Machine Kinematics
solved, then drawn
Field

Out of the plane

In space a body has six freedoms and a pin takes five away, so a closed loop needs seven joints before it moves at all. The mechanisms that move with four are not curiosities — one of them is in every car ever built.
Kutzbach's count against the measured mobility. Five closed loops of revolute joints. The count is 6(L − 1) − 5j, a statement about how many links and joints there are; the measurement is the number of joints minus the rank of the loop's screw system, which knows only where the axes point. They disagree for four of the five, and the one they agree on is the generic seven-joint loop — so the formula is not broken, it is blind to the special geometry that makes the other four work. The universal joint is counted at -2 degrees of freedom and is in every car built.

Six freedoms, not three

Kutzbach's mobility criterion for spatial mechanisms, why a closed loop needs seven revolute joints before it moves, and the measured screw-system rank that says otherwise for four loops out of five.

 A universal joint at 40° input, shafts 25° apart. Two shafts meeting at 25°, joined by a cross whose two pins are at right angles to each other and each at right angles to the shaft it carries. That is the whole geometry, and everything else follows from it. This is a four-joint spatial loop with all four axes through one point: Kutzbach counts −2 and the screw system has rank 3, so it has one degree of freedom and turns. At this instant the output shaft is at 42.79° while the input is at 40°, and the output is turning 1.0124 times as fast — which is not 1, and never is except at the four points of each turn where the curves cross.

The joint that is not constant velocity

The Hooke joint as a spherical four-bar, its velocity ratio measured by solving the loop and compared with Cardan's closed form, and the phasing condition that makes a pair of them constant-velocity.

 The Sarrus linkage at 0°. Two three-joint chains in perpendicular planes, joining a fixed plate to a moving one. The left chain's three axes are all parallel, so it allows the plate to move in its plane; the right chain's are parallel to a perpendicular direction and allow the plate to move in that one. What both permit is a straight line, and only a straight line. Six revolute joints in a single loop: Kutzbach says 0 degrees of freedom, the screw system has rank 5 and says 1, and the plate rises. Measured over 24 positions, its tilt never exceeds 2.5e-14 radians and it never leaves the axis by more than 6.8e-14 — exact, from pin joints, with no approximation anywhere in it.

Sarrus, and the straight line that is exact

The Sarrus linkage, its exactness measured against Watt's and Chebyshev's approximations, and why two three-joint chains in perpendicular planes give translation along a line with no approximation in it.

 Bennett's four-bar at 40°. Four bars, four revolute joints, and axes that are not parallel — a spatial four-bar, which Kutzbach counts at -2 degrees of freedom. Bennett's condition, sin α / a = sin β / b, makes the screw system rank 3 instead of 4, so the mechanism has 1. Driving the first joint through a full turn, 48 of 48 positions assemble. Orthographic projection, viewed from 40° azimuth and 24° elevation; dashed stubs mark the joint axes.

Bennett, and the condition that moves it

Bennett's overconstrained spatial four-bar, the condition sin α / a = sin β / b measured against a detuned copy, and what it means for a mechanism to work on a condition rather than approximately near one.

 A screw of pitch 0.25. A single screw drawn from its own decomposition. The six numbers of the twist go into a decomposition that returns a direction, a point on the axis and a pitch of 0.2500; the helix is then the path of a point at radius 0.5 about that axis, advancing 0.250 along it per radian turned. Over 1.15 turns the point advances 1.806 — the pitch times the angle, which is what pitch means. This screw is a screw.

Every motion is a screw

Chasles's theorem computed rather than quoted — the screw axis and pitch extracted from a rigid displacement, the instant centre shown to be the planar special case, and the six numbers that hold a motion.

 What each loop's constraint system is made of. For each mechanism: the order of the screw system its joints span, the order of the reciprocal system — the wrenches it carries without moving, which is always six minus the first — and what those wrenches are. A planar four-bar carries one force and two couples; a mechanism whose motion lies in no subgroup carries screws of finite pitch instead, and 2 of these 6 do.

What a mechanism cannot do

The reciprocal screw system computed from the same Jacobian that positions a mechanism, showing that a planar linkage carries a force and two bending couples, a spherical one three forces, and a Sarrus linkage a single couple.

