In my work with heavy-duty reducers and large transmission components, I have repeatedly seen how the choice of a gear blank determines the final quality, cost, and service life of the gear. A forged gear blank is often the first choice for small and medium gears because forging closes internal porosity, refines the grain structure, and produces a favorable metal flow line along the tooth profile. However, when the gear diameter exceeds roughly 2 m, the required forging equipment, ingot size, and heat-treatment capacity become severe constraints. At that scale, a fully forged gear blank can become uneconomical or even impossible to obtain within a reasonable delivery period. A cast gear blank, on the other hand, offers excellent shape freedom and low initial cost, but it often suffers from shrinkage, porosity, and inclusions in the rim region. Because the tooth root is usually located near the mid-thickness of the rim, these casting defects can be exposed after gear cutting and can drastically reduce bending fatigue strength. To solve this contradiction, I developed and applied a composite structure in which the rim is a forged gear blank and the hub is a cast component. The two parts are joined by a combination of cylindrical pins and a full-penetration welded connection. This composite forged gear blank approach preserves the high integrity of the tooth region while keeping the overall blank cost and manufacturing risk under control.
The application that first forced me to adopt this solution was a large secondary reducer gear for a 120 t converter tilting drive. The gear diameter was greater than 3 m, the material specification was a medium-carbon low-alloy cast steel, and the ultrasonic testing acceptance level for the original monolithic casting was extremely strict. Very few foundries could deliver such a casting with a acceptable first-pass yield. After discussing the problem with the manufacturing team, I proposed replacing the monolithic casting with a composite forged gear blank. The tooth ring would be produced by ring rolling and forging, while the hub would remain a casting. The two parts would be machined, assembled, pinned, welded, stress relieved, and then final machined. This article presents the design logic, process route, calculations, and quality controls that I used. It is written from my own engineering perspective and reflects the decisions that proved successful in production.
Service conditions and failure modes of large gears
A large gear in a converter tilting drive is not a simple power transmission element. It operates at low speed, carries heavy torque, and experiences shock loads during charging, tapping, and emergency braking. The tooth flanks are subjected to alternating contact stress, while the tooth roots are subjected to alternating bending stress. Simultaneous sliding and rolling contact generates friction and heat. If the load distribution is uneven because of misalignment, shaft deflection, or poor assembly, the local contact stress can exceed the allowable value and initiate pitting, spalling, or plastic deformation. If the root fillet contains a casting defect, an inclusion, or a sharp machining mark, a fatigue crack can start and propagate until the tooth breaks. Therefore, the quality of the forged gear blank in the tooth region is directly linked to the reliability of the entire drive.
The fundamental gear force relationships can be summarized as follows. For a cylindrical gear with torque \(T\), pitch diameter \(d\), normal pressure angle \(\alpha_n\), and helix angle \(\beta\), the tangential force is
$$ F_t = \frac{2T}{d} $$
The radial and axial components are
$$ F_r = \frac{F_t \tan \alpha_n}{\cos \beta} $$
$$ F_a = F_t \tan \beta $$
The tooth root bending stress can be estimated by
$$ \sigma_F = \frac{F_t}{b m_n} Y_F Y_S Y_\beta K_A K_V K_{F\beta} K_{F\alpha} \le \sigma_{FP} $$
where \(b\) is the face width, \(m_n\) is the normal module, \(Y_F\) is the form factor, \(Y_S\) is the stress correction factor, \(Y_\beta\) is the helix angle factor, and the \(K\) terms are load application, dynamic, face load, and transverse load factors. The contact stress can be estimated by
$$ \sigma_H = Z_E Z_H Z_\varepsilon \sqrt{ \frac{F_t}{b d_1} \frac{u+1}{u} K_A K_V K_{H\beta} K_{H\alpha} } \le \sigma_{HP} $$
where \(Z_E\) is the elastic coefficient, \(Z_H\) is the zone factor, \(Z_\varepsilon\) is the contact ratio factor, \(d_1\) is the pinion pitch diameter, and \(u\) is the gear ratio. These equations show that any internal defect in the rim directly reduces the allowable bending and contact stresses. A forged gear blank with a sound, homogeneous rim therefore provides a much higher safety margin than a cast gear blank with uncertain internal quality.
Table 1 summarizes the common failure modes and the corresponding material or manufacturing remedies that I consider when designing a large gear blank.
| Failure mode | Primary cause | Effect on gear life | Remedy in blank design |
|---|---|---|---|
| Tooth root fatigue fracture | Alternating bending stress, defect at root fillet | Sudden loss of drive, secondary damage | Use a forged gear blank for the rim; UT class II or better; smooth fillet |
| Pitting and spalling | High contact stress, poor lubrication, surface defects | Noise, vibration, progressive tooth loss | Clean forged material, proper heat treatment, adequate hardness |
| Tooth flank wear | Sliding friction, contamination, insufficient hardness | Reduced accuracy, increased backlash | Through-hardened or surface-hardened rim, controlled microstructure |
| Rim cracking | Casting porosity, shrinkage, low ductility | Catastrophic rim failure | Replace cast rim with a forged gear blank ring |
| Hub cracking | Stress concentration, casting defects, keyway stress | Loss of torque transmission | Cast hub with conservative fillets and UT class III acceptance |
| Weld or joint failure | Poor fusion, residual stress, insufficient pinning | Relative movement between rim and hub | Qualified welding procedure, pins, stress relief, NDT |
Conventional manufacturing routes and their limitations
For smaller gears, the conventional route is straightforward: cut an ingot, forge the blank, normalize, rough machine, quench and temper, finish machine, and cut the teeth. A forged gear blank made in this way has a continuous grain flow that follows the tooth profile, especially when the forging is performed by upsetting and punching followed by ring rolling. This flow improves fatigue resistance and impact toughness. For large gears, however, the forging press capacity, manipulator capacity, furnace size, and quenching tank size all become limiting factors. The cost of a large forging die and the risk of inadequate forging penetration also rise sharply. As a result, many large gears are made from castings.
