Investigation of the explicit cutoff filtering in Large Eddy Simulation

Abstract

Implicit filtering is a common procedure in Large Eddy Simulation (LES). Problems with implicit filtering have led to the development of an alternate method called explicit filtering. Most explicit filtering methods are expensive, but filtering the entire flow field is one of the least expensive methods. Closeness of the filter shape to the sharp cut-off is a significant parameter in this kind of filtering. The aim of this paper is to study the effects of the closeness of filter shape to the sharp cut-off on flow statistics and to compare the Pade-type filter with different discrete filters that have shapes close to the sharp cut-off. Turbulent channel flow was chosen for the simulations. The finite-volume method was used along with the PISO algorithm. Moreover, the second-order central difference scheme was applied for spatial derivations and the Crank-Nicolson method for time integration. The problem was solved by three different types of discrete filters, all of which had shapes close to the cut-off filter. The results suggest that the use of explicit filtering by sharp cut-off filters improves the specific statistics of the wall-bounded turbulent channel flow. Pade-type filters produced the most desirable results.