (D.f[DIR_M0M ])[kbw ];
- f_TW = (D.f[DIR_P0M ])[kbe ];
- f_BE = (D.f[DIR_M0P ])[ktw ];
- f_BS = (D.f[DIR_0PP ])[ktn ];
- f_TN = (D.f[DIR_0MM ])[kbs ];
- f_TS = (D.f[DIR_0PM ])[kbn ];
- f_BN = (D.f[DIR_0MP ])[kts ];
- f_BSW = (D.f[DIR_PPP ])[ktne ];
- f_BNE = (D.f[DIR_MMP ])[ktsw ];
- f_BNW = (D.f[DIR_PMP ])[ktse ];
- f_BSE = (D.f[DIR_MPP ])[ktnw ];
- f_TSW = (D.f[DIR_PPM ])[kbne ];
- f_TNE = (D.f[DIR_MMM ])[kbsw ];
- f_TNW = (D.f[DIR_PMM ])[kbse ];
- f_TSE = (D.f[DIR_MPM ])[kbnw ];
+ f_W = (D.f[DIR_P00])[ke ];
+ f_E = (D.f[DIR_M00])[kw ];
+ f_S = (D.f[DIR_0P0])[kn ];
+ f_N = (D.f[DIR_0M0])[ks ];
+ f_B = (D.f[DIR_00P])[kt ];
+ f_T = (D.f[DIR_00M])[kb ];
+ f_SW = (D.f[DIR_PP0])[kne ];
+ f_NE = (D.f[DIR_MM0])[ksw ];
+ f_NW = (D.f[DIR_PM0])[kse ];
+ f_SE = (D.f[DIR_MP0])[knw ];
+ f_BW = (D.f[DIR_P0P])[kte ];
+ f_TE = (D.f[DIR_M0M])[kbw ];
+ f_TW = (D.f[DIR_P0M])[kbe ];
+ f_BE = (D.f[DIR_M0P])[ktw ];
+ f_BS = (D.f[DIR_0PP])[ktn ];
+ f_TN = (D.f[DIR_0MM])[kbs ];
+ f_TS = (D.f[DIR_0PM])[kbn ];
+ f_BN = (D.f[DIR_0MP])[kts ];
+ f_BSW = (D.f[DIR_PPP])[ktne ];
+ f_BNE = (D.f[DIR_MMP])[ktsw ];
+ f_BNW = (D.f[DIR_PMP])[ktse ];
+ f_BSE = (D.f[DIR_MPP])[ktnw ];
+ f_TSW = (D.f[DIR_PPM])[kbne ];
+ f_TNE = (D.f[DIR_MMM])[kbsw ];
+ f_TNW = (D.f[DIR_PMM])[kbse ];
+ f_TSE = (D.f[DIR_MPM])[kbnw ];
////////////////////////////////////////////////////////////////////////////////
real vx1, vx2, vx3, drho, feq, q;
drho = f_TSE + f_TNW + f_TNE + f_TSW + f_BSE + f_BNW + f_BNE + f_BSW +
@@ -5408,63 +5417,63 @@ __global__ void QVelDevice27(int inx,
//////////////////////////////////////////////////////////////////////////
if (isEvenTimestep==false)
{
- D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
+ D.f[DIR_M00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_MPM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_PMP * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
//Test
@@ -5723,19 +5732,20 @@ __global__ void QVelDevice27(int inx,
////////////////////////////////////////////////////////////////////////////////
-__global__ void PropellerBC(unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- real* rho,
- real* ux,
- real* uy,
- real* uz,
- int* k_Q,
- unsigned int size_Prop,
- unsigned int size_Mat,
- unsigned int* bcMatD,
- real* DD,
- bool EvenOrOdd)
+__global__ void PropellerBC(
+ unsigned int* neighborX,
+ unsigned int* neighborY,
+ unsigned int* neighborZ,
+ real* rho,
+ real* ux,
+ real* uy,
+ real* uz,
+ int* k_Q,
+ unsigned int size_Prop,
+ unsigned long long numberOfLBnodes,
+ unsigned int* bcMatD,
+ real* DD,
+ bool EvenOrOdd)
{
////////////////////////////////////////////////////////////////////////////////
const unsigned x = threadIdx.x; // Globaler x-Index
@@ -5754,63 +5764,63 @@ __global__ void PropellerBC(unsigned int* neighborX,
Distributions27 D;
if (EvenOrOdd==true)
{
- D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
+ D.f[DIR_M00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_MPM * numberOfLBnodes];
}
//////////////////////////////////////////////////////////////////////////
unsigned int KQK = k_Q[k];
@@ -5859,58 +5869,58 @@ __global__ void PropellerBC(unsigned int* neighborX,
f_TW, f_TN, f_BS, f_BN, f_TS, f_TNE, f_TSW, f_TSE, f_TNW, f_BNE, f_BSW, f_BSE, f_BNW, f_ZERO;
f_ZERO= (D.f[DIR_000])[kzero];
- f_E = (D.f[DIR_P00 ])[ke ];
- f_W = (D.f[DIR_M00 ])[kw ];
- f_N = (D.f[DIR_0P0 ])[kn ];
- f_S = (D.f[DIR_0M0 ])[ks ];
- f_T = (D.f[DIR_00P ])[kt ];
- f_B = (D.f[DIR_00M ])[kb ];
- f_NE = (D.f[DIR_PP0 ])[kne ];
- f_SW = (D.f[DIR_MM0 ])[ksw ];
- f_SE = (D.f[DIR_PM0 ])[kse ];
- f_NW = (D.f[DIR_MP0 ])[knw ];
- f_TE = (D.f[DIR_P0P ])[kte ];
- f_BW = (D.f[DIR_M0M ])[kbw ];
- f_BE = (D.f[DIR_P0M ])[kbe ];
- f_TW = (D.f[DIR_M0P ])[ktw ];
- f_TN = (D.f[DIR_0PP ])[ktn ];
- f_BS = (D.f[DIR_0MM ])[kbs ];
- f_BN = (D.f[DIR_0PM ])[kbn ];
- f_TS = (D.f[DIR_0MP ])[kts ];
- f_TNE = (D.f[DIR_PPP ])[ktne ];
- f_BSW = (D.f[DIR_MMM ])[kbsw ];
- f_BNE = (D.f[DIR_PPM ])[kbne ];
- f_TSW = (D.f[DIR_MMP ])[ktsw ];
- f_TSE = (D.f[DIR_PMP ])[ktse ];
- f_BNW = (D.f[DIR_MPM ])[kbnw ];
- f_BSE = (D.f[DIR_PMM ])[kbse ];
- f_TNW = (D.f[DIR_MPP ])[ktnw ];
- //f_W = (D.f[DIR_P00 ])[ke ];
- //f_E = (D.f[DIR_M00 ])[kw ];
- //f_S = (D.f[DIR_0P0 ])[kn ];
- //f_N = (D.f[DIR_0M0 ])[ks ];
- //f_B = (D.f[DIR_00P ])[kt ];
- //f_T = (D.f[DIR_00M ])[kb ];
- //f_SW = (D.f[DIR_PP0 ])[kne ];
- //f_NE = (D.f[DIR_MM0 ])[ksw ];
- //f_NW = (D.f[DIR_PM0 ])[kse ];
- //f_SE = (D.f[DIR_MP0 ])[knw ];
- //f_BW = (D.f[DIR_P0P ])[kte ];
- //f_TE = (D.f[DIR_M0M ])[kbw ];
- //f_TW = (D.f[DIR_P0M ])[kbe ];
- //f_BE = (D.f[DIR_M0P ])[ktw ];
- //f_BS = (D.f[DIR_0PP ])[ktn ];
- //f_TN = (D.f[DIR_0MM ])[kbs ];
- //f_TS = (D.f[DIR_0PM ])[kbn ];
- //f_BN = (D.f[DIR_0MP ])[kts ];
- //f_BSW = (D.f[DIR_PPP ])[ktne ];
- //f_TNE = (D.f[DIR_MMM ])[kbsw ];
- //f_TSW = (D.f[DIR_PPM ])[kbne ];
- //f_BNE = (D.f[DIR_MMP ])[ktsw ];
- //f_BNW = (D.f[DIR_PMP ])[ktse ];
- //f_TSE = (D.f[DIR_MPM ])[kbnw ];
- //f_TNW = (D.f[DIR_PMM ])[kbse ];
- //f_BSE = (D.f[DIR_MPP ])[ktnw ];
+ f_E = (D.f[DIR_P00])[ke ];
+ f_W = (D.f[DIR_M00])[kw ];
+ f_N = (D.f[DIR_0P0])[kn ];
+ f_S = (D.f[DIR_0M0])[ks ];
+ f_T = (D.f[DIR_00P])[kt ];
+ f_B = (D.f[DIR_00M])[kb ];
+ f_NE = (D.f[DIR_PP0])[kne ];
+ f_SW = (D.f[DIR_MM0])[ksw ];
+ f_SE = (D.f[DIR_PM0])[kse ];
+ f_NW = (D.f[DIR_MP0])[knw ];
+ f_TE = (D.f[DIR_P0P])[kte ];
+ f_BW = (D.f[DIR_M0M])[kbw ];
+ f_BE = (D.f[DIR_P0M])[kbe ];
+ f_TW = (D.f[DIR_M0P])[ktw ];
+ f_TN = (D.f[DIR_0PP])[ktn ];
+ f_BS = (D.f[DIR_0MM])[kbs ];
+ f_BN = (D.f[DIR_0PM])[kbn ];
+ f_TS = (D.f[DIR_0MP])[kts ];
+ f_TNE = (D.f[DIR_PPP])[ktne ];
+ f_BSW = (D.f[DIR_MMM])[kbsw ];
+ f_BNE = (D.f[DIR_PPM])[kbne ];
+ f_TSW = (D.f[DIR_MMP])[ktsw ];
+ f_TSE = (D.f[DIR_PMP])[ktse ];
+ f_BNW = (D.f[DIR_MPM])[kbnw ];
+ f_BSE = (D.f[DIR_PMM])[kbse ];
+ f_TNW = (D.f[DIR_MPP])[ktnw ];
+ //f_W = (D.f[DIR_P00])[ke ];
+ //f_E = (D.f[DIR_M00])[kw ];
+ //f_S = (D.f[DIR_0P0])[kn ];
+ //f_N = (D.f[DIR_0M0])[ks ];
+ //f_B = (D.f[DIR_00P])[kt ];
+ //f_T = (D.f[DIR_00M])[kb ];
+ //f_SW = (D.f[DIR_PP0])[kne ];
+ //f_NE = (D.f[DIR_MM0])[ksw ];
+ //f_NW = (D.f[DIR_PM0])[kse ];
+ //f_SE = (D.f[DIR_MP0])[knw ];
+ //f_BW = (D.f[DIR_P0P])[kte ];
+ //f_TE = (D.f[DIR_M0M])[kbw ];
+ //f_TW = (D.f[DIR_P0M])[kbe ];
+ //f_BE = (D.f[DIR_M0P])[ktw ];
+ //f_BS = (D.f[DIR_0PP])[ktn ];
+ //f_TN = (D.f[DIR_0MM])[kbs ];
+ //f_TS = (D.f[DIR_0PM])[kbn ];
+ //f_BN = (D.f[DIR_0MP])[kts ];
+ //f_BSW = (D.f[DIR_PPP])[ktne ];
+ //f_TNE = (D.f[DIR_MMM])[kbsw ];
+ //f_TSW = (D.f[DIR_PPM])[kbne ];
+ //f_BNE = (D.f[DIR_MMP])[ktsw ];
+ //f_BNW = (D.f[DIR_PMP])[ktse ];
+ //f_TSE = (D.f[DIR_MPM])[kbnw ];
+ //f_TNW = (D.f[DIR_PMM])[kbse ];
+ //f_BSE = (D.f[DIR_MPP])[ktnw ];
//////////////////////////////////////////////////////////////////////////////////
real vxo1, vxo2, vxo3, drho;
drho = /*zero;*/f_TSE + f_TNW + f_TNE + f_TSW + f_BSE + f_BNW + f_BNE + f_BSW +
@@ -5992,88 +6002,88 @@ __global__ void PropellerBC(unsigned int* neighborX,
f_TNW = f_TNW + ((c1o1+drho) * (- c1o216*(c3o1*(-vxo1+vxo2+vxo3)+c9o2*(-vxo1+vxo2+vxo3)*(-vxo1+vxo2+vxo3)-cusq) + c1o216*(c3o1*(-vx1+vx2+vx3)+c9o2*(-vx1+vx2+vx3)*(-vx1+vx2+vx3)-cusq2)));
(D.f[DIR_000])[kzero] = f_ZERO;
- (D.f[DIR_P00 ])[ke ] = f_E ; // f_W ;//
- (D.f[DIR_M00 ])[kw ] = f_W ; // f_E ;//
- (D.f[DIR_0P0 ])[kn ] = f_N ; // f_S ;//
- (D.f[DIR_0M0 ])[ks ] = f_S ; // f_N ;//
- (D.f[DIR_00P ])[kt ] = f_T ; // f_B ;//
- (D.f[DIR_00M ])[kb ] = f_B ; // f_T ;//
- (D.f[DIR_PP0 ])[kne ] = f_NE ; // f_SW ;//
- (D.f[DIR_MM0 ])[ksw ] = f_SW ; // f_NE ;//
- (D.f[DIR_PM0 ])[kse ] = f_SE ; // f_NW ;//
- (D.f[DIR_MP0 ])[knw ] = f_NW ; // f_SE ;//
- (D.f[DIR_P0P ])[kte ] = f_TE ; // f_BW ;//
- (D.f[DIR_M0M ])[kbw ] = f_BW ; // f_TE ;//
- (D.f[DIR_P0M ])[kbe ] = f_BE ; // f_TW ;//
- (D.f[DIR_M0P ])[ktw ] = f_TW ; // f_BE ;//
- (D.f[DIR_0PP ])[ktn ] = f_TN ; // f_BS ;//
- (D.f[DIR_0MM ])[kbs ] = f_BS ; // f_TN ;//
- (D.f[DIR_0PM ])[kbn ] = f_BN ; // f_TS ;//
- (D.f[DIR_0MP ])[kts ] = f_TS ; // f_BN ;//
- (D.f[DIR_PPP ])[ktne ] = f_TNE ; // f_BSW ;//
- (D.f[DIR_MMM ])[kbsw ] = f_BSW ; // f_BNE ;//
- (D.f[DIR_PPM ])[kbne ] = f_BNE ; // f_BNW ;//
- (D.f[DIR_MMP ])[ktsw ] = f_TSW ; // f_BSE ;//
- (D.f[DIR_PMP ])[ktse ] = f_TSE ; // f_TSW ;//
- (D.f[DIR_MPM ])[kbnw ] = f_BNW ; // f_TNE ;//
- (D.f[DIR_PMM ])[kbse ] = f_BSE ; // f_TNW ;//
- (D.f[DIR_MPP ])[ktnw ] = f_TNW ; // f_TSE ;//
+ (D.f[DIR_P00])[ke ] = f_E ; // f_W ;//
+ (D.f[DIR_M00])[kw ] = f_W ; // f_E ;//
+ (D.f[DIR_0P0])[kn ] = f_N ; // f_S ;//
+ (D.f[DIR_0M0])[ks ] = f_S ; // f_N ;//
+ (D.f[DIR_00P])[kt ] = f_T ; // f_B ;//
+ (D.f[DIR_00M])[kb ] = f_B ; // f_T ;//
+ (D.f[DIR_PP0])[kne ] = f_NE ; // f_SW ;//
+ (D.f[DIR_MM0])[ksw ] = f_SW ; // f_NE ;//
+ (D.f[DIR_PM0])[kse ] = f_SE ; // f_NW ;//
+ (D.f[DIR_MP0])[knw ] = f_NW ; // f_SE ;//
+ (D.f[DIR_P0P])[kte ] = f_TE ; // f_BW ;//
+ (D.f[DIR_M0M])[kbw ] = f_BW ; // f_TE ;//
+ (D.f[DIR_P0M])[kbe ] = f_BE ; // f_TW ;//
+ (D.f[DIR_M0P])[ktw ] = f_TW ; // f_BE ;//
+ (D.f[DIR_0PP])[ktn ] = f_TN ; // f_BS ;//
+ (D.f[DIR_0MM])[kbs ] = f_BS ; // f_TN ;//
+ (D.f[DIR_0PM])[kbn ] = f_BN ; // f_TS ;//
+ (D.f[DIR_0MP])[kts ] = f_TS ; // f_BN ;//
+ (D.f[DIR_PPP])[ktne ] = f_TNE ; // f_BSW ;//
+ (D.f[DIR_MMM])[kbsw ] = f_BSW ; // f_BNE ;//
+ (D.f[DIR_PPM])[kbne ] = f_BNE ; // f_BNW ;//
+ (D.f[DIR_MMP])[ktsw ] = f_TSW ; // f_BSE ;//
+ (D.f[DIR_PMP])[ktse ] = f_TSE ; // f_TSW ;//
+ (D.f[DIR_MPM])[kbnw ] = f_BNW ; // f_TNE ;//
+ (D.f[DIR_PMM])[kbse ] = f_BSE ; // f_TNW ;//
+ (D.f[DIR_MPP])[ktnw ] = f_TNW ; // f_TSE ;//
//////////////////////////////////////////////////////////////////////////
////(D.f[DIR_000])[kzero] = c8over27* (drho-cu_sq);
- //(D.f[DIR_P00 ])[ke ] = three*c2over27* ( vx1 ); //six
- //(D.f[DIR_M00 ])[kw ] = three*c2over27* (-vx1 ); //six
- //(D.f[DIR_0P0 ])[kn ] = three*c2over27* ( vx2 ); //six
- //(D.f[DIR_0M0 ])[ks ] = three*c2over27* ( -vx2 ); //six
- //(D.f[DIR_00P ])[kt ] = three*c2over27* ( vx3); //six
- //(D.f[DIR_00M ])[kb ] = three*c2over27* ( -vx3); //six
- //(D.f[DIR_PP0 ])[kne ] = three*c1over54* ( vx1+vx2 ); //six
- //(D.f[DIR_MM0 ])[ksw ] = three*c1over54* (-vx1-vx2 ); //six
- //(D.f[DIR_PM0 ])[kse ] = three*c1over54* ( vx1-vx2 ); //six
- //(D.f[DIR_MP0 ])[knw ] = three*c1over54* (-vx1+vx2 ); //six
- //(D.f[DIR_P0P ])[kte ] = three*c1over54* ( vx1 +vx3); //six
- //(D.f[DIR_M0M ])[kbw ] = three*c1over54* (-vx1 -vx3); //six
- //(D.f[DIR_P0M ])[kbe ] = three*c1over54* ( vx1 -vx3); //six
- //(D.f[DIR_M0P ])[ktw ] = three*c1over54* (-vx1 +vx3); //six
- //(D.f[DIR_0PP ])[ktn ] = three*c1over54* ( vx2+vx3); //six
- //(D.f[DIR_0MM ])[kbs ] = three*c1over54* ( -vx2-vx3); //six
- //(D.f[DIR_0PM ])[kbn ] = three*c1over54* ( vx2-vx3); //six
- //(D.f[DIR_0MP ])[kts ] = three*c1over54* ( -vx2+vx3); //six
- //(D.f[DIR_PPP ])[ktne ] = three*c1over216*( vx1+vx2+vx3); //six
- //(D.f[DIR_MMM ])[kbsw ] = three*c1over216*(-vx1-vx2-vx3); //six
- //(D.f[DIR_PPM ])[kbne ] = three*c1over216*( vx1+vx2-vx3); //six
- //(D.f[DIR_MMP ])[ktsw ] = three*c1over216*(-vx1-vx2+vx3); //six
- //(D.f[DIR_PMP ])[ktse ] = three*c1over216*( vx1-vx2+vx3); //six
- //(D.f[DIR_MPM ])[kbnw ] = three*c1over216*(-vx1+vx2-vx3); //six
- //(D.f[DIR_PMM ])[kbse ] = three*c1over216*( vx1-vx2-vx3); //six
- //(D.f[DIR_MPP ])[ktnw ] = three*c1over216*(-vx1+vx2+vx3); //six
+ //(D.f[DIR_P00])[ke ] = three*c2over27* ( vx1 ); //six
+ //(D.f[DIR_M00])[kw ] = three*c2over27* (-vx1 ); //six
+ //(D.f[DIR_0P0])[kn ] = three*c2over27* ( vx2 ); //six
+ //(D.f[DIR_0M0])[ks ] = three*c2over27* ( -vx2 ); //six
+ //(D.f[DIR_00P])[kt ] = three*c2over27* ( vx3); //six
+ //(D.f[DIR_00M])[kb ] = three*c2over27* ( -vx3); //six
+ //(D.f[DIR_PP0])[kne ] = three*c1over54* ( vx1+vx2 ); //six
+ //(D.f[DIR_MM0])[ksw ] = three*c1over54* (-vx1-vx2 ); //six
+ //(D.f[DIR_PM0])[kse ] = three*c1over54* ( vx1-vx2 ); //six
+ //(D.f[DIR_MP0])[knw ] = three*c1over54* (-vx1+vx2 ); //six
+ //(D.f[DIR_P0P])[kte ] = three*c1over54* ( vx1 +vx3); //six
+ //(D.f[DIR_M0M])[kbw ] = three*c1over54* (-vx1 -vx3); //six
+ //(D.f[DIR_P0M])[kbe ] = three*c1over54* ( vx1 -vx3); //six
+ //(D.f[DIR_M0P])[ktw ] = three*c1over54* (-vx1 +vx3); //six
+ //(D.f[DIR_0PP])[ktn ] = three*c1over54* ( vx2+vx3); //six
+ //(D.f[DIR_0MM])[kbs ] = three*c1over54* ( -vx2-vx3); //six
+ //(D.f[DIR_0PM])[kbn ] = three*c1over54* ( vx2-vx3); //six
+ //(D.f[DIR_0MP])[kts ] = three*c1over54* ( -vx2+vx3); //six
+ //(D.f[DIR_PPP])[ktne ] = three*c1over216*( vx1+vx2+vx3); //six
+ //(D.f[DIR_MMM])[kbsw ] = three*c1over216*(-vx1-vx2-vx3); //six
+ //(D.f[DIR_PPM])[kbne ] = three*c1over216*( vx1+vx2-vx3); //six
+ //(D.f[DIR_MMP])[ktsw ] = three*c1over216*(-vx1-vx2+vx3); //six
+ //(D.f[DIR_PMP])[ktse ] = three*c1over216*( vx1-vx2+vx3); //six
+ //(D.f[DIR_MPM])[kbnw ] = three*c1over216*(-vx1+vx2-vx3); //six
+ //(D.f[DIR_PMM])[kbse ] = three*c1over216*( vx1-vx2-vx3); //six
+ //(D.f[DIR_MPP])[ktnw ] = three*c1over216*(-vx1+vx2+vx3); //six
//(D.f[DIR_000])[kzero] = c8over27* (drho-cu_sq);
- //(D.f[DIR_P00 ])[ke ] = c2over27* (drho+three*( vx1 )+c9over2*( vx1 )*( vx1 )-cu_sq);
- //(D.f[DIR_M00 ])[kw ] = c2over27* (drho+three*(-vx1 )+c9over2*(-vx1 )*(-vx1 )-cu_sq);
- //(D.f[DIR_0P0 ])[kn ] = c2over27* (drho+three*( vx2 )+c9over2*( vx2 )*( vx2 )-cu_sq);
- //(D.f[DIR_0M0 ])[ks ] = c2over27* (drho+three*( -vx2 )+c9over2*( -vx2 )*( -vx2 )-cu_sq);
- //(D.f[DIR_00P ])[kt ] = c2over27* (drho+three*( vx3)+c9over2*( vx3)*( vx3)-cu_sq);
- //(D.f[DIR_00M ])[kb ] = c2over27* (drho+three*( -vx3)+c9over2*( -vx3)*( -vx3)-cu_sq);
- //(D.f[DIR_PP0 ])[kne ] = c1over54* (drho+three*( vx1+vx2 )+c9over2*( vx1+vx2 )*( vx1+vx2 )-cu_sq);
- //(D.f[DIR_MM0 ])[ksw ] = c1over54* (drho+three*(-vx1-vx2 )+c9over2*(-vx1-vx2 )*(-vx1-vx2 )-cu_sq);
- //(D.f[DIR_PM0 ])[kse ] = c1over54* (drho+three*( vx1-vx2 )+c9over2*( vx1-vx2 )*( vx1-vx2 )-cu_sq);
- //(D.f[DIR_MP0 ])[knw ] = c1over54* (drho+three*(-vx1+vx2 )+c9over2*(-vx1+vx2 )*(-vx1+vx2 )-cu_sq);
- //(D.f[DIR_P0P ])[kte ] = c1over54* (drho+three*( vx1 +vx3)+c9over2*( vx1 +vx3)*( vx1 +vx3)-cu_sq);
- //(D.f[DIR_M0M ])[kbw ] = c1over54* (drho+three*(-vx1 -vx3)+c9over2*(-vx1 -vx3)*(-vx1 -vx3)-cu_sq);
- //(D.f[DIR_P0M ])[kbe ] = c1over54* (drho+three*( vx1 -vx3)+c9over2*( vx1 -vx3)*( vx1 -vx3)-cu_sq);
- //(D.f[DIR_M0P ])[ktw ] = c1over54* (drho+three*(-vx1 +vx3)+c9over2*(-vx1 +vx3)*(-vx1 +vx3)-cu_sq);
- //(D.f[DIR_0PP ])[ktn ] = c1over54* (drho+three*( vx2+vx3)+c9over2*( vx2+vx3)*( vx2+vx3)-cu_sq);
- //(D.f[DIR_0MM ])[kbs ] = c1over54* (drho+three*( -vx2-vx3)+c9over2*( -vx2-vx3)*( -vx2-vx3)-cu_sq);
- //(D.f[DIR_0PM ])[kbn ] = c1over54* (drho+three*( vx2-vx3)+c9over2*( vx2-vx3)*( vx2-vx3)-cu_sq);
- //(D.f[DIR_0MP ])[kts ] = c1over54* (drho+three*( -vx2+vx3)+c9over2*( -vx2+vx3)*( -vx2+vx3)-cu_sq);
- //(D.f[DIR_PPP ])[ktne ] = c1over216*(drho+three*( vx1+vx2+vx3)+c9over2*( vx1+vx2+vx3)*( vx1+vx2+vx3)-cu_sq);
- //(D.f[DIR_MMM ])[kbsw ] = c1over216*(drho+three*(-vx1-vx2-vx3)+c9over2*(-vx1-vx2-vx3)*(-vx1-vx2-vx3)-cu_sq);
- //(D.f[DIR_PPM ])[kbne ] = c1over216*(drho+three*( vx1+vx2-vx3)+c9over2*( vx1+vx2-vx3)*( vx1+vx2-vx3)-cu_sq);
- //(D.f[DIR_MMP ])[ktsw ] = c1over216*(drho+three*(-vx1-vx2+vx3)+c9over2*(-vx1-vx2+vx3)*(-vx1-vx2+vx3)-cu_sq);
- //(D.f[DIR_PMP ])[ktse ] = c1over216*(drho+three*( vx1-vx2+vx3)+c9over2*( vx1-vx2+vx3)*( vx1-vx2+vx3)-cu_sq);
- //(D.f[DIR_MPM ])[kbnw ] = c1over216*(drho+three*(-vx1+vx2-vx3)+c9over2*(-vx1+vx2-vx3)*(-vx1+vx2-vx3)-cu_sq);
- //(D.f[DIR_PMM ])[kbse ] = c1over216*(drho+three*( vx1-vx2-vx3)+c9over2*( vx1-vx2-vx3)*( vx1-vx2-vx3)-cu_sq);
- //(D.f[DIR_MPP ])[ktnw ] = c1over216*(drho+three*(-vx1+vx2+vx3)+c9over2*(-vx1+vx2+vx3)*(-vx1+vx2+vx3)-cu_sq);
+ //(D.f[DIR_P00])[ke ] = c2over27* (drho+three*( vx1 )+c9over2*( vx1 )*( vx1 )-cu_sq);
+ //(D.f[DIR_M00])[kw ] = c2over27* (drho+three*(-vx1 )+c9over2*(-vx1 )*(-vx1 )-cu_sq);
+ //(D.f[DIR_0P0])[kn ] = c2over27* (drho+three*( vx2 )+c9over2*( vx2 )*( vx2 )-cu_sq);
+ //(D.f[DIR_0M0])[ks ] = c2over27* (drho+three*( -vx2 )+c9over2*( -vx2 )*( -vx2 )-cu_sq);
+ //(D.f[DIR_00P])[kt ] = c2over27* (drho+three*( vx3)+c9over2*( vx3)*( vx3)-cu_sq);
+ //(D.f[DIR_00M])[kb ] = c2over27* (drho+three*( -vx3)+c9over2*( -vx3)*( -vx3)-cu_sq);
+ //(D.f[DIR_PP0])[kne ] = c1over54* (drho+three*( vx1+vx2 )+c9over2*( vx1+vx2 )*( vx1+vx2 )-cu_sq);
+ //(D.f[DIR_MM0])[ksw ] = c1over54* (drho+three*(-vx1-vx2 )+c9over2*(-vx1-vx2 )*(-vx1-vx2 )-cu_sq);
+ //(D.f[DIR_PM0])[kse ] = c1over54* (drho+three*( vx1-vx2 )+c9over2*( vx1-vx2 )*( vx1-vx2 )-cu_sq);
+ //(D.f[DIR_MP0])[knw ] = c1over54* (drho+three*(-vx1+vx2 )+c9over2*(-vx1+vx2 )*(-vx1+vx2 )-cu_sq);
+ //(D.f[DIR_P0P])[kte ] = c1over54* (drho+three*( vx1 +vx3)+c9over2*( vx1 +vx3)*( vx1 +vx3)-cu_sq);
+ //(D.f[DIR_M0M])[kbw ] = c1over54* (drho+three*(-vx1 -vx3)+c9over2*(-vx1 -vx3)*(-vx1 -vx3)-cu_sq);
+ //(D.f[DIR_P0M])[kbe ] = c1over54* (drho+three*( vx1 -vx3)+c9over2*( vx1 -vx3)*( vx1 -vx3)-cu_sq);
+ //(D.f[DIR_M0P])[ktw ] = c1over54* (drho+three*(-vx1 +vx3)+c9over2*(-vx1 +vx3)*(-vx1 +vx3)-cu_sq);
+ //(D.f[DIR_0PP])[ktn ] = c1over54* (drho+three*( vx2+vx3)+c9over2*( vx2+vx3)*( vx2+vx3)-cu_sq);
+ //(D.f[DIR_0MM])[kbs ] = c1over54* (drho+three*( -vx2-vx3)+c9over2*( -vx2-vx3)*( -vx2-vx3)-cu_sq);
+ //(D.f[DIR_0PM])[kbn ] = c1over54* (drho+three*( vx2-vx3)+c9over2*( vx2-vx3)*( vx2-vx3)-cu_sq);
+ //(D.f[DIR_0MP])[kts ] = c1over54* (drho+three*( -vx2+vx3)+c9over2*( -vx2+vx3)*( -vx2+vx3)-cu_sq);
+ //(D.f[DIR_PPP])[ktne ] = c1over216*(drho+three*( vx1+vx2+vx3)+c9over2*( vx1+vx2+vx3)*( vx1+vx2+vx3)-cu_sq);
+ //(D.f[DIR_MMM])[kbsw ] = c1over216*(drho+three*(-vx1-vx2-vx3)+c9over2*(-vx1-vx2-vx3)*(-vx1-vx2-vx3)-cu_sq);
+ //(D.f[DIR_PPM])[kbne ] = c1over216*(drho+three*( vx1+vx2-vx3)+c9over2*( vx1+vx2-vx3)*( vx1+vx2-vx3)-cu_sq);
+ //(D.f[DIR_MMP])[ktsw ] = c1over216*(drho+three*(-vx1-vx2+vx3)+c9over2*(-vx1-vx2+vx3)*(-vx1-vx2+vx3)-cu_sq);
+ //(D.f[DIR_PMP])[ktse ] = c1over216*(drho+three*( vx1-vx2+vx3)+c9over2*( vx1-vx2+vx3)*( vx1-vx2+vx3)-cu_sq);
+ //(D.f[DIR_MPM])[kbnw ] = c1over216*(drho+three*(-vx1+vx2-vx3)+c9over2*(-vx1+vx2-vx3)*(-vx1+vx2-vx3)-cu_sq);
+ //(D.f[DIR_PMM])[kbse ] = c1over216*(drho+three*( vx1-vx2-vx3)+c9over2*( vx1-vx2-vx3)*( vx1-vx2-vx3)-cu_sq);
+ //(D.f[DIR_MPP])[ktnw ] = c1over216*(drho+three*(-vx1+vx2+vx3)+c9over2*(-vx1+vx2+vx3)*(-vx1+vx2+vx3)-cu_sq);
}
}
}
diff --git a/src/gpu/VirtualFluids_GPU/GPU/WaleCumulant27.cu b/src/gpu/VirtualFluids_GPU/GPU/WaleCumulant27.cu
index 16028e2f9f87716f43ed60f82ed513289e381b7c..cbb892296322bc164241ad18c8ab63201d34647e 100644
--- a/src/gpu/VirtualFluids_GPU/GPU/WaleCumulant27.cu
+++ b/src/gpu/VirtualFluids_GPU/GPU/WaleCumulant27.cu
@@ -23,7 +23,7 @@ __global__ void LB_Kernel_WaleBySoniMalav_Cum_AA2016_Comp_SP_27(
real* veloZ,
real* DDStart,
real* turbulentViscosity,
- int size_Mat,
+ unsigned long long numberOfLBnodes,
int level,
real* forces,
bool EvenOrOdd)
@@ -39,7 +39,7 @@ __global__ void LB_Kernel_WaleBySoniMalav_Cum_AA2016_Comp_SP_27(
const unsigned k = nx*(ny*z + y) + x;
//////////////////////////////////////////////////////////////////////////
- if (k<size_Mat)
+ if (k<numberOfLBnodes)
{
////////////////////////////////////////////////////////////////////////////////
unsigned int BC;
@@ -50,63 +50,63 @@ __global__ void LB_Kernel_WaleBySoniMalav_Cum_AA2016_Comp_SP_27(
Distributions27 D;
if (EvenOrOdd == true)
{
- D.f[DIR_P00] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_M00] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0P0] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0M0] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00P] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00M] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_PP0] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_MM0] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_PM0] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_MP0] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_P0P] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_M0M] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_P0M] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_M0P] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0PP] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0MM] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0PM] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0MP] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_PPP] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_MMP] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_PMP] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_MPP] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_PPM] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_MMM] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_PMM] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_MPM] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DDStart[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DDStart[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DDStart[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DDStart[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DDStart[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DDStart[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DDStart[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DDStart[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DDStart[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DDStart[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DDStart[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DDStart[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DDStart[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DDStart[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DDStart[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DDStart[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DDStart[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DDStart[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DDStart[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DDStart[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DDStart[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DDStart[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DDStart[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DDStart[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DDStart[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DDStart[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DDStart[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_P00] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0M0] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0P0] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00M] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00P] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_MM0] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_PP0] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_MP0] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_PM0] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_M0M] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_P0P] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_M0P] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_P0M] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0MM] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0PP] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0MP] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0PM] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_MMM] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_PPM] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_MPM] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_PMM] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_MMP] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_PPP] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_MPP] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_PMP] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_M00] = &DDStart[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DDStart[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DDStart[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DDStart[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DDStart[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DDStart[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DDStart[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DDStart[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DDStart[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DDStart[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DDStart[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DDStart[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DDStart[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DDStart[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DDStart[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DDStart[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DDStart[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DDStart[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DDStart[DIR_000 * numberOfLBnodes];
+ D.f[DIR_MMM] = &DDStart[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DDStart[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DDStart[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DDStart[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DDStart[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PPP] = &DDStart[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DDStart[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DDStart[DIR_MPM * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////
diff --git a/src/gpu/VirtualFluids_GPU/GPU/WallFunction.cu b/src/gpu/VirtualFluids_GPU/GPU/WallFunction.cu
index d48fa80fd14ce15f4a380ed46403654b43c805e8..d2fe5935af9b2d3ad78f492e3a9d182873d20808 100644
--- a/src/gpu/VirtualFluids_GPU/GPU/WallFunction.cu
+++ b/src/gpu/VirtualFluids_GPU/GPU/WallFunction.cu
@@ -20,69 +20,69 @@ __global__ void WallFunction27(
unsigned int* neighborX,
unsigned int* neighborY,
unsigned int* neighborZ,
- unsigned int size_Mat,
+ unsigned long long numberOfLBnodes,
bool isEvenTimestep)
{
Distributions27 D;
if (isEvenTimestep==true)
{
- D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DD[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
- D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
- D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
- D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
- D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
- D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
- D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
- D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
+ D.f[DIR_M00] = &DD[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DD[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DD[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DD[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DD[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DD[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DD[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DD[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DD[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DD[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DD[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DD[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DD[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DD[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DD[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DD[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DD[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DD[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DD[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DD[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MMP] = &DD[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DD[DIR_MPM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DD[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PPM] = &DD[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MMM] = &DD[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DD[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DD[DIR_PMP * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////
const unsigned x = threadIdx.x; // Globaler x-Index
@@ -107,24 +107,24 @@ __global__ void WallFunction27(
// *q_dirBE, *q_dirTW, *q_dirTN, *q_dirBS, *q_dirBN, *q_dirTS,
// *q_dirTNE, *q_dirTSW, *q_dirTSE, *q_dirTNW, *q_dirBNE, *q_dirBSW,
// *q_dirBSE, *q_dirBNW;
- //q_dirE = &QQ[DIR_P00 * numberOfBCnodes];
- //q_dirW = &QQ[DIR_M00 * numberOfBCnodes];
- //q_dirN = &QQ[DIR_0P0 * numberOfBCnodes];
- //q_dirS = &QQ[DIR_0M0 * numberOfBCnodes];
- //q_dirT = &QQ[DIR_00P * numberOfBCnodes];
- //q_dirB = &QQ[DIR_00M * numberOfBCnodes];
- //q_dirNE = &QQ[DIR_PP0 * numberOfBCnodes];
- //q_dirSW = &QQ[DIR_MM0 * numberOfBCnodes];
- //q_dirSE = &QQ[DIR_PM0 * numberOfBCnodes];
- //q_dirNW = &QQ[DIR_MP0 * numberOfBCnodes];
- //q_dirTE = &QQ[DIR_P0P * numberOfBCnodes];
- //q_dirBW = &QQ[DIR_M0M * numberOfBCnodes];
- //q_dirBE = &QQ[DIR_P0M * numberOfBCnodes];
- //q_dirTW = &QQ[DIR_M0P * numberOfBCnodes];
- //q_dirTN = &QQ[DIR_0PP * numberOfBCnodes];
- //q_dirBS = &QQ[DIR_0MM * numberOfBCnodes];
- //q_dirBN = &QQ[DIR_0PM * numberOfBCnodes];
- //q_dirTS = &QQ[DIR_0MP * numberOfBCnodes];
+ //q_dirE = &QQ[DIR_P00 * numberOfBCnodes];
+ //q_dirW = &QQ[DIR_M00 * numberOfBCnodes];
+ //q_dirN = &QQ[DIR_0P0 * numberOfBCnodes];
+ //q_dirS = &QQ[DIR_0M0 * numberOfBCnodes];
+ //q_dirT = &QQ[DIR_00P * numberOfBCnodes];
+ //q_dirB = &QQ[DIR_00M * numberOfBCnodes];
+ //q_dirNE = &QQ[DIR_PP0 * numberOfBCnodes];
+ //q_dirSW = &QQ[DIR_MM0 * numberOfBCnodes];
+ //q_dirSE = &QQ[DIR_PM0 * numberOfBCnodes];
+ //q_dirNW = &QQ[DIR_MP0 * numberOfBCnodes];
+ //q_dirTE = &QQ[DIR_P0P * numberOfBCnodes];
+ //q_dirBW = &QQ[DIR_M0M * numberOfBCnodes];
+ //q_dirBE = &QQ[DIR_P0M * numberOfBCnodes];
+ //q_dirTW = &QQ[DIR_M0P * numberOfBCnodes];
+ //q_dirTN = &QQ[DIR_0PP * numberOfBCnodes];
+ //q_dirBS = &QQ[DIR_0MM * numberOfBCnodes];
+ //q_dirBN = &QQ[DIR_0PM * numberOfBCnodes];
+ //q_dirTS = &QQ[DIR_0MP * numberOfBCnodes];
//q_dirTNE = &QQ[DIR_PPP * numberOfBCnodes];
//q_dirTSW = &QQ[DIR_MMP * numberOfBCnodes];
//q_dirTSE = &QQ[DIR_PMP * numberOfBCnodes];
@@ -167,32 +167,32 @@ __global__ void WallFunction27(
real f_E, f_W, f_N, f_S, f_T, f_B, f_NE, f_SW, f_SE, f_NW, f_TE, f_BW, f_BE,
f_TW, f_TN, f_BS, f_BN, f_TS, f_TNE, f_TSW, f_TSE, f_TNW, f_BNE, f_BSW, f_BSE, f_BNW;
- f_W = (D.f[DIR_P00 ])[ke ];
- f_E = (D.f[DIR_M00 ])[kw ];
- f_S = (D.f[DIR_0P0 ])[kn ];
- f_N = (D.f[DIR_0M0 ])[ks ];
- f_B = (D.f[DIR_00P ])[kt ];
- f_T = (D.f[DIR_00M ])[kb ];
- f_SW = (D.f[DIR_PP0 ])[kne ];
- f_NE = (D.f[DIR_MM0 ])[ksw ];
- f_NW = (D.f[DIR_PM0 ])[kse ];
- f_SE = (D.f[DIR_MP0 ])[knw ];
- f_BW = (D.f[DIR_P0P ])[kte ];
- f_TE = (D.f[DIR_M0M ])[kbw ];
- f_TW = (D.f[DIR_P0M ])[kbe ];
- f_BE = (D.f[DIR_M0P ])[ktw ];
- f_BS = (D.f[DIR_0PP ])[ktn ];
- f_TN = (D.f[DIR_0MM ])[kbs ];
- f_TS = (D.f[DIR_0PM ])[kbn ];
- f_BN = (D.f[DIR_0MP ])[kts ];
- f_BSW = (D.f[DIR_PPP ])[ktne ];
- f_BNE = (D.f[DIR_MMP ])[ktsw ];
- f_BNW = (D.f[DIR_PMP ])[ktse ];
- f_BSE = (D.f[DIR_MPP ])[ktnw ];
- f_TSW = (D.f[DIR_PPM ])[kbne ];
- f_TNE = (D.f[DIR_MMM ])[kbsw ];
- f_TNW = (D.f[DIR_PMM ])[kbse ];
- f_TSE = (D.f[DIR_MPM ])[kbnw ];
+ f_W = (D.f[DIR_P00])[ke ];
+ f_E = (D.f[DIR_M00])[kw ];
+ f_S = (D.f[DIR_0P0])[kn ];
+ f_N = (D.f[DIR_0M0])[ks ];
+ f_B = (D.f[DIR_00P])[kt ];
+ f_T = (D.f[DIR_00M])[kb ];
+ f_SW = (D.f[DIR_PP0])[kne ];
+ f_NE = (D.f[DIR_MM0])[ksw ];
+ f_NW = (D.f[DIR_PM0])[kse ];
+ f_SE = (D.f[DIR_MP0])[knw ];
+ f_BW = (D.f[DIR_P0P])[kte ];
+ f_TE = (D.f[DIR_M0M])[kbw ];
+ f_TW = (D.f[DIR_P0M])[kbe ];
+ f_BE = (D.f[DIR_M0P])[ktw ];
+ f_BS = (D.f[DIR_0PP])[ktn ];
+ f_TN = (D.f[DIR_0MM])[kbs ];
+ f_TS = (D.f[DIR_0PM])[kbn ];
+ f_BN = (D.f[DIR_0MP])[kts ];
+ f_BSW = (D.f[DIR_PPP])[ktne ];
+ f_BNE = (D.f[DIR_MMP])[ktsw ];
+ f_BNW = (D.f[DIR_PMP])[ktse ];
+ f_BSE = (D.f[DIR_MPP])[ktnw ];
+ f_TSW = (D.f[DIR_PPM])[kbne ];
+ f_TNE = (D.f[DIR_MMM])[kbsw ];
+ f_TNW = (D.f[DIR_PMM])[kbse ];
+ f_TSE = (D.f[DIR_MPM])[kbnw ];
////////////////////////////////////////////////////////////////////////////////
// real vx2, vx3, feq, q;
real vx1, drho;
@@ -234,63 +234,63 @@ __global__ void WallFunction27(
// //////////////////////////////////////////////////////////////////////////
// if (isEvenTimestep==false)
// {
- // D.f[DIR_P00 ] = &DD[DIR_P00 *size_Mat];
- // D.f[DIR_M00 ] = &DD[DIR_M00 *size_Mat];
- // D.f[DIR_0P0 ] = &DD[DIR_0P0 *size_Mat];
- // D.f[DIR_0M0 ] = &DD[DIR_0M0 *size_Mat];
- // D.f[DIR_00P ] = &DD[DIR_00P *size_Mat];
- // D.f[DIR_00M ] = &DD[DIR_00M *size_Mat];
- // D.f[DIR_PP0 ] = &DD[DIR_PP0 *size_Mat];
- // D.f[DIR_MM0 ] = &DD[DIR_MM0 *size_Mat];
- // D.f[DIR_PM0 ] = &DD[DIR_PM0 *size_Mat];
- // D.f[DIR_MP0 ] = &DD[DIR_MP0 *size_Mat];
- // D.f[DIR_P0P ] = &DD[DIR_P0P *size_Mat];
- // D.f[DIR_M0M ] = &DD[DIR_M0M *size_Mat];
- // D.f[DIR_P0M ] = &DD[DIR_P0M *size_Mat];
- // D.f[DIR_M0P ] = &DD[DIR_M0P *size_Mat];
- // D.f[DIR_0PP ] = &DD[DIR_0PP *size_Mat];
- // D.f[DIR_0MM ] = &DD[DIR_0MM *size_Mat];
- // D.f[DIR_0PM ] = &DD[DIR_0PM *size_Mat];
- // D.f[DIR_0MP ] = &DD[DIR_0MP *size_Mat];
- // D.f[DIR_000] = &DD[DIR_000*size_Mat];
- // D.f[DIR_PPP ] = &DD[DIR_PPP *size_Mat];
- // D.f[DIR_MMP ] = &DD[DIR_MMP *size_Mat];
- // D.f[DIR_PMP ] = &DD[DIR_PMP *size_Mat];
- // D.f[DIR_MPP ] = &DD[DIR_MPP *size_Mat];
- // D.f[DIR_PPM ] = &DD[DIR_PPM *size_Mat];
- // D.f[DIR_MMM ] = &DD[DIR_MMM *size_Mat];
- // D.f[DIR_PMM ] = &DD[DIR_PMM *size_Mat];
- // D.f[DIR_MPM ] = &DD[DIR_MPM *size_Mat];
+ // D.f[DIR_P00] = &DD[DIR_P00 * size_Mat];
+ // D.f[DIR_M00] = &DD[DIR_M00 * size_Mat];
+ // D.f[DIR_0P0] = &DD[DIR_0P0 * size_Mat];
+ // D.f[DIR_0M0] = &DD[DIR_0M0 * size_Mat];
+ // D.f[DIR_00P] = &DD[DIR_00P * size_Mat];
+ // D.f[DIR_00M] = &DD[DIR_00M * size_Mat];
+ // D.f[DIR_PP0] = &DD[DIR_PP0 * size_Mat];
+ // D.f[DIR_MM0] = &DD[DIR_MM0 * size_Mat];
+ // D.f[DIR_PM0] = &DD[DIR_PM0 * size_Mat];
+ // D.f[DIR_MP0] = &DD[DIR_MP0 * size_Mat];
+ // D.f[DIR_P0P] = &DD[DIR_P0P * size_Mat];
+ // D.f[DIR_M0M] = &DD[DIR_M0M * size_Mat];
+ // D.f[DIR_P0M] = &DD[DIR_P0M * size_Mat];
+ // D.f[DIR_M0P] = &DD[DIR_M0P * size_Mat];
+ // D.f[DIR_0PP] = &DD[DIR_0PP * size_Mat];
+ // D.f[DIR_0MM] = &DD[DIR_0MM * size_Mat];
+ // D.f[DIR_0PM] = &DD[DIR_0PM * size_Mat];
+ // D.f[DIR_0MP] = &DD[DIR_0MP * size_Mat];
+ // D.f[DIR_000] = &DD[DIR_000 * size_Mat];
+ // D.f[DIR_PPP] = &DD[DIR_PPP * size_Mat];
+ // D.f[DIR_MMP] = &DD[DIR_MMP * size_Mat];
+ // D.f[DIR_PMP] = &DD[DIR_PMP * size_Mat];
+ // D.f[DIR_MPP] = &DD[DIR_MPP * size_Mat];
+ // D.f[DIR_PPM] = &DD[DIR_PPM * size_Mat];
+ // D.f[DIR_MMM] = &DD[DIR_MMM * size_Mat];
+ // D.f[DIR_PMM] = &DD[DIR_PMM * size_Mat];
+ // D.f[DIR_MPM] = &DD[DIR_MPM * size_Mat];
// }
// else
// {
- // D.f[DIR_M00 ] = &DD[DIR_P00 *size_Mat];
- // D.f[DIR_P00 ] = &DD[DIR_M00 *size_Mat];
- // D.f[DIR_0M0 ] = &DD[DIR_0P0 *size_Mat];
- // D.f[DIR_0P0 ] = &DD[DIR_0M0 *size_Mat];
- // D.f[DIR_00M ] = &DD[DIR_00P *size_Mat];
- // D.f[DIR_00P ] = &DD[DIR_00M *size_Mat];
- // D.f[DIR_MM0 ] = &DD[DIR_PP0 *size_Mat];
- // D.f[DIR_PP0 ] = &DD[DIR_MM0 *size_Mat];
- // D.f[DIR_MP0 ] = &DD[DIR_PM0 *size_Mat];
- // D.f[DIR_PM0 ] = &DD[DIR_MP0 *size_Mat];
- // D.f[DIR_M0M ] = &DD[DIR_P0P *size_Mat];
- // D.f[DIR_P0P ] = &DD[DIR_M0M *size_Mat];
- // D.f[DIR_M0P ] = &DD[DIR_P0M *size_Mat];
- // D.f[DIR_P0M ] = &DD[DIR_M0P *size_Mat];
- // D.f[DIR_0MM ] = &DD[DIR_0PP *size_Mat];
- // D.f[DIR_0PP ] = &DD[DIR_0MM *size_Mat];
- // D.f[DIR_0MP ] = &DD[DIR_0PM *size_Mat];
- // D.f[DIR_0PM ] = &DD[DIR_0MP *size_Mat];
- // D.f[DIR_000] = &DD[DIR_000*size_Mat];
- // D.f[DIR_PPP ] = &DD[DIR_MMM *size_Mat];
- // D.f[DIR_MMP ] = &DD[DIR_PPM *size_Mat];
- // D.f[DIR_PMP ] = &DD[DIR_MPM *size_Mat];
- // D.f[DIR_MPP ] = &DD[DIR_PMM *size_Mat];
- // D.f[DIR_PPM ] = &DD[DIR_MMP *size_Mat];
- // D.f[DIR_MMM ] = &DD[DIR_PPP *size_Mat];
- // D.f[DIR_PMM ] = &DD[DIR_MPP *size_Mat];
- // D.f[DIR_MPM ] = &DD[DIR_PMP *size_Mat];
+ // D.f[DIR_M00] = &DD[DIR_P00 * size_Mat];
+ // D.f[DIR_P00] = &DD[DIR_M00 * size_Mat];
+ // D.f[DIR_0M0] = &DD[DIR_0P0 * size_Mat];
+ // D.f[DIR_0P0] = &DD[DIR_0M0 * size_Mat];
+ // D.f[DIR_00M] = &DD[DIR_00P * size_Mat];
+ // D.f[DIR_00P] = &DD[DIR_00M * size_Mat];
+ // D.f[DIR_MM0] = &DD[DIR_PP0 * size_Mat];
+ // D.f[DIR_PP0] = &DD[DIR_MM0 * size_Mat];
+ // D.f[DIR_MP0] = &DD[DIR_PM0 * size_Mat];
+ // D.f[DIR_PM0] = &DD[DIR_MP0 * size_Mat];
+ // D.f[DIR_M0M] = &DD[DIR_P0P * size_Mat];
+ // D.f[DIR_P0P] = &DD[DIR_M0M * size_Mat];
+ // D.f[DIR_M0P] = &DD[DIR_P0M * size_Mat];
+ // D.f[DIR_P0M] = &DD[DIR_M0P * size_Mat];
+ // D.f[DIR_0MM] = &DD[DIR_0PP * size_Mat];
+ // D.f[DIR_0PP] = &DD[DIR_0MM * size_Mat];
+ // D.f[DIR_0MP] = &DD[DIR_0PM * size_Mat];
+ // D.f[DIR_0PM] = &DD[DIR_0MP * size_Mat];
+ // D.f[DIR_000] = &DD[DIR_000 * size_Mat];
+ // D.f[DIR_PPP] = &DD[DIR_MMM * size_Mat];
+ // D.f[DIR_MMP] = &DD[DIR_PPM * size_Mat];
+ // D.f[DIR_PMP] = &DD[DIR_MPM * size_Mat];
+ // D.f[DIR_MPP] = &DD[DIR_PMM * size_Mat];
+ // D.f[DIR_PPM] = &DD[DIR_MMP * size_Mat];
+ // D.f[DIR_MMM] = &DD[DIR_PPP * size_Mat];
+ // D.f[DIR_PMM] = &DD[DIR_MPP * size_Mat];
+ // D.f[DIR_MPM] = &DD[DIR_PMP * size_Mat];
// }
// ////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
// //Test
diff --git a/src/gpu/VirtualFluids_GPU/Init/InitLattice.cpp b/src/gpu/VirtualFluids_GPU/Init/InitLattice.cpp
index 2c85de9e3ec57d50a66fde2c49d3e703676fbf04..508e4498c36d352761c3ecaf24abaa52a5f84bbe 100644
--- a/src/gpu/VirtualFluids_GPU/Init/InitLattice.cpp
+++ b/src/gpu/VirtualFluids_GPU/Init/InitLattice.cpp
@@ -45,26 +45,44 @@ void initLattice(SPtr<Parameter> para, SPtr<PreProcessor> preProcessor, SPtr<Cud
preProcessor->init(para, lev);
CalcMacCompSP27(
- para->getParD(lev)->velocityX, para->getParD(lev)->velocityY, para->getParD(lev)->velocityZ, para->getParD(lev)->rho,
- para->getParD(lev)->pressure, para->getParD(lev)->typeOfGridNode, para->getParD(lev)->neighborX,
- para->getParD(lev)->neighborY, para->getParD(lev)->neighborZ, para->getParD(lev)->numberOfNodes,
- para->getParD(lev)->numberofthreads, para->getParD(lev)->distributions.f[0], para->getParD(lev)->isEvenTimestep);
+ para->getParD(lev)->velocityX,
+ para->getParD(lev)->velocityY,
+ para->getParD(lev)->velocityZ,
+ para->getParD(lev)->rho,
+ para->getParD(lev)->pressure,
+ para->getParD(lev)->typeOfGridNode,
+ para->getParD(lev)->neighborX,
+ para->getParD(lev)->neighborY,
+ para->getParD(lev)->neighborZ,
+ para->getParD(lev)->numberOfNodes,
+ para->getParD(lev)->numberofthreads,
+ para->getParD(lev)->distributions.f[0],
+ para->getParD(lev)->isEvenTimestep);
if (para->getCalcMedian()) {
constexpr uint tdiff = 1;
- CalcMacMedSP27(para->getParD(lev)->vx_SP_Med, para->getParD(lev)->vy_SP_Med, para->getParD(lev)->vz_SP_Med,
- para->getParD(lev)->rho_SP_Med, para->getParD(lev)->press_SP_Med, para->getParD(lev)->typeOfGridNode,
- para->getParD(lev)->neighborX, para->getParD(lev)->neighborY,
- para->getParD(lev)->neighborZ, tdiff, para->getParD(lev)->numberOfNodes,
- para->getParD(lev)->numberofthreads, para->getParD(lev)->isEvenTimestep);
+ CalcMacMedSP27(
+ para->getParD(lev)->vx_SP_Med,
+ para->getParD(lev)->vy_SP_Med,
+ para->getParD(lev)->vz_SP_Med,
+ para->getParD(lev)->rho_SP_Med,
+ para->getParD(lev)->press_SP_Med,
+ para->getParD(lev)->typeOfGridNode,
+ para->getParD(lev)->neighborX,
+ para->getParD(lev)->neighborY,
+ para->getParD(lev)->neighborZ,
+ tdiff,
+ para->getParD(lev)->numberOfNodes,
+ para->getParD(lev)->numberofthreads,
+ para->getParD(lev)->isEvenTimestep);
}
// advection - diffusion
if (para->getDiffOn()) {
cudaMemoryManager->cudaAllocConcentration(lev);
- for (unsigned int i = 0; i < para->getParH(lev)->numberOfNodes; i++) {
- para->getParH(lev)->Conc[i] = para->getTemperatureInit();
+ for (size_t index = 0; index < para->getParH(lev)->numberOfNodes; index++) {
+ para->getParH(lev)->Conc[index] = para->getTemperatureInit();
}
initTemperatur(para.get(), cudaMemoryManager.get(), lev);
}
diff --git a/src/gpu/VirtualFluids_GPU/Init/PositionReader.cpp b/src/gpu/VirtualFluids_GPU/Init/PositionReader.cpp
index 4e5a862d3fd1ed19109073aae0fe4c731f7f3e91..6eaa0b17653aaf5257c00e674c87e2844c26cf5d 100644
--- a/src/gpu/VirtualFluids_GPU/Init/PositionReader.cpp
+++ b/src/gpu/VirtualFluids_GPU/Init/PositionReader.cpp
@@ -3,6 +3,7 @@
#include "Parameter/Parameter.h"
#include <basics/utilities/UbFileInputASCII.h>
+
using namespace vf::lbm::dir;
//////////////////////////////////////////////////////////////////////////
@@ -169,7 +170,7 @@ void PositionReader::definePropellerQs(Parameter* para)
//////////////////////////////////////////////////////////////////
for(uint u=0; u<para->getParH(para->getFine())->propellerBC.numberOfBCnodes; u++)
{
- for (int dir = DIR_P00; dir<=DIR_MMM; dir++)
+ for (size_t dir = DIR_P00; dir<=DIR_MMM; dir++)
{
if ((dir==DIR_P00) ||
(dir==DIR_PP0) || (dir==DIR_PM0) || (dir==DIR_P0P) || (dir==DIR_P0M) ||
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Kernel.h b/src/gpu/VirtualFluids_GPU/Kernel/Kernel.h
index 9f9f7539bc5a1e28612d956ca32234c5a3589f8a..50b4460d774010ea7d7b98cfa6fa505cdfeb88c2 100644
--- a/src/gpu/VirtualFluids_GPU/Kernel/Kernel.h
+++ b/src/gpu/VirtualFluids_GPU/Kernel/Kernel.h
@@ -3,8 +3,11 @@
#include <vector>
+#include "LBM/LB.h"
+
#include "Kernel/Utilities/KernelGroup.h"
#include "PreProcessor/PreProcessorType.h"
+#include "Parameter/CudaStreamManager.h"
#include <helper_cuda.h>
@@ -13,7 +16,7 @@ class Kernel
public:
virtual ~Kernel() = default;
virtual void run() = 0;
- virtual void runOnIndices(const unsigned int *indices, unsigned int size_indices, int stream = -1) = 0; //if stream == -1: run on default stream
+ virtual void runOnIndices(const unsigned int *indices, unsigned int size_indices, CollisionTemplate collisionTemplate, CudaStreamIndex streamIdx=CudaStreamIndex::Legacy) = 0;
virtual bool checkParameter() = 0;
virtual std::vector<PreProcessorType> getPreProcessorTypes() = 0;
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/KernelImp.cpp b/src/gpu/VirtualFluids_GPU/Kernel/KernelImp.cpp
index 630aaf7339afc2907ab6bfbf65bd5fc55f75e215..9bd3945aa81147d03be2b1eac3ddec7c24d71532 100644
--- a/src/gpu/VirtualFluids_GPU/Kernel/KernelImp.cpp
+++ b/src/gpu/VirtualFluids_GPU/Kernel/KernelImp.cpp
@@ -1,9 +1,11 @@
#include "KernelImp.h"
+#include "LBM/LB.h"
+
#include "Kernel/Utilities/CheckParameterStrategy/CheckParameterStrategy.h"
-void KernelImp::runOnIndices(const unsigned int *indices, unsigned int size_indices, int stream)
+void KernelImp::runOnIndices(const unsigned int *indices, unsigned int size_indices, CollisionTemplate collisionTemplate, CudaStreamIndex streamIndex)
{
printf("Method not implemented for this Kernel \n");
}
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/KernelImp.h b/src/gpu/VirtualFluids_GPU/Kernel/KernelImp.h
index 0141ddda7e9579cc84148d26727ed81c084ea0c5..a96c2c123472ca33f635273e06a5bf36a745654d 100644
--- a/src/gpu/VirtualFluids_GPU/Kernel/KernelImp.h
+++ b/src/gpu/VirtualFluids_GPU/Kernel/KernelImp.h
@@ -1,6 +1,8 @@
#ifndef KERNEL_IMP_H
#define KERNEL_IMP_H
+#include "LBM/LB.h"
+
#include "Kernel.h"
#include <memory>
@@ -9,12 +11,12 @@
class CheckParameterStrategy;
class Parameter;
-
+class CudaStreamManager;
class KernelImp : public Kernel
{
public:
virtual void run() = 0;
- virtual void runOnIndices(const unsigned int *indices, unsigned int size_indices, int stream = -1);
+ virtual void runOnIndices(const unsigned int *indices, unsigned int size_indices, CollisionTemplate collisionTemplate, CudaStreamIndex streamIndex=CudaStreamIndex::Legacy);
bool checkParameter();
std::vector<PreProcessorType> getPreProcessorTypes();
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/AdvectionDiffusion/Compressible/Mod27/ADComp27/ADComp27.cu b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/AdvectionDiffusion/Compressible/Mod27/ADComp27/ADComp27.cu
index 51b9e4537fa0857e9302aa638ae7729fa9adcdbe..d4d6307f688da4c8fa37c54fb4958681d5ec4941 100644
--- a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/AdvectionDiffusion/Compressible/Mod27/ADComp27/ADComp27.cu
+++ b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/AdvectionDiffusion/Compressible/Mod27/ADComp27/ADComp27.cu
@@ -2,6 +2,7 @@
#include "ADComp27_Device.cuh"
#include "Parameter/Parameter.h"
+#include "cuda/CudaGrid.h"
std::shared_ptr<ADComp27> ADComp27::getNewInstance(std::shared_ptr<Parameter> para, int level)
{
@@ -10,34 +11,19 @@ std::shared_ptr<ADComp27> ADComp27::getNewInstance(std::shared_ptr<Parameter> pa
void ADComp27::run()
{
- int size_Mat = para->getParD(level)->numberOfNodes;
- int numberOfThreads = para->getParD(level)->numberofthreads;
-
- int Grid = (size_Mat / numberOfThreads) + 1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid / Grid1) + 1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 grid(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1);
-
- LB_KERNEL_AD_COMP_27 << < grid, threads >> >( para->getParD(level)->diffusivity,
- para->getParD(level)->typeOfGridNode,
- para->getParD(level)->neighborX,
- para->getParD(level)->neighborY,
- para->getParD(level)->neighborZ,
- para->getParD(level)->distributions.f[0],
- para->getParD(level)->distributionsAD27.f[0],
- para->getParD(level)->numberOfNodes,
- para->getParD(level)->isEvenTimestep);
- getLastCudaError("LB_Kernel_ThS27 execution failed");
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(para->getParD(level)->numberofthreads, para->getParD(level)->numberOfNodes);
+
+ LB_KERNEL_AD_COMP_27<<< grid.grid, grid.threads >>>(
+ para->getParD(level)->diffusivity,
+ para->getParD(level)->typeOfGridNode,
+ para->getParD(level)->neighborX,
+ para->getParD(level)->neighborY,
+ para->getParD(level)->neighborZ,
+ para->getParD(level)->distributions.f[0],
+ para->getParD(level)->distributionsAD27.f[0],
+ para->getParD(level)->numberOfNodes,
+ para->getParD(level)->isEvenTimestep);
+ getLastCudaError("LB_KERNEL_AD_COMP_27 execution failed");
}
ADComp27::ADComp27(std::shared_ptr<Parameter> para, int level)
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/AdvectionDiffusion/Compressible/Mod27/ADComp27/ADComp27_Device.cu b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/AdvectionDiffusion/Compressible/Mod27/ADComp27/ADComp27_Device.cu
index b4c1236300bbb49fe2df1b3f458f506e989e142b..40adfff91713b7d6db1e861be9282d1f38516c22 100644
--- a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/AdvectionDiffusion/Compressible/Mod27/ADComp27/ADComp27_Device.cu
+++ b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/AdvectionDiffusion/Compressible/Mod27/ADComp27/ADComp27_Device.cu
@@ -38,125 +38,125 @@ __global__ void LB_KERNEL_AD_COMP_27(real diffusivity,
Distributions27 D;
if (EvenOrOdd == true)
{
- D.f[DIR_P00] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_M00] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0P0] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0M0] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00P] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00M] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_PP0] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_MM0] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_PM0] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_MP0] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_P0P] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_M0M] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_P0M] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_M0P] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0PP] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0MM] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0PM] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0MP] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_PPP] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_MMP] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_PMP] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_MPP] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_PPM] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_MMM] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_PMM] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_MPM] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DDStart[DIR_P00 * size_Mat];
+ D.f[DIR_M00] = &DDStart[DIR_M00 * size_Mat];
+ D.f[DIR_0P0] = &DDStart[DIR_0P0 * size_Mat];
+ D.f[DIR_0M0] = &DDStart[DIR_0M0 * size_Mat];
+ D.f[DIR_00P] = &DDStart[DIR_00P * size_Mat];
+ D.f[DIR_00M] = &DDStart[DIR_00M * size_Mat];
+ D.f[DIR_PP0] = &DDStart[DIR_PP0 * size_Mat];
+ D.f[DIR_MM0] = &DDStart[DIR_MM0 * size_Mat];
+ D.f[DIR_PM0] = &DDStart[DIR_PM0 * size_Mat];
+ D.f[DIR_MP0] = &DDStart[DIR_MP0 * size_Mat];
+ D.f[DIR_P0P] = &DDStart[DIR_P0P * size_Mat];
+ D.f[DIR_M0M] = &DDStart[DIR_M0M * size_Mat];
+ D.f[DIR_P0M] = &DDStart[DIR_P0M * size_Mat];
+ D.f[DIR_M0P] = &DDStart[DIR_M0P * size_Mat];
+ D.f[DIR_0PP] = &DDStart[DIR_0PP * size_Mat];
+ D.f[DIR_0MM] = &DDStart[DIR_0MM * size_Mat];
+ D.f[DIR_0PM] = &DDStart[DIR_0PM * size_Mat];
+ D.f[DIR_0MP] = &DDStart[DIR_0MP * size_Mat];
+ D.f[DIR_000] = &DDStart[DIR_000 * size_Mat];
+ D.f[DIR_PPP] = &DDStart[DIR_PPP * size_Mat];
+ D.f[DIR_MMP] = &DDStart[DIR_MMP * size_Mat];
+ D.f[DIR_PMP] = &DDStart[DIR_PMP * size_Mat];
+ D.f[DIR_MPP] = &DDStart[DIR_MPP * size_Mat];
+ D.f[DIR_PPM] = &DDStart[DIR_PPM * size_Mat];
+ D.f[DIR_MMM] = &DDStart[DIR_MMM * size_Mat];
+ D.f[DIR_PMM] = &DDStart[DIR_PMM * size_Mat];
+ D.f[DIR_MPM] = &DDStart[DIR_MPM * size_Mat];
}
else
{
- D.f[DIR_M00] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_P00] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0M0] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0P0] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00M] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00P] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_MM0] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_PP0] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_MP0] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_PM0] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_M0M] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_P0P] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_M0P] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_P0M] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0MM] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0PP] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0MP] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0PM] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_MMM] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_PPM] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_MPM] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_PMM] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_MMP] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_PPP] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_MPP] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_PMP] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_M00] = &DDStart[DIR_P00 * size_Mat];
+ D.f[DIR_P00] = &DDStart[DIR_M00 * size_Mat];
+ D.f[DIR_0M0] = &DDStart[DIR_0P0 * size_Mat];
+ D.f[DIR_0P0] = &DDStart[DIR_0M0 * size_Mat];
+ D.f[DIR_00M] = &DDStart[DIR_00P * size_Mat];
+ D.f[DIR_00P] = &DDStart[DIR_00M * size_Mat];
+ D.f[DIR_MM0] = &DDStart[DIR_PP0 * size_Mat];
+ D.f[DIR_PP0] = &DDStart[DIR_MM0 * size_Mat];
+ D.f[DIR_MP0] = &DDStart[DIR_PM0 * size_Mat];
+ D.f[DIR_PM0] = &DDStart[DIR_MP0 * size_Mat];
+ D.f[DIR_M0M] = &DDStart[DIR_P0P * size_Mat];
+ D.f[DIR_P0P] = &DDStart[DIR_M0M * size_Mat];
+ D.f[DIR_M0P] = &DDStart[DIR_P0M * size_Mat];
+ D.f[DIR_P0M] = &DDStart[DIR_M0P * size_Mat];
+ D.f[DIR_0MM] = &DDStart[DIR_0PP * size_Mat];
+ D.f[DIR_0PP] = &DDStart[DIR_0MM * size_Mat];
+ D.f[DIR_0MP] = &DDStart[DIR_0PM * size_Mat];
+ D.f[DIR_0PM] = &DDStart[DIR_0MP * size_Mat];
+ D.f[DIR_000] = &DDStart[DIR_000 * size_Mat];
+ D.f[DIR_MMM] = &DDStart[DIR_PPP * size_Mat];
+ D.f[DIR_PPM] = &DDStart[DIR_MMP * size_Mat];
+ D.f[DIR_MPM] = &DDStart[DIR_PMP * size_Mat];
+ D.f[DIR_PMM] = &DDStart[DIR_MPP * size_Mat];
+ D.f[DIR_MMP] = &DDStart[DIR_PPM * size_Mat];
+ D.f[DIR_PPP] = &DDStart[DIR_MMM * size_Mat];
+ D.f[DIR_MPP] = &DDStart[DIR_PMM * size_Mat];
+ D.f[DIR_PMP] = &DDStart[DIR_MPM * size_Mat];
}
Distributions27 D27;
if (EvenOrOdd == true)
{
- D27.f[DIR_P00] = &DD27[DIR_P00 *size_Mat];
- D27.f[DIR_M00] = &DD27[DIR_M00 *size_Mat];
- D27.f[DIR_0P0] = &DD27[DIR_0P0 *size_Mat];
- D27.f[DIR_0M0] = &DD27[DIR_0M0 *size_Mat];
- D27.f[DIR_00P] = &DD27[DIR_00P *size_Mat];
- D27.f[DIR_00M] = &DD27[DIR_00M *size_Mat];
- D27.f[DIR_PP0] = &DD27[DIR_PP0 *size_Mat];
- D27.f[DIR_MM0] = &DD27[DIR_MM0 *size_Mat];
- D27.f[DIR_PM0] = &DD27[DIR_PM0 *size_Mat];
- D27.f[DIR_MP0] = &DD27[DIR_MP0 *size_Mat];
- D27.f[DIR_P0P] = &DD27[DIR_P0P *size_Mat];
- D27.f[DIR_M0M] = &DD27[DIR_M0M *size_Mat];
- D27.f[DIR_P0M] = &DD27[DIR_P0M *size_Mat];
- D27.f[DIR_M0P] = &DD27[DIR_M0P *size_Mat];
- D27.f[DIR_0PP] = &DD27[DIR_0PP *size_Mat];
- D27.f[DIR_0MM] = &DD27[DIR_0MM *size_Mat];
- D27.f[DIR_0PM] = &DD27[DIR_0PM *size_Mat];
- D27.f[DIR_0MP] = &DD27[DIR_0MP *size_Mat];
- D27.f[DIR_000] = &DD27[DIR_000*size_Mat];
- D27.f[DIR_PPP] = &DD27[DIR_PPP *size_Mat];
- D27.f[DIR_MMP] = &DD27[DIR_MMP *size_Mat];
- D27.f[DIR_PMP] = &DD27[DIR_PMP *size_Mat];
- D27.f[DIR_MPP] = &DD27[DIR_MPP *size_Mat];
- D27.f[DIR_PPM] = &DD27[DIR_PPM *size_Mat];
- D27.f[DIR_MMM] = &DD27[DIR_MMM *size_Mat];
- D27.f[DIR_PMM] = &DD27[DIR_PMM *size_Mat];
- D27.f[DIR_MPM] = &DD27[DIR_MPM *size_Mat];
+ D27.f[DIR_P00] = &DD27[DIR_P00 * size_Mat];
+ D27.f[DIR_M00] = &DD27[DIR_M00 * size_Mat];
+ D27.f[DIR_0P0] = &DD27[DIR_0P0 * size_Mat];
+ D27.f[DIR_0M0] = &DD27[DIR_0M0 * size_Mat];
+ D27.f[DIR_00P] = &DD27[DIR_00P * size_Mat];
+ D27.f[DIR_00M] = &DD27[DIR_00M * size_Mat];
+ D27.f[DIR_PP0] = &DD27[DIR_PP0 * size_Mat];
+ D27.f[DIR_MM0] = &DD27[DIR_MM0 * size_Mat];
+ D27.f[DIR_PM0] = &DD27[DIR_PM0 * size_Mat];
+ D27.f[DIR_MP0] = &DD27[DIR_MP0 * size_Mat];
+ D27.f[DIR_P0P] = &DD27[DIR_P0P * size_Mat];
+ D27.f[DIR_M0M] = &DD27[DIR_M0M * size_Mat];
+ D27.f[DIR_P0M] = &DD27[DIR_P0M * size_Mat];
+ D27.f[DIR_M0P] = &DD27[DIR_M0P * size_Mat];
+ D27.f[DIR_0PP] = &DD27[DIR_0PP * size_Mat];
+ D27.f[DIR_0MM] = &DD27[DIR_0MM * size_Mat];
+ D27.f[DIR_0PM] = &DD27[DIR_0PM * size_Mat];
+ D27.f[DIR_0MP] = &DD27[DIR_0MP * size_Mat];
+ D27.f[DIR_000] = &DD27[DIR_000 * size_Mat];
+ D27.f[DIR_PPP] = &DD27[DIR_PPP * size_Mat];
+ D27.f[DIR_MMP] = &DD27[DIR_MMP * size_Mat];
+ D27.f[DIR_PMP] = &DD27[DIR_PMP * size_Mat];
+ D27.f[DIR_MPP] = &DD27[DIR_MPP * size_Mat];
+ D27.f[DIR_PPM] = &DD27[DIR_PPM * size_Mat];
+ D27.f[DIR_MMM] = &DD27[DIR_MMM * size_Mat];
+ D27.f[DIR_PMM] = &DD27[DIR_PMM * size_Mat];
+ D27.f[DIR_MPM] = &DD27[DIR_MPM * size_Mat];
}
else
{
- D27.f[DIR_M00] = &DD27[DIR_P00 *size_Mat];
- D27.f[DIR_P00] = &DD27[DIR_M00 *size_Mat];
- D27.f[DIR_0M0] = &DD27[DIR_0P0 *size_Mat];
- D27.f[DIR_0P0] = &DD27[DIR_0M0 *size_Mat];
- D27.f[DIR_00M] = &DD27[DIR_00P *size_Mat];
- D27.f[DIR_00P] = &DD27[DIR_00M *size_Mat];
- D27.f[DIR_MM0] = &DD27[DIR_PP0 *size_Mat];
- D27.f[DIR_PP0] = &DD27[DIR_MM0 *size_Mat];
- D27.f[DIR_MP0] = &DD27[DIR_PM0 *size_Mat];
- D27.f[DIR_PM0] = &DD27[DIR_MP0 *size_Mat];
- D27.f[DIR_M0M] = &DD27[DIR_P0P *size_Mat];
- D27.f[DIR_P0P] = &DD27[DIR_M0M *size_Mat];
- D27.f[DIR_M0P] = &DD27[DIR_P0M *size_Mat];
- D27.f[DIR_P0M] = &DD27[DIR_M0P *size_Mat];
- D27.f[DIR_0MM] = &DD27[DIR_0PP *size_Mat];
- D27.f[DIR_0PP] = &DD27[DIR_0MM *size_Mat];
- D27.f[DIR_0MP] = &DD27[DIR_0PM *size_Mat];
- D27.f[DIR_0PM] = &DD27[DIR_0MP *size_Mat];
- D27.f[DIR_000] = &DD27[DIR_000*size_Mat];
- D27.f[DIR_MMM] = &DD27[DIR_PPP *size_Mat];
- D27.f[DIR_PPM] = &DD27[DIR_MMP *size_Mat];
- D27.f[DIR_MPM] = &DD27[DIR_PMP *size_Mat];
- D27.f[DIR_PMM] = &DD27[DIR_MPP *size_Mat];
- D27.f[DIR_MMP] = &DD27[DIR_PPM *size_Mat];
- D27.f[DIR_PPP] = &DD27[DIR_MMM *size_Mat];
- D27.f[DIR_MPP] = &DD27[DIR_PMM *size_Mat];
- D27.f[DIR_PMP] = &DD27[DIR_MPM *size_Mat];
+ D27.f[DIR_M00] = &DD27[DIR_P00 * size_Mat];
+ D27.f[DIR_P00] = &DD27[DIR_M00 * size_Mat];
+ D27.f[DIR_0M0] = &DD27[DIR_0P0 * size_Mat];
+ D27.f[DIR_0P0] = &DD27[DIR_0M0 * size_Mat];
+ D27.f[DIR_00M] = &DD27[DIR_00P * size_Mat];
+ D27.f[DIR_00P] = &DD27[DIR_00M * size_Mat];
+ D27.f[DIR_MM0] = &DD27[DIR_PP0 * size_Mat];
+ D27.f[DIR_PP0] = &DD27[DIR_MM0 * size_Mat];
+ D27.f[DIR_MP0] = &DD27[DIR_PM0 * size_Mat];
+ D27.f[DIR_PM0] = &DD27[DIR_MP0 * size_Mat];
+ D27.f[DIR_M0M] = &DD27[DIR_P0P * size_Mat];
+ D27.f[DIR_P0P] = &DD27[DIR_M0M * size_Mat];
+ D27.f[DIR_M0P] = &DD27[DIR_P0M * size_Mat];
+ D27.f[DIR_P0M] = &DD27[DIR_M0P * size_Mat];
+ D27.f[DIR_0MM] = &DD27[DIR_0PP * size_Mat];
+ D27.f[DIR_0PP] = &DD27[DIR_0MM * size_Mat];
+ D27.f[DIR_0MP] = &DD27[DIR_0PM * size_Mat];
+ D27.f[DIR_0PM] = &DD27[DIR_0MP * size_Mat];
+ D27.f[DIR_000] = &DD27[DIR_000 * size_Mat];
+ D27.f[DIR_MMM] = &DD27[DIR_PPP * size_Mat];
+ D27.f[DIR_PPM] = &DD27[DIR_MMP * size_Mat];
+ D27.f[DIR_MPM] = &DD27[DIR_PMP * size_Mat];
+ D27.f[DIR_PMM] = &DD27[DIR_MPP * size_Mat];
+ D27.f[DIR_MMP] = &DD27[DIR_PPM * size_Mat];
+ D27.f[DIR_PPP] = &DD27[DIR_MMM * size_Mat];
+ D27.f[DIR_MPP] = &DD27[DIR_PMM * size_Mat];
+ D27.f[DIR_PMP] = &DD27[DIR_MPM * size_Mat];
}
////////////////////////////////////////////////////////////////////////////////
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/AdvectionDiffusion/Compressible/Mod7/ADComp7/ADComp7.cu b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/AdvectionDiffusion/Compressible/Mod7/ADComp7/ADComp7.cu
index ab9b0c444513455e0498d79614575e87c2afb6a0..3ee06a1e9ea77c8443d94f44ea54d11ffe7304ac 100644
--- a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/AdvectionDiffusion/Compressible/Mod7/ADComp7/ADComp7.cu
+++ b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/AdvectionDiffusion/Compressible/Mod7/ADComp7/ADComp7.cu
@@ -2,6 +2,7 @@
#include "ADComp7_Device.cuh"
#include "Parameter/Parameter.h"
+#include "cuda/CudaGrid.h"
std::shared_ptr<ADComp7> ADComp7::getNewInstance(std::shared_ptr<Parameter> para, int level)
{
@@ -10,34 +11,19 @@ std::shared_ptr<ADComp7> ADComp7::getNewInstance(std::shared_ptr<Parameter> para
void ADComp7::run()
{
- int size_Mat = para->getParD(level)->numberOfNodes;
- int numberOfThreads = para->getParD(level)->numberofthreads;
-
- int Grid = (size_Mat / numberOfThreads) + 1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid / Grid1) + 1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 grid(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1);
-
- LB_Kernel_AD_Comp_7 << < grid, threads >> >( para->getParD(level)->diffusivity,
- para->getParD(level)->typeOfGridNode,
- para->getParD(level)->neighborX,
- para->getParD(level)->neighborY,
- para->getParD(level)->neighborZ,
- para->getParD(level)->distributions.f[0],
- para->getParD(level)->distributionsAD7.f[0],
- para->getParD(level)->numberOfNodes,
- para->getParD(level)->isEvenTimestep);
- getLastCudaError("LB_Kernel_ThS7 execution failed");
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(para->getParD(level)->numberofthreads, para->getParD(level)->numberOfNodes);
+
+ LB_Kernel_AD_Comp_7<<< grid.grid, grid.threads >>>(
+ para->getParD(level)->diffusivity,
+ para->getParD(level)->typeOfGridNode,
+ para->getParD(level)->neighborX,
+ para->getParD(level)->neighborY,
+ para->getParD(level)->neighborZ,
+ para->getParD(level)->distributions.f[0],
+ para->getParD(level)->distributionsAD7.f[0],
+ para->getParD(level)->numberOfNodes,
+ para->getParD(level)->isEvenTimestep);
+ getLastCudaError("LB_Kernel_AD_Comp_7 execution failed");
}
ADComp7::ADComp7(std::shared_ptr<Parameter> para, int level)
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/AdvectionDiffusion/Compressible/Mod7/ADComp7/ADComp7_Device.cu b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/AdvectionDiffusion/Compressible/Mod7/ADComp7/ADComp7_Device.cu
index 52ab9ba6e968ec2293f0a1c4959323c43f328206..ddaed84703640cd9c7d12d142ccc1bf8f9ea7efc 100644
--- a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/AdvectionDiffusion/Compressible/Mod7/ADComp7/ADComp7_Device.cu
+++ b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/AdvectionDiffusion/Compressible/Mod7/ADComp7/ADComp7_Device.cu
@@ -39,63 +39,63 @@ __global__ void LB_Kernel_AD_Comp_7(real diffusivity,
Distributions27 D;
if (EvenOrOdd == true)
{
- D.f[DIR_P00] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_M00] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0P0] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0M0] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00P] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00M] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_PP0] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_MM0] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_PM0] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_MP0] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_P0P] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_M0M] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_P0M] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_M0P] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0PP] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0MM] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0PM] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0MP] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_PPP] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_MMP] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_PMP] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_MPP] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_PPM] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_MMM] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_PMM] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_MPM] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DDStart[DIR_P00 * size_Mat];
+ D.f[DIR_M00] = &DDStart[DIR_M00 * size_Mat];
+ D.f[DIR_0P0] = &DDStart[DIR_0P0 * size_Mat];
+ D.f[DIR_0M0] = &DDStart[DIR_0M0 * size_Mat];
+ D.f[DIR_00P] = &DDStart[DIR_00P * size_Mat];
+ D.f[DIR_00M] = &DDStart[DIR_00M * size_Mat];
+ D.f[DIR_PP0] = &DDStart[DIR_PP0 * size_Mat];
+ D.f[DIR_MM0] = &DDStart[DIR_MM0 * size_Mat];
+ D.f[DIR_PM0] = &DDStart[DIR_PM0 * size_Mat];
+ D.f[DIR_MP0] = &DDStart[DIR_MP0 * size_Mat];
+ D.f[DIR_P0P] = &DDStart[DIR_P0P * size_Mat];
+ D.f[DIR_M0M] = &DDStart[DIR_M0M * size_Mat];
+ D.f[DIR_P0M] = &DDStart[DIR_P0M * size_Mat];
+ D.f[DIR_M0P] = &DDStart[DIR_M0P * size_Mat];
+ D.f[DIR_0PP] = &DDStart[DIR_0PP * size_Mat];
+ D.f[DIR_0MM] = &DDStart[DIR_0MM * size_Mat];
+ D.f[DIR_0PM] = &DDStart[DIR_0PM * size_Mat];
+ D.f[DIR_0MP] = &DDStart[DIR_0MP * size_Mat];
+ D.f[DIR_000] = &DDStart[DIR_000 * size_Mat];
+ D.f[DIR_PPP] = &DDStart[DIR_PPP * size_Mat];
+ D.f[DIR_MMP] = &DDStart[DIR_MMP * size_Mat];
+ D.f[DIR_PMP] = &DDStart[DIR_PMP * size_Mat];
+ D.f[DIR_MPP] = &DDStart[DIR_MPP * size_Mat];
+ D.f[DIR_PPM] = &DDStart[DIR_PPM * size_Mat];
+ D.f[DIR_MMM] = &DDStart[DIR_MMM * size_Mat];
+ D.f[DIR_PMM] = &DDStart[DIR_PMM * size_Mat];
+ D.f[DIR_MPM] = &DDStart[DIR_MPM * size_Mat];
}
else
{
- D.f[DIR_M00] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_P00] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0M0] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0P0] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00M] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00P] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_MM0] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_PP0] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_MP0] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_PM0] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_M0M] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_P0P] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_M0P] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_P0M] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0MM] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0PP] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0MP] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0PM] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_MMM] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_PPM] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_MPM] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_PMM] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_MMP] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_PPP] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_MPP] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_PMP] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_M00] = &DDStart[DIR_P00 * size_Mat];
+ D.f[DIR_P00] = &DDStart[DIR_M00 * size_Mat];
+ D.f[DIR_0M0] = &DDStart[DIR_0P0 * size_Mat];
+ D.f[DIR_0P0] = &DDStart[DIR_0M0 * size_Mat];
+ D.f[DIR_00M] = &DDStart[DIR_00P * size_Mat];
+ D.f[DIR_00P] = &DDStart[DIR_00M * size_Mat];
+ D.f[DIR_MM0] = &DDStart[DIR_PP0 * size_Mat];
+ D.f[DIR_PP0] = &DDStart[DIR_MM0 * size_Mat];
+ D.f[DIR_MP0] = &DDStart[DIR_PM0 * size_Mat];
+ D.f[DIR_PM0] = &DDStart[DIR_MP0 * size_Mat];
+ D.f[DIR_M0M] = &DDStart[DIR_P0P * size_Mat];
+ D.f[DIR_P0P] = &DDStart[DIR_M0M * size_Mat];
+ D.f[DIR_M0P] = &DDStart[DIR_P0M * size_Mat];
+ D.f[DIR_P0M] = &DDStart[DIR_M0P * size_Mat];
+ D.f[DIR_0MM] = &DDStart[DIR_0PP * size_Mat];
+ D.f[DIR_0PP] = &DDStart[DIR_0MM * size_Mat];
+ D.f[DIR_0MP] = &DDStart[DIR_0PM * size_Mat];
+ D.f[DIR_0PM] = &DDStart[DIR_0MP * size_Mat];
+ D.f[DIR_000] = &DDStart[DIR_000 * size_Mat];
+ D.f[DIR_MMM] = &DDStart[DIR_PPP * size_Mat];
+ D.f[DIR_PPM] = &DDStart[DIR_MMP * size_Mat];
+ D.f[DIR_MPM] = &DDStart[DIR_PMP * size_Mat];
+ D.f[DIR_PMM] = &DDStart[DIR_MPP * size_Mat];
+ D.f[DIR_MMP] = &DDStart[DIR_PPM * size_Mat];
+ D.f[DIR_PPP] = &DDStart[DIR_MMM * size_Mat];
+ D.f[DIR_MPP] = &DDStart[DIR_PMM * size_Mat];
+ D.f[DIR_PMP] = &DDStart[DIR_MPM * size_Mat];
}
Distributions7 D7;
@@ -157,33 +157,33 @@ __global__ void LB_Kernel_AD_Comp_7(real diffusivity,
real fTNE = (D.f[DIR_MMM])[kbsw];
real fTNW = (D.f[DIR_PMM])[kbs];//kbse
real fTSE = (D.f[DIR_MPM])[kbw];//kbnw
- //real fE = (D.f[DIR_P00 ])[k ];//ke
- //real fW = (D.f[DIR_M00 ])[kw ];
- //real fN = (D.f[DIR_0P0 ])[k ];//kn
- //real fS = (D.f[DIR_0M0 ])[ks ];
- //real fT = (D.f[DIR_00P ])[k ];//kt
- //real fB = (D.f[DIR_00M ])[kb ];
- //real fNE = (D.f[DIR_PP0 ])[k ];//kne
- //real fSW = (D.f[DIR_MM0 ])[ksw];
- //real fSE = (D.f[DIR_PM0 ])[ks ];//kse
- //real fNW = (D.f[DIR_MP0 ])[kw ];//knw
- //real fTE = (D.f[DIR_P0P ])[k ];//kte
- //real fBW = (D.f[DIR_M0M ])[kbw];
- //real fBE = (D.f[DIR_P0M ])[kb ];//kbe
- //real fTW = (D.f[DIR_M0P ])[kw ];//ktw
- //real fTN = (D.f[DIR_0PP ])[k ];//ktn
- //real fBS = (D.f[DIR_0MM ])[kbs];
- //real fBN = (D.f[DIR_0PM ])[kb ];//kbn
- //real fTS = (D.f[DIR_0MP ])[ks ];//kts
+ //real fE = (D.f[DIR_P00])[k ];//ke
+ //real fW = (D.f[DIR_M00])[kw ];
+ //real fN = (D.f[DIR_0P0])[k ];//kn
+ //real fS = (D.f[DIR_0M0])[ks ];
+ //real fT = (D.f[DIR_00P])[k ];//kt
+ //real fB = (D.f[DIR_00M])[kb ];
+ //real fNE = (D.f[DIR_PP0])[k ];//kne
+ //real fSW = (D.f[DIR_MM0])[ksw];
+ //real fSE = (D.f[DIR_PM0])[ks ];//kse
+ //real fNW = (D.f[DIR_MP0])[kw ];//knw
+ //real fTE = (D.f[DIR_P0P])[k ];//kte
+ //real fBW = (D.f[DIR_M0M])[kbw];
+ //real fBE = (D.f[DIR_P0M])[kb ];//kbe
+ //real fTW = (D.f[DIR_M0P])[kw ];//ktw
+ //real fTN = (D.f[DIR_0PP])[k ];//ktn
+ //real fBS = (D.f[DIR_0MM])[kbs];
+ //real fBN = (D.f[DIR_0PM])[kb ];//kbn
+ //real fTS = (D.f[DIR_0MP])[ks ];//kts
//real fZERO = (D.f[DIR_000])[k ];//kzero
- //real fTNE = (D.f[DIR_PPP ])[k ];//ktne
- //real fTSW = (D.f[DIR_MMP ])[ksw];//ktsw
- //real fTSE = (D.f[DIR_PMP ])[ks ];//ktse
- //real fTNW = (D.f[DIR_MPP ])[kw ];//ktnw
- //real fBNE = (D.f[DIR_PPM ])[kb ];//kbne
- //real fBSW = (D.f[DIR_MMM ])[kbsw];
- //real fBSE = (D.f[DIR_PMM ])[kbs];//kbse
- //real fBNW = (D.f[DIR_MPM ])[kbw];//kbnw
+ //real fTNE = (D.f[DIR_PPP])[k ];//ktne
+ //real fTSW = (D.f[DIR_MMP])[ksw];//ktsw
+ //real fTSE = (D.f[DIR_PMP])[ks ];//ktse
+ //real fTNW = (D.f[DIR_MPP])[kw ];//ktnw
+ //real fBNE = (D.f[DIR_PPM])[kb ];//kbne
+ //real fBSW = (D.f[DIR_MMM])[kbsw];
+ //real fBSE = (D.f[DIR_PMM])[kbs];//kbse
+ //real fBNW = (D.f[DIR_MPM])[kbw];//kbnw
////////////////////////////////////////////////////////////////////////////////
real f7ZERO = (D7.f[0])[k];
real f7E = (D7.f[1])[k];
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/AdvectionDiffusion/Incompressible/Mod27/ADIncomp27/ADIncomp27.cu b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/AdvectionDiffusion/Incompressible/Mod27/ADIncomp27/ADIncomp27.cu
index 4ad8a4678ae2e4025a90f639ae366311a247e4b3..f2a9feaa998b628fb782844d1a7d946317e5af5f 100644
--- a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/AdvectionDiffusion/Incompressible/Mod27/ADIncomp27/ADIncomp27.cu
+++ b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/AdvectionDiffusion/Incompressible/Mod27/ADIncomp27/ADIncomp27.cu
@@ -2,6 +2,7 @@
#include "ADIncomp27_Device.cuh"
#include "Parameter/Parameter.h"
+#include "cuda/CudaGrid.h"
std::shared_ptr<ADIncomp27> ADIncomp27::getNewInstance(std::shared_ptr<Parameter> para, int level)
{
@@ -10,34 +11,19 @@ std::shared_ptr<ADIncomp27> ADIncomp27::getNewInstance(std::shared_ptr<Parameter
void ADIncomp27::run()
{
- int size_Mat = para->getParD(level)->numberOfNodes;
- int numberOfThreads = para->getParD(level)->numberofthreads;
-
- int Grid = (size_Mat / numberOfThreads) + 1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid / Grid1) + 1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 grid(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1);
-
- LB_Kernel_AD_Incomp_27 << < grid, threads >> >( para->getParD(level)->diffusivity,
- para->getParD(level)->typeOfGridNode,
- para->getParD(level)->neighborX,
- para->getParD(level)->neighborY,
- para->getParD(level)->neighborZ,
- para->getParD(level)->distributions.f[0],
- para->getParD(level)->distributionsAD27.f[0],
- para->getParD(level)->numberOfNodes,
- para->getParD(level)->isEvenTimestep);
- getLastCudaError("LB_Kernel_AD_Incomp_27 execution failed");
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(para->getParD(level)->numberofthreads, para->getParD(level)->numberOfNodes);
+
+ LB_Kernel_AD_Incomp_27<<< grid.grid, grid.threads >>>(
+ para->getParD(level)->diffusivity,
+ para->getParD(level)->typeOfGridNode,
+ para->getParD(level)->neighborX,
+ para->getParD(level)->neighborY,
+ para->getParD(level)->neighborZ,
+ para->getParD(level)->distributions.f[0],
+ para->getParD(level)->distributionsAD27.f[0],
+ para->getParD(level)->numberOfNodes,
+ para->getParD(level)->isEvenTimestep);
+ getLastCudaError("LB_Kernel_AD_Incomp_27 execution failed");
}
ADIncomp27::ADIncomp27(std::shared_ptr<Parameter> para, int level)
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/AdvectionDiffusion/Incompressible/Mod27/ADIncomp27/ADIncomp27_Device.cu b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/AdvectionDiffusion/Incompressible/Mod27/ADIncomp27/ADIncomp27_Device.cu
index e686825ed100417110b02360876dec076553d7de..f9fdcee0f34106b05da0edc16e3fdd89f859752e 100644
--- a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/AdvectionDiffusion/Incompressible/Mod27/ADIncomp27/ADIncomp27_Device.cu
+++ b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/AdvectionDiffusion/Incompressible/Mod27/ADIncomp27/ADIncomp27_Device.cu
@@ -13,7 +13,7 @@ __global__ void LB_Kernel_AD_Incomp_27(real diffusivity,
unsigned int* neighborZ,
real* DDStart,
real* DD27,
- int size_Mat,
+ unsigned long long numberOfLBnodes,
bool EvenOrOdd)
{
////////////////////////////////////////////////////////////////////////////////
@@ -27,7 +27,7 @@ __global__ void LB_Kernel_AD_Incomp_27(real diffusivity,
const unsigned k = nx*(ny*z + y) + x;
//////////////////////////////////////////////////////////////////////////
- if (k<size_Mat)
+ if (k<numberOfLBnodes)
{
////////////////////////////////////////////////////////////////////////////////
unsigned int BC;
@@ -38,125 +38,125 @@ __global__ void LB_Kernel_AD_Incomp_27(real diffusivity,
Distributions27 D;
if (EvenOrOdd == true)
{
- D.f[DIR_P00] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_M00] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0P0] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0M0] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00P] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00M] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_PP0] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_MM0] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_PM0] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_MP0] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_P0P] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_M0M] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_P0M] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_M0P] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0PP] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0MM] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0PM] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0MP] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_PPP] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_MMP] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_PMP] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_MPP] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_PPM] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_MMM] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_PMM] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_MPM] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DDStart[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DDStart[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DDStart[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DDStart[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DDStart[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DDStart[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DDStart[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DDStart[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DDStart[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DDStart[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DDStart[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DDStart[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DDStart[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DDStart[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DDStart[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DDStart[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DDStart[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DDStart[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DDStart[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DDStart[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DDStart[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DDStart[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DDStart[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DDStart[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DDStart[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DDStart[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DDStart[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_P00] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0M0] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0P0] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00M] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00P] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_MM0] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_PP0] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_MP0] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_PM0] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_M0M] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_P0P] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_M0P] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_P0M] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0MM] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0PP] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0MP] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0PM] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_MMM] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_PPM] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_MPM] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_PMM] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_MMP] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_PPP] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_MPP] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_PMP] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_M00] = &DDStart[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DDStart[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DDStart[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DDStart[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DDStart[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DDStart[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DDStart[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DDStart[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DDStart[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DDStart[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DDStart[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DDStart[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DDStart[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DDStart[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DDStart[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DDStart[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DDStart[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DDStart[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DDStart[DIR_000 * numberOfLBnodes];
+ D.f[DIR_MMM] = &DDStart[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DDStart[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DDStart[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DDStart[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DDStart[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PPP] = &DDStart[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DDStart[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DDStart[DIR_MPM * numberOfLBnodes];
}
Distributions27 D27;
if (EvenOrOdd == true)
{
- D27.f[DIR_P00] = &DD27[DIR_P00 *size_Mat];
- D27.f[DIR_M00] = &DD27[DIR_M00 *size_Mat];
- D27.f[DIR_0P0] = &DD27[DIR_0P0 *size_Mat];
- D27.f[DIR_0M0] = &DD27[DIR_0M0 *size_Mat];
- D27.f[DIR_00P] = &DD27[DIR_00P *size_Mat];
- D27.f[DIR_00M] = &DD27[DIR_00M *size_Mat];
- D27.f[DIR_PP0] = &DD27[DIR_PP0 *size_Mat];
- D27.f[DIR_MM0] = &DD27[DIR_MM0 *size_Mat];
- D27.f[DIR_PM0] = &DD27[DIR_PM0 *size_Mat];
- D27.f[DIR_MP0] = &DD27[DIR_MP0 *size_Mat];
- D27.f[DIR_P0P] = &DD27[DIR_P0P *size_Mat];
- D27.f[DIR_M0M] = &DD27[DIR_M0M *size_Mat];
- D27.f[DIR_P0M] = &DD27[DIR_P0M *size_Mat];
- D27.f[DIR_M0P] = &DD27[DIR_M0P *size_Mat];
- D27.f[DIR_0PP] = &DD27[DIR_0PP *size_Mat];
- D27.f[DIR_0MM] = &DD27[DIR_0MM *size_Mat];
- D27.f[DIR_0PM] = &DD27[DIR_0PM *size_Mat];
- D27.f[DIR_0MP] = &DD27[DIR_0MP *size_Mat];
- D27.f[DIR_000] = &DD27[DIR_000*size_Mat];
- D27.f[DIR_PPP] = &DD27[DIR_PPP *size_Mat];
- D27.f[DIR_MMP] = &DD27[DIR_MMP *size_Mat];
- D27.f[DIR_PMP] = &DD27[DIR_PMP *size_Mat];
- D27.f[DIR_MPP] = &DD27[DIR_MPP *size_Mat];
- D27.f[DIR_PPM] = &DD27[DIR_PPM *size_Mat];
- D27.f[DIR_MMM] = &DD27[DIR_MMM *size_Mat];
- D27.f[DIR_PMM] = &DD27[DIR_PMM *size_Mat];
- D27.f[DIR_MPM] = &DD27[DIR_MPM *size_Mat];
+ D27.f[DIR_P00] = &DD27[DIR_P00 * numberOfLBnodes];
+ D27.f[DIR_M00] = &DD27[DIR_M00 * numberOfLBnodes];
+ D27.f[DIR_0P0] = &DD27[DIR_0P0 * numberOfLBnodes];
+ D27.f[DIR_0M0] = &DD27[DIR_0M0 * numberOfLBnodes];
+ D27.f[DIR_00P] = &DD27[DIR_00P * numberOfLBnodes];
+ D27.f[DIR_00M] = &DD27[DIR_00M * numberOfLBnodes];
+ D27.f[DIR_PP0] = &DD27[DIR_PP0 * numberOfLBnodes];
+ D27.f[DIR_MM0] = &DD27[DIR_MM0 * numberOfLBnodes];
+ D27.f[DIR_PM0] = &DD27[DIR_PM0 * numberOfLBnodes];
+ D27.f[DIR_MP0] = &DD27[DIR_MP0 * numberOfLBnodes];
+ D27.f[DIR_P0P] = &DD27[DIR_P0P * numberOfLBnodes];
+ D27.f[DIR_M0M] = &DD27[DIR_M0M * numberOfLBnodes];
+ D27.f[DIR_P0M] = &DD27[DIR_P0M * numberOfLBnodes];
+ D27.f[DIR_M0P] = &DD27[DIR_M0P * numberOfLBnodes];
+ D27.f[DIR_0PP] = &DD27[DIR_0PP * numberOfLBnodes];
+ D27.f[DIR_0MM] = &DD27[DIR_0MM * numberOfLBnodes];
+ D27.f[DIR_0PM] = &DD27[DIR_0PM * numberOfLBnodes];
+ D27.f[DIR_0MP] = &DD27[DIR_0MP * numberOfLBnodes];
+ D27.f[DIR_000] = &DD27[DIR_000 * numberOfLBnodes];
+ D27.f[DIR_PPP] = &DD27[DIR_PPP * numberOfLBnodes];
+ D27.f[DIR_MMP] = &DD27[DIR_MMP * numberOfLBnodes];
+ D27.f[DIR_PMP] = &DD27[DIR_PMP * numberOfLBnodes];
+ D27.f[DIR_MPP] = &DD27[DIR_MPP * numberOfLBnodes];
+ D27.f[DIR_PPM] = &DD27[DIR_PPM * numberOfLBnodes];
+ D27.f[DIR_MMM] = &DD27[DIR_MMM * numberOfLBnodes];
+ D27.f[DIR_PMM] = &DD27[DIR_PMM * numberOfLBnodes];
+ D27.f[DIR_MPM] = &DD27[DIR_MPM * numberOfLBnodes];
}
else
{
- D27.f[DIR_M00] = &DD27[DIR_P00 *size_Mat];
- D27.f[DIR_P00] = &DD27[DIR_M00 *size_Mat];
- D27.f[DIR_0M0] = &DD27[DIR_0P0 *size_Mat];
- D27.f[DIR_0P0] = &DD27[DIR_0M0 *size_Mat];
- D27.f[DIR_00M] = &DD27[DIR_00P *size_Mat];
- D27.f[DIR_00P] = &DD27[DIR_00M *size_Mat];
- D27.f[DIR_MM0] = &DD27[DIR_PP0 *size_Mat];
- D27.f[DIR_PP0] = &DD27[DIR_MM0 *size_Mat];
- D27.f[DIR_MP0] = &DD27[DIR_PM0 *size_Mat];
- D27.f[DIR_PM0] = &DD27[DIR_MP0 *size_Mat];
- D27.f[DIR_M0M] = &DD27[DIR_P0P *size_Mat];
- D27.f[DIR_P0P] = &DD27[DIR_M0M *size_Mat];
- D27.f[DIR_M0P] = &DD27[DIR_P0M *size_Mat];
- D27.f[DIR_P0M] = &DD27[DIR_M0P *size_Mat];
- D27.f[DIR_0MM] = &DD27[DIR_0PP *size_Mat];
- D27.f[DIR_0PP] = &DD27[DIR_0MM *size_Mat];
- D27.f[DIR_0MP] = &DD27[DIR_0PM *size_Mat];
- D27.f[DIR_0PM] = &DD27[DIR_0MP *size_Mat];
- D27.f[DIR_000] = &DD27[DIR_000*size_Mat];
- D27.f[DIR_MMM] = &DD27[DIR_PPP *size_Mat];
- D27.f[DIR_PPM] = &DD27[DIR_MMP *size_Mat];
- D27.f[DIR_MPM] = &DD27[DIR_PMP *size_Mat];
- D27.f[DIR_PMM] = &DD27[DIR_MPP *size_Mat];
- D27.f[DIR_MMP] = &DD27[DIR_PPM *size_Mat];
- D27.f[DIR_PPP] = &DD27[DIR_MMM *size_Mat];
- D27.f[DIR_MPP] = &DD27[DIR_PMM *size_Mat];
- D27.f[DIR_PMP] = &DD27[DIR_MPM *size_Mat];
+ D27.f[DIR_M00] = &DD27[DIR_P00 * numberOfLBnodes];
+ D27.f[DIR_P00] = &DD27[DIR_M00 * numberOfLBnodes];
+ D27.f[DIR_0M0] = &DD27[DIR_0P0 * numberOfLBnodes];
+ D27.f[DIR_0P0] = &DD27[DIR_0M0 * numberOfLBnodes];
+ D27.f[DIR_00M] = &DD27[DIR_00P * numberOfLBnodes];
+ D27.f[DIR_00P] = &DD27[DIR_00M * numberOfLBnodes];
+ D27.f[DIR_MM0] = &DD27[DIR_PP0 * numberOfLBnodes];
+ D27.f[DIR_PP0] = &DD27[DIR_MM0 * numberOfLBnodes];
+ D27.f[DIR_MP0] = &DD27[DIR_PM0 * numberOfLBnodes];
+ D27.f[DIR_PM0] = &DD27[DIR_MP0 * numberOfLBnodes];
+ D27.f[DIR_M0M] = &DD27[DIR_P0P * numberOfLBnodes];
+ D27.f[DIR_P0P] = &DD27[DIR_M0M * numberOfLBnodes];
+ D27.f[DIR_M0P] = &DD27[DIR_P0M * numberOfLBnodes];
+ D27.f[DIR_P0M] = &DD27[DIR_M0P * numberOfLBnodes];
+ D27.f[DIR_0MM] = &DD27[DIR_0PP * numberOfLBnodes];
+ D27.f[DIR_0PP] = &DD27[DIR_0MM * numberOfLBnodes];
+ D27.f[DIR_0MP] = &DD27[DIR_0PM * numberOfLBnodes];
+ D27.f[DIR_0PM] = &DD27[DIR_0MP * numberOfLBnodes];
+ D27.f[DIR_000] = &DD27[DIR_000 * numberOfLBnodes];
+ D27.f[DIR_MMM] = &DD27[DIR_PPP * numberOfLBnodes];
+ D27.f[DIR_PPM] = &DD27[DIR_MMP * numberOfLBnodes];
+ D27.f[DIR_MPM] = &DD27[DIR_PMP * numberOfLBnodes];
+ D27.f[DIR_PMM] = &DD27[DIR_MPP * numberOfLBnodes];
+ D27.f[DIR_MMP] = &DD27[DIR_PPM * numberOfLBnodes];
+ D27.f[DIR_PPP] = &DD27[DIR_MMM * numberOfLBnodes];
+ D27.f[DIR_MPP] = &DD27[DIR_PMM * numberOfLBnodes];
+ D27.f[DIR_PMP] = &DD27[DIR_MPM * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////
@@ -197,33 +197,33 @@ __global__ void LB_Kernel_AD_Incomp_27(real diffusivity,
real fTNW = (D.f[DIR_PMM])[kbs];//kbse
real fTSE = (D.f[DIR_MPM])[kbw];//kbnw
////////////////////////////////////////////////////////////////////////////////
- //real f27E = (D27.f[DIR_P00 ])[k ];//ke
- //real f27W = (D27.f[DIR_M00 ])[kw ];
- //real f27N = (D27.f[DIR_0P0 ])[k ];//kn
- //real f27S = (D27.f[DIR_0M0 ])[ks ];
- //real f27T = (D27.f[DIR_00P ])[k ];//kt
- //real f27B = (D27.f[DIR_00M ])[kb ];
- //real f27NE = (D27.f[DIR_PP0 ])[k ];//kne
- //real f27SW = (D27.f[DIR_MM0 ])[ksw];
- //real f27SE = (D27.f[DIR_PM0 ])[ks ];//kse
- //real f27NW = (D27.f[DIR_MP0 ])[kw ];//knw
- //real f27TE = (D27.f[DIR_P0P ])[k ];//kte
- //real f27BW = (D27.f[DIR_M0M ])[kbw];
- //real f27BE = (D27.f[DIR_P0M ])[kb ];//kbe
- //real f27TW = (D27.f[DIR_M0P ])[kw ];//ktw
- //real f27TN = (D27.f[DIR_0PP ])[k ];//ktn
- //real f27BS = (D27.f[DIR_0MM ])[kbs];
- //real f27BN = (D27.f[DIR_0PM ])[kb ];//kbn
- //real f27TS = (D27.f[DIR_0MP ])[ks ];//kts
+ //real f27E = (D27.f[DIR_P00])[k ];//ke
+ //real f27W = (D27.f[DIR_M00])[kw ];
+ //real f27N = (D27.f[DIR_0P0])[k ];//kn
+ //real f27S = (D27.f[DIR_0M0])[ks ];
+ //real f27T = (D27.f[DIR_00P])[k ];//kt
+ //real f27B = (D27.f[DIR_00M])[kb ];
+ //real f27NE = (D27.f[DIR_PP0])[k ];//kne
+ //real f27SW = (D27.f[DIR_MM0])[ksw];
+ //real f27SE = (D27.f[DIR_PM0])[ks ];//kse
+ //real f27NW = (D27.f[DIR_MP0])[kw ];//knw
+ //real f27TE = (D27.f[DIR_P0P])[k ];//kte
+ //real f27BW = (D27.f[DIR_M0M])[kbw];
+ //real f27BE = (D27.f[DIR_P0M])[kb ];//kbe
+ //real f27TW = (D27.f[DIR_M0P])[kw ];//ktw
+ //real f27TN = (D27.f[DIR_0PP])[k ];//ktn
+ //real f27BS = (D27.f[DIR_0MM])[kbs];
+ //real f27BN = (D27.f[DIR_0PM])[kb ];//kbn
+ //real f27TS = (D27.f[DIR_0MP])[ks ];//kts
//real f27ZERO = (D27.f[DIR_000])[k ];//kzero
- //real f27TNE = (D27.f[DIR_PPP ])[k ];//ktne
- //real f27TSW = (D27.f[DIR_MMP ])[ksw];//ktsw
- //real f27TSE = (D27.f[DIR_PMP ])[ks ];//ktse
- //real f27TNW = (D27.f[DIR_MPP ])[kw ];//ktnw
- //real f27BNE = (D27.f[DIR_PPM ])[kb ];//kbne
- //real f27BSW = (D27.f[DIR_MMM ])[kbsw];
- //real f27BSE = (D27.f[DIR_PMM ])[kbs];//kbse
- //real f27BNW = (D27.f[DIR_MPM ])[kbw];//kbnw
+ //real f27TNE = (D27.f[DIR_PPP])[k ];//ktne
+ //real f27TSW = (D27.f[DIR_MMP])[ksw];//ktsw
+ //real f27TSE = (D27.f[DIR_PMP])[ks ];//ktse
+ //real f27TNW = (D27.f[DIR_MPP])[kw ];//ktnw
+ //real f27BNE = (D27.f[DIR_PPM])[kb ];//kbne
+ //real f27BSW = (D27.f[DIR_MMM])[kbsw];
+ //real f27BSE = (D27.f[DIR_PMM])[kbs];//kbse
+ //real f27BNW = (D27.f[DIR_MPM])[kbw];//kbnw
////////////////////////////////////////////////////////////////////////////////
//real vx1 = ((fTNE-fBSW)+(fBNE-fTSW)+(fTSE-fBNW)+(fBSE-fTNW) +(fNE-fSW)+(fSE-fNW)+(fTE-fBW)+(fBE-fTW)+(fE-fW));
//real vx2 = ((fTNE-fBSW)+(fBNE-fTSW)+(fBNW-fTSE)+(fTNW-fBSE) +(fNE-fSW)+(fNW-fSE)+(fTN-fBS)+(fBN-fTS)+(fN-fS));
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/AdvectionDiffusion/Incompressible/Mod27/ADIncomp27/ADIncomp27_Device.cuh b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/AdvectionDiffusion/Incompressible/Mod27/ADIncomp27/ADIncomp27_Device.cuh
index a6d94de4fadb9a93a9e5fed63d87731b12ec2a07..3abee563f676910f422bba0930060c2a0b0c0e21 100644
--- a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/AdvectionDiffusion/Incompressible/Mod27/ADIncomp27/ADIncomp27_Device.cuh
+++ b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/AdvectionDiffusion/Incompressible/Mod27/ADIncomp27/ADIncomp27_Device.cuh
@@ -11,7 +11,7 @@ __global__ void LB_Kernel_AD_Incomp_27(real diffusivity,
unsigned int* neighborZ,
real* DDStart,
real* DD27,
- int size_Mat,
+ unsigned long long numberOfLBnodes,
bool EvenOrOdd);
#endif
\ No newline at end of file
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/AdvectionDiffusion/Incompressible/Mod7/ADIncomp7/ADIncomp7.cu b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/AdvectionDiffusion/Incompressible/Mod7/ADIncomp7/ADIncomp7.cu
index 27da776eb7612307fa4f9af2886594fc0c75d90b..d0c6a6a24ab4d0ebebee9324bdafa1f9e3db51b9 100644
--- a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/AdvectionDiffusion/Incompressible/Mod7/ADIncomp7/ADIncomp7.cu
+++ b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/AdvectionDiffusion/Incompressible/Mod7/ADIncomp7/ADIncomp7.cu
@@ -2,6 +2,7 @@
#include "ADIncomp7_Device.cuh"
#include "Parameter/Parameter.h"
+#include "cuda/CudaGrid.h"
std::shared_ptr<ADIncomp7> ADIncomp7::getNewInstance(std::shared_ptr<Parameter> para, int level)
{
@@ -10,34 +11,19 @@ std::shared_ptr<ADIncomp7> ADIncomp7::getNewInstance(std::shared_ptr<Parameter>
void ADIncomp7::run()
{
- int size_Mat = para->getParD(level)->numberOfNodes;
- int numberOfThreads = para->getParD(level)->numberofthreads;
-
- int Grid = (size_Mat / numberOfThreads) + 1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid / Grid1) + 1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 grid(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1);
-
- LB_Kernel_AD_Incomp_7 << < grid, threads >> >( para->getParD(level)->diffusivity,
- para->getParD(level)->typeOfGridNode,
- para->getParD(level)->neighborX,
- para->getParD(level)->neighborY,
- para->getParD(level)->neighborZ,
- para->getParD(level)->distributions.f[0],
- para->getParD(level)->distributionsAD7.f[0],
- para->getParD(level)->numberOfNodes,
- para->getParD(level)->isEvenTimestep);
- getLastCudaError("LB_Kernel_AD_Incomp_7 execution failed");
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(para->getParD(level)->numberofthreads, para->getParD(level)->numberOfNodes);
+
+ LB_Kernel_AD_Incomp_7<<< grid.grid, grid.threads >>>(
+ para->getParD(level)->diffusivity,
+ para->getParD(level)->typeOfGridNode,
+ para->getParD(level)->neighborX,
+ para->getParD(level)->neighborY,
+ para->getParD(level)->neighborZ,
+ para->getParD(level)->distributions.f[0],
+ para->getParD(level)->distributionsAD7.f[0],
+ para->getParD(level)->numberOfNodes,
+ para->getParD(level)->isEvenTimestep);
+ getLastCudaError("LB_Kernel_AD_Incomp_7 execution failed");
}
ADIncomp7::ADIncomp7(std::shared_ptr<Parameter> para, int level)
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/AdvectionDiffusion/Incompressible/Mod7/ADIncomp7/ADIncomp7_Device.cu b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/AdvectionDiffusion/Incompressible/Mod7/ADIncomp7/ADIncomp7_Device.cu
index d49b0b48d20d976076a52f804d485b68da55348e..e0bcc4e515b1b2ccf71f1050e2d572b60a40d94b 100644
--- a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/AdvectionDiffusion/Incompressible/Mod7/ADIncomp7/ADIncomp7_Device.cu
+++ b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/AdvectionDiffusion/Incompressible/Mod7/ADIncomp7/ADIncomp7_Device.cu
@@ -13,7 +13,7 @@ __global__ void LB_Kernel_AD_Incomp_7(real diffusivity,
unsigned int* neighborZ,
real* DDStart,
real* DD7,
- int size_Mat,
+ unsigned long long numberOfLBnodes,
bool EvenOrOdd)
{
////////////////////////////////////////////////////////////////////////////////
@@ -27,7 +27,7 @@ __global__ void LB_Kernel_AD_Incomp_7(real diffusivity,
const unsigned k = nx*(ny*z + y) + x;
//////////////////////////////////////////////////////////////////////////
- if (k<size_Mat)
+ if (k<numberOfLBnodes)
{
////////////////////////////////////////////////////////////////////////////////
unsigned int BC;
@@ -38,85 +38,85 @@ __global__ void LB_Kernel_AD_Incomp_7(real diffusivity,
Distributions27 D;
if (EvenOrOdd == true)
{
- D.f[DIR_P00] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_M00] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0P0] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0M0] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00P] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00M] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_PP0] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_MM0] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_PM0] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_MP0] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_P0P] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_M0M] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_P0M] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_M0P] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0PP] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0MM] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0PM] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0MP] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_PPP] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_MMP] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_PMP] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_MPP] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_PPM] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_MMM] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_PMM] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_MPM] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DDStart[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DDStart[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DDStart[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DDStart[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DDStart[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DDStart[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DDStart[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DDStart[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DDStart[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DDStart[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DDStart[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DDStart[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DDStart[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DDStart[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DDStart[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DDStart[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DDStart[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DDStart[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DDStart[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DDStart[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DDStart[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DDStart[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DDStart[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DDStart[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DDStart[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DDStart[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DDStart[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_P00] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0M0] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0P0] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00M] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00P] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_MM0] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_PP0] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_MP0] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_PM0] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_M0M] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_P0P] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_M0P] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_P0M] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0MM] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0PP] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0MP] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0PM] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_MMM] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_PPM] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_MPM] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_PMM] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_MMP] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_PPP] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_MPP] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_PMP] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_M00] = &DDStart[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DDStart[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DDStart[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DDStart[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DDStart[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DDStart[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DDStart[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DDStart[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DDStart[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DDStart[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DDStart[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DDStart[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DDStart[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DDStart[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DDStart[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DDStart[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DDStart[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DDStart[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DDStart[DIR_000 * numberOfLBnodes];
+ D.f[DIR_MMM] = &DDStart[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DDStart[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DDStart[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DDStart[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DDStart[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PPP] = &DDStart[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DDStart[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DDStart[DIR_MPM * numberOfLBnodes];
}
Distributions7 D7;
if (EvenOrOdd == true)
{
- D7.f[0] = &DD7[0 * size_Mat];
- D7.f[1] = &DD7[1 * size_Mat];
- D7.f[2] = &DD7[2 * size_Mat];
- D7.f[3] = &DD7[3 * size_Mat];
- D7.f[4] = &DD7[4 * size_Mat];
- D7.f[5] = &DD7[5 * size_Mat];
- D7.f[6] = &DD7[6 * size_Mat];
+ D7.f[0] = &DD7[0 * numberOfLBnodes];
+ D7.f[1] = &DD7[1 * numberOfLBnodes];
+ D7.f[2] = &DD7[2 * numberOfLBnodes];
+ D7.f[3] = &DD7[3 * numberOfLBnodes];
+ D7.f[4] = &DD7[4 * numberOfLBnodes];
+ D7.f[5] = &DD7[5 * numberOfLBnodes];
+ D7.f[6] = &DD7[6 * numberOfLBnodes];
}
else
{
- D7.f[0] = &DD7[0 * size_Mat];
- D7.f[2] = &DD7[1 * size_Mat];
- D7.f[1] = &DD7[2 * size_Mat];
- D7.f[4] = &DD7[3 * size_Mat];
- D7.f[3] = &DD7[4 * size_Mat];
- D7.f[6] = &DD7[5 * size_Mat];
- D7.f[5] = &DD7[6 * size_Mat];
+ D7.f[0] = &DD7[0 * numberOfLBnodes];
+ D7.f[2] = &DD7[1 * numberOfLBnodes];
+ D7.f[1] = &DD7[2 * numberOfLBnodes];
+ D7.f[4] = &DD7[3 * numberOfLBnodes];
+ D7.f[3] = &DD7[4 * numberOfLBnodes];
+ D7.f[6] = &DD7[5 * numberOfLBnodes];
+ D7.f[5] = &DD7[6 * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////
@@ -156,33 +156,33 @@ __global__ void LB_Kernel_AD_Incomp_7(real diffusivity,
real fTNE = (D.f[DIR_MMM])[kbsw];
real fTNW = (D.f[DIR_PMM])[kbs];//kbse
real fTSE = (D.f[DIR_MPM])[kbw];//kbnw
- //real fE = (D.f[DIR_P00 ])[k ];//ke
- //real fW = (D.f[DIR_M00 ])[kw ];
- //real fN = (D.f[DIR_0P0 ])[k ];//kn
- //real fS = (D.f[DIR_0M0 ])[ks ];
- //real fT = (D.f[DIR_00P ])[k ];//kt
- //real fB = (D.f[DIR_00M ])[kb ];
- //real fNE = (D.f[DIR_PP0 ])[k ];//kne
- //real fSW = (D.f[DIR_MM0 ])[ksw];
- //real fSE = (D.f[DIR_PM0 ])[ks ];//kse
- //real fNW = (D.f[DIR_MP0 ])[kw ];//knw
- //real fTE = (D.f[DIR_P0P ])[k ];//kte
- //real fBW = (D.f[DIR_M0M ])[kbw];
- //real fBE = (D.f[DIR_P0M ])[kb ];//kbe
- //real fTW = (D.f[DIR_M0P ])[kw ];//ktw
- //real fTN = (D.f[DIR_0PP ])[k ];//ktn
- //real fBS = (D.f[DIR_0MM ])[kbs];
- //real fBN = (D.f[DIR_0PM ])[kb ];//kbn
- //real fTS = (D.f[DIR_0MP ])[ks ];//kts
+ //real fE = (D.f[DIR_P00])[k ];//ke
+ //real fW = (D.f[DIR_M00])[kw ];
+ //real fN = (D.f[DIR_0P0])[k ];//kn
+ //real fS = (D.f[DIR_0M0])[ks ];
+ //real fT = (D.f[DIR_00P])[k ];//kt
+ //real fB = (D.f[DIR_00M])[kb ];
+ //real fNE = (D.f[DIR_PP0])[k ];//kne
+ //real fSW = (D.f[DIR_MM0])[ksw];
+ //real fSE = (D.f[DIR_PM0])[ks ];//kse
+ //real fNW = (D.f[DIR_MP0])[kw ];//knw
+ //real fTE = (D.f[DIR_P0P])[k ];//kte
+ //real fBW = (D.f[DIR_M0M])[kbw];
+ //real fBE = (D.f[DIR_P0M])[kb ];//kbe
+ //real fTW = (D.f[DIR_M0P])[kw ];//ktw
+ //real fTN = (D.f[DIR_0PP])[k ];//ktn
+ //real fBS = (D.f[DIR_0MM])[kbs];
+ //real fBN = (D.f[DIR_0PM])[kb ];//kbn
+ //real fTS = (D.f[DIR_0MP])[ks ];//kts
//real fZERO = (D.f[DIR_000])[k ];//kzero
- //real fTNE = (D.f[DIR_PPP ])[k ];//ktne
- //real fTSW = (D.f[DIR_MMP ])[ksw];//ktsw
- //real fTSE = (D.f[DIR_PMP ])[ks ];//ktse
- //real fTNW = (D.f[DIR_MPP ])[kw ];//ktnw
- //real fBNE = (D.f[DIR_PPM ])[kb ];//kbne
- //real fBSW = (D.f[DIR_MMM ])[kbsw];
- //real fBSE = (D.f[DIR_PMM ])[kbs];//kbse
- //real fBNW = (D.f[DIR_MPM ])[kbw];//kbnw
+ //real fTNE = (D.f[DIR_PPP])[k ];//ktne
+ //real fTSW = (D.f[DIR_MMP])[ksw];//ktsw
+ //real fTSE = (D.f[DIR_PMP])[ks ];//ktse
+ //real fTNW = (D.f[DIR_MPP])[kw ];//ktnw
+ //real fBNE = (D.f[DIR_PPM])[kb ];//kbne
+ //real fBSW = (D.f[DIR_MMM])[kbsw];
+ //real fBSE = (D.f[DIR_PMM])[kbs];//kbse
+ //real fBNW = (D.f[DIR_MPM])[kbw];//kbnw
////////////////////////////////////////////////////////////////////////////////
real f7ZERO = (D7.f[0])[k];
real f7E = (D7.f[1])[k];
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/AdvectionDiffusion/Incompressible/Mod7/ADIncomp7/ADIncomp7_Device.cuh b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/AdvectionDiffusion/Incompressible/Mod7/ADIncomp7/ADIncomp7_Device.cuh
index 25a17ddbd7038635a2beb2c39212822cbf762034..845ecda946a4e45678082b72b5c74dc96e5810c5 100644
--- a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/AdvectionDiffusion/Incompressible/Mod7/ADIncomp7/ADIncomp7_Device.cuh
+++ b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/AdvectionDiffusion/Incompressible/Mod7/ADIncomp7/ADIncomp7_Device.cuh
@@ -11,7 +11,7 @@ __global__ void LB_Kernel_AD_Incomp_7(real diffusivity,
unsigned int* neighborZ,
real* DDStart,
real* DD7,
- int size_Mat,
+ unsigned long long numberOfLBnodes,
bool EvenOrOdd);
#endif
\ No newline at end of file
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/BGK/BGKCompSP27.cu b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/BGK/BGKCompSP27.cu
index d2f9f60890379d07ecc3d04f4a54d59a0754907a..8c99f3b030984aef6215d5479be4b321145ee54f 100644
--- a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/BGK/BGKCompSP27.cu
+++ b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/BGK/BGKCompSP27.cu
@@ -2,6 +2,7 @@
#include "BGKCompSP27_Device.cuh"
#include "Parameter/Parameter.h"
+#include "cuda/CudaGrid.h"
std::shared_ptr<BGKCompSP27> BGKCompSP27::getNewInstance(std::shared_ptr<Parameter> para, int level)
{
@@ -10,33 +11,18 @@ std::shared_ptr<BGKCompSP27> BGKCompSP27::getNewInstance(std::shared_ptr<Paramet
void BGKCompSP27::run()
{
- int numberOfThreads = para->getParD(level)->numberofthreads;
- int size_Mat = para->getParD(level)->numberOfNodes;
-
- int Grid = (size_Mat / numberOfThreads) + 1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid / Grid1) + 1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 grid(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1);
-
- LB_Kernel_BGK_Comp_SP_27 << < grid, threads >> >( para->getParD(level)->omega,
- para->getParD(level)->typeOfGridNode,
- para->getParD(level)->neighborX,
- para->getParD(level)->neighborY,
- para->getParD(level)->neighborZ,
- para->getParD(level)->distributions.f[0],
- para->getParD(level)->numberOfNodes,
- para->getParD(level)->isEvenTimestep);
- getLastCudaError("LB_Kernel_BGK_Comp_SP_27 execution failed");
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(para->getParD(level)->numberofthreads, para->getParD(level)->numberOfNodes);
+
+ LB_Kernel_BGK_Comp_SP_27<<< grid.grid, grid.threads >>>(
+ para->getParD(level)->omega,
+ para->getParD(level)->typeOfGridNode,
+ para->getParD(level)->neighborX,
+ para->getParD(level)->neighborY,
+ para->getParD(level)->neighborZ,
+ para->getParD(level)->distributions.f[0],
+ para->getParD(level)->numberOfNodes,
+ para->getParD(level)->isEvenTimestep);
+ getLastCudaError("LB_Kernel_BGK_Comp_SP_27 execution failed");
}
BGKCompSP27::BGKCompSP27(std::shared_ptr<Parameter> para, int level)
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/BGK/BGKCompSP27_Device.cu b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/BGK/BGKCompSP27_Device.cu
index 09196d13e94a2404ba280e8a8e9394f0a79e8211..3bdb65c455bd67d66e8b35961f2fa7e1de45f763 100644
--- a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/BGK/BGKCompSP27_Device.cu
+++ b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/BGK/BGKCompSP27_Device.cu
@@ -38,63 +38,63 @@ __global__ void LB_Kernel_BGK_Comp_SP_27( real omega,
Distributions27 D;
if (EvenOrOdd == true)
{
- D.f[DIR_P00] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_M00] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0P0] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0M0] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00P] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00M] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_PP0] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_MM0] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_PM0] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_MP0] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_P0P] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_M0M] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_P0M] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_M0P] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0PP] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0MM] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0PM] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0MP] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_PPP] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_MMP] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_PMP] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_MPP] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_PPM] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_MMM] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_PMM] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_MPM] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DDStart[DIR_P00 * size_Mat];
+ D.f[DIR_M00] = &DDStart[DIR_M00 * size_Mat];
+ D.f[DIR_0P0] = &DDStart[DIR_0P0 * size_Mat];
+ D.f[DIR_0M0] = &DDStart[DIR_0M0 * size_Mat];
+ D.f[DIR_00P] = &DDStart[DIR_00P * size_Mat];
+ D.f[DIR_00M] = &DDStart[DIR_00M * size_Mat];
+ D.f[DIR_PP0] = &DDStart[DIR_PP0 * size_Mat];
+ D.f[DIR_MM0] = &DDStart[DIR_MM0 * size_Mat];
+ D.f[DIR_PM0] = &DDStart[DIR_PM0 * size_Mat];
+ D.f[DIR_MP0] = &DDStart[DIR_MP0 * size_Mat];
+ D.f[DIR_P0P] = &DDStart[DIR_P0P * size_Mat];
+ D.f[DIR_M0M] = &DDStart[DIR_M0M * size_Mat];
+ D.f[DIR_P0M] = &DDStart[DIR_P0M * size_Mat];
+ D.f[DIR_M0P] = &DDStart[DIR_M0P * size_Mat];
+ D.f[DIR_0PP] = &DDStart[DIR_0PP * size_Mat];
+ D.f[DIR_0MM] = &DDStart[DIR_0MM * size_Mat];
+ D.f[DIR_0PM] = &DDStart[DIR_0PM * size_Mat];
+ D.f[DIR_0MP] = &DDStart[DIR_0MP * size_Mat];
+ D.f[DIR_000] = &DDStart[DIR_000 * size_Mat];
+ D.f[DIR_PPP] = &DDStart[DIR_PPP * size_Mat];
+ D.f[DIR_MMP] = &DDStart[DIR_MMP * size_Mat];
+ D.f[DIR_PMP] = &DDStart[DIR_PMP * size_Mat];
+ D.f[DIR_MPP] = &DDStart[DIR_MPP * size_Mat];
+ D.f[DIR_PPM] = &DDStart[DIR_PPM * size_Mat];
+ D.f[DIR_MMM] = &DDStart[DIR_MMM * size_Mat];
+ D.f[DIR_PMM] = &DDStart[DIR_PMM * size_Mat];
+ D.f[DIR_MPM] = &DDStart[DIR_MPM * size_Mat];
}
else
{
- D.f[DIR_M00] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_P00] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0M0] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0P0] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00M] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00P] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_MM0] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_PP0] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_MP0] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_PM0] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_M0M] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_P0P] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_M0P] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_P0M] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0MM] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0PP] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0MP] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0PM] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_MMM] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_PPM] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_MPM] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_PMM] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_MMP] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_PPP] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_MPP] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_PMP] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_M00] = &DDStart[DIR_P00 * size_Mat];
+ D.f[DIR_P00] = &DDStart[DIR_M00 * size_Mat];
+ D.f[DIR_0M0] = &DDStart[DIR_0P0 * size_Mat];
+ D.f[DIR_0P0] = &DDStart[DIR_0M0 * size_Mat];
+ D.f[DIR_00M] = &DDStart[DIR_00P * size_Mat];
+ D.f[DIR_00P] = &DDStart[DIR_00M * size_Mat];
+ D.f[DIR_MM0] = &DDStart[DIR_PP0 * size_Mat];
+ D.f[DIR_PP0] = &DDStart[DIR_MM0 * size_Mat];
+ D.f[DIR_MP0] = &DDStart[DIR_PM0 * size_Mat];
+ D.f[DIR_PM0] = &DDStart[DIR_MP0 * size_Mat];
+ D.f[DIR_M0M] = &DDStart[DIR_P0P * size_Mat];
+ D.f[DIR_P0P] = &DDStart[DIR_M0M * size_Mat];
+ D.f[DIR_M0P] = &DDStart[DIR_P0M * size_Mat];
+ D.f[DIR_P0M] = &DDStart[DIR_M0P * size_Mat];
+ D.f[DIR_0MM] = &DDStart[DIR_0PP * size_Mat];
+ D.f[DIR_0PP] = &DDStart[DIR_0MM * size_Mat];
+ D.f[DIR_0MP] = &DDStart[DIR_0PM * size_Mat];
+ D.f[DIR_0PM] = &DDStart[DIR_0MP * size_Mat];
+ D.f[DIR_000] = &DDStart[DIR_000 * size_Mat];
+ D.f[DIR_MMM] = &DDStart[DIR_PPP * size_Mat];
+ D.f[DIR_PPM] = &DDStart[DIR_MMP * size_Mat];
+ D.f[DIR_MPM] = &DDStart[DIR_PMP * size_Mat];
+ D.f[DIR_PMM] = &DDStart[DIR_MPP * size_Mat];
+ D.f[DIR_MMP] = &DDStart[DIR_PPM * size_Mat];
+ D.f[DIR_PPP] = &DDStart[DIR_MMM * size_Mat];
+ D.f[DIR_MPP] = &DDStart[DIR_PMM * size_Mat];
+ D.f[DIR_PMP] = &DDStart[DIR_MPM * size_Mat];
}
////////////////////////////////////////////////////////////////////////////////
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/BGKPlus/BGKPlusCompSP27.cu b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/BGKPlus/BGKPlusCompSP27.cu
index beebda2437ca4e7385ab812b9106edabe213227e..a4b136d1c21b1e4c68432eef5e21ff8c968bdfec 100644
--- a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/BGKPlus/BGKPlusCompSP27.cu
+++ b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/BGKPlus/BGKPlusCompSP27.cu
@@ -2,6 +2,7 @@
#include "BGKPlusCompSP27_Device.cuh"
#include "Parameter/Parameter.h"
+#include "cuda/CudaGrid.h"
std::shared_ptr<BGKPlusCompSP27> BGKPlusCompSP27::getNewInstance(std::shared_ptr<Parameter> para, int level)
{
@@ -10,33 +11,18 @@ std::shared_ptr<BGKPlusCompSP27> BGKPlusCompSP27::getNewInstance(std::shared_ptr
void BGKPlusCompSP27::run()
{
- int numberOfThreads = para->getParD(level)->numberofthreads;
- int size_Mat = para->getParD(level)->numberOfNodes;
-
- int Grid = (size_Mat / numberOfThreads) + 1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid / Grid1) + 1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 grid(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1);
-
- LB_Kernel_BGK_Plus_Comp_SP_27 << < grid, threads >> >( para->getParD(level)->omega,
- para->getParD(level)->typeOfGridNode,
- para->getParD(level)->neighborX,
- para->getParD(level)->neighborY,
- para->getParD(level)->neighborZ,
- para->getParD(level)->distributions.f[0],
- size_Mat,
- para->getParD(level)->isEvenTimestep);
- getLastCudaError("LB_Kernel_BGK_Plus_Comp_SP_27 execution failed");
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(para->getParD(level)->numberofthreads, para->getParD(level)->numberOfNodes);
+
