← Back to team overview

dolfin team mailing list archive

benchmarking

 

Has anyone done any benchmarking with dolfin?

I've been comparing some dolfin output to some problems I've previously done using COMSOL and I have to tell you I'm not crazy about the results. In particular, I'm interested in linear Stokes flow problems around objects that arise out of homogenization and I find that when I compute the domain averages of the variables ( pressure, velocity ), as one does in such problems, the convergence to a fixed value as the mesh is refined isn't so good. I've tried both iterative and direct solvers too.


-gideon


Follow ups