During the initial operational phase of a specific model of reducer designed for palm oil processing equipment, a significant and concerning failure mode emerged: a high-volume fracture of the input gear shafts. These failures occurred prematurely, some even during the commissioning or trial run stages, indicating a fundamental flaw rather than a wear-out mechanism. To prevent recurrence and ensure reliability, a systematic root cause analysis was imperative. This article details my investigation into the fracture causation of these gear shafts, examining material properties, metallurgical structure, assembly procedures, and critically, the heat treatment process.
The reducer in question was coupled to a 45 kW motor, with the input gear shaft rotating at a high speed of 11,470 rpm. The lubrication of the gearbox was confirmed to be normal. Macroscopic examination of the fractured gear shafts revealed a consistent pattern. The break consistently originated at the terminus of the keyway, precisely at the transition zone between the carburized and non-carburized sections of the shaft. The fracture surface displayed distinct regions: a relatively smooth and bright zone, indicative of stable crack propagation, and a larger rough, fibrous zone, signifying final fast fracture. This morphology is classic of fatigue failure under alternating loads. The overall fracture plane was oriented at approximately 45 degrees to the shaft axis, suggesting a significant torsional stress component.
The design of the failed gear shaft necessitated carburizing and quenching to achieve a hard, wear-resistant surface. However, a segment of the shaft with a diameter of Ø55m6 required machining for a keyway post-heat-treatment. To keep this area soft and machinable, a protective coating was applied to prevent carbon diffusion during carburizing. This created a deliberate boundary on the shaft: one region (to the left of the coating line) was hardened, and the adjacent region (to the right) remained in a softer, lower-strength condition.
Systematic Investigation into the Failure of Gear Shafts
1. Assessment of Mechanical Strength
The gear shafts were manufactured from 18CrNiMo7-6, a low-carbon alloy steel known for its high core toughness and excellent hardenability. A preliminary check against standard mechanical property databases confirmed its suitability for such demanding applications. To rigorously evaluate the shaft’s integrity under service conditions, a finite element analysis (FEA) was performed, simulating a worst-case scenario of motor stall (with torque amplified to 1.8 times the nominal value).
The FEA results are summarized below. The maximum von Mises stress was found at the shoulder fillet where a bearing was seated, not at the eventual fracture site near the keyway. The stress at the keyway end was significantly lower. The material’s minimum specified yield strength (Rp0.2) is 785 MPa and tensile strength (Rm) is 1080 MPa. Comparing these values with the calculated stresses, it was clear that the nominal mechanical loading alone should not have caused yielding or fracture, even under extreme transient conditions.
| Location | Calculated Stress (MPa) | Material Strength (MPa) | Safety Factor |
|---|---|---|---|
| Bearing Shoulder (Max Stress) | 448 | Rp0.2 = 785 | > 1.75 |
| Keyway End (Fracture Origin) | 167 | Rp0.2 = 785 | > 4.7 |
The shaft’s deflection under the combined radial belt load and torque was also analyzed. The maximum displacement, occurring at the free end, was calculated to be 0.324 mm. While not negligible, this deformation was within typical allowable limits for such gear shafts and did not suggest a primary bending failure mode.
2. Material and Metallurgical Examination
2.1 Chemical Composition
A fundamental step was to verify that the gear shaft material conformed to specifications. Spectroscopic analysis was conducted on samples from the failed components. The results, compared against the standard for 18CrNiMo7-6, are presented in the table below. All elemental concentrations were well within the prescribed ranges, effectively ruling out material chemistry as a contributing factor to the failure of these gear shafts.
