[deltaquarti]

Hi, I'm currently trying to mesh a Fsae car with Star Ccm+, but I have a problem with the volume mesh.
As you can see from the images, the cells on the curvature result in a strange shape on the top side of the undertray, while on the bottom side of the same part the surface is well approximated. I tried to add a custom surface control for this part, reducing the minimum surface size, but it just increased the number of cells without improvement of the geometry.
I cannot use polyhedral mesh due to limits on my pc power.
Any suggestions on how to get a better result?

[me3840]

Check the prisms in that area, sometimes collapsed prisms can leave marks in the surface like that.

The best way to deal with curvature like that and getting the trimmer to do what you want is to split the curved part into a new surface and define its size explicitly instead of letting curvature do it.

If you lowered the minimum size and that didn't change the surface, this means your curvature settings are too low or (in the case of no CAD projection enabled) your input surface is tessellated too coarse.

[deltaquarti]

Thanks for your reply.

I tried to split the curved part into a new surface and I added a custom control for that part only, increasing both the minimum size and the points of curvature, but what I obtained is a better curvature with horrible, stretched cells.

I don't understand why the surface is practically perfect under the part, while above it has all these problems.

[me3840]

I'm not sure, something very strange is happening. Can you post pictures of the remeshed surface and of the CAD surface? Are you using CAD projection and have CAD available?

You say you increased the minimum size and points of curvature? That's not what you'd want to do - you want to drop the minimum size and increase the number of points on a circle. If you increase the minimum, the mesh will coarsen, as you see there I think.

Are there any anisotropic volume controls active?

[deltaquarti]

Sorry, I meant that I reduced the minimum size. CAD projection is enabled and there are no volume controls.

The part is designed as an extruded surface.

I also noticed that the mesh gets worse under a small flap that is located there.

[me3840]

I don't think I've seen the trimmer do that before, I'm not sure what could cause it. What happens if you remove the prism layers?

[deltaquarti]

I didn't expect it: without the prism layers the mesh is practically perfect.

[me3840]

If you lower the prism height dramatically (like by 50%) does it improve a lot? My suspicion is there's something going on when the prism layer projects backward that makes it divide cells a lot.

[deltaquarti]

Yes, it really improves a lot.

[me3840]

That may be your answer then, I don't think you can have that fine of a surface with that thick of a prism layer. From the first section image you posted the final layer transition ratio is bad anyway, so you're really not losing anything.

[deltaquarti]

Thanks, I'll try to change the prism layers settings in the next days to get a good mesh on that part.

[deltaquarti]

I was trying to change some other parameters when I obtained this, just by reducing the gap fill percentage to 0.001. I don't really know the cause of this phenomenon, probably the flap near there has a part in it.

[me3840]

If you reduced the gap fill percentage that far it probably just collapsed your prisms a lot, which we already know that solves the problem. You should look at a cut of the prism layers in addition to just the surface in every mesh.