[David Cooper]

my internet connection is too slow to watch the videos so the answer may be in those.

You should really watch those first. This will spare you a lot of guesswork (pay attention to the blade angle vs. rotation).

It simply isn't possible to watch them from where I am, but I will try to view them some time when I am in a better location. My internet connection (3G, but it feels more like GPRS today) is cutting out after each page load and it's taking forever to get anything done online - I suspect the weather's the problem because it isn't normally quite this bad.

Once the vehicle starts moving faster than the wind, the turbine is from that point on moving the wrong way to act as a propeller.

The pitch and transmission are chosen such that it always acts as a propeller. Nothing changes at wind-speed about that.

Do they flip the blades to a different angle to get round this problem?

No. The prop pitch is always positive. On the prototype the pitch is even fixed, and it still accelerates from zero to  >2 windspeed.

It's all very well saying that, but I've just repeated my experiment and it appears to go directly against that. When I blow at the turbine it turns anticlockwise and when I power it anticlockwise (with it facing in the same direction) it blows the air away from me. When the vehicle's moving faster than the wind, the turbine cannot be being turned by the wind without turning in the opposite direction (if someone else blows the turbine from the other side I see the blades turn clockwise from my side), and if it's being powered anticlockwise it should push air forwards. I'm not trying to be difficult - I'm just trying to understand why my experiment directly conflicts with theirs on this point.

EDIT: I've just worked it out and switched the computer on again to add this, and I also see by rereading what you said that you already know the answer, but I misunderstood what you'd said because I was still seeing it as a turbine. There is no turbine - it never acts as a turbine, and that's the real reason the acceleration is so slow rather than being down to torque. I had thought that it was acting as a turbine to begin with, but it isn't - the whole vehicle is actually acting as a very ineffective sail, but as it gets blown along faster it turns the propeller faster and faster as a propeller, so instead of having air moving downwind through a turbine there is air being moved upwind through it.

You have to go the a reference frame that moves directly downwind faster than the wind, at the same speed as the downwind VMG of a boat going at TWA 135°. In this frame the boat slows down the moving water via keel, and accelerates the air via the sail. This frame is analogous to the rest frame of the DDWFTTW cart.

I'm still struggling to visualise it, and I can't see how a keel can slow down the water as an ice runner isn't going to be able to slow down the ice, but I'll take my time over this and keep studying the diagram you've provided.

This whole thing is a fabulous overthrow of parts of my model of reality though, and that's exciting. It's a pity it takes five years for something like this to get through - makes me wonder what other revelations I've been missing out on.

[MarkV]

There is no turbine - it never acts as a turbine, and that's the real reason the acceleration is so slow rather than being down to torque. I had thought that it was acting as a turbine to begin with, but it isn't - the whole vehicle is actually acting as a very ineffective sail, but as it gets blown along faster it turns the propeller faster and faster as a propeller, so instead of having air moving downwind through a turbine there is air being moved upwind through it.

Yes. To increase the intial acceleration you could use the rotor as a trubine below windspeed. This is what Andrew Bauer did on his DDWFTTW cart in the 60s:

http://projects.m-qp-m.us/donkeypuss/wp-content/uploads/2009/06/Bauer-Faster-Than-The-Wind-The-Ancient-Interface.pdf [nofollow]

From what I heave read this guys opted for a one way transmission for several reasons:
- To prove steady state you want to exclude the possibility that stored momentum in the spinning rotor can drive the wheels.
- Installing ratchet hubs at the wheel axle allows differential wheel speeds in turns. A ratchet hub at the propeller shaft allows to brake the vehicle, without having to brake the propeller rotation (quick emergency stop).
- When you switch the direction of the power transmission, the chain slack has to go from one side to the other. Bauer had problems with the chain jumping off here.
- It is simpler: You don't need to build a variable pitch mechanism (although they added one later on, but it doesn't go negative)
- It makes a much better brainteaser if it starts in a tailwind, with the rotor turning the opposite to what most people assume.

