Research On Weld Defect Monitoring Based On Magneto Optic Imaging

The processing of ferromagnetic materials is the basis of contemporary industry,which has been widely used in the fields of metallurgical energy,automation,mechanical automobile,railway,bridge,petrochemical industry,etc.Therefore,the detection of ferromagnetic materials related equipment is of great significance for the development of modern industry.In the welding process of ferromagnetic materials and before and after putting into operation,failure or aging will inevitably occur,and the weld is the most prone to produce defects and damage.When the current mainstream nondestructive testing technology is used to detect the weld defects of ferromagnetic workpiece,there are some problems,such as weak correlation between signals,poor resolution,poor antiinterference ability,and the detection effect of weld defects is poor.Magneto Optic Imaging Nondestructive testing technology can realize the visual detection of defects by converting the magnetic field information into photometric information.It has high detection resolution and strong correlation between single image pixels.It can be applied to the non-destructive testing of ferromagnetic workpieces and has strong applicability to the weld detection of ferromagnetic workpieces.This paper starts from the application difficulties of magneto-optical detection technology to the welding seam of ferromagnetic workpiece.Through the research on the characteristics of magnetic leakage field of weld defects,the recognition method of weld defects based on the characteristics of magneto-optical images,and the enhancement of magneto-optical image sequence,the paper conducts the research on the detection of ferromagnetic workpiece,especially the weld defects.The main research points and innovations are as follows:(1)Firstly,based on the principle of magneto-optical imaging,the problems that may appear in the application of magneto-optical imaging to the welding seam of ferromagnetic workpiece are discussed by analyzing the working conditions of the ferromagnetic workpiece.Based on the finite element analysis method,the simulation model of the weld seam of ferromagnetic workpiece is established,and the influence of the actual situation on the distribution of magnetic leakage field is discussed.(2)Based on the principle of magneto-optical detection,the platform structure of magneto-optical imaging detection system is designed.In order to obtain more high quality and more abundant magneto-optical image with more defects information,the parameters of magneto-optical detection system of ferromagnetic workpiece are discussed according to the magneto-optical detection principle,and the components of each part are optimized.Finally,the installation and construction of magneto-optical imaging detection system platform is completed.Then,according to the relationship between the gray value of pixel points and the spatial distribution of magnetic leakage caused by defects,a sliding window detection method based on the feature of defect extreme region is proposed,which can effectively suppress the interference caused by uneven brightness distribution and thin film interference in magneto-optical images,Finally,the defects identification of magneto-optical images under various conditions is completed effectively.(3)According to the movement characteristics of magneto-optical detector of ferromagnetic workpiece in actual detection,a magneto-optical dynamic detection system platform is designed and built.And the reconstruction and enhancement method of magneto-optical image sequence based on principal component analysis is proposed.Finally,based on the problem that the main part of the actual detection is to extract abnormal frames,another dimension principal component dimension reduction analysis method is proposed,and the defect detection is successfully completed by the detection method of the defect extreme value region.