I investigated a recurring surface-quality problem in a 20CrMnTi steel gear forging line. The problem appeared as small, approximately circular point defects on the surface of gear forging blanks. I observed these defects both on as-forged blanks and after turning. Because a gear forging is expected to have a clean, homogeneous surface before machining, the presence of these point defects raised concerns about raw-material quality, hot deformation practice, and furnace atmosphere. I therefore combined macroscopic observation, chemical composition analysis, scanning electron microscopy, energy-dispersive X-ray spectroscopy, and metallographic examination to identify the origin of the defects. The main finding of my study is that the point defects did not originate from melting, casting, or rolling. Instead, they formed during the heating and hot forging stages. Refractory particles from the heating furnace adhered to the steel surface, and the subsequent gear forging operation pressed these particles into the workpiece. The result was a localized surface pit containing alumina and aluminum-silicon oxide particles, accompanied by oxide scale and slight decarburization. This article presents the evidence, the governing mechanisms, and the process controls that I recommend for preventing similar gear forging surface defects.
Background and Scope
20CrMnTi is a low-carbon, low-alloy structural steel used extensively for gear forging. Its combination of hardenability, toughness, wear resistance, and fatigue resistance makes it suitable for transmission gears, drive shafts, and other dynamically loaded components. In a typical gear forging route, the steel is melted, refined, vacuum treated, continuously cast into square billets, hot charged into a continuous furnace, rolled into round bars, finished, sawed to length, heated in a resistance furnace, die forged into a gear forging blank, and finally turned or machined into the required geometry. Each step can influence surface quality, but the geometry of the observed defects was particularly informative. A point defect that remains nearly equiaxed after hot rolling and appears on both as-forged and turned surfaces is unlikely to be a deformation-induced inclusion stringer. I therefore focused my investigation on the heating and gear forging stages.
The purpose of my examination was not only to identify the defect type but also to establish a causal chain that can be used for corrective action. In gear forging, a surface defect can reduce fatigue life, cause machining rejection, or create a stress concentration under service loading. Even a small exogenous particle pressed into the surface can act as a notch. Therefore, I treated the point defects as a gear forging quality issue requiring both microstructural evidence and process diagnosis.
Gear Forging Process Route and Defect Location
I reconstructed the production route to determine where the defects could have formed. The route includes several high-temperature operations, and each one has a different ability to produce surface and subsurface features. Table 1 summarizes the route, the expected effect on the material, and the relevance to the observed gear forging defects.
| Process stage | Primary thermal or mechanical action | Expected defect signature | Relevance to observed gear forging defect |
|---|---|---|---|
| Electric arc furnace melting | Melting and alloying | Macro-inclusions, segregation, gas porosity | Low, because chemistry and microstructure were acceptable |
| Refining and vacuum treatment | Deoxidation, desulfurization, degassing | Residual oxide inclusions, sulfides | Low, because no typical inclusion stringers were found |
| Continuous casting | Solidification | Surface cracks, corner cracks, oscillation marks | Low, because subsequent rolling would elongate or remove them |
| Hot charging and continuous rolling | Large reduction, longitudinal extension | Elongated inclusions, rolled-in scale, seams | Low to moderate, but observed defects were not elongated |
| Finishing and sawing | Straightening, cutting, surface preparation | Mechanical marks, burrs, saw marks | Low, because defects were approximately circular |
| Resistance furnace heating | Reheating to forging temperature | Scale, decarburization, refractory adhesion | High, because refractory particles were identified |
| Die forging | Compression, flash formation, surface flow | Laps, folds, pressed-in foreign matter | High, because particles were embedded into the surface |
| Turning and machining | Material removal | Tool marks, chatter, machining-induced defects | Low, because defects existed before turning |
I observed point defects on the turned surface and on the unturned gear forging blank. The defects were randomly distributed and did not follow the rolling direction. This random, non-directional distribution was an early indication that the defects formed after continuous rolling. In gear forging, random surface defects are often associated with handling, furnace contact, or die-surface contamination rather than with intrinsic steel cleanliness.
