Discussion Forum

I am now working on a couping problem including 3 modes: n-s eq., Brinkman eq.(to describe the convection in porous media), and conduction and convection eq.. As the Reynolds number in the n-s eq. was quite large, I tried hard to get a converged solution of the velocity field. In soluting the couped problem, I chose the time dependent segregated solver: UMFPACK for the n-s eq., BiCGSTab for the Brinkman eq., and GMRES for the conduction and convection eq.. Then, something strange came up. When I soluted any single one of the three equations, the solver returned a convergenced solution. However, when I soluted them together, the solver never converged. Could anyone tell me why is this and what I can do now? Thank u!