I investigated a batch of 20CrMnTi forged gear blanks that showed small spot-like surface defects during turning. The same type of defects was also visible on some forged gear blanks before turning, which immediately indicated that the spots were not created by the turning operation itself. My aim was to identify the origin of the spots, establish the stage of the manufacturing route at which they formed, distinguish exogenous contamination from metallurgical inclusions, and propose practical controls for future production. The production route of the forged gear blanks included electric-arc furnace melting, refining, vacuum treatment, continuous casting into square billets, hot charging to a continuous reheating furnace, continuous rolling into round bars, finishing, sawing, heating in a resistance furnace, die forging, and final turning. Because the defects appeared as approximately circular spots on forged gear blanks and because their chemistry and cross-sectional morphology did not match ordinary steelmaking inclusions, I focused on the heating and hot-forging stages.
Investigation route and first-person observations
I examined the forged gear blanks in both the as-forged and turned conditions. In the turned condition, the spots had sharper boundaries against the steel matrix. In the as-forged condition, the same spots appeared mainly as color differences, with less distinct edges. This difference is important: turning removed the deformed surface layer and exposed the embedded material more clearly, but it did not create the spots. The defect distribution was random over the inspected surfaces, and I did not observe a preferred direction that would correspond to the rolling direction of the bar stock. The spots remained approximately equiaxed rather than elongated. This observation was one of the first indicators that the defects formed after continuous rolling, most probably during heating before hot forging or during the hot-forging operation itself.
I used macroscopic observation, chemical composition analysis, scanning electron microscopy, energy-dispersive X-ray spectroscopy, metallographic examination, and direct inspection of the resistance heating furnace. I also compared the morphology and composition of the surface spots with refractory particles collected from the furnace. The evidence chain was consistent: refractory particles detached from the furnace lining, adhered to the surface of the forged gear blanks during heating, and were pressed into the steel during hot forging. The resulting surface discontinuities are best described as exogenous refractory entrapment defects rather than intrinsic steelmaking inclusions.
| Method | Sample condition | Primary observation | Interpretation |
|---|---|---|---|
| Macroscopic examination | Turned and unturned forged gear blanks | Approximately circular spots; sharper boundaries after turning; color contrast before turning | Defects existed before turning and were exposed more clearly by machining |
| Chemical composition analysis | Representative forged gear blanks | All measured elements within the specified range for 20CrMnTi | No evidence of a chemistry-related origin |
| Scanning electron microscopy | Surface defects before and after turning | Irregular dark-grey particles in a shallow surface depression | Exogenous particulate matter was present at the defect sites |
| Energy-dispersive X-ray spectroscopy | Surface defects and cross-sections | Al2O3 and Al-Si oxide phases; mullite-like Al6Si2O13 signatures | Refractory material is the dominant contaminant |
| Metallography | Cross-section through defect | U-shaped depression; oxide layer; decarburization near the matrix edge | High-temperature exposure and mechanical embedding occurred at the surface |
| Furnace inspection | Resistance heating furnace lining | Spalled and broken refractory particles | The furnace lining is the likely particle source |
Chemical composition and baseline metallurgical quality
I first verified that the forged gear blanks complied with the required 20CrMnTi composition. The measured carbon, manganese, silicon, chromium, phosphorus, sulfur, copper, nickel, molybdenum, and titanium contents were all within the specified ranges. In particular, phosphorus and sulfur were low, and no abnormal segregation pattern was found in the examined sections. This eliminated a simple composition error as the cause of the spots. If the spots had been caused by a gross chemistry deviation, I would expect either a systematic distribution of defects, abnormal banding, or a clear relationship between the spot locations and segregation. None of these features was present. The forged gear blanks therefore had a sound chemical basis, and the defect mechanism had to be sought in the thermomechanical and environmental history of the surface.
