Wing Sails on Proas!
While a solid wing may have it’s best L/D at a low AoA, the lift coefficient at that AoA is also low, say about 0.3. What this means is that you need about 4 times as much area for the same thrust as you do if operating close to a CL max. of say 1.2. So for the same thrust at L/D max. the wing could weigh up to 4 times as much and the centre of effort is much higher. So even though the wing efficiency is higher, the net boat performance may be significantly lower. I’m not saying that this renders the system inoperable, but one does need to be aware of the compromises.
Mal.
True enough Mal, but misleading, don’t you think? Nobody in their right mind would size a wing for 4 degrees AoA and 0.3 CL. The salient points are these; at 4 degrees, a soft sail (unarig or main/jib) won’t draw at all, thus has a CL of zero. Much inferior to the wing sail’s 0.3 CL at that AoA. Now let’s move out to a slightly more useful AoA; perhaps 12 degrees. The soft sail is now working near it’s max L/D; about 6 or 8:1 and a CL of 0.8 (not 1.2-1.5; AR = 6; planform = fathead main plus conventional 3/4 jib). The wing is now near its max CL, between 1.8 and 2.2, and has dropped its L/D to “only” 12. Overall, the wing is both higher and faster, which was the intent of the exercise, no?
Dave
I was curious about Peter Worsley’s system with regard to how the flap angle changes relative to the wing angle (my concern is that there is not enough flap movement in the close hauled position). So I modelled up a 2D sketch of the system using CAD software and plotted the angles. I used a 100mm dia. mast and a 300mm dia. cam with 80mm of travel either way. I positioned the pivot point for the cam follower as close as practical to cam in the maximum travel position. I adjusted the lever lengths so as to get as large a flap movement I could get in practical terms (about 60 degrees). The results, plotted below, show that as I suspected there is not a lot of flap movement in relative terms at small wing angles (i.e. close hauled). This may be ok if, as I did, you configure it to give large maximum flap deflections. If not, you won’t get any decent thrust out of the system when close hauled. —Mal.
Sorry to push so far back in time (more than a week!), but I keep putting this off and don’t want to.
I think Mal may be caught up in the C-Cat/AC45/AC72 POV where wingsails are typically 3 elements and masts are stayed, so you’ve got limited rotation of the wing, thus a need for large flap deflections and highly cambered wings. This is a common enough trap to fall into.
Tom Speer’s wingsail pages (http://www.tspeer.com) show that maximum CL for a 2-element wing is achieved with about 20-25 degrees flap deflection and ~10-15 degree AoA (CL above 2.0 for several main/flap element chord ratios). Thus, *if* you have an unstayed rig and can rotate the wing through 360 degrees, you never need more range of deflection or AoA than this (max is max, after all), so can alter the leverage of the tail crank, yielding sufficient flap deflection force on any (hull) heading.
Put another way, once the wing/flap/tail are at max CL, the hull can take any course it can lay, without ever altering the setting on the rig. The rig only needs setting for flatter deflection/lower AoA when hard on the wind. In all other cases one sets the rig at max CL and just sails away, under that fixed rig. Please read that again, the wing need *never* work at AoA above 10-15 degrees and a flap deflection of ~20 degrees.
I’m probably missing something, which I’m certain Mal will elucidate, but it appears that Peter’s mechanism has both sufficient travel and sufficient leverage to achieve nearly 100% of the courses and conditions a small hobbiest boat might want.
Dave
True enough Mal, but misleading, don’t you think? Nobody in their right mind would size a wing for 4 degrees AoA and 0.3 CL.
That is precisely true Dave and my point is that with the cam system 4 degrees may be as much AoA as you can get when sailing upwind, so you have to size the wing accordingly to get equal thrust to any rig, wing or soft sail, which is set at the optimum AoA. I have watched the video’s of Peter Worsely’s model and the scenes of the model sailing upwind, compared to the downwind legs, seem to support my theory i.e. there is a big disparity between downwind and upwind performance and the model, to me, seems to struggle upwind. Again, I don’t wish to disparage Peter Worsely’s work, which I think is inspired, but I do see the cam action curve as an inherent flaw that needs attention.
