This avoids attempting to write temporary fields before they have been created.
The executeAtStart can be set to 'yes' in the writeObjects dictionary or the
functions
{
#includeFunc writeObjects(executeAtStart = yes, <field1>, <field2>...)
}
This tutorial now serves as an example of how to compute flow-rates
through zones defined by triangulated surfaces.
A small fix has also been added to searchableSurfaceToFaceZone to
improve robustness on ambiguous cases.
This set now produces a contiguous face zone even when cell centres fall
exactly on the plane of the searchable surface.
The set now calculates the intersections between the
cell-centre-to-cell-centre vectors for every face, and uses those to to
categorise the adjacent cells either side of the set as either above or
below the surface. Then, faces which border both an above and a below
cell are included. This guarantees a contiguous result.
Previously faces for which an intersection was found were included
directly, which resulted in a non-contiguous result when the
intersections became ambiguous; i.e., when the surface passed very close
to the cell centres.
Temporary fields returned from functions which are then reused by the calling
function can now be cached by declaring them as non-reusable to guarantee their
integrity. *Field::New functions have been updated to use this new
functionality and it is recommended to use these functions to create temporary
fields rather than "new *Field".
Function objects that write need to execute after the residual checks so
that the associated modification of the time state correctly makes them
write on the final iteration.
This reverts commit c6089ae3a6, meaning
that residual controls can no longer apply to fields solved for as part
of function object executions.
Creating a system which permits solutions done within function objects
to be subject to residual control, *and* which correctly causes function
objects to write on final iterations, would be significantly more
complex.
This is a slight modification of the previous commit. All cached
temporary fields are now written to disk with the name enclosed by
"tmp<...>". This still allows for automatically constructed
field names to be read by paraFoam and other post-processing tools. It
also creates a consistent convention for naming all cached temporary
fields that are written to disk.
for example
cacheTemporaryObjects
(
"((1|((1|(1|A(U)))-H(1)))-(1|A(U)))"
);
functions
{
#includeFunc writeObjects(regExp=off, "((1|((1|(1|A(U)))-H(1)))-(1|A(U)))")
}
writes the temporary field with the name
"expr((1|((1|(1|A(U)))-H(1)))-(1|A(U)))" so that it can be read by paraFoam and
other post-processing tools.
It is now possible to define coordinate systems in a central location and
selected them by name for any model requiring one, e.g. the
explicitPorositySource.
Description
Provides a centralized coordinateSystem collection.
For example with the porous region specified in \c constant/fvOptions as
\verbatim
porosity
{
type explicitPorositySource;
explicitPorositySourceCoeffs
{
selectionMode cellZone;
cellZone porousBlockage;
type DarcyForchheimer;
// D 100; // Very little blockage
// D 200; // Some blockage but steady flow
// D 500; // Slight waviness in the far wake
D 1000; // Fully shedding behavior
d ($D $D $D);
f (0 0 0);
coordinateSystem porousBlockage;
}
}
\endverbatim
the corresponding coordinate system \c porousBlockage is looked-up
automatically from the \c constant/coordinateSystems dictionary:
\verbatim
porousBlockage
{
type cartesian;
origin (0 0 0);
coordinateRotation
{
type axesRotation;
e1 (1 0 0);
e2 (0 1 0);
}
}
\endverbatim
See \c tutorials/incompressible/pisoFoam/laminar/porousBlockage
such that div(q()) = divq(...)
and div(j()) = divj(...)
to unsure consistency between the reported heat (e.g. by the wallHeatFlux
functionObject) and mass fluxes and those used in the energy and specie
mass-fraction equations.
and changed to be an energy implicit correction to a temperature gradient
based heat-flux. This formulation is both energy conservative and temperature
consistent.
The wallHeatFlux functionObject has been updated to use a consistent heat-flux
from the heSolidThermo.
Fourier, eddyDiffusivity and nonUnityLewisEddyDiffusivity thermophysical
transport models now apply an implicit energy correction to a temperature
gradient based heat-flux to provide computational stability and efficiency while
converging to temperature gradient based solution. This ensures consistent heat
exchange between fluid and solid regions in CHT cases and with heat-flux
boundaries.
The Fourier and eddyDiffusivity models support single specie systems only
whereas nonUnityLewisEddyDiffusivity supports specie diffusion with independent
specification of turbulent Prandtl and Schmidt numbers, i.e. non-unity Lewis
number.
The unityLewisFourier and unityLewisEddyDiffusivity thermophysical transport
models use an implicit energy gradient based heat-flux which is optimal for
numerical stability and convergence but does not guarantee consistent heat
exchange between fluid and solid regions and heat-flux boundaries in the
presence of gradients of heat capacity. Both of these models support specie
diffusion with the restriction that the laminar and turbulent Prandtl and
Schmidt numbers are equal, i.e. unity Lewis number.
The thermophysical transport model is specified in the optional
thermophysicalTransport dictionary; if this file is not present the
unityLewisFourier model is selected for laminar and unityLewisEddyDiffusivity
for turbulent cases for backward compatibility.
The chtMultiRegionFoam tutorial cases have been updated to use the most
appropriate of the new thermophysical transport models.
