A beam-driven crank and flywheel is a linkage that converts the oscillating motion of a rocking beam into continuous rotary motion at a flywheel-loaded crankshaft. The beam swings on a central pivot, and a connecting rod from the beam end drives a crank pin offset from the shaft axis, while the flywheel's rotational inertia carries the crank through dead centres. This arrangement let early steam engines drive mills, pumps, and looms with smooth shaft output instead of pure reciprocation. James Watt's 1782 rotative beam engine is the canonical example — it powered British industry for over a century.
Beam-driven Crank and Fly-wheel Interactive Calculator
Vary stroke, rod length, beam length, and shaft speed to see crank throw, linkage ratio, beam swing, and mean piston speed.
Equation Used
The article states that the crank pin offset is half the desired stroke. This calculator uses that throw to estimate the rod-to-crank ratio, the beam half-angle needed to provide the stroke, and the mean piston speed at the selected flywheel RPM.
- Crank throw is half the desired piston stroke.
- Beam is treated as a rigid rocker with small-angle motion.
- Lever ratio is 1:1 between piston end and crank end.
- Mean piston speed uses two strokes per revolution.
The Beam-driven Crank and Fly-wheel in Action
Picture the beam as a giant seesaw. One end gets pushed up and down by a piston, the other end carries a connecting rod that drops to a crank pin on a horizontal shaft. As the beam rocks, the connecting rod traces a near-vertical line at its top end and a circle at its bottom end. The crank pin, offset from the shaft centre by half the desired stroke, is dragged around that circle. The flywheel — a heavy iron disc keyed to the shaft — stores kinetic energy on the power stroke and releases it through the dead centres where the connecting rod is collinear with the crank and produces zero torque.
Why design it this way? Because a piston only pushes and pulls — it cannot rotate a shaft on its own. The beam acts as a force multiplier and a motion router, and the crank converts the linear stroke into rotation. The flywheel is non-negotiable. Without it, the shaft would stall every half-revolution at top and bottom dead centre. A typical Watt-era flywheel ran at 15 to 30 RPM with a moment of inertia sized so that speed variation stayed within ±2% per revolution.
If the geometry drifts, the engine tells you immediately. A connecting rod that is too short relative to the crank throw — say a rod-to-crank ratio below 3:1 — produces severe secondary forces and you'll hear knocking at the crosshead. If the beam pivot bushings wear past about 0.5 mm radial slop, the parallel motion linkage at the piston end loses straight-line tracking and the piston rod scuffs the gland packing. Crank pin journal wear above 0.3 mm causes hammering at every reversal — a sound any old-time engineer recognises.
Key Components
- Working beam: The cast-iron rocking lever pivoted at its centre on a massive trunnion. It transfers force from the piston end to the crank end with a typical lever ratio of 1:1, though some pumping engines used 2:1 to favour stroke over speed. Beam deflection under load must stay below 1/1000 of its length to keep the parallel motion accurate.
- Parallel motion linkage: Watt's three-bar linkage at the piston end that constrains the piston rod to near-straight-line travel. Deviation from true vertical is typically held under 1 mm over a 2 m stroke. Without it, the rod would bind in the gland.
- Connecting rod: The link between beam end and crank pin. Length is sized so the rod-to-crank ratio sits between 4:1 and 6:1 — shorter ratios produce harsh secondary inertia forces, longer ratios waste height. The big-end bearing typically uses bronze shells with 0.05 to 0.10 mm running clearance.
- Crank and crank pin: An offset arm on the main shaft that converts the rod's reciprocating pull into rotation. The throw equals half the piston stroke. The crank pin journal must be ground to within 0.02 mm of round — anything worse and the rod big end pounds at every reversal.
- Flywheel: A heavy rim-loaded disc, often 4 to 6 m in diameter on full-size beam engines, that stores rotational kinetic energy. It carries the crank through top and bottom dead centre where torque drops to zero. Sized so coefficient of fluctuation stays below 0.02 for textile drive, 0.04 for general mill work.
- Main shaft and bearings: Forged shaft running in bronze or white-metal bearings. Shaft diameter is sized for both torque transmission and beam-load reaction at the crank end. Bearing clearance held at 0.001 × shaft diameter for steady running.
Real-World Applications of the Beam-driven Crank and Fly-wheel
Beam-driven cranks with flywheels show up wherever a slow, heavy reciprocating prime mover needs to deliver smooth rotary shaft power. The combination dominated industrial drive from roughly 1782 to the late 1800s, then survived in pumping stations and preserved engines into the 20th century. The mechanism still answers a real engineering question — how do you smooth out a pulsing input torque without electronics — and that's why you see the same arrangement in modern reproductions, educational models, and a handful of niche industrial uses.
- Heritage steam engineering: Crofton Pumping Station in Wiltshire, UK — its 1812 Boulton & Watt beam engine still pumps water to the Kennet & Avon Canal using the original beam, crank, and 24-foot flywheel.
- Industrial heritage / preserved engines: Kempton Park Steam Engines in London — twin triple-expansion beam engines with rotative cranks, each driving a 62-tonne flywheel for water supply pumping.
