[newbie384]

Hi

I use 2d axisymmetric with swirl to simulate heat transfer of fluid between two concentric cylinders, with the inner cylinder rotating and outer cylinder stationary. When I ran the simulation with pressure based solver, it diverged imediately but when I switch to density based solver, the residuals go down slowly (but with oscillation). May I know why pressure based solver gave divergence results? Is it because it is less stable than density based solver?

Additional information of my settings.
1) pressure inlet
2) pressure outlet
3) moving reference frame
4) use stationary wall for rotating inner cylinder
5) use moving wall for stationary outer cylinder with absolute speed=0

Thanks in advance.

[markrenten]

If your fluid does not have a fixed density , so compress able , the you should choose density based solution .

[newbie384]

hi

thanks for the advice. my fluid is air and I defined the density to be compressible according to ideal gas law. Density based solver is more appropriate for my case.