Quantitative Analysis Of Debond Defects Detection In Composite Materials Using Infrared Thermography Nondestructive Testing

Carbon fiber reinforced plastic(CFRP)laminate as a kind of advanced composite materials,because of its superior performance such as low weight and high strength,has been more and more widely used in aerospace,military,automotive and other fields.During the course of manufacture and use,the composite materials may be produced various types of defects,such as delamination,unstick,crack,and so on.These defects can reduce the performance of composite materials and lead to potential safety hazard.Therefore,researching and developing non-destructive testing technology for effective detection and evaluation of molding defects in composite materials is the essential means to ensure the quality of materials and safe operation.Aimed at two different properties of defects in CFRP,the infrared thermography non-destructive testing(NDT)technology with two different incentive methods is used in this thesis.The main contents are as follows:(1)The research of impact damage detection based on eddy current pulsed thermography(ECPT).The ECPT is a kind of NDT technology which heats the specimen by means of electromagnetic excitation.Then,the video of the surface temperature filed recorded by infrared thermal imager is used for detecting and evaluating defects.It has high detection efficiency and results directly.In this study,the principal component analysis method is applied to extract features of the original image sequence,and the image features are used for analyzing the damage in the CFRP specimens made by different impact energy.In order to detect the small damage caused by 4J low impact energy,a multi-resolution method employing the wavelet transform combined with principal component analysis,is used to extract the defective characteristics.And the detectability of impact energy producing damage are also discussed.(2)The research of debonding flaw detection based on flash thermography.For the composite specimens which have the debonding flaws in the non-conducting cementing layer,the flashlight is used as a thermal excitation to give specimens a surface uniform temperature filed.The abnormal phenomenon of surface temperature distribution which caused by its internal debonding flaws impeding the longitudinal thermal diffusion is observed and analyzed the properties of defects.A variety of image feature extraction algorithm are applied to deal with the original data,and the F-score criterion is used for quantitative evaluating the test effect of all algorithms.Finally,an image segmentation method based on region growing is applied to separate defective information from the thermal image.And the relation of actual defective size and the defective size in thermal image has been analyzed.