This Article Will Introduce You to Gears

• 1
  How a gear pair transmits motion, and what to introduce you to gears means in practice
• 2
  Tooth forms: where each one stops being the right answer
• 3
  Backlash, runout and the errors that show up after assembly
• 4
  When to machine a gear instead of hobbing it
• 5
  Materials, heat treatment and the cost of getting it wrong
• 6
  Gear type and process selection at a glance

A gear is a wheel with evenly spaced teeth cut around a pitch circle. Two gears in mesh act as a chain of levers: the driving tooth pushes the driven tooth along the line of action, and the contact point rolls rather than slides. That rolling contact is the whole point. It lets one shaft turn another at a fixed speed ratio with very little loss, provided the teeth are cut to the same module and pressure angle.

When we introduce you to gears on the shop floor, the first number we check is the module. Module is the pitch diameter in millimeters divided by the number of teeth. Two gears only mesh if their modules match. A 20-tooth gear at module 2 has a pitch diameter of 40 mm; the mating 40-tooth gear is 80 mm. Mismatch the module by 0.1 mm and the teeth will bind or slap, no matter how good the rest of the part is.

The second number is the pressure angle, usually 20° for modern drives and 14.5° for older or lightly loaded ones. Pressure angle sets how much of the force pushes along the line of centers and how much pushes sideways. A 20° tooth carries more load per unit width than a 14.5° tooth, but it also pushes harder on the bearing. That trade-off shows up as shaft deflection and noise.

The ratio comes straight from tooth counts. A 60-tooth driven gear behind a 20-tooth pinion gives a 3:1 reduction. Speed drops by 3, torque rises by roughly 3, minus losses. For a compound train, multiply each pair's ratio. There is no magic in the gearbox; it is arithmetic.

• 1
  Module must matchPitch diameter ÷ tooth count; same value for both gears.
• 2
  Pressure angle sets load direction20° is the default for loaded drives.
• 3
  Ratio = driven teeth ÷ driver teethCompound trains multiply the individual ratios.

Spur gears are the simplest: straight teeth, parallel shafts, no axial thrust. They are cheap to cut and easy to inspect. The downside is noise and shock loading, because contact starts and ends abruptly across the full face width. For a low-speed indexing axis or a prototype reduction box, spur is usually the right first choice.

Helical gears lean the teeth at an angle. Contact starts at one end of the face and rolls across, so load transfer is gradual and the pair runs quieter. The price is axial thrust, which the bearings must absorb, plus a more complex cut. A helical gear needs either a bearing arrangement that takes thrust or a matched pair with opposite hands to cancel it.

Bevel gears turn the axis by 90°. Straight bevel teeth are cut on a cone and work well at moderate speed. Spiral bevel teeth are curved and engage more gradually, so they carry more torque with less noise, which is why differentials use them. Cutting a good spiral bevel requires a dedicated machine or a 5-axis setup with the right tool path.

Worm gears give a large reduction in one stage, often 20:1 to 100:1, with the axes at 90° and non-intersecting. The sliding contact makes them quiet and self-locking at low lead angles, but it also generates heat and wears the bronze wheel. Use worm drives for intermittent motion, not for continuous high-speed duty.

Rack and pinion converts rotation to straight travel. The rack is a gear with an infinite pitch diameter, so the same module and pressure angle rules apply. Backlash here translates directly into positioning error, which matters on a CNC axis or a linear stage.

• 1
  SpurParallel shafts, low cost, noisy at speed.
• 2
  HelicalQuieter and stronger, but creates axial thrust.
• 3
  Bevel and spiral bevel90° axes; spiral carries more torque quietly.
• 4
  WormBig single-stage reduction, heat and wear at speed.

Backlash is the gap between meshing teeth when the driving gear is held still. Some backlash is required. Teeth need room for lubricant, for thermal growth and for the small manufacturing errors that always exist. A gear pair with zero backlash at room temperature will jam once the housing warms up or a chip lodges in the mesh.

For a module 2 steel pair in a general drive, 0.05–0.15 mm of circumferential backlash is a normal starting band. Instrument gearing runs tighter, often 0.01–0.03 mm. The number should come from the application, not from habit. A positioning axis wants the low end; a high-torque drive that will heat up wants the high end.

