Compressible Backwards Facing Step Simulation

[JCFD4]

I am working on a compressible simulation of a backwards facing step, simulating at 30,000ft (Pressure 30kPa, velocity 180m/s, temp -44C).
I am struggling to setup the project for drag coefficient to be calculated.

I have tried ramping the velocity to 180m/s over 600secs, the solver timed out after 24hrs and only reached 0.02secs (see coefficients plot below).

The setup for the above run is a transient simulation. I have run different courant numbers, maximal time steps, relaxation factors and velocity ramping. But the simulation either fails or does not make any progress.
The meshing log shows no issues with the mesh. It has been refined through the inflated boundary layer, a surface custom refinement on the step surfaces and volume custom refinement on the region to the rear of the step.

Any help in the issues faced would be greatly appreciated.
Project: https://www.simscale.com/workbench/?pid=2313403631859990604&mi=spec%3A55d2bb17-2bbd-4b62-b74c-03eff8032174%2Cservice%3ASIMULATION%2Cstrategy%3A1&ps=analysis%2Fnumerics

[shenoy]

Hey @JCFD4 ,

Ramping definitely helps with higher speeds like this. This steady-state compressible tutorial does something similar as well to aid convergence.

I can see that for most of your unfinished / cancelled transient runs, your setting for adjustable time step is ‘true’:

This forces the time step to be too small sometimes depending on the other specified parameters and can take very long especially for higher speeds like this. Note that this would only be sensible if you are really interested in the transient effects of the flow and not for steady-state drag.

However, I see you were able to run a few steady-state simulations successfully. Since it’s been a while, do you still need any more support with this project?

Best,
Satvik