In the field of aeroengine design and maintenance, fatigue-related failures are a critical concern, accounting for over 40% of all engine issues. As a key component, the reducer plays a vital role in transmitting power from the turbine to the propeller or transmission system, and any failure in its elements, such as the gear shaft, can lead to significant safety risks. This article addresses a fracture incident in the spherical press sleeve of an input drive gear shaft within a reducer during a 150-hour endurance test. We investigate the root causes through comprehensive methods, including macroscopic and microscopic examinations, material testing, fatigue experiments, and simulation analyses. Our focus is on the gear shaft assembly, where the press sleeve, made of GCr15 bearing steel, experienced cracking and eventual fracture due to stress concentrations and misalignment. By detailing our analytical approach and proposing design and process improvements, we aim to enhance the reliability of similar spherical flexible connections in aeroengines, ensuring optimal performance and safety. Throughout this study, the term ‘gear shaft’ is emphasized to highlight its centrality in the system, and we incorporate tables and equations to summarize key findings and theoretical models.
The reducer in this aeroengine is a two-stage planetary system that connects the power turbine to the propeller, with an input speed of 44,000 rpm and an output of 2,200 rpm, achieving a total gear ratio of 20. The input drive gear shaft interfaces with the power turbine via a spherical flexible coupling, which allows for misalignment compensation and uniform load distribution. This gear shaft is critical for transmitting torque, and its press sleeve, a concave spherical component, is designed to engage with a convex spherical plug on the turbine side. During post-test disassembly, we observed that two segments of the press sleeve’s spherical surface had fractured, indicating potential issues in the gear shaft assembly. The fracture initiated in areas with severe wear, suggesting stress concentrations and improper loading. To understand the failure mechanism, we conducted a series of inspections and tests, focusing on the gear shaft’s geometry and material properties. The spherical press sleeve is subjected to complex loads, including axial forces up to 1,038 N, friction, and torsion, which can lead to fatigue if not properly managed. Our initial assessment pointed to possible causes such as inadequate design strength, material defects, or assembly errors, but further analysis was needed to pinpoint the exact factors.
We began with a macroscopic examination of the failed press sleeve, which revealed fractures in two spherical segments, labeled as 1# and 2#, with visible wear patterns and a reddish discoloration on segment 3#. The gear shaft’s internal surfaces showed circumferential scoring and metal debris, particularly at the junction between the small inner bore and the inner end-face, where the designed chamfer was smaller than specified (approximately 0.4 mm instead of 0.8 mm). This reduction in chamfer size likely contributed to stress concentration and misalignment during assembly, exacerbating wear on the press sleeve. Microscopic analysis using scanning electron microscopy (SEM) indicated that the fracture originated in the heavily worn regions, with features typical of fatigue crack propagation, such as radial marks and a mix of quasi-cleavage and dimple morphologies. The fracture surfaces appeared gray and flat, consistent with brittle failure, and no significant material defects were detected at the crack initiation sites. To evaluate the material integrity, we performed composition analysis via energy-dispersive spectroscopy, hardness testing, and non-metallic inclusion assessment. The results confirmed that the press sleeve material, GCr15, met the required specifications, with a carbon content of 0.96%, chromium at 1.41%, and a hardness of 740–749 HV0.5 (approximately 62 HRC). Thus, material quality was ruled out as a primary cause, directing our attention to geometric and operational factors.
Further investigation involved fatigue testing on specimens machined from the same GCr15 batch to assess the material’s performance under simulated conditions. We conducted high-frequency tension-compression and rotating bending fatigue tests, with notched samples having a root radius of 0.4 mm to mimic the press sleeve’s critical region. The fatigue life data, summarized in Table 1, showed that the material endured stress levels between 390 MPa and 540 MPa for thousands to tens of thousands of cycles, indicating satisfactory fatigue resistance under standard conditions. However, the fracture surfaces of these test specimens exhibited brittle characteristics, similar to the failed press sleeve, with crack initiation at the notch roots and propagation patterns involving quasi-cleavage and dimples. This consistency suggested that while the material itself was adequate, the actual service conditions imposed higher stresses due to geometric imperfections. For instance, in the rotating bending test, the fracture surface showed intergranular features, hinting at potential environmental or loading effects. These findings reinforced the need to analyze the stress distribution in the gear shaft assembly under operational loads.
