Foiler Design

Doug

I accept that Ian Ward was a year ahead of the work we did in getting the Int14 up on foils an will ammend my furture claim statements accordingly

Re flying high if you heal the boat until the boats weight vector and its sail force vector meet on centreline (about 40 degrees for a moth) then height out of the water does not change the balance of moments and as you say higher is less wetted surface therefore better and the distance between CP's does not change the answer.

If the balance state is at a lesser angle and the depth of centreboard adjusted so that when flying low the submerged length is the same as flying high then I think you will find that flying low comes out in front.

alans
http://home.kooee.com.au/zach/hydrofoil.htm

 

Flying: high or low?

Alans et al,I decided to check this out more accurately and see what I came up with. Both "models" had the same wetted surface; "Flying high" was with the bottom 2' above the waterline; "Flying low" was with the bottom 1' above the waterline.
The additional heeling moment from "flying high" was deducted from the "high" righting moment to acurately compare the two righting moments:
1) At 10° angle of windward heel the boat flying LOW produced 37% more RM.
2)At 20° the boat flying LOW produced 4% more RM.
3) At 25° the boat flying HIGH produced 8% more RM.
The crossover-where both "low" and "high" produce the SAME RM seems to be about 22.5°.
To me this suggests that a vertically adjustable daggerboard raised around one foot in lighter conditions and lowered in medium to heavy conditions would be a good answer. In other words fly low in conditions where a 20+ degree windward heel is not practical and fly high when it is....
What do you think?

 

Doug

Yes I agree, the actual numbers depend significantly on where the sums put the CofG and CP's. I believe you get the same performance , high or low, when the line of the weight vector and the line of the sail force vector (Fys) intersect at a line normal to the hull at its centre. i.e. no sideforce on the centreboard and the lift on the foil normal to its span and equally distributed across its span.

alans

 

Lorsail, as per your last message, are the foils submerged at equal depth in the two examples?

Tom speer said:

http://www.fluidlab.naoe.t.u-tokyo.ac.jp/Research/CavPictures/index.html.en

 

foil submergence

Sigurd, generally the rudder foil is a few inches higher than the main foil so it doesn't operate in the wake turbulence from the main foil. BUT- from some of the pictures I've seen of Moths I'm not sure that works out in practice- I think in some cases the rudder foil may be at or below the level of the main foil judging by the nose up attitude of the Moth in some pictures...
========================
Sigurd, I think I misunderstood your question: in the two examples flying high and flying low the immersed part of the foil was the same.
By havinga board that has two settings ,full down and partially retracted it might be possible to have the minimum wetted surface for all conditions and the maximum righting moment from windward heel.

 

Incidence vs flaps

A question for Tom Speer.
At take off, with relatively low speeds, if you have a choice between increasing incidence of the whole foil or using a flap on the back of the foil, to give the same lift, which would give greater efficiency, ie lowest drag?

Or is it better to have a hybrid arrangement, ie: some flap and some incidence.

Also,
Is there any significant difference in stall angle.
Does the flap size matter ie: 20%, 30% of chord etc

regds, Ian

 

Watching the top moth guys at the worlds, race 3, 8 knots of wind, They were climbing out with angle of a attack and flap on both main and rudder foils. this would seem to be appropriate

 

Alan,
John Ilet commented in Melbourne that at high incidence angles the flap was in fact fully up and the boat was still lifting. You see this situation in many of the photos where the wand leaves the surface. This means that the foil is operating rather inefficiently, although I must say very effectively as they are still travelling at high speeds.

My own original bifoiler had no aft flap on the main foil, the foil was fully pivoting with incidence contol via an aft trailing wand on the centreboard. It offers the advantages of simplicity, the ability to set the foil fully neutral in light airs and can overcome the "flap fully up" condition.

I am interested to explore if I should continue my experiments using this method, or whether a flap offers any inherent efficiency advantages eg: higher stall angle, hence the question. It is more about theory than practice at present, as the Ilet foiler is emminently practical.

I note that aircraft primarlily seem to use incidence for takeoff and flaps for landing.

regds, Ian

 

The ideal would be to still be operating in the drag bucket. Looking at the data, it looks like flaps rather than incidence are the way to go.

The Fastacraft foils are a NACA 63412 section with a 120mm chord - Reynolds number of 66,000 per kt; say Re = 500,000 for takeoff. The profile drag is pretty constant from a lift coefficient of 0.1 to 0.7. This is where you want to operate. Say you designed the foil to take off at a lift coefficient of 1.2. The profile drag coefficient there is 0.01894, but at a lift coefficient of 0.6, the profile drag coefficient is 0.00816. So it would be better to double the area of the foil - that would cut the profile drag at takeoff by 14%.

With flap you can extend that to a coefficient of 1.0 or better. But with the bigger flaps and bigger flap deflections, there's such a strong suction peak at the hinge line that I wonder how good the predictions are. The hickup in the lift curve occurs when the transition (via laminar separation bubble) jumps from the flap to just behind the leading edge. The larger flaps extend the drag bucket a bit more. The best section lift/drag ratio looks to be with a 30% chord flap, operating near zero angle of attack.

I rather suspect XFOIL tends to overpredict the maximum lift. But the max lift doesn't seem to be very sensitive to the size of the flap. Flaps definitely increase the max lift, however.

