I carried out a systematic failure investigation on a fractured 20CrMnTi pinion gear that had been operating in a mechanical transmission. The pinion gears in this application are expected to transmit torque, change speed, and alter the direction of motion between two shafts. Because pinion gears must simultaneously endure impact loading, alternating bending stress, and contact stress, their failure behavior is of great practical importance. In this study, I combined chemical composition analysis, macroscopic fracture examination, scanning electron microscopy of the fracture surface, macrostructural inspection, microstructural observation, and hardened case depth measurement to identify the root cause of the fracture. The evidence shows that the failure was a brittle fracture combining transgranular cracking and quasi-cleavage cracking. The hardened case depth, tooth surface hardness, and core hardness all fell short of the specified values. The core microstructure was dominated by free ferrite and lath martensite, which is softer and less load-bearing than acicular martensite. Under the applied torsional and bending loads, the pinion gears therefore fractured prematurely.
Pinion gears are among the most widely used power transmission elements in modern machinery. Compared with belt drives, chain drives, and friction wheel drives, metal pinion gears offer a compact structure, high transmission efficiency, stable operation, accurate transmission ratios, and smooth running. They can be used for both speed reduction and speed increase, and they can operate across a wide range of speeds and powers. At the same time, the meshing teeth of pinion gears are subjected to repeated impact loads, alternating bending stresses, and contact stresses. The common failure modes of pinion gears include tooth breakage, pitting, scuffing, and plastic deformation. Since pinion gears are fundamental components in mechanical transmission systems and are used extensively in the national economy and defense industry, rapid and accurate analysis of pinion gear fracture is essential for safe operation and for continued development of the gear industry.
Experimental Programme
I designed the investigation to cover both material quality and processing quality. The chemical composition was determined to verify that the raw material met the 20CrMnTi specification. Macroscopic and microscopic fracture surfaces were examined to establish the fracture mode and the crack initiation site. Macrostructural and microstructural examinations were performed to evaluate the internal quality of the pinion gears. Finally, case depth and hardness measurements were used to assess the effectiveness of the carburizing and quenching treatment. The test methods and standards I used are summarized in Table 1.
| Test item | Method or standard | Purpose |
|---|---|---|
| Chemical composition | Inductively coupled plasma optical emission spectrometry; high-frequency induction furnace combustion infrared absorption | Verify alloying and impurity elements in the pinion gears |
| Macroscopic fracture morphology | Visual examination and low-magnification imaging | Identify crack origin, fracture direction, and macroscopic mode |
| Fracture surface microscopy | Scanning electron microscopy | Determine transgranular, intergranular, or quasi-cleavage features |
| Macrostructure | Acid etching per GB/T 226-2015; rating per GB/T 1979-2005 | Detect segregation, porosity, cracks, and shrinkage cavities |
| Microstructure | Metallographic preparation per GB/T 13298-2015 | Examine case, transition, and core structures |
| Case depth | GB/T 9450-2005 | Measure carburized hardened case depth |
| Hardness | Microhardness and Rockwell hardness testing | Compare tooth surface and core hardness with requirements |
Chemical Composition Analysis
I cut the end face of a pinion gear and machined a sample for chemical analysis. The alloying elements and impurity elements were measured separately. The results are listed in Table 2, together with the specification limits for 20CrMnTi steel. All measured elements were within the required ranges. In particular, carbon, manganese, chromium, and titanium were all controlled appropriately for a carburized pinion gear. Phosphorus and sulfur were also below the specified maximum levels, so the fracture could not be attributed to improper chemical composition or harmful impurity content.
