Foiler Design

Discussion in 'Sailboats' started by tspeer, Nov 12, 2003.

  1. markdrela
    Joined: Jun 2004
    Posts: 307
    Likes: 30, Points: 28, Legacy Rep: 324
    Location: MIT Aero & Astro

    markdrela Senior Member

    For the glider airfoil in the picture, a simple bottom skin hinge would be used. Most gliders have a simple V-notch cut into the top, with the point of the V-notch at the bottom hingeline. This allows upward deflection of the flap. Some gliders have fancy wiper seals, but these don't seem to improve the airfoil noticably. I've verified that laminar flow persists behind the small open gap, at least for Re=100K or less.

    For the symmetrical AC keel, the flap has a partial-circle D-nose which nestles into a partial-circle C-recess at the back of the front element. The hinge axis is obviously at the center of the circle. The hinge bearings are held by lugs which cantilever back from the front element, and sit in slots cut into the D-nose of the flap. For the keel sections I've designed, I did not specify any special wiper seal on the surface. All you need is an internal wiper seal to make sure there's no flow leaking from the pressure side to the suction side. That would be bad. Tripping at the slight gap is not an issue, since the flow is always turbulent by the time it gets to the hinge.
     
  2. Doug Lord

    Doug Lord Guest

    hinge

    Mark,thanks! My main foil was built with a solid but flexible kevlar hinge recessed about .0938"from the top surface.The back end of the forward part of the foil is square and the front end of the flap is cut at an angle from the hinge back so that the flap can go down 20 degrees.When the flap is at neutral there is a gap of about .625"on the bottom side and a .0938" deep X .25" notch in the top. In normal sailing "up" deflection would be very minimal ; down deflection of 15-20 degrees for takeoff-controlled by a wand(or manually). How does this sound?
     
  3. markdrela
    Joined: Jun 2004
    Posts: 307
    Likes: 30, Points: 28, Legacy Rep: 324
    Location: MIT Aero & Astro

    markdrela Senior Member

    I think you don't have enough camber. You should have significant negative flap in the high-speed cruise configuration, like -5 degrees say. If the airfoil is designed for this (e.g. smooth bottom surface), there will be minimal drag penalty. You will then be able to get a larger CLmax, with a more modest positive flap deflection.

    The low-CL high-speed case is the most demanding for hingeline surface quality. Therefore the hinge wants to be right at the bottom surface, so that both the top and bottom gaps are minimal with a negative flap setting. A top surface gap with positive flap is mostly irrelevant, since at high CL the BL will be thick and turbulent at the hingeline.
     
  4. Doug Lord

    Doug Lord Guest

    hinge

    Thanks again, Mark!
     
  5. tspeer
    Joined: Feb 2002
    Posts: 2,319
    Likes: 303, Points: 83, Legacy Rep: 1673
    Location: Port Gamble, Washington, USA

    tspeer Senior Member

    I don't think the flap deflection has anything to do with whether the boat is in a steady-state condition or not. The flap deflection is controlled by the wand, and is only determined by the flying height. In the photo, if the flap was deflected upwards, it was because the boat was trying to fly out of the water!

    There's only an indirect correlation between flap angle and speed. Let's assume that the wand is set so the flap angle is zero when the boat is at its design flying height. With, say, a level pitch attitude, the lift on the foil will only be correct at one speed. At higher speed, that same angle of attack will produce more lift, which is more than the weight being carried by the foil. As the boat rises up, the wand will deflect the flap up to counter the rise. The boat will trim out at the height that corresponds to the flap deflection that brings the lift back down to equal the weight.

    The problem with this is speed and height become coupled, and the flap is not powerful enough to counter substantial changes in angle of attack due to changes in pitch attitude. The wand does a good job of regulating the height once the boat is trimmed, but asking it to do the trim function, too, is using up a lot of its control power.

