Alternating motion is a gearing arrangement that converts continuous one-direction rotation of an input shaft into back-and-forth (reciprocating) motion of an output member. The trick is a pinion with teeth on only part of its circumference, or a pair of pinions engaging opposite sides of a rack — so engagement alternates and the output reverses without ever reversing the motor. We use it where you need clean, repeatable reciprocation from a cheap single-direction drive, like shaper rams, weaving shuttles, and reciprocating saws, where reversing the prime mover would waste time and stress the drivetrain.
Alternating Motion Interactive Calculator
Vary pinion pitch diameter, toothed sector angle, and shaft speed to see rack stroke and reciprocation rate.
Equation Used
The one-way rack stroke is the pitch-circle arc length covered by the toothed sector of the mutilated pinion. Cycle rate comes directly from input shaft speed when one shaft revolution produces one complete forward-and-return rack cycle.
- Rack travel equals the pitch-arc length of the toothed sector.
- One shaft revolution produces one full extend-and-retract reciprocation cycle.
- No slip, backlash, tooth deflection, or changeover dwell is included.
- Input rpm is matched to bottle rate for the worked example.
Inside the Alternating Motion
The core idea is dead simple. You have a motor that only ever spins one way. You want the output to go forward, then back, then forward again, on a fixed cycle. So you cut the teeth off part of a pinion — what we call a mutilated pinion or half gear — and you arrange it so that as the input rotates, the toothed sector engages a rack (or a second gear) for half the revolution, then the bare side coasts past while a second engagement on the opposite side of the rack pulls the output the other way. The output reciprocates. The motor never reverses.
There are a few common layouts. The classic one uses a single rack with two pinions stacked above and below it — one drives the rack right, the other drives it left, and a transfer arm or sliding pinion shifts engagement at each end of stroke. Another layout uses a single mutilated pinion meshing with a curved rack that wraps around it, so the toothed sector pushes the rack one way going in and the other way coming out. In rotary-to-linear conversions you'll also see this combined with a rack reversal mechanism where a spring-loaded shoe flips the engagement at end of stroke.
Get the tooth-count math wrong and the mechanism jams or skips. The toothed sector angle, the gap angle, and the rack length must all close out at the same instant — if the pinion still has teeth engaged when the rack hits the end stop, you'll either shear teeth or stall the motor. Typical failure modes are tooth shear at changeover, rack-end impact damage from over-travel, and timing drift caused by a worn idler bushing letting the half-gear lag a few degrees. Build it with the changeover happening 2-3° before mechanical end-of-stroke and you give the system breathing room.
Key Components
- Mutilated pinion (half gear): A pinion with teeth cut on only a portion of its circumference, typically 180° toothed and 180° smooth, though sectors of 120° or 90° are common in shorter strokes. The tooth-to-gap transition must be chamfered — we run a 0.3 mm × 30° chamfer minimum on the leading and trailing teeth so they pick up the rack cleanly without clipping.
- Rack: The linear toothed bar that carries the reciprocating output. Length must equal stroke plus 2 to 3 extra teeth on each end as engagement runout. Module must match the pinion exactly — a module mismatch of even 0.05 will cause binding within the first hundred cycles.
- Transfer or reversing element: Depending on layout this is either a second pinion stacked opposite the first, a curved return rack, or a sliding shoe that flips the rack onto the opposite pinion at end of stroke. This part takes the highest impact load in the system and is usually hardened to 55-60 HRC.
- Guide rails or linear bearing: Constrains the rack to pure linear travel and prevents it from lifting off the pinion. Total runout across the stroke should stay under 0.1 mm — any more and the rack rises off the pinion at mid-stroke and skips a tooth.
- Input shaft and bearing pair: Carries the half-gear and resists the side-load reaction from the rack. Because the load reverses every half-revolution, the bearings see fully reversed loading — pick a bearing rated for dynamic radial load at least 3× the calculated peak rack force.
Where the Alternating Motion Is Used
Alternating motion shows up wherever a cheap one-way motor needs to drive a back-and-forth output without electrical reversing, gearbox shuttling, or hydraulic complexity. The mechanism is silent, repeatable, and scales from desk-toy size up to industrial planers. You'll see it any time the duty cycle is high enough that reversing the prime mover would burn out contactors or chew through a clutch.
