Learning Forum

[dipika berry]

Hi all, i have been trying to do mesh convergence check for the below. Here, the mesh is refined at the contact region. However, with the same contact sizing the number of mesh elements keep varying. It varies from 17.4k elements to 21k elements and is very random. Suspected due to tetrahedral mesh. Could you advise any way to overcome this issue. 

This is the mesh, with global element size of 2000 um and contact sizing 300 um.

[mohan.urs]

Hey dipika,

Yes, the problem of different number of elements is due to the use of tetra elements, to overcome this use hexa elements (the above geometry looks sweeepable so it should be easy, if not dive into shared topology or multizone method documents to make the best use of hexa meshing for complex geometries). Since hexa mesh produces a structured mesh the number of elements do not usually vary but in case of tetra, it is a unstructured mesh therefore the meshing algorithm might produce different number of elements each time.

Thankyou.

[dipika berry]

Hi Mohan,

Thanks for your reply. So, is there no way to enforce structured meshing algorithm with tetra mesh, through scripting or commands? As unfortunately, I have been using this mesh type and wanted to just check convergence.