| Element | Standard Range (%) | Measured Value (%) | Conformance |
|---|---|---|---|
| C | 0.15 – 0.20 | 0.17 | Yes |
| Si | ≤ 0.40 | 0.25 | Yes |
| Mn | 0.50 – 0.90 | 0.57 | Yes |
| Cr | 1.50 – 1.80 | 1.60 | Yes |
| Ni | 1.40 – 1.70 | 1.68 | Yes |
| Mo | 0.25 – 0.35 | 0.30 | Yes |
| S | ≤ 0.025 | 0.002 | Yes |
| P | ≤ 0.020 | 0.010 | Yes |
2.2 Microstructural Analysis
The steel 18CrNiMo7-6 derives its properties from a carefully controlled heat treatment cycle: carburizing to increase surface carbon, quenching to form a hard martensitic case, and low-temperature tempering to relieve stresses and improve toughness. Metallographic samples were taken from cross-sections of the failed gear shafts, specifically near the fracture origin. Examination under an optical microscope revealed a uniform microstructure in the core, consisting primarily of tempered martensite with a fine, acicular morphology. No evidence of excessive retained austenite, non-martensitic transformation products (like bainite or ferrite networks), or overtempering was observed. The case depth was consistent and met the drawing specifications. Therefore, the basic execution of the heat treatment process was deemed satisfactory and not the direct source of failure.
3. Evaluation of Assembly and Process Factors
Following the initial failures, the assembly process was scrutinized. To eliminate potential misalignment, a laser-alignment procedure was employed during the replacement of the fractured gear shafts. Despite this precision intervention, subsequent units continued to experience identical fractures. This empirical evidence strongly suggested that the root cause was intrinsic to the component itself, not an extrinsic assembly error.
The focus then shifted decisively to the manufacturing process sequence, particularly the interaction between design intent and heat treatment. The fracture consistently initiated at the boundary between the hardened and soft zones. While the FEA showed the highest stress concentration at a different geometric feature (a fillet), it did not account for material property gradients. The manufacturing sequence was as follows:
- Apply anti-carburizing coating to the Ø55m6 segment destined for keyway machining.
- Perform gas carburizing on the entire shaft assembly (coating protects one zone).
- Quench and temper the shaft, creating a high-hardness case (~58-63 HRC) on uncoated areas and a much softer core/coated zone (<45 HRC).
- Remove coating and machine the keyway in the soft zone.
This process creates an abrupt transition in mechanical properties. To quantify this, microhardness traverses were conducted across the A-A boundary (the theoretical coating line).
3.1 Quantifying the Hardness Gradient
Hardness measurements were taken in two critical directions: along the axis of the gear shaft (surface readings) and radially from the surface to the core at the transition boundary.
Axial Surface Hardness Profile: Traversing 1.8 mm on either side of the A-A boundary revealed a dramatic drop in surface hardness. The data is represented by the following relationship, where \( x \) is the axial distance from the boundary (mm) and \( H_{RC}(x) \) is the Rockwell C hardness:
$$ H_{RC}(x) \approx 62 – 25 \cdot \text{tanh}(5x) \quad \text{for } x \in [-1.8, 1.8] $$
This equation, derived from measured data points, models the steep transition. The hardness plunges from approximately 62 HRC in the carburized zone to about 37 HRC in the coated zone over a very short distance.
| Axial Position from Boundary (mm) | Average Surface Hardness (HRC) |
|---|---|
| -1.5 (Carburized Side) | 61.5 |
| -0.5 | 60.0 |
| 0 (Boundary) | 52.0 |
| +0.5 | 40.5 |
| +1.5 (Coated Side) | 37.0 |
Radial Hardness Profile at the Boundary: A traverse from the surface to the core at the A-A plane showed that the steep gradient penetrates deeply. The case hardness persists for a short depth before a rapid decline towards the softer core hardness of the low-carbon steel.
| Radial Depth from Surface (mm) | Hardness (HRC) at Boundary |
|---|---|
| 0 (Surface) | 52.0 |
| 1.0 | 48.0 |
| 2.0 | 40.0 |
| 5.0 | 35.0 |
| 10.0 (Core) | 32.0 |
3.2 The Mechanics of the Hardness Gradient Failure
The presence of such a severe hardness gradient is mechanically detrimental. Hardness is directly related to yield strength (\( \sigma_y \)). A sharp change in hardness implies an equally sharp change in local yield strength. When a component with this gradient is subjected to load, the strain must be continuous (compatibility condition), but the stress (\( \sigma \)) distribution will be discontinuous because \( \sigma = E \cdot \epsilon \) only holds elastically, and the local yield limits differ. This mismatch induces localized stress concentrations and, more critically, creates significant residual stresses due to differential volumetric changes during quenching.
