Foiler Design

Terry,

I did wade through all the posts and I think you covered all the high points on what needs to be accomplished. I think I am agreeing if I say that in the early pages people were misapplying aircraft stability concepts. The system has to be active when you are using a submerged foil and analyzing stability has to include that active system. I think Tom Speer, and/or others, posted some good, but pretty theoretical stuff on control system analysis and requirements.

I'm not sure it's optimum, but you probably know that a wand controlling the flap on the DB main foil seems to be becoming something of a standard. I think this does have the advantage of doing a good job of keeping tight control of flying height. Also, at higher speeds the same flap deflection creates more force, so the system gain tends to go up with speed, which is good.

I think your ideas are creative, but I worry if they have enough control authority. If you look at the skipper moving, waves and the widely varying forces from the sails, I think the control system has to generate quite a bit of force pretty quickly.

If you would like to see what I waded into on my first post on SA, you can look here:

http://forums.sailinganarchy.com/index.php?showtopic=100326&hl=moth revolution

I learned that "revolution" was a loaded word for some people.

Peter Raymond

 

Above or below is irrelevant. What matters is the nature of the surface. Solid ground presents one kind of boundary condition, while the compliance of the free surface of water at high speed presents a different boundary condition.

There are two aspects to consider. One is the effect of a free surface, which is mostly a performance issue, although Russian hydrofoils designed to operate on the flat water of rivers have used the drop-off in lift near the surface to help regulate height.

The other is the precision necessary to avoid broaching the surface. Flight path precision is a matter of specifying the the performance of the control system, and this is no different from aircraft.

Analytically, both kinds of surface condition can be handled with the method of images. The only difference between the solid surface and the linearized high-speed free surface is the sign of the singularity strengths used in the image system.

This is another precision of control issue. There is a quantitative difference, but not a qualitative difference.

That's not quite right. The rear surface must be more lightly loaded than the forward surface for static pitch stability. The incidence of the rear surface will depend on the camber of both surfaces, downwash from the forward surface, etc., and could be less than or greater than the forward surface. And stall doesn't enter into it. Even in the linear lift range, far from stall, the rear surface needs to be more lightly loaded for pitch stability. The relative loading of the two surfaces ensures that for a positive change in angle of attack there will be a negative change in the pitching moment.

In addition to the static pitch stability, there is an additional static stability criterion that has to do with pitch-heave coupling. This criterion requires that a positive change in height (at constant angle of attack) generates a negative pitching moment. Another way of stating this is the "heave stiffness" of the forward foil must be greater than the heave stiffness of the aft foil. There are several ways of achieving this. One way is to use a surface piercing forward foil and a fully submerged aft foil. Hydroptre is of this variety. Another way is to use height feedback to the forward foil, ala the Bradfield system used on the Rave and Moth classes.

The theory of stability and control of flying vehicles is quite mature. The International Hydrofoil Society publishes CDROMs with hydrofoil technical literature. CD#1 has a hydrofoil design manual by Hydronautics that lays out the theory of hydrofoil stability and control. There are also NACA reports on hydrofoil dynamics. Joddy Chapman has also published their analysis of sailing hydrofoil dynamics. The theory of foiler flight is out there, if one researches the literature.

The difficulty is not in the understanding of the dynamics or solving the equations of motion to determine the stability or transient motion. The hard part is coming up with the right numbers for the stability derivatives that accurately represent the characteristics of the configuration. This is why companies spend big bucks on wind tunnel and tank tests to get these numbers. Despite the expense, it is still cheaper than experimenting at full scale for large vessels. For small sailing hydrofoils, it's just the opposite - it's cheaper and faster to use cut-and-try methods than to use sophisticate analytical methods. Nevertheless, the Chapmans found a simple dynamic analysis gave a lot of insight into the problems and made it much easier to find a solution.

Fly-by-wire control does not ignore the requirement for static stability. Instead, FbW augments the stability using feedback control. The mechanical feedback of height to flap from the wand of the Bradfield control system augments the heave stiffness of the forward foil so that it meets the requirement for stable pitch-heave coupling. Instead of losing lift due to loss of area as the craft rises, as would be the case for a passive surface-piercing foil, the lift is reduced by the flap deflection. The effect of the feedback can be analyzed as an equivalent characteristic of the foil itself

The problem with this approach is control of height using pitch control has a great deal of time lag. This can destabilize the system and it reduces the bandwidth and precision of the height control. To see why, imagine how the system has to respond to an error in the flying height. In order to return to the correct height using pitch control, the elevator first has to be deflected in response to the height error. Then the craft has to accelerate in pitch, achieve a pitch rate, and finally get the desired change in angle of attack to change the lift on the forward foil. Then the pitch rate has to be arrested and pitch attitude maintained as the craft glides or climbs back to the desired height. The pitch control has to start leveling out before the desired height is reached. If the pitch control isn't changed until the height error goes to zero, the craft will overshoot the desired height until the height error is sufficient in the opposite direction to change the pitch control enough to level out. But now there's a new error in the opposite direction, and the whole process starts over again.