Chemical Composition and Material Verification
I first verified that the 20CrMnTi steel met the required composition. A deviation in carbon, manganese, chromium, or titanium could alter hardenability, forged microstructure, and surface behavior. Table 2 compares the measured composition with the specified range. The measured values were within the standard limits, so I excluded chemical composition as the primary cause of the gear forging point defects.
| Element | Measured mass fraction, % | Specified range, % | Assessment |
|---|---|---|---|
| C | 0.21 | 0.17-0.23 | Conforming |
| Mn | 1.02 | 0.80-1.10 | Conforming |
| Si | 0.28 | 0.17-0.37 | Conforming |
| Cr | 1.18 | 1.00-1.30 | Conforming |
| P | 0.012 | ≤0.030 | Conforming |
| S | 0.003 | ≤0.030 | Conforming |
| Cu | 0.05 | ≤0.30 | Conforming |
| Ni | 0.04 | ≤0.30 | Conforming |
| Mo | 0.01 | ≤0.10 | Conforming |
| Ti | 0.07 | 0.04-0.10 | Conforming |
I also estimated the carbon equivalent to check whether the steel was in a normal hardenability window for gear forging. A common expression is
$$CE = C + \frac{Mn}{6} + \frac{Cr + Mo + V}{5} + \frac{Ni + Cu}{15}$$
Using the measured values and neglecting V, I obtained
$$CE \approx 0.21 + \frac{1.02}{6} + \frac{1.18 + 0.01}{5} + \frac{0.04 + 0.05}{15} \approx 0.62$$
This value is consistent with a low-alloy carburizing steel and does not indicate an abnormal composition that would produce surface point defects during gear forging. The defect therefore had to be exogenous or process-related.
Macroscopic and Microscopic Observations
I examined the surface of the gear forging blanks before and after turning. On the turned surface, the defects appeared as distinct circular spots with clear boundaries against the matrix. On the unturned surface, the defects had a different color from the matrix, but the boundary was less sharp. I interpreted this difference as follows: turning removed the deformed surface layer and exposed the embedded particle and its surrounding oxide, making the boundary more distinct. Before turning, the defect was partly covered by a thin deformed and oxidized layer, so the contrast was more diffuse.
Under scanning electron microscopy, the turned gear forging surface showed a clear separation between the defect and the matrix. The matrix was bright, while the defect contained dark gray and black foreign particles. The unturned surface showed a less distinct boundary, with a black-gray transition zone and localized charging that produced a bright central region. Despite these surface differences, the internal morphology of the defect was the same before and after turning. This confirmed that the defect was present in the gear forging blank and was not created by the turning operation.
I further magnified the defect and found two main types of dark particles. Energy-dispersive X-ray spectroscopy identified them as aluminum oxide and aluminum-silicon oxide. These phases are typical of refractory materials used in high-temperature furnaces. Their presence in a surface pit on a gear forging is strong evidence of an external source.
| Observation location | Morphology | Boundary character | Dominant phases | Interpretation |
|---|---|---|---|---|
| Turned gear forging surface | Circular spot, distinct contrast | Clear and irregular | Alumina, aluminum-silicon oxide, iron oxide | Embedded refractory particle exposed by machining |
| Unturned gear forging surface | Circular spot, diffuse contrast | Gradual | Alumina, aluminum-silicon oxide, iron oxide | Particle covered by deformed and oxidized surface layer |
| Furnace refractory debris | Angular and granular particles | Not applicable | Alumina, aluminum-silicon oxide | Source material matching the defect particles |
| Defect cross-section | U-shaped pit | Sharp matrix-pit interface | Iron oxide, alumina, aluminum-silicon oxide | Foreign particle pressed into the surface during gear forging |
Energy-Dispersive X-ray Spectroscopy and Phase Identification
The energy-dispersive X-ray spectroscopy results were decisive. The dark particles in the defect contained aluminum and oxygen, or aluminum, silicon, and oxygen. The first phase is consistent with alumina,
$$\mathrm{Al_2O_3}$$
and the second is consistent with a mullite-type aluminum-silicon oxide, commonly represented as
$$3\mathrm{Al_2O_3} \cdot 2\mathrm{SiO_2} \rightarrow \mathrm{Al_6Si_2O_{13}}$$
These compounds are common constituents of refractory bricks, castables, and furnace linings. I then examined debris collected from the resistance heating furnace. The debris contained the same two phase types. This match provided a direct link between the furnace lining and the gear forging surface defects. The defect was therefore not a metallurgical inclusion formed during solidification; it was a refractory fragment transferred to the steel surface and embedded during hot deformation.