| Element | Measured mass fraction, % | Specified range, % | Compliance judgment |
|---|---|---|---|
| C | 0.21 | 0.17-0.23 | Within specification |
| Mn | 1.02 | 0.80-1.10 | Within specification |
| Si | 0.28 | 0.17-0.37 | Within specification |
| Cr | 1.18 | 1.00-1.30 | Within specification |
| P | 0.012 | Maximum 0.030 | Within specification |
| S | 0.003 | Maximum 0.030 | Within specification |
| Cu | 0.05 | Maximum 0.30 | Within specification |
| Ni | 0.04 | Maximum 0.30 | Within specification |
| Mo | 0.01 | Maximum 0.10 | Within specification |
| Ti | 0.07 | 0.04-0.10 | Within specification |
To express the composition compliance in a compact form, I used a normalized position index for each element. For an element with measured value \(x_i\), lower limit \(L_i\), and upper limit \(U_i\), the normalized position is
$$
C_i=\frac{x_i-L_i}{U_i-L_i}
$$
where \(0 \leq C_i \leq 1\) indicates compliance. For the forged gear blanks examined here, every element satisfied this condition. I also calculated a simple distance-to-limit ratio to check whether any element was close to the boundary in a way that could promote abnormal surface behavior:
$$
R_i=\min\left(\frac{x_i-L_i}{U_i-L_i},\frac{U_i-x_i}{U_i-L_i}\right)
$$
The smallest values of \(R_i\) were still sufficiently far from zero for the intended steel grade, and no correlation was found between the minor chemistry variation and the spatial distribution of the spots on the forged gear blanks. This reinforced the conclusion that the spots were exogenous surface contamination rather than a composition-driven metallurgical defect.
Macroscopic morphology of the spots
The spots on the turned forged gear blanks were approximately circular and had a clear boundary with the steel matrix. On the unturned forged gear blanks, the same defects appeared as localized color differences with a less obvious boundary. This is exactly the behavior expected when a hard or refractory particle is pressed into a hot surface and then partly covered by oxide or deformed metal. Turning removes the outermost deformed layer and makes the embedded particle and the surrounding cavity easier to distinguish. The spots were not aligned with the bar axis, and they did not form long stringers. This lack of directionality is critical because inclusions formed before or during rolling are normally elongated along the rolling direction. The spots on these forged gear blanks remained compact and approximately equiaxed, which points to a post-rolling origin.
I quantified the shape of the spots using the circularity parameter
$$
\Psi=\frac{4\pi A}{P^2}
$$
where \(A\) is the projected spot area and \(P\) is its perimeter. A value of \(\Psi\) close to 1 indicates a circular shape, while a low value indicates an elongated or irregular shape. The spots on the forged gear blanks were much closer to circular than typical rolled inclusion stringers. I also estimated the surface defect density as
$$
\lambda=\frac{N}{A_{\mathrm{inspected}}}
$$
where \(N\) is the number of spots in the inspected area. The distribution was random, and no banded or directional pattern was observed. For production control, both \(\Psi\) and \(\lambda\) are useful because they can be measured on turned surfaces and compared with an acceptance baseline.
