Back to Dave Culp SpeedSailing

"One Oar in The Water"

How I'm Spending My Summer Holiday
By Dave Culp

First appearing in AYRS #112, 1993

I've built a new boat. It is an aerodynamically balanced hydrofoil with automatic two axis control via surface sensors. It flys on a single hydrofoil (thus the title above), and uses aerodynamic elements to supply three axis control and overcome both heeling and pitchpoling moments from the conventional catamaran rig. The basic boat was designed by Greg Ketterman, designer/builder of Longshot and Trifoiler. My input was to do the construction design, subsystem design and actual construction The hydrofoil and some substructures were built by Larry Tuttle of Santa Cruz, California. Larry built the foils for Longshot and all Trifoiler prototypes.

The new boat is powered by conventional soft sails (no kites this time). It is innovative in that it uses only one hydrofoil; an inverted "J" foil similar to Longshot's. The boat gains three axis stability when flying through the use of aerofoil elements. Pitch, roll and heave are auto-controlled via surface sensors and yaw control is pilot induced via a bow mounted air rudder.

The boat is a 'one way' proa. Though it sails quite happily on the 'off' tack, it can do so only when hull-borne. The pilot sits in the windward ama, fully 24 ft. to windward of the main hull and rig. The main hull is 22 ft. long (plus an 8 ft. sensor arm) and the boat is 26 ft. wide (plus 8 ft. overhang at the canard wing) overall. The masthead is 26 ft above the deck and the mainsail (a stock Prindle 16 catamaran main, but set on a beefier cut-down Prindle 19 mast) is 170 square feet (sf). The boat carries an additional 32 sf. in the air rudder (jib?), and 128 sf. in horizontally mounted airfoil elements. All aerofoils are symmetrical sectioned rigid wings.

Here's how the auto-controls work: First roll control: There is a 4 ft. by 16 ft. wing element, mounted on and free to rotate about, the cross beam. Its center of effort is 15 ft. to windward of the main hull. This wing is actuated by a leading edge mounted surface sensor on an 8 ft. arm. This sensor gives the wing a nose up attitude when hullborne and a nose down attitude when the windward ama rises too high.

At low speeds, the upward lift from the wing helps ama lift-off. At higher speeds, downward lift from the wing counteracts heeling due to sail forces. Greg's VPP program indicates that best speed (at highest efficiency) will be achieved when this wing is nominally not loaded, either positively or negatively. The aerofoil elements aren't meant to carry significant load at speed (too much induced drag). Their main function is to auto-control heeling (and pitch), allowing the pilot to keep sail power 'full on' and concentrate on course keeping. Greg credits this auto-control with his successes with Longshot. We designed the rest of the boat's dimensions and weights around this parameter. The wing does see both positive and negative transient loads, of course, as the boat and pilot respond to wind and wave. The net design goal, however, is no lift.

Second, pitch: Greg has come up with a rather clever approach here. The main (only) hydrofoil is positioned well aft on the main hull, under the sail's center of effort. It is aft of the main hull's center of gravity, but coincides with the boat's overall C of G when the ama is flying. The foil actually carries 98-100% of the boat's weight at speed. There is a canard wing at the bow of the main hull (actually two wings--one on either side of the bow--but cross linked to move as one). The canard's center of lift is 16 ft forward of the hydrofoil. This wing is actuated by a second surface sensor, also on an 8 ft. arm. (Both sensor arms are somewhat flexible, to attenuate the sensors' being buffeted by small waves.)

The hydrofoil is permanently set at a slight positive angle of attack (it is also asymmetrical, using a NACA 63 series low-drag section), but at hullborne speeds, its lift is insufficient to raise the boat; also drag is fairly low. The aerofoil canard has a pre-set positive angle of attack set by the sensor. When boatspeed and thus apparent wind is sufficient for the canard to lift the hull's bow (we want about 12 kts boatspeed and 18 kts. apparent windspeed at this point), the bow-up hull pitch angle adds to the hydrofoil's angle of attack and the hull lifts out. If the bow rises too high, the sensor calls for a negative attack angle on the canard and the bow comes back down. The sensor thus controls the canard's attitude, the canard controls the bow's altitude (and thus the hull pitch angle), and the hydrofoil 'slaves' along after, doing all the real work.

