Fouling and mud-pumping problems in ballasted track significantly degrade serviceability and jeopardize train operational safety. The phenomenological approaches for post hoc forensic investigation and remedies of mud pumps have relatively been well studied, but there still lacks studies on inherent mechanisms and ex ante approaches for early-age detection of mud pumps. This paper was aimed to exploring the feasibility of using particle acceleration responses to diagnose and identify early-age mud-pumping risks in real-world field applications. The innovative wireless sensors with 3D-printed shells resembling real shape of ballast particles were instrumented in the problematic railway section to monitor ballast particle movement prior to, during, and after maintenance operations, respectively. The real-time particle-scale acceleration data of ballast bed under both degraded and maintenance-restored clean conditions were recorded. The time histories, power spectra, and marginal spectra of 3D acceleration were comparatively analyzed. The results showed the 3D acceleration of ballast particles underneath rail-supporting tie plates displayed relatively clear periodicity of about 0.8 s with adjacent bogies regarded as a loading unit. The tamping operation was effective for compacting ballast bed laterally and improving the lateral interlocking of ballast particles, whereas the stabilizing operation was effective mainly in the lateral direction and for ballast particles underneath the sleepers. The mud pumps caused intensive particle-scale acceleration, and ballast particles underneath the sleepers were affected more severely than those in between adjacent sleepers. The ballast particles directly underneath tie plates exhibit dramatic acceleration variations due to maintenance operations as compared to those in other positions studied; hence, it seems promising to use particle-scale acceleration underneath tie plates as readily-implementable indicators for smart in-service track health monitoring.

The majority of existing regression models for unbound granular materials (UGMs) consider only the effects of the number of loading cycles and stress levels on the permanent deformation characteristics and are thus unable to account for the complexity of plastic deformation accumulation behavior influenced by other factors, such as dry density, moisture content and gradation. In this study, research efforts were made to develop artificial-neural-network (ANN)-based prediction models for the permanent deformation of UGMs. A series of laboratory repeated load triaxial tests were conducted on UGM specimens with varying gradations to simulate realistic stress paths exerted by moving wheel loads and study permanent deformation characteristics. On the basis of the laboratory testing database, the ANN prediction models were established. Parametric sensitivity analyses were then performed to evaluate and rank the relative importance of each factor on permanent deformation behavior. The results indicated that the developed ANN prediction model is more accurate and reliable as compared to previously published regression models. The two major factors influencing the magnitude of accumulated plastic deformation of UGMs are the shear stress ratio (SSR) and the number of loading cycles, of which the calculated influence coefficients are 38% and 41%, respectively, while the degree of influence of gradation is twice that of the confining pressure.

Unbound permeable aggregate base (UPAB) materials with strong load-transmitting skeleton yet adequate inter-connected pores are desired for use in the sponge-city initiative. However, the micro-scale fabric evolution and instability mechanism of macroscopic strength behavior of such UPAB materials still remain unclear. In this study, virtual monotonic triaxial compression tests were conducted by using the discrete element method (DEM) modeling approach on specimens with different gradations quantified by the parameter of gravel-to-sand ratio (G/S). The realistic aggregate particle shape and inter-particle contact behavior were properly considered in the DEM model. The micromechanical mechanisms of the shearing failure of such UPAB materials and their evolution characteristics with G/S values were disclosed from contact force chains, microstructures, and particle motion. It was found that the proportion of rotating particles in the specimens decreased and the proportion of relative sliding between particles increased as the content of fine particles decreased. The plastic yielding of the specimens originated from the failure of contact force chains and the occurrence of the relative motion between particles, while the final instability was manifested by the large-scale relative motion among particles along the failure plane (i.e., changes in the internal particle topology). By comparing the macroscopic strength, microstructure evolution, and particle motion characteristics of the specimens with different G/S values, it was found that the specimens with G/S value of 1.8 performed the best, and that the G/S value of 1.8 could be regarded as the threshold for separating floating dense and skeletal gap type packing structures. The variation of Euler angles of rotating particles was significantly reduced in the particle size range of 4.75 mm to 9.50 mm, indicating that this size range separates most of the particles from rolling and sliding. Since particle rolling and sliding behavior are directly related to shear strength, this validates the rationality of the parameter G/S for controlling and optimizing gradations from the perspective of particle movement. The findings could provide theoretical basis and technical guidance for the effective design and efficient utilization of UPAB materials.