Alternative Excitation and Data Analysis Techniques for Damage Detection in Metallic Plates

Abstract

Strategies for reliable damage detection of thin-walled structural components (e.g. metallic shell structures or stiffened composite panels) are required in various industrial branches, e.g. aerospace, shipbuilding, tower construction, etc. In the last years, structural health monitoring (SHM) methods and systems, based on guided ultrasonic waves and especially Lamb waves, have been successfully investigated. On the other hand, vibration-based SHM approaches showed their damage detection capabilities in many application domains, including for instance the surveillance of wind turbine structures [1] and civil infrastructure. Given by the stochastic nature of the ambient excitation of such structures, a variety of statistical signal processing techniques have been developed, including for example Null Space based Fault Detection (NSFD) method [2] or methods based on the changes of measured signals energy (e.g. Root Mean Square (RMS) analysis). The main goal of the present paper is to study statistical signal processing techniques, originally developed in the context of vibration-based SHM, for guided wave-based damage detection. Therefore, an experiment has been developed consisting of a metallic plate and two surface bonded piezoelectric transducers. Deterministic and also stochastic excitation was used to identify damage at different spatial locations. Experimental findings will be presented and discussed in order to derive the respective strengths and weaknesses.