Space-time mesh adaptation for the VMS-Smagorinsky modeling of high Reynolds number flows

Abstract

Traditional methods, such as Reynolds-Averaged Navier-Stokes (RANS) equations and Large Eddy Simulations (LES), provide consolidated tools for the numerical approximation of high Reynolds number flows in a wide range of applications - from green energy to industrial design. In general, RANS modeling is practical when the main interest is the time-averaged flow behavior. LES equations offer detailed insights into flow dynamics and a more accurate solution, but the high computational demand necessitates innovative strategies to reduce costs while maintaining precision. In this study, we enhance the Variational MultiScale (VMS)-Smagorinsky LES model by relying on an adaptive discretization strategy in both space and time, driven by a recovery-based a posteriori error analysis. We assess the effectiveness of the approach in capturing flow characteristics across a wide range of Reynolds numbers through benchmark tests.