Open Access
Subscription or Fee Access
SETRIC—Sparsity Enforcing Time Reversal Imaging Concept for Ultrasonic Guided Waves
Abstract
Ultrasonic guided waves are used for the detection, localization and severity estimation of damage in technical structures. In the past few years multiple damage localization techniques were developed in the framework of structural health monitoring (SHM) that use Lamb waves to inspect thin structures. In this work a novel imaging concept for damage localization is proposed and compared with existing image reconstruction methods based on simulated signals. The so-called Sparsity Enforcing Time Reversal Imaging Concept (SETRIC) represents a combination of time reversal (TR) and model-based compressed sensing (CS). In TR the received signals, containing the scattering response from damage, is time reversed and send back into the structure. The energy accumulates at the position of the damage because of the reciprocity of the wave equation. On the other hand, CS models the ultrasound scattering for each pixel, stores the signals in a dictionary and then approximates the measured signal by solving a linear equation based on a sparsity assumption, e.g. orthogonal matching pursuit (OMP). In SETRIC, we use TR to extract the dominant features from the signals of the damaged structure, obtained from the right singular vectors of the transfer matrix, and then use CS to generate the image.
DOI
10.12783/shm2019/32177
10.12783/shm2019/32177