View Issue Details

Jump to NotesJump to History
IDProjectCategoryView StatusDate SubmittedLast Update
0002520OpenFOAMBugpublic2017-04-04 11:102017-04-19 16:02
ReporterSvensen Assigned Towill  
PrioritynormalSeverityminorReproducibilityalways
Status closedResolutionno change required 
PlatformGNU/LinuxOSUbuntuOS Version14.04
Product Versiondev 
Summary0002520: irregular cells on the patches during snap stage
DescriptionAfter the snap stage, I've found that the Courant number in some cells on the patch's surfaces is very high compared to surrounding cells.

I've checked the STL surface with checkSurface tool and found no obvious errors.

Then I've tried to modify some parameters in meshQualityDict, but it didn't help. Despite that snappyHexMesh reported that 'Mesh is OK', some cells on the inlet and outlet patches are really irregular.

I've attached the case here:
https://yadi.sk/d/4m_qhV1U3GdHLw

Some Paraview screenshots are also attached.
Steps To ReproduceJust open Paraview, go to the last time step, apply a 'Threshold filter' on Courant field using the range [2; 13]. You will see that the high Courant cells are located only on the inlet and outlet patches. You can extract some high Courant cells with a 'Extract selection' filter.
Additional InformationThis bug is correlated with https://bugs.openfoam.org/view.php?id=2099 and https://bugs.openfoam.org/view.php?id=1633. But these two bug reports was about irregular triangles on a surface of a model. In the present case we have irregular cells on a patch surface.
TagsPre-processing, snappyHexMesh

Activities

Svensen

2017-04-04 11:10

reporter  

overview.png (179,792 bytes)   
overview.png (179,792 bytes)   

Svensen

2017-04-04 11:11

reporter  

inlet-patch.png (196,904 bytes)   
inlet-patch.png (196,904 bytes)   

Svensen

2017-04-04 11:11

reporter  

inlet-patch-2.png (196,904 bytes)   
inlet-patch-2.png (196,904 bytes)   

Svensen

2017-04-17 21:35

reporter   ~0008028

Any idea ?

will

2017-04-19 16:02

manager   ~0008032

That's just how snappy works. Cells on boundaries get morphed to fit the geometry, which can make them smaller. Cell volume itself is not a quality criteria by itself. Not a bug.

If you can propose a better way of doing it then by all means raise a feature request.

Issue History

Date Modified Username Field Change
2017-04-04 11:10 Svensen New Issue
2017-04-04 11:10 Svensen File Added: overview.png
2017-04-04 11:10 Svensen Tag Attached: Pre-processing
2017-04-04 11:10 Svensen Tag Attached: snappyHexMesh
2017-04-04 11:11 Svensen File Added: inlet-patch.png
2017-04-04 11:11 Svensen File Added: inlet-patch-2.png
2017-04-17 21:35 Svensen Note Added: 0008028
2017-04-19 16:02 will Assigned To => will
2017-04-19 16:02 will Status new => closed
2017-04-19 16:02 will Resolution open => no change required
2017-04-19 16:02 will Note Added: 0008032