A Research On Infrared Nondestructive Testing Technology Of Carbon Fiber Reinforced Plastic

Carbon Fiber Reinforced Polymer/Plastic(CFRP)is one of the best high-temperature protective materials at present.The composite material composed of carbon fiber and epoxy resin has high strength,heat insulation,high temperature resistance and excellent radiation performance.In recent years,it has been widely used in high temperature and high heat flow environments such as power machinery and aerospace.Although CFRP has broad application prospects,carbon fiber composite materials are prone to debonding,voiding and other defects during production and use due to their complicated structure,which is enough to affect their performance and longevity,and even cause serious harm.Therefore,it is necessary to use real-time detection and evaluation by non-destructive testing technology.Infrared nondestructive testing technology(NDT)provides a new method for defect detection of CFRP flat bottom hole specimens due to its high detection efficiency,easy operation and non-contact.In this paper,the theoretical basis of CFRP pulse excitation infrared non-destructive testing is firstly described.The theoretical derivation of heat conduction and surface temperature field distribution under pulse excitation is given.Then,the carbon fiber composite material is excited by pulse excitation using three-dimensional finite element simulation software(COMSOL?).The(CFRP)laminate was simulated,and the temperature field distribution of the carbon fiber reinforced plastic(CFRP)surface at different times was obtained.The sequence of these images was analyzed,and the surface temperature difference was obtained under different defect diameters and defect depths.The graph of contrast versus time analyzes the effects of different defect depths and diameters on defect detection.Finally,this paper combines the common image sequence processing algorithm,and proposes a Markov-independent component analysis method,a new image sequence processing algorithm.The method mainly uses the Markov chain current state to be related to the previous state,and reconstructs the image sequence based on the properties unrelated to all previous states.The image sequence is then separated by the independent component method to separate useful information and noise,thereby improving the sensitivity of defect detection.In this paper,two different CFRP plates of test 1 and test 2 were selected and two different image sequences were obtained.Based on this,the Markov-independent component method proposed in this paper was verified.Compared with algorithms such as polynomial fitting,SVD decomposition,pulse phase and independent component analysis,the Markov-independent component analysis method proposed in this paper has the best visual effect,higher sensitivity to defect detection,and strong noise suppression ability.Quantitative analysis of the results of the test piece 2 also found that the image processed by Markov-independent component analysis has the highest signal-to-noise ratio among several algorithms.It is verified that the Markov-independent component analysis method proposed in this paper has certain advantages over algorithms such as polynomial fitting and SVD decomposition.