Search

The aim of this paper is to present a novel statistical model based on multiple linear regression. Although the independent variables of most statistical methods for predicting sound absorption are the specific flow resistivity and material thickness, the model presented here considers alternative parameters: specific acoustic impedance, density and specific flow resistivity. First, some materials are tested in an impedance tube in order to obtain their acoustic characteristics. Second, macroscopic empirical models (The Delany-Bazley's and Allard and Champoux models) were used to predict the sound absorption coefficient of distinct multilayered systems. Third, the statistical model proposed herein is developed and its accuracy verified. [ABSTRACT FROM AUTHOR]