Free Vibration-Based Delamination Detection in Fiber Metal Laminates Composite Beam

Abstract

Delamination is a typical flaw in fiber-metal laminated composite structures that, when hidden from view, can significantly lower structural stiffness and alter a structure's ability to respond dynamically to forces like natural frequencies. Delamination is undoubtedly an important topic as it causes the performance of the Fiber Metal Laminate structures in the service to worsen. Delamination detection and severity analysis are essential in the aerospace industry for both safety and cost reasons. Although natural frequencies do not directly reveal the location or extent of damage, they are the most dependable metrics for damage detection. Frequency shifts in various modes are used to solve the inverse problem to pinpoint the damage location and extent. The bending natural frequencies shift brought on by delamination is utilized as an input to predict the delamination parameters. This study used an approach based on machine learning and a regression model to find the delamination parameters, i.e., locations and severity, in fiber metal laminated cantilever beams. FEA simulations using ANSYS were used to get the dataset pertaining to the position, severity, and eigen values (or bending natural frequencies) of the delamination. The results of this study indicate that the delamination locations and severity predictions developed using machine learning and regression models are reasonably accurate and demonstrate good agreement with the observed data.