Acoustic Beamforming: innovative methods to overcome current limitations

Since its introduction acoustic beamforming has imposed itself as the principal technique for the mapping of acoustic sources in the far field. Even if some decades have passed since it was firstly used, important limitations still remain at the state of the art. In particular, the main shortcomings are relative to three aspects. The first one is the beamformer resolution capability, that strictly depends on the structure of the array, i.e. on the hardware available for the measurement. The second is the relative phase assessment amongst sources, that can not be achievable by the known state of the art techniques. The third is that the measurement background noise can be high and cover the phenomenon of interest, in particular in some applications as the aeroacoustic ones. The aim of the present research is to introduce new methods to overcome the limits above. At first the technique called VAMOS (Virtualized Array for MOving Sources) is introduced and treated. It can be used in all those cases when the object under investigation is moving, such as pass–by or fly–over measurements. Once the measurement is performed, VAMOS exploits the object motion to virtually extend the array used for the acquisition: in this way the array aperture, i.e. its resolution, is increased in the direction of motion. Numerical validations, performances and uncertainty analysis are treated in the dissertation. Doctoral School on Engineering Sciences Università Politecnica delle Marche Secondly, the problem of relative phase assessment is treated. Two new techniques are described. The first one is the Phase Mapping by Monopole Substitution (PMMS). It constitutes a post– processing step to be performed after beamforming calculations to complete the obtained sound power beamforming map with that of phase. The technique is presented with a complete uncertainty analysis and both numerical and experimental results. The second technique introduced is an inverse technique directly proceeding by Suzuki’s L1 Generalized Inverse Beamforming. The map of sources relative phase is obtained by modeling the sound field with a orthonormal base of monopoles. Finally, multi–sensor data integration is exploited to highlight and understand aeroacoustic noise generation mechanisms. In particular, the acoustic field due to a wind turbine will be analyzed by means of Coherent Beamforming: aeroacoustic contributions are separated by means of a reference signal sensitive to the other disturbing sources, such as mechanical ones. At the end, the acoustic field of a laminar separation bubble will be studied by means of beamforming and causality correlation approach.