Medium Test - 12.06.2018

All cases could be accessed at: /home/alex4/prashant/SUPPORT/ESI/MEDIUM_TEST/tutorials_1806

• DNS/dnsFoam/boxTurb16 - solver crash
• IO/fileHandler/machineA/fileHandler - re-verify if we need cp -r while copying constant, line 36; 90 of Allrun
• combustion/PDRFoam/flamePropagationWithObstacles - solver crash
• mesh/foamyHexMesh/mixerVessel - solver crash
• compressible/rhoSimpleFoam/gasMixing/injectorPipe - solver crash
• compressible/sonicFoam/RAS/nacaAirfoil - solver crash
• incompressible/pimpleFoam/LES/vortexShed - surface noise failed 'expected word error'
• multiphase/compressibleInterFoam/laminar/climbingRod - solver failure
• multiphase/driftFluxFoam/RAS/dahl - solver failure
• multiphase/reactingTwoPhaseEulerFoam/RAS/LBend - solver failure
• multiphase/compressibleInterFoam/laminar/depthCharge2D - solver failure
• multiphase/interFoam/RAS/DTCHullMoving - solver failure
• multiphase/interFoam/RAS/DTCHullWave - decomposePar failure 'missing entry value' REMOVED
• multiphase/interFoam/RAS/motorBike - rename solver in controlDict_nextWrite
• multiphase/interFoam/laminar/vofToLagrangian/lagrangian* - both cases replace combustionModel type noCombustion to none

@andy @Mattijs @Sergio @mark

• incompressible/pimpleFoam/LES/vortexShed failure due to ensight format.

• case file does miss the counter for number of geometries.

FORMAT type: ensight gold

GEOMETRY model: 1 data/00000000/geometry

Which type of failure do you see? Since this cut plane is part of a non-moving geometry, there is only a single static ensight geometry and thus no corresponding time-step information for it. It loads OK with paraview, and no complaints from ens-checker either: ens_checker101.log

• OK as discussed on telephone, the problem is the ensight surface reader - issue #868 (closed)

DNS/dnsFoam/boxTurb16 - solver crash. Try re-compiling boxTurb. I could run it ok after recompile boxTurb.

combustion/PDRFoam/flamePropagationWithObstacles: This cases crashes due to new implementation of ddtPhiCorr.

compressible/rhoSimpleFoam/gasMixing/injectorPipe: This is the FO surfMeshSampleDiscrete same as the one before. Is it possible to fix or we should take it out of this tutorial?

This is fixed (see #862 (closed)), but it was fixed it in develop but obviously hasn't been merged into pre-release yet.

mesh/foamyHexMesh/mixerVessel - solver crash: This case fails in the first iter in the same way in master, develop and pre-release. There seems to be something wrong in the mesh (?). The set up of the case is correct. The AMI is correct. Valgrind shows 'nan' with pcorr or any other matrix solution.

master generates a mesh with 4 regions and pre-release with 3 regions.

Attached the checkMesh for pre-release.

log.checkMesh Is foamyHexMesh generating more regions than it should?

compressible/sonicFoam/RAS/nacaAirfoil: This cases crashes due to new ddtPhiCorr. The new ddtPhiCorr is Co based correction which seems to be affecting case with large U (sonicFoam, PDRFoam so far).

multiphase/driftFluxFoam/RAS/dahl: The switch alphaApplyPrevCorr needs to be false for the case to end. This case crashes in master as well.

multiphase/compressibleInterFoam/laminar/climbingRod. Solved, changed linearUpwind to upwind in the Maxwell turb model.

multiphase/compressibleInterFoam/laminar/depthCharge2D: The new compressibleInterFOam has some (correct) extra terms in the TEq, which in this case produces a higher T than previously. The sigma model used is taken from the liquidProperties file (H2O) which seems doesn't cover the new range of temperature or the temperature is higher than evaporation, in both cases the model crashes. As this tutorial is not about phase change, I replaced the thermo by idealLiquid and constant sigma model. It runs now

multiphase/reactingTwoPhaseEulerFoam/RAS/LBend: This case seems to run fine for until the L-bend is full, when the injection of solid should stop. I'll reduce the endTime to 1.9 sec instead of 2.0 sec and it runs ok

multiphase/interFoam/RAS/DTCHullWave - decomposePar failure 'missing entry value. This case uses setWaves app which is not available

multiphase/interFoam/RAS/motorBike - rename solver in controlDict_nextWrite. Done

multiphase/interFoam/RAS/DTCHullMoving - solver failure. Still testing.

