Adaptive wall treatment for the elliptic blending Reynolds stress model

Jean-François Wald (1, 2) , Sofiane Benhamadouche (1) , Remi Manceau (3, 2)

1 EDF R&D MFEE - Mécanique des Fluides, Energies et Environnement
2 LMAP - Laboratoire de Mathématiques et de leurs Applications [Pau]
3 CAGIRE - Computational Approximation with discontinous Galerkin methods and compaRison with Experiments

Wall functions are widely used in CFD in order to significantly reduce the computational cost compared to so called Low-Reynolds number formulations. They are, however, particularly restrictive in terms of meshing as they require the first calculation point to fall into the logarithmic region. Industrial simulations of internal flows, such as the ones encountered in nuclear applications, are particularly challenging due to their inherent complexity that makes it difficult to satisfy those conditions everywhere. The present study focuses on a new algebraic adaptive wall treatment for the Elliptic Blending Reynolds Stress Model (EB-RSM) by extending some of those recently proposed approaches. Blending functions that ensure a correct asymptotic behaviour at the wall for the velocity and the turbulent variables are introduced and boundary conditions are prescribed at the first near-wall cell. The approach shows very promising results on fully developed channel flows, comparable to what is obtained using a numerical integration down to the wall.

Domaines: Milieux fluides et réactifs

Jean-François Wald, Sofiane Benhamadouche, Remi Manceau. Adaptive wall treatment for the elliptic blending Reynolds stress model. ETMM10: 10th International ERCOFTAC Symposium on Engineering Turbulence Modelling and Measurements, 2014, Marbella, Spain. ⟨hal-01051800⟩