EXTRACTING AND CHARACTERIZING BLADE-VORTEX INTERACTION NOISE WITH WAVELETS

Applications of a discrete implementation of tile wavelet transform (W T) to the analysis of blade-vortex interaction (BVI) noise are present ed, The method decomposes a signal into a series of orthogonal subband s logarithmically spaced in frequency while simultaneously preserving temporal information, A BVI detection algorithm was developed which ta kes advantage of the prominence of BVI noise in certain subbands, A me thod for extracting BVI noise from other noise sources is described. T hree metrics are tested for suitability as BVI estimators. One of thes e, a root-mean square (RMS) metric, characterizes BVI noise severity, hut requires the added computational effort of inverting the transform , The other two metrics are computed using the information available f rom the WT. The first of these, computed from the amplitudes of BVI ev ents in a single subband, correlates with peak-to-peak pressure and is relatively insensitive to noise, The second of these, based on the ex ponential behavior of peak amplitudes across subbands, is apparently u nable to characterize BVI noise severity and is highly sensitive to no ise.