code_saturne User's Forum

Hallo All,

I try to simulate the turbine via turbomachinery modul with combination of rotation periodicity. I found some old topic on forum about that problem. From this reason I created the simple case to check. If I set the 2 sets of rotation boundary without settings of turbomachinery then the process run without issue. Then if I set the turbomachinery in section of calculation features then I get the message. If I check the joining faces it seems to be OK.

SIGSEGV signal (forbidden memory area access) intercepted!

Call stack:
1: 0x7fd5092f3e30 <+0x5a1e30> (libsaturne-7.0.so)
2: 0x7fd5092fb355 <cs_mesh_get_perio_faces+0x1d5> (libsaturne-7.0.so)
3: 0x7fd50935fe5e <cs_mesh_to_builder_perio_faces+0x2e> (libsaturne-7.0.so)
4: 0x7fd508f1c666 <+0x1ca666> (libsaturne-7.0.so)
5: 0x7fd508f1d556 <cs_turbomachinery_initialize+0x2f6> (libsaturne-7.0.so)
6: 0x7fd50a2d7535 <main+0x2e5> (libcs_solver-7.0.so)

Hello,

There has not been any progress on this issue. We should probably just throw an error reporting this incompatibility earlier on, with a clear/explicit error message, but there could be cases where periodicity and turbomachinery could work together, (if there is no intersection/adjacency between the periodic boundaries and rotor/stator boundaries, which is not the most common case).

The frozen rotor model should work with periodicity (if it fails, try joining the rotors first as you did in your experiments). For the transient rotor, handling periodicity would be complex. If you look at the diagrams in the theory documentation, you will see that combining the principle of periodic interface matching and periodicity will not be trivial when changing rotor position.... Since this is a bit tricky, and the demand for this feature is not frequent, it has not been a priority so far...

Best regards,

Yvan

Hello Yvan,
My simulation was with sittings of frozen rotor.
I tried to simulate this case again and I also joined the rotors first. But I had the same error message as before.
Is there some possibilities to check if it is possible to simulate the turbomachinery with frozen rotor + rotation periodicity?

Hello,

In this case, if you can provide a small test case, this might not be too difficult to fix (no guarantees, but I can check). I probably won't have much time to do so before the end of this month, but should have more time after that.

Best regards,

Yvan