mirror of
https://github.com/OpenFOAM/OpenFOAM-6.git
synced 2025-12-08 06:57:46 +00:00
When an OpenFOAM simulation runs in parallel, the data for decomposed fields and
mesh(es) has historically been stored in multiple files within separate
directories for each processor. Processor directories are named 'processorN',
where N is the processor number.
This commit introduces an alternative "collated" file format where the data for
each decomposed field (and mesh) is collated into a single file, which is
written and read on the master processor. The files are stored in a single
directory named 'processors'.
The new format produces significantly fewer files - one per field, instead of N
per field. For large parallel cases, this avoids the restriction on the number
of open files imposed by the operating system limits.
The file writing can be threaded allowing the simulation to continue running
while the data is being written to file. NFS (Network File System) is not
needed when using the the collated format and additionally, there is an option
to run without NFS with the original uncollated approach, known as
"masterUncollated".
The controls for the file handling are in the OptimisationSwitches of
etc/controlDict:
OptimisationSwitches
{
...
//- Parallel IO file handler
// uncollated (default), collated or masterUncollated
fileHandler uncollated;
//- collated: thread buffer size for queued file writes.
// If set to 0 or not sufficient for the file size threading is not used.
// Default: 2e9
maxThreadFileBufferSize 2e9;
//- masterUncollated: non-blocking buffer size.
// If the file exceeds this buffer size scheduled transfer is used.
// Default: 2e9
maxMasterFileBufferSize 2e9;
}
When using the collated file handling, memory is allocated for the data in the
thread. maxThreadFileBufferSize sets the maximum size of memory in bytes that
is allocated. If the data exceeds this size, the write does not use threading.
When using the masterUncollated file handling, non-blocking MPI communication
requires a sufficiently large memory buffer on the master node.
maxMasterFileBufferSize sets the maximum size in bytes of the buffer. If the
data exceeds this size, the system uses scheduled communication.
The installation defaults for the fileHandler choice, maxThreadFileBufferSize
and maxMasterFileBufferSize (set in etc/controlDict) can be over-ridden within
the case controlDict file, like other parameters. Additionally the fileHandler
can be set by:
- the "-fileHandler" command line argument;
- a FOAM_FILEHANDLER environment variable.
A foamFormatConvert utility allows users to convert files between the collated
and uncollated formats, e.g.
mpirun -np 2 foamFormatConvert -parallel -fileHandler uncollated
An example case demonstrating the file handling methods is provided in:
$FOAM_TUTORIALS/IO/fileHandling
The work was undertaken by Mattijs Janssens, in collaboration with Henry Weller.
1266 lines
45 KiB
C
1266 lines
45 KiB
C
/*---------------------------------------------------------------------------*\
|
|
========= |
|
|
\\ / F ield | OpenFOAM: The Open Source CFD Toolbox
|
|
\\ / O peration |
|
|
\\ / A nd | Copyright (C) 2011-2017 OpenFOAM Foundation
|
|
\\/ M anipulation |
|
|
-------------------------------------------------------------------------------
|
|
License
|
|
This file is part of OpenFOAM.
|
|
|
|
OpenFOAM 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.
|
|
|
|
OpenFOAM 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 OpenFOAM. If not, see <http://www.gnu.org/licenses/>.
|
|
|
|
Application
|
|
decomposePar
|
|
|
|
Description
|
|
Automatically decomposes a mesh and fields of a case for parallel
|
|
execution of OpenFOAM.
|
|
|
|
Usage
|
|
\b decomposePar [OPTION]
|
|
|
|
Options:
|
|
- \par -cellDist
|
|
Write the cell distribution as a labelList, for use with 'manual'
|
|
decomposition method or as a volScalarField for post-processing.
|
|
|
|
- \par -region \<regionName\> \n
|
|
Decompose named region. Does not check for existence of processor*.
|
|
|
|
- \par -allRegions \n
|
|
Decompose all regions in regionProperties. Does not check for
|
|
existence of processor*.
|
|
|
|
- \par -copyZero \n
|
|
Copy \a 0 directory to processor* rather than decompose the fields.
|
|
|
|
- \par -copyUniform \n
|
|
Copy any \a uniform directories too.
|
|
|
|
- \par -constant
|
|
|
|
- \par -time xxx:yyy \n
|
|
Override controlDict settings and decompose selected times. Does not
|
|
re-decompose the mesh i.e. does not handle moving mesh or changing
|
|
mesh cases.
|
|
|
|
- \par -fields \n
|
|
Use existing geometry decomposition and convert fields only.
|
|
|
|
- \par -noSets \n
|
|
Skip decomposing cellSets, faceSets, pointSets.
|
|
|
|
- \par -force \n
|
|
Remove any existing \a processor subdirectories before decomposing the
|
|
geometry.
|
|
|
|
- \par -ifRequired \n
|
|
Only decompose the geometry if the number of domains has changed from a
|
|
previous decomposition. No \a processor subdirectories will be removed
|
|
unless the \a -force option is also specified. This option can be used
|
|
to avoid redundant geometry decomposition (eg, in scripts), but should
|
|
be used with caution when the underlying (serial) geometry or the
|
|
decomposition method etc. have been changed between decompositions.
|
|
|
|
- \par -dict \<filename\>
|
|
Specify alternative dictionary for the decomposition.
