Crashworthiness analysis and multi-objective optimization of multi-cell thin-walled crash boxes with different cell geometries

In this study, externally reinforced multi-cell thin-walled structures with square, pentagonal, hexagonal and heptagonal geometries are proposed. It was found that heptagonal thin-walled structures have a higher collision performance than other geometries. To improve the collision performance criteria of this thin-walled structure, optimizations were carried out using a multi-objective genetic algorithm (MOGA) with variables defined as four different wall thicknesses. In the initial design, the specific energy absorption (SEA) of the thin-walled structure was 21.26 kJ/kg; after optimization, this value increased by approximately 15.43% to 25.14 kJ/kg. In terms of peak crushing force (PCF), this value decreased by 19.02%, from 212.062 kN to 178.64 kN. When the optimum parameters obtained after optimization were verified using finite element analysis, the error rates in PCF and SEA values were 0.044% and 3.24%, respectively. The results demonstrate that MOGA can be effectively used in the design of thin-walled structures.