A cast gear blank has the advantage of near-net shape. The outer profile, hub, web, and mounting features can be formed directly in the mold. Machining allowance can be small, and the blank cost is often lower than that of a large forging. However, the casting process is statistically more likely to produce internal defects. The rim region is especially critical because the tooth cutting removes a significant amount of material and exposes the mid-rim section. If shrinkage porosity or a cluster of inclusions lies in that region, the tooth root will be weakened. In my experience, the first-pass yield of large cast steel gear blanks under strict UT requirements is often only about 60%, and in some cases much lower. The repair welding of such defects is difficult and may not restore fatigue performance. Table 2 compares the main characteristics of the two conventional routes.
| Characteristic | Forged gear blank | Cast gear blank |
|---|---|---|
| Material soundness | High, porosity closed by deformation | Variable, risk of shrinkage and porosity |
| Grain flow | Continuous and favorable | Random dendritic structure |
| Fatigue strength | High, especially in rim and root | Lower and more scatter |
| Shape freedom | Limited by dies and press capacity | High, complex features possible |
| Machining allowance | Larger, especially for large rings | Smaller, near-net shape |
| Tooling cost | High for large dies | Lower for patterns and molds |
| Size limit | Limited by press, ring mill, furnace, quench tank | Can be large if foundry capacity exists |
| Delivery risk | High for very large parts | High due to defect risk and rework |
| Cost for diameter > 2 m | Often very high or unavailable | Lower if quality can be met |
| Inspection | UT, MT, hardness, grain flow | UT, MT, radiography sometimes |
Composite forged gear blank concept
The composite design separates the gear blank into two functional regions: the hub and the rim. The rim, which contains the teeth, is made from a forged gear blank. The hub, which transmits torque from the shaft to the rim and carries the mounting features, is made from a casting. The two parts are joined at a cylindrical interface. Cylindrical pins provide initial location and carry part of the shear load. A welded connection along the interface provides continuity and prevents relative motion. The forged rim ensures that the tooth root and flank are free from casting defects. The cast hub allows complex internal geometry, large fillets, and a lower-cost blank for the less critical region. This is the essence of the composite forged gear blank strategy.
For the 120 t converter tilting reducer, the original gear was a single cast steel component. The material was a medium-carbon low-alloy cast steel, and the UT acceptance level was class I. The foundry could not guarantee this level with acceptable cost and delivery. I therefore proposed a split design. The rim would be a forged and ring-rolled gear blank of 35CrMo, and the hub would be a casting of ZG35CrMo. The two materials have similar carbon and alloy content, which makes them compatible for welding. The rim would be UT inspected to class II, and the hub to class III. The tooth region would thus be forged, while the non-tooth region could tolerate a slightly lower inspection class. This reduced the manufacturing risk without changing the gear rating.
The following equations describe the load path through the composite joint. If the gear transmits torque \(T\), the tangential force at the joint radius \(r_j\) is
$$ F_j = \frac{T}{r_j} $$
If the joint is provided by \(n\) cylindrical pins of diameter \(d_p\), the average shear stress in each pin is
$$ \tau_p = \frac{F_j}{n A_p} = \frac{T}{n \pi d_p^2 r_j / 4} \le \tau_{p,allow} $$
If the welded ring is treated as a thin circular weld of throat thickness \(t_w\), the nominal shear stress in the weld under torsion is approximately
$$ \tau_w = \frac{T}{2 \pi r_j^2 t_w} \le \tau_{w,allow} $$
These equations are conservative because the pins and weld share the load, but they provide a useful first check. In the final design, I used both a finite element model and a qualified welding procedure to confirm the joint capacity.
Material selection for the composite forged gear blank
The material selection must satisfy three requirements: the rim must have enough hardenability and fatigue strength, the hub must have enough castability and toughness, and the two materials must be weldable to each other. For the rim, I selected 35CrMo forged steel. It is a medium-carbon alloy steel with good strength, toughness, and fatigue resistance after quenching and tempering. It can be ring rolled into a large diameter without excessive segregation. For the hub, I selected ZG35CrMo cast steel. Its composition is close to that of the rim, which minimizes the formation of a hard, brittle weld interface. It also casts well in thick sections and can be normalized and tempered to a uniform hardness. Table 3 lists typical compositions and mechanical properties. These values are representative and should be confirmed by actual mill certificates and test coupons.
