Pass the correct temperature into the rate term.
2 lines of code changed in 1 file:
Add tools to allow one to easily manipulate simple graphs from the input file. Specifically:
- tree cleaving: allow cleaving by name tag from input file
- attach_dependency from input file
68 lines of code changed in 3 files:
Whoops...got the order of operations wrong.
2 lines of code changed in 1 file:
Adding access to commonSetup for the constant property model.
2 lines of code changed in 1 file:
Only add adiabatic and sensible enthalpy if the model is actually a heat
loss model.
2 lines of code changed in 1 file:
Eliminate many duplicate handlings of TaskList parsing for a consolidated treatment.
86 lines of code changed in 7 files:
Do not check external deps on SoleVariable.
Currently SoleVariable will never send out through MPI.
4 lines of code changed in 1 file:
Nuking all the changes regarding volumeFraction and
reverting back to cellType. Using volumeFraction
to detect boundaries fails at inlet,outlet,
and other pressure BCs because there is no
way to use volumeFraction to distinguish between
flow cells and these kinds of BCs.
This implies that if RMCRT is to be implemented
in other components then the cellType variable
must be present in that component.
Sorry Todd, but this is going to break all
RMCRT RTs. Do you mind updating GSs?
71 lines of code changed in 7 files:
Remove some superfluous includes.
0 lines of code changed in 2 files:
1) Separate the calculation of source terms from the transported scalars. This will allow a source term to be reused by multiple transport equations without recalculating the source term.
2) Removing the <addSources> option in <TransportEqn> because it was redundant. Simply don't add <src … > the <eqn> if you don't want it used.
3) Cleaning up the westbrook dryer model:
- If flammability limits are used, but the diluent is constant everywhere, then assign a constant value rather than transporting one
- Use the source weights (committed earlier) to add or subtract the source term depending on which equation is using it
4) Bringing input files up to spec.
119 lines of code changed in 7 files:
No longer creating/destroying the volumeFrac label. It's passed in when you register the labels
1 lines of code changed in 1 file:
added particle variables for rate dependence in Arenisca
133 lines of code changed in 2 files:
Call common problem setup for this model.
1 lines of code changed in 1 file:
Reorder the loops in initializePressureBC. The material loop needs to be
outside of the loop over BCs due to the allocation of p.externalForceCorners
used in CBDI, in case more than one BC is applied to a given material (e.g, on
a different part of that material's surface).
55 lines of code changed in 1 file:
Add some macros to magma_testdefs to further simplify interfacing with MAGMA.
Pull some Redstorm weeds.
120 lines of code changed in 3 files:
Add brackets for clarity
3 lines of code changed in 1 file:
Now successfully and accurately doing the sci_lapack invert on the GPU via MAGMA library.
This is used in ARCHES-DQMOM.
Also added some items to SVN ignore list.
80 lines of code changed in 1 file:
Do not use d_labelSchedInit to check if the label should be computed or modified.
1 lines of code changed in 1 file:
Changed from cellType to volumeFrac to
allow for cross-component generality.
This shouldn't change the radiation answers,
but all the rmcrt tests will fail since
the udas cannot be compared due to the
change in variable name.
Todd, can I get you to update the GSs?
68 lines of code changed in 6 files:
Sync the system up before any rank can use the output base dir during
output initiliaztion. This will hopefully fix some output issue.
5 lines of code changed in 1 file:
udaReader: add forceReload for multi-level mesh reading, support for nodeCentered meshes.
0 lines of code changed in 4 files:
This should fix the restart issue for massflow rate bc's. Previously, velocities were not being computed until the second timestep of thre restart causing a blip in the answers.
In this fix, we need to recompile the taskgraph after the first timestep to remove the velocity BC calculation. If a sched_restartInitialize is made available
to the component, then we should schedule this task rather than recompiling the taskgraph.
37 lines of code changed in 3 files:
add license information and documentation to the TimeDerivative expression.
32 lines of code changed in 1 file:
Add ability to calculate time derivatives of all quantities. This will NOT work on the bottom-level nodes in the Expression graph (i.e. transported variables) at the moment. Bullet-proof the OldVariable for a couple of subtle cases. Move the OldVariable setup before the requires-tasks to avoid Uintah task warnings.
