A Scott Russell linkage is a three-bar mechanism that converts rotary input into exact straight-line output by tying the midpoint of a coupler bar to a fixed pivot while constraining one end to a sliding path. Unlike approximate straight-line linkages such as the Watt or Chebyshev, it produces mathematically exact linear travel without any rail or guide on the output. We use it where a rolling guide would foul the workspace — laboratory stages, antenna deployers, and mirror translators — to deliver clean linear motion from a single rotating shaft.
Scott Russell Linkage Interactive Calculator
Vary the coupler length and slider clearance to see the required crank length, exact straight-line stroke, and tolerance effect.
Equation Used
The classical Scott Russell linkage uses a coupler of length C = 2L, where L is the crank from the ground pivot to the coupler midpoint. The free output end travels on an exact straight line with total stroke S = 4L, which is also S = 2C. Slider radial clearance is treated as a direct approximate output wobble.
- Classical Scott Russell layout with slider rail through the ground pivot.
- Coupler is exactly twice the crank length.
- Slider clearance maps approximately one-to-one to output wobble.
How the Scott Russell Linkage Actually Works
The geometry is dead simple but the result is surprising. Take a coupler bar of length 2L. Pin the middle of that bar to ground through a swinging crank of length L. Constrain one end of the coupler to slide along a straight line passing through the ground pivot. The opposite end then traces an exact straight line perpendicular to the first. That's it — three links, two pivots, one slider, and you get rotary-to-linear conversion with zero rail on the output side.
Why does it work? Because the midpoint of the chord of any circle, when one chord-end rides on a diameter, traces a path perpendicular to that diameter. The crank forces the midpoint onto a circle. The slider forces one end onto a line. The geometry leaves the far end no choice — it must travel perpendicular and dead straight. If you build it with the bore of the slider 0.2 mm oversize on a 10 mm shaft, you'll see the output end wobble by roughly that same 0.2 mm at full extension because the slider clearance maps directly onto output deviation. The tolerance on the crank-to-coupler midpoint pivot is even more critical — drift that pin position by 0.5 mm off the true centre and the output path bends into a shallow arc, not a line.
Common failure modes are predictable. Bushing wear at the midpoint pivot causes the output trace to develop a slight bow — you'll see it as a 1-2 mm cross-axis error at the end of stroke. A bent coupler bar shows up as asymmetric travel between the two stroke directions. And if the slider binds because the rail isn't truly collinear with the ground pivot, you'll feel torque spikes at the crank shaft twice per revolution. The kinematic inversion variant — where the slider is moved off the ground pivot and the crank pivot relocates — is sometimes used to package the linkage into tighter envelopes, but the exact straight-line property only holds for the classical layout.
Key Components
- Coupler bar (length 2L): The single rigid link that defines the entire mechanism. Length tolerance must be held to ±0.1% on precision builds — a 200 mm coupler off by 0.5 mm shifts the output trace by half that across the stroke. Material is usually ground steel or anodised aluminium plate.
- Crank link (length L): Connects the midpoint of the coupler bar to the fixed ground pivot. Its length must be exactly half the coupler length — not 49%, not 51%. Any deviation breaks the exact straight-line property and the output path becomes a shallow ellipse.
- Sliding pivot: Constrains one end of the coupler bar to travel along a line passing through the ground pivot. Typically a linear bearing or precision bushing on a hardened shaft. Radial play above 0.05 mm shows up directly as cross-axis output error.
- Ground pivot: The fixed reference around which the crank rotates. Must be co-linear with the slider rail to within 0.1 mm over the full stroke. Misalignment here causes the output to spiral instead of translating.
- Output point: The free end of the coupler bar — this is where you bolt the payload. It traces an exact straight line perpendicular to the slider rail. Stroke length equals 4L when the crank rotates from -90° to +90°.
Industries That Rely on the Scott Russell Linkage
The Scott Russell shows up wherever you need clean linear motion from a rotating input but can't tolerate a guide rail on the output side. That constraint sounds niche, but it covers a surprising amount of optical, aerospace, and laboratory hardware. The mechanism is also popular in deployable structures where the input shaft can sit deep in the body of a vehicle and the output extends into clean space. Stroke is limited — practical builds run 50-400 mm of travel — and the force capacity is low at the stroke extremes because the mechanical advantage drops off as the crank approaches ±90°. That's why you don't see it on press tools or heavy lifts.
- Optical metrology: Newport and Aerotech-style precision translation stages use Scott Russell geometry inside compact mirror translators where a rail-mounted guide would block the optical path.