|
|
|
|
\*---------------------------------------------------------------------------*/
|
|
|
|
#include "OSspecific.H"
|
|
#include "fvCFD.H"
|
|
#include "IOobjectList.H"
|
|
#include "domainDecomposition.H"
|
|
#include "labelIOField.H"
|
|
#include "labelFieldIOField.H"
|
|
#include "scalarIOField.H"
|
|
#include "scalarFieldIOField.H"
|
|
#include "vectorIOField.H"
|
|
#include "vectorFieldIOField.H"
|
|
#include "sphericalTensorIOField.H"
|
|
#include "sphericalTensorFieldIOField.H"
|
|
#include "symmTensorIOField.H"
|
|
#include "symmTensorFieldIOField.H"
|
|
#include "tensorIOField.H"
|
|
#include "tensorFieldIOField.H"
|
|
#include "pointFields.H"
|
|
#include "regionProperties.H"
|
|
|
|
#include "readFields.H"
|
|
#include "dimFieldDecomposer.H"
|
|
#include "fvFieldDecomposer.H"
|
|
#include "pointFieldDecomposer.H"
|
|
#include "lagrangianFieldDecomposer.H"
|
|
#include "decompositionModel.H"
|
|
#include "collatedFileOperation.H"
|
|
|
|
// * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * //
|
|
|
|
namespace Foam
|
|
{
|
|
|
|
const labelIOList& procAddressing
|
|
(
|
|
const PtrList<fvMesh>& procMeshList,
|
|
const label proci,
|
|
const word& name,
|
|
PtrList<labelIOList>& procAddressingList
|
|
)
|
|
{
|
|
const fvMesh& procMesh = procMeshList[proci];
|
|
|
|
if (!procAddressingList.set(proci))
|
|
{
|
|
procAddressingList.set
|
|
(
|
|
proci,
|
|
new labelIOList
|
|
(
|
|
IOobject
|
|
(
|
|
name,
|
|
procMesh.facesInstance(),
|
|
procMesh.meshSubDir,
|
|
procMesh,
|
|
IOobject::MUST_READ,
|
|
IOobject::NO_WRITE,
|
|
false
|
|
)
|
|
)
|
|
);
|
|
}
|
|
return procAddressingList[proci];
|
|
}
|
|
|
|
|
|
void decomposeUniform
|
|
(
|
|
const bool copyUniform,
|
|
const domainDecomposition& mesh,
|
|
const Time& processorDb,
|
|
const word& regionDir = word::null
|
|
)
|
|
{
|
|
const Time& runTime = mesh.time();
|
|
|
|
// Any uniform data to copy/link?
|
|
const fileName uniformDir(regionDir/"uniform");
|
|
|
|
if (fileHandler().isDir(runTime.timePath()/uniformDir))
|
|
{
|
|
Info<< "Detected additional non-decomposed files in "
|
|
<< runTime.timePath()/uniformDir
|
|
<< endl;
|
|
|
|
const fileName timePath =
|
|
fileHandler().filePath(processorDb.timePath());
|
|
|
|
if (copyUniform || mesh.distributed())
|
|
{
|
|
if (!fileHandler().exists(timePath/uniformDir))
|
|
{
|
|
fileHandler().cp
|
|
(
|
|
runTime.timePath()/uniformDir,
|
|
timePath/uniformDir
|
|
);
|
|
}
|
|
}
|
|
else
|
|
{
|
|
// link with relative paths
|
|
string parentPath = string("..")/"..";
|
|
|
|
if (regionDir != word::null)
|
|
{
|
|
parentPath = parentPath/"..";
|
|
}
|
|
|
|
fileName currentDir(cwd());
|
|
chDir(timePath);
|
|
|
|
if (!fileHandler().exists(uniformDir))
|
|
{
|
|
fileHandler().ln
|
|
(
|
|
parentPath/runTime.timeName()/uniformDir,
|
|
uniformDir
|
|
);
|
|
}
|
|
chDir(currentDir);
|
|
}
|
|
}
|
|
}
|
|
|
|
}
|
|
|
|
|
|
// * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * //
|
|
|
|
int main(int argc, char *argv[])
|
|
{
|
|
argList::addNote
|
|
(
|
|
"decompose a mesh and fields of a case for parallel execution"
|
|
);
|
|
|
|
argList::noParallel();
|
|
#include "addRegionOption.H"
|
|
argList::addBoolOption
|
|
(
|
|
"allRegions",
|
|
"operate on all regions in regionProperties"
|
|
);
|
|
argList::addBoolOption
|
|
(
|
|
"cellDist",
|
|
"write cell distribution as a labelList - for use with 'manual' "
|
|
"decomposition method or as a volScalarField for post-processing."