I'm still struggling to visualise it, and I can't see how a keel can slow down the water as an ice runner isn't going to be able to slow down the ice,

If you apply a force to the earth (or the ice) there is always some tiny acceleration. In a reference frame where the earth moves this can slow down the earth a bit and extract energy. You don't notice the acceleration, because the huge mass of the earth. With fluids only a small mass is affected, so it is more obvious where the energy comes from or where it goes to, because the acceleration is noticeable.

[David Cooper]

More things are beginning to fall into place for me. In my thought experiment with the two tracks in space where the vehicle can move without any drag, it can accelerate up to the speed of light in either direction by tapping into energy from the relative motion of the two tracks to each other. In our Earthly experiments we're working with two tracks of a different kind though where one is air and the other is either land or water, and because one of the tracks is air, we are going to have trouble putting the power down against it efficiently, added to which there will be more drag going upwind than downwind, so this suggests to me that the speeds should always be higher downwind, thereby explaining why there's more advantage in using a hydrofoil downwind while it costs speed through added drag going upwind if the boat can't go fast enough to rise up on the foils.

Another thing I realise now from looking at the diagram you provided (with four boats and lots of vectors shown) is that the trailing edge of the sail is actually upwind of the leading edge (going by the true wind), just as the propeller blades on the wheeled machine are set to act as propeller blades rather than as turbine blades. In the case of the wheeled vehicle it's easy to see how the wheels are being forced round and are forcing the propeller to turn to push air backwards, but it isn't at all easy to see how it works with the boat and I'm still not there.

With a beam reach (90 degrees to the wind), it's easy to work out what's going on - a continual acceleration force is being applied by the wind by pushing against the sail and being deflected to the side, so if all friction could be eliminated the boat could keep on accelerating to the speed of light.

Another way of looking at it is like a wet bar of soap. If the two opposite sides are exactly parallel, it can slip out of your hands, but pressing harder won't accelerate it, however if the sides aren't parallel the pressure will cause an acceleration, and the harder you press the faster it'll accelerate. The keel/daggerboard is one of the surfaces being pressed and the sail is the other, and they are angled such that the pressure on these two surfaces will clearly accelerate the boat forwards.

In the case of the boat travelling at 135 degrees to the wind, this bar of soap analogy still fits, but because the boat is making downwind progress faster than the wind, it's hard to see how it can still be being squeezed from both sides. The diagrams clearly show that it is still generating a forward force, and that's all it needs to accelerate beyond the point where progress downwind is faster than the wind, but I'm still struggling to see how the power can really still be on in this way - it still feels as if it should be impossible, even though it clearly isn't. The sail is not backing - it's still being pushed out in a downwind direction by the wind (even though the wind is moving slower downwind than the sail) while water is still pushing against the keel and making the whole thing act as a wet bar of soap, so clearly it is still undergoing a force that will try to accelerate it. The sideways movement of the boat across the wind is what keeps the sail from backing, and that keeps the acceleration force active, but it seems strange that the boat's existing movement can generate this additional acceleration force. This is the final barrier to me understanding this properly.

With the wheeled vehicle we have a similar situation, but it all adds up more clearly: the wind is the first part of the chain of causation, pushing the vehicle forwards, making its wheels turn which makes the propeller turn, and it goes on working beyond the speed of the wind because the propeller is always pushing against air which is moving downwind rather than starting out with the speed of the ground. With the boat it must be the same: the sail is being driven through air which is moving at the speed of the wind and not the speed of the stationary water, so it's very different from a sail pushing stationary air where there would be no gain of energy - that must be the key to how this force is still being tapped.

