Research On The Technology Of GFRP/Nomex Honeycomb Sandwich Structure Defects Detection Using Infrared Thermal Wave Testing Method

Honeycomb sandwich structure is one of the important structural forms in the field of composite materials.It has been widely used in ships,spacecrafts and other fields because of its advantages of vibration mitigation,microwave transmission,high specific strength and impact resistance.Due to the complex manufacturing process and poor service environment of honeycomb sandwich structure,it is easy to produce delamination,water accumulation and other defects,which seriously affect the performance of materials and have huge security risks.Therefore,it is very important to realize the reliable detection of defects.Linear frequency modulation infrared thermal wave detection technology can detect defects of different depths and types at one time,which provides a new method for defect detection of GFRP/Nomex honeycomb sandwich structure.In this paper,the theory and simulation of linear frequency modulation infrared thermal wave detection of honeycomb sandwich structure,the experimental study of linear frequency modulation detection,the thermal image sequence processing algorithm,defect recognition and edge extraction are studied.The heat transfer theory of honeycomb sandwich structure excited by linear frequency modulation is analyzed,and the heat transfer model is established;The influence of defect geometric characteristics and Chirp detection process parameters on the surface temperature signal of the specimen was explored by numerical simulation,The feasibility of linear frequency modulation infrared thermal wave detection technology for honeycomb internal defect detection was discussed.The linear frequency modulation infrared thermal wave detection system was built to realize the effective detection of GFRP/Nomex honeycomb sandwich structure defects;Through the experimental study,the influence of different defect types,geometric dimensions and detection process parameters on the defect detection effect was explored,and the correctness of the heat transfer model was verified by comparison with the numerical simulation,and the reasonable range of detection parameters was obtained.The inter-frame difference-multi-frame cumulative average method,polynomial fitting method,Fourier transform method,principal component analysis method,total harmonic distortion method and thermal signal reconstruction method are used to process the image sequence.The signal-to-noise ratio is defined and the processing effect of each algorithm is evaluated.The results show that the principal component analysis method provides excellent defect recognition effect while maintaining high processing efficiency,and can be used as a preferred method for image sequence processing.The edge detection method of GFRP/Nomex honeycomb sandwich structure defects is studied,and three edge detection hybrid algorithms are used,which are based on adaptive filtering-double threshold segmentation-Canny operator,based on linear transformation-OTSU segmentation-Canny operator,and based on linear transformation-superpixel segmentation-Canny operator,so as to realize the simultaneous detection of debonding and water accumulation defects.A hybrid algorithm based on region growing method and Canny operator is used to realize the effective extraction of ice defect edges.The defect area prediction method based on edge detection image is studied,and the defect area prediction of debonding,water and icing of honeycomb sandwich structure is realized.The minimum prediction error of debonding defect area is 0.8%.The minimum prediction error of water accumulation defect area is 2.3%;the prediction error range of icing defect area is about2%～10.1%.