A method and system for predicting long-term settlement of high-speed railway ballastless track subgrade

Abstract

The present invention relates to the field of railway subgrade engineering technology and discloses a method and system for predicting the long-term settlement of a high-speed railway ballastless track subgrade. The method comprises the following steps: determining relationship data between the cumulative plastic strain of subgrade filler and the number of cyclic dynamic load actions; establishing an empirical model for predicting the cumulative plastic strain; numerically simulating the long-term settlement of a target subgrade by equating the number of cyclic load actions to a time unit and the cumulative plastic strain to static creep; constructing a track-subgrade finite element model and a train-track multi-body dynamics model; and using iterative calculation to perform joint simulation to establish a train-track-subgrade coupled dynamics numerical simulation model; substituting train speed, operating time, initial subgrade moisture content, and the number of wetting times into the train-track-subgrade coupled dynamics numerical simulation model to calculate the vertical deformation of the subgrade top surface, and predicting the long-term settlement evolution law of the target subgrade based on the vertical deformation of the subgrade top surface.

Description

Technical Field

[0001] The present invention relates to the technical field of railway subgrade engineering, and in particular to a method and system for predicting the long-term settlement of a high-speed railway ballastless track subgrade. Background Art

[0002] During the operation of high-speed railways in hot, humid, and rainy regions, the subgrade is subject to long-term alternating wetting conditions due to natural factors such as rainfall and groundwater level fluctuations. This can lead to significant changes in the mechanical properties of the subgrade soil, which in turn affects the subgrade's settlement characteristics. With the rapid development of high-speed railways, subgrade settlement has become a key factor affecting the safety and comfort of high-speed rail operations. Due to its unique engineering site characteristics and operating environment, high-speed rail red clay subgrades are prone to significant cumulative settlement under the influence of wetting, which in turn affects the safety and stability of train operations. Therefore, to improve the safety and stability of train operations, it is necessary to understand the changes in subgrade settlement under the influence of wetting and conduct subgrade settlement prediction.

[0003] At present, in the existing technology, the commonly used roadbed settlement prediction methods mainly include the layered summation method, the curve fitting method and the finite element method. However, these methods have many shortcomings when predicting the settlement evolution of high-speed railway roadbeds with certain specific soil types, especially the high-speed railway red clay roadbed under the action of wetting. For example, the layered summation method is difficult to consider the influence of soil lateral deformation and time factors. Although the curve fitting method can establish a settlement and time relationship model, it is difficult to accurately describe the influence of wetting on settlement evolution. The finite element method requires the establishment of a complex mathematical model, which is computationally intensive and sensitive to parameters. The calculated long-term settlement is usually quite different from the actual value. It can be seen that the existing roadbed settlement prediction methods have problems such as incomplete consideration of factors and low calculation accuracy. Summary of the Invention

[0004] The present invention provides a method and system for predicting the long-term settlement of a high-speed railway ballastless track subgrade, so as to solve the problems of incomplete consideration of factors and low calculation accuracy in existing subgrade settlement prediction methods.

[0005] In order to achieve the above object, the present invention is implemented through the following technical solutions:

[0006] In a first aspect, the present invention provides a method for predicting long-term settlement of a high-speed railway ballastless track subgrade, comprising:

[0007] Collect target roadbed fillers and determine the relationship between the cumulative plastic strain of the roadbed fillers and the number of cyclic dynamic loads based on the moisture content, wetting times and dynamic stress characteristics of the target roadbed fillers;

[0008] Based on the relationship between the cumulative plastic strain of the dynamic roadbed filler and the number of cyclic dynamic loads, an empirical model for predicting the cumulative plastic strain is established, which takes into account the moisture content amplitude, the number of wetting times and the dynamic stress amplitude.

[0009] Numerical simulation of the long-term settlement of the target roadbed is performed by equating the number of cyclic loads to time units and the accumulated plastic strain to static creep. A track-roadbed finite element model is constructed based on the numerical simulation results.

