Crushing analysis and multi-objective crashworthiness optimization of GFRP honeycomb-filled energy absorption devices

• We improve the crashworthiness of a standard vehicle energy absorption device.
• The specimen is filled with a glass-fiber reinforced polymer honeycomb structure.
• We use surrogate-based multi-objective optimization techniques.
• Three objective functions are used as metrics: mass, absorbed energy and peak load.
• Final results better the specific energy absorbed by 40% and reduce the peak load by 37%.

Fuel efficiency and occupant safety are two of the most important concerns in the automotive industry nowadays. Encouraged by the importance of this field of study, this research attempts an improvement in the crashworthiness of a vehicle crash absorber. This component consists in a square hollow steel tube filled with a honeycomb structure made of glass-fiber reinforced polyamide. Surrogate-based optimization techniques are used. The three objective functions chosen — mass, absorbed energy and peak load — are approximated by two different models: multivariate adaptive regression splines and Gaussian process (kriging). The thickness of both parts, the shape of the honeycomb and its height are selected as design variables. Two preliminary analyses of the specimen are performed: the computation of the interaction effect and a comparison of a hollow tube with the specimen. From the results of multi-objective crashworthiness optimization two Pareto frontiers are obtained, one for the absorbed energy and mass, and another one for the absorbed energy and peak load. The results achieved show great improvements on all objective functions compared to the original design. The peak load is reduced by 37% on a specimen with similar mass and absorbed energy, and the specific energy absorbed is increased by 39.5% for a specimen with a similar peak load to the one from the initial model.