[sfepy] Re: Stress and strain convergence with mesh size

Hi Vladimir,

Thank you very much for the response. Regarding "Around sharp edges, there is a nonlinear distribution of stresses/strains and the results of FE simulations are sensitive to the mesh size. ", let's say I'd like to estimate the strength based on LEFM according to the local failure strain of the elements, ( i.e. once the per element strain reaches to a specific value during the loading, we say the material fails), then since the local strain/stress around the edge changes with the change of mesh size, it means that I cannot define a unique strength value independent of mesh size. In that case, is there another way I can define the strength or any other way to avoid/weaken this mesh dependency ?

Best, Ali