|
|
);
|
|
argList::addBoolOption
|
|
(
|
|
"copyZero",
|
|
"Copy \a 0 directory to processor* rather than decompose the fields"
|
|
);
|
|
argList::addBoolOption
|
|
(
|
|
"copyUniform",
|
|
"copy any uniform/ directories too"
|
|
);
|
|
argList::addBoolOption
|
|
(
|
|
"fields",
|
|
"use existing geometry decomposition and convert fields only"
|
|
);
|
|
argList::addBoolOption
|
|
(
|
|
"noSets",
|
|
"skip decomposing cellSets, faceSets, pointSets"
|
|
);
|
|
argList::addBoolOption
|
|
(
|
|
"force",
|
|
"remove existing processor*/ subdirs before decomposing the geometry"
|
|
);
|
|
argList::addBoolOption
|
|
(
|
|
"ifRequired",
|
|
"only decompose geometry if the number of domains has changed"
|
|
);
|
|
|
|
argList::addOption
|
|
(
|
|
"dict",
|
|
"dictionary file name",
|
|
"specify alternative decomposition dictionary"
|
|
);
|
|
|
|
// Include explicit constant options, have zero from time range
|
|
timeSelector::addOptions(true, false);
|
|
|
|
#include "setRootCase.H"
|
|
|
|
bool allRegions = args.optionFound("allRegions");
|
|
bool writeCellDist = args.optionFound("cellDist");
|
|
bool copyZero = args.optionFound("copyZero");
|
|
bool copyUniform = args.optionFound("copyUniform");
|
|
bool decomposeFieldsOnly = args.optionFound("fields");
|
|
bool decomposeSets = !args.optionFound("noSets");
|
|
bool forceOverwrite = args.optionFound("force");
|
|
bool ifRequiredDecomposition = args.optionFound("ifRequired");
|
|
|
|
const word dictName("decomposeParDict");
|
|
|
|
// Set time from database
|
|
#include "createTime.H"
|
|
|
|
fileName dictPath;
|
|
|
|
// Check if the dictionary is specified on the command-line
|
|
if (args.optionFound("dict"))
|
|
{
|
|
dictPath = args["dict"];
|
|
|
|
dictPath =
|
|
(
|
|
isDir(dictPath)
|
|
? dictPath/dictName
|
|
: dictPath
|
|
);
|
|
}
|
|
else
|
|
{
|
|
dictPath = runTime.path()/"system"/dictName;
|
|
}
|
|
|
|
// Allow override of time
|
|
instantList times = timeSelector::selectIfPresent(runTime, args);
|
|
|
|
wordList regionNames;
|
|
wordList regionDirs;
|
|
if (allRegions)
|
|
{
|
|
Info<< "Decomposing all regions in regionProperties" << nl << endl;
|
|
regionProperties rp(runTime);
|
|
forAllConstIter(HashTable<wordList>, rp, iter)
|
|
{
|
|
const wordList& regions = iter();
|
|
forAll(regions, i)
|
|
{
|
|
if (findIndex(regionNames, regions[i]) == -1)
|
|
{
|
|
regionNames.append(regions[i]);
|
|
}
|
|
}
|
|
}
|
|
regionDirs = regionNames;
|
|
}
|
|
else
|
|
{
|
|
word regionName;
|
|
if (args.optionReadIfPresent("region", regionName))
|
|
{
|
|
regionNames = wordList(1, regionName);
|
|
regionDirs = regionNames;
|
|
}
|
|
else
|
|
{
|
|
regionNames = wordList(1, fvMesh::defaultRegion);
|
|
regionDirs = wordList(1, word::null);
|
|
}
|
|
}
|
|
|
|
|
|
|
|
forAll(regionNames, regioni)
|
|
{
|
|
const word& regionName = regionNames[regioni];
|
|
const word& regionDir = regionDirs[regioni];
|
|
|
|
Info<< "\n\nDecomposing mesh " << regionName << nl << endl;
|
|
|
|
|
|
// Determine the existing processor count directly
|
|
label nProcs = fileHandler().nProcs(runTime.path(), regionDir);
|
|
|
|
// Get requested numberOfSubdomains. Note: have no mesh yet so
|
|
// cannot use decompositionModel::New
|
|
const label nDomains = readLabel
|
|
(
|
|
IOdictionary
|
|
(
|
|
IOobject
|
|
(
|
|
dictName,
|
|
runTime.time().system(),
|
|
regionDir, // use region if non-standard
|
|
runTime,
|
|
IOobject::MUST_READ_IF_MODIFIED,
|
|
IOobject::NO_WRITE,
|
|
false
|
|
)
|
|
).lookup("numberOfSubdomains")
|
|
);
|
|
|
|
if (decomposeFieldsOnly)
|
|
{
|
|
// Sanity check on previously decomposed case
|
|
if (nProcs != nDomains)
|
|
{
|
|
FatalErrorInFunction
|
|
<< "Specified -fields, but the case was decomposed with "
|
|
<< nProcs << " domains"
|
|
<< nl
|
|
<< "instead of " << nDomains
|
|
<< " domains as specified in " << dictName
|
|
<< nl
|
|
<< exit(FatalError);
|
|
}
|
|
}
|
|
else if (nProcs)
|
|
{
|
|
bool procDirsProblem = true;
|
|
|
|
if (ifRequiredDecomposition && nProcs == nDomains)
|
|
{
|
|
// we can reuse the decomposition
|
|
decomposeFieldsOnly = true;
|
|
procDirsProblem = false;
|
|
forceOverwrite = false;
|
|
|
|
Info<< "Using existing processor directories" << nl;
|
|
}
|
|
|
|
if (forceOverwrite)
|
|
{
|
|
Info<< "Removing " << nProcs
|
|
<< " existing processor directories" << endl;
|
|
|
|
fileHandler().rmDir(runTime.path()/word("processors"));
|
|
|
|
// remove existing processor dirs
|
|
// reverse order to avoid gaps if someone interrupts the process
|
|
for (label proci = nProcs-1; proci >= 0; --proci)
|
|
{
|
|
fileName procDir
|
|
(
|
|
runTime.path()/(word("processor") + name(proci))
|
|
);
|
|
|
|
fileHandler().rmDir(procDir);
|
|
}
|
|
|
|
procDirsProblem = false;
|
|
}
|
|
|
|
if (procDirsProblem)
|
|
{
|
|
FatalErrorInFunction
|
|
<< "Case is already decomposed with " << nProcs
|
|
<< " domains, use the -force option or manually" << nl
|
|
<< "remove processor directories before decomposing. e.g.,"
|
|
<< nl
|
|
<< " rm -rf " << runTime.path().c_str() << "/processor*"
|
|
<< nl
|
|
<< exit(FatalError);
|
|
}
|
|
}
|
|
|
|
Info<< "Create mesh" << endl;
|
|
domainDecomposition mesh
|
|
(
|
|
IOobject
|
|
(
|
|
regionName,
|
|
runTime.timeName(),
|
|
runTime,
|
|
IOobject::NO_READ,
|
|
IOobject::NO_WRITE,
|
|
false
|
|
),
|
|
dictPath
|
|
);
|
|
|
|
// Decompose the mesh
|
|
if (!decomposeFieldsOnly)
|
|
{
|
|
// Disable buffering when writing mesh since we need to read
|
|
// it later on when decomposing the fields
|
|
float bufSz =
|
|
fileOperations::collatedFileOperation::maxThreadFileBufferSize;
|
|
fileOperations::collatedFileOperation::maxThreadFileBufferSize = 0;
|
|
|
|
mesh.decomposeMesh(dictPath);
|
|
|
|
mesh.writeDecomposition(decomposeSets);
|
|
|
|
if (writeCellDist)
|
|
{
|
|
const labelList& procIds = mesh.cellToProc();
|
|
|
|
// Write the decomposition as labelList for use with 'manual'
|
|
// decomposition method.