- Metalworking: Mechanical shaper machines like the Atlas 7B and South Bend 7-inch shaper, where a bull gear and yoke convert constant rotation into the slow forward cutting stroke and faster return of the ram
- Textiles: Shuttle drives in older Northrop and Sulzer flat looms, where alternating motion throws the shuttle across the warp and back without reversing the drive
- Consumer power tools: Reciprocating saws and jigsaws — the Bosch GSA 18V-LI and Milwaukee Sawzall lineage use a wobble plate or scotch yoke variant of alternating motion to convert motor rotation into 28 mm blade stroke
- Packaging machinery: Pusher mechanisms on Bosch and Bartelt cartoners that transfer products into the carton at fixed cycle rate, driven by a half-gear and rack to keep timing locked to the main camshaft
- Toys and education: Mechanical drawing toys and Tamiya gearbox kits, where a mutilated pinion drives a rack to make a figure walk back and forth on a desk
- Agricultural equipment: Hay rake oscillating drives and small-square baler needle drives, where a single-direction PTO has to swing a mechanism through a fixed reciprocating arc
The Formula Behind the Alternating Motion
What the practitioner actually needs to know is the linear stroke produced per input revolution, and how that scales with input RPM to give you cycle time. At the low end of the typical operating range, you get long, slow, controlled reciprocation — useful for cutting strokes where chip load matters. At the high end, you get fast cycling but the changeover impact rises with the square of speed and starts hammering the teeth. The sweet spot for most builds sits at 60-80% of the maximum RPM the half-gear tooth strength can survive on impact.
Variables
| Symbol | Meaning | Unit (SI) | Unit (Imperial) |
|---|---|---|---|
| S | Linear stroke per half-cycle (one direction of travel) | m | in |
| Dp | Pitch diameter of the mutilated pinion | m | in |
| θt | Toothed sector angle of the pinion | degrees | degrees |
| fcyc | Full reciprocation cycles per second | Hz | cycles/s |
| Nin | Input shaft rotational speed | rev/min | RPM |
Worked Example: Alternating Motion in a glass-bottle labelling station pusher
You are designing the side-pusher on a glass-bottle labelling line that nudges each bottle 50 mm sideways onto the labelling drum, then retracts. The line runs at 90 bottles per minute nominal, with operators occasionally throttling down to 45 bpm or pushing to 120 bpm during catch-up. You want a single-direction 3-phase gearmotor driving a mutilated pinion against a rack — no electrical reversing. Pitch diameter of the half-gear is 32 mm with a 180° toothed sector.
Given
- Dp = 32 mm
- θt = 180 degrees
- Required stroke = 50 mm
- Nominal cycle rate = 90 cycles/min
Solution
Step 1 — calculate the stroke produced by one engagement of the toothed sector at nominal pitch diameter:
That lands within 0.6% of the 50 mm target, so the geometry is right. Now translate input RPM into bottle throughput. One full input revolution gives one out-stroke and one return-stroke, so input RPM equals cycles per minute directly.
Step 2 — at nominal 90 bpm, input shaft runs at 90 RPM:
That gives a 0.67-second full cycle, with the pusher spending 0.33 s out and 0.33 s back. Plenty of dwell margin for the bottle to clear.
Step 3 — at the low end of the operating range, 45 bpm:
The pusher creeps over in 0.67 s — visibly slow on the line, useful during product changeover or operator-assisted runs. Tooth contact stress is trivial here.
Step 4 — at the high end, 120 bpm:
Full cycle drops to 0.5 s. The changeover impact at each end of stroke rises with the square of velocity, so loads on the leading tooth are roughly (120/90)2 = 1.78× the nominal value. That's still inside the safety margin for a properly hardened steel half-gear, but you'll start to hear the impact at the rack ends.
Result
The mechanism delivers a nominal 50. 3 mm stroke at 1.5 Hz when the line runs at 90 bpm. At that speed the pusher feels controlled — bottle transfer is positive, no rattling at end of stroke, and the gearmotor draws steady current. Across the operating range the mechanism behaves predictably: at 45 bpm the motion is almost lazy and tooth loads are negligible, while at 120 bpm changeover impact climbs to nearly twice nominal and you'll hear the rack ends thumping. If you measure stroke shorter than 50 mm in the field, suspect three causes in this order: (1) the half-gear mounting grub screw has slipped on the input shaft and the toothed sector is now starting late, (2) the rack guide rails have worn enough that the rack lifts off mid-stroke and skips a tooth at changeover, or (3) the input gearmotor coupling has developed backlash exceeding 2° which delays engagement at every reversal.
When to Use a Alternating Motion and When Not To
Alternating motion isn't the only way to get reciprocation from rotation. Scotch yokes and crank-slider mechanisms do similar work with different trade-offs in stroke smoothness, cost, and impact loading. Pick based on whether you need constant velocity through the stroke, smooth acceleration, or just brutal simplicity.