 Whose constraints stay put. The largest principal angle between a mechanism's screw system at the start of its motion and at each later position. The 3 mechanisms whose motion lies in a subgroup of the rigid displacements — planar, spherical, translational — never leave the same subspace, and read between 1.9e-6 and 3.0e-6 degrees, which is the precision of an arccosine near one rather than a movement. The paradoxical ones turn through 49° and 57°. Every curve runs over its own range of motion, because Bricard's linkage assembles over 120 degrees and a shared axis would hide it.

Two ways to be overconstrained

Subgroup and paradoxical overconstraint separated by measuring how far a mechanism's screw system turns through its own motion, with a planar four-bar at 2e-6 degrees against Bennett's 89 and Bricard's 22.

 A spherical four-bar: 90°, 40°, 100°, 80°. Four revolute axes, all through one point, so the linkage lives on a sphere. Its link "lengths" are the angles between consecutive axes — 90°, 40°, 100°, 80° — and the arcs drawn between the axis directions measure 40.00°, 100.00°, 80.00°, 90.00°. The whole planar four-bar theory carries over with sines where lengths were, including Grashof's condition: sorted, the arcs give s + l = 140° against p + q = 170°, so the shortest arc does turn all the way round — and swept, the input reaches 36 of 36 positions over a driveable range of 360°. This is the mechanism a universal joint is a special case of, with two of the four arcs at 90°.

When the link lengths are angles

Spherical four-bar linkages built by solving on the sphere, with the spherical Grashof condition predicted from the arcs and measured by sweeping, and the universal joint identified as the case with two right angles.

 The cylindroid at 75° and 1.00 apart. Every screw in the two-system spanned by two revolute axes 1.00 apart along their common perpendicular and 75° out of parallel, drawn as its own axis. The axes sweep a ruled surface — the cylindroid — and each generator carries a pitch, running from -0.384 to 0.652 and reaching its two extremes on the two principal screws, which cross at a right angle at the centre. The surface is 1.035 long along its own axis, which is exactly the spread of the pitches: a cylindroid is as long as its pitches are far apart.

The smallest screw system has a shape

The cylindroid computed from two revolute axes, with the principal screws found from a two-by-two eigenproblem, the pitch law and the surface's width checked against their closed forms, and both degenerate two-systems refused.

 The framework Maxwell's count calls a structure. Six joints and twelve bars in space. Three coordinates each gives eighteen unknowns, six rigid motions come off, and twelve bars is exactly twelve constraints — Maxwell's count is 6 against six rigid motions, which is the definition of isostatic: no mechanism, no redundancy, every bar carrying its own share and nothing spare. The rank is 11, not twelve. There is one dependency among the bars and one freedom left over, and the freedom is a genuine finite motion: walked here with every bar held to 4.4e-16 of its own length. The reason is a symmetry — three pairs of joints exchanged by a half turn about one line — and it is built into the coordinates rather than assumed of the result. positioned by solving, not by drawing.

Twelve bars and a symmetry

The Bricard octahedron as a bar framework: an isostatic count, a rank one short, and a motion that exists because of a symmetry the count cannot see.

 How much of Bennett's turn survives a bar being wrong. The same four bars and the same four twists, with one bar's length changed by the amount on the left and nothing else touched. The bar shows the fraction of 48 sampled positions of the first joint at which the loop closes to within 10⁻⁹. Two parts in a thousand already costs most of the travel. This is what it means for a mechanism to work only on a condition rather than approximately near one — and it is why Bennett's linkage was a curiosity for eighty years before anyone could machine to it.

Bennett's condition is a ratio

What a scaling does to a spatial loop, why a mechanism with two kinds of parameter behaves differently from a planar one, and what an angle-only measurement recovers of it.

