Method for isolating the tonal components of counter-rotating turbomachinery phased array microphone data for beamforming

Counter-rotating turbomachinery systems can be used in numerous applications where fuel efficiency and load reduction are crucial, such as the engines of drones or aircrafts. These unducted rotors are associated with a large magnitude of noise emission. Beamforming combined with microphone array measurements provides an appropriate method for noise source localization. The tonal components generally dominate narrow frequency bands, but in some cases their levels are of the same order of magnitude or lower as the broadband components. Using pre-processing methods on the recorded noise signal, the broadband component can be generated. With the use of cross spectral matrix operations, the broadband components can be subtracted from the original signal, extracting the tonal components of the noise of the investigated turbomachinery. This was not possible earlier using conventional single microphone pre-processing methods, due to the counter-rotation of the rotors and slight variations in forward and aft rotor rpm values throughout the sample. On the beamforming maps of the tonal signal resulting from the cross spectral matrix based pre-processing method, many of the formerly hidden tonal noise sources are now visible, which helps in unambiguously separating apart and identifying the tonal and broadband noise sources which have similar amplitudes.