code_saturne User's Forum

Hi all,

Reading through the documentation of Code Saturne 4.0 I've came across following statement about the k-omega SST (page 152, user guide):
"The k − ω SST model provides correct results whatever the thickness of the first cell. Yet, it requires
the knowledge of the distance to the wall in every cell of the calculation domain."

Does it mean, that the implementation is a "all y+" one? I mean able to utilize wall functions in case of to large y+ to be able to efficiently integrate to the wall?

Best regards,
Mariusz

Hello,

I'm not sure what you mean by "all y+", but in the case of the k-omega model, the wall distance is computed for all cells, but y+ is not maintained explicitly on all cells). I'm not an expert in turbulence models, so I can't say much more without checking the source and documentation in detail.

I do have doubts about the statement relative to the k-omega model, as I am not sure how this interacts with the scalable wall functions option. Again, others could answer better than me.

Regards,

Yvan

Thank you Yvan,

I was curious because I've got pretty reasonable results for heat transfer simulations using k-omega SST on the meshes varying between y+ of 14 and 56 (so including the buffer layer).

I have found in Code Saturne BPG 1 – Basic phenomena the following:
The k-omega SST model (ITURB=60) may be used if it is not clear whether the mesh refinement at
the wall makes the mesh suitable for a high Reynolds or a low Reynolds approach. It is not advised to
use this model otherwise. If possible the v2f model should be used.

what would kind of also answer my question.

Anyways, thank you again Yvan,
Best regards,
Mariusz

Hello,

You can also find quite a few model comparisons on the University of Manchester website (http://cfd.mace.manchester.ac.uk/twiki/ ... /TestCases), though all may not be up to date, and navigating is not always easy. k-omega is often reputed quite good for dymamics, but not great for thermal aspects, but test results might help to know when/if this is the case. Unfortunately, the Code_Saturne validation suite is not publicly viewable (as it contains a mix of confidential and non-confidential cases).

Regards,

Yvan