Damage localization in holed Carbon Fiber composite laminates using FBG sensors based on Back-Propagation Neural Network

Carbon fiber reinforced plastic (CFRP) has better properties. Among all kinds of carbon fiber structural elements, carbon fiber laminates are the most widely used because of their simple structure and easy mass production. However, due to process defects and improper transportation, there will always be some unidentified damage to the carbon fiber composite laminate, and it will continue to expand after loading, which greatly weakens the mechanical properties of the carbon fiber composite laminate. Therefore, it is of very importance to locate the damages of CFRP laminated plates. Firstly, mechanical simulation analysis is made for the holed CFRP laminated plates to gain its strain distribution; the model for the mapping relation between the strain distribution of surface and damage location is built based on BP neural network method; Fiber Bragg Grating (FBG) sensor is stuck on the surface of holed CFRP laminated plates. The surface strain of CFRP laminated plates is measured by the FBG sensor and substituted into the above-mentioned model to predict the damage location of CFRP laminated plates. The experience results indicate that there is a small error between the predicted damage location and the actual one and verify the effectiveness and accuracy of the damage localization method in CFRP laminated plates using FBG sensors based on a neural network.