Twin Rudders on the Melges 40: Why Two Blades

A wide, powered-up planing hull carries two rudders because a single centreline blade rotates toward horizontal and ventilates as the boat heels — it loses grip at the exact moment the boat is fastest and hardest to steer. Canting one rudder outboard on each side keeps whichever is to leeward standing near-vertical and fully buried through a wide band of heel. That immersed leeward blade does the steering, in cleaner flow than a centreline rudder sees at the same helm angle, which is what lets a crew actually use the power the keel and sails create instead of being frightened of it.

The problem: geometry beats a single blade

Start with the hull, because the rudder problem is really a hull-shape problem. A modern Grand Prix hull is wide and carries that beam a long way aft to a broad, near-flat transom. That planform gives early planing and buoyant, stable aft sections when the boat is pressed — but it also has almost no buoyancy in the fine bow relative to the fat quarter. So as the boat heels, it does not simply lean about its centreline; it rolls its stern out. The immersed shape becomes asymmetric, the leeward quarter drives down, the narrow bow buries, and the centreline of the transom climbs toward the surface. A rudder hung on that centreline goes with it, tilting from vertical toward horizontal as heel builds.

Two failure mechanisms follow, and they compound. First, the blade's effective immersed span collapses. Rudder side-force scales with immersed planform area and, for a low-aspect blade, roughly with the square of its effective span, so as the tip approaches the surface the foil loses lift faster than it loses wetted length — and because it is now tilted, a growing share of its force points up, lifting the stern rather than pushing it sideways. It behaves less like a pure side-force generator and more like a partly lifting foil, so it makes less grip per degree of helm just when you are asking for more.

Second, and the one that actually bites, is ventilation. This is not cavitation and the distinction matters. Cavitation needs local pressure on the suction side to fall to the vapour pressure of water, so the water boils — it is a high-speed, low-cavitation-number problem you rarely reach on a keelboat rudder. Ventilation is far easier to trigger: once the low-pressure (suction) face of the blade has a path to the free surface — because the tip is near it, or because a spinning vortex or the wake connects the two — ambient air at atmospheric pressure is simply sucked down that face. Air replaces water over part of the foil, the pressure difference that made the lift disappears, the flow detaches, and the blade stalls. It is Froude-driven and heel-driven, and it arrives with almost no warning. On a fast, heeled, powered-up boat this is the mechanism behind the sudden uncommanded round-up: the helmsman feels the boat load up, pulls more helm to fight it, drives the already-marginal blade past its critical angle of attack, deepens the ventilated stall — and now has no rudder at all. The bow rounds up, the boat lies over, and only easing sheets and unloading the blade recovers it. That is the exact failure mode twin rudders are designed to remove.

The fix: cant a blade outboard each side

Twin rudders sit either side of centreline with their stocks leaning outboard, so the blades splay apart below the hull. The whole trick is what heel does to that splay. As the boat heels, the leeward stock rotates toward vertical and the blade drops deeper, presenting nearly its full span to solid, undisturbed water — it becomes the working blade. The windward stock rotates the other way, lifting its blade so that much of it skims at or above the surface, out of the load path and largely out of the way. The system self-selects: at any heel the boat sails, there is always a foil standing up straight and buried on the side where the steering load wants to go.

Cleaner flow is the second, less obvious win. A centreline rudder steers in "second-hand" water — flow already disturbed by the keel fin, bulb and canard directly ahead of it, and by the hull's own boundary layer running aft down the centreline. The leeward twin sits well outboard of that wake and bites into flow that has not been churned, so it keeps a higher effective lift-curve slope and a later stall. Practically, that shows up as steering that stays progressive and connected across the heel range rather than going vague and then letting go.

