In situ characterization technique to increase robustness of imaging approaches in structural health monitoring using guided waves

Abstract

The performance of guided wave imaging strategies used in Structural Health Monitoring relies on the accurate knowledge of mechanical properties for proper damage detection and localization. In order to increase the performance and robustness of such algorithms, it is desirable to implement autonomous approaches that can characterize the mechanical properties of the structure whatsoever the environmental and operational conditions. This article presents an innovative in situ and integrated characterization procedure based on guided waves that evaluates the thermo-mechanical properties of a structure when subjected to thermal variations prior to imaging using the same set of piezoceramic transducers used for imaging. These properties are then exploited in the damage imaging using a correlation-based algorithm (Excitelet) combined with the optimal baseline subtraction. The characterization strategy uses a genetic algorithm to identify the optimal set of mechanical properties leading to the best correlation between an analytical formulation of dispersed guided waves propagation and experimental measurements. The strategy is assessed experimentally on an aluminum plate with three sparse bonded piezoceramic transducers used for both characterization and imaging at various temperatures, representative of operational conditions of an aircraft. An artificial damage is subsequently introduced in the plate, and the effect of the accuracy of the mechanical properties estimation on imaging is assessed through the detection capability, positioning, accuracy, and correlation amplitude. The approach is then compared to three imaging methods, namely, baseline-free imaging, imaging without considering thermo-mechanical effects, and imaging using stretching methods traditionally used to compensate for temperature effects.