Crashworthiness optimisation of a vehicle crash box using a novel sinusoidal double curvature design

This study investigates enhancing the crashworthiness of vehicle crash boxes through experimental testing, numerical simulation, and geometry optimisation. Two identical commercial crash boxes were extracted from a small vehicle and tested under quasi-static loading at a strain rate of 150 mm/min, with an average energy absorption of 3.39 kJ. A model of the crash box was developed and validated using finite element analysis in Abaqus 2024, incorporating the Johnson–Cook material model to simulate the plastic deformation of DC01 steel. The numerical results closely matched experimental findings, with an average deviation of 7.13%. To enhance crash performance, a novel origami-based design featuring sinusoidal double curvature geometry was introduced and optimised using a genetic algorithm. The optimisation targeted increased energy absorption and reduced initial peak force, evaluating the crush force efficiency. Generation 15 stands out as the top performer in our optimisation process, achieving the highest fitness value and demonstrating the greatest energy absorption of 7.03 kJ, which revealed a 99.20% improvement. Additionally, analysis of crash force efficiency confirms that the 14th generation greatly enhanced crashworthiness by up to 92.07%, which demonstrates significantly greater energy absorption per unit of peak force. These findings confirm that the optimised design significantly improves crashworthiness and is a promising solution for advanced energy-absorbing systems.