|
|
labelIOList cellDecomposition
|
|
(
|
|
IOobject
|
|
(
|
|
"cellDecomposition",
|
|
mesh.facesInstance(),
|
|
mesh,
|
|
IOobject::NO_READ,
|
|
IOobject::NO_WRITE,
|
|
false
|
|
),
|
|
procIds
|
|
);
|
|
cellDecomposition.write();
|
|
|
|
Info<< nl << "Wrote decomposition to "
|
|
<< cellDecomposition.objectPath()
|
|
<< " for use in manual decomposition." << endl;
|
|
|
|
// Write as volScalarField for postprocessing.
|
|
volScalarField cellDist
|
|
(
|
|
IOobject
|
|
(
|
|
"cellDist",
|
|
runTime.timeName(),
|
|
mesh,
|
|
IOobject::NO_READ,
|
|
IOobject::AUTO_WRITE
|
|
),
|
|
mesh,
|
|
dimensionedScalar("cellDist", dimless, 0)
|
|
);
|
|
|
|
forAll(procIds, celli)
|
|
{
|
|
cellDist[celli] = procIds[celli];
|
|
}
|
|
|
|
cellDist.write();
|
|
|
|
Info<< nl << "Wrote decomposition as volScalarField to "
|
|
<< cellDist.name() << " for use in postprocessing."
|
|
<< endl;
|
|
}
|
|
|
|
fileOperations::collatedFileOperation::maxThreadFileBufferSize = bufSz;
|
|
}
|
|
|
|
|
|
if (copyZero)
|
|
{
|
|
// Copy the 0 directory into each of the processor directories
|
|
fileName prevTimePath;
|
|
for (label proci = 0; proci < mesh.nProcs(); proci++)
|
|
{
|
|
Time processorDb
|
|
(
|
|
Time::controlDictName,
|
|
args.rootPath(),
|
|
args.caseName()/fileName(word("processor") + name(proci))
|
|
);
|
|
processorDb.setTime(runTime);
|
|
|
|
if (fileHandler().isDir(runTime.timePath()))
|
|
{
|
|
// Get corresponding directory name (to handle processors/)
|
|
const fileName timePath
|
|
(
|
|
fileHandler().objectPath
|
|
(
|
|
IOobject
|
|
(
|
|
"",
|
|
processorDb.timeName(),
|
|
processorDb
|
|
),
|
|
word::null
|
|
)
|
|
);
|
|
|
|
if (timePath != prevTimePath)
|
|
{
|
|
Info<< "Processor " << proci
|
|
<< ": copying " << runTime.timePath() << nl
|
|
<< " to " << timePath << endl;
|
|
fileHandler().cp(runTime.timePath(), timePath);
|
|
|
|
prevTimePath = timePath;
|
|
}
|
|
}
|
|
}
|
|
}
|
|
else
|
|
{
|
|
// Decompose the field files
|
|
|
|
// Cached processor meshes and maps. These are only preserved if
|
|
// running with multiple times.
|
|
PtrList<Time> processorDbList(mesh.nProcs());
|
|
PtrList<fvMesh> procMeshList(mesh.nProcs());
|
|
PtrList<labelIOList> faceProcAddressingList(mesh.nProcs());
|
|
PtrList<labelIOList> cellProcAddressingList(mesh.nProcs());
|
|
PtrList<labelIOList> boundaryProcAddressingList(mesh.nProcs());
|
|
PtrList<fvFieldDecomposer> fieldDecomposerList(mesh.nProcs());
|
|
PtrList<dimFieldDecomposer> dimFieldDecomposerList(mesh.nProcs());
|
|
PtrList<labelIOList> pointProcAddressingList(mesh.nProcs());
|
|
PtrList<pointFieldDecomposer> pointFieldDecomposerList
|
|
(
|
|
mesh.nProcs()
|
|
);
|
|
|
|
|
|
// Loop over all times
|
|
forAll(times, timeI)
|
|
{
|
|
runTime.setTime(times[timeI], timeI);
|
|
|
|
Info<< "Time = " << runTime.timeName() << endl;
|
|
|
|
// Search for list of objects for this time
|
|
IOobjectList objects(mesh, runTime.timeName());
|
|
|
|
|
|
// Construct the vol fields
|
|
// ~~~~~~~~~~~~~~~~~~~~~~~~
|
|
PtrList<volScalarField> volScalarFields;
|
|
readFields(mesh, objects, volScalarFields);
|
|
PtrList<volVectorField> volVectorFields;
|
|
readFields(mesh, objects, volVectorFields);
|
|
PtrList<volSphericalTensorField> volSphericalTensorFields;
|
|
readFields(mesh, objects, volSphericalTensorFields);
|
|
PtrList<volSymmTensorField> volSymmTensorFields;
|
|
readFields(mesh, objects, volSymmTensorFields);
|
|
PtrList<volTensorField> volTensorFields;
|
|
readFields(mesh, objects, volTensorFields);
|
|
|
|
|
|
// Construct the dimensioned fields
|
|
// ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
|
|
PtrList<DimensionedField<scalar, volMesh>> dimScalarFields;
|
|
readFields(mesh, objects, dimScalarFields);
|
