A worm-and-pinion feed roll drive is a right-angle reduction stage where a single-start or multi-start worm screw drives a pinion-style worm wheel mounted directly on the feed roll shaft. The worm's helical thread sliding against the wheel teeth converts high-RPM low-torque motor input into low-RPM high-torque roll output, often at ratios from 20:1 to 80:1 in a single stage. The geometry resists back-driving, so the feed roll holds position when the motor stops — useful on paper, textile, and strip-metal lines where the web must not creep during a stop. Most lines see roll surface speeds of 0.1 to 3 m/s with positional repeatability inside 0.5 mm.
Worm-and-pinion Feed Roll Drive Interactive Calculator
Vary worm starts, wheel teeth, roll size, and backlash to see reduction ratio, roll motion, feed advance, and web slack.
Equation Used
The reduction ratio is the worm wheel tooth count divided by the number of worm starts. One worm revolution turns the wheel by 360/i degrees. With the roll mounted on the wheel shaft, feed per worm revolution is the roll circumference divided by the reduction ratio, and backlash slack is the roll surface arc length for the backlash angle.
- Single worm stage with wheel mounted directly to the feed roll.
- No slip between feed roll and web.
- Backlash slack is calculated as arc length at the roll surface.
- Worked example self-locking condition assumes lead angle below about 6 deg.
Inside the Worm-and-pinion Feed Roll Drive
The worm is a screw — usually hardened steel, ground or polished — and the pinion is a bronze or cast-iron worm wheel cut with teeth that match the worm's helix. When the worm rotates, each thread sweeps the wheel teeth sideways by exactly one tooth pitch per worm revolution if the worm is single-start, two teeth if double-start, and so on. That's why reduction ratios stack so quickly in one stage. A 40-tooth wheel driven by a single-start worm gives 40:1 reduction. The same 40-tooth wheel driven by a 4-start worm gives 10:1. You pick the start count to land on the torque and self-locking behaviour you want.
Self-locking comes from the lead angle. If the lead angle of the worm is below the friction angle of the contact pair (typically around 5° to 6° for steel-on-bronze running in oil), the wheel cannot back-drive the worm. The feed roll stays put when the motor de-energises. Push the lead angle above 12° to 15° and you trade self-locking for efficiency — a 20° lead angle worm runs at 80%+ efficiency but will coast and creep under web tension. This is the central design decision and you have to make it consciously.
If tolerances drift, you'll see it in the web. Centre distance must be held within ±0.05 mm on a typical 80 mm centre — go wider and the teeth bottom out, go tighter and the backlash opens up to 0.3° or more, which on a 200 mm-diameter feed roll translates to about 0.5 mm of web slack at every reversal. Common failure modes are bronze-wheel scoring from oil starvation, worm thread wear at the entry side from misaligned axial preload, and broken wheel teeth at the load side when an operator jogs the line under a jammed web. Heat is the giveaway — a healthy worm box runs 30 to 50°C above ambient, and anything above 80°C means the lubricant film has collapsed.
Key Components
- Hardened Steel Worm: The driving screw, typically 20MnCr5 or 16MnCr5 case-hardened to 58-62 HRC and ground to a thread profile tolerance of ±10 µm. Surface finish on the flanks must be Ra 0.4 µm or better — rougher surfaces tear the bronze wheel and shed metallic debris into the oil within 200 hours.
- Bronze Worm Wheel (Pinion): Usually centrifugally cast CuSn12 or CuSn12Ni2 bronze, machine-cut to AGMA Q10 or DIN 7. The bronze wears sacrificially against the harder steel worm — that's intentional. Wheel face width is typically 0.6 to 0.8 × worm diameter to keep contact patch within the central tooth zone.
- Feed Roll Shaft: The wheel mounts directly on the feed roll shaft via a keyed taper bushing or shrink disk. Shaft runout at the wheel must stay under 0.02 mm TIR — anything more and the contact patch walks across the tooth face every revolution, producing a once-per-rev pulsation in web tension that printers and coaters will see as banding.
