基于图像处理的疲劳裂纹扩展长度在线测量方法  

Online Measuring Method of Fatigue Crack Growing Length Based on Image Processing Technology

作　　者：高红俐[1] 郑欢斌 姜伟[1] 齐子诚[1,2] 

机构地区：[1]浙江工业大学特种装备制造与先进加工技术教育部/浙江省重点实验室,杭州310014 [2]中国兵器科学研究院宁波分院,宁波315103

出　　处：《中国机械工程》2016年第7期917-924,共8页China Mechanical Engineering

摘　　要：为了在金属疲劳裂纹扩展试验中采用直观、准确、可靠的方法获得高精度裂纹长度尺寸,提出了基于图像处理技术的疲劳裂纹扩展长度在线测量方法。首先,采用大尺寸黑白全帧面阵CCD和微距显微镜头对疲劳裂纹扩展过程中的图像进行采集,分析疲劳裂纹的图像形态特征,采用亚像素边缘定位方法对预制裂纹边缘进行检测,求出疲劳裂纹起点位置。然后,对裂纹扩展区域进行图像子区划分,对图像子区采用基于灰度统计数据的自适应分割方法进行裂纹分割,采用单区域增长算法进行二值化裂纹图像连接得到裂纹主干。最后,进行了系统标定和疲劳裂纹长度在线测量试验。试验结果表明:本方法测量精度高、抗干扰能力强、数据稳定可靠,裂纹长度测量精度为0.03mm,满足裂纹扩展试验国际标准的要求。Based on image processing technology,a fatigue crack growing length online measuring method was proposed herein to obtain high precision crack length data by applying direct,reliable,accurate method in the fatigue crack propagation test.Firstly,a series of digital images in the process of fatigue crack propagation was collected by high-speed switching industrial camera that consists of a large scale area array CCD and a macro micro lens,and the morphological characteristics of fatigue crack image were analyzed.Then,sub-pixel edge detection method was used to detect the prefabricated crack edge in order to obtain the fatigue start position.Secondly,the crack region was divided into multiple square sub-images from fatigue crack start point to right specimen edge.The sub-images were segmented to binary images by adaptive image segmentation method based on grayscale statistics.The crack skeleton could be obtained by connecting the binary crack images applying the single-domain growing algorithm.Finally,system calibration and online measuring experiments of crack lengths were done.The experimental results indicate that the proposed method is of non contact,high precision,with strong anti-interference ability,and experimental data is stable and reliable.The measuring precision of crack length reaches 0.03 mm,which can satisfy the international standard of fatigue crack propagation test.

关 键 词：图像采集与处理 疲劳裂纹扩展试验 裂纹扩展长度 灰度方差 在线测量 

分 类 号：TP394.1[自动化与计算机技术—计算机应用技术] TH691.9[自动化与计算机技术—计算机科学与技术]