The method of detecting ballast bed defects using ground penetrating radar (GPR) is an important method for guiding the maintenance of railway infrastructure. Currently, this technology primarily relies on time–frequency analysis to assess the condition of the ballast bed and manual interpretation of GPR images to identify defect areas and types, resulting in low automation levels. This paper proposes a bimodal deep learning classification model that enables intelligent classification of moisture and mud pumping defects in ballast beds. This model includes two channels, each processing a different data modality. One channel uses a Multilayer Perceptron (MLP) to extract features of A-scan data in the time domain. The other channel utilizes Short-Time Fourier Transform (STFT) to convert time domain signals into frequency domain signals, which are then processed by a ResNet18 to extract frequency domain features. By fusing the time and frequency features, the proposed Time-Frequency-Fusion ResNet model (TFF-ResNet) demonstrates superior performance. Experimental results show that TFF-ResNet outperforms the standalone MLP and ResNet18 models, with performance improvements of approximately 24% and 14% on the validation dataset, and 21% and 34% on the testing dataset, respectively.

China’s railway covers a broad area, the geological environment and climate along the line are complex and variable, the operation and maintenance of railway are facing tremendous challenges. As an integral part of railway, the maintenance of ballast bed has always attracted considerable attention, nevertheless, the variety and uneven distribution of rock materials, as well as the single ballast material standard and selection in China, which don’t take into consideration factors such as geological conditions and climatic environment, bring a host of issues to railway construction and operation and maintenance. To deal with this problem, this paper summarizes railway ballast selection methods and methodology around the world, compares the ballast material selection criteria for complex environments, summarizes novel ballast materials, and explains the indicators and test approaches used to quantify ballast performance in the standards. Comparing the ballast materials and their matching selection standards in various countries, the main conclusions of this paper are as follows: (1) For the existing ballast specification problems in China, the ballast materials can be considered to be selected according to the geological and climatic factors along the railway line; (2) There is still no method to accurately and rapidly measure the density of ballast bed without damaging it. (3) To achieve the goal of double carbon, novel ballast materials such as construction solid waste and industrial solid waste can be considered on new or modified lines where conditions are favorable.

The dynamic response of a railway bridge can be affected by the properties of its foundations, particularly if founded on soft soils. Thus, this work aims to establish a coupled dynamic model to investigate the vibration of train-track-bridge systems considering piled foundations embedded in soft soil. Firstly, to construct the simulation framework, the finite element and multi-body methods are used to model the dynamic behavior of a train-track-bridge interaction (TTBI) system and a pier-cap-pile-soil interaction system. The equilibrium of the two sub-systems is maintained through the bridge's bearing force and a multi-time-step integration strategy is introduced to improve computational efficiency. The proposed model is validated by comparing it to the results from commercial finite element software ABAQUS. Then results are computed using the proposed model and the conventional TTBI model without piles. It is concluded that when considering the piled foundations, the low-frequency vibration of the TTBI system is dominant. Moreover, the vibration energy in the track and bridge below 7 Hz is larger compared with the conventional TTBI model. The influence of train speed on the vibration characteristics of the pile and soil is analyzed. It is found that higher train speeds cause increased pile and soil accelerations at the frequencies associated with the train axle spacing. The novelty of the analysis is providing a new insight into the coupled vibration properties of TTBI systems considering the participation of piled foundations in soft soil.

Railway track health monitoring and maintenance are crucial stages in railway asset management, aiming to enhance the train operation quality and service life. For this aim, various inspection means (using diverse non-destructive testing techniques) have been applied, however, these means are mostly not able to monitor whole railway track network or track underlying layers (e.g., ballast and subgrade). The use of remote sensing techniques, such as Interferometric Synthetic Aperture Radar (InSAR), can expedite the defect diagnosis process for railway tracks, elevating the scope of health monitoring to a network-wide level. The Ground Penetrating Radar (GPR) has emerged as a particularly reliable method, especially for detecting structural deficiencies in underlying layers. As a result, combining the two distinct non-destructive testing techniques – GPR and InSAR – presents a promising strategy for efficient railway asset management. Recognizing the significance of embracing newer and more advanced monitoring strategies, this paper reviews the fusion of GPR and InSAR methodologies, and explores the potential integration of machine learning models to develop a predictive health monitoring and condition-based maintenance approach for railway tracks.

