A Watt topology six-bar linkage is a planar mechanism made of six rigid links and seven revolute joints arranged so that the two ternary links (the ones with three pivots each) are directly connected to each other. James Watt himself developed the parent four-bar Watt straight-line linkage in 1784, and the six-bar topology that carries his name extends that idea by stacking a second four-bar loop onto the first. The arrangement converts a single rotary input into a complex coupler-point path with built-in dwells or near-straight segments. Designers use it where a four-bar can't deliver the required motion law — automotive suspensions, packaging folders, and prosthetic knees all rely on it.
Watt Topology Six-bar Linkage Interactive Calculator
Vary the Watt six-bar link count, joint count, input speed, and dwell time to see mobility, cycle time, and required dwell angle.
Equation Used
The Kutzbach criterion gives planar linkage mobility from the number of links, lower-pair joints, and higher-pair joints. The speed calculation converts input RPM to cycle time, then converts the required dwell time into the equivalent crank angle.
- Planar linkage with rigid links.
- j1 counts lower-pair 1-DOF joints such as revolute pivots.
- j2 counts higher-pair joints such as cams or gear contacts.
- One input shaft revolution equals one linkage cycle.
Operating Principle of the Watt Topology Six-bar Linkage
A Watt six-bar is built from two four-bar loops sharing a common link, with the defining feature being that the two ternary links sit adjacent — pivot-to-pivot — rather than being separated by a binary link as they are in a Stephenson topology. You drive one binary input crank, and the motion propagates through the first loop into the second, where a chosen coupler point traces a path that simply isn't reachable with a single four-bar. The two named variants — Watt I and Watt II — differ only in which link is grounded, and that single choice changes the available coupler curves dramatically.
Why build it this way? A four-bar gives you one coupler curve per geometry. Stack a second loop on top and you gain four extra design parameters, which is enough to dial in dwells, near-straight segments, or symmetric figure-eights without resorting to cams. The Grashof condition still applies to each loop individually, so you have to check both — get one loop non-Grashof when you wanted full rotation and the input crank will lock at a dead point partway through its cycle.
Tolerances bite hard here because errors compound through the loops. If your pivot bushings have 0.1 mm of radial clearance per joint, by the time motion reaches the output coupler point you can see 0.4-0.6 mm of positional uncertainty. Common failure modes are bushing wear at the central ternary-to-ternary joint (it carries the highest cyclic load), fatigue cracks at the ternary link's three-pivot triangle if the link is thin, and binding at branch-point singularities if the link lengths drift through wear. Kinematic synthesis software like SAM or Linkage will flag these — but only if you actually run the full rotation, not just a snapshot.
Key Components
- Ground link (frame): The fixed reference link carrying two of the seven revolute pivots. In a Watt I the ground is a binary link; in a Watt II it's one of the ternary links. Pivot hole concentricity must hold within 0.05 mm or the whole motion law shifts.
- Input crank: The binary driven link, typically rotating a full 360°. Sized so that the Grashof inequality s + l ≤ p + q is satisfied for the first loop, otherwise you get a rocker rather than a crank and the mechanism can't run continuously.
- First coupler (ternary link A): A three-pivot link that connects the input loop to the second loop. Carries one of the design coupler points. The triangle's interior angles directly set the phase relationship between the two loops, typically held to ±0.2° on precision builds.
- Second ternary link (ternary link B): Pivots directly to ternary link A — this adjacency is what defines the Watt topology versus the Stephenson topology. The pivot connecting A to B sees the highest cyclic loads and is the first place to check for bushing wear.
- Output rocker or coupler: The final binary link that either oscillates as an output rocker or carries the output coupler point. Length sets the amplitude of the output motion; small length errors here scale linearly into output position error.
- Revolute pivots (7 total): Seven turning pairs — typically PTFE-lined bushings, needle bearings, or sealed deep-groove ball bearings depending on speed and load. Total degrees of freedom comes out to 1, which is what you want for a single-input mechanism.
Where the Watt Topology Six-bar Linkage Is Used
Watt six-bars show up wherever a four-bar's coupler curve is too crude but a cam-and-follower would be overkill or too noisy. The two practical advantages are a long usable dwell from a continuously rotating input, and the ability to package a complex motion in a flat plane with no sliding pairs — meaning no wear-prone prismatic joints. You'll find them across automotive, packaging, medical, and aerospace deployment hardware.
- Automotive suspension: Watts linkage rear axle location on the Ford Ranger PX and the Aston Martin DB7, replacing a Panhard rod to give symmetric lateral location and pure vertical wheel travel.
- Packaging machinery: Carton-flap folding heads on Bosch CUT 120 cartoners, where the dwell at the bottom of the stroke holds the flap closed against the glue line for ~90° of input rotation.
- Prosthetics: Polycentric knee joints such as the Otto Bock 3R60 use a Watt-derived six-bar to give a moving instantaneous centre of rotation through stance and swing phases.
- Aerospace deployment: Landing gear sidebrace mechanisms on regional aircraft like the Bombardier Q400 use Watt-style six-bars for over-centre lock with a single hydraulic actuator input.