176 lines of code changed in 8 files:
Fixed a memory leak
2 lines of code changed in 1 file:
Allow user defined task to change outputInterval and checkpointInterval
Example:
schedule:
task->computes(shareState->get_outputInterval_label())
shareState->updateOutputInterval(true);
task function:
newdw->put(min_vartype(newvalue), shareState->get_outputInterval_label());
76 lines of code changed in 6 files:
Check allowsMultipleComputes() instead of compare strings for system var reduction
5 lines of code changed in 1 file:
Another step toward generalization of device tasks/task queues.
81 lines of code changed in 6 files:
Beginnings of generalized "device" tasks and work queues.
This will ultimately provide a unified approach to both GPU and MIC tasks.
1112 lines of code changed in 10 files:
Removing the option to set wall and intrusion abskg's since it appears
that this wasn't even the correct way to do things. These values are
now hard coded to 1.0 until we improve the model.
8 lines of code changed in 3 files:
Allow the user to weight a source term with an arbitrary constant.
37 lines of code changed in 3 files:
Fixes for Kraken:
Don't use '-fPIC' if '--enable-static' is set to 'full'. Note,
our 'static' build for the RT (due to the way linux installs)
requires some dynamic (system) libs, so we don't really force
a full static build (thus the reason we have the distinction between
the two 'static' build options). However, on Cray machines, we
need to be fully static.
Allow for a 'built-in' version of Hypre to be used. Use:
--enable-hypre=built-in
117 lines of code changed in 2 files:
added mpmflag
13 lines of code changed in 3 files:
change debug output for schedule common
3 lines of code changed in 1 file:
Changed the calculation that terminates when
a ray will terminate for cases with non-black
walls. This should fix the
problem of infinite loops for cases that
have non-black surfaces and extremely low
optical thicknesses in early timesteps.
This will break the RT for RMCRT tests that
have non-black walls, specifically,
RMCRT_ML and RMCRT_1L_reflect.
Todd, can you update the GSs?
3 lines of code changed in 1 file:
add rank ID to scheduler common debug output
1 lines of code changed in 1 file:
makeCombinedIndex: changed script from sh -> bash.
makeMasterUda_index: added "" around inputs
0 lines of code changed in 4 files:
added more diagnostic output when SCI_DEBUG is turned on.
0 lines of code changed in 2 files:
Adding a time window on WD model.
6 lines of code changed in 2 files:
rename coarsenDelt task name to reduceSysVar
call MPI reduce directly from reduceSysVar task
17 lines of code changed in 6 files:
Make sure that the old mf initial value is set to zero to avoid getting
into dusty corners of the code.
1 lines of code changed in 1 file:
This allows you to compute heat loss, but use a constant value when
doing table look ups. This is useful for debugging and verification
work.
113 lines of code changed in 2 files:
Changed Wasatch file for inlet turbulence to be consistent with digital filter generator changes.
12 lines of code changed in 1 file:
Changed the way the digital filter generator works to only output data for cells in the inlet. This will decrease table size for non-box inlet geometires.
149 lines of code changed in 3 files:
Committing James Colovos Arenisca to repo. Most current version, compatible with kinematics modifications made to SerialMPM.
-Dave A.
2436 lines of code changed in 2 files:
Fixing an error with getting an iterator for the intrusion.
1 lines of code changed in 1 file:
Don't prefill on restarts.
Recompile taskgraph on restart to get refreshed boolean.
1 lines of code changed in 1 file:
Print out some useful information about the intrusions at startup.
178 lines of code changed in 3 files:
Add an error catch to not finding the correct input file for velocities when using the VelocityFileInput option.
Change the restart boolean on newBC's so that prefill (if used) is not re-applied every timestep.
19 lines of code changed in 2 files:
Silence some warnings coming out of improperly ordered constructor initializers.
2 lines of code changed in 1 file:
BRUTE FORCE HACK.
Instead of storing the file pointers in the DataWarehouse just open the
file and then close it when you're done writing to it. This will be slow but
it will get us moving for now.
28 lines of code changed in 1 file:
* Fixing a restart bug with inlet areas, velocities and mass flow rates. This information wasn't computed properly (or at all) for restarts.
* allow a tabulated option for scalar on intrusion inlets
93 lines of code changed in 6 files: