Foil design for windsurfing

Discussion in 'Hydrodynamics and Aerodynamics' started by markmal, Aug 21, 2017.

  1. markmal
    Joined: Nov 2012
    Posts: 15
    Likes: 0, Points: 1, Legacy Rep: 10
    Location: Toronto

    markmal Junior Member

    I see most of famous and knowlegeable people are in Foiler Design thread.
    There are mostly discussed big foilers that have mechanical control: flaps, incidence control etc.
    I do not think I should post my questions about a small foil without moving parts there.
    This is why this new thread is created.

    There are fair number of foils for windsurfing on the market today.
    I assume they have done big RnD. However their foils are different.
    Some have anhedral, some do not, some have dihedral with tips bended down.
    Some are low aspect, many today are high aspect.
    Plan forms are different. Some even have negative sweep angle.
    Section of profiles are secret, but I assume they are also different.

    I have foil for kite. However the sizes of its wing an stab are not suitable for windfoiling (foiling on windsurf board).
    I've made my simple "delta" foil with primitive profile: flat bottom, rounded top with max thickness 11 mm at 1/4, rounded leading edge.
    It is not a NACA-something, all done by eyes and hands.
    Total span 75cm, mid chord 25cm, tip chord 10cm. Thickness decreases to tips but I think not proportionally.
    I think at tips there is more camber and relative thickness.

    It is OK. I can fly, it is roll stable enough, I can easily accelerate and take off.
    (here it is how it looks like , . Blue arrow on dashboard shows wind direction and speed from local weather station.)

    It works well from 10kn (gusts) for acceleration and taking off (with pumping actively), no pumping acceleration needs ~12 kn wind, and to fly it needs like 7-8kn of wind.
    But I feel increasing drag with speed above 20 km/h (13.5 kn) that prevents further acceleration in light winds.
    I can reach 30 km/h only in high winds like 18-20 knots.
    I've put my foil into XFLR5, yes it has quite a drag especially at angles >2, where it operates to carry 110 kg of board and me.

    I want to make my another foil main wing for light winds (8kn +-3kn), that is not so draggy, also has good roll stability.
    It should have good Cl on low speeds, starting taking off somewhere at 15 km/h (~4kn)
    And have low drag bucket up to 30-35 km/h (I do not think I could go faster than wind speed times 2.5).

    So far I found that most suitable section for my application is H105 (by Tom Speer).
    I believe that span 0.9-1m should be enough, and chords should be 14-7cm.
    If it is not enough I can increase a little bit span, and mostly chord.
    I think span more than 1m will create problems with controlling roll (using my foots).

    Question #1 is twist.
    Most of airplanes use negative twist so stall happens in the middle not at tips, improving roll stability at low speed.
    I my case low speed is when I accelerate on surface of water. Flat board touches water and provides with forces that stabilize the system.
    After I take up to speed that is enough for take off (~18-20 km/h) I start gradually taking off.
    So I do not need that property when I am at cruising speed.
    Should I use twist, how much, will negative twist decrease roll stability? May be it is better to use some positive twist?

    Question #2 is anhedral/flat/dihedral.
    What is better for windfoil? Kiters use anhedral so tips do not pierce water surface when they tilt foils at 45 dgr or more.
    In windfoil we do not tilt/roll that steep, may be 10 dgr, 15 max. So piercing is not a problem.
    Does anhedral provide more Cl at lo speed and less Cl at high speed?
    Does anhedral provide more roll stability?
    Does dihedral provide more roll stability with center of mass 2.2m above the foil?

    Question #3 is winglets.
    Many foils have anhedral bending further down at tips, forming winglets.
    Some foils have dihedral wings with tips bended down, forming winglets.
    Airplanes have winglets bended up, increasing Cl at cruising speed.
    In windfoil we do not need increased Cl with cruising/high speed. We need less. I believe. Unless we go with negative AoA. Am I right?
    Will downward bended winglets decrease Cl at high speed low AoA, while increase it at low speed, high AoA?

