Trunk Engine Mechanism: How It Works, Diagram, Parts and Naval Uses Explained

← Back to Engineering Library

A trunk engine is a direct-acting reciprocating steam engine in which the connecting rod passes through a large-diameter hollow tube — the trunk — that is rigidly attached to the piston and slides through a stuffing box in the cylinder cover. It removes the separate piston rod and crosshead so the engine fits into a much shorter footprint, which is what made it the standard powerplant for mid-19th-century gunboats and screw warships that had to keep their engines below the waterline. Builders like John Penn & Sons drove thousands of installed horsepower with it across the Royal Navy fleet from the 1840s to the 1880s.

Trunk Engine Interactive Calculator

Vary cylinder size, trunk diameter, steam pressure, and piston speed to see the unequal working areas and indicated power of a trunk engine.

Back Area
--
Trunk-side Area
--
Indicated Power
--
Area Loss
--

Equation Used

A_back = pi*D^2/4; A_trunk_side = pi*(D^2 - d^2)/4; IP = p*v*(A_back + A_trunk_side)/2

The trunk removes area from one piston face. The back face area is based on the full cylinder bore, while the trunk-side area subtracts the trunk cross-section. For a double-acting engine, indicated power is estimated from mean pressure times the sum of both working areas times half the mean piston speed.

  • Double-acting trunk engine with equal mean effective pressure on both piston faces.
  • Trunk outside diameter is the area removed from the trunk-side piston face.
  • Piston speed is mean piston speed; friction, leakage, and cut-off effects are ignored.
  • If trunk OD exceeds cylinder bore, trunk-side area is limited to zero.
Watch the Trunk Engine in motion
Video: Rotary cylinder 4-stroke engine by Nguyen Duc Thang (thang010146) on YouTube. Used here to complement the diagram below.
Trunk Engine Cross-Section Diagram An animated cutaway diagram showing how a trunk engine works. Trunk Engine cutaway TRUNK (hollow tube) PISTON STUFFING BOX (seal + bearing) CONNECTING ROD CRANK PIN CYLINDER Small-end pin Side thrust Rotation Rod swings ±12° inside trunk as crank rotates. Eliminates crosshead for compact below-deck fit.
Trunk Engine Cross-Section Diagram.

Operating Principle of the Trunk Engine

The trunk engine solves one specific problem: how do you fit a powerful horizontal steam engine into a hull that has almost no headroom and a short fore-and-aft run, while still keeping the entire engine below the waterline so enemy shot can't disable it. A conventional horizontal engine needs a piston rod, a crosshead, slide bars, and only then a connecting rod to the crankshaft. That whole train is long. The trunk engine deletes the piston rod and crosshead by making the piston itself the guide. The connecting rod's small end pins directly inside the piston, and the piston carries a large hollow tube — the trunk — that extends out through the front cylinder cover. As the piston reciprocates, the trunk slides in and out through a stuffing box, and the connecting rod swings inside the trunk's bore.

That geometry is why every dimension matters. The trunk bore must be large enough that the connecting rod can swing through its full angular sweep without the rod fouling the inside wall of the trunk at mid-stroke — typically you want at least 5 mm clearance between the rod shank and the trunk bore at maximum obliquity. The trunk gland packing has to seal a sliding surface that is moving at full piston speed, often 3 to 4 m/s on a Royal Navy gunboat engine, and that gland is exposed to live steam on one face. If the gland packing hardens or the trunk surface scores, you get steam blowing past on every outstroke and you lose mean effective pressure fast. The other failure mode is unequal piston area on the two faces — because the trunk passes through one face only, the working area on the trunk side is smaller than on the back face by exactly the cross-sectional area of the trunk. That asymmetry is baked into the indicated power calculation, and getting it wrong by 10% on the trunk OD throws the power figures off by a similar margin.

The other thing builders had to manage was side-thrust on the trunk. The connecting rod's obliquity puts a transverse force on the small-end pin, and that force is reacted by the trunk sliding in its gland. So the trunk gland is doing double duty as a steam seal and as a guide bearing. Wear on the gland bushes is the single most common service issue on these engines.