| Phase family | Detected elements | Likely compound | Typical source | Role in the gear forging defect |
|---|---|---|---|---|
| Alumina | Al, O | Al2O3 | Refractory lining, castable, brick | Hard particle pressed into the surface |
| Mullite-type oxide | Al, Si, O | Al6Si2O13 | Aluminum-silicate refractory | Hard particle and brittle fragment in the pit |
| Iron oxide | Fe, O | FeO, Fe3O4, Fe2O3 | High-temperature oxidation of steel | Scale layer separating particle from matrix |
| Matrix | Fe, Mn, Cr, Ti, C | Low-alloy steel | 20CrMnTi gear forging stock | Deformed base material around the pit |
Cross-Sectional Evidence of Embedment
I sectioned the defect and examined the cross-section under scanning electron microscopy. The defect appeared as a U-shaped pit. The pit boundary was sharp, and a thin iron oxide layer separated the foreign particles from the matrix. I did not observe a particle stringer or a cluster of typical intrinsic inclusions. Instead, I observed a mechanically depressed surface feature filled with refractory fragments and oxidation products. The matrix adjacent to the pit showed slight deformation, which is consistent with local plastic flow around a hard particle during gear forging.
The presence of iron oxide inside the pit is important. It indicates that the particle adhered to the steel surface at high temperature, where oxidation was active, and that the particle and scale were subsequently pressed into the surface. The oxide layer did not form after forging as a simple surface scale; it was trapped between the refractory particle and the steel matrix. This entrapment is a characteristic signature of a foreign particle embedded during hot working.
| Cross-sectional feature | Observed condition | Physical interpretation |
|---|---|---|
| Pit geometry | U-shaped depression | Local indentation by a hard particle |
| Matrix-particle interface | Thin iron oxide layer | High-temperature oxidation before or during embedment |
| Particle distribution | Discrete fragments inside pit | Brittle refractory debris pressed into the surface |
| Matrix deformation | Slight local flow | Plastic deformation during gear forging |
| Inclusion stringers | Not observed | Unlikely to be an intrinsic steel inclusion |
| Rolling directionality | Not observed | Defect formed after continuous rolling |
The absence of rolling directionality can be expressed quantitatively by the aspect ratio of the defect:
$$\lambda = \frac{L}{D}$$
where $$L$$ is the length along the expected rolling direction and $$D$$ is the transverse dimension. For a rolled-in inclusion or a seam, $$\lambda$$ is typically much greater than 1. For the observed gear forging defects, $$\lambda \approx 1$$. This supports the conclusion that the defect formed during or after the final gear forging operation rather than during continuous rolling.
Metallographic Examination and Decarburization
After etching with nitric acid in ethanol, I examined the defect edge under an optical microscope. The matrix near the pit showed a ferrite-pearlite structure. Adjacent to the pit, I observed decarburization, with a higher ferrite fraction than in the bulk matrix. This decarburization indicates that the surface was exposed to a carbon-depleting atmosphere at elevated temperature. It also supports the conclusion that the particle adhered before or during reheating, not during final machining.
Carbon diffusion during heating can be described by the one-dimensional diffusion equation:
$$\frac{\partial C}{\partial t} = D \frac{\partial^2 C}{\partial x^2}$$
For a semi-infinite solid with a constant surface carbon concentration, the solution is
$$C(x,t) = C_s + (C_0 – C_s) \operatorname{erf}\left( \frac{x}{2\sqrt{Dt}} \right)$$
where $$C_0$$ is the initial carbon content, $$C_s$$ is the surface carbon content, $$D$$ is the carbon diffusion coefficient, $$x$$ is depth, and $$t$$ is time. The temperature dependence of $$D$$ follows an Arrhenius relation:
$$D = D_0 \exp\left( -\frac{Q}{RT} \right)$$
The presence of a decarburized layer at the pit edge shows that the local surface was not protected from the furnace atmosphere. A refractory particle attached to the surface can create a local shadow or crevice, but it does not prevent carbon diffusion around the pit. The resulting ferrite-enriched zone is consistent with a gear forging surface that experienced high-temperature exposure and subsequent deformation.