    Thanks to every one who has read this, and double thanks for replies and advice.
    Last edited: Aug 22, 2017
  2. Skyak
    Joined: Jul 2012
    Posts: 1,429
    Likes: 123, Points: 63, Legacy Rep: 152
    Location: United States

    Skyak Senior Member

    Well, it appears that the Doctors are out and I like to think that this forum should provide prompt service to proven builders so I'll take a whack at it. If I am wrong I trust that experts will come out of the woodwork to correct me.

    #3 Winglets don't provide anything more than a similar sized extension of the main foil would -they have lower lift and higher drag. Winglets can be used where span is limited or you want less rotational inertia. For a low speed takeoff, span rules.

    #2 anhedral/flat/dihedral My vote is to keep it simple with flat. My reasoning is that angling will couple slip in different directions with torque in another axis. If there are any changes you want, you won't know what to do. Flat is easier to build and easier to predict force diagrams.

    I am not saying it is not possible a bend would improve some dynamic response, I am saying it is difficult to find an advantage and even harder to debug problems.

    #1 Twist My advice is that you forget about twist because it is too difficult to know on tangible 3D wings. You can't locate a tangible leading edge and most likely don't know the zero lift angle of the profile at any section -so twist as a concept is pointless. Think instead about the lift distribution across the span -that we know is optimally elliptic. We know smooth transition will result from smooth span development. Instead of "less twist at the tips" we would make the profile of the tips more tolerant of high angles of attack without stall (bigger nose radius, max thickness further forward).

    Well that's 3 questions answered, and with that you get one free unsolicited opinion. You plan to make a high aspect foil for low speed takeoff using H105 profile. That's dandy, but it is a difficult, 360 degree profile that is designed to minimize peak pressure. Peak pressure is critical for pushing the limits of speed or loading for cavitation or ventilation. You have none of these conditions but are likely somewhat limited in build capability (I presume based on the one-off nature). Why not pick some conventional high lift profile with a flat bottom that you can build accurately in a single mold that goes from the stagnation point, around the leading edge to the trailing edge. Then just infuse a flat bottom on. It's dramatically less work, far more accurate, and higher performance over your range of use.

    When you want to make a little foil that takes off at 30+ kn to go after a speed record well over 40 kn, then you can go for the H105.
    markmal likes this.
  3. markmal
    Joined: Nov 2012
    Posts: 15
    Likes: 0, Points: 1, Legacy Rep: 10
    Location: Toronto

    markmal Junior Member

    Hi Skyak!

    Thanks a lot for your advice! It is very helpful!

    I had similar doubts regarding these anhedrals and winglets. I suspect it has ancestry from kite foils where piercing is problem, so span needs to be reduced and tips bent down. Some companies make these bent wings for windfoil because they look cool and futuristic. I suspect. :)

    I've currently chosen H105 because many Moths have it, as far as I know. Our conditions are similar.
    But I am not decided yet. I am continuing to model in XFLR5 and doing simulations with other profiles too.
    I hear what you are saying. Cavitation should not be a problem with my speed range. So I can choose something with more camber and with medium thickness.

    Regarding a mold, I do not use this advanced technology yet. I do not have CNC.
    I am going to make my wing from plywood again. I know it is not that precise, but still. To improve precision I wrote program that draws 3D wing (exported from XFLR5 as STL file) as if it is made of plywood. So I see pattern of layers and can reproduce it. It should give me some reasonable approximated precision. I hope.
    BTW, program is free (GitHub ), but I have not tested it in development of real wing yet. Just going to :) So 360 degree profile is not a big issue for me.
  4. Skyak
    Joined: Jul 2012
    Posts: 1,429
    Likes: 123, Points: 63, Legacy Rep: 152
    Location: United States

    Skyak Senior Member

    You're welcome! Always glad to help when I can.

