Defect Detection Of Pressure Vessel Shell Based On Eddy Current Pulsed Thermography

Pressure vessels are widely used in industrial applications,and indispensable in petrochemical,military,pharmaceutical,chemical smelting and so on.A problem associated with the wide application is the failure of the pressure vessel.Once it fails,it will cause economic losses,even lead to accidents such as explosions and casualties.Therefore,it has been an important issue to detect and eliminate the cause of the failure of the pressure vessel before the accident.Early defects before the failure of the pressure vessel may be just small cracks on the surface,which are often difficult to visually recognize.At present,there are many non-destructive testing methods in the industry,which can detect defects while avoiding damage to the pressure vessel.However,the conventional detection methods are usually labor-intensive and the detection efficiency is not high or the cost is high.In order to solve the above problems,this paper uses the eddy current pulse thermal imaging detection method to identify the defects through the temperature distribution of the pressure vessel,and combines the generalized morphological component analysis algorithm and the edge detection algorithm to detect the surface crack defects.The main research contents of this paper are as follows:Based on the principle of eddy current pulse thermal imaging detection,a simulation model was established to simulate the factors affecting the surface temperature distribution of the pressure vessel.The effects of crack on the eddy current field and temperature field of the pressure vessel and the relationship between temperature and crack size and surface curvature of the pressure vessel were analyzed..Based on the principle of eddy current pulse thermography,the experimental system was built.The effects of crack depth and heating time on temperature were analyzed by artificially fabricated cracks and naturally formed cracks.Combining generalized morphological component analysis and sobel edge detection algorithm,the thermal image is denoised and edge extracted by simulation and experiment.The denoising effect of the image and the relationship between the crack depth and the denoising effect of the generalized morphological component analysis algorithm are analyzed.And compared the difference whether thermal images are denoised or not and the advantage of the eight-direction sobel edge detection method compared with the traditional edge detection method.The relationship between the gray value of the image and the crack depth and heating time are analyzed.Through the simulation and experimental research on the pressure vessel,it is proved that the eddy current pulsed thermal imaging detection technology can be applied to the crack detection of the curved shell of the pressure vessel.The main factors affecting the temperature distribution at the crack of the pressure vessel are studied,and the method of using the temperature increase ratio to measure the crack depth was proposed.Furthermore,the generalized morphological component analysis is used to reduce the thermal image noise interference to highlight the defects.The relationship between the denoising effect and the crack depth is analyzed.Then the edge information of the high temperature region of the crack defect is extracted by the sobel edge detection,and a method of determining the crack depth by the characteristics of a sudden change in the gray value of the region is proposed.