The safe and efficient operation of high-speed trains relies heavily on the stable support of axle box bearings, a core component in the bogie system. Undertaking the critical responsibilities of transmitting traction, braking force, and supporting wheel rotation, axle box bearings act as the "load-bearing joints" of high-speed trains. Their load characteristics directly determine fatigue life, while reliability is closely linked to operational safety. Failures such as fatigue spalling or wear in axle box bearings can easily lead to serious accidents like wheel instability or derailment. Therefore, in-depth research on the load laws, accurate fatigue life prediction, and optimization schemes of high-speed train axle box bearings holds significant engineering value and practical significance for enhancing operational safety and reducing maintenance costs.

1. Load Characteristics and Test Scheme Design of Axle Box Bearings

High-speed train axle box bearings primarily adopt double-row tapered roller bearings, operating under extremely complex conditions. During acceleration, deceleration, cornering, and turnout passage, bearings must withstand radial loads, axial loads, and overturning moments simultaneously. Track irregularities induce transient impact loads, while frequent starts/stops and steering during low-speed depot entry/exit result in specific load accumulation. These dynamic load variations are the core incentives for bearing fatigue damage, making accurate load data acquisition the foundation of subsequent research.

To meet the core demand for load measurement, the research team innovatively designed two reliable and practical test schemes. The first is an indirect measurement method: by sensorizing axle box springs and swing arms, real-time force data is collected to inversely calculate the comprehensive bearing loads. This scheme requires no modifications to the bearing itself, minimizing impact on normal train operation and suitable for long-term tracking tests. The second is a direct measurement method: stress measurement points are arranged in precision slots on the bearing outer ring to capture stress signals for load identification, maximizing the fidelity of the bearing's actual force state with higher measurement accuracy. Verified through actual high-speed train operation tests, both schemes output stable and reliable data, providing solid support for subsequent mechanical analysis and life prediction.

Furthermore, research using a vehicle-track dynamics model revealed that without track disturbances, vehicle speed has little impact on the contact load between rollers and raceways, but the contact load of rollers in the non-load-bearing area exhibits a square relationship with speed. Track disturbances cause sharp fluctuations in instantaneous loads; higher speeds lead to greater standard deviations of outer ring contact loads and more significant load fluctuations. These findings provide important theoretical basis for a comprehensive understanding of axle box bearing load characteristics.

2. Construction of Mechanical Analysis Models for Axle Box Bearings

To deeply analyze the internal force mechanism of double-row tapered roller bearings, a combined approach of statics and finite element methods (FEM) was adopted to establish mechanical analysis models, enabling progressive analysis from preliminary estimation to precise simulation.

The statics method simplifies the bearing structure by ignoring non-critical details, allowing rapid acquisition of approximate solutions for load distribution, suitable for preliminary scheme evaluation and screening. In contrast, the FEM demonstrates significant advantages in detail analysis due to its adaptability to complex structures. Using FEM software, a refined bearing model was constructed to accurately simulate the contact state between rolling elements and inner/outer rings, as well as to clearly reveal the variation laws of key parameters such as contact load distribution, component deformation, and contact stiffness. A comparison of results from both methods showed consistent overall trends in load distribution, further verifying the reliability of the analysis.

In addition, a specialized contact model was established for the contact area between tapered rollers and raceways, combined with Hertzian theory for stress simulation analysis. This model clearly revealed the intrinsic correlation between bearing loads and stress transmission—for example, contact stress increases nonlinearly with radial load, and stress concentration tends to occur at both ends of the rollers. This finding provides key theoretical support for subsequent stress-based fatigue life assessment and points out directions for bearing structure optimization.

3. Establishment and Verification of Fatigue Life Prediction Models

Based on measured load data, combined with traditional bearing fatigue life theory and the linear cumulative damage criterion, a fatigue life prediction model adapted to the time-varying load conditions of high-speed trains was established, overcoming the limitations of excessive simplification in traditional models. The model fully considers load fluctuations across different speed ranges and the resulting periodic stress cycles, making prediction results more consistent with actual service conditions.

Comparative tests showed that under the same conditions, the fatigue life predicted by the new model is lower than that calculated by traditional theory. Although this conservative estimation may seem "stringent," it significantly enhances the safety margin of train operation. Meanwhile, the research paid special attention to the impact of low-speed special conditions, innovatively introducing the concept of "damage per kilometer" to quantify bearing wear during depot entry/exit. Data indicated that despite low speeds, frequent starts/stops and steering result in much higher damage per kilometer than during normal high-speed operation. This finding emphasizes the necessity of incorporating damage accumulation from such special conditions into life calculations; otherwise, significant deviations in life prediction will occur.

Notably, research on wheel polygonization evolution further improved the life prediction system. It was found that as operating mileage increases, the amplitude of wheel polygon wear increases, leading to a significant rise in the maximum contact load and load standard deviation of the bearing outer raceway. The maximum damage per kilometer can reach 1.4×10