CNC machining gears for high-precision applications

• 1
  Key takeaways
• 2
  What high-precision gear geometry actually demands
• 3
  Milling, turning or grinding: picking the right cut
• 4
  Material and heat treatment choices
• 5
  How gear accuracy is measured
• 6
  When CNC machining gears is the wrong call
• 7
  From drawing to inspected gear
• 8
  Which gear process fits your part

A gear is not a disc with teeth. Every flank carries load, so the shape of the involute, the depth of the root fillet and the lead of the tooth all change how the part behaves in a gearbox. When we quote CNC machining gears for high-precision applications, the first question is which of those three features is critical. A low-speed positioning gear may only need a good pitch diameter and a clean bore. A gear running at 8,000 rpm will also need lead correction and a controlled surface finish.

Module and pressure angle set the starting point. Modules from 0.5 to 6 mm are common in machined gears; below 0.5 mm the cutter diameter gets small and deflection becomes the limiting factor rather than the machine. Pressure angles of 20° are standard, and 14.5° still appears in older replacement parts. A change in pressure angle changes the whole profile, so it has to be confirmed before programming, not after the first article.

Backlash is a system property, not a single-part property. The gear you machine will mesh with a mating part that has its own tolerance, so the drawing should state the minimum and maximum backlash for the pair, not for one gear. We often see drawings that specify a tooth thickness with no reference to the mate, which makes the inspection result impossible to interpret. Give us the centre distance and the mating part, and the tooth thickness window can be calculated.

Fillet radius is where most failures start. A sharp root concentrates stress and a crack begins there under cyclic load. A full fillet with a radius of roughly 0.38 × module is the usual rule for bending strength, but it competes with the space needed for the mating tip. If the design is tight, tell us the bending load and we will suggest a radius that can actually be cut with a standard tool.

• 1
  Module and pressure angleConfirm both before programming; they define the entire profile.
• 2
  Lead and crownAdd only if the gearbox needs it; they raise inspection time.
• 3
  Root filletAim for about 0.38 × module unless space forces a smaller radius.

For small and medium batches, cutting the teeth on a machining centre is the fastest route. A 4-axis mill with the gear on a rotary table indexes to each tooth position and cuts the gap with a form tool or a small end mill. This works well for spur gears, internal gears and timing pulleys. The setup is simple, the first article comes off in days, and there is no hob to buy.

A 5-axis machine adds the ability to tilt the tool. That matters for helical gears, because the tool can follow the helix in one continuous pass instead of being indexed in steps. It also lets us reach the root fillet at the correct angle without a special cutter. Our shop runs 16 simultaneous 5-axis machining centres, and most helical gear work goes there. A Ø400 mm rotary table covers gears that would otherwise need a much larger machine.

Turning is the better choice for gear blanks where the bore, face and outside diameter share a single datum. A mill-turn centre cuts the blank and the teeth in one setup, so the pitch diameter and the bore stay concentric. That concentricity is often what limits gear noise more than the tooth profile itself. We have 16 mill-turn centres for this kind of work.

Grinding comes in after heat treatment. Once a gear is hardened to 58 HRC or above, carbide cannot cut it economically, so the flanks are ground. Grinding also gives the tightest lead and profile, and it produces the fine finish that quiet gearboxes need. The trade-off is cost per piece and lead time. If your gear does not need hardness, skip grinding and spend the money on a better inspection plan.

• 1
  4-axis millingSpur gears, internal gears, pulleys. Fast setup, no hob cost.
• 2
  5-axis millingHelical gears and fillets that need a tilted tool path.
• 3
  Mill-turnOne setup for bore, faces and teeth; best concentricity.
• 4
  GrindingAfter hardening, or when lead and finish are the tight calls.

The material sets both the cutting strategy and the finishing step. Aluminium 6061-T6 and 7075 are easy to cut and hold a good profile, which makes them the default for prototype gears and low-load drives. They are not good for high surface pressure. If a gearbox sees real torque, steel is the answer: 1045 and 4140 for general machinery, 4340 when the section is thick and toughness matters.