|
PtrList<DimensionedField<vector, volMesh>> dimVectorFields;
|
|
readFields(mesh, objects, dimVectorFields);
|
|
PtrList<DimensionedField<sphericalTensor, volMesh>>
|
|
dimSphericalTensorFields;
|
|
readFields(mesh, objects, dimSphericalTensorFields);
|
|
PtrList<DimensionedField<symmTensor, volMesh>>
|
|
dimSymmTensorFields;
|
|
readFields(mesh, objects, dimSymmTensorFields);
|
|
PtrList<DimensionedField<tensor, volMesh>> dimTensorFields;
|
|
readFields(mesh, objects, dimTensorFields);
|
|
|
|
|
|
// Construct the surface fields
|
|
// ~~~~~~~~~~~~~~~~~~~~~~~~~~~~
|
|
PtrList<surfaceScalarField> surfaceScalarFields;
|
|
readFields(mesh, objects, surfaceScalarFields);
|
|
PtrList<surfaceVectorField> surfaceVectorFields;
|
|
readFields(mesh, objects, surfaceVectorFields);
|
|
PtrList<surfaceSphericalTensorField>
|
|
surfaceSphericalTensorFields;
|
|
readFields(mesh, objects, surfaceSphericalTensorFields);
|
|
PtrList<surfaceSymmTensorField> surfaceSymmTensorFields;
|
|
readFields(mesh, objects, surfaceSymmTensorFields);
|
|
PtrList<surfaceTensorField> surfaceTensorFields;
|
|
readFields(mesh, objects, surfaceTensorFields);
|
|
|
|
|
|
// Construct the point fields
|
|
// ~~~~~~~~~~~~~~~~~~~~~~~~~~
|
|
const pointMesh& pMesh = pointMesh::New(mesh);
|
|
|
|
PtrList<pointScalarField> pointScalarFields;
|
|
readFields(pMesh, objects, pointScalarFields);
|
|
PtrList<pointVectorField> pointVectorFields;
|
|
readFields(pMesh, objects, pointVectorFields);
|
|
PtrList<pointSphericalTensorField> pointSphericalTensorFields;
|
|
readFields(pMesh, objects, pointSphericalTensorFields);
|
|
PtrList<pointSymmTensorField> pointSymmTensorFields;
|
|
readFields(pMesh, objects, pointSymmTensorFields);
|
|
PtrList<pointTensorField> pointTensorFields;
|
|
readFields(pMesh, objects, pointTensorFields);
|
|
|
|
|
|
// Construct the Lagrangian fields
|
|
// ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
|
|
|
|
fileNameList cloudDirs
|
|
(
|
|
fileHandler().readDir
|
|
(
|
|
runTime.timePath()/cloud::prefix,
|
|
fileName::DIRECTORY
|
|
)
|
|
);
|
|
|
|
// Particles
|
|
PtrList<Cloud<indexedParticle>> lagrangianPositions
|
|
(
|
|
cloudDirs.size()
|
|
);
|
|
// Particles per cell
|
|
PtrList<List<SLList<indexedParticle*>*>> cellParticles
|
|
(
|
|
cloudDirs.size()
|
|
);
|
|
|
|
PtrList<PtrList<labelIOField>> lagrangianLabelFields
|
|
(
|
|
cloudDirs.size()
|
|
);
|
|
PtrList<PtrList<labelFieldCompactIOField>>
|
|
lagrangianLabelFieldFields
|
|
(
|
|
cloudDirs.size()
|
|
);
|
|
PtrList<PtrList<scalarIOField>> lagrangianScalarFields
|
|
(
|
|
cloudDirs.size()
|
|
);
|
|
PtrList<PtrList<scalarFieldCompactIOField>>
|
|
lagrangianScalarFieldFields
|
|
(
|
|
cloudDirs.size()
|
|
);
|
|
PtrList<PtrList<vectorIOField>> lagrangianVectorFields
|
|
(
|
|
cloudDirs.size()
|
|
);
|
|
PtrList<PtrList<vectorFieldCompactIOField>>
|
|
lagrangianVectorFieldFields
|
|
(
|
|
cloudDirs.size()
|
|
);
|
|
PtrList<PtrList<sphericalTensorIOField>>
|
|
lagrangianSphericalTensorFields
|
|
(
|
|
cloudDirs.size()
|
|
);
|
|
PtrList<PtrList<sphericalTensorFieldCompactIOField>>
|
|
lagrangianSphericalTensorFieldFields(cloudDirs.size());
|
|
PtrList<PtrList<symmTensorIOField>> lagrangianSymmTensorFields
|
|
(
|
|
cloudDirs.size()
|
|
);
|
|
PtrList<PtrList<symmTensorFieldCompactIOField>>
|
|
lagrangianSymmTensorFieldFields
|
|
(
|
|
cloudDirs.size()
|
|
);
|
|
PtrList<PtrList<tensorIOField>> lagrangianTensorFields
|
|
(
|
|
cloudDirs.size()
|
|
);
|
|
PtrList<PtrList<tensorFieldCompactIOField>>
|
|
lagrangianTensorFieldFields
|
|
(
|
|
cloudDirs.size()
|
|
);
|
|
|
|
label cloudI = 0;
|
|
|
|
forAll(cloudDirs, i)
|
|
{
|
|
IOobjectList sprayObjs
|
|
(
|
|
mesh,
|
|
runTime.timeName(),
|
|
cloud::prefix/cloudDirs[i],
