One hull is built to survive the hurricane force winds and mountainous seas of the Southern Ocean. Setting up the hull to ‘surf’ on 30 foot+ seas for minutes at a time, is a common tactic. The other is built to maximize pure speed in ideal to nearly ideal conditions along a comparatively tiny closed course.
The book, The Godforsaken Sea, is an excellent look at what goes into one Vendee Globe race. Managing sleep deprivation is a large part of the ordeal.
Indeed. I think with the new crop of boats there is serious worry that they will have crossed the line in human capability for an extended race. These boats are by all accounts beasts, and even with the available very capable autopilots, how they cope on extended races will be a big question. A trans-atlantic seems to be the longest race they have done so far. But a few of the previous generation semi-foilers did the last Vendée. So there some experience.
The shape of those hulls - broad, flat, like an arrowhead - seems odd to me. I’m sure they’re optimized to the micrometer, but it’s clearly not for sitting in the water anymore. I get that they spend almost all of their time lifted up on foils, but why are they so wide in the stern? Better lever arm for something? Just because they can?
Good article. To be strictly accurate, hull speed applies only to displacement hulls and is directly a function of loaded waterline length (LWL) rather than other common measures of boat size like length overall (LOA) or deck length. It can be roughly calculated in knots as 1.34 times the square root of the LWL in feet.
Hull speed is a fascinating topic, and occurs at the point where the wavelength of the bow wave becomes equal to the boat’s LWL. It’s why you see some sailboats with reverse transoms – transoms that are raked forward from the waterline up to the deck, providing a greater LWL than a boat of that size would otherwise have, particularly when heeled. The hull speed is not necessarily the maximum possible speed of the boat (though for sailboats it’s the major limiting factor), but it’s somewhat analogous to breaking the sound barrier – it takes a lot of power to break through the point of maximum resistance. Typical recreational motorboats facilitate this by using planing hulls rather than displacement hulls, in which the shape of the hull encourages the bow to lift out of the water and the hull is basically planing on the surface.
A motorboat planing at moderate speed is actually running much more efficiently than one that is chugging at sub-planing speeds, though unfortunately the latter can be necessary in rough weather. Among the many great things about sailboats is that they handle rough water much better than similar sized motorboats. It’s true that a friend of mine turned an interesting shade of green once when I took him out sailing and the weather turned rough, but it would have been much worse in a motorboat! 
With respect to the IMOCA-60 class used in the various long ocean races, but mostly focussed on the Vendée Globe. The wide stern shape of those ocean racers is a very specific optimisation, and comes from the time before foils, although the foiling design has further requirements. The round the world races are done in the “right” direction, that is with the prevailing winds, travelling from west to east. For the main part of the race, the boats are sailing with the wind on the rear quarter, and it is actually very rare for them to be close hauled. So the boats are usually reaching. If you want to win the race you optimise for the longest time in the fastest conditions. This does end up requiring significant righting moment, and thus, indeed, the wide stern provides a wider lever. The boats have a canting keel, so the righting moment is pretty big. The addition of foils increases this even more, and has reduced the need for a wide stern somewhat, and the newest boats are narrower at the stern, but still pretty wide.
Another problem is that the rules don’t allow for proper lifting foils on the rudders. So the boats drag their sterns. The boats still need the stability provided by a wide stern. A lot of the boat design is dictated by methods to get what you want within a rule-set that is usually a generation behind. The boats can’t fully foil yet. The class is controlled by the boat owners, and there is a long history of avoiding obsoleting older boats by allowing changes to the design rules that are too dramatic in one go. Many expect the class to allow full foiling in the next iteration. Maybe then the sterns will narrow further.
The Vendée is pretty much the race that defines the class. It is the one race every owner wants to do, and the one everyone wants to win. Boats are super optimised for the race. Designers will run simulations of how the boat will perform using historical weather and other data along the course. Very precise optimisations will be made so that the boat covers that one race in the best time, even if it means the boat is sub-optimal in other races.
Here in Oz we have the opposite. Our one big ocean race is the Sydney Hobart. Most years the wind is behind the fleet, and a reaching monster design with a wide stern will be too draggy to compete with a slim running optimised design. So we have Wild Oats 11, a very slim, very fast running boat optimised to race just this one race, which it often wins. If the wind isn’t the usual direction it can be beaten by other super-maxies, such as the wide stern, Comanche, a boat that is designed to break records, and not designed for a specific race. Comanche is optimised to go as fast as possible, so long as it gets the right wind. Wrong wind, and it suffers badly. Wild Oats 11 and Comanche are optimised for different conditions, and so eventually, it can be simply the weather that decides the race.
