Characterization of impervious layers using scale models and an inverse method

Abstract

We describe a novel procedure that uses an inverse method to determine unknown parameters for impervious layers used in multilayer structures. The proposed model of the multilayer structure is limited to an ideal double plate separated by an unbonded, fibrous, sound-absorbing material. Experimental data were obtained by nearfield acoustic holography for the calculation of the transmission loss of various multilayer structures mounted in a window in a wooden box designed specifically for this purpose. We used the Trochidis and Kalaroutis forecast model of acoustic insulation for multilayer structures, which is based on a spatial Fourier transform. The experimental pressure and velocity data were used as input data in the inverse method. By applying the Trochidis and Kalaroutis model and using numerical methods to adjust the variables that define the impervious layers of the system, the values of the unknown parameters of the layers could then be calculated. For validation, the results were compared with results obtained using the Ookura and Saito model, based on impedance coupling between layers and using the statistical-energy-analysis model, which subdivides the system into subsystems. We evaluate the measurement errors associated with the construction of a hologram by nearfield acoustic holography, i.e., errors due to sensor mismatch and position mismatch, in terms of their probabilities.