• CN:11-2187/TH
  • ISSN:0577-6686

机械工程学报 ›› 2024, Vol. 60 ›› Issue (18): 24-31.doi: 10.3901/JME.2024.18.024

• 仪器科学与技术 • 上一篇    下一篇

扫码分享

基于图像对比的疲劳裂纹扩展动态检测方法

孙士勇1, 朱殿立1, 李佳琦2, 樊俊铃3, 杨睿1, 赵延广4   

  1. 1. 大连理工大学机械工程学院 大连 116024;
    2. 大连理工大学白俄罗斯国立大学联合学院 大连 116024;
    3. 中国飞机强度研究所 西安 710065;
    4. 大连理工大学工业装备结构分析优化与CAE软件全国重点实验室 大连 116024
  • 收稿日期:2023-07-21 修回日期:2023-12-05 出版日期:2024-09-20 发布日期:2024-11-15
  • 作者简介:孙士勇,男,1981年生,副教授。主要研究方向为先进材料结构设计制造与表征。E-mail:sunshy@dlut.edu.cn
    赵延广(通信作者),男,1981年生,博士,高级工程师。主要研究方向为疲劳、断裂及损伤容限性能测试与分析研究。E-mail:ygzhao81@dlut.edu.
  • 基金资助:
    国家自然科学基金(51975085)和中央高校基本科研业务费专项资金(DUT21GF405)资助项目。

Dynamic Detection Method of Fatigue Crack Growth Based on Image Comparison

SUN Shiyong1, ZHU Dianli1, LI Jiaqi2, FAN Junling3, YANG Rui1, ZHAO Yanguang4   

  1. 1. School of Mechanical Engineering, Dalian University of Technology, Dalian 116024;
    2. DUT-BSU Joint Institute, Dalian University of Technology, Dalian 116024;
    3. China Aircraft Strength Research Institute, Xi' an 710065;
    4. State Key Laboratory of Structural Analysis, Optimization and CAE Software for Industrial Equipment, Dalian University of Technology, Dalian 116024
  • Received:2023-07-21 Revised:2023-12-05 Online:2024-09-20 Published:2024-11-15

摘要: 针对航空铝合金等材料表面的疲劳裂纹,提出一种基于图像对比的动态裂纹扩展检测方法,通过N次灰度作差双阈值设定的方案实现复杂表面状态下裂尖位置的检出,获得裂纹扩展曲线;展开材料疲劳试验,采集不同表面状态下的样件裂纹图像,通过与传统的静态裂纹识别算法比较,验证动态裂纹检测方法检出复杂表面状态下裂纹的可行性;与直读法比较,相对误差不大于4%,验证检测算法的准确性;通过讨论不同表面状态下的灰度作差方案、阈值判定、作差次数等,为多表面状态下裂纹的自动化检测提供参考与借鉴。白色表面状态下作差次数N取2,黑色表面状态下作差次数N取1,含划痕及污渍表面状态下N取2,能够较好地反映裂纹扩展状态。

关键词: 图像对比, 疲劳裂纹, 动态检测, 灰度作差, 裂纹扩展

Abstract: A dynamic crack growth detection method based on image comparison is proposed to detect the fatigue cracks on the surface of aviation aluminum alloy. The crack tip location is detected by using the scheme of times N gray difference and double threshold setting, and the crack growth curve is obtained. Fatigue tests are carried out to collect crack images under different surface states. Compared with the traditional static crack detection algorithm, the feasibility of the dynamic crack detection method to detect cracks under complex surface states is verified. Compared with the direct reading method, the relative error is less than 4%, and the accuracy of the algorithm is verified. Furthermore, the gray difference scheme, threshold value determination and number of difference in different surface states are discussed, which provides a reference for automatic crack detection in multiple surface states. The difference times N of white surface state is 2, the difference times N of black surface state is 1, and the difference times N of scratches and stains surface state is 2, which can better reflect the crack growth state.

Key words: image comparison, fatigue crack, dynamic detection, grayscale deviation, crack growth

中图分类号: