A double-rocker mechanism is a four-bar linkage in which both the input and output links oscillate through limited arcs rather than rotating fully. Franz Grashof codified the condition that governs this behaviour in 1883, showing that when the sum of the shortest and longest links exceeds the sum of the other two, no link can complete a full revolution. The coupler link transfers motion between the two rockers, converting one swing into another swing of different angle and timing. You see it in windscreen wiper pairs, aircraft landing-gear retraction linkages, and dockside crane luffing gear.
Double-rocker Mechanism Interactive Calculator
Vary the input and output rocker swing angles and see the swing gain, difference, and animated four-bar motion.
Equation Used
The worked example diagram compares an input rocker swing of about 35 deg with an output rocker swing of about 40 deg. This calculator reports the swing ratio and percentage gain between the two rocker arcs.
- Swing angles are total peak-to-peak rocker arcs.
- The linkage is treated as a teaching model for comparing rocker motion.
- Default angles match the worked example diagram values.
How the Double-rocker Mechanism Works
A double-rocker is a four-bar linkage with four pin joints, four rigid links, and one important property — neither of the two links pinned to ground can spin all the way around. Both swing back and forth. The link opposite the ground (the coupler) connects them and carries the motion between the two rockers. If you drive one rocker with a motor or a hydraulic cylinder, the other rocker responds with its own arc, usually a different swing angle and a different angular velocity profile.
The geometry has to satisfy the non-Grashof condition: s + l > p + q, where s is the shortest link, l is the longest, and p and q are the other two. Get this wrong by even a millimetre on a small linkage and you'll find one rocker locks up at a dead point or — worse — the linkage flips into a different mode mid-cycle. The transmission angle (the angle between the coupler and the output rocker) should stay between roughly 40° and 140° across the working stroke. Drop below 40° and the force the input has to apply climbs sharply, the pins start hammering, and bushings wear oval within a few hundred cycles.
Common failure modes are predictable. Pivot pins worn beyond about 0.1 mm radial play will show up as backlash at the output, often misread as "the motor is loose." If the coupler is too flexible, you'll see a phase lag between input and output that grows with load — a stiffer coupler usually fixes it. And if you ever feel the linkage "go over centre," that means you've crossed a singularity, and the mechanism either jams or snaps to its alternate assembly configuration. Both are bad.
Key Components
- Ground link (frame): The fixed reference link between the two ground pivots. Its length sets the spacing between rockers and largely determines the swing-angle ratio. On a typical mid-size linkage we hold this dimension to ±0.05 mm because a 0.1 mm error here shifts the transmission angle by about 1° at the extremes of stroke.
- Input rocker: The driven link, oscillating through a defined arc — often 30° to 90°. It receives torque from a motor, cylinder, or manual lever. The pivot bushing must carry both radial load from the coupler and any side load from the actuator without exceeding 0.1 mm play.
- Coupler link: The floating link between the two rockers. It must be torsionally stiff because every micron of bending shows up as phase lag at the output. We size couplers in steel or aluminium with a section modulus that keeps deflection under 0.05 mm at peak load.
- Output rocker: The driven oscillating link that performs the work — wiping, clamping, luffing, whatever the application demands. Its swing angle is set by the link-length ratios and rarely matches the input swing exactly. The ratio is the design lever you tune.
- Pivot pins and bushings: Four hardened pins at the four joints. Use bronze or polymer bushings with a press-fit tolerance of H7/p6. Radial clearance must stay below 0.05 mm new and replaced before it reaches 0.15 mm — beyond that the linkage feels sloppy and the transmission angle drifts.
Where the Double-rocker Mechanism Is Used
Double-rocker linkages show up wherever you need to convert one constrained oscillation into another, with a defined motion profile and no full rotation. Builders pick them over cranks when full revolution would be wasted or dangerous, over cams when the load is too high or the duty cycle too long for cam followers, and over linear actuators when the motion needs to be a true arc rather than a straight line. The mechanism is cheap to build, reliable for millions of cycles when the transmission angle is respected, and tolerates dirt and weather better than almost any alternative.
- Automotive: Tandem windscreen wiper linkages on the Mercedes-Benz W124 and similar pantograph-style wiper systems, where one motor drives both wiper arms through a double-rocker.
- Aerospace: Landing gear sidestay and retraction linkages on aircraft like the Cessna 172 mainwheel system, where the gear leg swings through a fixed arc into the wheel well.
