The foundation conditions of piers for multi-span long-distance heavy-haul railway bridges inevitably vary at different locations, which may lead to non-uniform ground motions at each pier position, potentially causing adverse effects on the bridge's seismic response. To investigate the seismic response of bridges and the running safety of heavy-haul trains as they cross the bridge during an earthquake, a three-dimensional heavy-haul railway train-track-bridge (HRTTB) coupled system model was developed using ANSYS/LS-DYNA. This model incorporates the nonlinear behavior of critical components such as bearings, lateral restrainers, piers, and wheel-rail contact interactions, and it has been validated against field-measured data to ensure reliable dynamics parameters for seismic analysis. A multi-span simply supported girder bridge from a heavy-haul railway (HHR) was employed as a case study, in which a spatially correlated non-stationary ground motion field was generated based on spectral representation harmonic theory. Comparative analyses of the seismic responses under spatially varying ground motions (SVGM) and uniform seismic excitation conditions were performed for the coupled system. The results indicate that the presence of heavy-haul trains prolongs the natural period of the HRTTB system, thereby appreciably altering its seismic response. At lower apparent wave velocities, more piers exhibit a low-response state, and some pier bases enter the elastic-plastic stage under local site effects. Compared with the piers, the bearings show higher sensitivity to seismic inputs; fixed bearings experience damage when subjected to traveling wave effects and local site effects, which is subsequently followed by the failure of lateral restrainers. Train running safety is markedly reduced when crossing local soft soil site conditions. The conclusions drawn from this study can be applied in the seismic design and running safety assessment of HHR bridge systems under SVGM.
Research on the dynamic response of subgrades is essential for designing heavy-haul railway subgrades. Therefore, a dynamic stress field test was carried out on the Daqin Railway using a three-dimensional dynamic soil pressure box capable of measuring the total stress component of soil elements. Then, a train-track-subgrade coupling finite-element model (FEM) considering the track irregularity and infinite element boundary conditions was established, and the validity of the model was verified using field test results. Subsequently, based on the field test results, the actual three-dimensional dynamic response and stress path of the subgrade under a train load were analyzed. Based on the FEM results, the effects of the train axle load, train speed, subgrade stiffness, and subgrade thickness on the three-dimensional dynamic response of the subgrade were analyzed, and a prediction model of the vertical dynamic stress was proposed. Finally, the influence of the depth of the heavy-haul train loads on the subgrade was studied. Research has shown that the normal stress caused by two wheelsets under the same bogie has a superposition effect, and each peak value of the normal stress corresponds to the center position of the bogie. When the train passes through the test section, the stress path of the soil element directly below the track is fairly elliptical, and the principal stress axis of the soil element rotates by 180 degrees. The normal stresses sigma x, sigma y, and sigma z increase proportionally with the speed and axle load of the train but decrease inversely proportional to the thickness of the ballast layer. The subgrade stiffness significantly influences the normal stress sigma x and sigma y but has no apparent influence on the normal stress sigma z. The influence depth of the train load in the subgrade is related to the axle load, train speed, and thickness of the ballast layer, but is unrelated to the stiffness of the subgrade surface layer. This study provides practical and theoretical data for analyzing the dynamic performance of heavy-haul railway subgrades.