An ancient windmill is a wind-driven machine that captures airflow on inclined sails fixed to a horizontal or vertical shaft and converts it into rotary mechanical power for grinding grain, pumping water, or sawing timber. Unlike water mills, which need a flowing river and a fixed site, a windmill works anywhere the wind blows and the operator can yaw it into the breeze. Builders used it to mechanise heavy repetitive labour long before steam. A well-trimmed Dutch tower mill produced 10–15 kW of shaft power in a steady 7 m/s wind — enough to grind 200 kg of wheat per hour.
Ancient Windmill Gear Train Interactive Calculator
Vary sail speed and gear diameters to see how the brake wheel steps millstone RPM up from the slow windshaft.
Equation Used
The brake wheel drives the smaller wallower, so millstone speed equals sail speed multiplied by the diameter ratio. The default uses the worked diagram values: 2.4 m brake wheel, 0.3 m wallower, and 15 rpm sails.
- Gear tooth count is proportional to pitch diameter.
- Ideal gear mesh with no slip or efficiency loss.
- Brake wheel drives the smaller wallower gear directly.
Operating Principle of the Ancient Windmill
Wind hits the sails at an angle. Each sail is set at a pitch — typically 18° at the tip rising to 25° near the hub on a traditional Dutch common sail — so the wind exerts both a drag force and a lift force, and the lift component spins the windshaft. The windshaft is a massive oak or iron beam that runs through the cap of the mill and carries the brake wheel, a large face gear roughly 2.4 m in diameter on a typical English tower mill. The brake wheel meshes with the wallower, a smaller iron or wooden gear on the vertical main shaft, and that ratio (often 8:1 stepping up) takes the slow 15-20 RPM of the sails up to 120-150 RPM at the millstone.
The operator controls speed two ways. First, by reefing the sail cloth — rolling cloth on or off the lattice frame to expose more or less sail area. Second, by the brake wheel itself, which carries a wooden brake band the miller tightens with a lever from the stage. If the wind picks up unexpectedly and the miller cannot reef in time, the stones can over-speed, the upper runner stone lifts off the bed stone, and the mill runs without grinding — millers called this "the stones racing," and it scorches the wooden tentering gear in minutes.
Yaw alignment matters as much as sail trim. A post mill rotates its entire body on a central oak post, pushed around by a tail pole. A tower mill or smock mill rotates only the cap, driven either by a tail fanstaff (a small auxiliary windmill at right angles) or by hand. If the cap sits more than about 15° off the wind, output drops by half and the windshaft sees an asymmetric load that quickly chews out the front neck bearing — a hardwood block bedded in tallow, replaced every few years on a working mill.
Key Components
- Sails (sweeps): Four lattice frames, each 9-12 m long on a Dutch tower mill, covered in canvas the miller rolls in or out to control area. The pitch twists from about 25° at the hub down to 18° at the tip, which keeps the angle of attack roughly constant from root to tip as tangential speed increases.
- Windshaft: The main rotating beam carrying the sails. Traditionally English oak, later cast iron after about 1820. Sits at a 5-15° upward tilt so the sails clear the tower and so the weight of the sail assembly bears down on the rear thrust bearing rather than out on the neck.
- Brake wheel: A large face gear, typically 2.0-2.5 m diameter, mounted on the windshaft. Doubles as the speed-control surface — a wooden brake band wraps the rim, and the miller pulls a lever to friction-stop the mill. Wooden cogs ('cants') in the rim are individually replaceable when worn.
- Wallower: The driven gear at the top of the vertical main shaft. On an 8:1 step-up, a 0.3 m wallower meshes with a 2.4 m brake wheel. Tooth pitch must match within about 1.5 mm or the cogs hammer and split.
- Main shaft and stone nut: Vertical oak or iron shaft running down through the mill body to a small pinion (the stone nut) that drives the runner millstone. The stone nut can be lifted out of mesh by the 'jack ring' to disengage the stones without stopping the sails.
- Cap and curb: On a tower mill, the rotating timber cap sits on a circular curb of greased oak or cast iron rollers at the top of the masonry tower. The cap weighs 5-10 tonnes loaded; the curb must stay round to within roughly 10 mm or the cap binds when the fantail tries to yaw it.
- Fantail: Small 6-bladed auxiliary windmill mounted at right angles to the main sails on the rear of the cap. Whenever the wind shifts, the fantail catches it side-on and drives a worm gear that yaws the cap back into wind. Edmund Lee patented this in 1745 and it removed the need for a miller to manually winch the mill round.
Real-World Applications of the Ancient Windmill
Ancient windmills did three jobs: grinding, pumping, and sawing. The exact arrangement of gearing and machinery downstream of the windshaft changed by application, but the wind-capture front end stayed the same. Surviving working mills across the Netherlands, England, Greece, and Iran still demonstrate every variant.
