Progress in Aero-Acoustic Optimization Techniques in DLR's VicToria Project

Abstract

The methods of the aero-acoustic optimization toolchain developed by DLR within the VicToria project are further analyzed. From previous investiga- tions, it was settled that an unsteady panel code with a free-wake model poses a fast alternative to higher-order CFD simulations to predict blade-vortex interaction noise. In this paper, a brief validation of both methods against the HART II rotor is done. Design parametric studies with UPM are carried out for which single CFD simulations are placed to examine the correlation between both methods. In the end, an optimization with the panel code as well as CFD is conducted to see how good the final designs match. From this research it is observed that despite that CFD predicts too quiet noise carpets, the overall trends are plausible. They are contradicted by the panel code to a certain degree, especially for blade sweep and anhedral. It is discovered that using CFD directly with five design parameters and ten samples arrives at quieter rotors than using the panel method with seven parameters and 210 rotor samples.