CFD Online Discussion Forums

CFD Online Discussion Forums (
-   OpenFOAM Other Meshers: ICEM, Star, Ansys, Pointwise, GridPro, Ansa, ... (
-   -   Issues with mesh imported from Gambit, boundary layers (

kjetil November 12, 2009 18:48

Issues with mesh imported from Gambit, boundary layers
2 Attachment(s)
Dear foamers,

I have created a mesh in Gambit - it seems alright in terms of skewness (~0.8), cells close to boundary walls and total number of cells. But when running rather simple cases using solvers interFoam or simpleFoam, something fishy happens at the walls. Please see the attached two images. It can be seen from the graphics, but truly makes no sense to me whatsoever; on the way "up" (towards a siphon) the field close to the wall - which are structured hex cells - have a distinctively slower velocity. But what's even more peculiar, is that on the way down, it changes, so that the field in these cells have larger velocity than the center core of the pipe geometry. The inlet velocity is 1m/s. Do these problems occure because the mesh consists of hex and tet cells at the same time?

I would really appreciate feedback on how to set the properties right (if any), or how to make the mesh work in OpenFOAM.

Thank you.

bigphil November 27, 2009 21:50


You could try:
'polyDualMesh . . 90' or 'polyDualMesh 90' in OF-1.5 1.6
on your mesh.

It will convert the tets to polyhedra which might improve your solution.
You could give it a try and see.

I find that 'tet' cells give fishy results too.

Philip C

All times are GMT -4. The time now is 11:22.