Towards self-adaptivity in hybrid RANS/LES based on physical criteria

Hybrid RANS/LES methods can produce more reliable results than RANS with a reasonable computational cost. Thus, they have the potential to become the next workhorse in the industry. However, in continuous approaches, the location of the switching between the RANS and LES modes is based on the mesh and have a significant impact on the results. The present paper aims at developing a self-adaptive strategy based on physical criteria to mitigate the influence of the user's meshing choices on the results. The method is applied to the backward-facing step with the Hybrid Temporal LES (HTLES) model, but is applicable to any other hybrid approach. Starting from a RANS computation for initialization, successive HTLES are carried out and the mesh is refined according to the criteria. The results obtained show that the method converges and significantly improves the results when compared to RANS, with no intervention from the user. The comparison of the results with the DNS is very encouraging.