Research On Key Issues Of Electromagnetic Tomography And Its Application In High-Speed Rail Wheel Defect Detection

In the last decade,China’s high-speed railway has developed rapidly and has developed into national-level strategies.High-speed rail wheel is the only moving part that high-speed rail vehicle contacts with the rail.Therefore,high-speed rail wheel works under very complex stress conditions,which causes small defects under the fatigue conditions are prone to appear stress concentration phenomenon,accelerating the diffusion speed,which may lead to fatigue cracks and other faults.Therefore,the regular non-destructive testing of high-speed rail wheel is of great significance to the safety of driving and people’s lives and property.Electromagnetic tomography(EMT)technology has advantages of non-contact,non-intrusive,non-radiation,low cost,fast imaging speed and sensitivity to both conductive and magnetic permeability materials.This thesis explores a novel method for high-speed rail wheel defect detection.The feasibility and validity of applying EMT in high-speed rail wheel defect detection are studied.The main contents of this thesis are as follows:(1)The EMT forward problem and inverse problem and three methods for solving sensitivity matrix were studied.Besides,the theory and numerical analysis method of EMT defect detection sensor were researched.The finite element method and its application in EMT were mainly studied.(2)A novel image reconstruction algorithm named L1-Lp was proposed in this thesis.The novel algorithm constructed new objective functional with L1 norm on the data term and Lp norm on the regularization term,which transformed inverse problem of EMT into an optimization problem.The proposed algorithm improves the sensitivity of image reconstruction algorithm to data errors.Numerical simulation and experimental results demonstrated the effectiveness of the proposed algorithm.(3)A modified Landweber iterative algorithm was proposed in this thesis through appending an additional regularization term into the objective functional of EMT inverse problem,which improved the quality of reconstructed images and convergence process.Both numerical simulation and experimental results demonstrated the feasibility and effectiveness of the modified Landweber iterative algorithm.(4)The impact of number of coils in O-type EMT sensors on quality of reconstructed images was studied.The simulation research platform was built using finite element simulation software.Five kinds of sensors with different number of coils were involved to conduct numerical simulations for five kinds of typical conductivity distributions in the object field.According to the quality of reconstructed images,the optimal number of coils under a certain sensor size was determined.(5)The theory of direct 3D EMT imaging was studied,including 3D EMT forward problem,sensitivity matrix and 3D EMT inverse problem.Direct 3D EMT research mainly includes 3D sensor layout,measurement strategy,numerical simulation,3D image reconstruction algorithm and 3D display.The FEM simulation model of direct 3D EMT was built and numerical simulations were conducted.The simulation results of reconstructed images verified the feasibility and effectiveness of direct 3D EMT imaging technique.(6)According to the structural characteristics and electromagnetic properties of materials of high-speed rail wheel,a U-type EMT sensor was designed,and the finite element model of high-speed rail wheel defect detection system with EMT was built.Typical types of superficial defects were simulated and reconstructed images were obtained.A high-speed rail wheel defect detection experimental system with EMT was further built,including the design and processing of defect detection bench,the design and processing of U-type sensor,the design of signal generation and data acquisition module,and the design of signal conditioning and channel switching module.The experimental system was used to carry out experimental study on typical defects.