摘要
在直齿面齿轮齿面仿真的基础上,利用Matlab与ANSYS相结合实现了面齿轮副在有限元软件中从参数化建模到承载接触分析的自动化。为解决面齿轮强度设计的应力计算问题,研究了点接触面齿轮载荷与弯曲应力的确切关系,并提出了一种采用人工神经网络预测面齿轮弯曲应力的方法,该方法以有限元分析结果为样本,能够完成面齿轮弯曲应力的快速定量计算,在训练神经网络的样本参数范围内,该方法具有很高的精度,为面齿轮的强度设计奠定了基础。
Based on tooth surface simulation of spur face gear, the automation from parametric modeling to load tooth contact analysis of face gear pair in finite element software is achieved by using Matlab and ANSYS. To solve the problem of stress calculation while design face gear, the exact relationship between load and bending stress of face gear is researched, and a method to predict the bending stress by artificial neural network is presented which regard the re- sult of finite element analysis as samples, the method can rapidly calculate the bending stress with high accuracy while parameters are within sample scope. The foundation for strength design of face gear is laid.
出处
《机械传动》
CSCD
北大核心
2013年第3期1-4,共4页
Journal of Mechanical Transmission
基金
国家高技术研究发展计划项目(863计划2009AA04Z404)
关键词
面齿轮
弯曲应力
接触有限元
BP神经网络
Face gear Bending stress Contact finite element method BP neural network
