AUTHOR=Chang Hao-Liang , Ren Hao-Tian , Wang Gang , Yang Ming , Zhu Xin-Yu 

TITLE=Infrared defect recognition technology for composite materials

JOURNAL=Frontiers in Physics

VOLUME=Volume 11 - 2023

YEAR=2023

URL=https://www.frontiersin.org/journals/physics/articles/10.3389/fphy.2023.1203762

DOI=10.3389/fphy.2023.1203762

ISSN=2296-424X

ABSTRACT=Composite materials are widely used in general aircraft, and multiple aircraft types are composed by more than 90% of composite materials. As a result, the defects of composite materials will seriously affect the flight safety. In this paper, the feasibility of the experiment was verified by simulation. In simulation, the minimum accuracy of detectable defects is 4 mm radius under the mesh division accuracy with a correlation coefficient of 5. On this basis, an accurate detection method and prototype nondestructive testing system for defects of aircraft composite materials based on infrared imaging detection technology were designed, which can realize the identification and positioning of defects in aircraft composite material structures, including type, size and accurate depth of defects. Finally, the data collected by the infrared detection system was recognized through YOLO neural network. The test result shows the confidence level for water point defect is more than 0.9, while the confidence level for crack defect is about 0.5. This research result will expand the use case of infrared nondestructive testing technology at home and abroad, and the transformation of the results will help to solve the maintenance problems of aircraft in general aviation.