Mud pumping is an undesirable subgrade distress in ballastless high-speed railway, significantly affecting the ride comfort and posing a threat to train operation safety. In this study, a full-scale physical model of the ballastless slab track was developed. A rainfall simulator was installed, and various testing sensors were embedded in the trackbed to investigate the phenomenon of mud pumping in ballastless tracks. The results revealed a three-stage process for the intruded rainwater, including the initial vertical infiltration, the following horizontal infiltration, and the eventual roadbed saturation. A significant excess pore water pressure gradient (PWP) was created vertically in the roadbed due to the moving train loads. Similarly, a small longitudinal gradient was also observed. Both PWP gradients indicated the spatial migration of fine particles within the roadbed. The contact pressure distribution under the concrete base varied notably under different roadbed conditions. In the saturated state, the maximum dynamic soil stress, initially located at side of the concrete base transitioned toward the track center. The findings contribute to a deeper understanding of mud pumping mechanism in ballastless tracks.

In this paper, a dynamic model is established to investigate the dynamic behavior of train-tracksubgrade-soil (TTSS) interaction. The effects of interfacial damage of the track slab induced by soil settlement on the dynamic interaction system are considered. The model framework is established by the finite element method. The soil settlement-induced track deformation is calculated by a practical iteration algorithm, where the nonlinear interfacial damage is simulated by the cohesive zone model. The simulation model is verified by comparison with other models. In regards to the excitations of the dynamic TTSS interaction system, two types of track irregularities are considered, namely the conventional track irregularities generated by known spectrums, and the additional irregularities caused by soil settlement. In the numerical study, the dynamic performances of the TTSS interaction subjected to interfacial damage, and soil settlement are compared. Next, the short wavelength irregularity is discussed as well. From the results, the vibration enhancement can be observed in the time and frequency domain. The interfacial damage of the track enhances the vibration both in low- and high-frequency domains, while the impacts of settlement are only observed in the frequency band of 0 similar to 3 Hz. The frequency band of vibrations triggered by short wavelength irregularities is correlated with its wavelength range. Moreover, the settlement with different wavelengths and amplitudes is studied. It is shown that the increase of settlement amplitude and decrease of settlement wavelength lead to higher damage degree in amplitude and wider spatial distribution. In regards to the dynamic responses, the vehicle accelerations, wheel-rail contact forces, track displacement, and soil displacement are more sensitive to the settlement amplitudes varying from 10 to 80 mm, while the sensitive settlement wavelength is concentrated in 20 to 40 m.

Recent field case study shows that the roadbed of ballastless high-speed railway experienced water-induced defect such as excessive fines pumping and even local subgrade-track contact loss affecting the normal operation of high-speed train due to water immersion through gaps of waterproof materials in expansion joints between the concrete base, particularly in rainy seasons. However, the study about the dynamic behavior of high-speed railway subgrade involving water is currently rare. Based on the theory of fluid dynamics in porous medium and the vehicle-track coupling vibration theory, a numerical method of hydraulic-dynamic coupling was established to evaluate the dynamic responses of saturated roadbed surface layer under the high-speed train loading with the validation by comparing the calculated values and field data. The temporal and spatial characteristics of dynamic behaviors (stress, pore water pressure, seepage velocity, displacement) of saturated roadbed surface layer are fully discussed. Also, the effects of train velocity, permeability, on aforementioned dynamic responses of the saturated roadbed surface layer are evaluated. The study shows that improving the drainage of ballastless track roadbed has a significant effect on minimizing the mud pumping of ballastless track, and the influence zone of hydraulic-mechanical coupling is mainly within 0.1 m of the roadbed.