航空学报 > 2023, Vol. 44 Issue (6): 427203-427203   doi: 10.7527/S1000-6893.2022.27203

低压涡轮铸造叶片几何不确定性统计

罗佳奇1(), 陈泽帅1, 邹正平2,3, 曾飞4, 杜鹏程2   

  1. 1.浙江大学 航空航天学院,杭州 310027
    2.北京航空航天大学 航空发动机研究院,北京 102206
    3.北京航空航天大学 航空发动机气动热力国防科技重点实验室,北京 102206
    4.中国航发湖南动力机械研究所,株洲 412002
  • 收稿日期:2022-03-28 修回日期:2022-04-21 接受日期:2022-05-30 出版日期:2022-06-09 发布日期:2022-06-08
  • 通讯作者: 罗佳奇 E-mail:jiaqil@zju.edu.cn
  • 基金资助:
    国家科技重大专项(J2019-II-0012-0032);国家自然科学基金(51976183)

Statistics on geometric uncertainties of casting blades in low-pressure turbines

Jiaqi LUO1(), Zeshuai CHEN1, Zhengping ZOU2,3, Fei ZENG4, Pengcheng DU2   

  1. 1.School of Aeronautics and Astronautics,Zhejiang University,Hangzhou 310027,China
    2.Research Institute of Aero-Engine,Beihang University,Beijing 102206,China
    3.National Key Laboratory of Science and Technology on Aero-Engine and Aero-Thermodynamics,Beihang University,Beijing 102206,China
    4.AECC Hunan Aviation Powerplant Research Institute,Zhuzhou 412002,China
  • Received:2022-03-28 Revised:2022-04-21 Accepted:2022-05-30 Online:2022-06-09 Published:2022-06-08
  • Contact: Jiaqi LUO E-mail:jiaqil@zju.edu.cn
  • Supported by:
    National Science and Technology Major Project(J2019-II-0012-0032);National Natural Science Foundation of China(51976183)

摘要:

精铸涡轮叶片的几何偏差诱因较多,几何偏差的统计对认识偏差来源及统计建模、评估叶片几何精度等均具有重要意义。首先初步分析千套低压涡轮真实叶片几何偏差的基本特征,通过主元分析提取偏差模态并识别主要偏差来源,发现该涡轮叶片存在明显的偏移、扭转和叶型误差。之后介绍一种基于优化策略的高效高精度叶片几何偏差分解方法,分离出偏移误差、扭转误差和叶型误差,统计发现总体几何偏差的概率密度函数(PDFs)接近高斯分布。最后对叶型误差进行统计分析并发现相对于真实叶片,偏差分解后叶型轮廓度误差的统计均值和标准差均明显下降,叶片合格率明显上升;此外,通过叶表特殊位置的轮廓度统计发现叶型轮廓度的概率密度函数也近似满足高斯分布。

关键词: 几何变化, 不确定性, 低压涡轮, 偏差分解, 叶型误差, 概率统计

Abstract:

It has been well known that there are different sources for geometric deviations of casting turbine blades. Statistical study of geometric deviations benefits not only the discovery of deviation sources and statistical modelling, but also the evaluation of blade geometric accuracy. This paper gives a statistical study of on geometric uncertainties of casting blades in low-pressure turbines. First, the geometric deviations of one thousand casting blades in the low-pressure turbine are preliminary studied. By principal component analysis, the basic modes of geometric deviations are extracted, and the main sources are distinguished, including movement error, twist error and profile error. Then, an error decomposition method based on the optimization strategy is introduced, which can be used to extract different kinds of errors. Statistical study finds that the Probability Density Functions (PDFs) of global geometric deviations are close to Gauss distributions. Finally, blade profile deviations are statistically analyzed. The results demonstrate that in comparison with the original casting blades, the geometric dispersion degree and the profile tolerance error of the corrected blades without global geometric deviations are significantly decreased, and the percentage of qualified blades is significantly increased. Besides, statistical study on profile tolerance at some specified positions on the blade demonstrates that the probability density functions of profile tolerance are also close to Gauss distributions.

Key words: geometric variation, uncertainty, low-pressure turbine, deviation decomposition, blade profile error, probability statistics

中图分类号: