首页|PTFE唇形密封结构优化方法研究

PTFE唇形密封结构优化方法研究

扫码查看
基于参数化建模和神经网络技术提出一种PTFE双唇形油封结构的优化方法.以密封唇唇厚及唇尖坐标作为设计变量,建立双唇形油封的参数化模型,通过拉丁超立方抽样与有限元方法开展试验设计,引入神经网络拟合代理模型近似代替目标函数,综合分析各设计变量对接触压力、径向力及等效应力的影响;以接触压力最大,等效应力和径向力最小为优化目标,采用优化算法在设计空间内寻优,建立双唇形密封结构优化设计的数学模型.结果表明:优化后的唇封结构最大接触压力增大了 6.3%,且接触压力分布更符合良好密封的要求;密封唇最大等效应力和最大径向力均有明显减小,提高了油封的可靠性.唇形油封性能的改善证实了优化方法的有效性.
Research on Optimization Method of PTFE Lip Seal Structure
Based on parametric modeling and neural network technology,an optimization method for double-lip oil seal structure was proposed.With the thickness of the sealing lip and the coordinates of the lip tip as design variables,the para-metric model of the double lip oil seal was established.The optimal Latin hypercube sampling method was used to sample the design space and the finite element analysis was combined to carry out the design of experiment.The neural network was introduced to fit the response surface model and approximatively replace the objection function,the influence of each design parameter on contact pressure,radial force and equivalent stress was comprehensively analyzed.Taking the maximum contact pressure,the minimum equivalent stress and redial force as objectives,and using the optimization algorithm to find the optimal solution in the design space,a mathematical model was established for the optimization design of the double-lip seal.The results show that the contact pressure of the optimized oil seal is increased by 6.3%and the contact pressure dis-tribution is more in line with the requirements of a good seal.Both the equivalent stress and radial force are significantly re-duced,so as to improve the reliability of the oil seal.The improvement in the performance of the lip seal proves the effec-tiveness of this optimization method.

lip sealparameter optimizationneural networksurrogate modelparametric modeling

杨芳、张海涛、冯涛、黄宏艳、王虎

展开 >

中国航发西安动力控制科技有限公司,陕西西安 710077

北京合工仿真技术有限公司,北京 100192

唇型密封 参数优化 神经网络 代理模型 参数化建模

2024

润滑与密封
中国机械工程学会 广州机械科学研究院有限公司

润滑与密封

CSTPCD北大核心
影响因子:0.478
ISSN:0254-0150
年,卷(期):2024.49(7)
  • 14