Rudder failure – how safe is yours?

Rudder failure represents a serious emergency. Towards the end of last year, the Jubilee Trust’s magnificent three-masted barque, SV Tenacious, on passage from the Canaries to Antigua, responded to an alert from the Maritime Rescue Coordination Centre in Martinique. A French yacht called Zouk – a 43ft Jeanneau Sun Odyssey belonging to the Glenans Sailing School – had lost its rudder and has been adrift for nearly a week. Attempts were made to tow the stricken yacht but to no avail. Her crew were taken off and Zouk was scuttled so she would pose no threat to other mariners.

An isolated mishap? Not really. Rudder failure raised its embarrassing head again. In the Atlantic Rally for Cruisers (ARC) of 2002, a Hunter Legend 450, known as F2, met a similar fate and, after initial support from three other entrant yachts, reinforcements arrived in the shape of … would you credit it … Tenacious! Taking advantage of their saviour’s comprehensive workshop, a jury rudder was fitted and F2 was on her way.

Onwards to the ARC of 2006 and the story quickens. This time Y Not, a Contest 48 and Arnolf, a Bavaria 35, both found themselves rudderless. Y Not made it safely to St Lucia but unfortunately Arnolf proved unmanageable and was abandoned.

In the course of my research for this article, I heard of other rudder failures, including a Catalina 42, J44, Wylie 38, Hunter (Legend) 466, and a quartet of Cal 39s. I have personal knowledge of an Excalibur 36 (I built its replacement rudder), a Rival 38, a Dehler 34, a couple of Trident 24s, various Westerlies and some earlier Moodys. When you think of how many boats there are out there, this is hardly a mechanical epidemic, but considering how important rudders are in the general scheme of things it’s certainly a cause for concern. Most were spade rudders, and the most common failures saw stocks breaking where they emerged from the hulls – always areas of highest stress concentration. It’s this type we’re mainly concerned with here.

So, what were the causes? In most cases we’ll never know. It’s surprising how many skippers claim they hit something. ‘There was a hell of a bang,’ one once told me. ‘Whatever it was felt really solid.’

‘Just one bang?’ I queried. ‘Surely it would have hit the keel first.’

Of course, this sort of conjecture is very circumstantial. Boats do hit submerged objects, suffering consequent damage, but it may be simply that the rudder flailed about as it came adrift, banging up under the hull in the process. It seems to me that so dramatic is the transition from business as usual to ‘Houston we have a problem’ that in the absence on an indignant whale or a tree trunk in the wake, we may never identify the cause.

To question precisely how such an event occurred might seem like nit-picking – after all, the rudder vanished, one way or another. But it’s really very important. No one can fashion a rudder that will take extreme and unnatural punishment but, if it gives up the ghost in normal service, we have something more worrying to brood on. For what we are then talking about is structural inadequacy from the outset.

Safety in numbers?

My intention here is not to delve too deeply – to review the various points without too much reference to the maths. But not entirely. Any calculation must start by estimating the maximum side force (F) on a rudder, and here we can refer to the old Lloyd’s based formula which for our purposes can be written as:

F = 1.1 x VS² x A

Where VS is hull speed and A is blade area. It can be seen that V is squared to conform with the axiom that lift on a foil increases by the square of its speed, relative to the fluid in which it’s immersed. This leaves the designer to estimate – some would say ‘guess’ – the value of V.

But Lloyds lost interest in yachts years ago (my own copy is dated 1981) and most designers switched to the American Bureau of Shipping’s (ABS) ‘Guide for Building and Classing Racing Yachts’. Their formula for estimating side force took a different tack:

F = 984 x C x LWL x A x N

Both are imperial versions of the formulae but it doesn’t matter. Note that ABS have abandoned hull speed and are simply inferring it from the waterline length (LWL).  A is again the area and C is a lift coefficient which is 1.5 for rudders of moderate aspect ratio. The value of N is usually 1.0 – and therefore has no effect on the result – but rises for lighter than average boats to allow for their higher potential speeds.

Following Lloyd’s departure, ABS decamped (at least from involvement with yachts of less than 24m (79ft)) and the world now awaits the completion of a new ISO Standard, understood to be based generally on ABS.

Anyway, having estimated the side force, the next steps would be to determine the bending moments, then onwards to establish what the size stock would be up to the task. But this is a process too grindingly tedious to pursue – most designers use a spreadsheet, hardly riveting reading. Instead, let’s move towards the source of all those loads.