Mal.
Sorry to push so far back in time (more than a week!), but I keep putting this off and don’t want to.
I think Mal may be caught up in the C-Cat/AC45/AC72 POV where wingsails are typically 3 elements and masts are stayed, so you’ve got limited rotation of the wing, thus a need for large flap deflections and highly cambered wings. This is a common enough trap to fall into.
Tom Speer’s wingsail pages (http://www.tspeer.com) show that maximum CL for a 2-element wing is achieved with about 20-25 degrees flap deflection and ~10-15 degree AoA (CL above 2.0 for several main/flap element chord ratios). Thus, *if* you have an unstayed rig and can rotate the wing through 360 degrees, you never need more range of deflection or AoA than this (max is max, after all), so can alter the leverage of the tail crank, yielding sufficient flap deflection force on any (hull) heading.
Put another way, once the wing/flap/tail are at max CL, the hull can take any course it can lay, without ever altering the setting on the rig. The rig only needs setting for flatter deflection/lower AoA when hard on the wind. In all other cases one sets the rig at max CL and just sails away, under that fixed rig. Please read that again, the wing need *never* work at AoA above 10-15 degrees and a flap deflection of ~20 degrees.
I’m probably missing something, which I’m certain Mal will elucidate, but it appears that Peter’s mechanism has both sufficient travel and sufficient leverage to achieve nearly 100% of the courses and conditions a small hobbiest boat might want.
Dave
Firstly I should clarify my terminology. I’m assuming a single element wing with a movable tail. I used the word flap, when I probably should have said tailplane, or something similar, so there may be some confusion there. At no point have I considered multi element wings in my analysis.
Secondly, we should clarify how the tail controlled wing system works. For a stable system, the system is usually balanced so that the tail is set with with a slightly negative angle of attack when actuated. Thus the tail contributes no lift to the system and in fact reduces the total lift of the system as well as adding drag and weight. So one of the aims of the tail design is to make it as small as possible and to place it as far aft as possible as Tom Speer explained. It is theoretically possible to balance the system so that the tail is at a slightly positive angle of attack, but you run the risk of the system oscillating when in neutral (instability). Either way, what this means is that the tail (flap) deflection angle is approximately equal to the angle of attack of the wing, within a a degree or two for a well balanced system.
My concern with the Woresly cam is that in the upwind position the tail deflection angle is small, In my analysis I set up the model to give large deflection angles in the downwind position only so that I could get acceptable defection angles in the upwind position. I was aiming for around 10 degrees deflection in the upwind position which equates to about 8 degrees or so angle of attack for the wing. To achieve this I ended with maximum tail deflections of over 50 degrees, meaning that on the downwind legs the wing is way over-sheeted, so to speak, and probably stalled. This is not ideal but on downwind legs rig drag is not important and you will still get good thrust from a stalled airfoil.
If you decide to set up the system for a downwind tail deflection angle of say 12 degrees, which would maintain attached flow on downwind legs, the curves I have drawn are just scaled to suit so you only end up with only a few degrees of tail deflection for upwind legs. This is not enough (in my opinion) to get good thrust form the wing on upwind legs.
You could set up the system for large tail deflections to get the required upwind tail deflections and then manually back it off for downwind legs, but that kind of defeats the the purpose of the fully automated cam system.
None of this has anything to do with multi element “conventional” wings.
Mal.
All this talk of tail controlled wings has got me thinking about modifying the ArcSail:
All this talk of tail controlled wings has got me thinking about modifying the ArcSail:
Go for it.
Got me to thinking also .... bad idea in this case.
What if?
What if the wing sail on the Bionic Broomstick works really well? Easy to handle, points like crazy, creates enough apparent wind to sneer at those times the wind is light (variable almost a non issue by definition).
What’s going to happen to Nomad?
It’s early too really obsess about this stuff, but BB is almost done sailwise, rivet in a headboard, make something a little more elegant than a pair of visegrips for the roller furling crank and its time to assemble the thing and yard sail a bit before assembling the HF trailer just purchased.