Runout is the wobble of the pitch circle relative to the bore. If the bore and the pitch circle are not concentric, the center distance changes once per revolution. That shows up as a once-per-rev velocity ripple, audible as a whine and visible as position error. The usual cause is a soft-jaw turning setup that let the part shift, or a hobbing fixture clamped on a rough surface.

Lead error is the drift of the tooth along the face width. It makes the tooth contact concentrate at one end, which raises local stress and wear. On a machined gear, lead error usually comes from tool deflection or from a feed rate that was too high for the tooth depth.

We check bore-to-pitch concentricity, tooth-to-tooth spacing and total composite error. For a prototype gear, a simple functional mesh test against the mating part tells you more than a stack of numbers. If it turns smoothly by hand with light drag, the geometry is close.

• 1
  Backlash0.05–0.15 mm typical for module 2 steel; tighter for instruments.
• 2
  RunoutBore-to-pitch eccentricity causes once-per-rev ripple.
• 3
  Lead errorContact shifts to one end of the face; stress rises.

Hobbing is the production answer. A hob rolls through the blank and generates the tooth form in one continuous pass. It is fast, repeatable and cheap per part once the setup is amortized. For a run of a few hundred identical spur or helical gears, hobbing or shaping wins on cost every time.

Machining wins in three cases. The first is low quantity: one prototype or a handful of replacement gears does not justify a hob and a fixture. The second is geometry that a hob cannot reach, such as a tooth form blended into a pocket, an internal spline with a non-standard profile, or a gear cut into a part that also carries bearing bores and mounting faces. The third is speed: when the design is still moving, cutting from a block of 7075 or 17-4PH lets you test the ratio before committing to tooling.

On our 5-axis centers we can cut a module 2 spur gear in 6061 or 1045 with a 2 mm end mill and a small corner radius, then verify the mesh against the mating part. Tolerance on the bore and the pitch diameter typically holds at ±0.005 mm. Surface finish on the flanks lands around Ra 0.8–1.6 μm, which is fine for a running-in pair and adequate for many instrument drives.

The limit is tool access. A small module gear with a deep, narrow tooth space needs a cutter thin enough to enter and stiff enough not to chatter. Below about module 0.5 in steel, a machined tooth becomes slow and the finish suffers. That is the point to switch to hobbing, shaping or wire EDM, or to change the material to brass or a plastic like POM or PEEK.

Material choice follows the duty. 1045 and 4140 are the workhorses for steel gears; 17-4PH gives corrosion resistance and higher strength for aerospace and medical parts; 7075 aluminum is light and machines fast but wears quickly without a hard anodized or electroless nickel layer. Bronze and brass run quietly against steel and are forgiving of slight misalignment.

• 1
  Choose hobbingRuns of hundreds, standard spur or helical forms.
• 2
  Choose CNC machiningOne-offs, complex geometry, integrated features, fast iteration.
• 3
  Watch the limitBelow module 0.5 in steel, tool access and finish suffer.

A gear fails at the tooth, not in the middle of the blank. Bending stress at the root and contact stress on the flank set the life. Case hardening puts a hard, wear-resistant layer on the flank while keeping a tough core that resists cracking. For a loaded steel gear, that combination is usually better than through-hardening the whole part.

Through-hardened 4140 at 28–32 HRC is a reasonable general-purpose choice. It machines before heat treatment, so the tooth form is cut soft and then hardened. The catch is distortion: quenching moves the bore and the pitch circle, so a hardened gear often needs the bore ground after treatment. Budget for that step or design a bearing fit that tolerates it.

Plastic gears change the rules. POM and PA run without lubrication, absorb shock and are quiet, but they carry far less load and grow with temperature and moisture. They work well against a steel pinion if the plastic gear is the sacrificial part. PEEK handles higher temperature and load but costs more and is harder to cut cleanly.

Surface finish on the flank matters more than most people expect. A rough flank wears the mating tooth and raises friction, which raises temperature. Ra 0.8–1.6 μm is a practical target for a machined steel gear; finer finishes help high-speed pairs but add cost.

One more boundary: lubrication. A gear pair designed for oil will not survive on grease at the same load, and a pair designed to run dry will fail quickly if grease is added and traps debris. Match the lubrication plan to the duty cycle before the drawing is released.

• 1
  Case hardeningHard flank, tough core; watch distortion on the bore.
• 2
  PlasticsQuiet and self-lubricating, but low load and thermal growth.
• 3
  Flank finishRa 0.8–1.6 μm is a practical machined target.