Abnormal carbody vibrations can significantly affect the safety and stability of high-speed trains. In this study, we conducted field measurements to analyze low-frequency swaying of a high-speed train on the Chinese alpine railway line. The test results reveal that the tail vehicle of the train swayed laterally in certain sections of the track, resulting in vehicle lateral stability index exceeding the limit. The swaying frequency of the carbody was mainly distributed in the range of 1.4–1.5 Hz. The causes of abnormal swaying were analyzed by establishing a computational fluid calculation model. Computational fluid dynamics calculations revealed that lift force, side force, and yaw moment were higher for the tail vehicle compared to the head and middle vehicles. The train-track coupled dynamics simulation results demonstrate that aerodynamics disturbances may be the key factor to intensify the abnormal swaying of the high-speed train tail vehicle.

As a common defect of heavy-haul railway wagons, the wheel diameter difference (WDD) will deteriorate wheel/rail dynamic interaction, which severely threatens the running stability and safety of the wagons. Accordingly, it is of great importance to detect the WDD forms of wagons and take appropriate measures in time. In this article, a novel method for WDD form detection of the running wagons by analyzing axle box acceleration (ABA) signals is proposed. To solve the difficulty of weak feature extraction for the vibration signals, a novel feature extraction method combining adaptive chirp mode decomposition (ACMD) with fractal box dimension (FBD) is proposed. First, a 3-D feature space is constructed by the FBDs, which is calculated from the extracted chirp modes by ACMD. Then, a multiple kernel extreme learning machine (ELM) optimized by genetic mutation particle swarm optimization (GMPSO) is developed for the classification of the feature vectors. Both simulation and field test results indicate that the proposed detection method is powerful for accurate identification of the wheelsets with standard diameter, in-phase WDD, and anti-phase WDD, and the algorithm efficiency shows practicability in onboard monitoring.

Turnouts are the weak spot in high-speed rail systems, and it is simple for the phenomenon of the wheel–rail force and the carbody lateral acceleration over-limit to arise when the train passes through, which affects the service life of the rail and the running stability of the train. In this paper, the turnout with wheel–rail force over-limit and carbody lateral acceleration over-limit is selected for analysis, and the profiles of the wheel and rail are monitored. Then, the vehicle–turnout coupled multi-body dynamics model is simulated. Additionally, the portable vibration analyzer, the comprehensive inspection train, and the wheel–rail contact dynamic stress tester monitors the data and evaluates the impact of rail grinding on high-speed railway. The results of this study demonstrated that the turnout profiles are in good agreement with the standard wheel profiles following grinding, and the wheel–rail contact point and equivalent conicity both improved. When the train passes the ground turnout at high speed with and without the wheel polygonal wear, the wheel–rail force and the carbody acceleration were clearly improved. Using the wheel–rail contact dynamic stress tester, the comprehensive inspection train, and the portable vibration analyzer monitoring the changes in the carbody acceleration, the wheel–rail force and the carbody acceleration are definitely better after grinding. Similar to the pattern in the simulation, the train’s running steadiness increased by grinding.

In the wind tunnel test of trains, the scale size changes the Reynolds numbers of trains, which may affect the aerodynamic characteristics of the train. Based on computational fluid dynamics (CFD), numerical models of train aerodynamics with five different scale sizes are established. The five different scale sizes are λ=1/1, 1/2, 1/8, 1/16 and 1/25, respectively, and the aerodynamic characteristics of trains running in the open-air operating condition and crosswind operating condition with different scale sizes are numerically simulated. The results show that the pressure drag coefficients and pressure lift coefficients of the train tend to decrease with the decrease of the scale size. In the open-air operating condition, compared with the full-size train, the pressure drag coefficient of the 1/25th scaled train is less by 14.4%, and the pressure lift coefficients of the head car, middle car and tail car change 16.1%, 46.6% and 12.3%, respectively. The scale size affects the velocity gradient near the train surface and the position of flow separation changes. The decrease of the scale size leads to the decrease of Reynolds numbers and the increase of viscous drag coefficient. When the scale size is 1/25, the viscous drag coefficient of the train is 0.186, which is 48.6% larger than the one of the full-size train. Compared with the open-air operating condition, the trend of the pressure drag coefficients and viscous drag coefficients is consistent except for the head car in crosswind operating condition when the scale size decreases. In the range of scale size λ between 1/1 and 1/25, the aerodynamic drag coefficient of the head car, middle car and tail car increase with the decrease of scale size, and the difference in the aerodynamic drag coefficient of the train is 12.9%. In addition, the train’s aerodynamic lift coefficient shows an increasing trend with the decrease of scale size.

When trains pass through turnouts, the lateral vibration acceleration of the carbody (LVAC) often exceeds the permissible limit, with excessive wheel-rail wear being one of the causes of this phenomenon. Rail grinding is considered a potential solution, but there is currently a lack of a grinding strategy model to provide precise grinding strategies in engineering. To address this issue, wheel profiles associated with excessive LVAC and their corresponding turnout rail profiles are collected and analyzed. The worn rail profiles are fitted using cubic Non-uniform Rational B-splines (NURBS) curves based on value points. Based on the vehicle dynamics model and the Back Propagation Neural Network-based Non-dominated Sorting Genetic Algorithm II (BP-NSGA II), the optimized profiles of personalized grinding target are obtained. By integrating considerations of grinding efficiency and quality, the grinding modes and strategies are optimized, ultimately leading to the development of a personalized grinding strategy model (PGSM). Research results show that the optimized profiles effectively reduce the grinding depth and improve the wheel-rail contact interaction. The PGSM can calculate optimized grinding strategies in real-time for practical engineering. After applying this model, significant improvements are observed in LVAC, derailment coefficient, and wear index. In summary, the PGSM offer a robust solution for enhancing rail grinding efficiency and mitigating carbody vibration, thereby contributing to improved railway performance and safety.