DLR-RY / TwoPhaseFlow

GNU General Public License v3.0
77 stars 38 forks source link

Issue with convergence #34

Closed Bumble-7 closed 2 years ago

Bumble-7 commented 2 years ago

Dear Authors, Apologies for the previous issues raised. I am a bit confused regarding the pressure convergence in the benchmark cases. I ran the Scriven wedge simulation with the given setup; I observed a couple of things and requested clarification. I am raising this issue mainly because the same issue has been seen in other simulations, irrespective of variations in schemes and solution methodology from my end.

The PIMPLE loop works in the PISO loop. After checking the residual, I see that not at all the timesteps the residual is converged. From my understanding, the initial residual has to be converged before going to the next timestep. Additionally, I have implemented the PIMPLE control in the solver for my simulations; even after multiple outer and inner loops, I see there are few instances where the pressure is still not converged. I am attaching the image from the log file of the Scriven wedge benchmark problem for reference.

logfile

I kindly request some clarifications on the same issue. Thank you for your time.