Research Of Testing Method For Defects In Aerospace Composite Based On Continuous Terahertz Wave

In today’s growing aerospace industry, aerospace aircraft safety problem has become the focus of people’s attention. While the aerospace aircraft in a process of flight, due to the friction with the air in a high-speed situation, the device will produce a lot of heat. Heat insulation composite materials must be used to secure the safety and integrity of the aerospace aircraft fuselage and the internal structure. These heat insulation materials were covered on the surface of the fuselage by the method of bonding. If the bonding layer debonded, the aircraft could have serious consequences. Therefore, it’s practical significance to research the test of bonding quality between the aerospace aircraft and thermal insulation composite materials.Aerospace composite insulation material was bonded with metal substrate as the test object. Aerospace composite insulation bonding quality automatic detection experiments with certain generality were designed in this paper. Based on the technological characteristics of the test object, THz imaging was adopted as a test method and a backward wave reflective optical imaging system and a frequency modulated continuous terahertz wave imaging system were set up for testing. By choosing appropriate frequency of terahertz probe and the step precision of mechanical displacement stages to scan samples, the system realized the extraction of the THz reflection signal, bonding states’ tomographic images of the test objects were obtained. According to the characteristics of the image, using median filter to eliminate background noise in the image and through the Retinex image enhancement algorithm, the image brightness and contrast aspects in test images have been improved a lot. Processing results show that the method effectively reduced the noise of he scan images which may affect the defect characteristics judgment under the premise of good defect information protection.Shen Castan algorithm and adaptive Canny algorithm were applied in the edge detection of terahertz images and the Otsu algorithm and FCM algorithm were applied in the image segmentation of terahertz images.After comparing the segmentation results, the FCM clustering segmentation with better effect was chosen as basis. This method increased the accuracy of the area computation and the traditional debonding area statistical method is improved.