Hackworth valve gear is a radial steam-engine valve gear that derives valve motion from a single eccentric and a fixed inclined slotted link, with a die block sliding in the slot to set cut-off and direction. Unlike Stephenson link motion, it needs no expansion link or pair of eccentrics — one eccentric does the job. The purpose is fewer moving parts, lighter reciprocating mass, and simpler reversing. John Wesley Hackworth patented the design in 1859, and it has driven everything from marine engines to small industrial locomotives where simplicity beats theoretical valve-event perfection.
Hackworth Valve Gear Interactive Calculator
Vary eccentric throw, valve lap, slot angle, and crank angle to see valve displacement, port opening, peak travel, and estimated cut-off.
Equation Used
The equation estimates signed valve displacement from mid-position using eccentric throw e, crank angle theta, and slot angle alpha. The cut-off estimate solves the same motion curve for the point where displacement falls back to the valve lap after dead centre.
FIRGELLI Automations - Interactive Mechanism Calculators.
- Angles entered in degrees and converted to radians.
- Ideal rigid linkage with no die-block clearance or rod deflection.
- Cut-off is estimated from the falling crossing where valve displacement equals lap after dead centre.
- Negative slot angle indicates reverse gear with the same travel calculation sign reversed.
How the Hackworth Valve Gear Works
The gear takes its drive from one eccentric on the crankshaft, set 90° from the crank — pure quarter-throw, no offset for lap or lead. The eccentric rod runs out to a die block that slides inside a straight slotted link. That link is not horizontal — it tilts at an angle you set with the reversing lever. The valve rod picks up motion from the die block via a short connection, so the valve gets a vertical component proportional to the slot angle and a horizontal component coming straight from the eccentric throw.
Tilt the slot one way, the engine runs forward. Tilt it the other way, it runs in reverse. Hold it vertical and the valve barely moves — that's mid-gear, near zero cut-off. The geometry is what gives you variable cut-off without a second eccentric. Lap and lead come from the connection geometry rather than from eccentric offset, which is the trick that lets a single eccentric do the work of two.
If the slot pivot drifts out of position by even 2-3 mm on a small engine, you get unequal cut-off between forward and reverse strokes — the locomotive will pull harder one way than the other, and you'll hear it in the exhaust beat. Worn die blocks are the classic failure mode. A die block running with 0.5 mm of slop in a 25 mm slot will cause the valve to overtravel at full gear and skid the slide valve face, scoring the port bars within a few hundred hours of running. Bent eccentric rods, loose slot-pivot bolts, and worn valve-rod pin bushings show up the same way: uneven exhaust, late steam admission, and lost power at notch-up.
Key Components
- Eccentric and eccentric rod: A single eccentric keyed to the crankshaft at 90° from the crank pin transmits a sinusoidal motion to the eccentric rod. Throw is typically 1.0 to 1.5 times the desired valve travel — for a 4 inch valve travel you want a 2 inch eccentric throw on the radius, machined to within ±0.05 mm to keep cut-off symmetric.
- Slotted link (slide bar): A straight machined slot, usually 150-300 mm long, mounted on a pivot at one end. The angle of this slot from vertical is what controls cut-off and direction. The slot must be ground parallel within 0.02 mm over its length, otherwise the die block binds at one end of travel.
- Die block: A hardened steel block running in the slot, picking up motion from the eccentric rod. The fit must be 0.05-0.10 mm clearance — tighter and it galls, looser and you get cut-off scatter between forward and reverse.
- Reversing lever and slot-pivot link: The reversing lever rotates the slot through its pivot, changing slot angle from roughly +30° (full forward) through 0° (mid-gear) to -30° (full reverse). Notches on the reversing quadrant set discrete cut-off positions, typically 75%, 50%, 25% in either direction.
- Valve rod and combination lever: Connects die block motion to the slide valve or piston valve. On Hackworth the valve rod usually picks up directly from the die block end — there's no separate combination lever like Walschaerts uses for lead steam, which is why Hackworth lead is fixed by the slot geometry rather than independently adjustable.
- Slide or piston valve: The end-effector. Lap is typically 20-25% of port width, lead 1-2 mm at full gear. The valve must seat on its face with no more than 0.1 mm wear before steam leakage starts costing measurable power.
Industries That Rely on the Hackworth Valve Gear
Hackworth gear shows up wherever a designer wanted radial valve gear without the cost or maintenance burden of Walschaerts or Stephenson. It was never the dominant choice on mainline locomotives, but it earned its keep on small industrial engines, marine work, and specialised builds where the single-eccentric simplicity paid off. You'll still find it running today on heritage installations and on derivatives like Marshall and Klug gears, which solved Hackworth's geometric weaknesses while keeping the basic single-eccentric layout.
- Heritage railways: The Bowes Railway in County Durham preserved several Hackworth-pattern industrial locomotives originally built for colliery service, where the simple reversing gear suited shunting work.
- Marine steam engines: Mid-19th-century Royal Navy gunboats fitted Hackworth gear on auxiliary engines because a single eccentric was easier to fit in tight engine-room spaces than Stephenson link motion.
- Industrial shunters: Manning Wardle and similar builders used Hackworth-derivative Marshall gear on narrow-gauge industrial 0-4-0 saddle tanks where simplicity of overhaul mattered more than top-end efficiency.
- Stationary mill engines: Some late-Victorian rolling-mill engines used Hackworth gear to drive auxiliary slide valves on reversing duty, where the engine had to change direction every 30 seconds without operator fatigue from heavy reversing levers.
- Steam launches and tugs: Small Thames steam launches built between 1870 and 1900 occasionally fitted Hackworth gear on single-cylinder engines, particularly where the builder wanted easy reversing for tight river work without doubling up eccentrics.
- Educational and museum demonstrations: The Science Museum, London holds working Hackworth-gear models used to demonstrate radial valve gear principles to engineering students.
The Formula Behind the Hackworth Valve Gear
The useful equation predicts valve travel as a function of slot angle and eccentric throw. At low slot angle (near mid-gear) the valve barely moves and the engine runs efficiently at short cut-off but produces little power. At high slot angle (full gear) the valve gets near-full eccentric throw and you get long cut-off, high power, low efficiency — what you want for starting heavy loads. The sweet spot for cruising is around 40-60% of full slot angle, which gives roughly 25-35% cut-off and good steam economy.
Variables
| Symbol | Meaning | Unit (SI) | Unit (Imperial) |
|---|---|---|---|
| xv | Valve displacement from mid-position | mm | in |
| e | Eccentric throw (half the eccentric stroke) | mm | in |
| θ | Crank angle from dead centre | rad or ° | rad or ° |
| α | Slot angle from vertical (set by reversing lever) | rad or ° | rad or ° |
Worked Example: Hackworth Valve Gear in a 1885 Black Hawthorn industrial 0-4-0 saddle tank
You are setting the reversing-lever notches on the Hackworth valve gear of a recommissioned 1885 Black Hawthorn 0-4-0 saddle tank being returned to demonstration running at a heritage ironworks museum in Shropshire. Eccentric throw is 50 mm, valve lap is 20 mm, and you need to predict valve travel and approximate cut-off at three notch positions to mark the reversing quadrant correctly.
Given
- e = 50 mm
- lap = 20 mm
- αfull = 30 °
- αmid-notch = 20 °
- αshort = 10 °
Solution
Step 1 — at full gear (α = 30°), peak valve displacement occurs near θ = 90°. Compute peak travel:
Total valve travel from extreme to extreme is 2 × 25 = 50 mm. Subtract twice the lap to get port opening: 50 − 40 = 10 mm port opening at full gear. Cut-off at this notch works out to roughly 75% of stroke — long admission, big power, heavy steam consumption. This is your starting notch for a loaded train out of the platform.
Step 2 — at the mid notch (α = 20°), repeat:
Total travel 34.2 mm. Port opening = 34.2 − 40 = negative — the valve no longer fully clears the lap. In practice cut-off has dropped to around 35-40% of stroke, port opening is marginal at 0-2 mm, and the engine is running on expansion. This is your cruising notch — efficient, quiet exhaust beat, comfortable for a 15 minute run round the demonstration loop.
Step 3 — at the short notch (α = 10°):
Total travel 17.4 mm — well below the 40 mm needed to clear lap on both sides. The valve never fully opens the steam port. Cut-off is below 15%, the engine is essentially coasting on residual steam. Useful for descending a gradient with the regulator cracked open, useless for pulling load. Mark this notch but do not expect the locomotive to climb anything in it.
Result
At full gear with α = 30°, peak valve travel is 25 mm and port opening is 10 mm — that's 75% cut-off and the loud, sharp four-beat exhaust you hear when a tank engine drags a heavy train out of a station. At the 20° mid notch the engine settles to ~35% cut-off with a softer beat, comfortable cruising. At the 10° short notch the valve barely clears lap, cut-off is below 15%, and the locomotive coasts. If you set the gear up and find forward and reverse cut-off differ by more than 5% at the same notch, suspect the slot pivot is offset from the eccentric centreline — a 2 mm pivot offset gives roughly 4-6% cut-off asymmetry. If port opening measures 30%+ below predicted at full gear, check the eccentric rod for bending and the die block for clearance above 0.15 mm. If the valve drifts off-square on the seat, the valve-rod fork pin is worn and needs reaming.