[0010] Constructing a train-track multi-body dynamics model, and using iterative calculation to jointly simulate the track-roadbed finite element model and the train-track multi-body dynamics model to establish a train-track-roadbed coupled dynamics numerical simulation model;

[0011] The train speed, operating time, initial moisture content of the roadbed and the number of wetting times are substituted into the train-track-roadbed coupled dynamic numerical simulation model to calculate the vertical deformation of the roadbed top surface, and the long-term settlement evolution law of the target roadbed is predicted based on the vertical deformation of the roadbed top surface.

[0012] Optionally, the relationship data between the cumulative plastic strain of the roadbed filler and the number of cyclic dynamic loads is determined based on the moisture content, wetting times and dynamic stress characteristics of the target roadbed filler, including:

[0013] Based on the moisture content, wetting times and dynamic stress characteristics of the target roadbed filler, a dynamic triaxial test is conducted to obtain the relationship curve between the cumulative plastic strain of the target roadbed filler and the number of cyclic dynamic loads.

[0014] The relationship curve between the cumulative plastic strain and the number of cyclic dynamic loads is used as the relationship data between the cumulative plastic strain and the number of cyclic dynamic loads of the target roadbed filler.

[0015] Optionally, the cumulative plastic strain prediction empirical formula considering the moisture content amplitude, the wetting number and the dynamic stress amplitude is established based on the relationship data between the cumulative plastic strain of the roadbed filler and the number of cyclic dynamic loads, including:

[0016] Based on the relationship between the cumulative plastic strain of roadbed filler and the number of cyclic dynamic loads, empirical models for predicting cumulative plastic strain were constructed for stable and destructive roadbeds, respectively, with water content ratio, dynamic stress amplitude, number of wetting times, and number of vibrations as variables.

[0017] For the empirical model of cumulative plastic strain prediction of stable roadbed, the model satisfies the following relationship:

[0018] ;

[0019] For the empirical model of cumulative plastic strain prediction of damaged roadbed, the model satisfies the following relationship:

[0020] ;

[0021] Where, represents the predicted accumulated plastic strain, represents the soil moisture content, represents the optimal moisture content of the soil, represents the dynamic stress amplitude, Indicates the local atmospheric pressure, Indicates the number of humidification cycles, represents the number of dynamic load actions, k1, k2, k3, k4, k5, and b are fitting coefficients, and the fitting coefficients are obtained by nonlinear least squares curve fitting using the Levenberg-Marquardt algorithm.

[0022] Optionally, the numerical simulation of the long-term settlement of the target roadbed by equating the number of cyclic load actions to a time unit and the accumulated plastic strain to static creep includes:

[0023] The numerical simulation of the long-term settlement of the target roadbed is equivalent to the calculation of soil strain. The number of cyclic loads is equivalent to the time unit, and the accumulated plastic strain is equivalent to static creep to calculate the soil strain. The calculation of soil strain includes elastic strain calculation, plastic strain calculation and creep calculation.

[0024] The critical state theory is used to calculate the time-independent elastic strain. The stiffness constant expressed as modulus in the elastic strain calculation is:

[0025] ;

[0026] ;

[0027] Where K and G are the elastic bulk compression modulus and shear modulus respectively, E is the elastic modulus, μ is the Poisson's ratio, and e is the soil porosity. is the slope of the rebound line in the e-lnp space, is the effective mean stress;

[0028] The plastic strain is calculated by following the yield criterion of the modified Cambridge model and the associated flow law. Under the premise that the plastic potential function is the same as the yield surface function, the calculation of the plastic strain satisfies the following relationship:

[0029] ;

[0030] Where, is the slope of the critical state line, is the deviatoric stress, and here the formula is , J2 is the second invariant of the deviatoric stress tensor, is the initial consolidation pressure of the soil, is the effective mean stress;

[0031] The calculation for creep satisfies the following relationship:

[0032] ;

[0033] Where, is an empirical expression for the axial cyclic cumulative strain rate of soil samples under undrained triaxial conditions based on experimental or measured data. is the creep potential function, is the rate-dependent cyclic cumulative strain rate tensor, is the dynamic stress level, is the total strain rate, is the dynamic load loading time, are other internal variables, is the effective mean stress, is the symbol of partial derivative.