| Metallographic feature | Observation | Interpretation |
|---|---|---|
| Bulk microstructure | Ferrite and pearlite | Normal low-carbon alloy steel structure |
| Pit-edge microstructure | Increased ferrite fraction | Decarburization during heating |
| Oxide layer | Thin iron oxide film | High-temperature oxidation |
| Foreign particles | Alumina and aluminum-silicon oxide | Refractory contamination |
| Deformation pattern | Local flow without stringers | Particle pressed in during gear forging |
Why the Defects Did Not Originate from Rolling or Machining
I considered several alternative explanations before concluding that the defects were caused by refractory embedment during gear forging. The first alternative was that the defects were intrinsic nonmetallic inclusions. Intrinsic inclusions formed during melting or casting would normally be distributed in the interior or along deformation bands. After continuous rolling, they would be elongated. The observed defects were surface-connected, nearly circular, and composed of alumina and aluminum-silicon oxide, which are not the typical deoxidation products expected in this steel at such high local concentrations. The second alternative was that the defects were rolled-in scale from the continuous rolling process. Rolled-in scale usually appears as elongated dark bands or seams, often with iron oxides and base-metal deformation. The observed defects lacked elongation and contained refractory phases. The third alternative was that the defects were created by turning. However, I found identical defects on unturned gear forging blanks. This excluded machining as the origin.
I also considered die-related defects. A damaged die can create laps, folds, or indentations, but it would not deposit alumina and aluminum-silicon oxide particles. A die contaminant could transfer foreign matter, but the particle chemistry matched the furnace refractory, not the die material. The most consistent explanation is therefore the following chain: the resistance furnace lining degraded; refractory fragments spalled or detached; some fragments adhered to the steel surface during heating; the hot billet was transferred to the forging press; and the die forging operation pressed the fragments into the surface, forming point defects.
| Candidate cause | Expected evidence | Observed evidence | Assessment |
|---|---|---|---|
| Intrinsic inclusion | Interior location, elongated after rolling, typical oxide/sulfide chemistry | Surface pit, equiaxed, alumina and mullite | Rejected |
| Rolled-in scale | Elongated bands, iron oxide-rich, aligned with rolling direction | Circular spots, refractory particles, no alignment | Rejected |
| Machining damage | Only on machined surfaces, tool marks, no foreign particles | Present before turning, foreign particles present | Rejected |
| Die damage | Laps, folds, die-material transfer, repeated location | Random location, furnace refractory chemistry | Rejected |
| Furnace refractory embedment | Alumina and mullite particles, oxide layer, decarburization, local deformation | All observed | Accepted |
Mechanism of Refractory Embedment during Gear Forging
I reconstructed the mechanism in six stages. First, the 20CrMnTi billet was reheated in the resistance furnace to the gear forging temperature. Second, the furnace refractory lining experienced thermal cycling, mechanical vibration, and chemical attack by scale and furnace atmosphere. Third, fragments of alumina and aluminum-silicon refractory detached from the lining. Fourth, some fragments fell or were carried onto the steel surface and adhered to the hot billet. Fifth, the billet was transferred to the die forging station, where the surface oxide and the refractory fragments were compressed against the steel. Sixth, the particles were pressed into the surface, creating a U-shaped pit with trapped oxide and refractory debris. Later turning exposed the pit more clearly.
The adhesion and embedding processes can be described by contact mechanics. When a refractory particle of radius $$R$$ is pressed against a hot steel surface by a force $$F$$, the contact radius $$a$$ and maximum contact pressure $$p_0$$ are approximately
$$a = \left( \frac{3FR}{4E^*} \right)^{1/3}$$
and
$$p_0 = \frac{3F}{2\pi a^2}$$
where $$E^*$$ is the effective elastic modulus. Embedding occurs when the contact pressure exceeds the local flow stress or hardness of the steel:
$$p_0 \ge H$$
or
$$p_0 \ge \sigma_y$$
At gear forging temperatures, the flow stress of 20CrMnTi is reduced, but the refractory particle remains hard and brittle. Therefore, even a moderate forging pressure can indent the steel surface. The pit depth depends on particle size, forging load, temperature, and local strain. A larger particle or higher local pressure produces a deeper pit. The observed U-shaped geometry is consistent with this indentation process.
