基于疲劳寿命的驱动桥壳可靠性与轻量化设计

为提高驱动桥壳的轻量化水平和道路行驶疲劳可靠性，对驱动桥壳进行6-Sigma稳健性多目标轻量化设计。首先，建立驱动桥壳的虚拟台架仿真模型，并进行垂直弯曲刚性和垂直弯曲静强度的仿真分析，将仿真得到的桥壳本体各测点变形量和关键受力点应力值与试验结果进行对比，以验证桥壳虚拟台架仿真模型的可信性。其次，建立驱动桥壳的最大垂向力仿真模型，结合耐久性强化路面下驱动桥壳板簧座处的垂向载荷谱，基于名义应力法，对驱动桥壳进行了道路行驶工况下的疲劳寿命分析。然后，选取驱动桥壳本体各截面壁厚为设计变量，基于熵权法和TOPSIS（Technique for Ordering Preferences by Similarity to Ideal Solution，TOPSIS）方法研究各壁厚变量对桥壳综合性能的影响。结合RBF（Radial Basis Function，RBF）近似模型和NSGA-Ⅱ算法（Elitist Non-dominated Sorting Genetic Algorithm，NSGA-Ⅱ）对驱动桥壳进行基于疲劳寿命的多目标确定性轻量化设计，获取Pareto最优解集，选取桥壳的优化方案。最后，基于蒙特卡罗模拟抽样方法和微存档遗传算法（AMGA）对驱动桥壳进行了多目标6-Sigma稳健性轻量化设计，得到桥壳稳健性优化方案。研究结果表明：稳健性优化后，驱动桥壳本体的疲劳寿命降低了12.3%，但和初始结构的疲劳寿命相比，仍提升了117%；桥壳本体疲劳寿命正态分布的标准方差下降了72.1%，说明桥壳本体的疲劳可靠性得到了大幅提升；桥壳本体的质量升高了1.8%，但和优化前的桥壳原结构相比，仍实现减重5.9%。

Reliable and Lightweight Design for Drive Axle Housing Based on Fatigue Life

To improve the lightweight level and fatigue reliability of drive axle housing, the 6-Sigma robust and deterministic multi-objective lightweight design was established. Firstly, a simulation model for the drive axle housing bench test was developed, and the vertical bending stiffness and vertical bending static strength were virtually analyzed. The simulated deformation of the measuring points and the stress value of the key stress points of the drive axle housing's main body were compared with the experimental results; thus, verifying the reliability of the virtual bench test's simulation model. Secondly, a maximum vertical force simulation model for the drive axle housing was developed, followed by its combination with the vertical load spectrum on the spring seats of the drive axle housing under durability-enhanced road conditions. Further, the fatigue life analysis of the drive axle housing under road driving condition was conducted based on the nominal stress method. Subsequently, the wall thicknesses of four sections of the drive axle housing's main body were selected as the design variables. Further, the influences of the design variables on the comprehensive performance of the drive axle housing were studied based on the entropy weight method combined with the Technique for Order of Preference by Similarity to Ideal Solution (TOPSIS) method. Finally, the radial basis function (RBF) approximation model and the elitist non-dominated sorting genetic algorithm (NSGA-Ⅱ) were adopted to perform the multi-objective lightweight design of the drive axle housing based on fatigue life. The Pareto-optimal solution set was obtained and one of the optimal solutions was identified as the optimization program for the drive axle housing. Finally, based on the Monte Carlo simulation sampling method and an archive-based micro genetic algorithm (AMGA), the 6-Sigma robust multi-objective lightweight design of the drive axle housing was developed to obtain a robust optimal solution. The comparison of robust and deterministic multi-objective optimization results indicate that after robust optimization, although the fatigue life of the drive axle housing's main body reduces by 12.3%, it still increases by 117% compared to the fatigue life of the initial structure; the standard variance of the normal distribution of the fatigue life of the drive axle housing's main body decreases by 72.1%, indicating that the fatigue reliability of the drive axle housing's main body greatly improves; the weight of the drive axle housing's main body increases by 1.8%, but still remains 5.9% lighter than the initial structure of the drive axle housing prior to optimization.