Is PyFR suitable for supersonic cases?

General
1

Dear all,

I've tried a few of supersonic cases and failed each time. And there's no
supersonic cases in published posts. I'm wondering if PyFR is suitable for
supersonic cases
in normal working condition?

Best regards,
Ray

2

Hi Ray,

How did it fail? What Mach numbers were you trying?

We have run cases with relatively low Mach numbers (<2) with PyFR using artificial viscosity (see the User Guide):

  • shock-capturing — shock capturing scheme:

    none | artificial-viscosity

    Cheers

    Peter

3

Hi Peter,

I’ve tried this case, Mach 3 Wind Tunnel With a Step. It’s in the book, AN ANALYSIS OF STABILITY OF THE FLUX RECONSTRUCTION FORMULATION WITH APPLICATIONS TO SHOCK CAPTURING. I got RuntimeError: NaNs detected at t = 0.05… Is there any problem in my configure file?

Thanks!

Ray

step_str.pyfrm (1020 KB)

step_str.ini (997 Bytes)

4

Hi Ray,

I think Forward facing step problem with M=3 is one of the tough benchmark problems of shock-capturing methods.

It may yield negative value issues around the expansion corner and you may need some special cures (ex. refining mesh around the corner).

Furthermore, you didn’t turn on the artificial viscosity option in solver setion as mentioned the previous Peter’s mail.

The current shock capturing method (artificial viscoisty) of PyFR runs quite well for the flow with weak shock or oblique shock.

Please find the attached sample mesh and ini files. I think you can run similar supersonic cases.

Regards,

Jin Seok

doubleshock.pyfrm (114 KB)

doubleshock.ini (1013 Bytes)

5

Hi Jin Seok,

Thanks a lot! It works.

Regards,
Ray

6

A post was split to a new topic: Double Shock case