Lamb wave measurement method for in-plane elastic constants of balanced symmetric laminates based on machine learning

Abstract

Abstract To achieve in-situ measurement of mechanical properties in large-scale composite structures, this paper introduces a Lamb wave measurement based on machine learning to predict the in-plane engineering elastic constants of balanced symmetric laminates. Firstly, we consider that balanced symmetric laminates are equivalent to orthotropic single-layer plates with nine engineering elastic constants. Secondly, by varying these elastic constants and comparing the dispersion curves at different propagation angles, we conclude that, under low frequency-thickness products, the phase velocity of S0-mode Lamb waves in orthotropic single-layer plates is dependent on four engineering elastic constants: tensile modulus, in-plane shear modulus, and in-plane Poisson’s ratio. Subsequently, leveraging this correlation in dispersion curves, a BP neural network model is established using machine learning techniques. Using the neural network model, the goal of predicting engineering elastic constants using phase velocities is achieved. Finally, the effectiveness of this method is verified through theoretical calculations and numerical simulations.