code_saturne User's Forum

Hi,

I have recently been experimenting with the use of hybrid turbulence models (focusing on DDES, using the k-omega model as a basis). I suspect I may not be setting it completely correctly. The graph of the hybrid_blend variable is somewhat strange in my opinion, and I also notice that the almax variable is automatically initialized as the cube root of the domain volume (despite me setting another value).

To set the model, I use the source files cs_user_parameters.c and cs_user_initialization.c (I attach them). I also show details of the graph of the hybrid_blend variable (I note that it does not vary smoothly in space, and that the jump in the first cells from the wall could be large enough to bring inaccuracy).

Thanks in advance.

Hello,

Where do you set the value of almax that is overwritten ? In cs_user_parameters.c ? Depending on whether you do this too early or not, it might be a bug. The cube root of the volume is indeed the default value.

For the rest, I'll suggest to Jean-François to take a look at this.

Best regards,

Yvan

Yvan Fournier wrote: Sun Oct 01, 2023 6:47 pm Hello,

Where do you set the value of almax that is overwritten ? In cs_user_parameters.c ? Depending on whether you do this too early or not, it might be a bug. The cube root of the volume is indeed the default value.

For the rest, I'll suggest to Jean-François to take a look at this.

Best regards,

Yvan

Thank you Yvan,

First, I set the desired almax value (=2.47648095) in the GUI only. Seeing in run_solver.log that the default value (cube root of the domain volume, 4.24 in this case) was not modified, then I set the same desired value both in the GUI and in cs_user_parameters.c, but it did not make a change in the log either.
Thanks also for what you can get about the analysis of the value of the blending factor.

Hello,

I just fixed the almax issue in the Git "master" and "v8.0" branches (synced to GitHub).

Best regards,

Yvan

Thank you Yvan