The settlement of piers and subgrade bending deformation are widely recognized as common issues in the transition zones of high-speed railway bridges. This study aims to investigate the settlement behavior within these transition zones and its impact on the dynamic interaction between trains and the track. To achieve this, a vehicle-track-transition zone mapping relationship model is developed to analyze both the settlement behavior and the resulting dynamic response characteristics. The study employs the finite element method and multi-body dynamics to construct the simulation model. Settlement effects are simulated using the Newton-Raphson iterative method, with the additional rail deformation caused by foundation settlement serving as the excitation for the vehicle-track-transition zone dynamic interaction system. In the numerical analysis, the dynamic effects of three key factors—train speed, transition zone length, and the amplitude of foundation settlement—are examined based on the performance of the vehicle-track-transition zone interaction. The time-frequency technique is utilized to comprehensively reveal and clarify the spatial-frequency characteristics of system responses influenced by settlement excitation. Moreover, the relationship between the safety-based settlement threshold and these three factors is calibrated.

This study explores the use of infrared thermography (IRT) technology for the non-destructive evaluation of ballast fouling in railway tracks, focusing on the influence of parent rock types and fouling materials. Utilizing thermal imaging, the research investigates how variations in ballast conditions affect surface temperature, which serves as an indicator of structural integrity and health. The experimental setup involved ballast samples derived from three different rock types—basalt, limestone, and andesite—fouled with commonly encountered materials like sand and clay at varying percentages. Results demonstrate that fouling level and type significantly influence the thermal signatures captured by IRT passive camera. Notably, ballast derived from darker rocks exhibited higher temperatures, indicating greater emissivity, while fouled ballast showed distinct temperature patterns compared to clean samples, emphasizing the potential of thermal imaging in detecting and quantifying fouling in ballast layers. This research underscores the viability of IRT passive camera in the routine maintenance and monitoring of railway infrastructure, providing a foundation for further development of integrated diagnostic tools for railway management systems.

Building Information Modelling (BIM) is a digitalisation tool that is widely adopted in construction industry. It is a three-dimensional digital replica of asset(s) such as buildings, which contain architectural information and building details (e.g. dimensions, materials, parts, and components). It has evolved from 2D CAD models (or blueprints) in the past to 3D CAD models embedded with information layers (e.g., construction time sequence or 4D-BIM), resulting in automation in construction. BIM has now been essential in various countries; for example, new UK BIM standards require asset owners to keep and maintain building information. BIM adopts an interoperable concept that can benefit the whole life-cycle assessment (LCA) and circularity of the built environments. Its applications extend to six dimensions (6D) where time sequence, cost and carbon footprint can now be reported in real time. These attributes are essential to stakeholders and critically help reduce any unexpected consumption and waste over the life cycle of a project. This study builds on the development of 6D BIM of an existing building to enrich circular value chains and stakeholder engagement. This paper highlights the development of 6D BIM, and, subsequently, the stakeholder interviews to address challenges, barriers, benefits, and effectiveness of 6D-BIM applications for stakeholder engagements across circular value chains. Snowballing sampling method has been used to identify stakeholder interviews to obtain new insights into the digital valorisation for stakeholder engagement. The outcome of this study will exhibit new insights and practical paradigms for BIM applications in built environments.

The occurrence of ballast contamination or fouling frequently results in a sudden decline in the capacity of railway ballasted tracks. Considering the various sources of ballast fouling, clay is the most severe one for causing a drastic reduction in the drainage capacity of the ballast layer. In the current study, we utilized a large-scale flume test to measure the water height along the cross-section of the clay-fouled ballast. Subsequently, an analytical–numerical (A-N) approach was developed to simulate the movement of water through porous media under steady-state conditions, while also considering the flow regime. This A-N approach was validated using the results of flume tests. Finally, the validated A-N approach was employed to generate a dataset and develop machine learning models for predicting water height. The characterized machine learning models included random forest regression (RFR), support vector machine (SVM), and extreme gradient boosting (XGBoost). Various variables, such as ballast gradation, fouling ratio, bed slope, rainfall rate, and water height on the side ditch, were incorporated into the machine learning models to reveal the contribution of each individual variable. Results show that for clean ballast, the incorporation of a nonlinear model between flow velocity and hydraulic gradient in the A-N approach is crucial to properly estimate the experimental measurements. However, a comparison of the water height measured via the flume test and the water level estimated based on the A-N approach confirms the suitability of the linear model, i.e., Darcy's law, for the water flow regime through clay-fouled ballast. According to the machine learning results, particularly those from the XGBoost model, which was characterized as the elite model, the rainfall rate and the fouling index emerged as the most influential variables affecting the water height in the clay-fouled ballast layer of the railway track.