+ LB_Kernel_BGK_Plus_Comp_SP_27 <<< grid.grid, grid.threads >>>(
+ para->getParD(level)->omega,
+ para->getParD(level)->typeOfGridNode,
+ para->getParD(level)->neighborX,
+ para->getParD(level)->neighborY,
+ para->getParD(level)->neighborZ,
+ para->getParD(level)->distributions.f[0],
+ para->getParD(level)->numberOfNodes,
+ para->getParD(level)->isEvenTimestep);
+ getLastCudaError("LB_Kernel_BGK_Plus_Comp_SP_27 execution failed");
}
BGKPlusCompSP27::BGKPlusCompSP27(std::shared_ptr<Parameter> para, int level)
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/BGKPlus/BGKPlusCompSP27_Device.cu b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/BGKPlus/BGKPlusCompSP27_Device.cu
index 325f65ece9baddf88adc91baa753bdfc4bd0eced..1f44fee9ea8b20241f87bea6310c96db2b82d1c4 100644
--- a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/BGKPlus/BGKPlusCompSP27_Device.cu
+++ b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/BGKPlus/BGKPlusCompSP27_Device.cu
@@ -38,63 +38,63 @@ __global__ void LB_Kernel_BGK_Plus_Comp_SP_27(
Distributions27 D;
if (EvenOrOdd == true)
{
- D.f[DIR_P00] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_M00] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0P0] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0M0] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00P] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00M] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_PP0] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_MM0] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_PM0] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_MP0] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_P0P] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_M0M] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_P0M] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_M0P] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0PP] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0MM] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0PM] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0MP] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_PPP] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_MMP] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_PMP] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_MPP] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_PPM] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_MMM] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_PMM] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_MPM] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DDStart[DIR_P00 * size_Mat];
+ D.f[DIR_M00] = &DDStart[DIR_M00 * size_Mat];
+ D.f[DIR_0P0] = &DDStart[DIR_0P0 * size_Mat];
+ D.f[DIR_0M0] = &DDStart[DIR_0M0 * size_Mat];
+ D.f[DIR_00P] = &DDStart[DIR_00P * size_Mat];
+ D.f[DIR_00M] = &DDStart[DIR_00M * size_Mat];
+ D.f[DIR_PP0] = &DDStart[DIR_PP0 * size_Mat];
+ D.f[DIR_MM0] = &DDStart[DIR_MM0 * size_Mat];
+ D.f[DIR_PM0] = &DDStart[DIR_PM0 * size_Mat];
+ D.f[DIR_MP0] = &DDStart[DIR_MP0 * size_Mat];
+ D.f[DIR_P0P] = &DDStart[DIR_P0P * size_Mat];
+ D.f[DIR_M0M] = &DDStart[DIR_M0M * size_Mat];
+ D.f[DIR_P0M] = &DDStart[DIR_P0M * size_Mat];
+ D.f[DIR_M0P] = &DDStart[DIR_M0P * size_Mat];
+ D.f[DIR_0PP] = &DDStart[DIR_0PP * size_Mat];
+ D.f[DIR_0MM] = &DDStart[DIR_0MM * size_Mat];
+ D.f[DIR_0PM] = &DDStart[DIR_0PM * size_Mat];
+ D.f[DIR_0MP] = &DDStart[DIR_0MP * size_Mat];
+ D.f[DIR_000] = &DDStart[DIR_000 * size_Mat];
+ D.f[DIR_PPP] = &DDStart[DIR_PPP * size_Mat];
+ D.f[DIR_MMP] = &DDStart[DIR_MMP * size_Mat];
+ D.f[DIR_PMP] = &DDStart[DIR_PMP * size_Mat];
+ D.f[DIR_MPP] = &DDStart[DIR_MPP * size_Mat];
+ D.f[DIR_PPM] = &DDStart[DIR_PPM * size_Mat];
+ D.f[DIR_MMM] = &DDStart[DIR_MMM * size_Mat];
+ D.f[DIR_PMM] = &DDStart[DIR_PMM * size_Mat];
+ D.f[DIR_MPM] = &DDStart[DIR_MPM * size_Mat];
}
else
{
- D.f[DIR_M00] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_P00] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0M0] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0P0] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00M] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00P] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_MM0] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_PP0] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_MP0] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_PM0] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_M0M] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_P0P] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_M0P] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_P0M] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0MM] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0PP] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0MP] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0PM] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_MMM] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_PPM] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_MPM] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_PMM] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_MMP] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_PPP] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_MPP] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_PMP] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_M00] = &DDStart[DIR_P00 * size_Mat];
+ D.f[DIR_P00] = &DDStart[DIR_M00 * size_Mat];
+ D.f[DIR_0M0] = &DDStart[DIR_0P0 * size_Mat];
+ D.f[DIR_0P0] = &DDStart[DIR_0M0 * size_Mat];
+ D.f[DIR_00M] = &DDStart[DIR_00P * size_Mat];
+ D.f[DIR_00P] = &DDStart[DIR_00M * size_Mat];
+ D.f[DIR_MM0] = &DDStart[DIR_PP0 * size_Mat];
+ D.f[DIR_PP0] = &DDStart[DIR_MM0 * size_Mat];
+ D.f[DIR_MP0] = &DDStart[DIR_PM0 * size_Mat];
+ D.f[DIR_PM0] = &DDStart[DIR_MP0 * size_Mat];
+ D.f[DIR_M0M] = &DDStart[DIR_P0P * size_Mat];
+ D.f[DIR_P0P] = &DDStart[DIR_M0M * size_Mat];
+ D.f[DIR_M0P] = &DDStart[DIR_P0M * size_Mat];
+ D.f[DIR_P0M] = &DDStart[DIR_M0P * size_Mat];
+ D.f[DIR_0MM] = &DDStart[DIR_0PP * size_Mat];
+ D.f[DIR_0PP] = &DDStart[DIR_0MM * size_Mat];
+ D.f[DIR_0MP] = &DDStart[DIR_0PM * size_Mat];
+ D.f[DIR_0PM] = &DDStart[DIR_0MP * size_Mat];
+ D.f[DIR_000] = &DDStart[DIR_000 * size_Mat];
+ D.f[DIR_MMM] = &DDStart[DIR_PPP * size_Mat];
+ D.f[DIR_PPM] = &DDStart[DIR_MMP * size_Mat];
+ D.f[DIR_MPM] = &DDStart[DIR_PMP * size_Mat];
+ D.f[DIR_PMM] = &DDStart[DIR_MPP * size_Mat];
+ D.f[DIR_MMP] = &DDStart[DIR_PPM * size_Mat];
+ D.f[DIR_PPP] = &DDStart[DIR_MMM * size_Mat];
+ D.f[DIR_MPP] = &DDStart[DIR_PMM * size_Mat];
+ D.f[DIR_PMP] = &DDStart[DIR_MPM * size_Mat];
}
////////////////////////////////////////////////////////////////////////////////
@@ -127,33 +127,33 @@ __global__ void LB_Kernel_BGK_Plus_Comp_SP_27(
//unsigned int ktne = k;
unsigned int kbsw = neighborZ[ksw];
//////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
- real mfcbb = (D.f[DIR_P00])[k];//[ke ];// + c2over27 ;(D.f[DIR_P00 ])[k ];//ke
- real mfabb = (D.f[DIR_M00])[kw];//[kw ];// + c2over27 ;(D.f[DIR_M00 ])[kw ];
- real mfbcb = (D.f[DIR_0P0])[k];//[kn ];// + c2over27 ;(D.f[DIR_0P0 ])[k ];//kn
- real mfbab = (D.f[DIR_0M0])[ks];//[ks ];// + c2over27 ;(D.f[DIR_0M0 ])[ks ];
- real mfbbc = (D.f[DIR_00P])[k];//[kt ];// + c2over27 ;(D.f[DIR_00P ])[k ];//kt
- real mfbba = (D.f[DIR_00M])[kb];//[kb ];// + c2over27 ;(D.f[DIR_00M ])[kb ];
- real mfccb = (D.f[DIR_PP0])[k];//[kne ];// + c1over54 ;(D.f[DIR_PP0 ])[k ];//kne
- real mfaab = (D.f[DIR_MM0])[ksw];//[ksw ];// + c1over54 ;(D.f[DIR_MM0 ])[ksw];
- real mfcab = (D.f[DIR_PM0])[ks];//[kse ];// + c1over54 ;(D.f[DIR_PM0 ])[ks ];//kse
- real mfacb = (D.f[DIR_MP0])[kw];//[knw ];// + c1over54 ;(D.f[DIR_MP0 ])[kw ];//knw
- real mfcbc = (D.f[DIR_P0P])[k];//[kte ];// + c1over54 ;(D.f[DIR_P0P ])[k ];//kte
- real mfaba = (D.f[DIR_M0M])[kbw];//[kbw ];// + c1over54 ;(D.f[DIR_M0M ])[kbw];
- real mfcba = (D.f[DIR_P0M])[kb];//[kbe ];// + c1over54 ;(D.f[DIR_P0M ])[kb ];//kbe
- real mfabc = (D.f[DIR_M0P])[kw];//[ktw ];// + c1over54 ;(D.f[DIR_M0P ])[kw ];//ktw
- real mfbcc = (D.f[DIR_0PP])[k];//[ktn ];// + c1over54 ;(D.f[DIR_0PP ])[k ];//ktn
- real mfbaa = (D.f[DIR_0MM])[kbs];//[kbs ];// + c1over54 ;(D.f[DIR_0MM ])[kbs];
- real mfbca = (D.f[DIR_0PM])[kb];//[kbn ];// + c1over54 ;(D.f[DIR_0PM ])[kb ];//kbn
- real mfbac = (D.f[DIR_0MP])[ks];//[kts ];// + c1over54 ;(D.f[DIR_0MP ])[ks ];//kts
+ real mfcbb = (D.f[DIR_P00])[k];//[ke ];// + c2over27 ;(D.f[DIR_P00])[k ];//ke
+ real mfabb = (D.f[DIR_M00])[kw];//[kw ];// + c2over27 ;(D.f[DIR_M00])[kw ];
+ real mfbcb = (D.f[DIR_0P0])[k];//[kn ];// + c2over27 ;(D.f[DIR_0P0])[k ];//kn
+ real mfbab = (D.f[DIR_0M0])[ks];//[ks ];// + c2over27 ;(D.f[DIR_0M0])[ks ];
+ real mfbbc = (D.f[DIR_00P])[k];//[kt ];// + c2over27 ;(D.f[DIR_00P])[k ];//kt
+ real mfbba = (D.f[DIR_00M])[kb];//[kb ];// + c2over27 ;(D.f[DIR_00M])[kb ];
+ real mfccb = (D.f[DIR_PP0])[k];//[kne ];// + c1over54 ;(D.f[DIR_PP0])[k ];//kne
+ real mfaab = (D.f[DIR_MM0])[ksw];//[ksw ];// + c1over54 ;(D.f[DIR_MM0])[ksw];
+ real mfcab = (D.f[DIR_PM0])[ks];//[kse ];// + c1over54 ;(D.f[DIR_PM0])[ks ];//kse
+ real mfacb = (D.f[DIR_MP0])[kw];//[knw ];// + c1over54 ;(D.f[DIR_MP0])[kw ];//knw
+ real mfcbc = (D.f[DIR_P0P])[k];//[kte ];// + c1over54 ;(D.f[DIR_P0P])[k ];//kte
+ real mfaba = (D.f[DIR_M0M])[kbw];//[kbw ];// + c1over54 ;(D.f[DIR_M0M])[kbw];
+ real mfcba = (D.f[DIR_P0M])[kb];//[kbe ];// + c1over54 ;(D.f[DIR_P0M])[kb ];//kbe
+ real mfabc = (D.f[DIR_M0P])[kw];//[ktw ];// + c1over54 ;(D.f[DIR_M0P])[kw ];//ktw
+ real mfbcc = (D.f[DIR_0PP])[k];//[ktn ];// + c1over54 ;(D.f[DIR_0PP])[k ];//ktn
+ real mfbaa = (D.f[DIR_0MM])[kbs];//[kbs ];// + c1over54 ;(D.f[DIR_0MM])[kbs];
+ real mfbca = (D.f[DIR_0PM])[kb];//[kbn ];// + c1over54 ;(D.f[DIR_0PM])[kb ];//kbn
+ real mfbac = (D.f[DIR_0MP])[ks];//[kts ];// + c1over54 ;(D.f[DIR_0MP])[ks ];//kts
real mfbbb = (D.f[DIR_000])[k];//[kzero];// + c8over27 ;(D.f[DIR_000])[k ];//kzero
- real mfccc = (D.f[DIR_PPP])[k];//[ktne ];// + c1over216;(D.f[DIR_PPP ])[k ];//ktne
- real mfaac = (D.f[DIR_MMP])[ksw];//[ktsw ];// + c1over216;(D.f[DIR_MMP ])[ksw];//ktsw
- real mfcac = (D.f[DIR_PMP])[ks];//[ktse ];// + c1over216;(D.f[DIR_PMP ])[ks ];//ktse
- real mfacc = (D.f[DIR_MPP])[kw];//[ktnw ];// + c1over216;(D.f[DIR_MPP ])[kw ];//ktnw
- real mfcca = (D.f[DIR_PPM])[kb];//[kbne ];// + c1over216;(D.f[DIR_PPM ])[kb ];//kbne
- real mfaaa = (D.f[DIR_MMM])[kbsw];//[kbsw ];// + c1over216;(D.f[DIR_MMM ])[kbsw];
- real mfcaa = (D.f[DIR_PMM])[kbs];//[kbse ];// + c1over216;(D.f[DIR_PMM ])[kbs];//kbse
- real mfaca = (D.f[DIR_MPM])[kbw];//[kbnw ];// + c1over216;(D.f[DIR_MPM ])[kbw];//kbnw
+ real mfccc = (D.f[DIR_PPP])[k];//[ktne ];// + c1over216;(D.f[DIR_PPP])[k ];//ktne
+ real mfaac = (D.f[DIR_MMP])[ksw];//[ktsw ];// + c1over216;(D.f[DIR_MMP])[ksw];//ktsw
+ real mfcac = (D.f[DIR_PMP])[ks];//[ktse ];// + c1over216;(D.f[DIR_PMP])[ks ];//ktse
+ real mfacc = (D.f[DIR_MPP])[kw];//[ktnw ];// + c1over216;(D.f[DIR_MPP])[kw ];//ktnw
+ real mfcca = (D.f[DIR_PPM])[kb];//[kbne ];// + c1over216;(D.f[DIR_PPM])[kb ];//kbne
+ real mfaaa = (D.f[DIR_MMM])[kbsw];//[kbsw ];// + c1over216;(D.f[DIR_MMM])[kbsw];
+ real mfcaa = (D.f[DIR_PMM])[kbs];//[kbse ];// + c1over216;(D.f[DIR_PMM])[kbs];//kbse
+ real mfaca = (D.f[DIR_MPM])[kbw];//[kbnw ];// + c1over216;(D.f[DIR_MPM])[kbw];//kbnw
////////////////////////////////////////////////////////////////////////////////////
//slow
//real oMdrho = one - ((((mfccc+mfaaa) + (mfaca+mfcac)) + ((mfacc+mfcaa) + (mfaac+mfcca))) +
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/BGKUnified/BGKUnified.cu b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/BGKUnified/BGKUnified.cu
index 3d7f6fb9a8980454ebc83c51c7dd8865688fa166..1107d343801f8ac3626b03a93ca92415217732ac 100644
--- a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/BGKUnified/BGKUnified.cu
+++ b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/BGKUnified/BGKUnified.cu
@@ -6,6 +6,7 @@
#include "../RunLBMKernel.cuh"
#include <lbm/BGK.h>
+#include <lbm/KernelParameter.h>
namespace vf
@@ -31,15 +32,16 @@ BGKUnified::BGKUnified(std::shared_ptr<Parameter> para, int level)
void BGKUnified::run()
{
- GPUKernelParameter kernelParameter{ para->getParD(level)->omega,
- para->getParD(level)->typeOfGridNode,
- para->getParD(level)->neighborX,
- para->getParD(level)->neighborY,
- para->getParD(level)->neighborZ,
- para->getParD(level)->distributions.f[0],
- (int)para->getParD(level)->numberOfNodes,
- nullptr, /* forces not used in bgk kernel */
- para->getParD(level)->isEvenTimestep };
+ GPUKernelParameter kernelParameter{
+ para->getParD(level)->omega,
+ para->getParD(level)->typeOfGridNode,
+ para->getParD(level)->neighborX,
+ para->getParD(level)->neighborY,
+ para->getParD(level)->neighborZ,
+ para->getParD(level)->distributions.f[0],
+ (int)para->getParD(level)->numberOfNodes,
+ nullptr, /* forces not used in bgk kernel */
+ para->getParD(level)->isEvenTimestep };
auto lambda = [] __device__(lbm::KernelParameter parameter) {
return lbm::bgk(parameter);
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/Cascade/CascadeCompSP27.cu b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/Cascade/CascadeCompSP27.cu
index eca3a9953024e44fd91e7f9f98956e4329574d09..dcfda06db462fd83120751a32a40365445d659ba 100644
--- a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/Cascade/CascadeCompSP27.cu
+++ b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/Cascade/CascadeCompSP27.cu
@@ -2,6 +2,7 @@
#include "CascadeCompSP27_Device.cuh"
#include "Parameter/Parameter.h"
+#include "cuda/CudaGrid.h"
std::shared_ptr<CascadeCompSP27> CascadeCompSP27::getNewInstance(std::shared_ptr<Parameter> para, int level)
{
@@ -10,33 +11,18 @@ std::shared_ptr<CascadeCompSP27> CascadeCompSP27::getNewInstance(std::shared_ptr
void CascadeCompSP27::run()
{
- int size_Mat = para->getParD(level)->numberOfNodes;
- int numberOfThreads = para->getParD(level)->numberofthreads;
-
- int Grid = (size_Mat / numberOfThreads) + 1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid / Grid1) + 1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 grid(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1);
-
- LB_Kernel_Cascade_Comp_SP_27 << < grid, threads >> >( para->getParD(level)->omega,
- para->getParD(level)->typeOfGridNode,
- para->getParD(level)->neighborX,
- para->getParD(level)->neighborY,
- para->getParD(level)->neighborZ,
- para->getParD(level)->distributions.f[0],
- para->getParD(level)->numberOfNodes,
- para->getParD(level)->isEvenTimestep);
- getLastCudaError("LB_Kernel_Cascade_Comp_SP_27 execution failed");
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(para->getParD(level)->numberofthreads, para->getParD(level)->numberOfNodes);
+
+ LB_Kernel_Cascade_Comp_SP_27 <<< grid.grid, grid.threads >>>(
+ para->getParD(level)->omega,
+ para->getParD(level)->typeOfGridNode,
+ para->getParD(level)->neighborX,
+ para->getParD(level)->neighborY,
+ para->getParD(level)->neighborZ,
+ para->getParD(level)->distributions.f[0],
+ para->getParD(level)->numberOfNodes,
+ para->getParD(level)->isEvenTimestep);
+ getLastCudaError("LB_Kernel_Cascade_Comp_SP_27 execution failed");
}
CascadeCompSP27::CascadeCompSP27(std::shared_ptr<Parameter> para, int level)
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/Cascade/CascadeCompSP27_Device.cu b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/Cascade/CascadeCompSP27_Device.cu
index 3f69fa47288343fbdd91e77dbb7f154501349098..af0a7c118191243c80c420856a70711a1fc17d2b 100644
--- a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/Cascade/CascadeCompSP27_Device.cu
+++ b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/Cascade/CascadeCompSP27_Device.cu
@@ -37,63 +37,63 @@ __global__ void LB_Kernel_Cascade_Comp_SP_27(real omega,
Distributions27 D;
if (EvenOrOdd==true)
{
- D.f[DIR_P00 ] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DDStart[DIR_P00 * size_Mat];
+ D.f[DIR_M00] = &DDStart[DIR_M00 * size_Mat];
+ D.f[DIR_0P0] = &DDStart[DIR_0P0 * size_Mat];
+ D.f[DIR_0M0] = &DDStart[DIR_0M0 * size_Mat];
+ D.f[DIR_00P] = &DDStart[DIR_00P * size_Mat];
+ D.f[DIR_00M] = &DDStart[DIR_00M * size_Mat];
+ D.f[DIR_PP0] = &DDStart[DIR_PP0 * size_Mat];
+ D.f[DIR_MM0] = &DDStart[DIR_MM0 * size_Mat];
+ D.f[DIR_PM0] = &DDStart[DIR_PM0 * size_Mat];
+ D.f[DIR_MP0] = &DDStart[DIR_MP0 * size_Mat];
+ D.f[DIR_P0P] = &DDStart[DIR_P0P * size_Mat];
+ D.f[DIR_M0M] = &DDStart[DIR_M0M * size_Mat];
+ D.f[DIR_P0M] = &DDStart[DIR_P0M * size_Mat];
+ D.f[DIR_M0P] = &DDStart[DIR_M0P * size_Mat];
+ D.f[DIR_0PP] = &DDStart[DIR_0PP * size_Mat];
+ D.f[DIR_0MM] = &DDStart[DIR_0MM * size_Mat];
+ D.f[DIR_0PM] = &DDStart[DIR_0PM * size_Mat];
+ D.f[DIR_0MP] = &DDStart[DIR_0MP * size_Mat];
+ D.f[DIR_000] = &DDStart[DIR_000 * size_Mat];
+ D.f[DIR_PPP] = &DDStart[DIR_PPP * size_Mat];
+ D.f[DIR_MMP] = &DDStart[DIR_MMP * size_Mat];
+ D.f[DIR_PMP] = &DDStart[DIR_PMP * size_Mat];
+ D.f[DIR_MPP] = &DDStart[DIR_MPP * size_Mat];
+ D.f[DIR_PPM] = &DDStart[DIR_PPM * size_Mat];
+ D.f[DIR_MMM] = &DDStart[DIR_MMM * size_Mat];
+ D.f[DIR_PMM] = &DDStart[DIR_PMM * size_Mat];
+ D.f[DIR_MPM] = &DDStart[DIR_MPM * size_Mat];
}
else
{
- D.f[DIR_M00 ] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_MMM ] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_PPM ] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_MPM ] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_PMM ] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_MMP ] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_PPP ] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_MPP ] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_PMP ] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_M00] = &DDStart[DIR_P00 * size_Mat];
+ D.f[DIR_P00] = &DDStart[DIR_M00 * size_Mat];
+ D.f[DIR_0M0] = &DDStart[DIR_0P0 * size_Mat];
+ D.f[DIR_0P0] = &DDStart[DIR_0M0 * size_Mat];
+ D.f[DIR_00M] = &DDStart[DIR_00P * size_Mat];
+ D.f[DIR_00P] = &DDStart[DIR_00M * size_Mat];
+ D.f[DIR_MM0] = &DDStart[DIR_PP0 * size_Mat];
+ D.f[DIR_PP0] = &DDStart[DIR_MM0 * size_Mat];
+ D.f[DIR_MP0] = &DDStart[DIR_PM0 * size_Mat];
+ D.f[DIR_PM0] = &DDStart[DIR_MP0 * size_Mat];
+ D.f[DIR_M0M] = &DDStart[DIR_P0P * size_Mat];
+ D.f[DIR_P0P] = &DDStart[DIR_M0M * size_Mat];
+ D.f[DIR_M0P] = &DDStart[DIR_P0M * size_Mat];
+ D.f[DIR_P0M] = &DDStart[DIR_M0P * size_Mat];
+ D.f[DIR_0MM] = &DDStart[DIR_0PP * size_Mat];
+ D.f[DIR_0PP] = &DDStart[DIR_0MM * size_Mat];
+ D.f[DIR_0MP] = &DDStart[DIR_0PM * size_Mat];
+ D.f[DIR_0PM] = &DDStart[DIR_0MP * size_Mat];
+ D.f[DIR_000] = &DDStart[DIR_000 * size_Mat];
+ D.f[DIR_MMM] = &DDStart[DIR_PPP * size_Mat];
+ D.f[DIR_PPM] = &DDStart[DIR_MMP * size_Mat];
+ D.f[DIR_MPM] = &DDStart[DIR_PMP * size_Mat];
+ D.f[DIR_PMM] = &DDStart[DIR_MPP * size_Mat];
+ D.f[DIR_MMP] = &DDStart[DIR_PPM * size_Mat];
+ D.f[DIR_PPP] = &DDStart[DIR_MMM * size_Mat];
+ D.f[DIR_MPP] = &DDStart[DIR_PMM * size_Mat];
+ D.f[DIR_PMP] = &DDStart[DIR_MPM * size_Mat];
}
////////////////////////////////////////////////////////////////////////////////
@@ -126,33 +126,33 @@ __global__ void LB_Kernel_Cascade_Comp_SP_27(real omega,
//unsigned int ktne = k;
unsigned int kbsw = neighborZ[ksw];
//////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
- real mfcbb = (D.f[DIR_P00 ])[k ];//[ke ];
- real mfabb = (D.f[DIR_M00 ])[kw ];//[kw ];
- real mfbcb = (D.f[DIR_0P0 ])[k ];//[kn ];
- real mfbab = (D.f[DIR_0M0 ])[ks ];//[ks ];
- real mfbbc = (D.f[DIR_00P ])[k ];//[kt ];
- real mfbba = (D.f[DIR_00M ])[kb ];//[kb ];
- real mfccb = (D.f[DIR_PP0 ])[k ];//[kne ];
- real mfaab = (D.f[DIR_MM0 ])[ksw];//[ksw ];
- real mfcab = (D.f[DIR_PM0 ])[ks ];//[kse ];
- real mfacb = (D.f[DIR_MP0 ])[kw ];//[knw ];
- real mfcbc = (D.f[DIR_P0P ])[k ];//[kte ];
- real mfaba = (D.f[DIR_M0M ])[kbw];//[kbw ];
- real mfcba = (D.f[DIR_P0M ])[kb ];//[kbe ];
- real mfabc = (D.f[DIR_M0P ])[kw ];//[ktw ];
- real mfbcc = (D.f[DIR_0PP ])[k ];//[ktn ];
- real mfbaa = (D.f[DIR_0MM ])[kbs];//[kbs ];
- real mfbca = (D.f[DIR_0PM ])[kb ];//[kbn ];
- real mfbac = (D.f[DIR_0MP ])[ks ];//[kts ];
+ real mfcbb = (D.f[DIR_P00])[k ];//[ke ];
+ real mfabb = (D.f[DIR_M00])[kw ];//[kw ];
+ real mfbcb = (D.f[DIR_0P0])[k ];//[kn ];
+ real mfbab = (D.f[DIR_0M0])[ks ];//[ks ];
+ real mfbbc = (D.f[DIR_00P])[k ];//[kt ];
+ real mfbba = (D.f[DIR_00M])[kb ];//[kb ];
+ real mfccb = (D.f[DIR_PP0])[k ];//[kne ];
+ real mfaab = (D.f[DIR_MM0])[ksw];//[ksw ];
+ real mfcab = (D.f[DIR_PM0])[ks ];//[kse ];
+ real mfacb = (D.f[DIR_MP0])[kw ];//[knw ];
+ real mfcbc = (D.f[DIR_P0P])[k ];//[kte ];
+ real mfaba = (D.f[DIR_M0M])[kbw];//[kbw ];
+ real mfcba = (D.f[DIR_P0M])[kb ];//[kbe ];
+ real mfabc = (D.f[DIR_M0P])[kw ];//[ktw ];
+ real mfbcc = (D.f[DIR_0PP])[k ];//[ktn ];
+ real mfbaa = (D.f[DIR_0MM])[kbs];//[kbs ];
+ real mfbca = (D.f[DIR_0PM])[kb ];//[kbn ];
+ real mfbac = (D.f[DIR_0MP])[ks ];//[kts ];
real mfbbb = (D.f[DIR_000])[k ];//[kzero];
- real mfccc = (D.f[DIR_PPP ])[k ];//[ktne ];
- real mfaac = (D.f[DIR_MMP ])[ksw];//[ktsw ];
- real mfcac = (D.f[DIR_PMP ])[ks ];//[ktse ];
- real mfacc = (D.f[DIR_MPP ])[kw ];//[ktnw ];
- real mfcca = (D.f[DIR_PPM ])[kb ];//[kbne ];
- real mfaaa = (D.f[DIR_MMM ])[kbsw];//[kbsw ]
- real mfcaa = (D.f[DIR_PMM ])[kbs];//[kbse ];
- real mfaca = (D.f[DIR_MPM ])[kbw];//[kbnw ];
+ real mfccc = (D.f[DIR_PPP])[k ];//[ktne ];
+ real mfaac = (D.f[DIR_MMP])[ksw];//[ktsw ];
+ real mfcac = (D.f[DIR_PMP])[ks ];//[ktse ];
+ real mfacc = (D.f[DIR_MPP])[kw ];//[ktnw ];
+ real mfcca = (D.f[DIR_PPM])[kb ];//[kbne ];
+ real mfaaa = (D.f[DIR_MMM])[kbsw];//[kbsw ]
+ real mfcaa = (D.f[DIR_PMM])[kbs];//[kbse ];
+ real mfaca = (D.f[DIR_MPM])[kbw];//[kbnw ];
////////////////////////////////////////////////////////////////////////////////////
real rho = (mfccc+mfaaa + mfaca+mfcac + mfacc+mfcaa + mfaac+mfcca +
mfbac+mfbca + mfbaa+mfbcc + mfabc+mfcba + mfaba+mfcbc + mfacb+mfcab + mfaab+mfccb +
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/Cumulant/CumulantCompSP27.cu b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/Cumulant/CumulantCompSP27.cu
index 3f45c7ea71c385f948eac2e052a8d970010c413d..7817c398285dda131401bd14c3ccdd8c119c5680 100644
--- a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/Cumulant/CumulantCompSP27.cu
+++ b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/Cumulant/CumulantCompSP27.cu
@@ -2,6 +2,7 @@
#include "CumulantCompSP27_Device.cuh"
#include "Parameter/Parameter.h"
+#include "cuda/CudaGrid.h"
std::shared_ptr<CumulantCompSP27> CumulantCompSP27::getNewInstance(std::shared_ptr<Parameter> para, int level)
{
@@ -10,33 +11,18 @@ std::shared_ptr<CumulantCompSP27> CumulantCompSP27::getNewInstance(std::shared_p
void CumulantCompSP27::run()
{
- int size_Mat = para->getParD(level)->numberOfNodes;
- int numberOfThreads = para->getParD(level)->numberofthreads;
-
- int Grid = (size_Mat / numberOfThreads) + 1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid / Grid1) + 1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 grid(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1);
-
- LB_Kernel_Cum_Comp_SP_27 << < grid, threads >> >( para->getParD(level)->omega,
- para->getParD(level)->typeOfGridNode,
- para->getParD(level)->neighborX,
- para->getParD(level)->neighborY,
- para->getParD(level)->neighborZ,
- para->getParD(level)->distributions.f[0],
- para->getParD(level)->numberOfNodes,
- para->getParD(level)->isEvenTimestep);
- getLastCudaError("LB_Kernel_Kum_Comp_SP_27 execution failed");
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(para->getParD(level)->numberofthreads, para->getParD(level)->numberOfNodes);
+
+ LB_Kernel_Cum_Comp_SP_27 <<< grid.grid, grid.threads >>>(
+ para->getParD(level)->omega,
+ para->getParD(level)->typeOfGridNode,
+ para->getParD(level)->neighborX,
+ para->getParD(level)->neighborY,
+ para->getParD(level)->neighborZ,
+ para->getParD(level)->distributions.f[0],
+ para->getParD(level)->numberOfNodes,
+ para->getParD(level)->isEvenTimestep);
+ getLastCudaError("LB_Kernel_Cum_Comp_SP_27 execution failed");
}
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/Cumulant/CumulantCompSP27_Device.cu b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/Cumulant/CumulantCompSP27_Device.cu
index ad2ffdf4170d98125e6758c0e2f548122093cea6..1dfab5846795e61509cdba28478fe6ce623983b5 100644
--- a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/Cumulant/CumulantCompSP27_Device.cu
+++ b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/Cumulant/CumulantCompSP27_Device.cu
@@ -37,63 +37,63 @@ __global__ void LB_Kernel_Cum_Comp_SP_27(real omega,
Distributions27 D;
if (EvenOrOdd == true)
{
- D.f[DIR_P00] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_M00] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0P0] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0M0] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00P] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00M] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_PP0] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_MM0] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_PM0] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_MP0] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_P0P] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_M0M] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_P0M] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_M0P] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0PP] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0MM] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0PM] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0MP] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_PPP] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_MMP] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_PMP] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_MPP] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_PPM] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_MMM] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_PMM] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_MPM] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DDStart[DIR_P00 * size_Mat];
+ D.f[DIR_M00] = &DDStart[DIR_M00 * size_Mat];
+ D.f[DIR_0P0] = &DDStart[DIR_0P0 * size_Mat];
+ D.f[DIR_0M0] = &DDStart[DIR_0M0 * size_Mat];
+ D.f[DIR_00P] = &DDStart[DIR_00P * size_Mat];
+ D.f[DIR_00M] = &DDStart[DIR_00M * size_Mat];
+ D.f[DIR_PP0] = &DDStart[DIR_PP0 * size_Mat];
+ D.f[DIR_MM0] = &DDStart[DIR_MM0 * size_Mat];
+ D.f[DIR_PM0] = &DDStart[DIR_PM0 * size_Mat];
+ D.f[DIR_MP0] = &DDStart[DIR_MP0 * size_Mat];
+ D.f[DIR_P0P] = &DDStart[DIR_P0P * size_Mat];
+ D.f[DIR_M0M] = &DDStart[DIR_M0M * size_Mat];
+ D.f[DIR_P0M] = &DDStart[DIR_P0M * size_Mat];
+ D.f[DIR_M0P] = &DDStart[DIR_M0P * size_Mat];
+ D.f[DIR_0PP] = &DDStart[DIR_0PP * size_Mat];
+ D.f[DIR_0MM] = &DDStart[DIR_0MM * size_Mat];
+ D.f[DIR_0PM] = &DDStart[DIR_0PM * size_Mat];
+ D.f[DIR_0MP] = &DDStart[DIR_0MP * size_Mat];
+ D.f[DIR_000] = &DDStart[DIR_000 * size_Mat];
+ D.f[DIR_PPP] = &DDStart[DIR_PPP * size_Mat];
+ D.f[DIR_MMP] = &DDStart[DIR_MMP * size_Mat];
+ D.f[DIR_PMP] = &DDStart[DIR_PMP * size_Mat];
+ D.f[DIR_MPP] = &DDStart[DIR_MPP * size_Mat];
+ D.f[DIR_PPM] = &DDStart[DIR_PPM * size_Mat];
+ D.f[DIR_MMM] = &DDStart[DIR_MMM * size_Mat];
+ D.f[DIR_PMM] = &DDStart[DIR_PMM * size_Mat];
+ D.f[DIR_MPM] = &DDStart[DIR_MPM * size_Mat];
}
else
{
- D.f[DIR_M00] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_P00] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0M0] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0P0] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00M] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00P] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_MM0] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_PP0] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_MP0] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_PM0] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_M0M] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_P0P] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_M0P] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_P0M] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0MM] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0PP] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0MP] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0PM] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_MMM] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_PPM] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_MPM] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_PMM] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_MMP] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_PPP] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_MPP] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_PMP] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_M00] = &DDStart[DIR_P00 * size_Mat];
+ D.f[DIR_P00] = &DDStart[DIR_M00 * size_Mat];
+ D.f[DIR_0M0] = &DDStart[DIR_0P0 * size_Mat];
+ D.f[DIR_0P0] = &DDStart[DIR_0M0 * size_Mat];
+ D.f[DIR_00M] = &DDStart[DIR_00P * size_Mat];
+ D.f[DIR_00P] = &DDStart[DIR_00M * size_Mat];
+ D.f[DIR_MM0] = &DDStart[DIR_PP0 * size_Mat];
+ D.f[DIR_PP0] = &DDStart[DIR_MM0 * size_Mat];
+ D.f[DIR_MP0] = &DDStart[DIR_PM0 * size_Mat];
+ D.f[DIR_PM0] = &DDStart[DIR_MP0 * size_Mat];
+ D.f[DIR_M0M] = &DDStart[DIR_P0P * size_Mat];
+ D.f[DIR_P0P] = &DDStart[DIR_M0M * size_Mat];
+ D.f[DIR_M0P] = &DDStart[DIR_P0M * size_Mat];
+ D.f[DIR_P0M] = &DDStart[DIR_M0P * size_Mat];
+ D.f[DIR_0MM] = &DDStart[DIR_0PP * size_Mat];
+ D.f[DIR_0PP] = &DDStart[DIR_0MM * size_Mat];
+ D.f[DIR_0MP] = &DDStart[DIR_0PM * size_Mat];
+ D.f[DIR_0PM] = &DDStart[DIR_0MP * size_Mat];
+ D.f[DIR_000] = &DDStart[DIR_000 * size_Mat];
+ D.f[DIR_MMM] = &DDStart[DIR_PPP * size_Mat];
+ D.f[DIR_PPM] = &DDStart[DIR_MMP * size_Mat];
+ D.f[DIR_MPM] = &DDStart[DIR_PMP * size_Mat];
+ D.f[DIR_PMM] = &DDStart[DIR_MPP * size_Mat];
+ D.f[DIR_MMP] = &DDStart[DIR_PPM * size_Mat];
+ D.f[DIR_PPP] = &DDStart[DIR_MMM * size_Mat];
+ D.f[DIR_MPP] = &DDStart[DIR_PMM * size_Mat];
+ D.f[DIR_PMP] = &DDStart[DIR_MPM * size_Mat];
}
////////////////////////////////////////////////////////////////////////////////
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantAll4/CumulantAll4CompSP27.cu b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantAll4/CumulantAll4CompSP27.cu
index 9a84df86e41b3fdff75c2ebf580813afc5ee3feb..1518dcc209de1edf8a88dae72c1f10c3d4666610 100644
--- a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantAll4/CumulantAll4CompSP27.cu
+++ b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantAll4/CumulantAll4CompSP27.cu
@@ -1,8 +1,8 @@
#include "CumulantAll4CompSP27.h"
#include "CumulantAll4CompSP27_Device.cuh"
-
#include "Parameter/Parameter.h"
+#include "cuda/CudaGrid.h"
std::shared_ptr<CumulantAll4CompSP27> CumulantAll4CompSP27::getNewInstance(std::shared_ptr<Parameter> para, int level)
{
@@ -11,36 +11,21 @@ std::shared_ptr<CumulantAll4CompSP27> CumulantAll4CompSP27::getNewInstance(std::
void CumulantAll4CompSP27::run()
{
- int numberOfThreads = para->getParD(level)->numberofthreads;
- int size_Mat = para->getParD(level)->numberOfNodes;
-
- int Grid = (size_Mat / numberOfThreads) + 1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid / Grid1) + 1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 grid(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1);
-
- LB_Kernel_Cumulant_D3Q27All4 << < grid, threads >> >( para->getParD(level)->omega,
- para->getParD(level)->typeOfGridNode,
- para->getParD(level)->neighborX,
- para->getParD(level)->neighborY,
- para->getParD(level)->neighborZ,
- para->getParD(level)->distributions.f[0],
- size_Mat,
- level,
- para->getForcesDev(),
- para->getQuadricLimitersDev(),
- para->getParD(level)->isEvenTimestep);
- getLastCudaError("LB_Kernel_Cumulant_D3Q27All4 execution failed");
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(para->getParD(level)->numberofthreads, para->getParD(level)->numberOfNodes);
+
+ LB_Kernel_Cumulant_D3Q27All4 <<< grid.grid, grid.threads >>>(
+ para->getParD(level)->omega,
+ para->getParD(level)->typeOfGridNode,
+ para->getParD(level)->neighborX,
+ para->getParD(level)->neighborY,
+ para->getParD(level)->neighborZ,
+ para->getParD(level)->distributions.f[0],
+ para->getParD(level)->numberOfNodes,
+ level,
+ para->getForcesDev(),
+ para->getQuadricLimitersDev(),
+ para->getParD(level)->isEvenTimestep);
+ getLastCudaError("LB_Kernel_Cumulant_D3Q27All4 execution failed");
}
CumulantAll4CompSP27::CumulantAll4CompSP27(std::shared_ptr<Parameter> para, int level)
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantAll4/CumulantAll4CompSP27_Device.cu b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantAll4/CumulantAll4CompSP27_Device.cu
index 681dbff2ba37a1e0de56341b39cc2dec791f656b..3593b41c4c62c8a8b19719e22e9d65d6b5fd987d 100644
--- a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantAll4/CumulantAll4CompSP27_Device.cu
+++ b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantAll4/CumulantAll4CompSP27_Device.cu
@@ -42,63 +42,63 @@ __global__ void LB_Kernel_Cumulant_D3Q27All4( real omega,
Distributions27 D;
if (EvenOrOdd == true)
{
- D.f[DIR_P00] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_M00] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0P0] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0M0] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00P] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00M] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_PP0] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_MM0] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_PM0] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_MP0] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_P0P] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_M0M] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_P0M] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_M0P] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0PP] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0MM] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0PM] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0MP] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_PPP] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_MMP] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_PMP] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_MPP] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_PPM] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_MMM] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_PMM] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_MPM] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DDStart[DIR_P00 * size_Mat];
+ D.f[DIR_M00] = &DDStart[DIR_M00 * size_Mat];
+ D.f[DIR_0P0] = &DDStart[DIR_0P0 * size_Mat];
+ D.f[DIR_0M0] = &DDStart[DIR_0M0 * size_Mat];
+ D.f[DIR_00P] = &DDStart[DIR_00P * size_Mat];
+ D.f[DIR_00M] = &DDStart[DIR_00M * size_Mat];
+ D.f[DIR_PP0] = &DDStart[DIR_PP0 * size_Mat];
+ D.f[DIR_MM0] = &DDStart[DIR_MM0 * size_Mat];
+ D.f[DIR_PM0] = &DDStart[DIR_PM0 * size_Mat];
+ D.f[DIR_MP0] = &DDStart[DIR_MP0 * size_Mat];
+ D.f[DIR_P0P] = &DDStart[DIR_P0P * size_Mat];
+ D.f[DIR_M0M] = &DDStart[DIR_M0M * size_Mat];
+ D.f[DIR_P0M] = &DDStart[DIR_P0M * size_Mat];
+ D.f[DIR_M0P] = &DDStart[DIR_M0P * size_Mat];
+ D.f[DIR_0PP] = &DDStart[DIR_0PP * size_Mat];
+ D.f[DIR_0MM] = &DDStart[DIR_0MM * size_Mat];
+ D.f[DIR_0PM] = &DDStart[DIR_0PM * size_Mat];
+ D.f[DIR_0MP] = &DDStart[DIR_0MP * size_Mat];
+ D.f[DIR_000] = &DDStart[DIR_000 * size_Mat];
+ D.f[DIR_PPP] = &DDStart[DIR_PPP * size_Mat];
+ D.f[DIR_MMP] = &DDStart[DIR_MMP * size_Mat];
+ D.f[DIR_PMP] = &DDStart[DIR_PMP * size_Mat];
+ D.f[DIR_MPP] = &DDStart[DIR_MPP * size_Mat];
+ D.f[DIR_PPM] = &DDStart[DIR_PPM * size_Mat];
+ D.f[DIR_MMM] = &DDStart[DIR_MMM * size_Mat];
+ D.f[DIR_PMM] = &DDStart[DIR_PMM * size_Mat];
+ D.f[DIR_MPM] = &DDStart[DIR_MPM * size_Mat];
}
else
{
- D.f[DIR_M00] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_P00] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0M0] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0P0] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00M] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00P] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_MM0] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_PP0] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_MP0] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_PM0] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_M0M] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_P0P] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_M0P] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_P0M] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0MM] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0PP] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0MP] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0PM] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_MMM] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_PPM] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_MPM] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_PMM] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_MMP] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_PPP] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_MPP] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_PMP] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_M00] = &DDStart[DIR_P00 * size_Mat];
+ D.f[DIR_P00] = &DDStart[DIR_M00 * size_Mat];
+ D.f[DIR_0M0] = &DDStart[DIR_0P0 * size_Mat];
+ D.f[DIR_0P0] = &DDStart[DIR_0M0 * size_Mat];
+ D.f[DIR_00M] = &DDStart[DIR_00P * size_Mat];
+ D.f[DIR_00P] = &DDStart[DIR_00M * size_Mat];
+ D.f[DIR_MM0] = &DDStart[DIR_PP0 * size_Mat];
+ D.f[DIR_PP0] = &DDStart[DIR_MM0 * size_Mat];
+ D.f[DIR_MP0] = &DDStart[DIR_PM0 * size_Mat];
+ D.f[DIR_PM0] = &DDStart[DIR_MP0 * size_Mat];
+ D.f[DIR_M0M] = &DDStart[DIR_P0P * size_Mat];
+ D.f[DIR_P0P] = &DDStart[DIR_M0M * size_Mat];
+ D.f[DIR_M0P] = &DDStart[DIR_P0M * size_Mat];
+ D.f[DIR_P0M] = &DDStart[DIR_M0P * size_Mat];
+ D.f[DIR_0MM] = &DDStart[DIR_0PP * size_Mat];
+ D.f[DIR_0PP] = &DDStart[DIR_0MM * size_Mat];
+ D.f[DIR_0MP] = &DDStart[DIR_0PM * size_Mat];
+ D.f[DIR_0PM] = &DDStart[DIR_0MP * size_Mat];
+ D.f[DIR_000] = &DDStart[DIR_000 * size_Mat];
+ D.f[DIR_MMM] = &DDStart[DIR_PPP * size_Mat];
+ D.f[DIR_PPM] = &DDStart[DIR_MMP * size_Mat];
+ D.f[DIR_MPM] = &DDStart[DIR_PMP * size_Mat];
+ D.f[DIR_PMM] = &DDStart[DIR_MPP * size_Mat];
+ D.f[DIR_MMP] = &DDStart[DIR_PPM * size_Mat];
+ D.f[DIR_PPP] = &DDStart[DIR_MMM * size_Mat];
+ D.f[DIR_MPP] = &DDStart[DIR_PMM * size_Mat];
+ D.f[DIR_PMP] = &DDStart[DIR_MPM * size_Mat];
}
////////////////////////////////////////////////////////////////////////////////
@@ -160,33 +160,33 @@ __global__ void LB_Kernel_Cumulant_D3Q27All4( real omega,
//unsigned int kbsw = neighborZ[ksw];
//////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
//////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
- real mfcbb = (D.f[DIR_P00])[k];//[ke ];// + c2over27 ;(D.f[DIR_P00 ])[k ];//ke
- real mfabb = (D.f[DIR_M00])[kw];//[kw ];// + c2over27 ;(D.f[DIR_M00 ])[kw ];
- real mfbcb = (D.f[DIR_0P0])[k];//[kn ];// + c2over27 ;(D.f[DIR_0P0 ])[k ];//kn
- real mfbab = (D.f[DIR_0M0])[ks];//[ks ];// + c2over27 ;(D.f[DIR_0M0 ])[ks ];
- real mfbbc = (D.f[DIR_00P])[k];//[kt ];// + c2over27 ;(D.f[DIR_00P ])[k ];//kt
- real mfbba = (D.f[DIR_00M])[kb];//[kb ];// + c2over27 ;(D.f[DIR_00M ])[kb ];
- real mfccb = (D.f[DIR_PP0])[k];//[kne ];// + c1over54 ;(D.f[DIR_PP0 ])[k ];//kne
- real mfaab = (D.f[DIR_MM0])[ksw];//[ksw ];// + c1over54 ;(D.f[DIR_MM0 ])[ksw];
- real mfcab = (D.f[DIR_PM0])[ks];//[kse ];// + c1over54 ;(D.f[DIR_PM0 ])[ks ];//kse
- real mfacb = (D.f[DIR_MP0])[kw];//[knw ];// + c1over54 ;(D.f[DIR_MP0 ])[kw ];//knw
- real mfcbc = (D.f[DIR_P0P])[k];//[kte ];// + c1over54 ;(D.f[DIR_P0P ])[k ];//kte
- real mfaba = (D.f[DIR_M0M])[kbw];//[kbw ];// + c1over54 ;(D.f[DIR_M0M ])[kbw];
- real mfcba = (D.f[DIR_P0M])[kb];//[kbe ];// + c1over54 ;(D.f[DIR_P0M ])[kb ];//kbe
- real mfabc = (D.f[DIR_M0P])[kw];//[ktw ];// + c1over54 ;(D.f[DIR_M0P ])[kw ];//ktw
- real mfbcc = (D.f[DIR_0PP])[k];//[ktn ];// + c1over54 ;(D.f[DIR_0PP ])[k ];//ktn
- real mfbaa = (D.f[DIR_0MM])[kbs];//[kbs ];// + c1over54 ;(D.f[DIR_0MM ])[kbs];
- real mfbca = (D.f[DIR_0PM])[kb];//[kbn ];// + c1over54 ;(D.f[DIR_0PM ])[kb ];//kbn
- real mfbac = (D.f[DIR_0MP])[ks];//[kts ];// + c1over54 ;(D.f[DIR_0MP ])[ks ];//kts
+ real mfcbb = (D.f[DIR_P00])[k];//[ke ];// + c2over27 ;(D.f[DIR_P00])[k ];//ke
+ real mfabb = (D.f[DIR_M00])[kw];//[kw ];// + c2over27 ;(D.f[DIR_M00])[kw ];
+ real mfbcb = (D.f[DIR_0P0])[k];//[kn ];// + c2over27 ;(D.f[DIR_0P0])[k ];//kn
+ real mfbab = (D.f[DIR_0M0])[ks];//[ks ];// + c2over27 ;(D.f[DIR_0M0])[ks ];
+ real mfbbc = (D.f[DIR_00P])[k];//[kt ];// + c2over27 ;(D.f[DIR_00P])[k ];//kt
+ real mfbba = (D.f[DIR_00M])[kb];//[kb ];// + c2over27 ;(D.f[DIR_00M])[kb ];
+ real mfccb = (D.f[DIR_PP0])[k];//[kne ];// + c1over54 ;(D.f[DIR_PP0])[k ];//kne
+ real mfaab = (D.f[DIR_MM0])[ksw];//[ksw ];// + c1over54 ;(D.f[DIR_MM0])[ksw];
+ real mfcab = (D.f[DIR_PM0])[ks];//[kse ];// + c1over54 ;(D.f[DIR_PM0])[ks ];//kse
+ real mfacb = (D.f[DIR_MP0])[kw];//[knw ];// + c1over54 ;(D.f[DIR_MP0])[kw ];//knw
+ real mfcbc = (D.f[DIR_P0P])[k];//[kte ];// + c1over54 ;(D.f[DIR_P0P])[k ];//kte
+ real mfaba = (D.f[DIR_M0M])[kbw];//[kbw ];// + c1over54 ;(D.f[DIR_M0M])[kbw];
+ real mfcba = (D.f[DIR_P0M])[kb];//[kbe ];// + c1over54 ;(D.f[DIR_P0M])[kb ];//kbe
+ real mfabc = (D.f[DIR_M0P])[kw];//[ktw ];// + c1over54 ;(D.f[DIR_M0P])[kw ];//ktw
+ real mfbcc = (D.f[DIR_0PP])[k];//[ktn ];// + c1over54 ;(D.f[DIR_0PP])[k ];//ktn
+ real mfbaa = (D.f[DIR_0MM])[kbs];//[kbs ];// + c1over54 ;(D.f[DIR_0MM])[kbs];
+ real mfbca = (D.f[DIR_0PM])[kb];//[kbn ];// + c1over54 ;(D.f[DIR_0PM])[kb ];//kbn
+ real mfbac = (D.f[DIR_0MP])[ks];//[kts ];// + c1over54 ;(D.f[DIR_0MP])[ks ];//kts
real mfbbb = (D.f[DIR_000])[k];//[kzero];// + c8over27 ;(D.f[DIR_000])[k ];//kzero
- real mfccc = (D.f[DIR_PPP])[k];//[ktne ];// + c1over216;(D.f[DIR_PPP ])[k ];//ktne
- real mfaac = (D.f[DIR_MMP])[ksw];//[ktsw ];// + c1over216;(D.f[DIR_MMP ])[ksw];//ktsw
- real mfcac = (D.f[DIR_PMP])[ks];//[ktse ];// + c1over216;(D.f[DIR_PMP ])[ks ];//ktse
- real mfacc = (D.f[DIR_MPP])[kw];//[ktnw ];// + c1over216;(D.f[DIR_MPP ])[kw ];//ktnw
- real mfcca = (D.f[DIR_PPM])[kb];//[kbne ];// + c1over216;(D.f[DIR_PPM ])[kb ];//kbne
- real mfaaa = (D.f[DIR_MMM])[kbsw];//[kbsw ];// + c1over216;(D.f[DIR_MMM ])[kbsw];
- real mfcaa = (D.f[DIR_PMM])[kbs];//[kbse ];// + c1over216;(D.f[DIR_PMM ])[kbs];//kbse
- real mfaca = (D.f[DIR_MPM])[kbw];//[kbnw ];// + c1over216;(D.f[DIR_MPM ])[kbw];//kbnw
+ real mfccc = (D.f[DIR_PPP])[k];//[ktne ];// + c1over216;(D.f[DIR_PPP])[k ];//ktne
+ real mfaac = (D.f[DIR_MMP])[ksw];//[ktsw ];// + c1over216;(D.f[DIR_MMP])[ksw];//ktsw
+ real mfcac = (D.f[DIR_PMP])[ks];//[ktse ];// + c1over216;(D.f[DIR_PMP])[ks ];//ktse
+ real mfacc = (D.f[DIR_MPP])[kw];//[ktnw ];// + c1over216;(D.f[DIR_MPP])[kw ];//ktnw
+ real mfcca = (D.f[DIR_PPM])[kb];//[kbne ];// + c1over216;(D.f[DIR_PPM])[kb ];//kbne
+ real mfaaa = (D.f[DIR_MMM])[kbsw];//[kbsw ];// + c1over216;(D.f[DIR_MMM])[kbsw];
+ real mfcaa = (D.f[DIR_PMM])[kbs];//[kbse ];// + c1over216;(D.f[DIR_PMM])[kbs];//kbse
+ real mfaca = (D.f[DIR_MPM])[kbw];//[kbnw ];// + c1over216;(D.f[DIR_MPM])[kbw];//kbnw
////////////////////////////////////////////////////////////////////////////////////
real drho = ((((mfccc + mfaaa) + (mfaca + mfcac)) + ((mfacc + mfcaa) + (mfaac + mfcca))) +
(((mfbac + mfbca) + (mfbaa + mfbcc)) + ((mfabc + mfcba) + (mfaba + mfcbc)) + ((mfacb + mfcab) + (mfaab + mfccb))) +
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK15/CumulantK15Comp.cu b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK15/CumulantK15Comp.cu
index 1b6ba1a2278b68f085a4b7df699b7ca230811f39..5a480e5d9c97126e491655b4bbe2aeefef3e7161 100644
--- a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK15/CumulantK15Comp.cu
+++ b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK15/CumulantK15Comp.cu
@@ -12,7 +12,7 @@ std::shared_ptr<CumulantK15Comp> CumulantK15Comp::getNewInstance(std::shared_ptr
void CumulantK15Comp::run()
{
int numberOfThreads = para->getParD(level)->numberofthreads;
- int size_Mat = para->getParD(level)->numberOfNodes;
+ int size_Mat = (int)para->getParD(level)->numberOfNodes;
int Grid = (size_Mat / numberOfThreads) + 1;
int Grid1, Grid2;
@@ -29,16 +29,17 @@ void CumulantK15Comp::run()
dim3 grid(Grid1, Grid2, 1);
dim3 threads(numberOfThreads, 1, 1);
- LB_Kernel_CumulantK15Comp <<< grid, threads >>>(para->getParD(level)->omega,
- para->getParD(level)->typeOfGridNode,
- para->getParD(level)->neighborX,
- para->getParD(level)->neighborY,
- para->getParD(level)->neighborZ,
- para->getParD(level)->distributions.f[0],
- size_Mat,
- level,
- para->getForcesDev(),
- para->getParD(level)->isEvenTimestep);
+ LB_Kernel_CumulantK15Comp <<< grid, threads >>>(
+ para->getParD(level)->omega,
+ para->getParD(level)->typeOfGridNode,
+ para->getParD(level)->neighborX,
+ para->getParD(level)->neighborY,
+ para->getParD(level)->neighborZ,
+ para->getParD(level)->distributions.f[0],
+ para->getParD(level)->numberOfNodes,
+ level,
+ para->getForcesDev(),
+ para->getParD(level)->isEvenTimestep);
getLastCudaError("LB_Kernel_CumulantK15Comp execution failed");
}
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK15/CumulantK15Comp_Device.cu b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK15/CumulantK15Comp_Device.cu
index 93d57d6c9871d66537f25b9188467d46e3b3d05c..f7fb1f0a6441cfc6f38ad9684fd5bc8dd1be7135 100644
--- a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK15/CumulantK15Comp_Device.cu
+++ b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK15/CumulantK15Comp_Device.cu
@@ -39,63 +39,63 @@ __global__ void LB_Kernel_CumulantK15Comp(real omega,
Distributions27 D;
if (EvenOrOdd == true)
{
- D.f[DIR_P00] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_M00] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0P0] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0M0] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00P] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00M] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_PP0] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_MM0] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_PM0] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_MP0] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_P0P] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_M0M] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_P0M] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_M0P] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0PP] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0MM] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0PM] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0MP] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_PPP] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_MMP] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_PMP] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_MPP] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_PPM] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_MMM] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_PMM] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_MPM] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DDStart[DIR_P00 * size_Mat];
+ D.f[DIR_M00] = &DDStart[DIR_M00 * size_Mat];
+ D.f[DIR_0P0] = &DDStart[DIR_0P0 * size_Mat];
+ D.f[DIR_0M0] = &DDStart[DIR_0M0 * size_Mat];
+ D.f[DIR_00P] = &DDStart[DIR_00P * size_Mat];
+ D.f[DIR_00M] = &DDStart[DIR_00M * size_Mat];
+ D.f[DIR_PP0] = &DDStart[DIR_PP0 * size_Mat];
+ D.f[DIR_MM0] = &DDStart[DIR_MM0 * size_Mat];
+ D.f[DIR_PM0] = &DDStart[DIR_PM0 * size_Mat];
+ D.f[DIR_MP0] = &DDStart[DIR_MP0 * size_Mat];
+ D.f[DIR_P0P] = &DDStart[DIR_P0P * size_Mat];
+ D.f[DIR_M0M] = &DDStart[DIR_M0M * size_Mat];
+ D.f[DIR_P0M] = &DDStart[DIR_P0M * size_Mat];
+ D.f[DIR_M0P] = &DDStart[DIR_M0P * size_Mat];
+ D.f[DIR_0PP] = &DDStart[DIR_0PP * size_Mat];
+ D.f[DIR_0MM] = &DDStart[DIR_0MM * size_Mat];
+ D.f[DIR_0PM] = &DDStart[DIR_0PM * size_Mat];
+ D.f[DIR_0MP] = &DDStart[DIR_0MP * size_Mat];
+ D.f[DIR_000] = &DDStart[DIR_000 * size_Mat];
+ D.f[DIR_PPP] = &DDStart[DIR_PPP * size_Mat];
+ D.f[DIR_MMP] = &DDStart[DIR_MMP * size_Mat];
+ D.f[DIR_PMP] = &DDStart[DIR_PMP * size_Mat];
+ D.f[DIR_MPP] = &DDStart[DIR_MPP * size_Mat];
+ D.f[DIR_PPM] = &DDStart[DIR_PPM * size_Mat];
+ D.f[DIR_MMM] = &DDStart[DIR_MMM * size_Mat];
+ D.f[DIR_PMM] = &DDStart[DIR_PMM * size_Mat];
+ D.f[DIR_MPM] = &DDStart[DIR_MPM * size_Mat];
}
else
{
- D.f[DIR_M00] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_P00] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0M0] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0P0] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00M] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00P] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_MM0] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_PP0] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_MP0] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_PM0] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_M0M] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_P0P] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_M0P] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_P0M] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0MM] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0PP] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0MP] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0PM] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_MMM] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_PPM] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_MPM] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_PMM] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_MMP] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_PPP] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_MPP] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_PMP] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_M00] = &DDStart[DIR_P00 * size_Mat];
+ D.f[DIR_P00] = &DDStart[DIR_M00 * size_Mat];
+ D.f[DIR_0M0] = &DDStart[DIR_0P0 * size_Mat];
+ D.f[DIR_0P0] = &DDStart[DIR_0M0 * size_Mat];
+ D.f[DIR_00M] = &DDStart[DIR_00P * size_Mat];
+ D.f[DIR_00P] = &DDStart[DIR_00M * size_Mat];
+ D.f[DIR_MM0] = &DDStart[DIR_PP0 * size_Mat];
+ D.f[DIR_PP0] = &DDStart[DIR_MM0 * size_Mat];
+ D.f[DIR_MP0] = &DDStart[DIR_PM0 * size_Mat];
+ D.f[DIR_PM0] = &DDStart[DIR_MP0 * size_Mat];
+ D.f[DIR_M0M] = &DDStart[DIR_P0P * size_Mat];
+ D.f[DIR_P0P] = &DDStart[DIR_M0M * size_Mat];
+ D.f[DIR_M0P] = &DDStart[DIR_P0M * size_Mat];
+ D.f[DIR_P0M] = &DDStart[DIR_M0P * size_Mat];
+ D.f[DIR_0MM] = &DDStart[DIR_0PP * size_Mat];
+ D.f[DIR_0PP] = &DDStart[DIR_0MM * size_Mat];
+ D.f[DIR_0MP] = &DDStart[DIR_0PM * size_Mat];
+ D.f[DIR_0PM] = &DDStart[DIR_0MP * size_Mat];
+ D.f[DIR_000] = &DDStart[DIR_000 * size_Mat];
+ D.f[DIR_MMM] = &DDStart[DIR_PPP * size_Mat];
+ D.f[DIR_PPM] = &DDStart[DIR_MMP * size_Mat];
+ D.f[DIR_MPM] = &DDStart[DIR_PMP * size_Mat];
+ D.f[DIR_PMM] = &DDStart[DIR_MPP * size_Mat];
+ D.f[DIR_MMP] = &DDStart[DIR_PPM * size_Mat];
+ D.f[DIR_PPP] = &DDStart[DIR_MMM * size_Mat];
+ D.f[DIR_MPP] = &DDStart[DIR_PMM * size_Mat];
+ D.f[DIR_PMP] = &DDStart[DIR_MPM * size_Mat];
}
////////////////////////////////////////////////////////////////////////////////
@@ -156,33 +156,33 @@ __global__ void LB_Kernel_CumulantK15Comp(real omega,
//unsigned int ktne = k;
//unsigned int kbsw = neighborZ[ksw];
//////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
- real mfcbb = (D.f[DIR_P00])[k];//[ke ];// + c2over27 ;(D.f[DIR_P00 ])[k ];//ke
- real mfabb = (D.f[DIR_M00])[kw];//[kw ];// + c2over27 ;(D.f[DIR_M00 ])[kw ];
- real mfbcb = (D.f[DIR_0P0])[k];//[kn ];// + c2over27 ;(D.f[DIR_0P0 ])[k ];//kn
- real mfbab = (D.f[DIR_0M0])[ks];//[ks ];// + c2over27 ;(D.f[DIR_0M0 ])[ks ];
- real mfbbc = (D.f[DIR_00P])[k];//[kt ];// + c2over27 ;(D.f[DIR_00P ])[k ];//kt
- real mfbba = (D.f[DIR_00M])[kb];//[kb ];// + c2over27 ;(D.f[DIR_00M ])[kb ];
- real mfccb = (D.f[DIR_PP0])[k];//[kne ];// + c1over54 ;(D.f[DIR_PP0 ])[k ];//kne
- real mfaab = (D.f[DIR_MM0])[ksw];//[ksw ];// + c1over54 ;(D.f[DIR_MM0 ])[ksw];
- real mfcab = (D.f[DIR_PM0])[ks];//[kse ];// + c1over54 ;(D.f[DIR_PM0 ])[ks ];//kse
- real mfacb = (D.f[DIR_MP0])[kw];//[knw ];// + c1over54 ;(D.f[DIR_MP0 ])[kw ];//knw
- real mfcbc = (D.f[DIR_P0P])[k];//[kte ];// + c1over54 ;(D.f[DIR_P0P ])[k ];//kte
- real mfaba = (D.f[DIR_M0M])[kbw];//[kbw ];// + c1over54 ;(D.f[DIR_M0M ])[kbw];
- real mfcba = (D.f[DIR_P0M])[kb];//[kbe ];// + c1over54 ;(D.f[DIR_P0M ])[kb ];//kbe
- real mfabc = (D.f[DIR_M0P])[kw];//[ktw ];// + c1over54 ;(D.f[DIR_M0P ])[kw ];//ktw
- real mfbcc = (D.f[DIR_0PP])[k];//[ktn ];// + c1over54 ;(D.f[DIR_0PP ])[k ];//ktn
- real mfbaa = (D.f[DIR_0MM])[kbs];//[kbs ];// + c1over54 ;(D.f[DIR_0MM ])[kbs];
- real mfbca = (D.f[DIR_0PM])[kb];//[kbn ];// + c1over54 ;(D.f[DIR_0PM ])[kb ];//kbn
- real mfbac = (D.f[DIR_0MP])[ks];//[kts ];// + c1over54 ;(D.f[DIR_0MP ])[ks ];//kts
+ real mfcbb = (D.f[DIR_P00])[k];//[ke ];// + c2over27 ;(D.f[DIR_P00])[k ];//ke
+ real mfabb = (D.f[DIR_M00])[kw];//[kw ];// + c2over27 ;(D.f[DIR_M00])[kw ];
+ real mfbcb = (D.f[DIR_0P0])[k];//[kn ];// + c2over27 ;(D.f[DIR_0P0])[k ];//kn
+ real mfbab = (D.f[DIR_0M0])[ks];//[ks ];// + c2over27 ;(D.f[DIR_0M0])[ks ];
+ real mfbbc = (D.f[DIR_00P])[k];//[kt ];// + c2over27 ;(D.f[DIR_00P])[k ];//kt
+ real mfbba = (D.f[DIR_00M])[kb];//[kb ];// + c2over27 ;(D.f[DIR_00M])[kb ];
+ real mfccb = (D.f[DIR_PP0])[k];//[kne ];// + c1over54 ;(D.f[DIR_PP0])[k ];//kne
+ real mfaab = (D.f[DIR_MM0])[ksw];//[ksw ];// + c1over54 ;(D.f[DIR_MM0])[ksw];
+ real mfcab = (D.f[DIR_PM0])[ks];//[kse ];// + c1over54 ;(D.f[DIR_PM0])[ks ];//kse
+ real mfacb = (D.f[DIR_MP0])[kw];//[knw ];// + c1over54 ;(D.f[DIR_MP0])[kw ];//knw
+ real mfcbc = (D.f[DIR_P0P])[k];//[kte ];// + c1over54 ;(D.f[DIR_P0P])[k ];//kte
+ real mfaba = (D.f[DIR_M0M])[kbw];//[kbw ];// + c1over54 ;(D.f[DIR_M0M])[kbw];
+ real mfcba = (D.f[DIR_P0M])[kb];//[kbe ];// + c1over54 ;(D.f[DIR_P0M])[kb ];//kbe
+ real mfabc = (D.f[DIR_M0P])[kw];//[ktw ];// + c1over54 ;(D.f[DIR_M0P])[kw ];//ktw
+ real mfbcc = (D.f[DIR_0PP])[k];//[ktn ];// + c1over54 ;(D.f[DIR_0PP])[k ];//ktn
+ real mfbaa = (D.f[DIR_0MM])[kbs];//[kbs ];// + c1over54 ;(D.f[DIR_0MM])[kbs];
+ real mfbca = (D.f[DIR_0PM])[kb];//[kbn ];// + c1over54 ;(D.f[DIR_0PM])[kb ];//kbn
+ real mfbac = (D.f[DIR_0MP])[ks];//[kts ];// + c1over54 ;(D.f[DIR_0MP])[ks ];//kts
real mfbbb = (D.f[DIR_000])[k];//[kzero];// + c8over27 ;(D.f[DIR_000])[k ];//kzero
- real mfccc = (D.f[DIR_PPP])[k];//[ktne ];// + c1over216;(D.f[DIR_PPP ])[k ];//ktne
- real mfaac = (D.f[DIR_MMP])[ksw];//[ktsw ];// + c1over216;(D.f[DIR_MMP ])[ksw];//ktsw
- real mfcac = (D.f[DIR_PMP])[ks];//[ktse ];// + c1over216;(D.f[DIR_PMP ])[ks ];//ktse
- real mfacc = (D.f[DIR_MPP])[kw];//[ktnw ];// + c1over216;(D.f[DIR_MPP ])[kw ];//ktnw
- real mfcca = (D.f[DIR_PPM])[kb];//[kbne ];// + c1over216;(D.f[DIR_PPM ])[kb ];//kbne
- real mfaaa = (D.f[DIR_MMM])[kbsw];//[kbsw ];// + c1over216;(D.f[DIR_MMM ])[kbsw];
- real mfcaa = (D.f[DIR_PMM])[kbs];//[kbse ];// + c1over216;(D.f[DIR_PMM ])[kbs];//kbse
- real mfaca = (D.f[DIR_MPM])[kbw];//[kbnw ];// + c1over216;(D.f[DIR_MPM ])[kbw];//kbnw
+ real mfccc = (D.f[DIR_PPP])[k];//[ktne ];// + c1over216;(D.f[DIR_PPP])[k ];//ktne
+ real mfaac = (D.f[DIR_MMP])[ksw];//[ktsw ];// + c1over216;(D.f[DIR_MMP])[ksw];//ktsw
+ real mfcac = (D.f[DIR_PMP])[ks];//[ktse ];// + c1over216;(D.f[DIR_PMP])[ks ];//ktse
+ real mfacc = (D.f[DIR_MPP])[kw];//[ktnw ];// + c1over216;(D.f[DIR_MPP])[kw ];//ktnw
+ real mfcca = (D.f[DIR_PPM])[kb];//[kbne ];// + c1over216;(D.f[DIR_PPM])[kb ];//kbne
+ real mfaaa = (D.f[DIR_MMM])[kbsw];//[kbsw ];// + c1over216;(D.f[DIR_MMM])[kbsw];
+ real mfcaa = (D.f[DIR_PMM])[kbs];//[kbse ];// + c1over216;(D.f[DIR_PMM])[kbs];//kbse
+ real mfaca = (D.f[DIR_MPM])[kbw];//[kbnw ];// + c1over216;(D.f[DIR_MPM])[kbw];//kbnw
////////////////////////////////////////////////////////////////////////////////////
real drho = ((((mfccc + mfaaa) + (mfaca + mfcac)) + ((mfacc + mfcaa) + (mfaac + mfcca))) +
(((mfbac + mfbca) + (mfbaa + mfbcc)) + ((mfabc + mfcba) + (mfaba + mfcbc)) + ((mfacb + mfcab) + (mfaab + mfccb))) +
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK15Bulk/CumulantK15BulkComp.cu b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK15Bulk/CumulantK15BulkComp.cu
index 188984d001f89d72c967dd6390ca10ae5d2eab32..51876f30b8c8e37d8cb3355edde5dcf2b04675d0 100644
--- a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK15Bulk/CumulantK15BulkComp.cu
+++ b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK15Bulk/CumulantK15BulkComp.cu
@@ -2,6 +2,7 @@
#include "CumulantK15BulkComp_Device.cuh"
#include "Parameter/Parameter.h"
+#include "cuda/CudaGrid.h"
std::shared_ptr<CumulantK15BulkComp> CumulantK15BulkComp::getNewInstance(std::shared_ptr<Parameter> para, int level)
{
@@ -10,35 +11,20 @@ std::shared_ptr<CumulantK15BulkComp> CumulantK15BulkComp::getNewInstance(std::sh
void CumulantK15BulkComp::run()
{
- int size_Mat = para->getParD(level)->numberOfNodes;
- int numberOfThreads = para->getParD(level)->numberofthreads;
-
- int Grid = (size_Mat / numberOfThreads) + 1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid / Grid1) + 1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 grid(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1);
-
- LB_Kernel_CumulantK15BulkComp <<< grid, threads >>>(para->getParD(level)->omega,
- para->getParD(level)->typeOfGridNode,
- para->getParD(level)->neighborX,
- para->getParD(level)->neighborY,
- para->getParD(level)->neighborZ,
- para->getParD(level)->distributions.f[0],
- para->getParD(level)->numberOfNodes,
- level,
- para->getForcesDev(),
- para->getParD(level)->isEvenTimestep);
- getLastCudaError("LB_Kernel_CumulantK15BulkComp execution failed");
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(para->getParD(level)->numberofthreads, para->getParD(level)->numberOfNodes);
+
+ LB_Kernel_CumulantK15BulkComp <<< grid.grid, grid.threads >>>(
+ para->getParD(level)->omega,
+ para->getParD(level)->typeOfGridNode,
+ para->getParD(level)->neighborX,
+ para->getParD(level)->neighborY,
+ para->getParD(level)->neighborZ,
+ para->getParD(level)->distributions.f[0],
+ para->getParD(level)->numberOfNodes,
+ level,
+ para->getForcesDev(),
+ para->getParD(level)->isEvenTimestep);
+ getLastCudaError("LB_Kernel_CumulantK15BulkComp execution failed");
}
CumulantK15BulkComp::CumulantK15BulkComp(std::shared_ptr<Parameter> para, int level)
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK15Bulk/CumulantK15BulkComp_Device.cu b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK15Bulk/CumulantK15BulkComp_Device.cu
index d2a2f61df902cfd7c5ef52b09f8e7738a108615e..085775d324bf65d783afdd745c06429d697c3788 100644
--- a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK15Bulk/CumulantK15BulkComp_Device.cu
+++ b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK15Bulk/CumulantK15BulkComp_Device.cu
@@ -39,63 +39,63 @@ __global__ void LB_Kernel_CumulantK15BulkComp(real omega,
Distributions27 D;
if (EvenOrOdd == true)
{
- D.f[DIR_P00] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_M00] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0P0] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0M0] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00P] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00M] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_PP0] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_MM0] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_PM0] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_MP0] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_P0P] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_M0M] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_P0M] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_M0P] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0PP] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0MM] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0PM] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0MP] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_PPP] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_MMP] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_PMP] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_MPP] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_PPM] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_MMM] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_PMM] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_MPM] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DDStart[DIR_P00 * size_Mat];
+ D.f[DIR_M00] = &DDStart[DIR_M00 * size_Mat];
+ D.f[DIR_0P0] = &DDStart[DIR_0P0 * size_Mat];
+ D.f[DIR_0M0] = &DDStart[DIR_0M0 * size_Mat];
+ D.f[DIR_00P] = &DDStart[DIR_00P * size_Mat];
+ D.f[DIR_00M] = &DDStart[DIR_00M * size_Mat];
+ D.f[DIR_PP0] = &DDStart[DIR_PP0 * size_Mat];
+ D.f[DIR_MM0] = &DDStart[DIR_MM0 * size_Mat];
+ D.f[DIR_PM0] = &DDStart[DIR_PM0 * size_Mat];
+ D.f[DIR_MP0] = &DDStart[DIR_MP0 * size_Mat];
+ D.f[DIR_P0P] = &DDStart[DIR_P0P * size_Mat];
+ D.f[DIR_M0M] = &DDStart[DIR_M0M * size_Mat];
+ D.f[DIR_P0M] = &DDStart[DIR_P0M * size_Mat];
+ D.f[DIR_M0P] = &DDStart[DIR_M0P * size_Mat];
+ D.f[DIR_0PP] = &DDStart[DIR_0PP * size_Mat];
+ D.f[DIR_0MM] = &DDStart[DIR_0MM * size_Mat];
+ D.f[DIR_0PM] = &DDStart[DIR_0PM * size_Mat];
+ D.f[DIR_0MP] = &DDStart[DIR_0MP * size_Mat];
+ D.f[DIR_000] = &DDStart[DIR_000 * size_Mat];
+ D.f[DIR_PPP] = &DDStart[DIR_PPP * size_Mat];
+ D.f[DIR_MMP] = &DDStart[DIR_MMP * size_Mat];
+ D.f[DIR_PMP] = &DDStart[DIR_PMP * size_Mat];
+ D.f[DIR_MPP] = &DDStart[DIR_MPP * size_Mat];
+ D.f[DIR_PPM] = &DDStart[DIR_PPM * size_Mat];
+ D.f[DIR_MMM] = &DDStart[DIR_MMM * size_Mat];
+ D.f[DIR_PMM] = &DDStart[DIR_PMM * size_Mat];
+ D.f[DIR_MPM] = &DDStart[DIR_MPM * size_Mat];
}
else
{
- D.f[DIR_M00] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_P00] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0M0] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0P0] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00M] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00P] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_MM0] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_PP0] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_MP0] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_PM0] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_M0M] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_P0P] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_M0P] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_P0M] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0MM] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0PP] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0MP] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0PM] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_MMM] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_PPM] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_MPM] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_PMM] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_MMP] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_PPP] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_MPP] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_PMP] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_M00] = &DDStart[DIR_P00 * size_Mat];
+ D.f[DIR_P00] = &DDStart[DIR_M00 * size_Mat];
+ D.f[DIR_0M0] = &DDStart[DIR_0P0 * size_Mat];
+ D.f[DIR_0P0] = &DDStart[DIR_0M0 * size_Mat];
+ D.f[DIR_00M] = &DDStart[DIR_00P * size_Mat];
+ D.f[DIR_00P] = &DDStart[DIR_00M * size_Mat];
+ D.f[DIR_MM0] = &DDStart[DIR_PP0 * size_Mat];
+ D.f[DIR_PP0] = &DDStart[DIR_MM0 * size_Mat];
+ D.f[DIR_MP0] = &DDStart[DIR_PM0 * size_Mat];
+ D.f[DIR_PM0] = &DDStart[DIR_MP0 * size_Mat];
+ D.f[DIR_M0M] = &DDStart[DIR_P0P * size_Mat];
+ D.f[DIR_P0P] = &DDStart[DIR_M0M * size_Mat];
+ D.f[DIR_M0P] = &DDStart[DIR_P0M * size_Mat];
+ D.f[DIR_P0M] = &DDStart[DIR_M0P * size_Mat];
+ D.f[DIR_0MM] = &DDStart[DIR_0PP * size_Mat];
+ D.f[DIR_0PP] = &DDStart[DIR_0MM * size_Mat];
+ D.f[DIR_0MP] = &DDStart[DIR_0PM * size_Mat];
+ D.f[DIR_0PM] = &DDStart[DIR_0MP * size_Mat];
+ D.f[DIR_000] = &DDStart[DIR_000 * size_Mat];
+ D.f[DIR_MMM] = &DDStart[DIR_PPP * size_Mat];
+ D.f[DIR_PPM] = &DDStart[DIR_MMP * size_Mat];
+ D.f[DIR_MPM] = &DDStart[DIR_PMP * size_Mat];
+ D.f[DIR_PMM] = &DDStart[DIR_MPP * size_Mat];
+ D.f[DIR_MMP] = &DDStart[DIR_PPM * size_Mat];
+ D.f[DIR_PPP] = &DDStart[DIR_MMM * size_Mat];
+ D.f[DIR_MPP] = &DDStart[DIR_PMM * size_Mat];
+ D.f[DIR_PMP] = &DDStart[DIR_MPM * size_Mat];
}
////////////////////////////////////////////////////////////////////////////////
@@ -156,33 +156,33 @@ __global__ void LB_Kernel_CumulantK15BulkComp(real omega,
//unsigned int ktne = k;
//unsigned int kbsw = neighborZ[ksw];
//////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
- real mfcbb = (D.f[DIR_P00])[k];//[ke ];// + c2over27 ;(D.f[DIR_P00 ])[k ];//ke
- real mfabb = (D.f[DIR_M00])[kw];//[kw ];// + c2over27 ;(D.f[DIR_M00 ])[kw ];
- real mfbcb = (D.f[DIR_0P0])[k];//[kn ];// + c2over27 ;(D.f[DIR_0P0 ])[k ];//kn
- real mfbab = (D.f[DIR_0M0])[ks];//[ks ];// + c2over27 ;(D.f[DIR_0M0 ])[ks ];
- real mfbbc = (D.f[DIR_00P])[k];//[kt ];// + c2over27 ;(D.f[DIR_00P ])[k ];//kt
- real mfbba = (D.f[DIR_00M])[kb];//[kb ];// + c2over27 ;(D.f[DIR_00M ])[kb ];
- real mfccb = (D.f[DIR_PP0])[k];//[kne ];// + c1over54 ;(D.f[DIR_PP0 ])[k ];//kne
- real mfaab = (D.f[DIR_MM0])[ksw];//[ksw ];// + c1over54 ;(D.f[DIR_MM0 ])[ksw];
- real mfcab = (D.f[DIR_PM0])[ks];//[kse ];// + c1over54 ;(D.f[DIR_PM0 ])[ks ];//kse
- real mfacb = (D.f[DIR_MP0])[kw];//[knw ];// + c1over54 ;(D.f[DIR_MP0 ])[kw ];//knw
- real mfcbc = (D.f[DIR_P0P])[k];//[kte ];// + c1over54 ;(D.f[DIR_P0P ])[k ];//kte
- real mfaba = (D.f[DIR_M0M])[kbw];//[kbw ];// + c1over54 ;(D.f[DIR_M0M ])[kbw];
- real mfcba = (D.f[DIR_P0M])[kb];//[kbe ];// + c1over54 ;(D.f[DIR_P0M ])[kb ];//kbe
- real mfabc = (D.f[DIR_M0P])[kw];//[ktw ];// + c1over54 ;(D.f[DIR_M0P ])[kw ];//ktw
- real mfbcc = (D.f[DIR_0PP])[k];//[ktn ];// + c1over54 ;(D.f[DIR_0PP ])[k ];//ktn
- real mfbaa = (D.f[DIR_0MM])[kbs];//[kbs ];// + c1over54 ;(D.f[DIR_0MM ])[kbs];
- real mfbca = (D.f[DIR_0PM])[kb];//[kbn ];// + c1over54 ;(D.f[DIR_0PM ])[kb ];//kbn
- real mfbac = (D.f[DIR_0MP])[ks];//[kts ];// + c1over54 ;(D.f[DIR_0MP ])[ks ];//kts
+ real mfcbb = (D.f[DIR_P00])[k];//[ke ];// + c2over27 ;(D.f[DIR_P00])[k ];//ke
+ real mfabb = (D.f[DIR_M00])[kw];//[kw ];// + c2over27 ;(D.f[DIR_M00])[kw ];
+ real mfbcb = (D.f[DIR_0P0])[k];//[kn ];// + c2over27 ;(D.f[DIR_0P0])[k ];//kn
+ real mfbab = (D.f[DIR_0M0])[ks];//[ks ];// + c2over27 ;(D.f[DIR_0M0])[ks ];
+ real mfbbc = (D.f[DIR_00P])[k];//[kt ];// + c2over27 ;(D.f[DIR_00P])[k ];//kt
+ real mfbba = (D.f[DIR_00M])[kb];//[kb ];// + c2over27 ;(D.f[DIR_00M])[kb ];
+ real mfccb = (D.f[DIR_PP0])[k];//[kne ];// + c1over54 ;(D.f[DIR_PP0])[k ];//kne
+ real mfaab = (D.f[DIR_MM0])[ksw];//[ksw ];// + c1over54 ;(D.f[DIR_MM0])[ksw];
+ real mfcab = (D.f[DIR_PM0])[ks];//[kse ];// + c1over54 ;(D.f[DIR_PM0])[ks ];//kse
+ real mfacb = (D.f[DIR_MP0])[kw];//[knw ];// + c1over54 ;(D.f[DIR_MP0])[kw ];//knw
+ real mfcbc = (D.f[DIR_P0P])[k];//[kte ];// + c1over54 ;(D.f[DIR_P0P])[k ];//kte
+ real mfaba = (D.f[DIR_M0M])[kbw];//[kbw ];// + c1over54 ;(D.f[DIR_M0M])[kbw];
+ real mfcba = (D.f[DIR_P0M])[kb];//[kbe ];// + c1over54 ;(D.f[DIR_P0M])[kb ];//kbe
+ real mfabc = (D.f[DIR_M0P])[kw];//[ktw ];// + c1over54 ;(D.f[DIR_M0P])[kw ];//ktw
+ real mfbcc = (D.f[DIR_0PP])[k];//[ktn ];// + c1over54 ;(D.f[DIR_0PP])[k ];//ktn
+ real mfbaa = (D.f[DIR_0MM])[kbs];//[kbs ];// + c1over54 ;(D.f[DIR_0MM])[kbs];
+ real mfbca = (D.f[DIR_0PM])[kb];//[kbn ];// + c1over54 ;(D.f[DIR_0PM])[kb ];//kbn
+ real mfbac = (D.f[DIR_0MP])[ks];//[kts ];// + c1over54 ;(D.f[DIR_0MP])[ks ];//kts
real mfbbb = (D.f[DIR_000])[k];//[kzero];// + c8over27 ;(D.f[DIR_000])[k ];//kzero
- real mfccc = (D.f[DIR_PPP])[k];//[ktne ];// + c1over216;(D.f[DIR_PPP ])[k ];//ktne
- real mfaac = (D.f[DIR_MMP])[ksw];//[ktsw ];// + c1over216;(D.f[DIR_MMP ])[ksw];//ktsw
- real mfcac = (D.f[DIR_PMP])[ks];//[ktse ];// + c1over216;(D.f[DIR_PMP ])[ks ];//ktse
- real mfacc = (D.f[DIR_MPP])[kw];//[ktnw ];// + c1over216;(D.f[DIR_MPP ])[kw ];//ktnw
- real mfcca = (D.f[DIR_PPM])[kb];//[kbne ];// + c1over216;(D.f[DIR_PPM ])[kb ];//kbne
- real mfaaa = (D.f[DIR_MMM])[kbsw];//[kbsw ];// + c1over216;(D.f[DIR_MMM ])[kbsw];
- real mfcaa = (D.f[DIR_PMM])[kbs];//[kbse ];// + c1over216;(D.f[DIR_PMM ])[kbs];//kbse
- real mfaca = (D.f[DIR_MPM])[kbw];//[kbnw ];// + c1over216;(D.f[DIR_MPM ])[kbw];//kbnw
+ real mfccc = (D.f[DIR_PPP])[k];//[ktne ];// + c1over216;(D.f[DIR_PPP])[k ];//ktne
+ real mfaac = (D.f[DIR_MMP])[ksw];//[ktsw ];// + c1over216;(D.f[DIR_MMP])[ksw];//ktsw
+ real mfcac = (D.f[DIR_PMP])[ks];//[ktse ];// + c1over216;(D.f[DIR_PMP])[ks ];//ktse
+ real mfacc = (D.f[DIR_MPP])[kw];//[ktnw ];// + c1over216;(D.f[DIR_MPP])[kw ];//ktnw
+ real mfcca = (D.f[DIR_PPM])[kb];//[kbne ];// + c1over216;(D.f[DIR_PPM])[kb ];//kbne
+ real mfaaa = (D.f[DIR_MMM])[kbsw];//[kbsw ];// + c1over216;(D.f[DIR_MMM])[kbsw];
+ real mfcaa = (D.f[DIR_PMM])[kbs];//[kbse ];// + c1over216;(D.f[DIR_PMM])[kbs];//kbse
+ real mfaca = (D.f[DIR_MPM])[kbw];//[kbnw ];// + c1over216;(D.f[DIR_MPM])[kbw];//kbnw
////////////////////////////////////////////////////////////////////////////////////
real drho = ((((mfccc + mfaaa) + (mfaca + mfcac)) + ((mfacc + mfcaa) + (mfaac + mfcca))) +
(((mfbac + mfbca) + (mfbaa + mfbcc)) + ((mfabc + mfcba) + (mfaba + mfcbc)) + ((mfacb + mfcab) + (mfaab + mfccb))) +
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK15Sponge/CumulantK15SpongeComp.cu b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK15Sponge/CumulantK15SpongeComp.cu
index d28c077031ff9125d1cbc1187def1d1d8fe4d6e8..613464125bafc572fe7951b8c372e3455ea5b21d 100644
--- a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK15Sponge/CumulantK15SpongeComp.cu
+++ b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK15Sponge/CumulantK15SpongeComp.cu
@@ -2,6 +2,7 @@
#include "CumulantK15SpongeComp_Device.cuh"
#include "Parameter/Parameter.h"
+#include "cuda/CudaGrid.h"
std::shared_ptr<CumulantK15SpongeComp> CumulantK15SpongeComp::getNewInstance(std::shared_ptr<Parameter> para, int level)
{
@@ -10,36 +11,21 @@ std::shared_ptr<CumulantK15SpongeComp> CumulantK15SpongeComp::getNewInstance(std
void CumulantK15SpongeComp::run()
{
- int size_Mat = para->getParD(level)->numberOfNodes;
- int numberOfThreads = para->getParD(level)->numberofthreads;
-
- int Grid = (size_Mat / numberOfThreads) + 1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid / Grid1) + 1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 grid(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1);
-
- LB_Kernel_CumulantK15SpongeComp <<< grid, threads >>>( para->getParD(level)->omega,
- para->getParD(level)->typeOfGridNode,
- para->getParD(level)->neighborX,
- para->getParD(level)->neighborY,
- para->getParD(level)->neighborZ,
- para->getParD(level)->coordinateX,
- para->getParD(level)->coordinateY,
- para->getParD(level)->coordinateZ,
- para->getParD(level)->distributions.f[0],
- para->getParD(level)->numberOfNodes,
- para->getParD(level)->isEvenTimestep);
- getLastCudaError("LB_Kernel_CumulantK15SpongeComp execution failed");
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(para->getParD(level)->numberofthreads, para->getParD(level)->numberOfNodes);
+
+ LB_Kernel_CumulantK15SpongeComp <<< grid.grid, grid.threads >>>(
+ para->getParD(level)->omega,
+ para->getParD(level)->typeOfGridNode,
+ para->getParD(level)->neighborX,
+ para->getParD(level)->neighborY,
+ para->getParD(level)->neighborZ,
+ para->getParD(level)->coordinateX,
+ para->getParD(level)->coordinateY,
+ para->getParD(level)->coordinateZ,
+ para->getParD(level)->distributions.f[0],
+ para->getParD(level)->numberOfNodes,
+ para->getParD(level)->isEvenTimestep);
+ getLastCudaError("LB_Kernel_CumulantK15SpongeComp execution failed");
}
CumulantK15SpongeComp::CumulantK15SpongeComp(std::shared_ptr<Parameter> para, int level)
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK15Sponge/CumulantK15SpongeComp_Device.cu b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK15Sponge/CumulantK15SpongeComp_Device.cu
index c2144d324aa3378e8fc9fc5b511bbed385b48a84..13788e65e70eb30803111a39a70d39682648a006 100644
--- a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK15Sponge/CumulantK15SpongeComp_Device.cu
+++ b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK15Sponge/CumulantK15SpongeComp_Device.cu
@@ -40,63 +40,63 @@ __global__ void LB_Kernel_CumulantK15SpongeComp(real omegaIn,
Distributions27 D;
if (EvenOrOdd == true)
{
- D.f[DIR_P00] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_M00] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0P0] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0M0] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00P] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00M] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_PP0] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_MM0] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_PM0] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_MP0] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_P0P] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_M0M] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_P0M] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_M0P] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0PP] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0MM] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0PM] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0MP] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_PPP] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_MMP] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_PMP] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_MPP] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_PPM] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_MMM] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_PMM] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_MPM] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DDStart[DIR_P00 * size_Mat];
+ D.f[DIR_M00] = &DDStart[DIR_M00 * size_Mat];
+ D.f[DIR_0P0] = &DDStart[DIR_0P0 * size_Mat];
+ D.f[DIR_0M0] = &DDStart[DIR_0M0 * size_Mat];
+ D.f[DIR_00P] = &DDStart[DIR_00P * size_Mat];
+ D.f[DIR_00M] = &DDStart[DIR_00M * size_Mat];
+ D.f[DIR_PP0] = &DDStart[DIR_PP0 * size_Mat];
+ D.f[DIR_MM0] = &DDStart[DIR_MM0 * size_Mat];
+ D.f[DIR_PM0] = &DDStart[DIR_PM0 * size_Mat];
+ D.f[DIR_MP0] = &DDStart[DIR_MP0 * size_Mat];
+ D.f[DIR_P0P] = &DDStart[DIR_P0P * size_Mat];
+ D.f[DIR_M0M] = &DDStart[DIR_M0M * size_Mat];
+ D.f[DIR_P0M] = &DDStart[DIR_P0M * size_Mat];
+ D.f[DIR_M0P] = &DDStart[DIR_M0P * size_Mat];
+ D.f[DIR_0PP] = &DDStart[DIR_0PP * size_Mat];
+ D.f[DIR_0MM] = &DDStart[DIR_0MM * size_Mat];
+ D.f[DIR_0PM] = &DDStart[DIR_0PM * size_Mat];
+ D.f[DIR_0MP] = &DDStart[DIR_0MP * size_Mat];
+ D.f[DIR_000] = &DDStart[DIR_000 * size_Mat];
+ D.f[DIR_PPP] = &DDStart[DIR_PPP * size_Mat];
+ D.f[DIR_MMP] = &DDStart[DIR_MMP * size_Mat];
+ D.f[DIR_PMP] = &DDStart[DIR_PMP * size_Mat];
+ D.f[DIR_MPP] = &DDStart[DIR_MPP * size_Mat];
+ D.f[DIR_PPM] = &DDStart[DIR_PPM * size_Mat];
+ D.f[DIR_MMM] = &DDStart[DIR_MMM * size_Mat];
+ D.f[DIR_PMM] = &DDStart[DIR_PMM * size_Mat];
+ D.f[DIR_MPM] = &DDStart[DIR_MPM * size_Mat];
}
else
{
- D.f[DIR_M00] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_P00] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0M0] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0P0] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00M] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00P] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_MM0] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_PP0] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_MP0] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_PM0] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_M0M] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_P0P] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_M0P] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_P0M] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0MM] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0PP] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0MP] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0PM] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_MMM] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_PPM] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_MPM] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_PMM] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_MMP] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_PPP] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_MPP] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_PMP] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_M00] = &DDStart[DIR_P00 * size_Mat];
+ D.f[DIR_P00] = &DDStart[DIR_M00 * size_Mat];
+ D.f[DIR_0M0] = &DDStart[DIR_0P0 * size_Mat];
+ D.f[DIR_0P0] = &DDStart[DIR_0M0 * size_Mat];
+ D.f[DIR_00M] = &DDStart[DIR_00P * size_Mat];
+ D.f[DIR_00P] = &DDStart[DIR_00M * size_Mat];
+ D.f[DIR_MM0] = &DDStart[DIR_PP0 * size_Mat];
+ D.f[DIR_PP0] = &DDStart[DIR_MM0 * size_Mat];
+ D.f[DIR_MP0] = &DDStart[DIR_PM0 * size_Mat];
+ D.f[DIR_PM0] = &DDStart[DIR_MP0 * size_Mat];
+ D.f[DIR_M0M] = &DDStart[DIR_P0P * size_Mat];
+ D.f[DIR_P0P] = &DDStart[DIR_M0M * size_Mat];
+ D.f[DIR_M0P] = &DDStart[DIR_P0M * size_Mat];
+ D.f[DIR_P0M] = &DDStart[DIR_M0P * size_Mat];
+ D.f[DIR_0MM] = &DDStart[DIR_0PP * size_Mat];
+ D.f[DIR_0PP] = &DDStart[DIR_0MM * size_Mat];
+ D.f[DIR_0MP] = &DDStart[DIR_0PM * size_Mat];
+ D.f[DIR_0PM] = &DDStart[DIR_0MP * size_Mat];
+ D.f[DIR_000] = &DDStart[DIR_000 * size_Mat];
+ D.f[DIR_MMM] = &DDStart[DIR_PPP * size_Mat];
+ D.f[DIR_PPM] = &DDStart[DIR_MMP * size_Mat];
+ D.f[DIR_MPM] = &DDStart[DIR_PMP * size_Mat];
+ D.f[DIR_PMM] = &DDStart[DIR_MPP * size_Mat];
+ D.f[DIR_MMP] = &DDStart[DIR_PPM * size_Mat];
+ D.f[DIR_PPP] = &DDStart[DIR_MMM * size_Mat];
+ D.f[DIR_MPP] = &DDStart[DIR_PMM * size_Mat];
+ D.f[DIR_PMP] = &DDStart[DIR_MPM * size_Mat];
}
////////////////////////////////////////////////////////////////////////////////
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK15Unified/CumulantK15Unified.cu b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK15Unified/CumulantK15Unified.cu
index 0b72b46cf25f331172be4abb8dded6d8e5e2b9c5..24b0bbc6f43a63093da6b6dcb3ce401b8a614f75 100644
--- a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK15Unified/CumulantK15Unified.cu
+++ b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK15Unified/CumulantK15Unified.cu
@@ -30,15 +30,16 @@ CumulantK15Unified::CumulantK15Unified(std::shared_ptr<Parameter> para, int leve
void CumulantK15Unified::run()
{
- GPUKernelParameter kernelParameter{ para->getParD(level)->omega,
- para->getParD(level)->typeOfGridNode,
- para->getParD(level)->neighborX,
- para->getParD(level)->neighborY,
- para->getParD(level)->neighborZ,
- para->getParD(level)->distributions.f[0],
- (int)para->getParD(level)->numberOfNodes,
- para->getParD(level)->forcing,
- para->getParD(level)->isEvenTimestep };
+ GPUKernelParameter kernelParameter{
+ para->getParD(level)->omega,
+ para->getParD(level)->typeOfGridNode,
+ para->getParD(level)->neighborX,
+ para->getParD(level)->neighborY,
+ para->getParD(level)->neighborZ,
+ para->getParD(level)->distributions.f[0],
+ (int)para->getParD(level)->numberOfNodes,
+ para->getParD(level)->forcing,
+ para->getParD(level)->isEvenTimestep };
auto lambda = [] __device__(lbm::KernelParameter parameter) {
return lbm::cumulantChimera(parameter, lbm::setRelaxationRatesK15);
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK17/CumulantK17.cu b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK17/CumulantK17.cu
new file mode 100644
index 0000000000000000000000000000000000000000..ea3442fecca63fdcb45878d742a547ce492ab5c8
--- /dev/null
+++ b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK17/CumulantK17.cu
@@ -0,0 +1,140 @@
+#include "CumulantK17.h"
+#include <logger/Logger.h>
+#include "Parameter/Parameter.h"
+#include "Parameter/CudaStreamManager.h"
+#include "CumulantK17_Device.cuh"
+
+#include <cuda.h>
+
+template<TurbulenceModel turbulenceModel>
+std::shared_ptr< CumulantK17<turbulenceModel> > CumulantK17<turbulenceModel>::getNewInstance(std::shared_ptr<Parameter> para, int level)
+{
+ return std::shared_ptr<CumulantK17<turbulenceModel> >(new CumulantK17<turbulenceModel>(para,level));
+}
+
+template<TurbulenceModel turbulenceModel>
+void CumulantK17<turbulenceModel>::run()
+{
+ LB_Kernel_CumulantK17 < turbulenceModel, false, false > <<< cudaGrid.grid, cudaGrid.threads >>>( para->getParD(level)->omega,
+ para->getParD(level)->neighborX, para->getParD(level)->neighborY, para->getParD(level)->neighborZ,
+ para->getParD(level)->distributions.f[0],
+ para->getParD(level)->rho,
+ para->getParD(level)->velocityX, para->getParD(level)->velocityY, para->getParD(level)->velocityZ,
+ para->getParD(level)->turbViscosity,
+ para->getSGSConstant(),
+ para->getParD(level)->numberOfNodes,
+ level,
+ para->getForcesDev(),
+ para->getParD(level)->forceX_SP, para->getParD(level)->forceY_SP, para->getParD(level)->forceZ_SP,
+ para->getQuadricLimitersDev(),
+ para->getParD(level)->isEvenTimestep,
+ para->getParD(level)->taggedFluidNodeIndices[CollisionTemplate::Default],
+ para->getParD(level)->numberOfTaggedFluidNodes[CollisionTemplate::Default]);
+
+ getLastCudaError("LB_Kernel_CumulantK17 execution failed");
+}
+
+template<TurbulenceModel turbulenceModel>
+void CumulantK17<turbulenceModel>::runOnIndices( const unsigned int *indices, unsigned int size_indices, CollisionTemplate collisionTemplate, CudaStreamIndex streamIndex )
+{
+ cudaStream_t stream = para->getStreamManager()->getStream(streamIndex);
+
+ switch (collisionTemplate)
+ {
+ case CollisionTemplate::Default:
+ LB_Kernel_CumulantK17 < turbulenceModel, false, false > <<< cudaGrid.grid, cudaGrid.threads, 0, stream >>>(para->getParD(level)->omega,
+ para->getParD(level)->neighborX, para->getParD(level)->neighborY, para->getParD(level)->neighborZ,
+ para->getParD(level)->distributions.f[0],
+ para->getParD(level)->rho,
+ para->getParD(level)->velocityX, para->getParD(level)->velocityY, para->getParD(level)->velocityZ,
+ para->getParD(level)->turbViscosity,
+ para->getSGSConstant(),
+ para->getParD(level)->numberOfNodes,
+ level,
+ para->getForcesDev(),
+ para->getParD(level)->forceX_SP, para->getParD(level)->forceY_SP, para->getParD(level)->forceZ_SP,
+ para->getQuadricLimitersDev(),
+ para->getParD(level)->isEvenTimestep,
+ indices,
+ size_indices);
+ break;
+
+ case CollisionTemplate::WriteMacroVars:
+ LB_Kernel_CumulantK17 < turbulenceModel, true, false > <<< cudaGrid.grid, cudaGrid.threads, 0, stream >>>( para->getParD(level)->omega,
+ para->getParD(level)->neighborX, para->getParD(level)->neighborY, para->getParD(level)->neighborZ,
+ para->getParD(level)->distributions.f[0],
+ para->getParD(level)->rho,
+ para->getParD(level)->velocityX, para->getParD(level)->velocityY, para->getParD(level)->velocityZ,
+ para->getParD(level)->turbViscosity,
+ para->getSGSConstant(),
+ para->getParD(level)->numberOfNodes,
+ level,
+ para->getForcesDev(),
+ para->getParD(level)->forceX_SP, para->getParD(level)->forceY_SP, para->getParD(level)->forceZ_SP,
+ para->getQuadricLimitersDev(),
+ para->getParD(level)->isEvenTimestep,
+ indices,
+ size_indices);
+ break;
+
+ case CollisionTemplate::SubDomainBorder:
+ case CollisionTemplate::AllFeatures:
+ LB_Kernel_CumulantK17 < turbulenceModel, true, true > <<< cudaGrid.grid, cudaGrid.threads, 0, stream >>>( para->getParD(level)->omega,
+ para->getParD(level)->neighborX, para->getParD(level)->neighborY, para->getParD(level)->neighborZ,
+ para->getParD(level)->distributions.f[0],
+ para->getParD(level)->rho,
+ para->getParD(level)->velocityX, para->getParD(level)->velocityY, para->getParD(level)->velocityZ,
+ para->getParD(level)->turbViscosity,
+ para->getSGSConstant(),
+ para->getParD(level)->numberOfNodes,
+ level,
+ para->getForcesDev(),
+ para->getParD(level)->forceX_SP, para->getParD(level)->forceY_SP, para->getParD(level)->forceZ_SP,
+ para->getQuadricLimitersDev(),
+ para->getParD(level)->isEvenTimestep,
+ indices,
+ size_indices);
+ break; case CollisionTemplate::ApplyBodyForce:
+ LB_Kernel_CumulantK17 < turbulenceModel, false, true > <<< cudaGrid.grid, cudaGrid.threads, 0, stream >>>( para->getParD(level)->omega,
+ para->getParD(level)->neighborX, para->getParD(level)->neighborY, para->getParD(level)->neighborZ,
+ para->getParD(level)->distributions.f[0],
+ para->getParD(level)->rho,
+ para->getParD(level)->velocityX, para->getParD(level)->velocityY, para->getParD(level)->velocityZ,
+ para->getParD(level)->turbViscosity,
+ para->getSGSConstant(),
+ para->getParD(level)->numberOfNodes,
+ level,
+ para->getForcesDev(),
+ para->getParD(level)->forceX_SP, para->getParD(level)->forceY_SP, para->getParD(level)->forceZ_SP,
+ para->getQuadricLimitersDev(),
+ para->getParD(level)->isEvenTimestep,
+ indices,
+ size_indices);
+ break; default:
+ throw std::runtime_error("Invalid CollisionTemplate in CumulantK17::runOnIndices()");
+ break;
+ }
+
+ getLastCudaError("LB_Kernel_CumulantK17 execution failed");
+}
+
+template<TurbulenceModel turbulenceModel>
+CumulantK17<turbulenceModel>::CumulantK17(std::shared_ptr<Parameter> para, int level)
+{
+ this->para = para;
+ this->level = level;
+
+ myPreProcessorTypes.push_back(InitCompSP27);
+
+ myKernelGroup = BasicKernel;
+
+ this->cudaGrid = vf::cuda::CudaGrid(para->getParD(level)->numberofthreads, para->getParD(level)->numberOfNodes);
+ this->kernelUsesFluidNodeIndices = true;
+
+ VF_LOG_INFO("Using turbulence model: {}", turbulenceModel);
+}
+
+template class CumulantK17<TurbulenceModel::AMD>;
+template class CumulantK17<TurbulenceModel::Smagorinsky>;
+template class CumulantK17<TurbulenceModel::QR>;
+template class CumulantK17<TurbulenceModel::None>;
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK17/CumulantK17.h b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK17/CumulantK17.h
new file mode 100644
index 0000000000000000000000000000000000000000..00c79a30c9ccf9a89901165d020fc85d5a479c1d
--- /dev/null
+++ b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK17/CumulantK17.h
@@ -0,0 +1,20 @@
+#ifndef CUMULANT_K17_H
+#define CUMULANT_K17_H
+
+#include "Kernel/KernelImp.h"
+#include "Parameter/Parameter.h"
+
+template<TurbulenceModel turbulenceModel>
+class CumulantK17 : public KernelImp
+{
+public:
+ static std::shared_ptr< CumulantK17<turbulenceModel> > getNewInstance(std::shared_ptr< Parameter> para, int level);
+ void run() override;
+ void runOnIndices(const unsigned int *indices, unsigned int size_indices, CollisionTemplate collisionTemplate, CudaStreamIndex streamIndex) override;
+
+private:
+ CumulantK17();
+ CumulantK17(std::shared_ptr<Parameter> para, int level);
+};
+
+#endif
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK17/CumulantK17Comp.cu b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK17/CumulantK17Comp.cu
deleted file mode 100644
index b176b94d07e7f280d738a797d5bd853095e3caed..0000000000000000000000000000000000000000
--- a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK17/CumulantK17Comp.cu
+++ /dev/null
@@ -1,33 +0,0 @@
-#include "CumulantK17Comp.h"
-
-#include "Parameter/Parameter.h"
-#include "CumulantK17Comp_Device.cuh"
-#include "cuda/CudaGrid.h"
-
-std::shared_ptr<CumulantK17Comp> CumulantK17Comp::getNewInstance(std::shared_ptr<Parameter> para, int level)
-{
- return std::shared_ptr<CumulantK17Comp>(new CumulantK17Comp(para,level));
-}
-
-void CumulantK17Comp::run()
-{
- LB_Kernel_CumulantK17Comp <<< cudaGrid.grid, cudaGrid.threads >>>(para->getParD(level)->omega,
- para->getParD(level)->typeOfGridNode,
- para->getParD(level)->neighborX,
- para->getParD(level)->neighborY,
- para->getParD(level)->neighborZ,
- para->getParD(level)->distributions.f[0],
- para->getParD(level)->numberOfNodes,
- level,
- para->getForcesDev(),
- para->getQuadricLimitersDev(),
- para->getParD(level)->isEvenTimestep);
- getLastCudaError("LB_Kernel_CumulantK17Comp execution failed");
-}
-
-CumulantK17Comp::CumulantK17Comp(std::shared_ptr<Parameter> para, int level): KernelImp(para, level)
-{
- myPreProcessorTypes.push_back(InitCompSP27);
- myKernelGroup = BasicKernel;
- this->cudaGrid = vf::cuda::CudaGrid(para->getParD(level)->numberofthreads, para->getParD(level)->numberOfNodes);
-}
\ No newline at end of file
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK17/CumulantK17Comp.h b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK17/CumulantK17Comp.h
deleted file mode 100644
index 22a95a688e5d078d7b710f494bfea360c9af0d6b..0000000000000000000000000000000000000000
--- a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK17/CumulantK17Comp.h
+++ /dev/null
@@ -1,17 +0,0 @@
-#ifndef CUMULANT_K17_COMP_H
-#define CUMULANT_K17_COMP_H
-
-#include "Kernel/KernelImp.h"
-
-class CumulantK17Comp : public KernelImp
-{
-public:
- static std::shared_ptr<CumulantK17Comp> getNewInstance(std::shared_ptr< Parameter> para, int level);
- void run();
-
-private:
- CumulantK17Comp();
- CumulantK17Comp(std::shared_ptr< Parameter> para, int level);
-};
-
-#endif
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK17/CumulantK17Comp_Device.cu b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK17/CumulantK17Comp_Device.cu
deleted file mode 100644
index 7cf27aa883cbfd3a0e4a0a36fa61649a62d06eeb..0000000000000000000000000000000000000000
--- a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK17/CumulantK17Comp_Device.cu
+++ /dev/null
@@ -1,1040 +0,0 @@
-#include "LBM/LB.h"
-#include "lbm/constants/D3Q27.h"
-#include <lbm/constants/NumericConstants.h>
-
-using namespace vf::lbm::constant;
-using namespace vf::lbm::dir;
-#include "math.h"
-
-
-__global__ void LB_Kernel_CumulantK17Comp(real omega,
- unsigned int* bcMatD,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- real* DDStart,
- int size_Mat,
- int level,
- real* forces,
- real* quadricLimiters,
- bool EvenOrOdd)
-{
- ////////////////////////////////////////////////////////////////////////////////
- const unsigned x = threadIdx.x; // Globaler x-Index
- const unsigned y = blockIdx.x; // Globaler y-Index
- const unsigned z = blockIdx.y; // Globaler z-Index
-
- const unsigned nx = blockDim.x;
- const unsigned ny = gridDim.x;
-
- const unsigned k = nx*(ny*z + y) + x;
- //////////////////////////////////////////////////////////////////////////
-
- if (k<size_Mat)
- {
- ////////////////////////////////////////////////////////////////////////////////
- unsigned int BC;
- BC = bcMatD[k];
-
- if ((BC != GEO_SOLID) && (BC != GEO_VOID))
- {
- Distributions27 D;
- if (EvenOrOdd == true)
- {
- D.f[DIR_P00] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_M00] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0P0] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0M0] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00P] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00M] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_PP0] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_MM0] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_PM0] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_MP0] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_P0P] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_M0M] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_P0M] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_M0P] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0PP] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0MM] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0PM] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0MP] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_PPP] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_MMP] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_PMP] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_MPP] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_PPM] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_MMM] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_PMM] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_MPM] = &DDStart[DIR_MPM *size_Mat];
- }
- else
- {
- D.f[DIR_M00] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_P00] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0M0] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0P0] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00M] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00P] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_MM0] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_PP0] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_MP0] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_PM0] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_M0M] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_P0P] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_M0P] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_P0M] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0MM] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0PP] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0MP] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0PM] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_MMM] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_PPM] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_MPM] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_PMM] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_MMP] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_PPP] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_MPP] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_PMP] = &DDStart[DIR_MPM *size_Mat];
- }
-
- ////////////////////////////////////////////////////////////////////////////////
- //index
- //unsigned int kzero= k;
- //unsigned int ke = k;
- unsigned int kw = neighborX[k];
- //unsigned int kn = k;
- unsigned int ks = neighborY[k];
- //unsigned int kt = k;
- unsigned int kb = neighborZ[k];
- unsigned int ksw = neighborY[kw];
- //unsigned int kne = k;
- //unsigned int kse = ks;
- //unsigned int knw = kw;
- unsigned int kbw = neighborZ[kw];
- //unsigned int kte = k;
- //unsigned int kbe = kb;
- //unsigned int ktw = kw;
- unsigned int kbs = neighborZ[ks];
- //unsigned int ktn = k;
- //unsigned int kbn = kb;
- //unsigned int kts = ks;
- //unsigned int ktse = ks;
- //unsigned int kbnw = kbw;
- //unsigned int ktnw = kw;
- //unsigned int kbse = kbs;
- //unsigned int ktsw = ksw;
- //unsigned int kbne = kb;
- //unsigned int ktne = k;
- unsigned int kbsw = neighborZ[ksw];
-
- //unsigned int kzero= k;
- //unsigned int ke = k;
- //unsigned int kw = neighborX[k];
- //unsigned int kn = k;
- //unsigned int ks = neighborY[k];
- //unsigned int kt = k;
- //unsigned int kb = neighborZ[k];
- //unsigned int ksw = neighborY[kw];
- //unsigned int kne = k;
- //unsigned int kse = ks;
- //unsigned int knw = kw;
- //unsigned int kbw = neighborZ[kw];
- //unsigned int kte = k;
- //unsigned int kbe = kb;
- //unsigned int ktw = kw;
- //unsigned int kbs = neighborZ[ks];
- //unsigned int ktn = k;
- //unsigned int kbn = kb;
- //unsigned int kts = ks;
- //unsigned int ktse = ks;
- //unsigned int kbnw = kbw;
- //unsigned int ktnw = kw;
- //unsigned int kbse = kbs;
- //unsigned int ktsw = ksw;
- //unsigned int kbne = kb;
- //unsigned int ktne = k;
- //unsigned int kbsw = neighborZ[ksw];
- //////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
- real mfcbb = (D.f[DIR_P00])[k];//[ke ];// + c2over27 ;(D.f[DIR_P00 ])[k ];//ke
- real mfabb = (D.f[DIR_M00])[kw];//[kw ];// + c2over27 ;(D.f[DIR_M00 ])[kw ];
- real mfbcb = (D.f[DIR_0P0])[k];//[kn ];// + c2over27 ;(D.f[DIR_0P0 ])[k ];//kn
- real mfbab = (D.f[DIR_0M0])[ks];//[ks ];// + c2over27 ;(D.f[DIR_0M0 ])[ks ];
- real mfbbc = (D.f[DIR_00P])[k];//[kt ];// + c2over27 ;(D.f[DIR_00P ])[k ];//kt
- real mfbba = (D.f[DIR_00M])[kb];//[kb ];// + c2over27 ;(D.f[DIR_00M ])[kb ];
- real mfccb = (D.f[DIR_PP0])[k];//[kne ];// + c1over54 ;(D.f[DIR_PP0 ])[k ];//kne
- real mfaab = (D.f[DIR_MM0])[ksw];//[ksw ];// + c1over54 ;(D.f[DIR_MM0 ])[ksw];
- real mfcab = (D.f[DIR_PM0])[ks];//[kse ];// + c1over54 ;(D.f[DIR_PM0 ])[ks ];//kse
- real mfacb = (D.f[DIR_MP0])[kw];//[knw ];// + c1over54 ;(D.f[DIR_MP0 ])[kw ];//knw
- real mfcbc = (D.f[DIR_P0P])[k];//[kte ];// + c1over54 ;(D.f[DIR_P0P ])[k ];//kte
- real mfaba = (D.f[DIR_M0M])[kbw];//[kbw ];// + c1over54 ;(D.f[DIR_M0M ])[kbw];
- real mfcba = (D.f[DIR_P0M])[kb];//[kbe ];// + c1over54 ;(D.f[DIR_P0M ])[kb ];//kbe
- real mfabc = (D.f[DIR_M0P])[kw];//[ktw ];// + c1over54 ;(D.f[DIR_M0P ])[kw ];//ktw
- real mfbcc = (D.f[DIR_0PP])[k];//[ktn ];// + c1over54 ;(D.f[DIR_0PP ])[k ];//ktn
- real mfbaa = (D.f[DIR_0MM])[kbs];//[kbs ];// + c1over54 ;(D.f[DIR_0MM ])[kbs];
- real mfbca = (D.f[DIR_0PM])[kb];//[kbn ];// + c1over54 ;(D.f[DIR_0PM ])[kb ];//kbn
- real mfbac = (D.f[DIR_0MP])[ks];//[kts ];// + c1over54 ;(D.f[DIR_0MP ])[ks ];//kts
- real mfbbb = (D.f[DIR_000])[k];//[kzero];// + c8over27 ;(D.f[DIR_000])[k ];//kzero
- real mfccc = (D.f[DIR_PPP])[k];//[ktne ];// + c1over216;(D.f[DIR_PPP ])[k ];//ktne
- real mfaac = (D.f[DIR_MMP])[ksw];//[ktsw ];// + c1over216;(D.f[DIR_MMP ])[ksw];//ktsw
- real mfcac = (D.f[DIR_PMP])[ks];//[ktse ];// + c1over216;(D.f[DIR_PMP ])[ks ];//ktse
- real mfacc = (D.f[DIR_MPP])[kw];//[ktnw ];// + c1over216;(D.f[DIR_MPP ])[kw ];//ktnw
- real mfcca = (D.f[DIR_PPM])[kb];//[kbne ];// + c1over216;(D.f[DIR_PPM ])[kb ];//kbne
- real mfaaa = (D.f[DIR_MMM])[kbsw];//[kbsw ];// + c1over216;(D.f[DIR_MMM ])[kbsw];
- real mfcaa = (D.f[DIR_PMM])[kbs];//[kbse ];// + c1over216;(D.f[DIR_PMM ])[kbs];//kbse
- real mfaca = (D.f[DIR_MPM])[kbw];//[kbnw ];// + c1over216;(D.f[DIR_MPM ])[kbw];//kbnw
- ////////////////////////////////////////////////////////////////////////////////////
- real drho = ((((mfccc + mfaaa) + (mfaca + mfcac)) + ((mfacc + mfcaa) + (mfaac + mfcca))) +
- (((mfbac + mfbca) + (mfbaa + mfbcc)) + ((mfabc + mfcba) + (mfaba + mfcbc)) + ((mfacb + mfcab) + (mfaab + mfccb))) +
- ((mfabb + mfcbb) + (mfbab + mfbcb) + (mfbba + mfbbc))) + mfbbb;
-
- real rho = c1o1 + drho;
- ////////////////////////////////////////////////////////////////////////////////////
- //slow
- //real oMdrho = one - ((((mfccc+mfaaa) + (mfaca+mfcac)) + ((mfacc+mfcaa) + (mfaac+mfcca))) +
- // (((mfbac+mfbca) + (mfbaa+mfbcc)) + ((mfabc+mfcba) + (mfaba+mfcbc)) + ((mfacb+mfcab) + (mfaab+mfccb))) +
- // ((mfabb+mfcbb) + (mfbab+mfbcb) + (mfbba+mfbbc)));//fehlt mfbbb
- real vvx = ((((mfccc - mfaaa) + (mfcac - mfaca)) + ((mfcaa - mfacc) + (mfcca - mfaac))) +
- (((mfcba - mfabc) + (mfcbc - mfaba)) + ((mfcab - mfacb) + (mfccb - mfaab))) +
- (mfcbb - mfabb)) / rho;
- real vvy = ((((mfccc - mfaaa) + (mfaca - mfcac)) + ((mfacc - mfcaa) + (mfcca - mfaac))) +
- (((mfbca - mfbac) + (mfbcc - mfbaa)) + ((mfacb - mfcab) + (mfccb - mfaab))) +
- (mfbcb - mfbab)) / rho;
- real vvz = ((((mfccc - mfaaa) + (mfcac - mfaca)) + ((mfacc - mfcaa) + (mfaac - mfcca))) +
- (((mfbac - mfbca) + (mfbcc - mfbaa)) + ((mfabc - mfcba) + (mfcbc - mfaba))) +
- (mfbbc - mfbba)) / rho;
- ////////////////////////////////////////////////////////////////////////////////////
- //the force be with you
- real fx = forces[0] / (pow((double)c2o1, (double)level)); //zero;//0.0032653/(pow(two,level)); //0.000000005;//(two/1600000.0) / 120.0; //
- real fy = forces[1] / (pow((double)c2o1, (double)level)); //zero;
- real fz = forces[2] / (pow((double)c2o1, (double)level)); //zero;
- vvx += fx*c1o2;
- vvy += fy*c1o2;
- vvz += fz*c1o2;
- ////////////////////////////////////////////////////////////////////////////////////
- //real omega = omega_in;
- ////////////////////////////////////////////////////////////////////////////////////
- //fast
- real oMdrho = c1o1; // comp special
- //real oMdrho = one - (mfccc+mfaaa + mfaca+mfcac + mfacc+mfcaa + mfaac+mfcca +
- // mfbac+mfbca + mfbaa+mfbcc + mfabc+mfcba + mfaba+mfcbc + mfacb+mfcab + mfaab+mfccb +
- // mfabb+mfcbb + mfbab+mfbcb + mfbba+mfbbc + mfbbb);//fehlt mfbbb nicht mehr
- //real vvx =mfccc-mfaaa + mfcac-mfaca + mfcaa-mfacc + mfcca-mfaac +
- // mfcba-mfabc + mfcbc-mfaba + mfcab-mfacb + mfccb-mfaab +
- // mfcbb-mfabb;
- //real vvy =mfccc-mfaaa + mfaca-mfcac + mfacc-mfcaa + mfcca-mfaac +
- // mfbca-mfbac + mfbcc-mfbaa + mfacb-mfcab + mfccb-mfaab +
- // mfbcb-mfbab;
- //real vvz =mfccc-mfaaa + mfcac-mfaca + mfacc-mfcaa + mfaac-mfcca +
- // mfbac-mfbca + mfbcc-mfbaa + mfabc-mfcba + mfcbc-mfaba +
- // mfbbc-mfbba;
- ////////////////////////////////////////////////////////////////////////////////////
- // oMdrho assembler style -------> faaaaaastaaaa
- // or much sloooowaaaa ... it dep�ndssssss on sadaku
- real m0, m1, m2;
- //real oMdrho;
- //{
- // oMdrho=mfccc+mfaaa;
- // m0=mfaca+mfcac;
- // m1=mfacc+mfcaa;
- // m2=mfaac+mfcca;
- // oMdrho+=m0;
- // m1+=m2;
- // oMdrho+=m1;
- // m0=mfbac+mfbca;
- // m1=mfbaa+mfbcc;
- // m0+=m1;
- // m1=mfabc+mfcba;
- // m2=mfaba+mfcbc;
- // m1+=m2;
- // m0+=m1;
- // m1=mfacb+mfcab;
- // m2=mfaab+mfccb;
- // m1+=m2;
- // m0+=m1;
- // oMdrho+=m0;
- // m0=mfabb+mfcbb;
- // m1=mfbab+mfbcb;
- // m2=mfbba+mfbbc;
- // m0+=m1+m2;
- // m0+=mfbbb; //hat gefehlt
- // oMdrho = one - (oMdrho + m0);
- //}
- //real vvx;
- real vx2;
- //{
- // vvx = mfccc-mfaaa;
- // m0 = mfcac-mfaca;
- // m1 = mfcaa-mfacc;
- // m2 = mfcca-mfaac;
- // vvx+= m0;
- // m1 += m2;
- // vvx+= m1;
- // vx2 = mfcba-mfabc;
- // m0 = mfcbc-mfaba;
- // m1 = mfcab-mfacb;
- // m2 = mfccb-mfaab;
- // vx2+= m0;
- // m1 += m2;
- // vx2+= m1;
- // vvx+= vx2;
- // vx2 = mfcbb-mfabb;
- // vvx+= vx2;
- //}
- //real vvy;
- real vy2;
- //{
- // vvy = mfccc-mfaaa;
- // m0 = mfaca-mfcac;
- // m1 = mfacc-mfcaa;
- // m2 = mfcca-mfaac;
- // vvy+= m0;
- // m1 += m2;
- // vvy+= m1;
- // vy2 = mfbca-mfbac;
- // m0 = mfbcc-mfbaa;
- // m1 = mfacb-mfcab;
- // m2 = mfccb-mfaab;
- // vy2+= m0;
- // m1 += m2;
- // vy2+= m1;
- // vvy+= vy2;
- // vy2 = mfbcb-mfbab;
- // vvy+= vy2;
- //}
- //real vvz;
- real vz2;
- //{
- // vvz = mfccc-mfaaa;
- // m0 = mfcac-mfaca;
- // m1 = mfacc-mfcaa;
- // m2 = mfaac-mfcca;
- // vvz+= m0;
- // m1 += m2;
- // vvz+= m1;
- // vz2 = mfbac-mfbca;
- // m0 = mfbcc-mfbaa;
- // m1 = mfabc-mfcba;
- // m2 = mfcbc-mfaba;
- // vz2+= m0;
- // m1 += m2;
- // vz2+= m1;
- // vvz+= vz2;
- // vz2 = mfbbc-mfbba;
- // vvz+= vz2;
- //}
- vx2 = vvx*vvx;
- vy2 = vvy*vvy;
- vz2 = vvz*vvz;
- ////////////////////////////////////////////////////////////////////////////////////
- real wadjust;
- real qudricLimitP = quadricLimiters[0]; //0.01f; // * 0.0001f; // 1000000.0f; // 1000000.0f; //
- real qudricLimitM = quadricLimiters[1]; //0.01f; // * 0.0001f; // 1000000.0f; // 1000000.0f; //
- real qudricLimitD = quadricLimiters[2]; //0.01f; // * 0.001f; // 1000000.0f; // 1000000.0f; //
- ////////////////////////////////////////////////////////////////////////////////////
- //Hin
- ////////////////////////////////////////////////////////////////////////////////////
- // mit 1/36, 1/9, 1/36, 1/9, 4/9, 1/9, 1/36, 1/9, 1/36 Konditionieren
- ////////////////////////////////////////////////////////////////////////////////////
- // Z - Dir
- m2 = mfaaa + mfaac;
- m1 = mfaac - mfaaa;
- m0 = m2 + mfaab;
- mfaaa = m0;
- m0 += c1o36 * oMdrho;
- mfaab = m1 - m0 * vvz;
- mfaac = m2 - c2o1* m1 * vvz + vz2 * m0;
- ////////////////////////////////////////////////////////////////////////////////////
- m2 = mfaba + mfabc;
- m1 = mfabc - mfaba;
- m0 = m2 + mfabb;
- mfaba = m0;
- m0 += c1o9 * oMdrho;
- mfabb = m1 - m0 * vvz;
- mfabc = m2 - c2o1* m1 * vvz + vz2 * m0;
- ////////////////////////////////////////////////////////////////////////////////////
- m2 = mfaca + mfacc;
- m1 = mfacc - mfaca;
- m0 = m2 + mfacb;
- mfaca = m0;
- m0 += c1o36 * oMdrho;
- mfacb = m1 - m0 * vvz;
- mfacc = m2 - c2o1* m1 * vvz + vz2 * m0;
- ////////////////////////////////////////////////////////////////////////////////////
- ////////////////////////////////////////////////////////////////////////////////////
- m2 = mfbaa + mfbac;
- m1 = mfbac - mfbaa;
- m0 = m2 + mfbab;
- mfbaa = m0;
- m0 += c1o9 * oMdrho;
- mfbab = m1 - m0 * vvz;
- mfbac = m2 - c2o1* m1 * vvz + vz2 * m0;
- ////////////////////////////////////////////////////////////////////////////////////
- m2 = mfbba + mfbbc;
- m1 = mfbbc - mfbba;
- m0 = m2 + mfbbb;
- mfbba = m0;
- m0 += c4o9 * oMdrho;
- mfbbb = m1 - m0 * vvz;
- mfbbc = m2 - c2o1* m1 * vvz + vz2 * m0;
- ////////////////////////////////////////////////////////////////////////////////////
- m2 = mfbca + mfbcc;
- m1 = mfbcc - mfbca;
- m0 = m2 + mfbcb;
- mfbca = m0;
- m0 += c1o9 * oMdrho;
- mfbcb = m1 - m0 * vvz;
- mfbcc = m2 - c2o1* m1 * vvz + vz2 * m0;
- ////////////////////////////////////////////////////////////////////////////////////
- ////////////////////////////////////////////////////////////////////////////////////
- m2 = mfcaa + mfcac;
- m1 = mfcac - mfcaa;
- m0 = m2 + mfcab;
- mfcaa = m0;
- m0 += c1o36 * oMdrho;
- mfcab = m1 - m0 * vvz;
- mfcac = m2 - c2o1* m1 * vvz + vz2 * m0;
- ////////////////////////////////////////////////////////////////////////////////////
- m2 = mfcba + mfcbc;
- m1 = mfcbc - mfcba;
- m0 = m2 + mfcbb;
- mfcba = m0;
- m0 += c1o9 * oMdrho;
- mfcbb = m1 - m0 * vvz;
- mfcbc = m2 - c2o1* m1 * vvz + vz2 * m0;
- ////////////////////////////////////////////////////////////////////////////////////
- m2 = mfcca + mfccc;
- m1 = mfccc - mfcca;
- m0 = m2 + mfccb;
- mfcca = m0;
- m0 += c1o36 * oMdrho;
- mfccb = m1 - m0 * vvz;
- mfccc = m2 - c2o1* m1 * vvz + vz2 * m0;
- ////////////////////////////////////////////////////////////////////////////////////
- ////////////////////////////////////////////////////////////////////////////////////
- // mit 1/6, 0, 1/18, 2/3, 0, 2/9, 1/6, 0, 1/18 Konditionieren
- ////////////////////////////////////////////////////////////////////////////////////
- // Y - Dir
- m2 = mfaaa + mfaca;
- m1 = mfaca - mfaaa;
- m0 = m2 + mfaba;
- mfaaa = m0;
- m0 += c1o6 * oMdrho;
- mfaba = m1 - m0 * vvy;
- mfaca = m2 - c2o1* m1 * vvy + vy2 * m0;
- ////////////////////////////////////////////////////////////////////////////////////
- m2 = mfaab + mfacb;
- m1 = mfacb - mfaab;
- m0 = m2 + mfabb;
- mfaab = m0;
- mfabb = m1 - m0 * vvy;
- mfacb = m2 - c2o1* m1 * vvy + vy2 * m0;
- ////////////////////////////////////////////////////////////////////////////////////
- m2 = mfaac + mfacc;
- m1 = mfacc - mfaac;
- m0 = m2 + mfabc;
- mfaac = m0;
- m0 += c1o18 * oMdrho;
- mfabc = m1 - m0 * vvy;
- mfacc = m2 - c2o1* m1 * vvy + vy2 * m0;
- ////////////////////////////////////////////////////////////////////////////////////
- ////////////////////////////////////////////////////////////////////////////////////
- m2 = mfbaa + mfbca;
- m1 = mfbca - mfbaa;
- m0 = m2 + mfbba;
- mfbaa = m0;
- m0 += c2o3 * oMdrho;
- mfbba = m1 - m0 * vvy;
- mfbca = m2 - c2o1* m1 * vvy + vy2 * m0;
- ////////////////////////////////////////////////////////////////////////////////////
- m2 = mfbab + mfbcb;
- m1 = mfbcb - mfbab;
- m0 = m2 + mfbbb;
- mfbab = m0;
- mfbbb = m1 - m0 * vvy;
- mfbcb = m2 - c2o1* m1 * vvy + vy2 * m0;
- ////////////////////////////////////////////////////////////////////////////////////
- m2 = mfbac + mfbcc;
- m1 = mfbcc - mfbac;
- m0 = m2 + mfbbc;
- mfbac = m0;
- m0 += c2o9 * oMdrho;
- mfbbc = m1 - m0 * vvy;
- mfbcc = m2 - c2o1* m1 * vvy + vy2 * m0;
- ////////////////////////////////////////////////////////////////////////////////////
- ////////////////////////////////////////////////////////////////////////////////////
- m2 = mfcaa + mfcca;
- m1 = mfcca - mfcaa;
- m0 = m2 + mfcba;
- mfcaa = m0;
- m0 += c1o6 * oMdrho;
- mfcba = m1 - m0 * vvy;
- mfcca = m2 - c2o1* m1 * vvy + vy2 * m0;
- ////////////////////////////////////////////////////////////////////////////////////
- m2 = mfcab + mfccb;
- m1 = mfccb - mfcab;
- m0 = m2 + mfcbb;
- mfcab = m0;
- mfcbb = m1 - m0 * vvy;
- mfccb = m2 - c2o1* m1 * vvy + vy2 * m0;
- ////////////////////////////////////////////////////////////////////////////////////
- m2 = mfcac + mfccc;
- m1 = mfccc - mfcac;
- m0 = m2 + mfcbc;
- mfcac = m0;
- m0 += c1o18 * oMdrho;
- mfcbc = m1 - m0 * vvy;
- mfccc = m2 - c2o1* m1 * vvy + vy2 * m0;
- ////////////////////////////////////////////////////////////////////////////////////
- ////////////////////////////////////////////////////////////////////////////////////
- // mit 1, 0, 1/3, 0, 0, 0, 1/3, 0, 1/9 Konditionieren
- ////////////////////////////////////////////////////////////////////////////////////
- // X - Dir
- m2 = mfaaa + mfcaa;
- m1 = mfcaa - mfaaa;
- m0 = m2 + mfbaa;
- mfaaa = m0;
- m0 += c1o1* oMdrho;
- mfbaa = m1 - m0 * vvx;
- mfcaa = m2 - c2o1* m1 * vvx + vx2 * m0;
- ////////////////////////////////////////////////////////////////////////////////////
- m2 = mfaba + mfcba;
- m1 = mfcba - mfaba;
- m0 = m2 + mfbba;
- mfaba = m0;
- mfbba = m1 - m0 * vvx;
- mfcba = m2 - c2o1* m1 * vvx + vx2 * m0;
- ////////////////////////////////////////////////////////////////////////////////////
- m2 = mfaca + mfcca;
- m1 = mfcca - mfaca;
- m0 = m2 + mfbca;
- mfaca = m0;
- m0 += c1o3 * oMdrho;
- mfbca = m1 - m0 * vvx;
- mfcca = m2 - c2o1* m1 * vvx + vx2 * m0;
- ////////////////////////////////////////////////////////////////////////////////////
- ////////////////////////////////////////////////////////////////////////////////////
- m2 = mfaab + mfcab;
- m1 = mfcab - mfaab;
- m0 = m2 + mfbab;
- mfaab = m0;
- mfbab = m1 - m0 * vvx;
- mfcab = m2 - c2o1* m1 * vvx + vx2 * m0;
- ////////////////////////////////////////////////////////////////////////////////////
- m2 = mfabb + mfcbb;
- m1 = mfcbb - mfabb;
- m0 = m2 + mfbbb;
- mfabb = m0;
- mfbbb = m1 - m0 * vvx;
- mfcbb = m2 - c2o1* m1 * vvx + vx2 * m0;
- ////////////////////////////////////////////////////////////////////////////////////
- m2 = mfacb + mfccb;
- m1 = mfccb - mfacb;
- m0 = m2 + mfbcb;
- mfacb = m0;
- mfbcb = m1 - m0 * vvx;
- mfccb = m2 - c2o1* m1 * vvx + vx2 * m0;
- ////////////////////////////////////////////////////////////////////////////////////
- ////////////////////////////////////////////////////////////////////////////////////
- m2 = mfaac + mfcac;
- m1 = mfcac - mfaac;
- m0 = m2 + mfbac;
- mfaac = m0;
- m0 += c1o3 * oMdrho;
- mfbac = m1 - m0 * vvx;
- mfcac = m2 - c2o1* m1 * vvx + vx2 * m0;
- ////////////////////////////////////////////////////////////////////////////////////
- m2 = mfabc + mfcbc;
- m1 = mfcbc - mfabc;
- m0 = m2 + mfbbc;
- mfabc = m0;
- mfbbc = m1 - m0 * vvx;
- mfcbc = m2 - c2o1* m1 * vvx + vx2 * m0;
- ////////////////////////////////////////////////////////////////////////////////////
- m2 = mfacc + mfccc;
- m1 = mfccc - mfacc;
- m0 = m2 + mfbcc;
- mfacc = m0;
- m0 += c1o9 * oMdrho;
- mfbcc = m1 - m0 * vvx;
- mfccc = m2 - c2o1* m1 * vvx + vx2 * m0;
- ////////////////////////////////////////////////////////////////////////////////////
- ////////////////////////////////////////////////////////////////////////////////////
-
- ////////////////////////////////////////////////////////////////////////////////////
- // Cumulants
- ////////////////////////////////////////////////////////////////////////////////////
- real OxxPyyPzz = c1o1; //set the bulk viscosity one is high / two is very low and zero is (too) high ... (also called omega 2)
-
- ////////////////////////////////////////////////////////////
- //3.