- Aerospace deployables: Solar array hinge mechanisms on small satellites use the linkage to extend a panel cleanly perpendicular to the bus without needing an external guide rail.
- Laboratory automation: Sample-handling arms in Hamilton Microlab pipetting platforms use Scott Russell variants for vertical aspirate/dispense strokes where rail debris would contaminate the deck.
- Industrial weighing: Mettler-Toledo-style force restoration cells use the geometry to constrain a weighing pan to vertical motion without the friction of a linear bearing.
- Vehicle suspension testing: MTS road simulator rigs use the linkage in vertical actuator linkages to apply pure vertical loads to a wheel hub without parasitic side forces.
- Architectural hardware: Concealed cabinet door mechanisms — particularly the Blum Aventos lift system — use Scott Russell-type geometry to lift doors clear of the carcase along a near-linear path.
The Formula Behind the Scott Russell Linkage
The output displacement of a Scott Russell linkage is a clean trigonometric function of the crank angle. What matters in practice is how the output velocity changes across the stroke — at mid-stroke (crank at 0°) the output moves at maximum speed for a given input RPM, and at the stroke extremes (crank at ±90°) the output velocity drops to zero. That velocity profile means the linkage is well-suited to applications where you want a soft start and stop, but it's a poor choice if you need constant velocity through the full stroke. The sweet spot for most builds is operating between -60° and +60° of crank angle, where the velocity ratio stays within 50% of peak.
Variables
| Symbol | Meaning | Unit (SI) | Unit (Imperial) |
|---|---|---|---|
| y | Output displacement of the free end of the coupler from the ground pivot reference, measured along the straight-line path | m | in |
| L | Length of the crank link (equal to half the coupler bar length) | m | in |
| θ | Crank angle measured from the perpendicular to the slider rail | rad or ° | rad or ° |
| vout | Instantaneous output velocity at angle θ | m/s | in/s |
| ω | Angular velocity of the crank | rad/s | rad/s |
Worked Example: Scott Russell Linkage in a vertical mirror translator for a Fizeau interferometer
A precision optics workshop in Jena is building a vertical mirror translator for a Fizeau interferometer using a Scott Russell linkage. The crank length is L = 75 mm (so coupler length is 150 mm), driven by a stepper-geared shaft at a nominal 20 RPM. The team needs to know the output stroke, the velocity at mid-stroke, and how the velocity changes between the practical operating limits of θ = ±30° (low end) and θ = ±60° (high end of usable range).
Given
- L = 75 mm
- N = 20 RPM
- θnom = 0 °
- θlow = 30 °
- θhigh = 60 °
Solution
Step 1 — convert crank speed to angular velocity:
Step 2 — compute total stroke. The output moves from y = -2L to y = +2L as the crank sweeps -90° to +90°:
Step 3 — at the nominal mid-stroke point (θ = 0°), output velocity is at peak. Output velocity is the derivative of y = 2L sin(θ), so vout = 2L × ω × cos(θ):
That's about 314 mm/s — fast enough that you'd see the mirror move smoothly across the field of view of a fringe camera in roughly one second of stroke.
Step 4 — at the low end of the practical operating range (θ = 30°), velocity drops because cos(30°) = 0.866:
Step 5 — at the high end (θ = 60°), velocity drops further to cos(60°) = 0.5:
So between θ = 0° and θ = 60°, output velocity halves while input speed stays constant. Push past θ = 75° and velocity drops below 25% of peak — the mirror visibly decelerates and the fringe-counter loses temporal resolution at the stroke ends.
Result
The linkage delivers a 300 mm total stroke with a peak output velocity of 0. 314 m/s at mid-stroke, dropping to 0.272 m/s at θ = 30° and 0.157 m/s at θ = 60°. The mid-stroke region is where the mirror tracks smoothly with near-constant velocity — the practical sweet spot for fringe acquisition is between -45° and +45°, covering roughly 70% of total stroke at better than 70% of peak velocity. If your measured output velocity at mid-stroke is more than 5% below the predicted 0.314 m/s, check three things: stepper microstep loss under load (motor drops a step and undershoots commanded ω), crank-to-coupler midpoint pin clearance above 0.05 mm causing lost motion, or coupler bar length error — a 1 mm error on a 150 mm coupler shifts the entire velocity curve by 0.7%. Cross-axis error at the output above 0.1 mm almost always points to slider rail misalignment with the ground pivot rather than wear.