The quenching transformation from austenite to martensite involves a volume expansion. The carburized zone, with its high hardenability, transforms fully and expands considerably. The coated, low-carbon zone transforms to a mixture with less volume expansion. The constrained expansion of the carburized zone by the softer adjacent material leads to the development of tensile residual stresses in the transition region of the softer material. This can be conceptually framed. The residual stress (\( \sigma_{res} \)) in the transition zone can be related to the difference in volumetric strain (\( \Delta \epsilon_v \)) and the constraint:
$$ \sigma_{res} \propto E \cdot \Delta \epsilon_v \approx k \cdot (H_{case} – H_{core}) $$
where \( k \) is a proportionality constant incorporating material and geometric factors, and \( H \) represents hardness as a proxy for transformation strain.
For the fractured gear shafts, the high-cycle fatigue loading from belt forces and torque superimposes an alternating stress (\( \sigma_a \)) on this already elevated tensile mean stress (\( \sigma_m \)) from the residual stress field. The modified Goodman relation highlights the danger:
$$ \sigma_a = \sigma_e \left( 1 – \frac{\sigma_m}{\sigma_{UTS}} \right) $$
where \( \sigma_e \) is the endurance limit of the material in a zero-mean-stress test. A high tensile \( \sigma_m \) drastically reduces the allowable alternating stress amplitude \( \sigma_a \) for infinite life. The combination of:
- High tensile residual stress at the hardness transition.
- Stress concentration from the keyway’s terminal geometry.
- Cyclic service loads (bending from belt pull and torsion).
created a perfect environment for fatigue crack initiation and propagation. The crack initiated in the softer, stressed material at the boundary, propagated until the remaining cross-section could no longer support the load, resulting in the observed sudden fast fracture. The 45-degree fracture plane is consistent with failure dominated by shear stresses from torsion, exacerbated by the bending stresses.
Conclusion and Engineering Insight
The investigation conclusively identified the primary cause of failure for these gear shafts as high-cycle fatigue originating from a zone of severe property mismatch. While material quality, basic heat treatment metallurgy, and assembly were ruled out, the specific heat treatment process flow was the culprit. The practice of creating an abrupt, full-section hardness transition on a shaft subjected to significant combined bending and torsional loads introduced a potent source of stress concentration and detrimental residual stresses.
The key lesson is that for dynamically loaded components like gear shafts, the gradient of mechanical properties is as critical as the absolute property values. An abrupt hardness change acts as an internal stress riser. To mitigate this in future designs, several approaches could be considered:
- Design Modification: Relocate the keyway entirely outside the carburized zone if possible, avoiding the property transition in a highly stressed section.
- Process Alternative: Use a selective heat treatment method (e.g., induction hardening) only on the gear teeth and bearing journals, leaving the keyway section in a normalized or tempered condition with a less severe gradient.
- Material/Process Optimization: If carburizing is essential, employ a more gradual carbon profile or a post-carburizing differential tempering process to soften the gradient. Alternatively, specify a material and case depth that allows for a less drastic hardness drop.
In summary, the fracture of these gear shafts underscores a fundamental principle in mechanical design: components must be analyzed not just as homogeneous bodies under load, but as systems of interconnected material domains with potentially incompatible properties. The management of interfaces and gradients is paramount for durability, especially in demanding power transmission applications.