Foiler Design

Foil Construction

I'll start by saying the foils I've used on my 16'foiler are Fastacraft and were a custom design by me that they built very well-excellent workmanship.(see below)
I've built production RC model foils-1' span with a tapered planform similar to Dr. Bradfields foils mentioned here earlier. What I did differently to most designs I've seen is to use a center flap-the center 1/3rd of the foil span(same area as a 20% full span flap). The reason for this was to allow me to use a technique I had considered for some time: incorporating the flap and flap hinge in the layup of the foil. Since the flap hinge was to be thin mylar I was concerned that if I made the flap full span it would bind up when the foil was under load. In fact, on the Rave, Dr. Bradfield used a full span flap with a piano hinge type arrangement but the clearances had to be made fairly large to prevent binding. I wanted to come up with a quick method of construction that produced a foil ready to assemble to a strut with the flap already attached. I decided that if I kept the flap 33% of the span or so then I could use a solid state mylar hinge which incidently also seals the flap joint w/o the worries of binding.The negative to this type of arrangement, I've been told, varies from a 5% higher drag to much less than that-Tom Speer or Mark Drela could comment intelligently on that.
So I started with a .375" thick piece of aluminum cut 2 inches longer than the span and 1.5" bigger in the chordwise direction. To test the concept I made foil and flap halves out of thin plywood allowing a .093" gap between the foil and flap at the hinge; the ends of the flap touched the foil. The plywood allowed fairly accurate shaping. Note that the aluminum "plug mount" for the foil halves was cut sharp and square. Once the pieces were glued to the aluminum on each side(symetrical foil)and waxed each half was laid up with epoxy resin and glass and reinforced with a flat piece of .375" aluminum the full width of the "plug". On the first layup 6 tabs were created on one mold half so that when the second layup was complete the two pieces would match perfectly.
With completed ,matched female molds each half was laid up using uni and 45/45 woven carbon; the foil halves were then filled with epoxy and micro balloons and a piece of mylar overlapping the hinge joint by .5" with holes in it and roughed up was laid into one half.The two were then clamped together with at least 6 "C" clamps Thats why the molds were built with the thick aluminum backing plate included in the laminate-so clamping pressure would not distort the foil.
Next day the part was pulled and a small cut made at each end of the flap and voila!-the flap worked as if it was on ball bearings! Since these were small foils the joining of the vertical fin and foil was relatively simple: a 5/32" stainless rod about 2" long was molded(along with the pushrod tube) into the vertical fin and the thing assembled first with a touch of cyano and then with epoxy putty completly encapsulating the area where cyano was used.The resulting bond was quite strong-one has never broken.
During that time the biggest problem, in terms of labor intensiveness, was the assembly of these two parts ; I've done a lot of work on this and have finally come up with a solution I haven't tried yet but based on my previous experience I'm confident will work. When I can I'm going to try this the simple way and the more involved(but better) way: the best way to do this is to have a company cut three female molds 1) left side including the left half of the vertical fin and the left top half of the hydrofoil,2) the right version of the same thing and 3) the bottom half of the foil. These molds would use the same method as above with a center flap and a mylar hinge in the middle. I've talked to Vecterworks in Titusville ,Fla and they can make a "quicky" set of tooling like this out of a special material using a 5 axis router. The quicky tooling will allow a few parts to be pulled to test. Thats one way... The other way is to have an aluminum "T" welded using carefully cut untwisted aluminum and building the three molds in a manner similar to that described above. Thin plywood could be used for the "plugs" and then be glued precisely to the aluminum pieces. The bottom piece(bottom of the foil+bottom of the flap) will be difficult to position exactly because of the separation of the vertical fin aluminum "plug" spanwise but I'm confident it could be done.
What this would allow is a one piece fin +hydrofoil including the flap-all layed up at once and pulled as a single part from three molds! It woud be a great simplification of the way foils are currently built and relatively simple. One other thing: by using cyano to bond the wood foil halves to the aluminum they can be removed at any time with no damage and a different section attached to the aluminum structural part of the plug. Finish of the wood pieces is fairly simple and with the 5 axis system is not an issue.
So it's fairly simple to make a foil that includes the flap especially if you use a symetrical section; its much harder to use this system for a 63412 section although with a 5 axis machine it is no problem.
You can see the foils Fastacraft made for me on their site at: www.fastacraft.com under "aeroSKIFF". John uses a unique solidstate kevlar hinge near the top edge of the foil; my mylar hinge is in the middle of the foil....
and you can see the small foils I made using this system under the F3,"multihulls" at:
www.microsail.com

 

If you're going to make molds like these, I suggest you look into resin tranfer molding (RTM). It requires a strong female mold, typically machined from metal, and a pump that injects the resin.