- Grain milling: Jack and Jill mills at Clayton, West Sussex — a working post mill (Jill, built 1821) and tower mill (Jack, built 1866) that ground flour commercially into the 1900s and still grind on demonstration days.
- Land drainage: The Kinderdijk windmill complex in South Holland — 19 polder mills built from 1738-1761 that lifted water from the Alblasserwaard polder up into the Lek river, each moving roughly 30 m³/min in a moderate wind.
- Sawmilling: De Otter sawmill in Amsterdam (1631), a paltrok-type windmill that drove a reciprocating frame saw cutting Baltic pine for the Dutch shipbuilding industry.
- Oil pressing: De Zoeker on the Zaanse Schans — an octagonal smock mill built in 1672 that crushes linseed and rapeseed using stamper hammers driven from the main shaft.
- Vertical-axis grain milling: The Nashtifan windmills in eastern Iran — vertical-axis Persian windmills (asbads) dating to roughly the 9th century, still grinding grain today using woven reed sails on a vertical shaft.
- Spice and dye milling: De Kat on the Zaanse Schans (1781), a working paint mill that grinds chalk, ochre, and lapis pigments for restoration suppliers across Europe.
The Formula Behind the Ancient Windmill
The shaft power a windmill produces follows the standard wind-power equation modified by a power coefficient Cp. This matters because wind speed enters as a cube — double the wind, you get eight times the power, and that is exactly why mills need brakes and reefing gear. At the low end of the useful operating range (around 4 m/s) a tower mill barely turns the stones and the miller will not bother starting up. At the nominal working wind of 7-8 m/s the mill hits its design sweet spot, around 10-15 kW shaft. Above 12 m/s the miller must reef sail or face overspeed; above 18 m/s he furls completely and sets the brake.
Variables
| Symbol | Meaning | Unit (SI) | Unit (Imperial) |
|---|---|---|---|
| P | Shaft power delivered to the windshaft | W | hp |
| ρ | Air density (≈1.225 at sea level, 15°C) | kg/m³ | lb/ft³ |
| A | Swept area of the sails (π × R2 for the circle the sail tips trace) | m² | ft² |
| v | Free-stream wind speed at hub height | m/s | ft/s |
| Cp | Power coefficient — fraction of wind energy converted to shaft work. Ancient cloth-sail mills run 0.10-0.20; modern HAWTs reach 0.45. | dimensionless | dimensionless |
Worked Example: Ancient Windmill in a restored Dutch grain tower mill
You are estimating shaft power for a restored Dutch grain tower mill at Schiedam with a 26 m sail span, four common sails, fully unfurled cloth, and a measured Cp of 0.17 from on-shaft torque measurements. The miller wants to know what to expect across a normal Dutch coastal wind day, where wind shifts from 4 m/s at dawn to 8 m/s mid-morning and gusts to 12 m/s by afternoon.
Given
- Sail span (tip to tip) = 26 m
- Swept radius R = 13 m
- ρ = 1.225 kg/m³
- Cp = 0.17 —
- v range = 4 – 8 – 12 m/s
Solution
Step 1 — compute swept area from the sail radius:
Step 2 — compute nominal shaft power at the design-day wind of 8 m/s:
That is the sweet spot. The stones turn at roughly 130 RPM, the miller can dress two pairs of stones simultaneously, and grain feeds smoothly through the hopper at about 250 kg/hr.
Step 3 — at the low end of the useful range, 4 m/s:
Eight times less than nominal, because v is cubed. At 3.5 kW the mill turns but barely overcomes friction in the wallower and stone nut. The miller would disengage the stones with the jack ring and let the sails idle until the wind builds.
Step 4 — at the high end, 12 m/s:
Theoretically. In practice the miller reefs the sail cloth back to roughly 40% area before this point because the brake wheel cannot dissipate that much energy as friction heat — the wooden brake band would char and seize. Effective output stays around 30-35 kW with the sails reefed.
Result
Nominal shaft power at 8 m/s comes out to 28 kW. That is enough to drive two pairs of 1.4 m French burr stones at 130 RPM and produce roughly 250 kg of wheat flour per hour — a typical day's work for a commercial Dutch tower mill. At 4 m/s output collapses to 3.5 kW (sails turn but no useful grinding); at 12 m/s the formula predicts 95 kW but real output stays near 30-35 kW because the miller reefs cloth to protect the brake. If your measured shaft power runs 30%+ below predicted, check three things: cap yaw alignment (more than 15° off-wind halves output), sail cloth condition (rotted or torn cloth leaks pressure across the lattice), and brake wheel cog wear (worn cants slip on the wallower under torque and waste power as cog hammer rather than rotation).