- //////////////////////////////
- real OxyyPxzz = c8o1*(-c2o1 + omega)*(c1o1 + c2o1*omega) / (-c8o1 - c14o1*omega + c7o1*omega*omega);//one;
- real OxyyMxzz = c8o1*(-c2o1 + omega)*(-c7o1 + c4o1*omega) / (c56o1 - c50o1*omega + c9o1*omega*omega);//one;
- real Oxyz = c24o1*(-c2o1 + omega)*(-c2o1 - c7o1*omega + c3o1*omega*omega) / (c48o1 + c152o1*omega - c130o1*omega*omega + c29o1*omega*omega*omega);//one;
- ////////////////////////////////////////////////////////////
- //4.
- //////////////////////////////
- real O4 = c1o1;
- //////////////////////////////
- //real O4 = omega;//TRT
- ////////////////////////////////////////////////////////////
- //5.
- //////////////////////////////
- real O5 = c1o1;
- ////////////////////////////////////////////////////////////
- //6.
- //////////////////////////////
- real O6 = c1o1;
- ////////////////////////////////////////////////////////////
-
-
- //central moments to cumulants
- //4.
- real CUMcbb = mfcbb - ((mfcaa + c1o3) * mfabb + c2o1 * mfbba * mfbab) / rho;
- real CUMbcb = mfbcb - ((mfaca + c1o3) * mfbab + c2o1 * mfbba * mfabb) / rho;
- real CUMbbc = mfbbc - ((mfaac + c1o3) * mfbba + c2o1 * mfbab * mfabb) / rho;
-
- real CUMcca = mfcca - (((mfcaa * mfaca + c2o1 * mfbba * mfbba) + c1o3 * (mfcaa + mfaca)) / rho - c1o9*(drho / rho));
- real CUMcac = mfcac - (((mfcaa * mfaac + c2o1 * mfbab * mfbab) + c1o3 * (mfcaa + mfaac)) / rho - c1o9*(drho / rho));
- real CUMacc = mfacc - (((mfaac * mfaca + c2o1 * mfabb * mfabb) + c1o3 * (mfaac + mfaca)) / rho - c1o9*(drho / rho));
-
- //5.
- real CUMbcc = mfbcc - ((mfaac * mfbca + mfaca * mfbac + c4o1 * mfabb * mfbbb + c2o1 * (mfbab * mfacb + mfbba * mfabc)) + c1o3 * (mfbca + mfbac)) / rho;
- real CUMcbc = mfcbc - ((mfaac * mfcba + mfcaa * mfabc + c4o1 * mfbab * mfbbb + c2o1 * (mfabb * mfcab + mfbba * mfbac)) + c1o3 * (mfcba + mfabc)) / rho;
- real CUMccb = mfccb - ((mfcaa * mfacb + mfaca * mfcab + c4o1 * mfbba * mfbbb + c2o1 * (mfbab * mfbca + mfabb * mfcba)) + c1o3 * (mfacb + mfcab)) / rho;
-
- //6.
-
- real CUMccc = mfccc + ((-c4o1 * mfbbb * mfbbb
- - (mfcaa * mfacc + mfaca * mfcac + mfaac * mfcca)
- - c4o1 * (mfabb * mfcbb + mfbab * mfbcb + mfbba * mfbbc)
- - c2o1 * (mfbca * mfbac + mfcba * mfabc + mfcab * mfacb)) / rho
- + (c4o1 * (mfbab * mfbab * mfaca + mfabb * mfabb * mfcaa + mfbba * mfbba * mfaac)
- + c2o1 * (mfcaa * mfaca * mfaac)
- + c16o1 * mfbba * mfbab * mfabb) / (rho * rho)
- - c1o3 * (mfacc + mfcac + mfcca) / rho
- - c1o9 * (mfcaa + mfaca + mfaac) / rho
- + (c2o1 * (mfbab * mfbab + mfabb * mfabb + mfbba * mfbba)
- + (mfaac * mfaca + mfaac * mfcaa + mfaca * mfcaa) + c1o3 *(mfaac + mfaca + mfcaa)) / (rho * rho) * c2o3
- + c1o27*((drho * drho - drho) / (rho*rho)));
-
- //2.
- // linear combinations
- real mxxPyyPzz = mfcaa + mfaca + mfaac;
- real mxxMyy = mfcaa - mfaca;
- real mxxMzz = mfcaa - mfaac;
-
- ////////////////////////////////////////////////////////////////////////////
- real Dxy = -c3o1*omega*mfbba;
- real Dxz = -c3o1*omega*mfbab;
- real Dyz = -c3o1*omega*mfabb;
-
- //3.
- // linear combinations
-
- real mxxyPyzz = mfcba + mfabc;
- real mxxyMyzz = mfcba - mfabc;
-
- real mxxzPyyz = mfcab + mfacb;
- real mxxzMyyz = mfcab - mfacb;
-
- real mxyyPxzz = mfbca + mfbac;
- real mxyyMxzz = mfbca - mfbac;
-
- ///////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
- //incl. correction (hat noch nicht so gut funktioniert...Optimierungsbedarf??)
-
- real dxux = c1o2 * (-omega) *(mxxMyy + mxxMzz) + c1o2 * OxxPyyPzz * (mfaaa - mxxPyyPzz);
- real dyuy = dxux + omega * c3o2 * mxxMyy;
- real dzuz = dxux + omega * c3o2 * mxxMzz;
-
- //relax
- mxxPyyPzz += OxxPyyPzz*(mfaaa - mxxPyyPzz) - c3o1 * (c1o1 - c1o2 * OxxPyyPzz) * (vx2 * dxux + vy2 * dyuy + vz2 * dzuz);//-magicBulk*OxxPyyPzz;
- mxxMyy += omega * (-mxxMyy) - c3o1 * (c1o1 + c1o2 * (-omega)) * (vx2 * dxux - vy2 * dyuy);
- mxxMzz += omega * (-mxxMzz) - c3o1 * (c1o1 + c1o2 * (-omega)) * (vx2 * dxux - vz2 * dzuz);
-
- ///////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
-
- /////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
- ////no correction
- //mxxPyyPzz += OxxPyyPzz*(mfaaa-mxxPyyPzz);//-magicBulk*OxxPyyPzz;
- //mxxMyy += -(-omega) * (-mxxMyy);
- //mxxMzz += -(-omega) * (-mxxMzz);
- /////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
- mfabb += omega * (-mfabb);
- mfbab += omega * (-mfbab);
- mfbba += omega * (-mfbba);
- //////////////////////////////////////////////////////////////////////////
-
- // linear combinations back
- mfcaa = c1o3 * (mxxMyy + mxxMzz + mxxPyyPzz);
- mfaca = c1o3 * (-c2o1* mxxMyy + mxxMzz + mxxPyyPzz);
- mfaac = c1o3 * (mxxMyy - c2o1* mxxMzz + mxxPyyPzz);
-
-
- //relax
- //////////////////////////////////////////////////////////////////////////
- //das ist der limiter
- wadjust = Oxyz + (c1o1 - Oxyz)*abs(mfbbb) / (abs(mfbbb) + qudricLimitD);
- mfbbb += wadjust * (-mfbbb);
- wadjust = OxyyPxzz + (c1o1 - OxyyPxzz)*abs(mxxyPyzz) / (abs(mxxyPyzz) + qudricLimitP);
- mxxyPyzz += wadjust * (-mxxyPyzz);
- wadjust = OxyyMxzz + (c1o1 - OxyyMxzz)*abs(mxxyMyzz) / (abs(mxxyMyzz) + qudricLimitM);
- mxxyMyzz += wadjust * (-mxxyMyzz);
- wadjust = OxyyPxzz + (c1o1 - OxyyPxzz)*abs(mxxzPyyz) / (abs(mxxzPyyz) + qudricLimitP);
- mxxzPyyz += wadjust * (-mxxzPyyz);
- wadjust = OxyyMxzz + (c1o1 - OxyyMxzz)*abs(mxxzMyyz) / (abs(mxxzMyyz) + qudricLimitM);
- mxxzMyyz += wadjust * (-mxxzMyyz);
- wadjust = OxyyPxzz + (c1o1 - OxyyPxzz)*abs(mxyyPxzz) / (abs(mxyyPxzz) + qudricLimitP);
- mxyyPxzz += wadjust * (-mxyyPxzz);
- wadjust = OxyyMxzz + (c1o1 - OxyyMxzz)*abs(mxyyMxzz) / (abs(mxyyMxzz) + qudricLimitM);
- mxyyMxzz += wadjust * (-mxyyMxzz);
- //////////////////////////////////////////////////////////////////////////
- //ohne limiter
- //mfbbb += OxyyMxzz * (-mfbbb);
- //mxxyPyzz += OxyyPxzz * (-mxxyPyzz);
- //mxxyMyzz += OxyyMxzz * (-mxxyMyzz);
- //mxxzPyyz += OxyyPxzz * (-mxxzPyyz);
- //mxxzMyyz += OxyyMxzz * (-mxxzMyyz);
- //mxyyPxzz += OxyyPxzz * (-mxyyPxzz);
- //mxyyMxzz += OxyyMxzz * (-mxyyMxzz);
- //////////////////////////////////////////////////////////////////////////
-
- // linear combinations back
- mfcba = (mxxyMyzz + mxxyPyzz) * c1o2;
- mfabc = (-mxxyMyzz + mxxyPyzz) * c1o2;
- mfcab = (mxxzMyyz + mxxzPyyz) * c1o2;
- mfacb = (-mxxzMyyz + mxxzPyyz) * c1o2;
- mfbca = (mxyyMxzz + mxyyPxzz) * c1o2;
- mfbac = (-mxyyMxzz + mxyyPxzz) * c1o2;
-
- //4.
- //////////////////////////////////////////////////////////////////////////
- //mit limiter
- // wadjust = O4+(one-O4)*abs(CUMacc)/(abs(CUMacc)+qudricLimit);
- //CUMacc += wadjust * (-CUMacc);
- // wadjust = O4+(one-O4)*abs(CUMcac)/(abs(CUMcac)+qudricLimit);
- //CUMcac += wadjust * (-CUMcac);
- // wadjust = O4+(one-O4)*abs(CUMcca)/(abs(CUMcca)+qudricLimit);
- //CUMcca += wadjust * (-CUMcca);
-
- // wadjust = O4+(one-O4)*abs(CUMbbc)/(abs(CUMbbc)+qudricLimit);
- //CUMbbc += wadjust * (-CUMbbc);
- // wadjust = O4+(one-O4)*abs(CUMbcb)/(abs(CUMbcb)+qudricLimit);
- //CUMbcb += wadjust * (-CUMbcb);
- // wadjust = O4+(one-O4)*abs(CUMcbb)/(abs(CUMcbb)+qudricLimit);
- //CUMcbb += wadjust * (-CUMcbb);
- //////////////////////////////////////////////////////////////////////////
- real factorA = (c4o1 + c2o1*omega - c3o1*omega*omega) / (c2o1 - c7o1*omega + c5o1*omega*omega);
- real factorB = (c4o1 + c28o1*omega - c14o1*omega*omega) / (c6o1 - c21o1*omega + c15o1*omega*omega);
- //////////////////////////////////////////////////////////////////////////
- //ohne limiter
- //CUMacc += O4 * (-CUMacc);
- //CUMcac += O4 * (-CUMcac);
- //CUMcca += O4 * (-CUMcca);
- //CUMbbc += O4 * (-CUMbbc);
- //CUMbcb += O4 * (-CUMbcb);
- //CUMcbb += O4 * (-CUMcbb);
- CUMacc = -O4*(c1o1 / omega - c1o2) * (dyuy + dzuz) * c2o3 * factorA + (c1o1 - O4) * (CUMacc);
- CUMcac = -O4*(c1o1 / omega - c1o2) * (dxux + dzuz) * c2o3 * factorA + (c1o1 - O4) * (CUMcac);
- CUMcca = -O4*(c1o1 / omega - c1o2) * (dyuy + dxux) * c2o3 * factorA + (c1o1 - O4) * (CUMcca);
- CUMbbc = -O4*(c1o1 / omega - c1o2) * Dxy * c1o3 * factorB + (c1o1 - O4) * (CUMbbc);
- CUMbcb = -O4*(c1o1 / omega - c1o2) * Dxz * c1o3 * factorB + (c1o1 - O4) * (CUMbcb);
- CUMcbb = -O4*(c1o1 / omega - c1o2) * Dyz * c1o3 * factorB + (c1o1 - O4) * (CUMcbb);
- //////////////////////////////////////////////////////////////////////////
-
-
- //5.
- CUMbcc += O5 * (-CUMbcc);
- CUMcbc += O5 * (-CUMcbc);
- CUMccb += O5 * (-CUMccb);
-
- //6.
- CUMccc += O6 * (-CUMccc);
-
-
-
- //back cumulants to central moments
- //4.
- mfcbb = CUMcbb + ((mfcaa + c1o3) * mfabb + c2o1 * mfbba * mfbab) / rho;
- mfbcb = CUMbcb + ((mfaca + c1o3) * mfbab + c2o1 * mfbba * mfabb) / rho;
- mfbbc = CUMbbc + ((mfaac + c1o3) * mfbba + c2o1 * mfbab * mfabb) / rho;
-
- mfcca = CUMcca + (((mfcaa * mfaca + c2o1 * mfbba * mfbba) + c1o3 * (mfcaa + mfaca)) / rho - c1o9*(drho / rho));
- mfcac = CUMcac + (((mfcaa * mfaac + c2o1 * mfbab * mfbab) + c1o3 * (mfcaa + mfaac)) / rho - c1o9*(drho / rho));
- mfacc = CUMacc + (((mfaac * mfaca + c2o1 * mfabb * mfabb) + c1o3 * (mfaac + mfaca)) / rho - c1o9*(drho / rho));
-
- //5.
- mfbcc = CUMbcc + ((mfaac * mfbca + mfaca * mfbac + c4o1 * mfabb * mfbbb + c2o1 * (mfbab * mfacb + mfbba * mfabc)) + c1o3 * (mfbca + mfbac)) / rho;
- mfcbc = CUMcbc + ((mfaac * mfcba + mfcaa * mfabc + c4o1 * mfbab * mfbbb + c2o1 * (mfabb * mfcab + mfbba * mfbac)) + c1o3 * (mfcba + mfabc)) / rho;
- mfccb = CUMccb + ((mfcaa * mfacb + mfaca * mfcab + c4o1 * mfbba * mfbbb + c2o1 * (mfbab * mfbca + mfabb * mfcba)) + c1o3 * (mfacb + mfcab)) / rho;
-
- //6.
-
- mfccc = CUMccc - ((-c4o1 * mfbbb * mfbbb
- - (mfcaa * mfacc + mfaca * mfcac + mfaac * mfcca)
- - c4o1 * (mfabb * mfcbb + mfbab * mfbcb + mfbba * mfbbc)
- - c2o1 * (mfbca * mfbac + mfcba * mfabc + mfcab * mfacb)) / rho
- + (c4o1 * (mfbab * mfbab * mfaca + mfabb * mfabb * mfcaa + mfbba * mfbba * mfaac)
- + c2o1 * (mfcaa * mfaca * mfaac)
- + c16o1 * mfbba * mfbab * mfabb) / (rho * rho)
- - c1o3 * (mfacc + mfcac + mfcca) / rho
- - c1o9 * (mfcaa + mfaca + mfaac) / rho
- + (c2o1 * (mfbab * mfbab + mfabb * mfabb + mfbba * mfbba)
- + (mfaac * mfaca + mfaac * mfcaa + mfaca * mfcaa) + c1o3 *(mfaac + mfaca + mfcaa)) / (rho * rho) * c2o3
- + c1o27*((drho * drho - drho) / (rho*rho)));
- ////////////////////////////////////////////////////////////////////////////////////
-
- ////////////////////////////////////////////////////////////////////////////////////
- //the force be with you
- mfbaa = -mfbaa;
- mfaba = -mfaba;
- mfaab = -mfaab;
- ////////////////////////////////////////////////////////////////////////////////////
-
-
- ////////////////////////////////////////////////////////////////////////////////////
- //back
- ////////////////////////////////////////////////////////////////////////////////////
- //mit 1, 0, 1/3, 0, 0, 0, 1/3, 0, 1/9 Konditionieren
- ////////////////////////////////////////////////////////////////////////////////////
- // Z - Dir
- m0 = mfaac * c1o2 + mfaab * (vvz - c1o2) + (mfaaa + c1o1* oMdrho) * (vz2 - vvz) * c1o2;
- m1 = -mfaac - c2o1* mfaab * vvz + mfaaa * (c1o1 - vz2) - c1o1* oMdrho * vz2;
- m2 = mfaac * c1o2 + mfaab * (vvz + c1o2) + (mfaaa + c1o1* oMdrho) * (vz2 + vvz) * c1o2;
- mfaaa = m0;
- mfaab = m1;
- mfaac = m2;
- ////////////////////////////////////////////////////////////////////////////////////
- m0 = mfabc * c1o2 + mfabb * (vvz - c1o2) + mfaba * (vz2 - vvz) * c1o2;
- m1 = -mfabc - c2o1* mfabb * vvz + mfaba * (c1o1 - vz2);
- m2 = mfabc * c1o2 + mfabb * (vvz + c1o2) + mfaba * (vz2 + vvz) * c1o2;
- mfaba = m0;
- mfabb = m1;
- mfabc = m2;
- ////////////////////////////////////////////////////////////////////////////////////
- m0 = mfacc * c1o2 + mfacb * (vvz - c1o2) + (mfaca + c1o3 * oMdrho) * (vz2 - vvz) * c1o2;
- m1 = -mfacc - c2o1* mfacb * vvz + mfaca * (c1o1 - vz2) - c1o3 * oMdrho * vz2;
- m2 = mfacc * c1o2 + mfacb * (vvz + c1o2) + (mfaca + c1o3 * oMdrho) * (vz2 + vvz) * c1o2;
- mfaca = m0;
- mfacb = m1;
- mfacc = m2;
- ////////////////////////////////////////////////////////////////////////////////////
- ////////////////////////////////////////////////////////////////////////////////////
- m0 = mfbac * c1o2 + mfbab * (vvz - c1o2) + mfbaa * (vz2 - vvz) * c1o2;
- m1 = -mfbac - c2o1* mfbab * vvz + mfbaa * (c1o1 - vz2);
- m2 = mfbac * c1o2 + mfbab * (vvz + c1o2) + mfbaa * (vz2 + vvz) * c1o2;
- mfbaa = m0;
- mfbab = m1;
- mfbac = m2;
- /////////b//////////////////////////////////////////////////////////////////////////
- m0 = mfbbc * c1o2 + mfbbb * (vvz - c1o2) + mfbba * (vz2 - vvz) * c1o2;
- m1 = -mfbbc - c2o1* mfbbb * vvz + mfbba * (c1o1 - vz2);
- m2 = mfbbc * c1o2 + mfbbb * (vvz + c1o2) + mfbba * (vz2 + vvz) * c1o2;
- mfbba = m0;
- mfbbb = m1;
- mfbbc = m2;
- /////////b//////////////////////////////////////////////////////////////////////////
- m0 = mfbcc * c1o2 + mfbcb * (vvz - c1o2) + mfbca * (vz2 - vvz) * c1o2;
- m1 = -mfbcc - c2o1* mfbcb * vvz + mfbca * (c1o1 - vz2);
- m2 = mfbcc * c1o2 + mfbcb * (vvz + c1o2) + mfbca * (vz2 + vvz) * c1o2;
- mfbca = m0;
- mfbcb = m1;
- mfbcc = m2;
- ////////////////////////////////////////////////////////////////////////////////////
- ////////////////////////////////////////////////////////////////////////////////////
- m0 = mfcac * c1o2 + mfcab * (vvz - c1o2) + (mfcaa + c1o3 * oMdrho) * (vz2 - vvz) * c1o2;
- m1 = -mfcac - c2o1* mfcab * vvz + mfcaa * (c1o1 - vz2) - c1o3 * oMdrho * vz2;
- m2 = mfcac * c1o2 + mfcab * (vvz + c1o2) + (mfcaa + c1o3 * oMdrho) * (vz2 + vvz) * c1o2;
- mfcaa = m0;
- mfcab = m1;
- mfcac = m2;
- /////////c//////////////////////////////////////////////////////////////////////////
- m0 = mfcbc * c1o2 + mfcbb * (vvz - c1o2) + mfcba * (vz2 - vvz) * c1o2;
- m1 = -mfcbc - c2o1* mfcbb * vvz + mfcba * (c1o1 - vz2);
- m2 = mfcbc * c1o2 + mfcbb * (vvz + c1o2) + mfcba * (vz2 + vvz) * c1o2;
- mfcba = m0;
- mfcbb = m1;
- mfcbc = m2;
- /////////c//////////////////////////////////////////////////////////////////////////
- m0 = mfccc * c1o2 + mfccb * (vvz - c1o2) + (mfcca + c1o9 * oMdrho) * (vz2 - vvz) * c1o2;
- m1 = -mfccc - c2o1* mfccb * vvz + mfcca * (c1o1 - vz2) - c1o9 * oMdrho * vz2;
- m2 = mfccc * c1o2 + mfccb * (vvz + c1o2) + (mfcca + c1o9 * oMdrho) * (vz2 + vvz) * c1o2;
- mfcca = m0;
- mfccb = m1;
- mfccc = m2;
- ////////////////////////////////////////////////////////////////////////////////////
- ////////////////////////////////////////////////////////////////////////////////////
- //mit 1/6, 2/3, 1/6, 0, 0, 0, 1/18, 2/9, 1/18 Konditionieren
- ////////////////////////////////////////////////////////////////////////////////////
- // Y - Dir
- m0 = mfaca * c1o2 + mfaba * (vvy - c1o2) + (mfaaa + c1o6 * oMdrho) * (vy2 - vvy) * c1o2;
- m1 = -mfaca - c2o1* mfaba * vvy + mfaaa * (c1o1 - vy2) - c1o6 * oMdrho * vy2;
- m2 = mfaca * c1o2 + mfaba * (vvy + c1o2) + (mfaaa + c1o6 * oMdrho) * (vy2 + vvy) * c1o2;
- mfaaa = m0;
- mfaba = m1;
- mfaca = m2;
- ////////////////////////////////////////////////////////////////////////////////////
- m0 = mfacb * c1o2 + mfabb * (vvy - c1o2) + (mfaab + c2o3 * oMdrho) * (vy2 - vvy) * c1o2;
- m1 = -mfacb - c2o1* mfabb * vvy + mfaab * (c1o1 - vy2) - c2o3 * oMdrho * vy2;
- m2 = mfacb * c1o2 + mfabb * (vvy + c1o2) + (mfaab + c2o3 * oMdrho) * (vy2 + vvy) * c1o2;
- mfaab = m0;
- mfabb = m1;
- mfacb = m2;
- ////////////////////////////////////////////////////////////////////////////////////
- m0 = mfacc * c1o2 + mfabc * (vvy - c1o2) + (mfaac + c1o6 * oMdrho) * (vy2 - vvy) * c1o2;
- m1 = -mfacc - c2o1* mfabc * vvy + mfaac * (c1o1 - vy2) - c1o6 * oMdrho * vy2;
- m2 = mfacc * c1o2 + mfabc * (vvy + c1o2) + (mfaac + c1o6 * oMdrho) * (vy2 + vvy) * c1o2;
- mfaac = m0;
- mfabc = m1;
- mfacc = m2;
- ////////////////////////////////////////////////////////////////////////////////////
- ////////////////////////////////////////////////////////////////////////////////////
- m0 = mfbca * c1o2 + mfbba * (vvy - c1o2) + mfbaa * (vy2 - vvy) * c1o2;
- m1 = -mfbca - c2o1* mfbba * vvy + mfbaa * (c1o1 - vy2);
- m2 = mfbca * c1o2 + mfbba * (vvy + c1o2) + mfbaa * (vy2 + vvy) * c1o2;
- mfbaa = m0;
- mfbba = m1;
- mfbca = m2;
- /////////b//////////////////////////////////////////////////////////////////////////
- m0 = mfbcb * c1o2 + mfbbb * (vvy - c1o2) + mfbab * (vy2 - vvy) * c1o2;
- m1 = -mfbcb - c2o1* mfbbb * vvy + mfbab * (c1o1 - vy2);
- m2 = mfbcb * c1o2 + mfbbb * (vvy + c1o2) + mfbab * (vy2 + vvy) * c1o2;
- mfbab = m0;
- mfbbb = m1;
- mfbcb = m2;
- /////////b//////////////////////////////////////////////////////////////////////////
- m0 = mfbcc * c1o2 + mfbbc * (vvy - c1o2) + mfbac * (vy2 - vvy) * c1o2;
- m1 = -mfbcc - c2o1* mfbbc * vvy + mfbac * (c1o1 - vy2);
- m2 = mfbcc * c1o2 + mfbbc * (vvy + c1o2) + mfbac * (vy2 + vvy) * c1o2;
- mfbac = m0;
- mfbbc = m1;
- mfbcc = m2;
- ////////////////////////////////////////////////////////////////////////////////////
- ////////////////////////////////////////////////////////////////////////////////////
- m0 = mfcca * c1o2 + mfcba * (vvy - c1o2) + (mfcaa + c1o18 * oMdrho) * (vy2 - vvy) * c1o2;
- m1 = -mfcca - c2o1* mfcba * vvy + mfcaa * (c1o1 - vy2) - c1o18 * oMdrho * vy2;
- m2 = mfcca * c1o2 + mfcba * (vvy + c1o2) + (mfcaa + c1o18 * oMdrho) * (vy2 + vvy) * c1o2;
- mfcaa = m0;
- mfcba = m1;
- mfcca = m2;
- /////////c//////////////////////////////////////////////////////////////////////////
- m0 = mfccb * c1o2 + mfcbb * (vvy - c1o2) + (mfcab + c2o9 * oMdrho) * (vy2 - vvy) * c1o2;
- m1 = -mfccb - c2o1* mfcbb * vvy + mfcab * (c1o1 - vy2) - c2o9 * oMdrho * vy2;
- m2 = mfccb * c1o2 + mfcbb * (vvy + c1o2) + (mfcab + c2o9 * oMdrho) * (vy2 + vvy) * c1o2;
- mfcab = m0;
- mfcbb = m1;
- mfccb = m2;
- /////////c//////////////////////////////////////////////////////////////////////////
- m0 = mfccc * c1o2 + mfcbc * (vvy - c1o2) + (mfcac + c1o18 * oMdrho) * (vy2 - vvy) * c1o2;
- m1 = -mfccc - c2o1* mfcbc * vvy + mfcac * (c1o1 - vy2) - c1o18 * oMdrho * vy2;
- m2 = mfccc * c1o2 + mfcbc * (vvy + c1o2) + (mfcac + c1o18 * oMdrho) * (vy2 + vvy) * c1o2;
- mfcac = m0;
- mfcbc = m1;
- mfccc = m2;
- ////////////////////////////////////////////////////////////////////////////////////
- ////////////////////////////////////////////////////////////////////////////////////
- //mit 1/36, 1/9, 1/36, 1/9, 4/9, 1/9, 1/36, 1/9, 1/36 Konditionieren
- ////////////////////////////////////////////////////////////////////////////////////
- // X - Dir
- m0 = mfcaa * c1o2 + mfbaa * (vvx - c1o2) + (mfaaa + c1o36 * oMdrho) * (vx2 - vvx) * c1o2;
- m1 = -mfcaa - c2o1* mfbaa * vvx + mfaaa * (c1o1 - vx2) - c1o36 * oMdrho * vx2;
- m2 = mfcaa * c1o2 + mfbaa * (vvx + c1o2) + (mfaaa + c1o36 * oMdrho) * (vx2 + vvx) * c1o2;
- mfaaa = m0;
- mfbaa = m1;
- mfcaa = m2;
- ////////////////////////////////////////////////////////////////////////////////////
- m0 = mfcba * c1o2 + mfbba * (vvx - c1o2) + (mfaba + c1o9 * oMdrho) * (vx2 - vvx) * c1o2;
- m1 = -mfcba - c2o1* mfbba * vvx + mfaba * (c1o1 - vx2) - c1o9 * oMdrho * vx2;
- m2 = mfcba * c1o2 + mfbba * (vvx + c1o2) + (mfaba + c1o9 * oMdrho) * (vx2 + vvx) * c1o2;
- mfaba = m0;
- mfbba = m1;
- mfcba = m2;
- ////////////////////////////////////////////////////////////////////////////////////
- m0 = mfcca * c1o2 + mfbca * (vvx - c1o2) + (mfaca + c1o36 * oMdrho) * (vx2 - vvx) * c1o2;
- m1 = -mfcca - c2o1* mfbca * vvx + mfaca * (c1o1 - vx2) - c1o36 * oMdrho * vx2;
- m2 = mfcca * c1o2 + mfbca * (vvx + c1o2) + (mfaca + c1o36 * oMdrho) * (vx2 + vvx) * c1o2;
- mfaca = m0;
- mfbca = m1;
- mfcca = m2;
- ////////////////////////////////////////////////////////////////////////////////////
- ////////////////////////////////////////////////////////////////////////////////////
- m0 = mfcab * c1o2 + mfbab * (vvx - c1o2) + (mfaab + c1o9 * oMdrho) * (vx2 - vvx) * c1o2;
- m1 = -mfcab - c2o1* mfbab * vvx + mfaab * (c1o1 - vx2) - c1o9 * oMdrho * vx2;
- m2 = mfcab * c1o2 + mfbab * (vvx + c1o2) + (mfaab + c1o9 * oMdrho) * (vx2 + vvx) * c1o2;
- mfaab = m0;
- mfbab = m1;
- mfcab = m2;
- ///////////b////////////////////////////////////////////////////////////////////////
- m0 = mfcbb * c1o2 + mfbbb * (vvx - c1o2) + (mfabb + c4o9 * oMdrho) * (vx2 - vvx) * c1o2;
- m1 = -mfcbb - c2o1* mfbbb * vvx + mfabb * (c1o1 - vx2) - c4o9 * oMdrho * vx2;
- m2 = mfcbb * c1o2 + mfbbb * (vvx + c1o2) + (mfabb + c4o9 * oMdrho) * (vx2 + vvx) * c1o2;
- mfabb = m0;
- mfbbb = m1;
- mfcbb = m2;
- ///////////b////////////////////////////////////////////////////////////////////////
- m0 = mfccb * c1o2 + mfbcb * (vvx - c1o2) + (mfacb + c1o9 * oMdrho) * (vx2 - vvx) * c1o2;
- m1 = -mfccb - c2o1* mfbcb * vvx + mfacb * (c1o1 - vx2) - c1o9 * oMdrho * vx2;
- m2 = mfccb * c1o2 + mfbcb * (vvx + c1o2) + (mfacb + c1o9 * oMdrho) * (vx2 + vvx) * c1o2;
- mfacb = m0;
- mfbcb = m1;
- mfccb = m2;
- ////////////////////////////////////////////////////////////////////////////////////
- ////////////////////////////////////////////////////////////////////////////////////
- m0 = mfcac * c1o2 + mfbac * (vvx - c1o2) + (mfaac + c1o36 * oMdrho) * (vx2 - vvx) * c1o2;
- m1 = -mfcac - c2o1* mfbac * vvx + mfaac * (c1o1 - vx2) - c1o36 * oMdrho * vx2;
- m2 = mfcac * c1o2 + mfbac * (vvx + c1o2) + (mfaac + c1o36 * oMdrho) * (vx2 + vvx) * c1o2;
- mfaac = m0;
- mfbac = m1;
- mfcac = m2;
- ///////////c////////////////////////////////////////////////////////////////////////
- m0 = mfcbc * c1o2 + mfbbc * (vvx - c1o2) + (mfabc + c1o9 * oMdrho) * (vx2 - vvx) * c1o2;
- m1 = -mfcbc - c2o1* mfbbc * vvx + mfabc * (c1o1 - vx2) - c1o9 * oMdrho * vx2;
- m2 = mfcbc * c1o2 + mfbbc * (vvx + c1o2) + (mfabc + c1o9 * oMdrho) * (vx2 + vvx) * c1o2;
- mfabc = m0;
- mfbbc = m1;
- mfcbc = m2;
- ///////////c////////////////////////////////////////////////////////////////////////
- m0 = mfccc * c1o2 + mfbcc * (vvx - c1o2) + (mfacc + c1o36 * oMdrho) * (vx2 - vvx) * c1o2;
- m1 = -mfccc - c2o1* mfbcc * vvx + mfacc * (c1o1 - vx2) - c1o36 * oMdrho * vx2;
- m2 = mfccc * c1o2 + mfbcc * (vvx + c1o2) + (mfacc + c1o36 * oMdrho) * (vx2 + vvx) * c1o2;
- mfacc = m0;
- mfbcc = m1;
- mfccc = m2;
- ////////////////////////////////////////////////////////////////////////////////////
-
- ////////////////////////////////////////////////////////////////////////////////////
- (D.f[DIR_P00])[k] = mfabb;//(D.f[ DIR_P00 ])[ke ] = mfabb;// - c2over27 ; (D.f[ DIR_P00 ])[k ]
- (D.f[DIR_M00])[kw] = mfcbb;//(D.f[ DIR_M00 ])[kw ] = mfcbb;// - c2over27 ; (D.f[ DIR_M00 ])[kw ]
- (D.f[DIR_0P0])[k] = mfbab;//(D.f[ DIR_0P0 ])[kn ] = mfbab;// - c2over27 ; (D.f[ DIR_0P0 ])[k ]
- (D.f[DIR_0M0])[ks] = mfbcb;//(D.f[ DIR_0M0 ])[ks ] = mfbcb;// - c2over27 ; (D.f[ DIR_0M0 ])[ks ]
- (D.f[DIR_00P])[k] = mfbba;//(D.f[ DIR_00P ])[kt ] = mfbba;// - c2over27 ; (D.f[ DIR_00P ])[k ]
- (D.f[DIR_00M])[kb] = mfbbc;//(D.f[ DIR_00M ])[kb ] = mfbbc;// - c2over27 ; (D.f[ DIR_00M ])[kb ]
- (D.f[DIR_PP0])[k] = mfaab;//(D.f[ DIR_PP0 ])[kne ] = mfaab;// - c1over54 ; (D.f[ DIR_PP0 ])[k ]
- (D.f[DIR_MM0])[ksw] = mfccb;//(D.f[ DIR_MM0 ])[ksw ] = mfccb;// - c1over54 ; (D.f[ DIR_MM0 ])[ksw ]
- (D.f[DIR_PM0])[ks] = mfacb;//(D.f[ DIR_PM0 ])[kse ] = mfacb;// - c1over54 ; (D.f[ DIR_PM0 ])[ks ]
- (D.f[DIR_MP0])[kw] = mfcab;//(D.f[ DIR_MP0 ])[knw ] = mfcab;// - c1over54 ; (D.f[ DIR_MP0 ])[kw ]
- (D.f[DIR_P0P])[k] = mfaba;//(D.f[ DIR_P0P ])[kte ] = mfaba;// - c1over54 ; (D.f[ DIR_P0P ])[k ]
- (D.f[DIR_M0M])[kbw] = mfcbc;//(D.f[ DIR_M0M ])[kbw ] = mfcbc;// - c1over54 ; (D.f[ DIR_M0M ])[kbw ]
- (D.f[DIR_P0M])[kb] = mfabc;//(D.f[ DIR_P0M ])[kbe ] = mfabc;// - c1over54 ; (D.f[ DIR_P0M ])[kb ]
- (D.f[DIR_M0P])[kw] = mfcba;//(D.f[ DIR_M0P ])[ktw ] = mfcba;// - c1over54 ; (D.f[ DIR_M0P ])[kw ]
- (D.f[DIR_0PP])[k] = mfbaa;//(D.f[ DIR_0PP ])[ktn ] = mfbaa;// - c1over54 ; (D.f[ DIR_0PP ])[k ]
- (D.f[DIR_0MM])[kbs] = mfbcc;//(D.f[ DIR_0MM ])[kbs ] = mfbcc;// - c1over54 ; (D.f[ DIR_0MM ])[kbs ]
- (D.f[DIR_0PM])[kb] = mfbac;//(D.f[ DIR_0PM ])[kbn ] = mfbac;// - c1over54 ; (D.f[ DIR_0PM ])[kb ]
- (D.f[DIR_0MP])[ks] = mfbca;//(D.f[ DIR_0MP ])[kts ] = mfbca;// - c1over54 ; (D.f[ DIR_0MP ])[ks ]
- (D.f[DIR_000])[k] = mfbbb;//(D.f[ DIR_000])[kzero] = mfbbb;// - c8over27 ; (D.f[ DIR_000])[k ]
- (D.f[DIR_PPP])[k] = mfaaa;//(D.f[ DIR_PPP ])[ktne ] = mfaaa;// - c1over216; (D.f[ DIR_PPP ])[k ]
- (D.f[DIR_PMP])[ks] = mfaca;//(D.f[ DIR_PMP ])[ktse ] = mfaca;// - c1over216; (D.f[ DIR_PMP ])[ks ]
- (D.f[DIR_PPM])[kb] = mfaac;//(D.f[ DIR_PPM ])[kbne ] = mfaac;// - c1over216; (D.f[ DIR_PPM ])[kb ]
- (D.f[DIR_PMM])[kbs] = mfacc;//(D.f[ DIR_PMM ])[kbse ] = mfacc;// - c1over216; (D.f[ DIR_PMM ])[kbs ]
- (D.f[DIR_MPP])[kw] = mfcaa;//(D.f[ DIR_MPP ])[ktnw ] = mfcaa;// - c1over216; (D.f[ DIR_MPP ])[kw ]
- (D.f[DIR_MMP])[ksw] = mfcca;//(D.f[ DIR_MMP ])[ktsw ] = mfcca;// - c1over216; (D.f[ DIR_MMP ])[ksw ]
- (D.f[DIR_MPM])[kbw] = mfcac;//(D.f[ DIR_MPM ])[kbnw ] = mfcac;// - c1over216; (D.f[ DIR_MPM ])[kbw ]
- (D.f[DIR_MMM])[kbsw] = mfccc;//(D.f[ DIR_MMM ])[kbsw ] = mfccc;// - c1over216; (D.f[ DIR_MMM ])[kbsw]
- ////////////////////////////////////////////////////////////////////////////////////
- }
- }
-}
\ No newline at end of file
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK17/CumulantK17Comp_Device.cuh b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK17/CumulantK17Comp_Device.cuh
deleted file mode 100644
index f44842057d554498b0b5d4c733e2425e524a3b75..0000000000000000000000000000000000000000
--- a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK17/CumulantK17Comp_Device.cuh
+++ /dev/null
@@ -1,18 +0,0 @@
-#ifndef LB_Kernel_CUMULANT_K17_COMP_H
-#define LB_Kernel_CUMULANT_K17_COMP_H
-
-#include <DataTypes.h>
-#include <curand.h>
-
-__global__ void LB_Kernel_CumulantK17Comp( real omega,
- unsigned int* bcMatD,
- unsigned int* neighborX,
- unsigned int* neighborY,
- unsigned int* neighborZ,
- real* DDStart,
- int size_Mat,
- int level,
- real* forces,
- real* quadricLimiters,
- bool EvenOrOdd);
-#endif
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK17chimRedesigned/CumulantK17CompChimRedesignedDevice.cu b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK17/CumulantK17_Device.cu
similarity index 63%
rename from src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK17chimRedesigned/CumulantK17CompChimRedesignedDevice.cu
rename to src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK17/CumulantK17_Device.cu
index db8caf1b23c2087a4c5c76886fb4530bc6272a1d..1ffec96c255b7923f3ee39c01f756abd8cad8862 100644
--- a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK17chimRedesigned/CumulantK17CompChimRedesignedDevice.cu
+++ b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK17/CumulantK17_Device.cu
@@ -1,55 +1,78 @@
-//=======================================================================================
-// ____ ____ __ ______ __________ __ __ __ __
-// \ \ | | | | | _ \ |___ ___| | | | | / \ | |
-// \ \ | | | | | |_) | | | | | | | / \ | |
-// \ \ | | | | | _ / | | | | | | / /\ \ | |
-// \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
-// \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
-// \ \ | | ________________________________________________________________
-// \ \ | | | ______________________________________________________________|
-// \ \| | | | __ __ __ __ ______ _______
-// \ | | |_____ | | | | | | | | | _ \ / _____)
-// \ | | _____| | | | | | | | | | | \ \ \_______
+
+// ____ ____ __ ______ __________ __ __ __ __
+// \ \ | | | | | _ \ |___ ___| | | | | / \ | |
+// \ \ | | | | | |_) | | | | | | | / \ | |
+// \ \ | | | | | _ / | | | | | | / /\ \ | |
+// \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
+// \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
+// \ \ | | ________________________________________________________________
+// \ \ | | | ______________________________________________________________|
+// \ \| | | | __ __ __ __ ______ _______
+// \ | | |_____ | | | | | | | | | _ \ / _____)
+// \ | | _____| | | | | | | | | | | \ \ \_______
// \ | | | | |_____ | \_/ | | | | |_/ / _____ |
-// \ _____| |__| |________| \_______/ |__| |______/ (_______/
+// \ _____| |__| |________| \_______/ |__| |______/ (_______/
//
-// This file is part of VirtualFluids. VirtualFluids is free software: you can
+// This file is part of VirtualFluids. VirtualFluids is free software: you can
// redistribute it and/or modify it under the terms of the GNU General Public
-// License as published by the Free Software Foundation, either version 3 of
+// License as published by the Free Software Foundation, either version 3 of
// the License, or (at your option) any later version.
-//
-// VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
-// ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
-// FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
+//
+// VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
+// ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+// FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
// for more details.
-//
+//
// You should have received a copy of the GNU General Public License along
// with VirtualFluids (see COPYING.txt). If not, see <http://www.gnu.org/licenses/>.
//
-//! \file Cumulant27chimStream.cu
-//! \ingroup GPU
-//! \author Martin Schoenherr, Anna Wellmann
+//! \file CumlantK17_Device.cu
+//! \author Anna Wellmann, Martin Schönherr, Henry Korb, Henrik Asmuth
+//! \date 05/12/2022
+//! \brief Kernel for CumulantK17 including different turbulence models and options for local body forces and writing macroscopic variables
+//!
+//! CumulantK17 kernel using chimera transformations and quartic limiters as present in Geier et al. (2017). Additional options are three different
+//! eddy-viscosity turbulence models (Smagorinsky, AMD, QR) that can be set via the template parameter turbulenceModel (with default
+//! TurbulenceModel::None).
+//! The kernel is executed separately for each subset of fluid node indices with a different tag CollisionTemplate. For each subset, only the locally
+//! required options are switched on ( \param writeMacroscopicVariables and/or \param applyBodyForce) in order to minimize memory accesses. The default
+//! refers to the plain cumlant kernel (CollisionTemplate::Default).
+//! Nodes are added to subsets (taggedFluidNodes) in Simulation::init using a corresponding tag with different values of CollisionTemplate. These subsets
+//! are provided by the utilized PostCollisionInteractiors depending on they specific requirements (e.g. writeMacroscopicVariables for probes).
+
//=======================================================================================
-/* Device code */
-#include "LBM/LB.h"
+#include "LBM/LB.h"
#include "lbm/constants/D3Q27.h"
-#include <lbm/constants/NumericConstants.h>
-#include "Kernel/Utilities/DistributionHelper.cuh"
+#include "lbm/constants/NumericConstants.h"
+#include "LBM/GPUHelperFunctions/KernelUtilities.h"
+#include "LBM/GPUHelperFunctions/ChimeraTransformation.h"
+
+#include "GPU/TurbulentViscosityInlines.cuh"
using namespace vf::lbm::constant;
using namespace vf::lbm::dir;
-#include "Kernel/Utilities/ChimeraTransformation.h"
+using namespace vf::gpu;
////////////////////////////////////////////////////////////////////////////////
-__global__ void LB_Kernel_CumulantK17CompChimRedesigned(
- real omega,
+template<TurbulenceModel turbulenceModel, bool writeMacroscopicVariables, bool applyBodyForce>
+__global__ void LB_Kernel_CumulantK17(
+ real omega_in,
uint* neighborX,
uint* neighborY,
uint* neighborZ,
real* distributions,
- unsigned long numberOfLBnodes,
+ real* rho,
+ real* vx,
+ real* vy,
+ real* vz,
+ real* turbulentViscosity,
+ real SGSconstant,
+ unsigned long long numberOfLBnodes,
int level,
real* forces,
+ real* bodyForceX,
+ real* bodyForceY,
+ real* bodyForceZ,
real* quadricLimiters,
bool isEvenTimestep,
const uint *fluidNodeIndices,
@@ -64,19 +87,18 @@ __global__ void LB_Kernel_CumulantK17CompChimRedesigned(
//! The cumulant kernel is executed in the following steps
//!
////////////////////////////////////////////////////////////////////////////////
- //! - Get the thread index coordinates from threadIdx, blockIdx, blockDim and gridDim.
+ //! - Get node index coordinates from threadIdx, blockIdx, blockDim and gridDim.
//!
- const unsigned kThread = vf::gpu::getNodeIndex();
+ const unsigned nodeIndex = getNodeIndex();
//////////////////////////////////////////////////////////////////////////
- //! - Return for non-fluid nodes
- if (kThread >= numberOfFluidNodes)
+ // run for all indices in size_Mat and fluid nodes
+ if (nodeIndex >= numberOfFluidNodes)
return;
-
////////////////////////////////////////////////////////////////////////////////
//! - Get the node index from the array containing all indices of fluid nodes
//!
- const unsigned k_000 = fluidNodeIndices[kThread];
+ const unsigned k_000 = fluidNodeIndices[nodeIndex];
//////////////////////////////////////////////////////////////////////////
//! - Read distributions: style of reading and writing the distributions from/to stored arrays dependent on
@@ -84,11 +106,11 @@ __global__ void LB_Kernel_CumulantK17CompChimRedesigned(
//! href="https://doi.org/10.3390/computation5020019"><b>[ M. Geier et al. (2017),
//! DOI:10.3390/computation5020019 ]</b></a>
//!
- Distributions27 dist = vf::gpu::getDistributionReferences27(distributions, numberOfLBnodes, isEvenTimestep);
-
+ Distributions27 dist;
+ getPointersToDistributions(dist, distributions, numberOfLBnodes, isEvenTimestep);
+
////////////////////////////////////////////////////////////////////////////////
//! - Set neighbor indices (necessary for indirect addressing)
- //!
uint k_M00 = neighborX[k_000];
uint k_0M0 = neighborY[k_000];
uint k_00M = neighborZ[k_000];
@@ -96,9 +118,8 @@ __global__ void LB_Kernel_CumulantK17CompChimRedesigned(
uint k_M0M = neighborZ[k_M00];
uint k_0MM = neighborZ[k_0M0];
uint k_MMM = neighborZ[k_MM0];
-
////////////////////////////////////////////////////////////////////////////////////
- //! - Set local distributions (f's):
+ //! - Set local distributions
//!
real f_000 = (dist.f[DIR_000])[k_000];
real f_P00 = (dist.f[DIR_P00])[k_000];
@@ -159,28 +180,28 @@ __global__ void LB_Kernel_CumulantK17CompChimRedesigned(
real& m_200 = f_PMM;
real& m_000 = f_MMM;
- ////////////////////////////////////////////////////////////////////////////////////
+ //////////////////////////////////////////////////////(unsigned long)//////////////////////////////
//! - Calculate density and velocity using pyramid summation for low round-off errors as in Eq. (J1)-(J3) \ref
//! <a href="https://doi.org/10.1016/j.camwa.2015.05.001"><b>[ M. Geier et al. (2015),
//! DOI:10.1016/j.camwa.2015.05.001 ]</b></a>
//!
real drho = ((((f_PPP + f_MMM) + (f_MPM + f_PMP)) + ((f_MPP + f_PMM) + (f_MMP + f_PPM))) +
- (((f_0MP + f_0PM) + (f_0MM + f_0PP)) + ((f_M0P + f_P0M) + (f_M0M + f_P0P)) +
- ((f_MP0 + f_PM0) + (f_MM0 + f_PP0))) +
- ((f_M00 + f_P00) + (f_0M0 + f_0P0) + (f_00M + f_00P))) +
+ (((f_0MP + f_0PM) + (f_0MM + f_0PP)) + ((f_M0P + f_P0M) + (f_M0M + f_P0P)) +
+ ((f_MP0 + f_PM0) + (f_MM0 + f_PP0))) +
+ ((f_M00 + f_P00) + (f_0M0 + f_0P0) + (f_00M + f_00P))) +
f_000;
real oneOverRho = c1o1 / (c1o1 + drho);
real vvx = ((((f_PPP - f_MMM) + (f_PMP - f_MPM)) + ((f_PMM - f_MPP) + (f_PPM - f_MMP))) +
(((f_P0M - f_M0P) + (f_P0P - f_M0M)) + ((f_PM0 - f_MP0) + (f_PP0 - f_MM0))) + (f_P00 - f_M00)) *
- oneOverRho;
+ oneOverRho;
real vvy = ((((f_PPP - f_MMM) + (f_MPM - f_PMP)) + ((f_MPP - f_PMM) + (f_PPM - f_MMP))) +
(((f_0PM - f_0MP) + (f_0PP - f_0MM)) + ((f_MP0 - f_PM0) + (f_PP0 - f_MM0))) + (f_0P0 - f_0M0)) *
- oneOverRho;
+ oneOverRho;
real vvz = ((((f_PPP - f_MMM) + (f_PMP - f_MPM)) + ((f_MPP - f_PMM) + (f_MMP - f_PPM))) +
(((f_0MP - f_0PM) + (f_0PP - f_0MM)) + ((f_M0P - f_P0M) + (f_P0P - f_M0M))) + (f_00P - f_00M)) *
- oneOverRho;
+ oneOverRho;
////////////////////////////////////////////////////////////////////////////////////
//! - Add half of the acceleration (body force) to the velocity as in Eq. (42) \ref
@@ -188,16 +209,55 @@ __global__ void LB_Kernel_CumulantK17CompChimRedesigned(
//! DOI:10.1016/j.camwa.2015.05.001 ]</b></a>
//!
real factor = c1o1;
- // The factor has to be scaled for each level to get the correct acceleration.
for (size_t i = 1; i <= level; i++) {
factor *= c2o1;
}
- real fx = forces[0] / factor;
- real fy = forces[1] / factor;
- real fz = forces[2] / factor;
- vvx += fx * c1o2;
- vvy += fy * c1o2;
- vvz += fz * c1o2;
+
+ real fx = forces[0];
+ real fy = forces[1];
+ real fz = forces[2];
+
+ if( applyBodyForce ){
+ fx += bodyForceX[k_000];
+ fy += bodyForceY[k_000];
+ fz += bodyForceZ[k_000];
+
+ // real vx = vvx;
+ // real vy = vvy;
+ // real vz = vvz;
+ real acc_x = fx * c1o2 / factor;
+ real acc_y = fy * c1o2 / factor;
+ real acc_z = fz * c1o2 / factor;
+
+ vvx += acc_x;
+ vvy += acc_y;
+ vvz += acc_z;
+
+ // Reset body force. To be used when not using round-off correction.
+ bodyForceX[k_000] = 0.0f;
+ bodyForceY[k_000] = 0.0f;
+ bodyForceZ[k_000] = 0.0f;
+
+ ////////////////////////////////////////////////////////////////////////////////////
+ //!> Round-off correction
+ //!
+ //!> Similar to Kahan summation algorithm (https://en.wikipedia.org/wiki/Kahan_summation_algorithm)
+ //!> Essentially computes the round-off error of the applied force and adds it in the next time step as a compensation.
+ //!> Seems to be necesseary at very high Re boundary layers, where the forcing and velocity can
+ //!> differ by several orders of magnitude.
+ //!> \note 16/05/2022: Testing, still ongoing!
+ //!
+ // bodyForceX[k_000] = (acc_x-(vvx-vx))*factor*c2o1;
+ // bodyForceY[k_000] = (acc_y-(vvy-vy))*factor*c2o1;
+ // bodyForceZ[k_000] = (acc_z-(vvz-vz))*factor*c2o1;
+ }
+ else{
+ vvx += fx * c1o2 / factor;
+ vvy += fy * c1o2 / factor;
+ vvz += fz * c1o2 / factor;
+ }
+
+
////////////////////////////////////////////////////////////////////////////////////
// calculate the square of velocities for this lattice node
real vx2 = vvx * vvx;
@@ -272,15 +332,21 @@ __global__ void LB_Kernel_CumulantK17CompChimRedesigned(
//! - Fifth order cumulants \f$ C_{221}, C_{212}, C_{122}\f$: \f$\omega_9=O5=1.0\f$.
//! - Sixth order cumulant \f$ C_{222}\f$: \f$\omega_{10}=O6=1.0\f$.
//!
+ ////////////////////////////////////////////////////////////////////////////////////
+ //! - Calculate modified omega with turbulent viscosity
+ //!
+ real omega = omega_in;
+ if(turbulenceModel != TurbulenceModel::None){ omega /= (c1o1 + c3o1*omega_in*turbulentViscosity[k_000]); }
////////////////////////////////////////////////////////////
// 2.
real OxxPyyPzz = c1o1;
////////////////////////////////////////////////////////////
// 3.
- real OxyyPxzz = c8o1 * (-c2o1 + omega) * (c1o1 + c2o1 * omega) / (-c8o1 - c14o1 * omega + c7o1 * omega * omega);
- real OxyyMxzz = c8o1 * (-c2o1 + omega) * (-c7o1 + c4o1 * omega) / (c56o1 - c50o1 * omega + c9o1 * omega * omega);
- real Oxyz = c24o1 * (-c2o1 + omega) * (-c2o1 - c7o1 * omega + c3o1 * omega * omega) /
- (c48o1 + c152o1 * omega - c130o1 * omega * omega + c29o1 * omega * omega * omega);
+ real OxyyPxzz = c8o1 * (-c2o1 + omega) * (c1o1 + c2o1 * omega) / (-c8o1 - c14o1 * omega + c7o1 * omega * omega);
+ real OxyyMxzz =
+ c8o1 * (-c2o1 + omega) * (-c7o1 + c4o1 * omega) / (c56o1 - c50o1 * omega + c9o1 * omega * omega);
+ real Oxyz = c24o1 * (-c2o1 + omega) * (-c2o1 - c7o1 * omega + c3o1 * omega * omega) /
+ (c48o1 + c152o1 * omega - c130o1 * omega * omega + c29o1 * omega * omega * omega);
////////////////////////////////////////////////////////////
// 4.
real O4 = c1o1;
@@ -292,16 +358,16 @@ __global__ void LB_Kernel_CumulantK17CompChimRedesigned(
real O6 = c1o1;
////////////////////////////////////////////////////////////////////////////////////
- //! - A and B: parameters for fourth order convergence of the diffusion term according to Eq. (115) and (116)
+ //! - A and DIR_00M: parameters for fourth order convergence of the diffusion term according to Eq. (115) and (116)
//! <a href="https://doi.org/10.1016/j.jcp.2017.05.040"><b>[ M. Geier et al. (2017),
//! DOI:10.1016/j.jcp.2017.05.040 ]</b></a> with simplifications assuming \f$ \omega_2 = 1.0 \f$ (modify for
//! different bulk viscosity).
//!
- real factorA = (c4o1 + c2o1 * omega - c3o1 * omega * omega) / (c2o1 - c7o1 * omega + c5o1 * omega * omega);
+ real factorA = (c4o1 + c2o1 * omega - c3o1 * omega * omega) / (c2o1 - c7o1 * omega + c5o1 * omega * omega);
real factorB = (c4o1 + c28o1 * omega - c14o1 * omega * omega) / (c6o1 - c21o1 * omega + c15o1 * omega * omega);
////////////////////////////////////////////////////////////////////////////////////
- //! - Compute cumulants (c's) from central moments according to Eq. (20)-(23) in
+ //! - Compute cumulants from central moments according to Eq. (20)-(23) in
//! <a href="https://doi.org/10.1016/j.jcp.2017.05.040"><b>[ M. Geier et al. (2017),
//! DOI:10.1016/j.jcp.2017.05.040 ]</b></a>
//!
@@ -318,27 +384,27 @@ __global__ void LB_Kernel_CumulantK17CompChimRedesigned(
// 5.
real c_122 =
m_122 - ((m_002 * m_120 + m_020 * m_102 + c4o1 * m_011 * m_111 + c2o1 * (m_101 * m_021 + m_110 * m_012)) +
- c1o3 * (m_120 + m_102)) *
- oneOverRho;
+ c1o3 * (m_120 + m_102)) *
+ oneOverRho;
real c_212 =
m_212 - ((m_002 * m_210 + m_200 * m_012 + c4o1 * m_101 * m_111 + c2o1 * (m_011 * m_201 + m_110 * m_102)) +
- c1o3 * (m_210 + m_012)) *
- oneOverRho;
+ c1o3 * (m_210 + m_012)) *
+ oneOverRho;
real c_221 =
m_221 - ((m_200 * m_021 + m_020 * m_201 + c4o1 * m_110 * m_111 + c2o1 * (m_101 * m_120 + m_011 * m_210)) +
- c1o3 * (m_021 + m_201)) *
- oneOverRho;
+ c1o3 * (m_021 + m_201)) *
+ oneOverRho;
////////////////////////////////////////////////////////////
// 6.
real c_222 = m_222 + ((-c4o1 * m_111 * m_111 - (m_200 * m_022 + m_020 * m_202 + m_002 * m_220) -
c4o1 * (m_011 * m_211 + m_101 * m_121 + m_110 * m_112) -
c2o1 * (m_120 * m_102 + m_210 * m_012 + m_201 * m_021)) *
oneOverRho +
- (c4o1 * (m_101 * m_101 * m_020 + m_011 * m_011 * m_200 + m_110 * m_110 * m_002) +
+ (c4o1 * (m_101 * m_101 * m_020 + m_011 * m_011 * m_200 + m_110 * m_110 * m_002) +
c2o1 * (m_200 * m_020 * m_002) + c16o1 * m_110 * m_101 * m_011) *
oneOverRho * oneOverRho -
c1o3 * (m_022 + m_202 + m_220) * oneOverRho - c1o9 * (m_200 + m_020 + m_002) * oneOverRho +
- (c2o1 * (m_101 * m_101 + m_011 * m_011 + m_110 * m_110) +
+ (c2o1 * (m_101 * m_101 + m_011 * m_011 + m_110 * m_110) +
(m_002 * m_020 + m_002 * m_200 + m_020 * m_200) + c1o3 * (m_002 + m_020 + m_200)) *
oneOverRho * oneOverRho * c2o3 +
c1o27 * ((drho * drho - drho) * oneOverRho * oneOverRho));
@@ -378,6 +444,22 @@ __global__ void LB_Kernel_CumulantK17CompChimRedesigned(
real dxux = c1o2 * (-omega) * (mxxMyy + mxxMzz) + c1o2 * OxxPyyPzz * (m_000 - mxxPyyPzz);
real dyuy = dxux + omega * c3o2 * mxxMyy;
real dzuz = dxux + omega * c3o2 * mxxMzz;
+
+ ////////////////////////////////////////////////////////////////////////////////////
+ switch (turbulenceModel)
+ {
+ case TurbulenceModel::None:
+ case TurbulenceModel::AMD: //AMD is computed in separate kernel
+ break;
+ case TurbulenceModel::Smagorinsky:
+ turbulentViscosity[k_000] = calcTurbulentViscositySmagorinsky(SGSconstant, dxux, dyuy, dzuz, Dxy, Dxz , Dyz);
+ break;
+ case TurbulenceModel::QR:
+ turbulentViscosity[k_000] = calcTurbulentViscosityQR(SGSconstant, dxux, dyuy, dzuz, Dxy, Dxz , Dyz);
+ break;
+ default:
+ break;
+ }
////////////////////////////////////////////////////////////
//! - Relaxation of second order cumulants with correction terms according to Eq. (33)-(35) in
//! <a href="https://doi.org/10.1016/j.jcp.2017.05.040"><b>[ M. Geier et al. (2017),
@@ -386,7 +468,6 @@ __global__ void LB_Kernel_CumulantK17CompChimRedesigned(
mxxPyyPzz += OxxPyyPzz * (m_000 - mxxPyyPzz) - c3o1 * (c1o1 - c1o2 * OxxPyyPzz) * (vx2 * dxux + vy2 * dyuy + vz2 * dzuz);
mxxMyy += omega * (-mxxMyy) - c3o1 * (c1o1 + c1o2 * (-omega)) * (vx2 * dxux - vy2 * dyuy);
mxxMzz += omega * (-mxxMzz) - c3o1 * (c1o1 + c1o2 * (-omega)) * (vx2 * dxux - vz2 * dzuz);
- //////////////////////////////////////////////////////////////////////////
////////////////////////////////////////////////////////////////////////////////////
////no correction
@@ -394,18 +475,18 @@ __global__ void LB_Kernel_CumulantK17CompChimRedesigned(
// mxxMyy += -(-omega) * (-mxxMyy);
// mxxMzz += -(-omega) * (-mxxMzz);
//////////////////////////////////////////////////////////////////////////
-
m_011 += omega * (-m_011);
m_101 += omega * (-m_101);
m_110 += omega * (-m_110);
+ ////////////////////////////////////////////////////////////////////////////////////
+ // relax
//////////////////////////////////////////////////////////////////////////
+ // incl. limiter
//! - Relaxation of third order cumulants including limiter according to Eq. (116)-(123)
//! <a href="https://doi.org/10.1016/j.jcp.2017.05.040"><b>[ M. Geier et al. (2017),
//! DOI:10.1016/j.jcp.2017.05.040 ]</b></a>
//!
- //////////////////////////////////////////////////////////////////////////
- // incl. limiter
real wadjust = Oxyz + (c1o1 - Oxyz) * abs(m_111) / (abs(m_111) + quadricLimitD);
m_111 += wadjust * (-m_111);
wadjust = OxyyPxzz + (c1o1 - OxyyPxzz) * abs(mxxyPyzz) / (abs(mxxyPyzz) + quadricLimitP);
@@ -459,6 +540,7 @@ __global__ void LB_Kernel_CumulantK17CompChimRedesigned(
c_121 = -O4 * (c1o1 / omega - c1o2) * Dxz * c1o3 * factorB + (c1o1 - O4) * (c_121);
c_211 = -O4 * (c1o1 / omega - c1o2) * Dyz * c1o3 * factorB + (c1o1 - O4) * (c_211);
+
//////////////////////////////////////////////////////////////////////////
// 5.
c_122 += O5 * (-c_122);
@@ -503,17 +585,17 @@ __global__ void LB_Kernel_CumulantK17CompChimRedesigned(
//////////////////////////////////////////////////////////////////////////
// 6.
m_222 = c_222 - ((-c4o1 * m_111 * m_111 - (m_200 * m_022 + m_020 * m_202 + m_002 * m_220) -
- c4o1 * (m_011 * m_211 + m_101 * m_121 + m_110 * m_112) -
- c2o1 * (m_120 * m_102 + m_210 * m_012 + m_201 * m_021)) *
- oneOverRho +
- (c4o1 * (m_101 * m_101 * m_020 + m_011 * m_011 * m_200 + m_110 * m_110 * m_002) +
- c2o1 * (m_200 * m_020 * m_002) + c16o1 * m_110 * m_101 * m_011) *
- oneOverRho * oneOverRho -
- c1o3 * (m_022 + m_202 + m_220) * oneOverRho - c1o9 * (m_200 + m_020 + m_002) * oneOverRho +
- (c2o1 * (m_101 * m_101 + m_011 * m_011 + m_110 * m_110) +
- (m_002 * m_020 + m_002 * m_200 + m_020 * m_200) + c1o3 * (m_002 + m_020 + m_200)) *
- oneOverRho * oneOverRho * c2o3 +
- c1o27 * ((drho * drho - drho) * oneOverRho * oneOverRho));
+ c4o1 * (m_011 * m_211 + m_101 * m_121 + m_110 * m_112) -
+ c2o1 * (m_120 * m_102 + m_210 * m_012 + m_201 * m_021)) *
+ oneOverRho +
+ (c4o1 * (m_101 * m_101 * m_020 + m_011 * m_011 * m_200 + m_110 * m_110 * m_002) +
+ c2o1 * (m_200 * m_020 * m_002) + c16o1 * m_110 * m_101 * m_011) *
+ oneOverRho * oneOverRho -
+ c1o3 * (m_022 + m_202 + m_220) * oneOverRho - c1o9 * (m_200 + m_020 + m_002) * oneOverRho +
+ (c2o1 * (m_101 * m_101 + m_011 * m_011 + m_110 * m_110) +
+ (m_002 * m_020 + m_002 * m_200 + m_020 * m_200) + c1o3 * (m_002 + m_020 + m_200)) *
+ oneOverRho * oneOverRho * c2o3 +
+ c1o27 * ((drho * drho - drho) * oneOverRho * oneOverRho));
////////////////////////////////////////////////////////////////////////////////////
//! - Add acceleration (body force) to first order cumulants according to Eq. (85)-(87) in
@@ -524,6 +606,15 @@ __global__ void LB_Kernel_CumulantK17CompChimRedesigned(
m_010 = -m_010;
m_001 = -m_001;
+ //Write to array here to distribute read/write
+ if(writeMacroscopicVariables)
+ {
+ rho[k_000] = drho;
+ vx[k_000] = vvx;
+ vy[k_000] = vvy;
+ vz[k_000] = vvz;
+ }
+
////////////////////////////////////////////////////////////////////////////////////
//! - Chimera transform from central moments to well conditioned distributions as defined in Appendix J in
//! <a href="https://doi.org/10.1016/j.camwa.2015.05.001"><b>[ M. Geier et al. (2015),
@@ -573,31 +664,63 @@ __global__ void LB_Kernel_CumulantK17CompChimRedesigned(
//! <a href="https://doi.org/10.3390/computation5020019"><b>[ M. Geier et al. (2017),
//! DOI:10.3390/computation5020019 ]</b></a>
//!
- (dist.f[DIR_P00])[k_000] = f_M00;
- (dist.f[DIR_M00])[k_M00] = f_P00;
- (dist.f[DIR_0P0])[k_000] = f_0M0;
- (dist.f[DIR_0M0])[k_0M0] = f_0P0;
- (dist.f[DIR_00P])[k_000] = f_00M;
- (dist.f[DIR_00M])[k_00M] = f_00P;
- (dist.f[DIR_PP0])[k_000] = f_MM0;
- (dist.f[DIR_MM0])[k_MM0] = f_PP0;
- (dist.f[DIR_PM0])[k_0M0] = f_MP0;
- (dist.f[DIR_MP0])[k_M00] = f_PM0;
- (dist.f[DIR_P0P])[k_000] = f_M0M;
- (dist.f[DIR_M0M])[k_M0M] = f_P0P;
- (dist.f[DIR_P0M])[k_00M] = f_M0P;
- (dist.f[DIR_M0P])[k_M00] = f_P0M;
- (dist.f[DIR_0PP])[k_000] = f_0MM;
- (dist.f[DIR_0MM])[k_0MM] = f_0PP;
- (dist.f[DIR_0PM])[k_00M] = f_0MP;
- (dist.f[DIR_0MP])[k_0M0] = f_0PM;
+ (dist.f[DIR_P00])[k_000] = f_M00;
+ (dist.f[DIR_M00])[k_M00] = f_P00;
+ (dist.f[DIR_0P0])[k_000] = f_0M0;
+ (dist.f[DIR_0M0])[k_0M0] = f_0P0;
+ (dist.f[DIR_00P])[k_000] = f_00M;
+ (dist.f[DIR_00M])[k_00M] = f_00P;
+ (dist.f[DIR_PP0])[k_000] = f_MM0;
+ (dist.f[DIR_MM0])[k_MM0] = f_PP0;
+ (dist.f[DIR_PM0])[k_0M0] = f_MP0;
+ (dist.f[DIR_MP0])[k_M00] = f_PM0;
+ (dist.f[DIR_P0P])[k_000] = f_M0M;
+ (dist.f[DIR_M0M])[k_M0M] = f_P0P;
+ (dist.f[DIR_P0M])[k_00M] = f_M0P;
+ (dist.f[DIR_M0P])[k_M00] = f_P0M;
+ (dist.f[DIR_0PP])[k_000] = f_0MM;
+ (dist.f[DIR_0MM])[k_0MM] = f_0PP;
+ (dist.f[DIR_0PM])[k_00M] = f_0MP;
+ (dist.f[DIR_0MP])[k_0M0] = f_0PM;
(dist.f[DIR_000])[k_000] = f_000;
- (dist.f[DIR_PPP])[k_000] = f_MMM;
- (dist.f[DIR_PMP])[k_0M0] = f_MPM;
- (dist.f[DIR_PPM])[k_00M] = f_MMP;
- (dist.f[DIR_PMM])[k_0MM] = f_MPP;
- (dist.f[DIR_MPP])[k_M00] = f_PMM;
- (dist.f[DIR_MMP])[k_MM0] = f_PPM;
- (dist.f[DIR_MPM])[k_M0M] = f_PMP;
- (dist.f[DIR_MMM])[k_MMM] = f_PPP;
-}
\ No newline at end of file
+ (dist.f[DIR_PPP])[k_000] = f_MMM;
+ (dist.f[DIR_PMP])[k_0M0] = f_MPM;
+ (dist.f[DIR_PPM])[k_00M] = f_MMP;
+ (dist.f[DIR_PMM])[k_0MM] = f_MPP;
+ (dist.f[DIR_MPP])[k_M00] = f_PMM;
+ (dist.f[DIR_MMP])[k_MM0] = f_PPM;
+ (dist.f[DIR_MPM])[k_M0M] = f_PMP;
+ (dist.f[DIR_MMM])[k_MMM] = f_PPP;
+}
+
+template __global__ void LB_Kernel_CumulantK17 < TurbulenceModel::AMD, true, true > ( real omega_in, uint* neighborX, uint* neighborY, uint* neighborZ, real* distributions, real* rho, real* vx, real* vy, real* vz, real* turbulentViscosity, real SGSconstant, unsigned long long numberOfLBnodes, int level, real* forces, real* bodyForceX, real* bodyForceY, real* bodyForceZ, real* quadricLimiters, bool isEvenTimestep, const uint *fluidNodeIndices, uint numberOfFluidNodes);
+
+template __global__ void LB_Kernel_CumulantK17 < TurbulenceModel::Smagorinsky, true, true > ( real omega_in, uint* neighborX, uint* neighborY, uint* neighborZ, real* distributions, real* rho, real* vx, real* vy, real* vz, real* turbulentViscosity, real SGSconstant, unsigned long long numberOfLBnodes, int level, real* forces, real* bodyForceX, real* bodyForceY, real* bodyForceZ, real* quadricLimiters, bool isEvenTimestep, const uint *fluidNodeIndices, uint numberOfFluidNodes);
+
+template __global__ void LB_Kernel_CumulantK17 < TurbulenceModel::QR, true, true > ( real omega_in, uint* neighborX, uint* neighborY, uint* neighborZ, real* distributions, real* rho, real* vx, real* vy, real* vz, real* turbulentViscosity, real SGSconstant, unsigned long long numberOfLBnodes, int level, real* forces, real* bodyForceX, real* bodyForceY, real* bodyForceZ, real* quadricLimiters, bool isEvenTimestep, const uint *fluidNodeIndices, uint numberOfFluidNodes);
+
+template __global__ void LB_Kernel_CumulantK17 < TurbulenceModel::None, true, true > ( real omega_in, uint* neighborX, uint* neighborY, uint* neighborZ, real* distributions, real* rho, real* vx, real* vy, real* vz, real* turbulentViscosity, real SGSconstant, unsigned long long numberOfLBnodes, int level, real* forces, real* bodyForceX, real* bodyForceY, real* bodyForceZ, real* quadricLimiters, bool isEvenTimestep, const uint *fluidNodeIndices, uint numberOfFluidNodes);
+
+template __global__ void LB_Kernel_CumulantK17 < TurbulenceModel::AMD, true, false > ( real omega_in, uint* neighborX, uint* neighborY, uint* neighborZ, real* distributions, real* rho, real* vx, real* vy, real* vz, real* turbulentViscosity, real SGSconstant, unsigned long long numberOfLBnodes, int level, real* forces, real* bodyForceX, real* bodyForceY, real* bodyForceZ, real* quadricLimiters, bool isEvenTimestep, const uint *fluidNodeIndices, uint numberOfFluidNodes);
+
+template __global__ void LB_Kernel_CumulantK17 < TurbulenceModel::Smagorinsky, true, false > ( real omega_in, uint* neighborX, uint* neighborY, uint* neighborZ, real* distributions, real* rho, real* vx, real* vy, real* vz, real* turbulentViscosity, real SGSconstant, unsigned long long numberOfLBnodes, int level, real* forces, real* bodyForceX, real* bodyForceY, real* bodyForceZ, real* quadricLimiters, bool isEvenTimestep, const uint *fluidNodeIndices, uint numberOfFluidNodes);
+
+template __global__ void LB_Kernel_CumulantK17 < TurbulenceModel::QR, true, false > ( real omega_in, uint* neighborX, uint* neighborY, uint* neighborZ, real* distributions, real* rho, real* vx, real* vy, real* vz, real* turbulentViscosity, real SGSconstant, unsigned long long numberOfLBnodes, int level, real* forces, real* bodyForceX, real* bodyForceY, real* bodyForceZ, real* quadricLimiters, bool isEvenTimestep, const uint *fluidNodeIndices, uint numberOfFluidNodes);
+
+template __global__ void LB_Kernel_CumulantK17 < TurbulenceModel::None, true, false > ( real omega_in, uint* neighborX, uint* neighborY, uint* neighborZ, real* distributions, real* rho, real* vx, real* vy, real* vz, real* turbulentViscosity, real SGSconstant, unsigned long long numberOfLBnodes, int level, real* forces, real* bodyForceX, real* bodyForceY, real* bodyForceZ, real* quadricLimiters, bool isEvenTimestep, const uint *fluidNodeIndices, uint numberOfFluidNodes);
+
+template __global__ void LB_Kernel_CumulantK17 < TurbulenceModel::AMD, false, true > ( real omega_in, uint* neighborX, uint* neighborY, uint* neighborZ, real* distributions, real* rho, real* vx, real* vy, real* vz, real* turbulentViscosity, real SGSconstant, unsigned long long numberOfLBnodes, int level, real* forces, real* bodyForceX, real* bodyForceY, real* bodyForceZ, real* quadricLimiters, bool isEvenTimestep, const uint *fluidNodeIndices, uint numberOfFluidNodes);
+
+template __global__ void LB_Kernel_CumulantK17 < TurbulenceModel::Smagorinsky, false, true > ( real omega_in, uint* neighborX, uint* neighborY, uint* neighborZ, real* distributions, real* rho, real* vx, real* vy, real* vz, real* turbulentViscosity, real SGSconstant, unsigned long long numberOfLBnodes, int level, real* forces, real* bodyForceX, real* bodyForceY, real* bodyForceZ, real* quadricLimiters, bool isEvenTimestep, const uint *fluidNodeIndices, uint numberOfFluidNodes);
+
+template __global__ void LB_Kernel_CumulantK17 < TurbulenceModel::QR, false, true > ( real omega_in, uint* neighborX, uint* neighborY, uint* neighborZ, real* distributions, real* rho, real* vx, real* vy, real* vz, real* turbulentViscosity, real SGSconstant, unsigned long long numberOfLBnodes, int level, real* forces, real* bodyForceX, real* bodyForceY, real* bodyForceZ, real* quadricLimiters, bool isEvenTimestep, const uint *fluidNodeIndices, uint numberOfFluidNodes);
+
+template __global__ void LB_Kernel_CumulantK17 < TurbulenceModel::None, false, true > ( real omega_in, uint* neighborX, uint* neighborY, uint* neighborZ, real* distributions, real* rho, real* vx, real* vy, real* vz, real* turbulentViscosity, real SGSconstant, unsigned long long numberOfLBnodes, int level, real* forces, real* bodyForceX, real* bodyForceY, real* bodyForceZ, real* quadricLimiters, bool isEvenTimestep, const uint *fluidNodeIndices, uint numberOfFluidNodes);
+
+template __global__ void LB_Kernel_CumulantK17 < TurbulenceModel::AMD, false, false > ( real omega_in, uint* neighborX, uint* neighborY, uint* neighborZ, real* distributions, real* rho, real* vx, real* vy, real* vz, real* turbulentViscosity, real SGSconstant, unsigned long long numberOfLBnodes, int level, real* forces, real* bodyForceX, real* bodyForceY, real* bodyForceZ, real* quadricLimiters, bool isEvenTimestep, const uint *fluidNodeIndices, uint numberOfFluidNodes);
+
+template __global__ void LB_Kernel_CumulantK17 < TurbulenceModel::Smagorinsky, false, false > ( real omega_in, uint* neighborX, uint* neighborY, uint* neighborZ, real* distributions, real* rho, real* vx, real* vy, real* vz, real* turbulentViscosity, real SGSconstant, unsigned long long numberOfLBnodes, int level, real* forces, real* bodyForceX, real* bodyForceY, real* bodyForceZ, real* quadricLimiters, bool isEvenTimestep, const uint *fluidNodeIndices, uint numberOfFluidNodes);
+
+template __global__ void LB_Kernel_CumulantK17 < TurbulenceModel::QR, false, false > ( real omega_in, uint* neighborX, uint* neighborY, uint* neighborZ, real* distributions, real* rho, real* vx, real* vy, real* vz, real* turbulentViscosity, real SGSconstant, unsigned long long numberOfLBnodes, int level, real* forces, real* bodyForceX, real* bodyForceY, real* bodyForceZ, real* quadricLimiters, bool isEvenTimestep, const uint *fluidNodeIndices, uint numberOfFluidNodes);
+
+template __global__ void LB_Kernel_CumulantK17 < TurbulenceModel::None, false, false > ( real omega_in, uint* neighborX, uint* neighborY, uint* neighborZ, real* distributions, real* rho, real* vx, real* vy, real* vz, real* turbulentViscosity, real SGSconstant, unsigned long long numberOfLBnodes, int level, real* forces, real* bodyForceX, real* bodyForceY, real* bodyForceZ, real* quadricLimiters, bool isEvenTimestep, const uint *fluidNodeIndices, uint numberOfFluidNodes);
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK17/CumulantK17_Device.cuh b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK17/CumulantK17_Device.cuh
new file mode 100644
index 0000000000000000000000000000000000000000..da576618d1b08b55629c3c65fc115ceb822c8f7e
--- /dev/null
+++ b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK17/CumulantK17_Device.cuh
@@ -0,0 +1,29 @@
+#ifndef LB_Kernel_CUMULANT_K17_H
+#define LB_Kernel_CUMULANT_K17_H
+
+#include <DataTypes.h>
+#include <curand.h>
+
+template< TurbulenceModel turbulenceModel, bool writeMacroscopicVariables, bool applyBodyForce > __global__ void LB_Kernel_CumulantK17(
+ real omega_in,
+ uint* neighborX,
+ uint* neighborY,
+ uint* neighborZ,
+ real* distributions,
+ real* rho,
+ real* vx,
+ real* vy,
+ real* vz,
+ real* turbulentViscosity,
+ real SGSconstant,
+ unsigned long long numberOfLBnodes,
+ int level,
+ real* forces,
+ real* bodyForceX,
+ real* bodyForceY,
+ real* bodyForceZ,
+ real* quadricLimiters,
+ bool isEvenTimestep,
+ const uint *fluidNodeIndices,
+ uint numberOfFluidNodes);
+#endif
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK17Bulk/CumulantK17BulkComp.cu b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK17Bulk/CumulantK17BulkComp.cu
index 72d13282fc604dddcfa84682425a7a1829855ea0..b9e25494490507bde5a6aa7d6dd588ac1a1f6c87 100644
--- a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK17Bulk/CumulantK17BulkComp.cu
+++ b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK17Bulk/CumulantK17BulkComp.cu
@@ -17,17 +17,18 @@ void CumulantK17BulkComp::run()
dim3 grid(Grid, 1, 1);
dim3 threads(numberOfThreads, 1, 1);
- LB_Kernel_CumulantK17BulkComp << < grid, threads >> >( para->getParD(level)->omega,
- para->getParD(level)->typeOfGridNode,
- para->getParD(level)->neighborX,
- para->getParD(level)->neighborY,
- para->getParD(level)->neighborZ,
- para->getParD(level)->distributions.f[0],
- para->getParD(level)->numberOfNodes,
- level,
- para->getForcesDev(),
- para->getQuadricLimitersDev(),
- para->getParD(level)->isEvenTimestep);
+ LB_Kernel_CumulantK17BulkComp << < grid, threads >> >(
+ para->getParD(level)->omega,
+ para->getParD(level)->typeOfGridNode,
+ para->getParD(level)->neighborX,
+ para->getParD(level)->neighborY,
+ para->getParD(level)->neighborZ,
+ para->getParD(level)->distributions.f[0],
+ para->getParD(level)->numberOfNodes,
+ level,
+ para->getForcesDev(),
+ para->getQuadricLimitersDev(),
+ para->getParD(level)->isEvenTimestep);
getLastCudaError("LB_Kernel_CumulantK17BulkComp execution failed");
}
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK17Bulk/CumulantK17BulkComp_Device.cu b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK17Bulk/CumulantK17BulkComp_Device.cu
index cec04116ae4b411b1b3816ff4a8cab606c92491e..b33a3c251b5fb0cde8b1da0fcce097f955353d69 100644
--- a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK17Bulk/CumulantK17BulkComp_Device.cu
+++ b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK17Bulk/CumulantK17BulkComp_Device.cu
@@ -40,63 +40,63 @@ __global__ void LB_Kernel_CumulantK17BulkComp(real omega,
Distributions27 D;
if (EvenOrOdd == true)
{
- D.f[DIR_P00] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_M00] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0P0] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0M0] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00P] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00M] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_PP0] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_MM0] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_PM0] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_MP0] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_P0P] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_M0M] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_P0M] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_M0P] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0PP] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0MM] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0PM] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0MP] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_PPP] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_MMP] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_PMP] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_MPP] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_PPM] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_MMM] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_PMM] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_MPM] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DDStart[DIR_P00 * size_Mat];
+ D.f[DIR_M00] = &DDStart[DIR_M00 * size_Mat];
+ D.f[DIR_0P0] = &DDStart[DIR_0P0 * size_Mat];
+ D.f[DIR_0M0] = &DDStart[DIR_0M0 * size_Mat];
+ D.f[DIR_00P] = &DDStart[DIR_00P * size_Mat];
+ D.f[DIR_00M] = &DDStart[DIR_00M * size_Mat];
+ D.f[DIR_PP0] = &DDStart[DIR_PP0 * size_Mat];
+ D.f[DIR_MM0] = &DDStart[DIR_MM0 * size_Mat];
+ D.f[DIR_PM0] = &DDStart[DIR_PM0 * size_Mat];
+ D.f[DIR_MP0] = &DDStart[DIR_MP0 * size_Mat];
+ D.f[DIR_P0P] = &DDStart[DIR_P0P * size_Mat];
+ D.f[DIR_M0M] = &DDStart[DIR_M0M * size_Mat];
+ D.f[DIR_P0M] = &DDStart[DIR_P0M * size_Mat];
+ D.f[DIR_M0P] = &DDStart[DIR_M0P * size_Mat];
+ D.f[DIR_0PP] = &DDStart[DIR_0PP * size_Mat];
+ D.f[DIR_0MM] = &DDStart[DIR_0MM * size_Mat];
+ D.f[DIR_0PM] = &DDStart[DIR_0PM * size_Mat];
+ D.f[DIR_0MP] = &DDStart[DIR_0MP * size_Mat];
+ D.f[DIR_000] = &DDStart[DIR_000 * size_Mat];
+ D.f[DIR_PPP] = &DDStart[DIR_PPP * size_Mat];
+ D.f[DIR_MMP] = &DDStart[DIR_MMP * size_Mat];
+ D.f[DIR_PMP] = &DDStart[DIR_PMP * size_Mat];
+ D.f[DIR_MPP] = &DDStart[DIR_MPP * size_Mat];
+ D.f[DIR_PPM] = &DDStart[DIR_PPM * size_Mat];
+ D.f[DIR_MMM] = &DDStart[DIR_MMM * size_Mat];
+ D.f[DIR_PMM] = &DDStart[DIR_PMM * size_Mat];
+ D.f[DIR_MPM] = &DDStart[DIR_MPM * size_Mat];
}
else
{
- D.f[DIR_M00] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_P00] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0M0] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0P0] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00M] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00P] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_MM0] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_PP0] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_MP0] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_PM0] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_M0M] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_P0P] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_M0P] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_P0M] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0MM] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0PP] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0MP] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0PM] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_MMM] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_PPM] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_MPM] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_PMM] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_MMP] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_PPP] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_MPP] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_PMP] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_M00] = &DDStart[DIR_P00 * size_Mat];
+ D.f[DIR_P00] = &DDStart[DIR_M00 * size_Mat];
+ D.f[DIR_0M0] = &DDStart[DIR_0P0 * size_Mat];
+ D.f[DIR_0P0] = &DDStart[DIR_0M0 * size_Mat];
+ D.f[DIR_00M] = &DDStart[DIR_00P * size_Mat];
+ D.f[DIR_00P] = &DDStart[DIR_00M * size_Mat];
+ D.f[DIR_MM0] = &DDStart[DIR_PP0 * size_Mat];
+ D.f[DIR_PP0] = &DDStart[DIR_MM0 * size_Mat];
+ D.f[DIR_MP0] = &DDStart[DIR_PM0 * size_Mat];
+ D.f[DIR_PM0] = &DDStart[DIR_MP0 * size_Mat];
+ D.f[DIR_M0M] = &DDStart[DIR_P0P * size_Mat];
+ D.f[DIR_P0P] = &DDStart[DIR_M0M * size_Mat];
+ D.f[DIR_M0P] = &DDStart[DIR_P0M * size_Mat];
+ D.f[DIR_P0M] = &DDStart[DIR_M0P * size_Mat];
+ D.f[DIR_0MM] = &DDStart[DIR_0PP * size_Mat];
+ D.f[DIR_0PP] = &DDStart[DIR_0MM * size_Mat];
+ D.f[DIR_0MP] = &DDStart[DIR_0PM * size_Mat];
+ D.f[DIR_0PM] = &DDStart[DIR_0MP * size_Mat];
+ D.f[DIR_000] = &DDStart[DIR_000 * size_Mat];
+ D.f[DIR_MMM] = &DDStart[DIR_PPP * size_Mat];
+ D.f[DIR_PPM] = &DDStart[DIR_MMP * size_Mat];
+ D.f[DIR_MPM] = &DDStart[DIR_PMP * size_Mat];
+ D.f[DIR_PMM] = &DDStart[DIR_MPP * size_Mat];
+ D.f[DIR_MMP] = &DDStart[DIR_PPM * size_Mat];
+ D.f[DIR_PPP] = &DDStart[DIR_MMM * size_Mat];
+ D.f[DIR_MPP] = &DDStart[DIR_PMM * size_Mat];
+ D.f[DIR_PMP] = &DDStart[DIR_MPM * size_Mat];
}
////////////////////////////////////////////////////////////////////////////////
@@ -129,33 +129,33 @@ __global__ void LB_Kernel_CumulantK17BulkComp(real omega,
//unsigned int ktne = k;
unsigned int kbsw = neighborZ[ksw];
//////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
- real mfcbb = (D.f[DIR_P00])[k];//[ke ];// + c2over27 ;(D.f[DIR_P00 ])[k ];//ke
- real mfabb = (D.f[DIR_M00])[kw];//[kw ];// + c2over27 ;(D.f[DIR_M00 ])[kw ];
- real mfbcb = (D.f[DIR_0P0])[k];//[kn ];// + c2over27 ;(D.f[DIR_0P0 ])[k ];//kn
- real mfbab = (D.f[DIR_0M0])[ks];//[ks ];// + c2over27 ;(D.f[DIR_0M0 ])[ks ];
- real mfbbc = (D.f[DIR_00P])[k];//[kt ];// + c2over27 ;(D.f[DIR_00P ])[k ];//kt
- real mfbba = (D.f[DIR_00M])[kb];//[kb ];// + c2over27 ;(D.f[DIR_00M ])[kb ];
- real mfccb = (D.f[DIR_PP0])[k];//[kne ];// + c1over54 ;(D.f[DIR_PP0 ])[k ];//kne
- real mfaab = (D.f[DIR_MM0])[ksw];//[ksw ];// + c1over54 ;(D.f[DIR_MM0 ])[ksw];
- real mfcab = (D.f[DIR_PM0])[ks];//[kse ];// + c1over54 ;(D.f[DIR_PM0 ])[ks ];//kse
- real mfacb = (D.f[DIR_MP0])[kw];//[knw ];// + c1over54 ;(D.f[DIR_MP0 ])[kw ];//knw
- real mfcbc = (D.f[DIR_P0P])[k];//[kte ];// + c1over54 ;(D.f[DIR_P0P ])[k ];//kte
- real mfaba = (D.f[DIR_M0M])[kbw];//[kbw ];// + c1over54 ;(D.f[DIR_M0M ])[kbw];
- real mfcba = (D.f[DIR_P0M])[kb];//[kbe ];// + c1over54 ;(D.f[DIR_P0M ])[kb ];//kbe
- real mfabc = (D.f[DIR_M0P])[kw];//[ktw ];// + c1over54 ;(D.f[DIR_M0P ])[kw ];//ktw
- real mfbcc = (D.f[DIR_0PP])[k];//[ktn ];// + c1over54 ;(D.f[DIR_0PP ])[k ];//ktn
- real mfbaa = (D.f[DIR_0MM])[kbs];//[kbs ];// + c1over54 ;(D.f[DIR_0MM ])[kbs];
- real mfbca = (D.f[DIR_0PM])[kb];//[kbn ];// + c1over54 ;(D.f[DIR_0PM ])[kb ];//kbn
- real mfbac = (D.f[DIR_0MP])[ks];//[kts ];// + c1over54 ;(D.f[DIR_0MP ])[ks ];//kts
+ real mfcbb = (D.f[DIR_P00])[k];//[ke ];// + c2over27 ;(D.f[DIR_P00])[k ];//ke
+ real mfabb = (D.f[DIR_M00])[kw];//[kw ];// + c2over27 ;(D.f[DIR_M00])[kw ];
+ real mfbcb = (D.f[DIR_0P0])[k];//[kn ];// + c2over27 ;(D.f[DIR_0P0])[k ];//kn
+ real mfbab = (D.f[DIR_0M0])[ks];//[ks ];// + c2over27 ;(D.f[DIR_0M0])[ks ];
+ real mfbbc = (D.f[DIR_00P])[k];//[kt ];// + c2over27 ;(D.f[DIR_00P])[k ];//kt
+ real mfbba = (D.f[DIR_00M])[kb];//[kb ];// + c2over27 ;(D.f[DIR_00M])[kb ];
+ real mfccb = (D.f[DIR_PP0])[k];//[kne ];// + c1over54 ;(D.f[DIR_PP0])[k ];//kne
+ real mfaab = (D.f[DIR_MM0])[ksw];//[ksw ];// + c1over54 ;(D.f[DIR_MM0])[ksw];
+ real mfcab = (D.f[DIR_PM0])[ks];//[kse ];// + c1over54 ;(D.f[DIR_PM0])[ks ];//kse
+ real mfacb = (D.f[DIR_MP0])[kw];//[knw ];// + c1over54 ;(D.f[DIR_MP0])[kw ];//knw
+ real mfcbc = (D.f[DIR_P0P])[k];//[kte ];// + c1over54 ;(D.f[DIR_P0P])[k ];//kte
+ real mfaba = (D.f[DIR_M0M])[kbw];//[kbw ];// + c1over54 ;(D.f[DIR_M0M])[kbw];
+ real mfcba = (D.f[DIR_P0M])[kb];//[kbe ];// + c1over54 ;(D.f[DIR_P0M])[kb ];//kbe
+ real mfabc = (D.f[DIR_M0P])[kw];//[ktw ];// + c1over54 ;(D.f[DIR_M0P])[kw ];//ktw
+ real mfbcc = (D.f[DIR_0PP])[k];//[ktn ];// + c1over54 ;(D.f[DIR_0PP])[k ];//ktn
+ real mfbaa = (D.f[DIR_0MM])[kbs];//[kbs ];// + c1over54 ;(D.f[DIR_0MM])[kbs];
+ real mfbca = (D.f[DIR_0PM])[kb];//[kbn ];// + c1over54 ;(D.f[DIR_0PM])[kb ];//kbn
+ real mfbac = (D.f[DIR_0MP])[ks];//[kts ];// + c1over54 ;(D.f[DIR_0MP])[ks ];//kts
real mfbbb = (D.f[DIR_000])[k];//[kzero];// + c8over27 ;(D.f[DIR_000])[k ];//kzero
- real mfccc = (D.f[DIR_PPP])[k];//[ktne ];// + c1over216;(D.f[DIR_PPP ])[k ];//ktne
- real mfaac = (D.f[DIR_MMP])[ksw];//[ktsw ];// + c1over216;(D.f[DIR_MMP ])[ksw];//ktsw
- real mfcac = (D.f[DIR_PMP])[ks];//[ktse ];// + c1over216;(D.f[DIR_PMP ])[ks ];//ktse
- real mfacc = (D.f[DIR_MPP])[kw];//[ktnw ];// + c1over216;(D.f[DIR_MPP ])[kw ];//ktnw
- real mfcca = (D.f[DIR_PPM])[kb];//[kbne ];// + c1over216;(D.f[DIR_PPM ])[kb ];//kbne
- real mfaaa = (D.f[DIR_MMM])[kbsw];//[kbsw ];// + c1over216;(D.f[DIR_MMM ])[kbsw];
- real mfcaa = (D.f[DIR_PMM])[kbs];//[kbse ];// + c1over216;(D.f[DIR_PMM ])[kbs];//kbse
- real mfaca = (D.f[DIR_MPM])[kbw];//[kbnw ];// + c1over216;(D.f[DIR_MPM ])[kbw];//kbnw
+ real mfccc = (D.f[DIR_PPP])[k];//[ktne ];// + c1over216;(D.f[DIR_PPP])[k ];//ktne
+ real mfaac = (D.f[DIR_MMP])[ksw];//[ktsw ];// + c1over216;(D.f[DIR_MMP])[ksw];//ktsw
+ real mfcac = (D.f[DIR_PMP])[ks];//[ktse ];// + c1over216;(D.f[DIR_PMP])[ks ];//ktse
+ real mfacc = (D.f[DIR_MPP])[kw];//[ktnw ];// + c1over216;(D.f[DIR_MPP])[kw ];//ktnw
+ real mfcca = (D.f[DIR_PPM])[kb];//[kbne ];// + c1over216;(D.f[DIR_PPM])[kb ];//kbne
+ real mfaaa = (D.f[DIR_MMM])[kbsw];//[kbsw ];// + c1over216;(D.f[DIR_MMM])[kbsw];
+ real mfcaa = (D.f[DIR_PMM])[kbs];//[kbse ];// + c1over216;(D.f[DIR_PMM])[kbs];//kbse
+ real mfaca = (D.f[DIR_MPM])[kbw];//[kbnw ];// + c1over216;(D.f[DIR_MPM])[kbw];//kbnw
////////////////////////////////////////////////////////////////////////////////////
real drho = ((((mfccc + mfaaa) + (mfaca + mfcac)) + ((mfacc + mfcaa) + (mfaac + mfcca))) +
(((mfbac + mfbca) + (mfbaa + mfbcc)) + ((mfabc + mfcba) + (mfaba + mfcbc)) + ((mfacb + mfcab) + (mfaab + mfccb))) +
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK17Unified/CumulantK17Unified.cu b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK17Unified/CumulantK17Unified.cu
index 6ef6b40d3b7079579f54ca68734deb274d0c1c3a..295804887f9c451120d463c7fcdd968bd2f24d12 100644
--- a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK17Unified/CumulantK17Unified.cu
+++ b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK17Unified/CumulantK17Unified.cu
@@ -31,15 +31,16 @@ CumulantK17Unified::CumulantK17Unified(std::shared_ptr<Parameter> para, int leve
void CumulantK17Unified::run()
{
- GPUKernelParameter kernelParameter{ para->getParD(level)->omega,
- para->getParD(level)->typeOfGridNode,
- para->getParD(level)->neighborX,
- para->getParD(level)->neighborY,
- para->getParD(level)->neighborZ,
- para->getParD(level)->distributions.f[0],
- (int)para->getParD(level)->numberOfNodes,
- para->getParD(level)->forcing,
- para->getParD(level)->isEvenTimestep };
+ GPUKernelParameter kernelParameter{
+ para->getParD(level)->omega,
+ para->getParD(level)->typeOfGridNode,
+ para->getParD(level)->neighborX,
+ para->getParD(level)->neighborY,
+ para->getParD(level)->neighborZ,
+ para->getParD(level)->distributions.f[0],
+ (int)para->getParD(level)->numberOfNodes,
+ para->getParD(level)->forcing,
+ para->getParD(level)->isEvenTimestep };
auto lambda = [] __device__(lbm::KernelParameter parameter) {
return lbm::cumulantChimera(parameter, lbm::setRelaxationRatesK17);
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK17chim/CumulantK17CompChim_Device.cu b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK17chim/CumulantK17CompChim_Device.cu
index 3eea267e55fee45111fb11cf1258559e2c3c63f2..a0db78d27b00372feab8490111183481abbec8b9 100644
--- a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK17chim/CumulantK17CompChim_Device.cu
+++ b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK17chim/CumulantK17CompChim_Device.cu
@@ -33,11 +33,12 @@
/* Device code */
#include "LBM/LB.h"
#include "lbm/constants/D3Q27.h"
-#include <lbm/constants/NumericConstants.h>
+#include "lbm/constants/NumericConstants.h"
+#include "LBM/GPUHelperFunctions/ChimeraTransformation.h"
using namespace vf::lbm::constant;
using namespace vf::lbm::dir;
-#include "Kernel/Utilities/ChimeraTransformation.h"
+using namespace vf::gpu;
////////////////////////////////////////////////////////////////////////////////
__global__ void LB_Kernel_CumulantK17CompChim(
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK17chimRedesigned/CumulantK17CompChimRedesigned.cu b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK17chimRedesigned/CumulantK17CompChimRedesigned.cu
deleted file mode 100644
index 8c06b7117c8b1ef62b932a76bf5de0be2ae99b1c..0000000000000000000000000000000000000000
--- a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK17chimRedesigned/CumulantK17CompChimRedesigned.cu
+++ /dev/null
@@ -1,61 +0,0 @@
-#include "CumulantK17CompChimRedesigned.h"
-
-#include "Parameter/Parameter.h"
-#include "Parameter/CudaStreamManager.h"
-#include "CumulantK17CompChimRedesigned_Device.cuh"
-
-#include <cuda.h>
-
-std::shared_ptr<CumulantK17CompChimRedesigned> CumulantK17CompChimRedesigned::getNewInstance(std::shared_ptr<Parameter> para,
- int level)
-{
- return std::shared_ptr<CumulantK17CompChimRedesigned>(new CumulantK17CompChimRedesigned(para, level));
-}
-
-void CumulantK17CompChimRedesigned::run()
-{
- LB_Kernel_CumulantK17CompChimRedesigned <<< cudaGrid.grid, cudaGrid.threads >>>(
- para->getParD(level)->omega,
- para->getParD(level)->neighborX,
- para->getParD(level)->neighborY,
- para->getParD(level)->neighborZ,
- para->getParD(level)->distributions.f[0],
- para->getParD(level)->numberOfNodes,
- level,
- para->getForcesDev(),
- para->getQuadricLimitersDev(),
- para->getParD(level)->isEvenTimestep,
- para->getParD(level)->fluidNodeIndices,
- para->getParD(level)->numberOfFluidNodes);
- getLastCudaError("LB_Kernel_CumulantK17CompChim execution failed");
-}
-
-void CumulantK17CompChimRedesigned::runOnIndices(const unsigned int *indices, unsigned int size_indices, int streamIndex)
-{
- cudaStream_t stream = (streamIndex == -1) ? CU_STREAM_LEGACY : para->getStreamManager()->getStream(streamIndex);
-
- LB_Kernel_CumulantK17CompChimRedesigned<<< cudaGrid.grid, cudaGrid.threads, 0, stream>>>(
- para->getParD(level)->omega,
- para->getParD(level)->neighborX,
- para->getParD(level)->neighborY,
- para->getParD(level)->neighborZ,
- para->getParD(level)->distributions.f[0],
- para->getParD(level)->numberOfNodes,
- level,
- para->getForcesDev(),
- para->getQuadricLimitersDev(),
- para->getParD(level)->isEvenTimestep,
- indices,
- size_indices);
- getLastCudaError("LB_Kernel_CumulantK17CompChim execution failed");
-
-}
-
-CumulantK17CompChimRedesigned::CumulantK17CompChimRedesigned(std::shared_ptr<Parameter> para, int level): KernelImp(para, level)
-{
- myPreProcessorTypes.push_back(InitCompSP27);
- myKernelGroup = BasicKernel;
- this->cudaGrid = vf::cuda::CudaGrid(para->getParD(level)->numberofthreads, para->getParD(level)->numberOfNodes);
- this->kernelUsesFluidNodeIndices = true;
-}
-
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK17chimRedesigned/CumulantK17CompChimRedesigned.h b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK17chimRedesigned/CumulantK17CompChimRedesigned.h
deleted file mode 100644
index 4658075de330665fdba88a5ec8149a9b476d5ac7..0000000000000000000000000000000000000000
--- a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK17chimRedesigned/CumulantK17CompChimRedesigned.h
+++ /dev/null
@@ -1,18 +0,0 @@
-#ifndef CUMULANT_K17_COMP_CHIM_REDESIGN_H
-#define CUMULANT_K17_COMP_CHIM_REDESIGN_H
-
-#include "Kernel/KernelImp.h"
-
-class CumulantK17CompChimRedesigned : public KernelImp
-{
-public:
- static std::shared_ptr<CumulantK17CompChimRedesigned> getNewInstance(std::shared_ptr<Parameter> para, int level);
- void run() override;
- void runOnIndices(const unsigned int *indices, unsigned int size_indices, int stream = -1) override;
-
-private:
- CumulantK17CompChimRedesigned();
- CumulantK17CompChimRedesigned(std::shared_ptr<Parameter> para, int level);
-};
-
-#endif
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK17chimRedesigned/CumulantK17CompChimRedesigned_Device.cuh b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK17chimRedesigned/CumulantK17CompChimRedesigned_Device.cuh
deleted file mode 100644
index 00628efc76447a09504d2fd32a26a63a4d611c66..0000000000000000000000000000000000000000
--- a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK17chimRedesigned/CumulantK17CompChimRedesigned_Device.cuh
+++ /dev/null
@@ -1,20 +0,0 @@
-#ifndef LB_Kernel_CUMULANT_K17_COMP_CHIM_REDESIGN_H
-#define LB_Kernel_CUMULANT_K17_COMP_CHIM_REDESIGN_H
-
-#include <DataTypes.h>
-#include <curand.h>
-
-__global__ void LB_Kernel_CumulantK17CompChimRedesigned(
- real omega,
- uint* neighborX,
- uint* neighborY,
- uint* neighborZ,
- real* distributions,
- unsigned long numberOfLBnodes,
- int level,
- real* forces,
- real* quadricLimiters,
- bool isEvenTimestep,
- const uint* fluidNodeIndices,
- uint numberOfFluidNodes);
-#endif
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK17chimStream/CumulantK17CompChimStream.cu b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK17chimStream/CumulantK17CompChimStream.cu
deleted file mode 100644
index 6fae9f6d4845019afd363790eea0ee17c69a060f..0000000000000000000000000000000000000000
--- a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK17chimStream/CumulantK17CompChimStream.cu
+++ /dev/null
@@ -1,61 +0,0 @@
-#include "CumulantK17CompChimStream.h"
-
-#include "Parameter/Parameter.h"
-#include "Parameter/CudaStreamManager.h"
-#include "CumulantK17CompChimStream_Device.cuh"
-
-#include <cuda.h>
-
-std::shared_ptr<CumulantK17CompChimStream> CumulantK17CompChimStream::getNewInstance(std::shared_ptr<Parameter> para,
- int level)
-{
- return std::shared_ptr<CumulantK17CompChimStream>(new CumulantK17CompChimStream(para, level));
-}
-
-void CumulantK17CompChimStream::run()
-{
- LB_Kernel_CumulantK17CompChimStream <<< cudaGrid.grid, cudaGrid.threads >>>(
- para->getParD(level)->omega,
- para->getParD(level)->neighborX,
- para->getParD(level)->neighborY,
- para->getParD(level)->neighborZ,
- para->getParD(level)->distributions.f[0],
- para->getParD(level)->numberOfNodes,
- level,
- para->getForcesDev(),
- para->getQuadricLimitersDev(),
- para->getParD(level)->isEvenTimestep,
- para->getParD(level)->fluidNodeIndices,
- para->getParD(level)->numberOfFluidNodes);
- getLastCudaError("LB_Kernel_CumulantK17CompChim execution failed");
-}
-
-void CumulantK17CompChimStream::runOnIndices(const unsigned int *indices, unsigned int size_indices, int streamIndex)
-{
- cudaStream_t stream = (streamIndex == -1) ? CU_STREAM_LEGACY : para->getStreamManager()->getStream(streamIndex);
-
- LB_Kernel_CumulantK17CompChimStream<<< cudaGrid.grid, cudaGrid.threads, 0, stream>>>(
- para->getParD(level)->omega,
- para->getParD(level)->neighborX,
- para->getParD(level)->neighborY,
- para->getParD(level)->neighborZ,
- para->getParD(level)->distributions.f[0],
- para->getParD(level)->numberOfNodes,
- level,
- para->getForcesDev(),
- para->getQuadricLimitersDev(),
- para->getParD(level)->isEvenTimestep,
- indices,
- size_indices);
- getLastCudaError("LB_Kernel_CumulantK17CompChim execution failed");
-
-}
-
-CumulantK17CompChimStream::CumulantK17CompChimStream(std::shared_ptr<Parameter> para, int level): KernelImp(para, level)
-{
- myPreProcessorTypes.push_back(InitCompSP27);
- myKernelGroup = BasicKernel;
- this->cudaGrid = vf::cuda::CudaGrid(para->getParD(level)->numberofthreads, para->getParD(level)->numberOfNodes);
- this->kernelUsesFluidNodeIndices = true;
-}
-
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK17chimStream/CumulantK17CompChimStream.h b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK17chimStream/CumulantK17CompChimStream.h
deleted file mode 100644
index 325826e04c893b7c56b7f00bb2503a4eb1fda441..0000000000000000000000000000000000000000
--- a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK17chimStream/CumulantK17CompChimStream.h
+++ /dev/null
@@ -1,18 +0,0 @@
-#ifndef CUMULANT_K17_COMP_CHIM_SPARSE_H
-#define CUMULANT_K17_COMP_CHIM_SPARSE_H
-
-#include "Kernel/KernelImp.h"
-
-class CumulantK17CompChimStream : public KernelImp
-{
-public:
- static std::shared_ptr<CumulantK17CompChimStream> getNewInstance(std::shared_ptr<Parameter> para, int level);
- void run() override;
- void runOnIndices(const unsigned int *indices, unsigned int size_indices, int stream = -1) override;
-
-private:
- CumulantK17CompChimStream();
- CumulantK17CompChimStream(std::shared_ptr<Parameter> para, int level);
-};
-
-#endif
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK17chimStream/CumulantK17CompChimStreamDevice.cu b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK17chimStream/CumulantK17CompChimStreamDevice.cu
deleted file mode 100644
index 830fcc6c328f2ecd0f626539040868696065065f..0000000000000000000000000000000000000000
--- a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK17chimStream/CumulantK17CompChimStreamDevice.cu
+++ /dev/null
@@ -1,640 +0,0 @@
-//=======================================================================================
-// ____ ____ __ ______ __________ __ __ __ __
-// \ \ | | | | | _ \ |___ ___| | | | | / \ | |
-// \ \ | | | | | |_) | | | | | | | / \ | |
-// \ \ | | | | | _ / | | | | | | / /\ \ | |
-// \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
-// \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
-// \ \ | | ________________________________________________________________
-// \ \ | | | ______________________________________________________________|
-// \ \| | | | __ __ __ __ ______ _______
-// \ | | |_____ | | | | | | | | | _ \ / _____)
-// \ | | _____| | | | | | | | | | | \ \ \_______
-// \ | | | | |_____ | \_/ | | | | |_/ / _____ |
-// \ _____| |__| |________| \_______/ |__| |______/ (_______/
-//
-// This file is part of VirtualFluids. VirtualFluids is free software: you can
-// redistribute it and/or modify it under the terms of the GNU General Public
-// License as published by the Free Software Foundation, either version 3 of
-// the License, or (at your option) any later version.
-//
-// VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
-// ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
-// FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
-// for more details.
-//
-// You should have received a copy of the GNU General Public License along
-// with VirtualFluids (see COPYING.txt). If not, see <http://www.gnu.org/licenses/>.
-//
-//! \file Cumulant27chimStream.cu
-//! \ingroup GPU
-//! \author Martin Schoenherr, Anna Wellmann
-//=======================================================================================
-/* Device code */
-#include "LBM/LB.h"
-#include "lbm/constants/D3Q27.h"
-#include <lbm/constants/NumericConstants.h>
-
-using namespace vf::lbm::constant;
-using namespace vf::lbm::dir;
-#include "Kernel/Utilities/ChimeraTransformation.h"
-
-////////////////////////////////////////////////////////////////////////////////
-__global__ void LB_Kernel_CumulantK17CompChimStream(
- real omega,
- uint* neighborX,
- uint* neighborY,
- uint* neighborZ,
- real* distributions,
- unsigned long size_Mat,
- int level,
- real* forces,
- real* quadricLimiters,
- bool isEvenTimestep,
- const uint *fluidNodeIndices,
- uint numberOfFluidNodes)
-{
- //////////////////////////////////////////////////////////////////////////
- //! Cumulant K17 Kernel is based on \ref
- //! <a href="https://doi.org/10.1016/j.jcp.2017.05.040"><b>[ M. Geier et al. (2017), DOI:10.1016/j.jcp.2017.05.040
- //! ]</b></a> and \ref <a href="https://doi.org/10.1016/j.jcp.2017.07.004"><b>[ M. Geier et al. (2017),
- //! DOI:10.1016/j.jcp.2017.07.004 ]</b></a>
- //!
- //! The cumulant kernel is executed in the following steps
- //!
- ////////////////////////////////////////////////////////////////////////////////
- //! - Get node index coordinates from threadIdx, blockIdx, blockDim and gridDim.
- //!
- const unsigned x = threadIdx.x;
- const unsigned y = blockIdx.x;
- const unsigned z = blockIdx.y;
-
- const unsigned nx = blockDim.x;
- const unsigned ny = gridDim.x;
-
- const unsigned k_thread = nx * (ny * z + y) + x;
-
- //////////////////////////////////////////////////////////////////////////
- // run for all indices in fluidNodeIndices
- if (k_thread < numberOfFluidNodes) {
- //////////////////////////////////////////////////////////////////////////
- //! - Read distributions: style of reading and writing the distributions from/to stored arrays dependent on
- //! timestep is based on the esoteric twist algorithm \ref <a
- //! href="https://doi.org/10.3390/computation5020019"><b>[ M. Geier et al. (2017),
- //! DOI:10.3390/computation5020019 ]</b></a>
-
- const unsigned k = fluidNodeIndices[k_thread];
-
- Distributions27 dist;
- if (isEvenTimestep) {
- dist.f[DIR_P00] = &distributions[DIR_P00 * size_Mat];
- dist.f[DIR_M00] = &distributions[DIR_M00 * size_Mat];
- dist.f[DIR_0P0] = &distributions[DIR_0P0 * size_Mat];
- dist.f[DIR_0M0] = &distributions[DIR_0M0 * size_Mat];
- dist.f[DIR_00P] = &distributions[DIR_00P * size_Mat];
- dist.f[DIR_00M] = &distributions[DIR_00M * size_Mat];
- dist.f[DIR_PP0] = &distributions[DIR_PP0 * size_Mat];
- dist.f[DIR_MM0] = &distributions[DIR_MM0 * size_Mat];
- dist.f[DIR_PM0] = &distributions[DIR_PM0 * size_Mat];
- dist.f[DIR_MP0] = &distributions[DIR_MP0 * size_Mat];
- dist.f[DIR_P0P] = &distributions[DIR_P0P * size_Mat];
- dist.f[DIR_M0M] = &distributions[DIR_M0M * size_Mat];
- dist.f[DIR_P0M] = &distributions[DIR_P0M * size_Mat];
- dist.f[DIR_M0P] = &distributions[DIR_M0P * size_Mat];
- dist.f[DIR_0PP] = &distributions[DIR_0PP * size_Mat];
- dist.f[DIR_0MM] = &distributions[DIR_0MM * size_Mat];
- dist.f[DIR_0PM] = &distributions[DIR_0PM * size_Mat];
- dist.f[DIR_0MP] = &distributions[DIR_0MP * size_Mat];
- dist.f[DIR_000] = &distributions[DIR_000 * size_Mat];
- dist.f[DIR_PPP] = &distributions[DIR_PPP * size_Mat];
- dist.f[DIR_MMP] = &distributions[DIR_MMP * size_Mat];
- dist.f[DIR_PMP] = &distributions[DIR_PMP * size_Mat];
- dist.f[DIR_MPP] = &distributions[DIR_MPP * size_Mat];
- dist.f[DIR_PPM] = &distributions[DIR_PPM * size_Mat];
- dist.f[DIR_MMM] = &distributions[DIR_MMM * size_Mat];
- dist.f[DIR_PMM] = &distributions[DIR_PMM * size_Mat];
- dist.f[DIR_MPM] = &distributions[DIR_MPM * size_Mat];
- } else {
- dist.f[DIR_M00] = &distributions[DIR_P00 * size_Mat];
- dist.f[DIR_P00] = &distributions[DIR_M00 * size_Mat];
- dist.f[DIR_0M0] = &distributions[DIR_0P0 * size_Mat];
- dist.f[DIR_0P0] = &distributions[DIR_0M0 * size_Mat];
- dist.f[DIR_00M] = &distributions[DIR_00P * size_Mat];
- dist.f[DIR_00P] = &distributions[DIR_00M * size_Mat];
- dist.f[DIR_MM0] = &distributions[DIR_PP0 * size_Mat];
- dist.f[DIR_PP0] = &distributions[DIR_MM0 * size_Mat];
- dist.f[DIR_MP0] = &distributions[DIR_PM0 * size_Mat];
- dist.f[DIR_PM0] = &distributions[DIR_MP0 * size_Mat];
- dist.f[DIR_M0M] = &distributions[DIR_P0P * size_Mat];
- dist.f[DIR_P0P] = &distributions[DIR_M0M * size_Mat];
- dist.f[DIR_M0P] = &distributions[DIR_P0M * size_Mat];
- dist.f[DIR_P0M] = &distributions[DIR_M0P * size_Mat];
- dist.f[DIR_0MM] = &distributions[DIR_0PP * size_Mat];
- dist.f[DIR_0PP] = &distributions[DIR_0MM * size_Mat];
- dist.f[DIR_0MP] = &distributions[DIR_0PM * size_Mat];
- dist.f[DIR_0PM] = &distributions[DIR_0MP * size_Mat];
- dist.f[DIR_000] = &distributions[DIR_000 * size_Mat];
- dist.f[DIR_MMM] = &distributions[DIR_PPP * size_Mat];
- dist.f[DIR_PPM] = &distributions[DIR_MMP * size_Mat];
- dist.f[DIR_MPM] = &distributions[DIR_PMP * size_Mat];
- dist.f[DIR_PMM] = &distributions[DIR_MPP * size_Mat];
- dist.f[DIR_MMP] = &distributions[DIR_PPM * size_Mat];
- dist.f[DIR_PPP] = &distributions[DIR_MMM * size_Mat];
- dist.f[DIR_MPP] = &distributions[DIR_PMM * size_Mat];
- dist.f[DIR_PMP] = &distributions[DIR_MPM * size_Mat];
- }
- ////////////////////////////////////////////////////////////////////////////////
- //! - Set neighbor indices (necessary for indirect addressing)
- uint kw = neighborX[k];
- uint ks = neighborY[k];
- uint kb = neighborZ[k];
- uint ksw = neighborY[kw];
- uint kbw = neighborZ[kw];
- uint kbs = neighborZ[ks];
- uint kbsw = neighborZ[ksw];
- ////////////////////////////////////////////////////////////////////////////////////
- //! - Set local distributions
- //!
