[bentkj]

Hi everyone, I am having some trouble with sHM and would like some help.

Background:

1. I am trying to conduct an external flow over a wing (NASA CRM) and currently I'm trying to create the mesh with snappyHexMesh.
2. The .stl file that I'm using is prepared with salome, with the inspections for Detect self-intersections giving 'There are no self-intersections in the shape' and check free boundaries to return 0 for both open and close free boundaries. A single patch is created for the entire wing.
3. The .stl (which is scaled from mm to m using surfaceTransformPoints) seems to be alright with surfaceCheck reporting that the surface is closed.

Problem:
Upon generating the external mesh with blockMesh (cell aspect ratio =1) and snappyHexMesh, I get bumps over the surface of the wing as shown in the pictures. It seems like these tend to appear where the cells transit between refinements. I have tried playing around with te sMHDict and meshqualityDict files with not much improvement in the mesh. However, I have observed that if I set the min/max refinementSurfaces values to the same, i.e. (10 10), the bumps do not show up. However this leads to extremely high cell counts and I would like something more reasonable i.e. (5 10), to which the bumps appear again. There's a similar post on this which was suggested to use minTetQuality -1e30. This helps alittle, but the bumps are still present.

I hope I have provided enough information to receive help my geometry file is ~100mb so I'll have to share that separately if necessary. I have been stuck on this for about 3 weeks now so anything is REALLY appreciated!!

[Tobi]

Hi,

related to your request on my youtube-channel, some remarks.

• You do only 30 smoothing iterations (a bit less I think)
• You have problems in your cells as you can see in the output of sHM during snapping
• I only can guess based on the output of sHM. nCells should be set larger and the refinement method you are using should be corrected
• the feature edge refinement should be changed (see http://www.holzmann-cfd.de/index.php...h/feature-edge) and you should check your features edge mesh if you did not do that till now
• Depending on everything, the tolerance factor can be enlarged a bit (based on the warning you have - but I do not have any idea where the point is located, OpenFOAM tells you).

[schiffbauer]

Dear snappy-users,


I also face a problem with "bumps" on the snapped surface (NACA-wing), like:



And here you can see the underlying stl-file:




Further details:
- featureSnap is turned off here
- the blockMesh is isotropic

- for a coarser mesh, no bumps occur
- snapControls:

nSmoothPatch 1; // 3; 5;

tolerance 4.0; // 2.0; 6.0;

nSolveIter 0; // 30; 60;

nRelaxIter 10; //5;

I played around more than a week but w/o success. It seems NOT to depend on
- the resolution of the stl-file
- the Foam version (1806 or 5)


Any idea would be highly appreciated!