    Kite foils look too much like candy to me. I suspect that the two largest considerations are how it looks on the sales floor and how much it looks like something you should pay someone else to make as opposed to making one yourself -lots of pointless curvature. Foiling kites are racing now and it looks to me like the nonsense is being rung out of the design -flatter, higher aspect ratio. That's why everyone loves competition -it trumps marketing fluff.

    Speaking of competition, the Moth class is fierce. I know Tom presented to the builders, but I didn't know they were using H105 -how do you know? The major difference between a Moth and your board is the flap. Their need for high lift is a very short amount of time. 90+% of the time they want low drag and they have more than enough speed. The H105 might have more positive attributes -because it doesn't have the big suction peek recovery is easy -it would be very forgiving. And by forgiving I mean that you could go a long way with the flap without too much separation. Since you have the H105 in XFLR you might test my theory (and explain the Moths attraction to it).

    There is another thing I am wondering while you have H105 in XFLR5 -does it need the curvature on the bottom side at low speed? My interest in the profile is for a surface piercing foil, so all I care about is the suction peak causing ventilation. I suspect the bottom curvature is just for drag reduction at high speed and low lift. If I am right you are going to want the flatter bottom (higher lift) for your foil -shall we call it H106?

    A mold of conic surfaces is high technology? I can't comprehend that comment, let alone accept it. I could spend the next week abusing you for saying that but instead I will make you prove to yourself that every second you spend making a mold for a foil is worth more than you can get from an hour of filling and sanding. While you have the profile in XFLR5, take one or two of the points somewhere in the first third of the suction side and move them up or down by a reasonable estimate of your sanding accuracy. That is the loss from the smallest conceivable error you will make! The H105 is likely the most forgiving, try it on a high lift profile -it will turn to crap fast. You have access to 3D printing and CNC. Everyone does!

    I haven't looked at that program you pointed out but it sounds just like the slice operation every 3D printer software does. If you see the software that does this, I am interested (and I think you will be too).
  5. mm7
    Joined: Jan 2014
    Posts: 17
    Likes: 2, Points: 3, Legacy Rep: 30
    Location: Toronto

    mm7 Junior Member

    Hi Skyak, thanks for your ideas, realy helpful!

    regarding to test H105 with a flap, I do not know how to make flaps in XFLR5. I could try to make something but how it will help me? Like to make my solid wing with kind of flap that is turned down. Kind of to increase camber?

    Moths have flap, because they use it for automatic height control, and also because they cannot accelerate actively. In windsurfing/windfoiling we often accelerating by pumping, actively waving sail and pushing board with legs. This helps us to get up to a planing/takeoff speed (5-6kn). High cambered foil would be good for this phase but it will create more drag later. So here it needs to be some balance. Or a profile that has high Cl and low drag at low speed and AoA 4-6 dgr. And low Cl at high speed, AoA 1-0 dgr that is just enough to support 100-110 kg of board and body.

    About using H105 on Moths I read somewhere here. What is in reality I do not know.

    Regarding the curvature on the bottom side of H105 at low speed, may be it would be better to ask Tom.
    I think it needs to balance upper side curvature to keep overall camber as designed.
    Making flat bottom would increase camber. Or if we keep camber, it would be needed to decrease thickness.
    It is my understanding.

    High technology means that it requires CNC to make mold. I will try again to find some CNC service, but I'm afraid that it will increase costs to a degree that it would be less expensive to buy a wing from some reputable maker, like Neil Pride / F4.
    And I think, if to go with molds it should be 360 mold. 2 pieces or even better 3 pieces. One piece for front edge that covers top and bottom, and two others for middle and rear top and bottom sides of wing. Or at least 2 side pieces then glued together.
    I think that front edge is very important and must be done with high precision on both, upper and lower sides.
    One side mold will make the foil precise on upper side. But lower side has to be post-processed, sanded and filed. Or CNC-ed?