- real mfcbb = (dist.f[DIR_P00])[k];
- real mfabb = (dist.f[DIR_M00])[kw];
- real mfbcb = (dist.f[DIR_0P0])[k];
- real mfbab = (dist.f[DIR_0M0])[ks];
- real mfbbc = (dist.f[DIR_00P])[k];
- real mfbba = (dist.f[DIR_00M])[kb];
- real mfccb = (dist.f[DIR_PP0])[k];
- real mfaab = (dist.f[DIR_MM0])[ksw];
- real mfcab = (dist.f[DIR_PM0])[ks];
- real mfacb = (dist.f[DIR_MP0])[kw];
- real mfcbc = (dist.f[DIR_P0P])[k];
- real mfaba = (dist.f[DIR_M0M])[kbw];
- real mfcba = (dist.f[DIR_P0M])[kb];
- real mfabc = (dist.f[DIR_M0P])[kw];
- real mfbcc = (dist.f[DIR_0PP])[k];
- real mfbaa = (dist.f[DIR_0MM])[kbs];
- real mfbca = (dist.f[DIR_0PM])[kb];
- real mfbac = (dist.f[DIR_0MP])[ks];
- real mfbbb = (dist.f[DIR_000])[k];
- real mfccc = (dist.f[DIR_PPP])[k];
- real mfaac = (dist.f[DIR_MMP])[ksw];
- real mfcac = (dist.f[DIR_PMP])[ks];
- real mfacc = (dist.f[DIR_MPP])[kw];
- real mfcca = (dist.f[DIR_PPM])[kb];
- real mfaaa = (dist.f[DIR_MMM])[kbsw];
- real mfcaa = (dist.f[DIR_PMM])[kbs];
- real mfaca = (dist.f[DIR_MPM])[kbw];
- ////////////////////////////////////////////////////////////////////////////////////
- //! - Calculate density and velocity using pyramid summation for low round-off errors as in Eq. (J1)-(J3) \ref
- //! <a href="https://doi.org/10.1016/j.camwa.2015.05.001"><b>[ M. Geier et al. (2015),
- //! DOI:10.1016/j.camwa.2015.05.001 ]</b></a>
- //!
- real drho = ((((mfccc + mfaaa) + (mfaca + mfcac)) + ((mfacc + mfcaa) + (mfaac + mfcca))) +
- (((mfbac + mfbca) + (mfbaa + mfbcc)) + ((mfabc + mfcba) + (mfaba + mfcbc)) +
- ((mfacb + mfcab) + (mfaab + mfccb))) +
- ((mfabb + mfcbb) + (mfbab + mfbcb) + (mfbba + mfbbc))) +
- mfbbb;
-
- real rho = c1o1 + drho;
- real OOrho = c1o1 / rho;
-
- real vvx = ((((mfccc - mfaaa) + (mfcac - mfaca)) + ((mfcaa - mfacc) + (mfcca - mfaac))) +
- (((mfcba - mfabc) + (mfcbc - mfaba)) + ((mfcab - mfacb) + (mfccb - mfaab))) + (mfcbb - mfabb)) *
- OOrho;
- real vvy = ((((mfccc - mfaaa) + (mfaca - mfcac)) + ((mfacc - mfcaa) + (mfcca - mfaac))) +
- (((mfbca - mfbac) + (mfbcc - mfbaa)) + ((mfacb - mfcab) + (mfccb - mfaab))) + (mfbcb - mfbab)) *
- OOrho;
- real vvz = ((((mfccc - mfaaa) + (mfcac - mfaca)) + ((mfacc - mfcaa) + (mfaac - mfcca))) +
- (((mfbac - mfbca) + (mfbcc - mfbaa)) + ((mfabc - mfcba) + (mfcbc - mfaba))) + (mfbbc - mfbba)) *
- OOrho;
- ////////////////////////////////////////////////////////////////////////////////////
- //! - Add half of the acceleration (body force) to the velocity as in Eq. (42) \ref
- //! <a href="https://doi.org/10.1016/j.camwa.2015.05.001"><b>[ M. Geier et al. (2015),
- //! DOI:10.1016/j.camwa.2015.05.001 ]</b></a>
- //!
- real factor = c1o1;
- for (size_t i = 1; i <= level; i++) {
- factor *= c2o1;
- }
- real fx = forces[0] / factor;
- real fy = forces[1] / factor;
- real fz = forces[2] / factor;
- vvx += fx * c1o2;
- vvy += fy * c1o2;
- vvz += fz * c1o2;
- ////////////////////////////////////////////////////////////////////////////////////
- // calculate the square of velocities for this lattice node
- real vx2 = vvx * vvx;
- real vy2 = vvy * vvy;
- real vz2 = vvz * vvz;
- ////////////////////////////////////////////////////////////////////////////////////
- //! - Set relaxation limiters for third order cumulants to default value \f$ \lambda=0.001 \f$ according to
- //! section 6 in \ref <a href="https://doi.org/10.1016/j.jcp.2017.05.040"><b>[ M. Geier et al. (2017),
- //! DOI:10.1016/j.jcp.2017.05.040 ]</b></a>
- //!
- real wadjust;
- real qudricLimitP = quadricLimiters[0];
- real qudricLimitM = quadricLimiters[1];
- real qudricLimitD = quadricLimiters[2];
- ////////////////////////////////////////////////////////////////////////////////////
- //! - Chimera transform from well conditioned distributions to central moments as defined in Appendix J in \ref
- //! <a href="https://doi.org/10.1016/j.camwa.2015.05.001"><b>[ M. Geier et al. (2015),
- //! DOI:10.1016/j.camwa.2015.05.001 ]</b></a> see also Eq. (6)-(14) in \ref <a
- //! href="https://doi.org/10.1016/j.jcp.2017.05.040"><b>[ M. Geier et al. (2017), DOI:10.1016/j.jcp.2017.05.040
- //! ]</b></a>
- //!
- ////////////////////////////////////////////////////////////////////////////////////
- // Z - Dir
- forwardInverseChimeraWithK(mfaaa, mfaab, mfaac, vvz, vz2, c36o1, c1o36);
- forwardInverseChimeraWithK(mfaba, mfabb, mfabc, vvz, vz2, c9o1, c1o9);
- forwardInverseChimeraWithK(mfaca, mfacb, mfacc, vvz, vz2, c36o1, c1o36);
- forwardInverseChimeraWithK(mfbaa, mfbab, mfbac, vvz, vz2, c9o1, c1o9);
- forwardInverseChimeraWithK(mfbba, mfbbb, mfbbc, vvz, vz2, c9o4, c4o9);
- forwardInverseChimeraWithK(mfbca, mfbcb, mfbcc, vvz, vz2, c9o1, c1o9);
- forwardInverseChimeraWithK(mfcaa, mfcab, mfcac, vvz, vz2, c36o1, c1o36);
- forwardInverseChimeraWithK(mfcba, mfcbb, mfcbc, vvz, vz2, c9o1, c1o9);
- forwardInverseChimeraWithK(mfcca, mfccb, mfccc, vvz, vz2, c36o1, c1o36);
-
- ////////////////////////////////////////////////////////////////////////////////////
- // Y - Dir
- forwardInverseChimeraWithK(mfaaa, mfaba, mfaca, vvy, vy2, c6o1, c1o6);
- forwardChimera(mfaab, mfabb, mfacb, vvy, vy2);
- forwardInverseChimeraWithK(mfaac, mfabc, mfacc, vvy, vy2, c18o1, c1o18);
- forwardInverseChimeraWithK(mfbaa, mfbba, mfbca, vvy, vy2, c3o2, c2o3);
- forwardChimera(mfbab, mfbbb, mfbcb, vvy, vy2);
- forwardInverseChimeraWithK(mfbac, mfbbc, mfbcc, vvy, vy2, c9o2, c2o9);
- forwardInverseChimeraWithK(mfcaa, mfcba, mfcca, vvy, vy2, c6o1, c1o6);
- forwardChimera(mfcab, mfcbb, mfccb, vvy, vy2);
- forwardInverseChimeraWithK(mfcac, mfcbc, mfccc, vvy, vy2, c18o1, c1o18);
-
- ////////////////////////////////////////////////////////////////////////////////////
- // X - Dir
- forwardInverseChimeraWithK(mfaaa, mfbaa, mfcaa, vvx, vx2, c1o1, c1o1);
- forwardChimera(mfaba, mfbba, mfcba, vvx, vx2);
- forwardInverseChimeraWithK(mfaca, mfbca, mfcca, vvx, vx2, c3o1, c1o3);
- forwardChimera(mfaab, mfbab, mfcab, vvx, vx2);
- forwardChimera(mfabb, mfbbb, mfcbb, vvx, vx2);
- forwardChimera(mfacb, mfbcb, mfccb, vvx, vx2);
- forwardInverseChimeraWithK(mfaac, mfbac, mfcac, vvx, vx2, c3o1, c1o3);
- forwardChimera(mfabc, mfbbc, mfcbc, vvx, vx2);
- forwardInverseChimeraWithK(mfacc, mfbcc, mfccc, vvx, vx2, c3o1, c1o9);
-
- ////////////////////////////////////////////////////////////////////////////////////
- //! - Setting relaxation rates for non-hydrodynamic cumulants (default values). Variable names and equations
- //! according to <a href="https://doi.org/10.1016/j.jcp.2017.05.040"><b>[ M. Geier et al. (2017),
- //! DOI:10.1016/j.jcp.2017.05.040 ]</b></a>
- //! => [NAME IN PAPER]=[NAME IN CODE]=[DEFAULT VALUE].
- //! - Trace of second order cumulants \f$ C_{200}+C_{020}+C_{002} \f$ used to adjust bulk
- //! viscosity:\f$\omega_2=OxxPyyPzz=1.0 \f$.
- //! - Third order cumulants \f$ C_{120}+C_{102}, C_{210}+C_{012}, C_{201}+C_{021} \f$: \f$ \omega_3=OxyyPxzz
- //! \f$ set according to Eq. (111) with simplifications assuming \f$ \omega_2=1.0\f$.
- //! - Third order cumulants \f$ C_{120}-C_{102}, C_{210}-C_{012}, C_{201}-C_{021} \f$: \f$ \omega_4 = OxyyMxzz
- //! \f$ set according to Eq. (112) with simplifications assuming \f$ \omega_2 = 1.0\f$.
- //! - Third order cumulants \f$ C_{111} \f$: \f$ \omega_5 = Oxyz \f$ set according to Eq. (113) with
- //! simplifications assuming \f$ \omega_2 = 1.0\f$ (modify for different bulk viscosity).
- //! - Fourth order cumulants \f$ C_{220}, C_{202}, C_{022}, C_{211}, C_{121}, C_{112} \f$: for simplification
- //! all set to the same default value \f$ \omega_6=\omega_7=\omega_8=O4=1.0 \f$.
- //! - Fifth order cumulants \f$ C_{221}, C_{212}, C_{122}\f$: \f$\omega_9=O5=1.0\f$.
- //! - Sixth order cumulant \f$ C_{222}\f$: \f$\omega_{10}=O6=1.0\f$.
- //!
- ////////////////////////////////////////////////////////////
- // 2.
- real OxxPyyPzz = c1o1;
- ////////////////////////////////////////////////////////////
- // 3.
- real OxyyPxzz = c8o1 * (-c2o1 + omega) * (c1o1 + c2o1 * omega) / (-c8o1 - c14o1 * omega + c7o1 * omega * omega);
- real OxyyMxzz =
- c8o1 * (-c2o1 + omega) * (-c7o1 + c4o1 * omega) / (c56o1 - c50o1 * omega + c9o1 * omega * omega);
- real Oxyz = c24o1 * (-c2o1 + omega) * (-c2o1 - c7o1 * omega + c3o1 * omega * omega) /
- (c48o1 + c152o1 * omega - c130o1 * omega * omega + c29o1 * omega * omega * omega);
- ////////////////////////////////////////////////////////////
- // 4.
- real O4 = c1o1;
- ////////////////////////////////////////////////////////////
- // 5.
- real O5 = c1o1;
- ////////////////////////////////////////////////////////////
- // 6.
- real O6 = c1o1;
-
- ////////////////////////////////////////////////////////////////////////////////////
- //! - A and DIR_00M: parameters for fourth order convergence of the diffusion term according to Eq. (115) and (116)
- //! <a href="https://doi.org/10.1016/j.jcp.2017.05.040"><b>[ M. Geier et al. (2017),
- //! DOI:10.1016/j.jcp.2017.05.040 ]</b></a> with simplifications assuming \f$ \omega_2 = 1.0 \f$ (modify for
- //! different bulk viscosity).
- //!
- real factorA = (c4o1 + c2o1 * omega - c3o1 * omega * omega) / (c2o1 - c7o1 * omega + c5o1 * omega * omega);
- real factorB = (c4o1 + c28o1 * omega - c14o1 * omega * omega) / (c6o1 - c21o1 * omega + c15o1 * omega * omega);
-
- ////////////////////////////////////////////////////////////////////////////////////
- //! - Compute cumulants from central moments according to Eq. (20)-(23) in
- //! <a href="https://doi.org/10.1016/j.jcp.2017.05.040"><b>[ M. Geier et al. (2017),
- //! DOI:10.1016/j.jcp.2017.05.040 ]</b></a>
- //!
- ////////////////////////////////////////////////////////////
- // 4.
- real CUMcbb = mfcbb - ((mfcaa + c1o3) * mfabb + c2o1 * mfbba * mfbab) * OOrho;
- real CUMbcb = mfbcb - ((mfaca + c1o3) * mfbab + c2o1 * mfbba * mfabb) * OOrho;
- real CUMbbc = mfbbc - ((mfaac + c1o3) * mfbba + c2o1 * mfbab * mfabb) * OOrho;
-
- real CUMcca =
- mfcca - (((mfcaa * mfaca + c2o1 * mfbba * mfbba) + c1o3 * (mfcaa + mfaca)) * OOrho - c1o9 * (drho * OOrho));
- real CUMcac =
- mfcac - (((mfcaa * mfaac + c2o1 * mfbab * mfbab) + c1o3 * (mfcaa + mfaac)) * OOrho - c1o9 * (drho * OOrho));
- real CUMacc =
- mfacc - (((mfaac * mfaca + c2o1 * mfabb * mfabb) + c1o3 * (mfaac + mfaca)) * OOrho - c1o9 * (drho * OOrho));
- ////////////////////////////////////////////////////////////
- // 5.
- real CUMbcc =
- mfbcc - ((mfaac * mfbca + mfaca * mfbac + c4o1 * mfabb * mfbbb + c2o1 * (mfbab * mfacb + mfbba * mfabc)) +
- c1o3 * (mfbca + mfbac)) *
- OOrho;
- real CUMcbc =
- mfcbc - ((mfaac * mfcba + mfcaa * mfabc + c4o1 * mfbab * mfbbb + c2o1 * (mfabb * mfcab + mfbba * mfbac)) +
- c1o3 * (mfcba + mfabc)) *
- OOrho;
- real CUMccb =
- mfccb - ((mfcaa * mfacb + mfaca * mfcab + c4o1 * mfbba * mfbbb + c2o1 * (mfbab * mfbca + mfabb * mfcba)) +
- c1o3 * (mfacb + mfcab)) *
- OOrho;
- ////////////////////////////////////////////////////////////
- // 6.
- real CUMccc = mfccc + ((-c4o1 * mfbbb * mfbbb - (mfcaa * mfacc + mfaca * mfcac + mfaac * mfcca) -
- c4o1 * (mfabb * mfcbb + mfbab * mfbcb + mfbba * mfbbc) -
- c2o1 * (mfbca * mfbac + mfcba * mfabc + mfcab * mfacb)) *
- OOrho +
- (c4o1 * (mfbab * mfbab * mfaca + mfabb * mfabb * mfcaa + mfbba * mfbba * mfaac) +
- c2o1 * (mfcaa * mfaca * mfaac) + c16o1 * mfbba * mfbab * mfabb) *
- OOrho * OOrho -
- c1o3 * (mfacc + mfcac + mfcca) * OOrho - c1o9 * (mfcaa + mfaca + mfaac) * OOrho +
- (c2o1 * (mfbab * mfbab + mfabb * mfabb + mfbba * mfbba) +
- (mfaac * mfaca + mfaac * mfcaa + mfaca * mfcaa) + c1o3 * (mfaac + mfaca + mfcaa)) *
- OOrho * OOrho * c2o3 +
- c1o27 * ((drho * drho - drho) * OOrho * OOrho));
-
- ////////////////////////////////////////////////////////////////////////////////////
- //! - Compute linear combinations of second and third order cumulants
- //!
- ////////////////////////////////////////////////////////////
- // 2.
- real mxxPyyPzz = mfcaa + mfaca + mfaac;
- real mxxMyy = mfcaa - mfaca;
- real mxxMzz = mfcaa - mfaac;
- ////////////////////////////////////////////////////////////
- // 3.
- real mxxyPyzz = mfcba + mfabc;
- real mxxyMyzz = mfcba - mfabc;
-
- real mxxzPyyz = mfcab + mfacb;
- real mxxzMyyz = mfcab - mfacb;
-
- real mxyyPxzz = mfbca + mfbac;
- real mxyyMxzz = mfbca - mfbac;
-
- ////////////////////////////////////////////////////////////////////////////////////
- // incl. correction
- ////////////////////////////////////////////////////////////
- //! - Compute velocity gradients from second order cumulants according to Eq. (27)-(32)
- //! <a href="https://doi.org/10.1016/j.jcp.2017.05.040"><b>[ M. Geier et al. (2017),
- //! DOI:10.1016/j.jcp.2017.05.040 ]</b></a> Further explanations of the correction in viscosity in Appendix H of
- //! <a href="https://doi.org/10.1016/j.camwa.2015.05.001"><b>[ M. Geier et al. (2015),
- //! DOI:10.1016/j.camwa.2015.05.001 ]</b></a> Note that the division by rho is omitted here as we need rho times
- //! the gradients later.
- //!
- real Dxy = -c3o1 * omega * mfbba;
- real Dxz = -c3o1 * omega * mfbab;
- real Dyz = -c3o1 * omega * mfabb;
- real dxux = c1o2 * (-omega) * (mxxMyy + mxxMzz) + c1o2 * OxxPyyPzz * (mfaaa - mxxPyyPzz);
- real dyuy = dxux + omega * c3o2 * mxxMyy;
- real dzuz = dxux + omega * c3o2 * mxxMzz;
- ////////////////////////////////////////////////////////////
- //! - Relaxation of second order cumulants with correction terms according to Eq. (33)-(35) in
- //! <a href="https://doi.org/10.1016/j.jcp.2017.05.040"><b>[ M. Geier et al. (2017),
- //! DOI:10.1016/j.jcp.2017.05.040 ]</b></a>
- //!
- mxxPyyPzz +=
- OxxPyyPzz * (mfaaa - mxxPyyPzz) - c3o1 * (c1o1 - c1o2 * OxxPyyPzz) * (vx2 * dxux + vy2 * dyuy + vz2 * dzuz);
- mxxMyy += omega * (-mxxMyy) - c3o1 * (c1o1 + c1o2 * (-omega)) * (vx2 * dxux - vy2 * dyuy);
- mxxMzz += omega * (-mxxMzz) - c3o1 * (c1o1 + c1o2 * (-omega)) * (vx2 * dxux - vz2 * dzuz);
-
- ////////////////////////////////////////////////////////////////////////////////////
- ////no correction
- // mxxPyyPzz += OxxPyyPzz*(mfaaa - mxxPyyPzz);
- // mxxMyy += -(-omega) * (-mxxMyy);
- // mxxMzz += -(-omega) * (-mxxMzz);
- //////////////////////////////////////////////////////////////////////////
- mfabb += omega * (-mfabb);
- mfbab += omega * (-mfbab);
- mfbba += omega * (-mfbba);
-
- ////////////////////////////////////////////////////////////////////////////////////
- // relax
- //////////////////////////////////////////////////////////////////////////
- // incl. limiter
- //! - Relaxation of third order cumulants including limiter according to Eq. (116)-(123)
- //! <a href="https://doi.org/10.1016/j.jcp.2017.05.040"><b>[ M. Geier et al. (2017),
- //! DOI:10.1016/j.jcp.2017.05.040 ]</b></a>
- //!
- wadjust = Oxyz + (c1o1 - Oxyz) * abs(mfbbb) / (abs(mfbbb) + qudricLimitD);
- mfbbb += wadjust * (-mfbbb);
- wadjust = OxyyPxzz + (c1o1 - OxyyPxzz) * abs(mxxyPyzz) / (abs(mxxyPyzz) + qudricLimitP);
- mxxyPyzz += wadjust * (-mxxyPyzz);
- wadjust = OxyyMxzz + (c1o1 - OxyyMxzz) * abs(mxxyMyzz) / (abs(mxxyMyzz) + qudricLimitM);
- mxxyMyzz += wadjust * (-mxxyMyzz);
- wadjust = OxyyPxzz + (c1o1 - OxyyPxzz) * abs(mxxzPyyz) / (abs(mxxzPyyz) + qudricLimitP);
- mxxzPyyz += wadjust * (-mxxzPyyz);
- wadjust = OxyyMxzz + (c1o1 - OxyyMxzz) * abs(mxxzMyyz) / (abs(mxxzMyyz) + qudricLimitM);
- mxxzMyyz += wadjust * (-mxxzMyyz);
- wadjust = OxyyPxzz + (c1o1 - OxyyPxzz) * abs(mxyyPxzz) / (abs(mxyyPxzz) + qudricLimitP);
- mxyyPxzz += wadjust * (-mxyyPxzz);
- wadjust = OxyyMxzz + (c1o1 - OxyyMxzz) * abs(mxyyMxzz) / (abs(mxyyMxzz) + qudricLimitM);
- mxyyMxzz += wadjust * (-mxyyMxzz);
- //////////////////////////////////////////////////////////////////////////
- // no limiter
- // mfbbb += OxyyMxzz * (-mfbbb);
- // mxxyPyzz += OxyyPxzz * (-mxxyPyzz);
- // mxxyMyzz += OxyyMxzz * (-mxxyMyzz);
- // mxxzPyyz += OxyyPxzz * (-mxxzPyyz);
- // mxxzMyyz += OxyyMxzz * (-mxxzMyyz);
- // mxyyPxzz += OxyyPxzz * (-mxyyPxzz);
- // mxyyMxzz += OxyyMxzz * (-mxyyMxzz);
-
- ////////////////////////////////////////////////////////////////////////////////////
- //! - Compute inverse linear combinations of second and third order cumulants
- //!
- mfcaa = c1o3 * (mxxMyy + mxxMzz + mxxPyyPzz);
- mfaca = c1o3 * (-c2o1 * mxxMyy + mxxMzz + mxxPyyPzz);
- mfaac = c1o3 * (mxxMyy - c2o1 * mxxMzz + mxxPyyPzz);
-
- mfcba = (mxxyMyzz + mxxyPyzz) * c1o2;
- mfabc = (-mxxyMyzz + mxxyPyzz) * c1o2;
- mfcab = (mxxzMyyz + mxxzPyyz) * c1o2;
- mfacb = (-mxxzMyyz + mxxzPyyz) * c1o2;
- mfbca = (mxyyMxzz + mxyyPxzz) * c1o2;
- mfbac = (-mxyyMxzz + mxyyPxzz) * c1o2;
- //////////////////////////////////////////////////////////////////////////
-
- //////////////////////////////////////////////////////////////////////////
- // 4.
- // no limiter
- //! - Relax fourth order cumulants to modified equilibrium for fourth order convergence of diffusion according
- //! to Eq. (43)-(48) <a href="https://doi.org/10.1016/j.jcp.2017.05.040"><b>[ M. Geier et al. (2017),
- //! DOI:10.1016/j.jcp.2017.05.040 ]</b></a>
- //!
- CUMacc = -O4 * (c1o1 / omega - c1o2) * (dyuy + dzuz) * c2o3 * factorA + (c1o1 - O4) * (CUMacc);
- CUMcac = -O4 * (c1o1 / omega - c1o2) * (dxux + dzuz) * c2o3 * factorA + (c1o1 - O4) * (CUMcac);
- CUMcca = -O4 * (c1o1 / omega - c1o2) * (dyuy + dxux) * c2o3 * factorA + (c1o1 - O4) * (CUMcca);
- CUMbbc = -O4 * (c1o1 / omega - c1o2) * Dxy * c1o3 * factorB + (c1o1 - O4) * (CUMbbc);
- CUMbcb = -O4 * (c1o1 / omega - c1o2) * Dxz * c1o3 * factorB + (c1o1 - O4) * (CUMbcb);
- CUMcbb = -O4 * (c1o1 / omega - c1o2) * Dyz * c1o3 * factorB + (c1o1 - O4) * (CUMcbb);
-
- //////////////////////////////////////////////////////////////////////////
- // 5.
- CUMbcc += O5 * (-CUMbcc);
- CUMcbc += O5 * (-CUMcbc);
- CUMccb += O5 * (-CUMccb);
-
- //////////////////////////////////////////////////////////////////////////
- // 6.
- CUMccc += O6 * (-CUMccc);
-
- ////////////////////////////////////////////////////////////////////////////////////
- //! - Compute central moments from post collision cumulants according to Eq. (53)-(56) in
- //! <a href="https://doi.org/10.1016/j.jcp.2017.05.040"><b>[ M. Geier et al. (2017),
- //! DOI:10.1016/j.jcp.2017.05.040 ]</b></a>
- //!
-
- //////////////////////////////////////////////////////////////////////////
- // 4.
- mfcbb = CUMcbb + c1o3 * ((c3o1 * mfcaa + c1o1) * mfabb + c6o1 * mfbba * mfbab) * OOrho;
- mfbcb = CUMbcb + c1o3 * ((c3o1 * mfaca + c1o1) * mfbab + c6o1 * mfbba * mfabb) * OOrho;
- mfbbc = CUMbbc + c1o3 * ((c3o1 * mfaac + c1o1) * mfbba + c6o1 * mfbab * mfabb) * OOrho;
-
- mfcca =
- CUMcca +
- (((mfcaa * mfaca + c2o1 * mfbba * mfbba) * c9o1 + c3o1 * (mfcaa + mfaca)) * OOrho - (drho * OOrho)) * c1o9;
- mfcac =
- CUMcac +
- (((mfcaa * mfaac + c2o1 * mfbab * mfbab) * c9o1 + c3o1 * (mfcaa + mfaac)) * OOrho - (drho * OOrho)) * c1o9;
- mfacc =
- CUMacc +
- (((mfaac * mfaca + c2o1 * mfabb * mfabb) * c9o1 + c3o1 * (mfaac + mfaca)) * OOrho - (drho * OOrho)) * c1o9;
-
- //////////////////////////////////////////////////////////////////////////
- // 5.
- mfbcc = CUMbcc + c1o3 *
- (c3o1 * (mfaac * mfbca + mfaca * mfbac + c4o1 * mfabb * mfbbb +
- c2o1 * (mfbab * mfacb + mfbba * mfabc)) +
- (mfbca + mfbac)) *
- OOrho;
- mfcbc = CUMcbc + c1o3 *
- (c3o1 * (mfaac * mfcba + mfcaa * mfabc + c4o1 * mfbab * mfbbb +
- c2o1 * (mfabb * mfcab + mfbba * mfbac)) +
- (mfcba + mfabc)) *
- OOrho;
- mfccb = CUMccb + c1o3 *
- (c3o1 * (mfcaa * mfacb + mfaca * mfcab + c4o1 * mfbba * mfbbb +
- c2o1 * (mfbab * mfbca + mfabb * mfcba)) +
- (mfacb + mfcab)) *
- OOrho;
-
- //////////////////////////////////////////////////////////////////////////
- // 6.
- mfccc = CUMccc - ((-c4o1 * mfbbb * mfbbb - (mfcaa * mfacc + mfaca * mfcac + mfaac * mfcca) -
- c4o1 * (mfabb * mfcbb + mfbab * mfbcb + mfbba * mfbbc) -
- c2o1 * (mfbca * mfbac + mfcba * mfabc + mfcab * mfacb)) *
- OOrho +
- (c4o1 * (mfbab * mfbab * mfaca + mfabb * mfabb * mfcaa + mfbba * mfbba * mfaac) +
- c2o1 * (mfcaa * mfaca * mfaac) + c16o1 * mfbba * mfbab * mfabb) *
- OOrho * OOrho -
- c1o3 * (mfacc + mfcac + mfcca) * OOrho - c1o9 * (mfcaa + mfaca + mfaac) * OOrho +
- (c2o1 * (mfbab * mfbab + mfabb * mfabb + mfbba * mfbba) +
- (mfaac * mfaca + mfaac * mfcaa + mfaca * mfcaa) + c1o3 * (mfaac + mfaca + mfcaa)) *
- OOrho * OOrho * c2o3 +
- c1o27 * ((drho * drho - drho) * OOrho * OOrho));
-
- ////////////////////////////////////////////////////////////////////////////////////
- //! - Add acceleration (body force) to first order cumulants according to Eq. (85)-(87) in
- //! <a href="https://doi.org/10.1016/j.camwa.2015.05.001"><b>[ M. Geier et al. (2015),
- //! DOI:10.1016/j.camwa.2015.05.001 ]</b></a>
- //!
- mfbaa = -mfbaa;
- mfaba = -mfaba;
- mfaab = -mfaab;
-
- ////////////////////////////////////////////////////////////////////////////////////
- //! - Chimera transform from central moments to well conditioned distributions as defined in Appendix J in
- //! <a href="https://doi.org/10.1016/j.camwa.2015.05.001"><b>[ M. Geier et al. (2015),
- //! DOI:10.1016/j.camwa.2015.05.001 ]</b></a> see also Eq. (88)-(96) in <a
- //! href="https://doi.org/10.1016/j.jcp.2017.05.040"><b>[ M. Geier et al. (2017), DOI:10.1016/j.jcp.2017.05.040
- //! ]</b></a>
- //!
- ////////////////////////////////////////////////////////////////////////////////////
- // X - Dir
- backwardInverseChimeraWithK(mfaaa, mfbaa, mfcaa, vvx, vx2, c1o1, c1o1);
- backwardChimera(mfaba, mfbba, mfcba, vvx, vx2);
- backwardInverseChimeraWithK(mfaca, mfbca, mfcca, vvx, vx2, c3o1, c1o3);
- backwardChimera(mfaab, mfbab, mfcab, vvx, vx2);
- backwardChimera(mfabb, mfbbb, mfcbb, vvx, vx2);
- backwardChimera(mfacb, mfbcb, mfccb, vvx, vx2);
- backwardInverseChimeraWithK(mfaac, mfbac, mfcac, vvx, vx2, c3o1, c1o3);
- backwardChimera(mfabc, mfbbc, mfcbc, vvx, vx2);
- backwardInverseChimeraWithK(mfacc, mfbcc, mfccc, vvx, vx2, c9o1, c1o9);
-
- ////////////////////////////////////////////////////////////////////////////////////
- // Y - Dir
- backwardInverseChimeraWithK(mfaaa, mfaba, mfaca, vvy, vy2, c6o1, c1o6);
- backwardChimera(mfaab, mfabb, mfacb, vvy, vy2);
- backwardInverseChimeraWithK(mfaac, mfabc, mfacc, vvy, vy2, c18o1, c1o18);
- backwardInverseChimeraWithK(mfbaa, mfbba, mfbca, vvy, vy2, c3o2, c2o3);
- backwardChimera(mfbab, mfbbb, mfbcb, vvy, vy2);
- backwardInverseChimeraWithK(mfbac, mfbbc, mfbcc, vvy, vy2, c9o2, c2o9);
- backwardInverseChimeraWithK(mfcaa, mfcba, mfcca, vvy, vy2, c6o1, c1o6);
- backwardChimera(mfcab, mfcbb, mfccb, vvy, vy2);
- backwardInverseChimeraWithK(mfcac, mfcbc, mfccc, vvy, vy2, c18o1, c1o18);
-
- ////////////////////////////////////////////////////////////////////////////////////
- // Z - Dir
- backwardInverseChimeraWithK(mfaaa, mfaab, mfaac, vvz, vz2, c36o1, c1o36);
- backwardInverseChimeraWithK(mfaba, mfabb, mfabc, vvz, vz2, c9o1, c1o9);
- backwardInverseChimeraWithK(mfaca, mfacb, mfacc, vvz, vz2, c36o1, c1o36);
- backwardInverseChimeraWithK(mfbaa, mfbab, mfbac, vvz, vz2, c9o1, c1o9);
- backwardInverseChimeraWithK(mfbba, mfbbb, mfbbc, vvz, vz2, c9o4, c4o9);
- backwardInverseChimeraWithK(mfbca, mfbcb, mfbcc, vvz, vz2, c9o1, c1o9);
- backwardInverseChimeraWithK(mfcaa, mfcab, mfcac, vvz, vz2, c36o1, c1o36);
- backwardInverseChimeraWithK(mfcba, mfcbb, mfcbc, vvz, vz2, c9o1, c1o9);
- backwardInverseChimeraWithK(mfcca, mfccb, mfccc, vvz, vz2, c36o1, c1o36);
-
- ////////////////////////////////////////////////////////////////////////////////////
- //! - Write distributions: style of reading and writing the distributions from/to
- //! stored arrays dependent on timestep is based on the esoteric twist algorithm
- //! <a href="https://doi.org/10.3390/computation5020019"><b>[ M. Geier et al. (2017),
- //! DOI:10.3390/computation5020019 ]</b></a>
- //!
- (dist.f[DIR_P00])[k] = mfabb;
- (dist.f[DIR_M00])[kw] = mfcbb;
- (dist.f[DIR_0P0])[k] = mfbab;
- (dist.f[DIR_0M0])[ks] = mfbcb;
- (dist.f[DIR_00P])[k] = mfbba;
- (dist.f[DIR_00M])[kb] = mfbbc;
- (dist.f[DIR_PP0])[k] = mfaab;
- (dist.f[DIR_MM0])[ksw] = mfccb;
- (dist.f[DIR_PM0])[ks] = mfacb;
- (dist.f[DIR_MP0])[kw] = mfcab;
- (dist.f[DIR_P0P])[k] = mfaba;
- (dist.f[DIR_M0M])[kbw] = mfcbc;
- (dist.f[DIR_P0M])[kb] = mfabc;
- (dist.f[DIR_M0P])[kw] = mfcba;
- (dist.f[DIR_0PP])[k] = mfbaa;
- (dist.f[DIR_0MM])[kbs] = mfbcc;
- (dist.f[DIR_0PM])[kb] = mfbac;
- (dist.f[DIR_0MP])[ks] = mfbca;
- (dist.f[DIR_000])[k] = mfbbb;
- (dist.f[DIR_PPP])[k] = mfaaa;
- (dist.f[DIR_PMP])[ks] = mfaca;
- (dist.f[DIR_PPM])[kb] = mfaac;
- (dist.f[DIR_PMM])[kbs] = mfacc;
- (dist.f[DIR_MPP])[kw] = mfcaa;
- (dist.f[DIR_MMP])[ksw] = mfcca;
- (dist.f[DIR_MPM])[kbw] = mfcac;
- (dist.f[DIR_MMM])[kbsw] = mfccc;
- }
-}
\ No newline at end of file
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK17chimStream/CumulantK17CompChimStream_Device.cuh b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK17chimStream/CumulantK17CompChimStream_Device.cuh
deleted file mode 100644
index f74192c0423ba9dc96820d7f46eecb9d49a39ed4..0000000000000000000000000000000000000000
--- a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK17chimStream/CumulantK17CompChimStream_Device.cuh
+++ /dev/null
@@ -1,20 +0,0 @@
-#ifndef LB_Kernel_CUMULANT_K17_COMP_CHIM_SPARSE_H
-#define LB_Kernel_CUMULANT_K17_COMP_CHIM_SPARSE_H
-
-#include <DataTypes.h>
-#include <curand.h>
-
-__global__ void LB_Kernel_CumulantK17CompChimStream(
- real omega,
- uint* neighborX,
- uint* neighborY,
- uint* neighborZ,
- real* distributions,
- unsigned long size_Mat,
- int level,
- real* forces,
- real* quadricLimiters,
- bool isEvenTimestep,
- const uint* fluidNodeIndices,
- uint numberOfFluidNodes);
-#endif
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK18/CumulantK18Comp.cu b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK18/CumulantK18Comp.cu
index 54af306039585f3beb39b05f2f2e0a96ae784e12..2e0af0bdb85d3f008768f9f430e8b4e5d9719b0f 100644
--- a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK18/CumulantK18Comp.cu
+++ b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK18/CumulantK18Comp.cu
@@ -1,8 +1,8 @@
#include "CumulantK18Comp.h"
#include "CumulantK18Comp_Device.cuh"
-
#include "Parameter/Parameter.h"
+#include "cuda/CudaGrid.h"
std::shared_ptr<CumulantK18Comp> CumulantK18Comp::getNewInstance(std::shared_ptr<Parameter> para, int level)
{
@@ -11,37 +11,22 @@ std::shared_ptr<CumulantK18Comp> CumulantK18Comp::getNewInstance(std::shared_ptr
void CumulantK18Comp::run()
{
- int numberOfThreads = para->getParD(level)->numberofthreads;
- int size_Mat = para->getParD(level)->numberOfNodes;
-
- int Grid = (size_Mat / numberOfThreads) + 1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid / Grid1) + 1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 grid(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1);
-
- LB_Kernel_CumulantK18Comp << < grid, threads >> >( para->getParD(level)->omega,
- para->getParD(level)->typeOfGridNode,
- para->getParD(level)->neighborX,
- para->getParD(level)->neighborY,
- para->getParD(level)->neighborZ,
- para->getParD(level)->distributions.f[0],
- para->getParD(level)->g6.g[0],
- para->getParD(level)->numberOfNodes,
- level,
- para->getForcesDev(),
- para->getQuadricLimitersDev(),
- para->getParD(level)->isEvenTimestep);
- getLastCudaError("LB_Kernel_CumulantK18Comp execution failed");
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(para->getParD(level)->numberofthreads, para->getParD(level)->numberOfNodes);
+
+ LB_Kernel_CumulantK18Comp <<< grid.grid, grid.threads >>>(
+ para->getParD(level)->omega,
+ para->getParD(level)->typeOfGridNode,
+ para->getParD(level)->neighborX,
+ para->getParD(level)->neighborY,
+ para->getParD(level)->neighborZ,
+ para->getParD(level)->distributions.f[0],
+ para->getParD(level)->g6.g[0],
+ para->getParD(level)->numberOfNodes,
+ level,
+ para->getForcesDev(),
+ para->getQuadricLimitersDev(),
+ para->getParD(level)->isEvenTimestep);
+ getLastCudaError("LB_Kernel_CumulantK18Comp execution failed");
}
CumulantK18Comp::CumulantK18Comp(std::shared_ptr<Parameter> para, int level)
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK18/CumulantK18Comp_Device.cu b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK18/CumulantK18Comp_Device.cu
index bb42d113e47ce28f153ac295f2d9a934dd1b213a..0e4ae5caebb9bd4b1c889a78bfadb62487742c98 100644
--- a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK18/CumulantK18Comp_Device.cu
+++ b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK18/CumulantK18Comp_Device.cu
@@ -42,83 +42,83 @@ __global__ void LB_Kernel_CumulantK18Comp(
Distributions27 D;
if (EvenOrOdd == true)
{
- D.f[DIR_P00] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_M00] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0P0] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0M0] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00P] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00M] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_PP0] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_MM0] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_PM0] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_MP0] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_P0P] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_M0M] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_P0M] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_M0P] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0PP] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0MM] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0PM] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0MP] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_PPP] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_MMP] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_PMP] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_MPP] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_PPM] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_MMM] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_PMM] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_MPM] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DDStart[DIR_P00 * size_Mat];
+ D.f[DIR_M00] = &DDStart[DIR_M00 * size_Mat];
+ D.f[DIR_0P0] = &DDStart[DIR_0P0 * size_Mat];
+ D.f[DIR_0M0] = &DDStart[DIR_0M0 * size_Mat];
+ D.f[DIR_00P] = &DDStart[DIR_00P * size_Mat];
+ D.f[DIR_00M] = &DDStart[DIR_00M * size_Mat];
+ D.f[DIR_PP0] = &DDStart[DIR_PP0 * size_Mat];
+ D.f[DIR_MM0] = &DDStart[DIR_MM0 * size_Mat];
+ D.f[DIR_PM0] = &DDStart[DIR_PM0 * size_Mat];
+ D.f[DIR_MP0] = &DDStart[DIR_MP0 * size_Mat];
+ D.f[DIR_P0P] = &DDStart[DIR_P0P * size_Mat];
+ D.f[DIR_M0M] = &DDStart[DIR_M0M * size_Mat];
+ D.f[DIR_P0M] = &DDStart[DIR_P0M * size_Mat];
+ D.f[DIR_M0P] = &DDStart[DIR_M0P * size_Mat];
+ D.f[DIR_0PP] = &DDStart[DIR_0PP * size_Mat];
+ D.f[DIR_0MM] = &DDStart[DIR_0MM * size_Mat];
+ D.f[DIR_0PM] = &DDStart[DIR_0PM * size_Mat];
+ D.f[DIR_0MP] = &DDStart[DIR_0MP * size_Mat];
+ D.f[DIR_000] = &DDStart[DIR_000 * size_Mat];
+ D.f[DIR_PPP] = &DDStart[DIR_PPP * size_Mat];
+ D.f[DIR_MMP] = &DDStart[DIR_MMP * size_Mat];
+ D.f[DIR_PMP] = &DDStart[DIR_PMP * size_Mat];
+ D.f[DIR_MPP] = &DDStart[DIR_MPP * size_Mat];
+ D.f[DIR_PPM] = &DDStart[DIR_PPM * size_Mat];
+ D.f[DIR_MMM] = &DDStart[DIR_MMM * size_Mat];
+ D.f[DIR_PMM] = &DDStart[DIR_PMM * size_Mat];
+ D.f[DIR_MPM] = &DDStart[DIR_MPM * size_Mat];
}
else
{
- D.f[DIR_M00] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_P00] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0M0] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0P0] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00M] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00P] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_MM0] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_PP0] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_MP0] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_PM0] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_M0M] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_P0P] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_M0P] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_P0M] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0MM] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0PP] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0MP] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0PM] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_MMM] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_PPM] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_MPM] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_PMM] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_MMP] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_PPP] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_MPP] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_PMP] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_M00] = &DDStart[DIR_P00 * size_Mat];
+ D.f[DIR_P00] = &DDStart[DIR_M00 * size_Mat];
+ D.f[DIR_0M0] = &DDStart[DIR_0P0 * size_Mat];
+ D.f[DIR_0P0] = &DDStart[DIR_0M0 * size_Mat];
+ D.f[DIR_00M] = &DDStart[DIR_00P * size_Mat];
+ D.f[DIR_00P] = &DDStart[DIR_00M * size_Mat];
+ D.f[DIR_MM0] = &DDStart[DIR_PP0 * size_Mat];
+ D.f[DIR_PP0] = &DDStart[DIR_MM0 * size_Mat];
+ D.f[DIR_MP0] = &DDStart[DIR_PM0 * size_Mat];
+ D.f[DIR_PM0] = &DDStart[DIR_MP0 * size_Mat];
+ D.f[DIR_M0M] = &DDStart[DIR_P0P * size_Mat];
+ D.f[DIR_P0P] = &DDStart[DIR_M0M * size_Mat];
+ D.f[DIR_M0P] = &DDStart[DIR_P0M * size_Mat];
+ D.f[DIR_P0M] = &DDStart[DIR_M0P * size_Mat];
+ D.f[DIR_0MM] = &DDStart[DIR_0PP * size_Mat];
+ D.f[DIR_0PP] = &DDStart[DIR_0MM * size_Mat];
+ D.f[DIR_0MP] = &DDStart[DIR_0PM * size_Mat];
+ D.f[DIR_0PM] = &DDStart[DIR_0MP * size_Mat];
+ D.f[DIR_000] = &DDStart[DIR_000 * size_Mat];
+ D.f[DIR_MMM] = &DDStart[DIR_PPP * size_Mat];
+ D.f[DIR_PPM] = &DDStart[DIR_MMP * size_Mat];
+ D.f[DIR_MPM] = &DDStart[DIR_PMP * size_Mat];
+ D.f[DIR_PMM] = &DDStart[DIR_MPP * size_Mat];
+ D.f[DIR_MMP] = &DDStart[DIR_PPM * size_Mat];
+ D.f[DIR_PPP] = &DDStart[DIR_MMM * size_Mat];
+ D.f[DIR_MPP] = &DDStart[DIR_PMM * size_Mat];
+ D.f[DIR_PMP] = &DDStart[DIR_MPM * size_Mat];
}
Distributions6 G;
if (EvenOrOdd == true)
{
- G.g[DIR_P00] = &G6[DIR_P00 *size_Mat];
- G.g[DIR_M00] = &G6[DIR_M00 *size_Mat];
- G.g[DIR_0P0] = &G6[DIR_0P0 *size_Mat];
- G.g[DIR_0M0] = &G6[DIR_0M0 *size_Mat];
- G.g[DIR_00P] = &G6[DIR_00P *size_Mat];
- G.g[DIR_00M] = &G6[DIR_00M *size_Mat];
+ G.g[DIR_P00] = &G6[DIR_P00 * size_Mat];
+ G.g[DIR_M00] = &G6[DIR_M00 * size_Mat];
+ G.g[DIR_0P0] = &G6[DIR_0P0 * size_Mat];
+ G.g[DIR_0M0] = &G6[DIR_0M0 * size_Mat];
+ G.g[DIR_00P] = &G6[DIR_00P * size_Mat];
+ G.g[DIR_00M] = &G6[DIR_00M * size_Mat];
}
else
{
- G.g[DIR_M00] = &G6[DIR_P00 *size_Mat];
- G.g[DIR_P00] = &G6[DIR_M00 *size_Mat];
- G.g[DIR_0M0] = &G6[DIR_0P0 *size_Mat];
- G.g[DIR_0P0] = &G6[DIR_0M0 *size_Mat];
- G.g[DIR_00M] = &G6[DIR_00P *size_Mat];
- G.g[DIR_00P] = &G6[DIR_00M *size_Mat];
+ G.g[DIR_M00] = &G6[DIR_P00 * size_Mat];
+ G.g[DIR_P00] = &G6[DIR_M00 * size_Mat];
+ G.g[DIR_0M0] = &G6[DIR_0P0 * size_Mat];
+ G.g[DIR_0P0] = &G6[DIR_0M0 * size_Mat];
+ G.g[DIR_00M] = &G6[DIR_00P * size_Mat];
+ G.g[DIR_00P] = &G6[DIR_00M * size_Mat];
}
////////////////////////////////////////////////////////////////////////////////
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK20/CumulantK20Comp.cu b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK20/CumulantK20Comp.cu
index 0c1778dc39496c6564dedcbe1f6e818bee147191..d0d81eaac711d4d80284b66a1040e0e8404f5d4d 100644
--- a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK20/CumulantK20Comp.cu
+++ b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK20/CumulantK20Comp.cu
@@ -1,8 +1,8 @@
#include "CumulantK20Comp.h"
#include "CumulantK20Comp_Device.cuh"
-
#include "Parameter/Parameter.h"
+#include "cuda/CudaGrid.h"
std::shared_ptr<CumulantK20Comp> CumulantK20Comp::getNewInstance(std::shared_ptr<Parameter> para, int level)
{
@@ -11,37 +11,22 @@ std::shared_ptr<CumulantK20Comp> CumulantK20Comp::getNewInstance(std::shared_ptr
void CumulantK20Comp::run()
{
- int numberOfThreads = para->getParD(level)->numberofthreads;
- int size_Mat = para->getParD(level)->numberOfNodes;
-
- int Grid = (size_Mat / numberOfThreads) + 1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid / Grid1) + 1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 grid(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1);
-
- LB_Kernel_CumulantK20Comp << < grid, threads >> >( para->getParD(level)->omega,
- para->getParD(level)->typeOfGridNode,
- para->getParD(level)->neighborX,
- para->getParD(level)->neighborY,
- para->getParD(level)->neighborZ,
- para->getParD(level)->distributions.f[0],
- para->getParD(level)->g6.g[0],
- para->getParD(level)->numberOfNodes,
- level,
- para->getForcesDev(),
- para->getQuadricLimitersDev(),
- para->getParD(level)->isEvenTimestep);
- getLastCudaError("LB_Kernel_CumulantK20Comp execution failed");
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(para->getParD(level)->numberofthreads, para->getParD(level)->numberOfNodes);
+
+ LB_Kernel_CumulantK20Comp <<< grid.grid, grid.threads >>>(
+ para->getParD(level)->omega,
+ para->getParD(level)->typeOfGridNode,
+ para->getParD(level)->neighborX,
+ para->getParD(level)->neighborY,
+ para->getParD(level)->neighborZ,
+ para->getParD(level)->distributions.f[0],
+ para->getParD(level)->g6.g[0],
+ para->getParD(level)->numberOfNodes,
+ level,
+ para->getForcesDev(),
+ para->getQuadricLimitersDev(),
+ para->getParD(level)->isEvenTimestep);
+ getLastCudaError("LB_Kernel_CumulantK20Comp execution failed");
}
CumulantK20Comp::CumulantK20Comp(std::shared_ptr<Parameter> para, int level)
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK20/CumulantK20Comp_Device.cu b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK20/CumulantK20Comp_Device.cu
index c805fc293aeb8b182bb0e01df82b584da69d0175..2dbe0bb62412f9363fdd0e714f5da296f81ae5b3 100644
--- a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK20/CumulantK20Comp_Device.cu
+++ b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK20/CumulantK20Comp_Device.cu
@@ -42,83 +42,83 @@ __global__ void LB_Kernel_CumulantK20Comp(
Distributions27 D;
if (EvenOrOdd == true)
{
- D.f[DIR_P00] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_M00] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0P0] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0M0] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00P] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00M] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_PP0] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_MM0] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_PM0] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_MP0] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_P0P] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_M0M] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_P0M] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_M0P] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0PP] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0MM] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0PM] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0MP] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_PPP] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_MMP] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_PMP] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_MPP] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_PPM] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_MMM] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_PMM] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_MPM] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DDStart[DIR_P00 * size_Mat];
+ D.f[DIR_M00] = &DDStart[DIR_M00 * size_Mat];
+ D.f[DIR_0P0] = &DDStart[DIR_0P0 * size_Mat];
+ D.f[DIR_0M0] = &DDStart[DIR_0M0 * size_Mat];
+ D.f[DIR_00P] = &DDStart[DIR_00P * size_Mat];
+ D.f[DIR_00M] = &DDStart[DIR_00M * size_Mat];
+ D.f[DIR_PP0] = &DDStart[DIR_PP0 * size_Mat];
+ D.f[DIR_MM0] = &DDStart[DIR_MM0 * size_Mat];
+ D.f[DIR_PM0] = &DDStart[DIR_PM0 * size_Mat];
+ D.f[DIR_MP0] = &DDStart[DIR_MP0 * size_Mat];
+ D.f[DIR_P0P] = &DDStart[DIR_P0P * size_Mat];
+ D.f[DIR_M0M] = &DDStart[DIR_M0M * size_Mat];
+ D.f[DIR_P0M] = &DDStart[DIR_P0M * size_Mat];
+ D.f[DIR_M0P] = &DDStart[DIR_M0P * size_Mat];
+ D.f[DIR_0PP] = &DDStart[DIR_0PP * size_Mat];
+ D.f[DIR_0MM] = &DDStart[DIR_0MM * size_Mat];
+ D.f[DIR_0PM] = &DDStart[DIR_0PM * size_Mat];
+ D.f[DIR_0MP] = &DDStart[DIR_0MP * size_Mat];
+ D.f[DIR_000] = &DDStart[DIR_000 * size_Mat];
+ D.f[DIR_PPP] = &DDStart[DIR_PPP * size_Mat];
+ D.f[DIR_MMP] = &DDStart[DIR_MMP * size_Mat];
+ D.f[DIR_PMP] = &DDStart[DIR_PMP * size_Mat];
+ D.f[DIR_MPP] = &DDStart[DIR_MPP * size_Mat];
+ D.f[DIR_PPM] = &DDStart[DIR_PPM * size_Mat];
+ D.f[DIR_MMM] = &DDStart[DIR_MMM * size_Mat];
+ D.f[DIR_PMM] = &DDStart[DIR_PMM * size_Mat];
+ D.f[DIR_MPM] = &DDStart[DIR_MPM * size_Mat];
}
else
{
- D.f[DIR_M00] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_P00] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0M0] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0P0] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00M] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00P] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_MM0] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_PP0] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_MP0] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_PM0] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_M0M] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_P0P] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_M0P] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_P0M] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0MM] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0PP] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0MP] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0PM] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_MMM] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_PPM] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_MPM] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_PMM] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_MMP] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_PPP] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_MPP] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_PMP] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_M00] = &DDStart[DIR_P00 * size_Mat];
+ D.f[DIR_P00] = &DDStart[DIR_M00 * size_Mat];
+ D.f[DIR_0M0] = &DDStart[DIR_0P0 * size_Mat];
+ D.f[DIR_0P0] = &DDStart[DIR_0M0 * size_Mat];
+ D.f[DIR_00M] = &DDStart[DIR_00P * size_Mat];
+ D.f[DIR_00P] = &DDStart[DIR_00M * size_Mat];
+ D.f[DIR_MM0] = &DDStart[DIR_PP0 * size_Mat];
+ D.f[DIR_PP0] = &DDStart[DIR_MM0 * size_Mat];
+ D.f[DIR_MP0] = &DDStart[DIR_PM0 * size_Mat];
+ D.f[DIR_PM0] = &DDStart[DIR_MP0 * size_Mat];
+ D.f[DIR_M0M] = &DDStart[DIR_P0P * size_Mat];
+ D.f[DIR_P0P] = &DDStart[DIR_M0M * size_Mat];
+ D.f[DIR_M0P] = &DDStart[DIR_P0M * size_Mat];
+ D.f[DIR_P0M] = &DDStart[DIR_M0P * size_Mat];
+ D.f[DIR_0MM] = &DDStart[DIR_0PP * size_Mat];
+ D.f[DIR_0PP] = &DDStart[DIR_0MM * size_Mat];
+ D.f[DIR_0MP] = &DDStart[DIR_0PM * size_Mat];
+ D.f[DIR_0PM] = &DDStart[DIR_0MP * size_Mat];
+ D.f[DIR_000] = &DDStart[DIR_000 * size_Mat];
+ D.f[DIR_MMM] = &DDStart[DIR_PPP * size_Mat];
+ D.f[DIR_PPM] = &DDStart[DIR_MMP * size_Mat];
+ D.f[DIR_MPM] = &DDStart[DIR_PMP * size_Mat];
+ D.f[DIR_PMM] = &DDStart[DIR_MPP * size_Mat];
+ D.f[DIR_MMP] = &DDStart[DIR_PPM * size_Mat];
+ D.f[DIR_PPP] = &DDStart[DIR_MMM * size_Mat];
+ D.f[DIR_MPP] = &DDStart[DIR_PMM * size_Mat];
+ D.f[DIR_PMP] = &DDStart[DIR_MPM * size_Mat];
}
Distributions6 G;
if (EvenOrOdd == true)
{
- G.g[DIR_P00] = &G6[DIR_P00 *size_Mat];
- G.g[DIR_M00] = &G6[DIR_M00 *size_Mat];
- G.g[DIR_0P0] = &G6[DIR_0P0 *size_Mat];
- G.g[DIR_0M0] = &G6[DIR_0M0 *size_Mat];
- G.g[DIR_00P] = &G6[DIR_00P *size_Mat];
- G.g[DIR_00M] = &G6[DIR_00M *size_Mat];
+ G.g[DIR_P00] = &G6[DIR_P00 * size_Mat];
+ G.g[DIR_M00] = &G6[DIR_M00 * size_Mat];
+ G.g[DIR_0P0] = &G6[DIR_0P0 * size_Mat];
+ G.g[DIR_0M0] = &G6[DIR_0M0 * size_Mat];
+ G.g[DIR_00P] = &G6[DIR_00P * size_Mat];
+ G.g[DIR_00M] = &G6[DIR_00M * size_Mat];
}
else
{
- G.g[DIR_M00] = &G6[DIR_P00 *size_Mat];
- G.g[DIR_P00] = &G6[DIR_M00 *size_Mat];
- G.g[DIR_0M0] = &G6[DIR_0P0 *size_Mat];
- G.g[DIR_0P0] = &G6[DIR_0M0 *size_Mat];
- G.g[DIR_00M] = &G6[DIR_00P *size_Mat];
- G.g[DIR_00P] = &G6[DIR_00M *size_Mat];
+ G.g[DIR_M00] = &G6[DIR_P00 * size_Mat];
+ G.g[DIR_P00] = &G6[DIR_M00 * size_Mat];
+ G.g[DIR_0M0] = &G6[DIR_0P0 * size_Mat];
+ G.g[DIR_0P0] = &G6[DIR_0M0 * size_Mat];
+ G.g[DIR_00M] = &G6[DIR_00P * size_Mat];
+ G.g[DIR_00P] = &G6[DIR_00M * size_Mat];
}
////////////////////////////////////////////////////////////////////////////////
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/MRT/MRTCompSP27.cu b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/MRT/MRTCompSP27.cu
index be94791572f739fb2eef7c049702caeedb6641fc..b576333f50304f5628e073d2eee16cf5b82c9d34 100644
--- a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/MRT/MRTCompSP27.cu
+++ b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/MRT/MRTCompSP27.cu
@@ -2,6 +2,7 @@
#include "MRTCompSP27_Device.cuh"
#include "Parameter/Parameter.h"
+#include "cuda/CudaGrid.h"
std::shared_ptr<MRTCompSP27> MRTCompSP27::getNewInstance(std::shared_ptr<Parameter> para, int level)
{
@@ -10,33 +11,18 @@ std::shared_ptr<MRTCompSP27> MRTCompSP27::getNewInstance(std::shared_ptr<Paramet
void MRTCompSP27::run()
{
- int size_Mat = para->getParD(level)->numberOfNodes;
- int numberOfThreads = para->getParD(level)->numberofthreads;
-
- int Grid = (size_Mat / numberOfThreads) + 1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid / Grid1) + 1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 grid(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1);
-
- LB_Kernel_MRT_Comp_SP_27 << < grid, threads >> >( para->getParD(level)->omega,
- para->getParD(level)->typeOfGridNode,
- para->getParD(level)->neighborX,
- para->getParD(level)->neighborY,
- para->getParD(level)->neighborZ,
- para->getParD(level)->distributions.f[0],
- para->getParD(level)->numberOfNodes,
- para->getParD(level)->isEvenTimestep);
- getLastCudaError("LB_Kernel_MRT_Comp_SP_27 execution failed");
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(para->getParD(level)->numberofthreads, para->getParD(level)->numberOfNodes);
+
+ LB_Kernel_MRT_Comp_SP_27 <<< grid.grid, grid.threads >>>(
+ para->getParD(level)->omega,
+ para->getParD(level)->typeOfGridNode,
+ para->getParD(level)->neighborX,
+ para->getParD(level)->neighborY,
+ para->getParD(level)->neighborZ,
+ para->getParD(level)->distributions.f[0],
+ para->getParD(level)->numberOfNodes,
+ para->getParD(level)->isEvenTimestep);
+ getLastCudaError("LB_Kernel_MRT_Comp_SP_27 execution failed");
}
MRTCompSP27::MRTCompSP27(std::shared_ptr<Parameter> para, int level)
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/MRT/MRTCompSP27_Device.cu b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/MRT/MRTCompSP27_Device.cu
index a9aefa2d62a962766470c93a62adeefa4f19570e..c3eb51a114e5c4a3be7605765d0889a7bae25cf0 100644
--- a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/MRT/MRTCompSP27_Device.cu
+++ b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/MRT/MRTCompSP27_Device.cu
@@ -37,63 +37,63 @@ __global__ void LB_Kernel_MRT_Comp_SP_27(real omega,
Distributions27 D;
if (EvenOrOdd == true)
{
- D.f[DIR_P00] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_M00] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0P0] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0M0] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00P] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00M] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_PP0] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_MM0] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_PM0] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_MP0] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_P0P] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_M0M] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_P0M] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_M0P] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0PP] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0MM] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0PM] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0MP] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_PPP] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_MMP] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_PMP] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_MPP] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_PPM] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_MMM] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_PMM] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_MPM] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DDStart[DIR_P00 * size_Mat];
+ D.f[DIR_M00] = &DDStart[DIR_M00 * size_Mat];
+ D.f[DIR_0P0] = &DDStart[DIR_0P0 * size_Mat];
+ D.f[DIR_0M0] = &DDStart[DIR_0M0 * size_Mat];
+ D.f[DIR_00P] = &DDStart[DIR_00P * size_Mat];
+ D.f[DIR_00M] = &DDStart[DIR_00M * size_Mat];
+ D.f[DIR_PP0] = &DDStart[DIR_PP0 * size_Mat];
+ D.f[DIR_MM0] = &DDStart[DIR_MM0 * size_Mat];
+ D.f[DIR_PM0] = &DDStart[DIR_PM0 * size_Mat];
+ D.f[DIR_MP0] = &DDStart[DIR_MP0 * size_Mat];
+ D.f[DIR_P0P] = &DDStart[DIR_P0P * size_Mat];
+ D.f[DIR_M0M] = &DDStart[DIR_M0M * size_Mat];
+ D.f[DIR_P0M] = &DDStart[DIR_P0M * size_Mat];
+ D.f[DIR_M0P] = &DDStart[DIR_M0P * size_Mat];
+ D.f[DIR_0PP] = &DDStart[DIR_0PP * size_Mat];
+ D.f[DIR_0MM] = &DDStart[DIR_0MM * size_Mat];
+ D.f[DIR_0PM] = &DDStart[DIR_0PM * size_Mat];
+ D.f[DIR_0MP] = &DDStart[DIR_0MP * size_Mat];
+ D.f[DIR_000] = &DDStart[DIR_000 * size_Mat];
+ D.f[DIR_PPP] = &DDStart[DIR_PPP * size_Mat];
+ D.f[DIR_MMP] = &DDStart[DIR_MMP * size_Mat];
+ D.f[DIR_PMP] = &DDStart[DIR_PMP * size_Mat];
+ D.f[DIR_MPP] = &DDStart[DIR_MPP * size_Mat];
+ D.f[DIR_PPM] = &DDStart[DIR_PPM * size_Mat];
+ D.f[DIR_MMM] = &DDStart[DIR_MMM * size_Mat];
+ D.f[DIR_PMM] = &DDStart[DIR_PMM * size_Mat];
+ D.f[DIR_MPM] = &DDStart[DIR_MPM * size_Mat];
}
else
{
- D.f[DIR_M00] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_P00] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0M0] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0P0] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00M] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00P] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_MM0] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_PP0] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_MP0] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_PM0] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_M0M] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_P0P] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_M0P] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_P0M] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0MM] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0PP] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0MP] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0PM] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_MMM] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_PPM] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_MPM] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_PMM] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_MMP] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_PPP] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_MPP] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_PMP] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_M00] = &DDStart[DIR_P00 * size_Mat];
+ D.f[DIR_P00] = &DDStart[DIR_M00 * size_Mat];
+ D.f[DIR_0M0] = &DDStart[DIR_0P0 * size_Mat];
+ D.f[DIR_0P0] = &DDStart[DIR_0M0 * size_Mat];
+ D.f[DIR_00M] = &DDStart[DIR_00P * size_Mat];
+ D.f[DIR_00P] = &DDStart[DIR_00M * size_Mat];
+ D.f[DIR_MM0] = &DDStart[DIR_PP0 * size_Mat];
+ D.f[DIR_PP0] = &DDStart[DIR_MM0 * size_Mat];
+ D.f[DIR_MP0] = &DDStart[DIR_PM0 * size_Mat];
+ D.f[DIR_PM0] = &DDStart[DIR_MP0 * size_Mat];
+ D.f[DIR_M0M] = &DDStart[DIR_P0P * size_Mat];
+ D.f[DIR_P0P] = &DDStart[DIR_M0M * size_Mat];
+ D.f[DIR_M0P] = &DDStart[DIR_P0M * size_Mat];
+ D.f[DIR_P0M] = &DDStart[DIR_M0P * size_Mat];
+ D.f[DIR_0MM] = &DDStart[DIR_0PP * size_Mat];
+ D.f[DIR_0PP] = &DDStart[DIR_0MM * size_Mat];
+ D.f[DIR_0MP] = &DDStart[DIR_0PM * size_Mat];
+ D.f[DIR_0PM] = &DDStart[DIR_0MP * size_Mat];
+ D.f[DIR_000] = &DDStart[DIR_000 * size_Mat];
+ D.f[DIR_MMM] = &DDStart[DIR_PPP * size_Mat];
+ D.f[DIR_PPM] = &DDStart[DIR_MMP * size_Mat];
+ D.f[DIR_MPM] = &DDStart[DIR_PMP * size_Mat];
+ D.f[DIR_PMM] = &DDStart[DIR_MPP * size_Mat];
+ D.f[DIR_MMP] = &DDStart[DIR_PPM * size_Mat];
+ D.f[DIR_PPP] = &DDStart[DIR_MMM * size_Mat];
+ D.f[DIR_MPP] = &DDStart[DIR_PMM * size_Mat];
+ D.f[DIR_PMP] = &DDStart[DIR_MPM * size_Mat];
}
////////////////////////////////////////////////////////////////////////////////
@@ -126,33 +126,33 @@ __global__ void LB_Kernel_MRT_Comp_SP_27(real omega,
//unsigned int ktne = k;
unsigned int kbsw = neighborZ[ksw];
//////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
- real mfcbb = (D.f[DIR_P00])[k];//[ke ];// + c2over27 ;(D.f[DIR_P00 ])[k ];//ke
- real mfabb = (D.f[DIR_M00])[kw];//[kw ];// + c2over27 ;(D.f[DIR_M00 ])[kw ];
- real mfbcb = (D.f[DIR_0P0])[k];//[kn ];// + c2over27 ;(D.f[DIR_0P0 ])[k ];//kn
- real mfbab = (D.f[DIR_0M0])[ks];//[ks ];// + c2over27 ;(D.f[DIR_0M0 ])[ks ];
- real mfbbc = (D.f[DIR_00P])[k];//[kt ];// + c2over27 ;(D.f[DIR_00P ])[k ];//kt
- real mfbba = (D.f[DIR_00M])[kb];//[kb ];// + c2over27 ;(D.f[DIR_00M ])[kb ];
- real mfccb = (D.f[DIR_PP0])[k];//[kne ];// + c1over54 ;(D.f[DIR_PP0 ])[k ];//kne
- real mfaab = (D.f[DIR_MM0])[ksw];//[ksw ];// + c1over54 ;(D.f[DIR_MM0 ])[ksw];
- real mfcab = (D.f[DIR_PM0])[ks];//[kse ];// + c1over54 ;(D.f[DIR_PM0 ])[ks ];//kse
- real mfacb = (D.f[DIR_MP0])[kw];//[knw ];// + c1over54 ;(D.f[DIR_MP0 ])[kw ];//knw
- real mfcbc = (D.f[DIR_P0P])[k];//[kte ];// + c1over54 ;(D.f[DIR_P0P ])[k ];//kte
- real mfaba = (D.f[DIR_M0M])[kbw];//[kbw ];// + c1over54 ;(D.f[DIR_M0M ])[kbw];
- real mfcba = (D.f[DIR_P0M])[kb];//[kbe ];// + c1over54 ;(D.f[DIR_P0M ])[kb ];//kbe
- real mfabc = (D.f[DIR_M0P])[kw];//[ktw ];// + c1over54 ;(D.f[DIR_M0P ])[kw ];//ktw
- real mfbcc = (D.f[DIR_0PP])[k];//[ktn ];// + c1over54 ;(D.f[DIR_0PP ])[k ];//ktn
- real mfbaa = (D.f[DIR_0MM])[kbs];//[kbs ];// + c1over54 ;(D.f[DIR_0MM ])[kbs];
- real mfbca = (D.f[DIR_0PM])[kb];//[kbn ];// + c1over54 ;(D.f[DIR_0PM ])[kb ];//kbn
- real mfbac = (D.f[DIR_0MP])[ks];//[kts ];// + c1over54 ;(D.f[DIR_0MP ])[ks ];//kts
+ real mfcbb = (D.f[DIR_P00])[k];//[ke ];// + c2over27 ;(D.f[DIR_P00])[k ];//ke
+ real mfabb = (D.f[DIR_M00])[kw];//[kw ];// + c2over27 ;(D.f[DIR_M00])[kw ];
+ real mfbcb = (D.f[DIR_0P0])[k];//[kn ];// + c2over27 ;(D.f[DIR_0P0])[k ];//kn
+ real mfbab = (D.f[DIR_0M0])[ks];//[ks ];// + c2over27 ;(D.f[DIR_0M0])[ks ];
+ real mfbbc = (D.f[DIR_00P])[k];//[kt ];// + c2over27 ;(D.f[DIR_00P])[k ];//kt
+ real mfbba = (D.f[DIR_00M])[kb];//[kb ];// + c2over27 ;(D.f[DIR_00M])[kb ];
+ real mfccb = (D.f[DIR_PP0])[k];//[kne ];// + c1over54 ;(D.f[DIR_PP0])[k ];//kne
+ real mfaab = (D.f[DIR_MM0])[ksw];//[ksw ];// + c1over54 ;(D.f[DIR_MM0])[ksw];
+ real mfcab = (D.f[DIR_PM0])[ks];//[kse ];// + c1over54 ;(D.f[DIR_PM0])[ks ];//kse
+ real mfacb = (D.f[DIR_MP0])[kw];//[knw ];// + c1over54 ;(D.f[DIR_MP0])[kw ];//knw
+ real mfcbc = (D.f[DIR_P0P])[k];//[kte ];// + c1over54 ;(D.f[DIR_P0P])[k ];//kte
+ real mfaba = (D.f[DIR_M0M])[kbw];//[kbw ];// + c1over54 ;(D.f[DIR_M0M])[kbw];
+ real mfcba = (D.f[DIR_P0M])[kb];//[kbe ];// + c1over54 ;(D.f[DIR_P0M])[kb ];//kbe
+ real mfabc = (D.f[DIR_M0P])[kw];//[ktw ];// + c1over54 ;(D.f[DIR_M0P])[kw ];//ktw
+ real mfbcc = (D.f[DIR_0PP])[k];//[ktn ];// + c1over54 ;(D.f[DIR_0PP])[k ];//ktn
+ real mfbaa = (D.f[DIR_0MM])[kbs];//[kbs ];// + c1over54 ;(D.f[DIR_0MM])[kbs];
+ real mfbca = (D.f[DIR_0PM])[kb];//[kbn ];// + c1over54 ;(D.f[DIR_0PM])[kb ];//kbn
+ real mfbac = (D.f[DIR_0MP])[ks];//[kts ];// + c1over54 ;(D.f[DIR_0MP])[ks ];//kts
real mfbbb = (D.f[DIR_000])[k];//[kzero];// + c8over27 ;(D.f[DIR_000])[k ];//kzero
- real mfccc = (D.f[DIR_PPP])[k];//[ktne ];// + c1over216;(D.f[DIR_PPP ])[k ];//ktne
- real mfaac = (D.f[DIR_MMP])[ksw];//[ktsw ];// + c1over216;(D.f[DIR_MMP ])[ksw];//ktsw
- real mfcac = (D.f[DIR_PMP])[ks];//[ktse ];// + c1over216;(D.f[DIR_PMP ])[ks ];//ktse
- real mfacc = (D.f[DIR_MPP])[kw];//[ktnw ];// + c1over216;(D.f[DIR_MPP ])[kw ];//ktnw
- real mfcca = (D.f[DIR_PPM])[kb];//[kbne ];// + c1over216;(D.f[DIR_PPM ])[kb ];//kbne
- real mfaaa = (D.f[DIR_MMM])[kbsw];//[kbsw ];// + c1over216;(D.f[DIR_MMM ])[kbsw];
- real mfcaa = (D.f[DIR_PMM])[kbs];//[kbse ];// + c1over216;(D.f[DIR_PMM ])[kbs];//kbse
- real mfaca = (D.f[DIR_MPM])[kbw];//[kbnw ];// + c1over216;(D.f[DIR_MPM ])[kbw];//kbnw
+ real mfccc = (D.f[DIR_PPP])[k];//[ktne ];// + c1over216;(D.f[DIR_PPP])[k ];//ktne
+ real mfaac = (D.f[DIR_MMP])[ksw];//[ktsw ];// + c1over216;(D.f[DIR_MMP])[ksw];//ktsw
+ real mfcac = (D.f[DIR_PMP])[ks];//[ktse ];// + c1over216;(D.f[DIR_PMP])[ks ];//ktse
+ real mfacc = (D.f[DIR_MPP])[kw];//[ktnw ];// + c1over216;(D.f[DIR_MPP])[kw ];//ktnw
+ real mfcca = (D.f[DIR_PPM])[kb];//[kbne ];// + c1over216;(D.f[DIR_PPM])[kb ];//kbne
+ real mfaaa = (D.f[DIR_MMM])[kbsw];//[kbsw ];// + c1over216;(D.f[DIR_MMM])[kbsw];
+ real mfcaa = (D.f[DIR_PMM])[kbs];//[kbse ];// + c1over216;(D.f[DIR_PMM])[kbs];//kbse
+ real mfaca = (D.f[DIR_MPM])[kbw];//[kbnw ];// + c1over216;(D.f[DIR_MPM])[kbw];//kbnw
////////////////////////////////////////////////////////////////////////////////////
real rho = (mfccc + mfaaa + mfaca + mfcac + mfacc + mfcaa + mfaac + mfcca +
mfbac + mfbca + mfbaa + mfbcc + mfabc + mfcba + mfaba + mfcbc + mfacb + mfcab + mfaab + mfccb +
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/RunLBMKernel.cuh b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/RunLBMKernel.cuh
index 558b4f333e7c92b372a5097aa4917dd6d1230a34..3be594e3e39a57cd71741cd060e9dddda15d6035 100644
--- a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/RunLBMKernel.cuh
+++ b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Compressible/RunLBMKernel.cuh
@@ -5,7 +5,7 @@
#include <DataTypes.h>
#include <cuda_runtime.h>
-#include <lbm/KernelParameter.h>
+#include "lbm/KernelParameter.h"
#include "Kernel/Utilities/DistributionHelper.cuh"
@@ -23,7 +23,7 @@ struct GPUKernelParameter
unsigned int* neighborY;
unsigned int* neighborZ;
real* distributions;
- int size_Mat;
+ int numberOfLBnodes;
real* forces;
bool isEvenTimestep;
};
@@ -31,19 +31,22 @@ struct GPUKernelParameter
template<typename KernelFunctor>
__global__ void runKernel(KernelFunctor kernel, GPUKernelParameter kernelParameter)
{
- const uint k = getNodeIndex();
+ ////////////////////////////////////////////////////////////////////////////////
+ //! - Get node index coordinates from threadIdx, blockIdx, blockDim and gridDim.
+ //!
+ const unsigned nodeIndex = getNodeIndex();
- if(k >= kernelParameter.size_Mat)
+ if(nodeIndex >= kernelParameter.numberOfLBnodes)
return;
- if (!isValidFluidNode(kernelParameter.typeOfGridNode[k]))
+ if (!isValidFluidNode(kernelParameter.typeOfGridNode[nodeIndex]))
return;
DistributionWrapper distributionWrapper {
kernelParameter.distributions,
- (unsigned int)kernelParameter.size_Mat,
+ (unsigned int)kernelParameter.numberOfLBnodes,
kernelParameter.isEvenTimestep,
- k,
+ nodeIndex,
kernelParameter.neighborX,
kernelParameter.neighborY,
kernelParameter.neighborZ
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Incompressible/BGK/BGKIncompSP27.cu b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Incompressible/BGK/BGKIncompSP27.cu
index 81655fac9cfd0b562ba60a5ee289fb64da5c1fba..3fb9be28654f83a7a98bb7d6b3a8a46e9170e7a8 100644
--- a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Incompressible/BGK/BGKIncompSP27.cu
+++ b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Incompressible/BGK/BGKIncompSP27.cu
@@ -2,6 +2,7 @@
#include "BGKIncompSP27_Device.cuh"
#include "Parameter/Parameter.h"
+#include "cuda/CudaGrid.h"
std::shared_ptr<BGKIncompSP27> BGKIncompSP27::getNewInstance(std::shared_ptr<Parameter> para, int level)
{
@@ -10,33 +11,18 @@ std::shared_ptr<BGKIncompSP27> BGKIncompSP27::getNewInstance(std::shared_ptr<Par
void BGKIncompSP27::run()
{
- int size_Mat = para->getParD(level)->numberOfNodes;
- int numberOfThreads = para->getParD(level)->numberofthreads;
-
- int Grid = (size_Mat / numberOfThreads) + 1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid / Grid1) + 1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 grid(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1);
-
- LB_Kernel_BGK_Incomp_SP_27 << < grid, threads >> >( para->getParD(level)->omega,
- para->getParD(level)->typeOfGridNode,
- para->getParD(level)->neighborX,
- para->getParD(level)->neighborY,
- para->getParD(level)->neighborZ,
- para->getParD(level)->distributions.f[0],
- para->getParD(level)->numberOfNodes,
- para->getParD(level)->isEvenTimestep);
- getLastCudaError("LB_Kernel_BGK_SP_27 execution failed");
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(para->getParD(level)->numberofthreads, para->getParD(level)->numberOfNodes);
+
+ LB_Kernel_BGK_Incomp_SP_27 <<< grid.grid, grid.threads >>>(
+ para->getParD(level)->omega,
+ para->getParD(level)->typeOfGridNode,
+ para->getParD(level)->neighborX,
+ para->getParD(level)->neighborY,
+ para->getParD(level)->neighborZ,
+ para->getParD(level)->distributions.f[0],
+ para->getParD(level)->numberOfNodes,
+ para->getParD(level)->isEvenTimestep);
+ getLastCudaError("LB_Kernel_BGK_Incomp_SP_27 execution failed");
}
BGKIncompSP27::BGKIncompSP27(std::shared_ptr<Parameter> para, int level)
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Incompressible/BGK/BGKIncompSP27_Device.cu b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Incompressible/BGK/BGKIncompSP27_Device.cu
index 9a94006b8a1be745fc2bcfdd80e454152347139d..233595656720f5c84cf5be9e555565af0e9c95d0 100644
--- a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Incompressible/BGK/BGKIncompSP27_Device.cu
+++ b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Incompressible/BGK/BGKIncompSP27_Device.cu
@@ -37,63 +37,63 @@ __global__ void LB_Kernel_BGK_Incomp_SP_27(real omega,
Distributions27 D;
if (EvenOrOdd == true)
{
- D.f[DIR_P00] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_M00] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0P0] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0M0] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00P] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00M] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_PP0] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_MM0] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_PM0] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_MP0] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_P0P] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_M0M] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_P0M] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_M0P] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0PP] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0MM] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0PM] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0MP] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_PPP] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_MMP] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_PMP] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_MPP] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_PPM] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_MMM] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_PMM] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_MPM] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DDStart[DIR_P00 * size_Mat];
+ D.f[DIR_M00] = &DDStart[DIR_M00 * size_Mat];
+ D.f[DIR_0P0] = &DDStart[DIR_0P0 * size_Mat];
+ D.f[DIR_0M0] = &DDStart[DIR_0M0 * size_Mat];
+ D.f[DIR_00P] = &DDStart[DIR_00P * size_Mat];
+ D.f[DIR_00M] = &DDStart[DIR_00M * size_Mat];
+ D.f[DIR_PP0] = &DDStart[DIR_PP0 * size_Mat];
+ D.f[DIR_MM0] = &DDStart[DIR_MM0 * size_Mat];
+ D.f[DIR_PM0] = &DDStart[DIR_PM0 * size_Mat];
+ D.f[DIR_MP0] = &DDStart[DIR_MP0 * size_Mat];
+ D.f[DIR_P0P] = &DDStart[DIR_P0P * size_Mat];
+ D.f[DIR_M0M] = &DDStart[DIR_M0M * size_Mat];
+ D.f[DIR_P0M] = &DDStart[DIR_P0M * size_Mat];
+ D.f[DIR_M0P] = &DDStart[DIR_M0P * size_Mat];
+ D.f[DIR_0PP] = &DDStart[DIR_0PP * size_Mat];
+ D.f[DIR_0MM] = &DDStart[DIR_0MM * size_Mat];
+ D.f[DIR_0PM] = &DDStart[DIR_0PM * size_Mat];
+ D.f[DIR_0MP] = &DDStart[DIR_0MP * size_Mat];
+ D.f[DIR_000] = &DDStart[DIR_000 * size_Mat];
+ D.f[DIR_PPP] = &DDStart[DIR_PPP * size_Mat];
+ D.f[DIR_MMP] = &DDStart[DIR_MMP * size_Mat];
+ D.f[DIR_PMP] = &DDStart[DIR_PMP * size_Mat];
+ D.f[DIR_MPP] = &DDStart[DIR_MPP * size_Mat];
+ D.f[DIR_PPM] = &DDStart[DIR_PPM * size_Mat];
+ D.f[DIR_MMM] = &DDStart[DIR_MMM * size_Mat];
+ D.f[DIR_PMM] = &DDStart[DIR_PMM * size_Mat];
+ D.f[DIR_MPM] = &DDStart[DIR_MPM * size_Mat];
}
else
{
- D.f[DIR_M00] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_P00] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0M0] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0P0] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00M] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00P] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_MM0] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_PP0] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_MP0] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_PM0] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_M0M] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_P0P] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_M0P] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_P0M] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0MM] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0PP] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0MP] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0PM] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_MMM] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_PPM] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_MPM] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_PMM] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_MMP] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_PPP] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_MPP] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_PMP] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_M00] = &DDStart[DIR_P00 * size_Mat];
+ D.f[DIR_P00] = &DDStart[DIR_M00 * size_Mat];
+ D.f[DIR_0M0] = &DDStart[DIR_0P0 * size_Mat];
+ D.f[DIR_0P0] = &DDStart[DIR_0M0 * size_Mat];
+ D.f[DIR_00M] = &DDStart[DIR_00P * size_Mat];
+ D.f[DIR_00P] = &DDStart[DIR_00M * size_Mat];
+ D.f[DIR_MM0] = &DDStart[DIR_PP0 * size_Mat];
+ D.f[DIR_PP0] = &DDStart[DIR_MM0 * size_Mat];
+ D.f[DIR_MP0] = &DDStart[DIR_PM0 * size_Mat];
+ D.f[DIR_PM0] = &DDStart[DIR_MP0 * size_Mat];
+ D.f[DIR_M0M] = &DDStart[DIR_P0P * size_Mat];
+ D.f[DIR_P0P] = &DDStart[DIR_M0M * size_Mat];
+ D.f[DIR_M0P] = &DDStart[DIR_P0M * size_Mat];
+ D.f[DIR_P0M] = &DDStart[DIR_M0P * size_Mat];
+ D.f[DIR_0MM] = &DDStart[DIR_0PP * size_Mat];
+ D.f[DIR_0PP] = &DDStart[DIR_0MM * size_Mat];
+ D.f[DIR_0MP] = &DDStart[DIR_0PM * size_Mat];
+ D.f[DIR_0PM] = &DDStart[DIR_0MP * size_Mat];
+ D.f[DIR_000] = &DDStart[DIR_000 * size_Mat];
+ D.f[DIR_MMM] = &DDStart[DIR_PPP * size_Mat];
+ D.f[DIR_PPM] = &DDStart[DIR_MMP * size_Mat];
+ D.f[DIR_MPM] = &DDStart[DIR_PMP * size_Mat];
+ D.f[DIR_PMM] = &DDStart[DIR_MPP * size_Mat];
+ D.f[DIR_MMP] = &DDStart[DIR_PPM * size_Mat];
+ D.f[DIR_PPP] = &DDStart[DIR_MMM * size_Mat];
+ D.f[DIR_MPP] = &DDStart[DIR_PMM * size_Mat];
+ D.f[DIR_PMP] = &DDStart[DIR_MPM * size_Mat];
}
////////////////////////////////////////////////////////////////////////////////
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Incompressible/BGKPlus/BGKPlusIncompSP27.cu b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Incompressible/BGKPlus/BGKPlusIncompSP27.cu
index 86b513f1252f2787abee637819e64606d111c4fa..f274f576a14fc193bcabd44d2c9078a2c98055bc 100644
--- a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Incompressible/BGKPlus/BGKPlusIncompSP27.cu
+++ b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Incompressible/BGKPlus/BGKPlusIncompSP27.cu
@@ -2,6 +2,7 @@
#include "BGKPlusIncompSP27_Device.cuh"
#include "Parameter/Parameter.h"
+#include "cuda/CudaGrid.h"
std::shared_ptr<BGKPlusIncompSP27> BGKPlusIncompSP27::getNewInstance(std::shared_ptr<Parameter> para, int level)
{
@@ -10,33 +11,18 @@ std::shared_ptr<BGKPlusIncompSP27> BGKPlusIncompSP27::getNewInstance(std::shared
void BGKPlusIncompSP27::run()
{
- int size_Mat = para->getParD(level)->numberOfNodes;
- int numberOfThreads = para->getParD(level)->numberofthreads;
-
- int Grid = (size_Mat / numberOfThreads) + 1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid / Grid1) + 1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 grid(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1);
-
- LB_Kernel_BGK_Plus_Incomp_SP_27 << < grid, threads >> >( para->getParD(level)->omega,
- para->getParD(level)->typeOfGridNode,
- para->getParD(level)->neighborX,
- para->getParD(level)->neighborY,
- para->getParD(level)->neighborZ,
- para->getParD(level)->distributions.f[0],
- para->getParD(level)->numberOfNodes,
- para->getParD(level)->isEvenTimestep);
- getLastCudaError("LB_Kernel_BGK_Plus_SP_27 execution failed");
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(para->getParD(level)->numberofthreads, para->getParD(level)->numberOfNodes);
+
+ LB_Kernel_BGK_Plus_Incomp_SP_27 <<< grid.grid, grid.threads >>>(
+ para->getParD(level)->omega,
+ para->getParD(level)->typeOfGridNode,
+ para->getParD(level)->neighborX,
+ para->getParD(level)->neighborY,
+ para->getParD(level)->neighborZ,
+ para->getParD(level)->distributions.f[0],
+ para->getParD(level)->numberOfNodes,
+ para->getParD(level)->isEvenTimestep);
+ getLastCudaError("LB_Kernel_BGK_Plus_Incomp_SP_27 execution failed");
}
BGKPlusIncompSP27::BGKPlusIncompSP27(std::shared_ptr<Parameter> para, int level)
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Incompressible/BGKPlus/BGKPlusIncompSP27_Device.cu b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Incompressible/BGKPlus/BGKPlusIncompSP27_Device.cu
index 9355e42aa5b05190f063f5247d8d6c0dea787a02..b49b76c6224be4b3543c01647a6553e6fc64b74e 100644
--- a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Incompressible/BGKPlus/BGKPlusIncompSP27_Device.cu
+++ b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Incompressible/BGKPlus/BGKPlusIncompSP27_Device.cu
@@ -37,63 +37,63 @@ __global__ void LB_Kernel_BGK_Plus_Incomp_SP_27(real omega,
Distributions27 D;
if (EvenOrOdd == true)
{
- D.f[DIR_P00] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_M00] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0P0] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0M0] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00P] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00M] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_PP0] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_MM0] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_PM0] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_MP0] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_P0P] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_M0M] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_P0M] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_M0P] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0PP] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0MM] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0PM] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0MP] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_PPP] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_MMP] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_PMP] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_MPP] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_PPM] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_MMM] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_PMM] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_MPM] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DDStart[DIR_P00 * size_Mat];
+ D.f[DIR_M00] = &DDStart[DIR_M00 * size_Mat];
+ D.f[DIR_0P0] = &DDStart[DIR_0P0 * size_Mat];
+ D.f[DIR_0M0] = &DDStart[DIR_0M0 * size_Mat];
+ D.f[DIR_00P] = &DDStart[DIR_00P * size_Mat];
+ D.f[DIR_00M] = &DDStart[DIR_00M * size_Mat];
+ D.f[DIR_PP0] = &DDStart[DIR_PP0 * size_Mat];
+ D.f[DIR_MM0] = &DDStart[DIR_MM0 * size_Mat];
+ D.f[DIR_PM0] = &DDStart[DIR_PM0 * size_Mat];
+ D.f[DIR_MP0] = &DDStart[DIR_MP0 * size_Mat];
+ D.f[DIR_P0P] = &DDStart[DIR_P0P * size_Mat];
+ D.f[DIR_M0M] = &DDStart[DIR_M0M * size_Mat];
+ D.f[DIR_P0M] = &DDStart[DIR_P0M * size_Mat];
+ D.f[DIR_M0P] = &DDStart[DIR_M0P * size_Mat];
+ D.f[DIR_0PP] = &DDStart[DIR_0PP * size_Mat];
+ D.f[DIR_0MM] = &DDStart[DIR_0MM * size_Mat];
+ D.f[DIR_0PM] = &DDStart[DIR_0PM * size_Mat];
+ D.f[DIR_0MP] = &DDStart[DIR_0MP * size_Mat];
+ D.f[DIR_000] = &DDStart[DIR_000 * size_Mat];
+ D.f[DIR_PPP] = &DDStart[DIR_PPP * size_Mat];
+ D.f[DIR_MMP] = &DDStart[DIR_MMP * size_Mat];
+ D.f[DIR_PMP] = &DDStart[DIR_PMP * size_Mat];
+ D.f[DIR_MPP] = &DDStart[DIR_MPP * size_Mat];
+ D.f[DIR_PPM] = &DDStart[DIR_PPM * size_Mat];
+ D.f[DIR_MMM] = &DDStart[DIR_MMM * size_Mat];
+ D.f[DIR_PMM] = &DDStart[DIR_PMM * size_Mat];
+ D.f[DIR_MPM] = &DDStart[DIR_MPM * size_Mat];
}
else
{
- D.f[DIR_M00] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_P00] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0M0] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0P0] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00M] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00P] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_MM0] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_PP0] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_MP0] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_PM0] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_M0M] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_P0P] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_M0P] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_P0M] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0MM] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0PP] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0MP] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0PM] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_MMM] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_PPM] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_MPM] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_PMM] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_MMP] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_PPP] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_MPP] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_PMP] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_M00] = &DDStart[DIR_P00 * size_Mat];
+ D.f[DIR_P00] = &DDStart[DIR_M00 * size_Mat];
+ D.f[DIR_0M0] = &DDStart[DIR_0P0 * size_Mat];
+ D.f[DIR_0P0] = &DDStart[DIR_0M0 * size_Mat];
+ D.f[DIR_00M] = &DDStart[DIR_00P * size_Mat];
+ D.f[DIR_00P] = &DDStart[DIR_00M * size_Mat];
+ D.f[DIR_MM0] = &DDStart[DIR_PP0 * size_Mat];
+ D.f[DIR_PP0] = &DDStart[DIR_MM0 * size_Mat];
+ D.f[DIR_MP0] = &DDStart[DIR_PM0 * size_Mat];
+ D.f[DIR_PM0] = &DDStart[DIR_MP0 * size_Mat];
+ D.f[DIR_M0M] = &DDStart[DIR_P0P * size_Mat];
+ D.f[DIR_P0P] = &DDStart[DIR_M0M * size_Mat];
+ D.f[DIR_M0P] = &DDStart[DIR_P0M * size_Mat];
+ D.f[DIR_P0M] = &DDStart[DIR_M0P * size_Mat];
+ D.f[DIR_0MM] = &DDStart[DIR_0PP * size_Mat];
+ D.f[DIR_0PP] = &DDStart[DIR_0MM * size_Mat];
+ D.f[DIR_0MP] = &DDStart[DIR_0PM * size_Mat];
+ D.f[DIR_0PM] = &DDStart[DIR_0MP * size_Mat];
+ D.f[DIR_000] = &DDStart[DIR_000 * size_Mat];
+ D.f[DIR_MMM] = &DDStart[DIR_PPP * size_Mat];
+ D.f[DIR_PPM] = &DDStart[DIR_MMP * size_Mat];
+ D.f[DIR_MPM] = &DDStart[DIR_PMP * size_Mat];
+ D.f[DIR_PMM] = &DDStart[DIR_MPP * size_Mat];
+ D.f[DIR_MMP] = &DDStart[DIR_PPM * size_Mat];
+ D.f[DIR_PPP] = &DDStart[DIR_MMM * size_Mat];
+ D.f[DIR_MPP] = &DDStart[DIR_PMM * size_Mat];
+ D.f[DIR_PMP] = &DDStart[DIR_MPM * size_Mat];
}
////////////////////////////////////////////////////////////////////////////////
@@ -126,33 +126,33 @@ __global__ void LB_Kernel_BGK_Plus_Incomp_SP_27(real omega,
//unsigned int ktne = k;
unsigned int kbsw = neighborZ[ksw];
//////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
- real mfcbb = (D.f[DIR_P00])[k];//[ke ];// + c2over27 ;(D.f[DIR_P00 ])[k ];//ke
- real mfabb = (D.f[DIR_M00])[kw];//[kw ];// + c2over27 ;(D.f[DIR_M00 ])[kw ];
- real mfbcb = (D.f[DIR_0P0])[k];//[kn ];// + c2over27 ;(D.f[DIR_0P0 ])[k ];//kn
- real mfbab = (D.f[DIR_0M0])[ks];//[ks ];// + c2over27 ;(D.f[DIR_0M0 ])[ks ];
- real mfbbc = (D.f[DIR_00P])[k];//[kt ];// + c2over27 ;(D.f[DIR_00P ])[k ];//kt
- real mfbba = (D.f[DIR_00M])[kb];//[kb ];// + c2over27 ;(D.f[DIR_00M ])[kb ];
- real mfccb = (D.f[DIR_PP0])[k];//[kne ];// + c1over54 ;(D.f[DIR_PP0 ])[k ];//kne
- real mfaab = (D.f[DIR_MM0])[ksw];//[ksw ];// + c1over54 ;(D.f[DIR_MM0 ])[ksw];
- real mfcab = (D.f[DIR_PM0])[ks];//[kse ];// + c1over54 ;(D.f[DIR_PM0 ])[ks ];//kse
- real mfacb = (D.f[DIR_MP0])[kw];//[knw ];// + c1over54 ;(D.f[DIR_MP0 ])[kw ];//knw
- real mfcbc = (D.f[DIR_P0P])[k];//[kte ];// + c1over54 ;(D.f[DIR_P0P ])[k ];//kte
- real mfaba = (D.f[DIR_M0M])[kbw];//[kbw ];// + c1over54 ;(D.f[DIR_M0M ])[kbw];
- real mfcba = (D.f[DIR_P0M])[kb];//[kbe ];// + c1over54 ;(D.f[DIR_P0M ])[kb ];//kbe
- real mfabc = (D.f[DIR_M0P])[kw];//[ktw ];// + c1over54 ;(D.f[DIR_M0P ])[kw ];//ktw
- real mfbcc = (D.f[DIR_0PP])[k];//[ktn ];// + c1over54 ;(D.f[DIR_0PP ])[k ];//ktn
- real mfbaa = (D.f[DIR_0MM])[kbs];//[kbs ];// + c1over54 ;(D.f[DIR_0MM ])[kbs];
- real mfbca = (D.f[DIR_0PM])[kb];//[kbn ];// + c1over54 ;(D.f[DIR_0PM ])[kb ];//kbn
- real mfbac = (D.f[DIR_0MP])[ks];//[kts ];// + c1over54 ;(D.f[DIR_0MP ])[ks ];//kts
+ real mfcbb = (D.f[DIR_P00])[k];//[ke ];// + c2over27 ;(D.f[DIR_P00])[k ];//ke
+ real mfabb = (D.f[DIR_M00])[kw];//[kw ];// + c2over27 ;(D.f[DIR_M00])[kw ];
+ real mfbcb = (D.f[DIR_0P0])[k];//[kn ];// + c2over27 ;(D.f[DIR_0P0])[k ];//kn
+ real mfbab = (D.f[DIR_0M0])[ks];//[ks ];// + c2over27 ;(D.f[DIR_0M0])[ks ];
+ real mfbbc = (D.f[DIR_00P])[k];//[kt ];// + c2over27 ;(D.f[DIR_00P])[k ];//kt
+ real mfbba = (D.f[DIR_00M])[kb];//[kb ];// + c2over27 ;(D.f[DIR_00M])[kb ];
+ real mfccb = (D.f[DIR_PP0])[k];//[kne ];// + c1over54 ;(D.f[DIR_PP0])[k ];//kne
+ real mfaab = (D.f[DIR_MM0])[ksw];//[ksw ];// + c1over54 ;(D.f[DIR_MM0])[ksw];
+ real mfcab = (D.f[DIR_PM0])[ks];//[kse ];// + c1over54 ;(D.f[DIR_PM0])[ks ];//kse
+ real mfacb = (D.f[DIR_MP0])[kw];//[knw ];// + c1over54 ;(D.f[DIR_MP0])[kw ];//knw
+ real mfcbc = (D.f[DIR_P0P])[k];//[kte ];// + c1over54 ;(D.f[DIR_P0P])[k ];//kte
+ real mfaba = (D.f[DIR_M0M])[kbw];//[kbw ];// + c1over54 ;(D.f[DIR_M0M])[kbw];
+ real mfcba = (D.f[DIR_P0M])[kb];//[kbe ];// + c1over54 ;(D.f[DIR_P0M])[kb ];//kbe
+ real mfabc = (D.f[DIR_M0P])[kw];//[ktw ];// + c1over54 ;(D.f[DIR_M0P])[kw ];//ktw
+ real mfbcc = (D.f[DIR_0PP])[k];//[ktn ];// + c1over54 ;(D.f[DIR_0PP])[k ];//ktn
+ real mfbaa = (D.f[DIR_0MM])[kbs];//[kbs ];// + c1over54 ;(D.f[DIR_0MM])[kbs];
+ real mfbca = (D.f[DIR_0PM])[kb];//[kbn ];// + c1over54 ;(D.f[DIR_0PM])[kb ];//kbn
+ real mfbac = (D.f[DIR_0MP])[ks];//[kts ];// + c1over54 ;(D.f[DIR_0MP])[ks ];//kts
real mfbbb = (D.f[DIR_000])[k];//[kzero];// + c8over27 ;(D.f[DIR_000])[k ];//kzero
- real mfccc = (D.f[DIR_PPP])[k];//[ktne ];// + c1over216;(D.f[DIR_PPP ])[k ];//ktne
- real mfaac = (D.f[DIR_MMP])[ksw];//[ktsw ];// + c1over216;(D.f[DIR_MMP ])[ksw];//ktsw
- real mfcac = (D.f[DIR_PMP])[ks];//[ktse ];// + c1over216;(D.f[DIR_PMP ])[ks ];//ktse
- real mfacc = (D.f[DIR_MPP])[kw];//[ktnw ];// + c1over216;(D.f[DIR_MPP ])[kw ];//ktnw
- real mfcca = (D.f[DIR_PPM])[kb];//[kbne ];// + c1over216;(D.f[DIR_PPM ])[kb ];//kbne
- real mfaaa = (D.f[DIR_MMM])[kbsw];//[kbsw ];// + c1over216;(D.f[DIR_MMM ])[kbsw];
- real mfcaa = (D.f[DIR_PMM])[kbs];//[kbse ];// + c1over216;(D.f[DIR_PMM ])[kbs];//kbse
- real mfaca = (D.f[DIR_MPM])[kbw];//[kbnw ];// + c1over216;(D.f[DIR_MPM ])[kbw];//kbnw
+ real mfccc = (D.f[DIR_PPP])[k];//[ktne ];// + c1over216;(D.f[DIR_PPP])[k ];//ktne
+ real mfaac = (D.f[DIR_MMP])[ksw];//[ktsw ];// + c1over216;(D.f[DIR_MMP])[ksw];//ktsw
+ real mfcac = (D.f[DIR_PMP])[ks];//[ktse ];// + c1over216;(D.f[DIR_PMP])[ks ];//ktse
+ real mfacc = (D.f[DIR_MPP])[kw];//[ktnw ];// + c1over216;(D.f[DIR_MPP])[kw ];//ktnw
+ real mfcca = (D.f[DIR_PPM])[kb];//[kbne ];// + c1over216;(D.f[DIR_PPM])[kb ];//kbne
+ real mfaaa = (D.f[DIR_MMM])[kbsw];//[kbsw ];// + c1over216;(D.f[DIR_MMM])[kbsw];
+ real mfcaa = (D.f[DIR_PMM])[kbs];//[kbse ];// + c1over216;(D.f[DIR_PMM])[kbs];//kbse
+ real mfaca = (D.f[DIR_MPM])[kbw];//[kbnw ];// + c1over216;(D.f[DIR_MPM])[kbw];//kbnw
////////////////////////////////////////////////////////////////////////////////////
//slow
//real oMdrho = one - ((((mfccc+mfaaa) + (mfaca+mfcac)) + ((mfacc+mfcaa) + (mfaac+mfcca))) +
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Incompressible/Cascade/CascadeIncompSP27.cu b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Incompressible/Cascade/CascadeIncompSP27.cu
index 05f374096c9c5da2460b32cf5ae8cb59cfa78382..3a6760b619d2ca1a7eb19771478eb9e5989ead0c 100644
--- a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Incompressible/Cascade/CascadeIncompSP27.cu
+++ b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Incompressible/Cascade/CascadeIncompSP27.cu
@@ -2,6 +2,7 @@
#include "CascadeIncompSP27_Device.cuh"
#include "Parameter/Parameter.h"
+#include "cuda/CudaGrid.h"
std::shared_ptr<CascadeIncompSP27> CascadeIncompSP27::getNewInstance(std::shared_ptr<Parameter> para, int level)
{
@@ -10,33 +11,18 @@ std::shared_ptr<CascadeIncompSP27> CascadeIncompSP27::getNewInstance(std::shared
void CascadeIncompSP27::run()
{
- int size_Mat = para->getParD(level)->numberOfNodes;
- int numberOfThreads = para->getParD(level)->numberofthreads;
-
- int Grid = (size_Mat / numberOfThreads) + 1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid / Grid1) + 1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 grid(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1);
-
- LB_Kernel_Cascade_Incomp_SP_27 << < grid, threads >> >( para->getParD(level)->omega,
- para->getParD(level)->typeOfGridNode,
- para->getParD(level)->neighborX,
- para->getParD(level)->neighborY,
- para->getParD(level)->neighborZ,
- para->getParD(level)->distributions.f[0],
- para->getParD(level)->numberOfNodes,
- para->getParD(level)->isEvenTimestep);
- getLastCudaError("LB_Kernel_Cascade_SP_27 execution failed");
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(para->getParD(level)->numberofthreads, para->getParD(level)->numberOfNodes);
+
+ LB_Kernel_Cascade_Incomp_SP_27 <<< grid.grid, grid.threads >>>(
+ para->getParD(level)->omega,
+ para->getParD(level)->typeOfGridNode,
+ para->getParD(level)->neighborX,
+ para->getParD(level)->neighborY,
+ para->getParD(level)->neighborZ,
+ para->getParD(level)->distributions.f[0],
+ para->getParD(level)->numberOfNodes,
+ para->getParD(level)->isEvenTimestep);
+ getLastCudaError("LB_Kernel_Cascade_Incomp_SP_27 execution failed");
}
CascadeIncompSP27::CascadeIncompSP27(std::shared_ptr<Parameter> para, int level)
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Incompressible/Cascade/CascadeIncompSP27_Device.cu b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Incompressible/Cascade/CascadeIncompSP27_Device.cu
index 92cc749b135739d5f38c9916c4ee0da7497e5f2d..8e607cabb4cc40bbb22c5ad3ec6db2c63154add6 100644
--- a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Incompressible/Cascade/CascadeIncompSP27_Device.cu
+++ b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Incompressible/Cascade/CascadeIncompSP27_Device.cu
@@ -37,63 +37,63 @@ __global__ void LB_Kernel_Cascade_Incomp_SP_27(real omega,
Distributions27 D;
if (EvenOrOdd == true)
{
- D.f[DIR_P00] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_M00] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0P0] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0M0] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00P] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00M] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_PP0] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_MM0] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_PM0] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_MP0] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_P0P] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_M0M] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_P0M] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_M0P] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0PP] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0MM] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0PM] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0MP] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_PPP] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_MMP] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_PMP] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_MPP] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_PPM] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_MMM] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_PMM] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_MPM] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DDStart[DIR_P00 * size_Mat];
+ D.f[DIR_M00] = &DDStart[DIR_M00 * size_Mat];
+ D.f[DIR_0P0] = &DDStart[DIR_0P0 * size_Mat];
+ D.f[DIR_0M0] = &DDStart[DIR_0M0 * size_Mat];
+ D.f[DIR_00P] = &DDStart[DIR_00P * size_Mat];
+ D.f[DIR_00M] = &DDStart[DIR_00M * size_Mat];
+ D.f[DIR_PP0] = &DDStart[DIR_PP0 * size_Mat];
+ D.f[DIR_MM0] = &DDStart[DIR_MM0 * size_Mat];
+ D.f[DIR_PM0] = &DDStart[DIR_PM0 * size_Mat];
+ D.f[DIR_MP0] = &DDStart[DIR_MP0 * size_Mat];
+ D.f[DIR_P0P] = &DDStart[DIR_P0P * size_Mat];
+ D.f[DIR_M0M] = &DDStart[DIR_M0M * size_Mat];
+ D.f[DIR_P0M] = &DDStart[DIR_P0M * size_Mat];
+ D.f[DIR_M0P] = &DDStart[DIR_M0P * size_Mat];
+ D.f[DIR_0PP] = &DDStart[DIR_0PP * size_Mat];
+ D.f[DIR_0MM] = &DDStart[DIR_0MM * size_Mat];
+ D.f[DIR_0PM] = &DDStart[DIR_0PM * size_Mat];
+ D.f[DIR_0MP] = &DDStart[DIR_0MP * size_Mat];
+ D.f[DIR_000] = &DDStart[DIR_000 * size_Mat];
+ D.f[DIR_PPP] = &DDStart[DIR_PPP * size_Mat];
+ D.f[DIR_MMP] = &DDStart[DIR_MMP * size_Mat];
+ D.f[DIR_PMP] = &DDStart[DIR_PMP * size_Mat];
+ D.f[DIR_MPP] = &DDStart[DIR_MPP * size_Mat];
+ D.f[DIR_PPM] = &DDStart[DIR_PPM * size_Mat];
+ D.f[DIR_MMM] = &DDStart[DIR_MMM * size_Mat];
+ D.f[DIR_PMM] = &DDStart[DIR_PMM * size_Mat];
+ D.f[DIR_MPM] = &DDStart[DIR_MPM * size_Mat];
}
else
{
- D.f[DIR_M00] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_P00] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0M0] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0P0] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00M] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00P] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_MM0] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_PP0] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_MP0] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_PM0] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_M0M] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_P0P] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_M0P] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_P0M] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0MM] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0PP] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0MP] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0PM] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_MMM] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_PPM] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_MPM] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_PMM] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_MMP] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_PPP] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_MPP] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_PMP] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_M00] = &DDStart[DIR_P00 * size_Mat];
+ D.f[DIR_P00] = &DDStart[DIR_M00 * size_Mat];
+ D.f[DIR_0M0] = &DDStart[DIR_0P0 * size_Mat];
+ D.f[DIR_0P0] = &DDStart[DIR_0M0 * size_Mat];
+ D.f[DIR_00M] = &DDStart[DIR_00P * size_Mat];
+ D.f[DIR_00P] = &DDStart[DIR_00M * size_Mat];
+ D.f[DIR_MM0] = &DDStart[DIR_PP0 * size_Mat];
+ D.f[DIR_PP0] = &DDStart[DIR_MM0 * size_Mat];
+ D.f[DIR_MP0] = &DDStart[DIR_PM0 * size_Mat];
+ D.f[DIR_PM0] = &DDStart[DIR_MP0 * size_Mat];
+ D.f[DIR_M0M] = &DDStart[DIR_P0P * size_Mat];
+ D.f[DIR_P0P] = &DDStart[DIR_M0M * size_Mat];
+ D.f[DIR_M0P] = &DDStart[DIR_P0M * size_Mat];
+ D.f[DIR_P0M] = &DDStart[DIR_M0P * size_Mat];
+ D.f[DIR_0MM] = &DDStart[DIR_0PP * size_Mat];
+ D.f[DIR_0PP] = &DDStart[DIR_0MM * size_Mat];
+ D.f[DIR_0MP] = &DDStart[DIR_0PM * size_Mat];
+ D.f[DIR_0PM] = &DDStart[DIR_0MP * size_Mat];
+ D.f[DIR_000] = &DDStart[DIR_000 * size_Mat];
+ D.f[DIR_MMM] = &DDStart[DIR_PPP * size_Mat];
+ D.f[DIR_PPM] = &DDStart[DIR_MMP * size_Mat];
+ D.f[DIR_MPM] = &DDStart[DIR_PMP * size_Mat];
+ D.f[DIR_PMM] = &DDStart[DIR_MPP * size_Mat];
+ D.f[DIR_MMP] = &DDStart[DIR_PPM * size_Mat];
+ D.f[DIR_PPP] = &DDStart[DIR_MMM * size_Mat];
+ D.f[DIR_MPP] = &DDStart[DIR_PMM * size_Mat];
+ D.f[DIR_PMP] = &DDStart[DIR_MPM * size_Mat];
}
////////////////////////////////////////////////////////////////////////////////
@@ -126,33 +126,33 @@ __global__ void LB_Kernel_Cascade_Incomp_SP_27(real omega,
//unsigned int ktne = k;
unsigned int kbsw = neighborZ[ksw];
//////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
- real mfcbb = (D.f[DIR_P00])[k];//[ke ];// + c2over27 ;(D.f[DIR_P00 ])[k ];//ke
- real mfabb = (D.f[DIR_M00])[kw];//[kw ];// + c2over27 ;(D.f[DIR_M00 ])[kw ];
- real mfbcb = (D.f[DIR_0P0])[k];//[kn ];// + c2over27 ;(D.f[DIR_0P0 ])[k ];//kn
- real mfbab = (D.f[DIR_0M0])[ks];//[ks ];// + c2over27 ;(D.f[DIR_0M0 ])[ks ];
- real mfbbc = (D.f[DIR_00P])[k];//[kt ];// + c2over27 ;(D.f[DIR_00P ])[k ];//kt
- real mfbba = (D.f[DIR_00M])[kb];//[kb ];// + c2over27 ;(D.f[DIR_00M ])[kb ];
- real mfccb = (D.f[DIR_PP0])[k];//[kne ];// + c1over54 ;(D.f[DIR_PP0 ])[k ];//kne
- real mfaab = (D.f[DIR_MM0])[ksw];//[ksw ];// + c1over54 ;(D.f[DIR_MM0 ])[ksw];
- real mfcab = (D.f[DIR_PM0])[ks];//[kse ];// + c1over54 ;(D.f[DIR_PM0 ])[ks ];//kse
- real mfacb = (D.f[DIR_MP0])[kw];//[knw ];// + c1over54 ;(D.f[DIR_MP0 ])[kw ];//knw
- real mfcbc = (D.f[DIR_P0P])[k];//[kte ];// + c1over54 ;(D.f[DIR_P0P ])[k ];//kte
- real mfaba = (D.f[DIR_M0M])[kbw];//[kbw ];// + c1over54 ;(D.f[DIR_M0M ])[kbw];
- real mfcba = (D.f[DIR_P0M])[kb];//[kbe ];// + c1over54 ;(D.f[DIR_P0M ])[kb ];//kbe
- real mfabc = (D.f[DIR_M0P])[kw];//[ktw ];// + c1over54 ;(D.f[DIR_M0P ])[kw ];//ktw
- real mfbcc = (D.f[DIR_0PP])[k];//[ktn ];// + c1over54 ;(D.f[DIR_0PP ])[k ];//ktn
- real mfbaa = (D.f[DIR_0MM])[kbs];//[kbs ];// + c1over54 ;(D.f[DIR_0MM ])[kbs];
- real mfbca = (D.f[DIR_0PM])[kb];//[kbn ];// + c1over54 ;(D.f[DIR_0PM ])[kb ];//kbn
- real mfbac = (D.f[DIR_0MP])[ks];//[kts ];// + c1over54 ;(D.f[DIR_0MP ])[ks ];//kts
+ real mfcbb = (D.f[DIR_P00])[k];//[ke ];// + c2over27 ;(D.f[DIR_P00])[k ];//ke
+ real mfabb = (D.f[DIR_M00])[kw];//[kw ];// + c2over27 ;(D.f[DIR_M00])[kw ];
+ real mfbcb = (D.f[DIR_0P0])[k];//[kn ];// + c2over27 ;(D.f[DIR_0P0])[k ];//kn
+ real mfbab = (D.f[DIR_0M0])[ks];//[ks ];// + c2over27 ;(D.f[DIR_0M0])[ks ];
+ real mfbbc = (D.f[DIR_00P])[k];//[kt ];// + c2over27 ;(D.f[DIR_00P])[k ];//kt
+ real mfbba = (D.f[DIR_00M])[kb];//[kb ];// + c2over27 ;(D.f[DIR_00M])[kb ];
+ real mfccb = (D.f[DIR_PP0])[k];//[kne ];// + c1over54 ;(D.f[DIR_PP0])[k ];//kne
+ real mfaab = (D.f[DIR_MM0])[ksw];//[ksw ];// + c1over54 ;(D.f[DIR_MM0])[ksw];
+ real mfcab = (D.f[DIR_PM0])[ks];//[kse ];// + c1over54 ;(D.f[DIR_PM0])[ks ];//kse
+ real mfacb = (D.f[DIR_MP0])[kw];//[knw ];// + c1over54 ;(D.f[DIR_MP0])[kw ];//knw
+ real mfcbc = (D.f[DIR_P0P])[k];//[kte ];// + c1over54 ;(D.f[DIR_P0P])[k ];//kte
+ real mfaba = (D.f[DIR_M0M])[kbw];//[kbw ];// + c1over54 ;(D.f[DIR_M0M])[kbw];
+ real mfcba = (D.f[DIR_P0M])[kb];//[kbe ];// + c1over54 ;(D.f[DIR_P0M])[kb ];//kbe
+ real mfabc = (D.f[DIR_M0P])[kw];//[ktw ];// + c1over54 ;(D.f[DIR_M0P])[kw ];//ktw
+ real mfbcc = (D.f[DIR_0PP])[k];//[ktn ];// + c1over54 ;(D.f[DIR_0PP])[k ];//ktn
+ real mfbaa = (D.f[DIR_0MM])[kbs];//[kbs ];// + c1over54 ;(D.f[DIR_0MM])[kbs];
+ real mfbca = (D.f[DIR_0PM])[kb];//[kbn ];// + c1over54 ;(D.f[DIR_0PM])[kb ];//kbn
+ real mfbac = (D.f[DIR_0MP])[ks];//[kts ];// + c1over54 ;(D.f[DIR_0MP])[ks ];//kts
real mfbbb = (D.f[DIR_000])[k];//[kzero];// + c8over27 ;(D.f[DIR_000])[k ];//kzero
- real mfccc = (D.f[DIR_PPP])[k];//[ktne ];// + c1over216;(D.f[DIR_PPP ])[k ];//ktne
- real mfaac = (D.f[DIR_MMP])[ksw];//[ktsw ];// + c1over216;(D.f[DIR_MMP ])[ksw];//ktsw
- real mfcac = (D.f[DIR_PMP])[ks];//[ktse ];// + c1over216;(D.f[DIR_PMP ])[ks ];//ktse
- real mfacc = (D.f[DIR_MPP])[kw];//[ktnw ];// + c1over216;(D.f[DIR_MPP ])[kw ];//ktnw
- real mfcca = (D.f[DIR_PPM])[kb];//[kbne ];// + c1over216;(D.f[DIR_PPM ])[kb ];//kbne
- real mfaaa = (D.f[DIR_MMM])[kbsw];//[kbsw ];// + c1over216;(D.f[DIR_MMM ])[kbsw];
- real mfcaa = (D.f[DIR_PMM])[kbs];//[kbse ];// + c1over216;(D.f[DIR_PMM ])[kbs];//kbse
- real mfaca = (D.f[DIR_MPM])[kbw];//[kbnw ];// + c1over216;(D.f[DIR_MPM ])[kbw];//kbnw
+ real mfccc = (D.f[DIR_PPP])[k];//[ktne ];// + c1over216;(D.f[DIR_PPP])[k ];//ktne
+ real mfaac = (D.f[DIR_MMP])[ksw];//[ktsw ];// + c1over216;(D.f[DIR_MMP])[ksw];//ktsw
+ real mfcac = (D.f[DIR_PMP])[ks];//[ktse ];// + c1over216;(D.f[DIR_PMP])[ks ];//ktse
+ real mfacc = (D.f[DIR_MPP])[kw];//[ktnw ];// + c1over216;(D.f[DIR_MPP])[kw ];//ktnw
+ real mfcca = (D.f[DIR_PPM])[kb];//[kbne ];// + c1over216;(D.f[DIR_PPM])[kb ];//kbne
+ real mfaaa = (D.f[DIR_MMM])[kbsw];//[kbsw ];// + c1over216;(D.f[DIR_MMM])[kbsw];
+ real mfcaa = (D.f[DIR_PMM])[kbs];//[kbse ];// + c1over216;(D.f[DIR_PMM])[kbs];//kbse
+ real mfaca = (D.f[DIR_MPM])[kbw];//[kbnw ];// + c1over216;(D.f[DIR_MPM])[kbw];//kbnw
////////////////////////////////////////////////////////////////////////////////////
//slow
//real oMdrho = one - ((((mfccc+mfaaa) + (mfaca+mfcac)) + ((mfacc+mfcaa) + (mfaac+mfcca))) +
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Incompressible/Cumulant1hSP27/Cumulant1hIncompSP27.cu b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Incompressible/Cumulant1hSP27/Cumulant1hIncompSP27.cu
index 62768ef9948b6c259c5ad4005237081f4d255e73..44beb8507d5664f01283130dd3087a788e4491ed 100644
--- a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Incompressible/Cumulant1hSP27/Cumulant1hIncompSP27.cu
+++ b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Incompressible/Cumulant1hSP27/Cumulant1hIncompSP27.cu
@@ -2,6 +2,7 @@
#include "Cumulant1hIncompSP27_Device.cuh"
#include "Parameter/Parameter.h"
+#include "cuda/CudaGrid.h"
std::shared_ptr<Cumulant1hIncompSP27> Cumulant1hIncompSP27::getNewInstance(std::shared_ptr<Parameter> para, int level)
{
@@ -10,38 +11,23 @@ std::shared_ptr<Cumulant1hIncompSP27> Cumulant1hIncompSP27::getNewInstance(std::
void Cumulant1hIncompSP27::run()
{
- int size_Mat = para->getParD(level)->numberOfNodes;
- int numberOfThreads = para->getParD(level)->numberofthreads;
-
- int Grid = (size_Mat / numberOfThreads) + 1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid / Grid1) + 1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 grid(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1);
-
- LB_Kernel_Cum_1h_Incomp_SP_27 << < grid, threads >> >( para->getParD(level)->omega,
- para->getParD(level)->deltaPhi,
- para->getAngularVelocity(),
- para->getParD(level)->typeOfGridNode,
- para->getParD(level)->neighborX,
- para->getParD(level)->neighborY,
- para->getParD(level)->neighborZ,
- para->getParD(level)->coordinateX,
- para->getParD(level)->coordinateY,
- para->getParD(level)->coordinateZ,
- para->getParD(level)->distributions.f[0],
- para->getParD(level)->numberOfNodes,
- para->getParD(level)->isEvenTimestep);
- getLastCudaError("LB_Kernel_Cum_1h_SP_27 execution failed");
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(para->getParD(level)->numberofthreads, para->getParD(level)->numberOfNodes);
+
+ LB_Kernel_Cum_1h_Incomp_SP_27 <<< grid.grid, grid.threads >>>(
+ para->getParD(level)->omega,
+ para->getParD(level)->deltaPhi,
+ para->getAngularVelocity(),
+ para->getParD(level)->typeOfGridNode,
+ para->getParD(level)->neighborX,
+ para->getParD(level)->neighborY,
+ para->getParD(level)->neighborZ,
+ para->getParD(level)->coordinateX,
+ para->getParD(level)->coordinateY,
+ para->getParD(level)->coordinateZ,
+ para->getParD(level)->distributions.f[0],
+ para->getParD(level)->numberOfNodes,
+ para->getParD(level)->isEvenTimestep);
+ getLastCudaError("LB_Kernel_Cum_1h_Incomp_SP_27 execution failed");
}
Cumulant1hIncompSP27::Cumulant1hIncompSP27(std::shared_ptr<Parameter> para, int level)
