Modal analysis using fiber optic sensors and neural networks for prediction of composite beam delamination

Extrinsic Fabry-Perot interferometric (EFPI) fiber optic sensors and a neural network provided a health-monitoring capability for laminated glass/epoxy composite beams. The EFPI sensors experimentally determined the first five modal frequencies of the cantilever beams. The feedforward backpropagation neural network used these modal frequencies to predict the size and location of delaminations in the composite beams. Beam modal frequencies shift as a function of delamination size and location. Five beams with prescribed delaminations, as well as a `healthy' beam with no delaminations, were excited by a surface-mounted piezoelectric actuator at frequencies up to 1 kHz. All beams had an eight-ply symmetric glass/epoxy composite design, were fabricated simultaneously, and had length and width dimensions of 26.04 and 2.33 cm, respectively. The beams with flaws had different delamination sizes ranging from 1.27-6.35 cm long prescribed in the mid-plane, i.e. between the fourth and fifth plies. The neural network was trained using classical-beam theory and tested using the experimental EFPI data. The delamination size and location predictions resulting from the neural network simulation had an average error of 5.9 and 4.7%, respectively. Also, analytical classical-beam theory, finite element methods, and ceramic piezoelectric sensors validated the EFPI modal frequency measurements.