| Feature | Observation on forged gear blanks | Expected for rolled inclusion | Expected for machining damage | Diagnostic meaning |
|---|---|---|---|---|
| Shape | Approximately circular | Usually elongated along rolling direction | Tool marks, scratches, or periodic pattern | Supports post-rolling foreign-particle entrapment |
| Directionality | No preferred direction | Strong directionality | Related to tool feed or chatter | Does not support rolling-origin inclusions |
| Presence before turning | Yes, visible on unturned forged gear blanks | May be exposed after turning | No, created during turning | Excludes turning as the root cause |
| Boundary after turning | Sharp and irregular | Usually smooth and elongated | Cutting marks or smeared metal | Supports an embedded particulate phase |
| Color contrast | Dark grey to black phases within a shallow pit | Dark or grey inclusion stringers | Bright metallic smearing | Supports oxide or refractory contamination |
Scanning electron microscopy and energy-dispersive spectroscopy
I cut specimens from the defect regions, cleaned them ultrasonically, and examined them by scanning electron microscopy. On the turned surfaces, the defects showed a distinct but irregular boundary between the contaminated region and the steel matrix. The matrix appeared bright, while the defect interior contained dark-grey and black particulate matter. On the unturned surfaces, the boundary was less distinct, and the central part of some spots showed charge accumulation, which is common for nonconductive oxide or refractory particles. At higher magnification, two main types of dark particles were present. Energy-dispersive spectroscopy identified them as aluminum oxide and aluminum-silicon oxide. The aluminum-silicon oxide signatures were consistent with mullite-type \(\mathrm{Al_6Si_2O_{13}}\), a common constituent of refractory materials used in heat-treatment and reheating furnaces.
The presence of both \(\mathrm{Al_2O_3}\) and \(\mathrm{Al_6Si_2O_{13}}\) is important. These phases are not typical deoxidation products in the amounts and morphologies observed here, and they do not match the expected inclusion population of a clean 20CrMnTi steel. More importantly, the same phases were found in the loose refractory material collected from the resistance heating furnace. The composition match between the surface spots and the furnace lining strongly indicates that the spots on the forged gear blanks are exogenous refractory contamination rather than endogenous steelmaking inclusions.
| Particle type | Dominant elemental signature | Identified phase | Typical morphology | Source significance |
|---|---|---|---|---|
| Type A | Al and O | Aluminum oxide, Al2O3 | Angular to blocky dark particles | Refractory lining constituent |
| Type B | Al, Si, and O | Aluminum-silicon oxide, consistent with mullite Al6Si2O13 | Irregular dark-grey particles | Refractory lining constituent |
| Matrix oxide | Fe and O | Iron oxide | Thin light-grey layer adjacent to the steel | High-temperature oxidation at the defect interface |
| Steel matrix | Fe with alloying elements | 20CrMnTi matrix | Bright metallic background | Reference matrix |
I also normalized the major oxide-forming cation fractions to compare the particle chemistry with the furnace refractory. For an EDS spectrum containing Al, Si, and O, the cation ratio can be expressed as
$$
R_{\mathrm{Al/Si}}=\frac{n_{\mathrm{Al}}}{n_{\mathrm{Si}}}
$$
and the oxygen balance can be checked against the stoichiometric expectation
$$
n_{\mathrm{O}}^{\mathrm{expected}}=\frac{3}{2}n_{\mathrm{Al}}+2n_{\mathrm{Si}}
$$
for a mixture of \(\mathrm{Al_2O_3}\) and \(\mathrm{SiO_2}\). Although EDS is not a replacement for quantitative phase analysis, the observed cation ratios and oxygen signatures were consistent with an alumina-mullite refractory system. The same comparison performed on the loose furnace particles produced the same phase families. This compositional fingerprint is the strongest evidence for the source of the spots on the forged gear blanks.
Cross-sectional morphology and high-temperature surface reactions
I sectioned several spots perpendicular to the surface and polished the cross-sections for scanning electron microscopy and metallography. The defects were shallow, U-shaped depressions rather than deep internal cavities. A distinct interface separated the depression from the steel matrix. At the matrix edge, I observed a thin layer of iron oxide. No dense rows of granular inclusions were found in the matrix adjacent to the depression, and there was no elongated inclusion stringer extending into the steel. Inside the depression, I found light-grey iron oxide together with dark-grey and black refractory particles. The composition of these dark particles matched the particles observed on the original surfaces and the loose furnace refractory.
