DES ( SA+LES) vs DES ( kwSST+LES) for bluff body

Phil

flow separation / vortex shedding.

Does anyone have any experiance with the accuracy of flow separation prediction when using SA for the URANS portion of DES vs. using kw SST for the URANS portion of DES?

My understanding is that the kw SST generally performs well in separation for pure URANS problems.

Does anyone have experience on the separation prediction performance of SA vs kw SST in a DES situation?

I would have guessed that since separation must initiate in the URANS portion of the domain the kw SST model would be preferable?

As it is CFX's DES model which uses kw SST is hopelessly over predicting mu_turb in my low Reynolds number shedding case which is supressing all vortex shedding. I am hence thinking of trying the case with Fluent which gives both options for the URANS portion of the DES model (SA and kw SST).

Thank you.