The Sarrus Linkage: Straight-Line Motion Without Guides

The problem: drawing a straight line with rotating joints

Every revolute joint — a hinge or a pin — turns in a circle. So how do you build a machine whose output moves in an exact straight line using nothing but hinges? For most of the 19th century engineers couldn't, and it mattered: James Watt's beam engines needed straight guidance for the piston rod, and precision machine tools needed to slide without a wearing track. Watt (1784) and Chebyshev produced clever four-bar linkages that approximate a straight line over a short arc, but they always drift off by a small error.

The Sarrus linkage was the first mechanism to solve the problem exactly. Its trick is to stop thinking in a plane. Instead of one flat linkage, Sarrus used two folding panels arranged in 3D at right angles to each other. Each panel, on its own, still lets the top plate wobble and swing. But when you combine two panels whose fold-axes point in two different (orthogonal) directions, each panel forbids the exact motions the other one would allow. What survives that mutual veto is a single clean motion: straight-up-and-down translation. This is why the Sarrus is called a spatial or three-dimensional linkage — the straightness is a consequence of geometry in space, not of any rail or guide.

How it works: two orthogonal hinge chains cancel every rotation

The canonical two-sided Sarrus consists of a fixed lower plate, a moving upper plate, and two folding panels connecting them. Each panel is a chain of two rigid links (plates) joined to the plates and to each other by parallel hinges — three revolute joints per panel, six in all. Count the parts: 2 plates + 4 links = 6 bodies, joined by 6 revolute joints. That is the 6R spatial linkage.

Consider one panel in isolation. Its three hinge axes are all parallel — say, all pointing along the x-direction. Because those axes are parallel, that panel behaves like a little planar mechanism: it will only permit the upper plate to move within planes perpendicular to x. In other words, panel A kills any velocity component and any rotation that isn't consistent with sliding in its own plane. It allows straight motion along one axis (z, up-down) plus some unwanted freedoms in its plane.

Now add panel B, built identically but rotated 90°, so its hinge axes point along y. By the same logic, panel B only permits motion consistent with its plane. The two permitted motion sets overlap in exactly one shared direction — the vertical z-axis common to both planes. Intersecting the two constraint sets leaves a single translational freedom:

allowed(A) ∩ allowed(B) = pure translation along z

The upper plate cannot rotate (each panel forbids two rotation axes), cannot slide sideways (each panel forbids the other's in-plane slide), and cannot tilt. It can only rise and fall, tracing a perfect vertical line. The panels fold and unfold like the sides of a bellows; the platform on top goes exactly straight.

The mobility paradox: why Grübler says it shouldn't move

Here is the subtle part that makes the Sarrus a favorite in kinematics courses. The standard Chebychev–Grübler–Kutzbach mobility formula for a spatial mechanism is:

M = 6(n − 1) − Σ(6 − f_i)

where n is the number of links (including ground), and each joint of freedom f_i removes 6 − f_i degrees of freedom. A revolute joint has f = 1, so it removes 5. Plug in the Sarrus: n = 6 and 6 revolute joints:

M = 6 × (6 − 1) − 6 × 5 = 30 − 30 = 0

The formula flatly predicts zero mobility — a rigid, immovable structure. Yet the mechanism clearly moves with one degree of freedom. This is not a bug in the math; it is the signature of an overconstrained (paradoxical) mechanism. The general formula assumes the constraints are independent and generically placed. In the Sarrus, the special geometry — parallel axes within each panel, panels exactly orthogonal — makes several constraint equations redundant. Because those redundant constraints don't each subtract a fresh degree of freedom, one net freedom survives. The Sarrus is historically the first published overconstrained linkage, predating Bennett's 4R (1903) by half a century, and it taught mechanism theorists that mobility must be checked with screw theory or a rank analysis of the constraint Jacobian, not the naive count.

The practical consequence: the mechanism only works if the geometry is held to tolerance. If the two panels are not truly orthogonal, or a hinge axis is skewed, the redundancy breaks, the count reasserts itself, and the linkage locks up into a rigid frame or binds.

