I want to perform convergence rate study of my FEM code for any benchmark problem. Code can solve Linearized NS but i have to select a problem which has analytic (exact) solution and at the same time it is not too simple that Numerical results are exactly equal to exact results even for coarse grid. The idea behind the convergence rate study is to prove that as we refine the mesh we get closer to the exact solution at a rate which depends on the order of polynomial (Shape function) and it is give in the literature. I want activate only the stokes operator and i will switch off the convective term for this study. So I am looking for a stokes problem which has exact solution. I tried Couette flow but it is too simple problem. Does Drive Cavity flow has analytical solution? Does any body know about Engleman Problem? Thanks

May i suggest to take a look at "Code Verification by the Method of Manufactured Solutions" by Patrick Roache, Journal of Turbomachinery March 2002? It's an overview of an alternative way to validate a computational code. Hope it helps