首页 | 本学科首页   官方微博 | 高级检索  
     

基于太赫兹图像的航天复合材料粘接缺陷检测方法研究
引用本文:潘钊,温银堂,郑晓康,崔耀耀.基于太赫兹图像的航天复合材料粘接缺陷检测方法研究[J].计量学报,2018,39(4):471-475.
作者姓名:潘钊  温银堂  郑晓康  崔耀耀
作者单位:1. 燕山大学 测试计量技术及仪器河北省重点实验室, 河北 秦皇岛 066004
2. 燕山大学 国防科学技术学院, 河北 秦皇岛 066004
3. 天津航空机电有限公司 防火环控系统产品制造分厂, 天津 300300
基金项目:国家自然科学基金(61403333)
摘    要:采用频率范围为0.23~0.32THz的反射式连续太赫兹波检测系统,对航天复合材料与基板的粘接缺陷进行了无损检测。采用环氧树脂作为粘接剂,制作了隔热复合材料与基板的粘接样件。使太赫兹波聚焦在粘接层处,获得层析图像,根据焦平面附近的图像,可直观判断出粘接面有无缺陷。结合中值滤波,采用模糊c均值(FCM)聚类算法对缺陷图像进行分割得到含有目标缺陷信息的二值图像,计算像素点得到缺陷面积。通过与已知缺陷面积样件对比,对粘接缺陷面积的预测误差在±9%以内。

关 键 词:计量学  太赫兹成像  粘接缺陷  复合材料  无损检测  模糊c均值算法  
收稿时间:2016-12-19

Study on Nondestructive Testing for Bonding Defects in Aerospace Composite Based on Terahertz Image
PAN Zhao,WEN Yin-tang,ZHENG Xiao-kang,CUI Yao-yao.Study on Nondestructive Testing for Bonding Defects in Aerospace Composite Based on Terahertz Image[J].Acta Metrologica Sinica,2018,39(4):471-475.
Authors:PAN Zhao  WEN Yin-tang  ZHENG Xiao-kang  CUI Yao-yao
Affiliation:1. Key Laboratory of Measurement Technology & Instruments of Hebei Province, Yanshan University,  Qinhuangdao, Hebei 066004, China
2. Institute of Defense Science and Technology, Yanshan University, Qinhuangdao, Hebei 066004, China
3. Fire Equipment & Environmental Control Equipment Brunch Factory, Tianjin Aviation Electro-Mechanical Co. Ltd., Tianjin 300300, China
Abstract:Thermal insulation aerospace composite and metal substrates’ bonding defects are tested by the reflection type frequency modulated continuous terahertz wave testing system which the frequency range is 0.23-0.32THz. Thermal insulation composite and metal substrates are bonded by epoxy resin adhesive as the experimental sample. Bonding defects of the adhesive interface can be detected visualized by image neighbor focal plane. Otsu’s threshold algorithm and FCM clustering algorithm is applied to get the segment of defect binary images combined with median filtering, then have the defect area by calculating the number of pixels. Compared with the samples which have known defect area, the FCM clustering algorithm is finally used because of the better segmentation effect and the error is about ±5%.
Keywords:metrology  terahertz imaging  bonding defect  composite  nondestructive testing  FCM algorithm  
本文献已被 CNKI 等数据库收录!
点击此处可从《计量学报》浏览原始摘要信息
点击此处可从《计量学报》下载全文
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号