Abstract
Computational inverse techniques are presented for inverse problems in the areas of material characterization using dynamic response in anisotropic inhomogeneous plates. In these techniques, the inverse problems are formulated into parameter identification problems, in which a set of parameters corresponding to the characteristics of material property can be found by minimizing error functions formulated using the measured displacement responses and that computed using forward solvers based on projected candidates of parameters. The forward solvers used in this work are the hybrid numerical method (HNM) and the finite element method (FEM). The HNM has been proven very effective for transient wave response analysis. The high efficiency of these two forward solvers paves the way for the inverse procedure to solve the inverse problems using waves. Three types of optimization algorithms: nonlinear least square method, evolutional method (genetic algorithm), and their combined method, as well as an identification technique: artificial intelligence method (neural network), are employed as the inverse procedure. Several application examples are presented. It is demonstrated in this work that these inverse problems in material characterization can be solved to a desired accuracy with high efficiency through innovative use of advanced computational techniques.