| Property | Forged rim 35CrMo | Cast hub ZG35CrMo |
|---|---|---|
| Carbon, wt% | 0.32–0.40 | 0.30–0.37 |
| Silicon, wt% | 0.17–0.37 | 0.30–0.60 |
| Manganese, wt% | 0.40–0.70 | 0.50–0.80 |
| Chromium, wt% | 0.80–1.10 | 0.80–1.20 |
| Molybdenum, wt% | 0.15–0.25 | 0.15–0.25 |
| Phosphorus, max wt% | 0.025 | 0.025 |
| Sulfur, max wt% | 0.025 | 0.025 |
| Tensile strength, MPa | ≥ 850 | ≥ 700 |
| Yield strength, MPa | ≥ 650 | ≥ 500 |
| Elongation, % | ≥ 14 | ≥ 12 |
| Impact energy, J | ≥ 40 at 20 °C | ≥ 30 at 20 °C |
| Hardness after Q&T | 250–280 HBW | 250–280 HBW |
I chose a quench and temper hardness range of 250–280 HBW for both parts. This range provides a good balance between strength and machinability. It also keeps the weld heat-affected zone from becoming excessively hard. If the hardness were allowed to exceed 320 HBW, the risk of cold cracking and machining difficulty would increase. If it were below 230 HBW, the tooth contact fatigue resistance would be reduced.
Hub casting process for the composite forged gear blank
The hub is the cast portion of the composite forged gear blank. Because it does not contain gear teeth, its inspection acceptance level can be lower than that of the rim. However, it still carries torque and must be sound enough to support the welded joint and the shaft connection. The casting process begins with a pattern that includes machining allowance, draft, and fillets. I specified a draft angle of 1:20 and transition fillets of \(R\) 17–25 mm. These fillets reduce stress concentration and help the metal flow during pouring. The gating and risering system must be designed to feed the thick sections and to avoid shrinkage porosity near the weld preparation. Chills may be used where the section thickness changes abruptly.
After casting, the hub is annealed or normalized to relieve casting stress and to homogenize the microstructure. It is then rough machined with approximately 5 mm per side allowance. Rough machining removes the skin and allows the subsequent quench and temper treatment to be more effective. The hub is then quenched and tempered to 250–280 HBW. After heat treatment, it is inspected by ultrasonic testing to class III. The surfaces that will mate with the rim are then finish machined. A U-shaped weld groove of approximately 30 mm depth and 8° included angle is prepared on the hub side. This groove provides good access for the welding torch and ensures full fusion at the root. Table 4 lists the hub casting and machining parameters.
| Step | Parameter or requirement | Purpose |
|---|---|---|
| Pattern draft | 1:20 | Easy pattern removal |
| Fillet radius | 17–25 mm | Reduce stress concentration |
| Machining allowance | 5 mm per side, rough stage | Remove skin and permit heat treatment |
| Heat treatment | Normalize or anneal, then quench and temper | Relieve stress, homogenize structure |
| Hardness | 250–280 HBW | Strength and machinability balance |
| Ultrasonic testing | Class III | Detect gross internal defects |
| Weld preparation | U-groove, 30 mm × 8° | Full penetration welding |
Rim forging process for the composite forged gear blank
The rim is the most critical part of the composite forged gear blank. It must be forged from a clean, fully killed steel ingot. I required a forging ratio of not less than 3. The forging ratio is defined as the ratio of the original cross-sectional area to the final cross-sectional area after deformation. A ratio of 3 or higher ensures that the dendritic structure is broken up and that internal porosity is closed. The forging sequence typically includes upsetting, punching, and ring rolling. Upsetting refines the axial structure and closes central porosity. Punching creates the central hole. Ring rolling reduces the wall thickness and increases the diameter while maintaining a circumferential grain flow. This flow is highly desirable because the tooth loads create circumferential and radial stresses in the rim.
After forging, the rim is heat treated to reduce forging stress and to prepare for machining. It is then rough turned, quenched and tempered to 250–280 HBW, and finish turned on the mating surfaces. The weld preparation is machined on the rim side to match the hub side. The rim is then inspected by ultrasonic testing to class II. This is stricter than the hub because the rim will contain the gear teeth. Any indication that could interact with the tooth root is unacceptable. Table 5 summarizes the rim forging and preparation steps.
| Step | Parameter or requirement | Purpose |
|---|---|---|
| Raw material | Clean, killed steel ingot | Low inclusion content |
| Forging ratio | ≥ 3 | Close porosity, refine grains |
| Forging operations | Upset, punch, ring roll | Circumferential grain flow |
| Post-forge heat treatment | Annealing or normalizing | Relieve forging stress |
| Rough machining | 5 mm per side allowance | Prepare for Q&T |
| Quench and temper | 250–280 HBW | Strength and toughness |
| Ultrasonic testing | Class II | Ensure rim soundness |
| Weld preparation | Matched U-groove | Full penetration weld |
Assembly and pinning of the composite forged gear blank
After the hub and rim have been independently inspected and machined, they are assembled. The fit between the hub and rim is a cylindrical clearance fit that allows assembly but limits eccentricity. The mating surfaces are cleaned and degreased. The two parts are aligned by their machined bores and faces. Cylindrical pins are then installed to locate the parts and to carry part of the shear load. In my design, I used eight pins on each side of the web, for a total of sixteen pins. The pins were positioned by layout and drilled and reamed in place. The pin locations were chosen so that no pin would fall under a tooth root after gear cutting. This is important because a pin hole near the root would create a stress concentration and could become a fatigue initiation site.