- Axial Thrust Bearing on Worm: The worm sees high axial thrust — often 60-70% of the tangential force on the wheel. A back-to-back angular contact pair or a tapered roller pair takes this load. Preload must be set to 5-15 µm of axial play; loose preload lets the worm walk and changes timing every cycle.
- Oil Bath or Forced Lubrication: Worm gears live or die by lubrication. ISO VG 460 mineral oil or PAO synthetic at sump temperature 60-80°C is standard. Oil level must reach the worm centreline at rest — too low starves the mesh, too high churns and overheats the box above 90°C.
- Housing with Cooling Surface: Worm boxes shed heat through the housing. A 100 W continuous loss in a 250 mm box raises sump temperature about 30°C above ambient at still air. Forced ventilation or an oil cooler is required above roughly 1.5 kW continuous, otherwise viscosity collapses and wear rate triples.
Where the Worm-and-pinion Feed Roll Drive Is Used
Worm-and-pinion feed roll drives show up wherever a web or strip needs steady, controlled feed with the holding torque to resist back-tension during stops. The combination of high reduction in one stage, right-angle layout, and inherent self-locking makes them a natural fit on lines where servo drives would be overkill or where the operator wants the roll to hold position when power drops.
- Paper Converting: Bobst Mastercut die-cutters use worm-driven feed rolls on the sheet-feeding section to hold register within 0.1 mm sheet-to-sheet at speeds up to 8,000 sheets/hour.
- Textile Finishing: Monforts Montex stenter frames drive the entry feed rolls through worm gearboxes to maintain fabric tension across the pin chain entry, typically running 20 to 80 m/min.
- Metal Strip Processing: Bruderer BSTA high-speed presses pair a roll-feed unit with a worm-and-pinion advance drive for strip pitches of 10-300 mm per stroke at up to 1,500 strokes/minute on the smaller models.
- Corrugated Board: BHS Corrugated single-facer wet-end feed rolls run worm boxes at 80:1 reduction to keep liner board feed steady at 250-400 m/min while resisting back-tension from the glue station.
- Label and Tag Printing: Mark Andy P5 narrow-web presses use worm-driven unwind nip rolls on auxiliary stations where the operator needs the web to lock instantly when an emergency stop fires.
- Wire and Cable: Niehoff multi-wire drawing lines use worm-driven capstan feed rolls upstream of the take-up bobbin to set the line speed reference, typically 5-25 m/s wire speed reduced from 1,750 RPM motor input.
The Formula Behind the Worm-and-pinion Feed Roll Drive
The core sizing calculation for a worm-and-pinion feed roll drive is the relationship between motor speed, gear ratio, and feed roll surface speed. This determines what web speed you actually deliver to the line. At the low end of typical operation — say 30% of nameplate motor speed — the worm box runs cool and efficient but you may be below the speed where the oil film is fully hydrodynamic, increasing wear. At nominal speed the box hits its design sweet spot for both efficiency and film thickness. Push to the high end and heat generation rises with the square of sliding velocity, so you'll see sump temperature climb and efficiency fall.
Variables
| Symbol | Meaning | Unit (SI) | Unit (Imperial) |
|---|---|---|---|
| vroll | Feed roll surface speed (web speed delivered) | m/min | ft/min |
| Droll | Feed roll outer diameter | m | in |
| Nmotor | Motor input speed at the worm | RPM | RPM |
| Zworm | Number of starts on the worm (1, 2, or 4 typical) | dimensionless | dimensionless |
| Zwheel | Number of teeth on the worm wheel | dimensionless | dimensionless |
Worked Example: Worm-and-pinion Feed Roll Drive in a pharmaceutical foil pouch line
Sizing the worm-and-pinion drive on the lidding-foil feed roll of a Bosch SVE 2520 horizontal form-fill-seal machine running aluminium-laminate lidding stock for unit-dose pouches. The motor is a 1,750 RPM 4-pole asynchronous unit driving a single-start worm. The worm wheel has 50 teeth. The feed roll is 120 mm in outer diameter and the line target is roughly 40 m/min foil speed at nominal conditions, with operating range 20-60 m/min depending on pouch size.