When to Use a Ancient Windmill and When Not To
Ancient windmills compete with water wheels, animal-driven mills, and (later) steam engines. Each has a niche, and the choice came down to local geography and the cost of the alternative.
| Property | Ancient Windmill | Water Wheel | Horse Mill |
|---|---|---|---|
| Typical shaft power | 5–25 kW | 2–40 kW | 0.5–1 kW per horse |
| Site requirement | Open exposure to prevailing wind | Flowing watercourse with head | Anywhere with feed and stable |
| Operating availability | 50–70% of days (wind dependent) | Near 100% (river dependent) | On demand |
| Output stability | Highly variable — v³ dependence | Stable except in flood/drought | Stable but limited by animal fatigue |
| Capital cost (period) | High — large carpentry build | Very high if dam/race needed | Low |
| Operating cost | Low — wind is free, sail cloth wears | Low — water is free | High — fodder and stable labour |
| Typical lifespan | 100+ years with cap and sail rebuilds | 30–50 years before wheel rebuild | Limited by animal working life |
| Speed control | Sail reefing + brake wheel | Sluice gate | Drover's pace |
Frequently Asked Questions About Ancient Windmill
A post mill sits lower to the ground than a tower mill of equivalent sail size, so its hub height typically lands in slower, more turbulent air. Wind speed scales roughly with the 1/7 power of height above ground, which sounds small until you cube it for power. Drop your hub from 18 m on a tower mill to 8 m on a post mill and you lose around 25% of available wind power before any other inefficiency.
The second factor is body blockage. The post mill's body sits directly behind the sails and creates a stagnation zone that disturbs the rear half of the sail sweep. Tower mills only have the tower base behind the sails, much further down.
Common sails (cloth on lattice) give you the highest Cp at low wind speeds, around 0.18–0.20, but the miller has to climb out and physically reef them when the wind builds. Andrew Meikle's spring sails (1772) replaced cloth with hinged shutters held closed by a spring — they self-feather in gusts but lose 10-15% peak output to shutter blockage. William Cubitt's patent sails (1807) added a central rod through the windshaft so the miller could adjust shutter tension from inside the cap without stopping.
Rule of thumb: for a demonstration mill where staff can reef on schedule, use common sails and accept the higher attendance. For a working production mill on a windy site, patent sails. Spring sails are rarely the right answer today — they combine the worst of both.
Two causes that aren't sail-related. First, the brake band itself: traditional brake bands are oak or elm faced with a leather or hessian liner. If the liner has glazed (gone hard and shiny from heat), the coefficient of friction drops by half and the band slips on the brake wheel rim no matter how hard you pull the lever. Sand the liner back to fresh fibre.
Second, the brake wheel rim itself. After decades of service the rim wears smooth and slightly conical. A conical rim acts like a wedge that pushes the brake band sideways instead of squeezing it radially. A millwright re-faces the rim by adzing it back to a flat cylindrical surface — a half-day job that restores braking immediately.
Below about 8 m sail span the fantail is overkill. The cap is light enough that one person on a tail pole can yaw it in 30 seconds, and you avoid the gear train of a fantail (worm, spur, and rack on the curb) which adds maintenance points. Most surviving small Greek island mills on Mykonos and Lasithi use tail poles for exactly this reason.
Above 10 m sail span a tail pole becomes a two-person job and dangerous in gusty conditions, so a fantail earns its keep. The threshold isn't sail span exactly — it's cap weight and the leverage available at the tail pole's end. A loaded tower mill cap at 8 tonnes needs a fantail.
Cog hammer under load almost always means the cog pitch is correct on average but the individual cogs are not all the same length end-to-end. Wooden cogs ('cants') are turned individually from hornbeam or apple, and if even two adjacent cogs sit 1-2 mm proud of their neighbours they take the full tooth load while the others coast. Under no load the gears mesh quietly because the lash absorbs the variation; under load the proud cogs hammer and the others ring.
Diagnostic check: rotate the brake wheel slowly by hand with the sails braked and listen for a periodic tick once per revolution of the brake wheel. That tick is your proud cog. Mark it, pull it, and dress it down 1 mm at a time until the noise evens out across the rotation.
Three things eat Cp on real heritage builds. The first is sail twist. Authentic sails twist from 25° pitch at the hub to 18° at the tip — if a restoration carpenter built straight (constant-pitch) sails to save labour, the tips stall and inboard sections under-load, and you lose 30-40% of theoretical Cp.
The second is canvas porosity. New canvas leaks air through the weave until it has been wetted and dried a few cycles to shrink the fibres. Expect Cp to climb 15-20% over the first few months of operation as the cloth seasons.
The third is sail-to-tower clearance. If your sail tips pass closer than about 0.5 m to the tower wall, tower interference creates a once-per-revolution pressure pulse that bleeds shaft torque. Move the windshaft tilt up by another 2-3° to clear the tower better.
References & Further Reading
- Wikipedia contributors. Windmill. Wikipedia
Building or designing a mechanism like this?
Explore the precision-engineered motion control hardware used by mechanical engineers, makers, and product designers.
