Energy absorption optimization of parametric plate lattices

Abstract

This study introduces an optimization method for generating novel plate lattices with enhanced energy absorption characteristics, focusing on achieving high specific energy absorption and low peak crushing forces. While previous research has primarily analyzed the energy absorption behavior of conventional plate structures, we employ a multilayer perceptron (MLP) model to map the relationship between energy absorption indicators and shape parameters of parametric plate lattices. The optimization process utilizes the nondominated sorting genetic algorithm II (NSGA-II) to compute the Pareto optimal set, balancing high energy absorption with low peak crushing force to identify the optimal lattice design. The study also explores the impact of shape parameters on energy absorption properties. Results demonstrate that the proposed framework successfully generates plate lattices with superior energy absorption performance, with the PPL_B lattice significantly outperforming the SC-BCC-FCC plate lattice and D-type Triply Periodic Minimal Surface (TPMS) lattice in both simulations and physical experiments. These findings provide a valuable guideline for selecting ideal energy-absorbing lattices in practical applications.