- Educational models: Stuart Models 'Beam Engine' kit (Stuart No. 9) — a working 1:24 scale beam-driven crank and flywheel built by hobby machinists for teaching slider-crank kinematics.
- Textile mills (historical): Quarry Bank Mill, Cheshire — beam engine drove the entire weaving floor through a crank, flywheel, and overhead lineshaft from 1810 onwards.
- Mine drainage (historical): Cornish beam engines at South Crofty tin mine used the beam-crank-flywheel layout to drive both the pumping rod and a winding drum from a single shaft.
- Museum demonstration drives: The Henry Ford Museum's 1855 Corliss-type beam engine — used as a hands-on display of how rotative motion was generated before electric drives.
The Formula Behind the Beam-driven Crank and Fly-wheel
The most useful formula for a beam-driven crank and flywheel is the coefficient of fluctuation, which tells you how much the shaft speed wobbles between the power stroke and the dead-centre coast. At the low end of the range — say a heavy pumping engine running at 15 RPM with a massive flywheel — fluctuation might sit at 0.01 (1%), which feels rock-steady. At the nominal mill-drive sweet spot of 30 RPM with a properly sized flywheel, you'll see 0.02 to 0.03. Push the same engine to 60 RPM with the original flywheel and fluctuation drops in absolute terms but the engine starts vibrating because the rim stresses climb with the square of speed. The formula tells you what flywheel mass moment of inertia you need to keep the shaft speed within tolerance for the load.
Variables
| Symbol | Meaning | Unit (SI) | Unit (Imperial) |
|---|---|---|---|
| I | Required mass moment of inertia of the flywheel about the shaft axis | kg·m² | lb·ft² |
| ΔE | Maximum fluctuation of energy per cycle (energy stored and released by the flywheel) | J | ft·lbf |
| Cs | Coefficient of fluctuation of speed (allowable speed variation, dimensionless) | — | — |
| ω | Mean angular velocity of the crankshaft | rad/s | rad/s |
Worked Example: Beam-driven Crank and Fly-wheel in a restored Cornish-style beam engine driving a flour mill
You are sizing the flywheel for a 1:6 scale working replica of a Cornish beam engine that will drive a small demonstration flour mill at a heritage site. The engine produces a maximum energy fluctuation of 1,200 J per revolution at the crank, and you need shaft speed steady enough that the millstones grind evenly. Target nominal speed is 45 RPM, with a typical operating range of 30 to 60 RPM depending on grain feed rate. Allowable coefficient of fluctuation for grain milling is 0.03.
Given
- ΔE = 1200 J
- Cs = 0.03 dimensionless
- Nnom = 45 RPM
- Nlow = 30 RPM
- Nhigh = 60 RPM
Solution
Step 1 — convert nominal shaft speed from RPM to angular velocity in rad/s:
Step 2 — apply the flywheel inertia formula at the nominal 45 RPM operating point:
That sets the design target. For a rim-loaded cast-iron flywheel of 1.8 m radius, the required rim mass works out to roughly 555 kg — a serious lump of iron, but completely typical for a beam engine of this scale.
Step 3 — at the low end of the operating range, 30 RPM (ω = 3.142 rad/s), the same flywheel gives a tighter speed regulation:
Wait — that's worse, not better. Lower speed means lower stored kinetic energy for the same inertia, so fluctuation climbs. At 30 RPM you'd see the millstones pulse visibly with each power stroke, and the flour would come out unevenly ground. At the high end, 60 RPM (ω = 6.283 rad/s):
At 60 RPM the shaft runs glass-smooth — fluctuation is almost halved compared to nominal. But the rim hoop stress scales with ω2, so you've doubled it from the 45 RPM design point, and a cast-iron rim above about 30 m/s rim speed starts cracking. Check the rim velocity: π × 3.6 × 60 / 60 = 11.3 m/s, well within safe limits. So this flywheel is happy across the full 30–60 RPM range, but the milling quality demands you stay above about 40 RPM.
Result
The nominal flywheel mass moment of inertia required is approximately 1,800 kg·m², which translates to a 3. 6 m diameter cast-iron rim weighing about 555 kg. At 45 RPM you'll see steady millstone rotation and even flour output. The range comparison shows the catch with beam engines — at 30 RPM the same flywheel only achieves Cs ≈ 0.067 (visible pulsing), while at 60 RPM you get Cs ≈ 0.017 (effectively silk-smooth), so the sweet spot for grinding sits at the upper half of the speed range. If your real-world measured fluctuation comes in worse than 0.03 at nominal speed, the most common causes are: (1) crank pin journal out of round by more than 0.05 mm, which adds a torque ripple the flywheel can't smooth; (2) loose flywheel hub key, letting the rim lag the shaft slightly on each power stroke; or (3) governor valve sticking, which shifts ΔE upward beyond the 1,200 J design figure.
