View Issue Details Jump to Notes ] Issue History ] Print ]
IDProjectCategoryView StatusDate SubmittedLast Update
0000494OpenFOAM[All Projects] Bugpublic2012-04-03 23:372012-06-01 19:14
Reportermturcios 
Assigned To 
PriorityhighSeveritymajorReproducibilityalways
StatusnewResolutionopen 
PlatformLinuxOSOpenSUSEOS Version11.4
Product Version2.1.x 
Target VersionFixed in Version 
Summary0000494: topoChangerFvMesh and not working properly with pimpleDyMFoam
DescriptionThe pressure and velocity fields do not match the results from the movingCone tutorial for identical mesh motion with topology changes (in this case, layer addition and removal).
Steps To Reproduce1. Unpack the attached file
2. Modify $FOAM_SRC/topoChangerFvMesh/movingConeTopoFvMesh/movingConeTopoFvMesh.C to look like the included version
3. Run ./Allrun in the included movingConeTopo case
Additional InformationThe movingCone tutorial doesn't require pRef entries while the topologically changing one does.
TagsNo tags attached.
Attached Filesgz file icon movingConeTopo.tar.gz [^] (9,209 bytes) 2012-04-03 23:37

- Relationships

-  Notes
(0001246)
mturcios (reporter)
2012-04-13 22:20

In a related study, I was testing layer addition/removal with an internal combustion engine mesh and it appears the problem occurs with layer addition. This is consistent with the movingConeTopoFvMesh as the pressure/velocity inconsistencies seem to originate at the rear of the cone where the layers are added.
(0001290)
mturcios (reporter)
2012-04-26 23:34

Some further observations

1) All fields are mapped and have values associated with them

2) All the fields appear to be mapped properly, with the exception of phi (and possibly meshPhi)

With this in mind I believe the error could be in layerAdditionRemoval/addlayer.C with the flux face flipping.
(0001356)
mturcios (reporter)
2012-06-01 19:14
edited on: 2012-06-01 23:31

Some further tinkering into the source code still leads me to believe that the problem is solely with the meshPhi field. In particular, the value of meshPhi on the faces where extrusion is occuring is about an order of magnitude larger than it should be.

Also of note is that the mesh motion fluxes in general are not working for topological changes in 2.1.x. If this were to be resolved there would be a whole raft of possibilities opened up!


- Issue History
Date Modified Username Field Change
2012-04-03 23:37 mturcios New Issue
2012-04-03 23:37 mturcios File Added: movingConeTopo.tar.gz
2012-04-13 22:20 mturcios Note Added: 0001246
2012-04-26 23:34 mturcios Note Added: 0001290
2012-06-01 19:14 mturcios Note Added: 0001356
2012-06-01 23:31 mturcios Note Edited: 0001356 View Revisions