Magneto-optical Imaging Characteristics And Recognition Method For Welded Defects Under Alternating/Rotating Magnetic Fields

As a key technology for advanced industrial manufacturing,welding has been widely used in various industrial fields.In the continuous welding process,the welding process is sometimes unstable because of the influence of welding power,welding speed,and surface condition of the weldment,which leads to welded defects and directly affects welded quality.These welded defects can lead to serious safety accidents.Therefore,it is necessary to carry out effective non-destructive testing(NDT)to ensure welding quality,in which welded crack detection and incomplete fusion detection are the most important.As a major test technology in industrial manufacturing,NDT plays a key role in the detection of welded defects.There is a future development tendency to integrate nondestructive testing technologies such as ultrasonic testing,radiographic testing,magnetic flux leakage,and eddy-current testing with machine vision to detect welded defects.However,each NDT method has its own advantages and limitations.In this paper,a magneto-optical(MO)imaging welded defect detection method based on Faraday rotation effect is studied.Compared with the traditional NDT method,MO imaging technology can realize the visualization of welded defects without complicated signal processing.This technique has high sensitivity and can be used to detect surface and subsurface defects.This paper studies MO imaging law of welded defect under direct-current(DC)magnetic field excitation.A finite element simulation model of welded defect is set up,and the leakage magnetic field distribution of different lift-off heights and different excitation voltages are analyzed,and the methods to improve the detection accuracy of welded defects are explored.A MO imaging inspection system for welded defect excited by DC magnetic field is established.According to the finite element simulation results and MO imaging experiments,the optimal parameters for welded defect detection are obtained.Aiming at the problem that the MO image of welded defect is easily saturated under the excitation of DC magnetic field,MO imaging law of welded cracks under alternating magnetic field excitation is studied.Welded surface and subsurface cracks are detected by a MO sensor,and the relationship between the MO images'characteristics and the magnetic field strength is analyzed based on the Faraday MO effect.A magnetic dipole model is used to study the magnetic field distribution of the welded crack.A finite element analysis(FEA)model of the welded crack is established,and the relationship between the magnetic flux leakage signal and the crack width is analyzed,which is useful for identifying cracks either on the surface or on the subsurface of the weld.A MO imaging NDT experiment is carried out to detect welded cracks under alternating magnetic field excitation,and the difference among welded cracks is obtained by analyzing the gray values of the welded cracks' MO images.The proposed finite element model and MO imaging experiment can effectively describe the influence of different welded defects on the distribution of magnetic flux leakage signal and image gray value,which are helpful to improve the welded defect assessment.MO imaging can be used to detect weld defects under alternating magnetic field excitation.However,in an alternating magnetic field,a MO sensor is more sensitive to defects perpendicular to the excitation direction than defects with other directions.Therefore,a multidirectional MO imaging NDT system based on rotating magnetic field excitation is studied to detect welded defects.A FEA model of cruciform welded cracks is established to study the distribution of rotating magnetic fields at different transient time.The finite element model is proved by MO imaging experiments of cruciform welded cracks.MO imaging NDT experiments were performed on welded defects including surface crack,subsurface crack,and incomplete fusion under the excitation of alternating magnetic field and rotating magnetic field.Defect information of the MO images is extracted through fusion image technology.The MO imaging effects of welded defects under alternating magnetic field excitation and rotating magnetic field excitation are compared.The maximum difference of image gray value is analyzed to determine the effect difference between the two magnetic fields.In order to study the detection and classification of welded defects under the excitation of rotating magnetic field,the MO imaging NDT system excited by rotating magnetic field is proposed for feature extraction and detection classification of invisible arbitrary-angle welded defects.The magnetic field direction of the rotating magnetic field changes periodically with time,and the magnitude and direction of the leakage magnetic field on the welded defect also changes periodically.MO imaging NDT experiments are performed on invisible welded defects,including surface crack,subsurface crack,and incomplete fusion under the excitation of rotating magnetic field.The gray-level co-occurrence matrix(GLCM)method is used to extract texture features of the MO image of welded defects,and the texture features of the images can reflect the leakage magnetic field characteristics of the defects.These texture features of MO images are used as input data for the defect classification model established using support vector machine(SVM).This model is evaluated by welded defect detection experiments and shows that it can effectively and accurately identify and classify the invisible arbitrary-angle welded defect.Aiming at the difficult to detect arbitrary-angle welded defects,a magneto-optical MO imaging NDT system for welded defects excited by different magnetic fields was studied.The mechanism of the alternating magnetic field generated by the U-shaped yoke and the rotating magnetic field produced by the plane cross yoke was analyzed.The MO imaging effects of different welded defects excited by alternating/rotating magnetic field were compared.The principal component analysis method was used to extract the grayscale features of the fused image column pixels and the texture features of the MO image were extracted by the GLCM.A classification based on backpropagation(BP)neural network was established to identify these defect features.