Peaucellier's parallel motion is an eight-bar linkage that converts circular input motion into exact straight-line output without sliding contact. The heart of the mechanism is the rhombus of four equal links pinned to two longer anchor links — that rhombus inverts the input point's path through the geometry of inversion, forcing the output pin to track a true straight line. Engineers used it on early steam engines and machine tools where a precise guide path was needed but high-quality flat slides were not yet manufacturable. The 1864 design was the first proven exact straight-line linkage in history.
Peaucellier's Parallel Motion Interactive Calculator
Vary the anchor links, rhombus links, input distance, and joint clearance to see the inversion constant, output position, sensitivity, and linkage geometry.
Equation Used
The Peaucellier cell inverts the driven point P about the fixed pivot O. For equal anchor links La and equal rhombus links Lr, the product OP x OQ is constant, so the output distance is OQ = (La² - Lr²) / OP.
- Two anchor links are equal length.
- Four rhombus links are equal length.
- Input point P lies on the inversion ray from O.
- Pin-clearance wander is estimated from the article rule of thumb: 0.1 mm per joint gives about 0.3 mm output wander.
The Peaucellier's Parallel Motion in Action
The mechanism uses 8 rigid links and 6 pin joints. Two long anchor links of equal length, La, pivot from a fixed point O. Four shorter links of equal length, Lr, form a rhombus pinned between the free ends of those anchors. The driven point P sits on one rhombus corner and is driven on a circular arc by a separate input crank, while the output point Q sits on the diagonally opposite rhombus corner. The geometry of inversion guarantees that the product OP × OQ is constant — and when P moves on a circle that passes through O, the inverse image Q moves on an exact straight line perpendicular to the line connecting O to the centre of P's circle.
This is why the linkage is also called the Peaucellier-Lipkin cell — Lipman Lipkin derived it independently a few years after Peaucellier. The straight-line motion is geometrically exact, not approximate like Watt parallel motion or the Chebyshev linkage. That distinction matters when you are guiding a piston rod into a stuffing box on a high-pressure steam engine: an approximate straight line means the rod sees side load on every stroke, which scrubs the gland packing and pulls the rod oval inside 2,000 hours of running.
Get the link lengths wrong and the cell stops being a cell. The two anchor links must match each other to within roughly 0.1% of length — not 1%, not 0.5%. The four rhombus links must match each other to the same tolerance. If La1 ≠ La2, the output traces a shallow arc instead of a straight line, and you will see the guided point wandering ±0.5 mm off the intended path over a 200 mm stroke. The other common failure is pin-joint slop — every pin needs a clearance under 0.05 mm in a precision build, because each joint's clearance compounds along the chain and the final output gets the sum of all six.
Key Components
- Fixed pivot O: The grounded reference point from which the two anchor links swing. The straight line traced by output Q lies perpendicular to a line drawn from O through the centre of input P's driving circle, so positioning O accurately on the machine frame within ±0.2 mm sets the orientation of the entire output path.
- Anchor links (2 long links of length La): Equal-length links pivoting from O to two opposite rhombus corners. Their lengths must match each other to within 0.1% — a 100 mm anchor pair must agree to ±0.1 mm, otherwise the inversion identity OP × OQ = La2 − Lr2 breaks and the output bows.
- Rhombus links (4 short links of length Lr): Four equal links forming a deformable rhombus between the anchor tips and the input/output points P and Q. They must all four be cut from the same stock and matched to ±0.1% length. Their length sets the working stroke window.
- Input point P: The driven pin, constrained to move on a circle that passes through O. A separate crank of radius equal to half the distance from O to the circle's centre drives P. If P's circle does not pass exactly through O, Q traces a circular arc of finite radius rather than a straight line.
- Output point Q: The opposite corner of the rhombus from P. Q is the point that traces the exact straight line. It carries the guided load — typically a piston rod, valve spindle, or machine slide — and the line it traces is perpendicular to OC, where C is the centre of P's circle.
- Pin joints (6 total): Five moving pins plus the fixed pivot at O. Each pin needs a radial clearance under 0.05 mm in a precision build. Slop compounds: 0.1 mm clearance per joint multiplied across the chain shows up as roughly 0.3 mm of unintended wander at the output Q.
Real-World Applications of the Peaucellier's Parallel Motion
The Peaucellier cell solved a real problem in the 1860s — how to guide a reciprocating shaft along a true straight line when nobody could yet machine a long flat slide accurately. It found its way into steam engines, marine ventilation gear, and later into demonstration apparatus and modern precision instruments where sliding-contact guides are unwelcome. Several of the surviving applications below still run today in heritage and laboratory settings.
- Steam Power: Royal Navy ventilation engines used Peaucellier linkages to guide blower rods on HMS Inflexible (1881), where reliable straight-line guidance was needed without the maintenance burden of crosshead slides in a confined engine room.
- Marine Engineering: Prony's experimental marine engine demonstrators in the late 19th century used the linkage to drive valve spindles where guide-bar lubrication was impractical.
