Why do the Vendée Globe solo designs, the IMOCA 60s, keep having keel failures? There have been two keel failures on this race: Safran and now Paprec-Virbac 3, and the breakage of a keel ram on Maître Coq.
If you look more broadly at the class, you're bound to reach the conclusion that the flaws leading to these catastrophes have been endemic. There were six keel failures in the race four years ago and 17 in total dating back to the shocking images of Tony Bullimore's boat upside-down in 1997.
How can it be that these keep happening and what's the reason? How come designers and engineers still cannot manage to ensure keels stay on for the duration of a round the world race?
The subject is immensely complicated, but the simple place to start is that no one - no designer, no engineer and certainly no sailor - wishes to take big risks in this area.
While they are trying not to be handicapped in terms of performance, most teams would be prepared to accept some compromise if it provided a guaranteed solution and over the years many teams have swapped materials back and forth several times in search of it.
The other issue is that keel fins are made out of one of a variety of materials, so there is no one answer. (Safran's was from fabricated titanium; JP Dick's of fabricated steel.) Each fin material has its pros and cons, and the problem is that there have been failures in every material. The table below illustrates some of the choices and trade-offs:
Keel fin types
Material Carbon Fabricated steel Milled steel
Typical weight 450kg 650-kg 1,100kg
Typical max width 120mm 115mm 100mm
Pros Light weight Lighter than milled steel Thin section
Heavier bulb Less expensive than carbon Less drag
'Fit and forget' Cheapest Durable and strong
Cons Flutter problems Finite life Manufacturing time
Expensive Regular maintance Expensive
Poor impact resistance Possible voids in casting
Damage can be invisible Heavy
More wetted surface
There are many areas where things can go wrong with keels. There can be material problems - for example voids in steel - and human error in manufacturing of fabricated steel or carbon keels. There can be problems with the canting mechanism, the hydraulic rams that push and pull the keel head, and in the attachment between or bonding of the keel head and the rams.
Then there are underlying structural problems and engineering miscalculations and, together with manufacturing issues, these are probably at the bottom of the majority of failures.
Besides these, some teams have blamed subsequent keel failures on earlier collisions with marine mammals. Jérémie Beyou voiced the opinion that a collision with a mammal might have bent his keel ram, putting him out of the race on the way down the North Atlantic. Such impacts can do serious damage, though admittedly more serious to the poor animal. Keel calculations are usually for a longitudinal grounding, whereas on a canted keel these can impose a big torsional load.
So coming up with a simple solution that fits all these types is near impossible. For this reason, the class has been debating for over five years the possibility of a one-design keel, or options of a solid forged steel one-design keel. In the last few years I've written about this several times, including here and here.
I prematurely concluded four years ago that the class would have to come up with a plan. The rule has been tweaked, but teams still take individual options, even quite leftfield ones like Safran's super-expensive titanium fin, which failed early on in this race. The team blamed fatigue.
As a group the class hasn't made the leap to a more failsafe option. Interestingly, Alex Thomson comments today: "When I joined the class in 2003 I was a little surprised that I had to change the keel on my first boat because it had exceeded its mileage of 80,000 miles. Since then people have been building keels that last only one round the world race to save a few kilos of weight. I came from the world that a keel lasted for the life of the boat and that is where we need to get to."
This time could a bold move be that much closer? The one-design keel/rig debate is set to be heard again against a counter-argument for a full-on one design for the Vendée Globe. Although Michel Desjoyeaux is not a particular fan of a move away from the development class, it so happens he has a candidate that could be considered: the Oceans50 one-design. More on that soon.
So an interesting crossroads is approaching for the skippers and class. And it may not even be entirely theirs to have. Desjoyeaux is interesting on the subject, pointing out that the Vendée Globe race and boats eligible for it is entirely down to the event management. "IMOCA won't be able to decide," he told me last week. "It is the organisers of the Vendée Globe who will decide. They could say: ‘For the interests of our race, we don't want this.'"
Ian H Bane,
January 23 06:36
The reasons for the many keel and structural failures in modern ocean racing is the evolution of design away from seaworthiness in the pursuit of pure speed. The engineering factors causing the failures fall into three main areas. These are vibration and the phenomena of resonance, lightweight eggshell design, and the inevitability in the big wide ocean of hitting objects.
The phenomenon of resonance when a structure is excited at its natural frequency, can wildly magnify the vibrations, and rapidly fatigue key components. Every engineering student has seen the famous video of the "Galloping Gertie" Tacoma Narrows Bridge failure, which is a classic example of resonance on a structure.
Vibration and resonance was never a problem when the primary stability of racing yachts was from short, heavy fixed keels. However modern designs rely extensively on shape stability, as well as long pendulum type ballast keels. Anyone who has sailed a trailer-sailer with predominantly form or shape stability (relying on beam rather than a heavy keel), will have noted how the hull moves in response to the smallest wavelet. The wide beam dart shape of these modern vessels similarly responds to all the complex beam direction waveforms encountered in a seaway. Add to this the much higher speeds of the lightweight flyers, and you now have vibration loads at frequencies which can resonate with both the long keel structure and the mast rig. It is impossible to calculate all of the potential vibration excitation which a seaway can produce, and particularly for the rigging, each component will have its own natural frequency.
The natural frequency of a pendulum reduces with the length of the pendulum, and the extended pendulum keels of the modern designs have moved their natural frequency closer to that which may be created by the vibration induced by the form stability of the hull form. Long and narrow produces lots of flexibility and classic fatigue problems, hence the perfect storm of fatigue and resonance, combining to induce failure.
Modern structural composite materials are extremely strong and light and have allowed the development of lightweight planing hulls. However, whilst they are strong under design loads, they are vulnerable to impact by objects, and there are many objects which can be encountered in a high-speed ocean passage, so hull failures are no surprise. This vulnerability also compounds the difficulties when things go wrong, such as losing a mast and having the broken mast punch a hole in the hull.
In my opinion the solution is to change the design rules to produce more seaworthy traditional vessels which may sail slower, but which will be safer, more comfortable for the crew, and lower cost, which in turn will make the "sport" of ocean racing more of a real sport than a high-tech business marketing exercise.