| Element | Measured value (wt.%) | Specification (wt.%) | Assessment |
|---|---|---|---|
| C | 0.206 | 0.17–0.22 | Conforming |
| Si | 0.20 | ≤ 0.37 | Conforming |
| Mn | 1.05 | 0.80–1.10 | Conforming |
| Cr | 1.21 | 1.00–1.30 | Conforming |
| Ti | 0.064 | 0.04–0.10 | Conforming |
| P | < 0.010 | ≤ 0.030 | Conforming |
| S | 0.015 | ≤ 0.035 | Conforming |
I also calculated the carbon equivalent to assess the hardenability tendency of the material. The carbon equivalent can be expressed as:
$$ CE = C + \frac{Mn}{6} + \frac{Cr + Mo + V}{5} + \frac{Ni + Cu}{15} $$
Using the measured values, I obtained:
$$ CE = 0.206 + \frac{1.05}{6} + \frac{1.21}{5} = 0.206 + 0.175 + 0.242 = 0.623 $$
This value indicates good hardenability, which is consistent with the intended use of 20CrMnTi for carburized pinion gears. The martensite start temperature can be estimated as:
$$ M_s = 561 – 474C – 33Mn – 17Cr – 17Ni – 21Mo $$
Substituting the measured composition gives:
$$ M_s = 561 – 474(0.206) – 33(1.05) – 17(1.21) = 561 – 97.6 – 34.7 – 20.6 = 408.1^\circ C $$
This relatively high martensite start temperature means that the core can transform to martensite over a wide temperature range during quenching. However, if the cooling rate is insufficient or if the carbon content in the core is too low, ferrite may form before or during martensitic transformation. That possibility is important for the present failure because the core microstructure contained a large amount of free ferrite.
Macroscopic Fracture Morphology
The fracture surface of the pinion gear showed clear radial patterns. By tracing the radial lines backward, I located the crack origin at the tooth root region. The fracture propagated from that origin until final separation occurred. The presence of radial marks and the overall orientation of the fracture indicated that the pinion gear had experienced a significant torsional stress component. The fracture surface had a silver-gray metallic luster and showed no obvious plastic deformation. When I rotated the fracture surface under light, I observed small shiny facets, which are characteristic of a macroscopic brittle fracture. These features already suggested that the pinion gears had not undergone substantial plastic deformation before breaking.
Macroscopically, the fracture surface could be divided into three regions: the crack origin region, the fibrous zone, and the final fracture zone. The crack origin was located at the tooth root, where the geometry changes abruptly. In such a location, machining marks or wire-cut traces can further intensify stress concentration. The radial pattern diverged from the origin, and the final fracture zone was located on the opposite side. The absence of necking and the bright faceted appearance confirmed that the pinion gears failed in a brittle manner rather than by ductile overload.
Microscopic Fracture Characteristics
I examined the crack origin region at the tooth root, the fibrous zone, the final fracture zone, and the tooth groove near the origin using scanning electron microscopy. The tooth root fracture surface at the crack origin was not a smooth plane. On both sides of the tooth root, I observed wavy small ledges. At different positions along the tooth root, there was no obvious wear trace, no distinct crack, and no abnormal inclusion defect. The fibrous zone and the final fracture zone exhibited typical transgranular fracture and quasi-cleavage morphology. This confirmed that the fracture mode was brittle.
At the tooth groove near the crack origin, several small cracks were visible. The number of longitudinal cracks was clearly greater than the number of transverse cracks. The tooth groove surface showed some slight wear marks, but no other machining痕迹 and no abnormal inclusions were found. The presence of multiple microcracks in the tooth groove indicates that local stress concentration and surface damage had already developed before final fracture. These microcracks could act as initiation sites for the final brittle fracture of the pinion gears.
The quasi-cleavage feature can be described in terms of the local fracture stress. For a cleavage crack to propagate, the local tensile stress must exceed a critical value:
$$ \sigma_c = \sqrt{\frac{4E\gamma_s}{\pi a}} $$
where E is Young’s modulus, γs is the surface energy, and a is the crack length. A lower core hardness and a shallower case depth both reduce the resistance of the pinion gears to this brittle propagation process. The transgranular path also indicates that the crack did not follow prior austenite grain boundaries; instead, it propagated through the martensitic and ferritic microstructure. This is consistent with a combination of quasi-cleavage and transgranular cracking in a carburized pinion gear.
