Reynolds-averaged Navier–Stokes initialization and benchmarking in shock-driven turbulent mixing

Abstract

We investigate a strategy for benchmarking Reynolds-averaged Navier–Stokes (RANS) models by comparing moments extracted from averaged large eddy simulation (LES) data and those predicted directly by RANS. We consider the Besnard–Harlow–Rauenzahn (BHR) RANS approach designed for variable-density compressible flows, which has been previously applied to a wide variety of turbulence problems of interest. We focus on the model's ability to predict moments relevant to shock-driven material mixing. A prototypical inverse chevron shock tube configuration is considered, for which laboratory and previous LES studies are available for comparison and validation. We show that when appropriately initialized, BHR is capable of accurately capturing various characteristic integral measures of the flow; strategies for initialization are demonstrated while addressing sensitivity of BHR predictions to closure and initialization specifics, initial material interface conditions, and grid resolution. The reference simulations are performed using implicit LES based on the Los Alamos National Laboratory RAGE hydrodynamics code.

Keywords:

• large eddy simulation
• Reynolds-averaged Navier–Stokes
• turbulence modeling: Reynolds averaged
• turbulent mixing
• compressible turbulence

Acknowledgements

Los Alamos National Laboratory is operated by the Los Alamos National Security, LLC for the U.S. Department of Energy, NNSA under Contract No. DE-AC52-06NA25396.