In the operation of wind turbines, the gearbox plays a critical role in transmitting mechanical energy, and the gear shaft is a core component responsible for meshing with gears and bearing significant bending stresses. During service, a gear shaft in a wind turbine gearbox experienced tooth breakage after approximately 10 months of operation. This study investigates the root cause of this failure through a comprehensive analysis using various techniques. The gear shaft material was 18CrNiMo7-6 steel, manufactured through processes including steelmaking, forging, heat treatment, and precision machining. The failure involved a single tooth fracture, with no evidence of wear, corrosion, or abnormal assembly, prompting an in-depth examination to prevent future occurrences.
The primary objective of this analysis is to identify the factors leading to the gear shaft tooth breakage, focusing on material defects, mechanical properties, and operational conditions. We employed macroscopic observation, fracture analysis, energy-dispersive spectroscopy (EDS), hardness testing, and metallographic examination to assess the gear shaft. The gear shaft’s design involves meshing at one end and interference fit with a large gear at the other, making it susceptible to high stress concentrations. By integrating these methods, we aimed to correlate the observed failure with underlying material inconsistencies and propose effective corrective measures.
Macroscopic observation of the broken gear shaft revealed that only one tooth had fractured, with the fracture surface displaying distinct fatigue marks, such as beach lines, indicating a single crack origin. The remaining teeth were intact, and the meshing patterns showed over 95% contact area, ruling out misalignment or uneven loading as potential causes. The crack initiation site was located near the subsurface, suggesting that internal defects might have played a role. This initial assessment guided further microscopic and chemical analyses to pinpoint the exact failure mechanism.
Chemical composition analysis of the gear shaft core was conducted using spark discharge atomic emission spectrometry, adhering to standard methods. The results, summarized in Table 1, confirm that the material complies with the specifications for 18CrNiMo7-6 steel, with no deviations in key elements like carbon, chromium, nickel, or molybdenum that could contribute to premature failure.
| Element | Measured Value (%) | Standard Requirement (%) |
|---|---|---|
| C | 0.18 | 0.15–0.21 |
| Si | 0.26 | ≤0.40 |
| Mn | 0.66 | 0.50–0.90 |
| P | 0.010 | ≤0.025 |
| S | 0.004 | ≤0.035 |
| Cr | 1.59 | 1.50–1.80 |
| Ni | 1.44 | 1.40–1.70 |
| Mo | 0.28 | 0.25–0.35 |
Scanning electron microscopy (SEM) examination of the fracture surface uncovered a significant inclusion at the crack origin, approximately 2–3 mm in length, which differed markedly from the surrounding matrix. The area exhibited fatigue striations and quasi-cleavage features, indicative of progressive crack propagation under cyclic loading. Energy-dispersive spectroscopy (EDS) analysis of this region, as shown in Figure 5 (refer to the spectral data), revealed elevated levels of aluminum and oxygen, identifying the inclusion as a B-type alumina (Al₂O₃) inclusion. Such inclusions are known for their high hardness and ability to disrupt material continuity, leading to stress concentration and fatigue initiation.
To evaluate potential manufacturing issues, grinding burn tests were performed on both the fractured and intact teeth surfaces, following standardized etching procedures. No evidence of thermal damage was found, eliminating grinding-induced defects as a cause. Hardness measurements, conducted using Rockwell hardness tests, yielded average values of 60.3 HRC at the surface and 33.7 HRC at the core, both within design specifications. Additionally, the carburized layer depth and hardness gradient were assessed, with results presented in Table 2 and Figure 7, showing a case depth of 2.0 mm and a gradual decrease in hardness, consistent with requirements.