Stainless grades 303, 304 and 17-4PH appear often in food, medical and marine equipment. Grade 303 machines cleanly but is not ideal for high load. Grade 17-4PH is the better structural choice and can be aged to around 40 HRC, which keeps the teeth dimensionally stable after heat treatment. Grade 316L is chosen for corrosion, not for strength.

Brass and bronze are used for worm wheels, bushings and low-noise drives. C36000 cuts fast and gives a smooth flank. Beryllium copper is occasionally specified for springs and contacts inside a gear assembly, but it needs its own handling controls. Titanium TC4 (Ti-6Al-4V) and Inconel are available for aerospace and high-temperature parts; both cut slowly and the cost reflects that.

Heat treatment is a separate decision from material. Case hardening gives a hard skin with a tough core and is common for 1018 and 4140 gears. Through hardening to 45–58 HRC suits smaller gears. Nitriding adds a hard surface with almost no distortion, which helps when the gear is already finished to size. Any hardening step changes dimensions, so the drawing should say which surfaces are finished before and after.

• 1
  AluminiumPrototypes and light drives; cut to final size, no grinding.
• 2
  Alloy steel4140 and 4340 for real torque; plan for hardening and grind.
• 3
  17-4PHCorrosion plus strength; age to about 40 HRC.
• 4
  Brass and bronzeWorm wheels and quiet drives; excellent machinability.

A gear is hard to inspect because the features that matter are not straight lines. Pitch diameter is usually measured over wires or pins, which is fast and repeatable. Tooth thickness is measured the same way, or with a span gauge across several teeth. Both checks need the correct wire diameter for the module, so the inspection plan depends on the drawing values.

Runout and concentricity are checked on a CMM or between centres with a dial indicator. A gear with a good profile but a bad bore will still run loud, so bore-to-pitch-diameter concentricity is often the first number we look at. For tighter work, a profile scanner traces the flank and reports deviation from the theoretical involute along the whole tooth.

Surface finish matters more than most drawings admit. A flank at Ra 1.6 μm and a flank at Ra 0.4 μm can have the same nominal profile and behave differently in a gearbox. Finer finishes reduce friction and noise, and they also reduce the chance of a crack starting at a machining mark. We hold Ra 0.8–1.6 μm as standard and can reach Ra 0.2–0.8 μm when the drawing calls for it.

Inspection reports are only useful if they answer the question the designer asked. Send the functional requirement with the drawing, and the report will show the numbers that prove or disprove it. A stack of data with no reference to function wastes everyone's time.

• 1
  Pitch diameterMeasured over wires or pins; needs the correct wire size.
• 2
  RunoutCMM or dial indicator; checks bore-to-pitch concentricity.
• 3
  Profile traceScans the involute for tight-tolerance gears.
• 4
  Surface finishAffects noise and fatigue life, not just appearance.

CNC machining is not always the cheapest way to make a gear. If you need 50,000 identical spur gears a year, hobbing or injection moulding will beat us on unit price. Machining wins when volume is low, when the geometry is unusual, or when the part must be made from a material that cannot be cast or moulded to tolerance.

Very small modules are another limit. Below 0.5 mm module, the cutter is thin, deflection is hard to control and the inspection equipment becomes the bottleneck. If your design is at 0.3 mm module, talk to a specialist gear shop before assuming a machining centre is the right tool.

Internal gears with a long bore and a small opening are difficult to reach. The tool has to pass through the bore and still cut the root, which limits the length-to-diameter ratio. In those cases we usually suggest wire EDM for the tooth form, or a split housing that lets us cut from the open side.

Large gears are possible up to 4,000 mm of processing size, but the tolerance has to be realistic. Holding ±0.005 mm on a 1 m diameter gear is a different problem from holding it on a 60 mm gear, because temperature and clamping distortion grow with size. Tell us the operating temperature and we will recommend a tolerance that survives it.

• 1
  High volumeHobbing or moulding will be cheaper above roughly 50,000 pcs.
• 2
  Module below 0.5 mmCutter deflection and metrology become the limit.
• 3
  Deep internal gearsTool reach may force wire EDM or a split design.