[0034] Optionally, the adopting of an iterative calculation method to jointly simulate the track-roadbed finite element model and the train-track multi-body dynamics model to establish a train-track-roadbed coupled dynamics numerical simulation model includes:

[0035] deriving wheel-rail force data from the train-track multi-body dynamics model, and applying the wheel-rail force data to the rail surface of the track-roadbed finite element model to obtain rail surface lateral deformation data and rail surface longitudinal deformation data;

[0036] Applying the rail surface transverse deformation data and the rail surface longitudinal deformation data to the track structure in the train-track multi-body dynamics model to derive new wheel-rail force data, forming an iterative loop;

[0037] extracting mechanical response data after a preset cycle time is reached, and using the mechanical response data to eliminate the influence of initial instability in the train-track multi-body dynamics model and boundary conditions in the track-roadbed finite element model;

[0038] A numerical simulation model of train-track-roadbed coupled dynamics is established based on the train-track multi-body dynamic model that eliminates initial instability and the track-roadbed finite element model that eliminates the influence of boundary conditions.

[0039] Optionally, the wheel-rail force data includes vertical force and lateral force, and the rail surface lateral deformation data and the rail surface longitudinal deformation data are obtained by superimposing a random irregularity spectrum of a high-speed railway and vehicle-induced additional track irregularity.

[0040] Optionally, the step of substituting the train speed, operating time, initial moisture content of the roadbed, and number of wetting times into the train-track-roadbed coupled dynamic numerical simulation model to calculate the vertical deformation of the roadbed top surface includes:

[0041] By inputting parameters such as roadbed moisture content, number of wetting cycles, train speed, and prediction duration, we can obtain the joint simulation results of the train-track multi-body dynamics model and the track-roadbed finite element model. Based on the simulation results, we can draw time-history curves of the vertical deformation and settlement rate of the roadbed top surface.

[0042] In a second aspect, an embodiment of the present application provides a high-speed railway ballastless track subgrade long-term settlement prediction system, comprising a fitter and a predictor;

[0043] A fitter for fitting a curve to a function including the number of wetting times, the moisture content amplitude, the train speed, and the prediction time based on the time history data of the vertical deformation of the roadbed top surface obtained by the method of the first aspect;

[0044] The predictor is used to input the number of wetting times, moisture content amplitude, train speed, and prediction time, calculate and draw the time history curve of the vertical deformation of the roadbed top surface within the prediction time range, and compare it with the pre-set roadbed settlement threshold to provide an over-limit warning message.

[0045] Beneficial effects:

[0046] The method for predicting the long-term settlement of the ballastless track subgrade of a high-speed railway provided by the present invention establishes a more accurate empirical formula for predicting cumulative plastic strain by considering multiple factors such as the wetting characteristics, including the moisture content amplitude, the number of wetting times, and the dynamic stress amplitude. This makes it possible to be closer to the actual situation when predicting the settlement evolution law of the high-speed railway red clay subgrade during long-term operation, and significantly improves the accuracy of the prediction. At the same time, by establishing a train-track-subgrade coupled dynamic numerical simulation model, the model can fully consider the dynamic response of the subgrade under the dynamic load of the train. By performing joint simulation in a cyclic iterative manner, it is possible to more realistically simulate the actual situation of the train running on the high-speed railway line, and provide more reliable data support for settlement prediction;

[0047] The creep static equivalence theory was introduced into the calculation, and the cumulative deformation under long-term cyclic loads was regarded as static creep. This can more accurately simulate the mechanical behavior under long-term dynamic wetting, further improving the accuracy of settlement prediction. In the settlement evolution prediction stage, it not only provides time-series curves of top surface settlement and settlement rate under different combinations of roadbed wetting times and moisture content amplitudes, but also can realize rapid prediction and evaluation and early warning of long-term roadbed settlement, and can more comprehensively evaluate the impact of roadbed settlement on train operation performance. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 This is one of the flow charts of the method for predicting long-term settlement of high-speed railway ballastless track subgrade according to a preferred embodiment of the present invention;

[0049] Figure 2 This is the second flow chart of the method for predicting long-term settlement of high-speed railway ballastless track subgrade according to the preferred embodiment of the present invention;

