Hi, there,

I am looking for a robust scheme for mesh smoothing (unstructured quadrilateral mesh). I tried the so-called Laplacian smoothing. It works but sometimes generates invalid grid. It there any scheme that guarantees valid elements? I heard of the ¡°Optimization-based smoothing"? How does it work? Any input is appreciated.

Tony

Hi Tony,

It may sound old fashioned, but the best unstructured quadrilateral mesher I have seen is Qmesh. It was developed by Jones in Sandia in the 70's (including several Laplacian smoothing variants), and had very clear documentation. It was sold as source code. Maybe it is still out there.