[ebertmp]

I have been working on a simple test problem involving laminar flow between a rotating and non-rotating cylinder (see attached file). Specifically I used simpleFoam while setting the turbulence model to laminar. This problem was chosen because it has an exact analytical solution with which to make comparisons. Grid convergence studies were done using a series of 4 hexahedral and 4 tetrahedral meshes. A comparison is also made with results from Fluent using the exact same meshes.

In summary the results show that the OpenFOAM prediction of the torque using the series of tet meshes DOES NOT appear to converge to the exact solution even though Fluent does. Of course "user error" can not be completely ruled out and so I would be glad to provide my OpenFOAM case directories to anyone interested.

Does anyone else have an example of this kind of behavior?

[ebertmp]

Is anyone regularly using multi-element meshes (tets + triangular prisms for boundary layer resolution) in OpenFOAM for incompressible flow simulations with success?

Thanks

--- Please participate in the poll at the beginning of this thread ---

[phsieh2005]

Hi, ebertmp,

Could you please send me the OpenFOAM case? I am interested in this case and would like to play with it. Thanks!

Pei
---------------
phsieh2005@yahoo.com

[ebertmp]

For anyone interested, the following links can be used to download my case directories. If you have more success than I did getting the correct answer using the tetrahedral meshes I would really appreciate feedback.

HexMesh_Coarse
HexMesh_Medium
HexMesh_Fine
HexMesh_Finest

TetMesh_Coarse
TetMesh_Medium
TetMesh_Fine
TetMesh_Finest

Thanks!

P.S. - These are setup for version 1.6.

[wiedangel]

Hi Ebertmp.

I can´t get access to your cases, I would like also to test a transient solver which I am developing on your case. Could you please provide your OF cases as I can compare with an analytic solution.

Best regards