Macrostructural Examination
I sectioned the fracture surface and prepared a low-magnification specimen. After acid etching, the macrostructure showed a clear hardened case zone, a transition zone, and a matrix zone. No obvious segregation, cracks, shrinkage cavities, or other structural defects were observed. The macrostructure ratings are given in Table 3. Center porosity was rated 1.5, general porosity was rated 0.5, and center segregation was rated 0.0. All values were within the acceptance limit of 2.0. Therefore, the fracture was not caused by a gross metallurgical defect such as severe segregation, porosity, or a shrinkage cavity in the pinion gears.
| Macrostructural feature | Measured rating | Acceptance limit | Assessment |
|---|---|---|---|
| Center porosity | 1.5 | ≤ 2.0 | Conforming |
| General porosity | 0.5 | ≤ 2.0 | Conforming |
| Center segregation | 0.0 | ≤ 2.0 | Conforming |
The presence of a distinct case, transition, and matrix zone is normal for carburized pinion gears. The absence of macrosegregation also means that the chemical composition was uniformly distributed at the macroscopic scale. However, macrostructural acceptance does not guarantee that the case depth and hardness are adequate. The subsequent hardness and case depth measurements showed that the heat treatment was insufficient even though the macrostructure appeared acceptable.
Microstructural Observation
I prepared metallographic specimens from the cross section of the failed tooth and observed the case zone, transition zone, and core zone. The results are summarized in Table 4. In the case zone of the failed tooth, the microstructure consisted of fine acicular martensite and a small amount of retained austenite. In the transition zone, the structure was mainly plate-like or lamellar martensite. In the core, the microstructure consisted of a large amount of free ferrite and lath martensite. This core microstructure is unfavorable because free ferrite is soft and lath martensite is less hard than acicular martensite.
| Zone | Observed microstructure | Desirable microstructure | Assessment |
|---|---|---|---|
| Case | Fine acicular martensite +少量 retained austenite | Fine acicular martensite + dispersed carbides | Generally acceptable but shallow |
| Transition | Plate-like martensite | Transition martensite with gradual hardness change | Acceptable |
| Core | Free ferrite + lath martensite | Lath martensite or low-carbon martensite with minimal ferrite | Unfavorable |
The hardness difference between lath martensite and acicular martensite is significant. In general, the hardness of martensite increases with carbon content and with the degree of carbon supersaturation. Acicular martensite contains more carbon in solid solution and has a higher dislocation density and finer substructure, so it provides higher hardness and better resistance to plastic deformation. Lath martensite, by contrast, forms in low-carbon regions and is relatively soft. When free ferrite is also present, the local hardness drops further. The approximate relationship between hardness and tensile strength for steel can be written as:
$$ \sigma_b \approx 3.3 \, HV $$
For the measured core hardness of 237 HV, 245 HV, and 223 HV, the estimated tensile strengths are approximately 782 MPa, 809 MPa, and 736 MPa. For the required core hardness range of 370–450 HV, the corresponding tensile strength range is approximately 1221–1485 MPa. Therefore, the actual core strength was only about 50–66% of the required level. This large shortfall helps explain why the pinion gears could not carry the applied torsional and bending loads.
The core microstructure also affects the fatigue crack growth resistance. The fatigue crack growth rate can be described by the Paris relationship:
$$ \frac{da}{dN} = C(\Delta K)^m $$
where da/dN is the crack growth rate per cycle, ΔK is the stress intensity factor range, and C and m are material constants. A softer core with free ferrite provides less crack-tip shielding and allows faster crack growth. In the present pinion gears, the combination of a shallow case and a soft core meant that the crack could initiate at the tooth root and propagate rapidly through the core once the local stress exceeded the reduced load-bearing capacity.
Case Depth and Hardness Measurements
I measured the hardened case depth at the root and at the middle of the tooth tip of the failed tooth, and also at the tip of a normal tooth. The measurements followed the method specified in GB/T 9450-2005. The case depth results are listed in Table 5. The failed tooth root values were 0.646 mm, 0.747 mm, and 0.687 mm. The normal tooth tip value was 0.831 mm. All of these values were below the required range of 1.2–1.4 mm. The case depth deficiency was therefore present not only in the failed tooth but also in the normal tooth, indicating a systematic heat treatment problem rather than an isolated local defect.