[0050] Figure 3 This is a flow chart of numerical calculation of cumulative plastic deformation according to a preferred embodiment of the present invention;

[0051] Figure 4 A schematic diagram of a train-track model provided for a preferred embodiment of the present invention;

[0052] Figure 5 A schematic diagram of a track-roadbed model provided for a preferred embodiment of the present invention;

[0053] Figure 6 A flow chart of the joint simulation of the train-track-roadbed coupled dynamics model provided in a preferred embodiment of the present invention;

[0054] Figure 7 A schematic diagram showing the comparison between the predicted and measured values ​​of roadbed settlement provided in a preferred embodiment of the present invention. DETAILED DESCRIPTION

[0055] The following is a clear and complete description of the technical solutions of the present invention. It should be understood that the embodiments described are only a portion of the embodiments of the present invention, not all of them. All other embodiments derived by persons of ordinary skill in the art based on the embodiments of the present invention without inventive effort are intended to fall within the scope of protection of the present invention.

[0056] Unless otherwise defined, the technical or scientific terms used in the present invention shall have the usual meanings understood by persons of ordinary skill in the field to which the present invention belongs. The words "first", "second" and similar terms used in the present invention do not indicate any order, quantity or importance, but are only used to distinguish different components. Similarly, words such as "one" or "a" do not indicate a quantity limitation, but rather indicate the existence of at least one. Words such as "connected" or "connected" are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. "Up", "down", "left", "right" and the like are only used to indicate relative positional relationships. When the absolute position of the object being described changes, the relative positional relationship also changes accordingly.

[0057] See Figure 1 The present application provides a method for predicting long-term settlement of a high-speed railway ballastless track subgrade, comprising:

[0058] Collect target roadbed fillers and determine the relationship between the cumulative plastic strain of the roadbed fillers and the number of cyclic dynamic loads based on the moisture content, wetting times and dynamic stress characteristics of the target roadbed fillers;

[0059] Based on the relationship between the cumulative plastic strain of the dynamic roadbed filler and the number of cyclic dynamic loads, an empirical model for predicting the cumulative plastic strain is established, which takes into account the moisture content amplitude, the number of wetting times and the dynamic stress amplitude.

[0060] Numerical simulation of the long-term settlement of the target roadbed is performed by equating the number of cyclic loads to time units and the accumulated plastic strain to static creep. A track-roadbed finite element model is constructed based on the numerical simulation results.

[0061] Constructing a train-track multi-body dynamics model, and using iterative calculation to jointly simulate the track-roadbed finite element model and the train-track multi-body dynamics model to establish a train-track-roadbed coupled dynamics numerical simulation model;

[0062] The train speed, operating time, initial moisture content of the roadbed and the number of wetting times are substituted into the train-track-roadbed coupled dynamic numerical simulation model to calculate the vertical deformation of the roadbed top surface, and the long-term settlement evolution law of the target roadbed is predicted based on the vertical deformation of the roadbed top surface.

[0063] In the above embodiment, by incorporating the number of roadbed wetting times and the moisture content amplitude into the roadbed settlement prediction system, the problem of only considering the temperature effect but not the wetting effect in the existing settlement prediction is improved, thereby improving the accuracy of the prediction.

[0064] like Figure 2 As shown, the above embodiment can be further refined into the following steps:

[0065] Step 1: Prepare samples with moisture contents of 20.5%, 22.0%, 23.5%, and 25.0%. For each group of moisture contents, prepare several samples with humidification times n = 0, 1, 2, 3, and 4.

[0066] Step 2: Conduct a dynamic triaxial test using the following process: Dynamic stresses of 30 kPa, 50 kPa, 70 kPa, and 90 kPa were selected. A half-sine waveform was used for dynamic load cyclic loading, with a loading frequency f of 1 Hz, a loading time of 0.1 s, and an interval of 0.9 s. At the same time, considering the vertical load generated by the track structure overlying the subgrade soil, a static deviatoric stress of 0.5 times the confining pressure was applied after the confining pressure. Cumulative strain-time history curves for the specimens under a total of 80 test conditions were obtained, using five different numbers of cycles, four different moisture content amplitudes, and four dynamic stress amplitudes.