Metallographic examination after etching with nitric acid in ethanol showed decarburization near the matrix side of the depression. The decarburized zone consisted of ferrite and pearlite, which is consistent with high-temperature exposure in an oxidizing or decarburizing atmosphere. The presence of a thin oxide layer between the refractory particle and the steel matrix is significant. It indicates that the particle was in contact with the hot steel surface at high temperature and that oxidation occurred at the interface before or during embedding. If the particle had been an intrinsic inclusion formed during solidification, such a surface oxide layer and localized decarburization would not be expected in this configuration.
| Cross-sectional feature | Observation | Interpretation |
|---|---|---|
| Depression geometry | Shallow U-shaped pit | Mechanical indentation of a foreign particle into a hot surface |
| Boundary | Distinct but irregular interface | Exogenous particle in contact with the steel matrix |
| Matrix edge oxide | Thin light-grey iron oxide layer | High-temperature oxidation at the defect interface |
| Pit interior | Iron oxide plus dark refractory particles | Oxidized wear or spall products trapped in the depression |
| Adjacent matrix | No elongated inclusion stringer | Not a rolled-in steelmaking inclusion |
| Decarburization | Ferrite and pearlite near the edge | Heating in an oxidizing or decarburizing atmosphere before or during forging |
The oxidation behavior can be described by a parabolic growth law. If \(x\) is the oxide thickness, \(t\) is time, and \(k_p\) is the parabolic rate constant, then
$$
x^2=k_p t
$$
and the temperature dependence of \(k_p\) is
$$
k_p=k_0\exp\left(-\frac{Q}{RT}\right)
$$
where \(k_0\) is a pre-exponential factor, \(Q\) is the activation energy, \(R\) is the gas constant, and \(T\) is the absolute temperature. I do not use these equations to assign a unique furnace temperature history, because the local oxygen potential and contact conditions are complex. I use them to support the qualitative conclusion that a thin oxide layer can form rapidly on a hot surface and that the layer will be preserved when a refractory particle is pressed into the surface during forging. Similarly, the decarburization depth \(d\) can be approximated by a diffusion-controlled relation of the form
$$
d=K\sqrt{t}\exp\left(-\frac{Q_d}{RT}\right)
$$
where \(K\) is a material-dependent constant and \(Q_d\) is an effective activation energy for carbon diffusion. The observed decarburization is shallow and localized, which is consistent with a heating and forging sequence rather than a prolonged bulk heat treatment.
Furnace inspection and source confirmation
I inspected the resistance heating furnace used before die forging. The furnace lining showed regions of cracking, spalling, and loose refractory debris. I collected several loose particles, mounted them in resin, ground and polished the mounts, coated them with carbon, and examined them by scanning electron microscopy and energy-dispersive spectroscopy. The particles displayed different colors and morphologies, but their dominant phase families were the same as those found in the surface spots: aluminum oxide and aluminum-silicon oxide with mullite-type signatures. This direct comparison completed the source identification. The refractory material was not merely a background contaminant; it was the specific material family present inside the surface defects on the forged gear blanks.
The detachment and adhesion of refractory particles can be discussed using a simple force balance. For a particle on the furnace wall or on the steel surface, detachment requires that the disturbing force exceed the adhesion force:
$$
F_{\mathrm{disturb}}>F_{\mathrm{adhesion}}
$$
The disturbing force can include thermal stress, gas flow, mechanical vibration, friction between the forged gear blanks and the furnace hearth, and impact during charging or discharging. Once a particle has detached, adhesion to the steel surface depends on the local contact area, surface roughness, oxide scale, and temperature. A simplified adhesion condition is
$$
F_{\mathrm{adhesion}}\geq F_{\mathrm{detach,local}}
$$
where \(F_{\mathrm{detach,local}}\) represents the local force tending to remove the particle from the steel surface before forging. If the particle remains on the surface until the forging stroke, it can be embedded. The embedding condition can be expressed using the mean contact pressure \(p_m\) and the temperature-dependent flow strength or hardness \(H(T)\) of the steel:
$$
p_m=\frac{F_{\mathrm{forge}}}{\pi a^2}
$$
and
$$
p_m>H(T)
$$
where \(F_{\mathrm{forge}}\) is the local forging force and \(a\) is the contact radius. When this condition is satisfied, the refractory particle penetrates the hot surface and creates a local depression. The surrounding metal flows plastically and partially covers the particle, which explains why the spots are less distinct before turning and more distinct after the deformed surface layer is removed.
