Microphone Array Method for the Characterization of Rotating Sound Sources in Axial Fans

B1 - Signal processing for noise source location/characterization

To successfully reduce the noise emitted by an axial fan it is mandatory to gain as much knowledge as possible about the source mechanisms on the fan. Methods based on microphone array measurements provide a powerful tool for the characterization of acoustic sources. In order to determine source positions and respective magnitudes of the emitted noise, sophisticated algorithms working in the frequency domain are being used widely.
These methods, evaluating the measured cross-spectral matrix, have proven to be able to successfully and accurately reconstruct the source parameters. However, their application is limited to stationary sources, whereas for moving sources less accurate time-domain methods have to be applied.
When dealing with rotating sources, as is the case with axial fans, it is often possible to overcome this constraint by arranging the microphones such that the measured signals can easily be interpolated between the microphone positions, simulating an array rotating at the exact same rate as the presumed sources.
The method presented in this contribution implements a virtual rotating microphone array in order to regard the rotating sources as being non-moving and thus allow the application of algorithms based on cross-spectral matrix evaluation. The possibility of separating stationary and rotating sources using this method will be discussed and prerequisites for its application will be specified.
For experimental validation, microphone array measurements with a four-bladed common household axial fan were conducted and evaluated using the new method. The presented results include the localization of different sources at various frequencies as well as their quantitative contribution to the overall noise emission.