Due to the relative characteristics of the fluids (air and water), those sail upwind, not downwind. It needs to have an air speed greater than the water speed.
Depends on how you gear it. You could make such a craft that sailed upwind or downwind. Though the upwind version may well be easier to do.
In principle all you are ever doing is extracting energy from the difference in speed between the wind and the water. It then becomes an impedance matching problem. Design an efficient screw in the water, and an efficient screw in the air, couple them with a variable ratio drive (where the ratio is allowed to go less than unity) and you can go anywhere you like. If there is enough energy in the air/water interface to overcome the losses in the system at any given wind angle you can go faster than the wind speed over water in that direction. The trick is that as you accelerate you need to constantly vary the ratio of your drive system, and at some points it actually needs to start directing energy in the opposite direction to intuition, hence the need to provide for ratios grater than and less than unity. But all the time the energy is being extracted from the air/water motion difference, not the difference between the boat and the air.
The problem with performing this trick on water rather on land are the ruinous losses into the water. Even on foils the losses into a screw are likely to be trivially enough to kill the idea off.
If you watch video of the America’s Cup catamarans you will notice that they only ever seem to be going upwind. Even on a downwind leg of the course, they seem to be heading upwind. When they gybe something really weird happens. The skipper pushes the helm down to gybe, and actual wind passes behind the boat, but because the boat is still moving fast enough it stays on its foils, and the apparent wind passes in front of the boat. So from the point of view of the crew on board, they tacked. If they mess up and it drops off its foils everything gets a bit sad, and they need to sail at a bout 90 degrees to the desired course to get enough speed back to foil, build apparent wind and get moving downwind fast again. That pretty much loses a race in one go.
The airscrew extracts power depending on the energy difference between the air and the boat, and water screw expends power depending on the speed difference between the boat and the water.
The airscrew has to be angled to the wind in the same way that a sail has to be angled to the wind. (The same idea, not the same angle) Also, the water screw has to be angled to the water, in the same way that the keel has to be angled to the water.
A straight (geared or sized) connection will accelerate upwind, due to the difference in Reynolds number. It will drift downwind, regardless of gearing. (Note that to do otherwise gets you into perpetual motion territory).
With the fan and the screw almost at right angles to the wind (and almost at right angle to the direction of travel), you can in theory extract energy while going almost downwind. (Similar to a helicopter). I’ve seen no suggestion that there is a fan shape and screw shape that is efficient enough to do this, although there are things that do not look like sails or wings, and screws and paddles and keels all look very different from each other.
No, that’s false. What matters is that there is a relative velocity between wind and water. If there is, then energy can be extracted. The velocity of the boat is irrelevant except inasmuch as there is drag to overcome at higher velocities, and the props need to extract enough energy to compensate.
A thought experiment to convince yourself that there’s no perpetual motion nonsense going on:
You have a boat that is divided into two parts. One part is a wind turbine thing with a big high-drag fin in the water, so that it hardly moves relative to the water, and generates power from the wind. It connects to the other part with a mechanical transmission cable.
The other part is a very long, skinny, low-drag boat with a (water) prop powered by the cable. It’s so low drag that it can go at almost any speed with arbitrarily low power input. The wind turbine power is easily able to move it faster than the wind speed.
The skinny boat races past the turbine boat. When the turbine boat reaches the stern of the skinny boat, the sailors grab it from the water, run to the front of the boat (feathering the blades so there is no extra drag), and plop it back into the water. The turbine boat weighs almost nothing so there are no momentum issues to consider. The process repeats indefinitely.
The system as a whole moves faster than the wind without any issue. The gearbox does the same thing but more simply.
No, that’s false. The method proposed for extracting energy from air (an airscrew), depends on the relative velocity between wind and boat.
Yes, if you propose a system that does not consist of an actual boat with an airscrew and a waterscrew, you can do something different.
For example, consider the system:
A hydro-electric generator, mounted in the bed of a river, connected by a wire to an airplane. You can make the airplane go arbitrarily fast by using electricity from the relative motion of the water and the river bed!