- Industrial automation: Indexing pick-and-place arms on Bosch Rexroth assembly cells where a 60° dwell is needed at both endpoints of the stroke for grip and release.
- Agricultural equipment: Round-baler pickup tine mechanisms on John Deere 560M balers, where the tine must follow a near-cycloidal path to lift hay cleanly without scuffing.
The Formula Behind the Watt Topology Six-bar Linkage
The Kutzbach criterion tells you the mobility (degrees of freedom) of any planar linkage from link count and joint count alone. For a Watt six-bar, this matters because if you mis-count joints during a redesign — say you add a watch-link without realising it adds a constraint — your mechanism's DOF can drop to 0 and the whole thing locks solid. At the low end of complexity (a basic four-bar, n=4), Kutzbach gives M=1 and you have exactly the freedom you want. At nominal Watt six-bar complexity (n=6, j=7) it still gives M=1 — that's the design sweet spot, single input drives single coupler path. Push to an eight-bar (n=8, j=10) and you still get M=1 but with two extra design parameters; go beyond that and you start fighting compounding tolerance stack rather than gaining useful design freedom.
Variables
| Symbol | Meaning | Unit (SI) | Unit (Imperial) |
|---|---|---|---|
| M | Mobility — degrees of freedom of the linkage | dimensionless | dimensionless |
| n | Total number of links including the ground link | count | count |
| j1 | Number of lower-pair (1-DOF) joints — revolutes and prismatics | count | count |
| j2 | Number of higher-pair (2-DOF) joints — cams, gear contacts | count | count |
Worked Example: Watt Topology Six-bar Linkage in a glass-bottle labeller infeed star
A craft-brewery equipment builder in Asheville is designing a Watt II six-bar to drive the label-applicator pad on a 6,000-bottle-per-hour rotary labeller. The pad must dwell against the bottle for 80 ms while pressure-sensitive label adhesive bonds, then retract through a coupler curve that clears the next bottle by 12 mm. They want to verify the mobility before machining the ternary links from 6082-T6 plate and need to confirm the linkage delivers the dwell at the planned 100 RPM input shaft speed.
Given
- n = 6 links
- j1 = 7 revolute joints
- j2 = 0 higher pairs
- Ninput = 100 RPM nominal
Solution
Step 1 — apply the Kutzbach criterion at the nominal Watt six-bar configuration to confirm single-input behaviour:
One degree of freedom — exactly what we want. Drive the input crank and the entire mechanism follows a deterministic path.
Step 2 — convert input speed to cycle time at nominal 100 RPM to size the dwell window:
For an 80 ms dwell requirement, the dwell arc must occupy at least 0.080 / 0.600 = 13.3% of the input rotation, or roughly 48° of crank angle. A well-synthesised Watt II easily delivers 60-90° of dwell, so we have margin.
Step 3 — check the low end of the labeller's typical operating range, 60 RPM (machine warm-up and changeover speed):
At 60 RPM the 48° geometric dwell becomes a generous 133 ms in real time — the adhesive bonds well before the pad releases. The pad motion looks slow and deliberate to the eye.
Step 4 — check the high end of the typical range, 150 RPM (a 9,000 bph push speed):
Now the dwell drops to 53 ms — below the 80 ms adhesive bonding spec. In practice you'll see labels lifting at the trailing edge as bottles leave the applicator. 100-110 RPM is the upper sweet spot before adhesive performance degrades.
Result
Mobility comes out to M = 1, confirming the Watt II design is kinematically correct as a single-input mechanism. At nominal 100 RPM the 48° geometric dwell delivers 64 ms of pad-on-bottle contact, which combined with the approach and retract phases gives the full 80 ms adhesive window. The 60 RPM low end gives a comfortable 133 ms contact time with no labelling defects, while the 150 RPM high end falls to 53 ms and labels start lifting at the trailing edge — so plan production speed at 100-110 RPM for clean output. If you machine the parts and measure mobility behaviour different from M=1, the usual culprits are: (1) a dimensional error in one of the ternary links pushing the mechanism into a branch-point singularity where it locks momentarily each cycle, (2) a missed Grashof check on the second loop turning your intended crank-rocker into a double-rocker that won't complete a full rotation, or (3) a parallelism error in the two ternary links creating an over-constrained loop that binds when the frame flexes under load.
