CFD Online Discussion Forums

CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)
-   OpenFOAM Meshing & Mesh Conversion (https://www.cfd-online.com/Forums/openfoam-meshing/)
-   -   [snappyHexMesh] symmetryPlane in sHM (https://www.cfd-online.com/Forums/openfoam-meshing/185640-symmetryplane-shm.html)

Tobi March 30, 2017 07:43

symmetryPlane in sHM
 
Hi everybody,

I have a question to you. Maybe one got to the same problem. I am meshing a complex geometry with sHM but I have problems with the symmetry planes. The faces are not 100% aligned to the patch normal and therefore I get e.g. displacements in y direction based on the surface normal direction.

Question 1
  • Is there a way to flatten the symmetry patches in order that all points are aligned within one plane (like flattenMesh for 2D).
Note 1
  • symmetry vs. symmetryPlane
  • I read that people just change it without thinking about the differences. The difference should be:
    • symmetryPlane has to be really flat and one single patch (no multi patch assembly)
    • symmetry can be used for non-100% aligned flat patches and multi patch assemblies
  • I got this from https://openfoam.org/release/2-3-0/numerics/
However, using the symmetry BC in my simulation lead to the problem that I get motion in the direction I want to constrain (like y-direction). This is based on the fact that I do not get a 100% aligned flat patch after snapping with snappyHexMesh.



Question 2
  • Does anybody know how to get a flat patch using sHM
  • Maybe there is a tool out which can be used to flatten a patch (like flattenMesh for 3D)
  • I was expecting that after defining the patchType in the snappyHexMeshDict to be symmetry or symmetryPlane, the meshing application takes care of adjusting the points.




assign the symmetryPlane patch type. I get the problem, that the application tells me the following:

Code:

Symmetry plane 'sym1' is not planar.
At local face at (0.00702909 0.00026875 0.0282749) the normal (1 3.22739e-15 1.60222e-15) differs from the average normal (0.428571 5.18688e-16 -6.34413e-17) by 0.326531
Either split the patch into planar parts or use the symmetry patch type

However, the normal vector at point (0.00702909 0.00026875 0.0282749) is fine. It should direct to the x direction which it probably do but I have no idea why it is complaining about that point because it is not on the sym1 patch.

KaLium April 25, 2017 06:58

You can use the outer limits of your blockMesh to create very flat surfaces.

Tobi April 25, 2017 07:59

Hello Kai,

this will not help, or if you think, please offer an example. very flat surfaces are not flat at all. Using snappy will produce non-flat surfaces at the patches during the snap process. The solution is to move the patch points to a plane by yourself. The program which can handle that could be found in my repositories.

KaLium April 25, 2017 08:46

I ment something like this https://sites.google.com/site/snappy.../cylinder-case
(inlet-boundaries, with "small blockMesh")

Tobi April 25, 2017 08:59

Did you ever checked if the patches (inlet/outlet) could be set for symmetryPlane? In addition my geometry is a chainsaw motor block. So no simple geometry. However, thanks for your reply.

Tobi April 25, 2017 09:11

Okay I checked your geometry ...
  • inlet and outlet can be specified as symmetryPlane
  • However, I tried a lot of things (even similar to yours) without success for the complex geometry

KaLium April 25, 2017 09:19

That is not my geometry. Just something I found.


All times are GMT -4. The time now is 12:45.