中国民航大学学报

• 民用航空 • 上一篇    下一篇

基于SVR 的高压涡轮一级转叶裂纹发展规律研究

曹惠玲,王晓兵   

  1. (中国民航大学航空工程学院,天津300300)
  • 收稿日期:2018-01-08 修回日期:2018-03-05 出版日期:2018-08-25 发布日期:2018-10-25
  • 作者简介:曹惠玲(1962—),女,河北唐山人,教授,博士,研究方向为航空发动机性能分析与故障诊断.
  • 基金资助:
    中央高校基本科研业务费专项(3122014D010)

Crack development law of HPT rotor first stage blades based on SVR

CAO Huiling, WANG Xiaobing   

  1. (College of Aeronautical Engineering, CAUC, Tianjin 300300, China)
  • Received:2018-01-08 Revised:2018-03-05 Online:2018-08-25 Published:2018-10-25

摘要: 针对航空发动机高压涡轮一级转子叶片10~12 孔区域裂纹损伤的潜在安全问题,首先利用ImageJ 软件获取裂纹损伤数据,然后建立基于SVR 的高压涡轮一级转子叶片裂纹损伤长度预测模型并检验模型的可靠性,最后利用模型对各相关因素进行深入分析。结果表明,裂纹发展长度随空气质量指数变化呈指数规律分布,随CSN、EGT 和N1 变化呈对数规律分布,且各相关因素的变化对裂纹损伤发展的贡献率有很大差异。

关键词: 预测, 支持向量回归机, 裂纹, QAR 数据, 高压涡轮一级转子叶片

Abstract: Aiming at the potential safety problems caused by crack damage in Holes No.10~12 of first-stage rotor blades of aero-engine high pressure turbine, experiments are conducted. Firstly, crack damage data are obtained through ImageJ. Then, prediction model of crack damage length is established based on SVR, reliability of the model is tested. Finally, in-depth analysis on related influencing factors is made using the model. Results show that the crack length varies exponentially with the change of AQI(air quality index), and it distributes in logarithmic order with the variation of CSN, EGT and N1. Moreover, the contribution rate of each related factor is quite different to the development of crack damage.

Key words: forecast, SVR, crack, QAR data, HPT rotor first stage blades

中图分类号: