Source identification and acoustic modeling of enclosures from experimental data

This paper presents an experimental method to analyze the acoustical characteristics of enclosures with respect to the noise sources. The method is based on the application of a multiple input/two output frequency domain model. Characterestic signals from each of the candidate noise sources represent the inputs, while the sound pressure signals from two closely spaced microphones are the outputs. By using a finite difference approximation with microphone signals the specific acoustic impedance and acoustic particle acceleration are estimated at an array of positions throughout the cavity. The impedance is used to separate the sound pressure into its progressive and standing wave components. The standing wave component is further described in a normal modes fashion by application of structural modal analysis curve fitting routines to the acoustic particle accelerance frequency response functions measured throughout the cavity. The contribution that the various sources have on the progressive and standing wave components is evaluated through interpretation of the multiple input/two output model. The underlying theory of the method is presented and applied to several laboratory tests to demonstrate the capabilities and limitations of the analysis. The results illustrate the separation of the progressive and standing wave components and the relative contributions from each source.