Direct lift control for regulating height does not have to wait for the pitch attitude to change in order to generate a change in lift. The vertical acceleration is immediate, as soon as the feedback of height error moves the flap. Flap deflection is reduced as the craft approaches the desired height, again without having to wait for a change in pitch angle.

Control of height with pitch does have its place. But it is best used for low-frequency trim to optimize performance, rather than high-frequency regulation of height due to waves and disturbances. Direct lift is limited to the flap deflection available, and this is soon saturated. So direct lift should be washed out with time and the steady-state equilibrium achieved with pitch control. This returns the flap to its neutral position, making maximum control power available for transient height regulation. Washout of the flap can happen naturally through the dynamics, as pitch trim changes the lift. Because the lift has to equal the weight in the steady state, as more of the lift is produced by the change in angle of attack, the feedback to the flap will naturally reduce the flap deflection.

This is why it is useful to have a flap or incidence control on the stern foil to adjust pitch trim, in addition to wand feedback to the forward foil. Because it is a low-frequency function, it can be manually controlled.

In a purely automated system, one could get the same effect by using proportional control from height feedback to flap deflection, but integral compensation from height feedback to pitch control. Mounting the height sensor forward of the forward foil adds lead to the pitch control that improves the damping of the system, and anticipates the change in height at the foil as waves approach.

The ideal split between the load carried by the forward foil and the aft foil is a performance issue that depends a great deal on the combined spanloading of the two surfaces. Although the aft foil has to carry less load per unit area than the forward foil to be statically stable in pitch, that doesn't mean it can't carry the majority of the total load if it is larger than the forward foil. Good examples of this approach are Don Nigg's Flying Fish, the Miller foiling sailboard and Sam Sutt's hydrofoils.

The pitching moment from the rig is an issue of pitch trim. Most of what we have been discussing has to do with stability. Stability and trim are two separate issues. Stability has to do with the changes in the forces and moments as the craft is disturbed from equilibrium. Trim has to do with the forces and moments at equilibrium.

Trimming the moments from the rig is mainly an issue of having enough control authority to handle the aerodynamic moments and still have enough control left to handle the transient disturbances and maneuver the boat.

Feedback control with integral compensation is also very useful in achieving the control deflections necessary to reach and maintain the equilibrium.

I think there would be severe performance implications from trying to produce a rig that is pitch-neutral. There are no performance penalties from moving weight aft relative to foil center of lift to trim out the pitching moments from the rig. However, aft centers of gravity make it more difficult to achieve positive static stability.

Dragging the stern can also lead to unstable pitch-heave coupling. The "stiffness" of the stern buoyancy is high compared to heave stiffness of the forward foil. This is the opposite of what's required of stable pitch-heave coupling. As the foil lifts the craft with the stern dragging, it will tend to rotate about the stern, increasing the pitch attitude and angle of attack, and increasing the lift. Which makes the foil lift the craft up even more. Which increases the pitch more, and the whole process feeds on itself. The completely stable craft would wheel-barrow, lifting the stern first and balancing on top of the foil as it took off.

However, the unstable situation of lifting by the foil and dragging the stern doesn't have to last very long. As soon as the stern lifts out, stability is restored. It's quite possible to exploit this behavior. When sailing hullborne, the foils can be set to produce a low level of lift, minimizing the drag while accelerating to takeoff speed. When the speed is high enough for the forward foil to start lifting the bow (while still carrying part of the weight buoyantly in the stern), the unstable pitch-heave coupling will cause a momentary runaway of the pitch angle, generating enough lift to raise the whole craft. As the stern leaves the water, the normal pitch stability can reduce the pitch attitude and level the craft out. This makes the boat pop out of the water when it reaches takeoff speed. The trick is to manage the transient so that it doesn't become a sudden shoot-for-the-sky-and-crash-back kind of behavior.

This is the principle behind the spring tab - another aeronautical invention whose theory is well worked out.

You might find it useful to use a spring to hold the flap down on the forward foil. This will cause it to blow up and reduce the lift at high speed, and extend down at low speed when you want more lift.

Flap control has the advantage that it shifts the drag bucket toward the new operating condition. Pure incidence control does not shift the drag bucket, and the foil may end up operating outside the low-drag region. This can double or triple the profile drag. Depending on the section, size of flap, etc, the optimum may be a combination of incidence and flap deflection. This can be done by gearing the flap deflection to the incidence and controlling the two with a single feedback.

In order to minimize the impact to induced drag, it's important that full-span flaps be used.

There have been a number of different systems used. Greg Ketterman uses planing floats to rotate the entire ama of the Trifoiler and change the incidence of the foils for height control. The Shutt strut invented by Sam Shutt uses a small planing float attached to a floating forward foil to change the incidence of the foil, and the foil then rotates the entire craft to change the incidence of the aft foil, which is the main lifting foil. The Hook hydrofoil system uses planing surface out ahead of the boat that were linked to the hydrofoils.

If one wants to use electronic sensors and powered actuators, ultrasound sensors (like the ones used in early auto-focus cameras) can look down to measure the height above water. Laser altimeters use two converging beams and the distance between the reflected spots of light is proportional to the height.