    Adjusting the incidence of the main foil would be an option. For example, the incidence and flap deflection could be linked together like an antiservo tab. This would reduce the incidence at the same time it deflected the flap up. One problem with this is it increases the hinge moment compared to just using incidence control alone.

    I think the answer is to start controlling pitch as well as heave. The aft foil could be given a flap of its own that the pilot could use to adjust the pitch trim, and thus the flying height and main foil flap deflection. Just what a Moth sailor needs - another thing to operate! Or maybe the aft foil could be run with a wand of its own.

    I think a more promising solution is to add a forward/aft foil interconnect. The aft flap would be linked to the forward flap so that they moved in opposite directions. As the boat rose up, the wand would deflect the forward flap up to reduce the lift on the main foil. But it would also deflect the aft flap down which would reduce the pitch attitude and also reduce the lift on the main foil. The flap acts immediately, but the rudder foil takes effect a little later because it's not until the pitch attitude has changed that change in lift is seen. However, in the long term, the net effect is much more powerful.

    It would be possible to sum a manual trim input with the interconnect to the aft flap. The attached sketch shows a schematic of how it can be done. When the trim input is fixed, the two flaps move in opposite directions in response to the wand movement. If one considers the wand to be fixed in position, the trim input moves the rudder flap. Adjusting the ratio of aft flap to forward flap movement provides a means to tune the system for best performance. Naturally, Morse cables are probably a better way of mechanizing the system, rather than the pushrods shown. But the action would be the same.

    Friction and having two surfaces instead of one may be tough for a wand to drive. After all, the energy to activate the controls comes from drag on the wand and you don't want any more of that than absolutely necessary. So it maybe desirable to put some kind of hydrodynamic balance on the flaps.

    I wouldn't rule out external flap foils, either. It's a lot easier to arrange the pivot on the flap to make it balanced and it can form a slotted flap for high lift on takeoff if that's desired. At low flap angles, the main difference between an external flap and a single element section of the same total area is the drag of the brackets needed to support the flap. There may be enough other advantages to make the bracket drag worthwhile.
     

    Attached Files:

  6. koen
    Joined: Aug 2004
    Posts: 10
    Likes: 0, Points: 0, Legacy Rep: 10
    Location: the Netherlands

    koen Junior Member

    What kind of strategy is used in designing a flapped air/hydro foil? Do you start with a maximum Cl shape and see how it behaves with a negative flap setting or the other way around?
     
  7. Doug Lord

    Doug Lord Guest

    foil control

    Thanks for the ideas Tom. On the Moth with Ilett's system the flap on the rudder foil is controlled by a twist grip on the extension tiller; on the Rave you can order a lever control for the aft flap.
    On my boat, currenty, the main foil is controlled by a twist grip on each extension tiller and the rudder flap can be "dialed" up or down by a knob located on the tiller.
    I'm getting ready to convert the boat to a wand system since manual control at this point in my education is too hard to use....I'll give some serious thought to the coupling idea as well as to being able to change angle of incidence of the main foil/board combo.
     
  8. alans
    Joined: May 2004
    Posts: 50
    Likes: 2, Points: 8, Legacy Rep: 27
    Location: Australia

    alans Alan

    I have just completed a 3 degree of freedom (speed, height and pitch attitude) dynamic model which nominal represents the Fastacraft moth, foils and wand system.

    At a boat speed of 6 m/s crossing a single triangular shaped wave at 12 m/s. The boat pitches bow up nicely to climb over the wave courtesy the bow mounted wand and then heaves about half the wave height. The height apogee occurs considerably after entering the wave by which time the boat has almost passed completely over the wave.

    At first sight, a remedy is to keep the foil as deeply submerged as possible in order that it does not come out of the water as the boat traverse the wave trough. This can be achieved by changing the off set angle between wand and foil.
     
  9. tspeer
    Joined: Feb 2002
    Posts: 2,319
    Likes: 303, Points: 83, Legacy Rep: 1673
    Location: Port Gamble, Washington, USA

    tspeer Senior Member

    Step #1 is to define your requirements. You need to know the range of speeds, the lift required from the foil, and what the range of operating conditions is expected to be.

