Comparative Assessment of Synthetic Turbulence Methods in an Unstructured Compressible Flow Solver

Kurzfassung

The industrial use of scale-resolving methods in high-Reynolds-number applications (e.g., airplanes) requires efficient hybrid approaches, in which flow regions of interest are resolved by local LES embedded in a global RANS simulation. The transition from RANS to LES has to be augmented by synthetic turbulence to quickly transform the modelled turbulence into realistic velocity fluctuations. For compressible flow solvers such as the DLR-TAU code, the consistent treatment of density and temperature, and the generation of spurious noise need to be considered, as well. Over the last years, different methods for synthetic turbulence Generation have been implemented in the unstructured TAU code: the Synthetic-Eddy Method (SEM), the Divergence-Free SEM, and an extended variant of the Synthetic Turbulence Generator of Shur et al., denoted by Random Velocity Field Generator (RVFG). Their implementations in TAU allow for flexible grid-independent turbulence injection in multiple planes or volumetric regions via forcing source terms in the flow equations. While individual assessments for different flow cases have been conducted before, this paper presents the first systematic comparison of all These methods in TAU for the same test cases in a unified framework, ruling out uncertainties due to numerical and implementation details.