|
|
IOobject::MUST_READ,
|
|
IOobject::NO_WRITE,
|
|
false
|
|
);
|
|
|
|
IOobject* positionsPtr = sprayObjs.lookup
|
|
(
|
|
word("positions")
|
|
);
|
|
|
|
if (positionsPtr)
|
|
{
|
|
// Read lagrangian particles
|
|
// ~~~~~~~~~~~~~~~~~~~~~~~~~
|
|
|
|
Info<< "Identified lagrangian data set: "
|
|
<< cloudDirs[i] << endl;
|
|
|
|
lagrangianPositions.set
|
|
(
|
|
cloudI,
|
|
new Cloud<indexedParticle>
|
|
(
|
|
mesh,
|
|
cloudDirs[i],
|
|
false
|
|
)
|
|
);
|
|
|
|
|
|
// Sort particles per cell
|
|
// ~~~~~~~~~~~~~~~~~~~~~~~
|
|
|
|
cellParticles.set
|
|
(
|
|
cloudI,
|
|
new List<SLList<indexedParticle*>*>
|
|
(
|
|
mesh.nCells(),
|
|
static_cast<SLList<indexedParticle*>*>(nullptr)
|
|
)
|
|
);
|
|
|
|
label i = 0;
|
|
|
|
forAllIter
|
|
(
|
|
Cloud<indexedParticle>,
|
|
lagrangianPositions[cloudI],
|
|
iter
|
|
)
|
|
{
|
|
iter().index() = i++;
|
|
|
|
label celli = iter().cell();
|
|
|
|
// Check
|
|
if (celli < 0 || celli >= mesh.nCells())
|
|
{
|
|
FatalErrorInFunction
|
|
<< "Illegal cell number " << celli
|
|
<< " for particle with index "
|
|
<< iter().index()
|
|
<< " at position "
|
|
<< iter().position() << nl
|
|
<< "Cell number should be between 0 and "
|
|
<< mesh.nCells()-1 << nl
|
|
<< "On this mesh the particle should"
|
|
<< " be in cell "
|
|
<< mesh.findCell(iter().position())
|
|
<< exit(FatalError);
|
|
}
|
|
|
|
if (!cellParticles[cloudI][celli])
|
|
{
|
|
cellParticles[cloudI][celli] =
|
|
new SLList<indexedParticle*>();
|
|
}
|
|
|
|
cellParticles[cloudI][celli]->append(&iter());
|
|
}
|
|
|
|
// Read fields
|
|
// ~~~~~~~~~~~
|
|
|
|
IOobjectList lagrangianObjects
|
|
(
|
|
mesh,
|
|
runTime.timeName(),
|
|
cloud::prefix/cloudDirs[cloudI],
|
|
IOobject::MUST_READ,
|
|
IOobject::NO_WRITE,
|
|
false
|
|
);
|
|
|
|
lagrangianFieldDecomposer::readFields
|
|
(
|
|
cloudI,
|
|
lagrangianObjects,
|
|
lagrangianLabelFields
|
|
);
|
|
|
|
lagrangianFieldDecomposer::readFieldFields
|
|
(
|
|
cloudI,
|
|
lagrangianObjects,
|
|
lagrangianLabelFieldFields
|
|
);
|
|
|
|
lagrangianFieldDecomposer::readFields
|
|
(
|
|
cloudI,
|
|
lagrangianObjects,
|
|
lagrangianScalarFields
|
|
);
|
|
|
|
lagrangianFieldDecomposer::readFieldFields
|
|
(
|
|
cloudI,
|
|
lagrangianObjects,
|
|
lagrangianScalarFieldFields
|
|
);
|
|
|
|
lagrangianFieldDecomposer::readFields
|
|
(
|
|
cloudI,
|
|
lagrangianObjects,
|
|
lagrangianVectorFields
|
|
);
|
|
|
|
lagrangianFieldDecomposer::readFieldFields
|
|
(
|
|
cloudI,
|
|
lagrangianObjects,
|
|
lagrangianVectorFieldFields
|
|
);
|
|
|
|
lagrangianFieldDecomposer::readFields
|
|
(
|
|
cloudI,
|
|
lagrangianObjects,
|
|
lagrangianSphericalTensorFields
|
|
);
|
|
|
|
lagrangianFieldDecomposer::readFieldFields
|
|
(
|
|
cloudI,
|
|
lagrangianObjects,
|
|
lagrangianSphericalTensorFieldFields
|
|
);
|
|
|
|
lagrangianFieldDecomposer::readFields
|
|
(
|
|
cloudI,
|
|
lagrangianObjects,
|
|
lagrangianSymmTensorFields
|
|
);
|
|
|
|
lagrangianFieldDecomposer::readFieldFields
|
|
(
|
|
cloudI,
|
|
lagrangianObjects,
|
|
lagrangianSymmTensorFieldFields
|
|
);
|
|
|
|
lagrangianFieldDecomposer::readFields
|
|
(
|
|
cloudI,
|
|
lagrangianObjects,
|
|
lagrangianTensorFields
|
|
);
|
|
|
|
lagrangianFieldDecomposer::readFieldFields
|
|
(
|
|
cloudI,
|
|
lagrangianObjects,
|
|
lagrangianTensorFieldFields
|
|
);
|
|
|
|
cloudI++;
|
|
}
|
|
}
|
|
|
|
lagrangianPositions.setSize(cloudI);
|
|
cellParticles.setSize(cloudI);
|
|
lagrangianLabelFields.setSize(cloudI);
|
|
lagrangianLabelFieldFields.setSize(cloudI);
|
|
lagrangianScalarFields.setSize(cloudI);
|
|
lagrangianScalarFieldFields.setSize(cloudI);
|
|
lagrangianVectorFields.setSize(cloudI);
|
|
lagrangianVectorFieldFields.setSize(cloudI);
|
|
lagrangianSphericalTensorFields.setSize(cloudI);