    Step #2 is to start with an existing candidate design and modify its pressure distribution at the key operating points to determine the foil's shape.

    If you are using a fully submerged flapped foil to compensate for waves, the ideal response of the boat is to fly straight and level as the foil encounters changes in angle of attack due to waves' orbital velocities. This requires the lift on the foil to be essentially constant, despite the changing angle of attack. So there are three conditions to consider: zero flap deflection at the design lift coefficient, positive angle of attack with negative flap deflection - at the design lift coefficient, and negative angle of attack with positive flap deflection - at the design lift coefficient. The range of angles of attack for which the flap can compensate for the change in lift determines the maximum amplitude of the waves for which "platforming" is possible at the design speed. Naturally, you don't want any separation or excessive drag in this range of flap deflections.

    Another use of the flap is to optimize the profile drag by centering the low-drag region on the operating lift coefficient as the speed changes. This a bit different problem, since you're not likely to have angle of attack and flap deflection going opposite ways. Instead, you'll probably want positive flap with positive angle of attack, etc. The idea here is to look at the drag polars for each flap deflection and pick the flap angle that gives the minimum drag for a given lift coefficient so that the foil operates on the outside envelope of the polar curves.

    As for designing the section itslef, like Prof Drela pointed out, it's a good idea to design the surface so that it has a smooth contour on the suction side with the flap deflected. So you'll need to concentrate on one side with one set of coordinates representing the deflected flap. Then change the coordinates to represent the opposite flap deflection and redesign the other side.

    You may need to do some touching up of the coordinates in between design iterations (Step #3). For example, when I was designing a wingmast/sail combination, the mast has to be symmetrical about its centerline, but an inverse design method usually doesn't have the capability do enforce that. And the inverse methods have a hard time capturing the sharp crease at the mast/sail junction. I usually pull the coordinates into Excel for this sort of work because it gives you the flexibility to program things. Then write out the coordinates as a text file to be read back into XFOIL and analyzed (Step #4).
     
  10. alans
    Joined: May 2004
    Posts: 50
    Likes: 2, Points: 8, Legacy Rep: 27
    Location: Australia

    alans Alan

    See my post 8:36 yesterday. My math model as active flaps on both centre board and rudder foils. Gearing and trim combinations can considerable tailor the pitch motions but the peak of the heave always occurs when the wave almost completely passed under the boat. The heave appears to be generated mainly by the main foil flap angle as a function of the gearing, not foil AofA. I will add to this post as i gain more understanding of the model results
     
  11. tspeer
    Joined: Feb 2002
    Posts: 2,319
    Likes: 303, Points: 83, Legacy Rep: 1673
    Location: Port Gamble, Washington, USA

    tspeer Senior Member

    That makes sense. You're seeing the difference in phase between the heave disturbance and the heave response. The boat can't instantaneously respond to the sharp peak in the sawtooth wave, so it's response will peak after the wave peak passes.

    If you use sine waves instead of sawtooth waves, you can put in different frequencies and see how the peak response and the timing of the peak response changes with the frequency of the wave. If you run your simulation for a while the response of the boat will settle down into a sinusoidal motion that will have a different amplitude and be lagged somewhat behind the input wave. At low frequencies, the time lag will be small because the boat can follow a swell with no problem. But a high frequencies/short wavelengths, the time lag will be a large proportion of the wave period.

    If you plot the amplitude of the boat's pitch and heave response divided by the wave amplitude, and the phase angle (phase = time lag / wave period * 360 deg), vs wave frequency, that will give you the Response Amplitude Operator for the pitch and heave response. You'll probably see two peaks in the RAO, corresponding the pitch and heave mode natural frequencies. You'll also see a rapid change in the phase angle at these frequencies. At low frequencies the RAO will be flat and the heave amplitude/wave amplitude will be near 1. But at high frequencies, the RAO will roll off and the boat will hardly respond to the wave at all.