    My Fanera program is much simpler than 3d-printer slicers. It does not need to calculate support "legs" (not the case with foils though). And it also can print patterns on paper in real size. :)

    Sorry I am not sure I understand why this board example is here?
    Is it about to 3d-print a honeycomb core for a foil and then to laminate it?
  6. Skyak
    Joined: Jul 2012
    Posts: 1,429
    Likes: 123, Points: 63, Legacy Rep: 152
    Location: United States

    Skyak Senior Member

    To make a deflected flap for XFLR5 you just rotate the points in the tail end of a copy. For small angles you can just add an increment times the distance from the flap start. I only made the suggestion as a reason the moth might use H105. If someone said the moth used H105 on this forum that would be easy to find.

    It is always better to ask Tom -it's just not better for Tom so I try to limit questions to what I can't figure out myself.

    Your problem is not that CNC costs too much, it is that you are not using a cheap one to its full capability. All the cheap 3D printers can be fitted with laser cutters. You could use one to cut two sizes of your profile representing the middle and the tip. Then you just cut the mold out of foam with a hot wire like RC modelers do.

    Just a suggestion. Do what you think is right.

    That surfboard link is to a guy that uses a laser cutter on cardboard that interlocks to make a 3D plug or core. It could also be used on wood or plastic.
    markmal likes this.
  7. markmal
    Joined: Nov 2012
    Posts: 15
    Likes: 0, Points: 1, Legacy Rep: 10
    Location: Toronto

    markmal Junior Member

    It is very nice method! Even without a laser, I can print a section on paper, glue it on aluminum sheet, cut and file to the required shape. Then hotwire a foam wing, make fiberglass mold from it. Then make shell halves and glue them together. I am afraid that foam mold will not withstand vacuum pressure.

    I am currently simulating different profiles and planforms. H105, S3010, E195, NACA63412...
    About a planform. My current wing has quite steep delta shape. It is for roll stability. Sweep works like dihedral in this regard.
    I did not want to make dihedral from plywood. Sweep is much simpler.
    What do you think is better, sweep of the front edge or dihedral, or moderate amount of both?
  8. markmal
    Joined: Nov 2012
    Posts: 15
    Likes: 0, Points: 1, Legacy Rep: 10
    Location: Toronto

    markmal Junior Member

    Simulated H105, S3010, E205, E817, NACA63412. Method 3d panels, viscous. RN 100K,500K,1M,1.5M,2M,2.5M,3M. NCrit 2. AoA: -2 to 8. Type: Constant Lift, for 110 kg mass.
    E817 failed on AoA less than 5 dgr.
    H105, S3010, E205, NACA63412 all show very close results for drag (CD).
    H105 needs a little bit more speed to support 110 kg and has visually longer transition.
    If the simulation is correct it means that profile precision does not make much difference for my lightwind foil.
  9. sigurd
    Joined: Jun 2004
    Posts: 827
    Likes: 8, Points: 18, Legacy Rep: 65
    Location: norway

    sigurd Pompuous Pangolin

    "Or a profile that has high Cl and low drag at low speed and AoA 4-6 dgr. And low Cl at high speed, AoA 1-0 dgr that is just enough to support 100-110 kg of board and body."

    Not sure you need a low drag profile at low speed, because that is when induced drag is so large.
    I think what you are saying is that you want a high max CL at the takeoff reynolds numbers and then a drag bucket which covers the speeds (meaning CL and Rn) where profile drag is actually noticeable.
    All foils can have low CL and the lift vs AoA slope is the same for each profile (0.11/degree in 2D). Just the zero lift AoA is not the same.
    Did you calculate the relative size of Di and Dp around takeoff speed?

