Search for a command to run...
Acoustic beamforming is a fundamental phased microphone array technique used for early stage noise source identification in rotating machines. However, its performance is limited by the array’s size and the spatial density of microphones. A cost-efficient approach to synthetically enlarge and/or densify microphone arrays involves multi-pass measurements, where a smaller prototype array is moved sequentially in front of or around the machine to effectively form the final array. The acoustic fields recorded during multiple measurement passes are resynchronized to estimate the final beamforming map, which is superior to the individual ones. Traditional resynchronization techniques designed for stationary measurements are no longer suitable, as the acoustic field radiated by a rotating machine operating at variable-speeds is nonstationary and contains orders which are modulated tones. This paper introduces a novel approach called Multi-Pass Order-Based Beamforming (MPOBBF), which resynchronizes tonal sound fields sequentially captured from variable-speed measurements. The complex envelopes of an order of interest are extracted from multiple sequential measurements and aligned into a unified dataset, over which Order-Based Beamforming (OBBF) is applied for source identification. OBBF is a microphone array technique that can localize and quantify sources emitting tonal noise in rotating machines operating at variable-speeds. The proposed approach is theoretically detailed and experimentally verified on an induction motor. • Allows resynchronization of tonal sound fields from multiple variable-speed measurements. • Virtually enlarges the microphone array and increases its spatial density. • Significantly lowers instrumentation costs compared to simultaneous measurements.