I pushed all the changes which are concerned with set up problems. We need to discuss about the ddtPhiCorr new implementation pros/cons

changed the description

marked the checklist item multiphase/driftFluxFoam/RAS/dahl - solver failure as completed

marked the checklist item multiphase/compressibleInterFoam/laminar/climbingRod - solver failure as completed

marked the checklist item multiphase/reactingTwoPhaseEulerFoam/RAS/LBend - solver failure as completed

marked the checklist item IO/fileHandler/machineA/fileHandler - re-verify if we need cp -r while copying constant, line 36; 90 of Allrun as completed

It explicitly should not copy the mesh, only the dictionaries in the constant/ directory. You will get a warning about ommitting directory etc.

mixerVessel: foamyHexMesh creates faceZone for all faces of some cells (see green and purple cells in attached pic). createBaffles will convert these into baffles and they'll show up in checkMesh as:

  <<Writing region 2 with 1 cells to cellSet region2
  <<Writing region 3 with 1 cells to cellSet region3
  <<Writing region 4 with 2 cells to cellSet region4

These cause the solver to crash with a sigfpe.

marked the checklist item mesh/foamyHexMesh/mixerVessel - solver crash as completed

marked the checklist item multiphase/compressibleInterFoam/laminar/depthCharge2D - solver failure as completed

marked the checklist item multiphase/interFoam/RAS/motorBike - rename solver in controlDict_nextWrite as completed

marked the checklist item incompressible/pimpleFoam/LES/vortexShed - surface noise failed 'expected word error' as completed

changed the description

mentioned in commit 6d01bc6616e042d3d6f9b5088616d9c7919de82e

multiphase/interFoam/RAS/DTCHullMoving - solver failure: This case run fine for me

mentioned in commit 5a127183c8fe53192afe94e9f671a3356a02468e

By Andrew Heather on 2018-06-14T13:07:12 (imported from GitLab project)

Re: ddtPhiCorr I've updated ddtScheme to use the v1712 version by default. The updated version from the org version is available by setting the info switch in the controlDict

InfoSwitches
{
    experimentalDdtCorr 1;
}

See commit 5a127183

Need to re-test to ensure all necessary changes are included. For the sonicFoam naca case the results look the same as v1712

By Andrew Heather on 2018-06-14T13:10:40 (imported from GitLab project)

marked the task compressible/sonicFoam/RAS/nacaAirfoil - solver crash as completed

By Andrew Heather on 2018-06-14T13:32:12 (imported from GitLab project)

marked the task multiphase/interFoam/RAS/DTCHullMoving - solver failure as completed

By Andrew Heather on 2018-06-14T13:32:16 (imported from GitLab project)

changed the description

By Andrew Heather on 2018-06-14T13:33:22 (imported from GitLab project)

marked the checklist item DNS/dnsFoam/boxTurb16 - solver crash as completed

mentioned in commit 75fee21999ba652e8d946d6b7e62f6b39ffec378

By Andrew Heather on 2018-06-14T13:45:36 (imported from GitLab project)

marked the task multiphase/interFoam/laminar/vofToLagrangian/lagrangian - both cases replace combustionModel type noCombustion to none* as completed

By Andrew Heather on 2018-06-14T13:45:46 (imported from GitLab project)

marked the checklist item combustion/PDRFoam/flamePropagationWithObstacles - solver crash as completed

marked the checklist item compressible/rhoSimpleFoam/gasMixing/injectorPipe - solver crash as completed

closed

mentioned in commit 528d8a7e