- Marine: Dockside crane luffing gear at the Port of Rotterdam container terminals, where the jib arm rocks through a controlled arc to position loads.
- Agricultural machinery: Hay-baler pickup tine linkages on John Deere round balers, oscillating the tine bar to lift cut crop into the chamber.
- Industrial automation: Pick-and-place transfer arms on glass-bottle filling lines such as the Krones Modulfill, swinging bottles between filling and capping stations.
- Heritage engineering: Beam-engine valve gear on the preserved Crofton Pumping Station Cornish engine in Wiltshire, where eccentric-driven rockers operate steam admission valves.
The Formula Behind the Double-rocker Mechanism
The most useful design formula for a double-rocker is the swing-angle ratio between input and output rockers, derived from the link lengths. At the low end of the typical operating range — short input swings of 20° or so — the ratio behaves almost linearly and the output tracks the input cleanly. Push the input swing to 60° or 80° and the ratio becomes strongly nonlinear, with the output accelerating and decelerating across the stroke. The sweet spot for most industrial designs sits around 30°-50° input swing, where you keep transmission angle above 45° throughout and avoid the ugly velocity spikes near the dead points.
Variables
| Symbol | Meaning | Unit (SI) | Unit (Imperial) |
|---|---|---|---|
| θout | Output rocker angle measured from ground link | degrees or radians | degrees |
| θin | Input rocker angle measured from ground link | degrees or radians | degrees |
| Lg | Ground link length (fixed frame) | mm | in |
| Lin | Input rocker length | mm | in |
| Lc | Coupler link length | mm | in |
| θoffset | Geometric offset constant from initial assembly | degrees | degrees |
Worked Example: Double-rocker Mechanism in a textile-loom shedding rocker linkage
A weaving-machinery rebuilder in Bursa is replacing the worn shedding linkage on a Picanol GTM-AS air-jet loom. The double-rocker drives the heald frame through a 40° nominal output arc from a 35° input swing on the cam-driven input rocker. Link lengths: ground Lg = 180 mm, input rocker Lin = 95 mm, coupler Lc = 165 mm, output rocker Lout = 110 mm. Check the output arc at low, nominal, and high input swings to confirm the transmission angle stays acceptable across the working range.
Given
- Lg = 180 mm
- Lin = 95 mm
- Lc = 165 mm
- Lout = 110 mm
- θin nominal = 35 degrees
Solution
Step 1 — confirm the non-Grashof condition. Shortest s = 95 mm (input), longest l = 180 mm (ground), other two p = 165, q = 110.
This sits right on the Grashof boundary — a change-point linkage. We need to nudge one length to push it firmly into double-rocker territory, so we'll treat Lin as 94 mm in practice (s + l = 274 < p + q = 275). Step 2 — compute the nominal output angle at θin = 35°.
Step 3 — at the low end of the working range, θin = 20°, the output tracks almost linearly:
The heald frame moves a small, controlled distance — fine for fabric inspection or slow indexing, but the shed opening is too small for the air-jet to clear the warp at production speed. Step 4 — at the high end, θin = 55°:
That sub-40° transmission angle is the warning sign — the input rocker has to push hard to drive the coupler, pin loads spike by roughly 1.6×, and bushings will wear oval within months at loom production cycles of around 600 picks per minute.
Result
Nominal output swing comes out to 40. 2°, matching the design intent for the heald-frame shed opening. At 20° input the output is only 23.5° — too small a shed for reliable weft insertion, the air-jet won't clear the warp cleanly. At 55° input the 62° output is geometrically achievable but the transmission angle drops to 38°, which means heavy pin loading and accelerated bushing wear at production speed. If you measure the actual output and find it 2°-3° below 40.2°, suspect one of three things: (1) coupler-pin radial clearance worn beyond 0.15 mm, adding lost motion, (2) the Lg dimension drifted because the ground-pivot brackets loosened on their mounting bolts — re-torque to spec and re-measure, or (3) the input rocker bushing is hammering oval, which you'll feel as a sharp click at each stroke reversal.
