Wing Sails on Proas!
Rather than drive the flap by a linkage from the tail - why not drive it from the main cam follower ?
If the flap were spring-biased to its ‘feathered condition’ then a rotary cam on the axis of, and driven by, the main cam follower could be used to actuate the flap in the desired non-linear manner. If no spring tension is an essential feature then a dual cam, akin to that in a Ducati Desmo motorcycle engine, could be used.
Peter H
Your suggestion above is smilar in princple to what I had in mind i.e. drive a second lever or cam system off the first. Of course, it is all extra complication.
Mal.
Your suggestion above is smilar in princple to what I had in mind i.e. drive a second lever or cam system off the first. Of course, it is all extra complication.
Mal.
It need not necessarily be much more complicated. The diagram below shows a simple Geneva mechanism which would operate the flap in the desired manner. The green flap-actuator arm is pivoted at the front of the wing and is driven by a ball-ended pin attached to the frame which actuates the tail. To minimise friction the ball could be some sort of rolling bearing. The pin sits in a slot in the flap actuator at mid-position as shown. The flap is feathered in that condition. It can be seen that as the frame pivots it rotates the flap actuator in the opposite sense - if the tail goes down then the flap goes up. The pin can leave the slot when the flap reaches its maximum angle and the frame is free to travel further to provide more tail movement. The ‘horns’ on either side of the slot prevent the flap from moving once it is at full travel by bearing against the end of the pin. As the frame swings back towards its neutral position the flap remains at full travel until the pin enters the slot and starts to drive the flap back towards feathered. Although shown here as a straight-sided slot and arc-shaped horns it would be possible to give the flap a different movement with respect to the tail by changing the shape of those surfaces - effectively rounding off the transition between slot and horn.
Peter H
Ah, now I see what you are doing! I though you were talking about driving the tailplane with Geneva mechanism, hence my reservations. Even so, the flap actuation is still an on/off affair, whereas Peter Worsely’s system gives a smooth throttling effect. I see what you are trying to do though.
Well, I drew up a system with an additional slot and pin bell crank arrangement. I used fairly extreme values for the lever lengths and got some interesting results (see below). I was certainly able to flatten out the curve at the full throttle position, in fact I configured it with a small reduction of flap angle in the mid range, in order to steepen the initial and final ends of the curve as much as possible. However, this had the unintended effect that as you reduce throttle initially, there is little change to the maximum flap angle, only a flattening of the ends of the curve. Only by throttling back a lot could the mid range flap angle be reduced, and at low throttle values the shape of the curve approaches that of the system without the slot and pin bell crank.
Anyway, it shows that you can do things mechanically to change the wing angle/flap angle relationship, but it’s obviously not such an easy task.
Mal.
...In other news, Peter shared this drawing with me (see attached). It’s possible to have both a main wing AND a flap/second element, BOTH controlled by a single tail vane. With this, on offwind courses you can camber up the wing and get much higher lift coefficient, or “fine” it out and minimize power. It’s not optimized, but it’s gonna be a bunch better than no second element, and hey, it’s all automatic, so what’s not to like?
Dave Culp
We used exactly this linkage on the landyacht Lydia. The flap and tail were linked so as to move in opposite directions. In order to provide for maximum flexibility, both the flap and tail were controlled by Morse cables that ran down from the wing to a T bar that the pilot used to control the wing. One side of the T controlled the flap and the other side of the T controlled the tail. There was a gate that would lock the two sides of the T together so as to operate both as a single control. In addition to the gate, there cables dead ended on blocks with worm screws running through them, and there was a small crank on the ends of the screws. So the pilot could run the blocks in and out to vary the mechanical advantage and thus the gearing between the T bar and the flap. Thus the pilot could elect to have independent controls for flap and tail, or combined control, and could vary the sensitivity for either surface.
Combining the flap and tail into one control worked very well. A given tail angle corresponded to a particular trim angle of attack for the wing. The wing would automatically adjust to the changing apparent wind angle as the yacht accelerated. It responded immediately to gusts as well, sheeting out as soon as the gust hit. Actually, most pilots felt the gust alleviation was too good - they missed the acceleration they normally got in a gust.
Aerodynamic control also improved the performance in light wind. This is always a very difficult condition for landyachts because the static friction means they have a definite wind threshold below which they simply won’t sail. As soon as there was enough wind for the yacht to start swishing its tail around, there was enough wind to sail. One could set the tail to trim the wing just below stall and the yacht would sail along hands free. If one pulled the tail gently to weather, the telltales flipped immediately, indicating the wing was tracking the set angle of attack very closely
It turned out for racing, the pilots’ preferred method of controlling the wing was to set the flap, neutralize the tail, and simply reach up to position the wing manually where they wanted it to be. With the tail providing the aerodynamic balance, it only took fingertip pressure to move the wing.
