Discussion Forum

Hi

the mesh quality is a factor giving an indication how regular your mesh element is, i.e if it fits nicely into a sphere, with regular shapes it has a high quality, if one node is stretched very far off the elongation wll be grat nd the quality is poor.

But thi si not really enough to tell if a mes is adequate or not. If you have an anisotropic object (material properties and geoemtry rather regular, then a symmetric "high quality mesh" is adequate and better than a distorted one, but if your dependet variables ahve a rather anisotropic solution, i.e. laminar flow in a thin long tube, many physical properties can be derived precisely with a "poor quality" mesh, as based on the default definition given above. with very elongated mesh elements along the axial direction you can still get good results.

Again as often, if you know the final results you know better how to define mesh size, shape density etc, that tells you that one should not be reluctant to remesh and resolve, once a first result has been found, if the results indicates that one can strongly reduce the mesh density in some little critical regions, just to be better prepared to increase the mesh where it seems more critical

Another issue is with ALE, there one should start with a rather high quality mesh as it will be distorted and loose quality as you solve, but here too, by knowing how the solution will distort, one can mesh accordingly to alow for a longer deforming mesh cycle, before one need to stop and remesh

--
Good luck
Ivar

thamks

David