• If you are citizen of an European Union member nation, you may not use this service unless you are at least 16 years old.

  • You already know Dokkio is an AI-powered assistant to organize & manage your digital files & messages. Very soon, Dokkio will support Outlook as well as One Drive. Check it out today!


Doug Taylor's variable geometry proa from 1977

Page history last edited by Kevin 15 years, 4 months ago

 All content property of Doug Taylor.


Proa Notes


by Doug Taylor


I wrote the following in 1977 to document my experiences when building and testing an experimental proa.  The design followed about a year of conceptual sketches and ruminations.  I decided that it was time to check my excursions of the imagination with a real test platform.



May, 1977




LOA 16’-7"

LWL 16’-1"

BOA 10’-9"

BWL  1'-1"



SA 130 SF


This design is of the Micronesian proa type with the ama, (outrigger float), always kept upwind. Tacking upwind is accomplished by changing ends; (the term "shunt" has been adopted).

The single mainsail is like a genoa in that the luff contains a jibstay. The tack of the sail slides from one end of the main hull to the other on a small block pulled along a standing cable traveller. Two mainsheets are attached to the clew, one leading to each end of the hull. If a whisker pole is used, it must be disengaged when shunting.

The hydrofoil, a surface piercing foil, is carried on the weather side of the ama. In section, the weather side is an arc; the lee side is flat. Steering is accomplished by shifting the center of lateral resistance. Longitudinal placement of the foil ranges the length of the hull by pulling the ama fore and aft on pivoting float booms (aka), like a pantograph.

The hydrofoil tilts by means of universal joints at the connections of the booms to ama. Tilted inwards towards the hull at speed, the hydrofoil provides downward resistance to heeling. Let go out from the hull, the hydrofoil may provide upward lift to reduce ama resistance in light air.

The float booms pivot on the weather side of the hull, beneath the weather side deck, which acts as an upper stop to their vertical travel. Connections at the ama, as mentioned above, permit the foil to intersect the waterline at 30 to 120 degrees.





Sailed Proa this afternoon after modifying the foil tilt arrangement. Temporarily abandoned the tilt adjustment line, and lashed a prominent tilt angle, (15 to 20 degrees in).

Sailed about 2 hours. Can be steered as a regular proa with a big paddle, with the aka in perpendicular position. Steers well using only ama swing, could swing smoother, though. The akas bump along under the weather deck. Wind was about 6-7 knots. Noticed appreciable luff bend. Tack shift and clew shift sail works very well. A solid pole for foil tilt control would be good.




In the design I did not much concern myself with the drag of the ama and foil. I had thought that the ama and foil pulled in toward the bow would result in a close hauled course. However, due to the drag, the actual effect is that a close hauled course is achieved with the akas about perpendicular to the hull. It can even pinch to luffing. Pulling the ama aft of midships slightly produces a steady windward beat without use of the paddle. Pulled way aft it produces a broad reach. Steering closer to a dead run requires steering with the paddle.


The aka thump and slide under the platform when shifting. Sitting in the canoe, on the after deck, helps, as opposed to sitting on the windward platform when sliding the aka forward and aft.

The arrangement of the foil tilting lines did not work due to the pull being led to both outrigger shrouds. The ama would flop to either side rather than hold the desired angle. I abandoned this arrangement and simply lashed the outrigger to a selected angle. Lashed so, it will still slide fore and aft without appreciably changing the angle.


To shunt, I found that the best results were obtained by first pulling the ama forward (in the initial direction), and letting go the sheet before it could point up very high. Then I backed to the opposite end of the craft, pulling the sail tack line with me, securing it as soon as I sat down. Then, after setting the sheet for the new direction, I would adjust the ama for the new course. All can be done in about 20 seconds. This could be bettered with practice. The sail tack traveler works well, as does the double, two-purchase sheet arrangement.

The strongest wind encountered was 6-7 knots. The ama lifted to its ski shaped bottom but no higher. The foil maintained a steady grip on the water, and I never had to hike out to counterbalance. In this mild wind the mast did bend to leeward some. On the next occasion that I sailed I first tightened the outboard (upper) shroud one notch and slacked the inboard (lower) shroud one notch. I tightened both (somewhat leeward) stays one notch.




Devise some sort of double tracks & cars for the aka to ride smoothly under the platform when being shifted.

Devise a foil angle adjuster other than the original.

Lighten overall weight.

Put mast and hull / aka pivot locations further outboard, to windward. Enhancing working room, sail lift, strength to weight ratio of configuration?

