Sound Source Localization on a Fast Rotating Fan Using Rotational Beamforming

Delay and Sum Beamforming with spaced microphone arrays is a common and well-known method for sound source localization. Nevertheless sound sources on fast rotating fans, rotors, blades and wheels are difficult to be identified due to the high rotational speed. Furthermore optical cameras with very high speed shutters must be used to have a chance to optically resolve the acoustic image. In air flow systems fans have a mayor contribution to the overall sound emission. Therefore it is of utmost interest to understand their acoustical behavior as a starting point for sound design and noise minimization. This paper introduces a rotational beamforming method using a virtual array rotation corresponding to the rotation speed of the source. It works as a filter on the temporal data and therefore allows for the usage of all known beamforming algorithms in the time and frequency domain. This makes the method highly flexible and powerful. Its usage is shown and evaluated on a fan test stand with two counter-rotating fans. Air flow impedances can be changed by using different end pieces, tubes and blades. The objective is to maximize the efficiency while minimizing the noise emission. Measurements with a multi-channel ring array have been performed on the test stand. They show that the method is capable of resolving various sound sources on the fan at different frequencies and at different locations on the blades like leading and tailing edge.