Multi-physics Based Electromagnetic And Thermography For Non-destructive Testing Of Rail-track Cracks

The rail-head crack damages induced by wheel-rail rolling contact fatigue will reduce the performance and usability of rail-track.It is one of the significant factors threat the safety operation of high speed railway.The online nondestructive testing(NDT)used for in-service rail-track inspection in high speed is an effective approach to find cracks for timely repair and maintenance,which can prevent the operational failure or structural fracture of rail-track.The electromagnetic NDT is a non-contact way to find rail-head cracks via scanning detection in high speed,which can make up the limitations of conventional ultrasonic testing on surface and subsurface cracks inspection.The electromagnetic induction thermography has advantages e.g.high resolution,large area detection within a short time and abundant features from multiple physical responses.It is potential to be used for early stages smaller and multiple cracks detection and characterization.Based on the magnetic flux leakage(MFL)and eddy current pulsed thermography(ECPT),this thesis investigates the multi-physics based electromagnetic and thermography techniques for the NDT of rail-track cracks in particular surface complex shaped angular crack and natural fatigue cracks.The main research works are as follows.(1)To solve the problem of the scanning speed influence on the MFL for rail-head cracks inspection,the magnetization mechanism of ferromagnetic material and the speed effect of motional magnetic field are investigated with physical analysis.Two models including local resistance model of magnetic circuit and skin effect model of motional magnetic field are proposed to understand the speed effect mechanism of the MFL in high speed.Numerical simulations have been carried out to analyze the relationship of the MFL signals of surface open crack with respect to material permeability decrease variation induced by speed effect.Experimental studies have been performed on various types of surface open cracks and subsurface defect for validation using the MFL with different speeds ranging from 0m/s to 50m/s.In contrast to conventional eddy current model,the research achievements provide theoretical foundation to understand the speed effect mechanism of motional magnetic field and its influence on the MFL signals.It also can provide technical support for the MFL detection structure optimization and its online real applications for rail-head cracks inspection.(2)To break through the limitation of single magnetic sensors to collect comprehensive information of leakage magnetic field for rail-head complex shaped cracks quantitative characterisation,a 3D sensor array based MFL scanning visualization method is proposed based on the spatial distribution characters of leakage magnetic field.Numerical simulations have been carried out to find the relationship of 3D MFL imaging characters and cracks geometrical shape.It can be known that the lift-off and sensing area of magnetic array sensors influence the detection sensitivity and imaging resolution of smaller and multiple cracks.A leakage magnetic field feature analysis method is proposed for complex shaped angular cracks characterization.Through the simulation analysis of MFL components along with different magnetic field vector directions from three coordinate plane,the relationship of MFL components features response to defect propagation direction and angles of complex shaped angular cracks is found.Experimental validation and online application test have been carried out on rail-track sample including various types of surface manmade cracks(minimum defect width 200um)and natural fatigue cracks(width about 20um)using the 3D sensor array MFL method via scanning visualization in different speeds.The research achievements can provide theoretical foundation and technical guidance for the MFL imaging inspection of rail-head complex shaped angular cracks and signal features extraction.(3)To realize the accurate imaging inspection of rail-head natural fatigue cracks,a multi-physics based electromagnetic pulsed thermography method is proposed based on the advantages integration of the MFL and ECPT.A ferrite-yoke core is used to design a new excitation configuration,which can produce uniformly distributed electromagnetic-thermal field on rail-head.A thermal camera is used to measure surface temperature distribution of rail-head to realize cracks geometry profile visualization via high resolution thermography.Numerical simulations have been carried out for the multiple physics analysis via electromagnetic-thermal field interactions.The directional distributions of magnetic flux and eddy current and related hysteresis loss and Joule heating effect have been analyzed and used for crack orientations characterization.Experimental validations have been performed on rail-track samples including complex shaped angular cracks and natural fatigue cracks using electromagnetic pulsed thermography.The multi-physics response features are extracted for cracks geometry profile,propagation direction and depth characterization.Comparison studies have been conducted to analyze the ferrite-yoke lift-off,speed and excitation configurations influence on rail-head cracks inspection using electromagnetic thermography.The research achievements can provide theoretical foundation and technical guidance for quantitative evaluation of rail-head natural fatigue cracks and the real applications of the multi-physics based electromagnetic and thermography.(4)To solve the emissivity problem induced by surface condition and environmental factors,a method with infrared and visible images fusion is proposed for emissivity influence correction during rail-head cracks inspection using electromagnetic thermography.The infrared and optical cameras are used to capture infrared and visible images simultaneously.The spectral correlation of the above two images are analyzed for emissivity correction.It is based on a hypothesis that the reflectivity is able to predict surface emissivity.Case study-1: A normalized difference vegetation index(NDVI)method is proposed for emissivity correction.It is based on the normalized ratio transformation of reflectivity obtained from infrared and visible band.The emissivity correction coefficient is calculated based on the ratio of local reflectivity values to global one.It has been verified that this method is effective to reduce emissivity influence.Case study-2: A method using spectrum correlation of infrared and visible images is proposed for emissivity correction.An invariant coefficient feature is calculated with suggested algorithms link to infrared image with respect to emissivity,and visible image with respect to reflectivity.Experimental results and comparison analysis demonstrate the corrected infrared images in different heating and cooling stages give accurate temperature distribution with correct spatial and transient responses,which is independent of surface emissivity.The main contributions and novelty of this thesis are as follows.(1)The speed effect of motional magnetic field and magnetization process of ferromagnetic material are investigated.Two models including a local resistance model of magnetic circuit and a skin effect model of motional magnetic field are established based on the above theoretical studies.Through simulation and experimental studies,the speed influence mechanism on the MFL test is verified.(2)The 3D sensor array based MFL scanning visualization method and the multiphysics based electromagnetic pulsed thermography method are proposed respectively for rail-head complex shaped angular cracks and natural fatigue cracks detection and characterization.Based on simulation,experimental and real application test studies,the above two methods demonstrate improved capability and efficiency for smaller cracks detection and characterization.(3)A method using infrared and visible images fusion is proposed to reduce the influence of surface emissivity on electromagnetic thermography for rail-head cracks inspection.The above research achievements provide theoretical foundations and technical guidance for rail-track NDT instrument development and the real online applications using multi-physics based electromagnetic and thermography techniques.