[Jonathan]

Hi everyone,

i hope someone has some experience with the following:

I have a rotationally cyclic mesh, which i import from from Fluent format, and use createPatch to recreate the cyclics etc.

However, i have just discovered renumberMesh to reduce bandwidth, and so gave this a try - how, it seems to mess up the face numbering on the cyclic faces - and then i get a face area mismatch warning when i try to run the solver.

Does anyone know how / if it is possible to use renumberMesh, without destroying the face numbering sequence on cyclic faces, or what i need to do to re-recreate the cyclic faces so they are correctly matched with their opposite faces?

best regards
jon

[kornickel]

I know this post is old, but did you try to run "createPatch" again after renumbering your mesh? That might fix mismatching faces.

[Jonathan]

hi Frank,

yes, i tried pretty much all the permutations, including renumberMesh after the renumbering, but had no luck. The mismatch warning you get which says something like: "there is a difference in area between face x and face y of xx.xxx% etc" popped up and i couldnt get past that if i remember correctly. I tried reducing the tolerances as is normally suggested, but no luck.

i have put it aside for the time being!

cheers for the reply
jonathan

[wyldckat]

Greetings to all!

This looks to me that this was a bug in an older version of OpenFOAM. I expect the latest 2.2.2 and 2.2.x to no longer have this problem, due to this bug report: http://www.openfoam.org/mantisbt/view.php?id=1086

Nonetheless, if a simple test case could be provided, I can give it a try and ascertain if this is indeed a bug or not, so that it can be reported on the aforementioned bug tracker!

Best regards,
Bruno