WingHopper · Designer · Guide · Sections · Reef · Methods · Pricing · FAQ
The wing you draw has no load on it. On the water it bends and twists, and the shape doing the work changes with speed. Seawater is over 800 times denser than air, so at 20 knots a foil carries the dynamic pressure of an aircraft wing at about 575 knots.
| Effect | Comes from | Grows with | Tips go |
|---|---|---|---|
| Lift twist | Lift acting ahead of the elastic axis | Stays about the same, since lift equals weight | Nose up, and it feeds itself |
| Camber twist | The section's own pitching moment | Speed squared | Nose down |
| Bending washout | Tips bending up on a swept wing | Load | Nose down, shedding tip load |
At low speed lift twist wins. At high speed camber twist wins. Sweep turns bending into washout on every wing that has it, whatever the layup.
| Takeoff | Top speed | Pumping | Gliding | |
|---|---|---|---|---|
| Condition | Low speed, high angle of attack | High speed, low lift coefficient | Load cycles every stroke | Moderate speed, near best L/D |
| What wins | Lift twist | Camber twist | Bending | Whatever drifts from the design twist |
| Tips | Nose up, load moves outboard | Nose down, tips unload | Swept: shed load at the peak of each stroke | Off their design angle |
| You notice | Tip stall, ventilation as the board comes up | More drag, back foot creeping in, twitchy at the top end | Soft wing feels smooth, stiff wing feels direct | Glide shorter than it should be |
| Lever | Sweep, elastic axis forward | ±45° plies for torsion | Stiffness sets feel against efficiency | Design the loaded shape, not the drawn one |
Takeoff speed itself barely moves. It is set by area and maximum lift. Flex decides where along the span the wing runs out of lift, which decides how clean the stall is.
When lift acts ahead of the elastic axis, nose-up twist increases lift, which adds more twist. Divergence is the speed where that feedback overwhelms torsional stiffness. The linear model then has no stable equilibrium. This is a reason to revise the design before building, rather than a speed to test on the water.
Where to see it. Build > Plies shows a Divergence check as a percent: the check speed divided by the predicted divergence speed. The check speed is the higher of 10 m/s and the top of the predicted operating band; read it in the tooltip. Through 75% the check shows more margin; amber above 75% means less, and red above 100% means the check speed exceeds predicted divergence. These bands describe this check, not the safety of the complete foil.
The curve illustrates the feedback with 1/(1 − r²), where r is speed divided by divergence speed. At 75% it gives 2.3× amplification. It is a simplified single-mode model, not a prediction of total tip twist or the response to chop.
The dots show example riding speeds for the Glide template at 85 kg all-up in seawater. WingHopper predicts divergence at about 49 knots with Carbon standard and 35 knots with Glass cruiser. At 20 knots those ratios are 41% and 57%; glass at 30 knots reaches 86%. The dots use those ratios on the simplified curve.
If yours is amber or red, raise the divergence speed. Measured in WingHopper:
The layup rows are the Glide template. The geometry rows are the stock wing from the effects tables on the design pages. The extra-ply example adds one full-span ±45° ply per skin to Carbon standard. It raises predicted divergence speed without changing the wing shape, but adds material and weight. Shortening span changes the hydrodynamics too. Recheck bending, skin compression and performance after any change; these gains belong to the examples, not every design.
Pick the condition that matters most, usually the glide for a pump or downwind wing and top speed for a race wing. Design the twist you want there under load, then work back to the unloaded shape. Rider weight moves the target: a 60 kg and a 100 kg rider deform the same wing differently.
The mast and fuselage flex too, and on a strut-mounted foil the root joint can add as much rotation as the wing itself.
Build > Plies solves load and twist together at 24 spanwise stations, using a vortex-lattice solve for the aerodynamics and your ply schedule for torsional stiffness. It reports converged twist, load amplification and a divergence ratio.
It does not model bend-twist coupling from an unbalanced layup, root flex at the mast, twist past about 5°, or more than one speed at a time.
Next: Finger foil