The experimental synthesis of random pressure fields: Practical feasibility

A previous paper discussed the methodology for the synthesis of partially correlated random pressure fields using a near-field array of loudspeakers. The acoustic sources are optimally driven so that various random excitations are reproduced over a test surface, namely an acoustic diffuse field, a grazing incident plane wave, and turbulent boundary layer fluctuating loads. This paper shows the physical limitation performances and the practical feasibility of synthesizing these random pressure fields in a series of loudspeakers array simulation experiments. Spatial error criteria are proposed on the number of acoustic sources per unit correlation length. They are more representative than mean-square error criteria to quantify the accuracy with which the assumed correlation structures are experimentally reconstructed. Furthermore, structural and acoustic models are formulated to investigate how sensitive is the panel vibroacoustic response to inaccuracies in the synthesized excitations. It is discussed how the direct reproduction of the panel vibroacoustic response with a limited number of loudspeakers should be feasible within the frequency bandwidth for which the modes at resonance well couple with the excitation.