Thermal stress in the furnace lining also contributes to refractory spalling. For a brittle refractory layer subjected to a temperature change $$\Delta T$$, the thermal stress can be approximated as
$$\sigma_{th} = \frac{E \alpha \Delta T}{1-\nu}$$
where $$E$$ is Young’s modulus, $$\alpha$$ is the coefficient of thermal expansion, and $$\nu$$ is Poisson’s ratio. If
$$\sigma_{th} \ge \sigma_f$$
where $$\sigma_f$$ is the fracture strength of the refractory, cracks form and fragments detach. This explains why the furnace lining supplied the foreign particles found in the gear forging defects.
Oxidation of the steel surface is another important factor. At high temperature, iron oxide grows according to a parabolic law:
$$x^2 = k_p t$$
where $$x$$ is oxide thickness, $$k_p$$ is the parabolic rate constant, and $$t$$ is time. The oxide layer can act as a bonding layer between the refractory particle and the steel. During gear forging, the oxide and particle are pressed together into the surface. This is why I observed a thin iron oxide film between the refractory fragments and the matrix.
| Stage | Physical event | Controlling variable | Result |
|---|---|---|---|
| Heating | Billet reaches forging temperature | Temperature, time, atmosphere | Scale growth and surface softening |
| Refractory degradation | Lining cracks and spalls | Thermal cycling, vibration, chemical attack | Alumina and mullite fragments |
| Particle transfer | Fragments contact billet surface | Furnace geometry, billet motion, draft | Particles adhere to hot steel |
| Transfer to press | Billet moves to die | Handling time, temperature loss | Particles remain on surface |
| Gear forging | Compression under die load | Forging pressure, strain, friction | Particles pressed into surface |
| Cooling and machining | Scale and deformed layer form | Cooling rate, machining depth | Point defects exposed |
Quantitative Interpretation of the Defect Signature
I used several quantitative relationships to support the diagnosis. The defect aspect ratio showed no rolling elongation. The cross-sectional geometry showed a U-shaped pit rather than an internal crack. The energy-dispersive X-ray spectroscopy results showed refractory chemistry rather than steelmaking inclusion chemistry. The metallographic results showed decarburization and oxidation. Together, these observations form a consistent signature. Table 8 summarizes the key equations and their diagnostic meaning.
| Equation | Expression | Diagnostic use |
|---|---|---|
| Heat conduction | $$\rho c_p \frac{\partial T}{\partial t} = \nabla \cdot (k \nabla T) + Q$$ | Describes billet heating and temperature gradients |
| Thermal stress | $$\sigma_{th} = \frac{E \alpha \Delta T}{1-\nu}$$ | Explains refractory spalling |
| Hertz contact | $$p_0 = \frac{3F}{2\pi a^2}$$ | Estimates contact pressure during embedding |
| Embedding criterion | $$p_0 \ge H$$ | Predicts whether a particle indents the surface |
| Oxide growth | $$x^2 = k_p t$$ | Explains the iron oxide layer around the particle |
| Decarburization | $$C(x,t) = C_s + (C_0 – C_s)\operatorname{erf}\left(\frac{x}{2\sqrt{Dt}}\right)$$ | Explains the ferrite-enriched pit edge |
| Diffusion coefficient | $$D = D_0 \exp\left(-\frac{Q}{RT}\right)$$ | Shows temperature dependence of carbon diffusion |
| Aspect ratio | $$\lambda = \frac{L}{D}$$ | Distinguishes rolled defects from post-rolling defects |
The contact-pressure criterion is particularly useful for gear forging. At elevated temperature, the steel flow stress decreases, so the same die pressure produces more plastic strain. A hard refractory particle acts as a rigid indenter. The local pressure under the particle can exceed the yield strength of the steel, causing the particle to sink into the surface. The surrounding steel flows around the particle, creating a pit with raised or deformed edges. Later machining removes the raised edge and exposes the pit as a circular point defect.
