基于CAE与遗传算法的汽车B柱外饰板轻量化设计
陈毅超1，王辉123
1.武汉理工大学汽车工程学院，武汉 430070； 2.武汉理工大学现代汽车零部件技术湖北省重点实验室，武汉 430070；
3.湖北隆中实验室，湖北襄阳 441000
Lightweight Design of Automotive B-pillar Exterior Panel based on CAE and Genetic Algorithm
Chen Yichao1 , Wang Hui123
1.School of Automotive and Engineering,Wuhan University of Technology,Wuhan 430070,China; 2.Hubei Key Laboratory of Advanced Technologyfor Automotive Components,Wuhan University of Technology,Wuhan 430070,China; 3.Hubei Longzhong Laboratory,Xiangyang 441000,China
摘要：以某品牌汽车 B 柱外饰板为研究对象，基于 CAE 分析技术对 B 柱外饰板结构进行轻量化设计。 首先利用ANSYS 软件对原始模型进行仿真 ，在此基础上将 B 柱外饰板受力工况等效为简支梁模型并计算得到壁厚范围为2.67～2.72 mm，结合实际生产特点最终确定饰板件主要厚度为 2.7 mm，随后在 CATIA 软件中建立轻量化设计后的模型并再次进行有限元仿真 ，得到饰板在设定的受力工况下的形变量为 1.485 1 mm，符合轻量化设计的刚度要求 。为了进一步验证设计的合理性，从实际生产出发 ，将翘曲变形量作为成型质量评价指标，运用正交试验法设计了六因素五水平的工艺组合方案，并采用 Moldflow 软件进行模流仿真，接着基于 Matlab 平台建立 BP 神经网络模型 ，将其作为适应度函数 ，通过遗传算法(NSGA-Ⅱ)进行全局寻优，得到最优工艺参数组合下轻量化设计前后两模型的最大翘曲变形量分别为 2.568 mm 和 2.353 mm，制件质量分别为 104.6 g 和 94.44 g。结果表明，该轻量化设计减重达 9.71% 并使翘曲变形量减小了 8.37%，较好地实现了轻量化的目标，有效提高了制件的成型质量。
Abstract：The B-pillar exterior panel of a certain brand automobile was chosen as the research subject for conducting a lightweight design of its structure using CAE analysis technology. Firstly, ANSYS software was used to simulate the original model, based on which the force condition of the B-pillar exterior panel was equivalently transformed into a simple supported beam model and the range of the wall thickness was calculated to be 2.67-2.72 mm. Taking into account the characteristics of actual production, the final determination for the main thickness of the panel was 2.7 mm. Afterwards, the lightweight design model was created in CATIA software, followed by another round of finite element simulation. The obtained deformation value of the panel under the prescribed loading condition was 1.485 1 mm, which satisfied the stiffness requirements of the lightweight design. To further validate the rationality of the design, starting from actual production, the warping deformation was taken as the evaluation indicator for forming quality. A six-factor and five-level process combination scheme was designed using the orthogonal test method, and Moldflow software was used to carry out the mold flow simulation. Then a BP neural network model was constructed based on the Matlab platform as the fitness function, and global optimization was carried out by the genetic algorithm (NSGAⅡ). Under the optimum combination of process parameters, the maximum warpage of the two models before and after the lightweight design is 2.568 mm and 2.353 mm, and the weight of the part is 104.6 g and 94.44 g. The results indicate that the lightweight design achieves a weight reduction of 9.71% and results in an 8.37% reduction in warping, effectively realizing the goal of lightweight design and significantly enhancing the molding quality of the parts.
关键词：轻量化；CAE分析；正交试验；BP神经网络；遗传算法
Keywords：lightweight; CAE analysis; orthogonal test; BP neural network; genetic algorithm
基金：国家自然科学基金项目 (51775398，52175360)，湖北省重点研发计划项目 (2022BAA073)，中国科协青年人才托举工程项目(2021QNRC001)，中国博士后科学基金项目 (2022T150278，2022M710062)，武汉理工大学襄阳技术转移中心科技产业化资金项目(WXCJ-20220005)，襄阳高新区科技计划项目(202203)，湖北省科技人才服务企业项目(2023DJC055)
本文引用格式：
陈毅超，王辉.基于CAE与遗传算法的汽车B柱外饰板轻量化设计[J].工程塑料应用，2023，51(10)：85-91.
Chen Yichao，Wang Hui. Lightweight design of automotive B-pillar exterior panel based on CAE and genetic algorithm[J]. EngineeringPlastics Application，2023，51(10)：85-91.
原文：
相似文章：