The angular spacing of the pins can be calculated as follows. If there are \(n\) pins on one side, the nominal angular spacing is
$$ \Delta \theta = \frac{360^\circ}{n} $$
If the pins are arranged in two rows, the rows can be offset by half a pitch to distribute the load more evenly:
$$ \theta_{i,row2} = \theta_{i,row1} + \frac{\Delta \theta}{2} $$
In my case, with eight pins per side, the spacing was 45°, and the second row was offset by 22.5°. This arrangement avoids alignment of pins with the same radial line and reduces the risk of a continuous crack path. Table 6 lists the pin and assembly parameters.
| Parameter | Value or requirement |
|---|---|
| Number of pins per side | 8 |
| Total number of pins | 16 |
| Pin type | Cylindrical, reamed fit |
| Pin material | Medium-carbon alloy steel, quenched and tempered |
| Pin diameter | Sized by shear calculation |
| Angular spacing per side | 45° |
| Second-row offset | 22.5° |
| Pin location rule | No pin under tooth root |
| Surface condition | Clean, dry, free of grease |
Welding of the composite forged gear blank
The welded connection is the most delicate step in the manufacture of the composite forged gear blank. The weld must transfer torque, resist bending, and remain sound after stress relief. Because both the rim and the hub are heat treated to 250–280 HBW, the weld procedure must control preheat, interpass temperature, heat input, and cooling rate. I specified flame preheating to 250–300 °C. The flame must not be concentrated on one spot, because local overheating can change the microstructure and create stress concentrations. The preheat should be uniform around the joint. Temperature-indicating crayons or contact thermocouples can be used to verify the preheat.
For the welding process, I selected Ar-CO\(_2\) mixed gas metal arc welding. The filler wire was H08Mn2SiA, which is a manganese-silicon steel wire suitable for medium-carbon low-alloy steels. The welding current was approximately 200 A, and the arc voltage was 27–30 V. These parameters produce a stable arc and sufficient penetration without excessive heat input. The heat input can be estimated by
$$ Q = \eta \frac{U I}{v} $$
where \(Q\) is the heat input in kJ/mm, \(\eta\) is the process efficiency, \(U\) is the arc voltage in volts, \(I\) is the current in amperes, and \(v\) is the travel speed in mm/s. For gas metal arc welding, \(\eta\) is typically 0.8–0.9. I kept the heat input within a qualified range to avoid cold cracking and excessive distortion. Table 7 lists the welding parameters.
| Parameter | Value or requirement |
|---|---|
| Welding process | Ar-CO\(_2\) gas metal arc welding |
| Filler wire | H08Mn2SiA |
| Current | 200 A |
| Voltage | 27–30 V |
| Preheat temperature | 250–300 °C |
| Interpass temperature | 250–300 °C |
| Welding position | All-position |
| Welding sequence | Balanced, alternating sides |
| Number of welders | Two or more simultaneously |
| Post-weld cooling | Cover with asbestos blanket, slow cooling |
Because the joint is circular, welding on only one side would create severe angular distortion and residual stress. I therefore required two or more welders to work simultaneously on opposite sides of the circumference. They followed a balanced sequence, depositing short beads and alternating positions. This method keeps the heat input symmetrical and controls distortion. After welding, the joint was covered with an asbestos blanket to slow cooling. Slow cooling helps avoid martensite formation and cold cracking. Once the joint reached room temperature, the weld surface was cleaned and inspected visually and by magnetic particle testing. Any surface indication was removed before stress relief.
Post-weld heat treatment of the composite forged gear blank
The post-weld heat treatment is intended to relieve welding residual stress without degrading the quenched and tempered properties of the rim and hub. The stress relief temperature must be lower than the tempering temperature of the base materials. Since both parts were tempered at a higher temperature during their original heat treatment, I selected a stress relief temperature of \(520 \pm 20\) °C. The holding time was 3 h. The cooling rate was controlled by cooling in the furnace to 350 °C and then air cooling. This practice reduces residual stress while maintaining the hardness and strength of the forged rim and cast hub.
The Larson-Miller parameter can be used to compare stress relief treatments:
$$ P_{LM} = T(\log_{10} t + 20) \times 10^{-3} $$
where \(T\) is the absolute temperature in kelvin and \(t\) is the time in hours. For my stress relief treatment, \(T = 793\) K and \(t = 3\) h, which gives a parameter suitable for carbon and low-alloy steels. I used this parameter only as a guide; the final procedure was qualified by hardness and impact testing. Table 8 lists the post-weld heat treatment parameters.
| Parameter | Value or requirement |
|---|---|
| Purpose | Relieve welding residual stress |
| Temperature | 520 ± 20 °C |
| Holding time | 3 h |
| Cooling | Furnace cool to 350 °C, then air cool |
| Constraint | Below original tempering temperature |
| Verification | Hardness, magnetic particle, dimensional check |
Final machining of the composite forged gear blank
After stress relief, the composite forged gear blank is ready for final machining. The machining sequence begins with semi-finish turning of the outer diameter, bore, and faces. The part is then finish turned to establish the gear cutting datum. The gear teeth are cut by hobbing or milling, depending on the module and accuracy requirement. The keyway is then cut by slotting or broaching. Because the rim is a forged gear blank, the tooth cutting operation exposes sound, dense material. This is the main quality advantage of the composite design. The hub, being a casting, is not subjected to tooth cutting, so any minor internal discontinuities remain in a low-stress region.
