Wind Tunnel vs. Free Flying Aircraft: Comparison of UHBR Jet Installation Noise Using Zonal LES on Octree-Cartesian Grids

Abstract

This paper presents compressible Large Eddy Simulations (LES) of Jet Installation Noise using a dispersion relation preserving (DRP) finite-difference method on octree-cartesian grids. A zonal large-eddy simulation is realized by solving the Navier-Stokes equations in non-linear disturbance equation (NLDE) formulation. The steady mean-flow part was taken from a predecessor RANS simulation using the DLR TAU code with unstructured CENTAUR meshes. For the subsequent scale resolving simulation of aeroacoustics, octree-grids with the mean-flow interpolated from the RANS grid are used. The geometry is resolved with different resolution levels with highest resolution in the jet and close to the surface. The far-field resolution is sufficient to maintain acoustic propagation up to St=6.7. As validation case a wind tunnel experiment is simulated and compared against a 1:19 scaled aircraft under approach conditions. The complex geometries of the experimental setup and the aircraft are represented by means of an immersed/embedded boundary condition and accounts for details such as split deployed flaps with gaps and all details of the wing-nacelle-pylon setup. The experiment consists of a semi-wing mounted in the open test section of the aeroacoustics wind tunnel at DLR Braunschweig (AWB) that is combined with an UHBR nozzle simulator and pylon model. Based on the octree-grids, complete meshing of both cases is accomplished automatically within the order of minutes. As the major difference between both cases considered, the experimental high-lift wing operates at a smaller lift coefficient relative to the approaching aircraft to avoid a wind tunnel flow deflection too large. Simulated sound characteristics of the wind tunnel case are juxtaposed with the free flying aircraft setup.