Inverse source strength reconstruction techniques for ducted acoustic sources

Abstract

The aeronautic, aerospace and automotive industries require improved noise reduction techniques to meet the current normative and increase the acoustic comfort at acceptable cost. The aim of this work is to investigate the capabilities of inverse methods for the noise source reconstruction of ducted acoustic sources in a laboratory acoustic facility installed in a semi anechoic chamber. Model-based inverse methods for nacelle noise identification build on the sound analysis of the in-duct pressure field. A model of suitably driven sources is assumed to reconstruct the pressure field measured at discrete points over a virtual surface. Useful results can be obtained with the use of appropriate regularization techniques. However, the performance results are found to be very dependent on the assumed reconstruction model. In this work, we have proposed an alternative methodology for the inverse source strength reconstruction that does not require a priori information on the original source distribution. It is based on an angular decomposition of the source strengths as Fourier series in combination with the minimisation of a mean square error criterion. The performance and limitations of this method is analysed in terms of physical parameter configurations and compared with the equivalent source method.