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Incompressible/Cumulant1hSP27/Cumulant1hIncompSP27_Device.cu b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Incompressible/Cumulant1hSP27/Cumulant1hIncompSP27_Device.cu
index 0243046082ce1853011c6632d5a2f80364ebe0db..5130017acc642c92b064a500e79ff685ec2f6d97 100644
--- a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Incompressible/Cumulant1hSP27/Cumulant1hIncompSP27_Device.cu
+++ b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Incompressible/Cumulant1hSP27/Cumulant1hIncompSP27_Device.cu
@@ -42,63 +42,63 @@ __global__ void LB_Kernel_Cum_1h_Incomp_SP_27(real omega,
Distributions27 D;
if (EvenOrOdd == true)
{
- D.f[DIR_P00] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_M00] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0P0] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0M0] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00P] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00M] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_PP0] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_MM0] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_PM0] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_MP0] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_P0P] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_M0M] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_P0M] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_M0P] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0PP] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0MM] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0PM] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0MP] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_PPP] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_MMP] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_PMP] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_MPP] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_PPM] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_MMM] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_PMM] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_MPM] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DDStart[DIR_P00 * size_Mat];
+ D.f[DIR_M00] = &DDStart[DIR_M00 * size_Mat];
+ D.f[DIR_0P0] = &DDStart[DIR_0P0 * size_Mat];
+ D.f[DIR_0M0] = &DDStart[DIR_0M0 * size_Mat];
+ D.f[DIR_00P] = &DDStart[DIR_00P * size_Mat];
+ D.f[DIR_00M] = &DDStart[DIR_00M * size_Mat];
+ D.f[DIR_PP0] = &DDStart[DIR_PP0 * size_Mat];
+ D.f[DIR_MM0] = &DDStart[DIR_MM0 * size_Mat];
+ D.f[DIR_PM0] = &DDStart[DIR_PM0 * size_Mat];
+ D.f[DIR_MP0] = &DDStart[DIR_MP0 * size_Mat];
+ D.f[DIR_P0P] = &DDStart[DIR_P0P * size_Mat];
+ D.f[DIR_M0M] = &DDStart[DIR_M0M * size_Mat];
+ D.f[DIR_P0M] = &DDStart[DIR_P0M * size_Mat];
+ D.f[DIR_M0P] = &DDStart[DIR_M0P * size_Mat];
+ D.f[DIR_0PP] = &DDStart[DIR_0PP * size_Mat];
+ D.f[DIR_0MM] = &DDStart[DIR_0MM * size_Mat];
+ D.f[DIR_0PM] = &DDStart[DIR_0PM * size_Mat];
+ D.f[DIR_0MP] = &DDStart[DIR_0MP * size_Mat];
+ D.f[DIR_000] = &DDStart[DIR_000 * size_Mat];
+ D.f[DIR_PPP] = &DDStart[DIR_PPP * size_Mat];
+ D.f[DIR_MMP] = &DDStart[DIR_MMP * size_Mat];
+ D.f[DIR_PMP] = &DDStart[DIR_PMP * size_Mat];
+ D.f[DIR_MPP] = &DDStart[DIR_MPP * size_Mat];
+ D.f[DIR_PPM] = &DDStart[DIR_PPM * size_Mat];
+ D.f[DIR_MMM] = &DDStart[DIR_MMM * size_Mat];
+ D.f[DIR_PMM] = &DDStart[DIR_PMM * size_Mat];
+ D.f[DIR_MPM] = &DDStart[DIR_MPM * size_Mat];
}
else
{
- D.f[DIR_M00] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_P00] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0M0] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0P0] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00M] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00P] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_MM0] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_PP0] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_MP0] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_PM0] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_M0M] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_P0P] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_M0P] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_P0M] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0MM] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0PP] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0MP] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0PM] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_MMM] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_PPM] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_MPM] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_PMM] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_MMP] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_PPP] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_MPP] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_PMP] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_M00] = &DDStart[DIR_P00 * size_Mat];
+ D.f[DIR_P00] = &DDStart[DIR_M00 * size_Mat];
+ D.f[DIR_0M0] = &DDStart[DIR_0P0 * size_Mat];
+ D.f[DIR_0P0] = &DDStart[DIR_0M0 * size_Mat];
+ D.f[DIR_00M] = &DDStart[DIR_00P * size_Mat];
+ D.f[DIR_00P] = &DDStart[DIR_00M * size_Mat];
+ D.f[DIR_MM0] = &DDStart[DIR_PP0 * size_Mat];
+ D.f[DIR_PP0] = &DDStart[DIR_MM0 * size_Mat];
+ D.f[DIR_MP0] = &DDStart[DIR_PM0 * size_Mat];
+ D.f[DIR_PM0] = &DDStart[DIR_MP0 * size_Mat];
+ D.f[DIR_M0M] = &DDStart[DIR_P0P * size_Mat];
+ D.f[DIR_P0P] = &DDStart[DIR_M0M * size_Mat];
+ D.f[DIR_M0P] = &DDStart[DIR_P0M * size_Mat];
+ D.f[DIR_P0M] = &DDStart[DIR_M0P * size_Mat];
+ D.f[DIR_0MM] = &DDStart[DIR_0PP * size_Mat];
+ D.f[DIR_0PP] = &DDStart[DIR_0MM * size_Mat];
+ D.f[DIR_0MP] = &DDStart[DIR_0PM * size_Mat];
+ D.f[DIR_0PM] = &DDStart[DIR_0MP * size_Mat];
+ D.f[DIR_000] = &DDStart[DIR_000 * size_Mat];
+ D.f[DIR_MMM] = &DDStart[DIR_PPP * size_Mat];
+ D.f[DIR_PPM] = &DDStart[DIR_MMP * size_Mat];
+ D.f[DIR_MPM] = &DDStart[DIR_PMP * size_Mat];
+ D.f[DIR_PMM] = &DDStart[DIR_MPP * size_Mat];
+ D.f[DIR_MMP] = &DDStart[DIR_PPM * size_Mat];
+ D.f[DIR_PPP] = &DDStart[DIR_MMM * size_Mat];
+ D.f[DIR_MPP] = &DDStart[DIR_PMM * size_Mat];
+ D.f[DIR_PMP] = &DDStart[DIR_MPM * size_Mat];
}
////////////////////////////////////////////////////////////////////////////////
@@ -159,33 +159,33 @@ __global__ void LB_Kernel_Cum_1h_Incomp_SP_27(real omega,
//unsigned int ktne = k;
//unsigned int kbsw = neighborZ[ksw];
//////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
- real mfcbb = (D.f[DIR_P00])[k];//[ke ];// + c2over27 ;(D.f[DIR_P00 ])[k ];//ke
- real mfabb = (D.f[DIR_M00])[kw];//[kw ];// + c2over27 ;(D.f[DIR_M00 ])[kw ];
- real mfbcb = (D.f[DIR_0P0])[k];//[kn ];// + c2over27 ;(D.f[DIR_0P0 ])[k ];//kn
- real mfbab = (D.f[DIR_0M0])[ks];//[ks ];// + c2over27 ;(D.f[DIR_0M0 ])[ks ];
- real mfbbc = (D.f[DIR_00P])[k];//[kt ];// + c2over27 ;(D.f[DIR_00P ])[k ];//kt
- real mfbba = (D.f[DIR_00M])[kb];//[kb ];// + c2over27 ;(D.f[DIR_00M ])[kb ];
- real mfccb = (D.f[DIR_PP0])[k];//[kne ];// + c1over54 ;(D.f[DIR_PP0 ])[k ];//kne
- real mfaab = (D.f[DIR_MM0])[ksw];//[ksw ];// + c1over54 ;(D.f[DIR_MM0 ])[ksw];
- real mfcab = (D.f[DIR_PM0])[ks];//[kse ];// + c1over54 ;(D.f[DIR_PM0 ])[ks ];//kse
- real mfacb = (D.f[DIR_MP0])[kw];//[knw ];// + c1over54 ;(D.f[DIR_MP0 ])[kw ];//knw
- real mfcbc = (D.f[DIR_P0P])[k];//[kte ];// + c1over54 ;(D.f[DIR_P0P ])[k ];//kte
- real mfaba = (D.f[DIR_M0M])[kbw];//[kbw ];// + c1over54 ;(D.f[DIR_M0M ])[kbw];
- real mfcba = (D.f[DIR_P0M])[kb];//[kbe ];// + c1over54 ;(D.f[DIR_P0M ])[kb ];//kbe
- real mfabc = (D.f[DIR_M0P])[kw];//[ktw ];// + c1over54 ;(D.f[DIR_M0P ])[kw ];//ktw
- real mfbcc = (D.f[DIR_0PP])[k];//[ktn ];// + c1over54 ;(D.f[DIR_0PP ])[k ];//ktn
- real mfbaa = (D.f[DIR_0MM])[kbs];//[kbs ];// + c1over54 ;(D.f[DIR_0MM ])[kbs];
- real mfbca = (D.f[DIR_0PM])[kb];//[kbn ];// + c1over54 ;(D.f[DIR_0PM ])[kb ];//kbn
- real mfbac = (D.f[DIR_0MP])[ks];//[kts ];// + c1over54 ;(D.f[DIR_0MP ])[ks ];//kts
+ real mfcbb = (D.f[DIR_P00])[k];//[ke ];// + c2over27 ;(D.f[DIR_P00])[k ];//ke
+ real mfabb = (D.f[DIR_M00])[kw];//[kw ];// + c2over27 ;(D.f[DIR_M00])[kw ];
+ real mfbcb = (D.f[DIR_0P0])[k];//[kn ];// + c2over27 ;(D.f[DIR_0P0])[k ];//kn
+ real mfbab = (D.f[DIR_0M0])[ks];//[ks ];// + c2over27 ;(D.f[DIR_0M0])[ks ];
+ real mfbbc = (D.f[DIR_00P])[k];//[kt ];// + c2over27 ;(D.f[DIR_00P])[k ];//kt
+ real mfbba = (D.f[DIR_00M])[kb];//[kb ];// + c2over27 ;(D.f[DIR_00M])[kb ];
+ real mfccb = (D.f[DIR_PP0])[k];//[kne ];// + c1over54 ;(D.f[DIR_PP0])[k ];//kne
+ real mfaab = (D.f[DIR_MM0])[ksw];//[ksw ];// + c1over54 ;(D.f[DIR_MM0])[ksw];
+ real mfcab = (D.f[DIR_PM0])[ks];//[kse ];// + c1over54 ;(D.f[DIR_PM0])[ks ];//kse
+ real mfacb = (D.f[DIR_MP0])[kw];//[knw ];// + c1over54 ;(D.f[DIR_MP0])[kw ];//knw
+ real mfcbc = (D.f[DIR_P0P])[k];//[kte ];// + c1over54 ;(D.f[DIR_P0P])[k ];//kte
+ real mfaba = (D.f[DIR_M0M])[kbw];//[kbw ];// + c1over54 ;(D.f[DIR_M0M])[kbw];
+ real mfcba = (D.f[DIR_P0M])[kb];//[kbe ];// + c1over54 ;(D.f[DIR_P0M])[kb ];//kbe
+ real mfabc = (D.f[DIR_M0P])[kw];//[ktw ];// + c1over54 ;(D.f[DIR_M0P])[kw ];//ktw
+ real mfbcc = (D.f[DIR_0PP])[k];//[ktn ];// + c1over54 ;(D.f[DIR_0PP])[k ];//ktn
+ real mfbaa = (D.f[DIR_0MM])[kbs];//[kbs ];// + c1over54 ;(D.f[DIR_0MM])[kbs];
+ real mfbca = (D.f[DIR_0PM])[kb];//[kbn ];// + c1over54 ;(D.f[DIR_0PM])[kb ];//kbn
+ real mfbac = (D.f[DIR_0MP])[ks];//[kts ];// + c1over54 ;(D.f[DIR_0MP])[ks ];//kts
real mfbbb = (D.f[DIR_000])[k];//[kzero];// + c8over27 ;(D.f[DIR_000])[k ];//kzero
- real mfccc = (D.f[DIR_PPP])[k];//[ktne ];// + c1over216;(D.f[DIR_PPP ])[k ];//ktne
- real mfaac = (D.f[DIR_MMP])[ksw];//[ktsw ];// + c1over216;(D.f[DIR_MMP ])[ksw];//ktsw
- real mfcac = (D.f[DIR_PMP])[ks];//[ktse ];// + c1over216;(D.f[DIR_PMP ])[ks ];//ktse
- real mfacc = (D.f[DIR_MPP])[kw];//[ktnw ];// + c1over216;(D.f[DIR_MPP ])[kw ];//ktnw
- real mfcca = (D.f[DIR_PPM])[kb];//[kbne ];// + c1over216;(D.f[DIR_PPM ])[kb ];//kbne
- real mfaaa = (D.f[DIR_MMM])[kbsw];//[kbsw ];// + c1over216;(D.f[DIR_MMM ])[kbsw];
- real mfcaa = (D.f[DIR_PMM])[kbs];//[kbse ];// + c1over216;(D.f[DIR_PMM ])[kbs];//kbse
- real mfaca = (D.f[DIR_MPM])[kbw];//[kbnw ];// + c1over216;(D.f[DIR_MPM ])[kbw];//kbnw
+ real mfccc = (D.f[DIR_PPP])[k];//[ktne ];// + c1over216;(D.f[DIR_PPP])[k ];//ktne
+ real mfaac = (D.f[DIR_MMP])[ksw];//[ktsw ];// + c1over216;(D.f[DIR_MMP])[ksw];//ktsw
+ real mfcac = (D.f[DIR_PMP])[ks];//[ktse ];// + c1over216;(D.f[DIR_PMP])[ks ];//ktse
+ real mfacc = (D.f[DIR_MPP])[kw];//[ktnw ];// + c1over216;(D.f[DIR_MPP])[kw ];//ktnw
+ real mfcca = (D.f[DIR_PPM])[kb];//[kbne ];// + c1over216;(D.f[DIR_PPM])[kb ];//kbne
+ real mfaaa = (D.f[DIR_MMM])[kbsw];//[kbsw ];// + c1over216;(D.f[DIR_MMM])[kbsw];
+ real mfcaa = (D.f[DIR_PMM])[kbs];//[kbse ];// + c1over216;(D.f[DIR_PMM])[kbs];//kbse
+ real mfaca = (D.f[DIR_MPM])[kbw];//[kbnw ];// + c1over216;(D.f[DIR_MPM])[kbw];//kbnw
////////////////////////////////////////////////////////////////////////////////////
//Ship
real coord0X = 281.125f;//7.5f;
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Incompressible/CumulantIsoSP27/CumulantIsoIncompSP27.cu b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Incompressible/CumulantIsoSP27/CumulantIsoIncompSP27.cu
index 6551e1bde300e3a4d2a4f50cefdfff258edfacee..3a740bef6d7fbaa2883b3d36930d49bf9bf0bb3e 100644
--- a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Incompressible/CumulantIsoSP27/CumulantIsoIncompSP27.cu
+++ b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Incompressible/CumulantIsoSP27/CumulantIsoIncompSP27.cu
@@ -2,6 +2,7 @@
#include "CumulantIsoIncompSP27_Device.cuh"
#include "Parameter/Parameter.h"
+#include "cuda/CudaGrid.h"
std::shared_ptr<CumulantIsoIncompSP27> CumulantIsoIncompSP27::getNewInstance(std::shared_ptr<Parameter> para, int level)
{
@@ -10,36 +11,21 @@ std::shared_ptr<CumulantIsoIncompSP27> CumulantIsoIncompSP27::getNewInstance(std
void CumulantIsoIncompSP27::run()
{
- int size_Mat = para->getParD(level)->numberOfNodes;
- int numberOfThreads = para->getParD(level)->numberofthreads;
-
- int Grid = (size_Mat / numberOfThreads) + 1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid / Grid1) + 1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 grid(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1);
-
- LB_Kernel_Cum_IsoTest_Incomp_SP_27 << < grid, threads >> >(para->getParD(level)->omega,
- para->getParD(level)->typeOfGridNode,
- para->getParD(level)->neighborX,
- para->getParD(level)->neighborY,
- para->getParD(level)->neighborZ,
- para->getParD(level)->distributions.f[0],
- para->getParD(level)->dxxUx,
- para->getParD(level)->dyyUy,
- para->getParD(level)->dzzUz,
- para->getParD(level)->numberOfNodes,
- para->getParD(level)->isEvenTimestep);
- getLastCudaError("LB_Kernel_Kum_IsoTest_SP_27 execution failed");
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(para->getParD(level)->numberofthreads, para->getParD(level)->numberOfNodes);
+
+ LB_Kernel_Cum_IsoTest_Incomp_SP_27 <<< grid.grid, grid.threads >>>(
+ para->getParD(level)->omega,
+ para->getParD(level)->typeOfGridNode,
+ para->getParD(level)->neighborX,
+ para->getParD(level)->neighborY,
+ para->getParD(level)->neighborZ,
+ para->getParD(level)->distributions.f[0],
+ para->getParD(level)->dxxUx,
+ para->getParD(level)->dyyUy,
+ para->getParD(level)->dzzUz,
+ para->getParD(level)->numberOfNodes,
+ para->getParD(level)->isEvenTimestep);
+ getLastCudaError("LB_Kernel_Cum_IsoTest_Incomp_SP_27 execution failed");
}
CumulantIsoIncompSP27::CumulantIsoIncompSP27(std::shared_ptr<Parameter> para, int level)
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Incompressible/CumulantIsoSP27/CumulantIsoIncompSP27_Device.cu b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Incompressible/CumulantIsoSP27/CumulantIsoIncompSP27_Device.cu
index 64d697f2b0953cee75f4397e399a0e6128e486a2..1f0ef2ec84c8d4b9b4be57548bde396c3316a80d 100644
--- a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Incompressible/CumulantIsoSP27/CumulantIsoIncompSP27_Device.cu
+++ b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Incompressible/CumulantIsoSP27/CumulantIsoIncompSP27_Device.cu
@@ -40,63 +40,63 @@ __global__ void LB_Kernel_Cum_IsoTest_Incomp_SP_27(real omega,
Distributions27 D;
if (EvenOrOdd == true)
{
- D.f[DIR_P00] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_M00] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0P0] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0M0] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00P] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00M] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_PP0] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_MM0] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_PM0] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_MP0] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_P0P] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_M0M] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_P0M] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_M0P] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0PP] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0MM] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0PM] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0MP] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_PPP] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_MMP] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_PMP] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_MPP] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_PPM] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_MMM] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_PMM] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_MPM] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DDStart[DIR_P00 * size_Mat];
+ D.f[DIR_M00] = &DDStart[DIR_M00 * size_Mat];
+ D.f[DIR_0P0] = &DDStart[DIR_0P0 * size_Mat];
+ D.f[DIR_0M0] = &DDStart[DIR_0M0 * size_Mat];
+ D.f[DIR_00P] = &DDStart[DIR_00P * size_Mat];
+ D.f[DIR_00M] = &DDStart[DIR_00M * size_Mat];
+ D.f[DIR_PP0] = &DDStart[DIR_PP0 * size_Mat];
+ D.f[DIR_MM0] = &DDStart[DIR_MM0 * size_Mat];
+ D.f[DIR_PM0] = &DDStart[DIR_PM0 * size_Mat];
+ D.f[DIR_MP0] = &DDStart[DIR_MP0 * size_Mat];
+ D.f[DIR_P0P] = &DDStart[DIR_P0P * size_Mat];
+ D.f[DIR_M0M] = &DDStart[DIR_M0M * size_Mat];
+ D.f[DIR_P0M] = &DDStart[DIR_P0M * size_Mat];
+ D.f[DIR_M0P] = &DDStart[DIR_M0P * size_Mat];
+ D.f[DIR_0PP] = &DDStart[DIR_0PP * size_Mat];
+ D.f[DIR_0MM] = &DDStart[DIR_0MM * size_Mat];
+ D.f[DIR_0PM] = &DDStart[DIR_0PM * size_Mat];
+ D.f[DIR_0MP] = &DDStart[DIR_0MP * size_Mat];
+ D.f[DIR_000] = &DDStart[DIR_000 * size_Mat];
+ D.f[DIR_PPP] = &DDStart[DIR_PPP * size_Mat];
+ D.f[DIR_MMP] = &DDStart[DIR_MMP * size_Mat];
+ D.f[DIR_PMP] = &DDStart[DIR_PMP * size_Mat];
+ D.f[DIR_MPP] = &DDStart[DIR_MPP * size_Mat];
+ D.f[DIR_PPM] = &DDStart[DIR_PPM * size_Mat];
+ D.f[DIR_MMM] = &DDStart[DIR_MMM * size_Mat];
+ D.f[DIR_PMM] = &DDStart[DIR_PMM * size_Mat];
+ D.f[DIR_MPM] = &DDStart[DIR_MPM * size_Mat];
}
else
{
- D.f[DIR_M00] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_P00] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0M0] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0P0] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00M] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00P] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_MM0] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_PP0] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_MP0] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_PM0] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_M0M] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_P0P] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_M0P] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_P0M] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0MM] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0PP] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0MP] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0PM] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_MMM] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_PPM] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_MPM] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_PMM] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_MMP] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_PPP] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_MPP] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_PMP] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_M00] = &DDStart[DIR_P00 * size_Mat];
+ D.f[DIR_P00] = &DDStart[DIR_M00 * size_Mat];
+ D.f[DIR_0M0] = &DDStart[DIR_0P0 * size_Mat];
+ D.f[DIR_0P0] = &DDStart[DIR_0M0 * size_Mat];
+ D.f[DIR_00M] = &DDStart[DIR_00P * size_Mat];
+ D.f[DIR_00P] = &DDStart[DIR_00M * size_Mat];
+ D.f[DIR_MM0] = &DDStart[DIR_PP0 * size_Mat];
+ D.f[DIR_PP0] = &DDStart[DIR_MM0 * size_Mat];
+ D.f[DIR_MP0] = &DDStart[DIR_PM0 * size_Mat];
+ D.f[DIR_PM0] = &DDStart[DIR_MP0 * size_Mat];
+ D.f[DIR_M0M] = &DDStart[DIR_P0P * size_Mat];
+ D.f[DIR_P0P] = &DDStart[DIR_M0M * size_Mat];
+ D.f[DIR_M0P] = &DDStart[DIR_P0M * size_Mat];
+ D.f[DIR_P0M] = &DDStart[DIR_M0P * size_Mat];
+ D.f[DIR_0MM] = &DDStart[DIR_0PP * size_Mat];
+ D.f[DIR_0PP] = &DDStart[DIR_0MM * size_Mat];
+ D.f[DIR_0MP] = &DDStart[DIR_0PM * size_Mat];
+ D.f[DIR_0PM] = &DDStart[DIR_0MP * size_Mat];
+ D.f[DIR_000] = &DDStart[DIR_000 * size_Mat];
+ D.f[DIR_MMM] = &DDStart[DIR_PPP * size_Mat];
+ D.f[DIR_PPM] = &DDStart[DIR_MMP * size_Mat];
+ D.f[DIR_MPM] = &DDStart[DIR_PMP * size_Mat];
+ D.f[DIR_PMM] = &DDStart[DIR_MPP * size_Mat];
+ D.f[DIR_MMP] = &DDStart[DIR_PPM * size_Mat];
+ D.f[DIR_PPP] = &DDStart[DIR_MMM * size_Mat];
+ D.f[DIR_MPP] = &DDStart[DIR_PMM * size_Mat];
+ D.f[DIR_PMP] = &DDStart[DIR_MPM * size_Mat];
}
////////////////////////////////////////////////////////////////////////////////
@@ -129,33 +129,33 @@ __global__ void LB_Kernel_Cum_IsoTest_Incomp_SP_27(real omega,
//unsigned int ktne = k;
unsigned int kbsw = neighborZ[ksw];
//////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
- real mfcbb = (D.f[DIR_P00])[k];//[ke ];// + c2over27 ;(D.f[DIR_P00 ])[k ];//ke
- real mfabb = (D.f[DIR_M00])[kw];//[kw ];// + c2over27 ;(D.f[DIR_M00 ])[kw ];
- real mfbcb = (D.f[DIR_0P0])[k];//[kn ];// + c2over27 ;(D.f[DIR_0P0 ])[k ];//kn
- real mfbab = (D.f[DIR_0M0])[ks];//[ks ];// + c2over27 ;(D.f[DIR_0M0 ])[ks ];
- real mfbbc = (D.f[DIR_00P])[k];//[kt ];// + c2over27 ;(D.f[DIR_00P ])[k ];//kt
- real mfbba = (D.f[DIR_00M])[kb];//[kb ];// + c2over27 ;(D.f[DIR_00M ])[kb ];
- real mfccb = (D.f[DIR_PP0])[k];//[kne ];// + c1over54 ;(D.f[DIR_PP0 ])[k ];//kne
- real mfaab = (D.f[DIR_MM0])[ksw];//[ksw ];// + c1over54 ;(D.f[DIR_MM0 ])[ksw];
- real mfcab = (D.f[DIR_PM0])[ks];//[kse ];// + c1over54 ;(D.f[DIR_PM0 ])[ks ];//kse
- real mfacb = (D.f[DIR_MP0])[kw];//[knw ];// + c1over54 ;(D.f[DIR_MP0 ])[kw ];//knw
- real mfcbc = (D.f[DIR_P0P])[k];//[kte ];// + c1over54 ;(D.f[DIR_P0P ])[k ];//kte
- real mfaba = (D.f[DIR_M0M])[kbw];//[kbw ];// + c1over54 ;(D.f[DIR_M0M ])[kbw];
- real mfcba = (D.f[DIR_P0M])[kb];//[kbe ];// + c1over54 ;(D.f[DIR_P0M ])[kb ];//kbe
- real mfabc = (D.f[DIR_M0P])[kw];//[ktw ];// + c1over54 ;(D.f[DIR_M0P ])[kw ];//ktw
- real mfbcc = (D.f[DIR_0PP])[k];//[ktn ];// + c1over54 ;(D.f[DIR_0PP ])[k ];//ktn
- real mfbaa = (D.f[DIR_0MM])[kbs];//[kbs ];// + c1over54 ;(D.f[DIR_0MM ])[kbs];
- real mfbca = (D.f[DIR_0PM])[kb];//[kbn ];// + c1over54 ;(D.f[DIR_0PM ])[kb ];//kbn
- real mfbac = (D.f[DIR_0MP])[ks];//[kts ];// + c1over54 ;(D.f[DIR_0MP ])[ks ];//kts
+ real mfcbb = (D.f[DIR_P00])[k];//[ke ];// + c2over27 ;(D.f[DIR_P00])[k ];//ke
+ real mfabb = (D.f[DIR_M00])[kw];//[kw ];// + c2over27 ;(D.f[DIR_M00])[kw ];
+ real mfbcb = (D.f[DIR_0P0])[k];//[kn ];// + c2over27 ;(D.f[DIR_0P0])[k ];//kn
+ real mfbab = (D.f[DIR_0M0])[ks];//[ks ];// + c2over27 ;(D.f[DIR_0M0])[ks ];
+ real mfbbc = (D.f[DIR_00P])[k];//[kt ];// + c2over27 ;(D.f[DIR_00P])[k ];//kt
+ real mfbba = (D.f[DIR_00M])[kb];//[kb ];// + c2over27 ;(D.f[DIR_00M])[kb ];
+ real mfccb = (D.f[DIR_PP0])[k];//[kne ];// + c1over54 ;(D.f[DIR_PP0])[k ];//kne
+ real mfaab = (D.f[DIR_MM0])[ksw];//[ksw ];// + c1over54 ;(D.f[DIR_MM0])[ksw];
+ real mfcab = (D.f[DIR_PM0])[ks];//[kse ];// + c1over54 ;(D.f[DIR_PM0])[ks ];//kse
+ real mfacb = (D.f[DIR_MP0])[kw];//[knw ];// + c1over54 ;(D.f[DIR_MP0])[kw ];//knw
+ real mfcbc = (D.f[DIR_P0P])[k];//[kte ];// + c1over54 ;(D.f[DIR_P0P])[k ];//kte
+ real mfaba = (D.f[DIR_M0M])[kbw];//[kbw ];// + c1over54 ;(D.f[DIR_M0M])[kbw];
+ real mfcba = (D.f[DIR_P0M])[kb];//[kbe ];// + c1over54 ;(D.f[DIR_P0M])[kb ];//kbe
+ real mfabc = (D.f[DIR_M0P])[kw];//[ktw ];// + c1over54 ;(D.f[DIR_M0P])[kw ];//ktw
+ real mfbcc = (D.f[DIR_0PP])[k];//[ktn ];// + c1over54 ;(D.f[DIR_0PP])[k ];//ktn
+ real mfbaa = (D.f[DIR_0MM])[kbs];//[kbs ];// + c1over54 ;(D.f[DIR_0MM])[kbs];
+ real mfbca = (D.f[DIR_0PM])[kb];//[kbn ];// + c1over54 ;(D.f[DIR_0PM])[kb ];//kbn
+ real mfbac = (D.f[DIR_0MP])[ks];//[kts ];// + c1over54 ;(D.f[DIR_0MP])[ks ];//kts
real mfbbb = (D.f[DIR_000])[k];//[kzero];// + c8over27 ;(D.f[DIR_000])[k ];//kzero
- real mfccc = (D.f[DIR_PPP])[k];//[ktne ];// + c1over216;(D.f[DIR_PPP ])[k ];//ktne
- real mfaac = (D.f[DIR_MMP])[ksw];//[ktsw ];// + c1over216;(D.f[DIR_MMP ])[ksw];//ktsw
- real mfcac = (D.f[DIR_PMP])[ks];//[ktse ];// + c1over216;(D.f[DIR_PMP ])[ks ];//ktse
- real mfacc = (D.f[DIR_MPP])[kw];//[ktnw ];// + c1over216;(D.f[DIR_MPP ])[kw ];//ktnw
- real mfcca = (D.f[DIR_PPM])[kb];//[kbne ];// + c1over216;(D.f[DIR_PPM ])[kb ];//kbne
- real mfaaa = (D.f[DIR_MMM])[kbsw];//[kbsw ];// + c1over216;(D.f[DIR_MMM ])[kbsw];
- real mfcaa = (D.f[DIR_PMM])[kbs];//[kbse ];// + c1over216;(D.f[DIR_PMM ])[kbs];//kbse
- real mfaca = (D.f[DIR_MPM])[kbw];//[kbnw ];// + c1over216;(D.f[DIR_MPM ])[kbw];//kbnw
+ real mfccc = (D.f[DIR_PPP])[k];//[ktne ];// + c1over216;(D.f[DIR_PPP])[k ];//ktne
+ real mfaac = (D.f[DIR_MMP])[ksw];//[ktsw ];// + c1over216;(D.f[DIR_MMP])[ksw];//ktsw
+ real mfcac = (D.f[DIR_PMP])[ks];//[ktse ];// + c1over216;(D.f[DIR_PMP])[ks ];//ktse
+ real mfacc = (D.f[DIR_MPP])[kw];//[ktnw ];// + c1over216;(D.f[DIR_MPP])[kw ];//ktnw
+ real mfcca = (D.f[DIR_PPM])[kb];//[kbne ];// + c1over216;(D.f[DIR_PPM])[kb ];//kbne
+ real mfaaa = (D.f[DIR_MMM])[kbsw];//[kbsw ];// + c1over216;(D.f[DIR_MMM])[kbsw];
+ real mfcaa = (D.f[DIR_PMM])[kbs];//[kbse ];// + c1over216;(D.f[DIR_PMM])[kbs];//kbse
+ real mfaca = (D.f[DIR_MPM])[kbw];//[kbnw ];// + c1over216;(D.f[DIR_MPM])[kbw];//kbnw
////////////////////////////////////////////////////////////////////////////////////
//slow
//real oMdrho = one - ((((mfccc+mfaaa) + (mfaca+mfcac)) + ((mfacc+mfcaa) + (mfaac+mfcca))) +
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Incompressible/CumulantK15/CumulantK15Incomp.cu b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Incompressible/CumulantK15/CumulantK15Incomp.cu
index 40cde56b007f70f98db13d5962f3e746b97637ef..7ae17b97170b4d8474acd6777f7c27411a962681 100644
--- a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Incompressible/CumulantK15/CumulantK15Incomp.cu
+++ b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Incompressible/CumulantK15/CumulantK15Incomp.cu
@@ -2,6 +2,7 @@
#include "CumulantK15Incomp_Device.cuh"
#include "Parameter/Parameter.h"
+#include "cuda/CudaGrid.h"
std::shared_ptr<CumulantK15Incomp> CumulantK15Incomp::getNewInstance(std::shared_ptr<Parameter> para, int level)
{
@@ -10,33 +11,18 @@ std::shared_ptr<CumulantK15Incomp> CumulantK15Incomp::getNewInstance(std::shared
void CumulantK15Incomp::run()
{
- int size_Mat = para->getParD(level)->numberOfNodes;
- int numberOfThreads = para->getParD(level)->numberofthreads;
-
- int Grid = (size_Mat / numberOfThreads) + 1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid / Grid1) + 1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 grid(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1);
-
- LB_Kernel_CumulantK15Incomp <<< grid, threads >>>( para->getParD(level)->omega,
- para->getParD(level)->typeOfGridNode,
- para->getParD(level)->neighborX,
- para->getParD(level)->neighborY,
- para->getParD(level)->neighborZ,
- para->getParD(level)->distributions.f[0],
- para->getParD(level)->numberOfNodes,
- para->getParD(level)->isEvenTimestep);
- getLastCudaError("LB_Kernel_CumulantK15Incomp execution failed");
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(para->getParD(level)->numberofthreads, para->getParD(level)->numberOfNodes);
+
+ LB_Kernel_CumulantK15Incomp <<< grid.grid, grid.threads >>>(
+ para->getParD(level)->omega,
+ para->getParD(level)->typeOfGridNode,
+ para->getParD(level)->neighborX,
+ para->getParD(level)->neighborY,
+ para->getParD(level)->neighborZ,
+ para->getParD(level)->distributions.f[0],
+ para->getParD(level)->numberOfNodes,
+ para->getParD(level)->isEvenTimestep);
+ getLastCudaError("LB_Kernel_CumulantK15Incomp execution failed");
}
CumulantK15Incomp::CumulantK15Incomp(std::shared_ptr<Parameter> para, int level)
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Incompressible/CumulantK15/CumulantK15Incomp_Device.cu b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Incompressible/CumulantK15/CumulantK15Incomp_Device.cu
index fc108ef1ef109a40735e250bd9a0f21491e4f977..01b60b3bf8067a81f99b912c4c0c700963f5448c 100644
--- a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Incompressible/CumulantK15/CumulantK15Incomp_Device.cu
+++ b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Incompressible/CumulantK15/CumulantK15Incomp_Device.cu
@@ -37,63 +37,63 @@ __global__ void LB_Kernel_CumulantK15Incomp(real omega,
Distributions27 D;
if (EvenOrOdd == true)
{
- D.f[DIR_P00] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_M00] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0P0] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0M0] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00P] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00M] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_PP0] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_MM0] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_PM0] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_MP0] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_P0P] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_M0M] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_P0M] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_M0P] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0PP] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0MM] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0PM] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0MP] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_PPP] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_MMP] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_PMP] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_MPP] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_PPM] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_MMM] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_PMM] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_MPM] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DDStart[DIR_P00 * size_Mat];
+ D.f[DIR_M00] = &DDStart[DIR_M00 * size_Mat];
+ D.f[DIR_0P0] = &DDStart[DIR_0P0 * size_Mat];
+ D.f[DIR_0M0] = &DDStart[DIR_0M0 * size_Mat];
+ D.f[DIR_00P] = &DDStart[DIR_00P * size_Mat];
+ D.f[DIR_00M] = &DDStart[DIR_00M * size_Mat];
+ D.f[DIR_PP0] = &DDStart[DIR_PP0 * size_Mat];
+ D.f[DIR_MM0] = &DDStart[DIR_MM0 * size_Mat];
+ D.f[DIR_PM0] = &DDStart[DIR_PM0 * size_Mat];
+ D.f[DIR_MP0] = &DDStart[DIR_MP0 * size_Mat];
+ D.f[DIR_P0P] = &DDStart[DIR_P0P * size_Mat];
+ D.f[DIR_M0M] = &DDStart[DIR_M0M * size_Mat];
+ D.f[DIR_P0M] = &DDStart[DIR_P0M * size_Mat];
+ D.f[DIR_M0P] = &DDStart[DIR_M0P * size_Mat];
+ D.f[DIR_0PP] = &DDStart[DIR_0PP * size_Mat];
+ D.f[DIR_0MM] = &DDStart[DIR_0MM * size_Mat];
+ D.f[DIR_0PM] = &DDStart[DIR_0PM * size_Mat];
+ D.f[DIR_0MP] = &DDStart[DIR_0MP * size_Mat];
+ D.f[DIR_000] = &DDStart[DIR_000 * size_Mat];
+ D.f[DIR_PPP] = &DDStart[DIR_PPP * size_Mat];
+ D.f[DIR_MMP] = &DDStart[DIR_MMP * size_Mat];
+ D.f[DIR_PMP] = &DDStart[DIR_PMP * size_Mat];
+ D.f[DIR_MPP] = &DDStart[DIR_MPP * size_Mat];
+ D.f[DIR_PPM] = &DDStart[DIR_PPM * size_Mat];
+ D.f[DIR_MMM] = &DDStart[DIR_MMM * size_Mat];
+ D.f[DIR_PMM] = &DDStart[DIR_PMM * size_Mat];
+ D.f[DIR_MPM] = &DDStart[DIR_MPM * size_Mat];
}
else
{
- D.f[DIR_M00] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_P00] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0M0] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0P0] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00M] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00P] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_MM0] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_PP0] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_MP0] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_PM0] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_M0M] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_P0P] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_M0P] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_P0M] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0MM] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0PP] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0MP] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0PM] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_MMM] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_PPM] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_MPM] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_PMM] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_MMP] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_PPP] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_MPP] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_PMP] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_M00] = &DDStart[DIR_P00 * size_Mat];
+ D.f[DIR_P00] = &DDStart[DIR_M00 * size_Mat];
+ D.f[DIR_0M0] = &DDStart[DIR_0P0 * size_Mat];
+ D.f[DIR_0P0] = &DDStart[DIR_0M0 * size_Mat];
+ D.f[DIR_00M] = &DDStart[DIR_00P * size_Mat];
+ D.f[DIR_00P] = &DDStart[DIR_00M * size_Mat];
+ D.f[DIR_MM0] = &DDStart[DIR_PP0 * size_Mat];
+ D.f[DIR_PP0] = &DDStart[DIR_MM0 * size_Mat];
+ D.f[DIR_MP0] = &DDStart[DIR_PM0 * size_Mat];
+ D.f[DIR_PM0] = &DDStart[DIR_MP0 * size_Mat];
+ D.f[DIR_M0M] = &DDStart[DIR_P0P * size_Mat];
+ D.f[DIR_P0P] = &DDStart[DIR_M0M * size_Mat];
+ D.f[DIR_M0P] = &DDStart[DIR_P0M * size_Mat];
+ D.f[DIR_P0M] = &DDStart[DIR_M0P * size_Mat];
+ D.f[DIR_0MM] = &DDStart[DIR_0PP * size_Mat];
+ D.f[DIR_0PP] = &DDStart[DIR_0MM * size_Mat];
+ D.f[DIR_0MP] = &DDStart[DIR_0PM * size_Mat];
+ D.f[DIR_0PM] = &DDStart[DIR_0MP * size_Mat];
+ D.f[DIR_000] = &DDStart[DIR_000 * size_Mat];
+ D.f[DIR_MMM] = &DDStart[DIR_PPP * size_Mat];
+ D.f[DIR_PPM] = &DDStart[DIR_MMP * size_Mat];
+ D.f[DIR_MPM] = &DDStart[DIR_PMP * size_Mat];
+ D.f[DIR_PMM] = &DDStart[DIR_MPP * size_Mat];
+ D.f[DIR_MMP] = &DDStart[DIR_PPM * size_Mat];
+ D.f[DIR_PPP] = &DDStart[DIR_MMM * size_Mat];
+ D.f[DIR_MPP] = &DDStart[DIR_PMM * size_Mat];
+ D.f[DIR_PMP] = &DDStart[DIR_MPM * size_Mat];
}
////////////////////////////////////////////////////////////////////////////////
@@ -154,33 +154,33 @@ __global__ void LB_Kernel_CumulantK15Incomp(real omega,
//unsigned int ktne = k;
//unsigned int kbsw = neighborZ[ksw];
//////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
- real mfcbb = (D.f[DIR_P00])[k];//[ke ];// + c2over27 ;(D.f[DIR_P00 ])[k ];//ke
- real mfabb = (D.f[DIR_M00])[kw];//[kw ];// + c2over27 ;(D.f[DIR_M00 ])[kw ];
- real mfbcb = (D.f[DIR_0P0])[k];//[kn ];// + c2over27 ;(D.f[DIR_0P0 ])[k ];//kn
- real mfbab = (D.f[DIR_0M0])[ks];//[ks ];// + c2over27 ;(D.f[DIR_0M0 ])[ks ];
- real mfbbc = (D.f[DIR_00P])[k];//[kt ];// + c2over27 ;(D.f[DIR_00P ])[k ];//kt
- real mfbba = (D.f[DIR_00M])[kb];//[kb ];// + c2over27 ;(D.f[DIR_00M ])[kb ];
- real mfccb = (D.f[DIR_PP0])[k];//[kne ];// + c1over54 ;(D.f[DIR_PP0 ])[k ];//kne
- real mfaab = (D.f[DIR_MM0])[ksw];//[ksw ];// + c1over54 ;(D.f[DIR_MM0 ])[ksw];
- real mfcab = (D.f[DIR_PM0])[ks];//[kse ];// + c1over54 ;(D.f[DIR_PM0 ])[ks ];//kse
- real mfacb = (D.f[DIR_MP0])[kw];//[knw ];// + c1over54 ;(D.f[DIR_MP0 ])[kw ];//knw
- real mfcbc = (D.f[DIR_P0P])[k];//[kte ];// + c1over54 ;(D.f[DIR_P0P ])[k ];//kte
- real mfaba = (D.f[DIR_M0M])[kbw];//[kbw ];// + c1over54 ;(D.f[DIR_M0M ])[kbw];
- real mfcba = (D.f[DIR_P0M])[kb];//[kbe ];// + c1over54 ;(D.f[DIR_P0M ])[kb ];//kbe
- real mfabc = (D.f[DIR_M0P])[kw];//[ktw ];// + c1over54 ;(D.f[DIR_M0P ])[kw ];//ktw
- real mfbcc = (D.f[DIR_0PP])[k];//[ktn ];// + c1over54 ;(D.f[DIR_0PP ])[k ];//ktn
- real mfbaa = (D.f[DIR_0MM])[kbs];//[kbs ];// + c1over54 ;(D.f[DIR_0MM ])[kbs];
- real mfbca = (D.f[DIR_0PM])[kb];//[kbn ];// + c1over54 ;(D.f[DIR_0PM ])[kb ];//kbn
- real mfbac = (D.f[DIR_0MP])[ks];//[kts ];// + c1over54 ;(D.f[DIR_0MP ])[ks ];//kts
+ real mfcbb = (D.f[DIR_P00])[k];//[ke ];// + c2over27 ;(D.f[DIR_P00])[k ];//ke
+ real mfabb = (D.f[DIR_M00])[kw];//[kw ];// + c2over27 ;(D.f[DIR_M00])[kw ];
+ real mfbcb = (D.f[DIR_0P0])[k];//[kn ];// + c2over27 ;(D.f[DIR_0P0])[k ];//kn
+ real mfbab = (D.f[DIR_0M0])[ks];//[ks ];// + c2over27 ;(D.f[DIR_0M0])[ks ];
+ real mfbbc = (D.f[DIR_00P])[k];//[kt ];// + c2over27 ;(D.f[DIR_00P])[k ];//kt
+ real mfbba = (D.f[DIR_00M])[kb];//[kb ];// + c2over27 ;(D.f[DIR_00M])[kb ];
+ real mfccb = (D.f[DIR_PP0])[k];//[kne ];// + c1over54 ;(D.f[DIR_PP0])[k ];//kne
+ real mfaab = (D.f[DIR_MM0])[ksw];//[ksw ];// + c1over54 ;(D.f[DIR_MM0])[ksw];
+ real mfcab = (D.f[DIR_PM0])[ks];//[kse ];// + c1over54 ;(D.f[DIR_PM0])[ks ];//kse
+ real mfacb = (D.f[DIR_MP0])[kw];//[knw ];// + c1over54 ;(D.f[DIR_MP0])[kw ];//knw
+ real mfcbc = (D.f[DIR_P0P])[k];//[kte ];// + c1over54 ;(D.f[DIR_P0P])[k ];//kte
+ real mfaba = (D.f[DIR_M0M])[kbw];//[kbw ];// + c1over54 ;(D.f[DIR_M0M])[kbw];
+ real mfcba = (D.f[DIR_P0M])[kb];//[kbe ];// + c1over54 ;(D.f[DIR_P0M])[kb ];//kbe
+ real mfabc = (D.f[DIR_M0P])[kw];//[ktw ];// + c1over54 ;(D.f[DIR_M0P])[kw ];//ktw
+ real mfbcc = (D.f[DIR_0PP])[k];//[ktn ];// + c1over54 ;(D.f[DIR_0PP])[k ];//ktn
+ real mfbaa = (D.f[DIR_0MM])[kbs];//[kbs ];// + c1over54 ;(D.f[DIR_0MM])[kbs];
+ real mfbca = (D.f[DIR_0PM])[kb];//[kbn ];// + c1over54 ;(D.f[DIR_0PM])[kb ];//kbn
+ real mfbac = (D.f[DIR_0MP])[ks];//[kts ];// + c1over54 ;(D.f[DIR_0MP])[ks ];//kts
real mfbbb = (D.f[DIR_000])[k];//[kzero];// + c8over27 ;(D.f[DIR_000])[k ];//kzero
- real mfccc = (D.f[DIR_PPP])[k];//[ktne ];// + c1over216;(D.f[DIR_PPP ])[k ];//ktne
- real mfaac = (D.f[DIR_MMP])[ksw];//[ktsw ];// + c1over216;(D.f[DIR_MMP ])[ksw];//ktsw
- real mfcac = (D.f[DIR_PMP])[ks];//[ktse ];// + c1over216;(D.f[DIR_PMP ])[ks ];//ktse
- real mfacc = (D.f[DIR_MPP])[kw];//[ktnw ];// + c1over216;(D.f[DIR_MPP ])[kw ];//ktnw
- real mfcca = (D.f[DIR_PPM])[kb];//[kbne ];// + c1over216;(D.f[DIR_PPM ])[kb ];//kbne
- real mfaaa = (D.f[DIR_MMM])[kbsw];//[kbsw ];// + c1over216;(D.f[DIR_MMM ])[kbsw];
- real mfcaa = (D.f[DIR_PMM])[kbs];//[kbse ];// + c1over216;(D.f[DIR_PMM ])[kbs];//kbse
- real mfaca = (D.f[DIR_MPM])[kbw];//[kbnw ];// + c1over216;(D.f[DIR_MPM ])[kbw];//kbnw
+ real mfccc = (D.f[DIR_PPP])[k];//[ktne ];// + c1over216;(D.f[DIR_PPP])[k ];//ktne
+ real mfaac = (D.f[DIR_MMP])[ksw];//[ktsw ];// + c1over216;(D.f[DIR_MMP])[ksw];//ktsw
+ real mfcac = (D.f[DIR_PMP])[ks];//[ktse ];// + c1over216;(D.f[DIR_PMP])[ks ];//ktse
+ real mfacc = (D.f[DIR_MPP])[kw];//[ktnw ];// + c1over216;(D.f[DIR_MPP])[kw ];//ktnw
+ real mfcca = (D.f[DIR_PPM])[kb];//[kbne ];// + c1over216;(D.f[DIR_PPM])[kb ];//kbne
+ real mfaaa = (D.f[DIR_MMM])[kbsw];//[kbsw ];// + c1over216;(D.f[DIR_MMM])[kbsw];
+ real mfcaa = (D.f[DIR_PMM])[kbs];//[kbse ];// + c1over216;(D.f[DIR_PMM])[kbs];//kbse
+ real mfaca = (D.f[DIR_MPM])[kbw];//[kbnw ];// + c1over216;(D.f[DIR_MPM])[kbw];//kbnw
////////////////////////////////////////////////////////////////////////////////////
//slow
//real oMdrho = one - ((((mfccc+mfaaa) + (mfaca+mfcac)) + ((mfacc+mfcaa) + (mfaac+mfcca))) +
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Incompressible/MRT/MRTIncompSP27.cu b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Incompressible/MRT/MRTIncompSP27.cu
index c4311309e4653f2862e303dacb3e2d07646a5061..7645703e0d40176b136762d6b48633f4a9c0d950 100644
--- a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Incompressible/MRT/MRTIncompSP27.cu
+++ b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Incompressible/MRT/MRTIncompSP27.cu
@@ -2,6 +2,7 @@
#include "MRTIncompSP27_Device.cuh"
#include "Parameter/Parameter.h"
+#include "cuda/CudaGrid.h"
std::shared_ptr<MRTIncompSP27> MRTIncompSP27::getNewInstance(std::shared_ptr<Parameter> para, int level)
{
@@ -10,33 +11,18 @@ std::shared_ptr<MRTIncompSP27> MRTIncompSP27::getNewInstance(std::shared_ptr<Par
void MRTIncompSP27::run()
{
- int size_Mat = para->getParD(level)->numberOfNodes;
- int numberOfThreads = para->getParD(level)->numberofthreads;
-
- int Grid = (size_Mat / numberOfThreads) + 1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid / Grid1) + 1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 grid(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1);
-
- LB_Kernel_MRT_Incomp_SP_27 << < grid, threads >> >( para->getParD(level)->omega,
- para->getParD(level)->typeOfGridNode,
- para->getParD(level)->neighborX,
- para->getParD(level)->neighborY,
- para->getParD(level)->neighborZ,
- para->getParD(level)->distributions.f[0],
- para->getParD(level)->numberOfNodes,
- para->getParD(level)->isEvenTimestep);
- getLastCudaError("LB_Kernel_MRT_SP_27 execution failed");
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(para->getParD(level)->numberofthreads, para->getParD(level)->numberOfNodes);
+
+ LB_Kernel_MRT_Incomp_SP_27 <<< grid.grid, grid.threads >>>(
+ para->getParD(level)->omega,
+ para->getParD(level)->typeOfGridNode,
+ para->getParD(level)->neighborX,
+ para->getParD(level)->neighborY,
+ para->getParD(level)->neighborZ,
+ para->getParD(level)->distributions.f[0],
+ para->getParD(level)->numberOfNodes,
+ para->getParD(level)->isEvenTimestep);
+ getLastCudaError("LB_Kernel_MRT_Incomp_SP_27 execution failed");
}
MRTIncompSP27::MRTIncompSP27(std::shared_ptr<Parameter> para, int level)
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Incompressible/MRT/MRTIncompSP27_Device.cu b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Incompressible/MRT/MRTIncompSP27_Device.cu
index f6a283c2f9ba3c15729061ebeabcf34edd0abe97..a6663cc3c72696fda2ce9819203cd19195088730 100644
--- a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Incompressible/MRT/MRTIncompSP27_Device.cu
+++ b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/BasicKernels/FluidFlow/Incompressible/MRT/MRTIncompSP27_Device.cu
@@ -37,63 +37,63 @@ __global__ void LB_Kernel_MRT_Incomp_SP_27(real omega,
Distributions27 D;
if (EvenOrOdd == true)
{
- D.f[DIR_P00] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_M00] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0P0] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0M0] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00P] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00M] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_PP0] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_MM0] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_PM0] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_MP0] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_P0P] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_M0M] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_P0M] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_M0P] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0PP] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0MM] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0PM] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0MP] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_PPP] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_MMP] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_PMP] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_MPP] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_PPM] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_MMM] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_PMM] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_MPM] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DDStart[DIR_P00 * size_Mat];
+ D.f[DIR_M00] = &DDStart[DIR_M00 * size_Mat];
+ D.f[DIR_0P0] = &DDStart[DIR_0P0 * size_Mat];
+ D.f[DIR_0M0] = &DDStart[DIR_0M0 * size_Mat];
+ D.f[DIR_00P] = &DDStart[DIR_00P * size_Mat];
+ D.f[DIR_00M] = &DDStart[DIR_00M * size_Mat];
+ D.f[DIR_PP0] = &DDStart[DIR_PP0 * size_Mat];
+ D.f[DIR_MM0] = &DDStart[DIR_MM0 * size_Mat];
+ D.f[DIR_PM0] = &DDStart[DIR_PM0 * size_Mat];
+ D.f[DIR_MP0] = &DDStart[DIR_MP0 * size_Mat];
+ D.f[DIR_P0P] = &DDStart[DIR_P0P * size_Mat];
+ D.f[DIR_M0M] = &DDStart[DIR_M0M * size_Mat];
+ D.f[DIR_P0M] = &DDStart[DIR_P0M * size_Mat];
+ D.f[DIR_M0P] = &DDStart[DIR_M0P * size_Mat];
+ D.f[DIR_0PP] = &DDStart[DIR_0PP * size_Mat];
+ D.f[DIR_0MM] = &DDStart[DIR_0MM * size_Mat];
+ D.f[DIR_0PM] = &DDStart[DIR_0PM * size_Mat];
+ D.f[DIR_0MP] = &DDStart[DIR_0MP * size_Mat];
+ D.f[DIR_000] = &DDStart[DIR_000 * size_Mat];
+ D.f[DIR_PPP] = &DDStart[DIR_PPP * size_Mat];
+ D.f[DIR_MMP] = &DDStart[DIR_MMP * size_Mat];
+ D.f[DIR_PMP] = &DDStart[DIR_PMP * size_Mat];
+ D.f[DIR_MPP] = &DDStart[DIR_MPP * size_Mat];
+ D.f[DIR_PPM] = &DDStart[DIR_PPM * size_Mat];
+ D.f[DIR_MMM] = &DDStart[DIR_MMM * size_Mat];
+ D.f[DIR_PMM] = &DDStart[DIR_PMM * size_Mat];
+ D.f[DIR_MPM] = &DDStart[DIR_MPM * size_Mat];
}
else
{
- D.f[DIR_M00] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_P00] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0M0] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0P0] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00M] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00P] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_MM0] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_PP0] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_MP0] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_PM0] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_M0M] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_P0P] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_M0P] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_P0M] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0MM] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0PP] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0MP] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0PM] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_MMM] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_PPM] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_MPM] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_PMM] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_MMP] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_PPP] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_MPP] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_PMP] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_M00] = &DDStart[DIR_P00 * size_Mat];
+ D.f[DIR_P00] = &DDStart[DIR_M00 * size_Mat];
+ D.f[DIR_0M0] = &DDStart[DIR_0P0 * size_Mat];
+ D.f[DIR_0P0] = &DDStart[DIR_0M0 * size_Mat];
+ D.f[DIR_00M] = &DDStart[DIR_00P * size_Mat];
+ D.f[DIR_00P] = &DDStart[DIR_00M * size_Mat];
+ D.f[DIR_MM0] = &DDStart[DIR_PP0 * size_Mat];
+ D.f[DIR_PP0] = &DDStart[DIR_MM0 * size_Mat];
+ D.f[DIR_MP0] = &DDStart[DIR_PM0 * size_Mat];
+ D.f[DIR_PM0] = &DDStart[DIR_MP0 * size_Mat];
+ D.f[DIR_M0M] = &DDStart[DIR_P0P * size_Mat];
+ D.f[DIR_P0P] = &DDStart[DIR_M0M * size_Mat];
+ D.f[DIR_M0P] = &DDStart[DIR_P0M * size_Mat];
+ D.f[DIR_P0M] = &DDStart[DIR_M0P * size_Mat];
+ D.f[DIR_0MM] = &DDStart[DIR_0PP * size_Mat];
+ D.f[DIR_0PP] = &DDStart[DIR_0MM * size_Mat];
+ D.f[DIR_0MP] = &DDStart[DIR_0PM * size_Mat];
+ D.f[DIR_0PM] = &DDStart[DIR_0MP * size_Mat];
+ D.f[DIR_000] = &DDStart[DIR_000 * size_Mat];
+ D.f[DIR_MMM] = &DDStart[DIR_PPP * size_Mat];
+ D.f[DIR_PPM] = &DDStart[DIR_MMP * size_Mat];
+ D.f[DIR_MPM] = &DDStart[DIR_PMP * size_Mat];
+ D.f[DIR_PMM] = &DDStart[DIR_MPP * size_Mat];
+ D.f[DIR_MMP] = &DDStart[DIR_PPM * size_Mat];
+ D.f[DIR_PPP] = &DDStart[DIR_MMM * size_Mat];
+ D.f[DIR_MPP] = &DDStart[DIR_PMM * size_Mat];
+ D.f[DIR_PMP] = &DDStart[DIR_MPM * size_Mat];
}
////////////////////////////////////////////////////////////////////////////////
@@ -126,33 +126,33 @@ __global__ void LB_Kernel_MRT_Incomp_SP_27(real omega,
//unsigned int ktne = k;
unsigned int kbsw = neighborZ[ksw];
//////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
- real mfcbb = (D.f[DIR_P00])[k];//[ke ];// + c2over27 ;(D.f[DIR_P00 ])[k ];//ke
- real mfabb = (D.f[DIR_M00])[kw];//[kw ];// + c2over27 ;(D.f[DIR_M00 ])[kw ];
- real mfbcb = (D.f[DIR_0P0])[k];//[kn ];// + c2over27 ;(D.f[DIR_0P0 ])[k ];//kn
- real mfbab = (D.f[DIR_0M0])[ks];//[ks ];// + c2over27 ;(D.f[DIR_0M0 ])[ks ];
- real mfbbc = (D.f[DIR_00P])[k];//[kt ];// + c2over27 ;(D.f[DIR_00P ])[k ];//kt
- real mfbba = (D.f[DIR_00M])[kb];//[kb ];// + c2over27 ;(D.f[DIR_00M ])[kb ];
- real mfccb = (D.f[DIR_PP0])[k];//[kne ];// + c1over54 ;(D.f[DIR_PP0 ])[k ];//kne
- real mfaab = (D.f[DIR_MM0])[ksw];//[ksw ];// + c1over54 ;(D.f[DIR_MM0 ])[ksw];
- real mfcab = (D.f[DIR_PM0])[ks];//[kse ];// + c1over54 ;(D.f[DIR_PM0 ])[ks ];//kse
- real mfacb = (D.f[DIR_MP0])[kw];//[knw ];// + c1over54 ;(D.f[DIR_MP0 ])[kw ];//knw
- real mfcbc = (D.f[DIR_P0P])[k];//[kte ];// + c1over54 ;(D.f[DIR_P0P ])[k ];//kte
- real mfaba = (D.f[DIR_M0M])[kbw];//[kbw ];// + c1over54 ;(D.f[DIR_M0M ])[kbw];
- real mfcba = (D.f[DIR_P0M])[kb];//[kbe ];// + c1over54 ;(D.f[DIR_P0M ])[kb ];//kbe
- real mfabc = (D.f[DIR_M0P])[kw];//[ktw ];// + c1over54 ;(D.f[DIR_M0P ])[kw ];//ktw
- real mfbcc = (D.f[DIR_0PP])[k];//[ktn ];// + c1over54 ;(D.f[DIR_0PP ])[k ];//ktn
- real mfbaa = (D.f[DIR_0MM])[kbs];//[kbs ];// + c1over54 ;(D.f[DIR_0MM ])[kbs];
- real mfbca = (D.f[DIR_0PM])[kb];//[kbn ];// + c1over54 ;(D.f[DIR_0PM ])[kb ];//kbn
- real mfbac = (D.f[DIR_0MP])[ks];//[kts ];// + c1over54 ;(D.f[DIR_0MP ])[ks ];//kts
+ real mfcbb = (D.f[DIR_P00])[k];//[ke ];// + c2over27 ;(D.f[DIR_P00])[k ];//ke
+ real mfabb = (D.f[DIR_M00])[kw];//[kw ];// + c2over27 ;(D.f[DIR_M00])[kw ];
+ real mfbcb = (D.f[DIR_0P0])[k];//[kn ];// + c2over27 ;(D.f[DIR_0P0])[k ];//kn
+ real mfbab = (D.f[DIR_0M0])[ks];//[ks ];// + c2over27 ;(D.f[DIR_0M0])[ks ];
+ real mfbbc = (D.f[DIR_00P])[k];//[kt ];// + c2over27 ;(D.f[DIR_00P])[k ];//kt
+ real mfbba = (D.f[DIR_00M])[kb];//[kb ];// + c2over27 ;(D.f[DIR_00M])[kb ];
+ real mfccb = (D.f[DIR_PP0])[k];//[kne ];// + c1over54 ;(D.f[DIR_PP0])[k ];//kne
+ real mfaab = (D.f[DIR_MM0])[ksw];//[ksw ];// + c1over54 ;(D.f[DIR_MM0])[ksw];
+ real mfcab = (D.f[DIR_PM0])[ks];//[kse ];// + c1over54 ;(D.f[DIR_PM0])[ks ];//kse
+ real mfacb = (D.f[DIR_MP0])[kw];//[knw ];// + c1over54 ;(D.f[DIR_MP0])[kw ];//knw
+ real mfcbc = (D.f[DIR_P0P])[k];//[kte ];// + c1over54 ;(D.f[DIR_P0P])[k ];//kte
+ real mfaba = (D.f[DIR_M0M])[kbw];//[kbw ];// + c1over54 ;(D.f[DIR_M0M])[kbw];
+ real mfcba = (D.f[DIR_P0M])[kb];//[kbe ];// + c1over54 ;(D.f[DIR_P0M])[kb ];//kbe
+ real mfabc = (D.f[DIR_M0P])[kw];//[ktw ];// + c1over54 ;(D.f[DIR_M0P])[kw ];//ktw
+ real mfbcc = (D.f[DIR_0PP])[k];//[ktn ];// + c1over54 ;(D.f[DIR_0PP])[k ];//ktn
+ real mfbaa = (D.f[DIR_0MM])[kbs];//[kbs ];// + c1over54 ;(D.f[DIR_0MM])[kbs];
+ real mfbca = (D.f[DIR_0PM])[kb];//[kbn ];// + c1over54 ;(D.f[DIR_0PM])[kb ];//kbn
+ real mfbac = (D.f[DIR_0MP])[ks];//[kts ];// + c1over54 ;(D.f[DIR_0MP])[ks ];//kts
real mfbbb = (D.f[DIR_000])[k];//[kzero];// + c8over27 ;(D.f[DIR_000])[k ];//kzero
- real mfccc = (D.f[DIR_PPP])[k];//[ktne ];// + c1over216;(D.f[DIR_PPP ])[k ];//ktne
- real mfaac = (D.f[DIR_MMP])[ksw];//[ktsw ];// + c1over216;(D.f[DIR_MMP ])[ksw];//ktsw
- real mfcac = (D.f[DIR_PMP])[ks];//[ktse ];// + c1over216;(D.f[DIR_PMP ])[ks ];//ktse
- real mfacc = (D.f[DIR_MPP])[kw];//[ktnw ];// + c1over216;(D.f[DIR_MPP ])[kw ];//ktnw
- real mfcca = (D.f[DIR_PPM])[kb];//[kbne ];// + c1over216;(D.f[DIR_PPM ])[kb ];//kbne
- real mfaaa = (D.f[DIR_MMM])[kbsw];//[kbsw ];// + c1over216;(D.f[DIR_MMM ])[kbsw];
- real mfcaa = (D.f[DIR_PMM])[kbs];//[kbse ];// + c1over216;(D.f[DIR_PMM ])[kbs];//kbse
- real mfaca = (D.f[DIR_MPM])[kbw];//[kbnw ];// + c1over216;(D.f[DIR_MPM ])[kbw];//kbnw
+ real mfccc = (D.f[DIR_PPP])[k];//[ktne ];// + c1over216;(D.f[DIR_PPP])[k ];//ktne
+ real mfaac = (D.f[DIR_MMP])[ksw];//[ktsw ];// + c1over216;(D.f[DIR_MMP])[ksw];//ktsw
+ real mfcac = (D.f[DIR_PMP])[ks];//[ktse ];// + c1over216;(D.f[DIR_PMP])[ks ];//ktse
+ real mfacc = (D.f[DIR_MPP])[kw];//[ktnw ];// + c1over216;(D.f[DIR_MPP])[kw ];//ktnw
+ real mfcca = (D.f[DIR_PPM])[kb];//[kbne ];// + c1over216;(D.f[DIR_PPM])[kb ];//kbne
+ real mfaaa = (D.f[DIR_MMM])[kbsw];//[kbsw ];// + c1over216;(D.f[DIR_MMM])[kbsw];
+ real mfcaa = (D.f[DIR_PMM])[kbs];//[kbse ];// + c1over216;(D.f[DIR_PMM])[kbs];//kbse
+ real mfaca = (D.f[DIR_MPM])[kbw];//[kbnw ];// + c1over216;(D.f[DIR_MPM])[kbw];//kbnw
////////////////////////////////////////////////////////////////////////////////////
//slow
//real oMdrho = one - ((((mfccc+mfaaa) + (mfaca+mfcac)) + ((mfacc+mfcaa) + (mfaac+mfcca))) +
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/PorousMediaKernels/FluidFlow/Compressible/CumulantOne/PMCumulantOneCompSP27.cu b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/PorousMediaKernels/FluidFlow/Compressible/CumulantOne/PMCumulantOneCompSP27.cu
index 77527d5bedab08fdcacb3a103727ae25274b2aa4..43724f9165e2bb8dca1705ae0053612df92413ec 100644
--- a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/PorousMediaKernels/FluidFlow/Compressible/CumulantOne/PMCumulantOneCompSP27.cu
+++ b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/PorousMediaKernels/FluidFlow/Compressible/CumulantOne/PMCumulantOneCompSP27.cu
@@ -11,7 +11,7 @@ std::shared_ptr<PMCumulantOneCompSP27> PMCumulantOneCompSP27::getNewInstance(std
void PMCumulantOneCompSP27::run()
{
- int size_Mat = para->getParD(level)->numberOfNodes;
+ int size_Mat = (int)para->getParD(level)->numberOfNodes;
int numberOfThreads = para->getParD(level)->numberofthreads;
int Grid = (size_Mat / numberOfThreads) + 1;
@@ -30,7 +30,8 @@ void PMCumulantOneCompSP27::run()
dim3 threads(numberOfThreads, 1, 1);
for (int i = 0; i < pm.size(); i++) {
- LB_Kernel_PM_Cum_One_Comp_SP_27 << < grid, threads >> >(para->getParD(level)->omega,
+ LB_Kernel_PM_Cum_One_Comp_SP_27 <<< grid, threads >>>(
+ para->getParD(level)->omega,
para->getParD(level)->neighborX,
para->getParD(level)->neighborY,
para->getParD(level)->neighborZ,
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/PorousMediaKernels/FluidFlow/Compressible/CumulantOne/PMCumulantOneCompSP27_Device.cu b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/PorousMediaKernels/FluidFlow/Compressible/CumulantOne/PMCumulantOneCompSP27_Device.cu
index 89975d1663fb236295c22b81af4b0544ffc489bb..4f5f61f9d7a61fee8fd3438de5c588c861d8604c 100644
--- a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/PorousMediaKernels/FluidFlow/Compressible/CumulantOne/PMCumulantOneCompSP27_Device.cu
+++ b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/PorousMediaKernels/FluidFlow/Compressible/CumulantOne/PMCumulantOneCompSP27_Device.cu
@@ -11,7 +11,7 @@ __global__ void LB_Kernel_PM_Cum_One_Comp_SP_27(real omega,
unsigned int* neighborY,
unsigned int* neighborZ,
real* DDStart,
- int size_Mat,
+ unsigned long long numberOfLBnodes,
int level,
real* forces,
real porosity,
@@ -24,63 +24,63 @@ __global__ void LB_Kernel_PM_Cum_One_Comp_SP_27(real omega,
Distributions27 D;
if (EvenOrOdd == true)
{
- D.f[DIR_P00] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_M00] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0P0] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0M0] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00P] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00M] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_PP0] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_MM0] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_PM0] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_MP0] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_P0P] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_M0M] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_P0M] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_M0P] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0PP] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0MM] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0PM] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0MP] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_PPP] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_MMP] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_PMP] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_MPP] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_PPM] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_MMM] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_PMM] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_MPM] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DDStart[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_M00] = &DDStart[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DDStart[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DDStart[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00P] = &DDStart[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00M] = &DDStart[DIR_00M * numberOfLBnodes];
+ D.f[DIR_PP0] = &DDStart[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_MM0] = &DDStart[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DDStart[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DDStart[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_P0P] = &DDStart[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_M0M] = &DDStart[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DDStart[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DDStart[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0PP] = &DDStart[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0MM] = &DDStart[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DDStart[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DDStart[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DDStart[DIR_000 * numberOfLBnodes];
+ D.f[DIR_PPP] = &DDStart[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DDStart[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_PMP] = &DDStart[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_MPP] = &DDStart[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DDStart[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_MMM] = &DDStart[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_PMM] = &DDStart[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_MPM] = &DDStart[DIR_MPM * numberOfLBnodes];
}
else
{
- D.f[DIR_M00] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_P00] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0M0] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0P0] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00M] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00P] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_MM0] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_PP0] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_MP0] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_PM0] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_M0M] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_P0P] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_M0P] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_P0M] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0MM] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0PP] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0MP] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0PM] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_MMM] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_PPM] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_MPM] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_PMM] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_MMP] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_PPP] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_MPP] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_PMP] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_M00] = &DDStart[DIR_P00 * numberOfLBnodes];
+ D.f[DIR_P00] = &DDStart[DIR_M00 * numberOfLBnodes];
+ D.f[DIR_0M0] = &DDStart[DIR_0P0 * numberOfLBnodes];
+ D.f[DIR_0P0] = &DDStart[DIR_0M0 * numberOfLBnodes];
+ D.f[DIR_00M] = &DDStart[DIR_00P * numberOfLBnodes];
+ D.f[DIR_00P] = &DDStart[DIR_00M * numberOfLBnodes];
+ D.f[DIR_MM0] = &DDStart[DIR_PP0 * numberOfLBnodes];
+ D.f[DIR_PP0] = &DDStart[DIR_MM0 * numberOfLBnodes];
+ D.f[DIR_MP0] = &DDStart[DIR_PM0 * numberOfLBnodes];
+ D.f[DIR_PM0] = &DDStart[DIR_MP0 * numberOfLBnodes];
+ D.f[DIR_M0M] = &DDStart[DIR_P0P * numberOfLBnodes];
+ D.f[DIR_P0P] = &DDStart[DIR_M0M * numberOfLBnodes];
+ D.f[DIR_M0P] = &DDStart[DIR_P0M * numberOfLBnodes];
+ D.f[DIR_P0M] = &DDStart[DIR_M0P * numberOfLBnodes];
+ D.f[DIR_0MM] = &DDStart[DIR_0PP * numberOfLBnodes];
+ D.f[DIR_0PP] = &DDStart[DIR_0MM * numberOfLBnodes];
+ D.f[DIR_0MP] = &DDStart[DIR_0PM * numberOfLBnodes];
+ D.f[DIR_0PM] = &DDStart[DIR_0MP * numberOfLBnodes];
+ D.f[DIR_000] = &DDStart[DIR_000 * numberOfLBnodes];
+ D.f[DIR_MMM] = &DDStart[DIR_PPP * numberOfLBnodes];
+ D.f[DIR_PPM] = &DDStart[DIR_MMP * numberOfLBnodes];
+ D.f[DIR_MPM] = &DDStart[DIR_PMP * numberOfLBnodes];
+ D.f[DIR_PMM] = &DDStart[DIR_MPP * numberOfLBnodes];
+ D.f[DIR_MMP] = &DDStart[DIR_PPM * numberOfLBnodes];
+ D.f[DIR_PPP] = &DDStart[DIR_MMM * numberOfLBnodes];
+ D.f[DIR_MPP] = &DDStart[DIR_PMM * numberOfLBnodes];
+ D.f[DIR_PMP] = &DDStart[DIR_MPM * numberOfLBnodes];
}
////////////////////////////////////////////////////////////////////////////////
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/PorousMediaKernels/FluidFlow/Compressible/CumulantOne/PMCumulantOneCompSP27_Device.cuh b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/PorousMediaKernels/FluidFlow/Compressible/CumulantOne/PMCumulantOneCompSP27_Device.cuh
index 6533c604f32a478cdc6a097e4dd7d0b56e48150d..f2cf530b5d331c71d4a13bd5882a3657a3bbddea 100644
--- a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/PorousMediaKernels/FluidFlow/Compressible/CumulantOne/PMCumulantOneCompSP27_Device.cuh
+++ b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/PorousMediaKernels/FluidFlow/Compressible/CumulantOne/PMCumulantOneCompSP27_Device.cuh
@@ -9,7 +9,7 @@ __global__ void LB_Kernel_PM_Cum_One_Comp_SP_27(real omega,
unsigned int* neighborY,
unsigned int* neighborZ,
real* DDStart,
- int size_Mat,
+ unsigned long long numberOfLBnodes,
int level,
real* forces,
real porosity,
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/TurbulentViscosityKernels/FluidFlow/Compressible/CumulantK17chim/TurbulentViscosityCumulantK17CompChim.cu b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/TurbulentViscosityKernels/FluidFlow/Compressible/CumulantK17chim/TurbulentViscosityCumulantK17CompChim.cu
deleted file mode 100644
index a9d518d14a286ae3f6b565176969162994afa269..0000000000000000000000000000000000000000
--- a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/TurbulentViscosityKernels/FluidFlow/Compressible/CumulantK17chim/TurbulentViscosityCumulantK17CompChim.cu
+++ /dev/null
@@ -1,57 +0,0 @@
-#include "TurbulentViscosityCumulantK17CompChim.h"
-#include "cuda/CudaGrid.h"
-#include <logger/Logger.h>
-#include "Parameter/Parameter.h"
-#include "TurbulentViscosityCumulantK17CompChim_Device.cuh"
-
-template<TurbulenceModel turbulenceModel>
-std::shared_ptr< TurbulentViscosityCumulantK17CompChim<turbulenceModel> > TurbulentViscosityCumulantK17CompChim<turbulenceModel>::getNewInstance(std::shared_ptr<Parameter> para, int level)
-{
- return std::shared_ptr<TurbulentViscosityCumulantK17CompChim<turbulenceModel> >(new TurbulentViscosityCumulantK17CompChim<turbulenceModel>(para,level));
-}
-
-template<TurbulenceModel turbulenceModel>
-void TurbulentViscosityCumulantK17CompChim<turbulenceModel>::run()
-{
- vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(para->getParH(level)->numberofthreads, para->getParH(level)->numberOfNodes);
-
- LB_Kernel_TurbulentViscosityCumulantK17CompChim < turbulenceModel > <<< grid.grid, grid.threads >>>( para->getParD(level)->omega,
- para->getParD(level)->typeOfGridNode, para->getParD(level)->neighborX,
- para->getParD(level)->neighborY,
- para->getParD(level)->neighborZ,
- para->getParD(level)->distributions.f[0],
- para->getParD(level)->rho,
- para->getParD(level)->velocityX,
- para->getParD(level)->velocityY,
- para->getParD(level)->velocityZ,
- para->getParD(level)->turbViscosity,
- para->getSGSConstant(),
- (unsigned long)para->getParD(level)->numberOfNodes,
- level,
- para->getIsBodyForce(),
- para->getForcesDev(),
- para->getParD(level)->forceX_SP,
- para->getParD(level)->forceY_SP,
- para->getParD(level)->forceZ_SP,
- para->getQuadricLimitersDev(),
- para->getParD(level)->isEvenTimestep);
-
- getLastCudaError("LB_Kernel_TurbulentViscosityCumulantK17CompChim execution failed");
-}
-
-template<TurbulenceModel turbulenceModel>
-TurbulentViscosityCumulantK17CompChim<turbulenceModel>::TurbulentViscosityCumulantK17CompChim(std::shared_ptr<Parameter> para, int level)
-{
- this->para = para;
- this->level = level;
-
- myPreProcessorTypes.push_back(InitCompSP27);
-
- myKernelGroup = BasicKernel;
-
- VF_LOG_INFO("Using turbulence model: {}", turbulenceModel);
-}
-
-template class TurbulentViscosityCumulantK17CompChim<TurbulenceModel::AMD>;
-template class TurbulentViscosityCumulantK17CompChim<TurbulenceModel::Smagorinsky>;
-template class TurbulentViscosityCumulantK17CompChim<TurbulenceModel::QR>;
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/TurbulentViscosityKernels/FluidFlow/Compressible/CumulantK17chim/TurbulentViscosityCumulantK17CompChim.h b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/TurbulentViscosityKernels/FluidFlow/Compressible/CumulantK17chim/TurbulentViscosityCumulantK17CompChim.h
deleted file mode 100644
index 0d35b68c916e54c6ec6eeeacd7189fe4d9a33c10..0000000000000000000000000000000000000000
--- a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/TurbulentViscosityKernels/FluidFlow/Compressible/CumulantK17chim/TurbulentViscosityCumulantK17CompChim.h
+++ /dev/null
@@ -1,19 +0,0 @@
-#ifndef TurbulentViscosityCUMULANT_K17_COMP_CHIM_H
-#define TurbulentViscosityCUMULANT_K17_COMP_CHIM_H
-
-#include "Kernel/KernelImp.h"
-#include "Parameter/Parameter.h"
-
-template<TurbulenceModel turbulenceModel>
-class TurbulentViscosityCumulantK17CompChim : public KernelImp
-{
-public:
- static std::shared_ptr< TurbulentViscosityCumulantK17CompChim<turbulenceModel> > getNewInstance(std::shared_ptr< Parameter> para, int level);
- void run();
-
-private:
- TurbulentViscosityCumulantK17CompChim();
- TurbulentViscosityCumulantK17CompChim(std::shared_ptr<Parameter> para, int level);
-};
-
-#endif
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/TurbulentViscosityKernels/FluidFlow/Compressible/CumulantK17chim/TurbulentViscosityCumulantK17CompChim_Device.cu b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/TurbulentViscosityKernels/FluidFlow/Compressible/CumulantK17chim/TurbulentViscosityCumulantK17CompChim_Device.cu
deleted file mode 100644
index 32350b95107b68103af0f238fefe095882919092..0000000000000000000000000000000000000000
--- a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/TurbulentViscosityKernels/FluidFlow/Compressible/CumulantK17chim/TurbulentViscosityCumulantK17CompChim_Device.cu
+++ /dev/null
@@ -1,687 +0,0 @@
-//=======================================================================================
-// ____ ____ __ ______ __________ __ __ __ __
-// \ \ | | | | | _ \ |___ ___| | | | | / \ | |
-// \ \ | | | | | |_) | | | | | | | / \ | |
-// \ \ | | | | | _ / | | | | | | / /\ \ | |
-// \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
-// \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
-// \ \ | | ________________________________________________________________
-// \ \ | | | ______________________________________________________________|
-// \ \| | | | __ __ __ __ ______ _______
-// \ | | |_____ | | | | | | | | | _ \ / _____)
-// \ | | _____| | | | | | | | | | | \ \ \_______
-// \ | | | | |_____ | \_/ | | | | |_/ / _____ |
-// \ _____| |__| |________| \_______/ |__| |______/ (_______/
-//
-// This file is part of VirtualFluids. VirtualFluids is free software: you can
-// redistribute it and/or modify it under the terms of the GNU General Public
-// License as published by the Free Software Foundation, either version 3 of
-// the License, or (at your option) any later version.
-//
-// VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
-// ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
-// FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
-// for more details.
-//
-// You should have received a copy of the GNU General Public License along
-// with VirtualFluids (see COPYING.txt). If not, see <http://www.gnu.org/licenses/>.
-//
-//! \file TurbulentViscosityCumulantK17CompChim_Device.cu
-//! \author Henry Korb, Henrik Asmuth
-//! \date 16/05/2022
-//! \brief CumulantK17CompChim kernel by Martin Schönherr that inlcudes turbulent viscosity and other small mods.
-//!
-//! Additions to CumulantK17CompChim:
-//! - can incorporate local body force
-//! - when applying a local body force, the total round of error of forcing+bodyforce is saved and added in next time step
-//! - uses turbulent viscosity that is computed in separate kernel (as of now AMD)
-//! - saves macroscopic values (needed for instance for probes, AMD, and actuator models)
-//!
-//=======================================================================================
-/* Device code */
-#include "LBM/LB.h"
-#include "lbm/constants/D3Q27.h"
-#include <lbm/constants/NumericConstants.h>
-#include "Kernel/Utilities/DistributionHelper.cuh"
-
-#include "GPU/TurbulentViscosityInlines.cuh"
-
-using namespace vf::lbm::constant;
-using namespace vf::lbm::dir;
-#include "Kernel/Utilities/ChimeraTransformation.h"
-
-
-////////////////////////////////////////////////////////////////////////////////
-template<TurbulenceModel turbulenceModel>
-__global__ void LB_Kernel_TurbulentViscosityCumulantK17CompChim(
- real omega_in,
- uint* typeOfGridNode,
- uint* neighborX,
- uint* neighborY,
- uint* neighborZ,
- real* distributions,
- real* rho,
- real* vx,
- real* vy,
- real* vz,
- real* turbulentViscosity,
- real SGSconstant,
- unsigned long size_Mat,
- int level,
- bool bodyForce,
- real* forces,
- real* bodyForceX,
- real* bodyForceY,
- real* bodyForceZ,
- real* quadricLimiters,
- bool isEvenTimestep)
-{
- //////////////////////////////////////////////////////////////////////////
- //! Cumulant K17 Kernel is based on \ref
- //! <a href="https://doi.org/10.1016/j.jcp.2017.05.040"><b>[ M. Geier et al. (2017), DOI:10.1016/j.jcp.2017.05.040
- //! ]</b></a> and \ref <a href="https://doi.org/10.1016/j.jcp.2017.07.004"><b>[ M. Geier et al. (2017),
- //! DOI:10.1016/j.jcp.2017.07.004 ]</b></a>
- //!
- //! The cumulant kernel is executed in the following steps
- //!
- ////////////////////////////////////////////////////////////////////////////////
- //! - Get node index coordinates from threadIdx, blockIdx, blockDim and gridDim.
- //!
- const unsigned k_000 = vf::gpu::getNodeIndex();
-
- //////////////////////////////////////////////////////////////////////////
- // run for all indices in size_Mat and fluid nodes
- if ((k_000 < size_Mat) && (typeOfGridNode[k_000] == GEO_FLUID)) {
- //////////////////////////////////////////////////////////////////////////
- //! - Read distributions: style of reading and writing the distributions from/to stored arrays dependent on
- //! timestep is based on the esoteric twist algorithm \ref <a
- //! href="https://doi.org/10.3390/computation5020019"><b>[ M. Geier et al. (2017),
- //! DOI:10.3390/computation5020019 ]</b></a>
- //!
- Distributions27 dist = vf::gpu::getDistributionReferences27(distributions, size_Mat, isEvenTimestep);
-
- ////////////////////////////////////////////////////////////////////////////////
- //! - Set neighbor indices (necessary for indirect addressing)
- uint k_M00 = neighborX[k_000];
- uint k_0M0 = neighborY[k_000];
- uint k_00M = neighborZ[k_000];
- uint k_MM0 = neighborY[k_M00];
- uint k_M0M = neighborZ[k_M00];
- uint k_0MM = neighborZ[k_0M0];
- uint k_MMM = neighborZ[k_MM0];
- ////////////////////////////////////////////////////////////////////////////////////
- //! - Set local distributions
- //!
- real f_000 = (dist.f[DIR_000])[k_000];
- real f_P00 = (dist.f[DIR_P00])[k_000];
- real f_M00 = (dist.f[DIR_M00])[k_M00];
- real f_0P0 = (dist.f[DIR_0P0])[k_000];
- real f_0M0 = (dist.f[DIR_0M0])[k_0M0];
- real f_00P = (dist.f[DIR_00P])[k_000];
- real f_00M = (dist.f[DIR_00M])[k_00M];
- real f_PP0 = (dist.f[DIR_PP0])[k_000];
- real f_MM0 = (dist.f[DIR_MM0])[k_MM0];
- real f_PM0 = (dist.f[DIR_PM0])[k_0M0];
- real f_MP0 = (dist.f[DIR_MP0])[k_M00];
- real f_P0P = (dist.f[DIR_P0P])[k_000];
- real f_M0M = (dist.f[DIR_M0M])[k_M0M];
- real f_P0M = (dist.f[DIR_P0M])[k_00M];
- real f_M0P = (dist.f[DIR_M0P])[k_M00];
- real f_0PP = (dist.f[DIR_0PP])[k_000];
- real f_0MM = (dist.f[DIR_0MM])[k_0MM];
- real f_0PM = (dist.f[DIR_0PM])[k_00M];
- real f_0MP = (dist.f[DIR_0MP])[k_0M0];
- real f_PPP = (dist.f[DIR_PPP])[k_000];
- real f_MPP = (dist.f[DIR_MPP])[k_M00];
- real f_PMP = (dist.f[DIR_PMP])[k_0M0];
- real f_MMP = (dist.f[DIR_MMP])[k_MM0];
- real f_PPM = (dist.f[DIR_PPM])[k_00M];
- real f_MPM = (dist.f[DIR_MPM])[k_M0M];
- real f_PMM = (dist.f[DIR_PMM])[k_0MM];
- real f_MMM = (dist.f[DIR_MMM])[k_MMM];
-
- ////////////////////////////////////////////////////////////////////////////////////
- //! - Define aliases to use the same variable for the moments (m's):
- //!
- real& m_111 = f_000;
- real& m_211 = f_P00;
- real& m_011 = f_M00;
- real& m_121 = f_0P0;
- real& m_101 = f_0M0;
- real& m_112 = f_00P;
- real& m_110 = f_00M;
- real& m_221 = f_PP0;
- real& m_001 = f_MM0;
- real& m_201 = f_PM0;
- real& m_021 = f_MP0;
- real& m_212 = f_P0P;
- real& m_010 = f_M0M;
- real& m_210 = f_P0M;
- real& m_012 = f_M0P;
- real& m_122 = f_0PP;
- real& m_100 = f_0MM;
- real& m_120 = f_0PM;
- real& m_102 = f_0MP;
- real& m_222 = f_PPP;
- real& m_022 = f_MPP;
- real& m_202 = f_PMP;
- real& m_002 = f_MMP;
- real& m_220 = f_PPM;
- real& m_020 = f_MPM;
- real& m_200 = f_PMM;
- real& m_000 = f_MMM;
-
- //////////////////////////////////////////////////////(unsigned long)//////////////////////////////
- //! - Calculate density and velocity using pyramid summation for low round-off errors as in Eq. (J1)-(J3) \ref
- //! <a href="https://doi.org/10.1016/j.camwa.2015.05.001"><b>[ M. Geier et al. (2015),
- //! DOI:10.1016/j.camwa.2015.05.001 ]</b></a>
- //!
- real drho = ((((f_PPP + f_MMM) + (f_MPM + f_PMP)) + ((f_MPP + f_PMM) + (f_MMP + f_PPM))) +
- (((f_0MP + f_0PM) + (f_0MM + f_0PP)) + ((f_M0P + f_P0M) + (f_M0M + f_P0P)) +
- ((f_MP0 + f_PM0) + (f_MM0 + f_PP0))) +
- ((f_M00 + f_P00) + (f_0M0 + f_0P0) + (f_00M + f_00P))) +
- f_000;
-
- real oneOverRho = c1o1 / (c1o1 + drho);
-
- real vvx = ((((f_PPP - f_MMM) + (f_PMP - f_MPM)) + ((f_PMM - f_MPP) + (f_PPM - f_MMP))) +
- (((f_P0M - f_M0P) + (f_P0P - f_M0M)) + ((f_PM0 - f_MP0) + (f_PP0 - f_MM0))) + (f_P00 - f_M00)) *
- oneOverRho;
- real vvy = ((((f_PPP - f_MMM) + (f_MPM - f_PMP)) + ((f_MPP - f_PMM) + (f_PPM - f_MMP))) +
- (((f_0PM - f_0MP) + (f_0PP - f_0MM)) + ((f_MP0 - f_PM0) + (f_PP0 - f_MM0))) + (f_0P0 - f_0M0)) *
- oneOverRho;
- real vvz = ((((f_PPP - f_MMM) + (f_PMP - f_MPM)) + ((f_MPP - f_PMM) + (f_MMP - f_PPM))) +
- (((f_0MP - f_0PM) + (f_0PP - f_0MM)) + ((f_M0P - f_P0M) + (f_P0P - f_M0M))) + (f_00P - f_00M)) *
- oneOverRho;
-
- ////////////////////////////////////////////////////////////////////////////////////
- //! - Add half of the acceleration (body force) to the velocity as in Eq. (42) \ref
- //! <a href="https://doi.org/10.1016/j.camwa.2015.05.001"><b>[ M. Geier et al. (2015),
- //! DOI:10.1016/j.camwa.2015.05.001 ]</b></a>
- //!
- real factor = c1o1;
- for (size_t i = 1; i <= level; i++) {
- factor *= c2o1;
- }
-
- real fx = forces[0];
- real fy = forces[1];
- real fz = forces[2];
-
- if( bodyForce ){
- fx += bodyForceX[k_000];
- fy += bodyForceY[k_000];
- fz += bodyForceZ[k_000];
-
- real vx = vvx;
- real vy = vvy;
- real vz = vvz;
- real acc_x = fx * c1o2 / factor;
- real acc_y = fy * c1o2 / factor;
- real acc_z = fz * c1o2 / factor;
-
- vvx += acc_x;
- vvy += acc_y;
- vvz += acc_z;
-
- // // Reset body force. To be used when not using round-off correction.
- // bodyForceX[k] = 0.0f;
- // bodyForceY[k] = 0.0f;
- // bodyForceZ[k] = 0.0f;
-
- ////////////////////////////////////////////////////////////////////////////////////
- //!> Round-off correction
- //!
- //!> Similar to Kahan summation algorithm (https://en.wikipedia.org/wiki/Kahan_summation_algorithm)
- //!> Essentially computes the round-off error of the applied force and adds it in the next time step as a compensation.
- //!> Seems to be necesseary at very high Re boundary layers, where the forcing and velocity can
- //!> differ by several orders of magnitude.
- //!> \note 16/05/2022: Testing, still ongoing!
- //!
- bodyForceX[k_000] = (acc_x-(vvx-vx))*factor*c2o1;
- bodyForceY[k_000] = (acc_y-(vvy-vy))*factor*c2o1;
- bodyForceZ[k_000] = (acc_z-(vvz-vz))*factor*c2o1;
- }
- else{
- vvx += fx * c1o2 / factor;
- vvy += fy * c1o2 / factor;
- vvz += fz * c1o2 / factor;
- }
-
-
- ////////////////////////////////////////////////////////////////////////////////////
- // calculate the square of velocities for this lattice node
- real vx2 = vvx * vvx;
- real vy2 = vvy * vvy;
- real vz2 = vvz * vvz;
- ////////////////////////////////////////////////////////////////////////////////////
- //! - Set relaxation limiters for third order cumulants to default value \f$ \lambda=0.001 \f$ according to
- //! section 6 in \ref <a href="https://doi.org/10.1016/j.jcp.2017.05.040"><b>[ M. Geier et al. (2017),
- //! DOI:10.1016/j.jcp.2017.05.040 ]</b></a>
- //!
- real quadricLimitP = quadricLimiters[0];
- real quadricLimitM = quadricLimiters[1];
- real quadricLimitD = quadricLimiters[2];
- ////////////////////////////////////////////////////////////////////////////////////
- //! - Chimera transform from well conditioned distributions to central moments as defined in Appendix J in \ref
- //! <a href="https://doi.org/10.1016/j.camwa.2015.05.001"><b>[ M. Geier et al. (2015),
- //! DOI:10.1016/j.camwa.2015.05.001 ]</b></a> see also Eq. (6)-(14) in \ref <a
- //! href="https://doi.org/10.1016/j.jcp.2017.05.040"><b>[ M. Geier et al. (2017), DOI:10.1016/j.jcp.2017.05.040
- //! ]</b></a>
- //!
- ////////////////////////////////////////////////////////////////////////////////////
- // Z - Dir
- forwardInverseChimeraWithK(f_MMM, f_MM0, f_MMP, vvz, vz2, c36o1, c1o36);
- forwardInverseChimeraWithK(f_M0M, f_M00, f_M0P, vvz, vz2, c9o1, c1o9);
- forwardInverseChimeraWithK(f_MPM, f_MP0, f_MPP, vvz, vz2, c36o1, c1o36);
- forwardInverseChimeraWithK(f_0MM, f_0M0, f_0MP, vvz, vz2, c9o1, c1o9);
- forwardInverseChimeraWithK(f_00M, f_000, f_00P, vvz, vz2, c9o4, c4o9);
- forwardInverseChimeraWithK(f_0PM, f_0P0, f_0PP, vvz, vz2, c9o1, c1o9);
- forwardInverseChimeraWithK(f_PMM, f_PM0, f_PMP, vvz, vz2, c36o1, c1o36);
- forwardInverseChimeraWithK(f_P0M, f_P00, f_P0P, vvz, vz2, c9o1, c1o9);
- forwardInverseChimeraWithK(f_PPM, f_PP0, f_PPP, vvz, vz2, c36o1, c1o36);
-
- ////////////////////////////////////////////////////////////////////////////////////
- // Y - Dir
- forwardInverseChimeraWithK(f_MMM, f_M0M, f_MPM, vvy, vy2, c6o1, c1o6);
- forwardChimera( f_MM0, f_M00, f_MP0, vvy, vy2);
- forwardInverseChimeraWithK(f_MMP, f_M0P, f_MPP, vvy, vy2, c18o1, c1o18);
- forwardInverseChimeraWithK(f_0MM, f_00M, f_0PM, vvy, vy2, c3o2, c2o3);
- forwardChimera( f_0M0, f_000, f_0P0, vvy, vy2);
- forwardInverseChimeraWithK(f_0MP, f_00P, f_0PP, vvy, vy2, c9o2, c2o9);
- forwardInverseChimeraWithK(f_PMM, f_P0M, f_PPM, vvy, vy2, c6o1, c1o6);
- forwardChimera( f_PM0, f_P00, f_PP0, vvy, vy2);
- forwardInverseChimeraWithK(f_PMP, f_P0P, f_PPP, vvy, vy2, c18o1, c1o18);
-
- ////////////////////////////////////////////////////////////////////////////////////
- // X - Dir
- forwardInverseChimeraWithK(f_MMM, f_0MM, f_PMM, vvx, vx2, c1o1, c1o1);
- forwardChimera( f_M0M, f_00M, f_P0M, vvx, vx2);
- forwardInverseChimeraWithK(f_MPM, f_0PM, f_PPM, vvx, vx2, c3o1, c1o3);
- forwardChimera( f_MM0, f_0M0, f_PM0, vvx, vx2);
- forwardChimera( f_M00, f_000, f_P00, vvx, vx2);
- forwardChimera( f_MP0, f_0P0, f_PP0, vvx, vx2);
- forwardInverseChimeraWithK(f_MMP, f_0MP, f_PMP, vvx, vx2, c3o1, c1o3);
- forwardChimera( f_M0P, f_00P, f_P0P, vvx, vx2);
- forwardInverseChimeraWithK(f_MPP, f_0PP, f_PPP, vvx, vx2, c3o1, c1o9);
-
- ////////////////////////////////////////////////////////////////////////////////////
- //! - Setting relaxation rates for non-hydrodynamic cumulants (default values). Variable names and equations
- //! according to <a href="https://doi.org/10.1016/j.jcp.2017.05.040"><b>[ M. Geier et al. (2017),
- //! DOI:10.1016/j.jcp.2017.05.040 ]</b></a>
- //! => [NAME IN PAPER]=[NAME IN CODE]=[DEFAULT VALUE].
- //! - Trace of second order cumulants \f$ C_{200}+C_{020}+C_{002} \f$ used to adjust bulk
- //! viscosity:\f$\omega_2=OxxPyyPzz=1.0 \f$.
- //! - Third order cumulants \f$ C_{120}+C_{102}, C_{210}+C_{012}, C_{201}+C_{021} \f$: \f$ \omega_3=OxyyPxzz
- //! \f$ set according to Eq. (111) with simplifications assuming \f$ \omega_2=1.0\f$.
- //! - Third order cumulants \f$ C_{120}-C_{102}, C_{210}-C_{012}, C_{201}-C_{021} \f$: \f$ \omega_4 = OxyyMxzz
- //! \f$ set according to Eq. (112) with simplifications assuming \f$ \omega_2 = 1.0\f$.
- //! - Third order cumulants \f$ C_{111} \f$: \f$ \omega_5 = Oxyz \f$ set according to Eq. (113) with
- //! simplifications assuming \f$ \omega_2 = 1.0\f$ (modify for different bulk viscosity).
- //! - Fourth order cumulants \f$ C_{220}, C_{202}, C_{022}, C_{211}, C_{121}, C_{112} \f$: for simplification
- //! all set to the same default value \f$ \omega_6=\omega_7=\omega_8=O4=1.0 \f$.
- //! - Fifth order cumulants \f$ C_{221}, C_{212}, C_{122}\f$: \f$\omega_9=O5=1.0\f$.
- //! - Sixth order cumulant \f$ C_{222}\f$: \f$\omega_{10}=O6=1.0\f$.
- //!
- ////////////////////////////////////////////////////////////////////////////////////
- //! - Calculate modified omega with turbulent viscosity
- //!
- real omega = omega_in / (c1o1 + c3o1*omega_in*turbulentViscosity[k_000]);
- ////////////////////////////////////////////////////////////
- // 2.
- real OxxPyyPzz = c1o1;
- ////////////////////////////////////////////////////////////
- // 3.
- real OxyyPxzz = c8o1 * (-c2o1 + omega) * (c1o1 + c2o1 * omega) / (-c8o1 - c14o1 * omega + c7o1 * omega * omega);
- real OxyyMxzz =
- c8o1 * (-c2o1 + omega) * (-c7o1 + c4o1 * omega) / (c56o1 - c50o1 * omega + c9o1 * omega * omega);
- real Oxyz = c24o1 * (-c2o1 + omega) * (-c2o1 - c7o1 * omega + c3o1 * omega * omega) /
- (c48o1 + c152o1 * omega - c130o1 * omega * omega + c29o1 * omega * omega * omega);
- ////////////////////////////////////////////////////////////
- // 4.
- real O4 = c1o1;
- ////////////////////////////////////////////////////////////
- // 5.
- real O5 = c1o1;
- ////////////////////////////////////////////////////////////
- // 6.
- real O6 = c1o1;
-
- ////////////////////////////////////////////////////////////////////////////////////
- //! - A and DIR_00M: parameters for fourth order convergence of the diffusion term according to Eq. (115) and (116)
- //! <a href="https://doi.org/10.1016/j.jcp.2017.05.040"><b>[ M. Geier et al. (2017),
- //! DOI:10.1016/j.jcp.2017.05.040 ]</b></a> with simplifications assuming \f$ \omega_2 = 1.0 \f$ (modify for
- //! different bulk viscosity).
- //!
- real factorA = (c4o1 + c2o1 * omega - c3o1 * omega * omega) / (c2o1 - c7o1 * omega + c5o1 * omega * omega);
- real factorB = (c4o1 + c28o1 * omega - c14o1 * omega * omega) / (c6o1 - c21o1 * omega + c15o1 * omega * omega);
-
- ////////////////////////////////////////////////////////////////////////////////////
- //! - Compute cumulants from central moments according to Eq. (20)-(23) in
- //! <a href="https://doi.org/10.1016/j.jcp.2017.05.040"><b>[ M. Geier et al. (2017),
- //! DOI:10.1016/j.jcp.2017.05.040 ]</b></a>
- //!
- ////////////////////////////////////////////////////////////
- // 4.
- real c_211 = m_211 - ((m_200 + c1o3) * m_011 + c2o1 * m_110 * m_101) * oneOverRho;
- real c_121 = m_121 - ((m_020 + c1o3) * m_101 + c2o1 * m_110 * m_011) * oneOverRho;
- real c_112 = m_112 - ((m_002 + c1o3) * m_110 + c2o1 * m_101 * m_011) * oneOverRho;
-
- real c_220 = m_220 - (((m_200 * m_020 + c2o1 * m_110 * m_110) + c1o3 * (m_200 + m_020)) * oneOverRho - c1o9 * (drho * oneOverRho));
- real c_202 = m_202 - (((m_200 * m_002 + c2o1 * m_101 * m_101) + c1o3 * (m_200 + m_002)) * oneOverRho - c1o9 * (drho * oneOverRho));
- real c_022 = m_022 - (((m_002 * m_020 + c2o1 * m_011 * m_011) + c1o3 * (m_002 + m_020)) * oneOverRho - c1o9 * (drho * oneOverRho));
- ////////////////////////////////////////////////////////////
- // 5.
- real c_122 =
- m_122 - ((m_002 * m_120 + m_020 * m_102 + c4o1 * m_011 * m_111 + c2o1 * (m_101 * m_021 + m_110 * m_012)) +
- c1o3 * (m_120 + m_102)) *
- oneOverRho;
- real c_212 =
- m_212 - ((m_002 * m_210 + m_200 * m_012 + c4o1 * m_101 * m_111 + c2o1 * (m_011 * m_201 + m_110 * m_102)) +
- c1o3 * (m_210 + m_012)) *
- oneOverRho;
- real c_221 =
- m_221 - ((m_200 * m_021 + m_020 * m_201 + c4o1 * m_110 * m_111 + c2o1 * (m_101 * m_120 + m_011 * m_210)) +
- c1o3 * (m_021 + m_201)) *
- oneOverRho;
- ////////////////////////////////////////////////////////////
- // 6.
- real c_222 = m_222 + ((-c4o1 * m_111 * m_111 - (m_200 * m_022 + m_020 * m_202 + m_002 * m_220) -
- c4o1 * (m_011 * m_211 + m_101 * m_121 + m_110 * m_112) -
- c2o1 * (m_120 * m_102 + m_210 * m_012 + m_201 * m_021)) *
- oneOverRho +
- (c4o1 * (m_101 * m_101 * m_020 + m_011 * m_011 * m_200 + m_110 * m_110 * m_002) +
- c2o1 * (m_200 * m_020 * m_002) + c16o1 * m_110 * m_101 * m_011) *
- oneOverRho * oneOverRho -
- c1o3 * (m_022 + m_202 + m_220) * oneOverRho - c1o9 * (m_200 + m_020 + m_002) * oneOverRho +
- (c2o1 * (m_101 * m_101 + m_011 * m_011 + m_110 * m_110) +
- (m_002 * m_020 + m_002 * m_200 + m_020 * m_200) + c1o3 * (m_002 + m_020 + m_200)) *
- oneOverRho * oneOverRho * c2o3 +
- c1o27 * ((drho * drho - drho) * oneOverRho * oneOverRho));
-
- ////////////////////////////////////////////////////////////////////////////////////
- //! - Compute linear combinations of second and third order cumulants
- //!
- ////////////////////////////////////////////////////////////
- // 2.
- real mxxPyyPzz = m_200 + m_020 + m_002;
- real mxxMyy = m_200 - m_020;
- real mxxMzz = m_200 - m_002;
- ////////////////////////////////////////////////////////////
- // 3.
- real mxxyPyzz = m_210 + m_012;
- real mxxyMyzz = m_210 - m_012;
-
- real mxxzPyyz = m_201 + m_021;
- real mxxzMyyz = m_201 - m_021;
-
- real mxyyPxzz = m_120 + m_102;
- real mxyyMxzz = m_120 - m_102;
-
- ////////////////////////////////////////////////////////////////////////////////////
- // incl. correction
- ////////////////////////////////////////////////////////////
- //! - Compute velocity gradients from second order cumulants according to Eq. (27)-(32)
- //! <a href="https://doi.org/10.1016/j.jcp.2017.05.040"><b>[ M. Geier et al. (2017),
- //! DOI:10.1016/j.jcp.2017.05.040 ]</b></a> Further explanations of the correction in viscosity in Appendix H of
- //! <a href="https://doi.org/10.1016/j.camwa.2015.05.001"><b>[ M. Geier et al. (2015),
- //! DOI:10.1016/j.camwa.2015.05.001 ]</b></a> Note that the division by rho is omitted here as we need rho times
- //! the gradients later.
- //!
- real Dxy = -c3o1 * omega * m_110;
- real Dxz = -c3o1 * omega * m_101;
- real Dyz = -c3o1 * omega * m_011;
- real dxux = c1o2 * (-omega) * (mxxMyy + mxxMzz) + c1o2 * OxxPyyPzz * (m_000 - mxxPyyPzz);
- real dyuy = dxux + omega * c3o2 * mxxMyy;
- real dzuz = dxux + omega * c3o2 * mxxMzz;
-
- ////////////////////////////////////////////////////////////////////////////////////
- switch (turbulenceModel)
- {
- case TurbulenceModel::AMD: //AMD is computed in separate kernel
- break;
- case TurbulenceModel::Smagorinsky:
- turbulentViscosity[k_000] = calcTurbulentViscositySmagorinsky(SGSconstant, dxux, dyuy, dzuz, Dxy, Dxz , Dyz);
- break;
- case TurbulenceModel::QR:
- turbulentViscosity[k_000] = calcTurbulentViscosityQR(SGSconstant, dxux, dyuy, dzuz, Dxy, Dxz , Dyz);
- break;
- default:
- break;
- }
- ////////////////////////////////////////////////////////////
- //! - Relaxation of second order cumulants with correction terms according to Eq. (33)-(35) in
- //! <a href="https://doi.org/10.1016/j.jcp.2017.05.040"><b>[ M. Geier et al. (2017),
- //! DOI:10.1016/j.jcp.2017.05.040 ]</b></a>
- //!
- mxxPyyPzz += OxxPyyPzz * (m_000 - mxxPyyPzz) - c3o1 * (c1o1 - c1o2 * OxxPyyPzz) * (vx2 * dxux + vy2 * dyuy + vz2 * dzuz);
- mxxMyy += omega * (-mxxMyy) - c3o1 * (c1o1 + c1o2 * (-omega)) * (vx2 * dxux - vy2 * dyuy);
- mxxMzz += omega * (-mxxMzz) - c3o1 * (c1o1 + c1o2 * (-omega)) * (vx2 * dxux - vz2 * dzuz);
-
- ////////////////////////////////////////////////////////////////////////////////////
- ////no correction
- // mxxPyyPzz += OxxPyyPzz*(mfaaa - mxxPyyPzz);
- // mxxMyy += -(-omega) * (-mxxMyy);
- // mxxMzz += -(-omega) * (-mxxMzz);
- //////////////////////////////////////////////////////////////////////////
- m_011 += omega * (-m_011);
- m_101 += omega * (-m_101);
- m_110 += omega * (-m_110);
-
- ////////////////////////////////////////////////////////////////////////////////////
- // relax
- //////////////////////////////////////////////////////////////////////////
- // incl. limiter
- //! - Relaxation of third order cumulants including limiter according to Eq. (116)-(123)
- //! <a href="https://doi.org/10.1016/j.jcp.2017.05.040"><b>[ M. Geier et al. (2017),
- //! DOI:10.1016/j.jcp.2017.05.040 ]</b></a>
- //!
- real wadjust = Oxyz + (c1o1 - Oxyz) * abs(m_111) / (abs(m_111) + quadricLimitD);
- m_111 += wadjust * (-m_111);
- wadjust = OxyyPxzz + (c1o1 - OxyyPxzz) * abs(mxxyPyzz) / (abs(mxxyPyzz) + quadricLimitP);
- mxxyPyzz += wadjust * (-mxxyPyzz);
- wadjust = OxyyMxzz + (c1o1 - OxyyMxzz) * abs(mxxyMyzz) / (abs(mxxyMyzz) + quadricLimitM);
- mxxyMyzz += wadjust * (-mxxyMyzz);
- wadjust = OxyyPxzz + (c1o1 - OxyyPxzz) * abs(mxxzPyyz) / (abs(mxxzPyyz) + quadricLimitP);
- mxxzPyyz += wadjust * (-mxxzPyyz);
- wadjust = OxyyMxzz + (c1o1 - OxyyMxzz) * abs(mxxzMyyz) / (abs(mxxzMyyz) + quadricLimitM);
- mxxzMyyz += wadjust * (-mxxzMyyz);
- wadjust = OxyyPxzz + (c1o1 - OxyyPxzz) * abs(mxyyPxzz) / (abs(mxyyPxzz) + quadricLimitP);
- mxyyPxzz += wadjust * (-mxyyPxzz);
- wadjust = OxyyMxzz + (c1o1 - OxyyMxzz) * abs(mxyyMxzz) / (abs(mxyyMxzz) + quadricLimitM);
- mxyyMxzz += wadjust * (-mxyyMxzz);
- //////////////////////////////////////////////////////////////////////////
- // no limiter
- // mfbbb += OxyyMxzz * (-mfbbb);
- // mxxyPyzz += OxyyPxzz * (-mxxyPyzz);
- // mxxyMyzz += OxyyMxzz * (-mxxyMyzz);
- // mxxzPyyz += OxyyPxzz * (-mxxzPyyz);
- // mxxzMyyz += OxyyMxzz * (-mxxzMyyz);
- // mxyyPxzz += OxyyPxzz * (-mxyyPxzz);
- // mxyyMxzz += OxyyMxzz * (-mxyyMxzz);
-
- ////////////////////////////////////////////////////////////////////////////////////
- //! - Compute inverse linear combinations of second and third order cumulants
- //!
- m_200 = c1o3 * (mxxMyy + mxxMzz + mxxPyyPzz);
- m_020 = c1o3 * (-c2o1 * mxxMyy + mxxMzz + mxxPyyPzz);
- m_002 = c1o3 * (mxxMyy - c2o1 * mxxMzz + mxxPyyPzz);
-
- m_210 = ( mxxyMyzz + mxxyPyzz) * c1o2;
- m_012 = (-mxxyMyzz + mxxyPyzz) * c1o2;
- m_201 = ( mxxzMyyz + mxxzPyyz) * c1o2;
- m_021 = (-mxxzMyyz + mxxzPyyz) * c1o2;
- m_120 = ( mxyyMxzz + mxyyPxzz) * c1o2;
- m_102 = (-mxyyMxzz + mxyyPxzz) * c1o2;
- //////////////////////////////////////////////////////////////////////////
-
- //////////////////////////////////////////////////////////////////////////
- // 4.
- // no limiter
- //! - Relax fourth order cumulants to modified equilibrium for fourth order convergence of diffusion according
- //! to Eq. (43)-(48) <a href="https://doi.org/10.1016/j.jcp.2017.05.040"><b>[ M. Geier et al. (2017),
- //! DOI:10.1016/j.jcp.2017.05.040 ]</b></a>
- //!
- c_022 = -O4 * (c1o1 / omega - c1o2) * (dyuy + dzuz) * c2o3 * factorA + (c1o1 - O4) * (c_022);
- c_202 = -O4 * (c1o1 / omega - c1o2) * (dxux + dzuz) * c2o3 * factorA + (c1o1 - O4) * (c_202);
- c_220 = -O4 * (c1o1 / omega - c1o2) * (dyuy + dxux) * c2o3 * factorA + (c1o1 - O4) * (c_220);
- c_112 = -O4 * (c1o1 / omega - c1o2) * Dxy * c1o3 * factorB + (c1o1 - O4) * (c_112);
- c_121 = -O4 * (c1o1 / omega - c1o2) * Dxz * c1o3 * factorB + (c1o1 - O4) * (c_121);
- c_211 = -O4 * (c1o1 / omega - c1o2) * Dyz * c1o3 * factorB + (c1o1 - O4) * (c_211);
-
-
- //////////////////////////////////////////////////////////////////////////
- // 5.
- c_122 += O5 * (-c_122);
- c_212 += O5 * (-c_212);
- c_221 += O5 * (-c_221);
-
- //////////////////////////////////////////////////////////////////////////
- // 6.
- c_222 += O6 * (-c_222);
-
- ////////////////////////////////////////////////////////////////////////////////////
- //! - Compute central moments from post collision cumulants according to Eq. (53)-(56) in
- //! <a href="https://doi.org/10.1016/j.jcp.2017.05.040"><b>[ M. Geier et al. (2017),
- //! DOI:10.1016/j.jcp.2017.05.040 ]</b></a>
- //!
-
- //////////////////////////////////////////////////////////////////////////
- // 4.
- m_211 = c_211 + c1o3 * ((c3o1 * m_200 + c1o1) * m_011 + c6o1 * m_110 * m_101) * oneOverRho;
- m_121 = c_121 + c1o3 * ((c3o1 * m_020 + c1o1) * m_101 + c6o1 * m_110 * m_011) * oneOverRho;
- m_112 = c_112 + c1o3 * ((c3o1 * m_002 + c1o1) * m_110 + c6o1 * m_101 * m_011) * oneOverRho;
-
- m_220 =
- c_220 + (((m_200 * m_020 + c2o1 * m_110 * m_110) * c9o1 + c3o1 * (m_200 + m_020)) * oneOverRho - (drho * oneOverRho)) * c1o9;
- m_202 =
- c_202 + (((m_200 * m_002 + c2o1 * m_101 * m_101) * c9o1 + c3o1 * (m_200 + m_002)) * oneOverRho - (drho * oneOverRho)) * c1o9;
- m_022 =
- c_022 + (((m_002 * m_020 + c2o1 * m_011 * m_011) * c9o1 + c3o1 * (m_002 + m_020)) * oneOverRho - (drho * oneOverRho)) * c1o9;
-
- //////////////////////////////////////////////////////////////////////////
- // 5.
- m_122 = c_122 + c1o3 *
- (c3o1 * (m_002 * m_120 + m_020 * m_102 + c4o1 * m_011 * m_111 + c2o1 * (m_101 * m_021 + m_110 * m_012)) +
- (m_120 + m_102)) * oneOverRho;
- m_212 = c_212 + c1o3 *
- (c3o1 * (m_002 * m_210 + m_200 * m_012 + c4o1 * m_101 * m_111 + c2o1 * (m_011 * m_201 + m_110 * m_102)) +
- (m_210 + m_012)) * oneOverRho;
- m_221 = c_221 + c1o3 *
- (c3o1 * (m_200 * m_021 + m_020 * m_201 + c4o1 * m_110 * m_111 + c2o1 * (m_101 * m_120 + m_011 * m_210)) +
- (m_021 + m_201)) * oneOverRho;
-
- //////////////////////////////////////////////////////////////////////////
- // 6.
- m_222 = c_222 - ((-c4o1 * m_111 * m_111 - (m_200 * m_022 + m_020 * m_202 + m_002 * m_220) -
- c4o1 * (m_011 * m_211 + m_101 * m_121 + m_110 * m_112) -
- c2o1 * (m_120 * m_102 + m_210 * m_012 + m_201 * m_021)) *
- oneOverRho +
- (c4o1 * (m_101 * m_101 * m_020 + m_011 * m_011 * m_200 + m_110 * m_110 * m_002) +
- c2o1 * (m_200 * m_020 * m_002) + c16o1 * m_110 * m_101 * m_011) *
- oneOverRho * oneOverRho -
- c1o3 * (m_022 + m_202 + m_220) * oneOverRho - c1o9 * (m_200 + m_020 + m_002) * oneOverRho +
- (c2o1 * (m_101 * m_101 + m_011 * m_011 + m_110 * m_110) +
- (m_002 * m_020 + m_002 * m_200 + m_020 * m_200) + c1o3 * (m_002 + m_020 + m_200)) *
- oneOverRho * oneOverRho * c2o3 +
- c1o27 * ((drho * drho - drho) * oneOverRho * oneOverRho));
-
- ////////////////////////////////////////////////////////////////////////////////////
- //! - Add acceleration (body force) to first order cumulants according to Eq. (85)-(87) in
- //! <a href="https://doi.org/10.1016/j.camwa.2015.05.001"><b>[ M. Geier et al. (2015),
- //! DOI:10.1016/j.camwa.2015.05.001 ]</b></a>
- //!
- m_100 = -m_100;
- m_010 = -m_010;
- m_001 = -m_001;
-
- //Write to array here to distribute read/write
- rho[k_000] = drho;
- vx[k_000] = vvx;
- vy[k_000] = vvy;
- vz[k_000] = vvz;
-
- ////////////////////////////////////////////////////////////////////////////////////
- //! - Chimera transform from central moments to well conditioned distributions as defined in Appendix J in
- //! <a href="https://doi.org/10.1016/j.camwa.2015.05.001"><b>[ M. Geier et al. (2015),
- //! DOI:10.1016/j.camwa.2015.05.001 ]</b></a> see also Eq. (88)-(96) in <a
- //! href="https://doi.org/10.1016/j.jcp.2017.05.040"><b>[ M. Geier et al. (2017), DOI:10.1016/j.jcp.2017.05.040
- //! ]</b></a>
- //!
- ////////////////////////////////////////////////////////////////////////////////////
- // X - Dir
- backwardInverseChimeraWithK(m_000, m_100, m_200, vvx, vx2, c1o1, c1o1);
- backwardChimera( m_010, m_110, m_210, vvx, vx2);
- backwardInverseChimeraWithK(m_020, m_120, m_220, vvx, vx2, c3o1, c1o3);
- backwardChimera( m_001, m_101, m_201, vvx, vx2);
- backwardChimera( m_011, m_111, m_211, vvx, vx2);
- backwardChimera( m_021, m_121, m_221, vvx, vx2);
- backwardInverseChimeraWithK(m_002, m_102, m_202, vvx, vx2, c3o1, c1o3);
- backwardChimera( m_012, m_112, m_212, vvx, vx2);
- backwardInverseChimeraWithK(m_022, m_122, m_222, vvx, vx2, c9o1, c1o9);
-
- ////////////////////////////////////////////////////////////////////////////////////
- // Y - Dir
- backwardInverseChimeraWithK(m_000, m_010, m_020, vvy, vy2, c6o1, c1o6);
- backwardChimera( m_001, m_011, m_021, vvy, vy2);
- backwardInverseChimeraWithK(m_002, m_012, m_022, vvy, vy2, c18o1, c1o18);
- backwardInverseChimeraWithK(m_100, m_110, m_120, vvy, vy2, c3o2, c2o3);
- backwardChimera( m_101, m_111, m_121, vvy, vy2);
- backwardInverseChimeraWithK(m_102, m_112, m_122, vvy, vy2, c9o2, c2o9);
- backwardInverseChimeraWithK(m_200, m_210, m_220, vvy, vy2, c6o1, c1o6);
- backwardChimera( m_201, m_211, m_221, vvy, vy2);
- backwardInverseChimeraWithK(m_202, m_212, m_222, vvy, vy2, c18o1, c1o18);
-
- ////////////////////////////////////////////////////////////////////////////////////
- // Z - Dir
- backwardInverseChimeraWithK(m_000, m_001, m_002, vvz, vz2, c36o1, c1o36);
- backwardInverseChimeraWithK(m_010, m_011, m_012, vvz, vz2, c9o1, c1o9);
- backwardInverseChimeraWithK(m_020, m_021, m_022, vvz, vz2, c36o1, c1o36);
- backwardInverseChimeraWithK(m_100, m_101, m_102, vvz, vz2, c9o1, c1o9);
- backwardInverseChimeraWithK(m_110, m_111, m_112, vvz, vz2, c9o4, c4o9);
- backwardInverseChimeraWithK(m_120, m_121, m_122, vvz, vz2, c9o1, c1o9);
- backwardInverseChimeraWithK(m_200, m_201, m_202, vvz, vz2, c36o1, c1o36);
- backwardInverseChimeraWithK(m_210, m_211, m_212, vvz, vz2, c9o1, c1o9);
- backwardInverseChimeraWithK(m_220, m_221, m_222, vvz, vz2, c36o1, c1o36);
-
- ////////////////////////////////////////////////////////////////////////////////////
- //! - Write distributions: style of reading and writing the distributions from/to
- //! stored arrays dependent on timestep is based on the esoteric twist algorithm
- //! <a href="https://doi.org/10.3390/computation5020019"><b>[ M. Geier et al. (2017),
- //! DOI:10.3390/computation5020019 ]</b></a>
- //!
