Evaluation of Enhanced Gray Area Mitigation Approaches Based on Jet Aeroacoustics

Abstract

An investigation of various eddy viscosity models for accurate scale-resolving simulation of aerodynamics and aeroacoustics of a turbulent jet is presented. The models are considered from the perspective of solving the so-called gray area problem. This problem is related to nonphysical delay in the transition from Reynolds-averaged Navier–Stokes (RANS) to large-eddy simulation (LES) in hybrid RANS–LES approaches such as detached eddy simulation (DES) when applied to shear-layer flows. The performance of recently developed dynamic adapting subgrid length scales ([Formula: see text], [Formula: see text], and [Formula: see text]) and subgrid LES models ([Formula: see text] and S3QR) separately and together is demonstrated. The object of the investigation is an immersed subsonic turbulent jet. The simulations are carried out on a set of refining meshes using two different scale-resolving numerical algorithms with basic and higher-accuracy schemes on unstructured meshes. The evaluation of various eddy-viscosity models within the DES approach is mostly focused on the analysis of far-field noise. The study has clearly demonstrated the importance of both a numerical scheme and a subgrid turbulence model. The results show that some of the considered techniques provide the necessary accuracy to predict the noise generated by a turbulent jet. The features of the considered approaches are identified and discussed. Recommendations are formulated for the choice of the LES model and subgrid length scale for this kind of problem.