A dry preform consisting of core plus dry fabric reinforcement is laid in the mold, and the mold clamped shut. Then resin is injected under pressure. Small tubes placed at strategic locations around the mold allow air to escape - when resin comes out of a tube, it's clamped off.

The cost would be high for a one-off design, but is modest for limited production run. The pumping equipment is easily within the range of small shop. The cost of the mold is probably the biggest factor.

RTM technology has advanced considerably over the past 20 years. For example, on Boeing's Joint Strike Fighter prototype, the wing spar caps and integral sine-wave webs were made using RTM. So it's possible to produce high-strength parts.

 

As for foil joints, one might consider some sort of fairing to act as a structural fillet. Dave Keiper found bullet fairings to be very effective at eliminating ventilation on Williwaw's bow foil, which probably indicates they were eliminating flow separation there.

The local velocities at the junction are amplified, so cavitation will begin there as well. In effect, the velocities from the strut and foil are added together at the junction. So not only is the maximum velocity higher, the velocity gradient coming down from the peak is also higher, leading to earlier flow separation.

One approach I've looked at for designing a fairing is to try to achieve some cancellation of the exaggerated velocity gradients in the junction area. If you consider a torpedo-shaped axisymmetric body, there will be a low pressure peak at the junction between the nose section and cylindrical center body, and another low pressure peak at the junction between the body and tail cone. If you pinch in the center body to make it more dumb-bell shaped, the pressure peaks are even bigger and there's a high pressure region in the center.

So now combine the dumb-bell body's velocity distribution with that of the foil junction. The foil is going to have a high pressure area at the leading edge stagnation line, a comparatively high pressure (a little higher than ambient) at the trailing edge, and a low pressure region in the middle near maximum thickness. By making the dumb-bell longer than the foil chord, you can position the low pressure peaks of the body on top of the high pressure regions of the foil, and put the high pressure region in the middle over the low pressure area of the junction. The result should smooth out the pressure gradients in the junction area and reduce the maximum velocity.

Finally, instead of keeping the axisymmetric tail cone, pinch it in to form sharp trailing edges that will help carry the lift through the junction region so there's not a big increase in induced drag at the junction due to a notch in the spanwise lift distribution there. In order to maintain the same cross sectional area, the pinched in parts of the fairing will have to extend farther out on the strut and foil.

This early design for a Williwaw-like bow foil shows such fairings used where the foils join at acute angles. This was just my first cut at a fairing design - I've not done any design refinement or tried to estimate the drag of the fairing compared to the bare junction. But it illustrates the concept. It may not look like it from the panel shapes, but the noses are smoothly rounded (but not circular) in cross section.

I think it's evident such a fairing would help to reinforce the joint structurally, as well as provide a place to hide fasteners and control linkages.

 

foil joints/mold technique

I really like the foil joint idea ,Tom; seems like there might be a problem making it right if the pushrod for flap control comes out of the vertical fin at the back of the joint and intersecting the flap at that point-I need to think about that.
There are a number of ways that the foils could be built using the technique I described; I don't know much about resin transfer molding but intend to learn!
The technique I mentioned when hand laid is very simple and the section shape can be changed at will. In addition a foil including a flap or a complete vertical fin/foil assembly can be built with a minimum investment in tooling. Using a version of the foil joint you suggest could be included in the plug making the resultant joint the strongest possible-way stronger than assembling two separate pieces.
Tom, what is your assesment of using a flap on the center 1/3 of the hydrofoil equivalent in area to a 20% full span flap in area? For structural integrity one half the area of the center flap would lie behind the trailing edge of the rest of the foil....

 

By "center 1/3 of the hydrofoil" I take it you mean a partial span flap, not a split flap in the middle of the chord.

I got a lesson in partial span flaps when we used one on a rigid wing landyacht rig. The problem is the spanwise lift distribution. Near the end of the flap you have a big drop in lift, which sheds a trailing vortex there and causes lots of induced drag. Immediately outboard of the flap, the local angle of attack is increased, so that part of the surface stalls before the rest. So while the partial span flap increased the lift, from a drag standpoint, it was a real horror show.

You see partial span flaps used all the time on aircraft, especially to reduce the landing speed. But in that case, the extra drag is welcome because it steepens the approach path. It effectively twists the outer wing panels because the whole wing operates at a lower angle of attack for the same total lift coefficient. So that ensures the outer wing is effective for roll control.