To use the previous example, you could use a 30% chord flap to take off at a lift coefficient of 0.9. That would be a foil 1/3 larger than the unflapped foil that used incidence to take off at a lift coefficient of 1.2. But the drag coefficient would be 0.00872, saving 40% in profile drag once you took the different foil sizes into account. Using the flap to take off at a lift coefficient of 0.9 vice using incidence and no flap to take off at the same lift coefficient (same sized foils) would save 35% in profile drag.

Of course, the induced drag will be big contribution. But if you use full span flaps, the induced drag of the flapped and incidence cases will be pretty much the same.

Judging by the photos I've seen, the pitch attitude seems to be high. So I suspect the main foil is undersized. A bigger stern foil won't help because the load is being taken by the forward foil, and it has to trim at the same angle of attack to carry the load. Increasing the span would significantly reduce the induced drag at takeoff, too. So you might try a longer main foil and see if that is a worthwhile trade of takeoff performance versus extra drag at high speed or when hullborne.

 

p.s. If one were designing a section to be used with a flap, it would be easy to reshape the leading edge to eliminate the suction peak you can see there with the large flap.

And if you put the hinge at the bottom of the foil, and have a radiused section on the leading edge of the flap that slides inside the main foil upper skin, that would form a smooth, rounded upper contour at the hinge line that would significantly reduce the size of the suction peak you see there. That would help with the drag and allow you to use more flap or angle of attack before separation began on the flap.

 

Ian

Sorry I missed you in Melbourne. I will let Tom answer your question from a theoretical position. The aircraft anaology may be miss leading, some flap for take off and more to full flap for landing is driven by drag /speed profile considerations and to some extent the improved view of the ground with flaps down. John Ilett's observation is factual, trailing edge up with the wand out of the water. However I suspect that the flap hinge moments are that high that while the foil is in the water the trailing edge may not be sugnificantly up at all. A heavier lower drag wand may help the situation so the wand stays deeper in the water and is capable of applying sufficient trailing edge up hinge moment.

Unfortunately I was unable to get down to Blackrock on any of the strong wind days but I know that area very well having sailed 14 foot skiffs from Blackrock for 25 years. I was amazed that Rohan and some of the better guys could successfull foil to windward in that sort of steep chop. Many is the time that we had the fourteen full airbourne in 18 to 20 knots from the southwest and thats straight dynanics , not foils.

 

Moths/ foils

John Ilett told me that the main foil on his system is set to 0° angle of incidence with respect to the static waterline(rudder foil at zero as well-but symetric I think). From the nose up way the boats sail especially upwind and in lighter wind it appears they are sailing with a foil angle of attack between 2 and 4 degrees.
That means with the flap in neutral-I think you can assume more or less neutral because the boat is in a steady state-neither rising or settling.
According to Abbott and Doenhoff then the foil(63412) is sailing within the "drag bucket"
with CL between .55 and .77 .
From what I can tell based on Johns foil area of 1.03 sq.ft.(main foil Rohans boat) at 15mph boat speed the foil is still sailing with almost one degree angle of attack but at 20mph boat speed the area is too great by about 20% meaning the flap will be deflected up and at 25mph the area is too great by 50%+ -meaning more "Up" flap (downforce) to keep the boat in contact with the water.
Wouldn't it be possible to literally "shift gears" at around 15mph boat speed by manually adjusting the board in the trunk with some sort of robust lever arrangement that would reduce the angle of incidence of the foil relative to the boat to eliminate "Up" flap? (rudder t-foil stays the same).Just as an aside, from what I've read some of the guys tilt the whole boat bow up to aid in takeoff.
It seems that because this foil(63412) is still producing lift at 0° angle of attack that a better idea might be to use a symetrical foil of, perhaps, slightly more area that would start off at an angle of incidence relative to the static waterline of 2.5° then at the right time ,say as speed approaches 15mph be 'shifted' to a lower angle of incidence to allow higher top end speed than possible with the 63412 section.Or at least lower drag at speeds of 15mph and above.In 'shifting' the rudder t-foil would not be changed. Does that make any sense?
--------
Note that 'shifting' would be done underway with an altitude controlled system(wand). The wand wouldn't be affected by shifting the daggerboard/foil angle -it would just continue to control the flap based on the preset altitude.

 

There are other tricks to improve the hinge knuckle problem. In general, a concave corner is far more benign than the convex corner. By pre-setting concave corners at the hingeline, the severity of the convex corner due to flap deflection is reduced. The first figure shows such an airfoil (for a large RC glider). The top surface is smooth with 0 flap, while the bottom surface is smooth with -5 flap. This considerably increases the usable flap deflection range, since the convex corners at the extreme +5 and -10 flap settings are alleviated. The other figures show that the Cp spikes at the hingeline are modest at these flap settings. This produces a very large CL range, as the polars indicate (this airfoil is really intended for Re=200K and Ncrit=9, in which case the Cd creep at high CL does not appear).

This approach to flapped airfoil design is even more effective at very large Re. I've used it with good results for AC keel airfoils, which operate at Re=2-6M. The intentional convex corners can be made more pronounced, and usable low-drag flap ranges up to +/-12 deg can be achieved.

 

One other benefit of the convex corners is that they make fairing of the hinge gap unnecessary. With the flap at neutral, the hingeline has a slight convex corner on each side, which gives a relatively small local velocity. Since excrescence drag scales as the local velocity^3, a slight glitch at the hingeline is relatively inconsequential.

 

hinge

Mark, how would you actually build such a hinge?