| Distance from Surface (mm) | Hardness (HV) |
|---|---|
| 0.1 | 659 |
| 0.3 | 662 |
| 0.5 | 662 |
| 0.7 | 646 |
| 0.9 | 633 |
| 1.1 | 625 |
| 1.3 | 628 |
| 1.5 | 615 |
| 1.6 | 606 |
| 1.7 | 594 |
| 1.8 | 578 |
| 1.9 | 556 |
| 2.0 | 552 |
| 2.1 | 529 |
| 2.2 | 525 |
| 2.3 | 487 |
| 2.4 | 499 |
| 2.5 | 463 |
| 2.7 | 470 |
| 2.9 | 442 |
| 3.1 | 445 |
| 3.3 | 430 |
| 3.5 | 428 |
The hardness gradient can be modeled using an exponential decay function, which describes the decrease in hardness with depth from the surface. This relationship is given by:
$$ H(d) = H_s + (H_c – H_s) \cdot e^{-k \cdot d} $$
where \( H(d) \) is the hardness at distance \( d \) from the surface, \( H_s \) is the surface hardness, \( H_c \) is the core hardness, and \( k \) is a decay constant specific to the material and heat treatment. For this gear shaft, substituting measured values: \( H_s = 659 \, \text{HV} \), \( H_c = 428 \, \text{HV} \), and \( k \approx 0.5 \, \text{mm}^{-1} \) provides a good fit to the data, indicating proper carburization.
Metallographic examination of the gear shaft revealed a surface microstructure consisting of fine needle-like martensite, retained austenite, and carbides, rated at level 2, while the core exhibited lath martensite with minimal ferrite. Grain size analysis showed a value of 7.5, meeting the standard requirement of ≥6. These findings confirm that the heat treatment processes were executed correctly, and no abnormalities in microstructure contributed to the failure. The presence of the large alumina inclusion, however, created a localized stress concentration that initiated fatigue cracking under operational loads.
Fatigue failure in gear shafts often follows the Paris-Erdoğan law for crack propagation, expressed as:
$$ \frac{da}{dN} = C (\Delta K)^m $$
where \( \frac{da}{dN} \) is the crack growth rate per cycle, \( \Delta K \) is the stress intensity factor range, and \( C \) and \( m \) are material constants. For this gear shaft, the inclusion acted as an initial flaw, reducing the fatigue limit and accelerating crack growth. The stress concentration factor \( K_t \) due to the inclusion can be estimated using:
$$ K_t = 1 + 2 \sqrt{\frac{a}{\rho}} $$
where \( a \) is the inclusion size and \( \rho \) is the root radius of the defect. Given the inclusion dimensions, \( K_t \) values exceeding 3 were possible, significantly amplifying the applied stresses and leading to premature failure.
Further analysis involved assessing the influence of inclusion size on fatigue strength. The relationship between fatigue limit \( \sigma_f \) and inclusion size \( \sqrt{\text{area}} \) is described by the Murakami equation:
$$ \sigma_f = \frac{C}{( \sqrt{\text{area}} )^{1/6}} $$
where \( C \) is a constant dependent on material and stress ratio. For the observed alumina inclusion with \( \sqrt{\text{area}} \approx 2.5 \, \text{mm} \), the calculated fatigue limit drops substantially, explaining the early failure despite nominal design adequacy. This underscores the criticality of controlling inclusion size in high-stress components like the gear shaft.
To mitigate such issues, corrective measures were implemented in the steelmaking process. During the melting phase, furnace lining drying and control of charge purity were emphasized to reduce oxide formation. In the oxidation stage, enhanced boiling duration, slag removal, and optimized oxygen and argon flow rates were adopted. The reduction phase involved careful alloy addition and desulfurization, while vacuum degassing with prolonged argon stirring helped minimize gaseous inclusions. During casting, argon protection prevented reoxidation, and thorough channel cleaning ensured cleaner steel. Forging processes were optimized with adequate reduction ratios to break down and disperse any remaining inclusions.
The effectiveness of these measures was validated through subsequent production of nearly 100 gear shafts, with no reported tooth breakages, demonstrating that improved steel cleanliness directly enhances the durability of the gear shaft. This case highlights the importance of integrated quality control from raw material to finished product, particularly for critical components in renewable energy applications.
In conclusion, the gear shaft tooth breakage was attributed to fatigue fracture originating from a large B-type alumina inclusion, which induced stress concentration under cyclic loading. Comprehensive analysis confirmed that material composition, hardness, and microstructure were within specifications, but the inherent defect compromised fatigue resistance. The implemented corrective actions in steelmaking and processing successfully eliminated recurrence, emphasizing the need for stringent inclusion control in gear shaft manufacturing. Future work could focus on predictive modeling of inclusion effects and real-time monitoring techniques to further enhance reliability in wind turbine gearboxes.