- (dist.f[DIR_P00])[k_000] = f_M00;
- (dist.f[DIR_M00])[k_M00] = f_P00;
- (dist.f[DIR_0P0])[k_000] = f_0M0;
- (dist.f[DIR_0M0])[k_0M0] = f_0P0;
- (dist.f[DIR_00P])[k_000] = f_00M;
- (dist.f[DIR_00M])[k_00M] = f_00P;
- (dist.f[DIR_PP0])[k_000] = f_MM0;
- (dist.f[DIR_MM0])[k_MM0] = f_PP0;
- (dist.f[DIR_PM0])[k_0M0] = f_MP0;
- (dist.f[DIR_MP0])[k_M00] = f_PM0;
- (dist.f[DIR_P0P])[k_000] = f_M0M;
- (dist.f[DIR_M0M])[k_M0M] = f_P0P;
- (dist.f[DIR_P0M])[k_00M] = f_M0P;
- (dist.f[DIR_M0P])[k_M00] = f_P0M;
- (dist.f[DIR_0PP])[k_000] = f_0MM;
- (dist.f[DIR_0MM])[k_0MM] = f_0PP;
- (dist.f[DIR_0PM])[k_00M] = f_0MP;
- (dist.f[DIR_0MP])[k_0M0] = f_0PM;
- (dist.f[DIR_000])[k_000] = f_000;
- (dist.f[DIR_PPP])[k_000] = f_MMM;
- (dist.f[DIR_PMP])[k_0M0] = f_MPM;
- (dist.f[DIR_PPM])[k_00M] = f_MMP;
- (dist.f[DIR_PMM])[k_0MM] = f_MPP;
- (dist.f[DIR_MPP])[k_M00] = f_PMM;
- (dist.f[DIR_MMP])[k_MM0] = f_PPM;
- (dist.f[DIR_MPM])[k_M0M] = f_PMP;
- (dist.f[DIR_MMM])[k_MMM] = f_PPP;
- }
-}
-
-template __global__ void LB_Kernel_TurbulentViscosityCumulantK17CompChim < TurbulenceModel::AMD > ( real omega_in, uint* typeOfGridNode, uint* neighborX, uint* neighborY, uint* neighborZ, real* distributions, real* rho, real* vx, real* vy, real* vz, real* turbulentViscosity, real SGSconstant, unsigned long size_Mat, int level, bool bodyForce, real* forces, real* bodyForceX, real* bodyForceY, real* bodyForceZ, real* quadricLimiters, bool isEvenTimestep);
-
-template __global__ void LB_Kernel_TurbulentViscosityCumulantK17CompChim < TurbulenceModel::Smagorinsky > ( real omega_in, uint* typeOfGridNode, uint* neighborX, uint* neighborY, uint* neighborZ, real* distributions, real* rho, real* vx, real* vy, real* vz, real* turbulentViscosity, real SGSconstant, unsigned long size_Mat, int level, bool bodyForce, real* forces, real* bodyForceX, real* bodyForceY, real* bodyForceZ, real* quadricLimiters, bool isEvenTimestep);
-
-template __global__ void LB_Kernel_TurbulentViscosityCumulantK17CompChim < TurbulenceModel::QR > ( real omega_in, uint* typeOfGridNode, uint* neighborX, uint* neighborY, uint* neighborZ, real* distributions, real* rho, real* vx, real* vy, real* vz, real* turbulentViscosity, real SGSconstant, unsigned long size_Mat, int level, bool bodyForce, real* forces, real* bodyForceX, real* bodyForceY, real* bodyForceZ, real* quadricLimiters, bool isEvenTimestep);
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/TurbulentViscosityKernels/FluidFlow/Compressible/CumulantK17chim/TurbulentViscosityCumulantK17CompChim_Device.cuh b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/TurbulentViscosityKernels/FluidFlow/Compressible/CumulantK17chim/TurbulentViscosityCumulantK17CompChim_Device.cuh
deleted file mode 100644
index 5ef37557399f263d25edf03b02b00f6a03c6e1cb..0000000000000000000000000000000000000000
--- a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/TurbulentViscosityKernels/FluidFlow/Compressible/CumulantK17chim/TurbulentViscosityCumulantK17CompChim_Device.cuh
+++ /dev/null
@@ -1,29 +0,0 @@
-#ifndef LB_Kernel_TURBULENT_VISCOSITY_CUMULANT_K17_COMP_CHIM_H
-#define LB_Kernel_TURBULENT_VISCOSITY_CUMULANT_K17_COMP_CHIM_H
-
-#include <DataTypes.h>
-#include <curand.h>
-
-template< TurbulenceModel turbulenceModel > __global__ void LB_Kernel_TurbulentViscosityCumulantK17CompChim(
- real omega_in,
- uint* typeOfGridNode,
- uint* neighborX,
- uint* neighborY,
- uint* neighborZ,
- real* distributions,
- real* rho,
- real* vx,
- real* vy,
- real* vz,
- real* turbulentViscosity,
- real SGSconstant,
- unsigned long size_Mat,
- int level,
- bool bodyForce,
- real* forces,
- real* bodyForceX,
- real* bodyForceY,
- real* bodyForceZ,
- real* quadricLimiters,
- bool isEvenTimestep);
-#endif
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/TurbulentViscosityKernels/FluidFlow/Compressible/TurbulentViscosityFluidFlowCompStrategy.cpp b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/TurbulentViscosityKernels/FluidFlow/Compressible/TurbulentViscosityFluidFlowCompStrategy.cpp
deleted file mode 100644
index f3615a89994f0ca1fafdc1eda905d3c3b615d478..0000000000000000000000000000000000000000
--- a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/TurbulentViscosityKernels/FluidFlow/Compressible/TurbulentViscosityFluidFlowCompStrategy.cpp
+++ /dev/null
@@ -1,23 +0,0 @@
-#include "TurbulentViscosityFluidFlowCompStrategy.h"
-
-#include "Parameter/Parameter.h"
-
-std::shared_ptr<TurbulentViscosityFluidFlowCompStrategy> TurbulentViscosityFluidFlowCompStrategy::getInstance()
-{
- static std::shared_ptr<TurbulentViscosityFluidFlowCompStrategy> uniqueInstance;
- if (!uniqueInstance)
- uniqueInstance = std::shared_ptr<TurbulentViscosityFluidFlowCompStrategy>(new TurbulentViscosityFluidFlowCompStrategy());
- return uniqueInstance;
-}
-
-bool TurbulentViscosityFluidFlowCompStrategy::checkParameter(std::shared_ptr<Parameter> para)
-{
- if (!para->getUseTurbulentViscosity())
- return false;
- else if (!para->getCompOn())
- return false;
- else
- return true;
-}
-
-TurbulentViscosityFluidFlowCompStrategy::TurbulentViscosityFluidFlowCompStrategy() {}
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/TurbulentViscosityKernels/FluidFlow/Compressible/TurbulentViscosityFluidFlowCompStrategy.h b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/TurbulentViscosityKernels/FluidFlow/Compressible/TurbulentViscosityFluidFlowCompStrategy.h
deleted file mode 100644
index 95eff17777f7f0d1c3e05fe1b0d93892a88646a4..0000000000000000000000000000000000000000
--- a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/TurbulentViscosityKernels/FluidFlow/Compressible/TurbulentViscosityFluidFlowCompStrategy.h
+++ /dev/null
@@ -1,18 +0,0 @@
-#ifndef AMD_FLUID_FLOW_COMP_STRATEGY_H
-#define AMD_FLUID_FLOW_COMP_STRATEGY_H
-
-#include "Kernel/Utilities/CheckParameterStrategy/CheckParameterStrategy.h"
-
-
-class TurbulentViscosityFluidFlowCompStrategy : public CheckParameterStrategy
-{
-public:
- static std::shared_ptr<TurbulentViscosityFluidFlowCompStrategy> getInstance();
-
- bool checkParameter(std::shared_ptr<Parameter> para);
-
-private:
- TurbulentViscosityFluidFlowCompStrategy();
-
-};
-#endif
\ No newline at end of file
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/WaleKernels/FluidFlow/Compressible/CumulantK15/WaleCumulantK15Comp.cu b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/WaleKernels/FluidFlow/Compressible/CumulantK15/WaleCumulantK15Comp.cu
index cfcb70cd2bd6f3cc8ec4349650c44b7d3b0619fc..2b8a7d61e8966e2ed00022986311ae68ac0ca6d6 100644
--- a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/WaleKernels/FluidFlow/Compressible/CumulantK15/WaleCumulantK15Comp.cu
+++ b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/WaleKernels/FluidFlow/Compressible/CumulantK15/WaleCumulantK15Comp.cu
@@ -10,7 +10,7 @@ std::shared_ptr<WaleCumulantK15Comp> WaleCumulantK15Comp::getNewInstance(std::sh
void WaleCumulantK15Comp::run()
{
- int size_Mat = para->getParD(level)->numberOfNodes;
+ int size_Mat = (int)para->getParD(level)->numberOfNodes;
int numberOfThreads = para->getParD(level)->numberofthreads;
int Grid = (size_Mat / numberOfThreads) + 1;
@@ -28,22 +28,23 @@ void WaleCumulantK15Comp::run()
dim3 grid(Grid1, Grid2, 1);
dim3 threads(numberOfThreads, 1, 1);
- LB_Kernel_WaleCumulantK15Comp << < grid, threads >> >( para->getParD(level)->omega,
- para->getParD(level)->typeOfGridNode,
- para->getParD(level)->neighborX,
- para->getParD(level)->neighborY,
- para->getParD(level)->neighborZ,
- para->getParD(level)->neighborInverse,
- para->getParD(level)->velocityX,
- para->getParD(level)->velocityY,
- para->getParD(level)->velocityZ,
- para->getParD(level)->distributions.f[0],
- para->getParD(level)->turbViscosity,
- para->getParD(level)->numberOfNodes,
- level,
- para->getTimestepOfCoarseLevel(),
- para->getForcesDev(),
- para->getParD(level)->isEvenTimestep);
+ LB_Kernel_WaleCumulantK15Comp <<< grid, threads >>>(
+ para->getParD(level)->omega,
+ para->getParD(level)->typeOfGridNode,
+ para->getParD(level)->neighborX,
+ para->getParD(level)->neighborY,
+ para->getParD(level)->neighborZ,
+ para->getParD(level)->neighborInverse,
+ para->getParD(level)->velocityX,
+ para->getParD(level)->velocityY,
+ para->getParD(level)->velocityZ,
+ para->getParD(level)->distributions.f[0],
+ para->getParD(level)->turbViscosity,
+ para->getParD(level)->numberOfNodes,
+ level,
+ para->getTimestepOfCoarseLevel(),
+ para->getForcesDev(),
+ para->getParD(level)->isEvenTimestep);
getLastCudaError("LB_Kernel_WaleCumulantK15Comp execution failed");
}
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/WaleKernels/FluidFlow/Compressible/CumulantK15/WaleCumulantK15Comp_Device.cu b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/WaleKernels/FluidFlow/Compressible/CumulantK15/WaleCumulantK15Comp_Device.cu
index 3da25060e6c82ea685a1659fecc8cf66eeaf44c4..a7018d1246c0832753df144ffbf2625b55f5508e 100644
--- a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/WaleKernels/FluidFlow/Compressible/CumulantK15/WaleCumulantK15Comp_Device.cu
+++ b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/WaleKernels/FluidFlow/Compressible/CumulantK15/WaleCumulantK15Comp_Device.cu
@@ -46,63 +46,63 @@ __global__ void LB_Kernel_WaleCumulantK15Comp(
Distributions27 D;
if (EvenOrOdd==true)
{
- D.f[DIR_P00 ] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DDStart[DIR_P00 * size_Mat];
+ D.f[DIR_M00] = &DDStart[DIR_M00 * size_Mat];
+ D.f[DIR_0P0] = &DDStart[DIR_0P0 * size_Mat];
+ D.f[DIR_0M0] = &DDStart[DIR_0M0 * size_Mat];
+ D.f[DIR_00P] = &DDStart[DIR_00P * size_Mat];
+ D.f[DIR_00M] = &DDStart[DIR_00M * size_Mat];
+ D.f[DIR_PP0] = &DDStart[DIR_PP0 * size_Mat];
+ D.f[DIR_MM0] = &DDStart[DIR_MM0 * size_Mat];
+ D.f[DIR_PM0] = &DDStart[DIR_PM0 * size_Mat];
+ D.f[DIR_MP0] = &DDStart[DIR_MP0 * size_Mat];
+ D.f[DIR_P0P] = &DDStart[DIR_P0P * size_Mat];
+ D.f[DIR_M0M] = &DDStart[DIR_M0M * size_Mat];
+ D.f[DIR_P0M] = &DDStart[DIR_P0M * size_Mat];
+ D.f[DIR_M0P] = &DDStart[DIR_M0P * size_Mat];
+ D.f[DIR_0PP] = &DDStart[DIR_0PP * size_Mat];
+ D.f[DIR_0MM] = &DDStart[DIR_0MM * size_Mat];
+ D.f[DIR_0PM] = &DDStart[DIR_0PM * size_Mat];
+ D.f[DIR_0MP] = &DDStart[DIR_0MP * size_Mat];
+ D.f[DIR_000] = &DDStart[DIR_000 * size_Mat];
+ D.f[DIR_PPP] = &DDStart[DIR_PPP * size_Mat];
+ D.f[DIR_MMP] = &DDStart[DIR_MMP * size_Mat];
+ D.f[DIR_PMP] = &DDStart[DIR_PMP * size_Mat];
+ D.f[DIR_MPP] = &DDStart[DIR_MPP * size_Mat];
+ D.f[DIR_PPM] = &DDStart[DIR_PPM * size_Mat];
+ D.f[DIR_MMM] = &DDStart[DIR_MMM * size_Mat];
+ D.f[DIR_PMM] = &DDStart[DIR_PMM * size_Mat];
+ D.f[DIR_MPM] = &DDStart[DIR_MPM * size_Mat];
}
else
{
- D.f[DIR_M00 ] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_MMM ] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_PPM ] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_MPM ] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_PMM ] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_MMP ] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_PPP ] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_MPP ] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_PMP ] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_M00] = &DDStart[DIR_P00 * size_Mat];
+ D.f[DIR_P00] = &DDStart[DIR_M00 * size_Mat];
+ D.f[DIR_0M0] = &DDStart[DIR_0P0 * size_Mat];
+ D.f[DIR_0P0] = &DDStart[DIR_0M0 * size_Mat];
+ D.f[DIR_00M] = &DDStart[DIR_00P * size_Mat];
+ D.f[DIR_00P] = &DDStart[DIR_00M * size_Mat];
+ D.f[DIR_MM0] = &DDStart[DIR_PP0 * size_Mat];
+ D.f[DIR_PP0] = &DDStart[DIR_MM0 * size_Mat];
+ D.f[DIR_MP0] = &DDStart[DIR_PM0 * size_Mat];
+ D.f[DIR_PM0] = &DDStart[DIR_MP0 * size_Mat];
+ D.f[DIR_M0M] = &DDStart[DIR_P0P * size_Mat];
+ D.f[DIR_P0P] = &DDStart[DIR_M0M * size_Mat];
+ D.f[DIR_M0P] = &DDStart[DIR_P0M * size_Mat];
+ D.f[DIR_P0M] = &DDStart[DIR_M0P * size_Mat];
+ D.f[DIR_0MM] = &DDStart[DIR_0PP * size_Mat];
+ D.f[DIR_0PP] = &DDStart[DIR_0MM * size_Mat];
+ D.f[DIR_0MP] = &DDStart[DIR_0PM * size_Mat];
+ D.f[DIR_0PM] = &DDStart[DIR_0MP * size_Mat];
+ D.f[DIR_000] = &DDStart[DIR_000 * size_Mat];
+ D.f[DIR_MMM] = &DDStart[DIR_PPP * size_Mat];
+ D.f[DIR_PPM] = &DDStart[DIR_MMP * size_Mat];
+ D.f[DIR_MPM] = &DDStart[DIR_PMP * size_Mat];
+ D.f[DIR_PMM] = &DDStart[DIR_MPP * size_Mat];
+ D.f[DIR_MMP] = &DDStart[DIR_PPM * size_Mat];
+ D.f[DIR_PPP] = &DDStart[DIR_MMM * size_Mat];
+ D.f[DIR_MPP] = &DDStart[DIR_PMM * size_Mat];
+ D.f[DIR_PMP] = &DDStart[DIR_MPM * size_Mat];
}
////////////////////////////////////////////////////////////////////////////////
@@ -136,33 +136,33 @@ __global__ void LB_Kernel_WaleCumulantK15Comp(
unsigned int kbsw = neighborZ[ksw];
//////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
- real mfcbb = (D.f[DIR_P00 ])[k ];
- real mfabb = (D.f[DIR_M00 ])[kw ];
- real mfbcb = (D.f[DIR_0P0 ])[k ];
- real mfbab = (D.f[DIR_0M0 ])[ks ];
- real mfbbc = (D.f[DIR_00P ])[k ];
- real mfbba = (D.f[DIR_00M ])[kb ];
- real mfccb = (D.f[DIR_PP0 ])[k ];
- real mfaab = (D.f[DIR_MM0 ])[ksw];
- real mfcab = (D.f[DIR_PM0 ])[ks ];
- real mfacb = (D.f[DIR_MP0 ])[kw ];
- real mfcbc = (D.f[DIR_P0P ])[k ];
- real mfaba = (D.f[DIR_M0M ])[kbw];
- real mfcba = (D.f[DIR_P0M ])[kb ];
- real mfabc = (D.f[DIR_M0P ])[kw ];
- real mfbcc = (D.f[DIR_0PP ])[k ];
- real mfbaa = (D.f[DIR_0MM ])[kbs];
- real mfbca = (D.f[DIR_0PM ])[kb ];
- real mfbac = (D.f[DIR_0MP ])[ks ];
+ real mfcbb = (D.f[DIR_P00])[k ];
+ real mfabb = (D.f[DIR_M00])[kw ];
+ real mfbcb = (D.f[DIR_0P0])[k ];
+ real mfbab = (D.f[DIR_0M0])[ks ];
+ real mfbbc = (D.f[DIR_00P])[k ];
+ real mfbba = (D.f[DIR_00M])[kb ];
+ real mfccb = (D.f[DIR_PP0])[k ];
+ real mfaab = (D.f[DIR_MM0])[ksw];
+ real mfcab = (D.f[DIR_PM0])[ks ];
+ real mfacb = (D.f[DIR_MP0])[kw ];
+ real mfcbc = (D.f[DIR_P0P])[k ];
+ real mfaba = (D.f[DIR_M0M])[kbw];
+ real mfcba = (D.f[DIR_P0M])[kb ];
+ real mfabc = (D.f[DIR_M0P])[kw ];
+ real mfbcc = (D.f[DIR_0PP])[k ];
+ real mfbaa = (D.f[DIR_0MM])[kbs];
+ real mfbca = (D.f[DIR_0PM])[kb ];
+ real mfbac = (D.f[DIR_0MP])[ks ];
real mfbbb = (D.f[DIR_000])[k ];
- real mfccc = (D.f[DIR_PPP ])[k ];
- real mfaac = (D.f[DIR_MMP ])[ksw];
- real mfcac = (D.f[DIR_PMP ])[ks ];
- real mfacc = (D.f[DIR_MPP ])[kw ];
- real mfcca = (D.f[DIR_PPM ])[kb ];
+ real mfccc = (D.f[DIR_PPP])[k ];
+ real mfaac = (D.f[DIR_MMP])[ksw];
+ real mfcac = (D.f[DIR_PMP])[ks ];
+ real mfacc = (D.f[DIR_MPP])[kw ];
+ real mfcca = (D.f[DIR_PPM])[kb ];
real mfaaa = (D.f[DIR_MMM])[kbsw];
- real mfcaa = (D.f[DIR_PMM ])[kbs];
- real mfaca = (D.f[DIR_MPM ])[kbw];
+ real mfcaa = (D.f[DIR_PMM])[kbs];
+ real mfaca = (D.f[DIR_MPM])[kbw];
////////////////////////////////////////////////////////////////////////////////////
real drho = ((((mfccc+mfaaa) + (mfaca+mfcac)) + ((mfacc+mfcaa) + (mfaac+mfcca))) +
(((mfbac+mfbca) + (mfbaa+mfbcc)) + ((mfabc+mfcba) + (mfaba+mfcbc)) + ((mfacb+mfcab) + (mfaab+mfccb))) +
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/WaleKernels/FluidFlow/Compressible/CumulantK15BySoniMalav/WaleBySoniMalavCumulantK15Comp.cu b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/WaleKernels/FluidFlow/Compressible/CumulantK15BySoniMalav/WaleBySoniMalavCumulantK15Comp.cu
index 15b808279a4c9dc771531f118cb369b7c5380a84..49ee20b44f37b01cd9bc837024a47c1428c00a18 100644
--- a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/WaleKernels/FluidFlow/Compressible/CumulantK15BySoniMalav/WaleBySoniMalavCumulantK15Comp.cu
+++ b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/WaleKernels/FluidFlow/Compressible/CumulantK15BySoniMalav/WaleBySoniMalavCumulantK15Comp.cu
@@ -10,7 +10,7 @@ std::shared_ptr<WaleBySoniMalavCumulantK15Comp> WaleBySoniMalavCumulantK15Comp::
void WaleBySoniMalavCumulantK15Comp::run()
{
- int size_Mat = para->getParD(level)->numberOfNodes;
+ int size_Mat = (int)para->getParD(level)->numberOfNodes;
int numberOfThreads = para->getParD(level)->numberofthreads;
//int Grid = size_Array / numberOfThreads;
@@ -32,21 +32,22 @@ void WaleBySoniMalavCumulantK15Comp::run()
dim3 grid(Grid1, Grid2, 1);
dim3 threads(numberOfThreads, 1, 1);
- LB_Kernel_WaleBySoniMalavCumulantK15Comp << < grid, threads >> >( para->getParD(level)->omega,
- para->getParD(level)->typeOfGridNode,
- para->getParD(level)->neighborX,
- para->getParD(level)->neighborY,
- para->getParD(level)->neighborZ,
- para->getParD(level)->neighborInverse,
- para->getParD(level)->velocityX,
- para->getParD(level)->velocityY,
- para->getParD(level)->velocityZ,
- para->getParD(level)->distributions.f[0],
- para->getParD(level)->turbViscosity,
- para->getParD(level)->numberOfNodes,
- level,
- para->getForcesDev(),
- para->getParD(level)->isEvenTimestep);
+ LB_Kernel_WaleBySoniMalavCumulantK15Comp <<< grid, threads >>>(
+ para->getParD(level)->omega,
+ para->getParD(level)->typeOfGridNode,
+ para->getParD(level)->neighborX,
+ para->getParD(level)->neighborY,
+ para->getParD(level)->neighborZ,
+ para->getParD(level)->neighborInverse,
+ para->getParD(level)->velocityX,
+ para->getParD(level)->velocityY,
+ para->getParD(level)->velocityZ,
+ para->getParD(level)->distributions.f[0],
+ para->getParD(level)->turbViscosity,
+ para->getParD(level)->numberOfNodes,
+ level,
+ para->getForcesDev(),
+ para->getParD(level)->isEvenTimestep);
getLastCudaError("LB_Kernel_WaleBySoniMalavCumulantK15Comp execution failed");
}
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/WaleKernels/FluidFlow/Compressible/CumulantK15BySoniMalav/WaleBySoniMalavCumulantK15Comp_Device.cu b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/WaleKernels/FluidFlow/Compressible/CumulantK15BySoniMalav/WaleBySoniMalavCumulantK15Comp_Device.cu
index 511219c352c4d156428565f718191a70b9cc6c32..6258c72c36cafa27b06b2934db42a5813ed74f99 100644
--- a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/WaleKernels/FluidFlow/Compressible/CumulantK15BySoniMalav/WaleBySoniMalavCumulantK15Comp_Device.cu
+++ b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/WaleKernels/FluidFlow/Compressible/CumulantK15BySoniMalav/WaleBySoniMalavCumulantK15Comp_Device.cu
@@ -45,63 +45,63 @@ __global__ void LB_Kernel_WaleBySoniMalavCumulantK15Comp(
Distributions27 D;
if (EvenOrOdd==true)
{
- D.f[DIR_P00 ] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DDStart[DIR_P00 * size_Mat];
+ D.f[DIR_M00] = &DDStart[DIR_M00 * size_Mat];
+ D.f[DIR_0P0] = &DDStart[DIR_0P0 * size_Mat];
+ D.f[DIR_0M0] = &DDStart[DIR_0M0 * size_Mat];
+ D.f[DIR_00P] = &DDStart[DIR_00P * size_Mat];
+ D.f[DIR_00M] = &DDStart[DIR_00M * size_Mat];
+ D.f[DIR_PP0] = &DDStart[DIR_PP0 * size_Mat];
+ D.f[DIR_MM0] = &DDStart[DIR_MM0 * size_Mat];
+ D.f[DIR_PM0] = &DDStart[DIR_PM0 * size_Mat];
+ D.f[DIR_MP0] = &DDStart[DIR_MP0 * size_Mat];
+ D.f[DIR_P0P] = &DDStart[DIR_P0P * size_Mat];
+ D.f[DIR_M0M] = &DDStart[DIR_M0M * size_Mat];
+ D.f[DIR_P0M] = &DDStart[DIR_P0M * size_Mat];
+ D.f[DIR_M0P] = &DDStart[DIR_M0P * size_Mat];
+ D.f[DIR_0PP] = &DDStart[DIR_0PP * size_Mat];
+ D.f[DIR_0MM] = &DDStart[DIR_0MM * size_Mat];
+ D.f[DIR_0PM] = &DDStart[DIR_0PM * size_Mat];
+ D.f[DIR_0MP] = &DDStart[DIR_0MP * size_Mat];
+ D.f[DIR_000] = &DDStart[DIR_000 * size_Mat];
+ D.f[DIR_PPP] = &DDStart[DIR_PPP * size_Mat];
+ D.f[DIR_MMP] = &DDStart[DIR_MMP * size_Mat];
+ D.f[DIR_PMP] = &DDStart[DIR_PMP * size_Mat];
+ D.f[DIR_MPP] = &DDStart[DIR_MPP * size_Mat];
+ D.f[DIR_PPM] = &DDStart[DIR_PPM * size_Mat];
+ D.f[DIR_MMM] = &DDStart[DIR_MMM * size_Mat];
+ D.f[DIR_PMM] = &DDStart[DIR_PMM * size_Mat];
+ D.f[DIR_MPM] = &DDStart[DIR_MPM * size_Mat];
}
else
{
- D.f[DIR_M00 ] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_MMM ] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_PPM ] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_MPM ] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_PMM ] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_MMP ] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_PPP ] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_MPP ] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_PMP ] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_M00] = &DDStart[DIR_P00 * size_Mat];
+ D.f[DIR_P00] = &DDStart[DIR_M00 * size_Mat];
+ D.f[DIR_0M0] = &DDStart[DIR_0P0 * size_Mat];
+ D.f[DIR_0P0] = &DDStart[DIR_0M0 * size_Mat];
+ D.f[DIR_00M] = &DDStart[DIR_00P * size_Mat];
+ D.f[DIR_00P] = &DDStart[DIR_00M * size_Mat];
+ D.f[DIR_MM0] = &DDStart[DIR_PP0 * size_Mat];
+ D.f[DIR_PP0] = &DDStart[DIR_MM0 * size_Mat];
+ D.f[DIR_MP0] = &DDStart[DIR_PM0 * size_Mat];
+ D.f[DIR_PM0] = &DDStart[DIR_MP0 * size_Mat];
+ D.f[DIR_M0M] = &DDStart[DIR_P0P * size_Mat];
+ D.f[DIR_P0P] = &DDStart[DIR_M0M * size_Mat];
+ D.f[DIR_M0P] = &DDStart[DIR_P0M * size_Mat];
+ D.f[DIR_P0M] = &DDStart[DIR_M0P * size_Mat];
+ D.f[DIR_0MM] = &DDStart[DIR_0PP * size_Mat];
+ D.f[DIR_0PP] = &DDStart[DIR_0MM * size_Mat];
+ D.f[DIR_0MP] = &DDStart[DIR_0PM * size_Mat];
+ D.f[DIR_0PM] = &DDStart[DIR_0MP * size_Mat];
+ D.f[DIR_000] = &DDStart[DIR_000 * size_Mat];
+ D.f[DIR_MMM] = &DDStart[DIR_PPP * size_Mat];
+ D.f[DIR_PPM] = &DDStart[DIR_MMP * size_Mat];
+ D.f[DIR_MPM] = &DDStart[DIR_PMP * size_Mat];
+ D.f[DIR_PMM] = &DDStart[DIR_MPP * size_Mat];
+ D.f[DIR_MMP] = &DDStart[DIR_PPM * size_Mat];
+ D.f[DIR_PPP] = &DDStart[DIR_MMM * size_Mat];
+ D.f[DIR_MPP] = &DDStart[DIR_PMM * size_Mat];
+ D.f[DIR_PMP] = &DDStart[DIR_MPM * size_Mat];
}
////////////////////////////////////////////////////////////////////////////////
@@ -115,33 +115,33 @@ __global__ void LB_Kernel_WaleBySoniMalavCumulantK15Comp(
unsigned int kbsw = neighborZ[ksw];
//////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
- real mfcbb = (D.f[DIR_P00 ])[k ];
- real mfabb = (D.f[DIR_M00 ])[kw ];
- real mfbcb = (D.f[DIR_0P0 ])[k ];
- real mfbab = (D.f[DIR_0M0 ])[ks ];
- real mfbbc = (D.f[DIR_00P ])[k ];
- real mfbba = (D.f[DIR_00M ])[kb ];
- real mfccb = (D.f[DIR_PP0 ])[k ];
- real mfaab = (D.f[DIR_MM0 ])[ksw];
- real mfcab = (D.f[DIR_PM0 ])[ks ];
- real mfacb = (D.f[DIR_MP0 ])[kw ];
- real mfcbc = (D.f[DIR_P0P ])[k ];
- real mfaba = (D.f[DIR_M0M ])[kbw];
- real mfcba = (D.f[DIR_P0M ])[kb ];
- real mfabc = (D.f[DIR_M0P ])[kw ];
- real mfbcc = (D.f[DIR_0PP ])[k ];
- real mfbaa = (D.f[DIR_0MM ])[kbs];
- real mfbca = (D.f[DIR_0PM ])[kb ];
- real mfbac = (D.f[DIR_0MP ])[ks ];
+ real mfcbb = (D.f[DIR_P00])[k ];
+ real mfabb = (D.f[DIR_M00])[kw ];
+ real mfbcb = (D.f[DIR_0P0])[k ];
+ real mfbab = (D.f[DIR_0M0])[ks ];
+ real mfbbc = (D.f[DIR_00P])[k ];
+ real mfbba = (D.f[DIR_00M])[kb ];
+ real mfccb = (D.f[DIR_PP0])[k ];
+ real mfaab = (D.f[DIR_MM0])[ksw];
+ real mfcab = (D.f[DIR_PM0])[ks ];
+ real mfacb = (D.f[DIR_MP0])[kw ];
+ real mfcbc = (D.f[DIR_P0P])[k ];
+ real mfaba = (D.f[DIR_M0M])[kbw];
+ real mfcba = (D.f[DIR_P0M])[kb ];
+ real mfabc = (D.f[DIR_M0P])[kw ];
+ real mfbcc = (D.f[DIR_0PP])[k ];
+ real mfbaa = (D.f[DIR_0MM])[kbs];
+ real mfbca = (D.f[DIR_0PM])[kb ];
+ real mfbac = (D.f[DIR_0MP])[ks ];
real mfbbb = (D.f[DIR_000])[k ];
- real mfccc = (D.f[DIR_PPP ])[k ];
- real mfaac = (D.f[DIR_MMP ])[ksw];
- real mfcac = (D.f[DIR_PMP ])[ks ];
- real mfacc = (D.f[DIR_MPP ])[kw ];
- real mfcca = (D.f[DIR_PPM ])[kb ];
+ real mfccc = (D.f[DIR_PPP])[k ];
+ real mfaac = (D.f[DIR_MMP])[ksw];
+ real mfcac = (D.f[DIR_PMP])[ks ];
+ real mfacc = (D.f[DIR_MPP])[kw ];
+ real mfcca = (D.f[DIR_PPM])[kb ];
real mfaaa = (D.f[DIR_MMM])[kbsw];
- real mfcaa = (D.f[DIR_PMM ])[kbs];
- real mfaca = (D.f[DIR_MPM ])[kbw];
+ real mfcaa = (D.f[DIR_PMM])[kbs];
+ real mfaca = (D.f[DIR_MPM])[kbw];
////////////////////////////////////////////////////////////////////////////////////
real drho = ((((mfccc+mfaaa) + (mfaca+mfcac)) + ((mfacc+mfcaa) + (mfaac+mfcca))) +
(((mfbac+mfbca) + (mfbaa+mfbcc)) + ((mfabc+mfcba) + (mfaba+mfcbc)) + ((mfacb+mfcab) + (mfaab+mfccb))) +
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/WaleKernels/FluidFlow/Compressible/CumulantK17/WaleCumulantK17Comp.cu b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/WaleKernels/FluidFlow/Compressible/CumulantK17/WaleCumulantK17Comp.cu
index 5eeea51301c666cf17546c85a444413111bebf2c..c9c16e2d2d2259656248948f3f10977c8f18fd24 100644
--- a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/WaleKernels/FluidFlow/Compressible/CumulantK17/WaleCumulantK17Comp.cu
+++ b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/WaleKernels/FluidFlow/Compressible/CumulantK17/WaleCumulantK17Comp.cu
@@ -10,7 +10,7 @@ std::shared_ptr<WaleCumulantK17Comp> WaleCumulantK17Comp::getNewInstance(std::sh
void WaleCumulantK17Comp::run()
{
- int size_Mat = para->getParD(level)->numberOfNodes;
+ int size_Mat = (int)para->getParD(level)->numberOfNodes;
int numberOfThreads = para->getParD(level)->numberofthreads;
//int Grid = size_Array / numberOfThreads;
@@ -32,23 +32,24 @@ void WaleCumulantK17Comp::run()
dim3 grid(Grid1, Grid2, 1);
dim3 threads(numberOfThreads, 1, 1);
- LB_Kernel_WaleCumulantK17Comp <<< grid, threads >>>(para->getParD(level)->omega,
- para->getParD(level)->typeOfGridNode,
- para->getParD(level)->neighborX,
- para->getParD(level)->neighborY,
- para->getParD(level)->neighborZ,
- para->getParD(level)->neighborInverse,
- para->getParD(level)->velocityX,
- para->getParD(level)->velocityY,
- para->getParD(level)->velocityZ,
- para->getParD(level)->distributions.f[0],
- para->getParD(level)->turbViscosity,
- para->getParD(level)->numberOfNodes,
- level,
- para->getTimestepOfCoarseLevel(),
- para->getForcesDev(),
- para->getQuadricLimitersDev(),
- para->getParD(level)->isEvenTimestep);
+ LB_Kernel_WaleCumulantK17Comp <<< grid, threads >>>(
+ para->getParD(level)->omega,
+ para->getParD(level)->typeOfGridNode,
+ para->getParD(level)->neighborX,
+ para->getParD(level)->neighborY,
+ para->getParD(level)->neighborZ,
+ para->getParD(level)->neighborInverse,
+ para->getParD(level)->velocityX,
+ para->getParD(level)->velocityY,
+ para->getParD(level)->velocityZ,
+ para->getParD(level)->distributions.f[0],
+ para->getParD(level)->turbViscosity,
+ para->getParD(level)->numberOfNodes,
+ level,
+ para->getTimestepOfCoarseLevel(),
+ para->getForcesDev(),
+ para->getQuadricLimitersDev(),
+ para->getParD(level)->isEvenTimestep);
getLastCudaError("LB_Kernel_WaleCumulantK17Comp execution failed");
}
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/WaleKernels/FluidFlow/Compressible/CumulantK17/WaleCumulantK17Comp_Device.cu b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/WaleKernels/FluidFlow/Compressible/CumulantK17/WaleCumulantK17Comp_Device.cu
index 8aaa13ab1d868e15ea5707d1566ba653b44c645d..e3161e0d26efe8993bb4b6c34bda32bf15af5d3d 100644
--- a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/WaleKernels/FluidFlow/Compressible/CumulantK17/WaleCumulantK17Comp_Device.cu
+++ b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/WaleKernels/FluidFlow/Compressible/CumulantK17/WaleCumulantK17Comp_Device.cu
@@ -47,63 +47,63 @@ __global__ void LB_Kernel_WaleCumulantK17Comp(
Distributions27 D;
if (EvenOrOdd==true)
{
- D.f[DIR_P00 ] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_M00 ] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0P0 ] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0M0 ] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00P ] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00M ] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_PP0 ] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_MM0 ] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_PM0 ] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_MP0 ] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_P0P ] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_M0M ] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_P0M ] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_M0P ] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0PP ] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0MM ] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0PM ] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0MP ] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_PPP ] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_MMP ] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_PMP ] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_MPP ] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_PPM ] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_MMM ] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_PMM ] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_MPM ] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DDStart[DIR_P00 * size_Mat];
+ D.f[DIR_M00] = &DDStart[DIR_M00 * size_Mat];
+ D.f[DIR_0P0] = &DDStart[DIR_0P0 * size_Mat];
+ D.f[DIR_0M0] = &DDStart[DIR_0M0 * size_Mat];
+ D.f[DIR_00P] = &DDStart[DIR_00P * size_Mat];
+ D.f[DIR_00M] = &DDStart[DIR_00M * size_Mat];
+ D.f[DIR_PP0] = &DDStart[DIR_PP0 * size_Mat];
+ D.f[DIR_MM0] = &DDStart[DIR_MM0 * size_Mat];
+ D.f[DIR_PM0] = &DDStart[DIR_PM0 * size_Mat];
+ D.f[DIR_MP0] = &DDStart[DIR_MP0 * size_Mat];
+ D.f[DIR_P0P] = &DDStart[DIR_P0P * size_Mat];
+ D.f[DIR_M0M] = &DDStart[DIR_M0M * size_Mat];
+ D.f[DIR_P0M] = &DDStart[DIR_P0M * size_Mat];
+ D.f[DIR_M0P] = &DDStart[DIR_M0P * size_Mat];
+ D.f[DIR_0PP] = &DDStart[DIR_0PP * size_Mat];
+ D.f[DIR_0MM] = &DDStart[DIR_0MM * size_Mat];
+ D.f[DIR_0PM] = &DDStart[DIR_0PM * size_Mat];
+ D.f[DIR_0MP] = &DDStart[DIR_0MP * size_Mat];
+ D.f[DIR_000] = &DDStart[DIR_000 * size_Mat];
+ D.f[DIR_PPP] = &DDStart[DIR_PPP * size_Mat];
+ D.f[DIR_MMP] = &DDStart[DIR_MMP * size_Mat];
+ D.f[DIR_PMP] = &DDStart[DIR_PMP * size_Mat];
+ D.f[DIR_MPP] = &DDStart[DIR_MPP * size_Mat];
+ D.f[DIR_PPM] = &DDStart[DIR_PPM * size_Mat];
+ D.f[DIR_MMM] = &DDStart[DIR_MMM * size_Mat];
+ D.f[DIR_PMM] = &DDStart[DIR_PMM * size_Mat];
+ D.f[DIR_MPM] = &DDStart[DIR_MPM * size_Mat];
}
else
{
- D.f[DIR_M00 ] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_P00 ] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0M0 ] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0P0 ] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00M ] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00P ] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_MM0 ] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_PP0 ] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_MP0 ] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_PM0 ] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_M0M ] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_P0P ] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_M0P ] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_P0M ] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0MM ] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0PP ] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0MP ] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0PM ] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_MMM ] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_PPM ] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_MPM ] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_PMM ] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_MMP ] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_PPP ] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_MPP ] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_PMP ] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_M00] = &DDStart[DIR_P00 * size_Mat];
+ D.f[DIR_P00] = &DDStart[DIR_M00 * size_Mat];
+ D.f[DIR_0M0] = &DDStart[DIR_0P0 * size_Mat];
+ D.f[DIR_0P0] = &DDStart[DIR_0M0 * size_Mat];
+ D.f[DIR_00M] = &DDStart[DIR_00P * size_Mat];
+ D.f[DIR_00P] = &DDStart[DIR_00M * size_Mat];
+ D.f[DIR_MM0] = &DDStart[DIR_PP0 * size_Mat];
+ D.f[DIR_PP0] = &DDStart[DIR_MM0 * size_Mat];
+ D.f[DIR_MP0] = &DDStart[DIR_PM0 * size_Mat];
+ D.f[DIR_PM0] = &DDStart[DIR_MP0 * size_Mat];
+ D.f[DIR_M0M] = &DDStart[DIR_P0P * size_Mat];
+ D.f[DIR_P0P] = &DDStart[DIR_M0M * size_Mat];
+ D.f[DIR_M0P] = &DDStart[DIR_P0M * size_Mat];
+ D.f[DIR_P0M] = &DDStart[DIR_M0P * size_Mat];
+ D.f[DIR_0MM] = &DDStart[DIR_0PP * size_Mat];
+ D.f[DIR_0PP] = &DDStart[DIR_0MM * size_Mat];
+ D.f[DIR_0MP] = &DDStart[DIR_0PM * size_Mat];
+ D.f[DIR_0PM] = &DDStart[DIR_0MP * size_Mat];
+ D.f[DIR_000] = &DDStart[DIR_000 * size_Mat];
+ D.f[DIR_MMM] = &DDStart[DIR_PPP * size_Mat];
+ D.f[DIR_PPM] = &DDStart[DIR_MMP * size_Mat];
+ D.f[DIR_MPM] = &DDStart[DIR_PMP * size_Mat];
+ D.f[DIR_PMM] = &DDStart[DIR_MPP * size_Mat];
+ D.f[DIR_MMP] = &DDStart[DIR_PPM * size_Mat];
+ D.f[DIR_PPP] = &DDStart[DIR_MMM * size_Mat];
+ D.f[DIR_MPP] = &DDStart[DIR_PMM * size_Mat];
+ D.f[DIR_PMP] = &DDStart[DIR_MPM * size_Mat];
}
////////////////////////////////////////////////////////////////////////////////
@@ -137,33 +137,33 @@ __global__ void LB_Kernel_WaleCumulantK17Comp(
unsigned int kbsw = neighborZ[ksw];
//////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
- real mfcbb = (D.f[DIR_P00 ])[k ];
- real mfabb = (D.f[DIR_M00 ])[kw ];
- real mfbcb = (D.f[DIR_0P0 ])[k ];
- real mfbab = (D.f[DIR_0M0 ])[ks ];
- real mfbbc = (D.f[DIR_00P ])[k ];
- real mfbba = (D.f[DIR_00M ])[kb ];
- real mfccb = (D.f[DIR_PP0 ])[k ];
- real mfaab = (D.f[DIR_MM0 ])[ksw];
- real mfcab = (D.f[DIR_PM0 ])[ks ];
- real mfacb = (D.f[DIR_MP0 ])[kw ];
- real mfcbc = (D.f[DIR_P0P ])[k ];
- real mfaba = (D.f[DIR_M0M ])[kbw];
- real mfcba = (D.f[DIR_P0M ])[kb ];
- real mfabc = (D.f[DIR_M0P ])[kw ];
- real mfbcc = (D.f[DIR_0PP ])[k ];
- real mfbaa = (D.f[DIR_0MM ])[kbs];
- real mfbca = (D.f[DIR_0PM ])[kb ];
- real mfbac = (D.f[DIR_0MP ])[ks ];
+ real mfcbb = (D.f[DIR_P00])[k ];
+ real mfabb = (D.f[DIR_M00])[kw ];
+ real mfbcb = (D.f[DIR_0P0])[k ];
+ real mfbab = (D.f[DIR_0M0])[ks ];
+ real mfbbc = (D.f[DIR_00P])[k ];
+ real mfbba = (D.f[DIR_00M])[kb ];
+ real mfccb = (D.f[DIR_PP0])[k ];
+ real mfaab = (D.f[DIR_MM0])[ksw];
+ real mfcab = (D.f[DIR_PM0])[ks ];
+ real mfacb = (D.f[DIR_MP0])[kw ];
+ real mfcbc = (D.f[DIR_P0P])[k ];
+ real mfaba = (D.f[DIR_M0M])[kbw];
+ real mfcba = (D.f[DIR_P0M])[kb ];
+ real mfabc = (D.f[DIR_M0P])[kw ];
+ real mfbcc = (D.f[DIR_0PP])[k ];
+ real mfbaa = (D.f[DIR_0MM])[kbs];
+ real mfbca = (D.f[DIR_0PM])[kb ];
+ real mfbac = (D.f[DIR_0MP])[ks ];
real mfbbb = (D.f[DIR_000])[k ];
- real mfccc = (D.f[DIR_PPP ])[k ];
- real mfaac = (D.f[DIR_MMP ])[ksw];
- real mfcac = (D.f[DIR_PMP ])[ks ];
- real mfacc = (D.f[DIR_MPP ])[kw ];
- real mfcca = (D.f[DIR_PPM ])[kb ];
+ real mfccc = (D.f[DIR_PPP])[k ];
+ real mfaac = (D.f[DIR_MMP])[ksw];
+ real mfcac = (D.f[DIR_PMP])[ks ];
+ real mfacc = (D.f[DIR_MPP])[kw ];
+ real mfcca = (D.f[DIR_PPM])[kb ];
real mfaaa = (D.f[DIR_MMM])[kbsw];
- real mfcaa = (D.f[DIR_PMM ])[kbs];
- real mfaca = (D.f[DIR_MPM ])[kbw];
+ real mfcaa = (D.f[DIR_PMM])[kbs];
+ real mfaca = (D.f[DIR_MPM])[kbw];
////////////////////////////////////////////////////////////////////////////////////
real drho = ((((mfccc+mfaaa) + (mfaca+mfcac)) + ((mfacc+mfcaa) + (mfaac+mfcca))) +
(((mfbac+mfbca) + (mfbaa+mfbcc)) + ((mfabc+mfcba) + (mfaba+mfcbc)) + ((mfacb+mfcab) + (mfaab+mfccb))) +
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/WaleKernels/FluidFlow/Compressible/CumulantK17Debug/WaleCumulantK17DebugComp.cu b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/WaleKernels/FluidFlow/Compressible/CumulantK17Debug/WaleCumulantK17DebugComp.cu
index 98dca58f522bf02ce66328819e42c717f0ceef28..b3cdd494c02c6649d60818b6b264b8db8b79d426 100644
--- a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/WaleKernels/FluidFlow/Compressible/CumulantK17Debug/WaleCumulantK17DebugComp.cu
+++ b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/WaleKernels/FluidFlow/Compressible/CumulantK17Debug/WaleCumulantK17DebugComp.cu
@@ -10,7 +10,7 @@ std::shared_ptr<WaleCumulantK17DebugComp> WaleCumulantK17DebugComp::getNewInstan
void WaleCumulantK17DebugComp::run()
{
- int size_Mat = para->getParD(level)->numberOfNodes;
+ int size_Mat = (int)para->getParD(level)->numberOfNodes;
int numberOfThreads = para->getParD(level)->numberofthreads;
//int Grid = size_Array / numberOfThreads;
@@ -32,34 +32,34 @@ void WaleCumulantK17DebugComp::run()
dim3 grid(Grid1, Grid2, 1);
dim3 threads(numberOfThreads, 1, 1);
- LB_Kernel_WaleCumulantK17DebugComp << < grid, threads >> >(
- para->getParD(level)->omega,
- para->getParD(level)->typeOfGridNode,
- para->getParD(level)->neighborX,
- para->getParD(level)->neighborY,
- para->getParD(level)->neighborZ,
- para->getParD(level)->neighborInverse,
- para->getParD(level)->velocityX,
- para->getParD(level)->velocityY,
- para->getParD(level)->velocityZ,
- para->getParD(level)->distributions.f[0],
- para->getParD(level)->turbViscosity,
- para->getParD(level)->gSij,
- para->getParD(level)->gSDij,
- para->getParD(level)->gDxvx,
- para->getParD(level)->gDyvx,
- para->getParD(level)->gDzvx,
- para->getParD(level)->gDxvy,
- para->getParD(level)->gDyvy,
- para->getParD(level)->gDzvy,
- para->getParD(level)->gDxvz,
- para->getParD(level)->gDyvz,
- para->getParD(level)->gDzvz,
- para->getParD(level)->numberOfNodes,
- level,
- para->getForcesDev(),
- para->getQuadricLimitersDev(),
- para->getParD(level)->isEvenTimestep);
+ LB_Kernel_WaleCumulantK17DebugComp <<< grid, threads >>>(
+ para->getParD(level)->omega,
+ para->getParD(level)->typeOfGridNode,
+ para->getParD(level)->neighborX,
+ para->getParD(level)->neighborY,
+ para->getParD(level)->neighborZ,
+ para->getParD(level)->neighborInverse,
+ para->getParD(level)->velocityX,
+ para->getParD(level)->velocityY,
+ para->getParD(level)->velocityZ,
+ para->getParD(level)->distributions.f[0],
+ para->getParD(level)->turbViscosity,
+ para->getParD(level)->gSij,
+ para->getParD(level)->gSDij,
+ para->getParD(level)->gDxvx,
+ para->getParD(level)->gDyvx,
+ para->getParD(level)->gDzvx,
+ para->getParD(level)->gDxvy,
+ para->getParD(level)->gDyvy,
+ para->getParD(level)->gDzvy,
+ para->getParD(level)->gDxvz,
+ para->getParD(level)->gDyvz,
+ para->getParD(level)->gDzvz,
+ para->getParD(level)->numberOfNodes,
+ level,
+ para->getForcesDev(),
+ para->getQuadricLimitersDev(),
+ para->getParD(level)->isEvenTimestep);
getLastCudaError("LB_Kernel_WaleCumulantK17DebugComp execution failed");
}
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/WaleKernels/FluidFlow/Compressible/CumulantK17Debug/WaleCumulantK17DebugComp_Device.cu b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/WaleKernels/FluidFlow/Compressible/CumulantK17Debug/WaleCumulantK17DebugComp_Device.cu
index a1feba477a6555ea728311a6e99d5302652813ff..63f4ecc8716fcd606fb6a75709408b0885d781e9 100644
--- a/src/gpu/VirtualFluids_GPU/Kernel/Kernels/WaleKernels/FluidFlow/Compressible/CumulantK17Debug/WaleCumulantK17DebugComp_Device.cu
+++ b/src/gpu/VirtualFluids_GPU/Kernel/Kernels/WaleKernels/FluidFlow/Compressible/CumulantK17Debug/WaleCumulantK17DebugComp_Device.cu
@@ -57,63 +57,63 @@ __global__ void LB_Kernel_WaleCumulantK17DebugComp(
Distributions27 D;
if (EvenOrOdd == true)
{
- D.f[DIR_P00] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_M00] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0P0] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0M0] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00P] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00M] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_PP0] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_MM0] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_PM0] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_MP0] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_P0P] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_M0M] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_P0M] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_M0P] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0PP] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0MM] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0PM] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0MP] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_PPP] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_MMP] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_PMP] = &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_MPP] = &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_PPM] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_MMM] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_PMM]= &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_MPM]= &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_P00] = &DDStart[DIR_P00 * size_Mat];
+ D.f[DIR_M00] = &DDStart[DIR_M00 * size_Mat];
+ D.f[DIR_0P0] = &DDStart[DIR_0P0 * size_Mat];
+ D.f[DIR_0M0] = &DDStart[DIR_0M0 * size_Mat];
+ D.f[DIR_00P] = &DDStart[DIR_00P * size_Mat];
+ D.f[DIR_00M] = &DDStart[DIR_00M * size_Mat];
+ D.f[DIR_PP0] = &DDStart[DIR_PP0 * size_Mat];
+ D.f[DIR_MM0] = &DDStart[DIR_MM0 * size_Mat];
+ D.f[DIR_PM0] = &DDStart[DIR_PM0 * size_Mat];
+ D.f[DIR_MP0] = &DDStart[DIR_MP0 * size_Mat];
+ D.f[DIR_P0P] = &DDStart[DIR_P0P * size_Mat];
+ D.f[DIR_M0M] = &DDStart[DIR_M0M * size_Mat];
+ D.f[DIR_P0M] = &DDStart[DIR_P0M * size_Mat];
+ D.f[DIR_M0P] = &DDStart[DIR_M0P * size_Mat];
+ D.f[DIR_0PP] = &DDStart[DIR_0PP * size_Mat];
+ D.f[DIR_0MM] = &DDStart[DIR_0MM * size_Mat];
+ D.f[DIR_0PM] = &DDStart[DIR_0PM * size_Mat];
+ D.f[DIR_0MP] = &DDStart[DIR_0MP * size_Mat];
+ D.f[DIR_000] = &DDStart[DIR_000 * size_Mat];
+ D.f[DIR_PPP] = &DDStart[DIR_PPP * size_Mat];
+ D.f[DIR_MMP] = &DDStart[DIR_MMP * size_Mat];
+ D.f[DIR_PMP] = &DDStart[DIR_PMP * size_Mat];
+ D.f[DIR_MPP] = &DDStart[DIR_MPP * size_Mat];
+ D.f[DIR_PPM] = &DDStart[DIR_PPM * size_Mat];
+ D.f[DIR_MMM] = &DDStart[DIR_MMM * size_Mat];
+ D.f[DIR_PMM]= &DDStart[DIR_PMM * size_Mat];
+ D.f[DIR_MPM]= &DDStart[DIR_MPM * size_Mat];
}
else
{
- D.f[DIR_M00] = &DDStart[DIR_P00 *size_Mat];
- D.f[DIR_P00] = &DDStart[DIR_M00 *size_Mat];
- D.f[DIR_0M0] = &DDStart[DIR_0P0 *size_Mat];
- D.f[DIR_0P0] = &DDStart[DIR_0M0 *size_Mat];
- D.f[DIR_00M] = &DDStart[DIR_00P *size_Mat];
- D.f[DIR_00P] = &DDStart[DIR_00M *size_Mat];
- D.f[DIR_MM0] = &DDStart[DIR_PP0 *size_Mat];
- D.f[DIR_PP0] = &DDStart[DIR_MM0 *size_Mat];
- D.f[DIR_MP0] = &DDStart[DIR_PM0 *size_Mat];
- D.f[DIR_PM0] = &DDStart[DIR_MP0 *size_Mat];
- D.f[DIR_M0M] = &DDStart[DIR_P0P *size_Mat];
- D.f[DIR_P0P] = &DDStart[DIR_M0M *size_Mat];
- D.f[DIR_M0P] = &DDStart[DIR_P0M *size_Mat];
- D.f[DIR_P0M] = &DDStart[DIR_M0P *size_Mat];
- D.f[DIR_0MM] = &DDStart[DIR_0PP *size_Mat];
- D.f[DIR_0PP] = &DDStart[DIR_0MM *size_Mat];
- D.f[DIR_0MP] = &DDStart[DIR_0PM *size_Mat];
- D.f[DIR_0PM] = &DDStart[DIR_0MP *size_Mat];
- D.f[DIR_000] = &DDStart[DIR_000*size_Mat];
- D.f[DIR_MMM] = &DDStart[DIR_PPP *size_Mat];
- D.f[DIR_PPM] = &DDStart[DIR_MMP *size_Mat];
- D.f[DIR_MPM]= &DDStart[DIR_PMP *size_Mat];
- D.f[DIR_PMM]= &DDStart[DIR_MPP *size_Mat];
- D.f[DIR_MMP] = &DDStart[DIR_PPM *size_Mat];
- D.f[DIR_PPP] = &DDStart[DIR_MMM *size_Mat];
- D.f[DIR_MPP] = &DDStart[DIR_PMM *size_Mat];
- D.f[DIR_PMP] = &DDStart[DIR_MPM *size_Mat];
+ D.f[DIR_M00] = &DDStart[DIR_P00 * size_Mat];
+ D.f[DIR_P00] = &DDStart[DIR_M00 * size_Mat];
+ D.f[DIR_0M0] = &DDStart[DIR_0P0 * size_Mat];
+ D.f[DIR_0P0] = &DDStart[DIR_0M0 * size_Mat];
+ D.f[DIR_00M] = &DDStart[DIR_00P * size_Mat];
+ D.f[DIR_00P] = &DDStart[DIR_00M * size_Mat];
+ D.f[DIR_MM0] = &DDStart[DIR_PP0 * size_Mat];
+ D.f[DIR_PP0] = &DDStart[DIR_MM0 * size_Mat];
+ D.f[DIR_MP0] = &DDStart[DIR_PM0 * size_Mat];
+ D.f[DIR_PM0] = &DDStart[DIR_MP0 * size_Mat];
+ D.f[DIR_M0M] = &DDStart[DIR_P0P * size_Mat];
+ D.f[DIR_P0P] = &DDStart[DIR_M0M * size_Mat];
+ D.f[DIR_M0P] = &DDStart[DIR_P0M * size_Mat];
+ D.f[DIR_P0M] = &DDStart[DIR_M0P * size_Mat];
+ D.f[DIR_0MM] = &DDStart[DIR_0PP * size_Mat];
+ D.f[DIR_0PP] = &DDStart[DIR_0MM * size_Mat];
+ D.f[DIR_0MP] = &DDStart[DIR_0PM * size_Mat];
+ D.f[DIR_0PM] = &DDStart[DIR_0MP * size_Mat];
+ D.f[DIR_000] = &DDStart[DIR_000 * size_Mat];
+ D.f[DIR_MMM] = &DDStart[DIR_PPP * size_Mat];
+ D.f[DIR_PPM] = &DDStart[DIR_MMP * size_Mat];
+ D.f[DIR_MPM]= &DDStart[DIR_PMP * size_Mat];
+ D.f[DIR_PMM]= &DDStart[DIR_MPP * size_Mat];
+ D.f[DIR_MMP] = &DDStart[DIR_PPM * size_Mat];
+ D.f[DIR_PPP] = &DDStart[DIR_MMM * size_Mat];
+ D.f[DIR_MPP] = &DDStart[DIR_PMM * size_Mat];
+ D.f[DIR_PMP] = &DDStart[DIR_MPM * size_Mat];
}
////////////////////////////////////////////////////////////////////////////////
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Utilities/ChimeraTransformation.h b/src/gpu/VirtualFluids_GPU/Kernel/Utilities/ChimeraTransformation.h
deleted file mode 100644
index f7822d63fa0efd34b27773dffdeebddf521a8792..0000000000000000000000000000000000000000
--- a/src/gpu/VirtualFluids_GPU/Kernel/Utilities/ChimeraTransformation.h
+++ /dev/null
@@ -1,68 +0,0 @@
-#ifndef CHIMERA_TRANSFORMATION_H
-#define CHIMERA_TRANSFORMATION_H
-
-#include <lbm/constants/NumericConstants.h>
-
-using namespace vf::lbm::constant;
-
-////////////////////////////////////////////////////////////////////////////////
-//! \brief forward chimera transformation \ref forwardInverseChimeraWithK
-//! Transformation from distributions to central moments according to Eq. (6)-(14) in \ref
-//! <a href="https://doi.org/10.1016/j.jcp.2017.05.040"><b>[ M. Geier et al. (2017), DOI:10.1016/j.jcp.2017.05.040
-//! ]</b></a> Modified for lower round-off errors.
-inline __device__ void forwardInverseChimeraWithK(real &mfa, real &mfb, real &mfc, real vv, real v2, real Kinverse, real K)
-{
- real m2 = mfa + mfc;
- real m1 = mfc - mfa;
- real m0 = m2 + mfb;
- mfa = m0;
- m0 *= Kinverse;
- m0 += c1o1;
- mfb = (m1 * Kinverse - m0 * vv) * K;
- mfc = ((m2 - c2o1 * m1 * vv) * Kinverse + v2 * m0) * K;
-}
-
-////////////////////////////////////////////////////////////////////////////////
-//! \brief backward chimera transformation \ref backwardInverseChimeraWithK
-//! Transformation from central moments to distributions according to Eq. (57)-(65) in \ref
-//! <a href="https://doi.org/10.1016/j.jcp.2017.05.040"><b>[ M. Geier et al. (2017), DOI:10.1016/j.jcp.2017.05.040
-//! ]</b></a> Modified for lower round-off errors.
-inline __device__ void backwardInverseChimeraWithK(real &mfa, real &mfb, real &mfc, real vv, real v2, real Kinverse, real K)
-{
- real m0 = (((mfc - mfb) * c1o2 + mfb * vv) * Kinverse + (mfa * Kinverse + c1o1) * (v2 - vv) * c1o2) * K;
- real m1 = (((mfa - mfc) - c2o1 * mfb * vv) * Kinverse + (mfa * Kinverse + c1o1) * (-v2)) * K;
- mfc = (((mfc + mfb) * c1o2 + mfb * vv) * Kinverse + (mfa * Kinverse + c1o1) * (v2 + vv) * c1o2) * K;
- mfa = m0;
- mfb = m1;
-}
-
-////////////////////////////////////////////////////////////////////////////////
-//! \brief forward chimera transformation \ref forwardChimera
-//! Transformation from distributions to central moments according to Eq. (6)-(14) in \ref
-//! <a href="https://doi.org/10.1016/j.jcp.2017.05.040"><b>[ M. Geier et al. (2017), DOI:10.1016/j.jcp.2017.05.040
-//! ]</b></a> for \f$ K_{abc}=0 \f$. This is to avoid unnessary floating point operations. Modified for lower round-off
-//! errors.
-inline __device__ void forwardChimera(real &mfa, real &mfb, real &mfc, real vv, real v2)
-{
- real m1 = (mfa + mfc) + mfb;
- real m2 = mfc - mfa;
- mfc = (mfc + mfa) + (v2 * m1 - c2o1 * vv * m2);
- mfb = m2 - vv * m1;
- mfa = m1;
-}
-
-////////////////////////////////////////////////////////////////////////////////
-//! \brief backward chimera transformation \ref backwardChimera
-//! Transformation from central moments to distributions according to Eq. (57)-(65) in \ref
-//! <a href="https://doi.org/10.1016/j.jcp.2017.05.040"><b>[ M. Geier et al. (2017), DOI:10.1016/j.jcp.2017.05.040
-//! ]</b></a> for \f$ K_{abc}=0 \f$. This is to avoid unnessary floating point operations. Modified for lower round-off
-//! errors.
-inline __device__ void backwardChimera(real &mfa, real &mfb, real &mfc, real vv, real v2)
-{
- real ma = (mfc + mfa * (v2 - vv)) * c1o2 + mfb * (vv - c1o2);
- real mb = ((mfa - mfc) - mfa * v2) - c2o1 * mfb * vv;
- mfc = (mfc + mfa * (v2 + vv)) * c1o2 + mfb * (vv + c1o2);
- mfb = mb;
- mfa = ma;
-}
-#endif
\ No newline at end of file
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Utilities/DistributionHelper.cu b/src/gpu/VirtualFluids_GPU/Kernel/Utilities/DistributionHelper.cu
index 7c477c539dc3526389dc22563b50501e778a63f3..240a6ffbace64147aa67224fe72c946761fdc452 100644
--- a/src/gpu/VirtualFluids_GPU/Kernel/Utilities/DistributionHelper.cu
+++ b/src/gpu/VirtualFluids_GPU/Kernel/Utilities/DistributionHelper.cu
@@ -2,8 +2,7 @@
#include <cuda_runtime.h>
-
-#include <lbm/constants/NumericConstants.h>
+#include "lbm/constants/NumericConstants.h"
#include "lbm/constants/D3Q27.h"
using namespace vf::lbm::dir;
@@ -80,10 +79,4 @@ __device__ void DistributionWrapper::write()
(distribution_references.f[DIR_000])[k] = distribution.f[vf::lbm::dir::ZZZ];
}
-__device__ bool isValidFluidNode(uint nodeType)
-{
- return (nodeType == GEO_FLUID || nodeType == GEO_PM_0 || nodeType == GEO_PM_1 || nodeType == GEO_PM_2);
-}
-
-
}
\ No newline at end of file
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Utilities/DistributionHelper.cuh b/src/gpu/VirtualFluids_GPU/Kernel/Utilities/DistributionHelper.cuh
index 1009ecfa92f31e821d825ad72ba681bc3ae96d1b..599f3f46668c07da49725770177d77239f8ef9df 100644
--- a/src/gpu/VirtualFluids_GPU/Kernel/Utilities/DistributionHelper.cuh
+++ b/src/gpu/VirtualFluids_GPU/Kernel/Utilities/DistributionHelper.cuh
@@ -37,76 +37,13 @@
#include "lbm/KernelParameter.h"
#include "lbm/constants/D3Q27.h"
+#include "LBM/GPUHelperFunctions/KernelUtilities.h"
using namespace vf::lbm::dir;
namespace vf::gpu
{
-__inline__ __device__ __host__ void getPointersToDistributions(Distributions27 &dist, real *distributionArray, const uint numberOfLBnodes, const bool isEvenTimestep)
-{
- if (isEvenTimestep)
- {
- dist.f[DIR_000] = &distributionArray[DIR_000 * numberOfLBnodes];
- dist.f[DIR_P00] = &distributionArray[DIR_P00 * numberOfLBnodes];
- dist.f[DIR_M00] = &distributionArray[DIR_M00 * numberOfLBnodes];
- dist.f[DIR_0P0] = &distributionArray[DIR_0P0 * numberOfLBnodes];
- dist.f[DIR_0M0] = &distributionArray[DIR_0M0 * numberOfLBnodes];
- dist.f[DIR_00P] = &distributionArray[DIR_00P * numberOfLBnodes];
- dist.f[DIR_00M] = &distributionArray[DIR_00M * numberOfLBnodes];
- dist.f[DIR_PP0] = &distributionArray[DIR_PP0 * numberOfLBnodes];
- dist.f[DIR_MM0] = &distributionArray[DIR_MM0 * numberOfLBnodes];
- dist.f[DIR_PM0] = &distributionArray[DIR_PM0 * numberOfLBnodes];
- dist.f[DIR_MP0] = &distributionArray[DIR_MP0 * numberOfLBnodes];
- dist.f[DIR_P0P] = &distributionArray[DIR_P0P * numberOfLBnodes];
- dist.f[DIR_M0M] = &distributionArray[DIR_M0M * numberOfLBnodes];
- dist.f[DIR_P0M] = &distributionArray[DIR_P0M * numberOfLBnodes];
- dist.f[DIR_M0P] = &distributionArray[DIR_M0P * numberOfLBnodes];
- dist.f[DIR_0PP] = &distributionArray[DIR_0PP * numberOfLBnodes];
- dist.f[DIR_0MM] = &distributionArray[DIR_0MM * numberOfLBnodes];
- dist.f[DIR_0PM] = &distributionArray[DIR_0PM * numberOfLBnodes];
- dist.f[DIR_0MP] = &distributionArray[DIR_0MP * numberOfLBnodes];
- dist.f[DIR_PPP] = &distributionArray[DIR_PPP * numberOfLBnodes];
- dist.f[DIR_MMP] = &distributionArray[DIR_MMP * numberOfLBnodes];
- dist.f[DIR_PMP] = &distributionArray[DIR_PMP * numberOfLBnodes];
- dist.f[DIR_MPP] = &distributionArray[DIR_MPP * numberOfLBnodes];
- dist.f[DIR_PPM] = &distributionArray[DIR_PPM * numberOfLBnodes];
- dist.f[DIR_MMM] = &distributionArray[DIR_MMM * numberOfLBnodes];
- dist.f[DIR_PMM] = &distributionArray[DIR_PMM * numberOfLBnodes];
- dist.f[DIR_MPM] = &distributionArray[DIR_MPM * numberOfLBnodes];
- }
- else
- {
- dist.f[DIR_M00] = &distributionArray[DIR_P00 * numberOfLBnodes];
- dist.f[DIR_P00] = &distributionArray[DIR_M00 * numberOfLBnodes];
- dist.f[DIR_0M0] = &distributionArray[DIR_0P0 * numberOfLBnodes];
- dist.f[DIR_0P0] = &distributionArray[DIR_0M0 * numberOfLBnodes];
- dist.f[DIR_00M] = &distributionArray[DIR_00P * numberOfLBnodes];
- dist.f[DIR_00P] = &distributionArray[DIR_00M * numberOfLBnodes];
- dist.f[DIR_MM0] = &distributionArray[DIR_PP0 * numberOfLBnodes];
- dist.f[DIR_PP0] = &distributionArray[DIR_MM0 * numberOfLBnodes];
- dist.f[DIR_MP0] = &distributionArray[DIR_PM0 * numberOfLBnodes];
- dist.f[DIR_PM0] = &distributionArray[DIR_MP0 * numberOfLBnodes];
- dist.f[DIR_M0M] = &distributionArray[DIR_P0P * numberOfLBnodes];
- dist.f[DIR_P0P] = &distributionArray[DIR_M0M * numberOfLBnodes];
- dist.f[DIR_M0P] = &distributionArray[DIR_P0M * numberOfLBnodes];
- dist.f[DIR_P0M] = &distributionArray[DIR_M0P * numberOfLBnodes];
- dist.f[DIR_0MM] = &distributionArray[DIR_0PP * numberOfLBnodes];
- dist.f[DIR_0PP] = &distributionArray[DIR_0MM * numberOfLBnodes];
- dist.f[DIR_0MP] = &distributionArray[DIR_0PM * numberOfLBnodes];
- dist.f[DIR_0PM] = &distributionArray[DIR_0MP * numberOfLBnodes];
- dist.f[DIR_000] = &distributionArray[DIR_000 * numberOfLBnodes];
- dist.f[DIR_PPP] = &distributionArray[DIR_MMM * numberOfLBnodes];
- dist.f[DIR_MMP] = &distributionArray[DIR_PPM * numberOfLBnodes];
- dist.f[DIR_PMP] = &distributionArray[DIR_MPM * numberOfLBnodes];
- dist.f[DIR_MPP] = &distributionArray[DIR_PMM * numberOfLBnodes];
- dist.f[DIR_PPM] = &distributionArray[DIR_MMP * numberOfLBnodes];
- dist.f[DIR_MMM] = &distributionArray[DIR_PPP * numberOfLBnodes];
- dist.f[DIR_PMM] = &distributionArray[DIR_MPP * numberOfLBnodes];
- dist.f[DIR_MPM] = &distributionArray[DIR_PMP * numberOfLBnodes];
- }
-}
-
/**
* Getting references to the 27 directions.
* @params distributions 1D real* array containing all data (number of elements = 27 * matrix_size)
@@ -114,7 +51,7 @@ __inline__ __device__ __host__ void getPointersToDistributions(Distributions27 &
* @params isEvenTimestep: stored data dependent on timestep is based on the esoteric twist algorithm
* @return a data struct containing the addresses to the 27 directions within the 1D distribution array
*/
-__inline__ __device__ __host__ DistributionReferences27 getDistributionReferences27(real* distributions, unsigned int numberOfLBnodes, bool isEvenTimestep){
+__inline__ __device__ __host__ DistributionReferences27 getDistributionReferences27(real* distributions, const unsigned long long numberOfLBnodes, const bool isEvenTimestep){
DistributionReferences27 distribution_references;
getPointersToDistributions(distribution_references, distributions, numberOfLBnodes, isEvenTimestep);
return distribution_references;
@@ -157,20 +94,6 @@ struct DistributionWrapper
const uint kbsw;
};
-__inline__ __device__ unsigned int getNodeIndex()
-{
- const unsigned x = threadIdx.x;
- const unsigned y = blockIdx.x;
- const unsigned z = blockIdx.y;
-
- const unsigned nx = blockDim.x;
- const unsigned ny = gridDim.x;
-
- return nx * (ny * z + y) + x;
-}
-
-__device__ bool isValidFluidNode(uint nodeType);
-
}
#endif
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Utilities/KernelFactory/KernelFactoryImp.cpp b/src/gpu/VirtualFluids_GPU/Kernel/Utilities/KernelFactory/KernelFactoryImp.cpp
index 53ec240f096080097416e640fdd095c3812fb34c..5a2d8c9a426e5cb23ca75f91aaf6fbff75cba72b 100644
--- a/src/gpu/VirtualFluids_GPU/Kernel/Utilities/KernelFactory/KernelFactoryImp.cpp
+++ b/src/gpu/VirtualFluids_GPU/Kernel/Utilities/KernelFactory/KernelFactoryImp.cpp
@@ -8,11 +8,9 @@
#include "Kernel/Kernels/BasicKernels/FluidFlow/Compressible/BGKPlus/BGKPlusCompSP27.h"
#include "Kernel/Kernels/BasicKernels/FluidFlow/Compressible/Cascade/CascadeCompSP27.h"
#include "Kernel/Kernels/BasicKernels/FluidFlow/Compressible/Cumulant/CumulantCompSP27.h"
-#include "Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK17/CumulantK17Comp.h"
#include "Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK17Unified/CumulantK17Unified.h"
#include "Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK17chim/CumulantK17CompChim.h"
-#include "Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK17chimStream/CumulantK17CompChimStream.h"
-#include "Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK17chimRedesigned/CumulantK17CompChimRedesigned.h"
+#include "Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK17/CumulantK17.h"
#include "Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK17Bulk/CumulantK17BulkComp.h"
#include "Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantAll4/CumulantAll4CompSP27.h"
#include "Kernel/Kernels/BasicKernels/FluidFlow/Compressible/CumulantK18/CumulantK18Comp.h"
@@ -49,9 +47,6 @@
#include "Kernel/Kernels/WaleKernels/FluidFlow/Compressible/CumulantK15/WaleCumulantK15Comp.h"
#include "Kernel/Kernels/WaleKernels/FluidFlow/Compressible/CumulantK15BySoniMalav/WaleBySoniMalavCumulantK15Comp.h"
-//turbulent viscosity kernel
-#include "Kernel/Kernels/TurbulentViscosityKernels/FluidFlow/Compressible/CumulantK17chim/TurbulentViscosityCumulantK17CompChim.h"
-
//strategies
#include "Kernel/Kernels/BasicKernels/FluidFlow/Compressible/FluidFlowCompStrategy.h"
#include "Kernel/Kernels/BasicKernels/FluidFlow/Incompressible/FluidFlowIncompStrategy.h"
@@ -61,7 +56,6 @@
#include "Kernel/Kernels/BasicKernels/AdvectionDiffusion/Incompressible/Mod7/ADMod7IncompStrategy.h"
#include "Kernel/Kernels/PorousMediaKernels/FluidFlow/Compressible/PMFluidFlowCompStrategy.h"
#include "Kernel/Kernels/WaleKernels/FluidFlow/Compressible/WaleFluidFlowCompStrategy.h"
-#include "Kernel/Kernels/TurbulentViscosityKernels/FluidFlow/Compressible/TurbulentViscosityFluidFlowCompStrategy.h"
std::vector<std::shared_ptr<Kernel>> KernelFactoryImp::makeKernels(std::shared_ptr<Parameter> para)
{
@@ -118,9 +112,6 @@ std::shared_ptr<Kernel> KernelFactoryImp::makeKernel(std::shared_ptr<Parameter>
} else if (kernel == "CumulantCompSP27") {
newKernel = CumulantCompSP27::getNewInstance(para, level);
checkStrategy = FluidFlowCompStrategy::getInstance();
- } else if (kernel == "CumulantK17Comp") {
- newKernel = CumulantK17Comp::getNewInstance(para, level);
- checkStrategy = FluidFlowCompStrategy::getInstance();
} else if (kernel == "CumulantK15Unified") {
newKernel = std::make_shared<vf::gpu::CumulantK15Unified>(para, level);
checkStrategy = FluidFlowCompStrategy::getInstance();
@@ -133,12 +124,26 @@ std::shared_ptr<Kernel> KernelFactoryImp::makeKernel(std::shared_ptr<Parameter>
} else if (kernel == "CumulantK17CompChim") {
newKernel = CumulantK17CompChim::getNewInstance(para, level);
checkStrategy = FluidFlowCompStrategy::getInstance();
- } else if (kernel == "CumulantK17CompChimStream") {
- newKernel = CumulantK17CompChimStream::getNewInstance(para, level);
- checkStrategy = FluidFlowCompStrategy::getInstance();
- } else if (kernel == "CumulantK17CompChimRedesigned") {
- newKernel = CumulantK17CompChimRedesigned::getNewInstance(para, level);
- checkStrategy = FluidFlowCompStrategy::getInstance();
+ } else if (kernel == "CumulantK17"){
+ switch(para->getTurbulenceModel())
+ {
+ case TurbulenceModel::AMD:
+ newKernel = CumulantK17<TurbulenceModel::AMD>::getNewInstance(para, level);
+ break;
+ case TurbulenceModel::Smagorinsky:
+ newKernel = CumulantK17<TurbulenceModel::Smagorinsky>::getNewInstance(para, level);
+ break;
+ case TurbulenceModel::QR:
+ newKernel = CumulantK17<TurbulenceModel::QR>::getNewInstance(para, level);
+ break;
+ case TurbulenceModel::None:
+ newKernel = CumulantK17<TurbulenceModel::None>::getNewInstance(para, level);
+ break;
+ default:
+ throw std::runtime_error("Unknown turbulence model!");
+ break;
+ }
+ checkStrategy = FluidFlowCompStrategy::getInstance();
} else if (kernel == "CumulantAll4CompSP27") {
newKernel = CumulantAll4CompSP27::getNewInstance(para, level);
checkStrategy = FluidFlowCompStrategy::getInstance();
@@ -197,35 +202,9 @@ std::shared_ptr<Kernel> KernelFactoryImp::makeKernel(std::shared_ptr<Parameter>
newKernel = WaleBySoniMalavCumulantK15Comp::getNewInstance(para, level);// ||
checkStrategy = WaleFluidFlowCompStrategy::getInstance(); // wale model
} //===============
- else if (kernel == "TurbulentViscosityCumulantK17CompChim"){ // compressible with turbulent viscosity
- switch(para->getTurbulenceModel()) // ||
- { // \/ //
- case TurbulenceModel::AMD:
- newKernel = TurbulentViscosityCumulantK17CompChim<TurbulenceModel::AMD>::getNewInstance(para, level);
- break;
- case TurbulenceModel::Smagorinsky:
- newKernel = TurbulentViscosityCumulantK17CompChim<TurbulenceModel::Smagorinsky>::getNewInstance(para, level);
- break;
- case TurbulenceModel::QR:
- newKernel = TurbulentViscosityCumulantK17CompChim<TurbulenceModel::QR>::getNewInstance(para, level);
- break;
- case TurbulenceModel::None:
- throw std::runtime_error("TurbulentViscosityCumulantK17CompChim currently not implemented for TurbulenceModel::None!");
- break;
- default:
- throw std::runtime_error("Unknown turbulence model!");
- break;
- }
- checkStrategy = TurbulentViscosityFluidFlowCompStrategy::getInstance();
- // /\ //
- // ||
- // compressible with turbulent viscosity
- //===============
- }
else {
throw std::runtime_error("KernelFactory does not know the KernelType.");
}
-
newKernel->setCheckParameterStrategy(checkStrategy);
para->setKernelNeedsFluidNodeIndicesToRun(newKernel->getKernelUsesFluidNodeIndices());
return newKernel;
diff --git a/src/gpu/VirtualFluids_GPU/Kernel/Utilities/ScalingHelperFunctions.h b/src/gpu/VirtualFluids_GPU/Kernel/Utilities/ScalingHelperFunctions.h
deleted file mode 100644
index 13ce5d88aaa7cb49225fa914c1f59c2de05802f5..0000000000000000000000000000000000000000
--- a/src/gpu/VirtualFluids_GPU/Kernel/Utilities/ScalingHelperFunctions.h
+++ /dev/null
@@ -1,148 +0,0 @@
-//=======================================================================================
-// ____ ____ __ ______ __________ __ __ __ __
-// \ \ | | | | | _ \ |___ ___| | | | | / \ | |
-// \ \ | | | | | |_) | | | | | | | / \ | |
-// \ \ | | | | | _ / | | | | | | / /\ \ | |
-// \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
-// \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
-// \ \ | | ________________________________________________________________
-// \ \ | | | ______________________________________________________________|
-// \ \| | | | __ __ __ __ ______ _______
-// \ | | |_____ | | | | | | | | | _ \ / _____)
-// \ | | _____| | | | | | | | | | | \ \ \_______
-// \ | | | | |_____ | \_/ | | | | |_/ / _____ |
-// \ _____| |__| |________| \_______/ |__| |______/ (_______/
-//
-// This file is part of VirtualFluids. VirtualFluids is free software: you can
-// redistribute it and/or modify it under the terms of the GNU General Public
-// License as published by the Free Software Foundation, either version 3 of
-// the License, or (at your option) any later version.
-//
-// VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
-// ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
-// FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
-// for more details.
-//
-// You should have received a copy of the GNU General Public License along
-// with VirtualFluids (see COPYING.txt). If not, see <http://www.gnu.org/licenses/>.
-//
-//! \file scalingHelperFunctions.h
-//! \ingroup GPU/Kernel/Utilities
-//! \author Martin Schoenherr, Anna Wellmann
-//=======================================================================================
-
-#ifndef SCALING_HELPER_FUNCTIONS_H
-#define SCALING_HELPER_FUNCTIONS_H
-
-#include "LBM/LB.h"
-#include "lbm/constants/D3Q27.h"
-#include "lbm/constants/NumericConstants.h"
-
-using namespace vf::lbm::constant;
-using namespace vf::lbm::dir;
-
-__device__ __inline__ void calculateMomentsOnSourceNodes(
- Distributions27& dist,
- real& omega,
- unsigned int& k_000,
- unsigned int& k_M00,
- unsigned int& k_0M0,
- unsigned int& k_00M,
- unsigned int& k_MM0,
- unsigned int& k_M0M,
- unsigned int& k_0MM,
- unsigned int& k_MMM,
- real& drho,
- real& velocityX,
- real& velocityY,
- real& velocityZ,
- real& kxyFromfcNEQ,
- real& kyzFromfcNEQ,
- real& kxzFromfcNEQ,
- real& kxxMyyFromfcNEQ,
- real& kxxMzzFromfcNEQ
- ){
- ////////////////////////////////////////////////////////////////////////////////////
- //! - Set local distributions (f's) on source nodes:
- //!
- real f_000 = (dist.f[DIR_000])[k_000];
- real f_P00 = (dist.f[DIR_P00])[k_000];
- real f_M00 = (dist.f[DIR_M00])[k_M00];
- real f_0P0 = (dist.f[DIR_0P0])[k_000];
- real f_0M0 = (dist.f[DIR_0M0])[k_0M0];
- real f_00P = (dist.f[DIR_00P])[k_000];
- real f_00M = (dist.f[DIR_00M])[k_00M];
- real f_PP0 = (dist.f[DIR_PP0])[k_000];
- real f_MM0 = (dist.f[DIR_MM0])[k_MM0];
- real f_PM0 = (dist.f[DIR_PM0])[k_0M0];
- real f_MP0 = (dist.f[DIR_MP0])[k_M00];
- real f_P0P = (dist.f[DIR_P0P])[k_000];
- real f_M0M = (dist.f[DIR_M0M])[k_M0M];
- real f_P0M = (dist.f[DIR_P0M])[k_00M];
- real f_M0P = (dist.f[DIR_M0P])[k_M00];
- real f_0PP = (dist.f[DIR_0PP])[k_000];
- real f_0MM = (dist.f[DIR_0MM])[k_0MM];
- real f_0PM = (dist.f[DIR_0PM])[k_00M];
- real f_0MP = (dist.f[DIR_0MP])[k_0M0];
- real f_PPP = (dist.f[DIR_PPP])[k_000];
- real f_MPP = (dist.f[DIR_MPP])[k_M00];
- real f_PMP = (dist.f[DIR_PMP])[k_0M0];
- real f_MMP = (dist.f[DIR_MMP])[k_MM0];
- real f_PPM = (dist.f[DIR_PPM])[k_00M];
- real f_MPM = (dist.f[DIR_MPM])[k_M0M];
- real f_PMM = (dist.f[DIR_PMM])[k_0MM];
- real f_MMM = (dist.f[DIR_MMM])[k_MMM];
-
- ////////////////////////////////////////////////////////////////////////////////////
- //! - Calculate density and velocity using pyramid summation for low round-off errors as in Eq. (J1)-(J3) \ref
- //! <a href="https://doi.org/10.1016/j.camwa.2015.05.001"><b>[ M. Geier et al. (2015),
- //! DOI:10.1016/j.camwa.2015.05.001 ]</b></a>
- //!
- drho = ((((f_PPP + f_MMM) + (f_MPM + f_PMP)) + ((f_MPP + f_PMM) + (f_MMP + f_PPM))) +
- (((f_0MP + f_0PM) + (f_0MM + f_0PP)) + ((f_M0P + f_P0M) + (f_M0M + f_P0P)) +
- ((f_MP0 + f_PM0) + (f_MM0 + f_PP0))) +
- ((f_M00 + f_P00) + (f_0M0 + f_0P0) + (f_00M + f_00P))) +
- f_000;
-
- real oneOverRho = c1o1 / (c1o1 + drho);
-
- velocityX = ((((f_PPP - f_MMM) + (f_PMP - f_MPM)) + ((f_PMM - f_MPP) + (f_PPM - f_MMP))) +
- (((f_P0M - f_M0P) + (f_P0P - f_M0M)) + ((f_PM0 - f_MP0) + (f_PP0 - f_MM0))) + (f_P00 - f_M00)) *
- oneOverRho;
- velocityY = ((((f_PPP - f_MMM) + (f_MPM - f_PMP)) + ((f_MPP - f_PMM) + (f_PPM - f_MMP))) +
- (((f_0PM - f_0MP) + (f_0PP - f_0MM)) + ((f_MP0 - f_PM0) + (f_PP0 - f_MM0))) + (f_0P0 - f_0M0)) *
- oneOverRho;
- velocityZ = ((((f_PPP - f_MMM) + (f_PMP - f_MPM)) + ((f_MPP - f_PMM) + (f_MMP - f_PPM))) +
- (((f_0MP - f_0PM) + (f_0PP - f_0MM)) + ((f_M0P - f_P0M) + (f_P0P - f_M0M))) + (f_00P - f_00M)) *
- oneOverRho;
-
- ////////////////////////////////////////////////////////////////////////////////////
- //! - Calculate second order moments for interpolation
- //!
- // example: kxxMzz: moment, second derivative in x direction minus the second derivative in z direction
- kxyFromfcNEQ =
- -c3o1 * omega *
- ((f_MM0 + f_MMM + f_MMP - f_MP0 - f_MPM - f_MPP - f_PM0 - f_PMM - f_PMP + f_PP0 + f_PPM + f_PPP) /
- (c1o1 + drho) -
- ((velocityX * velocityY)));
- kyzFromfcNEQ =
- -c3o1 * omega *
- ((f_0MM + f_PMM + f_MMM - f_0MP - f_PMP - f_MMP - f_0PM - f_PPM - f_MPM + f_0PP + f_PPP + f_MPP) /
- (c1o1 + drho) -
- ((velocityY * velocityZ)));
- kxzFromfcNEQ =
- -c3o1 * omega *
- ((f_M0M + f_MMM + f_MPM - f_M0P - f_MMP - f_MPP - f_P0M - f_PMM - f_PPM + f_P0P + f_PMP + f_PPP) /
- (c1o1 + drho) -
- ((velocityX * velocityZ)));
- kxxMyyFromfcNEQ =
- -c3o2 * omega *
- ((f_M0M + f_M00 + f_M0P - f_0MM - f_0M0 - f_0MP - f_0PM - f_0P0 - f_0PP + f_P0M + f_P00 + f_P0P) / (c1o1 + drho) -
- ((velocityX * velocityX - velocityY * velocityY)));
- kxxMzzFromfcNEQ =
- -c3o2 * omega *
- ((f_MM0 + f_M00 + f_MP0 - f_0MM - f_0MP - f_00M - f_00P - f_0PM - f_0PP + f_PM0 + f_P00 + f_PP0) / (c1o1 + drho) -
- ((velocityX * velocityX - velocityZ * velocityZ)));
-}
-
-#endif
\ No newline at end of file
diff --git a/src/gpu/VirtualFluids_GPU/KernelManager/ADKernelManager.cpp b/src/gpu/VirtualFluids_GPU/KernelManager/ADKernelManager.cpp
index 9ca813ac4987af618491422acb60207b7fee543c..5a36daecd5a82fc8a052bf51fedc1cb35b94a960 100644
--- a/src/gpu/VirtualFluids_GPU/KernelManager/ADKernelManager.cpp
+++ b/src/gpu/VirtualFluids_GPU/KernelManager/ADKernelManager.cpp
@@ -90,10 +90,10 @@ void ADKernelManager::initAD(const int level) const
////////////////////////////////////////////////////////////////////////////////
void ADKernelManager::setInitialNodeValuesAD(const int level, SPtr<CudaMemoryManager> cudaMemoryManager) const
{
- for (uint j = 1; j <= para->getParH(level)->numberOfNodes; j++) {
- const real coordX = para->getParH(level)->coordinateX[j];
- const real coordY = para->getParH(level)->coordinateY[j];
- const real coordZ = para->getParH(level)->coordinateZ[j];
+ for (size_t index = 1; index <= para->getParH(level)->numberOfNodes; index++) {
+ const real coordX = para->getParH(level)->coordinateX[index];
+ const real coordY = para->getParH(level)->coordinateY[index];
+ const real coordZ = para->getParH(level)->coordinateZ[index];
real concentration;
@@ -104,7 +104,7 @@ void ADKernelManager::setInitialNodeValuesAD(const int level, SPtr<CudaMemoryMan
concentration = real(0.0);
}
- para->getParH(level)->concentration[j] = concentration;
+ para->getParH(level)->concentration[index] = concentration;
}
cudaMemoryManager->cudaCopyConcentrationHostToDevice(level);
diff --git a/src/gpu/VirtualFluids_GPU/KernelManager/BCKernelManager.cpp b/src/gpu/VirtualFluids_GPU/KernelManager/BCKernelManager.cpp
index cc945ea225a28c58dca4ceefdb80fffb76228b21..e8fc3f318c920be36be7861a28659124a7b1e977 100644
--- a/src/gpu/VirtualFluids_GPU/KernelManager/BCKernelManager.cpp
+++ b/src/gpu/VirtualFluids_GPU/KernelManager/BCKernelManager.cpp
@@ -38,6 +38,7 @@
#include "BCKernelManager.h"
#include "Factories/BoundaryConditionFactory.h"
+#include "GridGenerator/TransientBCSetter/TransientBCSetter.h"
#include "Calculation/Cp.h"
#include "Calculation/DragLift.h"
#include "GPU/GPU_Interface.h"
@@ -51,6 +52,7 @@ BCKernelManager::BCKernelManager(SPtr<Parameter> parameter, BoundaryConditionFac
this->pressureBoundaryConditionPre = bcFactory->getPressureBoundaryConditionPre();
this->geometryBoundaryConditionPost = bcFactory->getGeometryBoundaryConditionPost();
this->stressBoundaryConditionPost = bcFactory->getStressBoundaryConditionPost();
+ this->precursorBoundaryConditionPost = bcFactory->getPrecursorBoundaryConditionPost();
checkBoundaryCondition(this->velocityBoundaryConditionPost, this->para->getParD(0)->velocityBC,
"velocityBoundaryConditionPost");
@@ -64,6 +66,8 @@ BCKernelManager::BCKernelManager(SPtr<Parameter> parameter, BoundaryConditionFac
"geometryBoundaryConditionPost");
checkBoundaryCondition(this->stressBoundaryConditionPost, this->para->getParD(0)->stressBC,
"stressBoundaryConditionPost");
+ checkBoundaryCondition(this->precursorBoundaryConditionPost, this->para->getParD(0)->precursorBC,
+ "precursorBoundaryConditionPost");
}
void BCKernelManager::runVelocityBCKernelPre(const int level) const
@@ -387,3 +391,41 @@ void BCKernelManager::runNoSlipBCKernelPost(const int level) const{
noSlipBoundaryConditionPost(para->getParD(level).get(), &(para->getParD(level)->noSlipBC));
}
}
+
+void BCKernelManager::runPrecursorBCKernelPost(int level, uint t, CudaMemoryManager* cudaMemoryManager)
+{
+ if(para->getParH(level)->precursorBC.numberOfBCnodes == 0) return;
+
+ uint t_level = para->getTimeStep(level, t, true);
+
+ uint lastTime = (para->getParD(level)->precursorBC.nPrecursorReads-2)*para->getParD(level)->precursorBC.timeStepsBetweenReads; // timestep currently loaded into last arrays
+ uint currentTime = (para->getParD(level)->precursorBC.nPrecursorReads-1)*para->getParD(level)->precursorBC.timeStepsBetweenReads; // timestep currently loaded into current arrays
+ uint nextTime = para->getParD(level)->precursorBC.nPrecursorReads *para->getParD(level)->precursorBC.timeStepsBetweenReads; // timestep currently loaded into next arrays
+
+ if(t_level>=currentTime)
+ {
+ //cycle time
+ lastTime = currentTime;
+ currentTime = nextTime;
+ nextTime += para->getParD(level)->precursorBC.timeStepsBetweenReads;
+
+ //cycle pointers
+ real* tmp = para->getParD(level)->precursorBC.last;
+ para->getParD(level)->precursorBC.last = para->getParD(level)->precursorBC.current;
+ para->getParD(level)->precursorBC.current = para->getParD(level)->precursorBC.next;
+ para->getParD(level)->precursorBC.next = tmp;
+
+ real loadTime = nextTime*pow(2,-level)*para->getTimeRatio();
+
+ for(auto reader : para->getParH(level)->transientBCInputFileReader)
+ {
+ reader->getNextData(para->getParH(level)->precursorBC.next, para->getParH(level)->precursorBC.numberOfPrecursorNodes, loadTime);
+ }
+ cudaMemoryManager->cudaCopyPrecursorData(level);
+ para->getParD(level)->precursorBC.nPrecursorReads++;
+ para->getParH(level)->precursorBC.nPrecursorReads++;
+ }
+
+ real tRatio = real(t_level-lastTime)/para->getParD(level)->precursorBC.timeStepsBetweenReads;
+ precursorBoundaryConditionPost(para->getParD(level).get(), ¶->getParD(level)->precursorBC, tRatio, para->getVelocityRatio());
+}
diff --git a/src/gpu/VirtualFluids_GPU/KernelManager/BCKernelManager.h b/src/gpu/VirtualFluids_GPU/KernelManager/BCKernelManager.h
index 423a9cc9056281a3a2a135ae32fa26cc47f93967..339100e6b5307e8e60f8d0846560bf89c6eea1a1 100644
--- a/src/gpu/VirtualFluids_GPU/KernelManager/BCKernelManager.h
+++ b/src/gpu/VirtualFluids_GPU/KernelManager/BCKernelManager.h
@@ -41,6 +41,7 @@
#include "PointerDefinitions.h"
#include "VirtualFluids_GPU_export.h"
+
class CudaMemoryManager;
class BoundaryConditionFactory;
class Parameter;
@@ -48,6 +49,7 @@ struct LBMSimulationParameter;
using boundaryCondition = std::function<void(LBMSimulationParameter *, QforBoundaryConditions *)>;
using boundaryConditionWithParameter = std::function<void(Parameter *, QforBoundaryConditions *, const int level)>;
+using precursorBoundaryCondition = std::function<void(LBMSimulationParameter *, QforPrecursorBoundaryConditions *, real tRatio, real velocityRatio)>;
//! \class BCKernelManager
//! \brief manage the cuda kernel calls to boundary conditions
@@ -84,7 +86,10 @@ public:
//! \brief calls the device function of the pressure boundary condition (post-collision)
void runPressureBCKernelPost(const int level) const;
- //! \brief calls the device function of the outflow boundary condition (pre-collision)
+ //! \brief calls the device function of the precursor boundary condition
+ void runPrecursorBCKernelPost(int level, uint t, CudaMemoryManager* cudaMemoryManager);
+
+ //! \brief calls the device function of the outflow boundary condition
void runOutflowBCKernelPre(const int level) const;
//! \brief calls the device function of the stress wall model (post-collision)
@@ -96,13 +101,16 @@ private:
//! \param boundaryCondition: a kernel function for the boundary condition
//! \param bcStruct: a struct containing the grid nodes which are part of the boundary condition
//! \param bcName: the name of the checked boundary condition
- template <typename bcFunction>
- void checkBoundaryCondition(const bcFunction &boundaryCondition, const QforBoundaryConditions &bcStruct, const std::string &bcName)
+ template <typename bcFunction, typename QforBC>
+ void checkBoundaryCondition(const bcFunction &boundaryCondition, const QforBC &bcStruct, const std::string &bcName)
{
if (!boundaryCondition && bcStruct.numberOfBCnodes > 0)
throw std::runtime_error("The boundary condition " + bcName + " was not set!");
}
+ void runDistributionPrecursorBCKernelPost(int level, uint t, CudaMemoryManager* cudaMemoryManager);
+ void runVelocityPrecursorBCKernelPost(int level, uint t, CudaMemoryManager* cudaMemoryManager);
+
SPtr<Parameter> para;
boundaryCondition velocityBoundaryConditionPost = nullptr;
@@ -111,5 +119,6 @@ private:
boundaryCondition pressureBoundaryConditionPre = nullptr;
boundaryCondition geometryBoundaryConditionPost = nullptr;
boundaryConditionWithParameter stressBoundaryConditionPost = nullptr;
+ precursorBoundaryCondition precursorBoundaryConditionPost = nullptr;
};
#endif
diff --git a/src/gpu/VirtualFluids_GPU/KernelManager/BCKernelManagerTest.cpp b/src/gpu/VirtualFluids_GPU/KernelManager/BCKernelManagerTest.cpp
index d55fa51bd8a225dd4e89e684bc81cd56f3f450c0..a0e02112e821eedcfeb013d3465529f668309529 100644
--- a/src/gpu/VirtualFluids_GPU/KernelManager/BCKernelManagerTest.cpp
+++ b/src/gpu/VirtualFluids_GPU/KernelManager/BCKernelManagerTest.cpp
@@ -53,3 +53,9 @@ TEST_F(BCKernelManagerTest_BCsNotSpecified, stressBoundaryConditionPost_NotSpeci
para->getParD(0)->stressBC.numberOfBCnodes = 1;
EXPECT_THROW(BCKernelManager(para, &bcFactory), std::runtime_error);
}
+
+TEST_F(BCKernelManagerTest_BCsNotSpecified, precursorBoundaryConditionPost_NotSpecified)
+{
+ para->getParD(0)->precursorBC.numberOfBCnodes = 1;
+ EXPECT_THROW(BCKernelManager(para, &bcFactory), std::runtime_error);
+}
diff --git a/src/gpu/VirtualFluids_GPU/KernelManager/GridScalingKernelManager.cpp b/src/gpu/VirtualFluids_GPU/KernelManager/GridScalingKernelManager.cpp
index c3129e31a9c750a012a26d58961062eaf3f40add..2b6a266c0d4e5f523091fa4982eee5d83b2ec675 100644
--- a/src/gpu/VirtualFluids_GPU/KernelManager/GridScalingKernelManager.cpp
+++ b/src/gpu/VirtualFluids_GPU/KernelManager/GridScalingKernelManager.cpp
@@ -59,8 +59,9 @@ GridScalingKernelManager::GridScalingKernelManager(SPtr<Parameter> parameter, Gr
VF_LOG_TRACE("Function for scalingCoarseToFine is nullptr");
}
-void GridScalingKernelManager::runFineToCoarseKernelLB(const int level, InterpolationCellFC *icellFC, OffFC &offFC, int streamIndex) const{
- cudaStream_t stream = (streamIndex == -1) ? CU_STREAM_LEGACY : para->getStreamManager()->getStream(streamIndex);
+void GridScalingKernelManager::runFineToCoarseKernelLB(const int level, InterpolationCellFC *icellFC, OffFC &offFC, CudaStreamIndex streamIndex) const
+{
+ cudaStream_t stream = para->getStreamManager()->getStream(streamIndex);
this->scalingFineToCoarse(para->getParD(level).get(), para->getParD(level+1).get(), icellFC, offFC, stream);
@@ -327,9 +328,9 @@ void GridScalingKernelManager::runFineToCoarseKernelAD(const int level) const
}
}
-void GridScalingKernelManager::runCoarseToFineKernelLB(const int level, InterpolationCellCF* icellCF, OffCF &offCF, int streamIndex) const
+void GridScalingKernelManager::runCoarseToFineKernelLB(const int level, InterpolationCellCF* icellCF, OffCF &offCF, CudaStreamIndex streamIndex) const
{
- cudaStream_t stream = (streamIndex == -1) ? CU_STREAM_LEGACY : para->getStreamManager()->getStream(streamIndex);
+ cudaStream_t stream = para->getStreamManager()->getStream(streamIndex);
this->scalingCoarseToFine(para->getParD(level).get(), para->getParD(level+1).get(), icellCF, offCF, stream);
// ScaleCF_comp_D3Q27F3(
diff --git a/src/gpu/VirtualFluids_GPU/KernelManager/GridScalingKernelManager.h b/src/gpu/VirtualFluids_GPU/KernelManager/GridScalingKernelManager.h
index 85cdd88ec2e3a6622108026ce8f53c5c770f8afe..3c78ee7f9db254556e8ec6dbbafaf51cd995f10b 100644
--- a/src/gpu/VirtualFluids_GPU/KernelManager/GridScalingKernelManager.h
+++ b/src/gpu/VirtualFluids_GPU/KernelManager/GridScalingKernelManager.h
@@ -44,6 +44,7 @@
class Parameter;
class CudaMemoryManager;
class GridScalingFactory;
+enum class CudaStreamIndex;
struct LBMSimulationParameter;
struct CUstream_st;
@@ -62,14 +63,14 @@ public:
//! \throws std::runtime_error when the user forgets to specify a scaling function
GridScalingKernelManager(SPtr<Parameter> parameter, GridScalingFactory *gridScalingFactory);
- //! \brief calls the device function of the fine to coarse grid interpolation kernel
- void runFineToCoarseKernelLB(const int level, InterpolationCellFC *icellFC, OffFC &offFC, int streamIndex) const;
+ //! \brief calls the device function of the fine to coarse grid interpolation kernelH
+ void runFineToCoarseKernelLB(const int level, InterpolationCellFC *icellFC, OffFC &offFC, CudaStreamIndex streamIndex) const;
//! \brief calls the device function of the fine to coarse grid interpolation kernel (advection diffusion)
void runFineToCoarseKernelAD(const int level) const;
//! \brief calls the device function of the coarse to fine grid interpolation kernel
- void runCoarseToFineKernelLB(const int level, InterpolationCellCF *icellCF, OffCF &offCF, int streamIndex) const;
+ void runCoarseToFineKernelLB(const int level, InterpolationCellCF *icellCF, OffCF &offCF, CudaStreamIndex streamIndex) const;
//! \brief calls the device function of the coarse to fine grid interpolation kernel (advection diffusion)
void runCoarseToFineKernelAD(const int level) const;
diff --git a/src/gpu/VirtualFluids_GPU/LBM/GPUHelperFunctions/ChimeraTransformation.h b/src/gpu/VirtualFluids_GPU/LBM/GPUHelperFunctions/ChimeraTransformation.h
new file mode 100644
index 0000000000000000000000000000000000000000..225f615ec3ad2d8ef11ec295f8d9e8a4166d99fe
--- /dev/null
+++ b/src/gpu/VirtualFluids_GPU/LBM/GPUHelperFunctions/ChimeraTransformation.h
@@ -0,0 +1,108 @@
+//=======================================================================================
+// ____ ____ __ ______ __________ __ __ __ __
+// \ \ | | | | | _ \ |___ ___| | | | | / \ | |
+// \ \ | | | | | |_) | | | | | | | / \ | |
+// \ \ | | | | | _ / | | | | | | / /\ \ | |
+// \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
+// \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
+// \ \ | | ________________________________________________________________
+// \ \ | | | ______________________________________________________________|
+// \ \| | | | __ __ __ __ ______ _______
+// \ | | |_____ | | | | | | | | | _ \ / _____)
+// \ | | _____| | | | | | | | | | | \ \ \_______
+// \ | | | | |_____ | \_/ | | | | |_/ / _____ |
+// \ _____| |__| |________| \_______/ |__| |______/ (_______/
+//
+// This file is part of VirtualFluids. VirtualFluids is free software: you can
+// redistribute it and/or modify it under the terms of the GNU General Public
+// License as published by the Free Software Foundation, either version 3 of
+// the License, or (at your option) any later version.
+//
+// VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
+// ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+// FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
+// for more details.
+//
+// You should have received a copy of the GNU General Public License along
+// with VirtualFluids (see COPYING.txt). If not, see <http://www.gnu.org/licenses/>.
+//
+//! \file ChimeraTransformation.h
+//! \ingroup LBM/GPUHelperFunctions
+//! \author Martin Schoenherr, Anna Wellmann, Soeren Peters
+//=======================================================================================
+#ifndef CHIMERA_TRANSFORMATION_H
+#define CHIMERA_TRANSFORMATION_H
+
+#include "LBM/LB.h"
+
+#include <lbm/constants/NumericConstants.h>
+
+using namespace vf::lbm::constant;
+
+namespace vf::gpu
+{
+
+////////////////////////////////////////////////////////////////////////////////
+//! \brief forward chimera transformation \ref forwardInverseChimeraWithK
+//! Transformation from distributions to central moments according to Eq. (6)-(14) in \ref
+//! <a href="https://doi.org/10.1016/j.jcp.2017.05.040"><b>[ M. Geier et al. (2017), DOI:10.1016/j.jcp.2017.05.040
+//! ]</b></a> Modified for lower round-off errors.
+__inline__ __device__ void forwardInverseChimeraWithK(real &mfa, real &mfb, real &mfc, real vv, real v2, real Kinverse, real K)
+{
+ real m2 = mfa + mfc;
+ real m1 = mfc - mfa;
+ real m0 = m2 + mfb;
+ mfa = m0;
+ m0 *= Kinverse;
+ m0 += c1o1;
+ mfb = (m1 * Kinverse - m0 * vv) * K;
+ mfc = ((m2 - c2o1 * m1 * vv) * Kinverse + v2 * m0) * K;
+}
+
+////////////////////////////////////////////////////////////////////////////////
+//! \brief backward chimera transformation \ref backwardInverseChimeraWithK
+//! Transformation from central moments to distributions according to Eq. (57)-(65) in \ref
+//! <a href="https://doi.org/10.1016/j.jcp.2017.05.040"><b>[ M. Geier et al. (2017), DOI:10.1016/j.jcp.2017.05.040
+//! ]</b></a> Modified for lower round-off errors.
+__inline__ __device__ void backwardInverseChimeraWithK(real &mfa, real &mfb, real &mfc, real vv, real v2, real Kinverse, real K)
+{
+ real m0 = (((mfc - mfb) * c1o2 + mfb * vv) * Kinverse + (mfa * Kinverse + c1o1) * (v2 - vv) * c1o2) * K;
+ real m1 = (((mfa - mfc) - c2o1 * mfb * vv) * Kinverse + (mfa * Kinverse + c1o1) * (-v2)) * K;
+ mfc = (((mfc + mfb) * c1o2 + mfb * vv) * Kinverse + (mfa * Kinverse + c1o1) * (v2 + vv) * c1o2) * K;
+ mfa = m0;
+ mfb = m1;
+}
+
+////////////////////////////////////////////////////////////////////////////////
+//! \brief forward chimera transformation \ref forwardChimera
+//! Transformation from distributions to central moments according to Eq. (6)-(14) in \ref
+//! <a href="https://doi.org/10.1016/j.jcp.2017.05.040"><b>[ M. Geier et al. (2017), DOI:10.1016/j.jcp.2017.05.040
+//! ]</b></a> for \f$ K_{abc}=0 \f$. This is to avoid unnessary floating point operations. Modified for lower round-off
+//! errors.
+__inline__ __device__ void forwardChimera(real &mfa, real &mfb, real &mfc, real vv, real v2)
+{
+ real m1 = (mfa + mfc) + mfb;
+ real m2 = mfc - mfa;
+ mfc = (mfc + mfa) + (v2 * m1 - c2o1 * vv * m2);
+ mfb = m2 - vv * m1;
+ mfa = m1;
+}
+
+////////////////////////////////////////////////////////////////////////////////
+//! \brief backward chimera transformation \ref backwardChimera
+//! Transformation from central moments to distributions according to Eq. (57)-(65) in \ref
+//! <a href="https://doi.org/10.1016/j.jcp.2017.05.040"><b>[ M. Geier et al. (2017), DOI:10.1016/j.jcp.2017.05.040
+//! ]</b></a> for \f$ K_{abc}=0 \f$. This is to avoid unnessary floating point operations. Modified for lower round-off
+//! errors.
+__inline__ __device__ void backwardChimera(real &mfa, real &mfb, real &mfc, real vv, real v2)
+{
+ real ma = (mfc + mfa * (v2 - vv)) * c1o2 + mfb * (vv - c1o2);
+ real mb = ((mfa - mfc) - mfa * v2) - c2o1 * mfb * vv;
+ mfc = (mfc + mfa * (v2 + vv)) * c1o2 + mfb * (vv + c1o2);
+ mfb = mb;
+ mfa = ma;
+}
+
+} // namespace vf::gpu
+
+#endif
diff --git a/src/gpu/VirtualFluids_GPU/LBM/GPUHelperFunctions/KernelUtilities.h b/src/gpu/VirtualFluids_GPU/LBM/GPUHelperFunctions/KernelUtilities.h
new file mode 100644
index 0000000000000000000000000000000000000000..37208ee59586533fa7f8ffbc269246826ed27fb8
--- /dev/null
+++ b/src/gpu/VirtualFluids_GPU/LBM/GPUHelperFunctions/KernelUtilities.h
@@ -0,0 +1,198 @@
+//=======================================================================================
+// ____ ____ __ ______ __________ __ __ __ __
+// \ \ | | | | | _ \ |___ ___| | | | | / \ | |
+// \ \ | | | | | |_) | | | | | | | / \ | |
+// \ \ | | | | | _ / | | | | | | / /\ \ | |
+// \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
+// \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
+// \ \ | | ________________________________________________________________
+// \ \ | | | ______________________________________________________________|
+// \ \| | | | __ __ __ __ ______ _______
+// \ | | |_____ | | | | | | | | | _ \ / _____)
+// \ | | _____| | | | | | | | | | | \ \ \_______
+// \ | | | | |_____ | \_/ | | | | |_/ / _____ |
+// \ _____| |__| |________| \_______/ |__| |______/ (_______/
+//
+// This file is part of VirtualFluids. VirtualFluids is free software: you can
+// redistribute it and/or modify it under the terms of the GNU General Public
+// License as published by the Free Software Foundation, either version 3 of
+// the License, or (at your option) any later version.
+//
+// VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
+// ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+// FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
+// for more details.
+//
+// You should have received a copy of the GNU General Public License along
+// with VirtualFluids (see COPYING.txt). If not, see <http://www.gnu.org/licenses/>.
+//
+//! \file KernelUtilities.h
+//! \ingroup LBM/GPUHelperFunctions
+//! \author Martin Schoenherr, Anna Wellmann, Soeren Peters
+//=======================================================================================
+#ifndef KERNEL_UTILITIES_H
+#define KERNEL_UTILITIES_H
+
+#include "LBM/LB.h"
+#include "lbm/constants/D3Q27.h"
+#include "lbm/constants/NumericConstants.h"
+
+using namespace vf::lbm::constant;
+using namespace vf::lbm::dir;
+
+namespace vf::gpu
+{
+
+__inline__ __device__ __host__ void getPointersToDistributions(Distributions27 &dist, real *distributionArray, const unsigned long long numberOfLBnodes, const bool isEvenTimestep)
+{
+ if (isEvenTimestep)
+ {
+ dist.f[DIR_000] = &distributionArray[DIR_000 * numberOfLBnodes];
+ dist.f[DIR_P00] = &distributionArray[DIR_P00 * numberOfLBnodes];
+ dist.f[DIR_M00] = &distributionArray[DIR_M00 * numberOfLBnodes];
+ dist.f[DIR_0P0] = &distributionArray[DIR_0P0 * numberOfLBnodes];
+ dist.f[DIR_0M0] = &distributionArray[DIR_0M0 * numberOfLBnodes];
+ dist.f[DIR_00P] = &distributionArray[DIR_00P * numberOfLBnodes];
+ dist.f[DIR_00M] = &distributionArray[DIR_00M * numberOfLBnodes];
+ dist.f[DIR_PP0] = &distributionArray[DIR_PP0 * numberOfLBnodes];
+ dist.f[DIR_MM0] = &distributionArray[DIR_MM0 * numberOfLBnodes];
+ dist.f[DIR_PM0] = &distributionArray[DIR_PM0 * numberOfLBnodes];
+ dist.f[DIR_MP0] = &distributionArray[DIR_MP0 * numberOfLBnodes];
+ dist.f[DIR_P0P] = &distributionArray[DIR_P0P * numberOfLBnodes];
+ dist.f[DIR_M0M] = &distributionArray[DIR_M0M * numberOfLBnodes];
+ dist.f[DIR_P0M] = &distributionArray[DIR_P0M * numberOfLBnodes];
+ dist.f[DIR_M0P] = &distributionArray[DIR_M0P * numberOfLBnodes];
+ dist.f[DIR_0PP] = &distributionArray[DIR_0PP * numberOfLBnodes];
+ dist.f[DIR_0MM] = &distributionArray[DIR_0MM * numberOfLBnodes];
+ dist.f[DIR_0PM] = &distributionArray[DIR_0PM * numberOfLBnodes];
+ dist.f[DIR_0MP] = &distributionArray[DIR_0MP * numberOfLBnodes];
+ dist.f[DIR_PPP] = &distributionArray[DIR_PPP * numberOfLBnodes];
+ dist.f[DIR_MMP] = &distributionArray[DIR_MMP * numberOfLBnodes];
+ dist.f[DIR_PMP] = &distributionArray[DIR_PMP * numberOfLBnodes];
+ dist.f[DIR_MPP] = &distributionArray[DIR_MPP * numberOfLBnodes];
+ dist.f[DIR_PPM] = &distributionArray[DIR_PPM * numberOfLBnodes];
+ dist.f[DIR_MMM] = &distributionArray[DIR_MMM * numberOfLBnodes];
+ dist.f[DIR_PMM] = &distributionArray[DIR_PMM * numberOfLBnodes];
+ dist.f[DIR_MPM] = &distributionArray[DIR_MPM * numberOfLBnodes];
+ }
+ else
+ {
+ dist.f[DIR_M00] = &distributionArray[DIR_P00 * numberOfLBnodes];
+ dist.f[DIR_P00] = &distributionArray[DIR_M00 * numberOfLBnodes];
+ dist.f[DIR_0M0] = &distributionArray[DIR_0P0 * numberOfLBnodes];
+ dist.f[DIR_0P0] = &distributionArray[DIR_0M0 * numberOfLBnodes];
+ dist.f[DIR_00M] = &distributionArray[DIR_00P * numberOfLBnodes];
+ dist.f[DIR_00P] = &distributionArray[DIR_00M * numberOfLBnodes];
+ dist.f[DIR_MM0] = &distributionArray[DIR_PP0 * numberOfLBnodes];
+ dist.f[DIR_PP0] = &distributionArray[DIR_MM0 * numberOfLBnodes];
+ dist.f[DIR_MP0] = &distributionArray[DIR_PM0 * numberOfLBnodes];
+ dist.f[DIR_PM0] = &distributionArray[DIR_MP0 * numberOfLBnodes];
+ dist.f[DIR_M0M] = &distributionArray[DIR_P0P * numberOfLBnodes];
+ dist.f[DIR_P0P] = &distributionArray[DIR_M0M * numberOfLBnodes];
+ dist.f[DIR_M0P] = &distributionArray[DIR_P0M * numberOfLBnodes];
+ dist.f[DIR_P0M] = &distributionArray[DIR_M0P * numberOfLBnodes];
+ dist.f[DIR_0MM] = &distributionArray[DIR_0PP * numberOfLBnodes];
+ dist.f[DIR_0PP] = &distributionArray[DIR_0MM * numberOfLBnodes];
+ dist.f[DIR_0MP] = &distributionArray[DIR_0PM * numberOfLBnodes];
+ dist.f[DIR_0PM] = &distributionArray[DIR_0MP * numberOfLBnodes];
+ dist.f[DIR_000] = &distributionArray[DIR_000 * numberOfLBnodes];
+ dist.f[DIR_PPP] = &distributionArray[DIR_MMM * numberOfLBnodes];
+ dist.f[DIR_MMP] = &distributionArray[DIR_PPM * numberOfLBnodes];
+ dist.f[DIR_PMP] = &distributionArray[DIR_MPM * numberOfLBnodes];
+ dist.f[DIR_MPP] = &distributionArray[DIR_PMM * numberOfLBnodes];
+ dist.f[DIR_PPM] = &distributionArray[DIR_MMP * numberOfLBnodes];
+ dist.f[DIR_MMM] = &distributionArray[DIR_PPP * numberOfLBnodes];
+ dist.f[DIR_PMM] = &distributionArray[DIR_MPP * numberOfLBnodes];
+ dist.f[DIR_MPM] = &distributionArray[DIR_PMP * numberOfLBnodes];
+ }
+}
+
+__inline__ __device__ void getPointersToSubgridDistances(SubgridDistances27& subgridD, real* subgridDistances, const unsigned int numberOfSubgridIndices)
+{
+ subgridD.q[DIR_P00] = &subgridDistances[DIR_P00 * numberOfSubgridIndices];
+ subgridD.q[DIR_M00] = &subgridDistances[DIR_M00 * numberOfSubgridIndices];
+ subgridD.q[DIR_0P0] = &subgridDistances[DIR_0P0 * numberOfSubgridIndices];
+ subgridD.q[DIR_0M0] = &subgridDistances[DIR_0M0 * numberOfSubgridIndices];
+ subgridD.q[DIR_00P] = &subgridDistances[DIR_00P * numberOfSubgridIndices];
+ subgridD.q[DIR_00M] = &subgridDistances[DIR_00M * numberOfSubgridIndices];
+ subgridD.q[DIR_PP0] = &subgridDistances[DIR_PP0 * numberOfSubgridIndices];
+ subgridD.q[DIR_MM0] = &subgridDistances[DIR_MM0 * numberOfSubgridIndices];
+ subgridD.q[DIR_PM0] = &subgridDistances[DIR_PM0 * numberOfSubgridIndices];
+ subgridD.q[DIR_MP0] = &subgridDistances[DIR_MP0 * numberOfSubgridIndices];
+ subgridD.q[DIR_P0P] = &subgridDistances[DIR_P0P * numberOfSubgridIndices];
+ subgridD.q[DIR_M0M] = &subgridDistances[DIR_M0M * numberOfSubgridIndices];
+ subgridD.q[DIR_P0M] = &subgridDistances[DIR_P0M * numberOfSubgridIndices];
+ subgridD.q[DIR_M0P] = &subgridDistances[DIR_M0P * numberOfSubgridIndices];
+ subgridD.q[DIR_0PP] = &subgridDistances[DIR_0PP * numberOfSubgridIndices];
+ subgridD.q[DIR_0MM] = &subgridDistances[DIR_0MM * numberOfSubgridIndices];
+ subgridD.q[DIR_0PM] = &subgridDistances[DIR_0PM * numberOfSubgridIndices];
+ subgridD.q[DIR_0MP] = &subgridDistances[DIR_0MP * numberOfSubgridIndices];
+ subgridD.q[DIR_000] = &subgridDistances[DIR_000 * numberOfSubgridIndices];
+ subgridD.q[DIR_PPP] = &subgridDistances[DIR_PPP * numberOfSubgridIndices];
+ subgridD.q[DIR_MMP] = &subgridDistances[DIR_MMP * numberOfSubgridIndices];
+ subgridD.q[DIR_PMP] = &subgridDistances[DIR_PMP * numberOfSubgridIndices];
+ subgridD.q[DIR_MPP] = &subgridDistances[DIR_MPP * numberOfSubgridIndices];
+ subgridD.q[DIR_PPM] = &subgridDistances[DIR_PPM * numberOfSubgridIndices];
+ subgridD.q[DIR_MMM] = &subgridDistances[DIR_MMM * numberOfSubgridIndices];
+ subgridD.q[DIR_PMM] = &subgridDistances[DIR_PMM * numberOfSubgridIndices];
+ subgridD.q[DIR_MPM] = &subgridDistances[DIR_MPM * numberOfSubgridIndices];
+}
+
+__inline__ __device__ real getEquilibriumForBC(const real& drho, const real& velocity, const real& cu_sq, const real weight)
+{
+ return weight * (drho + c9o2 * velocity * velocity * (c1o1 + drho) - cu_sq);
+}
+
+__inline__ __device__ real getInterpolatedDistributionForVeloBC(const real& q, const real& f, const real& fInverse, const real& feq,
+ const real& omega, const real& velocity, const real weight)
+{
+
+ return (c1o1-q) / (c1o1+q) * (f - fInverse + (f + fInverse - c2o1 * feq * omega) / (c1o1 - omega)) * c1o2
+ + (q * (f + fInverse) - c6o1 * weight * velocity) / (c1o1 + q);
+}
+
+__inline__ __device__ real getBounceBackDistributionForVeloBC( const real& f,
+ const real& velocity, const real weight)
+{
+
+ return f - (c6o1 * weight * velocity);
+}
+
+__inline__ __device__ real getInterpolatedDistributionForNoSlipBC(const real& q, const real& f, const real& fInverse, const real& feq,
+ const real& omega)
+{
+
+ return (c1o1-q) / (c1o1+q) * (f - fInverse + (f + fInverse - c2o1 * feq * omega) / (c1o1 - omega)) * c1o2
+ + (q * (f + fInverse)) / (c1o1 + q);
+}
+
+
+__inline__ __device__ real getInterpolatedDistributionForVeloWithPressureBC(const real& q, const real& f, const real& fInverse, const real& feq,
+ const real& omega, const real& drho, const real& velocity, const real weight)
+{
+
+ return (c1o1-q) / (c1o1+q) * (f - fInverse + (f + fInverse - c2o1 * feq * omega) / (c1o1 - omega)) * c1o2
+ + (q * (f + fInverse) - c6o1 * weight * velocity) / (c1o1 + q) - weight * drho;
+}
+
+__inline__ __device__ unsigned int getNodeIndex()
+{
+ const unsigned x = threadIdx.x;
+ const unsigned y = blockIdx.x;
+ const unsigned z = blockIdx.y;
+
+ const unsigned nx = blockDim.x;
+ const unsigned ny = gridDim.x;
+
+ return nx * (ny * z + y) + x;
+}
+
+__inline__ __device__ bool isValidFluidNode(uint nodeType)
+{
+ return (nodeType == GEO_FLUID || nodeType == GEO_PM_0 || nodeType == GEO_PM_1 || nodeType == GEO_PM_2);
+}
+
+
+}
+
+#endif
diff --git a/src/gpu/VirtualFluids_GPU/LBM/GPUHelperFunctions/ScalingUtilities.h b/src/gpu/VirtualFluids_GPU/LBM/GPUHelperFunctions/ScalingUtilities.h
new file mode 100644
index 0000000000000000000000000000000000000000..53990e452be06dc6840c801816e8231d26861e2e
--- /dev/null
+++ b/src/gpu/VirtualFluids_GPU/LBM/GPUHelperFunctions/ScalingUtilities.h
@@ -0,0 +1,136 @@
+//=======================================================================================
+// ____ ____ __ ______ __________ __ __ __ __
+// \ \ | | | | | _ \ |___ ___| | | | | / \ | |
+// \ \ | | | | | |_) | | | | | | | / \ | |
+// \ \ | | | | | _ / | | | | | | / /\ \ | |
+// \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
+// \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
+// \ \ | | ________________________________________________________________
+// \ \ | | | ______________________________________________________________|
+// \ \| | | | __ __ __ __ ______ _______
+// \ | | |_____ | | | | | | | | | _ \ / _____)
+// \ | | _____| | | | | | | | | | | \ \ \_______
+// \ | | | | |_____ | \_/ | | | | |_/ / _____ |
+// \ _____| |__| |________| \_______/ |__| |______/ (_______/
+//
+// This file is part of VirtualFluids. VirtualFluids is free software: you can
+// redistribute it and/or modify it under the terms of the GNU General Public
+// License as published by the Free Software Foundation, either version 3 of
+// the License, or (at your option) any later version.
+//
+// VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
+// ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+// FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
+// for more details.
+//
+// You should have received a copy of the GNU General Public License along
+// with VirtualFluids (see COPYING.txt). If not, see <http://www.gnu.org/licenses/>.
+//
+//! \file ScalingUtilities.h
+//! \ingroup LBM/GPUHelperFunctions
+//! \author Martin Schoenherr, Anna Wellmann
+//=======================================================================================
+#ifndef SCALING_HELPER_FUNCTIONS_H
+#define SCALING_HELPER_FUNCTIONS_H
+
+#include "LBM/LB.h"
+#include "lbm/constants/D3Q27.h"
+#include "lbm/constants/NumericConstants.h"
+
+using namespace vf::lbm::constant;
+using namespace vf::lbm::dir;
+
+namespace vf::gpu
+{
+
+__device__ __inline__ void calculateMomentsOnSourceNodes(Distributions27 &dist, real &omega, unsigned int &k_000,
+ unsigned int &k_M00, unsigned int &k_0M0, unsigned int &k_00M,
+ unsigned int &k_MM0, unsigned int &k_M0M, unsigned int &k_0MM,
+ unsigned int &k_MMM, real &drho, real &velocityX,
+ real &velocityY, real &velocityZ, real &kxyFromfcNEQ,
+ real &kyzFromfcNEQ, real &kxzFromfcNEQ, real &kxxMyyFromfcNEQ,
+ real &kxxMzzFromfcNEQ)
+{
+ ////////////////////////////////////////////////////////////////////////////////////
+ //! - Set local distributions (f's) on source nodes:
+ //!
+ real f_000 = (dist.f[DIR_000])[k_000];
+ real f_P00 = (dist.f[DIR_P00])[k_000];
+ real f_M00 = (dist.f[DIR_M00])[k_M00];
+ real f_0P0 = (dist.f[DIR_0P0])[k_000];
+ real f_0M0 = (dist.f[DIR_0M0])[k_0M0];
+ real f_00P = (dist.f[DIR_00P])[k_000];
+ real f_00M = (dist.f[DIR_00M])[k_00M];
+ real f_PP0 = (dist.f[DIR_PP0])[k_000];
+ real f_MM0 = (dist.f[DIR_MM0])[k_MM0];
+ real f_PM0 = (dist.f[DIR_PM0])[k_0M0];
+ real f_MP0 = (dist.f[DIR_MP0])[k_M00];
+ real f_P0P = (dist.f[DIR_P0P])[k_000];
+ real f_M0M = (dist.f[DIR_M0M])[k_M0M];
+ real f_P0M = (dist.f[DIR_P0M])[k_00M];
+ real f_M0P = (dist.f[DIR_M0P])[k_M00];
+ real f_0PP = (dist.f[DIR_0PP])[k_000];
+ real f_0MM = (dist.f[DIR_0MM])[k_0MM];
+ real f_0PM = (dist.f[DIR_0PM])[k_00M];
+ real f_0MP = (dist.f[DIR_0MP])[k_0M0];
+ real f_PPP = (dist.f[DIR_PPP])[k_000];
+ real f_MPP = (dist.f[DIR_MPP])[k_M00];
+ real f_PMP = (dist.f[DIR_PMP])[k_0M0];
+ real f_MMP = (dist.f[DIR_MMP])[k_MM0];
+ real f_PPM = (dist.f[DIR_PPM])[k_00M];
+ real f_MPM = (dist.f[DIR_MPM])[k_M0M];
+ real f_PMM = (dist.f[DIR_PMM])[k_0MM];
+ real f_MMM = (dist.f[DIR_MMM])[k_MMM];
+
+ ////////////////////////////////////////////////////////////////////////////////////
+ //! - Calculate density and velocity using pyramid summation for low round-off errors as in Eq. (J1)-(J3) \ref
+ //! <a href="https://doi.org/10.1016/j.camwa.2015.05.001"><b>[ M. Geier et al. (2015),
+ //! DOI:10.1016/j.camwa.2015.05.001 ]</b></a>
+ //!
+ drho = ((((f_PPP + f_MMM) + (f_MPM + f_PMP)) + ((f_MPP + f_PMM) + (f_MMP + f_PPM))) +
+ (((f_0MP + f_0PM) + (f_0MM + f_0PP)) + ((f_M0P + f_P0M) + (f_M0M + f_P0P)) +
+ ((f_MP0 + f_PM0) + (f_MM0 + f_PP0))) +
+ ((f_M00 + f_P00) + (f_0M0 + f_0P0) + (f_00M + f_00P))) +
+ f_000;
+
+ real oneOverRho = c1o1 / (c1o1 + drho);
+
+ velocityX = ((((f_PPP - f_MMM) + (f_PMP - f_MPM)) + ((f_PMM - f_MPP) + (f_PPM - f_MMP))) +
+ (((f_P0M - f_M0P) + (f_P0P - f_M0M)) + ((f_PM0 - f_MP0) + (f_PP0 - f_MM0))) + (f_P00 - f_M00)) *
+ oneOverRho;
+ velocityY = ((((f_PPP - f_MMM) + (f_MPM - f_PMP)) + ((f_MPP - f_PMM) + (f_PPM - f_MMP))) +
+ (((f_0PM - f_0MP) + (f_0PP - f_0MM)) + ((f_MP0 - f_PM0) + (f_PP0 - f_MM0))) + (f_0P0 - f_0M0)) *
+ oneOverRho;
+ velocityZ = ((((f_PPP - f_MMM) + (f_PMP - f_MPM)) + ((f_MPP - f_PMM) + (f_MMP - f_PPM))) +
+ (((f_0MP - f_0PM) + (f_0PP - f_0MM)) + ((f_M0P - f_P0M) + (f_P0P - f_M0M))) + (f_00P - f_00M)) *
+ oneOverRho;
+
+ ////////////////////////////////////////////////////////////////////////////////////
+ //! - Calculate second order moments for interpolation
+ //!
+ // example: kxxMzz: moment, second derivative in x direction minus the second derivative in z direction
+ kxyFromfcNEQ = -c3o1 * omega *
+ ((f_MM0 + f_MMM + f_MMP - f_MP0 - f_MPM - f_MPP - f_PM0 - f_PMM - f_PMP + f_PP0 + f_PPM + f_PPP) /
+ (c1o1 + drho) -
+ ((velocityX * velocityY)));
+ kyzFromfcNEQ = -c3o1 * omega *
+ ((f_0MM + f_PMM + f_MMM - f_0MP - f_PMP - f_MMP - f_0PM - f_PPM - f_MPM + f_0PP + f_PPP + f_MPP) /
+ (c1o1 + drho) -
+ ((velocityY * velocityZ)));
+ kxzFromfcNEQ = -c3o1 * omega *
+ ((f_M0M + f_MMM + f_MPM - f_M0P - f_MMP - f_MPP - f_P0M - f_PMM - f_PPM + f_P0P + f_PMP + f_PPP) /
+ (c1o1 + drho) -
+ ((velocityX * velocityZ)));
+ kxxMyyFromfcNEQ = -c3o2 * omega *
+ ((f_M0M + f_M00 + f_M0P - f_0MM - f_0M0 - f_0MP - f_0PM - f_0P0 - f_0PP + f_P0M + f_P00 + f_P0P) /
+ (c1o1 + drho) -
+ ((velocityX * velocityX - velocityY * velocityY)));
+ kxxMzzFromfcNEQ = -c3o2 * omega *
+ ((f_MM0 + f_M00 + f_MP0 - f_0MM - f_0MP - f_00M - f_00P - f_0PM - f_0PP + f_PM0 + f_P00 + f_PP0) /
+ (c1o1 + drho) -
+ ((velocityX * velocityX - velocityZ * velocityZ)));
+}
+
+} // namespace vf::gpu
+
+#endif
diff --git a/src/gpu/VirtualFluids_GPU/LBM/LB.h b/src/gpu/VirtualFluids_GPU/LBM/LB.h
index eea4adfda3c1ef0862f39ef58fc6e065af7bab1b..cfdbbbae040a13f94e97d40d702b93d5a1e19c86 100644
--- a/src/gpu/VirtualFluids_GPU/LBM/LB.h
+++ b/src/gpu/VirtualFluids_GPU/LBM/LB.h
@@ -15,9 +15,9 @@
//////////////////////////
//porous media
-#define GEO_PM_0 5
-#define GEO_PM_1 6
-#define GEO_PM_2 7
+#define GEO_PM_0 5
+#define GEO_PM_1 6
+#define GEO_PM_2 7
//////////////////////////
#define GEO_SOLID 15
@@ -53,17 +53,33 @@
//! \brief An enumeration for selecting a turbulence model
enum class TurbulenceModel {
//! - Smagorinsky
- Smagorinsky,
+ Smagorinsky,
//! - AMD (Anisotropic Minimum Dissipation) model, see e.g. Rozema et al., Phys. Fluids 27, 085107 (2015), https://doi.org/10.1063/1.4928700
- AMD,
+ AMD,
//! - QR model by Verstappen
- QR,
+ QR,
//! - TODO: move the WALE model here from the old kernels
//WALE
//! - No turbulence model
- None
+ None
};
+//! \brief An enumeration for selecting a template of the collision kernel (CumulantK17)
+enum class CollisionTemplate {
+ //! - Default: plain collision without additional read/write
+ Default,
+ //! - WriteMacroVars: collision \w write out macroscopic variables
+ WriteMacroVars,
+ //! - ApplyBodyForce: collision \w read and apply body force in the collision kernel
+ ApplyBodyForce,
+ //! - AllFeatures: collision \w write out macroscopic variables AND read and apply body force
+ AllFeatures,
+ //! - Border: collision on border nodes
+ SubDomainBorder
+};
+constexpr std::initializer_list<CollisionTemplate> all_CollisionTemplate = { CollisionTemplate::Default, CollisionTemplate::WriteMacroVars, CollisionTemplate::ApplyBodyForce, CollisionTemplate::AllFeatures, CollisionTemplate::SubDomainBorder};
+constexpr std::initializer_list<CollisionTemplate> bulk_CollisionTemplate = { CollisionTemplate::Default, CollisionTemplate::WriteMacroVars, CollisionTemplate::ApplyBodyForce, CollisionTemplate::AllFeatures};
+
struct InitCondition
{
real Re;
@@ -144,6 +160,7 @@ struct InitCondition
bool hasWallModelMonitor {false};
bool simulatePorousMedia {false};
bool streetVelocityFile {false};
+ real outflowPressureCorrectionFactor {0.0};
};
//Interface Cells
@@ -174,7 +191,7 @@ typedef struct OffFC{
// Distribution functions g 6
typedef struct Distri6 {
- real* g[6];
+ real* g[6];
} Distributions6;
// Distribution functions f 7
@@ -214,6 +231,21 @@ typedef struct QforBC{
real *normalX, *normalY, *normalZ;
}QforBoundaryConditions;
+typedef struct QforPrecursorBC{
+ int* k;
+ int numberOfBCnodes=0;
+ int sizeQ;
+ int numberOfPrecursorNodes=0;
+ uint nPrecursorReads=0;
+ uint timeStepsBetweenReads;
+ size_t numberOfQuantities;
+ real* q27[27];
+ uint* planeNeighbor0PP, *planeNeighbor0PM, *planeNeighbor0MP, *planeNeighbor0MM;
+ real* weights0PP, *weights0PM, *weights0MP, *weights0MM;
+ real* last, *current, *next;
+ real velocityX, velocityY, velocityZ;
+}QforPrecursorBoundaryConditions;
+
//BCTemp
typedef struct TempforBC{
int* k;
@@ -249,57 +281,56 @@ typedef struct WMparas{
real* Fz;
}WallModelParameters;
+
//measurePoints
typedef struct MeasP{
- std::string name;
- uint k;
- std::vector<real> Vx;
- std::vector<real> Vy;
- std::vector<real> Vz;
- std::vector<real> Rho;
- //real* Vx;
- //real* Vy;
- //real* Vz;
- //real* Rho;
+ std::string name;
+ uint k;
+ std::vector<real> Vx;
+ std::vector<real> Vy;
+ std::vector<real> Vz;
+ std::vector<real> Rho;
+ //real* Vx;
+ //real* Vy;
+ //real* Vz;
+ //real* Rho;
}MeasurePoints;
//Process Neighbors
typedef struct PN27{
- real* f[27];
- uint memsizeFs;
- int* index;
- uint memsizeIndex;
- uint rankNeighbor;
- int numberOfNodes;
- int numberOfFs;
+ real* f[27];
+ uint memsizeFs;
+ int* index;
+ uint memsizeIndex;
+ uint rankNeighbor;
+ int numberOfNodes;
+ int numberOfFs;
}ProcessNeighbor27;
typedef struct PN_F3 {
- real* g[6];
- uint memsizeGs;
- int* index;
- uint memsizeIndex;
- uint rankNeighbor;
- int numberOfNodes;
- int numberOfGs;
+ real* g[6];
+ uint memsizeGs;
+ int* index;
+ uint memsizeIndex;
+ uint rankNeighbor;
+ int numberOfNodes;
+ int numberOfGs;
}ProcessNeighborF3;
//path line particles
typedef struct PLP{
- bool *stuck, *hot;
- real *coordXabsolut, *coordYabsolut, *coordZabsolut;
- real *coordXlocal, *coordYlocal, *coordZlocal;
- real *veloX, *veloY, *veloZ;
- real *randomLocationInit;
- uint *timestep;
- uint *ID;
- uint *cellBaseID;
- uint numberOfParticles, numberOfTimestepsParticles;
- uint memSizeID, memSizeTimestep, memSizerealAll, memSizereal, memSizeBool, memSizeBoolBC;
+ bool *stuck, *hot;
+ real *coordXabsolut, *coordYabsolut, *coordZabsolut;
+ real *coordXlocal, *coordYlocal, *coordZlocal;
+ real *veloX, *veloY, *veloZ;
+ real *randomLocationInit;
+ uint *timestep;
+ uint *ID;
+ uint *cellBaseID;
+ uint numberOfParticles, numberOfTimestepsParticles;
+ uint memSizeID, memSizeTimestep, memSizerealAll, memSizereal, memSizeBool, memSizeBoolBC;
}PathLineParticles;
-
-
//////////////////////////////////////////////////////////////////////////
inline int vectorPosition(int i, int j, int k, int Lx, int Ly )
{
@@ -308,7 +339,4 @@ inline int vectorPosition(int i, int j, int k, int Lx, int Ly )
}
//////////////////////////////////////////////////////////////////////////
-
#endif
-
-
diff --git a/src/gpu/VirtualFluids_GPU/LBM/Simulation.cpp b/src/gpu/VirtualFluids_GPU/LBM/Simulation.cpp
index 703e935e6edb5676c7d6e045a38e3ec20d7a4b41..84ab84ff93fa7706bcc27d7e61a18f580f3c8dbe 100644
--- a/src/gpu/VirtualFluids_GPU/LBM/Simulation.cpp
+++ b/src/gpu/VirtualFluids_GPU/LBM/Simulation.cpp
@@ -97,11 +97,7 @@ void Simulation::init(GridProvider &gridProvider, BoundaryConditionFactory *bcFa
gridProvider.allocAndCopyForcing();
gridProvider.allocAndCopyQuadricLimiters();
- if (para->getKernelNeedsFluidNodeIndicesToRun()) {
- gridProvider.allocArrays_fluidNodeIndices();
- gridProvider.allocArrays_fluidNodeIndicesBorder();
- }
-
+
gridProvider.setDimensions();
gridProvider.setBoundingBox();
@@ -113,12 +109,7 @@ void Simulation::init(GridProvider &gridProvider, BoundaryConditionFactory *bcFa
para->setStartTurn((unsigned int)0); // 100000
restart_object = std::make_shared<ASCIIRestartObject>();
- //////////////////////////////////////////////////////////////////////////
- // CUDA streams
- if (para->getUseStreams()) {
- para->getStreamManager()->launchStreams(2u);
- para->getStreamManager()->createCudaEvents();
- }
+
//////////////////////////////////////////////////////////////////////////
VF_LOG_INFO("LB_Modell: D3Q{}", para->getD3Qxx());
VF_LOG_INFO("Re: {}", para->getRe());
@@ -134,14 +125,32 @@ void Simulation::init(GridProvider &gridProvider, BoundaryConditionFactory *bcFa
//////////////////////////////////////////////////////////////////////////
allocNeighborsOffsetsScalesAndBoundaries(gridProvider);
+ //! Get tagged fluid nodes with corresponding value for CollisionTemplate from interactors
for (SPtr<PreCollisionInteractor> actuator : para->getActuators()) {
actuator->init(para.get(), &gridProvider, cudaMemoryManager.get());
+ actuator->getTaggedFluidNodes( para.get(), &gridProvider );
}
for (SPtr<PreCollisionInteractor> probe : para->getProbes()) {
probe->init(para.get(), &gridProvider, cudaMemoryManager.get());
+ probe->getTaggedFluidNodes( para.get(), &gridProvider );
}
+ //////////////////////////////////////////////////////////////////////////
+ // CUDA streams
+ if (para->getUseStreams()) {
+ para->getStreamManager()->registerStream(CudaStreamIndex::SubDomainBorder);
+ para->getStreamManager()->registerStream(CudaStreamIndex::Bulk);
+ para->getStreamManager()->launchStreams();
+ para->getStreamManager()->createCudaEvents();
+ }
+ //////////////////////////////////////////////////////////////////////////
+
+ if (para->getKernelNeedsFluidNodeIndicesToRun())
+ {
+ gridProvider.sortFluidNodeTags();
+ gridProvider.allocArrays_taggedFluidNodes();
+ }
//////////////////////////////////////////////////////////////////////////
// Kernel init
//////////////////////////////////////////////////////////////////////////
diff --git a/src/gpu/VirtualFluids_GPU/Output/DistributionDebugInspector.cu b/src/gpu/VirtualFluids_GPU/Output/DistributionDebugInspector.cu
new file mode 100644
index 0000000000000000000000000000000000000000..f7bb2e680c0fb3ea597239ee0cbc1772f2efe81b
--- /dev/null
+++ b/src/gpu/VirtualFluids_GPU/Output/DistributionDebugInspector.cu
@@ -0,0 +1,179 @@
+//=======================================================================================
+// ____ ____ __ ______ __________ __ __ __ __
+// \ \ | | | | | _ \ |___ ___| | | | | / \ | |
+// \ \ | | | | | |_) | | | | | | | / \ | |
+// \ \ | | | | | _ / | | | | | | / /\ \ | |
+// \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
+// \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
+// \ \ | | ________________________________________________________________
+// \ \ | | | ______________________________________________________________|
+// \ \| | | | __ __ __ __ ______ _______
+// \ | | |_____ | | | | | | | | | _ \ / _____)
+// \ | | _____| | | | | | | | | | | \ \ \_______
+// \ | | | | |_____ | \_/ | | | | |_/ / _____ |
+// \ _____| |__| |________| \_______/ |__| |______/ (_______/
+//
+// This file is part of VirtualFluids. VirtualFluids is free software: you can
+// redistribute it and/or modify it under the terms of the GNU General Public
+// License as published by the Free Software Foundation, either version 3 of
+// the License, or (at your option) any later version.
+//
+// VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
+// ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+// FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
+// for more details.
+//
+// You should have received a copy of the GNU General Public License along
+// with VirtualFluids (see COPYING.txt). If not, see <http://www.gnu.org/licenses/>.
+//
+//! \file DistributionDebugInspector.cu
+//! \ingroup Output
+//! \author Henrik Asmuth, Henry Korb
+//======================================================================================
+#include "DistributionDebugInspector.h"
+
+#include "Parameter/Parameter.h"
+#include "lbm/constants/D3Q27.h"
+#include "lbm/constants/NumericConstants.h"
+#include "LBM/GPUHelperFunctions/KernelUtilities.h"
+
+#include <cuda/CudaGrid.h>
+#include <cuda.h>
+
+#include <iostream>
+
+using namespace vf::lbm::constant;
+using namespace vf::lbm::dir;
+using namespace vf::gpu;
+
+__global__ void printFs(
+ real* distributions,
+ bool isEvenTimestep,
+ unsigned long long numberOfFluidNodes,
+ uint* neighborX,
+ uint* neighborY,
+ uint* neighborZ,
+ uint* typeOfGridNode,
+ real* coordX,
+ real* coordY,
+ real* coordZ,
+ real minX,
+ real maxX,
+ real minY,
+ real maxY,
+ real minZ,
+ real maxZ)
+{
+ ////////////////////////////////////////////////////////////////////////////////
+ //! - Get node index coordinates from threadIdx, blockIdx, blockDim and gridDim.
+ //!
+ const unsigned k_000 = getNodeIndex();
+
+ if (k_000 >= numberOfFluidNodes || typeOfGridNode[k_000]!=GEO_FLUID )
+ return;
+
+ real coordNodeX = coordX[k_000];
+ real coordNodeY = coordY[k_000];
+ real coordNodeZ = coordZ[k_000];
+
+ if( coordNodeX>=minX && coordNodeX<=maxX &&
+ coordNodeY>=minY && coordNodeY<=maxY &&
+ coordNodeZ>=minZ && coordNodeZ<=maxZ )
+ {
+ Distributions27 dist;
+ getPointersToDistributions(dist, distributions, numberOfFluidNodes, isEvenTimestep);
+ ////////////////////////////////////////////////////////////////////////////////
+ //! - Set neighbor indices (necessary for indirect addressing)
+ uint k_M00 = neighborX[k_000];
+ uint k_0M0 = neighborY[k_000];
+ uint k_00M = neighborZ[k_000];
+ uint k_MM0 = neighborY[k_M00];
+ uint k_M0M = neighborZ[k_M00];
+ uint k_0MM = neighborZ[k_0M0];
+ uint k_MMM = neighborZ[k_MM0];
+ ////////////////////////////////////////////////////////////////////////////////////
+ //! - Set local distributions
+ //!
