航空学报 > 2023, Vol. 44 Issue (4): 426805-426805   doi: 10.7527/S1000-6893.2021.26805

基于GA⁃SVR的薄壁叶片辅助支撑布局优化方法

郑志阳1,2, 张阳1,2, 张钊1,2, 吴宝海1,2(), 张莹1,2   

  1. 1.西北工业大学 航空发动机高性能制造工信部重点实验室,西安 710072
    2.西北工业大学 航空发动机先进制造技术教育部工程研究中心,西安 710072
  • 收稿日期:2021-12-10 修回日期:2021-12-14 接受日期:2021-12-21 出版日期:2022-01-12 发布日期:2022-01-11
  • 通讯作者: 吴宝海 E-mail:wubaohai@nwpu.edu.cn
  • 基金资助:
    国家重点研发计划(2020YFB1710400);陕西省自然科学基础研究计划面上项目(2021JM-054);国家自然科学基金(52005413);陕西省自然科学基础研究计划(2020JQ-183)

Layout optimization of auxiliary support for thin-walled blade based on GA-SVR

Zhiyang ZHENG1,2, Yang ZHANG1,2, Zhao ZHANG1,2, Baohai WU1,2(), Ying ZHANG1,2   

  1. 1.Key Laboratory of High Performance Manufacturing for Aero Engine,Ministry of Industry and Information Technology,Northwestern Polytechnical University,Xi’an 710072,China
    2.Engineering Research Center of Advanced Manufacturing Technology for Aero Engine,Ministry of Education,Northwestern Polytechnical University,Xi’an 710072,China
  • Received:2021-12-10 Revised:2021-12-14 Accepted:2021-12-21 Online:2022-01-12 Published:2022-01-11
  • Contact: Baohai WU E-mail:wubaohai@nwpu.edu.cn
  • Supported by:
    National Key R&D Program of China(2020YFB1710400);Natural Science Basic Research Program of Shaanxi(2021JM-054);National Natural Science Foundation of China(52005413);Natural Science Basic Research Program of Shaanxi(2020JQ-183)

摘要:

作为一种典型的复杂曲面薄壁件,叶片在铣削加工过程中极易发生“让刀”,即弹性变形,严重影响着叶片的加工精度。针对该问题,设计了四自由度回转辅助支撑机构增加叶片的加工刚度,并提出了一种基于GA-SVR的薄壁叶片辅助支撑布局优化方法。首先,建立了综合考虑材料去除以及铣削力与弹性变形耦合效应的叶片铣削加工有限元仿真模型。其次,以辅助支撑布局作为设计变量,最大加工弹性变形和整体弹性变形均方差作为布局优劣评价指标,采用拉丁超立方试验设计和有限元仿真模型计算评价指标并生成样本集,再以支持向量机回归(SVR)对样本集进行训练获得评价指标的代理预测模型。然后,采用精英策略遗传算法(GA)优化薄壁叶片的辅助支撑布局。最后,进行了薄壁叶片铣削加工及三坐标测量实验,结果表明优化后的辅助支撑布局方案对叶片加工弹性变形的抑制程度可达57.6%。

关键词: 薄壁叶片, 弹性变形, 变形预测与控制, SVR代理模型, 支撑布局优化

Abstract:

As a typical thin-walled workpiece with complex curved surface, the blade is prone to deflection, that is, elastic deformation, which seriously affects the machining accuracy of the blade. As to this problem, a four degree-of-freedom rotary auxiliary support mechanism is designed to increase the machining rigidity of the blade, and a layout optimization method of auxiliary support for thin-wall blade based on GA is proposed. Firstly, a finite element simulation model for blade milling is established considering material removal and the coupling effect between milling force and elastic deformation. Secondly, the layout scheme of auxiliary support is taken as design variable, the largest value and standard deviation of overall machining elastic deformation are taken as the quality evaluation index of the layout. The sample set is constructed by Latin hypercube design and finite element simulation model. The surrogate forecast model is trained with Support Vector machine Regression (SVR). Then, elite strategy Genetic Algorithm (GA) is used to optimize the auxiliary support layout of thin-walled blade. Finally, the milling and CMM experiments of thin-walled blade are carried out, and the results show that the optimal auxiliary support layout scheme can suppress the blade elastic deformation by 57.6%.

Key words: thin-walled blade, elastic deformation, deformation prediction and control, SVR surrogate model, support layout optimization

中图分类号: