• CN:11-2187/TH
  • ISSN:0577-6686

机械工程学报 ›› 2021, Vol. 57 ›› Issue (18): 49-57.doi: 10.3901/JME.2021.18.049

• 特邀专栏:高速铁路钢轨检测监测与应用 • 上一篇    下一篇

扫码分享

基于磁巴克豪森噪声的钢轨白层无损检测研究

蒋生辉1, 丁昊昊1, 张翔2, 彭建平2, 王文健1, 刘启跃1, 郭俊1   

  1. 1. 西南交通大学牵引动力国家重点实验室 成都 610031;
    2. 西南交通大学物理科学与技术学院 成都 610031
  • 收稿日期:2020-12-30 修回日期:2021-06-11 出版日期:2021-09-20 发布日期:2021-11-30
  • 通讯作者: 王文健(通信作者),男,1980年出生,博士,研究员,博士研究生导师。主要研究方向为轨道交通轮轨关系与服役行为。E-mail:wwj527@163.com
  • 作者简介:蒋生辉,男,1995年出生。主要研究方向为基于巴克豪森效应的无损检测技术。E-mail:shenghui_jiang@163.com
  • 基金资助:
    国家重点研发计划政府间国际科技创新合作重点专项(2018YFE0109400)和国家自然科学基金(51775455,51975489)资助项目。

Research on Non-destructive Testing of Rail White Etching Layer Based on Magnetic Barkhausen Noise

JIANG Shenghui1, DING Haohao1, ZHANG Xiang2, PENG Jianping2, WANG Wenjian1, LIU Qiyue1, GUO Jun1   

  1. 1. State Key Laboratory of Traction Power, Southwest Jiaotong University, Chengdu 610031;
    2. School of Physical Science and Technology, Southwest Jiaotong University, Chengdu 610031
  • Received:2020-12-30 Revised:2021-06-11 Online:2021-09-20 Published:2021-11-30

摘要: 设计搭建了基于磁巴克豪森噪声(Magnetic Barkhausen noise,MBN)的钢轨白层无损检测系统,打磨U75V钢轨以预制白层,使用1 Hz,8 V的正弦信号磁化打磨钢轨试样并收集表面产生的MBN信号,研究了钢轨白层几何特征、微观组织与MBN信号特征值的关系,并分析了白层对MBN信号的影响机理。结果表明:白层作为磁硬性组织,会降低钢轨表面MBN信号强度。试样无明显白层时,打磨产生的残余拉应力使得MBN信号的峰峰值(VPP)和方均根值(VRMS)较高,当试样出现明显白层后,信号峰峰值和方均根值迅速减弱。白层中的马氏体晶粒细小,大量晶界的存在对磁畴有很强的钉扎作用,在有效检测区域内,当白层厚度增加,次表层可发生偏转的磁畴数量会减少,进而削弱传递到表面的MBN信号强度。MBN信号特征值与白层厚度(δWEL)、白层面积(SWEL)都呈现较好的线性负相关变化关系,其中方均根值与白层面积的线性拟合结果最佳,皮尔森系数为-0.896 9,线性拟合度达0.794 1。MBN信号特征值对钢轨白层的厚度、面积等几何特征具有定量分析能力,试验结果对服役钢轨的维护作业具有一定的指导意义。

关键词: 磁巴克豪森噪声, 钢轨, 白层, 微观组织

Abstract: A nondestructive inspection system for white etching layer (WEL) on rail is designed and built based on the Magnetic Barkhausen noise (MBN). The U75V rail is ground to prefabricate the WEL. The ground rail specimen is magnetized by sinusoidal signal with 1 Hz, 8 V, and the MBN signal generated on the surface is collected. The relationships between the geometric characteristics, microstructure of WEL on rail and the characteristic values of MBN signal are studied. The influence mechanism of WEL on MBN signal is analyzed. The results show that the WEL will reduce the MBN signal intensity on the rail surface because it is difficult to be magnetized. When no obvious WEL is produced on the specimen, the residual tensile stress caused by grinding makes the peak to peak value (VPP) and root mean square value (VRMS) of MBN signal higher, The VPP and VRMS of the signal decrease obviously when the WEL appears on the ground rail specimen. The martensite grains in the WEL are fine, and the existence of a large number of grain boundaries has a strong pinning effect on the magnetic domains. In the effective detection area, when the thickness of the WEL increases, the number of magnetic domains those can deflect in the subsurface will decrease, thus weakening the MBN signal strength transmitted from the subsurface to the surface. VPP and VRMS of MBN signal both show the linearly negative correlation with the WEL thickness (dWEL) and the WEL area (SWEL). The VRMS and WEL area has the best linear fitting result. Pearson coefficient is -0.896 9, and the linear fitting degree is 0.794 1. The characteristic value of MBN signal can quantitatively analyze the thickness and area of the WEL on rail. The experimental results have a certain guiding significance for the maintenance of the service rail.

Key words: magnetic Barkhausen noise, rail, white etching layer, microstructure

中图分类号: