Large-Scale Simulations of High Speed Turbulent Flows

Abstract

This paper briefly describes a new class of high-order Monotonicity-Preserving (MP) finite difference methods recently developed for direct numerical simulation (DNS) and large-eddy simulation (LES) of high-speed turbulent flows. The MP method has been implemented together with high-order compact (COMP) and weighted essentially nonoscillatory (WENO) methods in a generalized three-dimensional (3D) code and has been applied to various 1D, 2D and 3D problems. For the LES, compressible versions of the gradient-based subgrid-scale closures are employed. Detailed and extensive analysis of various flows indicates that MP schemes have less numerical dissipation and faster grid convergence than WENO schemes. Simulations conducted with high-order MP schemes preserve sharp changes in flow variables without spurious oscillations and capture the turbulence at the smallest simulated scales. The non-conservative form of the scalar equation solved with MP schemes are shown to generate the same results as COMP schemes for supersonic mixing problems involving shock waves.