Key Components

  • Trunk: A large-diameter hollow steel or wrought-iron tube bolted rigidly to the piston, extending through the front cylinder cover. The trunk's bore must clear the connecting rod swing — typically 5 mm minimum radial clearance at maximum obliquity — and its OD is ground to a 0.05 mm sliding fit in the gland.
  • Piston: Carries the trunk on its outer face and the connecting rod's small-end bearing on its inner face. The piston is a structural member, not just a pressure surface — it transmits all side-thrust from the connecting rod into the trunk and gland.
  • Trunk Gland and Stuffing Box: Seals the sliding trunk against steam leakage and provides the front guide bearing for the piston-trunk assembly. Packed with greased hemp or later metallic packing rings, it must be re-tensioned regularly because it carries side-thrust as well as steam pressure.
  • Connecting Rod: Pinned to the piston inside the trunk bore, swinging through its full obliquity inside the hollow trunk. Length is set so that the rod shank never touches the trunk wall at mid-stroke — obliquity angles of 12 to 15° are typical.
  • Crankshaft and Bearings: Sits directly under or behind the cylinder, no crosshead in between. Because there is no crosshead to absorb side-thrust, the main bearings see only vertical and rotational loads, but the trunk gland sees the full transverse component.
  • Cylinder: Bored straight through with covers on both ends, the front cover carrying the trunk stuffing box. Working area on the trunk side is reduced by the trunk cross-section, so indicated power on the two strokes is unequal by design.

Where the Trunk Engine Is Used

Trunk engines existed for one reason — to drive a screw propeller from a powerplant that fitted entirely below the waterline of a wooden or early ironclad warship. Once the Royal Navy committed to screw propulsion in the 1840s, the trunk engine became the workhorse for gunboats, sloops, and frigates, and John Penn & Sons of Greenwich built the majority of them. They also appeared in a handful of merchant and yacht installations, but the navy was always the dominant customer. By the 1880s, with the spread of triple-expansion engines and taller engine rooms in armoured ships, the trunk engine fell out of favour — but you still find them in preserved and replica vessels today.

  • Naval Steam Propulsion: John Penn & Sons supplied trunk engines to HMS Warrior (1860), the Royal Navy's first iron-hulled, armour-plated warship — a 2-cylinder horizontal trunk engine of 5,267 indicated horsepower preserved today at Portsmouth Historic Dockyard.
  • Royal Navy Gunboats: The Crimean War-era 'Albacore class' wooden screw gunboats used 60-nominal-horsepower Penn trunk engines to keep the entire powerplant below the load waterline.
  • Heritage Ship Preservation: HMS Gannet (1878), preserved at Chatham Historic Dockyard, retains her 2-cylinder horizontal trunk compound engine as a static heritage exhibit.
  • Imperial Russian Navy: Penn-pattern trunk engines were licence-built and installed in Russian Black Sea Fleet screw frigates of the 1850s and 60s.
  • Steam Yachts: Mid-Victorian private steam yachts built on the Clyde used scaled-down trunk engines where owners wanted a flush low deck profile over the engine bay.
  • Merchant Screw Steamers: Early P&O screw mail packets of the 1850s carried Penn trunk engines because they fitted under existing low-headroom main decks without major hull rework.

The Formula Behind the Trunk Engine

The most useful single calculation for a trunk engine is the indicated power, because it tells you what the engine is actually doing on the stroke — and crucially, the trunk-side and back-side strokes produce different power because the working areas are different. At the low end of typical operating speeds, around 30 RPM on a heritage demonstration run, indicated power is dominated by friction at the gland and you are barely above no-load. At nominal sea-going speed, typically 60 to 70 RPM on a mid-Victorian gunboat engine, the engine sits in its sweet spot — gland heat is manageable, mean effective pressure stays close to boiler pressure, and indicated power tracks linearly with RPM. Push beyond the high end, around 90 to 100 RPM, and you start losing MEP because the steam admission valve can't fill the cylinder fast enough and the trunk gland heats up from sliding friction.