| Comparison item | Surface spots on forged gear blanks | Loose furnace refractory particles | Match judgment |
|---|---|---|---|
| Al-O phase | Present | Present | Consistent |
| Al-Si-O phase | Present | Present | Consistent |
| Mullite-type signature | Observed | Observed | Consistent |
| Particle morphology | Angular to blocky fragments | Angular to blocky fragments | Consistent |
| Location relative to steel | Embedded in surface depression | Loose or adhered on furnace lining | Source-to-defect relationship supported |
| High-temperature oxide association | Iron oxide layer at interface | Refractory particles exposed to furnace atmosphere | Consistent with heating-stage exposure |
Why the defects formed after rolling and not during steelmaking
I considered three possible origins for the spots on the forged gear blanks: endogenous nonmetallic inclusions from steelmaking, rolled-in contamination from the rolling stage, and foreign-particle entrapment during heating or forging. The evidence favors the third mechanism. First, endogenous inclusions formed before or during rolling would be deformed along the rolling direction. The spots on the forged gear blanks were approximately circular and did not show a preferred orientation. Second, the composition of the particles was alumina and aluminum-silicon oxide with mullite-type signatures, which matches furnace refractory rather than the typical inclusion suite of a clean low-alloy steel. Third, the defects were present on unturned forged gear blanks, so they were not created by turning. Fourth, the cross-sections showed a U-shaped depression with an oxide layer and localized decarburization, which is a surface thermomechanical signature rather than an internal inclusion signature.
If the defect had been a rolled-in inclusion, I would expect a stringer aligned with the rolling direction, often with a length-to-width ratio much greater than one. I would also expect the inclusion to be located below the surface without a systematic surface depression, unless it happened to intersect the surface after machining. In contrast, the observed defects were open to the surface, contained refractory fragments, and were surrounded by a thin oxide layer. These features are characteristic of an exogenous particle that was pressed into the surface at high temperature and then partly oxidized.
| Hypothesis | Expected evidence | Observed evidence | Assessment |
|---|---|---|---|
| Steelmaking inclusion | Elongated stringers after rolling; internal location; inclusion chemistry matching deoxidation products | No elongation; surface depression; alumina and mullite-type refractory phases | Unlikely |
| Rolled-in contamination | Directional alignment with rolling; elongated shape; contamination distributed along the bar | No preferred direction; approximately circular spots; post-rolling surface morphology | Unlikely |
| Turning damage | Tool marks, periodic spacing, smeared metal, no spots before turning | Spots existed before turning; no periodic tool pattern | Excluded |
| Furnace refractory entrapment during hot forging | Refractory particles on the surface; U-shaped depression; oxide layer; decarburization; composition match with furnace lining | All features observed on the forged gear blanks | Strongly supported |
Process sequence and root-cause chain
The root-cause chain I established is as follows. The resistance heating furnace lining degraded by cracking and spalling. Loose refractory particles containing \(\mathrm{Al_2O_3}\) and \(\mathrm{Al_6Si_2O_{13}}\) detached from the lining. These particles adhered to the surface of the forged gear blanks during charging, heating, or transfer. At high temperature, the steel surface oxidized and a thin oxide layer formed. During the forging stroke, the localized pressure exceeded the high-temperature flow strength of the steel, so the refractory particles were pressed into the surface. The surrounding metal deformed plastically and formed a shallow U-shaped depression. After forging and cooling, the embedded particles remained at the surface and appeared as spot-like defects. Turning removed the outermost deformed layer and made the spots more visible, but it did not create them.
