Adaptively Tuned High-Order Methods for iLES

Abstract

[Abstract]: Implicit Large Eddy Simulation (iLES) is popular for modeling high-Reynolds-number turbulent flows due to its simplicity and efficiency. High-resolution numerical schemes in iLES mimic subgrid-scale models by using numerical dissipation/dispersion errors. Balancing these errors is crucial to avoid unphysical energy build-up or excessive diffusion, especially in compressible flows with discontinuities and smooth features. This work introduces an adaptive dissipation/dispersion adjustment (ADDA) algorithm for the CWENOZ scheme in a finite- volume framework for unstructured meshes, tested in subsonic, transonic, and supersonic regimes. The ADDA algorithm enhances robustness and scale-resolving capabilities, yielding efficient and physically meaningful results, and is available in the open-source UCNS3D CFD solver.