Subsurface damage identification and localization in PZT ceramics using point contact excitation and detection: An image processing framework

Abstract

Piezoelectric sensors hold immense potential in structural health monitoring (SHM) applications. However, their performance can be deteriorated by defects and extreme weathering. Therefore, diagnosing the sensor before implementation is very crucial. Unreliable experimental methods and inaccurate damage detection algorithms are major concerns that need addressing to develop a robust damage detection framework. In this work, we propose a subsurface anomaly detection framework that uses the evolution of ultrasonic waves in spatial and temporal domains. This framework comprises three key components: a novel Coulomb coupling-based experimental approach to visualize ultrasonic wave interactions with microscale Lead Zirconate Titanate (PZT) subsurface defects, an advanced denoising algorithm using block matching 3D (BM3D) filtering to reduce noise, and a multiresolution dynamic mode decomposition (mrDMD) algorithm to identify subsurface defects in PZT. The results conclude that the proposed framework is robust, efficient, and can provide reliable detection and localization of damage even with significant measurement noise and without any reference damage-free counterpart of the PZT material.