Warnings about initialisation of particles with locations outside of the
mesh and about the positional inaccuracy of NCC transfers are now
accumulated and printed once per time-step. This way, the log isn't
obscured by hundreds of such warnings.
Also, the pattern in which warnings are silenced after some arbitrary
number (typically 100) have been issued has been removed. This pattern
means that user viewing the log later in the run may be unaware that a
problem is still present. Accumulated warnings are concise enough that
they do not need to be silenced. They are generated every time-step, and
so remain visible throughout the log.
Cell-to-cell interpolation has been moved to a hierarchy separate from
meshToMesh, called cellsToCells. The meshToMesh class is now a
combination of a cellsToCells object and multiple patchToPatch objects.
This means that when only cell-to-cell interpolation is needed a basic
cellsToCells object can be selected.
Cell-to-cell and vol-field-to-vol-field interpolation now has two well
defined sets of functions, with a clear distinction in how weights that
do not sum to unity are handled. Non-unity weights are either
normalised, or a left-over field is provided with which to complete the
weighted sum.
The left-over approach is now consistently applied in mapFieldsPar,
across both the internal and patch fields, if mapping onto an existing
field in the target case. Warning are now generated for invalid
combinations of settings, such as mapping between inconsistent meshes
without a pre-existing target field.
All mapping functions now take fields as const references and return tmp
fields. This avoids the pattern in which non-const fields are provided
which relate to the source, and at some point in the function transfer
to the target. This pattern is difficult to reason about and does not
provide any actual computational advantage, as the fields invariably get
re-allocated as part of the process anyway.
MeshToMesh no longer stores the cutting patches. The set of cutting
patches is not needed anywhere except at the point of mapping a field,
so it is now supplied to the mapping functions as an argument.
The meshToMesh topology changer no longer supports cutting patch
information. This did not previously work. Cutting patches either get
generated as calculated, or they require a pre-existing field to specify
their boundary condition. Neither of these options is suitable for a
run-time mesh change.
More code has been shared with patchToPatch, reducing duplication.
The timeName() function simply returns the dimensionedScalar::name() which holds
the user-time name of the current time and now that timeName() is no longer
virtual the dimensionedScalar::name() can be called directly. The timeName()
function implementation is maintained for backward-compatibility.
Lagrangian is now compatible with the meshToMesh topology changer. If a
cloud is being simulated and this topology changer is active, then the
cloud data will be automatically mapped between the specified sequence
of meshes in the same way as the finite volume data. This works both for
serial and parallel simulations.
In addition, mapFieldsPar now also supports mapping of Lagrangian data
when run in parallel.
This reference represents unnecessary storage. The mesh can be obtained
from tracking data or passed to the particle evolution functions by
argument.
In addition, removing the mesh reference makes it possible to construct
as particle from an Istream without the need for an iNew class. This
simplifies stream-based transfer, and makes it possible for particles to
be communicated by a polyDistributionMap.
This reduces duplication and inconsistency between the List, Field, Map,
and PtrList variants. It also allows for future extension to other
container types such as DynamicList.
The template parameters were only ever polyBoundaryMesh and
processorPolyPatch. Un-templating makes mainteance and bug-fixing
quicker as it means minor modifications no longer cause a full rebuild
of OpenFOAM.
The subtract option in mapFieldsPar was not implemented correctly and the
significant complexity in meshToMesh required to support it creates an
unwarranted maintenance overhead. The equivalent functionality is now provided
by the more flexible, convenient and simpler subtract functionObject.
Mesh motion and topology change are now combinable run-time selectable options
within fvMesh, replacing the restrictive dynamicFvMesh which supported only
motion OR topology change.
All solvers which instantiated a dynamicFvMesh now instantiate an fvMesh which
reads the optional constant/dynamicFvMeshDict to construct an fvMeshMover and/or
an fvMeshTopoChanger. These two are specified within the optional mover and
topoChanger sub-dictionaries of dynamicFvMeshDict.
When the fvMesh is updated the fvMeshTopoChanger is first executed which can
change the mesh topology in anyway, adding or removing points as required, for
example for automatic mesh refinement/unrefinement, and all registered fields
are mapped onto the updated mesh. The fvMeshMover is then executed which moved
the points only and calculates the cell volume change and corresponding
mesh-fluxes for conservative moving mesh transport. If multiple topological
changes or movements are required these would be combined into special
fvMeshMovers and fvMeshTopoChangers which handle the processing of a list of
changes, e.g. solidBodyMotionFunctions:multiMotion.
