Guided wave SHM system for damage detection in complex composite structure

• Published in: Theoretical and applied fracture mechanics 2020-02, Vol.105, p.102408, Article 102408
• Main Authors: De Luca, A., Perfetto, D., De Fenza, A., Petrone, G., Caputo, F.
• Format: Article
• Language: English
• Subjects: Accuracy; Actuators; Composite; Composite structures; Computer simulation; Configurations; Damage detection; Damage index; Damage modelling; FE analysis; Fiber composites; Fiber reinforced polymers; Finite element method; Glass fiber reinforced plastics; Guided waves; Impact damage; Mathematical analysis; Mathematical models; Position sensing; Probability Ellipse method; Reliability analysis; SHM; Wave propagation; Winglets
• ISSN: 0167-8442; 1872-7638
• DOI: 10.1016/j.tafmec.2019.102408

•Paper deals with the simulation of guided wave propagation in a winglet.•Signals recorded on pristine and damaged winglets have been compared.•Damage causes variations of signals that can be quantified by damage indexes (DIs).•High DIs mean that damage is placed close the corresponding actuator-sensor path.•Probability Ellipse (PE) method has been used for damage detection purpose. The paper deals with the development of a Finite Element (FE) model aimed to simulate the propagation mechanisms of guided waves in a complex composite structure, such as a blended Glass Fibre Reinforced Polymers (GFRP) winglet, for real-time damage detection and monitoring purposes. The reliability of the herein presented FE model has been assessed against an experimental tests campaign, involving the winglet under both reference/pristine configuration and impacted/damaged configuration, also changing the actuator location. Because of the good level of accuracy shown by the numerical-experimental comparisons, the FE model has been used also for numerically investigating new actuators and damage position configurations, not experimentally analysed. Damage Indexes (DI’s) have been calculated to quantify the variations of the signals’ amplitude caused by the induced damage. Moreover, the Probability Ellipse (PE) method, which estimates the probability of the presence of the damage in the monitored area starting from the knowledge of selected DI’s for each sensors-path, has been used in both older and newer damage position. The accuracy of the PE method in detecting the damage location is herein discussed.