Wingsail’s not too far behind, got all the materials (I think) and the stub mast tripod & cam assembly is about half done.
There is a real possibility that the wing sail is going to work well enough that a quick cycle of development (mistake corrections) leaves me with a system that will be hard to ignore for Nomad. I can already visualize BB wing feathered in 10-12 mph wind over the float and stepping on the cam actuator bringing up thoughts of a seat belt 😉
The dilemma for Nomad and wingsails is ease of assembly and disassembly. Here’s my current thinking. Wings will be hotwired foam, probably 2 oz kevlar skin w/ carbon uni spar caps, rudder assemblies and float leading edges are already to be foam bits skinned with glass. Tripod assembly gives me a hard point to swing the stub mast into place, not that much different than how P52 worked. It may require an extra hand on the ground until a little stick time shows what possible and what’s not.
Cheers,
Skip
Skip,
Will you be able to use the BB wing as one of the wings on Nomad?
If I may ask a critical question, why are you using a tripod to raise the wings so far off the deck? Normally with a sail it is desirable to get it as low as possible, both to get the COE as low as possible and to benefit from end plate effect. If you were designing a sloop rig for instance, you would probably be trying to get the foot of the jib as close to the deck as possible.
Over the years I have read about many attempts at wingsails. Most of them fail. But if you analyse them, the reason they fail is usually in the execution of the idea, not in the idea itself. Most are too heavy, too small, too high above the deck and pay little or no attention to issues like planform shape, thickness distribution or twist. The assumption always seems to be that because it is a wing it will just work better than a soft sail. Really, the only advantage a wing has over a soft sail is a low drag section. But the efficiency of a sail is about much more than just section shape, you have to get all of the other factors right as well, or it will fail to impress. Soft sail systems have had a long time to evolve and mature. Solid wings are only in their infancy and we have to be wary of throwing out the baby with the bathwater, so to speak.
With tail controlled wings, there are operational benefits, but these must be balanced against the performance compromises. The tail automates incidence control and potentially reduces crew fatigue, but adds weight and drag and reduces the total lift of the wing system. I would not be expecting your wing to out perform a good soft sail, but would be balancing the operational benefits against the cost, complexity and performance loss.
One potential additional benefit of a tail controlled wing is that it can be used as a wind vane. Just couple the wing to the steering system and the boat will maintain it’s heading with respect to the wind direction, without the addition of a separate steering vane and with a fully mechanical system.
Having said all that, I’m really looking forward to seeing how your wing performs and I hope I’m being overly pessimistic.
Mal.
Skip,
A Will you be able to use the BB wing as one of the wings on Nomad?
B If I may ask a critical question, why are you using a tripod to raise the wings so far off the deck? Normally with a sail it is desirable to get it as low as possible, both to get the COE as low as possible and to benefit from end plate effect. If you were designing a sloop rig for instance, you would probably be trying to get the foot of the jib as close to the deck as possible.
C Over the years I have read about many attempts at wingsails. Most of them fail. But if you analyse them, the reason they fail is usually in the execution of the idea, not in the idea itself. Most are too heavy, too small, too high above the deck and pay little or no attention to issues like planform shape, thickness distribution or twist. The assumption always seems to be that because it is a wing it will just work better than a soft sail. Really, the only advantage a wing has over a soft sail is a low drag section. But the efficiency of a sail is about much more than just section shape, you have to get all of the other factors right as well, or it will fail to impress. Soft sail systems have had a long time to evolve and mature. Solid wings are only in their infancy and we have to be wary of throwing out the baby with the bathwater, so to speak.
D With tail controlled wings, there are operational benefits, but these must be balanced against the performance compromises. The tail automates incidence control and potentially reduces crew fatigue, but adds weight and drag and reduces the total lift of the wing system. I would not be expecting your wing to out perform a good soft sail, but would be balancing the operational benefits against the cost, complexity and performance loss.
