Prediction of the behaviour of CFRPs against high-velocity impact of solids employing an artificial neural network methodology

Abstract

A new methodology based on artificial neural networks has been developed to study the high velocity oblique impact of spheres into CFRP laminates. One multilayer perceptron (MLP) is employed to predict the occurrence of perforation of the laminate and a second MLP predicts the residual velocity, the obliquity of trajectory of the sphere after perforation and the damage extension in the laminate. In order to train and test the networks, multiple impact cases have been generated by finite-element numerical simulation covering different impact angles and impact velocities of the sphere for a given system sphere/laminate.