SnappyHexMesh - How to set-up to avoid skewfaces?
2 Attachment(s)
Hello forum!
I started using snappy a few weeks ago. So I haven't too much experience using it. I'm trying to mesh the air around a body. I changed almost every parameter I think can improve my mesh. No matter what I change I always get some skewfaces. These skewfaces always follow the same patern: It's an hex over the patch in the castellated mesh, in the snap step it become a skewfaces. What I did: I Changed the number of refinement steps in this patch, I changed the tolerance in the snap settings. I attached two screenshots and below there's my snappyHexMeshDict Code:
/*--------------------------------*- C++ -*----------------------------------*\ Regards Pablo |
Nice illustration
Thanks |
Dear Prashant, Thank you,
Have you had similar problems? I've used snappy a bit more, but I couldn't solve this problem yet. Regards Pablo |
I also have issues with skew faces and am not yet able to tell you how to avoid it.
-Louis |
Louis, thanks for your reply. This problem arises when your surface is almost aligned with the mesh. Then some nodes and/or elements are deleted. In my pictures the element is deleted so the surface has a cube inside the mesh, so in the snap stage the nodes are attracted by the stl surface. But, this is the problem, not the solution :( I don't have a solution either.
Some hints 1) Start with a coarse mesh and add layers. I don't have layer in my pictures. 2) Increase the tolerance in snapControls. In this case the resulting surface has a lot of bumps. These are just some ideas. I think there isn't a unique solution. Regards Pablo |
although if you reduce the max skewness setting in the snappyHexMeshDict to something very low (eg. 0.5 for both) it seems to help.
-Louis PS: thanks for your help also :) EDIT: It seems however that seeting my max skewneww really low makes the mesh really coarse.. perhaps not the best solution..! |
hi everybody,
may it be that the layer-addition phase is the problem here? i also have problems with "bumps" on my surface at the moment, and if i dont add layers and leave everything else the same, everything is smooth... regards, moritz |
Hi Moritz,
for me it seems that layers usually reduce the skewness, however, if I set the skewness tolerance in snappyhexmeshdict to a very low value (ex: below 0.5) that's when I see a very rough surface mesh, even without layers, but, skewness is reduced.. which is useless since my mesh is wrong! |
I agree
Hi Louis and Moritz
I agree with Louis, I test to reduce the skewness option recently. The result is a bumped surface :( . I solved it with a higher value in maxBoundarySkewness option, say 10. But with a lower value in maxInternalSkewness. I have no tested to add layers yet. So, I haven't experience using that option. Regards Pablo |
I had a lot of strange problems with snappy - not exactly as described above. The source was the fvSolution and fvSchemes files. These files in the motorbike tutorial are not the right ones for mesh generation with snappy (they are for running the wind tunnel simulation part of the motorbike tutorial, not the mesh part). Using the fvSolution and fvSchemes from the moveDynamicMesh tutorial fixed all of my mesh problems. Check both files and make sure they have settings for cellDisplacement, cellMotionU, etc.
Good luck, Alan |
Hi Alan,
thanks for pointing this out; however, changing to the fvSchemes and fvSolution files that do contain de CellMotion part does not change anything in the mesh that snappyHexMesh generates for me. My solution so far has been to tolerate the maximum skewness of 5-6 since the results still seem very reasonable and my simulation converges. Best regards, -Louis PS: as for trying a low internalSkew and high boundaryskew, it still does the rough mesh problem. |
Hi there,
I have the very same problem with skewed faces on a very complex geometry. ca. 300 SF with the max bound. skewness of 10 and 2. Does anyone have a solution yet for that problem or is it something you have to live with using sHM? How do you set it up the right way without making the mesh to coarse? Open for any answers. Bye cjm |
Hej all,
I found the same as Louis. MaxBoundSkew ~5 give the best results by keeping the shape of the grid (at least for the case I am working on). Another important thing is to make the blockMesh cells the way that the edge length in X,Y,Z are more or less the same or times 0.5 of each other, i.e. cube like or semi-cube like. This enables sHM to produce less skewed Faces as well. Bye cjm |
You may need to set surface refinement levels so that min and max are the same number.
|
solution for my case
i set up a few cases and raised always another setting. nSmoothPatch killed the skew faces at my case without decreasing the quality of my mesh. |
1 Attachment(s)
Hi everybody!
I have been reading your solutions, but I still have bad cells on surface (see attachement). I am using OpenFoam 2.3.0. I already changed the following parameters: maxNonOrtho maxBoundarySkewness maxInternalSkewness nSmoothPatch tolerance And it did not help. I also tried to run the three snappyHexMesh steps separately, but it did not help. Has anybody experienced such difficulty and could give me a tip? Best regards Benjamin |
Hi!
I have been changing other parameters in snapControls: nSolveIter nSmoothPatch nRelaxIter nFeatureSnapIter without any success. Has anybody any tip to avoid this kind of cell? Best regards BenJ |
Greetings BenJ,
I've had this thread on my to-do list for a while now and only now did I manage to come look into this. I don't know if you've found an answer for this or not, but if not, the problem is that although I think I might be able to help, there isn't much I can do without a test case to work with and test things myself :( Therefore, if you could provide a test case where this occurs, I can take a look into it. Best regards, Bruno |
Folder to solve high Skewness Faces
Dear Forum,
I'm facing this issue as well and I believe it might be the cause why the solver pimpleDyMFoam crashes. Following your recomendations I setted Quote:
Quote:
Quote:
Quote:
So if Bruno ,or anyone, still want like to get the bottom of this, I would be grateful that you help me out. You can find the case in the link: https://drive.google.com/file/d/0Bwm...ew?usp=sharing best regards, Werner ps.By the way, could you check the result when running pimpleDyMFoam ? What might be the origin of it ? Is it really the mesh ? Quote:
|
Dear Werner,
The skewFaces are likely not the cause of OF interrupting. Quote:
You can ignore the error by adding a directive such as Code:
AMI4 Regards, -Louis |
All times are GMT -4. The time now is 00:54. |