Research on optimization design method of CFRP/metal composite thin-walled square tube structure based on surrogate model

Abstract

CFRP/metal composite thin-walled structures show great potential for automotive energy absorbers. However, existing studies rarely address the co-optimization of key design variables—metal wall thickness, CFRP layup angle, and CFRP layup thickness—and conventional multi-objective algorithms are prone to local optimality when balancing specific energy absorption (SEA) and peak crushing force (PCF). To fill this gap, this paper focuses on CFRP/metal composite square tubes, selecting metal wall thickness (A), CFRP layup angle (B), and CFRP layup thickness (C) as design variables. A surrogate model is constructed via numerical simulation and experimental validation, and the improved multi-objective artificial hummingbird algorithm (MOAHA) is introduced to optimize SEA and PCF. Results show the improved algorithm achieves winning rates of 76.7%, 86.7%, and 80% on HV, IGD, and Spacing metrics, outperforming the original algorithm in 76.7–86.7% of test scenarios. The weighted distance method selects the optimal solution from the Pareto set, verified by Abaqus simulation and prototype experiments with relative errors all below 5%. This study illustrates the synergistic mechanism of the three variables on the SEA-PCF trade-off, mitigating their conflict, and provides a reference for engineering design of such energy absorbers and practical application of multi-objective optimization algorithms.