Real engineering numbers, trade-offs, and failure modes

• Stiffness varies with position. Near mid-stroke the folded panels stand at a steep angle and resist vertical load well. As the platform reaches full extension, the panels straighten toward a flat, near-singular pose where a small vertical force produces large hinge loads — mechanical advantage collapses and lateral stiffness drops sharply. Designers keep working strokes to roughly 60–80% of the fully-straightened height to stay clear of that toggle singularity.
• Side-load rejection is the weak point. A rail-and-carriage reacts bending moments directly; the Sarrus has to fight off-axis loads through the torsional stiffness of thin folding panels, so a moment of even a few N·m can twist the platform out of plane. It is excellent for concentric axial load, poor for cantilevered load.
• Tolerance stack-up is the failure mode. Because it is overconstrained, non-ideal geometry doesn't just add error — it can jam. A panel-orthogonality error of even ~0.5° or a hinge-axis skew introduces internal binding forces; hinges must therefore carry small clearances (a few hundredths of a millimeter) or use compliant/flexure hinges to absorb the mismatch.
• Backlash. Six pin joints in series means hinge clearances accumulate. Precision versions replace pin hinges with flexure (living) hinges — a single monolithic part with thin bending necks — eliminating backlash and lubrication entirely at the cost of limited stroke.

Typical hardware ranges from CubeSat solar-panel hinges only tens of millimeters across, up to warehouse vertical-lift platforms moving loads of tens to hundreds of kilograms over strokes of a meter or more.

Where it's used: from CubeSats to camera rigs

The Sarrus principle shows up wherever a straight lift is needed but a sliding track would be a liability:

• Space-deployable structures. The Sarrus is a standard construction unit for deployable arrays and polyhedral mechanisms. Folding solar-panel and antenna hinges in the Sarrus family deploy panels in vacuum without rails that could cold-weld or gall — a real hazard for sliding metal contacts in orbit. CubeSat and small-satellite panel hinges (e.g. of the type used on NASA's twin MarCO CubeSats that relayed the InSight Mars landing in 2018) rely on guideless folding hinge geometry rather than slides.
• Guideless vertical lifts. Warehouse vertical-lift modules and lab/medical lift tables use Sarrus-style or scissor-style folding to raise a platform straight up without a swaying rail, keeping the load centered.
• Camera and instrument platforms. Straight-line camera and focusing stages use the Sarrus to translate optics along one axis without introducing tilt from a worn slide.
• Compliant / MEMS versions. Because the topology maps cleanly onto flexures, the Sarrus is a common building block for compliant mechanisms and micro-stages that need precise, backlash-free straight-line travel etched into a single piece of silicon or spring steel.

Sarrus vs. Peaucellier, and a common misconception

The Sarrus is often mentioned alongside the Peaucellier–Lipkin linkage (1864), the other famous exact straight-line mechanism. The crucial difference is dimensionality. Peaucellier is a planar eight-bar mechanism that uses geometric inversion to convert a point moving on a circle into a point moving on a straight line — everything happens in one flat plane. The Sarrus is spatial: it achieves straightness by intersecting two three-dimensional constraint sets. Peaucellier gives elegant planar in-plane straightness; Sarrus gives a compact out-of-plane lift with fewer links. Historically, Sarrus came first (1853) but was largely overlooked, so Peaucellier long got credit for 'solving' the straight-line problem.

Common misconception: that the Sarrus is 'just an accordion' or a scissor lift and therefore approximate. It is not. A scissor lift's platform actually needs a roller in a slot at one end to stay level, and simple bellows have no defined path at all. The Sarrus produces an exact straight line purely from the intersection of hinge constraints — no guide, slot, or roller is involved. The related engineering pitfall is assuming that because it looks over-braced you can build it loosely: the same overconstraint that guarantees perfect straightness also means sloppy geometry causes it to bind rather than merely wobble. Precision in the hinge axes, not a precision rail, is what buys you the straight line.

Sarrus linkage vs. a conventional guided slider (rail + carriage) for vertical straight-line motion

Property	Sarrus linkage	Linear rail + carriage
Motion accuracy	Exact straight line by geometry	Depends on rail straightness/wear
Sliding surfaces	None — only rotating hinges	Ball or plain slide against a rail
Friction & wear source	Bearing/pivot only	Sliding contact + preload, galls in vacuum
Lubrication in vacuum/dust	Tolerant (dry bushings work)	Poor — galling, cold-welding, contamination
Load off axis / side loads	Weak — twists the folded panels	Strong — rail reacts moments directly
Stiffness at full extension	Low near straightened panels	High and roughly constant
Cost & part count	Low, sheet-metal + hinges	Higher (precision-ground rail, blocks)