| Measurement location | Case depth (mm) | Requirement (mm) | Assessment |
|---|---|---|---|
| Failed tooth root, location 1 | 0.646 | 1.2–1.4 | Below specification |
| Failed tooth tip, middle | 0.747 | 1.2–1.4 | Below specification |
| Failed tooth root, location 2 | 0.687 | 1.2–1.4 | Below specification |
| Normal tooth tip, middle | 0.831 | 1.2–1.4 | Below specification |
I also measured the hardness at the core and at the tooth surface. The results are given in Table 6. The core hardness values were 237 HV, 245 HV, and 223 HV, which are below the required range of 370–450 HV, equivalent to 38–45 HRC. The tooth surface hardness values were 698 HV, 703 HV, and 691 HV. These values lie within the specified 620–760 HV range, but they are close to the lower bound of the equivalent 56–62 HRC requirement. Taken together with the shallow case depth, the overall hardening state of the pinion gears was insufficient.
| Location | Measured hardness | Specification | Assessment |
|---|---|---|---|
| Gear core | 237 HV, 245 HV, 223 HV | 370–450 HV (38–45 HRC) | Below specification |
| Tooth surface | 698 HV, 703 HV, 691 HV | 620–760 HV (56–62 HRC) | Low end of range and variable |
The case depth is usually defined as the distance from the surface to the point where the hardness reaches 550 HV. This can be written as:
$$ CHD = x \quad \text{where} \quad HV(x) = 550 \, HV $$
A simple exponential approximation for the hardness profile is:
$$ HV(x) = HV_0 + (HV_s – HV_0)\exp(-kx) $$
where HVs is the surface hardness, HV0 is the core hardness, and k is a decay constant. If the case depth is too small, the effective hardened layer cannot support the bending and contact stresses at the tooth root. The carburizing process can be described by the diffusion equation:
$$ C(x,t) = C_s – (C_s – C_0)\operatorname{erf}\left(\frac{x}{2\sqrt{Dt}}\right) $$
where Cs is the surface carbon potential, C0 is the initial carbon content, D is the diffusion coefficient, and t is time. The effective case depth is proportional to the square root of time and to the square root of the diffusion coefficient. A case depth of 0.646–0.831 mm instead of 1.2–1.4 mm indicates that the carburizing time, temperature, carbon potential, or subsequent diffusion treatment was inadequate. The pinion gears therefore did not receive the intended hardened layer.
The diffusion coefficient itself is temperature-dependent:
$$ D = D_0 \exp\left(-\frac{Q}{RT}\right) $$
where D0 is the pre-exponential factor, Q is the activation energy, R is the gas constant, and T is the absolute temperature. A small deviation in temperature or time can produce a significant change in case depth. For the present pinion gears, the measured case depth corresponded to only about 50–65% of the required value. This is a major deficiency because the case depth controls the bending fatigue strength and the contact fatigue resistance of carburized pinion gears.
Root Geometry and Stress Concentration
The crack origin was located at the tooth root. This is a region where the cross section changes abruptly, so stress concentration is expected. If the transition fillet is too sharp, or if machining marks or wire-cut traces are present, the stress concentration factor increases further. I estimated the stress concentration using a simplified expression for a fillet:
$$ K_t = 1 + \sqrt{\frac{t}{r}} $$
where t is the tooth thickness at the root and r is the fillet radius. A smaller radius produces a larger Kt. For a typical pinion gear tooth, if t is 8 mm and r is 0.8 mm, then:
$$ K_t = 1 + \sqrt{\frac{8}{0.8}} = 1 + \sqrt{10} = 4.16 $$
Such a high stress concentration factor can raise the local stress far above the nominal stress. The nominal torsional shear stress in a shaft-like pinion gear can be approximated by:
$$ \tau_{nom} = \frac{16T}{\pi d^3} $$
For a torque of 500 N·m and a root diameter of 30 mm, the nominal shear stress is:
$$ \tau_{nom} = \frac{16(500)}{\pi (0.030)^3} = \frac{8000}{8.48 \times 10^{-5}} \approx 9.43 \times 10^7 \, Pa = 94.3 \, MPa $$
With a stress concentration factor of 4.16, the local stress would be:
$$ \tau_{max} = K_t \tau_{nom} = 4.16 \times 94.3 \approx 392 \, MPa $$
The core tensile strength estimated from 237 HV is about 782 MPa, but the local stress state is multiaxial and the tooth root also experiences bending stress. The combined effect can easily exceed the local load-bearing capacity of a soft core with free ferrite. The fracture mechanics approach gives additional insight:
