Large Eddy Simulation (LES) is considered as an efficient alternative to Direct Numerical Simulation (DNS) in order to investigate complex and high Reynolds number flows. In this regard, the evaluation of the LES resolution is still an ongoing challenge. In this paper, different criteria for the resolution evaluation are investigated, namely the ratio of resolved to total turbulent kinetic energy, Piomelli's criteria for the grid spacing and two-point correlations. LES of turbulent channel flows with 180 < Re_τ < 5200 are performed in OpenFOAM using the realizable dynamic LES model, LDMK. The results show that two-point correlations are the most promising method to assess the resolution of LES. Furthermore, by translating two-point correlations to the average eddy size, a clear indication of wall resolved LES is observed by the convergence of the calculated averaged structure size as a function of the grid spacing. The comparison between two-point correlations at different Reynolds numbers extends our knowledge on the scaling of the spatial correlations in stream-wise and span-wise directions. Such understanding of spatial two-point correlations provides a guidance to the estimation of the grid spacing and the number of grid points in the simulation of wall-bounded flows at high Reynolds numbers.
展开▼