[0067] Step 3: Combined with the above dynamic triaxial test results, refer to the commonly used form of cumulative strain prediction model and use the water content ratio w / w OMC , dynamic stress amplitude σ d , wetting times n, and vibration times N are used as variables, and cumulative strain prediction models are established for stable and destructive types respectively.

[0068] For the stable prediction model, the relationship is as follows:

[0069] ;

[0070] For the destructive prediction model, the relationship is as follows:

[0071] ;

[0072] Where, represents the predicted accumulated plastic strain, represents the soil moisture content, represents the optimal moisture content of the soil, represents the dynamic stress amplitude, Indicates the local atmospheric pressure, Indicates the number of humidification cycles, represents the number of dynamic load actions, k1, k2, k3, k4, k5, and b are fitting coefficients, and the fitting coefficients are obtained by nonlinear least squares curve fitting using the Levenberg-Marquardt algorithm.

[0073] Step 4: Use the Levenberg-Marquardt algorithm to perform nonlinear least squares curve fitting, and obtain the fitting coefficients as shown in Table 1 below:

[0074] Table 1: Fitting coefficients under different model types

[0075]

[0076] Step 5: Incorporating creep static equivalence theory, the accumulated deformation under long-term cyclic loading is considered static creep. The number of cyclic loading events is equated to a time unit. Numerical simulation of long-term subgrade settlement is performed in ABAQUS. The above prediction model is written as a UMAT subroutine. The soil strain under cyclic dynamic loading is divided into three components: elastic strain, plastic strain, and creep, and these components are calculated separately.

[0077] Furthermore, the critical state theory is used to calculate the time-independent elastic-plastic deformation. For the elastic deformation part, the stiffness constant is:

[0078] ;

[0079] ;

[0080] Where K and G are the elastic bulk compression modulus and shear modulus respectively, E is the elastic modulus, μ is the Poisson's ratio, and e is the soil porosity. is the slope of the rebound line in the e-lnp space, is the effective mean stress.

[0081] For plastic deformation, the yield criterion and associated flow law following the modified Cambridge model are adopted. The plastic potential function is the same as the yield surface function, that is:

[0082] ;

[0083] in, is the slope of the critical state line, is the deviatoric stress, and here the formula is , J2 is the second invariant of the deviatoric stress tensor, is the initial consolidation pressure of the soil, is the effective mean stress.

[0084] Furthermore, considering that the development pattern of the accumulated residual deformation of soil under long-term cyclic loading is similar to the creep deformation of soil under static load, the number of cyclic loading actions is equivalent to the unit of time. The creep potential function in creep calculation theory is introduced. Combined with the empirical model of cyclic accumulated deviatoric strain of soil under undrained conditions, a simplified calculation method that can describe the cyclic accumulated deformation of soil is established. The time-dependent cyclic accumulated strain rate tensor can be expressed as:

[0085] ;

[0086] in, is an empirical expression for the axial cyclic cumulative strain rate of soil samples under undrained triaxial conditions based on experimental or measured data. is the creep potential function, is the rate-dependent cyclic cumulative strain rate tensor, is the dynamic stress level, is the total strain rate, is the dynamic load loading time, are other internal variables, is the effective mean stress, is the symbol of partial derivative.

[0087] With the help of the user subroutine interface provided by the large commercial software Abaqus, the above algorithm is written into the UMAT subroutine. The numerical simulation workflow is as follows Figure 3 As shown in Figure 2, the input parameters include soil wetting times and moisture content amplitude.

[0088] Step 6: Establish a vehicle-track multi-body dynamics simulation model based on SIMPACK software. The physical and mechanical parameters of the vehicle body are based on the CRH380A EMU train, such as Figure 4 As shown, taking this embodiment as an example, the vehicle-track multi-body dynamics simulation model established adopts a CRH380A EMU train body, which has a length of 25000 mm, a body height of 3700 mm, a wheelset spacing of 2500 mm, a wheelset spacing of 17500 mm, a body center of gravity height of 1520 mm, and a bogie center of gravity height of 510 mm.