In practice, the assessment and treatment of rail corrugation are quantitatively based on the corrugation depth. Wheel–rail vertical forces (WRVF), as a direct reflection of wheel–rail interaction, can give expression to the corrugation depth and thus serve as a key parameter for assessing the corrugation. In this paper, we propose an evaluation method for rail corrugation based on the WRVF. First, a 3D wheel–rail dynamic finite element (FE) model was developed with typical parameters of CRTS II slab track and CRH3 vehicle for high-speed railways in China. The accuracy of the model was then validated with the measured WRVF data in the field. Second, using the validated model, the time–frequency domain distribution of WRVF (vehicle speed: 300 km/h) was obtained with consideration of the corrugation wavelength in the range of 40–180 mm. The non-linear least squares method and rational equation were used to fit the function between the large value of WRVF and the corrugation depth value under the conditions of different corrugation wavelengths. Next, effects of the Pinned–Pinned resonance frequency and vibration mode on the fitted parameters were analysed, by which an indicator for corrugation treatment (grinding) was proposed. Finally, the indicator was applied in the monitoring of rail corrugation for high-speed railway lines in the field. The results show that the misjudgement rate of rail grinding decisions (using the proposed indicator) is low with the accuracy at 92.5%. The proposed method can provide a basis for the rail corrugation evaluation and grinding decisions-making.

Differential railway track settlement can result in ballast voids, leading to sleepers that hang from the rail and are no longer supported by the ballast. These hanging sleepers are damage for track component. As a solution, this paper proposes and investigates a new concept sleeper with a wedge-shaped geometry, intended to stimulate the migration of ballast into any voids, thus reducing the occurrence of hanging sleepers. A series of scaled laboratory tests and 2D and 3D discrete element simulations are used to investigate different wedge-shaped geometries. The investigations include the wedge type (single long wedge versus multiple mini-wedges) and the wedge angle (30, 45, 60 degrees). First, the scaled laboratory tests are used to study the performance of different wedge geometries. Next, 3D DEM simulations are performed to analyse the contact forces in the ballast due to different wedge designs. Finally, 2D DEM simulations are performed to study the settlement behaviour. The main conclusions are that a single long wedge is preferable compared to multiple smaller wedges. when the wedge sleeper angle is larger than the ballast's angle of repose, particles have the freedom to migrate into the settlement induced voids. Also, an increased wedge sleeper angle stimulates greater particle migration and thus improves the support correction. However the longer wedge also leads to a decrease in effective ballast height under sleeper which may make retrofitting on existing lines challenging.

The railway ballast layer provides the function of bearing loading, resisting geometry degradation, and drainage. In those related research, the behaviour of ballast assembly can be obtained by laboratory (or in-situ) tests. Limited simulation methods can be used to analyse the behaviour of ballast particles at the mesoscopic level. The numerical simulations based on the Discrete Element Method (DEM) are employed, which treat every ballast particle as a calculation component. However, the efficiency of DEM simulation is very low due to the algorithm and a very large number of elements. This paper analysed the efficiency-related questions of the DEM modelling. The influence of particle shape and contact properties on the force behaviour is studied. Further, an optimised multi-layer ballast track model is introduced based on the most influential ballast areas. In such areas, particles are generated with an irregular shape to ensure the reliability of results, and particles except that area are generated with a rolling resisted ball shape to decrease the number of elements. A series of lateral resistance simulations are conducted to show and validate the accuracy and efficiency of this method in the dimension of the single sleeper section. Results show that this optimised multi-layer model building method largely improves efficiency, and it can provide accurate data.

Vehicle-mounted ground-penetrating radar (GPR) has been used to non-destructively inspect and evaluate railway subgrade conditions. However, existing GPR data processing and interpretation methods mostly rely on time-consuming manual interpretation, and limited studies have applied machine learning methods. GPR data are complex, high-dimensional, and redundant, in particular with non-negligible noises, for which traditional machine learning methods are not effective when applied to GPR data processing and interpretation. To solve this problem, deep learning is more suitable to process large amounts of training data, as well as to perform better data interpretation. In this study, we proposed a novel deep learning method to process GPR data, the CRNN network, which combines convolutional neural networks (CNN) and recurrent neural networks (RNN). The CNN processes raw GPR waveform data from signal channels, and the RNN processes features from multiple channels. The results show that the CRNN network achieves a higher precision at 83.4%, with a recall of 77.3%. Compared to the traditional machine learning method, the CRNN is 5.2 times faster and has a smaller size of 2.6 MB (traditional machine learning method: 104.0 MB). Our research output has demonstrated that the developed deep learning method improves the efficiency and accuracy of railway subgrade condition evaluation.