I summarize the process-stage behavior in Table 8. The table is useful because it separates the stages at which the defect could have formed from the stages at which it was merely revealed. The spots were already present on the as-forged surfaces of the forged gear blanks, and their chemistry matched the furnace refractory. This is the basis for assigning the origin to the heating and forging sequence.
| Process stage | Thermal or mechanical condition | Expected effect on a surface particle | Observed consequence |
|---|---|---|---|
| Steelmaking and refining | Liquid steel treatment | Endogenous inclusions may form and deform later | No evidence of an endogenous inclusion origin |
| Continuous casting | Solidification | Inclusions and segregation may develop | Chemistry and morphology do not support this route |
| Continuous rolling | High reduction and elongation | Pre-existing inclusions would elongate along the rolling direction | Spots are equiaxed and show no rolling directionality |
| Resistance furnace heating | High temperature and oxidizing atmosphere | Refractory particles can spall, adhere, and oxidize | Refractory phases and interface oxide are present |
| Die forging | High local pressure and plastic flow | Adhered particles can be pressed into the surface | U-shaped depressions and deformed rims are observed |
| Turning | Material removal | Defects can be exposed more clearly | Spots become more distinct but were already present before turning |
Quantitative interpretation and acceptance criteria
For routine production control, I recommend treating the spot defects as a surface cleanliness issue and monitoring them with simple quantitative metrics. The first metric is the surface spot density \(\lambda\). The second is the circularity \(\Psi\). The third is the surface coverage fraction \(\theta\), defined as
$$
\theta=\frac{1}{A_{\mathrm{inspected}}}\sum_{i=1}^{N}A_i
$$
where \(A_i\) is the projected area of the \(i\)-th spot. For forged gear blanks intended for gear machining, a low \(\theta\) value and a low \(\lambda\) value are desirable. The exact acceptance limits should be set from a statistically valid baseline of acceptable production, but the following framework is practical: if the spot density increases abruptly, the furnace lining should be inspected; if the circularity decreases and the spots become elongated, a rolling-origin inclusion problem should be investigated instead; if the spots contain refractory phases and are associated with a U-shaped depression, the heating furnace is the primary suspect.
| Indicator | Definition | Diagnostic use | Suggested action trigger |
|---|---|---|---|
| Spot density, \(\lambda\) | Number of spots per unit inspected area | Detects an increase in surface contamination | Investigate furnace condition when the value exceeds the baseline |
| Circularity, \(\Psi\) | Shape parameter based on area and perimeter | Distinguishes equiaxed entrapment from elongated inclusions | If values decrease sharply, check rolling and inclusion sources |
| Surface coverage, \(\theta\) | Total spot area divided by inspected area | Estimates the severity of surface contamination | Increase inspection frequency when coverage rises |
| Refractory phase fraction | Relative EDS frequency of Al-O and Al-Si-O particles | Confirms whether the spots are refractory-related | If positive, inspect and clean the heating furnace |
| Decarburization depth, \(d\) | Depth of the ferrite-rich zone at the defect edge | Indicates high-temperature exposure | If excessive, review heating time and atmosphere |
Prevention and process-control recommendations
Because the root cause is refractory spalling and particle entrapment, the most effective countermeasures are furnace maintenance and surface protection before forging. I recommend a structured inspection and cleaning program for the resistance heating furnace. Cracked or spalled refractory zones should be repaired before they release loose particles. The furnace hearth and the path of the forged gear blanks should be kept free of debris. Where possible, a suitable refractory coating or lining material with better spalling resistance should be selected. The heating atmosphere should be controlled to reduce excessive oxidation and decarburization, because a thick oxide scale can trap loose particles and make cleaning less effective. Before forging, the surface of the forged gear blanks should be visually inspected or cleaned by a controlled method such as brushing, air blowing, or another approved surface-cleaning practice. The cleaning method must not introduce new contamination.