The tutorials/multiphase/interFoam/laminar/sloshingTank3D3DoF case has been
updated to demonstrate this new functionality by combining solid-body motion
with mesh refinement/unrefinement:
/*--------------------------------*- C++ -*----------------------------------*\
========= |
\\ / F ield | OpenFOAM: The Open Source CFD Toolbox
\\ / O peration | Website: https://openfoam.org
\\ / A nd | Version: dev
\\/ M anipulation |
\*---------------------------------------------------------------------------*/
FoamFile
{
format ascii;
class dictionary;
location "constant";
object dynamicMeshDict;
}
// * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * //
mover
{
type motionSolver;
libs ("libfvMeshMovers.so" "libfvMotionSolvers.so");
motionSolver solidBody;
solidBodyMotionFunction SDA;
CofG (0 0 0);
lamda 50;
rollAmax 0.2;
rollAmin 0.1;
heaveA 4;
swayA 2.4;
Q 2;
Tp 14;
Tpn 12;
dTi 0.06;
dTp -0.001;
}
topoChanger
{
type refiner;
libs ("libfvMeshTopoChangers.so");
// How often to refine
refineInterval 1;
// Field to be refinement on
field alpha.water;
// Refine field in between lower..upper
lowerRefineLevel 0.001;
upperRefineLevel 0.999;
// Have slower than 2:1 refinement
nBufferLayers 1;
// Refine cells only up to maxRefinement levels
maxRefinement 1;
// Stop refinement if maxCells reached
maxCells 200000;
// Flux field and corresponding velocity field. Fluxes on changed
// faces get recalculated by interpolating the velocity. Use 'none'
// on surfaceScalarFields that do not need to be reinterpolated.
correctFluxes
(
(phi none)
(nHatf none)
(rhoPhi none)
(alphaPhi.water none)
(meshPhi none)
(meshPhi_0 none)
(ghf none)
);
// Write the refinement level as a volScalarField
dumpLevel true;
}
// ************************************************************************* //
Note that currently this is the only working combination of mesh-motion with
topology change within the new framework and further development is required to
update the set of topology changers so that topology changes with mapping are
separated from the mesh-motion so that they can be combined with any of the
other movements or topology changes in any manner.
All of the solvers and tutorials have been updated to use the new form of
dynamicMeshDict but backward-compatibility was not practical due to the complete
reorganisation of the mesh change structure.
Both AMIMethod and mapNearestMethod are run-time selectable using the standard
OpenFOAM constructor tables, they do not need a separate enumeration-based
selection method which requires duplicate constructors and a lot of other
clutter.
used to check the existence of and open an object file, read and check the
header without constructing the object.
'typeIOobject' operates in an equivalent and consistent manner to 'regIOobject'
but the type information is provided by the template argument rather than via
virtual functions for which the derived object would need to be constructed,
which is the case for 'regIOobject'.
'typeIOobject' replaces the previous separate functions 'typeHeaderOk' and
'typeFilePath' with a single consistent interface.
now all path functions in 'IOobject' are either templated on the type or require a
'globalFile' argument to specify if the type is case global e.g. 'IOdictionary' or
decomposed in parallel, e.g. almost everything else.
The 'global()' and 'globalFile()' virtual functions are now in 'regIOobject'
abstract base-class and overridden as required by derived classes. The path
functions using 'global()' and 'globalFile()' to differentiate between global
and processor local objects are now also in 'regIOobject' rather than 'IOobject'
to ensure the path returned is absolutely consistent with the type.
Unfortunately there is still potential for unexpected IO behaviour inconsistent
with the global/local nature of the type due to the 'fileOperation' classes
searching the processor directory for case global objects before searching the
case directory. This approach appears to be a work-around for incomplete
integration with and rationalisation of 'IOobject' but with the changes above it
is no longer necessary. Unfortunately this "up" searching is baked-in at a low
level and mixed-up with various complex ways to pick the processor directory
name out of the object path and will take some unravelling but this work will
undertaken as time allows.
for
db/functionObjects/timeControl/timeControl.H: timeControls
primitives/bools/Switch/Switch.H: class switchType
primitives/strings/fileName/fileName.H: fileType
primitives/strings/wordRe/wordRe.H: compOption
Tracking data classes are no longer templated on the derived cloud type.
The advantage of this is that they can now be passed to sub models. This
should allow continuous phase data to be removed from the parcel
classes. The disadvantage is that every function which once took a
templated TrackData argument now needs an additional TrackCloudType
argument in order to perform the necessary down-casting.
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.
terms of the local barycentric coordinates of the current tetrahedron,
rather than the global coordinate system.
Barycentric tracking works on any mesh, irrespective of mesh quality.
Particles do not get "lost", and tracking does not require ad-hoc
"corrections" or "rescues" to function robustly, because the calculation
of particle-face intersections is unambiguous and reproducible, even at
small angles of incidence.
Each particle position is defined by topology (i.e. the decomposed tet
cell it is in) and geometry (i.e. where it is in the cell). No search
operations are needed on restart or reconstruct, unlike when particle
positions are stored in the global coordinate system.
The particle positions file now contains particles' local coordinates
and topology, rather than the global coordinates and cell. This change
to the output format is not backwards compatible. Existing cases with
Lagrangian data will not restart, but they will still run from time
zero without any modification. This change was necessary in order to
guarantee that the loaded particle is valid, and therefore
fundamentally prevent "loss" and "search-failure" type bugs (e.g.,
2517, 2442, 2286, 1836, 1461, 1341, 1097).
The tracking functions have also been converted to function in terms
of displacement, rather than end position. This helps remove floating
point error issues, particularly towards the end of a tracking step.
Wall bounded streamlines have been removed. The implementation proved
incompatible with the new tracking algorithm. ParaView has a surface
LIC plugin which provides equivalent, or better, functionality.
Additionally, bug report <https://bugs.openfoam.org/view.php?id=2517>
is resolved by this change.
- patchFields now get mapped (instead of created)
- with -consistent it now maps all patches except for processor ones (they are
the only ones that are processor-local)
- all constraint patches get evaluated after mapping to bring them up to date.
Patch contributed by Mattijs Janssens
