A ship's steering gear is the mechanical and hydraulic system that turns the rudder against the water load to steer the vessel. Its core component is the rudder stock — a heavy forged shaft that transmits torque from the actuator down through the hull to the rudder blade. The system exists because no helmsman can move a 60-tonne rudder at 18 knots by hand. On a typical Panamax bulker the gear swings the rudder from 35° port to 30° starboard in 28 seconds, the limit set by SOLAS Chapter II-1.
Ship's Steering Gear Interactive Calculator
Vary the Rapson slide rudder angles and see the sec^2(theta) torque amplification and animated two-ram steering geometry.
Equation Used
The Rapson slide increases tiller torque by the factor sec^2(theta). The defaults mirror the article diagram: 0 deg gives 1.00x, 20 deg gives about 1.13x, and 35 deg gives about 1.49x, or roughly 49% more torque at hard-over.
- Mechanical advantage is normalized to torque at theta = 0 deg.
- Ram force F and tiller radius R are held constant.
- Angles are absolute rudder angles in degrees.
- Hydraulic losses, stock friction, and rudder water-load variation are not included.
Inside the Ship's Steering Gear
Pressure in the cylinders does the work. On most ships above 5,000 GT you'll find an electro-hydraulic ram-type steering gear: two or four hydraulic rams push a crosshead that bears against a tiller arm or Rapson slide keyed to the top of the rudder stock. A variable-displacement axial-piston pump (Hele-Shaw or Hatlapa style) sends oil to whichever pair of rams needs pressure, and the rudder turns. The Rapson slide geometry matters — torque on the stock is F × R × sec²θ, so as the rudder approaches hard-over the mechanical advantage actually increases, which is exactly when you need it because rudder torque peaks somewhere around 35°.
Why build it this way? Because rudder torque is huge and bidirectional. A 9,000 TEU container ship sees peak rudder stock torques in the 1,500–2,500 kNm range. You cannot do that with a screw or a chain — you need oil under 100–140 bar pressure feeding rams with bores of 250–400 mm. If your relief valves are set wrong, two things go bad fast. Set too low and the gear stalls before reaching hard-over in a following sea. Set too high and you overstress the rudder stock, and forged stock failure is the failure mode that has sunk ships — the loss of steering on the Maersk Honam fire response in 2018 is a reminder of why redundancy is mandatory.
SOLAS requires two independent power units and the ability to put the rudder from 35° one side to 30° the other in 28 seconds at maximum service speed and deepest draft. If you measure a slower swing during sea trials, the usual culprits are pump swash-plate wear, internal ram seal bypass, or air entrained in the hydraulic circuit from a low header tank. Follow-up control — where the helm angle commands a matching rudder angle through a feedback linkage or rotary encoder — is what makes the system actually steerable. Without follow-up you're running non-follow-up (NFU) tiller control, which is the emergency backup mode every Mate has to demonstrate.
Key Components
- Rudder Stock: The forged steel shaft (typically 42CrMo4 or equivalent) that connects the rudder blade to the steering gear above the waterline. Diameters run 300–900 mm depending on ship size, sized to ABS or DNV rules with a torsional safety factor of around 1.7 against yield.
- Hydraulic Rams: Two or four double-acting cylinders that push the crosshead. Bores of 200–400 mm at 100–140 bar working pressure deliver the linear force the tiller converts to rudder torque. Seal stack is usually a Hallite or Merkel U-cup with bronze guide bands.
- Tiller / Rapson Slide: The lever arm keyed to the rudder stock that converts ram thrust into stock torque. Rapson slide design gives sec²θ torque amplification at large rudder angles — at 35° hard-over you get roughly 49% more torque than at amidships for the same ram force.
- Variable-Displacement Pump: Axial-piston pump (Hele-Shaw, Hatlapa, or Rolls-Royce Kamewa) driven continuously at 1,500 RPM but delivering zero flow at neutral. Tilting the swash plate sends flow to whichever ram pair is commanded. Response time from neutral to full stroke is typically under 0.5 seconds.