Choosing the Watt Topology Six-bar Linkage: Pros and Cons
Watt six-bars sit between simple four-bars and full cam-follower systems. The decision usually comes down to whether you need a dwell, how much precision the application demands, and how much packaging space you have. Here's how the Watt topology stacks up against the two alternatives a designer typically considers.
| Property | Watt six-bar linkage | Four-bar linkage | Cam and follower |
|---|---|---|---|
| Maximum useful input speed | 300-600 RPM with bushed joints, up to 1,500 RPM with rolling-element bearings | Up to 3,000 RPM, fewer joints to fail | 500-1,200 RPM, limited by follower jump |
| Achievable dwell duration | 60-120° of input rotation, geometrically created | Effectively zero — coupler is always moving | Arbitrary, set by cam profile (0-300°) |
| Positional accuracy at output | ±0.1-0.5 mm typical, dominated by joint clearance stack-up | ±0.05-0.2 mm, only 4 joints to stack | ±0.02-0.1 mm with ground cam profile |
| Cost to build (small batch) | Moderate — 6 links, 7 pivots, mostly turned and milled parts | Low — 4 links, 4 pivots | High — precision cam grinding required |
| Service life before rebuild | 10-50 million cycles depending on bushing material | 20-100 million cycles | 5-20 million cycles, follower wear is limiting |
| Design effort | High — kinematic synthesis software essentially required | Low — closed-form synthesis well documented | Moderate — cam profile design is well-tooled |
| Best application fit | Dwell + complex coupler paths in a flat package, no sliding pairs | Simple oscillation, straight-line approximation, force amplification | Precise motion laws, programmable timing, high-cycle indexing |
Frequently Asked Questions About Watt Topology Six-bar Linkage
You've hit a branch-point singularity — a configuration where the second loop reaches a dead-centre alignment and the constraint Jacobian loses rank. Kutzbach mobility is a global count and doesn't catch instantaneous singularities. The usual cause is the second loop's geometry sitting too close to a Grashof boundary; small manufacturing variation in link lengths pushes one configuration over the edge.
Quick check: rotate the input slowly by hand and watch which joint goes still at the lock instant. That joint is the one whose link length needs adjustment. Lengthening or shortening it by 1-2% usually clears the singularity without redesigning the whole mechanism.
Watt topologies (the two ternary links touch) tend to give symmetric coupler curves and are the natural choice when you want a dwell at both ends of a stroke or a balanced figure-eight path. Stephenson topologies (a binary link separates the two ternary links) give asymmetric curves and are better when you want one long dwell and one fast return — like a quick-return mechanism with hold time at the cutting end.
Rule of thumb: if you can sketch the desired output motion and it's mirror-symmetric about a vertical axis, start with Watt. If it's asymmetric in time or space, start with Stephenson. Both can be forced to do the other's job, but you'll fight the synthesis the whole way.
That centre pivot sees the highest cyclic load in the entire mechanism because forces from both four-bar loops superimpose there. Treat it as the failure-critical joint. For 10 million cycles plus, use a sealed deep-groove ball bearing (6000 series or larger) sized so the dynamic load rating C is at least 8× the peak joint reaction force from your kinematic analysis.
Bushed joints work below 1 million cycles or below 100 RPM, but you need an oil-impregnated bronze or PTFE-lined steel bushing — never a dry steel-on-steel fit, which will gall in days at packaging-line speeds.
1.5 mm of offset on a typical 100-200 mm coupler curve is almost always one of two things. First, check the perpendicularity of your pivot pin axes to the link plate. Pin tilt of just 0.5° on a 50 mm-long ternary link offsets the pivot location by ~0.4 mm, and that error compounds through the loops.
Second, verify that the link plates are flat. 6082-T6 plate often comes with 0.3-0.8 mm of bow per 200 mm length, and if you didn't surface-grind both faces before drilling pivot holes, the assembled linkage will sit out of plane and the coupler point traces a trajectory tilted relative to the design plane.
Sometimes yes, sometimes no — and this is the key design decision. A four-bar coupler curve is a sextic algebraic curve with limited shape variability. If your target motion has a true dwell (zero output velocity for a finite arc) or two distinct stationary points, no four-bar coupler curve will match it. You need the extra two coupler-curve coefficients that the second loop in a Watt six-bar provides.
Practical test: try the four-bar synthesis first in software like Linkage or SAM. If the synthesised curve fits within ±5% of your target precision points but fails the velocity matching, you need the six-bar. If the position error itself exceeds 5%, definitely six-bar.
Inertia. At 60 RPM the centripetal acceleration on the coupler points is modest. At 200 RPM it's roughly 11× higher (acceleration scales with the square of speed), and unbalanced inertia forces from the two ternary links start exciting frame resonance. The ternary links are the worst offenders because their mass is offset from any rotation axis.
Two fixes: counterweight the ternary links so their centres of mass sit closer to the pivot axes, or stiffen the frame. A 2:1 increase in frame natural frequency typically buys you 2× usable speed before vibration becomes a problem. If you can't do either, accept the speed limit — running through the resonance peak will fatigue weld joints in months.
Build a paper or laser-cut acrylic prototype at 1:1 scale with brass pivot pins. Acrylic is forgiving — you can drill, redrill, and adjust link lengths in an afternoon. Run the input by hand through a full rotation and trace the coupler point with a fine pen. If the traced curve matches your synthesis software output to within 1 mm, the geometry is correct and you can move to metal.
This catches branch-point singularities, missed Grashof violations, and clearance interferences for the cost of a sheet of acrylic. We've seen builders skip this step and waste $3,000 of machined aluminium on a mechanism that locks at 270° of input.
References & Further Reading
- Wikipedia contributors. Six-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.
