Random Dynamic Analysis And Running Safety Reliability Evaluation Of Vehicle-Track(Bridge) Coupled Systems

With the increase of train speed and operating mileage,the increase of wheel-rail forces will reduce the smoothness of track structure and accelerate the wear of the wheels and the rails.At the same time,the condition of the vehicle suspension system will degrade due to wear,aging and for other reasons.The above factors directly affect the dynamic performance of the vehicle-track(bridge)coupled system.In addition,China is an earthquake-prone country.As the proportion of bridges in high-speed railway lines is high,the possibility of earthquake occurrence while a high-speed train is travelling on a bridge is big and derailment of trains on bridges during earthquakes will increase greatly.Therefore,it is of great engineering value to carry out the random dynamic analysis and running safety reliability evaluation of vehicle-track(bridge)coupled systems,which can provide reference for the design and improvement of high-speed trains,track or bridge systems.In this theoretical investigation,the influence of uncertainties,such as the wheel-rail profiles,suspension parameters,track irregularities and earthquakes,is analysed.The uncertainty factors affecting the dynamic behaviour and running safety performance of vehicle-track(bridge)coupled systems mainly include:1)the uncertainties of system parameters such as the wheel-rail profiles,vehicle suspension characteristics and so on;2)the model uncertainties of vehicle-track coupled systems due to the small-displacement assumption,different modelling methods of the connecting components and the neglect of small mass attachments;3)the uncertainties of earthquake excitations and occurrence time.In this dissertation,a random dynamic analysis and running safety reliability evaluation of vehicle-track(bridge)coupled systems are carried out from the aspects of the established coupled dynamic model of a vehicle-track(bridge)system,modelling uncertainties and dynamic reliability evaluation.The details are given as follows:1)A linear wheel-rail contact model considering the influence of the change of contact point positions and the vibration of rails is presented.The equations of motion of linear vehicle and track subsystems are established separately and the dynamic responses are solved with the iterative method.Random responses of the vehicle-track systems under the influence of random track irregularities are determined using the pseudo excitation method,and the power spectral density and variance in spatial domain of response indices,such as the wheel-rail forces and car body acceleration,are obtained.The applicability of the linear vehicle-track dynamic model is found by comparing it with the nonlinear vehicle-track dynamic model,in which the effect of practical wheel-rail profiles can be considered.2)In order to overcome the limitations of the linear vehicle-track dynamic model,a nonlinear wheel-rail contact model considering the effect of practical wheel-rail profiles is presented and an iteration solution method based on prediction of wheel-rail forces,called prediction iterative method,is proposed and applied to determine the dynamic responses of vehicle-track coupled systems.The Weighted Least-Squares Error(WLSE)predictor is introduced into the iterative solution process.By efficient prediction of wheel-rail forces and adoption of relaxation technique,the problems of difficultin convergence and the excessive number of iterations in the conventional iterative process can be avoided.The present method not only enhances the efficiency but also ensures the solution accuracy.3)An efficient method for dynamic reliability evaluation of the vehicle-track coupled systems is presented by combining Subset Simulation method with the prediction iterative method proposed by the authors.The failure probabilities of the lateral ride index on tangent track and the wheel derailment coefficient on curved track are calculated respectively,considering the influence of random factors such as the wheel-rail profiles,vehicle suspension parameters,track irregularities and wheel-rail friction coefficient.The accuracy and the efficiency of the present method for the dynamic reliability assessment of the vehicle-track coupled systems are verified by comparison with the Direct Monte Carlo Simulation(DMCS)method.4)Based on the non-parametic uncertainty modelling theory,a non-parametric probabilistic model of a vehicle-track coupled system is established using the random matrix theory.The non-parametric uncertainties of the vehicle and track subsystems are considered respectively.The dispersion level of each random matrix is controlled by the corresponding dissipation parameter.The number and the type of uncertainties in the subsystems need not be known accurately and the modelling error induced by inaccurate information or unknown details can well be characterized.5)The dynamic model of the train/ballast track/long-span bridge coupled systems under earthquakes is established and the effects of earthquake occurrence time on the safety performance of a train running on a long-span bridge are analysed from the perspective of dynamic reliability.The dynamic reliability of train/track(bridge)coupled systems is efficiently determined by combining Subset Simulation method with the prediction iterative method proposed by the authors,considering the randomness of train locations when earthquake occurs,the spatial variation and the parameter randomness of earthquake excitations.The intervals of the most unfavourable train position under various situations at the time of the earthquake are obtained.