But a hydrofoil, especially a sailing hydrofoil, has different requirements. A high L/D is essential. So I'd lean toward full span flaps to ensure a smooth spanwise lift distribution.

You can calculate the effect of different flap configurations using my lifting line spreadsheet. Just change the zero lift angle of attack for those sections that have the deflected flap. The amount of change in section zero lift angle of attack per degree of flap deflection for different flap chords can be found in any aerodynamics text book, or better yet, calculate it with XFOIL. You'll see the effect on the spanwise lift distribution, local lift coefficient, and induced drag in the spreadsheet.

 

foil flap

Thanks Tom; you're right I was describing a partial span flap. I wonder if small endplates at the ends of the flap might help. It seems worth considering given the structural advantages of this flap configuration. What do you think?

 

Small end plates will help a little bit, but not change the situation materially. There's going to be a big drop-off in lift, and that's going to create more drag. The drag comes fundamentally from the whole wake - you're deflecting the flow encountered by the flap more than the flow outboard of the flap. End plates aren't going to change that situation.

Look at it this way. You have the basic lift distribution for the un-deflected flap, so leave that alone. When you deflect the flap, it's like adding an extra foil with the span of the flap. Since the flap span is 1/3 that of the full foil, the lift produced by the flap is going to have roughly 9 times the induced drag of the same increment of lift produced by the full foil span.

Now maybe almost an order of magnitude increase in the induced drag of the flap lift is an acceptable penalty for the craft as a whole. If there are structural advantages, then it may allow you to increase the span of the foil itself and gain a net reduction in drag. Maybe the flap isn't deflected very much, or for very long if you're only using it for dynamic changes and not for trim. There are lots of good reasons for doing it, but you're going to take a hit to the drag if you do.

 

foil flaps

Thanks, Tom! Appreciate your insight....
------------------
This foil(symetrical version with tapered planform) has been used for some time on an rc model foiler with variable flap deflections around five degrees except at takeoff.The model takes off in 5-6mph of wind and the main foils are preset at +2.5° angle of incidence relative to the static waterline(which is nearly parallel to the flight waterline)-the flap deflection varies constantly since the foils also develop the RM for the boat; rudder foil at 0°. No deflection of the flap on the rudder foil...
An asymetical(63412) version is also being used on a 16' monofoiler but I don't have results that mean anything yet.
Hope to do some two boat testing in models and/or small monofoilers down the line...
------------------------

 

Moth + passenger

Check this pix out from the Moth website. Shows a Moth being sailed in 10k breeze with a passenger-a kid being taken for a ride. Interesting thing is that it really raises the foil loading but still she foils in relatively light air. Kid looks petrified to me..
http://www.moth-sailing.org/pictures/singapore/robosan.jpg

 

Mr. Speer, what's mo' better, a symmetric foil at some angle of incidence or a foil with a flexable (deflectable? deflexible? reflexable? deflectabulistic?) tail? (like a trim-tab)

Yokebutt.

 

flap vs incidence

YB, Tom answered that question on page 22 post number 324........

 

The best answer is to download XFOIL, and possibly Profili, and do your own comparisons vs your requirements.

And remember that the section design really has a comparatively minor contribution to the overall performance. Span, foil arrangement, control mechanization and operation can all have a much bigger impact.

 

Cp /volume distribution in foiler designs

I'm curious what designers such as Andy , Ian Ward ,John Ilett , Tom and others think about the Cp/ volume distribution for foiler designs that takeoff at speeds of 50 to 100% above "hull speed"? I'm assuming a fairly high beam to length ratio hull and I'm not completely satisfied that a low Cp is the way to go; any thoughts?

 

The trick is to get maximum speed in displacement mode as quickly as possible which then becomes the takeoff speed for foiling.

Prismatic coefficients of around .56 are fine for displacement hulls operating below "hull speed". V/^L =1.3 for a Moth gives a hull speed of around 4kts but in practice, normal upwind speed is around 6.5-7kts. These boats easily exceed theoretical hull speed and have little or no planing hump.

At these relative speeds, which are easily and consistently maintained, it seems best to have much higher prismatic coefficients. Today we use values around 0.68 or even higher. It seems that this is the best way to go for foiling designs. It may of course be a different story for foiling catamarans, I have not looked at this.

 

Cp /beam to length

Thanks, Ian; what kind of beam to length ratio do you think is optimum with that Cp? I've seen discussion where it seems a double ended hull might be appropriate to help with early take off since it would allow the whole boat to be tilted bow up with a fairly low impact drag wise; whats your take on this?