Recut sail for 2 1/2" shorter luff.

Stronger mast for San Francisco Bay.


Note: on a subsequent sail, with a passenger and 10 knots of wind, the mast did fail. It consisted of two 15' lengths of 3" aluminum electrical conduit connected with an aluminum sleeve. It bent above the sleeve about 30 degrees.




Trial Photos 1977


Me at the mast, unfurling the staysail main.  Patti O'Connor is passenger.  Foil is lashed to ama at a few degrees, flat side in, curved side out.





Reaching on the other tack.


Close reaching, I think, Ama really pulling down, as if ballasted, but not.


On the beach.  Foil has been unshipped from bracket on ama.

Hull is 17' long.  Mast is 30' long. Not a good idea for a mast fabricated of aluminum conduit.





Micronesian Sailing Canoes 

Douglas Taylor

Anthropology 3520
  25 January 1975
Prof. McKnight



In the study of Micronesia, the subject of traditional inter-island communication will attract the attention of Western sailors. The evolutionary development of the sailing craft of Micronesia was for the most part complete before the first European contact. This evolution produced a type of outrigger sailing canoe of an amazingly efficient design, which is radically different from Western craft.

The same basic design is found throughout Micronesia:

"…this mixture of races has, however, been so long in the melting pot that in all that concerns canoe design, Micronesia is as fully homogenous A Polynesia. What differences exist are due to local factors, and even there, are confined in the main to inshore craft. The voyaging canoes have a remarkably close family resemblance, bespeaking regular or frequent intercourse between the constituent island groups." (Hornell, 1936, p.345)

Among the most sophisticated of these canoes is the flying proa of the Carolina Islands. The designation "Flying Proa" is supposed to have been used since Magellan first sighted sailing outriggers in 1521, but it is thought that Magellan may have sighted Marianas canoes rather than Carolinas craft. (Hornell, 1936, p.375)

The speed and agility the canoes amazed the earliest visitors in Micronesia. Pigafetta, one of Magellan's crew, wrote, "the islanders delight to plough the sea with their small boats… they change stern and bow at will…and resemble dolphins which leap in the water from wave to wave." (Gibson, 1958, p. 66) Dampier, who visited in 1686, estimated a proa's speed to sometimes exceed 24 miles per hour.

The above mentioned 'changing from bow to stern at will' is the basis of the unique way in which the canoe is sailed. Discussion of the sailing technique will follow a description of the "Popo," or flying proa of Yap, (central Carolina's) and of variations found in other areas. This will provide a background for arguments on the sailing characteristics and some ideas on modern adaptation for further efficiency.

The Popo of Yap

The hull of the popo is built-up from a dugout keel to which are attached carefully fitted strakes. Traditionally these hull-forming strakes were secured with sennit twine through holes along their edges. Seams were caulked with sennit fibers and tree gum. As large trees are available in Yap, the dugout portion could be large, with only several strakes attached. On atolls with limited wood, however, the dugout part was wedge shaped, and many strake pieces were stitched together in a patchwork manner. These "patchwork" hulls fitted with great accuracy and skill.

The hull is longitudinally asymmetric. The outboard side, the side opposite the outrigger float, is straighter and flatter than the inboard side. The degree of asymmetry varies from nearly straight and flat to being only slightly less curved than the opposite side. The two basic functions of this asymmetry are the use of the steep side for lateral resistance, a "grip" on the water, like a Western sailboat’s keel, and to counteract the turning force from the drag of the outrigger float. An ideally shaped hull will cause the canoe to move straight through the water when pushed. (Hornell, p. 377)

Outrigger booms, "aka," attach to the topmost strakes on each side of the hull. They extend a few inches over the flat side of the hull, and eight or ten feet out from the weather side. The outrigger float, "ama," is solid wood and is shaped for little water resistance. It is attached to the booms with "Y" shaped wooden yokes. The weight of the outrigger is as important as important as is it’s buoyancy. When moving fast under sail, the float is often raised clear of the water, it’s weight countering the canoe’s heeling tendency.

Measurements of a representative Yapese popo are as follows: Overall length - 9.4m, Keel length - 7m, Depth amidships - 1.15m, Length of the booms to the axis of the float - 3.6m, Length of the float - 4m. (Muller, 1917, pp. 183, as quoted by Hornell, p. 182)

The sail is the shape of an isosceles triangle, with the two longer edges laced to two long poles. The mast is shorter than these spars, and pivots fore and aft from the middle of the hull. The upper spar is braced between the mast and a butt socket at either bow, The lower spar then hangs at the bottom of the sail and is controlled by a mainsheet. This is the oceanic lateen type of sail, of Southeast Asian origin.

