The LBend was set to run for 2 s, but at about 1.95 s the packed region
builds up to the inlet and the simulation diverges. The end time has
been reduced to 1.9 s so that this does not occur.
This change fixes an issue where mass transfer models were being looked
up on the wrong side of the interface. This also means that specifying
mass transfer just on one side of an interface is now possible without
generating errors.
When a part of the tetrahedral decomposition is inverted, tracking along
a straight line can result in a closed loop which never ends.
This change adds a limit to the number of tracks that are done that end
before or at the maximum distance already achieved. This breaks these
closed loops and prevents the simulation from hanging. The particles do,
however, end up in an incorrect position as a result of the tracking
being abandoned at an intermediate point in the step. A warning is
printed to indicate when this is occuring.
This resolves bug report https://bugs.openfoam.org/view.php?id=3056
Description
Fixed traction boundary condition for the standard linear elastic, fixed
coefficient displacement equation in which the traction is caused by
the hydrostatic pressure of an external liquid reservoir.
Description
Calculates the natural logarithm of the specified scalar field.
Performs \f$ln(max(x, a))\f$ where \f$x\f$ is the field and \f$a\f$ an
optional clip to handle 0 or negative \f$x\f$.
The etc/caseDicts/postProcessing/fields/log configuration file is provided so
that the simple #includeFunc can be used to execute this functionObject during
the run, e.g. for some dimensionless field x
functions
{
#includeFunc log(x)
}
or if x might be 0 or negative in some regions the optional clip may be applied:
functions
{
#includeFunc log(p,clip=1e-6)
}
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
For cases where the time-step varies in space ddtCorr is not needed with LTS but
in regions where the time-step is uniform it may be required. This change
ensures that the localEuler scheme operates equivalently to Euler in regions and
for cases where the time-step is uniform.
Addresses report https://bugs.openfoam.org/view.php?id=3020
In addition to the current syntax, particle forces can now also be
specified with a named dictionary and a type entry. This allows multiple
forces of the same type to be used. For example:
particleForces
{
// Existing syntax. The model typename, either on it's own
// or heading a dictionary of model parameters.
gravity;
ErgunWenYuDrag
{
alphac alpha.air;
}
// New syntax. An arbitrary name, followed by a dictionary
// containing a type entry, and any model parameters.
gravity2
{
type gravity;
}
drag2
{
type ErgunWenYuDrag;
alphac alpha.air;
}
}
A scaled force has also been added. This is a wrapper around another
particle force model, and applies a scaling factor to the force value.
It can be used to tune the models, or to blend two or more models.
particleForces
{
// Tuning the simulation by reducing the pressure gradient force
scaled1
{
type scaled;
factor 0.94;
forceType pressureGradient;
U U;
}
// Blending two drag models
scaled2
{
type scaled;
factor 0.2;
forceType sphereDrag;
}
scaled3
{
type scaled;
factor 0.8;
forceType ErgunWenYuDrag;
alphac alpha.air;
}
}
This work was supported by Moritz Hoefert, at Evonik