Influence of Gear Forging Parameters
I evaluated how gear forging parameters influence the formation and severity of the defect. Furnace temperature and time affect scale thickness and decarburization. A long residence time increases the probability that a refractory fragment will adhere to the billet. Forging temperature affects flow stress. Higher temperature lowers flow stress, making indentation easier. Forging pressure and strain affect the depth of embedment. Friction affects surface flow and can drag particles along the surface, but the observed defects were not elongated, so the particles were probably pressed in place rather than smeared. Cooling rate affects the final microstructure around the pit but does not change the fact that the particle is exogenous.
| Gear forging variable | Effect on defect formation | Practical control |
|---|---|---|
| Heating temperature | Higher temperature increases oxidation and lowers flow stress | Maintain specified forging temperature range |
| Heating time | Longer time increases scale and refractory contact opportunity | Minimize residence time |
| Furnace atmosphere | Oxidizing atmosphere promotes scale and decarburization | Control air-fuel ratio or use protective atmosphere |
| Refractory condition | Cracks and spalling supply particles | Inspect and repair lining regularly |
| Billet surface condition | Rough or scaled surfaces retain particles | Clean billets before heating when possible |
| Transfer time | Long transfer allows cooling and oxidation | Reduce transfer time to press |
| Forging pressure | Higher pressure increases embedding depth | Optimize die design and load |
| Die surface condition | Rough dies can trap particles | Maintain die surface and lubrication |
| Cooling practice | Affects oxide and microstructure around pit | Use controlled cooling |
I also considered the local stress state during gear forging. The equivalent stress can be expressed as
$$\sigma_{eq} = \sqrt{\frac{1}{2}\left[(\sigma_1-\sigma_2)^2 + (\sigma_2-\sigma_3)^2 + (\sigma_3-\sigma_1)^2\right]}$$
The equivalent plastic strain increment is
$$d\varepsilon_{eq}^p = \sqrt{\frac{2}{3} d\varepsilon_{ij}^p d\varepsilon_{ij}^p}$$
Near a hard particle, the local strain can be much higher than the nominal strain. This local concentration promotes indentation and void formation around the particle. If the particle is brittle, it may fracture into smaller fragments, as observed in the defect. The presence of multiple fragments in the pit supports this interpretation.
Comparison with Typical Gear Forging Surface Defects
To place the observed defects in context, I compared them with other common gear forging surface defects. Table 10 lists the distinguishing features.
| Defect type | Typical appearance | Location | Chemistry | Relationship to gear forging |
|---|---|---|---|---|
| Intrinsic inclusion | Elongated stringer or cluster | Interior or surface | Oxides, sulfides, nitrides | Formed before gear forging, deformed during rolling |
| Rolled-in scale | Dark elongated seam | Surface-connected | Iron oxides | Formed during hot rolling |
| Machining mark | Regular tool pattern | Machined surface | No foreign phase | Formed after gear forging |
| Die lap or fold | Irregular fold, overlap | Die-flow dependent | Base metal, oxide | Formed during gear forging |
| Refractory embedment | Circular pit with dark particles | Random surface location | Alumina, mullite, iron oxide | Formed during heating and gear forging |
The refractory embedment defect is distinctive because it combines a surface pit, a foreign refractory phase, an oxide layer, and local decarburization. No single feature alone would be conclusive, but together they define a clear process signature. In my assessment, the gear forging operation did not create the refractory particles; it only embedded them. The ultimate source was the heating furnace lining.
Preventive and Corrective Actions
Based on the evidence, I recommend a set of preventive actions focused on furnace maintenance, billet cleanliness, and gear forging process control. The most direct action is to inspect the resistance heating furnace regularly and repair or replace damaged refractory lining. Loose or spalled refractory material should be removed before it can contact the billets. The furnace atmosphere should be controlled to reduce excessive scale, because scale can act as a bonding layer between refractory particles and the steel. Billet handling should minimize contact with furnace walls and hearth debris. If possible, billets should be cleaned or descaled before entering the final heating furnace. The transfer time from furnace to press should be short and consistent. Dies should be maintained and cleaned to avoid secondary contamination. Finally, surface inspection should be performed on as-forged gear forging blanks, not only after machining, because early detection prevents defective parts from advancing.
