Identification of elastic parameters of an inclusion by a particle filter using ultrasonic waves

Abstract

Ultrasonic waves have been used as a nondestructive testing tool to evaluate defects in steel mill products. Although an ultrasonic wave has the capability to detect defects, the type classification of defects and material characterization of components continue to be challenging tasks. The aim of this paperis to develop an identification method to determine the elastic properties of an inclusion. A particle filter (PF) is applied to identify the wave velocity and density of an inclusion. The PF is a data assimilation technique that is based on a Bayesian approach, and requires a reliable simulation tool to describe the system and measurements. The estimated value by PF showed good convergence to the true value of the wave velocity anddensity of the inclusion. By using a finite integration technique as the simulation tool, the PF can estimate the wavevelocity and density of the inclusion in steel accurately and successfully.