- Precision Instruments: Optical comparators and early metrology stages used Peaucellier cells to translate a reading head along a true straight line without dovetail wear, including some Société Genevoise d'Instruments de Physique benches from the 1890s.
- Educational Apparatus: Cornell University's Reuleaux Kinematic Models collection holds a working brass Peaucellier linkage built by Gustav Voigt around 1882, still used to demonstrate exact straight-line motion.
- Aerospace Mechanisms: Hatch-deployment linkages on satellite payload bays occasionally use exact straight-line cells where a sliding mechanism would gall in vacuum — the linkage avoids any rubbing surface.
- Mechanical Watchmaking: High-end watch escapement testing rigs use miniature Peaucellier cells to translate a probe across a balance wheel's swing path with sub-micron straightness.
The Formula Behind the Peaucellier's Parallel Motion
The defining identity of the cell tells you the position of the output point Q given the position of the input point P. It expresses the product of the distances from O to P and from O to Q as a constant set by the link lengths. At the low end of the practical operating range — small input displacements near the centre of stroke — the geometry is robustly straight and tolerances are forgiving. At the high end of the range, near the limits of rhombus folding, the linkage approaches a singularity where small input errors amplify into large output errors. The sweet spot sits at roughly 60-70% of the geometric stroke limit, where output straightness stays under 0.05 mm per 100 mm of travel in a well-built cell.
Variables
| Symbol | Meaning | Unit (SI) | Unit (Imperial) |
|---|---|---|---|
| OP | Distance from fixed pivot O to driven input point P | mm | in |
| OQ | Distance from fixed pivot O to output point Q (the point tracing the straight line) | mm | in |
| La | Length of each of the two anchor links pinned at O | mm | in |
| Lr | Length of each of the four rhombus links | mm | in |
| k | Inversion constant, k = La2 − Lr2, which sets the scale of the straight-line output path | mm2 | in2 |
Worked Example: Peaucellier's Parallel Motion in a heritage paper-mill calender guide
You are sizing a Peaucellier cell to replace a worn crosshead slide on a recommissioned 1879 horizontal calender drive at a working paper-mill heritage site in Maine, where the original cast-iron slide bars have ovaled out beyond reclaim and the trustees want a sliding-contact-free guide for the 180 mm-stroke piston rod feeding the calender's pressure cylinder. You must pick anchor and rhombus link lengths that put the output Q on a true straight line at the rod's operating position, with the inversion constant k chosen so the output stroke clears the existing stuffing box without modification.
Given
- Required straight-line stroke = 180 mm
- La (anchor link length) = 240 mm
- Lr (rhombus link length) = 150 mm
- OP at mid-stroke (nominal) = 200 mm
Solution
Step 1 — compute the inversion constant k from the chosen link lengths. This is the fixed product OP × OQ that defines the cell:
Step 2 — at the nominal mid-stroke position, OP = 200 mm. Solve for OQ:
Step 3 — at the low end of the typical operating range, the input P swings closer to O. Take OP = 170 mm:
Step 4 — at the high end, P swings further from O. Take OP = 230 mm:
The output Q sweeps from 152.6 mm to 206.5 mm — a stroke of about 54 mm along the perpendicular straight line, comfortably bracketing the rod's working travel when scaled. Crucially, every value of OQ lies on the same straight line normal to the OC axis: that's the whole point of the cell. At the low-end position the rhombus is wide-open and tolerant — link-length errors of 0.5 mm barely move Q. At the high-end position the rhombus is closing toward its folded limit, and the same 0.5 mm length error pushes Q sideways by roughly 1.2 mm, visible as a curving output path. The sweet spot for steady running is the middle 60% of the OP range — here from about OP = 180 mm to OP = 220 mm — where straightness holds to under 0.05 mm.
Result
The nominal output position is OQ = 175. 5 mm, lying on the exact straight line perpendicular to the OC axis. In practice that means a draughtsman laying out the linkage on the engine frame can mark a single chalk line and the rod will track it within the build tolerance of the pins. Across the full operating range the output sweeps from 152.6 mm at high-end input to 206.5 mm at low-end input, with the centre 60% of that band giving the cleanest straightness — push outside it and you start trading geometric purity for stroke length. If you build the cell and measure 0.3 mm of lateral wander at Q instead of the expected sub-0.05 mm, the most common causes are: (1) anchor links La1 and La2 mismatched by more than 0.1% — re-measure with a vernier and re-cut, (2) the input P's driving circle does not pass exactly through O — shim the input crank pivot until it does, or (3) the four rhombus links cut from different stock with mixed length — match-grind all four from a single bar.
