[hz283]

Hi All,

In Openfoam, within one solver, how to realize that some of the governing equations are in one set of mesh (for instance, fine mesh), and others are solved in another set of coarse mesh? the data transfer between two meshes are done by interpolation during each time step. Does anybody know something about this problem? Could you please give me some hints about this?

best,
h

[wyldckat]

Hi H,

Quick answer: look into the solvers with the name patterns "chtMultiRegion*Foam" and "*MRF*". You can search for them with these commands:

find $FOAM_SOLVERS -name "chtMultiRegion*Foam.C"
find $FOAM_SOLVERS -name "*MRF*.C"

Best regards,
Bruno

[akidess]

Bruno, those solvers couple two or more adjacent meshes. I believe the question was about overlapping meshes. This would e.g. allow you to solve velocity on a fine mesh and temperature on a coarse mesh. Unfortunately I don't have an answer to that question, I don't think it's been done so far.

[hz283]

Originally Posted by akidess

Bruno, those solvers couple two or more adjacent meshes. I believe the question was about overlapping meshes. This would e.g. allow you to solve velocity on a fine mesh and temperature on a coarse mesh. Unfortunately I don't have an answer to that question, I don't think it's been done so far.

Hi Anton,

You are correct. But I do not how to realize it now.

best regards,
H

[codinging]

to implement this, you have to rewrite the main program without using the routine headfiles such as the setRootCase, createMesh, createTime, createFields. For interpolation between different meshes, you can refer to mapfield tool in the userguide, maybe just copy part of the code to your own program. It would be wise to write your own interpolation function if you know the detail how the coarse mesh is refined. Not difficult, but a time-consuming task.