When to Use a Double-rocker Mechanism and When Not To
A double-rocker is one of three obvious choices when you need oscillating output. The other two — a crank-rocker driven by a continuous motor, and a cam-and-follower — solve overlapping problems but with very different cost, complexity, and lifespan profiles. Pick on the application, not on theoretical elegance.
| Property | Double-rocker | Crank-rocker (Grashof) | Cam and follower |
|---|---|---|---|
| Input motion required | Oscillating (lever, cylinder, or driven rocker) | Full continuous rotation | Continuous rotation |
| Typical cycle speed | Up to ~600 cycles/min | Up to ~1500 RPM | Up to ~3000 RPM with light followers |
| Output motion accuracy | Geometry-defined, ±0.5° at fresh pin clearance | Geometry-defined, ±0.5° | Profile-defined, ±0.05° if cam ground precisely |
| Manufacturing cost | Low — four pinned links | Low to medium — adds a crankshaft and bearing | High — cam grinding and follower hardening |
| Maintenance interval (industrial duty) | Pin/bushing inspection every 2-5 million cycles | Crank bearing service every 5-10 million cycles | Cam-follower wear check every 10-50 million cycles |
| Tolerance to dirt and weather | Excellent — open pin joints accept grease | Good | Poor — cam profile damaged by debris |
| Best application fit | Wipers, landing gear, luffing, valve gear | Pumps, compressors, treadle drives | Engine valvetrains, indexing tables |
Frequently Asked Questions About Double-rocker Mechanism
You're hitting a singularity, not a Grashof violation. The non-Grashof condition guarantees no link rotates fully, but it doesn't prevent the input rocker from reaching a position where the coupler and output rocker line up — that's a dead point where the mechanism can't transmit force regardless of input torque.
Check the transmission angle across your full input range. If it touches 0° or 180° anywhere in the planned stroke, redesign the link ratios. A common fix is shortening the coupler by 5%-10% to push the dead point outside the working arc.
This is a real design lever. With the longest link as the ground (a Grashof double-rocker variant), you get a more symmetric output swing and the transmission angle stays in a friendlier range. With the longest link as the coupler, you get larger output amplification — more output swing per unit input swing — but you pay for it in pin loads near the extremes.
Rule of thumb: if your application is force-limited (clamping, valve actuation), put the long link on the ground. If it's motion-limited and force is plentiful (wiper sweep, light positioning), make the coupler the long link.
If new pins don't fix it, the link lengths themselves are wrong. The two most common offenders are: pivot-hole centres drilled from a CAD print without correcting for hole-to-hole tolerance stack, and a coupler that flexes under load — particularly thin-section sheet-metal couplers. A 0.5 mm error in any link length, multiplied across the geometry, easily shows up as a 3°-5° output discrepancy.
Measure each link pin-centre to pin-centre with calipers, not from CAD. Then load-test the coupler — clamp one end and apply expected operating load to the other. Any visible deflection means you need a stiffer section.
Yes, and it's often the cleanest choice. A linear actuator pinned between the ground frame and a point on the input rocker converts its linear stroke directly into the rocker's swing — no crank or rotary-to-oscillating conversion required. You get inherent end-stop limiting from the actuator's stroke length, which is a free safety feature.
The trick is positioning the actuator pin so the moment arm is acceptable across the full swing. Put the pin too close to the input pivot and you'll need huge actuator force; too far and the actuator runs out of stroke before the rocker completes its arc. Aim for a moment arm of 30%-50% of the rocker length at mid-stroke.
Every four-bar linkage has two valid assembly configurations — same link lengths, different mirror state of the coupler relative to the rockers. Crossing a singularity under load can snap the mechanism from one mode to the other, and the symptom is unmistakable: the output rocker suddenly reverses direction or jumps to a position 30°-60° away from where it should be.
Rebuild from scratch with the coupler oriented correctly and add a hard mechanical stop just outside the design swing range. The stop prevents the input rocker from reaching the singularity, which prevents mode-flipping entirely.
Below about 300 cycles/min on a typical mid-size industrial linkage, you can analyse it as a pure kinematic problem and the static transmission-angle calculations will predict reality within a couple of percent. Above 300 cycles/min, coupler inertia starts adding a sinusoidal load on top of the working load, and around 600 cycles/min that inertial component can equal or exceed the working load.
If your application sits above 600 cycles/min, you need to either lighten the coupler dramatically (aluminium or carbon-fibre tube), or accept that bushing life will drop by half or more compared to a low-speed build. The Picanol air-jet loom example in the worked solution sits right in this danger zone, which is why those linkages get rebuilt on a strict schedule.
References & Further Reading
- Wikipedia contributors. Four-bar linkage. Wikipedia
Building or designing a mechanism like this?
Explore the precision-engineered motion control hardware used by mechanical engineers, makers, and product designers.