I also recommend tracing the defect population by furnace campaign. If the spot density on forged gear blanks increases after a furnace repair, the repair material or curing procedure should be reviewed. If the spot density decreases after cleaning, the refractory source is confirmed. This campaign-based comparison is more reliable than a single visual inspection because refractory spalling can be intermittent. A simple control chart of \(\lambda\) versus production sequence is useful. When a sudden increase in \(\lambda\) occurs, the furnace should be inspected before the next production run. The same logic applies to the surface coverage \(\theta\): a slow upward trend suggests progressive lining degradation, while a step change suggests a local spall or a change in furnace operation.
| Control point | Recommended action | Acceptance criterion | Frequency |
|---|---|---|---|
| Furnace lining | Inspect for cracks, spalling, and loose refractory | No visible loose debris on the hearth or lining | Each campaign and after repair |
| Furnace cleaning | Remove broken refractory particles and oxide debris | Clean hearth and unobstructed transfer path | Start of campaign and as needed |
| Refractory material selection | Use spalling-resistant refractory with suitable phase stability | Reduced spall rate compared with the previous lining | At lining replacement |
| Heating atmosphere | Control oxygen potential and avoid excessive scale formation | Scale and decarburization within the approved range | Continuous monitoring where possible |
| Surface cleaning before forging | Remove loose particles from the forged gear blanks | No visible adhered particles before forging | Every batch |
| Post-forging inspection | Check for spot defects before turning | Spot density below the approved baseline | Every batch or sampling plan |
| Turning inspection | Record spot count and morphology after turning | No abnormal increase in \(\lambda\) or \(\theta\) | Every batch |
| Root-cause verification | Compare defect particles with furnace refractory by SEM and EDS | Phase identification matches the refractory source | When a new defect pattern appears |
Discussion of surface engineering and forging mechanics
The formation of these spots on forged gear blanks can be understood as a competition between particle adhesion, particle removal, and plastic embedding. During heating, the steel surface develops an oxide scale. The oxide scale may increase the roughness of the surface, which can increase mechanical interlocking between a refractory particle and the steel. At the same time, the high temperature reduces the flow strength of the steel, so a relatively small forging pressure can produce local plastic deformation. The particle does not need to be large to create a visible spot; a small refractory fragment can produce a shallow depression because the surrounding metal flows around it. After forging, the depression is partly filled by oxide and refractory fragments. Turning then cuts through the deformed rim and exposes the embedded particle more clearly.
I can express the local indentation depth in a simplified form. For a spherical or rounded particle of radius \(r\) pressed by a force \(F\) into a surface with resistance pressure \(H(T)\), an approximate contact radius \(a\) and penetration depth \(h\) satisfy
$$
F=\pi a^2 H(T)
$$
and for small penetration,
$$
a^2\approx 2rh-h^2
$$
Combining these relations gives
$$
F\approx \pi H(T)\left(2rh-h^2\right)
$$
This equation shows that for a given particle radius and forging force, the penetration depth increases as \(H(T)\) decreases. This is why the defect is associated with hot forging rather than cold handling. It also explains why the defect appears as a shallow U-shaped depression rather than a deep crack. The hot metal deforms locally around the particle and accommodates it without requiring a large macroscopic indentation.
Oxide scale and decarburization provide additional evidence of high-temperature exposure. The thin iron oxide layer at the interface separates the refractory particle from the steel matrix, which means that oxygen reached the interface before or during embedding. The decarburized zone near the matrix edge indicates that carbon diffused outward while the surface was hot. These two features are consistent with a heating operation in an oxidizing atmosphere. If the particle had been introduced after forging, for example during turning or handling, decarburization and interfacial oxide would not be expected in this configuration. The combination of refractory chemistry, surface depression, interfacial oxide, and decarburization is therefore a strong diagnostic signature.
