A Hybrid Time-Frequency Approach for the Noise Localization Analysis of Aircraft Fly-overs

A hybrid time-frequency approach based on acoustic beamforming has been successfully developed in order to determine the absolute contribution of the aircraft noise components measured during fly-overs. The method, derived from DAMAS, 1 accounts for the fact that the sources move relative to the microphones. Indeed, the motion is responsible for a frequency shift of the sidelobes and a strong modification of the point-spread functions compared to the static case. The method developed is hybrid in the sense that the beamforming algorithm is applied in the time domain while the point-spread functions are approximated in the frequency domain. When the sound sources are characterized by broadband spectra, it becomes possible to discard the frequency coupling between the sources and the sidelobes. This enables to reduce the computing time considerably. In the present publication, the method is applied to AIRBUS A340 fly-overs with the high-lift devices deployed, landing gear up, and the engines running at idle. The study is focused on the influence of some processing parameters (grid spacing, number of iterations, cut of the focusing grid, and frequency bandwidth) on the source breakdown. It turns out that the noise spectra of the aircraft components are almost independent of reasonable variations of these parameters. This indicates that the method is relatively robust. A detailed investigation of the noise sources (flaps, slats, and engines) with respect to the variations of the airspeed, the ECAM position, the engine rating, etc. is not part of this work.