| Specimen Type | Stress Range (MPa) | Cycles to Failure |
|---|---|---|
| Tension-Compression | 390–480 | 1,000–3,150 |
| Rotating Bending | 510–540 | 9,357–18,004 |
To quantify the stress conditions, we employed Kisssoft software for simulation calculations, modeling the press sleeve’s interaction with the gear shaft under various assembly scenarios. The gear shaft operates at high speeds and temperatures (200°C), with an axial load of 1,038 N applied during engine operation. We evaluated different cases: ideal assembly with no misalignment, normal assembly with 0.65 mm float, and misaligned assembly with reduced contact widths on the spherical surface. The equivalent stress (von Mises stress) was computed using the formula: $$\sigma_{eq} = \sqrt{\frac{1}{2}\left[(\sigma_1 – \sigma_2)^2 + (\sigma_2 – \sigma_3)^2 + (\sigma_3 – \sigma_1)^2\right]}$$ where $\sigma_1$, $\sigma_2$, and $\sigma_3$ are the principal stresses. The results, presented in Table 2, demonstrated that under ideal conditions, the maximum equivalent stress was around 205 MPa, well within the material’s strength limits. However, when the assembly was misaligned, with the load applied over narrower contact widths, the stress increased dramatically. For a contact width of 0.08 mm, the equivalent stress reached 1,000 MPa, exceeding the material’s fatigue strength and leading to crack initiation. This stress concentration was primarily due to the small chamfer at the gear shaft’s inner bore, which caused uneven load distribution and localized high stresses. The simulation output, visualized as stress contour plots, highlighted the critical regions where fractures were likely to occur, aligning with our observational data from the failed component.
| Assembly State | Float (mm) | Contact Width (mm) | Equivalent Stress (MPa) |
|---|---|---|---|
| Ideal | 0.00 | N/A | 204 |
| Normal | 0.65 | N/A | 205 |
| Misaligned | 0.65 | 0.32 | 279 |
| Misaligned | 0.65 | 0.16 | 517 |
| Misaligned | 0.65 | 0.08 | 1000 |
Based on these analyses, we identified the root cause of the press sleeve fracture as a combination of geometric stress risers and assembly-induced misalignment in the gear shaft. The small chamfer (0.4 mm) at the junction of the gear shaft’s small inner bore and inner end-face created a sharp transition, leading to high stress concentrations when mated with the spherical press sleeve. During assembly, this could cause the press sleeve to tilt, resulting in uneven loading and localized stress peaks during operation. The fatigue life of the component can be estimated using the Basquin equation: $$N_f = C \cdot S^{-b}$$ where $N_f$ is the number of cycles to failure, $S$ is the stress amplitude, and $C$ and $b$ are material constants. For GCr15, under high-stress conditions, the reduced contact width significantly shortens the fatigue life, explaining the premature failure. Additionally, micro-motion and fretting wear at the interface, evidenced by oxide layers on the press sleeve surface, further exacerbated the situation. To address these issues, we proposed design modifications: increasing the chamfer size to C1.5 (approximately 1.5 mm) and reducing the press sleeve’s outer fillet radius to R0.4 ± 0.1 mm to avoid interference fits. We also revised the assembly process by introducing specialized tooling to ensure vertical installation of the press sleeve into the gear shaft, minimizing tilt. Post-assembly checks included using a 0.01 mm feeler gauge to verify gap uniformity and measuring the spherical surface runout to within 0.02 mm for precision alignment.
Implementation of these improvements was validated through a subsequent 150-hour endurance test on a bench engine. After disassembly, the press sleeve and gear shaft showed no signs of abnormal wear or cracking, confirming the effectiveness of the changes. The gear shaft’s performance was stable, with even load distribution and no stress concentration points. This case underscores the importance of meticulous design and assembly in critical components like the gear shaft, where small geometric details can have profound impacts on reliability. In conclusion, the fracture of the spherical press sleeve was primarily due to stress concentrations from an undersized chamfer and assembly misalignment, leading to fatigue failure. Our multidisciplinary approach, combining experimental testing and simulation, provided a clear pathway for improvement. These insights are applicable to other spherical flexible connections in aeroengines, enhancing overall safety and durability. Future work could explore advanced materials or surface treatments to further optimize the gear shaft assembly under high-cycle fatigue conditions.
In summary, this study highlights the critical role of the gear shaft in aeroengine reducers and the need for robust design practices. By addressing geometric imperfections and refining assembly techniques, we can mitigate fatigue risks and extend component life. The integration of theoretical models, such as stress analysis and fatigue life predictions, with empirical data ensures a comprehensive understanding of failure mechanisms. As aeroengines continue to evolve, lessons from this analysis will contribute to more reliable and efficient propulsion systems, where the gear shaft remains a focal point for innovation and improvement.