2d incompressible polar problem
 

I am having trouble solving the polar form of the Navier Stokes Equations when applied to a donut shape physical grid.

The lower boundary is moving at 10 RPM and the upper boundary is stationary.

A book recommended that I solve the momentum equations sequentially and thus I did the following.

The method used per iteration is the point by point method or Gauss seidel.

I solve the angular momentum equations first until I get a linear result. Then I solve the radial velocities. As the solution solves an oscilation appears near the stationary boundary. No amount of relaxation factors will change it.

I am using this within a SIMPLE algorithm.

Any help would be appreciated.