Given
- Droll = 0.120 m
- Nmotor = 1,750 (nominal); 875 (low); 2,625 (high via VFD) RPM
- Zworm = 1 starts
- Zwheel = 50 teeth
Solution
Step 1 — compute the gear ratio first, since it sets everything else:
Step 2 — at nominal 1,750 RPM motor speed, calculate the feed roll output RPM:
Step 3 — convert to surface speed at the nominal operating point:
The 40 m/min target tells you a single-start worm at 1:50 won't reach the line speed at base motor RPM. You either move to a 3-start worm (giving 50/3 ≈ 16.7:1, raising v to 39.6 m/min at nominal) or you push the motor with a VFD. Step 4 — re-run with a 3-start worm at nominal:
That hits the target. Step 5 — at the low end of the operating range with the VFD at 875 RPM:
This is the slow-speed regime used for small 20 mm pouches. The worm box runs cool, around 35°C above ambient, but you're below the speed where hydrodynamic oil film fully forms — expect bronze wheel wear roughly 1.5× the rate at nominal. Step 6 — at the high end with VFD at 2,625 RPM (1.5× base):
Theoretically clean, but in practice the sump temperature on a 250 mm worm box climbs to 75-85°C at this speed because sliding velocity at the mesh is now around 8 m/s. You'll need forced cooling or a synthetic PAO oil to hold viscosity, otherwise efficiency drops from ~75% to under 60% within an hour and the box starts cycling on thermal protection.
Result
Nominal feed roll speed lands at 39. 6 m/min with a 3-start worm and 50-tooth wheel at 1,750 RPM motor input — within 1% of the 40 m/min line target. At the low end (875 RPM) you get 19.8 m/min, slow enough for small-format pouches but below the hydrodynamic film threshold so wear rate climbs; at the high end (2,625 RPM) you get 59.4 m/min on paper, but heat becomes the limit and you'll need synthetic oil and forced ventilation to hold the box below 80°C sump. If you measure 32 m/min instead of the predicted 39.6 m/min, three failure modes are most likely: (1) the worm-wheel bronze has worn enough to open backlash above 0.5°, eating effective stride per rev, (2) the keyed taper bushing on the feed roll shaft has slipped under repeated jam-jog cycles, letting the roll lag the wheel by a few degrees per revolution, or (3) the motor is dropping speed under load because the contactor wiring crossed two phases of an undersized supply transformer.
Worm-and-pinion Feed Roll Drive vs Alternatives
Worm-and-pinion isn't the only way to drive a feed roll. Helical gear reducers, cycloidal drives, and direct servo drives all show up on the same kinds of lines. The right choice depends on what you value most — holding torque at rest, efficiency under continuous load, positional accuracy, or capital cost.
| Property | Worm-and-Pinion Feed Roll Drive | Helical Gear Reducer | Direct Servo Drive |
|---|---|---|---|
| Single-stage reduction ratio | 20:1 to 80:1 | 3:1 to 8:1 typical | 1:1 (no reduction stage) |
| Efficiency at nominal load | 55-85% (lead-angle dependent) | 96-98% | 92-95% (motor + drive losses) |
| Self-locking when power off | Yes, below ~6° lead angle | No — back-drives freely | Yes, with brake; no without |
| Positional accuracy at roll | ±0.3-0.5 mm typical (backlash limited) | ±0.1-0.2 mm | ±0.01-0.05 mm with encoder |
| Maintenance interval (oil change) | 4,000-8,000 hours | 8,000-20,000 hours | Bearing-only, 20,000+ hours |
| Lifespan to bronze wheel replacement | 15,000-30,000 hours | 40,000+ hours (steel-on-steel) | 30,000+ hours (motor bearings limit) |
| Capital cost (relative) | 1.0× (baseline) | 1.3-1.6× | 2.5-4.0× including drive and feedback |
| Best application fit | High-reduction, hold-at-stop web feed | High-efficiency continuous parallel-shaft drive | High-accuracy registered feed with frequent reversals |
Frequently Asked Questions About Worm-and-pinion Feed Roll Drive
Don't trust the catalogue claim — measure it. With the input motor de-energised and uncoupled, apply the maximum expected back-torque at the wheel through a torque wrench on the feed roll shaft. If the worm rotates, you're not self-locking under that load. The reason catalogue claims fail in service is dynamic self-locking is weaker than static — vibration from the line (especially at frequencies near the worm's first torsional mode, often 80-150 Hz on a typical box) reduces the effective friction coefficient by 30-50%.
