Research On X-ray Digital Radiographic Technology For Composite Materials

Composite materials are wildly applied in many fields such as space industry, national defense and construction, for composite materials have characteristics of lightness, high intensity, being easy to be modeled, good anticorrosive properties, fatigue resistant and low expansion coefficient. But the production and application of the products made of composite materials may cause defects that could make economical loss and personal injuries. Consequently the nondestructive testing for composite materials is required. At present, X-ray testing is still one of the nondestructive testing methods of composite materials. But the X-ray images of composite materials have characteristics of poor consistency of gray, low contrast, fuzzy defects edge.Which make the testing and identification of inner defects of composite materials is difficult.Based on this background, this paper studied the X-ray digital radiographic technology for composite materials and the algorithm of X-ray image processing. The specific work done in this paper is as following:(1)The X-ray attenuation rule of composite materials was studied with taking the 4-axes centre DR/CT detection system as testing platform in this paper, which provided important basis for accurate parameters selecting and improving sensitivity of the system. (2)Aiming at the structure properties of composite materials and the characteristics of X-ray detection, the features of X-ray images was analyzed and the impact on images caused by the material properties was mainly studies in the paper. (3)In the image preprocessing, the emphasis is placed on the algorithm research of restricting the scattering noise and according to the deficiency of the traditional approach, the scattering noise reduction method based on the constructor is put forward. (4)On the aspect of defect inspection, the defection algorithm based on EMD(Empirical Mode Decomposition)and iteration threshold segmentation was proposed in the paper, and the extraction and segmentation of the internal faint defects of composite materials was realized.The experiment results showed that the above methods can not only improve the image quality fundamentally, but also enhance the image defect features significantly and extract and divide the defect area successfully, which provided an important theoretical basis for the automatic detection and identification of defects of composite materials in the future.