I wrote above how an /actual boat/ could try to do it: the airscrew and water screw have to be angled correctly to the direction of motion of the boat to get the correct direction of motion in the air and in the water. And yes, you can do this by using two independent keels and energy storage or an infinite wire. But, IMHO, two independent keels is not really just ‘correct gearing’.
The system I proposed has an airscrew and waterscrew.
The system I proposed is not anchored in the seabed. That would of course open up all kinds of other possibilities for energy generation, but it is not in the realm of boat-like things.
You claimed earlier that there were “perpetual motion” concerns. If that were the case, it should not have been possible for me to design a system of any kind, since that would violate physical law. We can reject all 1LoT-violating machines as unphysical, regardless of the details. The converse is that if a machine can extract energy from the environment, then it must not be violating the 1LoT.
Nevertheless, here’s another way of looking at it. Consider the initially proposed system, a boat with an airscrew and a waterscrew connected by a variable gearbox. We’ll start by assuming an infinite ratio between the two: the waterscrew can spin freely, while the airscrew is effectively fixed. Assuming negligible drag, the boat will drift downwind at wind speed.
We can ask now if changing the ratio will make the boat accelerate. We reduce it from infinity:1 to some-large-number:1. Because we’re assuming perfect efficiency, we can say that power input equals power output. As it happens, the power output spins up the airscrew, applying a positive force to the boat. And the power input is from the waterscrew, which acts as a drag force slowing the craft.
But the positive force is much greater than the drag. The airscrew was stationary, and now is moving slowly. In the boat’s frame of reference, it takes almost no input power to provide a large force, because it only has to accelerate the air itself by a small amount.
On the other side, the waterscrew doesn’t cause much drag at all. The water is moving by so quickly, we can extract a large amount of energy from a small change in momentum. This is just the converse of the effect that a propeller (water or air) takes more power to create some amount of thrust the faster it’s going relative to the fluid. Same idea, except that we’re extracting energy here.
So the slowly-moving airscrew provides more thrust than the waterscrew adds drag, and the boat will accelerate from exactly the wind speed to something greater (how much greater depends on the efficiency and the chosen gear ratio).
An AC72 catamaran can sail very significantly faster than windspeed nearly downwind. In 5m/s of wind at a true wind angle of 150 degrees it can manage about 13m/s. That is 2.6 times wind speed, with a cosine correction for wind angle of only 15% needed, the boat is still making a velocity made good exactly downwind of twice windspeed. If I were to attach a rope to the back of the catamaran of sufficient length that as the boat gybes side to side downwind I could be comfortably dragged downwind, I could waterski at twice wind speed, dead downwind.
Reference for predicted and measured polars on an AC72.
The only difference between an AC72 and coupled propellers in water and air is the way the flow directions are handled.
A rotating propeller in air achieves the same angle of attack as the gybing sail/wing, but is not required to flop back and forth gybing. The centerboards versus propeller in water have the same relationship.
What matters is that energy can be extracted from the air-water interface even when the boat is going faster than the speed of air over water in the same direction of travel.
There is a dead zone that has to be transited as the apparent wind shifts from the rear to the front. But it is not impenetrable. Inertia helps.
Without foiling this would never work. But it does, and there are now many existence proofs. There are affordable off the beach cats that can manage this now.
Interesting thread. I’ve raced as crew on a two man Hobie-16, and to me the most exciting moments in sailboat racing happen when the wind totally dies. That’s when you see the perfectly tuned boat moving smartly past other boats that are sitting dead in the water.
Running some example numbers, since I was a little vague above:
Let’s suppose our craft has an airscrew arranged such that it accelerates 1 kg/s of air from 0 m/s to 1 m/s (note again that the craft is initially stationary with respect to the wind). As such it produces 1 kg/s⋅1 m/s=1 N of thrust, and because KE=0.5mv2, it takes 0.5 J of energy to accelerate each 1 kg block of air, for 0.5 W of power.
Suppose the wind speed is 5 m/s, so the craft is already traveling at that rate with respect to the water. Our waterscrew needs to produce 0.5 W for the craft to function. To make things easy, suppose that it also processes 1 kg/s of water. Each kg of water has 12.5 J of kinetic energy, and we need to extract 0.5 J to power the airscrew. The water leaves the waterscrew at 4.9 m/s.
So the water slowed by only 0.1 m/s, and so produces only 1 kg/s⋅0.1 m/s=0.1 N of drag. There’s 0.9 N of net thrust and so the craft can accelerate.