View Issue Details

IDProjectCategoryView StatusLast Update
0001359OpenFOAMBugpublic2014-12-28 17:13
Reporterdhora Assigned Tohenry  
PrioritynormalSeveritymajorReproducibilityalways
Status closedResolutionunable to reproduce 
Summary0001359: Different fvc::div(phi,U) in staticFvMesh and rotatingMotion simulation with omega=0
DescriptionI got very different results in a pimpleDyMFoam simulation with a staticFvMesh and a solidBodyMotionFvMesh (rotatingMotion with omega=0). The staticFvMesh results were identical to the pimpleFoam solution.

The difference seemed to be caused in fvc::div(phi,U) with schemes like vanLeer or linearUpwind. The standard upwind or linear scheme give identical results in both cases.
Steps To ReproduceExecute the uploaded application for the attached case and compare the result of fvDivPhiU at the first time step (0.002) in the staticFvMesh and the solidBodyMotionFvMesh simulation. Especially cell 421 shows very different results.
TagsNo tags attached.

Activities

dhora

2014-07-28 23:45

reporter  

mixerVesselAMI2D.tar.gz (312,339 bytes)

dhora

2014-07-28 23:46

reporter  

dhora

2014-07-29 07:33

reporter   ~0003182

The results of the propeller tutorial seem to be identical for both setups. Maybe it's only a problem in the fvc and not in the fvm implementation.

henry

2014-07-29 09:55

manager   ~0003184

There appears to be an issue with the consistency between the U, phi and boundary specification in your 0 directory. If you remove the phi field
fvc::div(phi,U) evaluates the same for staticFvMesh and a solidBodyMotionFvMesh.

henry

2014-07-29 12:37

manager   ~0003186

Why do you want to update of the cyclicAMI connections at each time-step for a static mesh?

dhora

2014-07-29 14:09

reporter  

dhora

2014-07-30 09:05

reporter   ~0003189

Last edited: 2014-07-30 09:43

Difference comes probably from the snGrads at the walls. U.boundaryField() on the rotor is different in the two cases. It's almost 0 in the omega=0 case and !=0 in the staticFvMesh case.

henry

2014-12-28 17:13

manager   ~0003369

It is not clear what is being attempted here and the behavior is not reproducible with standard released solvers and cases. It is likely that you will need to arrange direct support with the writing of your application.

Issue History

Date Modified Username Field Change
2014-07-28 23:45 dhora New Issue
2014-07-28 23:45 dhora File Added: mixerVesselAMI2D.tar.gz
2014-07-28 23:46 dhora File Added: fvcDivPhiUDyMFoam.tar.gz
2014-07-29 07:33 dhora Note Added: 0003182
2014-07-29 09:55 henry Note Added: 0003184
2014-07-29 12:37 henry Note Added: 0003186
2014-07-29 14:09 dhora File Added: cyclicAMIPolyPatch.tar.gz
2014-07-30 09:05 dhora Note Added: 0003189
2014-07-30 09:14 dhora Note Edited: 0003189
2014-07-30 09:24 dhora Note Edited: 0003189
2014-07-30 09:37 dhora Note Edited: 0003189
2014-07-30 09:39 dhora Note Edited: 0003189
2014-07-30 09:43 dhora Note Edited: 0003189
2014-12-28 17:13 henry Note Added: 0003369
2014-12-28 17:13 henry Status new => closed
2014-12-28 17:13 henry Assigned To => henry
2014-12-28 17:13 henry Resolution open => unable to reproduce