摘要
以典型的轻质高强蛛网藻硅质壳结构为仿生模板,设计一种仿生汽车轮毂结构,并建立有限元参数模型进行仿真分析,发现其较普通辐条式轮毂结构的弯曲与疲劳性能更好,且具有进一步优化的空间。采用响应面法进行优化设计,获得仿生轮毂最优结构方案,优化后的整体结构较优化前减重8.4%,其中轮辐减重25.6%。进一步利用BP神经网络提高结构最大等效应力及疲劳寿命的预测精度。计算结果表明,BP神经网络对结构最大等效应力的预测精度与Workbench响应面法的预测精度相当,但对结构疲劳寿命预测精度则提高了56.97%。
A lightweight wheel inspired by the structure of the topical lightweight high-strength arachnoidiscus shell is designed.The finite element parameter model is established for the simulation analysis.It is found that the bionic wheel has batter bending and fatigue performance than the conventional spoke wheel,and further optimization space.Sampling points are obtained using Latin hypercube sampling method.In addition,the design aims to not only lower its mass,but also maintain the remarkable mechanical properties.The results indicates that the mass of the optimized wheel is dramatically decrease by 8.4%and that of spoke is by 25.6%.The BP method is further used to improve the prediction accuracy of the maximum equivalent stress and fatigue life of the structure.The result showes that BP method after training has the high accuracy,and it is the same as the prediction accuracy of the maximum equivalent stress of the BP neural network was comparable to that of the Workbench response surface method,and the predication accuracy of the structural the fatigue life is improved by 56.97%.
作者
陈依婷
郭策
李龙海
马耀鹏
马玉秋
CHEN Yiting;GUO Ce;LI Longhai;MA Yaopeng;MA Yuqiu(College of Astronautics,Nanjing University of Aeronautics and Astronautics,Nanjing 210016,China;College of Mechanical and Electrical Engineering,Nanjing University of Aeronautics and Astronautics,Nanjing 210016,China)
出处
《机械制造与自动化》
2020年第6期90-94,100,共6页
Machine Building & Automation
基金
国家自然科学基金资助项目(51875282)
国防基础科研计划资助项目(JCKY2018605C010)。
关键词
仿生轻质结构
轮毂
响应面方法
BP神经网络
bio-inspired structure
wheel
response surface method
BP neural network