| Action | Target | Frequency | Expected benefit |
|---|---|---|---|
| Inspect furnace lining | Resistance heating furnace | Each maintenance shutdown | Detect cracks and spalling before particle release |
| Remove loose refractory | Furnace hearth and walls | Weekly or as needed | Eliminate particle source |
| Control furnace atmosphere | Heating zone | Continuous monitoring | Reduce scale and decarburization |
| Clean billets before heating | Incoming gear forging stock | Each lot | Reduce surface contamination |
| Reduce furnace-to-press time | Transfer path | Continuous | Lower oxidation and particle adhesion |
| Maintain dies | Forging dies | Per production schedule | Prevent laps and secondary embedment |
| Inspect as-forged surfaces | Gear forging blanks | Each shift | Early detection and containment |
| Audit refractory quality | Furnace lining material | Per campaign | Improve thermal shock resistance |
I also recommend a more detailed acceptance criterion for gear forging surface quality. A point defect should be evaluated by location, size, depth, and chemistry. If a defect contains alumina or mullite, it should be classified as exogenous refractory embedment. If it is elongated and contains iron oxide, it is more likely to be rolled-in scale. If it is internal and elongated, it is more likely to be an intrinsic inclusion. This classification helps maintenance and quality teams assign the correct corrective action.
| Inspection method | Feature detected | Decision value |
|---|---|---|
| Visual inspection | Circular spots, color contrast | Initial screening of gear forging blanks |
| Scanning electron microscopy | Particle morphology and pit geometry | Distinguishes surface pit from inclusion |
| Energy-dispersive X-ray spectroscopy | Al, Si, O, Fe, Mn, Cr | Identifies refractory versus steelmaking phases |
| Metallography | Decarburization, oxide layer, microstructure | Confirms thermal history |
| Cross-section analysis | U-shaped pit, trapped oxide, local flow | Confirms embedment during gear forging |
| Aspect-ratio measurement | Rolling directionality | Separates pre-rolling and post-rolling defects |
Process Control Window for Gear Forging
A robust gear forging process should keep the billet surface clean and the furnace environment stable. I summarized the key control variables in Table 13. The specific numerical limits depend on the furnace design and product specification, but the logic is general: minimize time at high temperature, minimize scale, prevent refractory contact, and maintain consistent forging conditions.
| Control variable | Preferred condition | Reason |
|---|---|---|
| Furnace lining integrity | No cracks, spalling, or loose material | Eliminates refractory particle source |
| Furnace atmosphere | Controlled oxidizing or protective as required | Controls scale and decarburization |
| Heating time | Minimum required for uniform temperature | Reduces oxidation and particle adhesion |
| Billet surface | Clean and free of loose debris | Prevents trapped foreign matter |
| Transfer time | Short and repeatable | Maintains forging temperature and limits scale |
| Die condition | Smooth, clean, properly lubricated | Prevents laps and particle smearing |
| Forging load | Within validated range | Controls embedding depth |
| Cooling | Controlled after gear forging | Prevents additional oxidation and cracking |
Discussion of Root Cause and Responsibility
The root cause of the point defects was not the 20CrMnTi steel itself. The steel chemistry was conforming, and the defect chemistry was foreign to the steelmaking process. The root cause was a furnace-lining condition that allowed refractory fragments to contact the billet. The gear forging operation was the mechanism that embedded the fragments, but it was not the source. This distinction is important. If the problem were misclassified as an intrinsic inclusion, corrective action would focus on steel cleanliness, which would not solve the issue. If it were misclassified as a machining defect, action would focus on tooling, which would also be ineffective. The correct action is furnace maintenance and surface protection before gear forging.
I also note that the defect was detectable before final machining. As-forged inspection could have identified the circular spots earlier. A simple visual inspection after gear forging, followed by scanning electron microscopy and energy-dispersive X-ray spectroscopy when needed, can distinguish refractory embedment from other surface defects. This early detection is valuable because it prevents defective gear forging blanks from consuming machining capacity and reduces the risk of shipping a compromised component.
Implications for Gear Forging Quality Systems
For a gear forging quality system, the findings suggest several changes. First, the furnace should be treated as a critical process equipment item, not merely a heating device. Its refractory condition directly affects surface quality. Second, surface cleanliness should be specified for incoming billets and for billets after heating. Third, defect classification should include chemistry-based rules. A circular surface pit with alumina or mullite should trigger a furnace inspection. Fourth, maintenance records should be linked to quality data, so that refractory repairs can be correlated with defect rates. Fifth, operators should be trained to recognize the difference between a scale spot, a machining mark, and a refractory embedment. These steps are practical and can be implemented without major capital investment.
