Converting to unstructured mesh
 

Hey people,

I'm meshing a certain geometry using ICEM and I've managed to make a good hexa grid, checking the pre-mesh quality gives me determinant greater than 0.35.

I've converted to unstructured mesh (gonna use Fluent to simulate) and no problems at all after checking (no overlapping, no negative volumes, perfect).

However, I could not achieve convergence in my simulating, so I decided to come back for a mesh manipulation. My grid had quite low angle values, so I modified some blocks and I've managed to get angles greater than 13.5, which I reckon is ok. I know there is this suggestion to aim for angles greater than 18deg, but I don't think 13.5 will be a problem, would it?

After doing this modification in the pre-mesh, I've converted it to unstructured mesh again, replacing, not merging, the old one. Now I have some few penetrating elements in a region I didn't even touch and which does not have any influence from the region I've changed things. Why is this?

A also noted determinant values were a bit different from the pre-mesh values, which seems weird. Is it normal?

And finally, I exported the mesh to Fluent regardless of the penetrating elements and Fluent reads it fine, not diplaying any error message or bad grid warning when I check. No negative-volume elements. So, would it affect my simulation? Those few penetrating elements found in ICEM but not pointed out by Fluent?

Thanks guys, I'd really aprreciate some insight on this.
Cheers