|
|
lagrangianSphericalTensorFieldFields.setSize(cloudI);
|
|
lagrangianSymmTensorFields.setSize(cloudI);
|
|
lagrangianSymmTensorFieldFields.setSize(cloudI);
|
|
lagrangianTensorFields.setSize(cloudI);
|
|
lagrangianTensorFieldFields.setSize(cloudI);
|
|
|
|
Info<< endl;
|
|
|
|
// split the fields over processors
|
|
for (label proci = 0; proci < mesh.nProcs(); proci++)
|
|
{
|
|
Info<< "Processor " << proci << ": field transfer" << endl;
|
|
|
|
|
|
// open the database
|
|
if (!processorDbList.set(proci))
|
|
{
|
|
processorDbList.set
|
|
(
|
|
proci,
|
|
new Time
|
|
(
|
|
Time::controlDictName,
|
|
args.rootPath(),
|
|
args.caseName()
|
|
/fileName(word("processor") + name(proci))
|
|
)
|
|
);
|
|
}
|
|
Time& processorDb = processorDbList[proci];
|
|
|
|
|
|
processorDb.setTime(runTime);
|
|
|
|
// read the mesh
|
|
if (!procMeshList.set(proci))
|
|
{
|
|
procMeshList.set
|
|
(
|
|
proci,
|
|
new fvMesh
|
|
(
|
|
IOobject
|
|
(
|
|
regionName,
|
|
processorDb.timeName(),
|
|
processorDb
|
|
)
|
|
)
|
|
);
|
|
}
|
|
const fvMesh& procMesh = procMeshList[proci];
|
|
|
|
const labelIOList& faceProcAddressing = procAddressing
|
|
(
|
|
procMeshList,
|
|
proci,
|
|
"faceProcAddressing",
|
|
faceProcAddressingList
|
|
);
|
|
|
|
const labelIOList& cellProcAddressing = procAddressing
|
|
(
|
|
procMeshList,
|
|
proci,
|
|
"cellProcAddressing",
|
|
cellProcAddressingList
|
|
);
|
|
|
|
const labelIOList& boundaryProcAddressing = procAddressing
|
|
(
|
|
procMeshList,
|
|
proci,
|
|
"boundaryProcAddressing",
|
|
boundaryProcAddressingList
|
|
);
|
|
|
|
|
|
// FV fields
|
|
{
|
|
if (!fieldDecomposerList.set(proci))
|
|
{
|
|
fieldDecomposerList.set
|
|
(
|
|
proci,
|
|
new fvFieldDecomposer
|
|
(
|
|
mesh,
|
|
procMesh,
|
|
faceProcAddressing,
|
|
cellProcAddressing,
|
|
boundaryProcAddressing
|
|
)
|
|
);
|
|
}
|
|
const fvFieldDecomposer& fieldDecomposer =
|
|
fieldDecomposerList[proci];
|
|
|
|
fieldDecomposer.decomposeFields(volScalarFields);
|
|
fieldDecomposer.decomposeFields(volVectorFields);
|
|
fieldDecomposer.decomposeFields
|
|
(
|
|
volSphericalTensorFields
|
|
);
|
|
fieldDecomposer.decomposeFields(volSymmTensorFields);
|
|
fieldDecomposer.decomposeFields(volTensorFields);
|
|
|
|
fieldDecomposer.decomposeFields(surfaceScalarFields);
|
|
fieldDecomposer.decomposeFields(surfaceVectorFields);
|
|
fieldDecomposer.decomposeFields
|
|
(
|
|
surfaceSphericalTensorFields
|
|
);
|
|
fieldDecomposer.decomposeFields
|
|
(
|
|
surfaceSymmTensorFields
|
|
);
|
|
fieldDecomposer.decomposeFields(surfaceTensorFields);
|
|
|
|
if (times.size() == 1)
|
|
{
|
|
// Clear cached decomposer
|
|
fieldDecomposerList.set(proci, nullptr);
|
|
}
|
|
}
|
|
|
|
// Dimensioned fields
|
|
{
|
|
if (!dimFieldDecomposerList.set(proci))
|
|
{
|
|
dimFieldDecomposerList.set
|
|
(
|
|
proci,
|
|
new dimFieldDecomposer
|
|
(
|
|
mesh,
|
|
procMesh,
|
|
faceProcAddressing,
|
|
cellProcAddressing
|
|
)
|
|
);
|
|
}
|
|
const dimFieldDecomposer& dimDecomposer =
|
|
dimFieldDecomposerList[proci];
|
|
|
|
dimDecomposer.decomposeFields(dimScalarFields);
|
|
dimDecomposer.decomposeFields(dimVectorFields);
|
|
dimDecomposer.decomposeFields(dimSphericalTensorFields);
|
|
dimDecomposer.decomposeFields(dimSymmTensorFields);
|
|
dimDecomposer.decomposeFields(dimTensorFields);
|
|
|
|
if (times.size() == 1)
|
|
{
|
|
dimFieldDecomposerList.set(proci, nullptr);
|
|
}
|
|
}
|
|
|
|
|
|
// Point fields
|
|
if
|
|
(
|
|
pointScalarFields.size()
|
|
|| pointVectorFields.size()
|
|
|| pointSphericalTensorFields.size()
|
|
|| pointSymmTensorFields.size()
|
|
|| pointTensorFields.size()
|
|
)
|
|
{
|
|
const labelIOList& pointProcAddressing = procAddressing
|
|
(
|
|
procMeshList,