The geometry is set by two angles the designer chooses to match the hull's real sailing attitudes. Cant angle — how far outboard the stocks lean — sets how quickly the leeward blade reaches vertical as heel builds, and is tuned so that "vertical and fully buried" coincides with the heel angles the boat is actually sailed at upwind and reaching. Get it right and the crossover — the heel at which the leeward blade takes over cleanly — sits inside the normal sailing band; get it wrong and there is a slot of heel where neither blade is ideal. Toe is the horizontal alignment, and because the hull crabs to leeway, a small toe-in (leading edges nearer centreline) aligns each blade closer to the true local flow, stops the two foils from fighting, and damps the wander and flutter you get if they run dead parallel. The right figure is small — a fraction of a degree up to around a degree depending on design — and it is a drawing-office number, not a guess: too much toe just crabs both foils and scrubs speed. Both blades then connect through a single tie-bar to one tiller or wheel, so despite all this geometry the helm still steers one boat, with one feel, not two.

Balance, section and the numbers behind the feel

Under the fairing, each blade is an engineered foil, not a plank. The section is typically a symmetric NACA form in roughly the 9–15% thickness-to-chord range. That band is a deliberate compromise: a thinner section gives lower drag and a sharper entry but stalls earlier and offers less room for structure; a thicker section carries a higher stall angle and more internal depth for the stock — genuinely useful on a boat that must take big, sudden helm inputs down a wave — at the cost of a little more form drag. The blade is usually low-to-moderate aspect ratio, which trades some efficiency for a higher stall angle and more forgiving behaviour near the limit, exactly what you want in a steering foil that occasionally gets loaded hard and fast.

Balance sets the steering effort and the feel. Move the stock aft so that some area sits forward of the pivot and the water's pressure on that forward area helps turn the blade, cutting the load the helmsman fights. Race rudders typically run a modest balance — on the order of low tens of a percent of area ahead of the axis — deliberately under-balanced rather than neutral, because a fully balanced blade goes numb and a slightly loaded helm is what tells the driver, through the wheel, when a blade is about to let go. The same logic drives rig and hull trim toward a small steady weather helm (a few degrees of rudder), which keeps a whisker of lift working on the blade and preserves feel; run zero or lee helm and the first sign of trouble is the boat simply not answering. On a wide hull the useful cue is precisely this: the leeward blade only develops proper, progressive feel once it is standing vertical, so a helm that suddenly goes light is the leeward blade starting to ventilate, and a hum or shudder through the wheel is flow separating or a toe/alignment fault — both are "unload it now" signals, not "pull harder".

Stock, bearings and structure: the load path

The steering loads on a powered-up 12-metre boat are large and, in a broach or a hard bear-away, shock-loaded, so the stock and its bearings are a real structural system. The stock is typically pre-preg carbon/epoxy laid over a mandrel and oven-cured, often on a rectangular or blade-following section rather than a plain tube so it resists bending in the loaded plane. The laminate is tapered — thickest wall where the stock passes the lower bearing as it exits the hull, because that is where bending moment and shear peak, and lighter toward the tip and the upper bearing where the loads fall away. The stock is a cantilever: the water pushes on the blade, the two hull bearings react it as a couple, and the lower bearing takes the highest concentrated load.

Because a carbon stock bends under that load rather than staying dead straight, the bearings must accommodate the deflection or they bind and gall. That is why race installations use self-aligning bearings that can articulate — commonly on the order of a few degrees (self-aligning carriers of this type are rated to around ±3.5°) — so the bearing follows the stock's deflected axis and rotation stays smooth even fully loaded and hard over. A bearing that has lost that articulation, or an upper and lower bearing that have crept out of line, shows up as a stiff spot, a notch in the rotation, or accelerated wear. Where fitted, a kick-up or fuse in the system lets a blade rotate up or shear on grounding to protect the hull and the stock — worth confirming it releases and re-sets, and that both sides behave identically.

