Damage detection in hybrid metal-composite plates using ultrasonic guided waves based on outliers estimate

Abstract

The present research focusses on the development of a robust data-driven damage diagnosis technique to detect different types of damages in a hybrid metal-composite (HMC) plate specimen resulting from manufacturing processes, loading conditions, and ambient environmental conditions. These defects over a course of time deteriorate the load-bearing capacity of the HMC’s and in turn, their reliability in terms of safe operation. In this work, ultrasonic guided waves (UGW) are used for non-destructive evaluation (NDE) of the HMC. The use of UGW for NDE offers advantages such as long-range inspection and sensitivity to small-sized surface and sub surface damages. The ultrasonic tests are simulated using a pitch-catch active sensing technique at a typical frequency-mode pair best suited to detect and classify damages in the HMCs. The damage-sensitive feature is extracted from the received UGW using Hilbert transform-based feature extraction method. The damage indicator is classified in the damage-sensitive feature space using the root mean square technique identified as outliers, which is further used to classify the detected damages. The achieved results manifest the ability of the proposed technique to be a part of the industrial structural integrity inspection process typically for HMCs in detecting and classifying embedded damages with high accuracy.