any better way to solve 3D stokes flow?

Yangang Bao

Dear friends,

Anybody can tell me what is a better way to solve 3D stokes flow for tetrahedron elements? I want the pressure and velocity solved seperately(better not to solve velocity directly) since mixed FEM formulation will involve a huge linear system. I am currently using a pressure-based, SIMPLE-like algorithm, but it seems not working. And is it WRONG to use linear element for both pressure and velocity since gradient of velocity is involved in integral stokes equations? I rememberd in T. Coupez's papers, P1+/P1 elements are used and bubble function is added for velocity.

I am using pow law model for viscosity and inertia is neglected in momentum equation, so steady, incompressible stokes flow is assumed.

I appreciate your response.

bao