Damage detection in thick walled composites using surface mounted piezoelectric elements

Vibration signatures can be used as an efficient and quick way to detect, locate and quantify damage on large structures. While there have been considerable efforts by the modal analysis community to detect damage and evaluate its severity, little attention has been paid to finding a reliable way to excite the structure. The goal of the research presented in this paper was to investigate the feasibility of using piezoelectric actuators and sensors in a damage detection system for thick walled composite plates. It was demonstrated that a single piezoelectric actuator and sensor could sufficiently excite a thick walled composite to detect damage. The piezo sensor/actuator arrays were sized and positioned based on a finite element model. Frequency response functions between the actuators and sensors were acquired by exciting the actuators with a random signal and using appropriate signal conditioning for the sensors. Using frequency response functions for the healthy structure as a baseline transfer function, subsequent tests of this structure and/or other similar structures were compared to the baseline data using several damage detection algorithms. The algorithms include the Transfer Function, Group Delay and Transmissibility methods. These algorithms were used to compute metrics to distinguish between the healthy and the damaged structures. An analytical study to determine the effect of parameter variability on the damage metrics was conducted and will also be presented. It is shown that damage can be detected successfully by Group Delay and Transmissibility algorithms.