+ real f_000 = (dist.f[DIR_000])[k_000];
+ real f_P00 = (dist.f[DIR_P00])[k_000];
+ real f_M00 = (dist.f[DIR_M00])[k_M00];
+ real f_0P0 = (dist.f[DIR_0P0])[k_000];
+ real f_0M0 = (dist.f[DIR_0M0])[k_0M0];
+ real f_00P = (dist.f[DIR_00P])[k_000];
+ real f_00M = (dist.f[DIR_00M])[k_00M];
+ real f_PP0 = (dist.f[DIR_PP0])[k_000];
+ real f_MM0 = (dist.f[DIR_MM0])[k_MM0];
+ real f_PM0 = (dist.f[DIR_PM0])[k_0M0];
+ real f_MP0 = (dist.f[DIR_MP0])[k_M00];
+ real f_P0P = (dist.f[DIR_P0P])[k_000];
+ real f_M0M = (dist.f[DIR_M0M])[k_M0M];
+ real f_P0M = (dist.f[DIR_P0M])[k_00M];
+ real f_M0P = (dist.f[DIR_M0P])[k_M00];
+ real f_0PP = (dist.f[DIR_0PP])[k_000];
+ real f_0MM = (dist.f[DIR_0MM])[k_0MM];
+ real f_0PM = (dist.f[DIR_0PM])[k_00M];
+ real f_0MP = (dist.f[DIR_0MP])[k_0M0];
+ real f_PPP = (dist.f[DIR_PPP])[k_000];
+ real f_MPP = (dist.f[DIR_MPP])[k_M00];
+ real f_PMP = (dist.f[DIR_PMP])[k_0M0];
+ real f_MMP = (dist.f[DIR_MMP])[k_MM0];
+ real f_PPM = (dist.f[DIR_PPM])[k_00M];
+ real f_MPM = (dist.f[DIR_MPM])[k_M0M];
+ real f_PMM = (dist.f[DIR_PMM])[k_0MM];
+ real f_MMM = (dist.f[DIR_MMM])[k_MMM];
+
+ real drho = ((((f_PPP + f_MMM) + (f_MPM + f_PMP)) + ((f_MPP + f_PMM) + (f_MMP + f_PPM))) +
+ (((f_0MP + f_0PM) + (f_0MM + f_0PP)) + ((f_M0P + f_P0M) + (f_M0M + f_P0P)) +
+ ((f_MP0 + f_PM0) + (f_MM0 + f_PP0))) +
+ ((f_M00 + f_P00) + (f_0M0 + f_0P0) + (f_00M + f_00P))) +
+ f_000;
+
+ real oneOverRho = c1o1 / (c1o1 + drho);
+
+ real vvx = ((((f_PPP - f_MMM) + (f_PMP - f_MPM)) + ((f_PMM - f_MPP) + (f_PPM - f_MMP))) +
+ (((f_P0M - f_M0P) + (f_P0P - f_M0M)) + ((f_PM0 - f_MP0) + (f_PP0 - f_MM0))) + (f_P00 - f_M00)) *
+ oneOverRho;
+ real vvy = ((((f_PPP - f_MMM) + (f_MPM - f_PMP)) + ((f_MPP - f_PMM) + (f_PPM - f_MMP))) +
+ (((f_0PM - f_0MP) + (f_0PP - f_0MM)) + ((f_MP0 - f_PM0) + (f_PP0 - f_MM0))) + (f_0P0 - f_0M0)) *
+ oneOverRho;
+ real vvz = ((((f_PPP - f_MMM) + (f_PMP - f_MPM)) + ((f_MPP - f_PMM) + (f_MMP - f_PPM))) +
+ (((f_0MP - f_0PM) + (f_0PP - f_0MM)) + ((f_M0P - f_P0M) + (f_P0P - f_M0M))) + (f_00P - f_00M)) *
+ oneOverRho;
+
+ printf("Node %u \t (%f\t%f\t%f)\n rho: %f\t velo: %f\t %f \t %f\n\n" , k_000, coordNodeX, coordNodeY, coordNodeZ, drho, vvx, vvy, vvz);
+ printf("Node %u \t (%f\t%f\t%f)\n f_M00\t%f\t f_000\t%f\t f_P00\t%f\n f_MP0\t%f\t f_0P0\t%f\t f_PP0\t%f\n f_MM0\t%f\t f_0M0\t%f\t f_PM0\t%f\n f_M0P\t%f\t f_00P\t%f\t f_P0P\t%f\n f_M0M\t%f\t f_00M\t%f\t f_P0M\t%f\n f_MPP\t%f\t f_0PP\t%f\t f_PPP\t%f\n f_MPM\t%f\t f_0PM\t%f\t f_PPM\t%f\n f_MMP\t%f\t f_0MP\t%f\t f_PMP\t%f\n f_MMM\t%f\t f_0MM\t%f\t f_PMM\t%f\n\n\n" , k_000, coordNodeX, coordNodeY, coordNodeZ, f_M00, f_000, f_P00,f_MP0, f_0P0, f_PP0, f_MM0, f_0M0, f_PM0, f_M0P, f_00P, f_P0P, f_M0M, f_00M, f_P0M, f_MPP, f_0PP, f_PPP, f_MPM, f_0PM, f_PPM, f_MMP, f_0MP, f_PMP, f_MMM, f_0MM, f_PMM);
+
+ }
+
+}
+
+
+
+
+void DistributionDebugInspector::inspect(std::shared_ptr<Parameter> para, uint level, uint t)
+{
+ if(this->inspectionLevel!=level)
+ return;
+
+ std::cout << tag << ": distributions on level " << level << " at t " << t << std::endl;
+
+ vf::cuda::CudaGrid cudaGrid = vf::cuda::CudaGrid(para->getParD(level)->numberofthreads, para->getParD(level)->numberOfNodes);
+
+ printFs <<< cudaGrid.grid, cudaGrid.threads >>>( para->getParD(level)->distributions.f[0],
+ para->getParD(level)->isEvenTimestep,
+ para->getParD(level)->numberOfNodes,
+ para->getParD(level)->neighborX,
+ para->getParD(level)->neighborY,
+ para->getParD(level)->neighborZ,
+ para->getParD(level)->typeOfGridNode,
+ para->getParD(level)->coordinateX,
+ para->getParD(level)->coordinateY,
+ para->getParD(level)->coordinateZ,
+ minX,
+ maxX,
+ minY,
+ maxY,
+ minZ,
+ maxZ);
+
+}
\ No newline at end of file
diff --git a/src/gpu/VirtualFluids_GPU/Output/DistributionDebugInspector.h b/src/gpu/VirtualFluids_GPU/Output/DistributionDebugInspector.h
new file mode 100644
index 0000000000000000000000000000000000000000..95fea46d4eba0c2f2ff0846d22ee5da4f6c357ea
--- /dev/null
+++ b/src/gpu/VirtualFluids_GPU/Output/DistributionDebugInspector.h
@@ -0,0 +1,76 @@
+//=======================================================================================
+// ____ ____ __ ______ __________ __ __ __ __
+// \ \ | | | | | _ \ |___ ___| | | | | / \ | |
+// \ \ | | | | | |_) | | | | | | | / \ | |
+// \ \ | | | | | _ / | | | | | | / /\ \ | |
+// \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
+// \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
+// \ \ | | ________________________________________________________________
+// \ \ | | | ______________________________________________________________|
+// \ \| | | | __ __ __ __ ______ _______
+// \ | | |_____ | | | | | | | | | _ \ / _____)
+// \ | | _____| | | | | | | | | | | \ \ \_______
+// \ | | | | |_____ | \_/ | | | | |_/ / _____ |
+// \ _____| |__| |________| \_______/ |__| |______/ (_______/
+//
+// This file is part of VirtualFluids. VirtualFluids is free software: you can
+// redistribute it and/or modify it under the terms of the GNU General Public
+// License as published by the Free Software Foundation, either version 3 of
+// the License, or (at your option) any later version.
+//
+// VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
+// ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+// FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
+// for more details.
+//
+// You should have received a copy of the GNU General Public License along
+// with VirtualFluids (see COPYING.txt). If not, see <http://www.gnu.org/licenses/>.
+//
+//! \file DistributionDebugInspector.h
+//! \author Henrik Asmuth
+//! \date 13/012/2022
+//! \brief Basic debugging class to print out f's in a certain area of the domain
+//!
+//! Basic debugging class. Needs to be directly added in UpdateGrid (could potentially also be added as a proper Probe in the future)
+//! How to use: Define a part of the domain via min/max x, y, and z. The DistributionDebugInspector will print out all f's in that area.
+//!
+//=======================================================================================
+
+#ifndef DISTRIBUTION_INSPECTOR_H
+#define DISTRIBUTION_INSPECTOR_H
+
+#include "Parameter/Parameter.h"
+
+
+class DistributionDebugInspector
+{
+public:
+ DistributionDebugInspector(uint _inspectionLevel, real _minX, real _maxX, real _minY, real _maxY, real _minZ, real _maxZ, std::string _tag):
+ inspectionLevel(_inspectionLevel),
+ minX(_minX),
+ maxX(_maxX),
+ minY(_minY),
+ maxY(_maxY),
+ minZ(_minZ),
+ maxZ(_maxZ),
+ tag(_tag)
+ {};
+
+ ~DistributionDebugInspector() = default;
+
+ void inspect(std::shared_ptr<Parameter> para, uint level, uint t);
+
+
+private:
+uint inspectionLevel;
+real minX;
+real maxX;
+real minY;
+real maxY;
+real minZ;
+real maxZ;
+std::string tag;
+
+};
+
+#endif
diff --git a/src/gpu/VirtualFluids_GPU/Output/FileWriter.cpp b/src/gpu/VirtualFluids_GPU/Output/FileWriter.cpp
index c6e53ee3cbfb98f11e373ca014c7faf4e70a86f0..edf705421530bdbc9c2c9fd8c44eca6d3c5ab923 100644
--- a/src/gpu/VirtualFluids_GPU/Output/FileWriter.cpp
+++ b/src/gpu/VirtualFluids_GPU/Output/FileWriter.cpp
@@ -50,7 +50,7 @@ void FileWriter::writeTimestep(std::shared_ptr<Parameter> para, unsigned int tim
void FileWriter::writeTimestep(std::shared_ptr<Parameter> para, unsigned int timestep, int level)
{
- const unsigned int numberOfParts = para->getParH(level)->numberOfNodes / para->getlimitOfNodesForVTK() + 1;
+ const unsigned int numberOfParts = (uint)para->getParH(level)->numberOfNodes / para->getlimitOfNodesForVTK() + 1;
std::vector<std::string> fname;
std::vector<std::string> fnameMed;
@@ -217,8 +217,8 @@ void FileWriter::writeUnstrucuredGridLT(std::shared_ptr<Parameter> para, int lev
for (unsigned int part = 0; part < fname.size(); part++)
{
- if (((part + 1)*para->getlimitOfNodesForVTK()) > para->getParH(level)->numberOfNodes)
- sizeOfNodes = para->getParH(level)->numberOfNodes - (part * para->getlimitOfNodesForVTK());
+ if (((part + 1)*para->getlimitOfNodesForVTK()) > (uint)para->getParH(level)->numberOfNodes)
+ sizeOfNodes = (uint)para->getParH(level)->numberOfNodes - (part * para->getlimitOfNodesForVTK());
else
sizeOfNodes = para->getlimitOfNodesForVTK();
@@ -340,8 +340,8 @@ void FileWriter::writeUnstrucuredGridLTConc(std::shared_ptr<Parameter> para, int
for (unsigned int part = 0; part < fname.size(); part++)
{
- if (((part + 1)*para->getlimitOfNodesForVTK()) > para->getParH(level)->numberOfNodes)
- sizeOfNodes = para->getParH(level)->numberOfNodes - (part * para->getlimitOfNodesForVTK());
+ if (((part + 1) * para->getlimitOfNodesForVTK()) > (uint)para->getParH(level)->numberOfNodes)
+ sizeOfNodes = (uint)para->getParH(level)->numberOfNodes - (part * para->getlimitOfNodesForVTK());
else
sizeOfNodes = para->getlimitOfNodesForVTK();
@@ -449,9 +449,9 @@ void FileWriter::writeUnstrucuredGridMedianLT(std::shared_ptr<Parameter> para, i
{
//printf("\n test in if I... \n");
//////////////////////////////////////////////////////////////////////////
- if (((part + 1)*para->getlimitOfNodesForVTK()) > para->getParH(level)->numberOfNodes)
+ if (((part + 1) * para->getlimitOfNodesForVTK()) > (uint)para->getParH(level)->numberOfNodes)
{
- sizeOfNodes = para->getParH(level)->numberOfNodes - (part * para->getlimitOfNodesForVTK());
+ sizeOfNodes = (uint)para->getParH(level)->numberOfNodes - (part * para->getlimitOfNodesForVTK());
}
else
{
@@ -558,8 +558,8 @@ void FileWriter::writeUnstrucuredGridMedianLTConc(std::shared_ptr<Parameter> par
for (unsigned int part = 0; part < fname.size(); part++)
{
- if (((part + 1)*para->getlimitOfNodesForVTK()) > para->getParH(level)->numberOfNodes)
- sizeOfNodes = para->getParH(level)->numberOfNodes - (part * para->getlimitOfNodesForVTK());
+ if (((part + 1) * para->getlimitOfNodesForVTK()) > (uint)para->getParH(level)->numberOfNodes)
+ sizeOfNodes = (uint)para->getParH(level)->numberOfNodes - (part * para->getlimitOfNodesForVTK());
else
sizeOfNodes = para->getlimitOfNodesForVTK();
//////////////////////////////////////////////////////////////////////////
diff --git a/src/gpu/VirtualFluids_GPU/Output/InterfaceDebugWriter.hpp b/src/gpu/VirtualFluids_GPU/Output/InterfaceDebugWriter.hpp
index 0b1e9dc1c25457457eabe3013a288c4c93577dc3..4d5895b323efa1b94a5780a59c882fd5ce1be7eb 100644
--- a/src/gpu/VirtualFluids_GPU/Output/InterfaceDebugWriter.hpp
+++ b/src/gpu/VirtualFluids_GPU/Output/InterfaceDebugWriter.hpp
@@ -290,10 +290,10 @@ void writeNeighborXPointsDebug(Parameter *para)
nodesVec.resize(nodeNumberVec);
int nodeCount2 = 0;
for (int level = 0; level <= para->getMaxLevel(); level++) {
- for (unsigned int u = 0; u < para->getParH(level)->numberOfNodes; u++) {
- real x1 = para->getParH(level)->coordinateX[para->getParH(level)->neighborX[u]];
- real x2 = para->getParH(level)->coordinateY[para->getParH(level)->neighborX[u]];
- real x3 = para->getParH(level)->coordinateZ[para->getParH(level)->neighborX[u]];
+ for (size_t index = 0; index < para->getParH(level)->numberOfNodes; index++) {
+ real x1 = para->getParH(level)->coordinateX[para->getParH(level)->neighborX[index]];
+ real x2 = para->getParH(level)->coordinateY[para->getParH(level)->neighborX[index]];
+ real x3 = para->getParH(level)->coordinateZ[para->getParH(level)->neighborX[index]];
nodesVec[nodeCount2++] = (makeUbTuple((float)(x1), (float)(x2), (float)(x3)));
}
@@ -317,18 +317,18 @@ void writeNeighborXLinesDebug(Parameter *para)
nodesVec.resize(nodeNumberVec * 2);
int nodeCount = 0;
for (int level = 0; level < para->getMaxLevel(); level++) {
- for (unsigned int u = 0; u < para->getParH(level)->numberOfNodes; u++) {
- real x1 = para->getParH(level)->coordinateX[u];
- real x2 = para->getParH(level)->coordinateY[u];
- real x3 = para->getParH(level)->coordinateZ[u];
- real x1N = para->getParH(level)->coordinateX[para->getParH(level)->neighborX[u]];
- real x2N = para->getParH(level)->coordinateY[para->getParH(level)->neighborX[u]];
- real x3N = para->getParH(level)->coordinateZ[para->getParH(level)->neighborX[u]];
+ for (size_t index = 0; index < para->getParH(level)->numberOfNodes; index++) {
+ real x1 = para->getParH(level)->coordinateX[index];
+ real x2 = para->getParH(level)->coordinateY[index];
+ real x3 = para->getParH(level)->coordinateZ[index];
+ real x1N = para->getParH(level)->coordinateX[para->getParH(level)->neighborX[index]];
+ real x2N = para->getParH(level)->coordinateY[para->getParH(level)->neighborX[index]];
+ real x3N = para->getParH(level)->coordinateZ[para->getParH(level)->neighborX[index]];
nodesVec[nodeCount++] = (makeUbTuple((float)(x1), (float)(x2), (float)(x3)));
nodesVec[nodeCount++] = (makeUbTuple((float)(x1N), (float)(x2N), (float)(x3N)));
- if (para->getParH(level)->typeOfGridNode[u] == GEO_FLUID) {
+ if (para->getParH(level)->typeOfGridNode[index] == GEO_FLUID) {
cellsVec.push_back(makeUbTuple(nodeCount - 2, nodeCount - 1));
}
}
@@ -350,10 +350,10 @@ void writeNeighborYPointsDebug(Parameter *para)
nodesVec.resize(nodeNumberVec);
int nodeCount2 = 0;
for (int level = 0; level <= para->getMaxLevel(); level++) {
- for (unsigned int u = 0; u < para->getParH(level)->numberOfNodes; u++) {
- real x1 = para->getParH(level)->coordinateX[para->getParH(level)->neighborY[u]];
- real x2 = para->getParH(level)->coordinateY[para->getParH(level)->neighborY[u]];
- real x3 = para->getParH(level)->coordinateZ[para->getParH(level)->neighborY[u]];
+ for (size_t index = 0; index < para->getParH(level)->numberOfNodes; index++) {
+ real x1 = para->getParH(level)->coordinateX[para->getParH(level)->neighborY[index]];
+ real x2 = para->getParH(level)->coordinateY[para->getParH(level)->neighborY[index]];
+ real x3 = para->getParH(level)->coordinateZ[para->getParH(level)->neighborY[index]];
nodesVec[nodeCount2++] = (makeUbTuple((float)(x1), (float)(x2), (float)(x3)));
}
@@ -377,18 +377,18 @@ void writeNeighborYLinesDebug(Parameter *para)
nodesVec.resize(nodeNumberVec * 2);
int nodeCount = 0;
for (int level = 0; level < para->getMaxLevel(); level++) {
- for (unsigned int u = 0; u < para->getParH(level)->numberOfNodes; u++) {
- real x1 = para->getParH(level)->coordinateX[u];
- real x2 = para->getParH(level)->coordinateY[u];
- real x3 = para->getParH(level)->coordinateZ[u];
- real x1N = para->getParH(level)->coordinateX[para->getParH(level)->neighborY[u]];
- real x2N = para->getParH(level)->coordinateY[para->getParH(level)->neighborY[u]];
- real x3N = para->getParH(level)->coordinateZ[para->getParH(level)->neighborY[u]];
+ for (size_t index = 0; index < para->getParH(level)->numberOfNodes; index++) {
+ real x1 = para->getParH(level)->coordinateX[index];
+ real x2 = para->getParH(level)->coordinateY[index];
+ real x3 = para->getParH(level)->coordinateZ[index];
+ real x1N = para->getParH(level)->coordinateX[para->getParH(level)->neighborY[index]];
+ real x2N = para->getParH(level)->coordinateY[para->getParH(level)->neighborY[index]];
+ real x3N = para->getParH(level)->coordinateZ[para->getParH(level)->neighborY[index]];
nodesVec[nodeCount++] = (makeUbTuple((float)(x1), (float)(x2), (float)(x3)));
nodesVec[nodeCount++] = (makeUbTuple((float)(x1N), (float)(x2N), (float)(x3N)));
- if (para->getParH(level)->typeOfGridNode[u] == GEO_FLUID) {
+ if (para->getParH(level)->typeOfGridNode[index] == GEO_FLUID) {
cellsVec.push_back(makeUbTuple(nodeCount - 2, nodeCount - 1));
}
}
@@ -410,10 +410,10 @@ void writeNeighborZPointsDebug(Parameter *para)
nodesVec.resize(nodeNumberVec);
int nodeCount2 = 0;
for (int level = 0; level <= para->getMaxLevel(); level++) {
- for (unsigned int u = 0; u < para->getParH(level)->numberOfNodes; u++) {
- real x1 = para->getParH(level)->coordinateX[para->getParH(level)->neighborZ[u]];
- real x2 = para->getParH(level)->coordinateY[para->getParH(level)->neighborZ[u]];
- real x3 = para->getParH(level)->coordinateZ[para->getParH(level)->neighborZ[u]];
+ for (size_t index = 0; index < para->getParH(level)->numberOfNodes; index++) {
+ real x1 = para->getParH(level)->coordinateX[para->getParH(level)->neighborZ[index]];
+ real x2 = para->getParH(level)->coordinateY[para->getParH(level)->neighborZ[index]];
+ real x3 = para->getParH(level)->coordinateZ[para->getParH(level)->neighborZ[index]];
nodesVec[nodeCount2++] = (makeUbTuple((float)(x1), (float)(x2), (float)(x3)));
}
@@ -437,18 +437,18 @@ void writeNeighborZLinesDebug(Parameter *para)
nodesVec.resize(nodeNumberVec * 2);
int nodeCount = 0;
for (int level = 0; level < para->getMaxLevel(); level++) {
- for (unsigned int u = 0; u < para->getParH(level)->numberOfNodes; u++) {
- real x1 = para->getParH(level)->coordinateX[u];
- real x2 = para->getParH(level)->coordinateY[u];
- real x3 = para->getParH(level)->coordinateZ[u];
- real x1N = para->getParH(level)->coordinateX[para->getParH(level)->neighborZ[u]];
- real x2N = para->getParH(level)->coordinateY[para->getParH(level)->neighborZ[u]];
- real x3N = para->getParH(level)->coordinateZ[para->getParH(level)->neighborZ[u]];
+ for (size_t index = 0; index < para->getParH(level)->numberOfNodes; index++) {
+ real x1 = para->getParH(level)->coordinateX[index];
+ real x2 = para->getParH(level)->coordinateY[index];
+ real x3 = para->getParH(level)->coordinateZ[index];
+ real x1N = para->getParH(level)->coordinateX[para->getParH(level)->neighborZ[index]];
+ real x2N = para->getParH(level)->coordinateY[para->getParH(level)->neighborZ[index]];
+ real x3N = para->getParH(level)->coordinateZ[para->getParH(level)->neighborZ[index]];
nodesVec[nodeCount++] = (makeUbTuple((float)(x1), (float)(x2), (float)(x3)));
nodesVec[nodeCount++] = (makeUbTuple((float)(x1N), (float)(x2N), (float)(x3N)));
- if (para->getParH(level)->typeOfGridNode[u] == GEO_FLUID) {
+ if (para->getParH(level)->typeOfGridNode[index] == GEO_FLUID) {
cellsVec.push_back(makeUbTuple(nodeCount - 2, nodeCount - 1));
}
}
diff --git a/src/gpu/VirtualFluids_GPU/Output/NeighborDebugWriter.hpp b/src/gpu/VirtualFluids_GPU/Output/NeighborDebugWriter.hpp
index 83f0a677b0012153cf079b466a333acc58bda6be..57139d25ae4d046e1dd1be1f3ef5e179daf0872e 100644
--- a/src/gpu/VirtualFluids_GPU/Output/NeighborDebugWriter.hpp
+++ b/src/gpu/VirtualFluids_GPU/Output/NeighborDebugWriter.hpp
@@ -5,53 +5,57 @@
#include "Logger.h"
#include "Parameter/Parameter.h"
#include "basics/utilities/UbSystem.h"
-#include "grid/NodeValues.h"
+#include "gpu/GridGenerator/grid/NodeValues.h"
#include "lbm/constants/D3Q27.h"
#include <basics/writer/WbWriterVtkXmlBinary.h>
-#include "Utilities/FindNeighbors.h"
-#include "VirtualFluids_GPU/Communication/Communicator.h"
#include "Core/StringUtilities/StringUtil.h"
+#include "Utilities/FindNeighbors.h"
+#include "gpu/VirtualFluids_GPU/Communication/Communicator.h"
namespace NeighborDebugWriter
{
-inline void writeNeighborLinkLines(Parameter *para, const int level, const uint numberOfNodes, const int direction,
- const std::string &name)
+inline void writeNeighborLinkLines(LBMSimulationParameter *parH, int direction, const std::string &name,
+ WbWriter *writer)
{
VF_LOG_INFO("Write node links in direction {}.", direction);
- std::vector<UbTupleFloat3> nodes(numberOfNodes * 2);
- std::vector<UbTupleInt2> cells(numberOfNodes);
- for (uint position = 0; position < numberOfNodes; position++) {
- if (para->getParH(level)->typeOfGridNode[position] != GEO_FLUID)
+ const unsigned long long numberOfNodes = parH->numberOfNodes;
+ std::vector<UbTupleFloat3> nodes;
+ nodes.reserve(numberOfNodes);
+ std::vector<UbTupleInt2> cells;
+ cells.reserve(numberOfNodes/2);
+
+ for (size_t position = 0; position < numberOfNodes; position++) {
+ if (parH->typeOfGridNode[position] != GEO_FLUID)
continue;
- const double x1 = para->getParH(level)->coordinateX[position];
- const double x2 = para->getParH(level)->coordinateY[position];
- const double x3 = para->getParH(level)->coordinateZ[position];
+ const double x1 = parH->coordinateX[position];
+ const double x2 = parH->coordinateY[position];
+ const double x3 = parH->coordinateZ[position];
- const uint positionNeighbor = getNeighborIndex(para->getParH(level).get(), position, direction);
+ const uint positionNeighbor = getNeighborIndex(parH, (uint)position, direction);
- const double x1Neighbor = para->getParH(level)->coordinateX[positionNeighbor];
- const double x2Neighbor = para->getParH(level)->coordinateY[positionNeighbor];
- const double x3Neighbor = para->getParH(level)->coordinateZ[positionNeighbor];
+ const double x1Neighbor = parH->coordinateX[positionNeighbor];
+ const double x2Neighbor = parH->coordinateY[positionNeighbor];
+ const double x3Neighbor = parH->coordinateZ[positionNeighbor];
nodes.emplace_back(float(x1), float(x2), float(x3));
nodes.emplace_back(float(x1Neighbor), float(x2Neighbor), float(x3Neighbor));
cells.emplace_back((int)nodes.size() - 2, (int)nodes.size() - 1);
}
- WbWriterVtkXmlBinary::getInstance()->writeLines(name, nodes, cells);
+ writer->writeLines(name, nodes, cells);
}
inline void writeNeighborLinkLinesDebug(Parameter *para)
{
for (int level = 0; level <= para->getMaxLevel(); level++) {
- for (int direction = vf::lbm::dir::STARTDIR; direction <= vf::lbm::dir::ENDDIR; direction++) {
+ for (size_t direction = vf::lbm::dir::STARTDIR; direction <= vf::lbm::dir::ENDDIR; direction++) {
const std::string fileName = para->getFName() + "_" + StringUtil::toString<int>(level) + "_Link_" +
std::to_string(direction) + "_Debug.vtk";
- writeNeighborLinkLines(para, level, para->getParH(level)->numberOfNodes, direction, fileName);
+ writeNeighborLinkLines(para->getParH(level).get(), (int)direction, fileName, WbWriterVtkXmlBinary::getInstance());
}
}
}
diff --git a/src/gpu/VirtualFluids_GPU/Output/NeighborDebugWriterTest.cpp b/src/gpu/VirtualFluids_GPU/Output/NeighborDebugWriterTest.cpp
new file mode 100644
index 0000000000000000000000000000000000000000..a19ed3d723f28998f5d27cd15ebf4bab8ba061c4
--- /dev/null
+++ b/src/gpu/VirtualFluids_GPU/Output/NeighborDebugWriterTest.cpp
@@ -0,0 +1,79 @@
+#include <gmock/gmock.h>
+#include "NeighborDebugWriter.hpp"
+#include "gpu/VirtualFluids_GPU/Utilities/testUtilitiesGPU.h"
+
+class WbWriterSpy : public WbWriter
+{
+public:
+ std::string writeLines(const std::string & /*filename*/, std::vector<UbTupleFloat3> &nodes,
+ std::vector<UbTupleInt2> &lines) override
+ {
+ this->nodes = nodes;
+ this->lines = lines;
+ return "";
+ }
+ std::vector<UbTupleFloat3> nodes;
+ std::vector<UbTupleInt2> lines;
+
+ std::string getFileExtension() override { return ""; }
+};
+
+class NeighborDebugWriterTest : public testing::Test
+{
+protected:
+ void SetUp() override
+ {
+ typeOfGridNode = std::vector<uint>(numberOfNodes, GEO_FLUID);
+ neighbors = std::vector<uint>(numberOfNodes, 2);
+ coordinates = std::vector<real>(numberOfNodes, 1.0);
+ coordinates[2] = 3.0;
+
+ parH->numberOfNodes = numberOfNodes;
+ parH->coordinateX = coordinates.data();
+ parH->coordinateY = coordinates.data();
+ parH->coordinateZ = coordinates.data();
+ parH->neighborX = neighbors.data();
+ parH->typeOfGridNode = typeOfGridNode.data();
+ }
+
+ const int level = 0;
+ const unsigned long long numberOfNodes = 3;
+ const uint direction = vf::lbm::dir::DIR_P00; // x
+ std::unique_ptr<LBMSimulationParameter> parH = std::make_unique<LBMSimulationParameter>();
+ WbWriterSpy writerSpy;
+ std::vector<uint> typeOfGridNode;
+ std::vector<uint> neighbors;
+ std::vector<real> coordinates;
+};
+
+TEST_F(NeighborDebugWriterTest, writeNeighborLinkLines_onlyFLuidNodes_writesAllNodes)
+{
+ UbTupleFloat3 oneCoord(1.0, 1.0, 1.0);
+ UbTupleFloat3 threeCoord(3.0, 3.0, 3.0);
+ std::vector<UbTupleFloat3> expectedNodes = { oneCoord, threeCoord, oneCoord, threeCoord, threeCoord, threeCoord };
+ std::vector<UbTupleInt2> expectedLines = { UbTupleInt2(0, 1), UbTupleInt2(2, 3), UbTupleInt2(4, 5) };
+
+ NeighborDebugWriter::writeNeighborLinkLines(parH.get(), direction, "name", &writerSpy);
+
+ EXPECT_THAT(writerSpy.nodes.size(), testing::Eq(numberOfNodes * 2));
+ EXPECT_THAT(writerSpy.lines.size(), testing::Eq(numberOfNodes));
+ EXPECT_THAT(writerSpy.nodes, testing::Eq(expectedNodes));
+ EXPECT_THAT(writerSpy.lines, testing::Eq(expectedLines));
+}
+
+TEST_F(NeighborDebugWriterTest, writeNeighborLinkLines_fluidAndSolidNodes_writesOnlyFluidNodes)
+{
+ typeOfGridNode[2] = GEO_SOLID;
+
+ UbTupleFloat3 oneCoord(1.0, 1.0, 1.0);
+ UbTupleFloat3 threeCoord(3.0, 3.0, 3.0);
+ std::vector<UbTupleFloat3> expectedNodes = { oneCoord, threeCoord, oneCoord, threeCoord};
+ std::vector<UbTupleInt2> expectedLines = { UbTupleInt2(0, 1), UbTupleInt2(2, 3)};
+
+ NeighborDebugWriter::writeNeighborLinkLines(parH.get(), direction, "name", &writerSpy);
+
+ EXPECT_THAT(writerSpy.nodes.size(), testing::Eq((numberOfNodes-1) * 2));
+ EXPECT_THAT(writerSpy.lines.size(), testing::Eq(numberOfNodes-1));
+ EXPECT_THAT(writerSpy.nodes, testing::Eq(expectedNodes));
+ EXPECT_THAT(writerSpy.lines, testing::Eq(expectedLines));
+}
diff --git a/src/gpu/VirtualFluids_GPU/Output/PosWriter.hpp b/src/gpu/VirtualFluids_GPU/Output/PosWriter.hpp
index 456f9c148c75c27fb899f976ba4f99b109fc3d4b..ce611d25d1aa3f9e98840a0f04d9b2045d0a224f 100644
--- a/src/gpu/VirtualFluids_GPU/Output/PosWriter.hpp
+++ b/src/gpu/VirtualFluids_GPU/Output/PosWriter.hpp
@@ -33,9 +33,9 @@ public:
{
out.writeInteger(para->getParH(level)->numberOfNodes);
out.writeLine();
- for(unsigned int u=0; u<para->getParH(level)->numberOfNodes; u++)
+ for(size_t index = 0; index < para->getParH(level)->numberOfNodes; index++)
{
- out.writeInteger(para->getParH(level)->typeOfGridNode[u]);
+ out.writeInteger(para->getParH(level)->typeOfGridNode[index]);
}
out.writeLine();
} //end levelloop
@@ -46,9 +46,9 @@ public:
{
out.writeInteger(para->getParH(level)->numberOfNodes);
out.writeLine();
- for(unsigned int u=0; u<para->getParH(level)->numberOfNodes; u++)
+ for (size_t index = 0; index < para->getParH(level)->numberOfNodes; index++)
{
- out.writeInteger(para->getParH(level)->neighborX[u]);
+ out.writeInteger(para->getParH(level)->neighborX[index]);
}
out.writeLine();
} //end levelloop
@@ -59,9 +59,9 @@ public:
{
out.writeInteger(para->getParH(level)->numberOfNodes);
out.writeLine();
- for(unsigned int u=0; u<para->getParH(level)->numberOfNodes; u++)
+ for (size_t index = 0; index < para->getParH(level)->numberOfNodes; index++)
{
- out.writeInteger(para->getParH(level)->neighborY[u]);
+ out.writeInteger(para->getParH(level)->neighborY[index]);
}
out.writeLine();
} //end levelloop
@@ -72,9 +72,9 @@ public:
{
out.writeInteger(para->getParH(level)->numberOfNodes);
out.writeLine();
- for(unsigned int u=0; u<para->getParH(level)->numberOfNodes; u++)
+ for (size_t index = 0; index < para->getParH(level)->numberOfNodes; index++)
{
- out.writeInteger(para->getParH(level)->neighborZ[u]);
+ out.writeInteger(para->getParH(level)->neighborZ[index]);
}
out.writeLine();
} //end levelloop
diff --git a/src/gpu/VirtualFluids_GPU/Output/QDebugVtkWriter.hpp b/src/gpu/VirtualFluids_GPU/Output/QDebugVtkWriter.hpp
new file mode 100644
index 0000000000000000000000000000000000000000..d567c695a0e33b7a88c2c8cf3bcb88093ce5b802
--- /dev/null
+++ b/src/gpu/VirtualFluids_GPU/Output/QDebugVtkWriter.hpp
@@ -0,0 +1,96 @@
+#ifndef QVTKWRITER_HPP
+#define QVTKWRITER_HPP
+
+#include <array>
+#include <vector>
+
+#include "basics/Core/StringUtilities/StringUtil.h"
+#include "basics/utilities/UbSystem.h"
+#include "basics/writer/WbWriterVtkXmlBinary.h"
+#include "lbm/constants/D3Q27.h"
+#include "logger/Logger.h"
+
+#include "gpu/GridGenerator/grid/NodeValues.h"
+#include "gpu/VirtualFluids_GPU/Communication/Communicator.h"
+#include "gpu/VirtualFluids_GPU/LBM/LB.h"
+#include "gpu/VirtualFluids_GPU/Parameter/Parameter.h"
+#include "gpu/VirtualFluids_GPU/Utilities/FindNeighbors.h"
+
+namespace QDebugVtkWriter
+{
+
+using namespace vf::lbm::dir;
+
+namespace
+{
+inline void modifyLineLengthsForQs(const std::array<double, 3> &coords, std::array<double, 3> &neighborCoords, real q)
+{
+ if (q == 1.0 || q <= 0.0)
+ return;
+
+ const auto dx = neighborCoords[0] - coords[0];
+ const auto dy = neighborCoords[1] - coords[1];
+ const auto dz = neighborCoords[2] - coords[2];
+
+ neighborCoords[0] = coords[0] + q * dx;
+ neighborCoords[1] = coords[1] + q * dy;
+ neighborCoords[2] = coords[2] + q * dz;
+}
+
+inline void writeQLines(LBMSimulationParameter *parH, QforBoundaryConditions &boundaryQ, const std::string &filepath,
+ WbWriter *writer)
+{
+ VF_LOG_INFO("Write qs in for boundary condition to {}.", filepath);
+
+ const auto numberOfNodes = boundaryQ.numberOfBCnodes;
+ std::vector<UbTupleFloat3> nodes;
+ nodes.reserve(numberOfNodes * 8 * 2);
+ std::vector<UbTupleInt2> lines;
+ lines.reserve(numberOfNodes * 8);
+
+ std::vector<std::string> dataNames = { "nodeIndex", "q" };
+ std::vector<std::vector<float>> lineData(2);
+
+ for (size_t i = 0; i < numberOfNodes; i++) {
+ const auto nodeIndex = boundaryQ.k[i];
+ const std::array<double, 3> coords = { parH->coordinateX[nodeIndex], parH->coordinateY[nodeIndex],
+ parH->coordinateZ[nodeIndex] };
+
+ for (size_t direction = 1; direction < ENDDIR; direction++) {
+
+ const auto q = boundaryQ.q27[direction][i];
+ if (q <= (real)0.0) {
+ continue;
+ }
+
+ const auto positionNeighbor = getNeighborIndex(parH, (uint)nodeIndex, (int)direction);
+
+ std::array<double, 3> neighborCoords = { parH->coordinateX[positionNeighbor],
+ parH->coordinateY[positionNeighbor],
+ parH->coordinateZ[positionNeighbor] };
+
+ modifyLineLengthsForQs(coords, neighborCoords, q);
+
+ nodes.emplace_back(float(coords[0]), float(coords[1]), coords[2]);
+ nodes.emplace_back(float(neighborCoords[0]), float(neighborCoords[1]), float(neighborCoords[2]));
+
+ lines.emplace_back((int)nodes.size() - 2, (int)nodes.size() - 1);
+ lineData[0].push_back(nodeIndex);
+ lineData[1].push_back(q);
+ }
+ }
+
+ writer->writeLinesWithLineData(filepath, nodes, lines, dataNames, lineData);
+}
+} // namespace
+
+inline void writeQLinesDebug(Parameter *para, QforBoundaryConditions &boundaryQ, uint level, const std::string& fileName)
+{
+ const auto filePath = para->getFName() + "_" + fileName + ".vtk";
+ auto writer = WbWriterVtkXmlBinary::getInstance();
+ writeQLines(para->getParH(level).get(), boundaryQ, filePath, writer);
+}
+
+} // namespace QDebugVtkWriter
+
+#endif
diff --git a/src/gpu/VirtualFluids_GPU/Output/QDebugVtkWriterTest.cpp b/src/gpu/VirtualFluids_GPU/Output/QDebugVtkWriterTest.cpp
new file mode 100644
index 0000000000000000000000000000000000000000..9eecb25c663fcfc8fde353b76ccf20cbcb9cf272
--- /dev/null
+++ b/src/gpu/VirtualFluids_GPU/Output/QDebugVtkWriterTest.cpp
@@ -0,0 +1,60 @@
+#include "gmock/gmock.h"
+#include <cmath>
+#include <gmock/gmock.h>
+#include <gtest/gtest.h>
+#include "QDebugVtkWriter.hpp"
+#include <tuple>
+
+MATCHER(DoubleNear5, "") {
+ return abs(std::get<0>(arg) - std::get<1>(arg)) < 0.00001;
+}
+
+using namespace QDebugVtkWriter;
+
+double calcVectorLength(const std::array<double, 3> coords, const std::array<double, 3> neighborCoords)
+{
+ return std::sqrt(std::pow((neighborCoords[0] - coords[0]), 2) + std::pow((neighborCoords[1] - coords[1]), 2) +
+ std::pow((neighborCoords[2] - coords[2]), 2));
+}
+
+TEST(QDebugVtkWriterTest, modifyLineLengthsForQsSameCoords3)
+{
+ const std::array<double, 3> coords = { 0, 0, 0 };
+ std::array<double, 3> neighborCoords = { 1, 1, 1 };
+ const real q = 0.3;
+ const real initialLength = calcVectorLength(coords, neighborCoords);
+
+ modifyLineLengthsForQs(coords, neighborCoords, q);
+
+ std::array<double, 3> expectedNeighborCoords = { 0.3, 0.3, 0.3 };
+ EXPECT_THAT(neighborCoords,testing::Pointwise(DoubleNear5(), expectedNeighborCoords));
+ EXPECT_THAT(calcVectorLength(coords, neighborCoords), testing::DoubleNear(q*initialLength, 0.00001));
+}
+
+TEST(QDebugVtkWriterTest, modifyLineLengthDifferentCoords)
+{
+ const std::array<double, 3> coords = { 0, 0, 0 };
+ std::array<double, 3> neighborCoords = { 1, 2, 3 };
+ const real q = 0.3;
+ const real initialLength = calcVectorLength(coords, neighborCoords);
+
+ modifyLineLengthsForQs(coords, neighborCoords, q);
+
+ std::array<double, 3> expectedNeighborCoords = { 0.3, 0.6, 0.9 };
+ EXPECT_THAT(neighborCoords,testing::Pointwise(DoubleNear5(), expectedNeighborCoords));
+ EXPECT_THAT(calcVectorLength(coords, neighborCoords), testing::DoubleNear(q*initialLength, 0.00001));
+}
+
+TEST(QDebugVtkWriterTest, modifyLineLengthNegativeCoord)
+{
+ const std::array<double, 3> coords = { 0, 0, 0 };
+ std::array<double, 3> neighborCoords = { 1, 2, -3 };
+ const real q = 0.3;
+ const real initialLength = calcVectorLength(coords, neighborCoords);
+
+ modifyLineLengthsForQs(coords, neighborCoords, q);
+
+ std::array<double, 3> expectedNeighborCoords = { 0.3, 0.6, -0.9 };
+ EXPECT_THAT(neighborCoords,testing::Pointwise(DoubleNear5(), expectedNeighborCoords));
+ EXPECT_THAT(calcVectorLength(coords, neighborCoords), testing::DoubleNear(q*initialLength, 0.00001));
+}
diff --git a/src/gpu/VirtualFluids_GPU/Output/QDebugWriter.hpp b/src/gpu/VirtualFluids_GPU/Output/QDebugWriter.hpp
index d006636572377477aeb3599a8ae843ea2b1e31ff..c1a3658d318eb47e84530bf437afa0bb6ba91743 100644
--- a/src/gpu/VirtualFluids_GPU/Output/QDebugWriter.hpp
+++ b/src/gpu/VirtualFluids_GPU/Output/QDebugWriter.hpp
@@ -13,8 +13,6 @@
#include <basics/writer/WbWriterVtkXmlBinary.h>
#include "Core/StringUtilities/StringUtil.h"
-//using namespace std;
-
namespace QDebugWriter
{
void writeQValues(QforBoundaryConditions &Q, int* k, int kq, const std::string &name)
diff --git a/src/gpu/VirtualFluids_GPU/Output/UnstructuredGridWriter.hpp b/src/gpu/VirtualFluids_GPU/Output/UnstructuredGridWriter.hpp
index 81f2c028a6bbc7cd9c077571349f4f0465a08a05..f26b4e5795466a72aa1894de37bdb066b9ab9d04 100644
--- a/src/gpu/VirtualFluids_GPU/Output/UnstructuredGridWriter.hpp
+++ b/src/gpu/VirtualFluids_GPU/Output/UnstructuredGridWriter.hpp
@@ -33,7 +33,7 @@ namespace UnstructuredGridWriter
bool neighborsFluid;
- unsigned int allnodes = para->getParH(level)->numberOfNodes * 8;
+ unsigned long long allnodes = para->getParH(level)->numberOfNodes * 8;
nodes.resize(allnodes);
nodedata[0].resize(allnodes);
@@ -45,7 +45,7 @@ namespace UnstructuredGridWriter
unsigned int nodeCount = 0;
double nodeDeltaLevel = para->getParH(level)->dx;
- for (unsigned int pos=0;pos<para->getParH(level)->numberOfNodes;pos++)
+ for (size_t pos = 0; pos < para->getParH(level)->numberOfNodes; pos++)
{
if (para->getParH(level)->typeOfGridNode[pos] == GEO_FLUID /*!= GEO_VOID*/)
{
@@ -197,9 +197,9 @@ namespace UnstructuredGridWriter
vxmax = 0;
//printf("\n test in if I... \n");
//////////////////////////////////////////////////////////////////////////
- if ( ((part+1)*para->getlimitOfNodesForVTK()) > para->getParH(level)->numberOfNodes)
+ if ( ((part+1)*para->getlimitOfNodesForVTK()) > (uint)para->getParH(level)->numberOfNodes)
{
- sizeOfNodes = para->getParH(level)->numberOfNodes - (part * para->getlimitOfNodesForVTK());
+ sizeOfNodes = (uint)para->getParH(level)->numberOfNodes - (part * para->getlimitOfNodesForVTK());
}
else
{
@@ -340,9 +340,9 @@ namespace UnstructuredGridWriter
vxmax = 0;
//printf("\n test in if I... \n");
//////////////////////////////////////////////////////////////////////////
- if (((part + 1)*para->getlimitOfNodesForVTK()) > para->getParH(level)->numberOfNodes)
+ if (((part + 1) * para->getlimitOfNodesForVTK()) > (uint)para->getParH(level)->numberOfNodes)
{
- sizeOfNodes = para->getParH(level)->numberOfNodes - (part * para->getlimitOfNodesForVTK());
+ sizeOfNodes = (uint)para->getParH(level)->numberOfNodes - (part * para->getlimitOfNodesForVTK());
}
else
{
@@ -479,9 +479,9 @@ namespace UnstructuredGridWriter
vxmax = 0;
//printf("\n test in if I... \n");
//////////////////////////////////////////////////////////////////////////
- if (((part + 1)*para->getlimitOfNodesForVTK()) > para->getParH(level)->numberOfNodes)
+ if (((part + 1) * para->getlimitOfNodesForVTK()) > (uint)para->getParH(level)->numberOfNodes)
{
- sizeOfNodes = para->getParH(level)->numberOfNodes - (part * para->getlimitOfNodesForVTK());
+ sizeOfNodes = (uint)para->getParH(level)->numberOfNodes - (part * para->getlimitOfNodesForVTK());
}
else
{
@@ -628,9 +628,9 @@ namespace UnstructuredGridWriter
vxmax = 0;
//printf("\n test in if I... \n");
//////////////////////////////////////////////////////////////////////////
- if (((part + 1)*para->getlimitOfNodesForVTK()) > para->getParH(level)->numberOfNodes)
+ if (((part + 1) * para->getlimitOfNodesForVTK()) > (uint)para->getParH(level)->numberOfNodes)
{
- sizeOfNodes = para->getParH(level)->numberOfNodes - (part * para->getlimitOfNodesForVTK());
+ sizeOfNodes = (uint)para->getParH(level)->numberOfNodes - (part * para->getlimitOfNodesForVTK());
}
else
{
@@ -771,9 +771,9 @@ namespace UnstructuredGridWriter
vxmax = 0;
//printf("\n test in if I... \n");
//////////////////////////////////////////////////////////////////////////
- if ( ((part+1)*para->getlimitOfNodesForVTK()) > para->getParH(level)->numberOfNodes)
+ if (((part + 1) * para->getlimitOfNodesForVTK()) > (uint)para->getParH(level)->numberOfNodes)
{
- sizeOfNodes = para->getParH(level)->numberOfNodes - (part * para->getlimitOfNodesForVTK());
+ sizeOfNodes = (uint)para->getParH(level)->numberOfNodes - (part * para->getlimitOfNodesForVTK());
}
else
{
@@ -896,10 +896,10 @@ namespace UnstructuredGridWriter
vector< vector< double > > nodedata(nodedatanames.size());
//printf("\n test for if... \n");
- if (para->getParH(level)->numberOfNodes > limitOfNodes)
+ if ((uint)para->getParH(level)->numberOfNodes > limitOfNodes)
{
//printf("\n test in if I... \n");
- unsigned int restOfNodes = para->getParH(level)->numberOfNodes - limitOfNodes;
+ unsigned int restOfNodes = (uint)para->getParH(level)->numberOfNodes - limitOfNodes;
//////////////////////////////////////////////////////////////////////////
//PART I
nodes.resize(limitOfNodes);
@@ -984,7 +984,7 @@ namespace UnstructuredGridWriter
nodedata[5].resize(restOfNodes);
//printf("\n test in if IV... \n");
- for (unsigned int pos=limitOfNodes;pos<para->getParH(level)->numberOfNodes;pos++)
+ for (size_t pos = limitOfNodes; pos < para->getParH(level)->numberOfNodes; pos++)
{
if (para->getParH(level)->typeOfGridNode[pos] == GEO_FLUID)
{
@@ -1055,7 +1055,7 @@ namespace UnstructuredGridWriter
nodedata[5].resize(para->getParH(level)->numberOfNodes);
//printf("\n test in else II... \n");
- for (unsigned int pos=0;pos<para->getParH(level)->numberOfNodes;pos++)
+ for (size_t pos = 0; pos < para->getParH(level)->numberOfNodes; pos++)
{
if (para->getParH(level)->typeOfGridNode[pos] == GEO_FLUID)
{
@@ -1148,7 +1148,7 @@ namespace UnstructuredGridWriter
unsigned int number1,number2,number3,number4,number5,number6,number7,number8;
bool neighborsFluid;
double vxmax = 0;
- vector< vector< double > > nodedata(nodedatanames.size());
+ vector<vector<double>> nodedata(nodedatanames.size());
nodes.resize(para->getParH(level)->numberOfNodes);
nodedata[0].resize(para->getParH(level)->numberOfNodes);
@@ -1158,7 +1158,7 @@ namespace UnstructuredGridWriter
nodedata[4].resize(para->getParH(level)->numberOfNodes);
nodedata[5].resize(para->getParH(level)->numberOfNodes);
- for (unsigned int pos=0;pos<para->getParH(level)->numberOfNodes;pos++)
+ for (size_t pos = 0; pos < para->getParH(level)->numberOfNodes; pos++)
{
if (para->getParH(level)->typeOfGridNode[pos] == GEO_FLUID)
{
@@ -1244,7 +1244,7 @@ namespace UnstructuredGridWriter
nodedata[4].resize(para->getParH(level)->numberOfNodes);
nodedata[5].resize(para->getParH(level)->numberOfNodes);
- for (unsigned int pos=0;pos<para->getParH(level)->numberOfNodes;pos++)
+ for (size_t pos = 0; pos < para->getParH(level)->numberOfNodes; pos++)
{
if (para->getParH(level)->typeOfGridNode[pos] == GEO_FLUID)
{
@@ -1342,9 +1342,9 @@ namespace UnstructuredGridWriter
vxmax = 0;
//printf("\n test in if I... \n");
//////////////////////////////////////////////////////////////////////////
- if ( ((part+1)*para->getlimitOfNodesForVTK()) > para->getParH(level)->numberOfNodes)
+ if ( ((part+1)*para->getlimitOfNodesForVTK()) > (uint)para->getParH(level)->numberOfNodes)
{
- sizeOfNodes = para->getParH(level)->numberOfNodes - (part * para->getlimitOfNodesForVTK());
+ sizeOfNodes = (uint)para->getParH(level)->numberOfNodes - (part * para->getlimitOfNodesForVTK());
}
else
{
@@ -1364,7 +1364,7 @@ namespace UnstructuredGridWriter
nodedata[5].resize(sizeOfNodes);
//////////////////////////////////////////////////////////////////////////
//printf("\n test in if II... \n");
- for (unsigned int pos=startpos;pos<endpos;pos++)
+ for (size_t pos = startpos; pos < endpos; pos++)
{
if (para->getParH(level)->typeOfGridNode[pos] == GEO_FLUID)
{
@@ -1465,9 +1465,9 @@ namespace UnstructuredGridWriter
vxmax = 0;
//printf("\n test in if I... \n");
//////////////////////////////////////////////////////////////////////////
- if (((part + 1)*para->getlimitOfNodesForVTK()) > para->getParH(level)->numberOfNodes)
+ if (((part + 1) * para->getlimitOfNodesForVTK()) > (uint)para->getParH(level)->numberOfNodes)
{
- sizeOfNodes = para->getParH(level)->numberOfNodes - (part * para->getlimitOfNodesForVTK());
+ sizeOfNodes = (uint)para->getParH(level)->numberOfNodes - (part * para->getlimitOfNodesForVTK());
}
else
{
@@ -1595,9 +1595,9 @@ namespace UnstructuredGridWriter
vxmax = 0;
//printf("\n test in if I... \n");
//////////////////////////////////////////////////////////////////////////
- if (((part + 1)*para->getlimitOfNodesForVTK()) > para->getParH(level)->numberOfNodes)
+ if (((part + 1) * para->getlimitOfNodesForVTK()) > (uint)para->getParH(level)->numberOfNodes)
{
- sizeOfNodes = para->getParH(level)->numberOfNodes - (part * para->getlimitOfNodesForVTK());
+ sizeOfNodes = (uint)para->getParH(level)->numberOfNodes - (part * para->getlimitOfNodesForVTK());
}
else
{
@@ -1728,7 +1728,7 @@ namespace UnstructuredGridWriter
nodedatanames.push_back("geo");
unsigned int number1,number2,number3,number4,number5,number6,number7,number8;
bool neighborsFluid;
- vector< vector< double > > nodedata(nodedatanames.size());
+ vector< vector<double>> nodedata(nodedatanames.size());
nodes.resize(para->getParH(level)->numberOfNodes);
nodedata[0].resize(para->getParH(level)->numberOfNodes);
@@ -1738,7 +1738,7 @@ namespace UnstructuredGridWriter
nodedata[4].resize(para->getParH(level)->numberOfNodes);
nodedata[5].resize(para->getParH(level)->numberOfNodes);
- for (unsigned int pos=0;pos<para->getParH(level)->numberOfNodes;pos++)
+ for (size_t pos = 0; pos < para->getParH(level)->numberOfNodes; pos++)
{
if (para->getParH(level)->typeOfGridNode[pos] == GEO_FLUID)
{
@@ -1825,7 +1825,7 @@ namespace UnstructuredGridWriter
nodedata[4].resize(para->getParH(level)->numberOfNodes);
nodedata[5].resize(para->getParH(level)->numberOfNodes);
- for (unsigned int pos=0;pos<para->getParH(level)->numberOfNodes;pos++)
+ for (size_t pos = 0; pos < para->getParH(level)->numberOfNodes; pos++)
{
if (para->getParH(level)->typeOfGridNode[pos] == GEO_FLUID)
{
@@ -1975,9 +1975,9 @@ namespace UnstructuredGridWriter
vxmax = 0;
//printf("\n test in if I... \n");
//////////////////////////////////////////////////////////////////////////
- if ( ((part+1)*para->getlimitOfNodesForVTK()) > para->getParH(level)->numberOfNodes)
+ if (((part + 1) * para->getlimitOfNodesForVTK()) > (uint)para->getParH(level)->numberOfNodes)
{
- sizeOfNodes = para->getParH(level)->numberOfNodes - (part * para->getlimitOfNodesForVTK());
+ sizeOfNodes = (uint)para->getParH(level)->numberOfNodes - (part * para->getlimitOfNodesForVTK());
}
else
{
@@ -2080,9 +2080,9 @@ namespace UnstructuredGridWriter
vxmax = 0;
//printf("\n test in if I... \n");
//////////////////////////////////////////////////////////////////////////
- if ( ((part+1)*para->getlimitOfNodesForVTK()) > para->getParH(level)->numberOfNodes)
+ if (((part + 1) * para->getlimitOfNodesForVTK()) > (uint)para->getParH(level)->numberOfNodes)
{
- sizeOfNodes = para->getParH(level)->numberOfNodes - (part * para->getlimitOfNodesForVTK());
+ sizeOfNodes = (uint)para->getParH(level)->numberOfNodes - (part * para->getlimitOfNodesForVTK());
}
else
{
@@ -2192,9 +2192,9 @@ namespace UnstructuredGridWriter
vxmax = 0;
//printf("\n test in if I... \n");
//////////////////////////////////////////////////////////////////////////
- if ( ((part+1)*para->getlimitOfNodesForVTK()) > para->getParH(level)->numberOfNodes)
+ if (((part + 1) * para->getlimitOfNodesForVTK()) > (uint)para->getParH(level)->numberOfNodes)
{
- sizeOfNodes = para->getParH(level)->numberOfNodes - (part * para->getlimitOfNodesForVTK());
+ sizeOfNodes = (uint)para->getParH(level)->numberOfNodes - (part * para->getlimitOfNodesForVTK());
}
else
{
@@ -2319,7 +2319,7 @@ namespace UnstructuredGridWriter
wallX3 = 0.0;
q = 0.0;
//////////////////////////////////////////////////////////////////////////
- for (unsigned int typeOfQ = STARTDIR; typeOfQ <= ENDDIR; typeOfQ++)
+ for (size_t typeOfQ = vf::lbm::dir::STARTDIR; typeOfQ <= vf::lbm::dir::ENDDIR; typeOfQ++)
{
QQ = para->getParH(level)->geometryBC.q27[0];
Q.q27[typeOfQ] = &QQ[typeOfQ*sizeOfNodes];
@@ -2423,7 +2423,7 @@ namespace UnstructuredGridWriter
wallX3 = 0.0;
q = 0.0;
//////////////////////////////////////////////////////////////////////////
- for (unsigned int typeOfQ = STARTDIR; typeOfQ <= ENDDIR; typeOfQ++)
+ for (size_t typeOfQ = vf::lbm::dir::STARTDIR; typeOfQ <= vf::lbm::dir::ENDDIR; typeOfQ++)
{
QQ = para->getParH(level)->velocityBC.q27[0];
Q.q27[typeOfQ] = &QQ[typeOfQ*sizeOfNodes];
@@ -2528,7 +2528,7 @@ namespace UnstructuredGridWriter
wallX3 = 0.0;
q = 0.0;
//////////////////////////////////////////////////////////////////////////
- for (unsigned int typeOfQ = STARTDIR; typeOfQ <= ENDDIR; typeOfQ++)
+ for (size_t typeOfQ = vf::lbm::dir::STARTDIR; typeOfQ <= vf::lbm::dir::ENDDIR; typeOfQ++)
{
QQ = para->getParH(level)->pressureBC.q27[0];
Q.q27[typeOfQ] = &QQ[typeOfQ*sizeOfNodes];
diff --git a/src/gpu/VirtualFluids_GPU/Parameter/CudaStreamManager.cpp b/src/gpu/VirtualFluids_GPU/Parameter/CudaStreamManager.cpp
index 3731836f336d91c1bc4cc5f1a8f5ea0a10bee0a6..3cc771e413134e90b0d09d8eeb6dfee791f8a1e2 100644
--- a/src/gpu/VirtualFluids_GPU/Parameter/CudaStreamManager.cpp
+++ b/src/gpu/VirtualFluids_GPU/Parameter/CudaStreamManager.cpp
@@ -31,25 +31,39 @@
#include <helper_cuda.h>
#include <iostream>
-void CudaStreamManager::launchStreams(uint numberOfStreams)
+void CudaStreamManager::registerStream(CudaStreamIndex streamIndex)
+{
+ if(streamIndex != CudaStreamIndex::Legacy)
+ cudaStreams.emplace(streamIndex, nullptr);
+}
+void CudaStreamManager::launchStreams()
{
- cudaStreams.resize(numberOfStreams);
- for (cudaStream_t &stream : cudaStreams)
- cudaStreamCreate(&stream);
+ for (auto &stream : cudaStreams)
+ cudaStreamCreate(&stream.second);
}
void CudaStreamManager::terminateStreams()
{
- for (cudaStream_t &stream : cudaStreams)
- cudaStreamDestroy(stream);
+ for (auto &stream : cudaStreams)
+ cudaStreamDestroy(stream.second);
}
-cudaStream_t &CudaStreamManager::getStream(uint streamIndex)
-{ return cudaStreams[streamIndex]; }
-
-int CudaStreamManager::getBorderStreamIndex() { return borderStreamIndex; }
+cudaStream_t &CudaStreamManager::getStream(CudaStreamIndex streamIndex, uint multiStreamIndex)
+{
+ if(streamIndex == CudaStreamIndex::Legacy) return legacyStream;
+ if(streamIsRegistered(streamIndex))
+ {
+ auto it = cudaStreams.find(streamIndex);
+ for(uint idx=0; idx<multiStreamIndex; idx++) it++;
+ return it->second;
+ }
+ return legacyStream;
+}
-int CudaStreamManager::getBulkStreamIndex() { return bulkStreamIndex; }
+bool CudaStreamManager::streamIsRegistered(CudaStreamIndex streamIndex)
+{
+ return cudaStreams.count(streamIndex) > 0;
+}
void CudaStreamManager::createCudaEvents()
{
@@ -61,12 +75,12 @@ void CudaStreamManager::destroyCudaEvents()
checkCudaErrors(cudaEventDestroy(startBulkKernel));
}
-void CudaStreamManager::triggerStartBulkKernel(int streamIndex)
+void CudaStreamManager::triggerStartBulkKernel(CudaStreamIndex streamIndex, uint multiStreamIndex)
{
- checkCudaErrors(cudaEventRecord(startBulkKernel, cudaStreams[streamIndex]));
+ checkCudaErrors(cudaEventRecord(startBulkKernel, getStream(streamIndex, multiStreamIndex)));
}
-void CudaStreamManager::waitOnStartBulkKernelEvent(int streamIndex)
+void CudaStreamManager::waitOnStartBulkKernelEvent(CudaStreamIndex streamIndex, uint multiStreamIndex)
{
- checkCudaErrors(cudaStreamWaitEvent(cudaStreams[streamIndex], startBulkKernel));
+ checkCudaErrors(cudaStreamWaitEvent(getStream(streamIndex, multiStreamIndex), startBulkKernel));
}
diff --git a/src/gpu/VirtualFluids_GPU/Parameter/CudaStreamManager.h b/src/gpu/VirtualFluids_GPU/Parameter/CudaStreamManager.h
index c2d515ab5fe9c24388632a7ca9e1e4c78b7f1467..5c59bcd3a5e6178d6e70a63f803caf8e29f32604 100644
--- a/src/gpu/VirtualFluids_GPU/Parameter/CudaStreamManager.h
+++ b/src/gpu/VirtualFluids_GPU/Parameter/CudaStreamManager.h
@@ -30,32 +30,40 @@
#ifndef STREAM_MANAGER_H
#define STREAM_MANAGER_H
-#include <vector>
-#include "Core/DataTypes.h"
-
+#include <map>
+#include <cuda.h>
#include <cuda_runtime.h>
+#include "Core/DataTypes.h"
+enum class CudaStreamIndex
+ {
+ Legacy,
+ Bulk,
+ SubDomainBorder,
+ Precursor,
+ ActuatorFarm
+ };
class CudaStreamManager
-{
+{
private:
- std::vector<cudaStream_t> cudaStreams;
+ std::multimap<CudaStreamIndex, cudaStream_t> cudaStreams;
cudaEvent_t startBulkKernel = NULL;
- const int borderStreamIndex = 1;
- const int bulkStreamIndex = 0;
+ cudaStream_t legacyStream = CU_STREAM_LEGACY;
+
public:
- void launchStreams(uint numberOfStreams);
+ void registerStream(CudaStreamIndex streamIndex);
+ void launchStreams();
void terminateStreams();
- cudaStream_t &getStream(uint streamIndex);
-
- int getBorderStreamIndex();
- int getBulkStreamIndex();
+ cudaStream_t &getStream(CudaStreamIndex streamIndex, uint multiStreamIndex=0);
+ bool streamIsRegistered(CudaStreamIndex streamIndex);
// Events
void createCudaEvents();
void destroyCudaEvents();
- void triggerStartBulkKernel(int streamIndex);
- void waitOnStartBulkKernelEvent(int strteamIndex);
+
+ void triggerStartBulkKernel(CudaStreamIndex streamIndex, uint multiStreamIndex=0);
+ void waitOnStartBulkKernelEvent(CudaStreamIndex streamIndex, uint multiStreamIndex=0);
};
#endif
diff --git a/src/gpu/VirtualFluids_GPU/Parameter/Parameter.cpp b/src/gpu/VirtualFluids_GPU/Parameter/Parameter.cpp
index 7687ec926270f23e57608ca5f3084bd26d4de20e..e593d16d6ed1f69ca65a22606a157e7ea9e6b111 100644
--- a/src/gpu/VirtualFluids_GPU/Parameter/Parameter.cpp
+++ b/src/gpu/VirtualFluids_GPU/Parameter/Parameter.cpp
@@ -43,6 +43,7 @@
#include <basics/config/ConfigurationFile.h>
+#include "Logger.h"
#include "Parameter/CudaStreamManager.h"
Parameter::Parameter() : Parameter(1, 0, {}) {}
@@ -65,6 +66,8 @@ Parameter::Parameter(int numberOfProcesses, int myId, std::optional<const vf::ba
initGridPaths();
initGridBasePoints();
initDefaultLBMkernelAllLevels();
+
+ this->cudaStreamManager = std::make_unique<CudaStreamManager>();
}
Parameter::~Parameter() = default;
@@ -500,10 +503,10 @@ void Parameter::initLBMSimulationParameter()
parH[i]->sizePlaneXY = parH[i]->nx * parH[i]->ny;
parH[i]->sizePlaneYZ = parH[i]->ny * parH[i]->nz;
parH[i]->sizePlaneXZ = parH[i]->nx * parH[i]->nz;
- parH[i]->mem_size_real = sizeof(real) * parH[i]->size_Mat;
- parH[i]->mem_size_int = sizeof(unsigned int) * parH[i]->size_Mat;
- parH[i]->mem_size_bool = sizeof(bool) * parH[i]->size_Mat;
- parH[i]->mem_size_real_yz = sizeof(real) * parH[i]->ny * parH[i]->nz;
+// parH[i]->mem_size_real = sizeof(real) * parH[i]->size_Mat; //DEPRECATED: related to full matrix
+// parH[i]->mem_size_int = sizeof(unsigned int) * parH[i]->size_Mat; //DEPRECATED: related to full matrix
+// parH[i]->mem_size_bool = sizeof(bool) * parH[i]->size_Mat; //DEPRECATED: related to full matrix
+// parH[i]->mem_size_real_yz = sizeof(real) * parH[i]->ny * parH[i]->nz; //DEPRECATED: related to full matrix
parH[i]->isEvenTimestep = true;
parH[i]->startz = parH[i]->gridNZ * ic.myProcessId;
parH[i]->endz = parH[i]->gridNZ * ic.myProcessId + parH[i]->gridNZ;
@@ -568,10 +571,10 @@ void Parameter::initLBMSimulationParameter()
parD[i]->sizePlaneXY = parH[i]->sizePlaneXY;
parD[i]->sizePlaneYZ = parH[i]->sizePlaneYZ;
parD[i]->sizePlaneXZ = parH[i]->sizePlaneXZ;
- parD[i]->mem_size_real = sizeof(real) * parD[i]->size_Mat;
- parD[i]->mem_size_int = sizeof(unsigned int) * parD[i]->size_Mat;
- parD[i]->mem_size_bool = sizeof(bool) * parD[i]->size_Mat;
- parD[i]->mem_size_real_yz = sizeof(real) * parD[i]->ny * parD[i]->nz;
+ //parD[i]->mem_size_real = sizeof(real) * parD[i]->size_Mat; //DEPRECATED: related to full matrix
+ //parD[i]->mem_size_int = sizeof(unsigned int) * parD[i]->size_Mat; //DEPRECATED: related to full matrix
+ //parD[i]->mem_size_bool = sizeof(bool) * parD[i]->size_Mat; //DEPRECATED: related to full matrix
+ //parD[i]->mem_size_real_yz = sizeof(real) * parD[i]->ny * parD[i]->nz; //DEPRECATED: related to full matrix
parD[i]->isEvenTimestep = parH[i]->isEvenTimestep;
parD[i]->startz = parH[i]->startz;
parD[i]->endz = parH[i]->endz;
@@ -586,6 +589,30 @@ void Parameter::initLBMSimulationParameter()
parD[i]->distY = parH[i]->distY;
parD[i]->distZ = parH[i]->distZ;
}
+
+ checkParameterValidityCumulantK17();
+}
+
+void Parameter::checkParameterValidityCumulantK17() const
+{
+ if (this->mainKernel != "CumulantK17")
+ return;
+
+ const real viscosity = this->parH[maxlevel]->vis;
+ const real viscosityLimit = 1.0 / 42.0;
+ if (viscosity > viscosityLimit) {
+ VF_LOG_WARNING("The viscosity (in LB units) at level {} is {:1.3g}. It is recommended to keep it smaller than {:1.3g} "
+ "for the CumulantK17 collision kernel.",
+ maxlevel, viscosity, viscosityLimit);
+ }
+
+ const real velocity = this->ic.u0;
+ const real velocityLimit = 0.1;
+ if (velocity > velocityLimit) {
+ VF_LOG_WARNING("The velocity (in LB units) is {:1.4g}. It is recommended to keep it smaller than {:1.4g} for the "
+ "CumulantK17 collision kernel.",
+ velocity, velocityLimit);
+ }
}
void Parameter::copyMeasurePointsArrayToVector(int lev)
@@ -829,7 +856,7 @@ real Parameter::getLengthRatio()
}
real Parameter::getForceRatio()
{
- return this->getDensityRatio() * this->getVelocityRatio()/this->getTimeRatio();
+ return (this->getDensityRatio()+1.0) * this->getVelocityRatio()/this->getTimeRatio();
}
real Parameter::getScaledViscosityRatio(int level)
{
@@ -859,6 +886,10 @@ real Parameter::getScaledForceRatio(int level)
{
return this->getForceRatio()*(level+1);
}
+real Parameter::getScaledStressRatio(int level)
+{
+ return this->getVelocityRatio()*this->getVelocityRatio();
+}
void Parameter::setRealX(real RealX)
{
ic.RealX = RealX;
@@ -883,6 +914,10 @@ void Parameter::setPressOutZ(unsigned int PressOutZ)
{
ic.PressOutZ = PressOutZ;
}
+void Parameter::setOutflowPressureCorrectionFactor(real pressBCrhoCorrectionFactor)
+{
+ ic.outflowPressureCorrectionFactor = pressBCrhoCorrectionFactor;
+}
void Parameter::setMaxDev(int maxdev)
{
ic.maxdev = maxdev;
@@ -1607,7 +1642,7 @@ void Parameter::setOutflowBoundaryNormalZ(std::string outflowNormalZ)
void Parameter::setMainKernel(std::string kernel)
{
this->mainKernel = kernel;
- if (kernel.find("Stream") != std::string::npos || kernel.find("Redesigned") != std::string::npos)
+ if ( kernel.find("CumulantK17") != std::string::npos )
this->kernelNeedsFluidNodeIndicesToRun = true;
}
void Parameter::setMultiKernelOn(bool isOn)
@@ -1720,22 +1755,22 @@ unsigned int Parameter::getSizeMat(int level)
{
return parH[level]->size_Mat;
}
-unsigned int Parameter::getMemSizereal(int level)
-{
- return parH[level]->mem_size_real;
-}
-unsigned int Parameter::getMemSizeInt(int level)
-{
- return parH[level]->mem_size_int;
-}
-unsigned int Parameter::getMemSizeBool(int level)
-{
- return parH[level]->mem_size_bool;
-}
-unsigned int Parameter::getMemSizerealYZ(int level)
-{
- return parH[level]->mem_size_real_yz;
-}
+//unsigned int Parameter::getMemSizereal(int level) //DEPRECATED: related to full matrix
+//{
+// return parH[level]->mem_size_real;
+//}
+//unsigned int Parameter::getMemSizeInt(int level) //DEPRECATED: related to full matrix
+//{
+// return parH[level]->mem_size_int;
+//}
+//unsigned int Parameter::getMemSizeBool(int level) //DEPRECATED: related to full matrix
+//{
+// return parH[level]->mem_size_bool;
+//}
+//unsigned int Parameter::getMemSizerealYZ(int level) //DEPRECATED: related to full matrix
+//{
+// return parH[level]->mem_size_real_yz;
+//}
int Parameter::getFine()
{
return fine;
@@ -1916,6 +1951,10 @@ unsigned int Parameter::getPressOutZ()
{
return ic.PressOutZ;
}
+real Parameter::getOutflowPressureCorrectionFactor()
+{
+ return ic.outflowPressureCorrectionFactor;
+}
int Parameter::getMaxDev()
{
return ic.maxdev;
@@ -2657,8 +2696,7 @@ void Parameter::setUseStreams(bool useStreams)
if (useStreams) {
if (this->getNumprocs() != 1) {
this->useStreams = useStreams;
- this->cudaStreamManager = std::make_unique<CudaStreamManager>();
- return;
+ return;
} else {
std::cout << "Can't use streams with only one process!" << std::endl;
}
diff --git a/src/gpu/VirtualFluids_GPU/Parameter/Parameter.h b/src/gpu/VirtualFluids_GPU/Parameter/Parameter.h
index cbb8bfd68702bc2285947eb76e6d0adc54a5b6c1..fa45b1742f20e32258195c78b630ce95175af938 100644
--- a/src/gpu/VirtualFluids_GPU/Parameter/Parameter.h
+++ b/src/gpu/VirtualFluids_GPU/Parameter/Parameter.h
@@ -54,6 +54,8 @@ class ConfigurationFile;
}
class CudaStreamManager;
+class TransientBCInputFileReader;
+
//! \struct LBMSimulationParameter
//! \brief struct holds and manages the LB-parameter of the simulation
//! \brief For this purpose it holds structures and pointer for host and device data, respectively.
@@ -65,16 +67,78 @@ struct LBMSimulationParameter {
//////////////////////////////////////////////////////////////////////////
//! \brief stores the number of threads per GPU block
uint numberofthreads;
+ //! \brief store all distribution functions for the D3Q27
+ Distributions27 distributions;
+ //////////////////////////////////////////////////////////////////////////
+ //! \brief stores the type for every lattice node (f.e. fluid node)
+ uint *typeOfGridNode;
+ //////////////////////////////////////////////////////////////////////////
+ //! \brief store the neighbors in +X, +Y, +Z, and in diagonal negative direction
+ //! \brief this information is important because we use an indirect addressing scheme
+ uint *neighborX, *neighborY, *neighborZ, *neighborInverse;
+ //////////////////////////////////////////////////////////////////////////
+ //! \brief store the coordinates for every lattice node
+ real *coordinateX, *coordinateY, *coordinateZ;
+ //////////////////////////////////////////////////////////////////////////
+ //! \brief store the macroscopic values (velocity, density, pressure)
+ //! \brief for every lattice node
+ real *velocityX, *velocityY, *velocityZ, *rho, *pressure;
+ //! \brief stores the value for omega
+ real omega;
+ //////////////////////////////////////////////////////////////////////////
+ //! \brief stores the number of nodes (based on indirect addressing scheme)
+ unsigned long long numberOfNodes;
+ //! \brief stores the size of the memory consumption for real/int values of the arrays (e.g. coordinates, velocity)
+ unsigned long long memSizeRealLBnodes, memSizeLonglongLBnodes;
+
+
+
+
+
+
+ //////////////////////////////////////////////////////////////////////////
+ // DEPRECATED
+ //////////////////////////////////////////////////////////////////////////
// distributions///////////
// Distributions19 d0;
Distributions27 d0; // DEPRECATED: distribution functions for full matrix (not sparse)
- //! \brief store all distribution functions for the D3Q27
- Distributions27 distributions;
+
+ // typeOfGridNode (formerly known as "geo") /////////////////////
+ int *geo; // DEPRECATED: typeOfGridNode for full matrix (not sparse)
+
+ // k///////////////////////
+ unsigned int *k; // DEPRECATED: index for full matrix
+
+ // memsize/////////////////
+ //unsigned int mem_size_real_yz;
+ //unsigned int mem_size_bool;
+ //unsigned int mem_size_int;
+ //unsigned int mem_size_real;
+
+ //////////////////////////////////////////////////////////////////////////
+
+
+
+
+
+ //////////////////////////////////////////////////////////////////////////
+ // additional logic
+ //////////////////////////////////////////////////////////////////////////
// distributions F3////////
Distributions6 g6;
+ unsigned int size_Array_SP;
+
+
+ // memsizeSP/////////////////
+
+
+
+ //////////////////////////////////////////////////////////////////////////
+
+
// advection diffusion //////////////////
//! \brief store all distribution functions for the D3Q7 advection diffusion field
Distributions7 distributionsAD7;
@@ -104,22 +168,6 @@ struct LBMSimulationParameter {
real cStartx, cStarty, cStartz;
real cFx, cFy, cFz;
- // typeOfGridNode (formerly known as "geo") /////////////////////
- int *geo; // DEPRECATED: typeOfGridNode for full matrix (not sparse)
- //! \brief stores the type for every lattice node (f.e. fluid node)
- unsigned int *typeOfGridNode;
-
- // k///////////////////////
- unsigned int *k; // DEPRECATED: index for full matrix
-
- // neighbor///////////////////////////////////////////////////////////////
- //! \brief store the neighbors in +X, +Y, +Z, and in diagonal negative direction
- //! \brief this information is important because we use an indirect addressing scheme
- uint *neighborX, *neighborY, *neighborZ, *neighborInverse;
-
- // coordinates////////////////////////////////////////////////////////////
- //! \brief store the coordinates for every lattice node
- real *coordinateX, *coordinateY, *coordinateZ;
// body forces////////////
real *forceX_SP, *forceY_SP, *forceZ_SP;
@@ -138,11 +186,6 @@ struct LBMSimulationParameter {
// macroscopic values//////
// real *vx, *vy, *vz, *rho; // DEPRECATED: macroscopic values for full matrix
- //! \brief store the macroscopic values (velocity, density, pressure)
- //! \brief for every lattice node
- real *velocityX, *velocityY, *velocityZ, *rho, *pressure;
- //! \brief stores the value for omega
- real omega;
//! \brief stores the value for viscosity (on level 0)
real vis;
@@ -163,11 +206,6 @@ struct LBMSimulationParameter {
unsigned int size_Mat;
unsigned int sizePlaneXY, sizePlaneYZ, sizePlaneXZ;
- // size of sparse matrix//////////
- //! \brief stores the number of nodes (based on indirect addressing scheme)
- unsigned int numberOfNodes;
- unsigned int size_Array_SP;
-
// size of Plane btw. 2 GPUs//////
unsigned int sizePlaneSB, sizePlaneRB, startB, endB;
unsigned int sizePlaneST, sizePlaneRT, startT, endT;
@@ -180,16 +218,6 @@ struct LBMSimulationParameter {
unsigned int sizePlanePressOUT, startPOUT;
bool isSetPress;
- // memsizeSP/////////////////
- //! \brief stores the size of the memory consumption for real/int values of the arrays (e.g. coordinates, velocity)
- unsigned int mem_size_real_SP;
- unsigned int mem_size_int_SP;
-
- // memsize/////////////////
- unsigned int mem_size_real;
- unsigned int mem_size_int;
- unsigned int mem_size_bool;
- unsigned int mem_size_real_yz;
// print///////////////////
unsigned int startz, endz;
@@ -218,16 +246,16 @@ struct LBMSimulationParameter {
OffsetFC offFCBulk;
unsigned int mem_size_kCF_off;
unsigned int mem_size_kFC_off;
-
- // BC's////////////////////
+
//! \brief stores the boundary condition data
QforBoundaryConditions noSlipBC, velocityBC, outflowBC, slipBC, stressBC, pressureBC;
//! \brief number of lattice nodes for the boundary conditions
- unsigned int numberOfNoSlipBCnodesRead, numberOfVeloBCnodesRead, numberOfOutflowBCnodesRead, numberOfSlipBCnodesRead, numberOfStressBCnodesRead, numberOfPressureBCnodesRead;
+ unsigned int numberOfNoSlipBCnodesRead, numberOfVeloBCnodesRead, numberOfOutflowBCnodesRead, numberOfSlipBCnodesRead, numberOfStressBCnodesRead, numberOfPressureBCnodesRead, numberOfPrecursorBCnodesRead;
QforBoundaryConditions QpressX0, QpressX1, QpressY0, QpressY1, QpressZ0, QpressZ1; // DEPRECATED
QforBoundaryConditions propellerBC;
QforBoundaryConditions geometryBC;
+ QforPrecursorBoundaryConditions precursorBC;
QforBoundaryConditions geometryBCnormalX, geometryBCnormalY, geometryBCnormalZ;
QforBoundaryConditions inflowBCnormalX, inflowBCnormalY, inflowBCnormalZ;
QforBoundaryConditions outflowBCnormalX, outflowBCnormalY, outflowBCnormalZ;
@@ -235,6 +263,8 @@ struct LBMSimulationParameter {
unsigned int kInletQread, kOutletQread; // DEPRECATED
WallModelParameters wallModel;
+ std::vector<SPtr<TransientBCInputFileReader>> transientBCInputFileReader;
+ real outflowPressureCorrectionFactor;
// testRoundoffError
Distributions27 kDistTestRE;
@@ -367,10 +397,19 @@ struct LBMSimulationParameter {
std::vector<EdgeNodePositions> edgeNodesYtoZ;
///////////////////////////////////////////////////////
- uint *fluidNodeIndices;
- uint numberOfFluidNodes;
- uint *fluidNodeIndicesBorder;
- uint numberOfFluidNodesBorder;
+ std::map<CollisionTemplate, uint*> taggedFluidNodeIndices = {{CollisionTemplate::Default, nullptr},
+ {CollisionTemplate::SubDomainBorder,nullptr},
+ {CollisionTemplate::WriteMacroVars, nullptr},
+ {CollisionTemplate::ApplyBodyForce, nullptr},
+ {CollisionTemplate::AllFeatures, nullptr}};
+ std::map<CollisionTemplate, uint > numberOfTaggedFluidNodes = {{CollisionTemplate::Default, 0},
+ {CollisionTemplate::SubDomainBorder,0},
+ {CollisionTemplate::WriteMacroVars, 0},
+ {CollisionTemplate::ApplyBodyForce, 0},
+ {CollisionTemplate::AllFeatures, 0}};
+
+ std::vector<CollisionTemplate> allocatedBulkFluidNodeTags = {};
+
};
//! \brief Class for LBM-parameter management
@@ -471,6 +510,7 @@ public:
void setpressBcPos(std::string pressBcPos);
void setpressBcQs(std::string pressBcQs);
void setpressBcValue(std::string pressBcValue);
+ void setOutflowPressureCorrectionFactor(real correctionFactor);
void setpressBcValues(std::string pressBcValues);
void setvelBcQs(std::string velBcQs);
void setvelBcValues(std::string velBcValues);
@@ -527,7 +567,6 @@ public:
void setUseWale(bool useWale);
void setTurbulenceModel(TurbulenceModel turbulenceModel);
void setUseTurbulentViscosity(bool useTurbulentViscosity);
- void setUseAMD(bool useAMD);
void setSGSConstant(real SGSConstant);
void setHasWallModelMonitor(bool hasWallModelMonitor);
void setUseInitNeq(bool useInitNeq);
@@ -726,10 +765,10 @@ public:
unsigned int getPressOutID();
unsigned int getPressInZ();
unsigned int getPressOutZ();
- unsigned int getMemSizereal(int level);
- unsigned int getMemSizeInt(int level);
- unsigned int getMemSizeBool(int level);
- unsigned int getMemSizerealYZ(int level);
+// unsigned int getMemSizereal(int level); //DEPRECATED: related to full matrix
+// unsigned int getMemSizeInt(int level); //DEPRECATED: related to full matrix
+// unsigned int getMemSizeBool(int level); //DEPRECATED: related to full matrix
+// unsigned int getMemSizerealYZ(int level); //DEPRECATED: related to full matrix
unsigned int getSizeMat(int level);
unsigned int getTimestepStart();
unsigned int getTimestepInit();
@@ -765,6 +804,8 @@ public:
real getScaledDensityRatio(int level);
//! \returns the pressure ratio in SI/LB units scaled to the respective level
real getScaledPressureRatio(int level);
+ //! \returns the stress ratio in SI/LB units scaled to the respective level
+ real getScaledStressRatio(int level);
//! \returns the time ratio in SI/LB units scaled to the respective level
real getScaledTimeRatio(int level);
//! \returns the length ratio in SI/LB units scaled to the respective level
@@ -853,6 +894,7 @@ public:
std::string getOutflowBoundaryNormalX();
std::string getOutflowBoundaryNormalY();
std::string getOutflowBoundaryNormalZ();
+ real getOutflowPressureCorrectionFactor();
// CUDA random number
curandState *getRandomState();
// Kernel
@@ -896,6 +938,8 @@ private:
void setPathAndFilename(std::string fname);
+ void checkParameterValidityCumulantK17() const;
+
private:
bool compOn{ false };
bool diffOn{ false };
diff --git a/src/gpu/VirtualFluids_GPU/Parameter/ParameterTest.cpp b/src/gpu/VirtualFluids_GPU/Parameter/ParameterTest.cpp
index 4025acf7acad362e9f0f3702cb897b9c1b6dbf3b..72a12ae880556e6e257eb69dee4e806617252629 100644
--- a/src/gpu/VirtualFluids_GPU/Parameter/ParameterTest.cpp
+++ b/src/gpu/VirtualFluids_GPU/Parameter/ParameterTest.cpp
@@ -1,4 +1,3 @@
-#include <gmock/gmock.h>
#include "basics/tests/testUtilities.h"
#include <filesystem>
@@ -8,7 +7,6 @@
#include "Parameter.h"
#include "basics/config/ConfigurationFile.h"
-
TEST(ParameterTest, passingEmptyFileWithoutPath_ShouldNotThrow)
{
// assuming that the config files is stored parallel to this file.
@@ -37,7 +35,9 @@ TEST(ParameterTest, check_all_Parameter_CanBePassedToConstructor)
// test optional parameter
EXPECT_THAT(para.getOutputPath(), testing::Eq("/output/path/"));
- EXPECT_THAT(para.getGridPath(), testing::Eq("/path/to/grid/")); // ... all grid files (e.g. multi-gpu/ multi-level) could be tested as well
+ EXPECT_THAT(
+ para.getGridPath(),
+ testing::Eq("/path/to/grid/")); // ... all grid files (e.g. multi-gpu/ multi-level) could be tested as well
EXPECT_THAT(para.getgeoVec(), testing::Eq("/path/to/grid/geoVec.dat"));
EXPECT_THAT(para.getMaxDev(), testing::Eq(2));
EXPECT_THAT(para.getDevices(), testing::ElementsAreArray({ 2, 3 }));
@@ -163,7 +163,7 @@ TEST(ParameterTest, setGridPathOverridesDefaultGridPath)
Parameter para(2, 1);
para.setGridPath("gridPathTest");
- EXPECT_THAT( para.getGridPath(), testing::Eq("gridPathTest/1/"));
+ EXPECT_THAT(para.getGridPath(), testing::Eq("gridPathTest/1/"));
EXPECT_THAT(para.getConcentration(), testing::Eq("gridPathTest/1/conc.dat"));
}
@@ -177,9 +177,8 @@ TEST(ParameterTest, setGridPathOverridesConfigFile)
auto para = Parameter(2, 0, &config);
para.setGridPath("gridPathTest");
- EXPECT_THAT( para.getGridPath(), testing::Eq("gridPathTest/0/"));
+ EXPECT_THAT(para.getGridPath(), testing::Eq("gridPathTest/0/"));
EXPECT_THAT(para.getConcentration(), testing::Eq("gridPathTest/0/conc.dat"));
-
}
TEST(ParameterTest, userMissedSlash)
@@ -189,7 +188,6 @@ TEST(ParameterTest, userMissedSlash)
EXPECT_THAT(para.getGridPath(), testing::Eq("gridPathTest/"));
EXPECT_THAT(para.getConcentration(), testing::Eq("gridPathTest/conc.dat"));
-
}
TEST(ParameterTest, userMissedSlashMultiGPU)
@@ -199,4 +197,87 @@ TEST(ParameterTest, userMissedSlashMultiGPU)
EXPECT_THAT(para.getGridPath(), testing::Eq("gridPathTest/0/"));
EXPECT_THAT(para.getConcentration(), testing::Eq("gridPathTest/0/conc.dat"));
-}
\ No newline at end of file
+}
+
+class ParameterTestCumulantK17 : public testing::Test
+{
+protected:
+ void SetUp() override
+ {
+ }
+
+ bool stdoutContainsWarning()
+ {
+ std::string output = testing::internal::GetCapturedStdout();
+ return output.find("warning") != std::string::npos;
+ }
+
+ Parameter para;
+};
+
+TEST_F(ParameterTestCumulantK17, CumulantK17_VelocityIsTooHigh_expectWarning)
+{
+
+ para.setVelocityLB(0.11);
+ para.setMainKernel("CumulantK17");
+ testing::internal::CaptureStdout();
+
+ para.initLBMSimulationParameter();
+
+ EXPECT_TRUE(stdoutContainsWarning());
+}
+
+TEST_F(ParameterTestCumulantK17, CumulantK17_VelocityIsOk_expectNoWarning)
+{
+ para.setVelocityLB(0.09);
+ para.setMainKernel("CumulantK17");
+ testing::internal::CaptureStdout();
+
+ para.initLBMSimulationParameter();
+
+ EXPECT_FALSE(stdoutContainsWarning());
+}
+
+TEST_F(ParameterTestCumulantK17, NotCumulantK17_VelocityIsTooHigh_expectNoWarning)
+{
+ para.setVelocityLB(42);
+ para.setMainKernel("K");
+ testing::internal::CaptureStdout();
+
+ para.initLBMSimulationParameter();
+
+ EXPECT_FALSE(stdoutContainsWarning());
+}
+
+TEST_F(ParameterTestCumulantK17, CumulantK17_ViscosityIsTooHigh_expectWarning)
+{
+ para.setViscosityLB(0.024);
+ para.setMainKernel("CumulantK17");
+ testing::internal::CaptureStdout();
+
+ para.initLBMSimulationParameter();
+
+ EXPECT_TRUE(stdoutContainsWarning());
+}
+
+TEST_F(ParameterTestCumulantK17, CumulantK17_ViscosityIsOk_expectNoWarning)
+{
+ para.setViscosityLB(0.023);
+ para.setMainKernel("CumulantK17");
+ testing::internal::CaptureStdout();
+
+ para.initLBMSimulationParameter();
+
+ EXPECT_FALSE(stdoutContainsWarning());
+}
+
+TEST_F(ParameterTestCumulantK17, NotCumulantK17_ViscosityIsTooHigh_expectNoWarning)
+{
+ para.setViscosityLB(10);
+ para.setMainKernel("K");
+ testing::internal::CaptureStdout();
+
+ para.initLBMSimulationParameter();
+
+ EXPECT_FALSE(stdoutContainsWarning());
+}
diff --git a/src/gpu/VirtualFluids_GPU/Particles/Particles.cpp b/src/gpu/VirtualFluids_GPU/Particles/Particles.cpp
index 7c710f50afb0ae07edd53ef9d68e294c7af54ac1..e0156e3fbae46282baeb1359c719a077f021cf6b 100644
--- a/src/gpu/VirtualFluids_GPU/Particles/Particles.cpp
+++ b/src/gpu/VirtualFluids_GPU/Particles/Particles.cpp
@@ -141,12 +141,12 @@ void initParticles(Parameter* para)
para->getParH(lev)->plp.coordZabsolut[i] = (real)zCoordVec[i];
// find IDs
- for (unsigned int ii = 0; ii < para->getParH(lev)->numberOfNodes; ii++)
+ for (size_t index = 0; index < para->getParH(lev)->numberOfNodes; index++)
{
- if ((para->getParH(lev)->coordinateX[ii] <= para->getParH(lev)->plp.coordXabsolut[i]) &&
- ((para->getParH(lev)->plp.coordXabsolut[i] - para->getParH(lev)->coordinateX[ii]) <= dx))
+ if ((para->getParH(lev)->coordinateX[index] <= para->getParH(lev)->plp.coordXabsolut[i]) &&
+ ((para->getParH(lev)->plp.coordXabsolut[i] - para->getParH(lev)->coordinateX[index]) <= dx))
{
- tempID.push_back(ii);
+ tempID.push_back((int)index);
}
}
@@ -455,7 +455,7 @@ void rearrangeGeometry(Parameter* para, CudaMemoryManager* cudaMemoryManager)
int counter2 = 0;
//////////////////////////////////////////////////////////////////////////
//redefine fluid nodes
- for (uint index = 0; index < para->getParH(lev)->numberOfNodes; index++)
+ for (size_t index = 0; index < para->getParH(lev)->numberOfNodes; index++)
{
if (para->getParH(lev)->typeOfGridNode[index] == GEO_FLUID_OLD)
{
diff --git a/src/gpu/VirtualFluids_GPU/PreCollisionInteractor/ActuatorFarm.cu b/src/gpu/VirtualFluids_GPU/PreCollisionInteractor/ActuatorFarm.cu
new file mode 100644
index 0000000000000000000000000000000000000000..9447a8636e801c132df9cef2feced4b5ab4e68de
--- /dev/null
+++ b/src/gpu/VirtualFluids_GPU/PreCollisionInteractor/ActuatorFarm.cu
@@ -0,0 +1,629 @@
+//=======================================================================================
+// ____ ____ __ ______ __________ __ __ __ __
+// \ \ | | | | | _ \ |___ ___| | | | | / \ | |
+// \ \ | | | | | |_) | | | | | | | / \ | |
+// \ \ | | | | | _ / | | | | | | / /\ \ | |
+// \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
+// \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
+// \ \ | | ________________________________________________________________
+// \ \ | | | ______________________________________________________________|
+// \ \| | | | __ __ __ __ ______ _______
+// \ | | |_____ | | | | | | | | | _ \ / _____)
+// \ | | _____| | | | | | | | | | | \ \ \_______
+// \ | | | | |_____ | \_/ | | | | |_/ / _____ |
+// \ _____| |__| |________| \_______/ |__| |______/ (_______/
+//
+// This file is part of VirtualFluids. VirtualFluids is free software: you can
+// redistribute it and/or modify it under the terms of the GNU General Public
+// License as published by the Free Software Foundation, either version 3 of
+// the License, or (at your option) any later version.
+//
+// VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
+// ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+// FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
+// for more details.
+//
+// You should have received a copy of the GNU General Public License along
+// with VirtualFluids (see COPYING.txt). If not, see <http://www.gnu.org/licenses/>.
+//
+//! \file ActuatorFarm.cu
+//! \ingroup PreCollisionInteractor
+//! \author Henrik Asmuth, Henry Korb
+//======================================================================================
+#include "ActuatorFarm.h"
+
+#include <cuda.h>
+#include <cuda_runtime.h>
+#include <helper_cuda.h>
+
+#include "cuda/CudaGrid.h"
+#include "VirtualFluids_GPU/GPU/GeometryUtils.h"
+#include "LBM/GPUHelperFunctions/KernelUtilities.h"
+
+#include "Parameter/Parameter.h"
+#include "Parameter/CudaStreamManager.h"
+#include "DataStructureInitializer/GridProvider.h"
+#include "GPU/CudaMemoryManager.h"
+#include "lbm/constants/NumericConstants.h"
+#include "logger/Logger.h"
+
+using namespace vf::lbm::constant;
+
+
+__host__ __device__ __inline__ uint calcNode(uint bladeNode, uint numberOfBladeNodes, uint blade, uint numberOfBlades, uint turbine, uint numberOfTurbines)
+{
+
+ return bladeNode+numberOfBladeNodes*(blade+numberOfBlades*turbine);
+}
+
+__host__ __device__ __inline__ void calcTurbineBladeAndBladeNode(uint node, uint& bladeNode, uint numberOfBladeNodes, uint& blade, uint numberOfBlades, uint& turbine, uint numberOfTurbines)
+{
+ turbine = node/(numberOfBladeNodes*numberOfBlades);
+ uint x_off = turbine*numberOfBladeNodes*numberOfBlades;
+ blade = (node - x_off)/numberOfBlades;
+ uint y_off = numberOfBladeNodes*blade+x_off;
+ bladeNode = (node - y_off)/numberOfBladeNodes;
+}
+
+__host__ __device__ __forceinline__ real distSqrd(real distX, real distY, real distZ)
+{
+ return distX*distX+distY*distY+distZ*distZ;
+}
+
+void swapArrays(real* &arr1, real* &arr2)
+{
+ real* tmp = arr1;
+ arr1 = arr2;
+ arr2 = tmp;
+}
+
+__host__ __device__ __inline__ void rotateFromBladeToGlobal(
+ real& bladeCoordX_BF, real& bladeCoordY_BF, real& bladeCoordZ_BF,
+ real& bladeCoordX_GF, real& bladeCoordY_GF, real& bladeCoordZ_GF,
+ real& azimuth, real& yaw)
+{
+ real tmpX, tmpY, tmpZ;
+
+ rotateAboutX3D(azimuth, bladeCoordX_BF, bladeCoordY_BF, bladeCoordZ_BF, tmpX, tmpY, tmpZ);
+ rotateAboutZ3D(yaw, tmpX, tmpY, tmpZ, bladeCoordX_GF, bladeCoordY_GF, bladeCoordZ_GF);
+
+}
+
+__host__ __device__ __inline__ void rotateFromGlobalToBlade(
+ real& bladeCoordX_BF, real& bladeCoordY_BF, real& bladeCoordZ_BF,
+ real& bladeCoordX_GF, real& bladeCoordY_GF, real& bladeCoordZ_GF,
+ real& azimuth, real& yaw)
+{
+ real tmpX, tmpY, tmpZ;
+
+ invRotateAboutZ3D(yaw, bladeCoordX_GF, bladeCoordY_GF, bladeCoordZ_GF, tmpX, tmpY, tmpZ);
+ invRotateAboutX3D(azimuth, tmpX, tmpY, tmpZ, bladeCoordX_BF, bladeCoordY_BF, bladeCoordZ_BF);
+}
+
+__global__ void interpolateVelocities(real* gridCoordsX, real* gridCoordsY, real* gridCoordsZ,
+ uint* neighborsX, uint* neighborsY, uint* neighborsZ, uint* neighborsWSB,
+ real* vx, real* vy, real* vz,
+ real* bladeCoordsX, real* bladeCoordsY, real* bladeCoordsZ,
+ real* bladeVelocitiesX, real* bladeVelocitiesY, real* bladeVelocitiesZ,
+ uint numberOfTurbines, uint numberOfBlades, uint numberOfBladeNodes,
+ real* azimuths, real* yaws, real* omegas,
+ real* turbPosX, real* turbPosY, real* turbPosZ,
+ uint* bladeIndices, real velocityRatio, real invDeltaX)
+{
+
+ ////////////////////////////////////////////////////////////////////////////////
+ //! - Get node index coordinates from threadIdx, blockIdx, blockDim and gridDim.
+ //!