[0089] Step 7: Taking the high-speed railway CRTS-I type double-block ballastless track as an example, a track-roadbed finite element simulation model is established based on Abaqus software. The model is 65m long in the longitudinal direction, 30m wide and 10m high in the foundation. The cross-sectional dimensions are drawn according to the standard design drawing of the CRTS-I type double-block ballastless track roadbed section. The roadbed side slope is 1:1.5. Figure 5 As shown in Figure 1, the simulation model includes the track structure, roadbed structure, and intermediate fill area. The track structure consists of 60 kg / m rails 1, WG-I sleepers 2, track slabs 3, and a supporting layer 4. The roadbed structure comprises asphalt concrete 5, a subgrade surface layer 6, a subgrade bottom layer 7, a subgrade body 8, and foundation soil 9. The intermediate fill area is composed of C25 concrete 10 and graded crushed stone 11. The material property parameters of each component are based on measured values ​​from a section of the Wuhan-Guangzhou High-Speed ​​Railway. The subgrade uses a custom red clay material and is associated with the UMAT subroutine.

[0090] Step 8: Use Figure 6 The iterative simulation scheme shown co-simulates the vehicle-track multibody dynamics simulation model and the track-roadbed finite element simulation model.

[0091] In a single cycle, the wheel-rail force data derived from SIMPACK is applied to the rail surface of the ABAQUS model, and the transverse and longitudinal deformation data of the rail surface derived from ABAQUS are applied to the track structure of the SIMPACK model. The data migration is performed using a sampling (from SIMPACK simulation results) and repetition (longitudinal rail surface deformation data) approach.

[0092] The mechanical response data are extracted starting from the 10th second of the train-track model output to eliminate the influence of the initial instability of the train model and the boundary conditions of the track-subgrade finite element model.

[0093] The average wheel-rail force during the train passing period is used as the equivalent static load and applied at the longitudinal midpoint of the track-roadbed model;

[0094] The wheel-rail interaction force includes vertical force and lateral force, and the rail surface deformation is the superposition of the random irregularity spectrum of high-speed railway and the vehicle-induced additional track irregularity.

[0095] By simulating 1500 trains passing in one direction at a speed of 250 km / h, the cumulative plastic strain time history curve of the roadbed top surface on the track centerline after the wheelset cyclic load is applied 48,000 times is obtained, as shown in Figure 7 As shown in the figure, the cyclic iteration method simulates the cumulative deformation of the roadbed top surface after 1,500 trains have passed, resulting in a value of 1.15 mm, 2.22% smaller than the measured value. The traditional method, which does not consider cyclic iteration, achieves a value of 0.98 mm, 16.75% smaller than the measured value. This demonstrates that the static creep equivalent model and cyclic iteration method proposed in this paper can effectively simulate the development of roadbed settlement.

[0096] Step 9: Change the number of wetting times, moisture content amplitude, and train speed parameters input to the model to obtain settlement prediction results including but not limited to the settlement of the roadbed top surface and the settlement rate time history curve.

[0097] Table 2 shows the simulation results of the settlement evolution of the roadbed top surface when the red clay of the roadbed is at a moisture content of 20.5% and 10,000 pairs of trains pass through in both directions at different wetting times.

[0098] Table 2: Simulation results of the evolution of subgrade top surface settlement under different wetting times

[0099]

[0100] Table 3 shows the simulation results of the evolution of the subgrade top surface settlement when the subgrade filler material has not undergone cyclic wetting and the EMU passes 10,000 trains at 350 km / h under different soil moisture contents.

[0101] Table 3: Simulation results of subgrade top surface settlement evolution under different soil moisture contents

[0102]

[0103] The embodiment of the present invention further provides a high-speed railway ballastless track subgrade long-term settlement prediction system, comprising a fitter and a predictor;

[0104] A fitter for fitting a curve into a function including the number of wetting times, the moisture content amplitude, the train speed, and the prediction time based on the time history data of the vertical deformation of the roadbed top surface obtained from the long-term settlement prediction method of the high-speed railway ballastless track roadbed;

[0105] The predictor is used to input parameters such as the number of wetting times, moisture content amplitude, train speed, and prediction time, calculate and draw a time-history curve of the vertical deformation of the roadbed top surface within the prediction time range, and compare it with the pre-set roadbed settlement threshold to provide an over-limit warning message.

