考虑尺寸、拓扑与材料变量的客车骨架多响应优化设计

The structural design of bus skeleton is the core and sustentive technology, which can implement the transformation from large bus country to strong one. Beam element has the advantages of parameterization, being close to manufacture process and high efficiency. Therefore, the bus skeleton consisting of beam elements is adopted to solve the problems of low accuracy of joint stress and multi-variable and multi-response optimization of bus skeleton in this project, respectively..Firstly, the stress concentration coefficients of the joints are tested by stress gages to archive the response surface model of stress concentration coefficients with respect to cross-sectional sizes of rectangular tube. Moreover, the nominal stress is updated by the stress concentration coefficients to obtain the accurate and modified beam element..Then, the size, topology and material design variables are together introduced to create the optimization model, which can minimize the mass of bus skeleton, constrained with static bending and torsional stiffness, dynamic frequency stiffness, yielding stress and material costs. Ordered SIMP interpolation is used to transfer the combinatorial optimization problem into a continuous optimization one. In order to reduce the complexity and computational costs of this optimization model, the original optimization problem is divided into two subproblems by separating the local and global constraints. The sensitivities of responses with respect to design variables are derived. Sequential linear programming and optimal criteria method are adopted to solve this two subproblems, respectively.. Lastly, the CAE optimization platform of bus skeleton is developed to provide a comprehensive solution.

客车车身结构设计是实现我国由客车大国向客车强国转型的核心支撑技术。由于梁单元易于参数化、贴近制造工艺、求解效率高，所以本项目以梁单元构成的客车骨架为研究对象，解决接头处应力求解精度低、多变量多响应客车骨架优化难题。.首先，由应变花对接头处的应力集中系数进行实验测试，建立应力集中系数与矩形管断面尺寸变量的响应面预测模型，用其更新名义应力，得到修正的高精度梁单元。.其次，以尺寸、拓扑、材料为设计变量，整车质量为目标函数，弯扭静刚度、频率动刚度、屈服应力、材料成本为约束，建立客车结构优化模型。采用序列SIMP插值将拓扑和材料组合优化问题转化为连续变量优化问题。为了降低模型的复杂度与计算量，将全局与局部约束分离，原始优化问题解耦为2个优化子问题。推导响应关于设计变量的灵敏度，分别采用序列线性规划和最优准则法求解这2个子问题。.最后，自主开发CAE优化平台，为客车车身设计提供一整套解决方案。

我国正处在由客车大国向客车强国转型的关键阶段，客车车身骨架设计作为核心支撑技术之一，是确保客车向安全、节能与环保方向发展的关键因素。本项目研究了考虑尺寸、拓扑、材料变量的客车骨架多响应优化设计。.为了解决客车骨架接头应力集中问题，以矩形管梁单元截面尺寸高度、宽度、厚度、外载荷为试验因素，设计大型全因子试验表，测试客车骨架常见3种接头相交处的应力集中系数。并采用2阶多项式响应面建立应力集中系数关于截面尺寸的显式表达式，用其更新铁木辛柯梁的名义应力，得到修正的高精度梁单元。为下一步将应力约束加入客车骨架的优化模型中奠定基础。研究了序列SIMP（Ordered Solid Isotropic Material with Penalization）法，包括序列SIMP函数不同的惩罚方式与曲线类型对插值有效性的影响以及惩罚因子的取值范围等。在序列SIMP法基础上，提出了一种从多材料拓扑优化到截面尺寸优化的两步客车骨架优化方法。采用梁单元搭建客车骨架有限元模型，采用序列SIMP法实现多材料拓扑优化，建立了多种材料的弹性模量与密度的插值关系，用密度变量代替材料变量与拓扑变量，将材料优化纳入到拓扑优化的体系，从而可以使用铝合金、镁合金和高强钢等多种材料，理性、高效地设计客车骨架拓扑结构。客车骨架多材料拓扑优化之后，对最优拓扑结构构件的截面尺寸进一步优化。两步优化都考虑了弯扭静刚度与模态频率约束，相对于单一材料拓扑和尺寸优化，该两步优化法可以获得更轻量化、性能更好的客车骨架。采用序列线性规划法求解以上两步优化的数学模型，由于序列线性规划是梯度型算法，需要提供灵敏度信息，所以推导了弯扭静刚度、模态频率等客车骨架响应关于截面尺寸以及密度变量的解析灵敏度公式。.根据上述研究内容，开发了客车骨架优化软件CarFrame，实现“轻量化、高刚度、多材料”的客车骨架设计。软件使用权成果转让1项，在一汽集团技术中心取得了成功应用。发表SCI论文16篇、EI论文7篇，授权发明专利5项，毕业硕士生10名，在读博士生2名，吉林省自然科学二等奖1项。