E One potential additional benefit of a tail controlled wing is that it can be used as a wind vane. Just couple the wing to the steering system and the boat will maintain it’s heading with respect to the wind direction, without the addition of a separate steering vane and with a fully mechanical system.
F Having said all that, I’m really looking forward to seeing how your wing performs and I hope I’m being overly pessimistic.
Mal.
Excuse the format of my answer, haven’t mastered nested quotes on this forum.
A. No the section is the same, 30” chord, but Nomad will have more span and be built a little more substantially as would be appropriate for a boat meant to travel.
B. Main reason is to have the swinging bits out of the way of people and the tent for overnighting in mosquito territory. The end plate condition is open to interpretation depending on who you listen to, I don’t have any strong opinion either way on the issue. Other consideration is spreading the loads out and ease of raising the dang things.
C. I agree, that’s the reason I’m giving a go at a reasonable trial on a 14’ trial horse. I’m not really a proponent of wingsails but was blown away with the elegant simplicity of Worseleys’ cam and an interest in Elkaims paper on Atlantis and low Re # airfoils which resonated with my experience building a small RC deltawing glider almost 50 years ago. The plane was uncontrollable with the equipment of the day and my reactions but it had an unbelievable glide, flat and slow, never saw anything like it before or since, it had a root section that was virtually identical to what Elkaim has developed.
D. I agree, I’m not looking to set the world on fire, I just want to fly down the Laguna Madre in relative comfort at 12-14 mph typically and tack/shunt thru 90 degrees or so.
E. Maybe, Nomad’s going to have push pull steering like BB, the rudder lift/lower cordage on the tiller currently seems a really nice setup. Self steering isn’t high on my list, when P52 was oar steered it required 100% attention, plus just a bit; with rudder/ tiller I could do other things within reason, except downwind in rough weather. Sometime in the future I might fiddle with an Arduino based system and take input from the wingsail orientation.
F. I also am really looking forward to seeing how it all works out, have really mixed emotions about the whole deal, I won’t be heartbroken if the wingsail doesn’t make the cut as long as it works reasonably well.
Thanks for the thoughts,
Skip
Just in case anyone interested in wing sails has missed this site:
What about an inflatable wing-sail with some thin carbon-fiber tubes in pockets for added stiffness? It should be easy to build a lightweight wingsail in sections to make it modular and reefable?!?
Just bring a small footpowered air-pump for refilling the wing when needed.
Cheers,
Johannes
Just in case anyone interested in wing sails has missed this site:
Thanks for this, Rob. Pretty cool. They (seem) to offer anything from complete wings ~$800 to DIY kits at ~$200.
Here’s another article on inexpensive wings about this same size: http://www.duckworksmagazine.com/13/projects/rigid/index.htm#.Uu1lTXlt1XY
Dave
Firstly I should clarify my terminology. I’m assuming a single element wing with a movable tail. I used the word flap, when I probably should have said tailplane, or something similar, so there may be some confusion there. At no point have I considered multi element wings in my analysis.
Ah, that devil nomenclature! Can we perhaps standardize on something we all can understand? Perhaps “wing,” “flap,” and “tail”? This is almost universally used in the wingsailing community I believe.
I’ve spoken almost exclusively of 2-element wings, and have clearly said so at each juncture. Single element wingsails are not much better than soft sails, as you say (though “not much” remains in the range of 20-30% higher Cl and 50-70% lower Cd, numbers any soft sailor would kill for!)
In my opinion (and others, including yourself), single element wings ought not be used on small boats, especially when adding a flap is not a big deal either cost wise or controls wise. You, Tom Speer, myself and also Peter Worsley have *all* agreed on this subject, and yet you seem to persist in denigrating wings as so-so performers while solely basing your arguments on the poorest performing of wings.
Secondly, we should clarify how the tail controlled wing system works. For a stable system, the system is usually balanced so that the tail is set with with a slightly negative angle of attack when actuated. Thus the tail contributes no lift to the system and in fact reduces the total lift of the system as well as adding drag and weight.
Much like the tail surfaces on any of millions of airplanes. Without quantifying what you mean by “reduces” and “adding” we have no way of knowing how tiny you believe the consequences actually are.
