Characterization of the elastic moduli in composite plates via dispersive guided waves data and genetic algorithms

In this study, an inverse procedure based on stress guided waves is proposed for the characterization of the elastic moduli of composite plates. The characterization is carried out via genetic algorithms by minimizing the discrepancy between experimental and numerical group delay curves for different directions of propagation along the plate. Experimentally, for a given distance source-receiver, the group delay curves are obtained by processing the guided waves time-transient signals via a time-frequency transform. For the same distance, a fast and reliable semi-analytical finite element formulation is used for the forward computation of the group delay curves. Here, pseudo-experimental data, generated by means of the semi-analytical finite element model for an assumed known set of elastic coefficients, are used to test the reliability of the proposed procedure. The results obtained for three different plates are promising. Since semi-analytical finite element formulations can also handle plates with uniform curvature, this identification procedure could be extended to composite shells. The authors believe that this procedure could support the on-field nondestructive evaluation and structural health monitoring of composite plates and shells.