I maintained the following machining parameters for the large gear. The semi-finish turning allowance was 2–3 mm per side. The finish turning allowance was 0.3–0.5 mm per side. The gear cutting was performed on a large hobbing machine with a rigid setup and continuous coolant supply. The tooth accuracy was checked by chordal thickness, base tangent length, and cumulative pitch error. The keyway was checked for symmetry and width. Table 9 lists the final machining steps.
| Step | Operation | Key requirement |
|---|---|---|
| 1 | Semi-finish turning | Establish uniform allowance |
| 2 | Finish turning | Control bore, faces, and outer diameter |
| 3 | Gear cutting | Correct module, helix angle, and accuracy |
| 4 | Keyway machining | Width, depth, and symmetry |
| 5 | Deburring and chamfering | Remove sharp edges at tooth ends |
| 6 | Final inspection | Dimensional, hardness, and NDT |
Quality inspection and acceptance criteria
The quality plan for the composite forged gear blank includes inspection at every major stage. The raw materials are certified for composition and mechanical properties. The hub casting is inspected by ultrasonic testing to class III. The rim forging is inspected by ultrasonic testing to class II. After welding, the joint is inspected by magnetic particle testing and sometimes by ultrasonic testing. After stress relief, hardness is measured on the rim, hub, and weld heat-affected zone. After final machining, the gear is inspected for dimensions, tooth thickness, pitch error, runout, and surface finish. Table 10 summarizes the inspection plan.
| Stage | Inspection method | Acceptance criterion |
|---|---|---|
| Raw material | Chemical analysis, mechanical test | Within specification |
| Hub casting | Ultrasonic testing | Class III |
| Rim forging | Ultrasonic testing | Class II |
| After welding | Magnetic particle testing | No relevant surface indications |
| After stress relief | Hardness test | 250–280 HBW |
| Final gear | Dimensional and tooth inspection | Drawing and accuracy class |
| Final gear | Magnetic particle testing of teeth | No cracks or laps |
Economic and technical comparison
The composite forged gear blank offers a balanced solution between a fully forged blank and a fully cast blank. A fully forged gear blank of 3 m diameter requires a very large press and ring mill, a huge heat-treatment furnace, and extensive machining. The cost is high, and the delivery risk is significant. A fully cast gear blank has lower initial cost but a high risk of internal defects in the rim. If the casting fails inspection, the entire blank must be scrapped or repaired, which can delay the project. The composite design places the forged material only where it is needed—the tooth region—and uses cast material for the hub. This reduces the cost of the forged portion and the inspection burden. Table 11 compares the three options.
| Criterion | Fully forged gear blank | Fully cast gear blank | Composite forged gear blank |
|---|---|---|---|
| Rim soundness | Excellent | Variable | Excellent |
| Hub complexity | Limited | Excellent | Excellent |
| Tooth fatigue life | High | Lower | High |
| First-pass UT yield | High but costly | Low for class I | High for rim, moderate for hub |
| Manufacturing cost | Very high for large size | Low if no defects | Moderate |
| Delivery risk | High for large size | High due to defect risk | Moderate |
| Machining allowance | Large | Small | Moderate |
| Welding and joining | None | None | Required, qualified procedure |
| Inspection complexity | Moderate | High for rim | Moderate, staged |
Case study: large gear for a converter tilting reducer
The case that demonstrates the value of the composite forged gear blank was a secondary reducer gear for a 120 t converter tilting drive. The gear diameter was greater than 3 m. The original design was a monolithic casting with a strict class I ultrasonic acceptance requirement. The foundry trial castings repeatedly showed indications in the rim region. After several attempts, the delivery schedule was at risk. I proposed the composite design and prepared the manufacturing route. The rim was forged and ring rolled from 35CrMo with a forging ratio of at least 3. The hub was cast from ZG35CrMo. Both were quenched and tempered to 250–280 HBW. The rim was UT inspected to class II, and the hub to class III. The two parts were assembled with sixteen pins, preheated to 250–300 °C, welded with Ar-CO\(_2\) gas metal arc welding using H08Mn2SiA wire, and stress relieved at 520 ± 20 °C for 3 h. After final machining, the gear passed all dimensional and NDT requirements on the first submission.