Yes, I think I may be getting there. If you look at the diagram with the four boats and the vectors, imagine holding the boat in still air and moving the boat directly sideways through the air: you apply a sideways force to generate a forwards force and lose out overall to drag. Now do the same with the keel in water and you generate more forwards force while losing out to more drag. In both cases it isn't sustainable and the boat would slow to a halt if you let go of it. Now try it again though with the water moving while keeping the air still. The water is flowing towards you from ahead while the air is still. You now push the boat sideways again with the keel in the water. This time the keel is being pushed from the other side which you'd think would make things worse, and actually it does in a way, because the boat is now going to use the force it generates just to maintain it's position relative to you (in upwind/downwind terms, with the air being stationary) - previously the boat was being deflected ahead of you at high speed when the water was stationary with the air. That is one major difference, but the other is that the effect of the moving water keeping the boat level with you is that its sail is being driven through air which is moving relative to a boat in a very different way from in the case where the water wasn't moving relative to the air, because in that case the boat was deflected forwards and its sail was hitting air which was moving fast and backwards relative to it. In this new case with the water moving, more energy is necessarily tapped from the relative boat and air speeds, ensuring that an acceleration force remains in place continually.

That is it - a plain words explanation without any confusing abstract maths. I can expand further upon it by comparing it to what happens when you are swimming. When you start the length, you push away from the wall and it's easy to gain a lot of speed by doing so, but if the wall started to move away from you at a fixed speed just as you started to push against it, you'd get less energy from the push. If the wall moved with you as you pushed off, you'd get more energy from it. That is the mechanism by which the boat can continue to accelerate beyond the speed of the wind, because it is still tapping that extra energy that comes from the fact that the air is moving at a different speed from the water, keeping its speed closer to that of the boat.

Okay - I'm happy with it all now, though I wouldn't be surprised if I'm still missing something. What I'd particularly like to see now though are diagrams of tables showing different angles to the wind and boatspeed, ideally for many different kind of craft (including slow monohulls). If anyone stumbles upon anything of that kind, please could you post a link to it here.

[MarkV]

Once you have faster internet you should check the animations. They pretty much show what you are trying to grasp intuitively.

In the case of the boat travelling at 135 degrees to the wind, this bar of soap analogy still fits, but because the boat is making downwind progress faster than the wind, it's hard to see how it can still be being squeezed from both sides.

The squeezed wedge for downwind VMG > WS is shown here:

Now try it again though with the water moving while keeping the air still. The water is flowing towards you from ahead while the air is still.

This animated diagram shows downwind VMG > WS from 3 different reference frames:

Note how that air is still being slowed down in the frame of the surface. To extract wind energy you have to slow down the true wind (reduce the velocity difference between air and surface).

Here is an article explaining this as well:
http://rightnice.blogspot.de/2010/08/racing-wind.html [nofollow]

What I'd particularly like to see now though are diagrams of tables showing different angles to the wind and boatspeed, ideally for many different kind of craft (including slow monohulls).

The diagrams are called "polars". You can search google images for them.

Here is one for slow mono hulls:
http://www.sailonline.org/static/var/sphene/sphwiki/attachment/2011/06/25/TallShipPolars.png [nofollow]

Here for skiffs that achieve downwind VMG > WS:
http://www.yachtblick.de/www.yachtblick.de/wp-content/uploads/2012/02/BoatAndWindSpeed-custom-size-660-440.jpeg [nofollow]

Didn't find any for ice-boats or land-yachts, which easily achieve downwind VMG > 2WS.

[David Cooper]

Once you have faster internet you should check the animations. They pretty much show what you are trying to grasp intuitively.

Will do.

Here is an article explaining this as well:
http://rightnice.blogspot.de/2010/08/racing-wind.html

I see there are links to videos at the bottom of that page too - if they're different ones I'll check them too when I can get better Web access.

The diagrams are called "polars". You can search google images for them.

Thanks. That's the vital word I lacked, and this is one of the particular things I was looking for:-

http://www.thebeachcats.com/OnTheWire/westnet/_lpm/hobie/archives/v1-i3/feature3.htm

The top two are catamarans, and I particularly wanted to see how well the Tornado (until recently an Olympic class boat) races the wind - the answer is that it just about achieves the speed of a ten-knot wind at 135 degrees and will therefore fall far short overall, though at 130 degrees it goes 10% faster than the wind and actually makes slightly faster progress downwind. The lowest part of the line is the point where progress downwind is greatest.

