The purpose of this work is to improve the signal-to-noise ratio of acoustic measurements made in a wind tunnel with a microphone phased array. The functional beamforming method is improved to compensate for effects of acoustic propagation through turbulence, and to correct for inaccurate steering vectors. This increases the accuracy with which acoustic measurements can be made in a non-acoustic or acoustic wind tunnel.
The objective is to make accurate measurements of the acoustic source strength of a test model in a noisy wind tunnel. The prior art measured the loudest noise. One prior approach is to use a phased array of microphones and take two data sets: a background data set with the model missing or idle, and a full data set. The background array cross-spectral matrix is subtracted from the full cross-spectral matrix, and the difference is used for beamforming. This does nothing to address turbulent decorrelation, and is limited in the dynamic range improvement by the degree of sameness of the wind tunnel noise in the two datasets. A prior way to compensate for turbulent propagation through a boundary layer or shear layer is to reduce the size of the array depending on frequency. This requires extra microphones and makes the results dependent on an arbitrary parameter: the array diameter.
The estimation method of cos q is a novelty of this system. Separate determinations of s and cos q are made. In prior art, these factors could get confused with each other, causing errors. It will now be feasible to measure UAV noise in nonacoustic wind tunnels, filling an important market need.