So one of the aims of the tail design is to make it as small as possible and to place it as far aft as possible as Tom Speer explained. It is theoretically possible to balance the system so that the tail is at a slightly positive angle of attack, but you run the risk of the system oscillating when in neutral (instability). Either way, what this means is that the tail (flap) deflection angle is approximately equal to the angle of attack of the wing, within a a degree or two for a well balanced system.
My concern with the Woresly cam is that in the upwind position the tail deflection angle is small, In my analysis I set up the model to give large deflection angles in the downwind position only so that I could get acceptable defection angles in the upwind position. I was aiming for around 10 degrees deflection in the upwind position which equates to about 8 degrees or so angle of attack for the wing. To achieve this I ended with maximum tail deflections of over 50 degrees, meaning that on the downwind legs the wing is way over-sheeted, so to speak, and probably stalled. This is not ideal but on downwind legs rig drag is not important and you will still get good thrust from a stalled airfoil.
You’re (slightly) optimizing one part of the system while (greatly) de-optimizing another in this case. I agree with you that the AoA of the tail will often be about the same as the deflection of the wing. I disagree that the tail must be “as small as possible.” Induced drag varies with lift produced, not with wing area. “Wetted” surface drag does vary with wing size, but as has been said many times, zero AoA/zero induced drag of the entire wing and all its controls are less than those of a bare round mast alone. Carrying a slightly “too big” tail plane, then, offers very slight increase in drag whilst maintaining as large a reserve as you like with which to control the wing. Even more efficent is to increase the length of the tail boom, which carries no aerodynamic penalty at all.
If you decide to set up the system for a downwind tail deflection angle of say 12 degrees, which would maintain attached flow on downwind legs, the curves I have drawn are just scaled to suit so you only end up with only a few degrees of tail deflection for upwind legs. This is not enough (in my opinion) to get good thrust form the wing on upwind legs.
In your drawing attached, It appears you show just as you state above; about 12 degrees of tail deflection deflects the wing about 10 degrees. But that’s all that is necessary, for any course, at any boatspeed. It is never necessary to set the wing angle greater than 10 degrees, as we are speaking of the wing and not the hull. Even on broad reaching courses, the wing can remain at 10 degrees to the wind as the boat rotates under it.
You could set up the system for large tail deflections to get the required upwind tail deflections and then manually back it off for downwind legs, but that kind of defeats the the purpose of the fully automated cam system.
Agreed, however, you do make a reasonably cogent argument that the control of such a wing/flap and tail combination might profitably be “binary.” A 10 degree angle of incidence on the main wing will be a +\- optimal setting for almost every course the boat can physically sail. If the system had had no cam at all and only 2 settings; 0 degrees and 10 degrees, thus only “On” and “Off”, mightn’t you be able to sail anywhere on earth? Kinda takes your breath away, no? 😊
Yes, you need to be able to reef the system, but you could add a simple bias so that “on” becomes a smaller and smaller incidence angle, either controlled by wind speed or other over-power sensing, or just manually at the skipper’s pleasure.
None of this has anything to do with multi element “conventional” wings.
Perhaps it should. I thought we had broad agreement that multi-element wings were the only practical solution for small boat sailing.
Dave
Excuse the format of my answer, haven’t mastered nested quotes on this forum.
...
B. Main reason is to have the swinging bits out of the way of people and the tent for overnighting in mosquito territory. The end plate condition is open to interpretation depending on who you listen to, I don’t have any strong opinion either way on the issue. Other consideration is spreading the loads out and ease of raising the dang things.
