[mb.pejvak]

Hi friends;
I want to apply GMRES (m) relaxation method on the code I developed. I used FVM in structured grid with SIMPLER method for coupling and TVD (for convection term's discritization). I applied Gauss-Siedel iteration method, and the code converged, but when I apply GMRES (m), it diverges. I don't know why. I should mention that I applied the subroutine of GMRES(m) on a diffusion-based equation, and it worked out well. So I think it is problem in the way I apply it on NS equation or some important points should be considered in applying the method on NS equation.
SO I 'd really appreciate it if anyone could help me.

thanks in advance