Identification of anomalies in microphone array measurements of trailing edge noise by eigenstructure analysis

Abstract

An eigenstructure analysis is used to identify anomalies in large acoustic array measurements of airfoil trailing edge noise. Cross-spectra of 102 microphones distributed in a spiral configuration were inspected considering a range of inflow conditions and airfoil configurations. A measurement expectation is set by repeat measurements of non-anomalous data and is used for the analysis of additional data. This method is based solely on the comparison between the eigenstructure of cross-spectral matrices and can be used to efficiently inspect data for anomalies without having to beamform, as is conventional. Thus, processing and reviewing beamform maps can be relegated to post-processing after the experiment when more time is available.