|
|
proci,
|
|
"pointProcAddressing",
|
|
pointProcAddressingList
|
|
);
|
|
|
|
const pointMesh& procPMesh = pointMesh::New(procMesh);
|
|
|
|
if (!pointFieldDecomposerList.set(proci))
|
|
{
|
|
pointFieldDecomposerList.set
|
|
(
|
|
proci,
|
|
new pointFieldDecomposer
|
|
(
|
|
pMesh,
|
|
procPMesh,
|
|
pointProcAddressing,
|
|
boundaryProcAddressing
|
|
)
|
|
);
|
|
}
|
|
const pointFieldDecomposer& pointDecomposer =
|
|
pointFieldDecomposerList[proci];
|
|
|
|
pointDecomposer.decomposeFields(pointScalarFields);
|
|
pointDecomposer.decomposeFields(pointVectorFields);
|
|
pointDecomposer.decomposeFields
|
|
(
|
|
pointSphericalTensorFields
|
|
);
|
|
pointDecomposer.decomposeFields(pointSymmTensorFields);
|
|
pointDecomposer.decomposeFields(pointTensorFields);
|
|
|
|
|
|
if (times.size() == 1)
|
|
{
|
|
pointProcAddressingList.set(proci, nullptr);
|
|
pointFieldDecomposerList.set(proci, nullptr);
|
|
}
|
|
}
|
|
|
|
|
|
// If there is lagrangian data write it out
|
|
forAll(lagrangianPositions, cloudI)
|
|
{
|
|
if (lagrangianPositions[cloudI].size())
|
|
{
|
|
lagrangianFieldDecomposer fieldDecomposer
|
|
(
|
|
mesh,
|
|
procMesh,
|
|
faceProcAddressing,
|
|
cellProcAddressing,
|
|
cloudDirs[cloudI],
|
|
lagrangianPositions[cloudI],
|
|
cellParticles[cloudI]
|
|
);
|
|
|
|
// Lagrangian fields
|
|
{
|
|
fieldDecomposer.decomposeFields
|
|
(
|
|
cloudDirs[cloudI],
|
|
lagrangianLabelFields[cloudI]
|
|
);
|
|
fieldDecomposer.decomposeFieldFields
|
|
(
|
|
cloudDirs[cloudI],
|
|
lagrangianLabelFieldFields[cloudI]
|
|
);
|
|
fieldDecomposer.decomposeFields
|
|
(
|
|
cloudDirs[cloudI],
|
|
lagrangianScalarFields[cloudI]
|
|
);
|
|
fieldDecomposer.decomposeFieldFields
|
|
(
|
|
cloudDirs[cloudI],
|
|
lagrangianScalarFieldFields[cloudI]
|
|
);
|
|
fieldDecomposer.decomposeFields
|
|
(
|
|
cloudDirs[cloudI],
|
|
lagrangianVectorFields[cloudI]
|
|
);
|
|
fieldDecomposer.decomposeFieldFields
|
|
(
|
|
cloudDirs[cloudI],
|
|
lagrangianVectorFieldFields[cloudI]
|
|
);
|
|
fieldDecomposer.decomposeFields
|
|
(
|
|
cloudDirs[cloudI],
|
|
lagrangianSphericalTensorFields[cloudI]
|
|
);
|
|
fieldDecomposer.decomposeFieldFields
|
|
(
|
|
cloudDirs[cloudI],
|
|
lagrangianSphericalTensorFieldFields[cloudI]
|
|
);
|
|
fieldDecomposer.decomposeFields
|
|
(
|
|
cloudDirs[cloudI],
|
|
lagrangianSymmTensorFields[cloudI]
|
|
);
|
|
fieldDecomposer.decomposeFieldFields
|
|
(
|
|
cloudDirs[cloudI],
|
|
lagrangianSymmTensorFieldFields[cloudI]
|
|
);
|
|
fieldDecomposer.decomposeFields
|
|
(
|
|
cloudDirs[cloudI],
|
|
lagrangianTensorFields[cloudI]
|
|
);
|
|
fieldDecomposer.decomposeFieldFields
|
|
(
|
|
cloudDirs[cloudI],
|
|
lagrangianTensorFieldFields[cloudI]
|
|
);
|
|
}
|
|
}
|
|
}
|
|
|
|
// Decompose the "uniform" directory in the time region
|
|
// directory
|
|
decomposeUniform(copyUniform, mesh, processorDb, regionDir);
|
|
|
|
// For the first region of a multi-region case additionally
|
|
// decompose the "uniform" directory in the time directory
|
|
if (regionNames.size() > 1 && regioni == 0)
|
|
{
|
|
decomposeUniform(copyUniform, mesh, processorDb);
|
|
}
|
|
|
|
// We have cached all the constant mesh data for the current
|
|
// processor. This is only important if running with
|
|
// multiple times, otherwise it is just extra storage.
|
|
if (times.size() == 1)
|
|
{
|
|
boundaryProcAddressingList.set(proci, nullptr);
|
|
cellProcAddressingList.set(proci, nullptr);
|
|
faceProcAddressingList.set(proci, nullptr);
|
|
procMeshList.set(proci, nullptr);
|
|
processorDbList.set(proci, nullptr);
|
|
}
|
|
}
|
|
}
|
|
}
|
|
}
|
|
|
|
Info<< "\nEnd\n" << endl;
|
|
|
|
return 0;
|
|
}
|
|
|
|
|
|
// ************************************************************************* //
|