Optimization of an auxetic jounce bumper based on Gaussian process metamodel and series hybrid GA-SQP algorithm

Jounce bumpers are indispensable components in vehicle suspension systems. It can attenuate impact energy and improve the ride comfort performances during transient pulse impacts from roads. However, the traditional jounce bumpers cannot achieve ideal mechanical performance and are difficult for parametric optimization due to its various shapes which are hard for parameterization. In this paper, a cylinder auxetic structure was introduced and applied as the suspension jounce bumper. Auxetic jounce bumper can be primarily defined by several variables. With tunable structure parameters, a large region of mechanical performances can be realized. In order to evaluate and optimize the performance of auxetic jounce bumpers, they were assembled into vehicle models to conduct virtual ride comfort tests under bump road. The maximum vertical acceleration of vehicle was defined as the objective function of the optimization problem of auxetic jounce bumper. On the purpose of improving optimization efficiency, Gaussian process regression method was employed to establish the metamodel of auxetic jounce bumper based on the sample set created by orthogonal design of experiments method. A series hybrid GA-SQP (Genetic algorithm-Sequential quadratic programming) algorithm, provided with both global and fast local searching abilities, was used for the optimization of auxetic jounce bumper. For comparison, GA algorithm was also applied for auxetic jounce bumper optimization alone. It is proved that the proposed optimization process is accurate and can improve optimization efficiency. Then the optimized design through hybrid algorithm was compared with original design. The maximum vertical acceleration of vehicle when travelling through bump was reduced with optimized design, and the vehicle ride comfort performance was improved.