 In space the arithmetic allows almost nothing. A body in space has six freedoms and a revolute joint takes five, so a mobility of one needs (6n − 7)/5 joints — and that is an integer only when the link count leaves a remainder of two on division by five. The whole table is this: 7, 12, 17, 22 links, and nothing else. At seven links the degrees must sum to fourteen across seven links with none below two, so every link is binary and the graph is a single seven-cycle: there is exactly one spatial chain, and it is a loop. That is the census explanation for something the spatial field has lived with since it was written — every spatial mechanism in these essays is one closed loop — and it had never been stated as a count. The next admissible size is twelve links and thirteen joints, where two assortments are arithmetically possible, 157 candidates give 33 graphs, and 5 of them are chains — every one with ten binary links and two ternary, so the assortment with a quaternary link is empty exactly as four of the planar ones are.

In space there is one chain

The spatial census: why only some link counts are admissible, why the seven-link answer is a single loop, and what that says about a field whose every mechanism has been one loop since it was written.

 Four instruments, and only the last one names the group. Every instrument these essays have for an overconstrained loop, on the same six mechanisms. Kutzbach's count gives −2 for a planar four-bar and −2 for Bennett's. The rank of the constraint Jacobian gives three and three. Both are right and neither separates them. The last two columns are this field's: the span is how many dimensions the logarithms of the displacements the moving link actually reaches occupy, and closes at is the dimension after those are closed under the bracket. A planar four-bar closes at three and the three are planar motion; Sarrus closes at one, a translation, which is the exact straight line the spatial field measured by solving the mechanism sixty times. Bennett closes at six: its displacements occupy four dimensions and no group smaller than all of them contains those four. That is what "paradoxical" has meant in the earlier essays, stated as an integer.

A name for each overconstraint

The two kinds of spatial overconstraint distinguished by a closure dimension that names the subgroup, checked against the screw-system drift on the same six loops.

 Three joints, and the three screws that describe them. A leg of three revolute joints, drawn as its three axes, and the principal screws of the three-system they span, drawn through the system's own centre. The three principal axes are mutually perpendicular — worst cosine 1.1e-16 — and they meet at one point, missing it by 2.6e-16. Their pitches are -0.3766, -0.0338, 0.8075, and every screw the leg leaves free has a pitch the three of them give by h₁l² + h₂m² + h₃n². Nothing in the three joint axes looks like a right angle and the system's own frame is one.

What a leg of three joints leaves free

The three-system a three-revolute leg spans, with its principal screws found by a generalised eigenproblem, the pitch normal form checked on screws the eigenproblem never saw, and the cone of zero-pitch directions drawn.

 What a leg of three joints permits and what it resists, on one frame. The leg of three revolute joints in grey, and the principal screws of two three-systems drawn through one centre: the twists the leg lets its platform make, and the wrenches it holds without any joint turning. They are computed separately — the second is the reciprocal complement of the first, found by its own eigenproblem — and they come out with the same centre, 5e-16 apart, and the same three axes, parallel to 0e+0. Only the pitches differ, and each wrench pitch is its twist's with the sign reversed, to 1e-15: -0.3766 against 0.3766, -0.0338 against 0.0338, 0.8075 against -0.8075.

The lines a leg turns about and the lines it is pushed along

The reciprocal of a three-system found by its own eigenproblem and shown to share the system's principal frame with every pitch negated, the pitch quadric on which twists of one pitch and wrenches of the opposite pitch both lie, the hyperboloid of one sheet ruled by a leg's revolute axes and its lines of force, and a correction to where those axes are.

 Two legs that share a line of force. Two legs of three revolute joints each under one platform, drawn about their common centre: the first leg's joints in one colour and the second's in another. The second leg was built so that each of its joints passes through a point of one line, the dark one, which also crosses all three joints of the first leg — the six marked points, each met to 1e-16. A force along that line does no work on any joint of either leg, so both legs resist it: it is a load path the platform has twice over. The dashed line is what comes with it: the axis of the one twist both legs permit, a screw of pitch -0.049 that passes 0.032 from the shared line at 57.2° to it. Dragging moves the shared line along the first leg's first joint and builds the second leg afresh through each position.

A shared load is a shared motion

Two three-revolute legs on one platform: shared wrenches (redundant load paths) and shared twists (platform freedoms) counted from the stacked constraint systems and shown equal for any two three-joint legs; a second leg built through a line of force of the first, and one built on a revolute axis of the first's freedom; the shared twist's pitch and axis against the shared line; a finite-push test showing the shared twist is a first-order motion only; and the least singular value of the six wrenches as the pair's distance from a freedom.

