Study On Defect Detection Of Carbon Fiber Composites By Line Laser Scanning Thermography

Carbon fiber reinforced polymer(CFRP),as a representative of thermosetting resin-based composite materials,has superior properties and is extensively employed in aerospace,military industry,new energy and other fields.However,CFRP products are prone to defects during production and use,resulting in material strength and service life significantly reduced,therefore,nondestructive testing(NDT)is a crucial aspect to be integrated into the whole process of composite materials from production to use.The line laser scanning thermography NDT technology uses a line laser as the thermal excitation source and adopts the scanning heating method,which has a larger detection area per unit time and higher detection efficiency;It is especially suitable for large-size and long-distance carbon fiber composite materials NDT scenarios.The research on this technology is mainly carried out from the following aspects:(1)The existing NDT technology of carbon fiber composites is introduced,and the current research status of infrared thermography testing of carbon fiber composites at home and abroad is described,and the thermodynamic theory involved in line laser scanning thermography is briefly introduced,a line laser scanning thermography NDT system is built,and experiments on debonding and surface defect detection are conducted.(2)In order to improve the detection capability of the detection system,three parameters affecting the detection effect of the system were summarized,including scanning direction,laser power and scanning speed;The simulation model was established to investigate the relationship between the three parameters and the detection effect,and the results showed that the three parameters do have an impact on the detection effect of the system,and the principle of selecting parameters that take into account the detection effect and the detection efficiency was summarized in the detection.(3)For the disadvantages of weak contrast and blurred edges of infrared images that are not conducive to surface defect feature extraction,a edge feature extraction method based on the intuitionistic fuzzy C-mean clustering algorithm is introduced,compared with edge operator extraction results and the clustering results based on the K-Means algorithm,the method can improve the recognition and detection ability of defect fuzzy edges and retain more detail information while suppressing certain noise interference.The method can achieve the purpose of complete extraction of defective edges.(4)For the quantification of surface defects,a method of quantifying the area and length of defects by combining the object-image relationship with "finding the minimum enclosing circle" algorithm is used,which has a high quantification accuracy of length or diameter with a minimum error of 12.99%,but a large quantification error of area with a maximum error of37.26%.The simulation model was established to investigate the relationship between the depth of debonding defects and the temperature difference of the defect surface,and the results showed that the smaller the defect depth,the easier it is to be detected,and there is a limit to the detection of the depth of the laser scanning thermography method,beyond which the temperature curve of the defect surface overlaps,and it is impossible to distinguish the defect depth.