Scott Russell Linkage vs Alternatives
The Scott Russell competes with other rotary-to-linear converters whenever the design allows a guide rail on the output. Once you decide an output rail is acceptable, simpler mechanisms beat it on cost and stroke. The comparison below shows where it earns its place and where you should pick something else.
| Property | Scott Russell linkage | Slider-crank | Peaucellier-Lipkin linkage |
|---|---|---|---|
| Straight-line accuracy | Mathematically exact (limited only by tolerances) | Output is exact only because of slider rail | Mathematically exact (eight-bar) |
| Number of links | 3 links + slider | 3 links + slider | 8 links, no slider |
| Practical stroke range | 50-400 mm typical | 10 mm to 2 m+ | 20-200 mm typical |
| Output guide rail required | No — output is rail-free | Yes — output rides on a rail | No — output is rail-free |
| Velocity uniformity across stroke | Cosine profile, drops to 0 at ends | Near-sinusoidal | Better uniformity in mid-range |
| Load capacity at stroke end | Drops to zero at ±90° crank | High throughout stroke | Moderate, drops near limits |
| Build cost (relative) | Medium — needs precision pivot tolerances | Low — well-understood | High — eight precision pivots |
| Best application fit | Optical, deployable, contamination-sensitive systems | Engines, presses, general motion conversion | Precision metrology where no rail is acceptable |
Frequently Asked Questions About Scott Russell Linkage
The exact straight-line property depends on one geometric condition: the crank length must be exactly half the coupler length, and the crank pin must sit at the true geometric midpoint of the coupler. If either is off, the output traces an arc.
Measure the coupler bar end-to-end and divide by 2. The crank pin position must match that figure to within ±0.1 mm on a 150 mm coupler. The most common cause we see is builders measuring from the bar edge rather than the pin centre — a 3 mm pin diameter error gives you a 1.5 mm offset and a clearly bowed output path.
No. The instant the crank length deviates from L = coupler/2, the linkage stops producing exact straight-line motion. You'll get a longer stroke but the output traces an ellipse, not a line.
If you need more stroke, scale the entire linkage up — a 200 mm crank with a 400 mm coupler gives 800 mm of stroke while keeping the geometry. The constraint is packaging: coupler length doubles linearly with stroke, which is why most practical builds top out around 400 mm of travel before the swept volume becomes unworkable.
For most cases, yes. Scott Russell uses three links and one slider. Peaucellier uses eight links and seven pivots. At 100 mm stroke the accuracy difference is negligible if you hold pivot tolerances to ±0.02 mm on either design, but the Peaucellier has roughly seven times the cumulative pivot-clearance error budget.
Pick Peaucellier only if a slider is genuinely unacceptable in your application — for example, in a vacuum chamber where slider lubrication outgasses, or where particulate generation from a sliding contact would contaminate a wafer. Otherwise the Scott Russell wins on parts count, cost, and assembly tolerance stack-up.
The output velocity follows a cosine profile, which means output acceleration peaks twice per revolution — once at each stroke end. If your output is moving a real mass, that acceleration demands torque from the crank.
The fix is either to reduce input speed, reduce the moving mass, or limit operating range to ±45° crank angle where the acceleration is lower. If the spikes correlate with stroke direction reversal but not crank angle, the cause is different — it's almost always backlash in the gearbox driving the crank, not the linkage itself.
Three causes account for nearly every case. First, the slider has friction that loads the crank — if you're driving with a small stepper near its torque limit, it'll lose microsteps and the actual ω drops below commanded. Check by measuring crank speed directly with an encoder rather than trusting the command.
Second, pivot bushings with radial clearance above 0.1 mm let the coupler bar shift under load, effectively shortening L by a fraction of a millimetre and reducing output velocity proportionally. Third — and this catches a lot of people — if the slider rail isn't truly collinear with the ground pivot, the coupler end binds intermittently and the output velocity averages lower than the instantaneous theoretical value.
Yes, the kinematics are reversible. Push the output end along its straight line and the crank rotates. Some weighing-cell designs work exactly this way — a vertical force translates into crank rotation that's measured by an encoder or strain gauge.
The catch is that mechanical advantage reverses too. Near θ = ±90°, a small linear force at the output produces near-infinite torque at the crank, which is useful for force amplification but means small linear displacements at stroke end produce no measurable crank rotation. For sensing applications, stay within ±60° of crank angle to keep the response linear and stable.
References & Further Reading
- Wikipedia contributors. Scott Russell linkage. Wikipedia
Building or designing a mechanism like this?
Explore the precision-engineered motion control hardware used by mechanical engineers, makers, and product designers.
