Study On Longitudinal Thermal Stress Detection Algorithm Of In-Service Rail Based On Multi-mode Ultrasonic Guided Waves

Continuous welded rail(CWR)eliminates rail gaps,reduces wheel-rail impact and is widely used in various rail transportation systems such as high-speed railways,subways,and heavy-haul railways.When the temperature changes,CWR cannot expand and contract freely which could generate longitudinal thermal stress in rail.Excessive thermal stress will lead to runway or rail fracture,which seriously affects the safety of railway operation.CWR faces complex environment during long-term service,including alternating temperature and stress,as well as changes in profile caused by wear and railgrinding.At present,there is still a lack of a convenient and feasible method for the accurate estimation of longitudinal thermal stress of CWR.Aiming at this problem,research on the estimation method of longitudinal thermal stress of in-service rails based on multi-mode ultrasonic guided waves is carried out.This dissertation systematically analyzes the propagation characteristics of ultrasonic guided waves of in-service rail under the influence of the complex service environment.The thermal stress estimation algorithm for in-service rail based on the phase velocity measurement of multiple guided wave modes is proposed.It lays a good foundation for the engineering application of the longitudinal thermal stress detection technology of inservice rail.This dissertation includes the following main contents:First,the propagation characteristics of ultrasonic guided waves in rail under complex service environment are analyzed.Based on this,the rule of selecting the optimal guided wave mode set for the estimation of longitudinal thermal stress of in-service rail is proposed,which can improve the performance of longitudinal thermal stress estimation method based on ultrasonic guided waves.It also provides the theoretical basis for the high-precision longitudinal thermal stress estimation of in-service rails.Secondly,aiming at the problem of frequent changes in material elastic constants of in-service rail,a two-stage rail thermal stress estimation method based on the inversion of elastic constants of in-service rail is proposed.In the first stage,an inversion scheme of material elastic constants of in-service rail based on the phase velocity measurement of multiple ultrasonic guided waves is proposed.Accurate estimation of the elastic constants of in-service rail material is achieved based on the Simulated Annealing method and Nelder-Mead simplex method.In the second stage,single ultrasonic guided wave is used to estimate the longitudinal thermal stress of in-service rail based on the estimated elastic constants.The simulation results show that the algorithm can realize the accurate longitudinal thermal stress estimation with an accuracy of 4 MPa.For the in-service rails in specific area of operation lines,the quick detection of longitudinal thermal stress of inservice rail in multiple locations of the specific area can be achieved using one ultrasonic guided wave mode,which only needs to perform the inversion of the material elastic constants of in-service rail using multiple guided wave modes for once.Then,aiming at the problem of two-stage thermal stress estimation method which is not suitable for the rapid census of large-scale spare locations,a single-stage longitudinal thermal stress estimation algorithm of in-service rail based on the improved genetic algorithm(IGA)and iterative weighted least squares(IWLS)is proposed to further simplify the complexity of longitudinal thermal stress estimation algorithm based the ultrasonic guided waves.The algorithm takes the in-service rail profile and the phase velocities of multiple guided wave modes as input.First,IGA is used to realize the rough estimation of longitudinal thermal stress and Young's modulus,and then IWLS method is used to realize the accurate estimation of the longitudinal thermal stress of in-service rail.In order to verify the robustness and applicability of the longitudinal thermal stress estimation algorithm,the numerical simulation verification with multiple loading conditions under the influence of measurement error is carried out.The estimation results of the IGA-IWLS algorithm are compared with the IGA and IWLS method.The results show that the proposed IGA-IWLS method can accurately estimate the longitudinal thermal stress of in-service rail with different working conditions with an accuracy of 3 MPa.The proposed IGA-IWLS algorithm also has characteristics of large stress estimation range and fast convergence.Finally,aiming at the application scheme in actual detection process of the proposed longitudinal thermal stress estimation algorithm of in-service rail,the study of the longitudinal thermal stress detection method and the finite element simulation verification are carried out.A single ultrasonic guided wave mode excitation method suitable for online rail detection of in-service rails is proposed.The three-dimensional finite element model of in-service rail with absorption boundaries is established using the absorbing layers using increasing damping method.Based on the model,the simulation verification of the detection method in long-distance continuously welded rail is carried out.The results of simulation verification of proposed inversion method of material elastic constants of in-service rail show that the maximum relative estimation error of the elastic constants of in-service rail with different working conditions is less than 3%.The load simulation test of the longitudinal thermal stress of the in-service rail shows that the stress estimation error is less than 4 MPa.The dissertation fully explores the problems faced in the application of longitudinal thermal stress detection method of in-service rail based on the ultrasonic guided wave through the finite element simulation verification,which lays a good foundation for the engineering application of the method.