Mass Production of Large Gears: How the Process Actually Works

• 1
  What makes a large gear different to machine
• 2
  Process routes for mass production of large gears
• 3
  Where distortion comes from in gear blanks
• 4
  Cutting parameters that hold on large modules
• 5
  Heat treatment sequence and its effect on final size
• 6
  Inspection: what to measure and when
• 7
  Choosing a route for mass production of large gears

A large gear is not just a small gear scaled up. As diameter grows, the same tooth error sits at a bigger radius, so a 20 μm flank deviation on a Ø2,000 mm gear shows up as a larger linear error at the pitch circle than on a Ø200 mm gear. Cutting forces also rise with the chip cross-section, and those forces push the blank away from the tool. Stiffness of the setup becomes a cutting parameter, not a shop-floor detail.

Heat is the second difference. A large ring gear has a lot of material, and roughing it removes tens of kilograms of stock. The blank absorbs heat unevenly, cools unevenly, and moves. If the gear is finish-cut before that movement stops, the teeth come out fine on the machine and off-size a day later.

So the mass production of large gears is really a scheduling problem wrapped around a machining problem. Blank, stress relief, semi-finish, finish, inspection. Each stage exists to remove a specific source of error before the next stage can hold tolerance.

• 1
  Radius amplifies errorFlank deviations grow as linear error with pitch diameter.
• 2
  Chip load drives deflectionLarge modules mean deep cuts and high radial force.
• 3
  Thermal drift needs timeStress relief between roughing and finishing is not optional.

There are three practical routes. The first is hobbing followed by profile or generating grinding. It suits hardened gears where flank finish and profile tolerance matter, and it is the route most wind and industrial gearbox work takes. The second is hobbing plus shaving or honing for soft gears that do not need a hardened flank. It is faster and cheaper per piece, but the flank surface is rougher.

The third route is form milling or slotting for very large modules and internal teeth, where a hob cannot reach or the module is too coarse for economical hobbing. Slotting is slow. It is used when geometry leaves no alternative, such as internal ring gears in a planetary set.

On 5-axis machining centers we can mill the tooth flank directly for coarse modules and for repair work, then grind if the drawing calls for it. That route is common for one-off replacements and for gears where the lead time of a dedicated hob is longer than the machining time itself. For true mass production, dedicated hobbing and grinding still win on cost per piece.

• 1
  Hardened flankHob, heat treat, then grind for profile and lead.
• 2
  Soft flank, high volumeHob and shave; skip grinding if Ra and profile allow.
• 3
  Internal or coarse moduleSlot or form mill, then finish if required.

Distortion has three main sources: residual stress from the blank itself, heat from machining, and clamping force. Forged and cast blanks arrive with locked-in stress from cooling. The first roughing pass releases it, and the part bends. That is why roughing leaves 2–3 mm of stock per flank on large gears rather than finishing close to size.

Heat is more predictable. A heavy roughing cut on a 4140 ring gear can raise the local surface temperature enough to shift the part by tens of microns. Flood coolant and moderate depth of cut keep this manageable, but the part still needs to return to room temperature before a finishing pass. Measuring a warm gear is measuring the wrong part.

Clamping force matters most on thin rings. A ring gear with a wall thickness under 25 mm will ovalise under a strong three-jaw chuck. For those parts, we use a face plate with soft jaws machined to the actual diameter, or clamp on a sacrificial boss and cut it off after finishing.

• 1
  Leave stock for stress release2–3 mm per flank after roughing on large modules.
• 2
  Let the part coolBring it back to room temperature before finishing.
• 3
  Match clamping to wall thicknessThin rings need soft jaws or a sacrificial boss.

For hobbing a 4140 gear blank at 280–320 HB, surface speed usually lands between 60 and 90 m/min with coated carbide hobs, and feed per revolution sits near 1–2 mm depending on module. Push surface speed higher and the hob edge breaks down fast on interrupted cuts. Push feed higher and the flank lead error grows.

Grinding runs much slower in terms of material removal but much faster in surface speed. Vitrified CBN or aluminium oxide wheels run at 30–45 m/s for profile grinding of case-hardened gears. Depth of cut per pass stays small, typically 0.02–0.05 mm, with several spark-out passes at the end. Spark-out is what removes the elastic deflection that would otherwise leave a tapered flank.

For 5-axis flank milling of coarse teeth, we use trochoidal or high-efficiency milling paths with radial engagement around 5–10% of tool diameter. This keeps radial force low and lets a smaller tool reach into a deep tooth space without chatter. Cycle time is longer than hobbing, but setup is simpler and no dedicated hob is needed.

• 1
  Hobbing speed60–90 m/min on 4140 at 280–320 HB.
• 2
  Grinding depth0.02–0.05 mm per pass with spark-out at the end.
• 3
  5-axis milling5–10% radial engagement to control chatter.

Case carburising at 925–950 °C followed by quenching moves a large gear. The case grows, the core may grow or shrink depending on hardenability, and the part distorts in a pattern that depends on section thickness. The rule is simple: heat treat before final grinding, and leave enough stock for the grinder to correct the distortion.

How much stock? On a Ø800 mm carburised ring gear, 0.3–0.5 mm per flank is a normal allowance. On a Ø2,000 mm part with uneven sections, it can be more. If the allowance is too small, the grinder removes the case on one side before it cleans up the other, and the tooth is soft there. That is a scrap part, not a rework part.

Induction hardening is different. It heats only the flank surface, so overall distortion is small, but the hardened depth is shallow and the transition zone can crack if the part is ground aggressively afterwards. For large gears, carburising and grinding is the more predictable route, even though it costs more per piece.

• 1
  Grind after heat treatDistortion is corrected, not avoided.
• 2
  Allowance for case0.3–0.5 mm per flank on a Ø800 mm ring gear.
• 3
  Induction hardeningLow distortion, shallow case, watch the transition zone.

Gear inspection is not one measurement at the end. Blank dimensions are checked before hobbing, because a blank that is out of round will produce a gear that is out of round no matter how good the cutting is. After hobbing, we check tooth thickness, runout and lead. After grinding, we check profile, lead and pitch again on the finished flank.

For large gears, a dedicated gear measuring center is the practical tool. It traces the flank, reports profile and lead deviation, and gives a DIN or AGMA grade. For gears above 2,000 mm, the part may need to be measured on the machine or with a portable setup, because moving a part that size to a measuring room introduces its own error.

Runout is the measurement that catches most production problems. A gear can have perfect flanks and still run out if the bore and the pitch circle are not concentric. On mass production runs we check runout on every part, not on a sample, because runout drift usually comes from a fixture that has started to wear.

• 1
  Blank checkRoundness and bore before any tooth is cut.
• 2
  Post-hob checkTooth thickness, lead and runout.
• 3
  Final checkProfile, lead and pitch on the ground flank.