An Evaluation Method of Cohesive Quality in Bonded Plates Using Wavelet Transform and Neural Networks

A method based on a combination of wavelet transform (WT) and artificial neural networks (ANN) is presented to evaluate the cohesive quality in bonded plates using laser‐generated guided waves in two‐layer composite plates. The transient waveforms obtained by numerical simulations are taken as the sample database of ANN for training and learning, and the WT is used to extract the eigenvectors from the guided wave signals to simplify the structure of the ANN. The researches show that it is available to classify three kinds of interfaces, as the rigid, the weak and the slip interfaces of the adhesive layer. As to the weak interface situation, a more detailed quantitative work proves the effectiveness of the inversion of stiffness coefficients from the transient guided waves. This method provides a new promising way for the characterization of the cohesive quality in bonded plates by laser‐generated guided wave detection.