- Telemotor / Follow-Up Control: Electrical or hydraulic feedback link between the bridge helm and the steering gear pump. Modern ships use a synchro or rotary encoder feeding a PLC that drives the swash-plate solenoid. Position deadband must be below 1° to meet helm-feel requirements.
- Relief and Shock Valves: Pressure-limiting valves on each ram circuit set typically at 1.25× working pressure. Shock valves protect against rudder slamming from following seas — a wave hitting the blade can spike line pressure to 200 bar in milliseconds, and without shock valves you'd crack the ram end caps.
- Header Tank and Replenishment Line: Gravity-fed reservoir that keeps the suction side of the pump flooded and replaces seal leakage. Low-level alarm trips at 50% capacity per SOLAS — running dry causes pump cavitation and total loss of steering within minutes.
Where the Ship's Steering Gear Is Used
Every powered vessel above lifeboat scale has steering gear of some form, but the engineering varies massively with displacement, speed, and class rules. On small craft it's a cable-and-quadrant or a single hydraulic cylinder. On a VLCC it's a four-ram electro-hydraulic monster the size of a delivery van. The choice between ram-type, rotary-vane, and electric-mechanical comes down to torque, redundancy requirements, and machinery-space footprint.
- Commercial Shipping: Maersk Triple-E class container ships use four-ram electro-hydraulic steering gear from MacGregor-Porsgrunn, sized for roughly 2,200 kNm peak stock torque to handle the 165 m² semi-balanced rudder.
- Cruise & Ferry: Royal Caribbean's Oasis-class vessels run twin Rolls-Royce Promas units with rotary-vane steering gear, chosen for compact installation aft and the ability to swing rudders 70° for low-speed manoeuvring in port.
- Naval: Arleigh Burke-class destroyers use redundant electro-hydraulic steering with two completely independent power units, meeting US Navy survivability requirements that one hit cannot disable both systems.
- Tugs & Workboats: Damen ASD Tug 2810 series uses azimuth thruster steering rather than a conventional rudder, but the steering gear principle — hydraulic rotary actuators driven by feedback-controlled pumps — is identical.
- Offshore / DP Vessels: Subsea 7 construction vessels like Seven Borealis run high-redundancy steering integrated with DP2 dynamic positioning, where rudder angle commands come from the DP controller at 5 Hz update rate rather than a human helm.
- Inland & Coastal: Washington State Ferries Issaquah-class double-enders carry two complete steering gears — one at each end — because they reverse direction at every dock without turning the hull.
The Formula Behind the Ship's Steering Gear
The number you actually need to size a steering gear is the peak rudder stock torque the rams must overcome. This drives ram bore, working pressure, and pump displacement. The torque depends on rudder area, ship speed squared, rudder angle, and a non-dimensional torque coefficient that comes from the rudder profile. At low rudder angles (5–10°) torque is small and the gear loafs — you barely hear the pump. Around the design sweet spot of 30–35° you hit peak torque and that's what sizes the system. Push past 35° and you're outside the SOLAS-required envelope and into territory where flow separation on the rudder can actually reduce torque while killing turning effectiveness.
Variables
| Symbol | Meaning | Unit (SI) | Unit (Imperial) |
|---|---|---|---|
| TR | Rudder stock torque | Nm | lbf·ft |
| ρ | Seawater density (≈1025 kg/m³) | kg/m³ | slug/ft³ |
| AR | Rudder blade projected area | m² | ft² |
| V | Ship speed through water at the rudder | m/s | ft/s |
| CN | Normal force coefficient (function of rudder angle and aspect ratio) | dimensionless | dimensionless |
| xcp | Distance from rudder stock axis to centre of pressure | m | ft |
Worked Example: Ship's Steering Gear in a 50,000 DWT product tanker refit
Your shipyard project team in Ulsan is sizing the steering gear for a mid-life refit of a 183 m product tanker. Rudder area is 22 m², service speed is 14.5 knots (7.46 m/s), the rudder is a NACA-section semi-balanced spade with 18% balance ratio, and the chord at the stock is 3.2 m. You need peak stock torque at 35° to spec the rams.