 A third leg that keeps the platform's shake. The two legs of the earlier construction — the second built through a line of force of the first, the dark line — share one twist, whose axis is the dashed line: a screw of pitch -0.049. A third leg is drawn in a third colour, built so that its three joints span a family of motions containing that twist: two joints chosen freely and the third solved for. The platform on all three legs keeps its one freedom. The dotted line is the axis of the one wrench the third leg shares with the first: a screw of pitch 0.201, not a line of force, and 0.244 from the pair's shared line. Dragging picks another third leg from the same family.

A third leg that keeps the shake

A third three-revolute leg built so that its freedom contains the twist two legs sharing a line of force already share; 200 such legs and 60 built through the shared line, with the platform's freedom and redundancy and the wrench each third leg shares with each of the first two; the pitches of those shared wrenches; their reciprocity with the twist; the push test with three legs; and how the platform's nearness to a freedom grows as the third leg misses.

 Turning one joint about the shared line tunes the shake's first coefficient to nought. Two legs of three revolute joints each under one platform, sharing a line of force, the dark line, that crosses every joint of both — so they share one twist and the platform shakes along it. The second leg's first joint is turned about the shared line by 0°, which keeps it crossing the line and keeps the shared twist. The shake's second-order coefficient, the reciprocal product of the shared wrench with half the difference of the two legs' bracket terms, is 2.62e-2 here. It passes through nought at 36.488° and again at 109.612°; at those turns the platform still shakes, one order higher.

A shake has an order

Two three-revolute legs sharing a line of force and so a twist: the legs with the second's first joint turned about the shared line; the push test's second-order coefficient against the bracket of the shared wrench with half the difference of the legs' bracket terms, over 24 legs; the bracket as one joint turns, with its two roots; the gap against the push for four pairs, slopes two, three and four and none; the third-order coefficient along the curve on which the second is nought; and the four pairs tabulated.

 A pair moves until its gap reaches what its joints' play can take up. Two legs of three revolute joints each, sharing one twist, pushed along it: the gap left between their nearest poses, along the one wrench they share, for the pair built through the shared line, the pair with one joint tuned and the pair with two. The horizontal line is the gap the six joints' play can take up, 4.52 times a radial play of 10⁻⁶. Each pair travels until its curve meets the line: built through the shared line, 1.31e-2; one joint tuned, 6.63e-2; two joints tuned, 1.39e-1.

A shake travels as a root of the play

How far a pair of three-revolute legs sharing a twist travels when every joint has radial play: the gap along the shared wrench against the push for an untuned pair and pairs with one and two tuned joints, with the gap the play can absorb; travel against play, by bisection on the solved gap and by the root law; each joint's share of the play's capacity along the shared wrench; and how many times further the tuned pairs travel.

 Where two coefficients vanish, a curve; where three do, two points. The second leg's three joints each turned about the shared line by ψ₁, ψ₂ and ψ₃. Every point of the curve makes the shake's second- and third-order coefficients nought, σ₂ = σ₃ = 0, found by continuation from the earlier two-joint tuning (black) both ways, over an arc of 7.25 rad, until it reaches configurations where the two legs come close to sharing a second twist (crosses), where every coefficient runs off to infinity. Left: ψ₂ against ψ₁; right: ψ₃ against ψ₁. The curve is drawn in one colour where σ₄ is negative and another where it is positive; it changes sign at the two dots, (106.31°, 140.35°, 24.66°) and (136.39°, 57.24°, 18.34°), where σ₄ is nought too. Marked: 50% of the way along, at (107.0°, 141.4°, 25.2°), where σ₄ is positive.

Three tuned joints stop at the fifth order

Two shaking three-revolute legs with three joints of the second turned about their shared line: the curve where two coefficients vanish and the two points where three do; the fourth- and fifth-order coefficients along that curve; the push test at every tuning from none to three; every order's coefficient at every tuning; how far each travels in its joints' play; and the Taylor series against the bracket formula over twenty-four legs.

All essays