I don’t think the cam on the mast is necessary or desirable. It is simply a fact of life that the apparent wind will come from different directions relative to the hull, even on the same point of sail. The neutral angle will change constantly, and what is important is the angle of the wing from neutral. The force on the wing is proportional to the angle of the tail, and that’s all that’s needed to control the power. A pushrod or cables that go up the axis of rotation of the wing can transmit the tail angle from the pilot, who’s position is fixed relative to the hull, to the tail, which is fixed to the wing. The wing can rotate around the pushrod or cables without changing the angle of the tail. A pushrod with bearings on the ends would allow for unlimited rotation without interference.
Hi Tom,
Looking at the photo above, there does not appear to be a counterweight for the tail. Is that because it wasn’t necessary for a landyacht, as opposed to a yacht on a water surface that is constantly moving? Also the tail boom appears to be very long.
I note that the flap on the wing is short. Is that to provide a twist effect?
Mal.
Hi Tom,
Looking at the photo above, there does not appear to be a counterweight for the tail. Is that because it wasn’t necessary for a landyacht, as opposed to a yacht on a water surface that is constantly moving?
We started with a counterweight attached to the foot of the wing. It broke off when the yacht hit a bump. The wing didn’t seem to need it.
Also the tail boom appears to be very long.
Yes, it is. The first configuration we tried was a canard, with the arm short enough and the surface small enough to fit under the forestay. That was not successful because drag on the canard opposes lift when you’re trying to move back to zero angle of attack, and there will be an angle at which the canard has no control authority and the wing swings to high angle of attack, out of control.
Then we switched to an aft tail. We found we needed to make the tail larger. And the boom longer. And longer. The configuration shown was quite satisfactory, however. For the same tail volume (area times arm length), aerodynamic damping goes up by the square of the arm length. So a small surface on a long arm is better behaved than a large surface on a short arm. And, with a class that is determined by total sail area, the small tail doesn’t rob as much area from the wing.
FWIW, it may not be very visible in the photo, but there was a thin cable from the top of wing that supported the outer end of the tail boom.
I note that the flap on the wing is short. Is that to provide a twist effect?
Yes, it was. In part it was to concentrate the flap area where the wing chord was larger so as to get more for a given flap area. And in part so in high winds, the pilot could operate with more flap and lower angle of attack, essentially twisting off the top part. The twist aspect worked.
What didn’t work was the performance of the partial span flap. In addition to the increase in induced drag, the induced angle of attack on the upper part of the wing stalled a portion of the wing immediately above the flap. So there was extra drag from the separation as well. At the time, I didn’t have the means of computing these effects.
Hi Tom, welcome to the forum. Nice to see you around.
P.S.: I edited your post to correct the quotes.
Ah, now I see what you are doing! I though you were talking about driving the tailplane with Geneva mechanism, hence my reservations. Even so, the flap actuation is still an on/off affair, whereas Peter Worsely’s system gives a smooth throttling effect. I see what you are trying to do though.
It can be, as you put it, an on/off affair. As mentioned in the original posting the shape of the Geneva cam can be modified to give a different shape to the tail/flap curve.
Although shown here as a straight-sided slot and arc-shaped horns it would be possible to give the flap a different movement with respect to the tail by changing the shape of those surfaces - effectively rounding off the transition between slot and horn.
Equally, the pin (or roller) could run in a suitably-shaped slot to give positive control rather than, as here, on a single surface.
Peter H
What didn’t work was the performance of the partial span flap. In addition to the increase in induced drag, the induced angle of attack on the upper part of the wing stalled a portion of the wing immediately above the flap. So there was extra drag from the separation as well. At the time, I didn’t have the means of computing these effects.
Would it have worked better if the flap ran the full length of the trailing edge but tapered, say, to zero at the head? The idea being to retain some of the twist effect but smooth the change in lift along he span.
Mal.
Having had a closer look at the design, I’ve concluded that the cam should be on the tail-actuation frame and the follower on the flap-actuating arm.
Peter H
Once it warms up a bit (again!) I’ll post some pics of where I’ve headed on this deal. After settling on the 2” alum tube at hand (calc max stress 23,000 psi) it became obvious I was going to really need a few pieces of smaller alum tubing for the booms and counter weight and was going to have to order some stuff no matter what. Will still use the tubing at hand and brace with my favorite down and dirty pvc pipe heated and squashed on the ends. Parts for the cam started before this last norther, looks promising, the actual cam a piece of 6” pvc pipe follower of 1/2” thick recycled polyethylene and extruded fiberglass rod (stuff at hand).
Will be at least a month or more before it gets done, still need to finish the cambered panel staysail and satisfy wife and doctor I can go out and play on the water again.
Skip
Looks good Skip
Tink
I’ll say it looks good. Very exciting stuff.
What’s the area of your wing?
30 s.f. 10’ x 3’ to best use some of the 20"x 30” dollar store foam board for leading edge. It will probably be an absolute dog in really light stuff but may prove interesting once winds pick up to F3 or 4. Am thinking the cam may be foot operated to remind me of old gearhead days long long ago.