2015 conference paper

Estimating the mechanical properties of layered media using ultrasonics: Numerical study with high resolution wave propagation

Structural health monitoring 2015: system reliability for verification and implementation, vols. 1 and 2, 2559–2566.

By: R. Rashetnia* & M. Pour-Ghaz

Source: NC State University Libraries
Added: August 6, 2018

Estimating the mechanical properties of layered media using ultrasonics is challenging due to the ill-posed, nonconvex, and nonlinear nature of the problem. To address this challenge, we use Kurganov Tadmor (KT) central scheme for the forward model of wave propagation since it provides a robust model for layered media with sharp boundaries. KT also minimizes numerical challenges such as dispersion error propagation. In this study, we consider a one-dimensional stress wave propagation problem. To solve the inverse problem of estimating the mechanical properties, we utilize non-heuristic random search algorithm with surrogate model. The random search algorithm performs a global search within the feasible search area, and the surrogate model applies an artificial Neural Network and a genetic algorithm to produce optimal estimations. This method also decreases the computational cost. The results show that this method can be successfully used to estimate the mechanical properties of the layered one-dimensional media. doi: 10.12783/SHM2015/318