Beating to windward is thought to be the hardest point of sailing. There’s lots of spray, the boat heels abominably and the crew takes a hammering. The forces acting on the sails, spars and rigging can be savage and these are opposed by the keel and rudder below the waterline.

But, consider this: despite all the apparent drama, the boat’s speed will be low – maybe no more than three quarters maximum hull speed. It follows, therefore, that since the rate of waterflow over the rudder is moderate, according to all established methods of computation the side force will be moderate as well. Despite being counter-intuitive, the fact is that rudders are relatively at peace upwind and more stressed when running or broad reaching, as you might expect in a trade wind crossing. And it’s quite easy to understand why.

Life on the ocean wave

The Greek word for wheel – ‘trochus’ – lends its meaning to ‘trochoidal wave theory’, first advanced in 1862 by William Rankine. It describes a mechanism in which each molecule within a deep water wave follows an almost perfect circle, returning to pretty much its original position once the wave has passed. Actually, the artwork here illustrates a sine curve but it’s close enough to a trochoid for our purposes (trochoidal curves have flatter troughs). The diameter of the ‘wheel’ followed by the outermost molecules is equal to the wave height, and the time taken for it to turn a full circle will be the wave’s period – more commonly expressed as the time in seconds between two consecutive wave crests passing a fixed point. The yellow arrows show the direction of the surface flow and broad blue ones show the varying velocity. The wind direction should be obvious.

Of course, the whole body of water isn’t moving at the speed of the wave train. Waves are the manifestation of the advance of energy, not the horizontal flow of the sea. The orbital action is localized for all intents and purposes. But that doesn’t reduce its significance to sailors who can be dramatically effected by the fact that the water on each wave’s surface moves in different directions and velocities depending on which part of the wave you occupy.

Let’s take the case of a boat running before the trades towards the West Indies. It is December – a good time to cross. The average wave height in that region for that month is about 1.5m (5ft) but ‘averages’ are simply that, and much higher waves can be expected. Waves of over twice the average are common.

So, our boat is keeping up a good lick under twin headsails. It’s exhilarating sailing. She rises to the top of a particularly large wave and … the surface flow at the crest propels her forward over the lip. And, in case you think such a boost would be trivial, the maximum orbital current for a 3m wave with a period of 6 seconds would be nearly 4 knots. At this point, with the boat having first slowed, climbing the wave, she enters the orbital current and the helm goes light. The input velocity of water over the rudder blade has fallen almost to zero, the rudder is ineffective and remains so as our boat starts its descent down the wave face. The helmsman does what he can but yaws off track. Speed increases. The boost from the surface flow, plus the impetus provided by gravity – very considerable if the wave is steep – is now augmenting the boat’s own speed. A lightish 12m monohull could be well into double figures of knots by the time it reaches the trough when – lo and behold – it finds the orbital current rushing up to meet it – at 4 knots, to take our specimen wave.

By now those formulaic estimations of VS based on hull speed plus a bit are starting to look pathetic. In those brief moments before the counter current again slows the boat, the inflow velocity peaks alarmingly. But there could be worse to come, for this is the classic broach situation. Reaching the trough slightly askew, the bow now meets with strong resistance while the stern is still on a charge. The helmsman, naturally, senses he’s losing it and puts the wheel hard over in an attempt to avoid the inevitable. He fails. Instinctive though his actions are, this is not good news.

To understand why, we have to think a bit about foils. Symmetrical foils – such as rudders – obtain their lift by maintaining an angle of incidence with the fluid in which they’re acting – seawater in our case. If the angle of incidence is increased gradually, both lift and drag slowly build to the point where the foil stalls – that’s to say, the flow breaks down on the low pressure side of the foil. This is a comparatively mild event, usually with no serious consequences. However, if the angle of incidence is increased rapidly – as in putting a rudder hard over – the foil can experience what’s known as a ‘dynamic stall’ where the speed of the event almost seems to take it by surprise. This creates a momentary spike of lift, well above that generated in the steady state, before the flow breaks down as before.

It’s also been suggested that modern autopilots might have a hand in our rudder problems. The argument goes that, despite sophisticated control algorithms, they have no anticipatory capabilities, being machines that can only respond to events as they arise. Whereas, for example, a skilled helm would see a monster wave ahead and could prepare to mitigate its effects in advance, the autopilot would be oblivious until its sensors recorded a crisis, whereupon its tiny brain would go into overdrive to put matters right. It’s also possible that autopilots simply ‘work’ the rudder more. Even the best helmsmen don’t correct for every swerve off course whereas autopilots are constantly vigilant.

