基于电涡流热成像的钢轨滚动接触疲劳裂纹动态检测研究  

Research on Dynamic Detection of Rail Rolling Contact Fatigue Crack Based on Eddy Current Thermography

作　　者：苗玲 高斌[1] 石永生 李浩然 黎潇枫 吴同乐 张喜源 田贵云[1,3] MIAO Ling;GAO Bin;SHI Yongsheng;LI Haoran;LI Xiaofeng;WU Tongle;ZHANG Xiyuan;TIAN Guiyun(School of Automation Engineering,University of Electronic Science and Technology,Chengdu 611731;Railway Infrastructure Inspection Center,CHINA RAILWAY,Beijing 100081;School of Electrical and Electronic Engineering,Newcastle University,Newcastle NE17RU,UK)

机构地区：[1]电子科技大学自动化工程学院,成都611731 [2]中国国家铁路集团有限公司铁路基础设施检测中心,北京100081 [3]纽卡斯尔大学电气与电子工程学院,英国纽卡斯尔NE17RU

出　　处：《机械工程学报》2021年第18期86-97,共12页Journal of Mechanical Engineering

基　　金：国家自然科学基金(61971093,61527803,61960206010);四川省科技厅(2019YJ0208,2018JY0655,2018GZ0047);中央高校基本科研业务费专项资金(ZYGX2019J067)资助项目。

摘　　要：针对钢轨滚动接触疲劳裂纹检测挑战,提出了新型传感结构下的多物理电涡流热成像技术。通过COMSOL Multiphysics有限元仿真软件,分析了圆柱形磁芯结构、弧形磁轭结构和U形磁轭结构的电磁场特性,并研究了在不同传感架构下对疲劳裂纹的检测效果。针对实际应用条件,在U形磁轭结构的基础上研制了L形磁轭和梭形磁轭结构并集成开发了感应激励源系统。静止状态下铣平钢轨试块缺陷检测试验表明自主研制的系统对钢轨表面微小缺陷的检测灵敏度高,能实现0.5mm宽缺陷的定量分析,且最小能识别钢轨表面0.2 mm宽的缺陷。通过圆盘缺陷检测试验对速度效应下疲劳裂纹的检测进行了研究,提出的鲁棒低秩张量裂纹分解算法极大地消除了原始热图像的背景噪声并提升缺陷检测分辨率。最后搭建并集成了钢轨低速巡检试验平台,验证了对钢轨踏面复杂滚动接触疲劳裂纹检测的有效性。In response to the challenge of rail rolling contact fatigue crack detection, a multi-physics eddy current thermography technology under new sensor structures is proposed. Through COMSOL Multiphysics finite element simulation software, the electromagnetic field characteristics of the cylindrical magnetic core structure, the arc-shaped yoke structure and the U-shaped yoke structure are analysed, and the detection effect of fatigue crack under different sensing architectures is studied. According to actual application conditions, the L-shaped yoke and the shuttle-shaped yoke structures are developed based on the U-shaped yoke structure,and the induction excitation source system is integrated and developed. Defect detection experiment of milled rail test blocks in the static state shows that the self-developed system has high detection sensitivity for small defects on the rail surface. It can realize quantitative analysis of 0.5 mm wide defects and identify defects with a minimum width of 0.2 mm on the rail surface. Through the defect detection experiment of disc, the fatigue crack detection under speed effect is studied. The proposed robust low-rank tensor crack decomposition algorithm greatly eliminates the background noise of the original thermal image and improves the defect detection resolution. Finally, the low-speed rail inspection experimental platform is built and integrated to verify effectiveness of complex rolling contact fatigue crack detection on rail treads.

关 键 词：滚动接触疲劳裂纹 电涡流热成像 传感架构 定量检测 巡检 

分 类 号：TG142.15[一般工业技术—材料科学与工程] U216.3[金属学及工艺—金属材料]