Table 12 lists the main design and process parameters of the case study.
| Parameter | Value |
|---|---|
| Application | Converter tilting secondary reducer |
| Gear diameter | Greater than 3 m |
| Original blank | Monolithic cast steel |
| Original UT requirement | Class I |
| Revised design | Composite forged gear blank |
| Rim material | 35CrMo forged and ring rolled |
| Hub material | ZG35CrMo cast |
| Forging ratio | ≥ 3 |
| Hardness | 250–280 HBW |
| Rim UT class | II |
| Hub UT class | III |
| Pins | 8 per side, 16 total |
| Preheat | 250–300 °C |
| Welding process | Ar-CO\(_2\) GMAW |
| Filler wire | H08Mn2SiA |
| Current and voltage | 200 A, 27–30 V |
| Stress relief | 520 ± 20 °C, 3 h |
| Result | First-pass acceptance |
Detailed process control for the forged rim
Because the rim is the forged gear blank portion of the composite, its process control deserves special attention. I required the following steps. First, the ingot was inspected and cleaned. Surface defects such as cracks, laps, and heavy scale were removed before forging. Second, the ingot was heated slowly to the forging temperature. A slow heat-up avoids thermal shock and ensures uniform temperature through the cross-section. Third, the upsetting operation was performed with sufficient reduction to close central porosity. Fourth, the punched hole was centered and sized to allow ring rolling. Fifth, the ring rolling operation was controlled to maintain a uniform wall thickness and to avoid ovality. Sixth, after forging, the ring was cooled slowly, preferably in a pit or under an insulating blanket, to avoid cracking. Seventh, the ring was annealed or normalized before rough machining. Eighth, the ring was quenched and tempered to the final hardness. Ninth, the ring was UT inspected over the entire volume. Tenth, the weld preparation was machined and protected until assembly.
Table 13 lists the forging process controls that I used for the rim of the composite forged gear blank.
| Stage | Control item | Reason |
|---|---|---|
| Ingot preparation | Remove surface defects, clean | Prevent defect inheritance |
| Heating | Slow, uniform heat-up | Avoid thermal stress and cracking |
| Upsetting | Sufficient reduction | Close central porosity |
| Punching | Centered, proper diameter | Prepare for ring rolling |
| Ring rolling | Uniform wall thickness, controlled ovality | Dimensional consistency |
| Post-forge cooling | Slow cooling | Avoid cracking |
| Pre-machining heat treatment | Anneal or normalize | Relieve stress, improve machinability |
| Quench and temper | 250–280 HBW | Strength, toughness, fatigue resistance |
| Ultrasonic testing | Class II | Ensure rim soundness |
| Weld preparation | Machined U-groove | Full penetration weld |
Weld distortion control and residual stress management
Distortion is a major concern when joining a large forged ring to a cast hub. The weld shrinkage forces can pull the ring out of round and can tilt the hub relative to the rim. To control distortion, I used the following measures. The parts were preheated uniformly. The welding was performed by two or more welders working on opposite sides. The welding sequence was balanced, with short passes and alternating directions. The interpass temperature was maintained within the specified range. After welding, the part was covered and allowed to cool slowly. Before stress relief, the part was checked for runout and roundness. If localized distortion was found, it was corrected by machining allowance rather than by forcing. The stress relief treatment then reduced the residual stress without causing further distortion.
The residual stress after welding can be estimated conceptually by the restraint intensity and the yield strength of the weld metal. A simplified expression for the reaction stress in a fully restrained joint is
$$ \sigma_{res} \approx E \alpha \Delta T $$
where \(E\) is the elastic modulus, \(\alpha\) is the coefficient of thermal expansion, and \(\Delta T\) is the temperature change from the stress-free temperature. For steel, \(E \approx 210\) GPa and \(\alpha \approx 12 \times 10^{-6}\) per °C. If \(\Delta T = 250\) °C, the thermal stress would be on the order of 630 MPa, which exceeds the yield strength of the base metal. This is why preheat, balanced welding, and stress relief are essential. The actual residual stress is lower because of plastic deformation and creep, but the estimate shows the severity of the problem. Table 14 lists the distortion control measures.
| Measure | Purpose |
|---|---|
| Uniform preheat | Reduce temperature gradient |
| Multiple welders | Balance heat input |
| Short, alternating passes | Limit local shrinkage |
| Interpass temperature control | Avoid excessive heat buildup |
| Slow cooling under blanket | Prevent hard zones and cracking |
| Stress relief | Relax residual stress |
| Dimensional check before final machining | Ensure sufficient allowance |
Machining strategy after composite joining
After stress relief, the composite forged gear blank is a single structural unit. The final machining strategy must account for the fact that the rim is forged and the hub is cast. The cutting tools will encounter two materials with slightly different machinability. The forged rim is generally more uniform and may produce longer chips. The cast hub may have a skin or localized hard spots. I used coated carbide inserts with moderate speeds and feeds for turning. For gear cutting, I used high-speed steel or carbide hobs depending on the module. The cutting speed was selected to avoid chatter and to produce a good surface finish at the tooth root. Because the rim is a forged gear blank, the tooth root material is dense and free of porosity, which reduces the risk of breakout or edge chipping during cutting.
Table 15 lists the machining strategy for the composite forged gear blank.
| Operation | Tooling | Parameter consideration |
|---|---|---|
| Semi-finish turning | Coated carbide | Moderate speed, uniform allowance |
| Finish turning | Coated carbide or ceramic | High accuracy, low runout |
| Gear cutting | Hob or milling cutter | Rigid setup, proper coolant |
| Keyway machining | Slotter or broach | Straightness and symmetry |
| Deburring | Hand or robotic tools | Remove sharp edges |
Advantages of the composite forged gear blank
The composite forged gear blank provides several advantages that I have observed in production. First, the tooth region is made from a forged material with a continuous grain flow and no casting porosity. This directly improves bending fatigue and contact fatigue life. Second, the hub can be cast with complex internal features, large fillets, and a lower inspection class, which reduces cost. Third, the overall blank weight can be optimized because the rim and hub can have different section thicknesses without the constraints of a single casting. Fourth, the manufacturing risk is distributed: if the hub casting has a minor indication, it may still be acceptable, while the forged rim is inspected to a higher standard. Fifth, the composite design allows smaller forging equipment to be used because only the ring is forged, not the entire gear body. Sixth, the delivery time can be shorter because the hub and rim can be manufactured in parallel. Table 16 summarizes these advantages.