Actually there's a serious flaw with those diagrams which makes them very misleading. If you look at the one for the Soling (Olympic class monohull), it looks as if it can gain by tacking downwind, and the text below claims that all six of these boats do gain by doing that, but if you apply pythagoras/trig to the diagrams you should be puzzled at how a point where the wavy line hits 8 knots on th 135 degree line can be further down the diagram than the point at 6 knots on the 180 degree line. The answer turns out to be that the numbers appear to have been put on the wrong lines with 4 actually being 3, 5 being 4, etc., so either the speeds are wrong or the graph is distorted, and I suspect the former because if it was the latter there would be no advantage for the Soling in tacking downwind. Clearly you need to keep your wits about you when looking at polars.

Edit: I've just seen the links you've added. It's interesting to see the Class C ketch in the tall ships one - in 3 knots and 20 knots of wind it should tack downwind, but at 12 knots and 30 knots of wind it should go directly downwind. You really need to know your boat's polars.

[MarkV]

The answer turns out to be that the numbers appear to have been put on the wrong lines with 4 actually being 3, 5 being 4, etc., so either the speeds are wrong or the graph is distorted,

Yes, I see what you mean. If you extrapolate the scale, you end up with 1kts in the center, not 0kts.

The most impressive VMG preformance on water, that I'am aware of:
http://en.wikipedia.org/wiki/USA_17_%28yacht%29 [nofollow]
"USA 17 reached the windward mark in 1h29, so her velocity made good was about 13.5 knots (25.0 km/h; 15.5 mph), or about 1.8 times wind speed. USA 17 took 63 minutes to reach the downwind mark, so her velocity made good downwind was about 19 knots (35 km/h; 22 mph), or about 2.5 times wind speed."

[imatfaal]

Great posts MarkV - thanks

[David Cooper]

The most impressive VMG preformance on water, that I'am aware of:
http://en.wikipedia.org/wiki/USA_17_%28yacht%29

One particularly good link from there is to this:-

http://en.wikipedia.org/wiki/Sailing_faster_than_the_wind

As a child I used to sail Mirror dinghies in a handicap fleet which contained at the top end Nacra 5.2 catamarans (crewed by people who raced at national competition level). We all sailed more or less directly downwind, not knowing any better, though the p.y. handicap ratings for the boats would have been based on the way people sailed back in those days. Things have clearly moved on a lot since then, though I still can't find a polar for the Mirror dinghy. I imagine that it's quite hard to work them out though, and different people measuring speeds and angles in different and varying windspeeds are going to produce different polars. It would take GPS and windspeed and direction measurements to do the job properly by putting such devices an all boats and merging the data to create the most accurate polars, although they'd still vary just through weight and weight distribution of different crews, plus steering and sail handling style, so even then it won't be possible to get anything definitive. It will of course be able to distinguish between the top sailors and the rest, so it would help everyone up their game, indicating where they're doing something wrong.

[David Cooper]

[Edit: the following is a comment on two posts which have now been deleted as someone had posted them to the wrong thread.]

That's an unfortunate hazard with this forum if you open another thread in another tab and then try to reply to a previously opened tab.

[MarkV]

Things have clearly moved on a lot since then,

Yep. When you think about, what the USA17 VMG performance actually means: You can release a balloon at the upwind mark, when the boat starts at the downwind mark. The boat will go to the upwind mark, then back to the downwind mark, and arrive there before the balloon.

In terms of absolute speed, things have also move a lot recently. 121 km/h is the new record from last year:

203 km/h  for landyachts:

[dlorde]

Thanks MarkV, for the detailed clarification and explanations; I have only a simple understanding of the principles, not the details necessary for a clear explanation. Your contribution helped me too 

[MarkV]

Great posts MarkV - thanks

Thanks MarkV, for the detailed clarification and explanations; I have only a simple understanding of the principles, not the details necessary for a clear explanation. Your contribution helped me too 

You're welcome!

If you find high-performance windpowered stuff interesting, check out dynamic soaring:

http://en.wikipedia.org/wiki/Dynamic_soaring [nofollow]
"As of March 2012, the highest reported speed for radio control dynamic soaring was 498 mph (801 km/h)"

This is achieved using winds of about 50 mph.