Chapter 15 - Case Studies of Multi-Method SHM with PWAS Transducers: Damage ID in Experimental Signals

Abstract

This chapter illustrates the SHM methodology through case studies. In the first case study, a simple situation is considered in which a specimen of simple geometry is exposed to increasing amounts of damage. The simple geometry of the specimen allows the development of closed-form analytical solutions that permit the prediction of the SHM signals to be recorded on a pristine specimen. These predictions are compared with pristine specimen experimental data in order to validate the SHM measurement method using the E/M impedance technique. Subsequently, the validated SHM method is used to record data on specimens with increasing amounts of damage. Various damage-detection methods are applied to the recorded data and compared: (a) damage metrics; (b) statistical methods; (c) probabilistic neural nets. The second case study deals with a specimen of complicated geometry. This aircraft-like specimen is a built-up panel with construction features similar to those found in actual aging aircraft. Two specimens are considered, one pristine and the other damaged. Signal prediction analysis is not practical for this specimen because of its complicated geometry; hence, the study is performed entirely experimentally. Two SHM measurement methods are discussed and compared on these specimens: (a) the pulse-echo method; (b) the E/M impedance technique. The two techniques are used to detect the same damage, which is a crack emanating from a rivet hole.