[Boitshoko Baloyi]

Greetings!


I hope you are well.


I am trying to simulate the flow in an external-compression, supersonic inlet The basic components of such an inlet are illustrated in an attached image. To be more specific, I am trying to reproduce the results illustrated in the image titled 'Inlet_CFD' which is from a journal paper. The boundary conditions are free-stream all the way around apart from the engine face. The engine face is a pressure outlet. I am currently using the kappa-epsilon, realizable two layer turbulence model. I am also running the simulation steady state.



For some reason, the normal shock wave in my CFD simulation does not converge to a single location. It oscillates after a certain number of iterations as illustrated in the images titled 'Scalar_Scene_1_image_xxxxx'. The 'xxxxx' is the iteration number. The normal shock wave keeps going back and forth and doesn't seem to be settling at a single location. I am not sure why this is happening.


I would like to get the normal shock to settle at a single location. Is there any advice anyone could give me? Your help will be greatly appreciated.



Kind Regards,
Boitshoko

[JBeilke]

Experience says not to use k-epsilon for cases with shocks. Try laminar or k-omega sst.

[Boitshoko Baloyi]

Thank you for the advice. I tried using k-omega but the result was the same.

[bluebase]

Hi Boitschoko,

could you plot and show the spatially distributed continuity residual. Then, in the gradient node of your physics continuum, switch on the TVB limiter, and check the continuity again after further running the simulation.

Best regards,
Sebastian