Optimization of microphone array arrangements for wavenumber-frequency spectral measurements

Abstract

The wavenumber-frequency spectral (WFS) analysis is a well-established technique widely used for the separation of pressure fluctuations in turbulent flow fields into hydrodynamic pressure fluctuation (HPF) and acoustic pressure fluctuations (APF). The WFS analysis requires an array system comprising multiple microphones. The aim of this research is explored how the number and placement of microphones affect the accuracy of the analysis and determine an optimized microphone array layout for analyzing vehicle noise. We conducted separation analyses of HPF and APF using pressure fluctuations from a diffuse sound field generated by a turbulent boundary layer with varying numbers of microphones and placement patterns. The results showed that a well-balanced combination of spiral or heterogeneous random array arrangements with 96 channels or more is necessary to effectively separate APF and HPF in the frequency band from 500 to 5 kHz, which is crucial for automobile noise analysis. Based on simulation results, we constructed a compact and cost-effective system addressed issues related to wiring and complex data acquisition by utilizing a digital MEMS microphone and an FPGA board. Consequently, the developed system can be mounted on a vehicle and utilized for WFS analysis in the flow field surrounding the vehicle.