Choosing the Beam-driven Crank and Fly-wheel: Pros and Cons
The beam-driven crank and flywheel competes against simpler and more modern ways to convert reciprocating motion to rotation. Each alternative wins on specific axes — speed, footprint, cost, or maintenance — and loses on others. Here's how it stacks up against the two most common substitutes.
| Property | Beam-driven crank and flywheel | Direct-acting slider-crank (no beam) | Scotch yoke with flywheel |
|---|---|---|---|
| Typical operating speed (RPM) | 10–60 RPM | 100–6000 RPM | 50–1500 RPM |
| Speed regulation (coefficient of fluctuation, typical) | 0.02–0.04 | 0.01–0.03 | 0.03–0.05 |
| Footprint per kW output | Very large (multi-storey building) | Compact (engine block) | Medium |
| Capital cost (modern reproduction) | High — bespoke castings | Low — standard engine practice | Medium — custom yoke and slot |
| Maintenance interval (running hours between rebuilds) | 20,000+ hours | 2,000–10,000 hours | 5,000–15,000 hours |
| Lifespan (service life of major castings) | 100+ years documented | 20–40 years | 30–60 years |
| Best application fit | Heritage, demonstration, slow heavy pumping | Vehicles, generators, compressors | Slow heavy presses, valve drives |
| Mechanical complexity (linked parts) | High — beam, parallel motion, rod, crank | Low — rod and crank only | Medium — yoke and slot |
Frequently Asked Questions About Beam-driven Crank and Fly-wheel
Stalling at TDC almost always means the flywheel's stored kinetic energy at that instant is below the friction torque demand, not that the flywheel is too light overall. Three real causes show up repeatedly: the engine is being barred over too slowly to build up rim speed before the steam admits, the valve timing is letting steam in late so peak torque arrives after TDC instead of just past it, or the crosshead bushings are dry and friction has climbed.
Quick diagnostic — spin the flywheel by hand with the piston disconnected and time how long it freewheels. A well-set scale beam engine should coast for 30+ seconds. If it stops in under 10, you have friction, not inertia, problems.
The ratio sets the secondary inertia force amplitude and the height of your engine. At 4:1, secondary forces are about 25% of primary and you save vertical space — fine for a compact mill engine. At 6:1, secondaries drop to roughly 17% and the motion is closer to pure sinusoidal, which makes the flywheel's smoothing job easier and reduces beam-end side loads.
For a heritage replica running below 60 RPM, 5:1 is the practical sweet spot. Below 4:1 you'll feel a knock at every reversal because the rod angularity peaks too sharply. Above 6:1 you're just adding height and weight for diminishing returns.
Adding mass at the hub does almost nothing — moment of inertia scales with radius squared, so doubling the rim mass at the same diameter doubles I, but moving the existing mass outward by 40% has the same effect. If you bolted plates to the flywheel face near the hub, you barely shifted I.
Check where the added mass sits. The rim radius dominates. Also verify the flywheel hub key is tight — a sloppy key lets the rim lag the shaft on each power pulse, which masquerades as poor smoothing even though the inertia is correct on paper.
Yes, and many late-19th-century beam engines did exactly this — it's called a 'side-lever' or 'crosshead' modification. You lose a small amount of historical authenticity but you gain easier alignment and longer gland packing life because the piston rod travels in a guaranteed straight line.
The trade-off is the slide bar wears, and you now need to maintain its straightness within about 0.1 mm over the stroke length. Watt's parallel motion has no sliding contact, so it can run for decades with just bushing replacement. For a working demonstration engine that runs daily, the crosshead is usually the better choice.
This is counterintuitive but classic beam engine behaviour. Under load, the governor opens the throttle and admits steam through a longer portion of the stroke, which spreads the torque pulse and lets the flywheel smooth it easily. At no-load, the governor cuts off steam very early in the stroke, producing a short sharp torque spike followed by a long coast.
That short impulse excites the natural torsional frequency of the shaft-flywheel system, and you feel it as vibration. The fix is either tightening the governor response so it doesn't chase the load, or accepting that no-load running is hard on the engine and avoiding extended idle periods.
Conventional engineering practice caps cast-iron flywheel rim speed at about 30 m/s for plain rim construction, and below 25 m/s if the rim has spokes that introduce stress concentrations at the joins. Above these limits the hoop stress from centrifugal force approaches the tensile strength of grey cast iron (around 200 MPa) and you risk rim burst — historically a catastrophic failure mode that killed mill workers.
For a beam engine running at 45 RPM with a 1.8 m radius rim, rim speed is only 8.5 m/s — well within safe territory. You only run into the limit on small high-speed reproductions trying to squeeze too much speed from a too-small flywheel.
The knock location and timing tells you. A crank pin (big-end) knock comes at every reversal of rod load — twice per revolution — and you feel it at the connecting rod itself if you rest a hand on it. A main bearing knock is duller, lower-pitched, and seems to come from the engine bedplate.
Diagnostic check — bar the engine slowly through a full revolution with steam off and watch for visible movement at each joint. Big-end clearance above 0.15 mm is audible; main bearing clearance above 0.20 mm produces the bedplate knock. Re-shimming or re-metalling the bearing shells fixes both.
References & Further Reading
- Wikipedia contributors. Beam engine. Wikipedia
Building or designing a mechanism like this?
Explore the precision-engineered motion control hardware used by mechanical engineers, makers, and product designers.