$$ K_I = Y\sigma\sqrt{\pi a} $$
where KI is the mode I stress intensity factor, Y is a geometry factor, σ is the applied stress, and a is the crack length. For a brittle steel with a fracture toughness of about 25 MPa·m1/2, an applied stress of 400 MPa, and Y = 1.12, the critical crack size is:
$$ a_c = \frac{1}{\pi}\left(\frac{K_{IC}}{Y\sigma}\right)^2 = \frac{1}{\pi}\left(\frac{25}{1.12 \times 400}\right)^2 \approx 9.9 \times 10^{-4} \, m \approx 1.0 \, mm $$
A critical crack size of about 1 mm is small. This means that once a crack initiates at the tooth root, it can propagate to failure with little additional growth. The multiple microcracks observed in the tooth groove were therefore dangerous because they could readily link up and reach the critical size under the applied torsional and bending loads.
Integrated Discussion
The chemical composition of the pinion gears was fully conforming, and the macrostructure showed no significant segregation, porosity, or shrinkage defects. Therefore, the fracture was not caused by incorrect material selection or by a gross metallurgical defect. The fracture surface showed radial marks, a silver-gray appearance, shiny facets, and no plastic deformation, all of which are characteristic of brittle fracture. Microscopic examination confirmed transgranular fracture and quasi-cleavage cracking. The crack origin was at the tooth root, where stress concentration is highest. These observations establish that the pinion gears failed by brittle fracture initiated at the tooth root.
The proximate cause of the brittle fracture was inadequate heat treatment. The case depth at the failed tooth root was 0.646–0.687 mm, and even the normal tooth tip had a case depth of only 0.831 mm. All values were below the required 1.2–1.4 mm. The core hardness was 223–245 HV, far below the required 370–450 HV. The core microstructure contained a large amount of free ferrite and lath martensite. Free ferrite is soft and provides little resistance to plastic deformation, while lath martensite is less hard than acicular martensite. The combination of a shallow case and a soft core reduced the load-bearing capacity of the pinion gears.
The load-bearing capacity of a carburized pinion gear depends strongly on the case depth and the core hardness. A simple proportionality can be used to illustrate the shortfall:
$$ \frac{\sigma_{allow,actual}}{\sigma_{allow,required}} \approx \frac{CHD_{actual}}{CHD_{required}} = \frac{0.687}{1.30} \approx 0.53 $$
This estimate indicates that the actual load-bearing capacity was only about half of the required capacity. When the pinion gears were subjected to torsional and bending loads, the local stress at the tooth root exceeded the reduced capacity, and a brittle crack initiated. The crack then propagated through the soft core by transgranular and quasi-cleavage mechanisms. The presence of multiple longitudinal microcracks in the tooth groove further reduced the effective cross section and accelerated the final fracture.
It is also important to consider the role of the tooth root transition geometry. The crack origin was at the tooth root, and the root is a natural stress raiser. If the transition fillet is not optimized, or if machining marks remain on the surface, the stress concentration factor can increase significantly. The multiple microcracks observed in the tooth groove suggest that local stress concentration and surface damage had already developed. These microcracks could act as initiation sites for the final fracture. Therefore, both the heat treatment and the root geometry contributed to the failure of the pinion gears.
The pinion gears in this application must withstand alternating bending stress and contact stress. The tooth root experiences the highest bending stress, while the tooth flank experiences contact stress. A carburized case provides a hard, wear-resistant surface and a compressive residual stress that improves fatigue resistance. A tough core provides support for the case and resists crack propagation. When the case is too shallow and the core is too soft, the beneficial compressive residual stress is reduced, and the core cannot support the case. The result is a pinion gear that is prone to bending fatigue, cracking, and brittle fracture. The present failure is consistent with this mechanism.