| Advantage | Explanation |
|---|---|
| High tooth fatigue strength | Forged gear blank rim with favorable grain flow |
| Lower cost | Cast hub, forged ring only where needed |
| Design flexibility | Different materials and sections for hub and rim |
| Reduced manufacturing risk | Staged inspection and parallel production |
| Smaller equipment requirement | Only the ring must be forged |
| Shorter delivery | Hub and rim can be produced simultaneously |
| Better inspectability | Rim UT class II, hub UT class III |
Limitations and precautions
The composite forged gear blank is not a universal solution. It introduces a welded joint, which must be qualified and inspected. The welding procedure must be carefully controlled to avoid cold cracking, lack of fusion, and distortion. The pin holes must be located away from the tooth roots. The stress relief temperature must be below the tempering temperature of both base materials. The fit between the hub and rim must be accurate enough to maintain concentricity. The manufacturing sequence must be planned so that the rim and hub are machined with sufficient allowance for welding distortion. If these precautions are not followed, the composite design can fail at the joint. Table 17 lists the main limitations and the corresponding precautions.
| Limitation | Precaution |
|---|---|
| Welded joint is a potential weak point | Qualified WPS, preheat, NDT, stress relief |
| Distortion during welding | Balanced welding, multiple welders, slow cooling |
| Pin holes near tooth root | Careful layout, no pin under root |
| Material compatibility | Similar carbon and alloy content, weldability test |
| Stress relief temperature | Below original tempering temperature |
| Machining allowance | Extra allowance for weld distortion |
| Inspection complexity | Staged UT, MT, hardness, dimensional checks |
Process route summary for the composite forged gear blank
For clarity, I summarize the complete process route that I used for the composite forged gear blank. The hub route is: pattern making, casting, annealing, rough machining, quench and temper, UT class III, finish machining of mating surfaces, weld groove preparation. The rim route is: ingot preparation, forging, ring rolling, post-forge heat treatment, rough machining, quench and temper, UT class II, finish machining of mating surfaces, weld groove preparation. The assembly route is: cleaning, alignment, pinning, preheating, welding, slow cooling, magnetic particle testing, stress relief, hardness testing, final machining, gear cutting, keyway machining, final inspection. Table 18 presents this route in a compact form.
| Phase | Hub | Rim | Joint and final |
|---|---|---|---|
| Forming | Casting | Forging and ring rolling | — |
| Heat treatment | Anneal, Q&T | Post-forge anneal, Q&T | Stress relief |
| Inspection | UT class III | UT class II | MT, hardness, dimensional |
| Machining | Rough and finish mating surfaces | Rough and finish mating surfaces | Final turning, gear cutting, keyway |
| Joining | Pin holes | Pin holes | Pinning, welding |
Quality assurance and documentation
For a critical gear of this size, documentation is as important as the hardware. I required the following records: material certificates for the rim forging and hub casting; forging process records including temperatures and reduction ratios; heat treatment charts for annealing, quenching, tempering, and stress relief; ultrasonic testing reports for the hub and rim; magnetic particle testing reports for the weld; hardness test reports for the rim, hub, and heat-affected zone; welding procedure specification and welder qualification records; dimensional inspection reports for the final gear; and tooth accuracy records. These documents provide traceability and support future maintenance decisions. Table 19 lists the documentation package.
| Document | Content |
|---|---|
| Material certificates | Composition, mechanical properties, heat number |
| Forging record | Temperature, reduction ratio, ring rolling data |
| Heat treatment charts | Annealing, Q&T, stress relief |
| UT reports | Hub class III, rim class II |
| MT reports | Weld and tooth surface |
| Hardness reports | Rim, hub, HAZ |
| Welding documents | WPS, PQR, welder qualification |
| Dimensional reports | Bore, faces, runout, tooth thickness |
| Tooth accuracy reports | Pitch error, profile, lead |
Practical lessons from the project
The most important lesson I learned from this project is that the forged gear blank should be used where the stress is highest, not necessarily for the entire component. By limiting the forged material to the rim, I obtained the fatigue benefits of forging without the cost and size limitations of a fully forged gear. The cast hub provided the shape freedom needed for the mounting and torque transmission features. The welded and pinned joint, although more complex than a monolithic blank, was manageable with a qualified procedure. The key was to control preheat, heat input, cooling rate, and stress relief. The use of multiple welders working simultaneously was essential for distortion control. The pin layout had to be coordinated with the tooth cutting plan so that no pin hole would fall under a tooth root. The UT acceptance classes had to be assigned according to the criticality of each region: class II for the rim and class III for the hub. This staged inspection saved cost and time.
