Simulation Of Magnetic Field Distribution And Analysis Of Magneto-Optical Imaging Characteristics Of Weld Defects

In the welding process,various effects from external random factors and welding parameters often occur to result in the weld defects of welding products in the manufacturing process,such as non-fusion,cracks,pits,pores,and slag.In order to ensure the safety of welding products including equipment and tool,the safe condition of the entire production system,and the benefits of the national economy and people's safety,it's necessary to conduct non-destructive testing of weld joints.Non-destructive testing is a technique to inspect and test the type,shape and positional characteristics of internal and surface defects of various materials,parts,test pieces,structures,etc.without damaging or destroying the performance of the object.In this paper,the research object is mainly ferromagnetic material.The magnetic field strength and magnetic induction intensity of the weld defects in the process of leakage magnetic field detection are analyzed by magnetizing the material.Besides,the magnetic field distribution and magneto-optical imaging characteristics of different weld defects are especially studied.Based on the working principle of the leakage magnetic field detection technology,the three-dimensional model of weld defect detection is built up and the magnetic field strength and magnetic induction intensity of the weld defect in the process of leakage magnetic field detection is analyzed.To be specific,the magnetic dipole model is used to theoretically analyze the leakage magnetic field of weld defects,and different types of weld defects are studied by different magnetic dipole models.On top of that,the distribution of magnetic charge factors is analyzed in theory.And according to the requirements of the actual tests,the mode of the defects with different shapes is established.The relationship between cracks,unfused and pits,depth,angle and width is studied by studying the leakage magnetic field signal.Besides,weld defect of the leakage magnetic field analysis model is established.Using the principle of leakage magnetic field,the magnetic field distribution of weld defects in different shapes studied by changing the shape parameter of the weld defects.In closing,the relationship between the depth of the rectangular defect,the angle of the triangular defect and the width of the elliptical defect and the leakage magnetic field signal of the defect is found.And the experimental results show that the finite element analysis results are consistent with the theoretical analysis results.The correlation between leakage magnetic field intensity in different magnetic field directions and the characteristics of welding defects is analyzed by alternating current detection technology.To be specific,the magnetic field distribution law of the electromagnet under alternating excitation is explored,and the magneto-optical images of defects on the weld surface under different magnetization directions are obtained.What's more,the simulation and experiment of different types of welding defects are carried out to study the correlation between the gray level co-occurrence matrix of magneto-optical images and the corresponding leakage magnetic field signals in different magnetization directions.Besides,large number of magneto-optical images corresponding to the weld defects in real objects are preprocessed,and the defect contour is extracted more accurately by using the segmentation algorithm based on color vector.The region of interest is denoised and extracted.The geometric features and gray co-occurrence matrix texture features of the magneto-optical images are extracted.Taking the feature as the input of fuzzy clustering algorithm,this paper studies the weld defect identification model based on fuzzy clustering by selecting appropriate fuzzy index,input key eigenvalues and increasing weld defect samples,so as to provide theoretical and experimental basis for weld defect identification.