IHP = (Pm × L × (Aback + Atrunk) × N) / 33,000

Variables

Symbol Meaning Unit (SI) Unit (Imperial)
IHP Indicated horsepower of the engine summed over both cylinder faces kW (× 0.7457) hp
Pm Mean effective pressure from the indicator card bar psi
L Piston stroke length m ft
Aback Full piston area on the non-trunk side in²
Atrunk Piston area on the trunk side, reduced by trunk cross-section: Aback − (π/4) × Dtrunk2 in²
N Crankshaft rotational speed rev/min RPM

Worked Example: Trunk Engine in a recommissioned Penn trunk engine on a heritage screw gunboat

Confirming indicated power across three operating points on a recommissioned 1862 John Penn & Sons 2-cylinder horizontal trunk engine being returned to demonstration steaming aboard a preserved wooden screw gunboat at the Chatham Historic Dockyard heritage harbour, where the engine drives a 2-bladed bronze screw at 65 RPM nominal, and the trustees want indicated power verified at slow trial running of 30 RPM, nominal cruising of 65 RPM, and a brisk demonstration burst of 90 RPM before the public open day.

Given

  • Cylinder bore = 0.610 m (24 in)
  • Stroke L = 0.610 m (24 in)
  • Trunk OD Dtrunk = 0.305 m (12 in)
  • MEP Pm at nominal load = 1.72 bar (25 psi)
  • Number of cylinders = 2 —

Solution

Step 1 — work out the two piston working areas. The back face sees full piston area; the trunk face is reduced by the trunk cross-section.

Aback = (π/4) × 0.6102 = 0.2922 m²
Atrunk = 0.2922 − (π/4) × 0.3052 = 0.2922 − 0.0731 = 0.2191 m²

Step 2 — compute indicated power per cylinder at nominal 65 RPM with MEP of 1.72 bar (172,000 Pa). Power = MEP × stroke × (Aback + Atrunk) × strokes-per-second.

Pcyl,nom = 172,000 × 0.610 × (0.2922 + 0.2191) × (65/60) = 58,100 W ≈ 58 kW
IHPnom = 2 × 58 kW = 116 kW ≈ 156 hp

Step 3 — at the low end, 30 RPM with MEP dropping to about 1.4 bar because the throttle is barely cracked, indicated power scales roughly linearly with RPM and slightly with MEP.

IHPlow ≈ 116 × (30/65) × (1.4/1.72) = 43.6 kW ≈ 58 hp

That is creep speed — the screw barely produces useful thrust, the engine note is a slow chuff, and the trustees would use this for engine-room photography passes. At the high end, 90 RPM, the steam chest can't refill fast enough and MEP falls to around 1.5 bar from valve throttling losses.

IHPhigh ≈ 116 × (90/65) × (1.5/1.72) = 140 kW ≈ 188 hp

That's the showpiece burst, but the trunk gland will be running noticeably hotter — expect surface temperatures climbing past 90°C at the gland brass and audible hissing if packing is even slightly loose.

Result

At nominal 65 RPM the engine produces 116 kW indicated, around 156 horsepower — entirely consistent with original 1862 trial figures for a Penn trunk engine of this size. At 30 RPM you get 58 hp and a barely-turning screw; at 90 RPM you push to 188 hp but pay for it in gland temperature and steam chest throttling, and the sweet spot for sustained running stays at the nominal 65 RPM mark. If the indicator card shows MEP 15% below the predicted 1.72 bar at nominal, check three things in order: trunk gland packing blowing past the trunk OD on the outstroke (you'll hear it as a soft puff in time with the stroke), worn or scored trunk surface dropping the seal contact area, or piston ring leakage between the two cylinder faces equalising pressure across the piston. Each of those drops MEP without changing geometry, and they're the dominant real-world causes of underperformance on a recommissioned trunk engine.

Choosing the Trunk Engine: Pros and Cons

The trunk engine was never the most efficient or the most refined steam engine — it was the most compact horizontal engine of its era, and that was the entire point. Compare it against the two engines that bracket it historically: the return connecting rod engine, which solved the same low-headroom problem differently, and the later horizontal triple-expansion engine, which displaced the trunk engine entirely once headroom became available again.

Property Trunk Engine Return Connecting Rod Engine Horizontal Triple-Expansion
Typical operating speed 50-90 RPM 50-90 RPM 80-150 RPM
Engine room headroom required Very low — fits below waterline Very low — fits below waterline Moderate to tall
Fore-and-aft footprint Short — no crosshead Long — rod returns past crank Long — three cylinders inline
Steam economy (lb/IHP·hr) 18-25 18-25 1.5-2.0
Gland maintenance interval Frequent — sliding trunk seal Standard piston-rod gland Standard piston-rod gland
Side-thrust handling Through trunk gland Through dedicated crosshead Through dedicated crosshead
Typical service period 1840s–1880s 1840s–1870s 1880s–1950s
Best application fit Mid-Victorian screw warships Paddle steamers, early screw vessels Late merchant and naval steamers

Frequently Asked Questions About Trunk Engine

The trunk passes through one face of the piston only, so the working area on that side is reduced by the trunk's cross-sectional area. On a 24-inch cylinder with a 12-inch trunk you lose 25% of the working area on every trunk-side stroke. The MEP on the indicator card is the same on both faces if the valve is set right, but the force × stroke product is smaller, so indicated power per stroke is lower.