When to Use a Hackworth Valve Gear and When Not To
Hackworth competes against Stephenson link motion and Walschaerts gear. Each one trades simplicity against valve-event accuracy and lead behaviour, and the right choice depends on what the engine is doing for a living.
| Property | Hackworth | Stephenson link motion | Walschaerts |
|---|---|---|---|
| Number of eccentrics required | 1 | 2 | 1 (or return crank) |
| Lead steam behaviour | Variable with cut-off (decreases at short cut-off) | Variable with cut-off | Constant lead at all cut-offs |
| Cut-off range typical | 10% to 80% | 15% to 75% | 5% to 85% |
| Reciprocating mass at gear | Low — 1 eccentric, 1 link | High — 2 eccentrics, expansion link | Medium — return crank, combination lever |
| Maintenance interval (heritage service) | ~1500 hours | ~1000 hours | ~2000 hours |
| Cost to manufacture (1860s baseline) | Low | High | Medium-high |
| Valve event accuracy | Moderate — geometric errors at full gear | Good | Excellent |
| Best application fit | Small industrial, marine auxiliary | Mid-Victorian mainline locomotives | 20th-century mainline locomotives |
Frequently Asked Questions About Hackworth Valve Gear
Asymmetric cut-off almost always traces to the slot pivot not lying on the line through the eccentric centre and the valve-rod end. The Hackworth geometry assumes that pivot point is on the centreline — if it's offset by even 2 mm vertically or horizontally, the slot tilts asymmetrically about its working position and gives you longer cut-off one way than the other.
Diagnostic check: set the lever to mid-gear, bar the engine over by hand, and measure valve displacement at θ = 90° and θ = 270°. They should be equal within 0.5 mm. If they're not, shim the slot pivot bracket until they match before you worry about anything else.
Hackworth lead is geometric — it falls out of the slot angle and connection geometry rather than being set independently like Walschaerts. As you reduce slot angle toward mid-gear, the horizontal component of valve motion drops faster than the vertical component, and lead shrinks with it. By 20% cut-off you may have effectively zero lead.
Whether it's a problem depends on running speed. At low industrial-shunter speeds (10-15 mph) it doesn't matter much. At 40+ mph it causes late admission, soft exhaust, and reduced power at speed — which is exactly why Marshall and Klug gears were developed to fix Hackworth's lead behaviour while keeping the single-eccentric layout.
Marshall is the better choice for almost any new build. It's a direct descendant of Hackworth that swaps the straight slotted link for a curved one, which corrects the geometric error that makes Hackworth lead fall away at short cut-off. You keep the single-eccentric simplicity but get cleaner valve events across the cut-off range.
Pick true Hackworth only if you're replicating a specific historic locomotive for heritage authenticity — for example, restoring an 1870s industrial saddle tank where Marshall would be anachronistic. For any working duty, Marshall or Klug.
Three things cause overtravel at full gear, in order of frequency. First, slot angle has been set beyond design — check the reversing-quadrant stops and confirm full-gear angle matches the original drawings (typically 25-30°, never more than 35°). Second, eccentric throw has been incorrectly machined or the eccentric has shifted on its key — measure throw against the design value with a dial gauge.
Third, and the one people miss: the valve-rod connection length has been adjusted to chase a steam leak somewhere else, and now the valve is geometrically biased toward one end of its travel. Fix the upstream problem (worn lap, bad seat) and reset rod length to design before assuming the gear itself is wrong.
Start from required peak valve travel. For a slide valve, peak travel = 2 × (lap + max port opening). On a small industrial engine with 25 mm port width and 20 mm lap, you want roughly 2 × (20 + 25) = 90 mm peak travel, so 45 mm half-travel. With Hackworth, peak valve displacement at full gear equals e × sin(αfull). Setting αfull to 30° gives sin(α) = 0.5, so eccentric throw e = 90 mm.
Rule of thumb: e ≈ 2 × peak valve displacement at the chosen full-gear angle. Always check that the resulting throw doesn't put excessive load on the eccentric strap — if the eccentric ends up larger than the crankshaft journal can comfortably carry, increase αfull instead.
15% deficit is a big number and it's almost never the eccentric. Check three things. First, slot wear — a slot that has worn oval at the die-block working zone effectively shortens the geometry. Hold a straight edge along the slot face and look for daylight. Second, eccentric-rod end-pin wear — every millimetre of slop in pin bushings costs you about 2-3% of valve travel.
Third, and the silent killer: the slot pivot bracket has loosened on its mounting and the entire slot has shifted. Grab the slot at its free end and try to move it perpendicular to its working plane. Any movement at all means the bracket bolts need pulling, the bracket faced off, and refitted with fresh dowels.
References & Further Reading
- Wikipedia contributors. Hackworth valve gear. Wikipedia
Building or designing a mechanism like this?
Explore the precision-engineered motion control hardware used by mechanical engineers, makers, and product designers.
