Reply To: Avoiding spurious peak in residual convergence plot

[sthavishtha]

Dear Mathias

Thanks for the suggestion. Here is the updated plot for the same Re and different grid sizes. The streamline profiles from 64 x 64 don’t exactly match with that at higher grid sizes due to grid dependence. However, the profiles at higher grid sizes (512 x 512) are still off from the literature.

I do agree that one cannot expect a steady state solution for higher Re. But Re = 1000 also is not very high enough to expect an unsteady solution state.

I am now planning to try this by adopting MRTdynamics2, as you state in the code that this uses relaxation times for higher stability. So, is replacing MRTdynamics with MRTdynamics2 in the code and using the same descriptor MRTD2Q9Descriptor the right approach to invoke this?

Regards
Sthavishtha