Damage Detection in Composite Plates with Ultrasonic Guided-waves and Nonlinear System Identification

• Published in: 2020 IEEE Symposium Series on Computational Intelligence (SSCI) pp. 2039 - 2046
• Main Authors: de Castro Ribeiro, Mateus Gheorghe, Kubrusly, Alan Conci, Ayala, Helon Vicente Hultmann
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01.12.2020
• Subjects: Acoustics; artificial neural networks; CFRP; Fault detection; Feature extraction; guided-waves; Monitoring; Recurrent neural networks; Resistance; structural health monitoring; System identification
• DOI: 10.1109/SSCI47803.2020.9308212

Carbon Fiber Reinforced Polymer Composites (CFRP) are materials widely used in many applications due to its interesting properties, such as high strength and stiffness, low weight and high fatigue resistance. The usage of this material demands a reliable damage detection strategy since flaws can cause economic losses and life-threatening situations. In this context, this paper proposes a structural health monitoring strategy to detect flaws in CFRP plates using ultrasonic guided waves signals, which are processed for feature extraction based on linear and nonlinear auto-regressive system identification techniques and artificial neural networks. Baseline and baseline-free signals are compared and results have shown that a perfect detection was able to be achieved in the former case. This shows that the present contribution is suitable to be used for structural health monitoring applications of CFRP plates subject to damage and encourages further research work to improve the baseline-free case.