Grumman developed a system in which a vertical strut was attached directly to the flap. Drag on the strut deflected the flap down as the hydrofoil was immersed, balanced against the hinge moment of the flap. It's hard to imagine a simpler, purely mechanical system, as the flap+strut was the only moving part. No buoyancy was involved. It was purely a balance between the hydrodynamic hinge moments - this made it self-compensating for changes in the speed of the craft. There are papers describing this system on the IHS AMV CDs.

[/QUOTE]Another depth sensing arrangement not yet tried as far as I know uses hydrostatic pressure feedback....[/QUOTE]
I think you'll find this has been tried. There are a couple of problems with it. First, there's the amount of control power available from the change in pressure. You're only talking a small force unless you let the pressure act on a very large surface area.

And then there's the change in static pressure due to the speed of the craft. You need to locate the static source at a location that naturally has a pressure coefficient near zero for the range of angles of attack, etc., or else you need to measure total pressure as well and compensate for the position error at the static source.

 

USA Foil Lift

Tom, if you can w/o violating any confidence's can you say what the percentage of total displ. the lee foil carried when the boat was flying two hulls?

 

I really have no idea. I don't think it was a major portion of the boat's weight, but I wasn't involved in that part of the design.

 

------------------
Thanks, Tom. I know the ORMA tris went up to 60-70% but my guesstimate here was 15-20%. If you ever hear anything about it I'd sure like to know.

 

Peter: thanks for the inputs - I obviously need to finish reading the thread. Automatic main foil flap control is a great idea. A beginner foiler should not go out in heavy wave conditions so attempting level flight seems reasonable at that level; soft feedback seemed like it would work but I don’t have the wet-foot experience to judge. I imagine following a big wave profile at 20k might be a rough ride though ...

Tom: thank you for your detailed reply. I over-simplified by refering to imminence of stall but I did not want to be too technical. The theory of flight and foil performance is well established, but the theory of foiler height control seems rather empirical. I am familiar with control systems but the linear math used does not handle discontinuities like boundary constraints well so, as you noted it is still cheaper to build it and see. I will look for the Chapman analysis.

We will have to agree to disagree on fly-by-wire for now - I think it substitutes dynamic stability for static stability. If I am correct it could broaden the possibilities for design. Having the CoG aft of the main foil may be safest in the event the forward foil gets dry. The boat should sit back on its transom and sink quietly back to the surface rather than "going down the mine" ...

I agree with you over the potential for performance loss using sail rigs with pitch moment compensation, but some contenders for the absolute sailing speed record use inclined rigs to minimize heeling moment, so I haven’t dismissed the concept yet. I plan some experiments.

United States Patent 2749869 describes the use of a flexible bulb to sense pressure for controlling immersed hydrofoil depth but it is part of a complex system and obviously intended for highspeed power craft. I haven’t found anything relating to foilers yet. If it works it would be simple and robust, ideal for the People’s Foiler.

-Terry

 

Here you go,Terry-I think this is the right paper:
http://homepages.rya-online.net/ejcchapman/default.htm Click on :"The Rise of the Hydrofoil and the Displacement of the Hull"

 

Thanks Doug, you are helpful as always! I have put the link in my favorites folder for later study. Has there been any progress on design/build of an off-the-beach boat with retractable foils since the earlier mentions?

 

==============
My boat has had almost no progress in the last few months but the RS600FF is just what you describe and appears to be slowly growing.

pix-RS600FF with new rig and retractable foils:
(3 by Peter Chinook
"2" and beach photos from Full Force site)

 

At the weekend I went to the dinghy show in London and saw this system for controlling the foil incidence on a National 12.The boat is not a fully foiling design and this should be kept in mind.The lower fitting is fitted to a telescoping tube that can be moved by a rope purchase.The upper fitting is a Rose/Heim joint mounted on a stainless V frame that is pivoted at each end.The engineering purists might dislike the idea of applying bending loads to the upper bearing in this manner,but it probably has enough safety margin to endure for a while.

 

Very interesting! Thanks. Any new foilers-or others of note?

 

================================
Mal, has posted pdf's on SA of his "foiling laser" using a unique(?) and very interesting swept back surface piercing mainfoil arrangement off the daggerboard. Looks like the aspect ratio of the foil would decrease with speed? I wonder what the take-off speed is? Mal, have you seen pix of Doug Culnanes "diamond foil"?


http://forums.sailinganarchy.com/index.php?showtopic=105914&pid=2785378&st=25&#entry2785378 see post 35

 

Doug,

Take off speed (boat speed) should be?? similar to Moth take off speed, as per my figures submitted earlier in this thread, with about 40% higher wind speed required than for a Moth.

I have seen Doug Culnanes foil, and noted his comments.

 

That's a very severe Dihedral Angle. Incidentally in the pic it looks closer to 120 deg since Dihedral Angle is generally understood to be the angle between the normals of the 2 planes but that's a quibble. What was the year of launch?

 

Terry, Doug is gone for the weekend- I think he's had the boat for about 7-8 years but I could be wrong. I think he has found that the foils are less likely to ventilate at the large dihedral.