[0106] The above describes in detail the preferred embodiments of the present invention. It should be understood that those skilled in the art can make numerous modifications and variations based on the concepts of the present invention without inventive effort. Therefore, any technical solutions that can be derived by those skilled in the art through logical analysis, reasoning, or limited experimentation based on the concepts of the present invention and the prior art should be within the scope of protection defined by the claims.

Claims

1. A method for predicting long-term settlement of high-speed railway ballastless track subgrade, characterized in that: include: Collect target roadbed fillers and determine the relationship between the cumulative plastic strain of the roadbed fillers and the number of cyclic dynamic loads based on the moisture content, wetting times and dynamic stress characteristics of the target roadbed fillers; Based on the relationship data between the cumulative plastic strain of roadbed filler and the number of cyclic dynamic loads, an empirical model for predicting cumulative plastic strain considering the moisture content amplitude, wetting times and dynamic stress amplitude is established. Numerical simulation of the long-term settlement of the target roadbed is performed by equating the number of cyclic loads to time units and the accumulated plastic strain to static creep. Based on the results of the numerical simulation, a track-roadbed finite element model is constructed. Constructing a train-track multi-body dynamics model, and using an iterative calculation method to jointly simulate the track-roadbed finite element model and the train-track multi-body dynamics model to establish a train-track-roadbed coupled dynamics numerical simulation model; Substituting the train speed, operating time, initial moisture content of the roadbed, and the number of wetting cycles into the train-track-roadbed coupled dynamic numerical simulation model to calculate the vertical deformation of the roadbed top surface, and predicting the long-term settlement evolution law of the target roadbed based on the vertical deformation of the roadbed top surface; The cumulative plastic strain prediction empirical formula considering the moisture content amplitude, wetting times and dynamic stress amplitude is established based on the relationship data between the cumulative plastic strain of the roadbed filler and the number of cyclic dynamic loads, including: Based on the relationship between the cumulative plastic strain of roadbed filler and the number of cyclic dynamic loads, empirical models for predicting cumulative plastic strain were constructed for stable and destructive roadbeds, respectively, with water content ratio, dynamic stress amplitude, number of wetting times, and number of vibrations as variables. For the empirical model for predicting the cumulative plastic strain of a stable roadbed, the empirical model for predicting the cumulative plastic strain of a stable roadbed satisfies the following relationship: ； For the empirical model for predicting the cumulative plastic strain of a damaged roadbed, the empirical model for predicting the cumulative plastic strain of a damaged roadbed satisfies the following relationship: ； Where, represents the predicted accumulated plastic strain, represents the soil moisture content, represents the optimal moisture content of the soil, represents the dynamic stress amplitude, Indicates the local atmospheric pressure, Indicates the number of humidification cycles, represents the number of dynamic load actions, k1, k2, k3, k4, k5, and b are fitting coefficients, and the fitting coefficients are obtained by nonlinear least squares curve fitting using the Levenberg-Marquardt algorithm.

2. The method for predicting long-term settlement of high-speed railway ballastless track subgrade according to claim 1, characterized in that: The relationship data between the cumulative plastic strain of the roadbed filler and the number of cyclic dynamic loads is determined based on the moisture content, wetting times and dynamic stress characteristics of the target roadbed filler, including: Based on the moisture content, wetting times and dynamic stress characteristics of the target roadbed filler, a dynamic triaxial test is conducted to obtain the relationship curve between the cumulative plastic strain of the target roadbed filler and the number of cyclic dynamic loads. The relationship curve between the cumulative plastic strain and the number of cyclic dynamic loads is used as the relationship data between the cumulative plastic strain and the number of cyclic dynamic loads of the target roadbed filler.