Another practical lesson is that the rim forging must have a sufficient forging ratio. A ratio below 3 may leave residual porosity or a cast dendritic structure that can reduce fatigue life. I specified a minimum ratio of 3 and verified it from the forging records. The ring rolling operation must produce a uniform wall thickness; otherwise, the subsequent machining will not clean up evenly. The weld groove must be machined after the final heat treatment of each part; otherwise, the groove may distort or oxidize. The preheat must be uniform and must not overheat any local area. The stress relief temperature must be selected with care. If it is too high, the base metal hardness will drop. If it is too low, the residual stress will remain. I found that \(520 \pm 20\) °C for 3 h was a good compromise for 35CrMo and ZG35CrMo at 250–280 HBW.
Design checks for the composite joint
In addition to the shear checks described earlier, I performed a simple torsional check on the welded ring. The polar moment of inertia of a thin circular weld is approximately
$$ J_w = 2 \pi r_j^3 t_w $$
The shear stress at the weld is then
$$ \tau_w = \frac{T r_j}{J_w} = \frac{T}{2 \pi r_j^2 t_w} $$
For a torque of \(T = 1.2 \times 10^6\) N·m, a joint radius of \(r_j = 1.2\) m, and a throat thickness of \(t_w = 25\) mm, the nominal shear stress is
$$ \tau_w = \frac{1.2 \times 10^6}{2 \pi (1.2)^2 (0.025)} \approx 5.3 \text{ MPa} $$
This value is low compared with the allowable shear strength of the weld metal, which is typically on the order of 200 MPa. The pins carry a significant portion of the load and provide redundancy. The actual stress distribution is not uniform because of the stiffness of the hub and rim, but the calculation shows that the joint has ample nominal capacity. The more critical issue is fatigue and stress concentration, which is why the weld toe must be smooth and the pins must be properly fitted. Table 20 lists the joint design checks.
| Check | Formula | Result or action |
|---|---|---|
| Pin shear | \(\tau_p = T / (n \pi d_p^2 r_j / 4)\) | Select pin diameter and number |
| Weld torsion | \(\tau_w = T / (2 \pi r_j^2 t_w)\) | Verify nominal shear stress |
| Weld throat | \(t_w = 0.707 a\) for fillet, or groove depth | Ensure sufficient throat |
| Pin location | \(\Delta \theta = 360^\circ / n\) | Avoid tooth root |
| Stress relief | \(P_{LM} = T(\log_{10} t + 20) \times 10^{-3}\) | Confirm tempering condition |
Comparison with alternative repair methods
When a large cast gear blank fails inspection because of rim defects, there are several possible responses. One is to scrap the casting and try again. This is expensive and delays the project. Another is to repair weld the defect. Repair welding of a large cast steel gear is risky because the heat-affected zone can become hard and brittle, and the repair may not restore fatigue strength. A third option is to use a composite forged gear blank. In my case, the composite approach was more reliable than repair welding and less expensive than a fully forged gear. Table 21 compares these options.
| Option | Cost | Quality risk | Delivery risk |
|---|---|---|---|
| Scrap and recast | High | May repeat defects | High |
| Repair weld cast rim | Moderate | High, fatigue strength uncertain | Moderate |
| Fully forged replacement | Very high | Low if available | High for large size |
| Composite forged gear blank | Moderate | Low with qualified procedure | Moderate |
Recommendations for future designs
Based on my experience, I recommend the composite forged gear blank for large gears where the diameter exceeds the economic or technical limit of a fully forged blank and where the rim quality is critical. The design should be considered early in the project so that the hub and rim can be procured in parallel. The material selection should prioritize weldability and similar hardenability. The welding procedure should be qualified before production. The pin layout should be coordinated with the tooth geometry. The inspection plan should assign the highest UT class to the rim and a lower class to the hub. The stress relief treatment should be validated by hardness and impact testing. The final machining should include enough allowance for weld distortion. Table 22 summarizes my recommendations.
| Recommendation | Reason |
|---|---|
| Use a forged gear blank for the rim | Highest stress and fatigue risk |
| Use a casting for the hub | Shape freedom and lower cost |
| Select similar weldable materials | Avoid hard, brittle interface |
| Qualify the welding procedure | Ensure repeatable quality |
| Use pins away from tooth roots | Avoid stress concentration |
| Stress relieve below tempering temperature | Preserve base metal properties |
| Inspect rim to class II and hub to class III | Balance quality and cost |
| Allow extra machining stock | Compensate for weld distortion |
Conclusion
The composite forged gear blank is a practical solution for large gears that cannot be produced economically as a single forging and cannot be produced reliably as a single casting. By separating the gear into a forged rim and a cast hub, I was able to place high-quality forged material in the tooth region while using a lower-cost casting for the less critical hub. The joint was secured by cylindrical pins and a full-penetration weld, and the entire assembly was stress relieved and final machined. The process required careful control of material selection, preheat, welding, cooling, stress relief, and inspection. In the converter tilting reducer application, the composite forged gear blank achieved first-pass acceptance and avoided the cost and delay of a fully forged or fully cast alternative. For future large gear projects, I will continue to consider the composite forged gear blank whenever rim quality, manufacturing capacity, and cost must be balanced. The forged gear blank rim provides the fatigue resistance needed for the teeth, while the cast hub provides the geometric freedom needed for the rest of the gear. With a qualified procedure and a disciplined quality plan, this composite design is a reliable and economical choice.