| Property | Alternating motion (mutilated pinion + rack) | Scotch yoke | Crank-slider |
|---|---|---|---|
| Stroke velocity profile | Near-constant velocity through stroke, sharp reversal at ends | Pure sinusoidal — zero velocity at ends, peak at midstroke | Near-sinusoidal with slight asymmetry from rod-length ratio |
| Typical cycle rate | 0.5 to 5 Hz comfortable, 8 Hz max with hardened gears | Up to 15 Hz with balanced yoke | Up to 50 Hz in IC engines, 5-10 Hz in industrial drives |
| End-of-stroke impact load | High — tooth changeover takes the hit | Zero — sinusoidal deceleration | Zero — sinusoidal deceleration |
| Stroke length adjustability | Fixed by pitch diameter and sector angle — change parts to change stroke | Fixed by crank radius — adjustable only by swapping crank | Fixed by crank radius |
| Manufacturing complexity | Moderate — gear cutting plus a rack, but tooth chamfering is fussy | Low — slot, pin, and crank | Low — proven design from every IC engine ever built |
| Typical service life | 5 to 20 million cycles before tooth wear shortens stroke | 20 to 100 million cycles with hardened slot insert | 100+ million cycles with proper bearing selection |
| Best application fit | Fixed-stroke industrial reciprocators where motor reversal isn't an option | Smooth high-speed reciprocation, test rigs, pumps | Engines, compressors, and any high-cycle reciprocator |
Frequently Asked Questions About Alternating Motion
This is almost always an asymmetric end-stop or a misaligned changeover. The tooth that shears is the one that is still partially engaged when the rack hits its mechanical limit — so all the inertia of the rack and load dumps into that single tooth.
Check the pinion clocking relative to the rack end stops with a dial indicator. The toothed sector should disengage 2-3° of input rotation before the rack reaches mechanical end-of-stroke. If one end shows zero margin and the other has 5°, you've found it. Loosen the half-gear, re-clock it to centre the margins, and torque to spec.
Yes — that's actually one of the strengths of this mechanism over a scotch yoke or crank-slider. You design the toothed sector to occupy less than 180° of pinion rotation. If the toothed sector is 150°, the rack moves during 150° of input rotation and dwells during the remaining 30° on each side. At 90 RPM input, a 30° dwell gives you about 55 ms of stationary time per side.
For longer dwells you go further — a 90° toothed sector with a 270° dwell gap gives you three-quarters of every cycle as dwell time. Just be aware that shorter sectors mean smaller stroke for the same pitch diameter, so you'll need a larger pinion to keep the stroke up.
Three millimetres is suspiciously close to the kind of loss you get from cumulative engagement slop. Check the chamfer on the leading and trailing teeth of the half-gear first. If somebody ground them aggressively to stop tooth-clipping, you can easily lose 1-2 teeth of effective engagement, which on a typical module-1 pinion is 3 mm of stroke.
Second suspect is the rack itself — if it's mounted on a slightly compliant carriage with the pinion pushing it, end-of-stroke compliance can swallow another millimetre. Put a dial indicator on the rack and slow-roll the input by hand. If the indicator reads 50.3 mm at quasi-static speed but you measure 47 mm running, the loss is dynamic compliance, not geometry.
For a pump at 3 Hz the scotch yoke wins. Pump efficiency hates the constant-velocity-then-slam profile of alternating motion — every reversal sends a pressure spike through the fluid column, and check valves chatter. The sinusoidal velocity of a scotch yoke gives you smooth acceleration at the ends of stroke and a clean pressure waveform.
Alternating motion is the right pick when you want most of the stroke at near-constant speed (like a cutting tool feed, a labelling pusher, or a paint spray reciprocator) and you don't care about the reversal impact because the load is rigid and well-supported. For fluid work, go scotch yoke or crank.
Don't average the two — size for the worst case plus the changeover impact. Take the higher of the two stroke loads, add the rack-and-pinion friction, and multiply by 1.5 to cover the inertia spike at the moment teeth re-engage on reversal.
The trap people fall into is sizing for steady-state running torque and getting blindsided by stall events at changeover. We've seen 80 W gearmotors stall on a 30 N load because the engineer didn't account for the 3× inertial peak when the half-gear tooth first contacts a stationary rack at 90 RPM. Always check the gearmotor's intermittent-duty peak torque rating, not just continuous.
Timing drift in alternating-motion mechanisms almost always traces back to one of two places. First, the input-shaft-to-half-gear interface — if it's a setscrew on a smooth shaft, the setscrew slowly burnishes a flat into the shaft and then walks. After a few thousand cycles the half-gear has rotated 1-3° relative to the shaft, which throws end-of-stroke timing.
The fix is to replace the setscrew mount with a keyed shaft or a pinned hub. Second possibility is the idler or transfer-pinion bushing wearing oval, which lets the engagement geometry shift over time. A bronze bushing in this duty typically lasts 2-5 million cycles before the oval wear becomes measurable; switch to a needle bearing if you need 10× that life.
References & Further Reading
- Wikipedia contributors. Reciprocating motion. Wikipedia
Building or designing a mechanism like this?
Explore the precision-engineered motion control hardware used by mechanical engineers, makers, and product designers.
