Computational fluid dynamics (CFD) is nowadays applied extensively in all aerodynamics-based topics of aircraft design, development and optimization. Since standard CFD approaches still lack accuracy in areas of highly nonlinear, unsteady flows close to the borders of the flight envelope, the aeronautical industry is increasingly willing to apply more costly scale-resolving methods, if such are able to provide a real predictive alternative for critical situations. While Large Eddy Simulation (LES) may be a viable option in certain areas, it is still far too costly-if not impossible-to apply it to high Reynolds number flows about even moderately complex configurations. Thus, the family of Hybrid RANS-LES Methods (HRLM) currently appears to be the best candidate for the next generation of CFD methods to increase solution fidelity at an industrially feasible expense. HRLM have been proven to perform considerably better than conventional Reynolds-Averaged Navier-Stokes (RANS or URANS) approaches in situations with strong flow separation, but they are less effective once they have to deal with weakly unstable 1 flows, e.g. thin separation regions or shear layers in general. In such cases, resolved structures develop only very slowly, resulting in areas where the total amount of turbulence (both in modeled and resolved terms) is unphysically low. These so-called "Grey Areas" often lead to results that are worse than those of RANS simulations. Unfortunately, such grey area situations appear in many of the flows important close to the borders of the flight envelope, e.g. near maximum lift. Accordingly, there is a strong necessity to mitigate the grey area in order to provide the industry with Hybrid RANS-LES approaches that are trustworthy for relevant flow situations. Precisely this was the primary objective of the Go4Hybrid project, which focused on two main aspects: to provide viable extensions to HRLM mitigating the grey area problem in non-zonal approaches and to improve embedded methods, such that they are applicable to arbitrarily complex geometries. 1 Note, that unstable refers here to an easy switching from an unresolved to a resolved modeling of turbulence.

No abstract

No abstract

No abstract

No abstract