+ const unsigned nodeIndex = vf::gpu::getNodeIndex();
+
+ if(nodeIndex>=numberOfBladeNodes*numberOfBlades*numberOfTurbines) return;
+
+ uint turbine, bladeNode, blade;
+
+ calcTurbineBladeAndBladeNode(nodeIndex, bladeNode, numberOfBladeNodes, blade, numberOfBlades, turbine, numberOfTurbines);
+
+ real bladeCoordX_BF = bladeCoordsX[nodeIndex];
+ real bladeCoordY_BF = bladeCoordsY[nodeIndex];
+ real bladeCoordZ_BF = bladeCoordsZ[nodeIndex];
+
+ real bladeCoordX_GF, bladeCoordY_GF, bladeCoordZ_GF;
+
+ real localAzimuth = azimuths[turbine]+blade*c2Pi/numberOfBlades;
+ real yaw = yaws[turbine];
+
+
+ rotateFromBladeToGlobal(bladeCoordX_BF, bladeCoordY_BF, bladeCoordZ_BF,
+ bladeCoordX_GF, bladeCoordY_GF, bladeCoordZ_GF,
+ localAzimuth, yaw);
+
+ bladeCoordX_GF += turbPosX[turbine];
+ bladeCoordY_GF += turbPosY[turbine];
+ bladeCoordZ_GF += turbPosZ[turbine];
+
+ uint k, ke, kn, kt;
+ uint kne, kte, ktn, ktne;
+
+ k = findNearestCellBSW(bladeIndices[nodeIndex],
+ gridCoordsX, gridCoordsY, gridCoordsZ,
+ bladeCoordX_GF, bladeCoordY_GF, bladeCoordZ_GF,
+ neighborsX, neighborsY, neighborsZ, neighborsWSB);
+
+ bladeIndices[nodeIndex] = k;
+
+ getNeighborIndicesOfBSW(k, ke, kn, kt, kne, kte, ktn, ktne, neighborsX, neighborsY, neighborsZ);
+
+ real dW, dE, dN, dS, dT, dB;
+
+ real distX = invDeltaX*(bladeCoordX_GF-gridCoordsX[k]);
+ real distY = invDeltaX*(bladeCoordY_GF-gridCoordsY[k]);
+ real distZ = invDeltaX*(bladeCoordZ_GF-gridCoordsZ[k]);
+
+ getInterpolationWeights(dW, dE, dN, dS, dT, dB, distX, distY, distZ);
+
+ real bladeVelX_GF = trilinearInterpolation(dW, dE, dN, dS, dT, dB, k, ke, kn, kt, kne, kte, ktn, ktne, vx)*velocityRatio;
+ real bladeVelY_GF = trilinearInterpolation(dW, dE, dN, dS, dT, dB, k, ke, kn, kt, kne, kte, ktn, ktne, vy)*velocityRatio;
+ real bladeVelZ_GF = trilinearInterpolation(dW, dE, dN, dS, dT, dB, k, ke, kn, kt, kne, kte, ktn, ktne, vz)*velocityRatio;
+
+ real bladeVelX_BF, bladeVelY_BF, bladeVelZ_BF;
+
+ rotateFromGlobalToBlade(bladeVelX_BF, bladeVelY_BF, bladeVelZ_BF,
+ bladeVelX_GF, bladeVelY_GF, bladeVelZ_GF,
+ localAzimuth, yaw);
+
+ bladeVelocitiesX[nodeIndex] = bladeVelX_BF;
+ bladeVelocitiesY[nodeIndex] = bladeVelY_BF+omegas[turbine]*bladeCoordZ_BF;
+ bladeVelocitiesZ[nodeIndex] = bladeVelZ_BF;
+}
+
+
+__global__ void applyBodyForces(real* gridCoordsX, real* gridCoordsY, real* gridCoordsZ,
+ real* gridForcesX, real* gridForcesY, real* gridForcesZ,
+ real* bladeCoordsX, real* bladeCoordsY, real* bladeCoordsZ,
+ real* bladeForcesX, real* bladeForcesY, real* bladeForcesZ,
+ const uint numberOfTurbines, const uint numberOfBlades, const uint numberOfBladeNodes,
+ real* azimuths, real* yaws, real* diameters,
+ real* turbPosX, real* turbPosY, real* turbPosZ,
+ uint* gridIndices, uint nIndices,
+ const real invEpsilonSqrd, const real factorGaussian)
+{
+
+ const uint index = vf::gpu::getNodeIndex();
+
+ if(index>=nIndices) return;
+
+
+ uint gridIndex = gridIndices[index];
+
+ real gridCoordX_GF = gridCoordsX[gridIndex];
+ real gridCoordY_GF = gridCoordsY[gridIndex];
+ real gridCoordZ_GF = gridCoordsZ[gridIndex];
+
+ real gridForceX_RF = c0o1;
+ real gridForceY_RF = c0o1;
+ real gridForceZ_RF = c0o1;
+
+ real dAzimuth = c2Pi/numberOfBlades;
+
+ for(uint turbine = 0; turbine<numberOfTurbines; turbine++)
+ {
+ real radius = c1o2*diameters[turbine];
+ real gridCoordX_RF = gridCoordX_GF - turbPosX[turbine];
+ real gridCoordY_RF = gridCoordY_GF - turbPosY[turbine];
+ real gridCoordZ_RF = gridCoordZ_GF - turbPosZ[turbine];
+
+ if(distSqrd(gridCoordX_RF, gridCoordY_RF, gridCoordZ_RF)*invEpsilonSqrd > radius*radius*invEpsilonSqrd+c7o1)
+ continue;
+
+ real azimuth = azimuths[turbine];
+ real yaw = yaws[turbine];
+
+ for( uint blade=0; blade<numberOfBlades; blade++)
+ {
+ real localAzimuth = azimuth+blade*dAzimuth;
+
+
+ real gridCoordX_BF, gridCoordY_BF, gridCoordZ_BF;
+
+ rotateFromGlobalToBlade(gridCoordX_BF, gridCoordY_BF, gridCoordZ_BF,
+ gridCoordX_RF, gridCoordY_RF, gridCoordZ_RF,
+ localAzimuth, yaw);
+
+ uint node;
+ uint nextNode = calcNode(0, numberOfBladeNodes, blade, numberOfBlades, turbine, numberOfTurbines);
+
+ real last_z = c0o1;
+ real current_z = c0o1;
+ real next_z = bladeCoordsZ[nextNode];
+
+ real x, y, dz, eta, forceX_RF, forceY_RF, forceZ_RF;
+
+ for( uint bladeNode=0; bladeNode<numberOfBladeNodes-1; bladeNode++)
+ {
+ node = nextNode;
+ nextNode = calcNode(bladeNode+1, numberOfBladeNodes, blade, numberOfBlades, turbine, numberOfTurbines);
+
+ x = bladeCoordsX[node];
+ y = bladeCoordsY[node];
+ last_z = current_z;
+ current_z = next_z;
+ next_z = bladeCoordsZ[nextNode];
+
+ dz = c1o2*(next_z-last_z);
+
+ eta = dz*factorGaussian*exp(-distSqrd(x-gridCoordX_BF, y-gridCoordY_BF, current_z-gridCoordZ_BF)*invEpsilonSqrd);
+ rotateFromBladeToGlobal(bladeForcesX[node], bladeForcesY[node], bladeForcesZ[node],
+ forceX_RF, forceY_RF, forceZ_RF,
+ localAzimuth, yaw);
+
+ gridForceX_RF += forceX_RF*eta;
+ gridForceY_RF += forceY_RF*eta;
+ gridForceZ_RF += forceZ_RF*eta;
+ }
+
+ //Handle last node separately
+
+ node = nextNode;
+
+ x = bladeCoordsX[node];
+ y = bladeCoordsY[node];
+ last_z = current_z;
+ current_z = next_z;
+
+ dz = c1o2*(radius-last_z);
+
+ eta = dz*factorGaussian*exp(-distSqrd(x-gridCoordX_BF, y-gridCoordY_BF, current_z-gridCoordZ_BF)*invEpsilonSqrd);
+
+ rotateFromBladeToGlobal(bladeForcesX[node], bladeForcesY[node], bladeForcesZ[node],
+ forceX_RF, forceY_RF, forceZ_RF,
+ localAzimuth, yaw);
+
+ gridForceX_RF += forceX_RF*eta;
+ gridForceY_RF += forceY_RF*eta;
+ gridForceZ_RF += forceZ_RF*eta;
+ }
+ }
+
+ gridForcesX[gridIndex] += gridForceX_RF;
+ gridForcesY[gridIndex] += gridForceY_RF;
+ gridForcesZ[gridIndex] += gridForceZ_RF;
+}
+
+void ActuatorFarm::addTurbine(real posX, real posY, real posZ, real diameter, real omega, real azimuth, real yaw, std::vector<real> bladeRadii)
+{
+ preInitPosX.push_back(posX);
+ preInitPosY.push_back(posY);
+ preInitPosZ.push_back(posZ);
+ preInitOmegas.push_back(omega);
+ preInitAzimuths.push_back(azimuth);
+ preInitYaws.push_back(yaw);
+ preInitDiameters.push_back(diameter);
+ preInitBladeRadii.push_back(bladeRadii);
+}
+
+void ActuatorFarm::init(Parameter* para, GridProvider* gridProvider, CudaMemoryManager* cudaMemoryManager)
+{
+ if(!para->getIsBodyForce()) throw std::runtime_error("try to allocate ActuatorFarm but BodyForce is not set in Parameter.");
+ this->forceRatio = para->getForceRatio();
+ this->initTurbineGeometries(cudaMemoryManager);
+ this->initBladeCoords(cudaMemoryManager);
+ this->initBladeIndices(para, cudaMemoryManager);
+ this->initBladeVelocities(cudaMemoryManager);
+ this->initBladeForces(cudaMemoryManager);
+ this->initBoundingSpheres(para, cudaMemoryManager);
+ this->streamIndex = 0;
+}
+
+void ActuatorFarm::interact(Parameter* para, CudaMemoryManager* cudaMemoryManager, int level, unsigned int t)
+{
+ if (level != this->level) return;
+
+ cudaStream_t stream = para->getStreamManager()->getStream(CudaStreamIndex::ActuatorFarm, this->streamIndex);
+
+ if(useHostArrays) cudaMemoryManager->cudaCopyBladeCoordsHtoD(this);
+
+ vf::cuda::CudaGrid bladeGrid = vf::cuda::CudaGrid(para->getParH(level)->numberofthreads, this->numberOfNodes);
+
+ interpolateVelocities<<< bladeGrid.grid, bladeGrid.threads, 0, stream >>>(
+ para->getParD(this->level)->coordinateX, para->getParD(this->level)->coordinateY, para->getParD(this->level)->coordinateZ,
+ para->getParD(this->level)->neighborX, para->getParD(this->level)->neighborY, para->getParD(this->level)->neighborZ, para->getParD(this->level)->neighborInverse,
+ para->getParD(this->level)->velocityX, para->getParD(this->level)->velocityY, para->getParD(this->level)->velocityZ,
+ this->bladeCoordsXDCurrentTimestep, this->bladeCoordsYDCurrentTimestep, this->bladeCoordsZDCurrentTimestep,
+ this->bladeVelocitiesXDCurrentTimestep, this->bladeVelocitiesYDCurrentTimestep, this->bladeVelocitiesZDCurrentTimestep,
+ this->numberOfTurbines, this->numberOfBlades, this->numberOfBladeNodes,
+ this->azimuthsD, this->yawsD, this->omegasD,
+ this->turbinePosXD, this->turbinePosYD, this->turbinePosZD,
+ this->bladeIndicesD, para->getVelocityRatio(), this->invDeltaX);
+
+ cudaStreamSynchronize(stream);
+ if(useHostArrays) cudaMemoryManager->cudaCopyBladeVelocitiesDtoH(this);
+ this->calcBladeForces();
+ this->swapDeviceArrays();
+
+ if(useHostArrays) cudaMemoryManager->cudaCopyBladeForcesHtoD(this);
+
+ vf::cuda::CudaGrid sphereGrid = vf::cuda::CudaGrid(para->getParH(level)->numberofthreads, this->numberOfIndices);
+
+ applyBodyForces<<<sphereGrid.grid, sphereGrid.threads, 0, stream>>>(
+ para->getParD(this->level)->coordinateX, para->getParD(this->level)->coordinateY, para->getParD(this->level)->coordinateZ,
+ para->getParD(this->level)->forceX_SP, para->getParD(this->level)->forceY_SP, para->getParD(this->level)->forceZ_SP,
+ this->bladeCoordsXDCurrentTimestep, this->bladeCoordsYDCurrentTimestep, this->bladeCoordsZDCurrentTimestep,
+ this->bladeForcesXDCurrentTimestep, this->bladeForcesYDCurrentTimestep, this->bladeForcesZDCurrentTimestep,
+ this->numberOfTurbines, this->numberOfBlades, this->numberOfBladeNodes,
+ this->azimuthsD, this->yawsD, this->diametersD,
+ this->turbinePosXD, this->turbinePosYD, this->turbinePosZD,
+ this->boundingSphereIndicesD, this->numberOfIndices,
+ this->invEpsilonSqrd, this->factorGaussian);
+ cudaMemoryManager->cudaCopyBladeOrientationsHtoD(this);
+ cudaStreamSynchronize(stream);
+}
+
+
+void ActuatorFarm::free(Parameter* para, CudaMemoryManager* cudaMemoryManager)
+{
+ cudaMemoryManager->cudaFreeBladeGeometries(this);
+ cudaMemoryManager->cudaFreeBladeOrientations(this);
+ cudaMemoryManager->cudaFreeBladeCoords(this);
+ cudaMemoryManager->cudaFreeBladeVelocities(this);
+ cudaMemoryManager->cudaFreeBladeForces(this);
+ cudaMemoryManager->cudaFreeBladeIndices(this);
+ cudaMemoryManager->cudaFreeSphereIndices(this);
+}
+
+
+void ActuatorFarm::calcForcesEllipticWing()
+{
+ real u_rel, v_rel, u_rel_sq;
+ real phi;
+ real Cl = c1o1;
+ real Cd = c0o1;
+ real c0 = 20*c1o10;
+ real c, Cn, Ct;
+ for(uint turbine=0; turbine<this->numberOfTurbines; turbine++)
+ {
+ real diameter = this->diametersH[turbine];
+ for( uint blade=0; blade<this->numberOfBlades; blade++)
+ {
+ for( uint bladeNode=0; bladeNode<this->numberOfBladeNodes; bladeNode++)
+ {
+ uint node = calcNode(bladeNode, this->numberOfBladeNodes, blade, this->numberOfBlades, turbine, this->numberOfTurbines);
+
+ u_rel = this->bladeVelocitiesXH[node];
+ v_rel = this->bladeVelocitiesYH[node];
+ u_rel_sq = u_rel*u_rel+v_rel*v_rel;
+ phi = atan2(u_rel, v_rel);
+
+ real tmp = c4o1*this->bladeRadiiH[bladeNode]/diameter-c1o1;
+ c = c0 * sqrt( c1o1- tmp*tmp );
+ Cn = Cl*cos(phi)+Cd*sin(phi);
+ Ct = Cl*sin(phi)-Cd*cos(phi);
+ real fx = c1o2*u_rel_sq*c*this->density*Cn;
+ real fy = c1o2*u_rel_sq*c*this->density*Ct;
+ this->bladeForcesXH[node] = -fx;
+ this->bladeForcesYH[node] = -fy;
+ this->bladeForcesZH[node] = c0o1;
+ // printf("u %f v %f fx %f fy %f \n", u_rel, v_rel, fx, fy);
+ }
+ }
+ azimuthsH[turbine] = azimuthsH[turbine]+deltaT*omegasH[turbine];
+ }
+}
+
+void ActuatorFarm::calcBladeForces()
+{
+ this->calcForcesEllipticWing();
+}
+
+void ActuatorFarm::getTaggedFluidNodes(Parameter *para, GridProvider* gridProvider)
+{
+ std::vector<uint> indicesInSphere(this->boundingSphereIndicesH, this->boundingSphereIndicesH+this->numberOfIndices);
+ gridProvider->tagFluidNodeIndices(indicesInSphere, CollisionTemplate::AllFeatures, this->level);
+}
+
+
+void ActuatorFarm::initTurbineGeometries(CudaMemoryManager* cudaMemoryManager)
+{
+ this->numberOfTurbines = uint(this->preInitDiameters.size());
+ this->numberOfNodes = numberOfTurbines*numberOfBladeNodes*numberOfBlades;
+
+ cudaMemoryManager->cudaAllocBladeGeometries(this);
+ cudaMemoryManager->cudaAllocBladeOrientations(this);
+
+ for(uint turbine=0; turbine<this->numberOfTurbines; turbine++)
+ {
+ for(uint node=0; node<this->numberOfBladeNodes; node++)
+ {
+ this->bladeRadiiH[calcNode(node, numberOfBladeNodes, 0, 1, turbine, numberOfTurbines)] = this->preInitBladeRadii[turbine][node];
+ }
+
+ }
+ std::copy(preInitPosX.begin(), preInitPosX.end(), turbinePosXH);
+ std::copy(preInitPosY.begin(), preInitPosY.end(), turbinePosYH);
+ std::copy(preInitPosZ.begin(), preInitPosZ.end(), turbinePosZH);
+ std::copy(preInitDiameters.begin(), preInitDiameters.end(), diametersH);
+
+ cudaMemoryManager->cudaCopyBladeGeometriesHtoD(this);
+ std::copy(preInitAzimuths.begin(), preInitAzimuths.end(), this->azimuthsH);
+ std::copy(preInitOmegas.begin(), preInitOmegas.end(), this->omegasH);
+ std::copy(preInitYaws.begin(), preInitYaws.end(), this->yawsH);
+
+ cudaMemoryManager->cudaCopyBladeOrientationsHtoD(this);
+ this->factorGaussian = pow(this->epsilon*sqrt(cPi),-c3o1)/this->forceRatio;
+}
+
+void ActuatorFarm::initBladeCoords(CudaMemoryManager* cudaMemoryManager)
+{
+ cudaMemoryManager->cudaAllocBladeCoords(this);
+
+ for(uint turbine=0; turbine<numberOfTurbines; turbine++)
+ {
+ for(uint blade=0; blade<this->numberOfBlades; blade++)
+ {
+ for(uint bladeNode=0; bladeNode<this->numberOfBladeNodes; bladeNode++)
+ {
+ uint node = calcNode(bladeNode, this->numberOfBladeNodes, blade, this->numberOfBlades, turbine, this->numberOfTurbines);
+
+ this->bladeCoordsXH[node] = c0o1;
+ this->bladeCoordsYH[node] = c0o1;
+ this->bladeCoordsZH[node] = this->bladeRadiiH[calcNode(bladeNode, numberOfBladeNodes, 0, 1, turbine, numberOfTurbines)];
+ }
+ }
+ }
+ cudaMemoryManager->cudaCopyBladeCoordsHtoD(this);
+ swapArrays(this->bladeCoordsXDCurrentTimestep, this->bladeCoordsXDPreviousTimestep);
+ swapArrays(this->bladeCoordsYDCurrentTimestep, this->bladeCoordsYDPreviousTimestep);
+ swapArrays(this->bladeCoordsZDCurrentTimestep, this->bladeCoordsZDPreviousTimestep);
+ cudaMemoryManager->cudaCopyBladeCoordsHtoD(this);
+}
+
+void ActuatorFarm::initBladeVelocities(CudaMemoryManager* cudaMemoryManager)
+{
+ cudaMemoryManager->cudaAllocBladeVelocities(this);
+
+ std::fill_n(this->bladeVelocitiesXH, this->numberOfNodes, c0o1);
+ std::fill_n(this->bladeVelocitiesYH, this->numberOfNodes, c0o1);
+ std::fill_n(this->bladeVelocitiesZH, this->numberOfNodes, c0o1);
+
+ cudaMemoryManager->cudaCopyBladeVelocitiesHtoD(this);
+ swapArrays(this->bladeVelocitiesXDCurrentTimestep, this->bladeVelocitiesXDPreviousTimestep);
+ swapArrays(this->bladeVelocitiesYDCurrentTimestep, this->bladeVelocitiesYDPreviousTimestep);
+ swapArrays(this->bladeVelocitiesZDCurrentTimestep, this->bladeVelocitiesZDPreviousTimestep);
+ cudaMemoryManager->cudaCopyBladeVelocitiesHtoD(this);
+}
+
+void ActuatorFarm::initBladeForces(CudaMemoryManager* cudaMemoryManager)
+{
+ cudaMemoryManager->cudaAllocBladeForces(this);
+
+ std::fill_n(this->bladeForcesXH, this->numberOfNodes, c0o1);
+ std::fill_n(this->bladeForcesYH, this->numberOfNodes, c0o1);
+ std::fill_n(this->bladeForcesZH, this->numberOfNodes, c0o1);
+
+ cudaMemoryManager->cudaCopyBladeForcesHtoD(this);
+ swapArrays(this->bladeForcesXDCurrentTimestep, this->bladeForcesXDPreviousTimestep);
+ swapArrays(this->bladeForcesYDCurrentTimestep, this->bladeForcesYDPreviousTimestep);
+ swapArrays(this->bladeForcesZDCurrentTimestep, this->bladeForcesZDPreviousTimestep);
+ cudaMemoryManager->cudaCopyBladeForcesHtoD(this);
+}
+
+void ActuatorFarm::initBladeIndices(Parameter* para, CudaMemoryManager* cudaMemoryManager)
+{
+ cudaMemoryManager->cudaAllocBladeIndices(this);
+
+ std::fill_n(this->bladeIndicesH, this->numberOfNodes, 1);
+
+ cudaMemoryManager->cudaCopyBladeIndicesHtoD(this);
+}
+
+void ActuatorFarm::initBoundingSpheres(Parameter* para, CudaMemoryManager* cudaMemoryManager)
+{
+ std::vector<int> nodesInSpheres;
+
+ for(uint turbine=0; turbine<this->numberOfTurbines; turbine++)
+ {
+ real sphereRadius = c1o2*this->diametersH[turbine]+c4o1*this->epsilon;
+
+ real posX = this->turbinePosXH[turbine];
+ real posY = this->turbinePosYH[turbine];
+ real posZ = this->turbinePosZH[turbine];
+
+ real sphereRadiusSqrd = sphereRadius*sphereRadius;
+
+ uint minimumNumberOfNodesPerSphere = (uint)(c4o3*cPi*pow(sphereRadius-this->deltaX, c3o1)/pow(this->deltaX, c3o1));
+ uint nodesInThisSphere = 0;
+
+ for (size_t pos = 1; pos <= para->getParH(this->level)->numberOfNodes; pos++)
+ {
+ const real distX = para->getParH(this->level)->coordinateX[pos]-posX;
+ const real distY = para->getParH(this->level)->coordinateY[pos]-posY;
+ const real distZ = para->getParH(this->level)->coordinateZ[pos]-posZ;
+ if(distSqrd(distX,distY,distZ) < sphereRadiusSqrd)
+ {
+ nodesInSpheres.push_back((int)pos);
+ nodesInThisSphere++;
+ }
+ }
+
+ if(nodesInThisSphere<minimumNumberOfNodesPerSphere)
+ {
+ VF_LOG_CRITICAL("Found only {} nodes in bounding sphere of turbine no. {}, expected at least {}!", nodesInThisSphere, turbine, minimumNumberOfNodesPerSphere);
+ throw std::runtime_error("ActuatorFarm::initBoundingSpheres: Turbine bounding sphere partially out of domain.");
+ }
+ }
+
+ this->numberOfIndices = uint(nodesInSpheres.size());
+
+ cudaMemoryManager->cudaAllocSphereIndices(this);
+ std::copy(nodesInSpheres.begin(), nodesInSpheres.end(), this->boundingSphereIndicesH);
+ cudaMemoryManager->cudaCopySphereIndicesHtoD(this);
+}
+
+void ActuatorFarm::setAllAzimuths(real* _azimuths)
+{
+ std::copy_n(_azimuths, this->numberOfTurbines, this->azimuthsH);
+}
+
+void ActuatorFarm::setAllOmegas(real* _omegas)
+{
+ std::copy_n(_omegas, this->numberOfTurbines, this->omegasH);
+}
+
+void ActuatorFarm::setAllYaws(real* _yaws)
+{
+ std::copy_n(_yaws, this->numberOfTurbines, this->yawsH);
+}
+
+void ActuatorFarm::setAllBladeCoords(real* _bladeCoordsX, real* _bladeCoordsY, real* _bladeCoordsZ)
+{
+ std::copy_n(_bladeCoordsX, this->numberOfNodes, this->bladeCoordsXH);
+ std::copy_n(_bladeCoordsY, this->numberOfNodes, this->bladeCoordsYH);
+ std::copy_n(_bladeCoordsZ, this->numberOfNodes, this->bladeCoordsZH);
+}
+
+void ActuatorFarm::setAllBladeVelocities(real* _bladeVelocitiesX, real* _bladeVelocitiesY, real* _bladeVelocitiesZ)
+{
+ std::copy_n(_bladeVelocitiesX, this->numberOfNodes, this->bladeVelocitiesXH);
+ std::copy_n(_bladeVelocitiesY, this->numberOfNodes, this->bladeVelocitiesYH);
+ std::copy_n(_bladeVelocitiesZ, this->numberOfNodes, this->bladeVelocitiesZH);
+}
+
+void ActuatorFarm::setAllBladeForces(real* _bladeForcesX, real* _bladeForcesY, real* _bladeForcesZ)
+{
+ std::copy_n(_bladeForcesX, this->numberOfNodes, this->bladeForcesXH);
+ std::copy_n(_bladeForcesY, this->numberOfNodes, this->bladeForcesYH);
+ std::copy_n(_bladeForcesZ, this->numberOfNodes, this->bladeForcesZH);
+
+}void ActuatorFarm::setTurbineBladeCoords(uint turbine, real* _bladeCoordsX, real* _bladeCoordsY, real* _bladeCoordsZ)
+{
+ std::copy_n(_bladeCoordsX, numberOfBladeNodes*numberOfBlades, &this->bladeCoordsXH[turbine*numberOfBladeNodes*numberOfBlades]);
+ std::copy_n(_bladeCoordsY, numberOfBladeNodes*numberOfBlades, &this->bladeCoordsYH[turbine*numberOfBladeNodes*numberOfBlades]);
+ std::copy_n(_bladeCoordsZ, numberOfBladeNodes*numberOfBlades, &this->bladeCoordsZH[turbine*numberOfBladeNodes*numberOfBlades]);
+}
+
+void ActuatorFarm::setTurbineBladeVelocities(uint turbine, real* _bladeVelocitiesX, real* _bladeVelocitiesY, real* _bladeVelocitiesZ)
+{
+ std::copy_n(_bladeVelocitiesX, numberOfBladeNodes*numberOfBlades, &this->bladeVelocitiesXH[turbine*numberOfBladeNodes*numberOfBlades]);
+ std::copy_n(_bladeVelocitiesY, numberOfBladeNodes*numberOfBlades, &this->bladeVelocitiesYH[turbine*numberOfBladeNodes*numberOfBlades]);
+ std::copy_n(_bladeVelocitiesZ, numberOfBladeNodes*numberOfBlades, &this->bladeVelocitiesZH[turbine*numberOfBladeNodes*numberOfBlades]);
+}
+
+void ActuatorFarm::setTurbineBladeForces(uint turbine, real* _bladeForcesX, real* _bladeForcesY, real* _bladeForcesZ)
+{
+ std::copy_n(_bladeForcesX, numberOfBladeNodes*numberOfBlades, &this->bladeForcesXH[turbine*numberOfBladeNodes*numberOfBlades]);
+ std::copy_n(_bladeForcesY, numberOfBladeNodes*numberOfBlades, &this->bladeForcesYH[turbine*numberOfBladeNodes*numberOfBlades]);
+ std::copy_n(_bladeForcesZ, numberOfBladeNodes*numberOfBlades, &this->bladeForcesZH[turbine*numberOfBladeNodes*numberOfBlades]);
+}
+
+void ActuatorFarm::swapDeviceArrays()
+{
+ swapArrays(this->bladeCoordsXDPreviousTimestep, this->bladeCoordsXDCurrentTimestep);
+ swapArrays(this->bladeCoordsYDPreviousTimestep, this->bladeCoordsYDCurrentTimestep);
+ swapArrays(this->bladeCoordsZDPreviousTimestep, this->bladeCoordsZDCurrentTimestep);
+
+ swapArrays(this->bladeVelocitiesXDPreviousTimestep, this->bladeVelocitiesXDCurrentTimestep);
+ swapArrays(this->bladeVelocitiesYDPreviousTimestep, this->bladeVelocitiesYDCurrentTimestep);
+ swapArrays(this->bladeVelocitiesZDPreviousTimestep, this->bladeVelocitiesZDCurrentTimestep);
+
+ swapArrays(this->bladeForcesXDPreviousTimestep, this->bladeForcesXDCurrentTimestep);
+ swapArrays(this->bladeForcesYDPreviousTimestep, this->bladeForcesYDCurrentTimestep);
+ swapArrays(this->bladeForcesZDPreviousTimestep, this->bladeForcesZDCurrentTimestep);
+}
\ No newline at end of file
diff --git a/src/gpu/VirtualFluids_GPU/PreCollisionInteractor/ActuatorFarm.h b/src/gpu/VirtualFluids_GPU/PreCollisionInteractor/ActuatorFarm.h
new file mode 100644
index 0000000000000000000000000000000000000000..8e21cdb6b21efd323f6723e21d6b28614109f1ec
--- /dev/null
+++ b/src/gpu/VirtualFluids_GPU/PreCollisionInteractor/ActuatorFarm.h
@@ -0,0 +1,197 @@
+#ifndef ActuatorFarm_H
+#define ActuatorFarm_H
+
+#include "PreCollisionInteractor.h"
+#include "PointerDefinitions.h"
+#include "lbm/constants/NumericConstants.h"
+#include <stdexcept>
+
+using namespace vf::lbm::constant;
+
+class Parameter;
+class GridProvider;
+using namespace vf::lbm::constant;
+
+class ActuatorFarm : public PreCollisionInteractor
+{
+public:
+ ActuatorFarm(
+ const uint _nBlades,
+ const real _density,
+ const uint _nBladeNodes,
+ const real _epsilon,
+ int _level,
+ const real _deltaT,
+ const real _deltaX,
+ const bool _useHostArrays
+ ) :
+ numberOfBlades(_nBlades),
+ density(_density),
+ numberOfBladeNodes(_nBladeNodes),
+ epsilon(_epsilon),
+ level(_level),
+ useHostArrays(_useHostArrays),
+ numberOfTurbines(0),
+ numberOfNodes(0),
+ PreCollisionInteractor()
+ {
+ this->deltaT = _deltaT*exp2(-this->level);
+ this->deltaX = _deltaX*exp2(-this->level);
+ this->invEpsilonSqrd = 1/(epsilon*epsilon);
+ this->invDeltaX = c1o1/this->deltaX;
+
+ if(this->epsilon<this->deltaX)
+ throw std::runtime_error("ActuatorFarm::ActuatorFarm: epsilon needs to be larger than dx!");
+ }
+
+ ~ActuatorFarm() override = default;
+ void addTurbine(real turbinePosX, real turbinePosY, real turbinePosZ, real diameter, real omega, real azimuth, real yaw, std::vector<real> bladeRadii);
+ void init(Parameter* para, GridProvider* gridProvider, CudaMemoryManager* cudaManager) override;
+ void interact(Parameter* para, CudaMemoryManager* cudaManager, int level, uint t) override;
+ void free(Parameter* para, CudaMemoryManager* cudaManager) override;
+ void getTaggedFluidNodes(Parameter *para, GridProvider* gridProvider) override;
+
+ void write(uint t);
+
+ real getDensity(){ return this->density; };
+ real getDeltaT(){ return this->deltaT; };
+ real getDeltaX(){ return this->deltaX; };
+
+ uint getNumberOfTurbines(){ return this->numberOfTurbines; };
+ uint getNumberOfNodesPerBlade(){ return this->numberOfBladeNodes; };
+ uint getNumberOfBladesPerTurbine(){ return this->numberOfBlades; };
+
+ uint getNumberOfIndices(){ return this->numberOfIndices; };
+ uint getNumberOfNodes(){ return this->numberOfNodes; };
+
+ real* getAllAzimuths(){ return azimuthsH; };
+ real* getAllOmegas(){ return omegasH; };
+ real* getAllYaws(){ return yawsH; };
+
+ real* getAllTurbinePosX(){ return turbinePosXH; };
+ real* getAllTurbinePosY(){ return turbinePosYH; };
+ real* getAllTurbinePosZ(){ return turbinePosZH; };
+
+ real getTurbineAzimuth(uint turbine){ return azimuthsH[turbine]; };
+ real getTurbineOmega (uint turbine){ return omegasH[turbine]; };
+ real getTurbineYaw (uint turbine){ return yawsH[turbine]; };
+
+ real getTurbinePosX(uint turbine){ return turbinePosXH[turbine]; };
+ real getTurbinePosY(uint turbine){ return turbinePosYH[turbine]; };
+ real getTurbinePosZ(uint turbine){ return turbinePosZH[turbine]; };
+
+ real* getAllBladeRadii(){ return this->bladeRadiiH; };
+ real* getAllBladeCoordsX(){ return this->bladeCoordsXH; };
+ real* getAllBladeCoordsY(){ return this->bladeCoordsYH; };
+ real* getAllBladeCoordsZ(){ return this->bladeCoordsZH; };
+ real* getAllBladeVelocitiesX(){ return this->bladeVelocitiesXH; };
+ real* getAllBladeVelocitiesY(){ return this->bladeVelocitiesYH; };
+ real* getAllBladeVelocitiesZ(){ return this->bladeVelocitiesZH; };
+ real* getAllBladeForcesX(){ return this->bladeForcesXH; };
+ real* getAllBladeForcesY(){ return this->bladeForcesYH; };
+ real* getAllBladeForcesZ(){ return this->bladeForcesZH; };
+
+ real* getTurbineBladeRadii(uint turbine){ return &this->bladeRadiiH[turbine*numberOfBladeNodes*numberOfBlades]; };
+ real* getTurbineBladeCoordsX(uint turbine){ return &this->bladeCoordsXH[turbine*numberOfBladeNodes*numberOfBlades]; };
+ real* getTurbineBladeCoordsY(uint turbine){ return &this->bladeCoordsYH[turbine*numberOfBladeNodes*numberOfBlades]; };
+ real* getTurbineBladeCoordsZ(uint turbine){ return &this->bladeCoordsZH[turbine*numberOfBladeNodes*numberOfBlades]; };
+ real* getTurbineBladeVelocitiesX(uint turbine){ return &this->bladeVelocitiesXH[turbine*numberOfBladeNodes*numberOfBlades]; };
+ real* getTurbineBladeVelocitiesY(uint turbine){ return &this->bladeVelocitiesYH[turbine*numberOfBladeNodes*numberOfBlades]; };
+ real* getTurbineBladeVelocitiesZ(uint turbine){ return &this->bladeVelocitiesZH[turbine*numberOfBladeNodes*numberOfBlades]; };
+ real* getTurbineBladeForcesX(uint turbine){ return &this->bladeForcesXH[turbine*numberOfBladeNodes*numberOfBlades]; };
+ real* getTurbineBladeForcesY(uint turbine){ return &this->bladeForcesYH[turbine*numberOfBladeNodes*numberOfBlades]; };
+ real* getTurbineBladeForcesZ(uint turbine){ return &this->bladeForcesZH[turbine*numberOfBladeNodes*numberOfBlades]; };
+
+ real* getAllBladeRadiiDevice(){ return this->bladeRadiiD; };
+ real* getAllBladeCoordsXDevice(){ return this->bladeCoordsXDCurrentTimestep; };
+ real* getAllBladeCoordsYDevice(){ return this->bladeCoordsYDCurrentTimestep; };
+ real* getAllBladeCoordsZDevice(){ return this->bladeCoordsZDCurrentTimestep; };
+ real* getAllBladeVelocitiesXDevice(){ return this->bladeVelocitiesXDCurrentTimestep; };
+ real* getAllBladeVelocitiesYDevice(){ return this->bladeVelocitiesYDCurrentTimestep; };
+ real* getAllBladeVelocitiesZDevice(){ return this->bladeVelocitiesZDCurrentTimestep; };
+ real* getAllBladeForcesXDevice(){ return this->bladeForcesXDCurrentTimestep; };
+ real* getAllBladeForcesYDevice(){ return this->bladeForcesYDCurrentTimestep; };
+ real* getAllBladeForcesZDevice(){ return this->bladeForcesZDCurrentTimestep; };
+
+ real* getTurbineBladeRadiiDevice(uint turbine){ return &this->bladeRadiiD[turbine*numberOfBladeNodes]; };
+ real* getTurbineBladeCoordsXDevice(uint turbine){ return &this->bladeCoordsXDCurrentTimestep[turbine*numberOfBladeNodes*numberOfBlades]; };
+ real* getTurbineBladeCoordsYDevice(uint turbine){ return &this->bladeCoordsYDCurrentTimestep[turbine*numberOfBladeNodes*numberOfBlades]; };
+ real* getTurbineBladeCoordsZDevice(uint turbine){ return &this->bladeCoordsZDCurrentTimestep[turbine*numberOfBladeNodes*numberOfBlades]; };
+ real* getTurbineBladeVelocitiesXDevice(uint turbine){ return &this->bladeVelocitiesXDCurrentTimestep[turbine*numberOfBladeNodes*numberOfBlades]; };
+ real* getTurbineBladeVelocitiesYDevice(uint turbine){ return &this->bladeVelocitiesYDCurrentTimestep[turbine*numberOfBladeNodes*numberOfBlades]; };
+ real* getTurbineBladeVelocitiesZDevice(uint turbine){ return &this->bladeVelocitiesZDCurrentTimestep[turbine*numberOfBladeNodes*numberOfBlades]; };
+ real* getTurbineBladeForcesXDevice(uint turbine){ return &this->bladeForcesXDCurrentTimestep[turbine*numberOfBladeNodes*numberOfBlades]; };
+ real* getTurbineBladeForcesYDevice(uint turbine){ return &this->bladeForcesYDCurrentTimestep[turbine*numberOfBladeNodes*numberOfBlades]; };
+ real* getTurbineBladeForcesZDevice(uint turbine){ return &this->bladeForcesZDCurrentTimestep[turbine*numberOfBladeNodes*numberOfBlades]; };
+
+ void setAllAzimuths(real* _azimuth);
+ void setAllOmegas(real* _omegas);
+ void setAllYaws(real* yaws);
+
+ void setTurbineAzimuth(uint turbine, real azimuth){ azimuthsH[turbine] = azimuth; };
+ void setTurbineYaw(uint turbine, real yaw){ yawsH[turbine] = yaw; };
+ void setTurbineOmega(uint turbine, real omega){ omegasH[turbine] = omega; };
+
+ void setAllBladeCoords(real* _bladeCoordsX, real* _bladeCoordsY, real* _bladeCoordsZ);
+ void setAllBladeVelocities(real* _bladeVelocitiesX, real* _bladeVelocitiesY, real* _bladeVelocitiesZ);
+ void setAllBladeForces(real* _bladeForcesX, real* _bladeForcesY, real* _bladeForcesZ);
+
+ void setTurbineBladeCoords(uint turbine, real* _bladeCoordsX, real* _bladeCoordsY, real* _bladeCoordsZ);
+ void setTurbineBladeVelocities(uint turbine, real* _bladeVelocitiesX, real* _bladeVelocitiesY, real* _bladeVelocitiesZ);
+ void setTurbineBladeForces(uint turbine, real* _bladeForcesX, real* _bladeForcesY, real* _bladeForcesZ);
+
+ virtual void calcBladeForces();
+
+private:
+ void initTurbineGeometries(CudaMemoryManager* cudaManager);
+ void initBoundingSpheres(Parameter* para, CudaMemoryManager* cudaManager);
+ void initBladeCoords(CudaMemoryManager* cudaManager);
+ void initBladeVelocities(CudaMemoryManager* cudaManager);
+ void initBladeForces(CudaMemoryManager* cudaManager);
+ void initBladeIndices(Parameter* para, CudaMemoryManager* cudaManager);
+
+ void calcForcesEllipticWing();
+ void rotateBlades(real angle, uint turbineID);
+
+ void writeBladeCoords(uint t);
+ void writeBladeForces(uint t);
+ void writeBladeVelocities(uint t);
+
+ void swapDeviceArrays();
+
+public:
+ real* bladeRadiiH;
+ real* bladeRadiiD;
+ real* bladeCoordsXH, * bladeCoordsYH, * bladeCoordsZH;
+ real* bladeCoordsXDPreviousTimestep, * bladeCoordsYDPreviousTimestep, * bladeCoordsZDPreviousTimestep;
+ real* bladeCoordsXDCurrentTimestep, * bladeCoordsYDCurrentTimestep, * bladeCoordsZDCurrentTimestep;
+ real* bladeVelocitiesXH, * bladeVelocitiesYH, * bladeVelocitiesZH;
+ real* bladeVelocitiesXDPreviousTimestep, * bladeVelocitiesYDPreviousTimestep, * bladeVelocitiesZDPreviousTimestep;
+ real* bladeVelocitiesXDCurrentTimestep, * bladeVelocitiesYDCurrentTimestep, * bladeVelocitiesZDCurrentTimestep;
+ real* bladeForcesXH, * bladeForcesYH, * bladeForcesZH;
+ real* bladeForcesXDPreviousTimestep, * bladeForcesYDPreviousTimestep, * bladeForcesZDPreviousTimestep;
+ real* bladeForcesXDCurrentTimestep, * bladeForcesYDCurrentTimestep, * bladeForcesZDCurrentTimestep;
+ uint* bladeIndicesH;
+ uint* bladeIndicesD;
+ uint* boundingSphereIndicesH;
+ uint* boundingSphereIndicesD;
+ real* turbinePosXH, *turbinePosYH, *turbinePosZH, *omegasH, *azimuthsH, *yawsH, *diametersH;
+ real* turbinePosXD, *turbinePosYD, *turbinePosZD, *omegasD, *azimuthsD, *yawsD, *diametersD;
+
+private:
+ std::vector<real> preInitPosX, preInitPosY, preInitPosZ, preInitDiameters, preInitOmegas, preInitAzimuths, preInitYaws;
+ std::vector<std::vector<real>> preInitBladeRadii;
+ const bool useHostArrays;
+ const real density;
+ real deltaT, deltaX;
+ const uint numberOfBladeNodes, numberOfBlades;
+ uint numberOfTurbines;
+ const real epsilon; // in m
+ const int level;
+ uint numberOfIndices;
+ uint numberOfNodes;
+ real forceRatio, factorGaussian, invEpsilonSqrd, invDeltaX;
+ int streamIndex;
+};
+
+#endif
diff --git a/src/gpu/VirtualFluids_GPU/PreCollisionInteractor/ActuatorLine.cu b/src/gpu/VirtualFluids_GPU/PreCollisionInteractor/ActuatorLine.cu
deleted file mode 100644
index 71897bd21ea4fb299d3cc0ffa385506d4503f360..0000000000000000000000000000000000000000
--- a/src/gpu/VirtualFluids_GPU/PreCollisionInteractor/ActuatorLine.cu
+++ /dev/null
@@ -1,423 +0,0 @@
-#include "ActuatorLine.h"
-
-#include <cuda.h>
-#include <cuda_runtime.h>
-#include <helper_cuda.h>
-
-#include <cuda/CudaGrid.h>
-#include "VirtualFluids_GPU/GPU/GeometryUtils.h"
-
-#include "Parameter/Parameter.h"
-#include "DataStructureInitializer/GridProvider.h"
-#include "GPU/CudaMemoryManager.h"
-
-__host__ __device__ __inline__ uint calcNode(uint bladeNode, uint nBladeNodes, uint blade, uint nBlades)
-{
- return bladeNode+blade*nBladeNodes;
-}
-
-__host__ __device__ __inline__ void calcBladeAndBladeNode(uint node, uint& bladeNode, uint nBladeNodes, uint& blade, uint nBlades)
-{
- blade = node/nBladeNodes;
- bladeNode = node - blade*nBladeNodes;
-}
-
-__host__ __device__ __forceinline__ real distSqrd(real distX, real distY, real distZ)
-{
- return distX*distX+distY*distY+distZ*distZ;
-}
-
-__host__ __device__ __inline__ void rotateFromBladeToGlobal(
- real& bladeCoordX_BF, real& bladeCoordY_BF, real& bladeCoordZ_BF,
- real& bladeCoordX_GF, real& bladeCoordY_GF, real& bladeCoordZ_GF,
- real& azimuth, real& yaw)
-{
- real tmpX, tmpY, tmpZ;
-
- rotateAboutX3D(azimuth, bladeCoordX_BF, bladeCoordY_BF, bladeCoordZ_BF, tmpX, tmpY, tmpZ);
- rotateAboutZ3D(yaw, tmpX, tmpY, tmpZ, bladeCoordX_GF, bladeCoordY_GF, bladeCoordZ_GF);
-
-}
-
-__host__ __device__ __inline__ void rotateFromGlobalToBlade(
- real& bladeCoordX_BF, real& bladeCoordY_BF, real& bladeCoordZ_BF,
- real& bladeCoordX_GF, real& bladeCoordY_GF, real& bladeCoordZ_GF,
- real& azimuth, real& yaw)
-{
- real tmpX, tmpY, tmpZ;
-
- invRotateAboutZ3D(yaw, bladeCoordX_GF, bladeCoordY_GF, bladeCoordZ_GF, tmpX, tmpY, tmpZ);
- invRotateAboutX3D(azimuth, tmpX, tmpY, tmpZ, bladeCoordX_BF, bladeCoordY_BF, bladeCoordZ_BF);
-}
-
-__global__ void interpolateVelocities(real* gridCoordsX, real* gridCoordsY, real* gridCoordsZ,
- uint* neighborsX, uint* neighborsY, uint* neighborsZ, uint* neighborsWSB,
- real* vx, real* vy, real* vz,
- real* bladeCoordsX, real* bladeCoordsY, real* bladeCoordsZ,
- real* bladeVelocitiesX, real* bladeVelocitiesY, real* bladeVelocitiesZ,
- uint nBlades, uint nBladeNodes,
- real azimuth, real yaw, real omega,
- real turbPosX, real turbPosY, real turbPosZ,
- uint* bladeIndices, real velocityRatio, real invDeltaX)
-{
- const uint x = threadIdx.x;
- const uint y = blockIdx.x;
- const uint z = blockIdx.y;
-
- const uint nx = blockDim.x;
- const uint ny = gridDim.x;
-
- const uint node = nx*(ny*z + y) + x;
-
- uint bladeNode, blade;
-
- calcBladeAndBladeNode(node, bladeNode, nBladeNodes, blade, nBlades);
-
- if(node>=nBladeNodes*nBlades) return;
-
- real bladeCoordX_BF = bladeCoordsX[node];
- real bladeCoordY_BF = bladeCoordsY[node];
- real bladeCoordZ_BF = bladeCoordsZ[node];
-
- real bladeCoordX_GF, bladeCoordY_GF, bladeCoordZ_GF;
-
- real localAzimuth = azimuth+blade*c2Pi/nBlades;
-
- rotateFromBladeToGlobal(bladeCoordX_BF, bladeCoordY_BF, bladeCoordZ_BF,
- bladeCoordX_GF, bladeCoordY_GF, bladeCoordZ_GF,
- localAzimuth, yaw);
-
- bladeCoordX_GF += turbPosX;
- bladeCoordY_GF += turbPosY;
- bladeCoordZ_GF += turbPosZ;
-
- uint k, ke, kn, kt;
- uint kne, kte, ktn, ktne;
-
- k = findNearestCellBSW(bladeIndices[node],
- gridCoordsX, gridCoordsY, gridCoordsZ,
- bladeCoordX_GF, bladeCoordY_GF, bladeCoordZ_GF,
- neighborsX, neighborsY, neighborsZ, neighborsWSB);
-
- bladeIndices[node] = k;
-
- getNeighborIndicesOfBSW(k, ke, kn, kt, kne, kte, ktn, ktne, neighborsX, neighborsY, neighborsZ);
-
- real dW, dE, dN, dS, dT, dB;
-
- real distX = invDeltaX*(bladeCoordX_GF-gridCoordsX[k]);
- real distY = invDeltaX*(bladeCoordY_GF-gridCoordsY[k]);
- real distZ = invDeltaX*(bladeCoordZ_GF-gridCoordsZ[k]);
-
- getInterpolationWeights(dW, dE, dN, dS, dT, dB, distX, distY, distZ);
-
- real bladeVelX_GF = trilinearInterpolation(dW, dE, dN, dS, dT, dB, k, ke, kn, kt, kne, kte, ktn, ktne, vx)*velocityRatio;
- real bladeVelY_GF = trilinearInterpolation(dW, dE, dN, dS, dT, dB, k, ke, kn, kt, kne, kte, ktn, ktne, vy)*velocityRatio;
- real bladeVelZ_GF = trilinearInterpolation(dW, dE, dN, dS, dT, dB, k, ke, kn, kt, kne, kte, ktn, ktne, vz)*velocityRatio;
-
- real bladeVelX_BF, bladeVelY_BF, bladeVelZ_BF;
-
- rotateFromGlobalToBlade(bladeVelX_BF, bladeVelY_BF, bladeVelZ_BF,
- bladeVelX_GF, bladeVelY_GF, bladeVelZ_GF,
- localAzimuth, yaw);
-
- bladeVelocitiesX[node] = bladeVelX_BF;
- bladeVelocitiesY[node] = bladeVelY_BF+omega*bladeCoordZ_BF;
- bladeVelocitiesZ[node] = bladeVelZ_BF;
-}
-
-
-__global__ void applyBodyForces(real* gridCoordsX, real* gridCoordsY, real* gridCoordsZ,
- real* gridForcesX, real* gridForcesY, real* gridForcesZ,
- real* bladeCoordsX, real* bladeCoordsY, real* bladeCoordsZ,
- real* bladeForcesX, real* bladeForcesY,real* bladeForcesZ,
- uint nBlades, uint nBladeNodes,
- real azimuth, real yaw, real omega,
- real turbPosX, real turbPosY, real turbPosZ,
- uint* gridIndices, uint nIndices,
- real invEpsilonSqrd, real factorGaussian)
-{
- const uint x = threadIdx.x;
- const uint y = blockIdx.x;
- const uint z = blockIdx.y;
-
- const uint nx = blockDim.x;
- const uint ny = gridDim.x;
-
- const uint index = nx*(ny*z + y) + x;
-
- if(index>=nIndices) return;
-
- uint gridIndex = gridIndices[index];
-
- real gridCoordX_RF = gridCoordsX[gridIndex] - turbPosX;
- real gridCoordY_RF = gridCoordsY[gridIndex] - turbPosY;
- real gridCoordZ_RF = gridCoordsZ[gridIndex] - turbPosZ;
-
- real gridForceX_RF = c0o1;
- real gridForceY_RF = c0o1;
- real gridForceZ_RF = c0o1;
-
- real dAzimuth = c2Pi/nBlades;
-
- for( uint blade=0; blade<nBlades; blade++)
- {
- real localAzimuth = azimuth+blade*dAzimuth;
-
- real gridCoordX_BF, gridCoordY_BF, gridCoordZ_BF;
-
- rotateFromGlobalToBlade(gridCoordX_BF, gridCoordY_BF, gridCoordZ_BF,
- gridCoordX_RF, gridCoordY_RF, gridCoordZ_RF,
- localAzimuth, yaw);
-
- for( uint bladeNode=0; bladeNode<nBladeNodes; bladeNode++)
- {
- uint node = calcNode(bladeNode, nBladeNodes, blade, nBlades);
-
- real eta = factorGaussian*exp(-distSqrd(bladeCoordsX[node]-gridCoordX_BF, bladeCoordsY[node]-gridCoordY_BF, bladeCoordsZ[node]-gridCoordZ_BF)*invEpsilonSqrd);
-
- real forceX_RF, forceY_RF, forceZ_RF;
-
- rotateFromBladeToGlobal(bladeForcesX[node], bladeForcesY[node], bladeForcesZ[node],
- forceX_RF, forceY_RF, forceZ_RF,
- localAzimuth, yaw);
-
- gridForceX_RF += forceX_RF*eta;
- gridForceY_RF += forceY_RF*eta;
- gridForceZ_RF += forceZ_RF*eta;
- }
- }
-
- atomicAdd(&gridForcesX[gridIndex], gridForceX_RF);
- atomicAdd(&gridForcesY[gridIndex], gridForceY_RF);
- atomicAdd(&gridForcesZ[gridIndex], gridForceZ_RF);
-}
-
-
-void ActuatorLine::init(Parameter* para, GridProvider* gridProvider, CudaMemoryManager* cudaMemoryManager)
-{
- if(!para->getIsBodyForce()) throw std::runtime_error("try to allocate ActuatorLine but BodyForce is not set in Parameter.");
- this->initBladeRadii(cudaMemoryManager);
- this->initBladeCoords(cudaMemoryManager);
- this->initBladeIndices(para, cudaMemoryManager);
- this->initBladeVelocities(cudaMemoryManager);
- this->initBladeForces(cudaMemoryManager);
- this->initBoundingSphere(para, cudaMemoryManager);
-}
-
-
-void ActuatorLine::interact(Parameter* para, CudaMemoryManager* cudaMemoryManager, int level, unsigned int t)
-{
- if (level != this->level) return;
-
- cudaMemoryManager->cudaCopyBladeCoordsHtoD(this);
-
- vf::cuda::CudaGrid bladeGrid = vf::cuda::CudaGrid(para->getParH(level)->numberofthreads, this->nNodes);
-
- interpolateVelocities<<< bladeGrid.grid, bladeGrid.threads >>>(
- para->getParD(this->level)->coordinateX, para->getParD(this->level)->coordinateY, para->getParD(this->level)->coordinateZ,
- para->getParD(this->level)->neighborX, para->getParD(this->level)->neighborY, para->getParD(this->level)->neighborZ, para->getParD(this->level)->neighborInverse,
- para->getParD(this->level)->velocityX, para->getParD(this->level)->velocityY, para->getParD(this->level)->velocityZ,
- this->bladeCoordsXD, this->bladeCoordsYD, this->bladeCoordsZD,
- this->bladeVelocitiesXD, this->bladeVelocitiesYD, this->bladeVelocitiesZD,
- this->nBlades, this->nBladeNodes,
- this->azimuth, this->yaw, this->omega,
- this->turbinePosX, this->turbinePosY, this->turbinePosZ,
- this->bladeIndicesD, para->getVelocityRatio(), this->invDeltaX);
-
- cudaMemoryManager->cudaCopyBladeVelocitiesDtoH(this);
-
- this->calcBladeForces();
-
- cudaMemoryManager->cudaCopyBladeForcesHtoD(this);
-
- vf::cuda::CudaGrid sphereGrid = vf::cuda::CudaGrid(para->getParH(level)->numberofthreads, this->nIndices);
-
- applyBodyForces<<<sphereGrid.grid, sphereGrid.threads>>>(
- para->getParD(this->level)->coordinateX, para->getParD(this->level)->coordinateY, para->getParD(this->level)->coordinateZ,
- para->getParD(this->level)->forceX_SP, para->getParD(this->level)->forceY_SP, para->getParD(this->level)->forceZ_SP,
- this->bladeCoordsXD, this->bladeCoordsYD, this->bladeCoordsZD,
- this->bladeForcesXD, this->bladeForcesYD, this->bladeForcesZD,
- this->nBlades, this->nBladeNodes,
- this->azimuth, this->yaw, this->omega,
- this->turbinePosX, this->turbinePosY, this->turbinePosZ,
- this->boundingSphereIndicesD, this->nIndices,
- this->invEpsilonSqrd, this->factorGaussian);
-
- this->azimuth = fmod(this->azimuth+this->omega*this->deltaT,c2Pi);
-}
-
-
-void ActuatorLine::free(Parameter* para, CudaMemoryManager* cudaMemoryManager)
-{
- cudaMemoryManager->cudaFreeBladeRadii(this);
- cudaMemoryManager->cudaFreeBladeCoords(this);
- cudaMemoryManager->cudaFreeBladeVelocities(this);
- cudaMemoryManager->cudaFreeBladeForces(this);
- cudaMemoryManager->cudaFreeBladeIndices(this);
- cudaMemoryManager->cudaFreeSphereIndices(this);
-}
-
-
-void ActuatorLine::calcForcesEllipticWing()
-{
- uint node;
- real u_rel, v_rel, u_rel_sq;
- real phi;
- real Cl = c1o1;
- real Cd = c0o1;
- real c0 = c1o1;
-
- real c, Cn, Ct;
-
- for( uint blade=0; blade<this->nBlades; blade++)
- {
- for( uint bladeNode=0; bladeNode<this->nBladeNodes; bladeNode++)
- {
- node = calcNode(bladeNode, this->nBladeNodes, blade, this->nBlades);
-
- u_rel = this->bladeVelocitiesXH[node];
- v_rel = this->bladeVelocitiesYH[node];
- u_rel_sq = u_rel*u_rel+v_rel*v_rel;
- phi = atan2(u_rel, v_rel);
-
- real tmp = c4o1*this->bladeRadiiH[bladeNode]/this->diameter-c1o1;
- c = c0 * sqrt( c1o1- tmp*tmp );
- Cn = Cl*cos(phi)+Cd*sin(phi);
- Ct = Cl*sin(phi)-Cd*cos(phi);
-
- this->bladeForcesXH[node] = -c1o2*u_rel_sq*c*this->density*Cn;
- this->bladeForcesYH[node] = -c1o2*u_rel_sq*c*this->density*Ct;
- this->bladeForcesZH[node] = c0o1;
- }
- }
-}
-
-void ActuatorLine::calcBladeForces()
-{
- this->calcForcesEllipticWing();
-}
-
-void ActuatorLine::initBladeRadii(CudaMemoryManager* cudaMemoryManager)
-{
- cudaMemoryManager->cudaAllocBladeRadii(this);
-
- real dr = c1o2*this->diameter/this->nBladeNodes;
-
- for(uint node=0; node<this->nBladeNodes; node++)
- {
- this->bladeRadiiH[node] = dr*(node+1);
- }
- cudaMemoryManager->cudaCopyBladeRadiiHtoD(this);
-
- real dxOPiSqrtEps = pow(this->deltaX/(this->epsilon*sqrt(cPi)),c3o1);
- this->factorGaussian = dr*dxOPiSqrtEps/this->forceRatio;
-}
-
-void ActuatorLine::initBladeCoords(CudaMemoryManager* cudaMemoryManager)
-{
- cudaMemoryManager->cudaAllocBladeCoords(this);
-
- for(uint blade=0; blade<this->nBlades; blade++)
- {
- for(uint bladeNode=0; bladeNode<this->nBladeNodes; bladeNode++)
- {
- uint node = calcNode(bladeNode, this->nBladeNodes, blade, this->nBlades);
-
- this->bladeCoordsXH[node] = c0o1;
- this->bladeCoordsYH[node] = c0o1;
- this->bladeCoordsZH[node] = this->bladeRadiiH[bladeNode];
- }
- }
- cudaMemoryManager->cudaCopyBladeCoordsHtoD(this);
-}
-
-void ActuatorLine::initBladeVelocities(CudaMemoryManager* cudaMemoryManager)
-{
- cudaMemoryManager->cudaAllocBladeVelocities(this);
-
- for(uint node=0; node<this->nNodes; node++)
- {
- this->bladeVelocitiesXH[node] = c0o1;
- this->bladeVelocitiesYH[node] = c0o1;
- this->bladeVelocitiesZH[node] = c0o1;
- }
- cudaMemoryManager->cudaCopyBladeVelocitiesHtoD(this);
-}
-
-void ActuatorLine::initBladeForces(CudaMemoryManager* cudaMemoryManager)
-{
- cudaMemoryManager->cudaAllocBladeForces(this);
-
- for(uint node=0; node<this->nNodes; node++)
- {
- this->bladeForcesXH[node] = c0o1;
- this->bladeForcesYH[node] = c0o1;
- this->bladeForcesZH[node] = c0o1;
- }
- cudaMemoryManager->cudaCopyBladeForcesHtoD(this);
-}
-
-void ActuatorLine::initBladeIndices(Parameter* para, CudaMemoryManager* cudaMemoryManager)
-{
- cudaMemoryManager->cudaAllocBladeIndices(this);
-
- for(uint node=0; node<this->nNodes; node++)
-
- {
- this->bladeIndicesH[node] = 1;
- }
- cudaMemoryManager->cudaCopyBladeIndicesHtoD(this);
-}
-
-void ActuatorLine::initBoundingSphere(Parameter* para, CudaMemoryManager* cudaMemoryManager)
-{
- // Actuator line exists only on 1 level
- std::vector<int> nodesInSphere;
- real sphereRadius = c1o2*this->diameter+c4o1*this->epsilon;
- real sphereRadiusSqrd = sphereRadius*sphereRadius;
-
- for (uint j = 1; j <= para->getParH(this->level)->numberOfNodes; j++)
- {
- const real distX = para->getParH(this->level)->coordinateX[j]-this->turbinePosX;
- const real distY = para->getParH(this->level)->coordinateY[j]-this->turbinePosY;
- const real distZ = para->getParH(this->level)->coordinateZ[j]-this->turbinePosZ;
- if(distSqrd(distX,distY,distZ) < sphereRadiusSqrd) nodesInSphere.push_back(j);
- }
-
- this->nIndices = uint(nodesInSphere.size());
- cudaMemoryManager->cudaAllocSphereIndices(this);
- std::copy(nodesInSphere.begin(), nodesInSphere.end(), this->boundingSphereIndicesH);
- cudaMemoryManager->cudaCopySphereIndicesHtoD(this);
-}
-
-void ActuatorLine::setBladeCoords(real* _bladeCoordsX, real* _bladeCoordsY, real* _bladeCoordsZ)
-{
-
- for(uint node=0; node<this->nNodes; node++)
- {
- this->bladeCoordsXH[node] = _bladeCoordsX[node];
- this->bladeCoordsYH[node] = _bladeCoordsY[node];
- this->bladeCoordsZH[node] = _bladeCoordsZ[node];
- }
-}
-
-void ActuatorLine::setBladeVelocities(real* _bladeVelocitiesX, real* _bladeVelocitiesY, real* _bladeVelocitiesZ)
-{
- for(uint node=0; node<this->nNodes; node++)
- {
- this->bladeVelocitiesXH[node] = _bladeVelocitiesX[node];
- this->bladeVelocitiesYH[node] = _bladeVelocitiesY[node];
- this->bladeVelocitiesZH[node] = _bladeVelocitiesZ[node];
- }
-}
-
-void ActuatorLine::setBladeForces(real* _bladeForcesX, real* _bladeForcesY, real* _bladeForcesZ)
-{
- for(uint node=0; node<this->nNodes; node++)
- {
- this->bladeForcesXH[node] = _bladeForcesX[node];
- this->bladeForcesYH[node] = _bladeForcesY[node];
- this->bladeForcesZH[node] = _bladeForcesZ[node];
- }
-}
\ No newline at end of file
diff --git a/src/gpu/VirtualFluids_GPU/PreCollisionInteractor/ActuatorLine.h b/src/gpu/VirtualFluids_GPU/PreCollisionInteractor/ActuatorLine.h
deleted file mode 100644
index b44c89c5020eb206baa3bba1994b1e45f760c3bb..0000000000000000000000000000000000000000
--- a/src/gpu/VirtualFluids_GPU/PreCollisionInteractor/ActuatorLine.h
+++ /dev/null
@@ -1,120 +0,0 @@
-#ifndef ActuatorLine_H
-#define ActuatorLine_H
-
-#include "PreCollisionInteractor.h"
-#include "PointerDefinitions.h"
-#include "VirtualFluids_GPU_export.h"
-#include "lbm/constants/NumericConstants.h"
-
-class Parameter;
-class GridProvider;
-
-using namespace vf::lbm::constant;
-class VIRTUALFLUIDS_GPU_EXPORT ActuatorLine : public PreCollisionInteractor
-{
-public:
- ActuatorLine(
- const uint _nBlades,
- const real _density,
- const uint _nBladeNodes,
- const real _epsilon,
- real _turbinePosX, real _turbinePosY, real _turbinePosZ,
- const real _diameter,
- int _level,
- const real _deltaT,
- const real _deltaX
- ) : nBlades(_nBlades),
- density(_density),
- nBladeNodes(_nBladeNodes),
- epsilon(_epsilon),
- turbinePosX(_turbinePosX), turbinePosY(_turbinePosY), turbinePosZ(_turbinePosZ),
- diameter(_diameter),
- level(_level),
- PreCollisionInteractor()
- {
- this->deltaT = _deltaT*exp2(-this->level);
- this->deltaX = _deltaX*exp2(-this->level);
- this->invDeltaX = c1o1/this->deltaX;
- this->forceRatio = this->density*pow(this->deltaX,4)*pow(this->deltaT,-2);
- this->invEpsilonSqrd = c1o1/(this->epsilon*this->epsilon);
- this->nNodes = this->nBladeNodes*this->nBlades;
- this->omega = c1o1;
- this->azimuth = c0o1;
- this->yaw = c0o1;
- };
-
- virtual ~ActuatorLine(){};
-
- void init(Parameter* para, GridProvider* gridProvider, CudaMemoryManager* cudaMemoryManager) override;
- void interact(Parameter* para, CudaMemoryManager* cudaMemoryManager, int level, uint t) override;
- void free(Parameter* para, CudaMemoryManager* cudaMemoryManager) override;
- void write(uint t);
-
- uint getNBladeNodes(){ return this->nBladeNodes; };
- uint getNBlades(){ return this->nBlades;};
- uint getNIndices(){ return this->nIndices; };
- uint getNNodes(){ return this->nNodes; };
- real getOmega(){ return this->omega; };
- real getAzimuth(){ return this->azimuth; };
- real getYaw(){ return this->yaw; };
- real getDensity(){ return this->density; };
- real getPositionX(){ return this->turbinePosX; };
- real getPositionY(){ return this->turbinePosY; };
- real getPositionZ(){ return this->turbinePosZ; };
- real* getBladeRadii(){ return this->bladeRadiiH; };
- real* getBladeCoordsX(){ return this->bladeCoordsXH; };
- real* getBladeCoordsY(){ return this->bladeCoordsYH; };
- real* getBladeCoordsZ(){ return this->bladeCoordsZH; };
- real* getBladeVelocitiesX(){ return this->bladeVelocitiesXH; };
- real* getBladeVelocitiesY(){ return this->bladeVelocitiesYH; };
- real* getBladeVelocitiesZ(){ return this->bladeVelocitiesZH; };
- real* getBladeForcesX(){ return this->bladeForcesXH; };
- real* getBladeForcesY(){ return this->bladeForcesYH; };
- real* getBladeForcesZ(){ return this->bladeForcesZH; };
-
- void setOmega(real _omega){ this->omega = _omega; };
- void setAzimuth(real _azimuth){ this->azimuth = _azimuth; };
- void setYaw(real _yaw){ this->yaw = _yaw; };
- void setBladeCoords(real* _bladeCoordsX, real* _bladeCoordsY, real* _bladeCoordsZ);
- void setBladeVelocities(real* _bladeVelocitiesX, real* _bladeVelocitiesY, real* _bladeVelocitiesZ);
- void setBladeForces(real* _bladeForcesX, real* _bladeForcesY, real* _bladeForcesZ);
- virtual void calcBladeForces();
-
-private:
- void initBoundingSphere(Parameter* para, CudaMemoryManager* cudaMemoryManager);
-
- void initBladeRadii(CudaMemoryManager* cudaMemoryManager);
- void initBladeCoords(CudaMemoryManager* cudaMemoryManager);
- void initBladeVelocities(CudaMemoryManager* cudaMemoryManager);
- void initBladeForces(CudaMemoryManager* cudaMemoryManager);
- void initBladeIndices(Parameter* para, CudaMemoryManager* cudaMemoryManager);
-
- void calcForcesEllipticWing();
-
-public:
- real* bladeRadiiH;
- real* bladeRadiiD;
- real* bladeCoordsXH, * bladeCoordsYH, * bladeCoordsZH;
- real* bladeCoordsXD, * bladeCoordsYD, * bladeCoordsZD;
- real* bladeVelocitiesXH, * bladeVelocitiesYH, * bladeVelocitiesZH;
- real* bladeVelocitiesXD, * bladeVelocitiesYD, * bladeVelocitiesZD;
- real* bladeForcesXH, * bladeForcesYH, * bladeForcesZH;
- real* bladeForcesXD, * bladeForcesYD, * bladeForcesZD;
- uint* bladeIndicesH;
- uint* bladeIndicesD;
- uint* boundingSphereIndicesH;
- uint* boundingSphereIndicesD;
-
-private:
- const real density;
- real turbinePosX, turbinePosY, turbinePosZ;
- real omega, azimuth, yaw, deltaT, deltaX, invDeltaX, forceRatio, factorGaussian, invEpsilonSqrd;
- const real diameter;
- const uint nBladeNodes;
- const uint nBlades;
- const real epsilon; // in m
- const int level;
- uint nIndices, nNodes;
-};
-
-#endif
\ No newline at end of file
diff --git a/src/gpu/VirtualFluids_GPU/PreCollisionInteractor/PreCollisionInteractor.h b/src/gpu/VirtualFluids_GPU/PreCollisionInteractor/PreCollisionInteractor.h
index a9b233f3035890c2617d3a00b639f995be6c218f..f9a87f613e7607301e59a7c1e67eb556418892e4 100644
--- a/src/gpu/VirtualFluids_GPU/PreCollisionInteractor/PreCollisionInteractor.h
+++ b/src/gpu/VirtualFluids_GPU/PreCollisionInteractor/PreCollisionInteractor.h
@@ -33,6 +33,7 @@ public:
virtual void init(Parameter *para, GridProvider *gridProvider, CudaMemoryManager *cudaMemoryManager) = 0;
virtual void interact(Parameter *para, CudaMemoryManager *cudaMemoryManager, int level, uint t) = 0;
virtual void free(Parameter *para, CudaMemoryManager *cudaMemoryManager) = 0;
+ virtual void getTaggedFluidNodes(Parameter *para, GridProvider* gridProvider) = 0;
protected:
uint updateInterval;
diff --git a/src/gpu/VirtualFluids_GPU/PreCollisionInteractor/PrecursorWriter.cu b/src/gpu/VirtualFluids_GPU/PreCollisionInteractor/PrecursorWriter.cu
new file mode 100644
index 0000000000000000000000000000000000000000..1a8260ef936e2707fb38fbbba71cdbfac692f350
--- /dev/null
+++ b/src/gpu/VirtualFluids_GPU/PreCollisionInteractor/PrecursorWriter.cu
@@ -0,0 +1,359 @@
+//=======================================================================================
+// ____ ____ __ ______ __________ __ __ __ __
+// \ \ | | | | | _ \ |___ ___| | | | | / \ | |
+// \ \ | | | | | |_) | | | | | | | / \ | |
+// \ \ | | | | | _ / | | | | | | / /\ \ | |
+// \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
+// \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
+// \ \ | | ________________________________________________________________
+// \ \ | | | ______________________________________________________________|
+// \ \| | | | __ __ __ __ ______ _______
+// \ | | |_____ | | | | | | | | | _ \ / _____)
+// \ | | _____| | | | | | | | | | | \ \ \_______
+// \ | | | | |_____ | \_/ | | | | |_/ / _____ |
+// \ _____| |__| |________| \_______/ |__| |______/ (_______/
+//
+// This file is part of VirtualFluids. VirtualFluids is free software: you can
+// redistribute it and/or modify it under the terms of the GNU General Public
+// License as published by the Free Software Foundation, either version 3 of
+// the License, or (at your option) any later version.
+//
+// VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
+// ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+// FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
+// for more details.
+//
+// You should have received a copy of the GNU General Public License along
+// with VirtualFluids (see COPYING.txt). If not, see <http://www.gnu.org/licenses/>.
+//
+//! \file PrecursorWriter.cu
+//! \ingroup PreCollisionInteractor
+//! \author Henrik Asmuth, Henry Korb
+//======================================================================================
+#include "PrecursorWriter.h"
+#include "basics/writer/WbWriterVtkXmlImageBinary.h"
+
+#include <cuda.h>
+#include <cuda_runtime.h>
+#include <helper_cuda.h>
+#include "cuda/CudaGrid.h"
+#include "LBM/GPUHelperFunctions/KernelUtilities.h"
+
+#include "Core/StringUtilities/StringUtil.h"
+
+#include "Parameter/Parameter.h"
+#include "DataStructureInitializer/GridProvider.h"
+#include "GPU/CudaMemoryManager.h"
+
+using namespace vf::lbm::dir;
+using namespace vf::gpu;
+
+////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
+//TODO check everything for multiple level
+void index1d(int& idx, int y, int z, int ny, int nz)
+{
+ idx = y+ny*z;
+}
+
+void index2d(int idx, int& y, int& z, int ny, int nz)
+{
+ z = idx/ny;
+ y = idx-ny*z;
+}
+
+__inline__ __host__ __device__ uint linearIdx(const uint component, const uint node, const uint timestep, const uint numberOfComponents, const uint numberOfNodes)
+{
+ return node+numberOfNodes*(component+numberOfComponents*timestep);
+}
+
+__inline__ __host__ __device__ uint linearIdx(const uint component, const uint node, const uint numberOfNodes)
+{
+ return node+component*numberOfNodes;
+}
+
+
+////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
+__global__ void fillArrayVelocities(const uint numberOfPrecursorNodes,
+ uint* indices,
+ real *precursorData,
+ real *vx,
+ real *vy,
+ real *vz,
+ real velocityRatio)
+
+
+{
+ ////////////////////////////////////////////////////////////////////////////////
+ //! - Get node index coordinates from threadIdx, blockIdx, blockDim and gridDim.
+ //!
+ const unsigned nodeIndex = vf::gpu::getNodeIndex();
+
+ if(nodeIndex>=numberOfPrecursorNodes) return;
+
+ precursorData[linearIdx(0u, nodeIndex, numberOfPrecursorNodes)] = vx[indices[nodeIndex]]*velocityRatio;
+ precursorData[linearIdx(1u, nodeIndex, numberOfPrecursorNodes)] = vy[indices[nodeIndex]]*velocityRatio;
+ precursorData[linearIdx(2u, nodeIndex, numberOfPrecursorNodes)] = vz[indices[nodeIndex]]*velocityRatio;
+}
+
+
+////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
+__global__ void fillArrayDistributions( uint numberOfPrecursorNodes,
+ uint* indices,
+ real* precursorData,
+ real* distributions,
+ uint* neighborX, uint* neighborY, uint* neighborZ,
+ bool isEvenTimestep,
+ unsigned long numberOfLBnodes)
+{
+ ////////////////////////////////////////////////////////////////////////////////
+ //! - Get node index coordinates from threadIdx, blockIdx, blockDim and gridDim.
+ //!
+ const unsigned nodeIndex = vf::gpu::getNodeIndex();
+
+ if(nodeIndex>=numberOfPrecursorNodes) return;
+
+ Distributions27 dist;
+ getPointersToDistributions(dist, distributions, numberOfLBnodes, isEvenTimestep);
+
+ ////////////////////////////////////////////////////////////////////////////////
+ // ! - Set neighbor indices (necessary for indirect addressing)
+ uint k_000 = indices[nodeIndex];
+ // uint k_M00 = neighborX[k_000];
+ uint k_0M0 = neighborY[k_000];
+ uint k_00M = neighborZ[k_000];
+ // uint k_MM0 = neighborY[k_M00];
+ // uint k_M0M = neighborZ[k_M00];
+ uint k_0MM = neighborZ[k_0M0];
+ // uint k_MMM = neighborZ[k_MM0];
+
+ ////////////////////////////////////////////////////////////////////////////////////
+ //! - Get local distributions in PX directions
+ //!
+ precursorData[linearIdx(PrecP00, nodeIndex, numberOfPrecursorNodes)] = (dist.f[DIR_P00])[k_000];
+ precursorData[linearIdx(PrecPP0, nodeIndex, numberOfPrecursorNodes)] = (dist.f[DIR_PP0])[k_000];
+ precursorData[linearIdx(PrecPM0, nodeIndex, numberOfPrecursorNodes)] = (dist.f[DIR_PM0])[k_0M0];
+ precursorData[linearIdx(PrecP0P, nodeIndex, numberOfPrecursorNodes)] = (dist.f[DIR_P0P])[k_000];
+ precursorData[linearIdx(PrecP0M, nodeIndex, numberOfPrecursorNodes)] = (dist.f[DIR_P0M])[k_00M];
+ precursorData[linearIdx(PrecPPP, nodeIndex, numberOfPrecursorNodes)] = (dist.f[DIR_PPP])[k_000];
+ precursorData[linearIdx(PrecPMP, nodeIndex, numberOfPrecursorNodes)] = (dist.f[DIR_PMP])[k_0M0];
+ precursorData[linearIdx(PrecPPM, nodeIndex, numberOfPrecursorNodes)] = (dist.f[DIR_PPM])[k_00M];
+ precursorData[linearIdx(PrecPMM, nodeIndex, numberOfPrecursorNodes)] = (dist.f[DIR_PMM])[k_0MM];
+}
+
+
+////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
+void PrecursorWriter::init(Parameter* para, GridProvider* gridProvider, CudaMemoryManager* cudaManager)
+{
+ VF_LOG_INFO("PrecursorWriter: Start initializing...");
+ VF_LOG_INFO("Writing yz-planes at x={}m every {}. timestep, starting at t={}", this->xPos, this->tSave, this->tStartOut);
+
+ precursorStructs.resize(para->getMaxLevel()+1);
+ for(int level=0; level<=para->getMaxLevel(); level++)
+ {
+
+ real dx = abs(para->getParH(level)->coordinateX[1]-para->getParH(level)->coordinateX[para->getParH(level)->neighborX[1]]);
+ int maxPoints = (int((yMax-yMin)/dx)+1)* (int((zMax-zMin)/dx)+1);
+
+ real lowestY, lowestZ, highestY, highestZ;
+
+ lowestY = para->getParH(level)->coordinateY[para->getParH(level)->numberOfNodes-1];
+ highestY = para->getParH(level)->coordinateY[1];
+
+ lowestZ = para->getParH(level)->coordinateZ[para->getParH(level)->numberOfNodes-1];
+ highestZ = para->getParH(level)->coordinateZ[1];
+
+ std::vector<uint> indicesOnGrid;
+ std::vector<int> indicesOnPlane;
+ std::vector<real> coordY, coordZ;
+
+ for(size_t pos = 1; pos < para->getParH(level)->numberOfNodes; pos++ )
+ {
+ real pointCoordX = para->getParH(level)->coordinateX[pos];
+ real pointCoordY = para->getParH(level)->coordinateY[pos];
+ real pointCoordZ = para->getParH(level)->coordinateZ[pos];
+ if( para->getParH(level)->typeOfGridNode[pos] == GEO_FLUID &&
+ pointCoordX < (dx+xPos) && pointCoordX >= xPos &&
+ pointCoordY<=yMax && pointCoordY>=yMin &&
+ pointCoordZ<=zMax && pointCoordZ>=zMin)
+ {
+ highestY = max(highestY, pointCoordY);
+ highestZ = max(highestZ, pointCoordZ);
+
+ lowestY = min(lowestY, pointCoordY);
+ lowestZ = min(lowestZ, pointCoordZ);
+ indicesOnGrid.push_back((uint)pos);
+ coordY.push_back(pointCoordY);
+ coordZ.push_back(pointCoordZ);
+ }
+ }
+ if(indicesOnGrid.size()==0)
+ throw std::runtime_error("PrecursorWriter did not find any points on the grid");
+
+ int ny = int((highestY-lowestY)/dx)+1;
+ int nz = int((highestZ-lowestZ)/dx)+1;
+
+ for(uint i=0;i<indicesOnGrid.size(); i++)
+ {
+ int idxY = int((coordY[i]-lowestY)/dx);
+ int idxZ = int((coordZ[i]-lowestZ)/dx);
+ int idx;
+ index1d(idx, idxY, idxZ, ny, nz);
+ indicesOnPlane.push_back(idx);
+ }
+
+ precursorStructs[level] = SPtr<PrecursorStruct>(new PrecursorStruct);
+ precursorStructs[level]->numberOfPointsInBC = (uint)indicesOnGrid.size();
+ precursorStructs[level]->indicesOnPlane = (int*) malloc(precursorStructs[level]->numberOfPointsInBC*sizeof(int));
+ precursorStructs[level]->spacing = makeUbTuple(dx, dx, tSave*para->getTimeRatio()*pow(2,-level));
+ precursorStructs[level]->origin = makeUbTuple(lowestY, lowestZ);
+ precursorStructs[level]->extent = makeUbTuple(0, ny-1, 0, nz-1);
+ precursorStructs[level]->numberOfPointsInData = ny*nz;
+ precursorStructs[level]->numberOfTimestepsPerFile = min(para->getlimitOfNodesForVTK()/(ny*nz), maxtimestepsPerFile);
+ precursorStructs[level]->numberOfFilesWritten = 0;
+ precursorStructs[level]->numberOfTimestepsBuffered = 0;
+
+ switch (outputVariable)
+ {
+ case OutputVariable::Velocities:
+ precursorStructs[level]->numberOfQuantities = 3;
+ break;
+ case OutputVariable::Distributions:
+ precursorStructs[level]->numberOfQuantities = 9;
+ break;
+
+ default:
+ break;
+ }
+
+ cudaManager->cudaAllocPrecursorWriter(this, level);
+
+ std::copy(indicesOnGrid.begin(), indicesOnGrid.end(), precursorStructs[level]->indicesH);
+ std::copy(indicesOnPlane.begin(), indicesOnPlane.end(), precursorStructs[level]->indicesOnPlane);
+
+ cudaManager->cudaCopyPrecursorWriterIndicesHtoD(this, level);
+
+ VF_LOG_INFO("Found {} points in precursor plane on level {}", precursorStructs[level]->numberOfPointsInBC, level);
+ }
+ VF_LOG_INFO("PrecursorWriter: Done initializing.");
+}
+
+
+void PrecursorWriter::interact(Parameter* para, CudaMemoryManager* cudaManager, int level, uint t)
+{
+ uint t_level = para->getTimeStep(level, t, true);
+ uint tStartOut_level = tStartOut*pow(2, level);
+ uint tEnd_level = para->getTimestepEnd()*pow(2, level);
+
+ if(t_level>tStartOut_level && ((t_level-tStartOut_level) % tSave)==0)
+ {
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(para->getParH(level)->numberofthreads, precursorStructs[level]->numberOfPointsInBC);
+
+ if(this->outputVariable==OutputVariable::Velocities)
+ {
+ fillArrayVelocities<<<grid.grid, grid.threads>>>( precursorStructs[level]->numberOfPointsInBC, precursorStructs[level]->indicesD,
+ precursorStructs[level]->bufferD,
+ para->getParD(level)->velocityX, para->getParD(level)->velocityY, para->getParD(level)->velocityZ,
+ para->getVelocityRatio());
+ getLastCudaError("In PrecursorWriter::interact fillArrayVelocities execution failed");
+ }
+ else if(this->outputVariable==OutputVariable::Distributions)
+ {
+ fillArrayDistributions<<<grid.grid, grid.threads>>>(precursorStructs[level]->numberOfPointsInBC, precursorStructs[level]->indicesD,
+ precursorStructs[level]->bufferD,
+ para->getParD(level)->distributions.f[0],
+ para->getParD(level)->neighborX, para->getParD(level)->neighborY, para->getParD(level)->neighborZ,
+ para->getEvenOrOdd(level), para->getParD(level)->numberOfNodes);
+ getLastCudaError("In PrecursorWriter::interact fillArrayDistributions execution failed");
+ }
+ cudaManager->cudaCopyPrecursorWriterOutputVariablesDtoH(this, level);
+
+ // switch device buffer and data pointer so precursor data is gathered in buffer and copied from bufferD to bufferH
+ real *tmp = precursorStructs[level]->bufferD;
+ precursorStructs[level]->bufferD = precursorStructs[level]->dataD;
+ precursorStructs[level]->dataD = tmp;
+
+ precursorStructs[level]->numberOfTimestepsBuffered++;
+
+ if(precursorStructs[level]->numberOfTimestepsBuffered >= precursorStructs[level]->numberOfTimestepsPerFile || t == para->getTimestepEnd())
+ {
+ // switch host buffer and data pointer so precursor data is copied in buffer and written from data
+
+ tmp = precursorStructs[level]->bufferH;
+ precursorStructs[level]->bufferH = precursorStructs[level]->dataH;
+ precursorStructs[level]->dataH = tmp;
+
+ writeFuture.wait();
+ writeFuture = std::async(std::launch::async, [this](Parameter* para, uint level, uint timesteps){ this->write(para, level, timesteps); }, para, level, precursorStructs[level]->numberOfTimestepsBuffered);
+ precursorStructs[level]->numberOfTimestepsBuffered = 0;
+ }
+ }
+}
+
+
+void PrecursorWriter::free(Parameter* para, CudaMemoryManager* cudaManager)
+{
+ writeFuture.wait();
+ for(int level=0; level<=para->getMaxLevel(); level++)
+ {
+ if(getPrecursorStruct(level)->numberOfTimestepsBuffered>0)
+ write(para, level, getPrecursorStruct(level)->numberOfTimestepsBuffered);
+
+ cudaManager->cudaFreePrecursorWriter(this, level);
+ }
+}
+
+
+void PrecursorWriter::write(Parameter* para, int level, uint numberOfTimestepsBuffered)
+{
+ std::string fname = this->makeFileName(fileName, level, para->getMyProcessID(), precursorStructs[level]->numberOfFilesWritten) + getWriter()->getFileExtension();
+ std::string wholeName = outputPath + "/" + fname;
+
+ uint numberOfPointsInData = precursorStructs[level]->numberOfPointsInData;
+
+ int startTime = precursorStructs[level]->numberOfFilesWritten*precursorStructs[level]->numberOfTimestepsPerFile;
+
+ UbTupleInt6 extent = makeUbTuple( val<1>(precursorStructs[level]->extent), val<2>(precursorStructs[level]->extent),
+ val<3>(precursorStructs[level]->extent), val<4>(precursorStructs[level]->extent),
+ startTime, startTime+(int)numberOfTimestepsBuffered-1);
+
+ UbTupleFloat3 origin = makeUbTuple( val<1>(precursorStructs[level]->origin), val<2>(precursorStructs[level]->origin), 0.f);
+
+ std::vector<std::vector<double>> nodedata;
+
+ for(uint quant=0; quant<precursorStructs[level]->numberOfQuantities; quant++)
+ {
+ std::vector<double> doubleArr(numberOfPointsInData*numberOfTimestepsBuffered, NAN);
+ for( uint timestep=0; timestep<numberOfTimestepsBuffered; timestep++)
+ {
+ for (uint pos=0; pos < precursorStructs[level]->numberOfPointsInBC; pos++)
+ {
+ int indexOnPlane = precursorStructs[level]->indicesOnPlane[pos]+timestep*numberOfPointsInData;
+ doubleArr[indexOnPlane] = double(precursorStructs[level]->dataH[linearIdx(quant, pos, timestep, precursorStructs[level]->numberOfQuantities, precursorStructs[level]->numberOfPointsInBC)]);
+ }
+ }
+ nodedata.push_back(doubleArr);
+ }
+
+ std::vector<std::vector<double>> celldata;
+ getWriter()->writeData(wholeName, nodedatanames, celldatanames, nodedata, celldata, extent, origin, precursorStructs[level]->spacing, extent, this->writePrecision);
+ precursorStructs[level]->numberOfFilesWritten++;
+}
+
+std::string PrecursorWriter::makeFileName(std::string fileName, int level, int id, uint numberOfFilesWritten)
+{
+ return fileName + "_lev_" + StringUtil::toString<int>(level)
+ + "_ID_" + StringUtil::toString<int>(id)
+ + "_File_" + StringUtil::toString<int>(numberOfFilesWritten);
+}
+
+void PrecursorWriter::getTaggedFluidNodes(Parameter *para, GridProvider* gridProvider)
+{
+ for(uint level=0; level<(uint)para->getMaxLevel(); level++)
+ {
+ if(outputVariable==OutputVariable::Velocities)
+ {
+ std::vector<uint> indices(precursorStructs[level]->indicesH, precursorStructs[level]->indicesH+precursorStructs[level]->numberOfPointsInBC);
+ gridProvider->tagFluidNodeIndices(indices, CollisionTemplate::WriteMacroVars, level);
+ }
+ }
+}
\ No newline at end of file
diff --git a/src/gpu/VirtualFluids_GPU/PreCollisionInteractor/PrecursorWriter.h b/src/gpu/VirtualFluids_GPU/PreCollisionInteractor/PrecursorWriter.h
new file mode 100644
index 0000000000000000000000000000000000000000..264023b58ba6db46b50f6a85b334c530864a0b8f
--- /dev/null
+++ b/src/gpu/VirtualFluids_GPU/PreCollisionInteractor/PrecursorWriter.h
@@ -0,0 +1,161 @@
+//=======================================================================================
+// ____ ____ __ ______ __________ __ __ __ __
+// \ \ | | | | | _ \ |___ ___| | | | | / \ | |
+// \ \ | | | | | |_) | | | | | | | / \ | |
+// \ \ | | | | | _ / | | | | | | / /\ \ | |
+// \ \ | | | | | | \ \ | | | \__/ | / ____ \ | |____
+// \ \ | | |__| |__| \__\ |__| \________/ /__/ \__\ |_______|
+// \ \ | | ________________________________________________________________
+// \ \ | | | ______________________________________________________________|
+// \ \| | | | __ __ __ __ ______ _______
+// \ | | |_____ | | | | | | | | | _ \ / _____)
+// \ | | _____| | | | | | | | | | | \ \ \_______
+// \ | | | | |_____ | \_/ | | | | |_/ / _____ |
+// \ _____| |__| |________| \_______/ |__| |______/ (_______/
+//
+// This file is part of VirtualFluids. VirtualFluids is free software: you can
+// redistribute it and/or modify it under the terms of the GNU General Public
+// License as published by the Free Software Foundation, either version 3 of
+// the License, or (at your option) any later version.
+//
+// VirtualFluids is distributed in the hope that it will be useful, but WITHOUT
+// ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+// FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
+// for more details.
+//
+// You should have received a copy of the GNU General Public License along
+// with VirtualFluids (see COPYING.txt). If not, see <http://www.gnu.org/licenses/>.
+//
+//! \file PrecursorWriter.h
+//! \author Henry Korb, Henrik Asmuth
+//! \date 05/12/2022
+//! \brief Probe writing planes of data to be used as inflow data in successor simulation using PrecursorBC
+//!
+//! The probe writes out yz-planes at a specific x position ( \param xPos ) of either velocity or distributions
+//! that can be read by PrecursorBC as inflow data.
+//=======================================================================================
+
+
+#ifndef PRECURSORPROBE_H_
+#define PRECURSORPROBE_H_
+
+#include "PreCollisionInteractor.h"
+#include "WbWriterVtkXmlImageBinary.h"
+#include "LBM/LB.h"
+#include <string>
+#include <vector>
+#include <future>
+#include "PointerDefinitions.h"
+#include "Logger.h"
+
+class Parameter;
+class CudaMemoryManager;
+class GridProvider;
+
+enum class OutputVariable {
+ //! - Velocities
+ Velocities,
+ //! - Distributions
+ Distributions
+};
+
+static constexpr uint PrecP00 = 0;
+static constexpr uint PrecPP0 = 1;
+static constexpr uint PrecPM0 = 2;
+static constexpr uint PrecP0P = 3;
+static constexpr uint PrecP0M = 4;
+static constexpr uint PrecPPP = 5;
+static constexpr uint PrecPMP = 6;
+static constexpr uint PrecPPM = 7;
+static constexpr uint PrecPMM = 8;
+
+struct PrecursorStruct
+{
+ uint numberOfPointsInBC, numberOfPointsInData, numberOfTimestepsPerFile, numberOfFilesWritten, numberOfTimestepsBuffered;
+ uint *indicesH, *indicesD;
+ real *dataH, *dataD;
+ real *bufferH, *bufferD;
+ uint numberOfQuantities;
+ UbTupleInt4 extent;
+ UbTupleFloat2 origin;
+ UbTupleFloat3 spacing;
+ int* indicesOnPlane;
+ cudaStream_t stream;
+};
+
+class PrecursorWriter : public PreCollisionInteractor
+{
+public:
+ PrecursorWriter(
+ const std::string _fileName,
+ const std::string _outputPath,
+ real _xPos,
+ real _yMin, real _yMax,
+ real _zMin, real _zMax,
+ uint _tStartOut,
+ uint _tSave,
+ OutputVariable _outputVariable,
+ uint _maxTimestepsPerFile=uint(1e4)
+ ):
+ fileName(_fileName),
+ outputPath(_outputPath),
+ xPos(_xPos),
+ yMin(_yMin),
+ yMax(_yMax),
+ zMin(_zMin),
+ zMax(_zMax),
+ tStartOut(_tStartOut),
+ tSave(_tSave),
+ outputVariable(_outputVariable),
+ maxtimestepsPerFile(_maxTimestepsPerFile)
+ {
+ nodedatanames = determineNodeDataNames();
+ writeFuture = std::async([](){});
+ };
+
+ void init(Parameter* para, GridProvider* gridProvider, CudaMemoryManager* cudaManager) override;
+ void interact(Parameter* para, CudaMemoryManager* cudaManager, int level, uint t) override;
+ void free(Parameter* para, CudaMemoryManager* cudaManager) override;
+ void getTaggedFluidNodes(Parameter *para, GridProvider* gridProvider) override;
+
+ OutputVariable getOutputVariable(){ return this->outputVariable; }
+
+ SPtr<PrecursorStruct> getPrecursorStruct(int level){return precursorStructs[level];}
+ static std::string makeFileName(std::string fileName, int level, int id, uint part);
+
+ void setWritePrecision(uint _writePrecision){ this->writePrecision=_writePrecision;}
+
+private:
+ WbWriterVtkXmlImageBinary* getWriter(){ return WbWriterVtkXmlImageBinary::getInstance(); };
+ void write(Parameter* para, int level, uint numberOfTimestepsBuffered);
+
+ std::vector<std::string> determineNodeDataNames()
+ {
+ switch (outputVariable)
+ {
+ case OutputVariable::Velocities:
+ return {"vx", "vy", "vz"};
+ break;
+ case OutputVariable::Distributions:
+ return {"fP00", "fPP0", "fPM0", "fP0P", "fP0M", "fPPP", "fPMP", "fPPM", "fPMM"};
+ break;
+
+ default:
+ throw std::runtime_error("Invalid OutputVariable for PrecursorWriter");
+ break;
+ }
+ }
+
+private:
+ std::vector<SPtr<PrecursorStruct>> precursorStructs;
+ std::string fileName, outputPath;
+ std::vector<std::string> nodedatanames;
+ std::vector<std::string> celldatanames;
+ uint tStartOut, tSave, maxtimestepsPerFile;
+ real xPos, yMin, yMax, zMin, zMax;
+ OutputVariable outputVariable;
+ std::future<void> writeFuture;
+ uint writePrecision = 8;
+};
+
+#endif //PRECURSORPROBE_H_
diff --git a/src/gpu/VirtualFluids_GPU/PreCollisionInteractor/Probes/PlanarAverageProbe.cu b/src/gpu/VirtualFluids_GPU/PreCollisionInteractor/Probes/PlanarAverageProbe.cu
index f5b520acfad74f6787e9e657fce3ccdceed9d539..e89d392b5d4bf5983f9bb47642fef81d0f06cc89 100644
--- a/src/gpu/VirtualFluids_GPU/PreCollisionInteractor/Probes/PlanarAverageProbe.cu
+++ b/src/gpu/VirtualFluids_GPU/PreCollisionInteractor/Probes/PlanarAverageProbe.cu
@@ -15,6 +15,7 @@
#include "Parameter/Parameter.h"
#include "DataStructureInitializer/GridProvider.h"
#include "GPU/CudaMemoryManager.h"
+#include "GPU/GPU_Interface.h"
#include <algorithm>
@@ -235,7 +236,7 @@ void PlanarAverageProbe::findPoints(Parameter* para, GridProvider* gridProvider,
}
// Find all points along the normal direction
- for(uint j=1; j<para->getParH(level)->numberOfNodes; j++ )
+ for(size_t j = 1; j < para->getParH(level)->numberOfNodes; j++ )
{
if(para->getParH(level)->typeOfGridNode[j] == GEO_FLUID)
{
@@ -250,16 +251,16 @@ void PlanarAverageProbe::findPoints(Parameter* para, GridProvider* gridProvider,
std::sort(pointCoordsNormal->begin(), pointCoordsNormal->end());
// Find all pointCoords in the first plane
- for(uint j=1; j<para->getParH(level)->numberOfNodes; j++ )
+ for(size_t pos = 1; pos < para->getParH(level)->numberOfNodes; pos++ )
{
- if( para->getParH(level)->typeOfGridNode[j] == GEO_FLUID && pointCoordsNormal_par[j] == pointCoordsNormal->at(0))
+ if( para->getParH(level)->typeOfGridNode[pos] == GEO_FLUID && pointCoordsNormal_par[pos] == pointCoordsNormal->at(0))
{
//not needed in current state, might become relevant for two-point correlations
// pointCoordsNormal->push_back( pointCoordsNormal_par[j] );
// pointCoordsInplane1->push_back( pointCoordsInplane1_par[j] );
// pointCoordsInplane2->push_back( pointCoordsInplane2_par[j] );
- probeIndices_level.push_back(j);
+ probeIndices_level.push_back((int)pos);
}
}
}
@@ -268,6 +269,23 @@ void PlanarAverageProbe::findPoints(Parameter* para, GridProvider* gridProvider,
void PlanarAverageProbe::calculateQuantities(SPtr<ProbeStruct> probeStruct, Parameter* para, uint t_level, int level)
{
+ // Compute macroscopic variables in entire domain
+ CalcMacCompSP27(
+ para->getParD(level)->velocityX,
+ para->getParD(level)->velocityY,
+ para->getParD(level)->velocityZ,
+ para->getParD(level)->rho,
+ para->getParD(level)->pressure,
+ para->getParD(level)->typeOfGridNode,
+ para->getParD(level)->neighborX,
+ para->getParD(level)->neighborY,
+ para->getParD(level)->neighborZ,
+ para->getParD(level)->numberOfNodes,
+ para->getParD(level)->numberofthreads,
+ para->getParD(level)->distributions.f[0],
+ para->getParD(level)->isEvenTimestep);
+ getLastCudaError("In PlanarAverageProbe Kernel CalcMacSP27 execution failed");
+
// Definition of normal and inplane directions for moveIndices kernels
uint *neighborNormal, *neighborInplane1, *neighborInplane2;
if( this->planeNormal == 'x' )
diff --git a/src/gpu/VirtualFluids_GPU/PreCollisionInteractor/Probes/PlanarAverageProbe.h b/src/gpu/VirtualFluids_GPU/PreCollisionInteractor/Probes/PlanarAverageProbe.h
index d11f8e76e4d13113b201af5494b7d0cfcfe18353..3d3533f74501e776f9150c83c9d9101a0be7ecbc 100644
--- a/src/gpu/VirtualFluids_GPU/PreCollisionInteractor/Probes/PlanarAverageProbe.h
+++ b/src/gpu/VirtualFluids_GPU/PreCollisionInteractor/Probes/PlanarAverageProbe.h
@@ -74,6 +74,7 @@ public:
planeNormal(_planeNormal)
{
+ if (_tStartTmpAvg<_tStartAvg) throw std::runtime_error("Probe: tStartTmpAvg must be larger than tStartAvg!");
if(!(_planeNormal == 'x' || _planeNormal == 'y' || _planeNormal == 'z'))
throw std::runtime_error("PlanarAverageProbe: planeNormal must be 'x', 'y' or 'z'!");
}
diff --git a/src/gpu/VirtualFluids_GPU/PreCollisionInteractor/Probes/PlaneProbe.cu b/src/gpu/VirtualFluids_GPU/PreCollisionInteractor/Probes/PlaneProbe.cu
index 7d1c0205219737e4b28acbb1a893a0a6071ae9de..f55045505bff0e3b5b0b1426be4e9e1a3832d088 100644
--- a/src/gpu/VirtualFluids_GPU/PreCollisionInteractor/Probes/PlaneProbe.cu
+++ b/src/gpu/VirtualFluids_GPU/PreCollisionInteractor/Probes/PlaneProbe.cu
@@ -76,11 +76,11 @@ void PlaneProbe::findPoints(Parameter* para, GridProvider* gridProvider, std::ve
int level)
{
real dx = abs(para->getParH(level)->coordinateX[1]-para->getParH(level)->coordinateX[para->getParH(level)->neighborX[1]]);
- for(uint j=1; j<para->getParH(level)->numberOfNodes; j++ )
+ for(size_t pos = 1; pos < para->getParH(level)->numberOfNodes; pos++ )
{
- real pointCoordX = para->getParH(level)->coordinateX[j];
- real pointCoordY = para->getParH(level)->coordinateY[j];
- real pointCoordZ = para->getParH(level)->coordinateZ[j];
+ real pointCoordX = para->getParH(level)->coordinateX[pos];
+ real pointCoordY = para->getParH(level)->coordinateY[pos];
+ real pointCoordZ = para->getParH(level)->coordinateZ[pos];
real distX = pointCoordX - this->posX;
real distY = pointCoordY - this->posY;
real distZ = pointCoordZ - this->posZ;
@@ -88,7 +88,7 @@ void PlaneProbe::findPoints(Parameter* para, GridProvider* gridProvider, std::ve
if( distX <= this->deltaX && distY <= this->deltaY && distZ <= this->deltaZ &&
distX >=0.f && distY >=0.f && distZ >=0.f)
{
- probeIndices_level.push_back(j);
+ probeIndices_level.push_back((int)pos);
distX_level.push_back( distX/dx );
distY_level.push_back( distY/dx );
distZ_level.push_back( distZ/dx );
@@ -106,4 +106,14 @@ void PlaneProbe::calculateQuantities(SPtr<ProbeStruct> probeStruct, Parameter* p
para->getParD(level)->velocityX, para->getParD(level)->velocityY, para->getParD(level)->velocityZ, para->getParD(level)->rho,
para->getParD(level)->neighborX, para->getParD(level)->neighborY, para->getParD(level)->neighborZ,
probeStruct->quantitiesD, probeStruct->arrayOffsetsD, probeStruct->quantitiesArrayD);
+}
+
+void PlaneProbe::getTaggedFluidNodes(Parameter *para, GridProvider* gridProvider)
+{
+ for(int level=0; level<=para->getMaxLevel(); level++)
+ {
+ SPtr<ProbeStruct> probeStruct = this->getProbeStruct(level);
+ std::vector<uint> probeIndices( probeStruct->pointIndicesH, probeStruct->pointIndicesH+probeStruct->nIndices);
+ gridProvider->tagFluidNodeIndices( probeIndices, CollisionTemplate::WriteMacroVars, level);
+ }
}
\ No newline at end of file
diff --git a/src/gpu/VirtualFluids_GPU/PreCollisionInteractor/Probes/PlaneProbe.h b/src/gpu/VirtualFluids_GPU/PreCollisionInteractor/Probes/PlaneProbe.h
index 3440c01020f9b3505be7148024e47373b76648ff..180169707a6d7f3f7975f6a2bc4009f7c0aba527 100644
--- a/src/gpu/VirtualFluids_GPU/PreCollisionInteractor/Probes/PlaneProbe.h
+++ b/src/gpu/VirtualFluids_GPU/PreCollisionInteractor/Probes/PlaneProbe.h
@@ -54,7 +54,7 @@ public:
): Probe(_probeName,
_outputPath,
_tStartAvg,
- 0,
+ _tStartAvg+1,
_tAvg,
_tStartOut,
_tOut,
@@ -72,6 +72,8 @@ public:
this->deltaZ = _deltaZ;
}
+ void getTaggedFluidNodes(Parameter *para, GridProvider* gridProvider) override;
+
private:
bool isAvailableStatistic(Statistic _variable) override;
diff --git a/src/gpu/VirtualFluids_GPU/PreCollisionInteractor/Probes/PointProbe.cu b/src/gpu/VirtualFluids_GPU/PreCollisionInteractor/Probes/PointProbe.cu
index e78a98f02ac2093fc46b4daa4a2485ed1395275b..89e1f6b87687ed42c079415a5340f1d385c8d62c 100644
--- a/src/gpu/VirtualFluids_GPU/PreCollisionInteractor/Probes/PointProbe.cu
+++ b/src/gpu/VirtualFluids_GPU/PreCollisionInteractor/Probes/PointProbe.cu
@@ -75,20 +75,20 @@ void PointProbe::findPoints(Parameter* para, GridProvider* gridProvider, std::ve
{
real dx = abs(para->getParH(level)->coordinateX[1]-para->getParH(level)->coordinateX[para->getParH(level)->neighborX[1]]);
- for(uint j=1; j<para->getParH(level)->numberOfNodes; j++ )
+ for(size_t pos = 1; pos < para->getParH(level)->numberOfNodes; pos++ )
{
for(uint point=0; point<this->pointCoordsX.size(); point++)
{
real pointCoordX = this->pointCoordsX[point];
real pointCoordY = this->pointCoordsY[point];
real pointCoordZ = this->pointCoordsZ[point];
- real distX = pointCoordX-para->getParH(level)->coordinateX[j];
- real distY = pointCoordY-para->getParH(level)->coordinateY[j];
- real distZ = pointCoordZ-para->getParH(level)->coordinateZ[j];
+ real distX = pointCoordX-para->getParH(level)->coordinateX[pos];
+ real distY = pointCoordY-para->getParH(level)->coordinateY[pos];
+ real distZ = pointCoordZ-para->getParH(level)->coordinateZ[pos];
if( distX <=dx && distY <=dx && distZ <=dx &&
distX >0.f && distY >0.f && distZ >0.f)
{
- probeIndices_level.push_back(j);
+ probeIndices_level.push_back((int)pos);
distX_level.push_back( distX/dx );
distY_level.push_back( distY/dx );
distZ_level.push_back( distZ/dx );
@@ -140,4 +140,14 @@ void PointProbe::addProbePointsFromXNormalPlane(real pos_x, real pos0_y, real po
}
printf("Added %u points \n", n_y*n_z);
+}
+
+void PointProbe::getTaggedFluidNodes(Parameter *para, GridProvider* gridProvider)
+{
+ for(int level=0; level<=para->getMaxLevel(); level++)
+ {
+ SPtr<ProbeStruct> probeStruct = this->getProbeStruct(level);
+ std::vector<uint> probeIndices( probeStruct->pointIndicesH, probeStruct->pointIndicesH+probeStruct->nIndices);
+ gridProvider->tagFluidNodeIndices( probeIndices, CollisionTemplate::WriteMacroVars, level);
+ }
}
\ No newline at end of file
diff --git a/src/gpu/VirtualFluids_GPU/PreCollisionInteractor/Probes/PointProbe.h b/src/gpu/VirtualFluids_GPU/PreCollisionInteractor/Probes/PointProbe.h
index 6a6fbe76f089acfafc22672dd3e9d71bd193a3b3..08c359705f03b20fbd3276fe209b6ff4d782a5e5 100644
--- a/src/gpu/VirtualFluids_GPU/PreCollisionInteractor/Probes/PointProbe.h
+++ b/src/gpu/VirtualFluids_GPU/PreCollisionInteractor/Probes/PointProbe.h
@@ -64,6 +64,7 @@ public:
void addProbePointsFromList(std::vector<real>& _pointCoordsX, std::vector<real>& _pointCoordsY, std::vector<real>& _pointCoordsZ);
void addProbePointsFromXNormalPlane(real pos_x, real pos0_y, real pos0_z, real pos1_y, real pos1_z, uint n_y, uint n_z);
+ void getTaggedFluidNodes(Parameter *para, GridProvider* gridProvider) override;
private:
bool isAvailableStatistic(Statistic _variable) override;
diff --git a/src/gpu/VirtualFluids_GPU/PreCollisionInteractor/Probes/Probe.cu b/src/gpu/VirtualFluids_GPU/PreCollisionInteractor/Probes/Probe.cu
index cc027b07bded01455437e65e08ccdcd51bcf7dc0..03c18f5a9a2133bec244053113209abc70469a2a 100644
--- a/src/gpu/VirtualFluids_GPU/PreCollisionInteractor/Probes/Probe.cu
+++ b/src/gpu/VirtualFluids_GPU/PreCollisionInteractor/Probes/Probe.cu
@@ -187,7 +187,7 @@ void Probe::init(Parameter* para, GridProvider* gridProvider, CudaMemoryManager*
this->velocityRatio = std::bind(&Parameter::getScaledVelocityRatio, para, _1);
this->densityRatio = std::bind(&Parameter::getScaledDensityRatio, para, _1);
this->forceRatio = std::bind(&Parameter::getScaledForceRatio, para, _1);
- this->stressRatio = std::bind(&Parameter::getScaledPressureRatio, para, _1);
+ this->stressRatio = std::bind(&Parameter::getScaledStressRatio, para, _1);
this->viscosityRatio = std::bind(&Parameter::getScaledViscosityRatio, para, _1);
this->nondimensional = std::bind(&Probe::getNondimensionalConversionFactor, this, _1);
@@ -315,6 +315,12 @@ void Probe::free(Parameter* para, CudaMemoryManager* cudaMemoryManager)
}
}
+void Probe::getTaggedFluidNodes(Parameter *para, GridProvider* gridProvider)
+{
+ // Do nothing
+};
+
+
void Probe::addStatistic(Statistic variable)
{
if (!this->isAvailableStatistic(variable)) throw std::runtime_error("Probe::addStatistic(): Statistic not available for this probe type!");
@@ -329,6 +335,22 @@ void Probe::addStatistic(Statistic variable)
}
}
+std::string Probe::makeParallelFileName(int id, int t)
+{
+ return this->probeName + "_bin_ID_" + StringUtil::toString<int>(id)
+ + "_t_" + StringUtil::toString<int>(t)
+ + ".vtk";
+}
+
+std::string Probe::makeGridFileName(int level, int id, int t, uint part)
+{
+ return this->probeName + "_bin_lev_" + StringUtil::toString<int>(level)
+ + "_ID_" + StringUtil::toString<int>(id)
+ + "_Part_" + StringUtil::toString<int>(part)
+ + "_t_" + StringUtil::toString<int>(t)
+ + ".vtk";
+}
+
void Probe::addAllAvailableStatistics()
{
for( int var=0; var < int(Statistic::LAST); var++)
@@ -347,119 +369,76 @@ void Probe::write(Parameter* para, int level, int t)
std::vector<std::string> fnames;
for (uint i = 1; i <= numberOfParts; i++)
{
- std::string fname = this->probeName + "_bin_lev_" + StringUtil::toString<int>(level)
- + "_ID_" + StringUtil::toString<int>(para->getMyProcessID())
- + "_Part_" + StringUtil::toString<int>(i);
- if(!this->outputTimeSeries) fname += "_t_" + StringUtil::toString<int>(t_write);
- fname += ".vtk";
- fnames.push_back(fname);
- this->fileNamesForCollectionFile.push_back(fname);
+ this->writeGridFile(para, level, t_write, i);
}
- this->writeGridFiles(para, level, fnames, t);
-
- if(level == 0 && !this->outputTimeSeries) this->writeCollectionFile(para, t);
+ if(level == 0&& !this->outputTimeSeries) this->writeParallelFile(para, t);
}
-void Probe::writeCollectionFile(Parameter* para, int t)
+void Probe::writeParallelFile(Parameter* para, int t)
{
int t_write = this->fileNameLU ? t: t/this->tOut;
- std::string filename = this->probeName + "_bin_ID_" + StringUtil::toString<int>(para->getMyProcessID())
- + "_t_" + StringUtil::toString<int>(t_write)
- + ".vtk";
-
- std::ofstream file;
-
- file.open(this->outputPath + "/" + filename + ".pvtu" );
-
- //////////////////////////////////////////////////////////////////////////
-
- file << "<VTKFile type=\"PUnstructuredGrid\" version=\"1.0\" byte_order=\"LittleEndian\" header_type=\"UInt64\">" << std::endl;
- file << " <PUnstructuredGrid GhostLevel=\"1\">" << std::endl;
-
- file << " <PPointData>" << std::endl;
-
- for(std::string varName: this->getVarNames()) //TODO
- {
- file << " <DataArray type=\"Float64\" Name=\""<< varName << "\" /> " << std::endl;
- }
- file << " </PPointData>" << std::endl;
-
- file << " <PPoints>" << std::endl;
- file << " <PDataArray type=\"Float32\" Name=\"Points\" NumberOfComponents=\"3\"/>" << std::endl;
- file << " </PPoints>" << std::endl;
-
- for( auto& fname : this->fileNamesForCollectionFile )
- {
- const auto filenameWithoutPath=fname.substr( fname.find_last_of('/') + 1 );
- file << " <Piece Source=\"" << filenameWithoutPath << ".bin.vtu\"/>" << std::endl;
- }
-
- file << " </PUnstructuredGrid>" << std::endl;
- file << "</VTKFile>" << std::endl;
+ std::string filename = this->outputPath + "/" + this->makeParallelFileName(para->getMyProcessID(), t_write);
- //////////////////////////////////////////////////////////////////////////
+ std::vector<std::string> nodedatanames = this->getVarNames();
+ std::vector<std::string> cellNames;
- file.close();
+ getWriter()->writeParallelFile(filename, fileNamesForCollectionFile, nodedatanames, cellNames);
this->fileNamesForCollectionFile.clear();
}
-void Probe::writeGridFiles(Parameter* para, int level, std::vector<std::string>& fnames, int t)
+void Probe::writeGridFile(Parameter* para, int level, int t, uint part)
{
+ std::string fname = this->outputPath + "/" + this->makeGridFileName(level, para->getMyProcessID(), t, part);
+
std::vector< UbTupleFloat3 > nodes;
std::vector< std::string > nodedatanames = this->getVarNames();
- uint startpos = 0;
- uint endpos = 0;
- uint sizeOfNodes = 0;
std::vector< std::vector< double > > nodedata(nodedatanames.size());
SPtr<ProbeStruct> probeStruct = this->getProbeStruct(level);
- for (uint part = 0; part < fnames.size(); part++)
- {
- startpos = part * para->getlimitOfNodesForVTK();
- uint nDataPoints = this->outputTimeSeries? this->tProbe: probeStruct->nPoints;
- sizeOfNodes = min(para->getlimitOfNodesForVTK(), nDataPoints - startpos);
- endpos = startpos + sizeOfNodes;
+ uint startpos = (part-1) * para->getlimitOfNodesForVTK();
+ uint sizeOfNodes = min(para->getlimitOfNodesForVTK(), probeStruct->nPoints - startpos);
+ uint endpos = startpos + sizeOfNodes;
- //////////////////////////////////////////////////////////////////////////
- nodes.resize(sizeOfNodes);
+ //////////////////////////////////////////////////////////////////////////
+ nodes.resize(sizeOfNodes);
- for (uint pos = startpos; pos < endpos; pos++)
- {
- nodes[pos-startpos] = makeUbTuple( float(probeStruct->pointCoordsX[pos]),
- float(probeStruct->pointCoordsY[pos]),
- float(probeStruct->pointCoordsZ[pos]));
- }
+ for (uint pos = startpos; pos < endpos; pos++)
+ {
+ nodes[pos-startpos] = makeUbTuple( float(probeStruct->pointCoordsX[pos]),
+ float(probeStruct->pointCoordsY[pos]),
+ float(probeStruct->pointCoordsZ[pos]));
+ }
- for( auto it=nodedata.begin(); it!=nodedata.end(); it++) it->resize(sizeOfNodes);
+ for( auto it=nodedata.begin(); it!=nodedata.end(); it++) it->resize(sizeOfNodes);
- for( int var=0; var < int(Statistic::LAST); var++){
- if(this->quantities[var])
- {
- Statistic statistic = static_cast<Statistic>(var);
- real coeff;
+ for( int var=0; var < int(Statistic::LAST); var++){
+ if(this->quantities[var])
+ {
+ Statistic statistic = static_cast<Statistic>(var);
+ real coeff;
- std::vector<PostProcessingVariable> postProcessingVariables = this->getPostProcessingVariables(statistic);
- uint n_arrs = uint(postProcessingVariables.size());
+ std::vector<PostProcessingVariable> postProcessingVariables = this->getPostProcessingVariables(statistic);
+ uint n_arrs = uint(postProcessingVariables.size());
- uint arrOff = probeStruct->arrayOffsetsH[var];
- uint arrLen = probeStruct->nPoints;
+ uint arrOff = probeStruct->arrayOffsetsH[var];
+ uint arrLen = probeStruct->nPoints;
+
+ for(uint arr=0; arr<n_arrs; arr++)
+ {
+ coeff = postProcessingVariables[arr].conversionFactor(level);
- for(uint arr=0; arr<n_arrs; arr++)
+ for (uint pos = startpos; pos < endpos; pos++)
{
- coeff = postProcessingVariables[arr].conversionFactor(level);
-
- for (uint pos = startpos; pos < endpos; pos++)
- {
- nodedata[arrOff+arr][pos-startpos] = double(probeStruct->quantitiesArrayH[(arrOff+arr)*arrLen+pos]*coeff);
- }
+ nodedata[arrOff+arr][pos-startpos] = double(probeStruct->quantitiesArrayH[(arrOff+arr)*arrLen+pos]*coeff);
}
}
}
- WbWriterVtkXmlBinary::getInstance()->writeNodesWithNodeData(this->outputPath + "/" + fnames[part], nodes, nodedatanames, nodedata);
}
+ std::string fullName = getWriter()->writeNodesWithNodeData(fname, nodes, nodedatanames, nodedata);
+ this->fileNamesForCollectionFile.push_back(fullName.substr(fullName.find_last_of('/') + 1));
}
std::vector<std::string> Probe::getVarNames()
diff --git a/src/gpu/VirtualFluids_GPU/PreCollisionInteractor/Probes/Probe.h b/src/gpu/VirtualFluids_GPU/PreCollisionInteractor/Probes/Probe.h
index 9cb0bd43e27fb7a28cae9c363ce245fbd9cc5677..aaf294e87d23c64707a16692b9337d6e9ff9c896 100644
--- a/src/gpu/VirtualFluids_GPU/PreCollisionInteractor/Probes/Probe.h
+++ b/src/gpu/VirtualFluids_GPU/PreCollisionInteractor/Probes/Probe.h
@@ -49,6 +49,7 @@
#include "PreCollisionInteractor/PreCollisionInteractor.h"
#include "PointerDefinitions.h"
+#include "WbWriterVtkXmlBinary.h"
//=======================================================================================
//! \note How to add new Statistics
@@ -153,12 +154,12 @@ public:
PreCollisionInteractor()
{
if (_tStartOut<_tStartAvg) throw std::runtime_error("Probe: tStartOut must be larger than tStartAvg!");
- if (_tStartTmpAvg<_tStartAvg) throw std::runtime_error("Probe: tStartTmpAvg must be larger than tStartAvg!");
}
void init(Parameter* para, GridProvider* gridProvider, CudaMemoryManager* cudaMemoryManager) override;
void interact(Parameter* para, CudaMemoryManager* cudaMemoryManager, int level, uint t) override;
void free(Parameter* para, CudaMemoryManager* cudaMemoryManager) override;
+ virtual void getTaggedFluidNodes(Parameter *para, GridProvider* gridProvider) override;
SPtr<ProbeStruct> getProbeStruct(int level){ return this->probeParams[level]; }
@@ -171,6 +172,8 @@ public:
void setFileNameToNOut(){this->fileNameLU = false;}
void setTStartTmpAveraging(uint _tStartTmpAveraging){this->tStartTmpAveraging = _tStartTmpAveraging;}
+protected:
+ virtual WbWriterVtkXmlBinary* getWriter(){ return WbWriterVtkXmlBinary::getInstance(); };
real getNondimensionalConversionFactor(int level);
private:
@@ -188,12 +191,15 @@ private:
int level);
virtual void calculateQuantities(SPtr<ProbeStruct> probeStruct, Parameter* para, uint t, int level) = 0;
- void write(Parameter* para, int level, int t);
- void writeCollectionFile(Parameter* para, int t);
- void writeGridFiles(Parameter* para, int level, std::vector<std::string >& fnames, int t);
+ virtual void write(Parameter* para, int level, int t);
+ virtual void writeParallelFile(Parameter* para, int t);
+ virtual void writeGridFile(Parameter* para, int level, int t, uint part);
+
std::vector<std::string> getVarNames();
-
-private:
+ std::string makeGridFileName(int level, int id, int t, uint part);
+ std::string makeParallelFileName(int id, int t);
+
+protected:
const std::string probeName;
const std::string outputPath;
@@ -202,7 +208,6 @@ private:
bool hasDeviceQuantityArray; //!> flag initiating memCopy in Point and PlaneProbe. Other probes are only based on thrust reduce functions and therefore dont need explict memCopy in interact()
bool outputTimeSeries; //!> flag initiating overwrite of output vtk files, skipping collection files and limiting the length of the written data to the current time step (currently only used for WallModelProbe)
std::vector<std::string> fileNamesForCollectionFile;
- std::vector<std::string> varNames;
bool fileNameLU = true; //!> if true, written file name contains time step in LU, else is the number of the written probe files
@@ -215,7 +220,6 @@ protected:
uint tProbe = 0; //!> counter for number of probe evaluations. Only used when outputting timeseries
-
std::function<real(int)> velocityRatio;
std::function<real(int)> densityRatio;
std::function<real(int)> forceRatio;
diff --git a/src/gpu/VirtualFluids_GPU/PreCollisionInteractor/Probes/WallModelProbe.cu b/src/gpu/VirtualFluids_GPU/PreCollisionInteractor/Probes/WallModelProbe.cu
index 81da15595baae55aa562bc77e24442a9258d992f..3341111c134ace7ca6ff64eeb7f87b38f8014656 100644
--- a/src/gpu/VirtualFluids_GPU/PreCollisionInteractor/Probes/WallModelProbe.cu
+++ b/src/gpu/VirtualFluids_GPU/PreCollisionInteractor/Probes/WallModelProbe.cu
@@ -171,11 +171,11 @@ void WallModelProbe::findPoints(Parameter* para, GridProvider* gridProvider, std
{
if (!para->getIsBodyForce()) throw std::runtime_error("WallModelProbe::findPoints(): bodyforce not allocated!");
// Find all fluid nodes
- for(uint j=1; j<para->getParH(level)->numberOfNodes; j++ )
+ for(size_t pos = 1; pos < para->getParH(level)->numberOfNodes; pos++ )
{
- if( para->getParH(level)->typeOfGridNode[j] == GEO_FLUID)
+ if( para->getParH(level)->typeOfGridNode[pos] == GEO_FLUID)
{
- probeIndices_level.push_back(j);
+ probeIndices_level.push_back((int)pos);
}
}
}
diff --git a/src/gpu/VirtualFluids_GPU/PreCollisionInteractor/Probes/WallModelProbe.h b/src/gpu/VirtualFluids_GPU/PreCollisionInteractor/Probes/WallModelProbe.h
index d6464c5ca2aa60310cc6bb7ca0a210bc12e755ff..4ea90f74c7a0d57af4995e1b5874234967f1e901 100644
--- a/src/gpu/VirtualFluids_GPU/PreCollisionInteractor/Probes/WallModelProbe.h
+++ b/src/gpu/VirtualFluids_GPU/PreCollisionInteractor/Probes/WallModelProbe.h
@@ -55,14 +55,17 @@ public:
uint _tStartOut,
uint _tOut
): Probe(_probeName,
- _outputPath,
- _tStartAvg,
- _tStartTmpAvg,
- _tAvg,
- _tStartOut,
- _tOut,
- false,
- true){}
+ _outputPath,
+ _tStartAvg,
+ _tStartTmpAvg,
+ _tAvg,
+ _tStartOut,
+ _tOut,
+ false,
+ true)
+ {
+ if (_tStartTmpAvg<_tStartAvg) throw std::runtime_error("Probe: tStartTmpAvg must be larger than tStartAvg!");
+ }
void setForceOutputToStress(bool _outputStress){ this->outputStress = _outputStress; }
diff --git a/src/gpu/VirtualFluids_GPU/PreProcessor/PreProcessorStrategy/InitCompAD27/InitCompAD27.cu b/src/gpu/VirtualFluids_GPU/PreProcessor/PreProcessorStrategy/InitCompAD27/InitCompAD27.cu
index e43fb54a6b56b4d9a501269544cea000df31cdb7..60dbb2228e6d01fdabf7a6e1bfca786e2104d5b0 100644
--- a/src/gpu/VirtualFluids_GPU/PreProcessor/PreProcessorStrategy/InitCompAD27/InitCompAD27.cu
+++ b/src/gpu/VirtualFluids_GPU/PreProcessor/PreProcessorStrategy/InitCompAD27/InitCompAD27.cu
@@ -2,6 +2,7 @@
#include "InitCompAD27_Device.cuh"
#include "Parameter/Parameter.h"
+#include "cuda/CudaGrid.h"
std::shared_ptr<PreProcessorStrategy> InitCompAD27::getNewInstance(std::shared_ptr<Parameter> para)
{
@@ -10,36 +11,21 @@ std::shared_ptr<PreProcessorStrategy> InitCompAD27::getNewInstance(std::shared_p
void InitCompAD27::init(int level)
{
- int numberOfThreads = para->getParD(level)->numberofthreads;
- int size_Mat = para->getParD(level)->numberOfNodes;
-
- int Grid = (size_Mat / numberOfThreads) + 1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid / Grid1) + 1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 grid(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1);
-
- LB_Init_Comp_AD_27 << < grid, threads >> >( para->getParD(level)->neighborX,
- para->getParD(level)->neighborY,
- para->getParD(level)->neighborZ,
- para->getParD(level)->typeOfGridNode,
- para->getParD(level)->Conc,
- para->getParD(level)->velocityX,
- para->getParD(level)->velocityY,
- para->getParD(level)->velocityZ,
- para->getParD(level)->numberOfNodes,
- para->getParD(level)->distributionsAD27.f[0],
- para->getParD(level)->isEvenTimestep);
- getLastCudaError("InitAD27 execution failed");
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(para->getParD(level)->numberofthreads, para->getParD(level)->numberOfNodes);
+
+ LB_Init_Comp_AD_27 <<< grid.grid, grid.threads >>>(
+ para->getParD(level)->neighborX,
+ para->getParD(level)->neighborY,
+ para->getParD(level)->neighborZ,
+ para->getParD(level)->typeOfGridNode,
+ para->getParD(level)->Conc,
+ para->getParD(level)->velocityX,
+ para->getParD(level)->velocityY,
+ para->getParD(level)->velocityZ,
+ para->getParD(level)->numberOfNodes,
+ para->getParD(level)->distributionsAD27.f[0],
+ para->getParD(level)->isEvenTimestep);
+ getLastCudaError("LB_Init_Comp_AD_27 execution failed");
}
bool InitCompAD27::checkParameter()
diff --git a/src/gpu/VirtualFluids_GPU/PreProcessor/PreProcessorStrategy/InitCompAD7/InitCompAD7.cu b/src/gpu/VirtualFluids_GPU/PreProcessor/PreProcessorStrategy/InitCompAD7/InitCompAD7.cu
index 8a53dff5c14adef69aa012bdf1d870d62a9749b2..8097ee13d9064c4104ead8cd8eb5ba529d8972fc 100644
--- a/src/gpu/VirtualFluids_GPU/PreProcessor/PreProcessorStrategy/InitCompAD7/InitCompAD7.cu
+++ b/src/gpu/VirtualFluids_GPU/PreProcessor/PreProcessorStrategy/InitCompAD7/InitCompAD7.cu
@@ -2,6 +2,7 @@
#include "InitCompAD7_Device.cuh"
#include "Parameter/Parameter.h"
+#include "cuda/CudaGrid.h"
std::shared_ptr<InitCompAD7> InitCompAD7::getNewInstance(std::shared_ptr<Parameter> para)
{
@@ -10,36 +11,21 @@ std::shared_ptr<InitCompAD7> InitCompAD7::getNewInstance(std::shared_ptr<Paramet
void InitCompAD7::init(int level)
{
- int numberOfThreads = para->getParD(level)->numberofthreads;
- int size_Mat = para->getParD(level)->numberOfNodes;
-
- int Grid = (size_Mat / numberOfThreads) + 1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid / Grid1) + 1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 grid(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1);
-
- LB_Init_Comp_AD_7 << < grid, threads >> >( para->getParD(level)->neighborX,
- para->getParD(level)->neighborY,
- para->getParD(level)->neighborZ,
- para->getParD(level)->typeOfGridNode,
- para->getParD(level)->Conc,
- para->getParD(level)->velocityX,
- para->getParD(level)->velocityY,
- para->getParD(level)->velocityZ,
- para->getParD(level)->numberOfNodes,
- para->getParD(level)->distributionsAD7.f[0],
- para->getParD(level)->isEvenTimestep);
- getLastCudaError("InitAD7 execution failed");
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(para->getParD(level)->numberofthreads, para->getParD(level)->numberOfNodes);
+
+ LB_Init_Comp_AD_7 <<< grid.grid, grid.threads >>>(
+ para->getParD(level)->neighborX,
+ para->getParD(level)->neighborY,
+ para->getParD(level)->neighborZ,
+ para->getParD(level)->typeOfGridNode,
+ para->getParD(level)->Conc,
+ para->getParD(level)->velocityX,
+ para->getParD(level)->velocityY,
+ para->getParD(level)->velocityZ,
+ para->getParD(level)->numberOfNodes,
+ para->getParD(level)->distributionsAD7.f[0],
+ para->getParD(level)->isEvenTimestep);
+ getLastCudaError("LB_Init_Comp_AD_7 execution failed");
}
bool InitCompAD7::checkParameter()
diff --git a/src/gpu/VirtualFluids_GPU/PreProcessor/PreProcessorStrategy/InitCompSP27/InitCompSP27.cu b/src/gpu/VirtualFluids_GPU/PreProcessor/PreProcessorStrategy/InitCompSP27/InitCompSP27.cu
index 23ec3e5293ec3a49bf632a720ab554d156dc9674..c4676f28f969e2db8ff7f1910ac784a1c0dab351 100644
--- a/src/gpu/VirtualFluids_GPU/PreProcessor/PreProcessorStrategy/InitCompSP27/InitCompSP27.cu
+++ b/src/gpu/VirtualFluids_GPU/PreProcessor/PreProcessorStrategy/InitCompSP27/InitCompSP27.cu
@@ -2,6 +2,7 @@
#include "InitCompSP27_Device.cuh"
#include "Parameter/Parameter.h"
+#include "cuda/CudaGrid.h"
std::shared_ptr<PreProcessorStrategy> InitCompSP27::getNewInstance(std::shared_ptr<Parameter> para)
{
@@ -10,27 +11,12 @@ std::shared_ptr<PreProcessorStrategy> InitCompSP27::getNewInstance(std::shared_p
void InitCompSP27::init(int level)
{
- int numberOfThreads = para->getParD(level)->numberofthreads;
- int size_Mat = para->getParD(level)->numberOfNodes;
-
- int Grid = (size_Mat / numberOfThreads) + 1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid / Grid1) + 1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 grid(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1);
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(para->getParD(level)->numberofthreads, para->getParD(level)->numberOfNodes);
if( ! para->getUseInitNeq() )
{
- LB_Init_Comp_SP_27 <<< grid, threads >>> (para->getParD(level)->neighborX,
+ LB_Init_Comp_SP_27 <<< grid.grid, grid.threads >>> (
+ para->getParD(level)->neighborX,
para->getParD(level)->neighborY,
para->getParD(level)->neighborZ,
para->getParD(level)->typeOfGridNode,
@@ -41,11 +27,12 @@ void InitCompSP27::init(int level)
para->getParD(level)->numberOfNodes,
para->getParD(level)->distributions.f[0],
para->getParD(level)->isEvenTimestep);
- getLastCudaError("LBInitSP27 execution failed");
+ getLastCudaError("LB_Init_Comp_SP_27 execution failed");
}
else
{
- LB_Init_Comp_Neq_SP_27 <<< grid, threads >>> (para->getParD(level)->neighborX,
+ LB_Init_Comp_Neq_SP_27 <<< grid.grid, grid.threads >>> (
+ para->getParD(level)->neighborX,
para->getParD(level)->neighborY,
para->getParD(level)->neighborZ,
para->getParD(level)->neighborInverse,
@@ -59,7 +46,7 @@ void InitCompSP27::init(int level)
para->getParD(level)->omega,
para->getParD(level)->isEvenTimestep);
cudaDeviceSynchronize();
- getLastCudaError("LBInitNeqSP27 execution failed");
+ getLastCudaError("LB_Init_Comp_Neq_SP_27 execution failed");
}
diff --git a/src/gpu/VirtualFluids_GPU/PreProcessor/PreProcessorStrategy/InitF3/InitF3.cu b/src/gpu/VirtualFluids_GPU/PreProcessor/PreProcessorStrategy/InitF3/InitF3.cu
index cb6b40b4371a206c6d1e031822338621c4907be1..14d6b725337aa8b9af279bf794ff1c0912516b64 100644
--- a/src/gpu/VirtualFluids_GPU/PreProcessor/PreProcessorStrategy/InitF3/InitF3.cu
+++ b/src/gpu/VirtualFluids_GPU/PreProcessor/PreProcessorStrategy/InitF3/InitF3.cu
@@ -2,6 +2,7 @@
#include "InitF3_Device.cuh"
#include "Parameter/Parameter.h"
+#include "cuda/CudaGrid.h"
std::shared_ptr<PreProcessorStrategy> InitF3::getNewInstance(std::shared_ptr<Parameter> para)
{
@@ -10,36 +11,21 @@ std::shared_ptr<PreProcessorStrategy> InitF3::getNewInstance(std::shared_ptr<Par
void InitF3::init(int level)
{
- int numberOfThreads = para->getParD(level)->numberofthreads;
- int size_Mat = para->getParD(level)->numberOfNodes;
-
- int Grid = (size_Mat / numberOfThreads) + 1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid / Grid1) + 1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 grid(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1);
-
- LB_Init_F3 << < grid, threads >> >( para->getParD(level)->neighborX,
- para->getParD(level)->neighborY,
- para->getParD(level)->neighborZ,
- para->getParD(level)->typeOfGridNode,
- para->getParD(level)->rho,
- para->getParD(level)->velocityX,
- para->getParD(level)->velocityY,
- para->getParD(level)->velocityZ,
- para->getParD(level)->numberOfNodes,
- para->getParD(level)->g6.g[0],
- para->getParD(level)->isEvenTimestep);
- getLastCudaError("LBInitF3 execution failed");
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(para->getParD(level)->numberofthreads, para->getParD(level)->numberOfNodes);
+
+ LB_Init_F3 <<< grid.grid, grid.threads >>>(
+ para->getParD(level)->neighborX,
+ para->getParD(level)->neighborY,
+ para->getParD(level)->neighborZ,
+ para->getParD(level)->typeOfGridNode,
+ para->getParD(level)->rho,
+ para->getParD(level)->velocityX,
+ para->getParD(level)->velocityY,
+ para->getParD(level)->velocityZ,
+ para->getParD(level)->numberOfNodes,
+ para->getParD(level)->g6.g[0],
+ para->getParD(level)->isEvenTimestep);
+ getLastCudaError("LB_Init_F3 execution failed");
}
bool InitF3::checkParameter()
diff --git a/src/gpu/VirtualFluids_GPU/PreProcessor/PreProcessorStrategy/InitIncompAD27/InitIncompAD27.cu b/src/gpu/VirtualFluids_GPU/PreProcessor/PreProcessorStrategy/InitIncompAD27/InitIncompAD27.cu
index 419ae80b96be57f8dc9c4ebecaccac0d435f00e0..ea700010960b11a1facdda18c35f220f43eb6a66 100644
--- a/src/gpu/VirtualFluids_GPU/PreProcessor/PreProcessorStrategy/InitIncompAD27/InitIncompAD27.cu
+++ b/src/gpu/VirtualFluids_GPU/PreProcessor/PreProcessorStrategy/InitIncompAD27/InitIncompAD27.cu
@@ -2,6 +2,7 @@
#include "InitIncompAD27_Device.cuh"
#include "Parameter/Parameter.h"
+#include "cuda/CudaGrid.h"
std::shared_ptr<PreProcessorStrategy> InitIncompAD27::getNewInstance(std::shared_ptr<Parameter> para)
{
@@ -10,36 +11,21 @@ std::shared_ptr<PreProcessorStrategy> InitIncompAD27::getNewInstance(std::shared
void InitIncompAD27::init(int level)
{
- int numberOfThreads = para->getParD(level)->numberofthreads;
- int size_Mat = para->getParD(level)->numberOfNodes;
-
- int Grid = (size_Mat / numberOfThreads) + 1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid / Grid1) + 1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 grid(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1);
-
- LB_Init_Incomp_AD_27 << < grid, threads >> >( para->getParD(level)->neighborX,
- para->getParD(level)->neighborY,
- para->getParD(level)->neighborZ,
- para->getParD(level)->typeOfGridNode,
- para->getParD(level)->Conc,
- para->getParD(level)->velocityX,
- para->getParD(level)->velocityY,
- para->getParD(level)->velocityZ,
- para->getParD(level)->numberOfNodes,
- para->getParD(level)->distributionsAD27.f[0],
- para->getParD(level)->isEvenTimestep);
- getLastCudaError("LBInitIncompAD27 execution failed");
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(para->getParD(level)->numberofthreads, para->getParD(level)->numberOfNodes);
+
+ LB_Init_Incomp_AD_27 <<< grid.grid, grid.threads >>>(
+ para->getParD(level)->neighborX,
+ para->getParD(level)->neighborY,
+ para->getParD(level)->neighborZ,
+ para->getParD(level)->typeOfGridNode,
+ para->getParD(level)->Conc,
+ para->getParD(level)->velocityX,
+ para->getParD(level)->velocityY,
+ para->getParD(level)->velocityZ,
+ para->getParD(level)->numberOfNodes,
+ para->getParD(level)->distributionsAD27.f[0],
+ para->getParD(level)->isEvenTimestep);
+ getLastCudaError("LB_Init_Incomp_AD_27 execution failed");
}
bool InitIncompAD27::checkParameter()
diff --git a/src/gpu/VirtualFluids_GPU/PreProcessor/PreProcessorStrategy/InitIncompAD7/InitIncompAD7.cu b/src/gpu/VirtualFluids_GPU/PreProcessor/PreProcessorStrategy/InitIncompAD7/InitIncompAD7.cu
index 795cd0496a207e0861e35e4f310481950a037caf..d7c08e6932cacf2fb5a946010c1855212f1631fc 100644
--- a/src/gpu/VirtualFluids_GPU/PreProcessor/PreProcessorStrategy/InitIncompAD7/InitIncompAD7.cu
+++ b/src/gpu/VirtualFluids_GPU/PreProcessor/PreProcessorStrategy/InitIncompAD7/InitIncompAD7.cu
@@ -2,6 +2,7 @@
#include "InitIncompAD7_Device.cuh"
#include "Parameter/Parameter.h"
+#include "cuda/CudaGrid.h"
std::shared_ptr<PreProcessorStrategy> InitIncompAD7::getNewInstance(std::shared_ptr<Parameter> para)
{
@@ -10,36 +11,21 @@ std::shared_ptr<PreProcessorStrategy> InitIncompAD7::getNewInstance(std::shared_
void InitIncompAD7::init(int level)
{
- int numberOfThreads = para->getParD(level)->numberofthreads;
- int size_Mat = para->getParD(level)->numberOfNodes;
-
- int Grid = (size_Mat / numberOfThreads) + 1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid / Grid1) + 1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 grid(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1);
-
- LB_Init_Incomp_AD_7 << < grid, threads >> >( para->getParD(level)->neighborX,
- para->getParD(level)->neighborY,
- para->getParD(level)->neighborZ,
- para->getParD(level)->typeOfGridNode,
- para->getParD(level)->Conc,
- para->getParD(level)->velocityX,
- para->getParD(level)->velocityY,
- para->getParD(level)->velocityZ,
- para->getParD(level)->numberOfNodes,
- para->getParD(level)->distributionsAD27.f[0],
- para->getParD(level)->isEvenTimestep);
- getLastCudaError("LBInitIncompAD7 execution failed");
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(para->getParD(level)->numberofthreads, para->getParD(level)->numberOfNodes);
+
+ LB_Init_Incomp_AD_7 <<< grid.grid, grid.threads >>>(
+ para->getParD(level)->neighborX,
+ para->getParD(level)->neighborY,
+ para->getParD(level)->neighborZ,
+ para->getParD(level)->typeOfGridNode,
+ para->getParD(level)->Conc,
+ para->getParD(level)->velocityX,
+ para->getParD(level)->velocityY,
+ para->getParD(level)->velocityZ,
+ para->getParD(level)->numberOfNodes,
+ para->getParD(level)->distributionsAD27.f[0],
+ para->getParD(level)->isEvenTimestep);
+ getLastCudaError("LB_Init_Incomp_AD_7 execution failed");
}
bool InitIncompAD7::checkParameter()
diff --git a/src/gpu/VirtualFluids_GPU/PreProcessor/PreProcessorStrategy/InitSP27/InitSP27.cu b/src/gpu/VirtualFluids_GPU/PreProcessor/PreProcessorStrategy/InitSP27/InitSP27.cu
index 0538c7ab89eb750a40cfc47486dc0891d4493976..078ad24f24659bf10a3dc9ed90bfd62b5e021187 100644
--- a/src/gpu/VirtualFluids_GPU/PreProcessor/PreProcessorStrategy/InitSP27/InitSP27.cu
+++ b/src/gpu/VirtualFluids_GPU/PreProcessor/PreProcessorStrategy/InitSP27/InitSP27.cu
@@ -2,6 +2,7 @@
#include "InitSP27_Device.cuh"
#include "Parameter/Parameter.h"
+#include "cuda/CudaGrid.h"
std::shared_ptr<PreProcessorStrategy> InitSP27::getNewInstance(std::shared_ptr<Parameter> para)
{
@@ -10,36 +11,21 @@ std::shared_ptr<PreProcessorStrategy> InitSP27::getNewInstance(std::shared_ptr<P
void InitSP27::init(int level)
{
- int numberOfThreads = para->getParD(level)->numberofthreads;
- int size_Mat = para->getParD(level)->numberOfNodes;
-
- int Grid = (size_Mat / numberOfThreads) + 1;
- int Grid1, Grid2;
- if (Grid>512)
- {
- Grid1 = 512;
- Grid2 = (Grid / Grid1) + 1;
- }
- else
- {
- Grid1 = 1;
- Grid2 = Grid;
- }
- dim3 grid(Grid1, Grid2);
- dim3 threads(numberOfThreads, 1, 1);
-
- LB_Init_SP_27 << < grid, threads >> >( para->getParD(level)->neighborX,
- para->getParD(level)->neighborY,
- para->getParD(level)->neighborZ,
- para->getParD(level)->typeOfGridNode,
- para->getParD(level)->rho,
- para->getParD(level)->velocityX,
- para->getParD(level)->velocityY,
- para->getParD(level)->velocityZ,
- para->getParD(level)->numberOfNodes,
- para->getParD(level)->distributions.f[0],
- para->getParD(level)->isEvenTimestep);
- getLastCudaError("LBInitSP27 execution failed");
+ vf::cuda::CudaGrid grid = vf::cuda::CudaGrid(para->getParD(level)->numberofthreads, para->getParD(level)->numberOfNodes);
+
+ LB_Init_SP_27 <<< grid.grid, grid.threads >>>(
+ para->getParD(level)->neighborX,
+ para->getParD(level)->neighborY,
+ para->getParD(level)->neighborZ,
+ para->getParD(level)->typeOfGridNode,
+ para->getParD(level)->rho,
+ para->getParD(level)->velocityX,
+ para->getParD(level)->velocityY,
+ para->getParD(level)->velocityZ,
+ para->getParD(level)->numberOfNodes,
+ para->getParD(level)->distributions.f[0],
+ para->getParD(level)->isEvenTimestep);
+ getLastCudaError("LB_Init_SP_27 execution failed");
}
bool InitSP27::checkParameter()
diff --git a/src/gpu/VirtualFluids_GPU/Restart/RestartObject.cpp b/src/gpu/VirtualFluids_GPU/Restart/RestartObject.cpp
index a38535f7bdff0d87a5af74a69f0ed8255c647382..15813b7967a84e45f44eb4d286c41aa99c4ff343 100644
--- a/src/gpu/VirtualFluids_GPU/Restart/RestartObject.cpp
+++ b/src/gpu/VirtualFluids_GPU/Restart/RestartObject.cpp
@@ -9,12 +9,12 @@ void RestartObject::deserialize(const std::string &filename, std::shared_ptr<Par
{
deserialize_internal(filename);
- for (int j = para->getCoarse(); j <= para->getFine(); j++) {
+ for (int index1 = para->getCoarse(); index1 <= para->getFine(); index1++) {
std::vector<real> vec;
fs.push_back(vec);
- for (unsigned int i = 0; i < (para->getD3Qxx() * para->getParH(j)->numberOfNodes); i++) {
- para->getParH(j)->distributions.f[0][i] = fs[j][i];
+ for (size_t index2 = 0; index2 < (para->getD3Qxx() * para->getParH(index1)->numberOfNodes); index2++) {
+ para->getParH(index1)->distributions.f[0][index2] = fs[index1][index2];
}
}
}
@@ -24,15 +24,15 @@ void RestartObject::serialize(const std::string &filename, const std::shared_ptr
if (fs.size() > 0) {
clear(para);
}
- for (int j = para->getCoarse(); j <= para->getFine(); j++) {
+ for (int index1 = para->getCoarse(); index1 <= para->getFine(); index1++) {
std::vector<real> vec;
fs.push_back(vec);
- for (unsigned int i = 0; i < (para->getD3Qxx() * para->getParH(j)->numberOfNodes); i++) {
- if (UbMath::isNaN(para->getParH(j)->distributions.f[0][i])) {
- fs[j].push_back((real)0.0);
+ for (size_t index2 = 0; index2 < (para->getD3Qxx() * para->getParH(index1)->numberOfNodes); index2++) {
+ if (UbMath::isNaN(para->getParH(index1)->distributions.f[0][index2])) {
+ fs[index1].push_back((real)0.0);
} else {
- fs[j].push_back(para->getParH(j)->distributions.f[0][i]);
+ fs[index1].push_back(para->getParH(index1)->distributions.f[0][index2]);
}
}
}
diff --git a/src/lbm/CMakeLists.txt b/src/lbm/CMakeLists.txt
index afa90bdd3f95bb71cf7f1eda6407f9b38766072a..7a9a96ace1c7377b7ad0c67937464d1f2c00cce6 100644
--- a/src/lbm/CMakeLists.txt
+++ b/src/lbm/CMakeLists.txt
@@ -1,12 +1,12 @@
-if(BUILD_VF_CPU)
- project(lbm LANGUAGES CXX)
- vf_add_library(NAME lbm PUBLIC_LINK basics)
- target_link_libraries(lbm PRIVATE project_warnings)
-
- vf_add_tests()
-endif()
+vf_add_library(PUBLIC_LINK basics)
if(BUILD_VF_GPU OR BUILD_VF_GKS)
- add_subdirectory(cuda)
+ set_target_properties(lbm PROPERTIES CUDA_SEPARABLE_COMPILATION ON POSITION_INDEPENDENT_CODE ON)
+
+ set_source_files_properties(KernelParameter.cpp PROPERTIES LANGUAGE CUDA)
+ set_source_files_properties(CumulantChimera.cpp PROPERTIES LANGUAGE CUDA)
+ set_source_files_properties(BGK.cpp PROPERTIES LANGUAGE CUDA)
endif()
+
+vf_add_tests()
\ No newline at end of file
diff --git a/src/lbm/KernelParameter.cpp b/src/lbm/KernelParameter.cpp
index e039214d218ef19f35e8adf927f36d3a6f1aa355..7bf5a369d0e5d4e673d79dcb30bc22fc2c330e68 100644
--- a/src/lbm/KernelParameter.cpp
+++ b/src/lbm/KernelParameter.cpp
@@ -5,11 +5,8 @@
#include "MacroscopicQuantities.h"
-namespace vf
+namespace vf::lbm
{
-namespace lbm
-{
-
inline __host__ __device__ real Distribution27::getDensity_() const
@@ -17,8 +14,6 @@ inline __host__ __device__ real Distribution27::getDensity_() const
return getDensity(f);
}
-
-
__host__ __device__ real abs_internal(real value)
{
#ifdef __CUDA_ARCH__
@@ -30,4 +25,3 @@ __host__ __device__ real abs_internal(real value)
}
-}
diff --git a/src/lbm/KernelParameter.h b/src/lbm/KernelParameter.h
index 95226628110637f3794c8a1f7e6f6c1f6dda937b..18c4f2a4b20b84d9d519993f3ddb54cf612d4306 100644
--- a/src/lbm/KernelParameter.h
+++ b/src/lbm/KernelParameter.h
@@ -11,9 +11,7 @@
#include <basics/Core/DataTypes.h>
-namespace vf
-{
-namespace lbm
+namespace vf::lbm
{
struct Distribution27
@@ -35,9 +33,6 @@ struct KernelParameter
};
-
-
-}
}
#endif
diff --git a/src/lbm/constants/D3Q27.h b/src/lbm/constants/D3Q27.h
index f923695f3756712748e638aee59121a5537456c5..c799331815ff92b41b3daf8433bcc10d026a8738 100644
--- a/src/lbm/constants/D3Q27.h
+++ b/src/lbm/constants/D3Q27.h
@@ -6,92 +6,92 @@
namespace vf::lbm::dir
{
- //real, double und float auf real
-static constexpr int STARTDIR = 0;
-static constexpr int ENDDIR = 26;
+
+static constexpr size_t STARTDIR = 0;
+static constexpr size_t ENDDIR = 26;
// used in the CPU and the GPU version
-static constexpr int DIR_000 = 0;
-static constexpr int DIR_P00 = 1;
-static constexpr int DIR_M00 = 2;
-static constexpr int DIR_0P0 = 3;
-static constexpr int DIR_0M0 = 4;
-static constexpr int DIR_00P = 5;
-static constexpr int DIR_00M = 6;
-static constexpr int DIR_PP0 = 7;
-static constexpr int DIR_MM0 = 8;
-static constexpr int DIR_PM0 = 9;
-static constexpr int DIR_MP0 = 10;
-static constexpr int DIR_P0P = 11;
-static constexpr int DIR_M0M = 12;
-static constexpr int DIR_P0M = 13;
-static constexpr int DIR_M0P = 14;
-static constexpr int DIR_0PP = 15;
-static constexpr int DIR_0MM = 16;
-static constexpr int DIR_0PM = 17;
-static constexpr int DIR_0MP = 18;
-static constexpr int DIR_PPP = 19;
-static constexpr int DIR_MPP = 20;
-static constexpr int DIR_PMP = 21;
-static constexpr int DIR_MMP = 22;
-static constexpr int DIR_PPM = 23;
-static constexpr int DIR_MPM = 24;
-static constexpr int DIR_PMM = 25;
-static constexpr int DIR_MMM = 26;
-
-static constexpr int INV_P00 = DIR_M00;
-static constexpr int INV_M00 = DIR_P00;
-static constexpr int INV_0P0 = DIR_0M0;
-static constexpr int INV_0M0 = DIR_0P0;
-static constexpr int INV_00P = DIR_00M;
-static constexpr int INV_00M = DIR_00P;
-static constexpr int INV_PP0 = DIR_MM0;
-static constexpr int INV_MM0 = DIR_PP0;
-static constexpr int INV_PM0 = DIR_MP0;
-static constexpr int INV_MP0 = DIR_PM0;
-static constexpr int INV_P0P = DIR_M0M;
-static constexpr int INV_M0M = DIR_P0P;
-static constexpr int INV_P0M = DIR_M0P;
-static constexpr int INV_M0P = DIR_P0M;
-static constexpr int INV_0PP = DIR_0MM;
-static constexpr int INV_0MM = DIR_0PP;
-static constexpr int INV_0PM = DIR_0MP;
-static constexpr int INV_0MP = DIR_0PM;
-static constexpr int INV_PPP = DIR_MMM;
-static constexpr int INV_MPP = DIR_PMM;
-static constexpr int INV_PMP = DIR_MPM;
-static constexpr int INV_MMP = DIR_PPM;
-static constexpr int INV_PPM = DIR_MMP;
-static constexpr int INV_MPM = DIR_PMP;
-static constexpr int INV_PMM = DIR_MPP;
-static constexpr int INV_MMM = DIR_PPP;
-
-static constexpr int SGD_P00 = 0;
-static constexpr int SGD_M00 = 1;
-static constexpr int SGD_0P0 = 2;
-static constexpr int SGD_0M0 = 3;
-static constexpr int SGD_00P = 4;
-static constexpr int SGD_00M = 5;
-static constexpr int SGD_PP0 = 6;
-static constexpr int SGD_MM0 = 7;
-static constexpr int SGD_PM0 = 8;
-static constexpr int SGD_MP0 = 9;
-static constexpr int SGD_P0P = 10;
-static constexpr int SGD_M0M = 11;
-static constexpr int SGD_P0M = 12;
-static constexpr int SGD_M0P = 13;
-static constexpr int SGD_0PP = 14;
-static constexpr int SGD_0MM = 15;
-static constexpr int SGD_0PM = 16;
-static constexpr int SGD_0MP = 17;
-static constexpr int SGD_PPP = 18;
-static constexpr int SGD_MPP = 19;
-static constexpr int SGD_PMP = 20;
-static constexpr int SGD_MMP = 21;
-static constexpr int SGD_PPM = 22;
-static constexpr int SGD_MPM = 23;
-static constexpr int SGD_PMM = 24;
-static constexpr int SGD_MMM = 25;
+static constexpr size_t DIR_000 = 0;
+static constexpr size_t DIR_P00 = 1;
+static constexpr size_t DIR_M00 = 2;
+static constexpr size_t DIR_0P0 = 3;
+static constexpr size_t DIR_0M0 = 4;
+static constexpr size_t DIR_00P = 5;
+static constexpr size_t DIR_00M = 6;
+static constexpr size_t DIR_PP0 = 7;
+static constexpr size_t DIR_MM0 = 8;
+static constexpr size_t DIR_PM0 = 9;
+static constexpr size_t DIR_MP0 = 10;
+static constexpr size_t DIR_P0P = 11;
+static constexpr size_t DIR_M0M = 12;
+static constexpr size_t DIR_P0M = 13;
+static constexpr size_t DIR_M0P = 14;
+static constexpr size_t DIR_0PP = 15;
+static constexpr size_t DIR_0MM = 16;
+static constexpr size_t DIR_0PM = 17;
+static constexpr size_t DIR_0MP = 18;
+static constexpr size_t DIR_PPP = 19;
+static constexpr size_t DIR_MPP = 20;
+static constexpr size_t DIR_PMP = 21;
+static constexpr size_t DIR_MMP = 22;
+static constexpr size_t DIR_PPM = 23;
+static constexpr size_t DIR_MPM = 24;
+static constexpr size_t DIR_PMM = 25;
+static constexpr size_t DIR_MMM = 26;
+
+static constexpr size_t INV_P00 = DIR_M00;
+static constexpr size_t INV_M00 = DIR_P00;
+static constexpr size_t INV_0P0 = DIR_0M0;
+static constexpr size_t INV_0M0 = DIR_0P0;
+static constexpr size_t INV_00P = DIR_00M;
+static constexpr size_t INV_00M = DIR_00P;
+static constexpr size_t INV_PP0 = DIR_MM0;
+static constexpr size_t INV_MM0 = DIR_PP0;
+static constexpr size_t INV_PM0 = DIR_MP0;
+static constexpr size_t INV_MP0 = DIR_PM0;
+static constexpr size_t INV_P0P = DIR_M0M;
+static constexpr size_t INV_M0M = DIR_P0P;
+static constexpr size_t INV_P0M = DIR_M0P;
+static constexpr size_t INV_M0P = DIR_P0M;
+static constexpr size_t INV_0PP = DIR_0MM;
+static constexpr size_t INV_0MM = DIR_0PP;
+static constexpr size_t INV_0PM = DIR_0MP;
+static constexpr size_t INV_0MP = DIR_0PM;
+static constexpr size_t INV_PPP = DIR_MMM;
+static constexpr size_t INV_MPP = DIR_PMM;
+static constexpr size_t INV_PMP = DIR_MPM;
+static constexpr size_t INV_MMP = DIR_PPM;
+static constexpr size_t INV_PPM = DIR_MMP;
+static constexpr size_t INV_MPM = DIR_PMP;
+static constexpr size_t INV_PMM = DIR_MPP;
+static constexpr size_t INV_MMM = DIR_PPP;
+
+static constexpr size_t SGD_P00 = 0;
+static constexpr size_t SGD_M00 = 1;
+static constexpr size_t SGD_0P0 = 2;
+static constexpr size_t SGD_0M0 = 3;
+static constexpr size_t SGD_00P = 4;
+static constexpr size_t SGD_00M = 5;
+static constexpr size_t SGD_PP0 = 6;
+static constexpr size_t SGD_MM0 = 7;
+static constexpr size_t SGD_PM0 = 8;
+static constexpr size_t SGD_MP0 = 9;
+static constexpr size_t SGD_P0P = 10;
+static constexpr size_t SGD_M0M = 11;
+static constexpr size_t SGD_P0M = 12;
+static constexpr size_t SGD_M0P = 13;
+static constexpr size_t SGD_0PP = 14;
+static constexpr size_t SGD_0MM = 15;
+static constexpr size_t SGD_0PM = 16;
+static constexpr size_t SGD_0MP = 17;
+static constexpr size_t SGD_PPP = 18;
+static constexpr size_t SGD_MPP = 19;
+static constexpr size_t SGD_PMP = 20;
+static constexpr size_t SGD_MMP = 21;
+static constexpr size_t SGD_PPM = 22;
+static constexpr size_t SGD_MPM = 23;
+static constexpr size_t SGD_PMM = 24;
+static constexpr size_t SGD_MMM = 25;
struct countersForPointerChasing{
uint counterInverse;
@@ -100,7 +100,7 @@ struct countersForPointerChasing{
uint counterZ;
};
-const std::map<const int, const countersForPointerChasing> mapForPointerChasing =
+const std::map<const size_t, const countersForPointerChasing> mapForPointerChasing =
{
{DIR_000, countersForPointerChasing{0, 0, 0, 0}},
{DIR_P00, countersForPointerChasing{0, 1, 0, 0}},
diff --git a/src/lbm/constants/NumericConstants.h b/src/lbm/constants/NumericConstants.h
index fb7764255201dbd31ac1134e756fc4bfd6e3d982..e642c2c95171927a7f8dc8f1a911d98117af66a6 100644
--- a/src/lbm/constants/NumericConstants.h
+++ b/src/lbm/constants/NumericConstants.h
@@ -18,6 +18,7 @@ static constexpr double c1o8 = 0.125;
static constexpr double c1o9 = 0.111111111111111;
static constexpr double c2o9 = 0.222222222222222;
static constexpr double c4o9 = 0.444444444444444;
+static constexpr double c4o10 = 0.4;
static constexpr double c1o10 = 0.1;
static constexpr double c1o12 = 0.083333333333333;
static constexpr double c1o16 = 0.0625;
@@ -48,6 +49,7 @@ static constexpr double c99o100 = 0.99;
static constexpr double c1o126 = 0.007936507936508;
static constexpr double c1o216 = 0.004629629629630;
static constexpr double c5o4 = 1.25;
+static constexpr double c4o3 = 1.333333333333333;
static constexpr double c9o4 = 2.25;
static constexpr double c5o2 = 2.5;
static constexpr double c9o2 = 4.5;
@@ -99,15 +101,15 @@ static constexpr double c72o1 = 72.;
static constexpr double c84o1 = 84.;
static constexpr double c88o1 = 88.;
static constexpr double c96o1 = 96.;
-static constexpr double c100o1 = 10.;
-static constexpr double c130o1 = 13.;
-static constexpr double c152o1 = 15.;
-static constexpr double c166o1 = 16.;
-static constexpr double c195o1 = 19.;
-static constexpr double c216o1 = 21.;
-static constexpr double c264o1 = 26.;
-static constexpr double c290o1 = 29.;
-static constexpr double c367o1 = 36.;
+static constexpr double c100o1 = 100.;
+static constexpr double c130o1 = 130.;
+static constexpr double c152o1 = 152.;
+static constexpr double c166o1 = 166.;
+static constexpr double c195o1 = 195.;
+static constexpr double c216o1 = 216.;
+static constexpr double c264o1 = 264.;
+static constexpr double c290o1 = 290.;
+static constexpr double c367o1 = 367.;
static constexpr double Op0000002 = 0.0000002;
static constexpr double c10eM30 = 1e-30;
@@ -137,6 +139,7 @@ static constexpr float c1o8 = 0.125f;
static constexpr float c1o9 = (1.0f / 9.0f);
static constexpr float c2o9 = (2.0f / 9.0f);
static constexpr float c4o9 = (4.0f / 9.0f);
+static constexpr float c4o10 = 0.4f;
static constexpr float c1o10 = 0.1f;
static constexpr float c1o12 = (1.0f / 12.0f);
static constexpr float c1o16 = 0.0625f;
@@ -167,6 +170,7 @@ static constexpr float c99o100 = 0.99f;
static constexpr float c1o126 = (1.0f / 126.0f);
static constexpr float c1o216 = (1.0f / 216.0f);
static constexpr float c5o4 = 1.25f;
+static constexpr float c4o3 = (4.0f / 3.0f);
static constexpr float c9o4 = 2.25f;
static constexpr float c5o2 = 2.5f;
static constexpr float c9o2 = 4.5f;
diff --git a/src/lbm/cuda/CMakeLists.txt b/src/lbm/cuda/CMakeLists.txt
deleted file mode 100644
index 4142b7c3b1c46275c3257e3dfd657cc6b30c841d..0000000000000000000000000000000000000000
--- a/src/lbm/cuda/CMakeLists.txt
+++ /dev/null
@@ -1,12 +0,0 @@
-project(lbmCuda LANGUAGES CUDA CXX)
-
-
-vf_add_library(NAME lbmCuda BUILDTYPE static PUBLIC_LINK basics FOLDER ../../lbm)
-
-
-set_target_properties(lbmCuda PROPERTIES CUDA_SEPARABLE_COMPILATION ON POSITION_INDEPENDENT_CODE ON)
-
-
-set_source_files_properties(../KernelParameter.cpp PROPERTIES LANGUAGE CUDA)
-set_source_files_properties(../CumulantChimera.cpp PROPERTIES LANGUAGE CUDA)
-set_source_files_properties(../BGK.cpp PROPERTIES LANGUAGE CUDA)
diff --git a/utilities/setup_builder.py b/utilities/setup_builder.py
new file mode 100644
index 0000000000000000000000000000000000000000..94e4f805b3e05024742ead72e3ffb28c477c282f
--- /dev/null
+++ b/utilities/setup_builder.py
@@ -0,0 +1,34 @@
+from setuptools import build_meta
+
+class builder(build_meta._BuildMetaBackend):
+
+ def run_setup(self, setup_script='setup.py'):
+ # Note that we can reuse our build directory between calls
+ # Correctness comes first, then optimization later
+ __file__ = setup_script
+ __name__ = '__main__'
+
+ with build_meta._open_setup_script(__file__) as f:
+ code = f.read().replace(r'\r\n', r'\n')
+ args = locals()
+ args["config_args"] = self.extra_args
+ exec(code, args)
+
+
+ def add_settings(self, config_settings):
+ self.extra_args = dict()
+ if config_settings:
+ self.extra_args = config_settings
+
+ def build_wheel(self, wheel_directory, config_settings=None,
+ metadata_directory=None):
+ self.add_settings(config_settings)
+ return super().build_wheel(wheel_directory, config_settings, metadata_directory)
+
+ def build_sdist(self, sdist_directory, config_settings=None):
+ self.add_settings(config_settings)
+ return super().build_wheel(sdist_directory, config_settings)
+
+build = builder()
+build_wheel = build.build_wheel
+build_sdist = build.build_sdist
\ No newline at end of file