When to Use a Peaucellier's Parallel Motion and When Not To
When you need to guide a reciprocating point along a straight line, you have several options. The Peaucellier cell is the only one that's geometrically exact, but it pays for that with link count and pin-joint complexity. The realistic alternatives are Watt's parallel motion (used on the original beam engines) and the Chebyshev linkage (a four-bar approximate straight-line generator).
| Property | Peaucellier-Lipkin cell | Watt parallel motion | Chebyshev linkage |
|---|---|---|---|
| Straightness accuracy over 200 mm stroke | Exact (geometric, ≤ 0.05 mm in good build) | Approximate (~0.3-0.5 mm bow) | Approximate (~0.2-0.4 mm bow) |
| Link count | 8 links, 6 pin joints | 3 links, 4 pin joints | 4 links, 4 pin joints |
| Build complexity and cost | High — six matched pins, two matched pairs of link lengths | Low — three links and a fixed pivot | Low — classic four-bar |
| Maximum useful stroke as fraction of envelope | ~60% of geometric limit before singularity effects appear | ~40% of beam length | ~25% of crank radius |
| Sensitivity to link-length error | High — 0.1% mismatch shows in output path | Moderate — small errors absorbed by approximate nature | Moderate — same |
| Typical historical application | Marine ventilation engines, precision metrology stages | Watt and Boulton beam engines, 1784 onward | Late 19th-century linkage demonstrations |
| Load capacity at output | Limited by pin joints — moderate | High — beam carries direct piston load | Moderate |
Frequently Asked Questions About Peaucellier's Parallel Motion
The most likely cause is that your input point P is not being driven on a circle that passes exactly through the fixed pivot O. The whole inversion identity only generates a straight line when P's path is a circle through O — if P's driving crank is offset by even 1-2 mm, P traces a circle that misses O and the inverse image Q traces a circle of finite radius rather than a true line.
Check this by clamping the cell, marking O, and tracing P with the crank disconnected. The traced circle's edge must touch the O point. If it doesn't, shim the crank-bearing housing until it does. This is a more common build error than mismatched links because it's harder to verify by inspection.
The force ceiling is set by pin-joint bearing pressure and link buckling, not by the geometry. A cell scaled to La = 600 mm with hardened steel pins and bronze bushings can handle several kN at the output, but you'll find the pins govern the design — pin diameter has to grow with load while pin clearance has to stay under 0.05 mm regardless. That's a hard combination to manufacture economically.
For piston rods above about 5 kN axial load, a crosshead slide with white-metal liners is generally cheaper and more reliable than scaling up the cell. The Peaucellier wins where sliding contact is forbidden — vacuum mechanisms, ultra-clean environments, or precision metrology — not where load alone is the deciding factor.
You choose Peaucellier when the application cannot tolerate the bow that approximate straight-line linkages produce. A Chebyshev or Hoekens linkage is fine for a walking-toy foot or a crude reciprocating drive, but the output deviates from a straight line by 0.5-1% of stroke length. On a 200 mm stroke that's 1-2 mm of lateral wander.
If you're guiding an optical probe, a stuffing-box rod that can't tolerate side-load, or a metrology stage, that wander is unacceptable and the eight-link complexity of Peaucellier earns its keep. If you're driving a child's toy or a non-critical demonstration, the four-bar linkage is the better engineering choice.
You're hitting the singularity where the rhombus folds flat. As OP grows, OQ shrinks toward zero (or vice versa), and the rhombus links approach colinearity with the anchor links. At that point the cell has zero mechanical advantage to push through, and any small disturbance can flip it into the mirror-image configuration — which is the snap-through you're seeing.
The fix is to add a hard stop on the input crank that limits OP to between roughly 0.7 × (La − Lr) and 0.95 × (La + Lr). Inside that window the rhombus stays well away from the singular configurations and the output is well-behaved. Don't try to use the full geometric range.
The output stroke is approximately k × (1/OPmin − 1/OPmax), where OPmin and OPmax are the input swing limits. To hit a 50 mm output stroke with OP swinging from 180 to 220 mm, you need k ≈ 50 / (1/180 − 1/220) ≈ 50 / 0.00101 ≈ 49500 mm2.
Then pick La and Lr so that La2 − Lr2 matches that k. There's a family of valid (La, Lr) pairs — pick one where Lr is roughly 0.6 × La for a balance of stroke length and singularity margin. Going thinner on the rhombus (smaller Lr) gives a stiffer cell but reduces the usable input swing.
Brass-on-brass pin joints wear quickly, especially if the cell is running dry. Each pin develops 0.1-0.2 mm of clearance after a few hundred cycles, and because there are six joints in the chain, the clearances add. By the time each joint has 0.15 mm slop you've got nearly a millimetre of unintended wander at Q.
The fix is to ream the pin holes oversize, fit hardened steel pins running in bronze or PTFE bushings, and keep a film of light oil on every joint. A demonstration cell built this way will hold straightness through tens of thousands of cycles. The Voigt models at Cornell were built this way in the 1880s and still work.
References & Further Reading
- Wikipedia contributors. Peaucellier–Lipkin linkage. Wikipedia
Building or designing a mechanism like this?
Explore the precision-engineered motion control hardware used by mechanical engineers, makers, and product designers.