Given
- AR = 22 m²
- V = 7.46 m/s
- ρ = 1025 kg/m³
- CN at 35° = 1.30 dimensionless
- xcp (chord × (0.33 − balance ratio)) = 3.2 × (0.33 − 0.18) = 0.48 m
Solution
Step 1 — at the nominal hard-over angle of 35° and full service speed, compute the normal force on the rudder:
Step 2 — multiply by the lever arm from stock axis to centre of pressure to get nominal stock torque:
Step 3 — at the low end of the operating range, slow steaming at 10 knots (5.14 m/s), torque scales with V². You'd expect roughly:
That's a comfortable load — the rams will see around 70 bar and the pump runs near the bottom of its swash range. The helmsman feels almost no lag. Step 4 — at the high end, design overload of 16 knots (8.23 m/s) plus a 20° drift angle from a beam sea, effective CN can climb to 1.5 and you must size for it:
This is the figure that actually drives the ram bore selection — you spec the gear for 550 kNm continuous and verify shock-valve settings protect against transient spikes that can hit 1.4× this in following seas.
Result
Peak nominal rudder stock torque is approximately 391 kNm at 35° hard-over and 14. 5 knots. That sets the four 280 mm bore rams running at 110 bar working pressure with comfortable margin. At the low-end 10-knot slow-steaming case the gear sees 186 kNm and barely works; at the 16-knot overload-with-drift case it climbs to 550 kNm, which is what really sizes the hardware. If your sea-trial swing time exceeds the SOLAS 28-second hard-over benchmark, three failure modes account for almost every case: pump swash-plate wear letting internal flow bypass the rams, a partially clogged suction strainer starving the pump and dropping volumetric efficiency below 85%, or air ingestion from a header tank running below the 60% level alarm. Any of those will show up as a slow, hesitant rudder rather than a clean swing.
When to Use a Ship's Steering Gear and When Not To
Three steering gear architectures dominate modern ships: ram-type electro-hydraulic, rotary-vane electro-hydraulic, and direct electric drive. The choice is mostly torque, footprint, and class-rule redundancy. Here's how they actually compare on the dimensions that matter when you're specifying a refit.
| Property | Ram-type Electro-Hydraulic | Rotary Vane | Electric-Mechanical (EMA) |
|---|---|---|---|
| Peak torque capacity | Up to 6,000 kNm (4-ram VLCC) | Up to 4,000 kNm | Up to ~250 kNm currently |
| Hard-over swing time (SOLAS limit 28 s) | 20–28 s typical | 18–25 s typical | 15–22 s typical |
| Machinery space footprint | Large — needs ram travel length | Compact — fits over the stock | Smallest |
| Working pressure | 100–140 bar | 70–100 bar | N/A (electric) |
| Maintenance interval (major overhaul) | 5 years / 25,000 hr | 5 years / 25,000 hr | 10+ years (no hydraulic seals) |
| Capital cost (relative) | 1.0× baseline | 1.1–1.3× | 1.5–2.0× |
| Best application fit | Cargo ships, tankers >10,000 GT | Cruise, ferries, naval — space-limited | Small craft, future large-ship development |
| Failure modes | Seal leak, pump wear, oil contamination | Vane seal wear, internal bypass | Motor brush wear, gearbox backlash |
Frequently Asked Questions About Ship's Steering Gear
Because rudder torque scales with V² but the pump output is roughly constant. At the dock with zero forward speed, the only load on the rams is the seal friction and the rudder's own weight — the pump has spare flow and the swing is fast. At full service speed the rams are working against full hydrodynamic torque and the pump may be hitting its full swash-plate displacement.
If you're failing the trial by 2–4 seconds, the usual fix is checking that both pumps are actually loading in parallel during the test. Many gears are spec'd as one-pump-meets-28s, two-pumps-much-faster, but if the second pump's solenoid is sticky or its swash response is sluggish you'll only get the single-pump performance. Class surveyors will specifically call this out.