Particle shape plays an essential role in deformation characteristics of railway ballast bed. The numerical reconstruction of ballast morphological features, including overall shape and angular distribution, remains a hot issue in research on ballast mechanical behavior simulation. A novel shape reconstruction method was adopted to generate ballast particles that met the desired probability density distribution of morphological indices. On this basis, the numerical model of ballast triaxial tests were established under different confining pressures. The results were compared with those obtained from indoor tests and simulations whose particles were generated from 3D scanning or non-statistical random generation. The results show that the particle shape has a growing effect on the mechanical response of ballast, with an increase in confining pressure. The relation between deviatoric stress and axial strain in the specimen which meets the probability density distribution is more consistent with the experimental results than that of the non-statistical randomly generated specimen. The lateral deformation of ballast is correlated with the adjustment of the packing structure. For non-statistical randomly generated specimen, both the lateral deformation and the particle adjustment are larger than those generated by 3D scanning. The ballast contact force evolution is less influenced by its morphological features. Nevertheless, the difference in the maximum contact force of specimens with various particle shapes is nearly 50%.

The dynamic performance of a railway track subjected to moving trains depends strongly on track support conditions. In reality, even for the well-constructed and well-maintained tracks, sleeper support stiffness and global track stiffness vary substantially along the track, which affects the train-track dynamic interactions, causing rapid track geometry degradation as well as the riding comfort and safety issues. Consequently, track stiffness irregularity (TSI, the spatial variation of track stiffness along the track) is important for railway construction and maintenance in addition to track geometry irregularities. So far, extensive research has been published on the TSI whereas the relevant issues have not been paid sufficient attention. In this paper, a summary and comments have been made in the field of TSI about the current research status and future trends from a critical point of view. Novel concepts of the critical values of TSIs and the integrated management of the track geometry and stiffness irregularities are proposed. The review presented in this work is valuable to advance the research on TSI and can help guide the design, construction and maintenance of railway tracks.

Ballast layer defects are the primary cause for rapid track geometry degradation. Detecting these defects in real-time during track inspections is urgently needed to ensure safe train operations. To achieve this, an indicator, the track degradation rate (TDR) was proposed. This rate is calculated using track geometry inspection data to locate and predict railway-line sections with ballast layer defects. The TDR is determined by the monthly standard deviation of the rail longitudinal level, which is one aspect of track geometry. The Ballast Layer Health Classification (BLHC) is designed by assessing the two successive TDRs before and after track geometry maintenance actions. The BLHC is used to categorize the conditions of the ballast layer, including normal periodic deterioration, abrupt deterioration, effective maintenance, rising deterioration, and severe deterioration. Both the TDR and BLHC were validated through field assessments of ballast layer conditions, where the two indicators were found to be effective in revealing defects. The results indicate that the TDR is sensitive to ballast layer defects, while the BLHC can quickly identify the location of these defects. Consequently, the BLHC can provide real-time guidance for ballast layer maintenance.

Parent rock strength and crumb rubber modification are two critical mechanical parameters that significantly decide the ballast layer degradation subjected to train dynamic loading. Using machine learning to predict ballast degradation considering these two parameters is helpful for deciding ballasted track maintenance cycle. In the current study, the ballast degradation process data (variables: parent rock types, loading types, ballast gradations and compositions of crumb rubber-ballast mixture) were used to train machine learning models. The drop-weight impact loading tests were performed to simulate different train dynamic loadings. Two well-established machine learning models, i.e., random forest (RF) and support vector regression (SVR) were trained and verified, to more effectively assess the importance of these variables. The results from the validated machine learning models confirm that the parent rock type is the most influential parameter, followed by the loading type (applied stress level), to control and predict the degradation of the ballast-CR mixture. The experimental assessment reveals that although the incorporation of CR suppresses degradation across all characterized rock types, the improvement in performance of the ballast-CR specimen against degradation is more noticeable for high-strength parent rock subjected to a considerable stress level. Meanwhile, this positive influence is also observed for ballast of weaker strength when the applied stress level is low.

Railway ballast performance is dictated by a complex mix of mechanical properties. These effect its performance at the particle level for example in terms of particle degradation, but also at the track system level in terms of settlement and stability. Therefore this paper seeks to develop new understandings of ballast behaviour and identify opportunities for future research directions. First, ballast particle size and size distribution curves are discussed, considering opportunities to improve breakage, settlement and drainage characteristics. Next, particle geometry is discussed, with a focus on form, angularity and surface texture. This is followed by a discussion on the degradation mechanisms of ballast particles and the effect of fouling on permeability. Next, techniques to assess and improve ballast bulk density are discussed, such as ground penetration radar and dynamic track stabilisation. Testing methods for studying ballast are also reviewed, first considering both smaller-scale tests such as direct shear tests and the Los Angeles abrasion test. Then larger-scale laboratory testing is discussed, including large-diameter dynamic triaxial testing and the use of full-scale laboratory tracks. Finally, conclusions are drawn and suggestions for future research directions are given.