在轿车侧面碰撞事故中,儿童乘员头、胸部是主要损伤部位。为最大限度地降低儿童乘员头、胸部的损伤风险,对儿童约束系统和侧面气囊的安全性能进行优化设计。采用多体动力学软件MADYMO建立儿童乘员侧面碰撞仿真模型,并通过台车试验结果验证仿真模型的有效性。以头、胸部加速度为优化目标,运用试验设计方法和方差分析进行相关设计变量的灵敏度分析,并确定主要的优化设计变量为侧面气囊的泄气率常数、儿童安全座椅的上固定点位置和下固定织带的刚度系数,进而构建优化目标的二阶多元回归代理模型,利用NSGA-Ⅱ多目标优化算法对构建的代理模型进行优化求解,获得包含头、胸部加速度最优解集的Pareto前沿曲线。通过权衡分析选取使头、胸部的损伤风险同时降低的最优解,最后确定侧面气囊的泄气率常数0.037,儿童座椅的上固定点x向位置-0.25 m,座椅下固定织带的刚度系数0.5为儿童约束系统主要设计参数的最优配置。在最优设计配置条件下得到的儿童头部合成加速度峰值比初始值下降了约24%,胸部合成加速度峰值比初始值下降约30%。
The head and thorax of the child occupant are mainly injured regions in passenger car side impacts.In order to reduce the injury risks of head and thorax,a multi-objective optimization design of child restraint system(CRS) and side airbag is conducted in terms of CRS safety performances.A mathematical model of a child occupant in side collision is developed by using MADYMO software and the validity of the model is evaluated by comparing results between simulation and sled test.For the purpose of minimizing accelerations of child head and thorax,a sensitivity analysis of the design variables is conducted using the orthogonal design of experiment(DOE) and analysis of variance.The primary design variables are therefore determined,including material permeability of side airbag,location of top tether in X-direction and stiffness coefficient of lower anchorage.The objective functions for accelerations of child head and thorax are generated with second-order multinomial regression equations in terms of the determined design variables.Optimization is carried out byusing non-dominated sorting genetic algorithm II(NSGA-II).The optimized configuration of the CRS is obtained by further analysis based on the Pareto-optimal solutions.The results indicate that main design parameters of child restraint system are material permeability 0.037 of side airbag,location ?0.25 m of top tether in X-direction,stiffness coefficient 0.5 of lower anchorage.Under the optimized condition,the peak value of head acceleration is minimized by about 24 percent.The peak value of thorax acceleration is declined by about 30 percent.