Hybrid Inversion technique for predicting geometrical parameters of Porous Materials

In prediction of acoustical behavior of porous materials, five geometrical parameters play a very important role, but some of these geometrical properties are very difficult to measure directly. So many authors have suggested different inversion strategies for getting these properties from directly measured acoustical properties of the material using standing wave tube. These approaches can be divided in two different categories: analytical (based on the limit behavior of the bulk properties) and minimization based methods (which make use of searching algorithms to determine the best solution that minimizes a cost function calculated by means of a prediction model). Recent studies have shown the reliability of the analytical methods for the determination of the airflow resistivity and the minimization based approach by using genetic algorithms for getting the other physical parameters. Consequently, a hybrid inversion technique can be proposed for the complete calculation of the geometrical quantities and here it is presented in detail. Moreover the paper compares the results from the hybrid approach with the experimentally measured parameters and the values of the five parameters obtained by using genetic algorithms. Finally, the paper presents the effect of both inversion techniques on acoustical properties using Johnson-Allard-Champoux model.