Which brings us to the matter of material fatigue. The majority of stocks are of stainless steel but some builders favour aluminium. Some Swans and Beneteaus use composite stocks and doubtless there are others that do likewise. But the choice of material is catered for in the formulae. When calculating stock diameter, the material’s ‘minimum tensile strength’ or the ‘minimum yield strength’ is used, whichever is the lesser. Safety factors are built in but these are based on their structural characteristics when new. Unfortunately, subjecting a material to repeated load cycles will weaken it with time, and will do so all the more rapidly if those loads approach its ultimate strength. In practical terms this means that marginally spec’ed stock will fail before a stronger one and that all of them will weaken with age.

If this catalogue of events seems alarmist, then I must admit it is. But it serves to demonstrate how a combination of factors can conspire to overload accepted practices. At least Lloyd’s method of estimating side loads allows the designer to specify the value of VS but the ABS formula – and, presumably the ISO standard that will replace it – decides this for itself in a rather simplistic way, perhaps without full regard to all that might befall. Yes, of course a formula can be tweaked to err grossly on the pessimistic side, but that’s only possible if all the issues are fully understood.

The dynamics behind these unforseen loads was first highlighted by the emminent aero- and hydrodynamicist CA 'Tony' Marchaj, whose work in this field deserves the warmest acknowledgement and applause, since he has shed light into many dark corners. In an article of his he cites the Chairman of a branch of Britain’s Institute for Structural Engineers as saying:

‘Structural engineering is the art of modeling materials we do not wholly understand into shapes we cannot precisely analyse as to withstand forces we cannot properly assess in such a way that the public at large has no reason to suspect the extent of our ignorance.’

So, now you know it all.

Taking precautions

 Understanding the problems associated with rudder failure, is one thing; doing something about them quite another. Since most of the responsibility rests with the designers and builders, it leaves us little more than passive bystanders in the process. But at least foreknowledge allows us to make informed choices.

Given that heavier, slower boats with their greater inertia are less prone to the sort of breakaway speeds demonstrated by their flightier brethren – could they be the better choice? Well, there’s certainly a body of opinion that would say so, but that seems to fly in the face of design development and – what’s more – you would have to pay for all the extra material that goes into them.

Then there’s the issue of spade rudders. Undoubtedly, these are the most vulnerable, so should we turn our back on them and choose skeg-hung rudders? Here I think there’s a stronger case. Manufacturers love spades because they’re efficient, inexpensive to make and can be installed without time-consuming alignment procedures. My own preference is for partial skegs, which add valuable support, allow the rudder to be semi-balanced and also improve directional stability – again, alas, at more expense and with some loss of nimbleness when maneuvering.

But no boat can be made entirely immune, so maybe we should think about back-up systems. Cobbled together jury rigs are a possibility – drogues and the like among them – but I wonder how many of these notions have been tested at sea and how many are just theoretical notions. Lying helplessly beam on to a large swell is no time to be scratching one’s head.

One proven way forward would be to fit a wind-vane self-steering gear that can be adapted for emergency steering. A couple of designs come to mind but there are no doubt others. The Hydrovane with its independent auxiliary rudder is one example and Scanmar’s Monitor pendulum servo gear is another, once converted to its manual steering mode by replacing the pendulum with the ‘M-rud’ blade, available as an extra. Indeed, Scanmar have gone further. Recognizing a growing concern amongst offshore sailors, they have recently introduced their ‘SOS Rudder’ – not a self-steering gear at all but a dedicated emergency rudder that mounts on the transom and can be hinged up out of the way until needed.

But it could be that it’s just unreasonable to expect our rudders to soldier on for ever. The Excalibur 36 mentioned earlier had given 30 years of staunch service before the stock let go. Time had simply taken its toll. We all know that metal fatigue limits the life of our standing rigging; indeed, most of us are reconciled to replacing it periodically. So why not our rudder stocks?

At the end of the day, it will all be down to personal choice. The risk of losing your rudder isn’t high and is probably a risk worth taking for the sort of inshore cruising many of us do. But for those proceeding offshore, where self reliance is everything, it’s certainly worth a pause for thought.