Damage Detection in Laminated Composites by Neural Networks and High Order Finite Elements

Abstract

In the aerospace industry, machine learning techniques are becoming more and more important for Structural Health Monitoring (SHM). In fact, they could be useful in giving a precise and complete mapping of damage distribution in a structure, including low-intensities or local defects, which cannot be detected via traditional tests. In this work, feedforward artificial neural networks (ANN) are employed for vibration-based damage detection in composite laminates. In the framework of Carrera Unified formulation (CUF), one-dimensional refined models in conjunction with layer-wise (LW) theory are adopted. CUF-based Monte Carlo simulations have been used for the creation of a dataset of damage scenarios for the training of the ANN. Therefore, the latter is fed with the vibrational characteristics of these structures. The trained ANN, given these dynamic parameters, is able to predict location and intensity of all damages in the laminated composite structures.