Overall width of the popo is increased about 20% beyond the hull and float width by the addition of the lee platform. This rectangular deck is secured to two poles lashed to the aka in the canoe hull, pointing out from the flat side of the hull at an upward angle. This platform ,opposite the ama, is used for cargo or passengers. There was often a low hut of woven fronds here for protection from the weather. A somewhat taller hut was sometimes sitting on the float booms, between the hull and the float.


Theabove description of a Yapese popo can generally be applied to other Micronesian sailing canoes. Regional variations are obvious enough that a canoe’s place of origin can be determined by its appearance. Variations are usually limited to the presence or lack of a lee platform, the method of securing the float to the hull, and the design of the stem ends. In the case of the canoes of Palau and Sonsorol, the hulls are symmetrical, but the sailing method is the same. The following discussions pertain only to deep water sailing craft:

The Gilbert Islands canoes have no lee platform, they use three float booms. Frames are inserted into the hull for strength, as they are built up of patchwork planking. There are no stem projections.

Marshall Island canoes have lee and weather platforms. The float is supported by two main and four to eight minor booms. Hornell describes this system as having been brought to relative perfection. (Hornell, p. 362) These canoes use long, narrow floats. Narrow stems extend beyond the ends of the hull.

The Marianas canoe has no lee or weather platform. They have three outrigger booms, and the hull has slightly projection stems.

Canoes of the Palau Islands, and Sonsorol are not built with the asymmetry common to the others. There is a small weather platform, and two booms secure the float. On the Palau canoe straight thwarts across the gunwales project a little over the outrigger side. These ends support a long straight pole parallel to the hull, looking like a grab rail.

From Hornell’s summary on Micronesian canoes:

"In all the Micronesian groups of islands the design of the outrigger canoe reached a higher level of development than in any part of Polynesia, as did the knowledge of the science of navigation possessed by certain of the Islanders. It is difficult to conceive of any possible great improvement on the sailing canoe design of the Micronesians, for it combines three inventions of the utmost utility in sailing: 1, the flattened lee side of the hull acting as a leeboard to reduce drift to leeward and compensating to some extent for the pull to the weather side of the outrigger float; 2, the use of a lee platform on the cantilever system, enabling a greater quantity of cargo to be carried; 3, the midships pivoting of the mast, where the canoe was able to sail either end forward and so to keep the outrigger on the weather side, whichever course she was on." (Hornell, p. 438)

Sailing an Outrigger Canoe

When sailing, the wind is always kept on the outrigger side. If the wind came over the flat side, the canoe would be taken aback, and as the mast is not supported from that side, could be dismasted. The sail yards are stepped in the bow with the mast raked forward. A steersman works a large steering paddle from the extreme stern of the hull. Another crew member controls the sheet.

To tack an outrigger requires at least two persons. The canoe is put on a course perpendicular to the wind. The butt end of the upper spar is disengaged from the bow, carried manually to the opposite end, and secured. Another crew member controls the mast stays to concurrently rake the mast to the new bow. The helmsman takes a position at the new stern, the sail is trimmed, and the canoe moves forward on the new tack. Sometimes the steering paddle is secured to a lanyard tied amidships so that it can be dropped overboard and allowed to drift toward the other end. Accompanying photographs show the sequence of this tacking method.

Micronesian canoes sail easiest across the wind or somewhat downwind. Despite the efficient implications of the asymmetric hull, they do not point very high into the wind when to windward. In practice they are sailed upwind with the sheet, not so much the steering paddle. When the sheet is hauled in, the wind pressure on the sail toward the front of the canoe is increased, so the bow is forced downwind. Conversely wind is spilled from the sail when the sheet is slacked away, so that due to the drag of the outrigger, the boat turns up into the wind. It therefore proceeds to windward in a series of curves between 50 and 70 degrees off the true wind. Leeway, drift to leeward, is about 15 degrees, which makes an average windward course of about 80 degrees. (Lewis, 1972, p. 270) This upwind sailing problem can be described as a severe lee helm imbalance that forces the bow off the wind.