End plate reduction of drag isn’t a belief system, Skip. Laws of physics don’t actually require our agreement. 😉 Your wings shown are wonderfully thin and elegant. From your sketch, each would only “sweep” an area on deck of perhaps 30 inch radius. Surely it wouldn’t be a great handicap to duck around them as they tack or jibe? On the big AC boats, it’s actually easier to duck around the trailing edge of the deck-sweeper than it would be to dive under a low boom with conventional sail. It would be no big deal to keep the tail surfaces up high where they wouldn’t be bopping folks on the head. Ditto the mass-balancing counterweights. In fact, Tom Speer finally got it through my thick skull that we *want* to have these weights as high as possible, counter-intuitive as that is. *
If you decide to bring the wings back down to the deck—or some deck-like sweepable end plate structure maybe integrated with the deckhouse roof?—not only will you reduce the height and heeling moment, youl also reduce the cost and weight of all the supporting structure.
—Quoted from Mal:
Over the years I have read about many attempts at wingsails. Most of them fail. But if you analyse them, the reason they fail is usually in the execution of the idea, not in the idea itself. Most are too heavy, too small, too high above the deck and pay little or no attention to issues like planform shape, thickness distribution or twist.
Mal is painting with a pretty broad brush here. Sure, not everyone is an aeronautical engineer, but when I search YouTube’s site for “wingsail” I get 3400 videos of wingsailed boats succeeding; searching “wingsail fail” yields 400 only. This doesn’t sound to me like “most of them fail.” I wonder how many erstwhile soft-sailboat designs succeed?
D. I agree, I’m not looking to set the world on fire, I just want to fly down the Laguna Madre in relative comfort at 12-14 mph typically and tack/shunt thru 90 degrees or so.
Don’t let the pessimists get you down Skip. Virtually every wingboat actually built, discounting the ones that fall over, blow up or suffer other adventures related to lack of engineering know-how, reports his boat sails faster, closer to wind, has less heeling force, is easier to sail and is just plain COOLer than anything owned prior. Check out the http://www.solidwingsails.com site. you could have both your rigs as a DIY kit for a few hundred bucks.
Dave
*From Tom Speer, regarding mass balancing of wingmasts:
Dave Culp: ...you mentioned that a mast-balancing counterweight cantilevered from the front of the wing is best carried as high above the deck as practical, to gain advantage from “lateral displacement of the wing.” This one escaped me altogether; what did you mean?
Tom Speer: The wing doesn’t stick straight upright all the time. Slack or stretch in the rigging allows it to move from side to side. There’s no translation at the mast step. The side-to-side translation is greatest at the head. Since the purpose of the counterweight is to react to the translational acceleration by creating a stabilizing rotational dynamic moment, you can get the required moment from the least weight by putting it as high as possible.
End plate reduction of drag isn’t a belief system, Skip. Laws of physics don’t actually require our agreement. 😉 Your wings shown are wonderfully thin and elegant. From your sketch, each would only “sweep” an area on deck of perhaps 30 inch radius. Surely it wouldn’t be a great handicap to duck around them as they tack or jibe? On the big AC boats, it’s actually easier to duck around the trailing edge of the deck-sweeper than it would be to dive under a low boom with conventional sail. It would be no big deal to keep the tail surfaces up high where they wouldn’t be bopping folks on the head. Ditto the mass-balancing counterweights. In fact, Tom Speer finally got it through my thick skull that we *want* to have these weights as high as possible, counter-intuitive as that is. *
Don’t let the pessimists get you down Skip. Virtually every wingboat actually built, discounting the ones that fall over, blow up or suffer other adventures related to lack of engineering know-how, reports his boat sails faster, closer to wind, has less heeling force, is easier to sail and is just plain COOLer than anything owned prior. Check out the http://www.solidwingsails.com site. you could have both your rigs as a DIY kit for a few hundred bucks.
Dave
There are some pretty well reasoned articles in AYRS journals and elsewhere that question some of the efficacy of end plate condition particularly at the boom end of sails. I’m fairly open either way but like the idea of tripod mounted wingsail from the basis of simplicity and simple mechanics of mounting the things. I’ll have more of an opinion once I’ve played with the Broomstick a bit.
Haven’t come across any pessimist on this forum but have learned a lot and snarfed a number of very excellent ideas, details and the like. Already done the cool deal https://www.youtube.com/watch?v=brld8jPK6oA
😉 now I just want to play in this little offbeat corner of the universe, really is fun.
Skip