Rule of thumb: if your lead angle is above 4.5°, plan on adding a brake. Self-locking only stays reliable below that with steel-on-bronze in oil.
Once-per-rev banding almost always traces to wheel runout, not the worm. Mount a dial indicator on the feed roll OD and rotate slowly by hand — if you see more than 0.02 mm TIR, that's your culprit. The bronze wheel may be cut to AGMA Q10 but if it sits on a shaft that's bent, or on a taper bushing that's seated unevenly, the contact patch walks across the tooth face once per revolution and modulates the instantaneous transmission ratio by 0.1-0.3%.
That's enough to show up as visible banding on a 4-colour print job. Re-seat the bushing with the manufacturer's specified torque sequence — usually a star pattern across 4-6 bolts at 30%, 70%, 100% of final torque.
Single-start, almost always. A single-start worm at typical module sizes lands at 3-5° lead angle, well inside the self-locking zone. A 2-start worm pushes you to 8-10°, and a 4-start worm to 16-20° — the latter back-drives freely under web tension.
The trade is efficiency. A single-start worm runs around 55-70% efficient; a 4-start runs 85%+. If holding torque at stop matters more than running efficiency — typical for unwind nip rolls and intermittent feeders — go single-start and accept the heat. If the line runs continuously and an external brake handles holding, go multi-start and harvest the efficiency.
Three things drive sump temperature above prediction: oil level too high (churning loss can double the heat input above the calculated mesh loss), oil grade wrong for the speed (VG 220 in a box specced for VG 460 will run 15-20°C hotter at the same load), or the housing has been painted with a thick decorative coat that cuts radiative heat transfer by 30-40%.
Check oil level cold first — it should sit at the worm centreline, not above. Then verify the oil grade against the nameplate. If both check out, look at airflow around the box; many install cabinets restrict natural convection and a small 24 V fan brings sump temp down 15°C in minutes.
Assume worst-case efficiency at the start of life and worse at end of life. For a single-start worm, design the motor for 55% efficiency at the mesh and add a 1.3× service factor on top for jog-and-jam events. For a 2-start worm, use 70% efficiency. The reason: catalogue efficiency assumes broken-in gears at nominal speed and temperature, and your real-world startup torque needs to clear cold-oil drag, which can spike to 2-3× the running torque on a box that's been sitting overnight at 5°C.
Practical check: calculate steady-state torque required at the roll, divide by ratio, divide by 0.55, multiply by 1.3. That's your motor minimum. If it lands close to a frame size, go up one frame — worm boxes punish under-sized motors with stalls during cold starts.
That's a centre-distance or worm-axial-position error. The contact pattern on a correctly aligned worm-and-pinion sits in the centre 60-70% of the tooth face. If wear is biased to one side, the worm sits axially off its theoretical position by more than the design tolerance — usually because the thrust bearing preload has loosened or one of the housing dowel pins has worked free.
Diagnose with engineer's blue: paint the worm threads, run by hand for 5-10 revs, and inspect the wheel teeth. The blue should land centrally. If it's biased to the entry side, shim the thrust bearing on the exit side; if biased to the exit side, do the opposite. Ignore this and the off-loaded edge of the tooth chips off within 500-1,000 hours.
References & Further Reading
- Wikipedia contributors. Worm drive. Wikipedia
Building or designing a mechanism like this?
Explore the precision-engineered motion control hardware used by mechanical engineers, makers, and product designers.
