Three-dimensional beamforming of dipolar aeroacoustic sources

• Published in: Journal of sound and vibration Vol. 355; pp. 117 - 134
• Main Authors: Porteous, Ric, Prime, Zebb, Doolan, Con.J., Moreau, Danielle.J., Valeau, Vincent
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 27.10.2015
• Subjects: Aeroacoustics; Algorithms; Arrays; Beamforming; Cleaning; Microphones; Three dimensional; Vibration
• ISSN: 0022-460X; 1095-8568
• DOI: 10.1016/j.jsv.2015.06.030

This paper outlines and compares four beamforming algorithms for accurately localising acoustic dipole sources in a three-dimensional domain, such as noise sources produced by flow–body interaction. These algorithms include conventional cross-spectral beamforming, conventional beamforming with deconvolution via CLEAN-SC, ‘multiplicative’ cross-spectral beamforming and multiplicative beamforming with CLEAN-SC. The latter two algorithms are novel to the field of aeroacoustics and rely on the mutual cancellation of spatially incoherent sources between orthogonally aligned microphone arrays to improve the quality of the source map. The algorithms were used on both synthetic and experimental data. By comparing the performance of each algorithm in terms of source localisation accuracy, source strength estimation and resolution, it was found that conventional beamforming with CLEAN-SC is the preferred method for beamforming aeroacoustic sources in three dimensions, albeit at a higher computational cost than the other three. The results also showed that multiplicative beamforming methods give source maps that are more interpretable than conventional cross-spectral beamforming methods at no extra computational expense. •We propose four 3D beamforming methods that use a small number of microphones.•These methods include ‘conventional’ and ‘multiplicative’ beamforming methods.•The algorithms were tested on both synthetic and experimental data.•Conventional beamforming methods with deconvolution gave the best results.•‘Multiplicative methods’ were a computationally cheap alternative.