Reliable optimisation design of vehicle structure crashworthiness under multiple impact cases

This paper presents the crashworthiness design optimisation of a typical vehicle structure subjected to offset deformable barrier and full rigid barrier impact cases. Several approaches involving sampling techniques, surrogate model, multi-objective optimisation algorithm and reliability analysis are introduced and applied. Four correlation functions of Kriging are adopted for the design optimisation. It shows that the accuracy of the correlation functions is different and the best one is selected for the objective or constraint function. Both the deterministic optimisation and reliability-based design optimisation (RBDO) are conducted. In RBDO, the reliability levels of constraints are computed through the first-order reliability method (FORM) and second-order reliability method (SORM), and are compared with Monte Carlo simulation. It is found that SORM is more accurate than FORM when the constraint functions are highly nonlinear. Finally, the results demonstrate that RBDO solutions are more reliable than that of the deterministic optimisation for real engineering application. RBDO solutions are further validated by finite element analysis model, which shows the effectiveness of the proposed RRDO approach in obtaining the reliable optimal design of the vehicle structures.