No abstract

No abstract

In the field of automotive safety, the lightweight design of crash absorbers is an important research topic with a direct effect on the occupant safety levels. The design of these absorbers usually requires an optimization of their crashworthiness, which can include multi-objective and reliability-based optimization techniques. This process is very time-consuming, and in spite of the continuous growing of computational power, the problem needs a reliable solving scheme. The use of surrogate models and parallel computing are suitable alternatives to deal with this issue. However, the strongly non-linear response functions obtained from the finite element simulations need careful treatment. This work contributes with the application of a surrogate-based reliability-based design optimization method to an original design of a crash absorber made of metal and a glass-fiber reinforced polymer which is subjected to a frontal impact. Multi-adaptive regression splines models are employed to emulate the original responses, and three different approaches in the sampling stage of M. Cid Montoya the method are compared. The absorbed energy and the mass of the element are considered as objective functions, while the peak value of the force transmitted to the occupants of the vehicle is the design constraint. A discussion of the employed materials is presented and the proposed approaches are compared. Finally, several Pareto fronts are obtained as a solution to the probabilistic problem. Results show that a combination of aluminum and glass fiber reinforced polymer is optimum for this problem, and some design rules are offered.

No abstract

No abstract

No abstract

No abstract

No abstract