CFD Online Discussion Forums

CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)
-   OpenFOAM Community Contributions (https://www.cfd-online.com/Forums/openfoam-community-contributions/)
-   -   [cfMesh] Refine mesh inside auxiliary STL (https://www.cfd-online.com/Forums/openfoam-community-contributions/210922-refine-mesh-inside-auxiliary-stl.html)

Dreoasteh November 7, 2018 08:42

Refine mesh inside auxiliary STL
 
Hi,

I'm trying to mesh a nozzle geometry with cfmesh and have a question regarding refinement with external STLs.

In snappy you can access an auxiliary STL and refine inside it using "refinementRegions" and "mode inside".

However in cfmesh I only see two options:

  1. Use the primitive shapes in "objectRefinements" from which I cannot produce the complex shape I'm looking for.

    OR
  2. Use "surfaceMeshRefinements" which only refines the region of the mesh intersected by the auxiliary surface.
Is there a way to do the same thing as in snappyHexMesh (importing a surface mesh in STL and refine inside it?).

MichaelXPS April 1, 2020 10:10

Have you found an answer to this? I am thinking of using patch for overlapping faces of different regions but am also trying to figure out how to get a refinement in an arbitrary region, e.g. described by another stl file.

Quote:

Originally Posted by Dreoasteh (Post 714473)
Hi,

I'm trying to mesh a nozzle geometry with cfmesh and have a question regarding refinement with external STLs.

In snappy you can access an auxiliary STL and refine inside it using "refinementRegions" and "mode inside".

However in cfmesh I only see two options:

  1. Use the primitive shapes in "objectRefinements" from which I cannot produce the complex shape I'm looking for.

    OR
  2. Use "surfaceMeshRefinements" which only refines the region of the mesh intersected by the auxiliary surface.
Is there a way to do the same thing as in snappyHexMesh (importing a surface mesh in STL and refine inside it?).


Dreoasteh April 1, 2020 10:31

Hi,

From out testing cfMesh is not ideal because, although it achieves great boundary layer coverage, it retains bad quality cells which causes cases to eventually diverge. My recommendation would be to go back to sHM which does not have this critical issue and enables you to have arbitary refinement regions

MichaelXPS April 1, 2020 14:06

Catchup discussion
 
Hi, Dreoasteh,

thanks for your comment on cfMesh. By bad quality cell you mean the degree of skewness? I don't know yet but maybe there's some options to regulate this like in blockMesh? Ok, I would try sHM that you suggested and the only problem is that I don't know what is a sHM? Could you inform a full name?

Quote:

Originally Posted by Dreoasteh (Post 763839)
Hi,

From out testing cfMesh is not ideal because, although it achieves great boundary layer coverage, it retains bad quality cells which causes cases to eventually diverge. My recommendation would be to go back to sHM which does not have this critical issue and enables you to have arbitary refinement regions


Dreoasteh April 1, 2020 14:17

The bad quality cells come from different types, skewness, cell volume, etc. There's no option to tighten tolerances or improve mesh quality as the whole idea of cfMesh is to have an automatic mesher with a small numer of degrees of freedom. In our experience it is always a good idea to try out cfMesh but be aware that your case can blow up with no aparent reason and with little in the way of fixing it.

By sHM I ment snappyHexMesh which is an automatic mesher that comes with OpenFOAM.


All times are GMT -4. The time now is 09:10.