Localizing aeroacoustic sources in complex geometries: A hybrid method coupling 3D microphone array and time-reversal

Flow-induced noise is widely encountered in engineering domains, and noise source localization is the first step to understand the generation mechanisms. A large part of aeroacoustic noise measurements is performed in wind tunnels by using microphone arrays. A widespread method to identify the noise sources based on these measurements is the beamforming technique. This relies on a Green function, which ideally accounts for source type, flow, and reflecting boundaries. This function is generally not known analytically. An alternative approach is the time-reversal technique considered here. This method consists in propagating the sound field from the microphones back to focus points interpreted as sources. The back propagation can be done numerically with an acoustic propagation solver, which replaces the Green function needed in the beamforming. To date time reversal has been used to localize sound sources with flow in two-dimensions, without considering any reflecting/diffracting boundary. The objective of this work is to consider a three-dimensional situation with flow, and to include the presence of such boundaries to see if this can improve source localization. A synthetic sound source enclosed in a streamlined body is placed in a wind tunnel, and its sound field is diffracted by a nearby airfoil. A three-dimensional antenna with 768 microphones is used to measure the sound field, and time-reversal is performed by solving the three-dimensional linearized Euler equations with a discontinuous Galerkin method. The interest of taking into account the airfoil geometry for source localization is discussed, and some comparisons with the beamforming method are performed.