Climate change amplifies temperature fluctuations and extreme weather events, potentially increasing the risk of buckling in continuously welded rail track. Current mitigation strategies, such as imposing conservative speed restrictions during hot weather, are mainly reactive and may be costly and disruptive. This Paper investigates the variation of stress-free temperature and its influence on buckling potential along railway track, by means of dynamic finite element analysis. A 3D Abaqus model, which accounts for permanent deformation of the sleeper positions and rail geometry, is developed to simulate the dynamic response of a track subjected to temperature cycles. A parametric study investigates the effects of key factors: rail lateral misalignment, longitudinal temperature gradient, fastening system strength, train speed, braking force and track curvature. The results give insights into the potential range of axial stress profiles, residual displacements and stress-free temperature variations along continuously welded railway track. Vulnerable track sections, such as those exposed to differential heating and cooling, unidirectional braking/acceleration, and with degraded fasteners, are identified. The findings can be used to support the development of more effective and proactive maintenance strategies, inform revisions to current standards, and help reduce operational disruptions. Additionally, the methodology provides a transferable framework for international rail operators facing similar climate change-induced risks.
The extensive operational mileage and heavy traffic loads of railways impose stringent requirements on rail profile quality, as it directly governs wheel-rail contact geometry and vehicle hunting stability. However, existing rail evaluation indicators focus primarily on geometric deviations without establishing a clear link to contact parameters, limiting their effectiveness for dynamics-oriented maintenance. This study first proposes a set of novel rail profile indicators from two complementary perspectives of wear morphology and geometric characteristics. Based on field measurements of rail profiles from high-speed railway lines, the correlation between each indicator and wheel-rail contact parameters is then analyzed under both symmetric and asymmetric rail wear conditions, revealing the mapping relationship from geometric features to equivalent conicity. Finally, an orthogonal experimental design is employed to eliminate confounding factors such as wheel profile, gauge, and rail cant, thereby validating the robustness of the proposed indicators. Results show that the rail head indentation exhibits the strongest correlation with nominal equivalent conicity, achieving a Spearman coefficient exceeding 0.95 and an R 2 value of 0.97, followed by other geometric indicators. The proposed indicators effectively capture the spatial heterogeneity and left-right asymmetry of rail wear, and the derived regression equations provide a quantitative mapping from geometric features to contact parameters. This research bridges the gap between rail profile geometry and wheel-rail contact behavior, offering a scientific basis for condition-based rail grinding decisions and contributing to enhanced operational stability and cost-effective maintenance of high-speed railways.
A series of mobile flash-butt welds in a combination of new and worn high strength 68 kg/m rail were manufactured with upset distances ranging from low (11 mm) to high (18.4 mm). Tensile test pieces were prepared from a number of positions in each weld to examine the influence of both rail condition (i.e. new vs worn) and upset distance on bond strength. Additional tests were performed to determine the parent rail compositions associated with each weld, as well as the hardness distribution in the web. The results of the test program showed that in the absence of flat-spot defects, variation of the forging distance did not measurably influence the tensile strength in a consistent manner across all test positions. However, forging distance did appear to be linked with the prevalence of specific types of flat-spots, including voids and flat spots, at the bond line. The experimental results showed that it is imperative that the upset cycle is sufficient to close up voids and extrude the bulk oxides responsible for dull type flat-spots.
The levitation system is essential for non-contact operation in maglev trains, with control performance being critical for stable long-term operation. In low-speed maglev trains with electromagnetic suspension (EMS), the coupling of multiple electromagnet modules enhances coordination but also transmits coupled vibrations, requiring vibration suppression for stability. This paper develops the dynamic model of a two-point levitation system based on the single-point levitation system of low-speed EMS maglev trains and analyzes its coupling characteristics. A Linear Active Disturbance Rejection Control (LADRC) levitation controller is designed for disturbance suppression, with a cross-coupling compensation loop to improve synchronization. A Takagi-Sugeno (T-S) fuzzy neural network-based synchronous error compensator is introduced to reduce gap deviation and ensure synchronization of electromagnets. Experimental validation shows that with the LADRC controller, the gap deviation stays within 0.15 mm, while the error compensator reduces it to under 0.1 mm, improving synchronization and preventing performance degradation from a fixed cross-coupling coefficient.
To investigate the safety implications of onboard hydrogen leakage in hydrogen-powered rail vehicles under tunnel conditions, this study established a test platform incorporating a full-scale hydrogen storage module and tunnel section, together with high-pressure hydrogen release, concentration and overpressure measurement, and video recording. Six tests, comprising two leakage-only tests and four delayed-ignition tests, examined the blow-down of a single 140 L cylinder initially pressurised to 35 MPa through a manually actuated 5 mm TPRD. The tests were conducted under a fans-off condition and a preset ventilated condition, with the latter corresponding to a fan-controller setting of 50 km/h. A source-term analysis based on the Abel–Noble equation of state and the choked-flow relation gave an initial mass-flow rate of 0.32 kg/s and a choked-flow duration of approximately 56 s. Under the fans-off condition, the highest measured hydrogen concentration was 31.6 vol.% at the sensor above the release region. Clear ignition-related overpressure was recorded at only one of the five pressure sensors, with a maximum of approximately 5.0 kPa; the other sensors showed no response exceeding the approximately ±0.3 kPa baseline fluctuations. Sustained ignition occurred in three of the four delayed-ignition tests; each produced a jet fire, and no deflagration-type pressure signature was identified at the five instrumented pressure locations within the 30 s pressure-recording window. Under the preset ventilated condition, hydrogen was detected at the downstream sensor locations within the instrumented region, and the flame was progressively deflected as the release weakened. These measurements and the documented source-term procedure provide condition-specific engineering information that may inform future ventilation design and hazardous-zone assessment for comparable roof-mounted hydrogen-storage configurations.