Upwind sailing was avoided whenever possible, though, because a wind change could more easily result in a gust from the wrong side that would capsize the canoe, and because upwind sailing subjected the craft to greater stress.Early canoes, bound entirely with vegetable fiber, were much more vulnerable to stress upwind than modern canoes made with iron tools, and fastened together with bolts and nails. (Sharp, 1963, pp. 56)

Authors’ opinions on the above matters do not all coincide. David Lewis refers to L. Mason and W. Draper to argue that the advantage of the flexibility of sennit lashings would be defeated by the use of iron fastenings. (Lewis, 1972, pp. 56) Alan Villiers states that the use of fiber lashings results in unsatisfactory seagoing vessels, subject to disintegration. (Villiers, 1949, p. 48)

It is my view that flexibility serves a shock absorbing function and is important in reduction of structural fatigue. However, the degree of optimal flexing could be determined for the various parts of the canoe. Modern materials, like metal alloys, plywood, and fiberglass could be used in such a way as to incorporate controlled flexing combined with strength. If Lewis is entirely correct, it follows that automobile suspensions would be made of fibrous materials.

The oceanic lateen sail plan enjoyed the advantages of maneuverability and upwind superiority over the earliest Western square sailed craft. Due to material advantages, however, Western technology has continually improved the efficiency of sails and hulls. Among the quickest of small sports sailboats today are the light planing sloops and catamarans. The latter bear only superficial resemblance to the Polynesian originals. A modernization of the Micronesian proa might result in a sports sail boat as fast as the best. If such a craft could outperform established designs of similar size, or if it could come close at less cost to the owner, then a new design could compete in the sailboat market.

Thomas Gladwin discusses recent innovations in the design and construction of Puluwat canoes. The three he lists are: replacement of the mat sail with cloth, the use of iron tools in fairing the hull, and dropping the taboos and rituals associated with canoe building. This last point results in smoother hulls because more people can lend a hand. (Gladwin, 1970, p. 123) Attempts at innovation outside Micronesia seem to miss the point of the importance of lateral symmetry. The Malibu Outrigger is a class of sailboat seen in Southern California which is an outrigger canoe, but which tacks upwind like other Western craft. The crew must try to counter the dragging outrigger when it is on the down wind side, by hiking their bodies out over the opposite side.

To approach the problem of modifying the Micronesian outrigger I first listed the assets and liabilities of the traditional design, then formed an idea of what a new design should be able to do. Superior aspects include the low water resistance of the extremely narrow hull, which also weighs less than a conventional hull of the same length. The flat side of the asymmetrical hull resists leeward drift without the depth or drag of a keel or centerboard. The curved windward side causes a water flow much like the aerodynamic flow on an airplane wing, which tends to give the hull upwind lift. (Gladwin, 1970, p. 94) The weight of the outrigger float performs the same function as a lead keel, but the ability of it's weight to counter wind pressure is immediate because it is attached at 90 degrees to the mast. A boat with a ballasted keel has to heel over quite a bit before the weight will counter heavy wind pressure. For these reasons, the weight of the outrigger float can be far less than that of an equivalent ballasted keel for the same amount of stability.

The disadvantages of these canoes include the necessity of manhandling the sail from one end to the other, making it impossible to sail single-handed. People are needed to sit on the outrigger booms to counter the wind if it grows heavier. The sail being attached so fat forward causes the previously mentioned lee helm tending to turn the canoe downwind when trying to tack. When running before the wind, the sail thrust is all outboard of the flat side and the float is in the water dragging on the other side. This necessitates the use of a large steering paddle to keep it on course, which constitutes a drag.

An ideal design, then, should provide a canoe that can be worked single-handed, including moving the rig from one end to the other. Either the sail plan to the center of lateral resistance should be continuously adjustable for steady tacking upwind, and to lessen the need for a heavy rudder / steering paddle when sailing off the wind. If it is to be sailed single-handedly, a way of spilling excess wind would be useful.



Construction of 17' Prototype, Sketches 



Here is how the boat articulates.  The aka(s) pivot at the hull and at the ama.  The windward platform keeps the aka from swinging up, shrouds to the aka assembly keep it from dropping.  The sail tack traveller can be seen.  Most of these ideas came out a little different from the sketch in actual construction.




This scale drawing did not scan so well, so you have to scroll horizontally to see stuff.  The stays and shrouds are shown here.



This scale drawing did not scan so well either, so you have to scroll horizontally to see stuff also.  The stays and shrouds are shown here, as well as the lashing of the hydrofoil angle that was adopted during trials.


(Editor's note:  I can't seem to make the Wayback Machine cough up sketch #3, not sure why.  I'll try to get it from Doug.)





1977 - Rigging Proa before first launch.  Andy, the best dog I ever knew, waiting patiently for a ride. 




web statistics















Comments (0)

You don't have permission to comment on this page.