Pressure-rise rates. A main relief valve is sized for steady-state overload — it opens when sustained pressure exceeds the setpoint. A wave slamming the rudder face creates a pressure spike that rises in 10–30 milliseconds, faster than the main relief poppet can react. By the time the main relief lifts, the spike has already passed through to the ram end caps and rudder stock.
Shock valves are direct-acting, fast-response valves with much smaller moving mass — they crack open in 2–5 ms and dump oil directly between the two ram chambers. Without them, you get cracked end-cap bolts and forged stock fatigue cracks at the keyway. Several rudder stock failures investigated by class societies trace back to shock valves seized open or set incorrectly.
Check rudder balance ratio first. The lever arm xcp is extremely sensitive to where the stock axis sits relative to the rudder's centre of pressure. If as-built drawings show 18% balance but the actual fabricated rudder ended up at 14% (a common shipyard tolerance creep), your xcp jumps from 0.48 m to 0.61 m — that alone is a 27% torque increase.
The other suspect is rudder profile. A NACA section will give CN ≈ 1.3 at 35°, but a flat-plate or poorly-faired rudder can hit 1.5–1.6 because of post-stall behaviour. Have a diver photograph the leading edge — yard repairs that fill in the leading-edge radius are surprisingly common and they wreck the CN curve.
Two conditions push you toward rotary vane: limited steering flat headroom and the need for rudder angles beyond 45°. Cruise ships and ferries with high-lift rudders or Promas integrated systems often need 70° rudder angles for harbour manoeuvring, and a ram-type tiller geometrically cannot reach that — the Rapson slide goes singular past about 40°.
Below 2,000 kNm peak torque rotary vane is competitive on cost and wins on footprint. Above that, ram-type tends to win because vane seals at high pressure and large diameters become a maintenance headache. For a tanker or bulker where you have plenty of steering flat space and never need more than 35°, ram-type is almost always the right answer.
You almost certainly have air entrained in the system that hasn't bled out. Hunting at small amplitude after an oil change is the signature symptom — tiny compressible bubbles in the ram lines act like soft springs in the feedback loop and the controller overshoots, corrects, overshoots again. It's not the controller PID, it's the fluid stiffness.
Run the rudder slowly hard-over to hard-over five or six times with the header tank vent open. The bubbles migrate up to the highest points in the ram circuit — usually the bleed screws on the ram end caps and the top of the four-way control block. Crack each bleed in turn until clear oil comes out. If the hunt persists after a full bleed, look at the feedback transducer mounting — a loose synchro coupling on the rudder stock will cause the same symptom.
Because the formula gives you operating torque and class rules give you design torque. ABS, DNV, and Lloyd's all apply factors to account for transient peaks (following seas, ice impact, manoeuvring overshoot) and they specify a minimum design torque that's typically 1.4–1.7× the steady-state hydrodynamic value at hard-over.
Use the formula to understand the physics and to size your rams for actual operating performance. Use the class rule for rudder stock diameter and the SOLAS-mandated time-to-hard-over verification. The two are answering different questions — one is 'what does the gear feel in normal service' and the other is 'what survives the worst credible event'.
No, and class will not let you. SOLAS Chapter II-1 Regulation 29 specifies the 28-second hard-over performance must be met at maximum ahead service speed and deepest seagoing draft — not at the operator's intended cruising speed. Slow steaming is a commercial decision; the regulatory case assumes the ship can and will run at full design speed.
The other reason is emergency manoeuvring. Even on a slow-steaming voyage, the moment a collision avoidance situation develops the bridge will call for full speed and full helm simultaneously. A steering gear sized for 10 knots will stall when it's most needed. The honest answer is the gear gets sized for design speed plus a margin, full stop.
References & Further Reading
- Wikipedia contributors. Steering gear. Wikipedia
Building or designing a mechanism like this?
Explore the precision-engineered motion control hardware used by mechanical engineers, makers, and product designers.