| Mechanistic factor | Physical meaning | Observable consequence |
|---|---|---|
| Refractory spalling | Lining fragments detach under thermal and mechanical stress | Loose Al2O3 and Al6Si2O13 particles in the furnace |
| Adhesion | Particles remain on the steel surface during transfer | Localized contamination before forging |
| Hot plasticity | Steel flow strength decreases at high temperature | Particles are pressed into the surface |
| Oxidation | Iron oxide forms at the hot interface | Thin light-grey oxide layer at the defect edge |
| Decarburization | Carbon diffuses outward in a decarburizing atmosphere | Ferrite and pearlite zone near the depression |
| Turning | Material removal reveals the embedded particle | Spots appear sharper after machining |
Implications for inspection and quality assurance
For quality assurance, the inspection of forged gear blanks should distinguish between as-forged spots and machining-induced marks. I recommend examining a sample of forged gear blanks before turning and comparing the spot locations with those found after turning. If the same spots are present before and after turning, the machining process is not responsible. If the spots contain refractory phases, the heating furnace should be inspected. If the spots are elongated and aligned with the bar axis, the rolling and inclusion history should be investigated instead. This decision logic prevents misclassification and avoids unnecessary changes to the turning process.
I also recommend retaining defect specimens for SEM and EDS analysis whenever a new spot pattern appears. The phase identification is fast and provides a direct link to the refractory source. The presence of both \(\mathrm{Al_2O_3}\) and \(\mathrm{Al_6Si_2O_{13}}\) is a useful fingerprint because it points to an alumina-mullite refractory system. When this fingerprint is found on forged gear blanks, the furnace lining should be inspected before the next campaign. When the fingerprint is absent and the particles are typical steelmaking inclusions, the investigation should shift to deoxidation, casting, and rolling. This distinction is important for efficient problem solving.
| Observation | Most likely origin | Next action |
|---|---|---|
| Spots present before turning and after turning | Formed before turning | Focus on heating, forging, and bar surface condition |
| Spots approximately circular with no rolling directionality | Post-rolling surface entrapment | Inspect furnace and forging environment |
| Al2O3 and Al6Si2O13 particles in the spot | Refractory contamination | Inspect and clean the resistance heating furnace |
| Elongated stringers aligned with the bar axis | Rolled inclusion or rolled-in contamination | Review steelmaking, casting, and rolling |
| Tool marks or periodic scratches | Machining damage | Review tool condition and cutting parameters |
| U-shaped depression with interfacial oxide and decarburization | Hot embedding of a foreign particle | Check particle source and pre-forging surface cleaning |
Main conclusions
I conclude that the spot-like surface defects on the 20CrMnTi forged gear blanks were caused by refractory particles that detached from the resistance heating furnace, adhered to the steel surface before hot forging, and were pressed into the surface during forging. The particles were primarily \(\mathrm{Al_2O_3}\) and aluminum-silicon oxide with mullite-type \(\mathrm{Al_6Si_2O_{13}}\) signatures. The same phases were found in the loose refractory material collected from the furnace. The defects were present before turning, were approximately circular, showed no rolling directionality, and had cross-sectional features of a U-shaped depression with interfacial iron oxide and localized decarburization. These features are consistent with high-temperature surface entrapment and inconsistent with endogenous steelmaking inclusions or turning damage.
The most effective prevention measures are to inspect and maintain the resistance heating furnace, remove loose refractory debris, select spalling-resistant lining materials, control the heating atmosphere, clean the surfaces of forged gear blanks before forging, and monitor surface spot density and morphology after forging and turning. I recommend using the refractory phase fingerprint as a rapid diagnostic tool. When \(\mathrm{Al_2O_3}\) and \(\mathrm{Al_6Si_2O_{13}}\) are identified in surface spots on forged gear blanks, the furnace lining should be treated as the primary source until proven otherwise. With these controls, the recurrence of spot defects on forged gear blanks can be reduced and the surface quality of the final gear components can be stabilized.