Predicting energy absorption parameters of aluminum lattice structures filled tubes via artificial neural networks

Abstract

In this paper, the energy absorption parameters of aluminum body centered cubic lattice structures filled thin-walled tubes under axial loading are predicted via artificial neural networks (ANNs). Different tube thickness, lattice member diameters and number of lattice unit cells are considered as design variables, the total amount of energy absorption (EA), the specific energy absorption (SEA), the mean crush force (MCF) and the peak crush force (PCF) are considered as design criteria (energy absorption parameters). The proposed approach is based on finite element simulations for construction of the sample design space and verification, ANNs for predicting the energy absorption parameters. The results showed that the proposed ANN approach is able to predict the energy absorption parameters with high accuracy.