The advantages here are several: 1) The highly loaded main foil doesn't need to be actuated and is rigidly bolted to the hull. 2) The main strut is vertical and thus resists ventilation. 3) Only one surface piercing strut minimizes spray loss and ventilation sites. 4) Wetted surface is minimized, in this case, exclusive of the sensors' 'footprints,' wetted area is about 3.73 sf. Third, yaw: Greg has specified an air rudder in order to reduce wetted surface and induced hydrodynamic drag. His VPP shows that aero drag at speed will be less than hydro drag of an equivalent water rudder.

It is significant to note that all aerofoil elements are providing minimal lift and drag at top overall boat efficiency. The sensors are contributing less than 10% of the total drag, and that designed boatspeed is 3.1 times true wind speed (46.8 kts boatspeed in 15 kts true windspeed). Lest one suppose these predictions are too extreme, I should note that Greg degraded efficiency figures from those used for Longshot. Foil L/D suppositions are from empirical data taken from in-the-water boats using very similar foils. A similar VPP run on Longshot predicted 2.3 times windspeed at 15 kts true and the boat has been measured at 2.5. Greg actually thinks that these figures are conservative.

Results to date: First, the boat is heavy. The VPP supposes the all-up weight with pilot to be 480 lbs, of which 280 is in the ama. Actual all-up weight is about 555 lbs, with 290 in the ama. This will surely increase take-off speed and lower top speed, but very little.

Construction went well. The ama is built of foam sandwich with 3/8 inch Kleegicell, plus one 8 oz. layer E-glass/polyester inside and two layers outside. It weighs less than 45 lbs empty. (Greg Ketterman has developed a very 'quick and dirty' one-off method for getting out foam sandwich hulls, and I've simplified it again. The ama is 11 ft long, by about 24 inches in cross section. I built in for about $125 in materials and not 50 hours of work. I'll try to write a future article about the technique.) The main hull is the weight culprit at 150 lbs. It is 3mm plywood over 12 x 40mm softwood stringers. It is covered with 2 layers of 4 oz E- glass/epoxy. The after third of this hull has an additional 3 layers of 8 oz glass set at +/- 45°, to resist torsion loads between the foil and mast socket. In addition, this hull has an interior strut and jackstay consisting of a 40mm x 75mm wooden compression strut 16 ft long under the deck and a doubled 5mm stainless stay from the forestay chainplate, under the mast socket, up to the main sheet chainplate. All this is to resist excessive bending of the hull due to mast compression. We anticipate sheet tensions of about 900 lbs and mast compression of over twice that in 50 kts of apparent wind.

The aerofoil elements were semi-mass produced, all 5 identical. They are 8 ft long and 4 ft in chord and use a 10% thickness/chord ratio NACA 00 series section. There are two elements coupled together in the cross beam wing, two in the canard, and one is the rudder. They are built of aircraft Dacron heat-shrinked over wood frames and finished with butyrate dope. Torsional rigidity is through Kevlar tows laid on diagonally under the fabric skins. They weigh just 16 lbs each. The supporting aluminum framework and spars account for the remainder of the all-up weight of the boat.

If I were to do it again, I'd make two changes. I'd build the main hull of foam sandwich also, eliminating the strut and jackstay in favor of additional glass thickness. We thought the plywood hull would be quick and cheap; it was neither, and heavy. Second, I would skin the aerofoils with 2mm foam and 'glass them. I had anticipated doing this on the second generation aerofoils (after expected destruction of the first set), but I wish I had done it originally. They would be heavier, but tougher.

The boat is complete and in the water, but we've only managed about 1 1/2 hours of sailing time this year, and all in winds under 12 kts. The boat is going through expected teething problems. The over square (wider than long) and asymmetrical geometry create helm balance challenges. The helm changes quite significantly from port to starboard tacks and also from hullborne to foilborne attitude. The boat has not yet flown and I expect it will need another season's tweaking before we get it right. Time and money considerations have limited sailing time this year. Nothing has broken yet and the boat sets up rather easily in about 1 1/2 hours with 3 people.

I welcome correspondence and criticism. Please e-mail me at:

Back to Dave Culp SpeedSailing