    You can shift the natural frequencies by changing the gearing in the control system. The more aggressive your control system is (large flaps, high gearing), the higher the natural frequency will be. But it will also lower the damping and with enough feedback will make the system unstable - it will have oscillations that grow with time.

    A simulation like yours is absolutely vital to doing serious control system design work.
     
  12. alans
    Joined: May 2004
    Posts: 50
    Likes: 2, Points: 8, Legacy Rep: 27
    Location: Australia

    alans Alan

    Tom

    I have the computational tools to go directly from the equations of motion to frequency domain and or root locus formats. However I can see from my simulation that there is a dominate pair of poles and varying the gearing and position of the wand hinges moves the response from unstable to critically damped. Hence in practice, irrespective of configuration, I believe wave frequencies will be encounted where the waves and the heave motion will be almost 180 degrees out of phase. Hence my opinion that a heavier wand with a low drag provile and a wand to flap off set angle that keeps the boat low but no hitting the tops of the waves is likely to be the best oprion.
     
  13. alans
    Joined: May 2004
    Posts: 50
    Likes: 2, Points: 8, Legacy Rep: 27
    Location: Australia

    alans Alan

    P.S. the wand needs to be heavy and low drag to keep it in the water at all time, once the the high drag profile wand is out of the water it will bounce on impact possibly bad enough to drive the flap trailing edge down and that is the last thing you want.
     
  14. Doug Lord

    Doug Lord Guest

    wands

    Alan, I'm not sure you know but the wand on Iletts sytem and on the Rave are kept in/on the water by shock cord tension which is variably adjustable..
     

  15. tspeer
    Joined: Feb 2002
    Posts: 2,319
    Likes: 303, Points: 83, Legacy Rep: 1673
    Location: Port Gamble, Washington, USA

    tspeer Senior Member

    If the boat can't respond to the wave frequency, then it sounds like you ought to be platforming the waves instead of contouring them. Maybe the wand skipping over the wave tops is a good thing - it would have the boat responding to the maximum water level instead of following the wave shape! Seriously, though I think I'd prefer a lightweight wand that followed the waves with minimum drag so I'd know where the real water surface is.

    But the flap motion doesn't have to correspond directly to the wand position. You can modify the wand "signal" before it gets sent to the flap. Swaping out components in the linkage may be easier than designing a whole new wand system for each change.

    I've been wondering whether acceleration feedback in addition to the wand feedback would be useful. A bobweight could be mechanically linked to add its motion to the wand angle in the linkage to the flap. Negative acceleration feedback would make the boat appear to be more massive, lowering its natural frequency and making it less responsive to high wave frequencies.

    Another approach would be a tuned mass damper. If you put a sprung mass in series with the wand and tuned the mass to have the same natural frequency as the boat's heave mode, it would change the phase angle of the flap deflection compared to the wand angle. Since the sprung mass is in the feedback path, its transfer function will tend to be inverted in the closed loop response of the boat. So the resonance of the mass will show up as a notch in the heave response to the wand. And at the higher frequencies, the mass will roll off the input from the wand, reducing the tendency of the flap to be excited by the short wave lengths.

    The bobweight and tuned mass damper could be combined by choosing the bobweight mass (or moment of inertia) and centering spring to be that of the tuned mass damper, but mounting the weight on a horizontal arm so it responds to vertical acceleration.

    Since you have a simulation, you can place various transfer functions in the feedback path and see what kind of compensation makes sense for best control system performance. Then design mechanical components that will provide that compensation. That way you can concentrate on ideas that are likely to work, and the form of the needed compensation may give you ideas you'd never come up with otherwise.
     
Loading...
Forum posts represent the experience, opinion, and view of individual users. Boat Design Net does not necessarily endorse nor share the view of each individual post.
When making potentially dangerous or financial decisions, always employ and consult appropriate professionals. Your circumstances or experience may be different.