You correct for it by computing IHP separately for each face — Aback for the back stroke, Atrunk for the trunk-side stroke — and summing. If you naively use full piston area for both strokes you'll over-report power by 10-15% on a typical Penn engine.

The trunk gland carries the side-thrust from the connecting rod's obliquity as well as sealing the steam, and that side-thrust is the dominant heat source — not packing friction from the seal alone. At 65 RPM with 12° rod obliquity, the transverse load on the gland is roughly tan(12°) × piston force, which on a 24-inch cylinder at 25 psi works out to about 1,500 lbf reacted on a sliding surface moving at 3 m/s.

If the gland is discolouring, check the trunk's parallel alignment with the cylinder bore first. Misalignment of even 0.5 mm over the trunk length concentrates the side-thrust on one quadrant of the gland bush and you get localised heat marks. Re-tensioning packing won't fix it — you'll just run hotter.

The rod swings through its full obliquity at mid-stroke, which is when the small end is at maximum lateral offset inside the trunk. Compute the offset as Lrod × sin(θmax) where θmax is the maximum obliquity angle, typically 12-15° on a trunk engine. For a 1.5 m connecting rod at 14°, lateral offset is about 360 mm.

Then add half the rod shank diameter and a minimum 5 mm working clearance. If the trunk inside diameter doesn't accommodate that with margin, the rod kisses the trunk wall at mid-stroke and you get a characteristic knock that sounds like a worn big-end but doesn't go away when you re-shim the bearing. Don't tighten the bore — you have to either lengthen the rod or open the trunk.

Both engine types solved the same low-headroom problem in mid-Victorian warships, but they suit different replica scenarios. Pick a trunk engine if your reference vessel is a Penn-engined screw warship from roughly 1845-1880 and you want historical fidelity to a specific named ship like HMS Warrior or HMS Gannet — the trunk engine is the recognisable, photographable artefact in those engine rooms.

Pick a return connecting rod engine if you have more fore-and-aft length to spare and you want easier maintenance, because the RCR engine uses a conventional crosshead and piston rod gland that's far less demanding to service than a sliding trunk gland. Trunk engines are demanding on packing skill, and that's a real factor for a small heritage trust running on volunteer engineers.

It's almost always asymmetric trunk gland wear. Because side-thrust direction reverses with rotation direction (ahead vs astern), the trunk slides against opposite gland quadrants on opposite stroke directions. If the engine spent its working life running predominantly ahead, the gland bush develops an oval wear pattern with the worn quadrant at the bottom-thrust position for ahead running.

When you run astern, the trunk now bears against the unworn quadrant and the alignment shifts slightly — you get more vibration and a different gland heat signature. The fix is to remachine the gland bush concentric or fit a new bush, not to adjust the packing. A quick diagnostic: run ahead for 5 minutes, measure gland temperature at 12, 3, 6, and 9 o'clock positions; uneven readings of more than 10°C between opposite positions confirm the wear pattern.

That asymmetric drop is a classic signature of trunk gland steam loss. On the back stroke, steam acts on the full piston face and is sealed by the back cylinder cover, which is a static joint. On the trunk-side stroke, steam acts on the reduced area and the seal is the sliding trunk gland — and if the gland is leaking, pressure escapes past the trunk OD as the piston approaches the end of stroke and pressure peaks.

Confirm it by holding the engine on the centre and pressurising the steam chest manually — listen at the trunk gland for hiss. Don't chase it as a valve timing problem; valve events affect both strokes equally, and a one-sided pressure loss on the indicator card is almost always a trunk gland issue.

References & Further Reading

  • Wikipedia contributors. Trunk engine. Wikipedia

Building or designing a mechanism like this?

Explore the precision-engineered motion control hardware used by mechanical engineers, makers, and product designers.

← Back to Mechanisms Index
Share This Article
Tags: