Online debonding detection in honeycomb sandwich structures using multi-frequency guided waves

Due to the complex nature of sandwich structures, development of the online structural health monitoring system to detect damages in honeycomb sandwich panels inherently imposes many challenges. In this study, the leaky guided wave propagation in the honeycomb sandwich structures generated by piezoelectric wafer actuators/sensors is first simulated numerically based on the finite element method (FEM). In the numerical model, the real geometry of the honeycomb core is considered. To accurately detect debonding in the honeycomb sandwich structures, signal processing based on continuous wavelet transform is adopted to filter out the unwanted noise in the leaky Lamb wave signals collected from the experimental testing. A correlation analysis between the benchmark signals at the normal condition and those recorded at the debonded condition is then performed to determine the differential features due to the presence of debonding. Finally, the image of the debonding is formed by using a probability analysis. Specifically, fusing images acquired from multi-frequency leaky Lamb waves are obtained to enhance the quality of the final image of the structure. The location and size of the debonding in the honeycomb sandwich structures are estimated quantitatively.