3. The method for predicting long-term settlement of high-speed railway ballastless track subgrade according to claim 1, characterized in that: The numerical simulation of the long-term settlement of the target roadbed by equating the number of cyclic load actions to a time unit and the accumulated plastic strain to static creep includes: The numerical simulation of the long-term settlement of the target roadbed is equivalent to the calculation of soil strain. The number of cyclic loads is equivalent to the time unit, and the accumulated plastic strain is equivalent to static creep to calculate the soil strain. The calculation of soil strain includes elastic strain calculation, plastic strain calculation and creep calculation. The critical state theory is used to calculate the time-independent elastic strain. The stiffness constant expressed as modulus in the elastic strain calculation is: ； ； Where K and G are the elastic bulk compression modulus and shear modulus respectively, E is the elastic modulus, μ is the Poisson's ratio, and e is the soil porosity. is the slope of the rebound line in the e-lnp space, is the effective mean stress; The plastic strain is calculated by following the yield criterion of the modified Cambridge model and the associated flow law. Under the premise that the plastic potential function is the same as the yield surface function, the calculation of the plastic strain satisfies the following relationship: ； Where, is the slope of the critical state line, is the deviatoric stress, and here in the calculation formula , J2 is the second invariant of the deviatoric stress tensor, is the initial consolidation pressure of the soil, is the effective mean stress; The calculation for creep satisfies the following relationship: ； Where, is an empirical expression for the axial cyclic cumulative strain rate of soil samples under undrained triaxial conditions based on experimental or measured data. is the creep potential function, is the rate-dependent cyclic cumulative strain rate tensor, is the dynamic stress level, is the total strain rate, is the dynamic load loading time, are other internal variables, is the effective mean stress, is the symbol of partial derivative.

4. The method for predicting long-term settlement of high-speed railway ballastless track subgrade according to claim 1, characterized in that: The method of using an iterative calculation method to jointly simulate the track-roadbed finite element model and the train-track multi-body dynamics model to establish a train-track-roadbed coupled dynamics numerical simulation model includes: deriving wheel-rail force data from the train-track multi-body dynamics model, and applying the wheel-rail force data to the rail surface of the track-roadbed finite element model to obtain rail surface lateral deformation data and rail surface longitudinal deformation data; Applying the rail surface transverse deformation data and the rail surface longitudinal deformation data to the track structure in the train-track multi-body dynamics model to derive new wheel-rail force data, forming an iterative loop; extracting mechanical response data after a preset cycle time is reached, and using the mechanical response data to eliminate the influence of initial instability in the train-track multi-body dynamics model and boundary conditions in the track-roadbed finite element model; A numerical simulation model of train-track-roadbed coupled dynamics is established based on the train-track multi-body dynamic model that eliminates initial instability and the track-roadbed finite element model that eliminates the influence of boundary conditions.

5. The method for predicting long-term settlement of high-speed railway ballastless track subgrade according to claim 4, characterized in that: The wheel-rail force data includes vertical force and lateral force. The rail surface lateral deformation data and the rail surface longitudinal deformation data are obtained by superimposing the high-speed railway random irregularity spectrum and the vehicle-induced track additional irregularity.

6. The method for predicting long-term settlement of high-speed railway ballastless track subgrade according to claim 1, characterized in that: Substituting the train speed, operating time, initial moisture content of the roadbed and the number of wetting times into the train-track-roadbed coupled dynamic numerical simulation model to calculate the vertical deformation of the roadbed top surface includes: By inputting the roadbed moisture content, number of wetting cycles, train speed, and predicted duration, we can obtain the joint simulation results of the train-track multi-body dynamics model and the track-roadbed finite element model. Based on the simulation results, we can draw time-history curves of the vertical deformation and settlement rate of the roadbed top surface.

7. A high-speed railway ballastless track long-term settlement prediction system, characterized in that: Includes fitters and predictors; A fitting device for fitting a curve to a function including the number of wetting times, the moisture content amplitude, the train speed, and the prediction time based on the time history data of the vertical deformation of the roadbed top surface obtained by the method according to any one of claims 1 to 6; The predictor is used to input the number of wetting times, moisture content amplitude, train speed, and prediction time, calculate and draw the time history curve of the vertical deformation of the roadbed top surface within the prediction time range, and compare it with the pre-set roadbed settlement threshold to provide an over-limit warning message.

Images