In terms of gear forging process capability, the defect can be modeled as a random event. If the probability of a refractory particle landing on a billet surface is $$p$$, and the probability that it remains until forging is $$q$$, then the probability of at least one embedded particle on a billet is
$$P_{defect} = 1 – (1 – p q)^N$$
where $$N$$ is the number of potential particle contacts. Reducing $$p$$ by maintaining the furnace lining and reducing $$q$$ by shortening transfer time both lower $$P_{defect}$$. This simple model supports the recommended actions: eliminate the source and reduce the opportunity for adhesion.
Additional Metallurgical Considerations
I also considered whether the titanium content in 20CrMnTi could influence the defect. Titanium is added to control grain size and improve toughness. Titanium carbonitrides can form in the steel, but they are typically fine and distributed. They do not produce large surface pits with alumina and mullite. The observed particles were much larger and chemically distinct. Therefore, titanium addition was not the cause. Similarly, chromium and manganese affect hardenability and oxidation behavior, but they do not produce exogenous refractory particles. The defect was a surface contamination phenomenon.
The presence of decarburization at the pit edge is also relevant to gear forging performance. A decarburized surface layer has lower hardness and fatigue resistance. If it remains after machining, it can reduce contact fatigue life. Therefore, the defect is not only a cosmetic issue. The combination of a pit, an oxide layer, and decarburization can create a local weak zone. For critical gear forgings, such defects should be removed or the affected parts should be rejected. The depth of decarburization can be estimated from the carbon diffusion equation if the time and temperature are known. In practice, metallographic measurement is more reliable.
Summary of Evidence
Table 14 condenses the evidence and the corresponding conclusions.
| Evidence | Observation | Conclusion |
|---|---|---|
| Chemical composition | Within specification | Steel grade is correct |
| Defect shape | Circular, equiaxed | No rolling elongation |
| Defect location | Random surface spots | Not a rolled-in stringer |
| Defect chemistry | Alumina and mullite | Refractory origin |
| Furnace debris | Same phases as defect | Furnace lining is the source |
| Cross-section | U-shaped pit, oxide layer | Particle pressed into surface |
| Metallography | Decarburization at pit edge | High-temperature exposure before forging |
| Machining comparison | Defects on unturned blanks | Not caused by turning |
Conclusions
I conclude that the point defects on the 20CrMnTi gear forging blanks were caused by refractory particles from the resistance heating furnace. The particles adhered to the steel surface during heating and were pressed into the surface during gear forging. The resulting defects contained alumina and aluminum-silicon oxide, with a thin iron oxide layer and local decarburization. The defects were not intrinsic inclusions, rolled-in scale, or machining damage. The gear forging operation was the embedding mechanism, but the source was the furnace lining.
To prevent recurrence, I recommend regular inspection and maintenance of the resistance heating furnace, removal of loose refractory material, control of the furnace atmosphere, minimization of heating time, cleaning of billets before heating, reduction of furnace-to-press transfer time, and maintenance of clean forging dies. I also recommend surface inspection of as-forged gear forging blanks and chemistry-based defect classification. These measures will reduce the probability of refractory embedment and improve the surface quality of 20CrMnTi gear forgings.
From a broader perspective, this case shows that gear forging quality depends on the entire thermal-mechanical chain. A conforming steel composition is necessary but not sufficient. The furnace, the atmosphere, the transfer path, and the die all participate in defining the final gear forging surface. When a surface defect appears, the most reliable approach is to combine macroscopic observation, chemical analysis, scanning electron microscopy, energy-dispersive X-ray spectroscopy, and metallography. In this case, that approach identified the root cause and provided a clear path for corrective action. The same approach can be applied to other gear forging defects where the source is not obvious from visual inspection alone.
Finally, I emphasize that the defect signature is repeatable and diagnosable. A circular surface pit containing alumina or mullite, with an oxide layer and decarburization, should be treated as a refractory embedment defect in gear forging. The presence of such a signature should trigger a furnace-lining audit and a review of billet handling. If these actions are taken promptly, the defect can be eliminated, and the gear forging process can return to a stable, high-quality condition.