How this plays out on the Melges 40

The Melges 40 is a Botin Partners-designed, canting-keel, twin-rudder, retractable-bowsprit one-design built by Premier Composite Technologies from epoxy-infused carbon over a foam core, aimed at windward/leewards and coastal racing. Published class figures put it at roughly 11.99 m LOA, about 11.10 m waterline, 3.53 m maximum beam and around 3,250 kg displacement with about 1,200 kg of ballast, driven by a large square-top mainsail (about 72 m²), a jib around 49 m² and a gennaker in the region of 200 m² — an unusually high sail-area-to-displacement number that makes it "heavily powered" in the precise sense that matters here: it generates far more heeling and driving force per kilo than a heavier boat, so the rudder problem above is acute rather than academic. Treat every one of those figures as needing verification against the current class rules and the individual boat's own documentation before you rely on it — published sources vary and the class controls the real numbers.

The canting keel cants up to 45 degrees to weather for righting moment, and that is precisely why the boat also carries a centreline daggerboard forward of the keel — a canard. Once the keel swings off centreline it supplies moment but little lateral resistance, so the board takes over the side-force job upwind and retracts downwind to shed drag. Splitting ballast from lateral resistance this way is the hallmark of canting-keel design and is covered further in canting keel explained and the Melges 40 systems guide. Rudder cant, toe, blade section and dimensions are boat- and design-specific — take them from the drawings, not from another hull.

Sailing, setup and what good looks like

The behaviour follows directly from the geometry. Because the leeward blade only works properly standing vertical, sailing the boat at roughly its designed heel keeps the steering light, progressive and connected; sail it too upright and you throw the crossover away — the blade is still canted and never gets vertical — while over-heeling loads the leeward blade, drags the windward one, and pushes both toward their limits. Good feel is a wheel that stays alive through gusts and down waves with the boat tracking rather than skating, answering small inputs without needing a handful of helm. The warning set is specific and worth drilling into the afterguard: a helm that goes suddenly light or soft (leeward blade ventilating), a shudder or hum through the wheel (separation or toe/alignment), or the boat wanting to round up despite eased sheets (the leeward blade at or past stall) — every one of which says depower and unload, not pull. The concepts in downwind mode basics and heavy air mode all quietly assume a rudder that keeps gripping when the boat is lit up; on this hull that assumes two of them, and common speed killers is the companion on keeping the boat in the groove that makes them work.

Maintenance and inspection

Twin rudders are a safety system and a performance system, so they earn scheduled checks, not just a glance. Work the bearings first: rotate each blade lock to lock and feel for radial play, notchiness or a stiff spot, and confirm any self-aligning bearing still articulates so it can track stock bend under load rather than binding. Check the tie-bar, tiller arms, quadrants and linkage for security, correct toe alignment and equal load-sharing — misalignment shows as drag, uneven bearing wear and a boat that steers noticeably differently on each tack. Examine the carbon stock and the stock-to-hull structure where the stock passes through the hull and around the lower bearing for hairline crazing, weeps, or resin-starved or delaminated patches — this is the highest-stress region and the first place a load-path problem announces itself. Go over each blade for leading-edge nicks, trailing-edge chips and impact cracks near the tip, and fair anything that disturbs flow: on a foil, a surprisingly small surface defect trips early transition and separation and is real, measurable speed lost. Finally, confirm both blades hit identical stops each way and, if the system kicks up, that both release and re-set cleanly. Fold these into your pre-race inspection checklist and periodic carbon inspection rather than leaving them to chance.

The takeaway

Twin rudders are the reason a wide, powered-up hull's speed is usable rather than dangerous. By canting a blade outboard each side, the boat guarantees itself a near-vertical, fully buried, cleanly-fed foil to leeward through the heel range it actually sails — sidestepping the ventilated stall that turns a single centreline blade into a passenger the instant the boat heels and lights up. The cost is honest and modest: a little parasitic drag from the half-immersed windward blade upwind. The payoff is a boat that keeps steering through gusts and down waves instead of rounding up. Together with the canting keel and the bowsprit, they are central to what makes the Melges 40 fast.