The themo tables used in wallBoiling have had their Cp/Cv values
corrected, and have been coarsened and reduced in size to bound only the
operating point of the wallBoiling tutorials. They have also been moved
to $FOAM_TUTORIALS/resources/thermoData.
The correction to thermophysical properties has improved the stability
of these cases. As a result it has been possible to reduce the amount of
under-relaxation used in the wall modelling.
The phase-fraction filtering of the compressibility terms is present to avoid
spurious phase-change due to numerical noise. The fvModels contribution may
cause physical phase-change due to cavitation, boiling, transfer from film or
VoF even where none of the phase is present and hence must be added after the
compressibility filtering.
With VoFClouds and VoFSurfaceFilm compressibleInterFoam supports Lagrangian
clouds which can impinge on walls forming a film which in turn can transfer to
the VoF when the film is thick enough to resolve. The new tutorial case
tutorials/multiphase/compressibleInterFoam/laminar/cylinder
is provided to demonstrate this functionality.
Direct transfer of droplets to the VoF phase is not yet supported but will be
added later.
A cloud may now contain species that do not exist within the carrier
thermodynamics. Correspondence between lagrangian and carrier components
is only required when there is phase change.
The thermo parcel now supports thermophysical property modelling. This
particle does not store phase or specie fractions so it only provides a
single phase with a uniform composition. Additional specification is
required in the cloud subModel configuration in order to select the
specie. For example:
compositionModel singlePhaseMixture;
singlePhaseMixtureCoeffs
{
phases
(
solid
{
CaCO3 1;
}
);
}
The reduced SLGThermo has been renamed parcelThermo to better represent the
purpose.
parcelThermo is not created and stored in the cloud that requires it rather than
requiring it to be created in the solver createFields and passed to the cloud on
construction.
This makes usage of transformPoints the same as for
surfaceTransformPoints. Transformations are supplied as a string and are
applied in sequence.
Usage
transformPoints "\<transformations\>" [OPTION]
Supported transformations:
- "translate=<translation vector>"
Translational transformation by given vector
- "rotate=(<n1 vector> <n2 vector>)"
Rotational transformation from unit vector n1 to n2
- "Rx=<angle [deg] about x-axis>"
Rotational transformation by given angle about x-axis
- "Ry=<angle [deg] about y-axis>"
Rotational transformation by given angle about y-axis
- "Rz=<angle [deg] about z-axis>"
Rotational transformation by given angle about z-axis
- "Ra=<axis vector> <angle [deg] about axis>"
Rotational transformation by given angle about given axis
- "scale=<x-y-z scaling vector>"
Anisotropic scaling by the given vector in the x, y, z
coordinate directions
Example usage:
transformPoints \
"translate=(-0.05 -0.05 0), \
Rz=45, \
translate=(0.05 0.05 0)"
Now the VoFSurfaceFilm library is optionally loaded at run-time for cases that
require surface film by adding the optional "libs" entry in controlDict:
libs ("libVoFSurfaceFilm.so");
See tutorials/multiphase/compressibleInterFoam/laminar/plateFilm
Now the VoFSurfaceFilm library is optionally loaded at run-time for cases that
require surface film by adding the optional "libs" entry in controlDict:
libs ("libVoFSurfaceFilm.so");
See tutorials/multiphase/compressibleInterFoam/laminar/plateFilm
The property entries created by the "none" models were not formatted
correctly, oweing to the lack of a base name or model name in the
sub-model base. These erroneous entries then resulted in restart
failures.
in particular for equations of state which do not support coefficient mixing
required by equilibrium constant evaluation in reactions.
Also improved the set of pre-compiled combinations of properties and mixing
rules.
Description
Deferred correction interpolation scheme derived from upwind
which returns upwind weighting factors and an explicit correction obtained
from the specified scheme.
This ensures that the transport matrix generated is at least diagonally
equal and avoids the need for relaxation of the matrix (which can affect
conservation) for transient running.
e.g.
divSchemes
{
default none;
div(phi,U) Gauss deferred limitedLinear 1;
.
.
.
and with this setting relaxation of the U-equation is not necessary irrespective
of the Courant number.
Unused entries have been removed, and documentation provided for
previously undocumented entries has been added.
Patch contributed by Francesco Contino
The chemical time-scale has been generalised so that it generates
identical results for equivalent sets of reactions. Otherwise, the
calculation is unchanged. The time scale is still based on the amount of
time a reaction would take to consume all moles currently present in the
system.
To provide more flexibility, extensibility, run-time modifiability and
consistency the handling of optional pressure limits has been moved from
pressureControl (settings in system/fvSolution) to the new limitPressure
fvConstraint (settings in system/fvConstraints).
All tutorials have been updated which provides guidance when upgrading cases but
also helpful error messages are generated for cases using the old settings
providing specific details as to how the case should be updated, e.g. for the
tutorials/compressible/rhoSimpleFoam/squareBend case which has the pressure
limit specification:
SIMPLE
{
...
pMinFactor 0.1;
pMaxFactor 2;
...
generates the error message
--> FOAM FATAL IO ERROR:
Pressure limits should now be specified in fvConstraints:
limitp
{
type limitPressure;
minFactor 0.1;
maxFactor 2;
}
file: /home/dm2/henry/OpenFOAM/OpenFOAM-dev/tutorials/compressible/rhoSimpleFoam/squareBend/system/fvSolution/SIMPLE from line 41 to line 54.
The linearisation of the reaction rate relative to the concentration of
a reference specie is not required anywhere. This data has been removed
from the rate method's output and the internals of the rate method
simplified accordingly. The clipping in the rate methods has also been
simplified and made consistent across the different rate and rate
derivative methods.
This change ensures that ISAT operates correctly as a result of changes
to the chemistry integration step, by putting it permenantly into what
was previously a variable-time-step mode. The constant-time-step mode is
no longer available.
The constant-time-step mode was not compatible with run-time changes,
certain types of restart, or models in which chemistry is integrated
over a time-scale which differs from the physical time-step. The level
of optimsation that the constant-time-step mode previously provided was
also found to be negligible. It's removal therefore simplifies the code
and improves the correctness of operation whilst not removing any
meaningful functionality.
Now that dynamic compilation of thermo and chemistry is available it is no
longer necessary or useful to instantiate vast numbers of thermo combinations
within the standard OpenFOAM libraries. This reduces compile time, library size
and simplifies maintenance.
This switch controls whether the model is corrected every outer
iteration, or just once per timestep.
This is the same as the corresponding switch in the laminar and PaSR
models, except that in the case of EDC it defaults to true. EDC folds
more information pertaining to transport into the chemistry integration,
so updating the chemistry on outer iteration is likely to have greater
effect than for the other models.
Properties have been removed that are set in the standard TDAC ".cfg"
file, and ".orig" files have been used to better ensure that cleanCase
restores the original state. Sandia has also had it's TDAC parameters
slightly tweaked for stability.
The "initialSet" and "fuelSpecies" settings for chemistry reduction
methods now have to be formatted as lists, rather than dictionaries.
This is so that the settings in the TDAC configuration files can be
overridden in a case without the dictionaries being merged.