    Here is a riddle for you:
    Say you found two foils.
    Foil A) has a max CL of 0.6 at your takeoff speed, and a L/D of 50 at CL=0.15
    Foil B) has a max CL of 1.2 at takeoff, and L/D of 50 at CL of 0.3
    Which one is better for your application?
    Last edited: Nov 13, 2017
  10. Mije
    Joined: Jan 2018
    Posts: 1
    Likes: 0, Points: 1
    Location: Alabama

    Mije New Member

    Hi Markal
    I have attempted two stabilizers with anhedral using a plywood core. Made a jig out of two 2 x 6 cut with a 8.5 deg angle and bent microwaved( steamed) plywood. Lots of spring back with loss of angle so decided to laminate thinner strips of plywood. Used 1/4 in for top and 1/8 for bottom and glued them together with a layer of 5.6 oz carbon between using West System. Really nice black cord line!!! for reference when sanding to shape. Printed mh60 sections with airfoiltools onto stiff photo paper and cut out templates at various cord lengths(3). Primitive but gets pretty close. Use a caliper to check max thickness at various cords. A Carpenters contour guage is very useful to remove asymmetry with light sanding. Lots of error from cutout and sanding. Contour gage improves precision , I think, but does nothing for the accumulated error and may just transfer error from side to side. Remarkedly stiff without laminate. Laminating without vacuum bag introduces even more error especially with stiff carbon. Very light glass would be much better. Lots of fiddeling but wing is not too bad. Have found that laminating top and bottom in two steps is best and it helps to press cloth to leading edge when very tacky. The leading edge may need a bit of sanding and fill.
    Chose mh60 because of Cm with little tendency to dive. Have wondered if similar sections might be good for main wing to prevent dive at ventilation. Also want anhedral in stabilizer and bottom mounted stabilizer to keep stabilizer under water at ventilation. First flight with first stabilizer was promising but stab angle was too low with resulting slightly deficient lift. Several ventilation episodes produced a gentile solash return to the surface and sailed away from every one. Only session so far as cold has driven me from water. Second stabilizer is curing. Hope to do a much better job with lamination on this one.
    Saw foil in Gorge this summer with a mh106 main wing that was cut with CMC from a block of G 10 like material that looked very stable with very good sailor. This is what got me interested in flying wing like foils.
    There is an article several years old that compares three Moths from three manufactures with specs and tank tests. One used the h105 and one a Naca. My windfoil has a h105 main wing and stabilizer. The countour gage can be used to copy foil section and solve the mystery of which section they are using.
    Hope some of this is helpful

  11. SVLB
    Joined: Sep 2016
    Posts: 6
    Likes: 0, Points: 1, Legacy Rep: 10
    Location: nz

    SVLB Junior Member

    I'm no expert in the field but after building several foils for surfboards/SUPs I have come to a few conclusions. dihedral is the absolute worst for a hydrofoil unless you design in some serious washout (twist) as the spanwise flow on the upper surface of the foil accelerates greatly as it heads up and outboard, also any angle of incidence in the outboard sections of the wing combined with dihedral work to cause a large 'hole' to be left in the water behind the tips thus creating huge tip vortices, a large overall wake and lots of drag. with anhedral the spanwise flow and tip vortices get directed downward and actually increase the lift, this effect was noticeable after making changes as small as changing the tiplets on a dihedral foil from curving up to curving down (the reason aeroplanes have upcurved tiplets is becuase if curved down they would hit the ground on rough and rolly landings). If anything I would shape a flat foil, cut 200mm off each tip and glue them back on angled down a few degrees, look at the front on picture of a great white shark or something and copy his anhedral angles.
    here is also a couple of my latest SUP foil shapes, there has been no numbers of calculating, only an eyeball and experience. the criteria is the mythical low speed, low drag and high lift

    Attached Files:

Similar Threads
  1. hashtag_laeuft
  2. 14berlin
  3. RumnCoke
  4. Ale98
  5. schakel
  6. Doug Lord
  7. ltp
  8. radityonugra
  9. Rohan
  10. radityonugra
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.