AN EFFICIENT PROCEDURE FOR VISUALIZING THE SOUND FIELD RADIATED BY VEHICLES DURING STANDARDIZED PASSBY TESTS

Abstract

Throughout the world, motor vehicles are required to conform to statutory passby noise requirements. These standards usually entail the measurement of the peak A-weighted sound level while the vehicle accelerates past a stationary microphone. This pass-or-fail test is of little diagnostic value when a vehicle exceeds the statutory level. In the work presented here it is shown that a stationary array of microphones may be used to visualize the location of the noise sources on a vehicle executing a standard passby test. By locating the positions of the sources that dominate the received sideline levels, noise control efforts may be directed appropriately and efficiently. In the present work, a delay-and-sum beamformer approach was used to detect noise source locations. Algorithms have been included to correct the apparent source frequency shifts that result from the source motion. The restoration and de-Dopplerization procedure is based on using a polynomial representation of the source velocity (inferred in the first instance from a radar signal). Forward and backward propagation procedures are compared in terms of computational expense. In addition, a spherical spreading correction factor is described, along with a maximum likelihood procedure for obtaining an optimal array weighting dependent on the relative distance between the microphones and the focus point. The new weighting factor is compared with a more “intuitive” weighting factor and it will be shown that the new approach can reduce the sidelobe levels by comparison with the more intuitive approach. Experimental results presented here were based on the use of a 16 element, sparse array. It will be shown that this array could be used to identify the location and directivity of tire noise sources and of other sources on accelerating vehicles.