In summary, the evidence from chemical analysis, fracture morphology, macrostructural examination, microstructural observation, and hardness testing all points to the same conclusion. The pinion gears were made from a conforming 20CrMnTi steel, but the carburizing and quenching treatment did not produce the required case depth or core hardness. The core microstructure contained free ferrite and lath martensite, which are softer and less load-bearing than acicular martensite. The crack initiated at the tooth root, where stress concentration is highest, and propagated in a brittle transgranular and quasi-cleavage mode. The pinion gears therefore fractured under the service loads.
Conclusions
Based on the investigation, I draw the following conclusions:
1. The chemical composition of the pinion gears met the 20CrMnTi specification, and the macrostructure showed no significant segregation, porosity, or shrinkage defects. The fracture was not caused by incorrect material composition or by a gross metallurgical defect.
2. The fracture mode was brittle fracture, combining transgranular cracking and quasi-cleavage cracking. The crack origin was located at the tooth root, where stress concentration is highest. Radial marks and shiny facets on the fracture surface confirmed the brittle nature of the failure.
3. The case depth of the pinion gears was insufficient. The measured values were 0.646 mm, 0.747 mm, and 0.687 mm at the failed tooth and 0.831 mm at the normal tooth, all below the required 1.2–1.4 mm. This indicates a systematic deficiency in the carburizing process.
4. The core hardness was 223–245 HV, below the required 370–450 HV. The core microstructure contained a large amount of free ferrite and lath martensite. These constituents are softer and provide lower load-bearing capacity than acicular martensite.
5. The tooth surface hardness was 691–703 HV, which is at the low end of the specified range and indicates insufficient hardening. The combination of a shallow case, a soft core, and a low tooth surface hardness reduced the load-bearing capacity of the pinion gears.
6. The combined effect of insufficient case depth and low core hardness reduced the load-bearing capacity to approximately half of the required value. Under torsional and bending loads, the pinion gears fractured at the tooth root by a brittle mechanism.
7. The tooth root transition geometry and the presence of multiple microcracks in the tooth groove contributed to the stress concentration and crack initiation. The root fillet should be optimized to avoid sharp notches and machining marks.
Recommendations
To prevent similar failures in pinion gears, I recommend the following actions:
| Area | Recommendation | Expected benefit |
|---|---|---|
| Carburizing process | Increase carburizing time, temperature control, and carbon potential to achieve 1.2–1.4 mm case depth | Improves bending fatigue strength and contact fatigue resistance |
| Quenching process | Improve cooling rate and quench uniformity to eliminate free ferrite and promote martensitic transformation in the core | Raises core hardness to 370–450 HV and increases load-bearing capacity |
| Core microstructure | Avoid free ferrite and lath martensite; target a uniform low-carbon martensite or bainitic structure | Increases toughness and resistance to crack propagation |
| Tooth root geometry | Optimize the root fillet radius and remove machining marks or wire-cut traces | Reduces stress concentration and delays crack initiation |
| Surface integrity | Control surface roughness and introduce compressive residual stress by shot peening if necessary | Improves fatigue life of the pinion gears |
| Quality control | Measure case depth, surface hardness, and core hardness on every batch of pinion gears | Ensures that heat treatment meets specification before service |
If these measures are implemented, the pinion gears should achieve the required case depth, core hardness, and surface hardness. The risk of brittle fracture at the tooth root would then be greatly reduced. Because pinion gears are critical components in mechanical transmissions, improving their heat treatment quality and root geometry is essential for reliable service. The present failure analysis shows that a conforming chemical composition alone is not sufficient; the carburizing and quenching process must also be controlled to produce the correct case depth and core microstructure. Only then can the pinion gears carry the torsional, bending, and contact loads for which they are designed.
In my assessment, the fractured pinion gears failed because the heat treatment did not meet the design requirements. The case was too shallow, the core was too soft, and the core microstructure contained free ferrite and lath martensite. These conditions reduced the load-bearing capacity and made the pinion gears susceptible to brittle fracture at the tooth root. The corrective actions described above should be applied to future production to ensure that the pinion gears meet the required performance and reliability standards.