Philip Goode at Pelican


In the matter of harnessing wind power to reduce fuel consumption in ocean carriers, it would seem that the high capital cost of some of the devices being proposed make them questionable financial propositions. These high costs arise from the design policy that these devices should require no extra crew.

But an unbiased long-term assessment of operating costs, including maintenance and replacement, might well show that a partial return to a simpler technology with much lesser initial cost would be a considerably less vulnerable economical proposition, even taking into account the  cost of the reduced number of crew which  present-day advances in sail handling permit.

If a ship fell on har d times, in the last resort the operators could stand down the crew;  but the same could not be done with the amortization of the huge loan involved if she had been  equipped with  ultr-costly telescoping or foldaway metal airfoils.

The drawing shows a proposed 25,000 ton dwt vessel which could be 75 to 95 percent wind-propelled according to route. She could carry bulk cargoes or uo to 800 20-foot TEUs below decks and the raised poop shown,  with its  600 sq.m. area could accommodate 24 passengers, their staff and facilities, which could be operated as a concession. Setting of sails and rotation of yards would all be managed from deck by no more than 6 sail handlers, out of a total navigating crew of perhaps 12, subject to flag requirenments.

The rig is a variant of the Dyna-rig which is being proposed for similar vessels, but because the masts are stayed instead of being unsupported this makes it a far more realistic economic proposition. The staying has been arranged for minimal interference with cargo handling gear at terminals.


An outstanding feature of modern square rig, one which has never been a part of windjammer course-plotting until now, is the new ability to  sail much closer to the wind than was ever thought possible. And here I wish to emphasize two things: one, is that the last statement doesn´t mean the ship would go looking for head winds to pit herself against, and  the other is that all the  ideas in this article are only offered as starting-points for the consideration of Engineers, Naval Architects and Riggers.

The ship would have moderate auxiliary power and a large wind generator with blades which would fold away in heavy weather, to provide most of the electrical needs on passage.


length bp, 204.0 m
beam 25. 6m
depth 16.2 m
load draught 12.6 m
load displacement, 36,000 tons.
deadweight, 25,000 tons
sail area, 7072 sq.m.
prismatic coefficient, 0.70


Up-to-date notions as to bracing angle of yards (18.5º from centreline) make for a square rigger which can sail much closer to the wind than was possible before. This very sharp bracing angle offers one other important gain: if forced by dire circumstances to motor into a strong wind with sails furled, a ship could do so on a wind heading of18.5º with her yards pointed straight into the wind, whereas a 25-knot headwind will bring a conventional square rigger to a standstill due to windage.

It has been shown in full-scale practice (S.T.V.”Pelican of Londin”, 2007) that a stack of square sails of modern cut (no more than 2% roach) will develop awesome power to windward. In moderate wind and with gentle handling, the airflow over the whole rig will at times appear to be contained within a sort of cocoon, forming a self-contained system which is capable of pointing the ship upwind well past the heading where sails would normally stall, without any evidence of this happening. In fact, I have photos of a sailing model with weatherly rig tethered by the stern in a pond, accepting a negative angle of incidence (as confirmed by the windvanes), a phenomenon which repeated itself as the gusting wind fluctuated in direction. And there is a story about the 109 m Russian sail training vessel “Mir”, a fully rigged three-masted ship, being teased up into an impossibly close-seeming wind angle in a light breeze, which would appear to bear out the cocoon theory, even if the question is purely academic.

Now, I´m not suggesting this should be the usual mode of progress, but given these observed facts it should not be beyond the aims of designers and operators to envisage a sailing vessel of sufficiently large size, steadily punching her way due east into the trade wind or equivalent course in a monsoon, if that should be a part of the most economical route.


This would aim at securing 95% wind propulsion, using the strongest following winds – where square rig has no peer – by rounding the great capes in the favourable direction as opposed to queueing, canal dues and lack of wind near Panama and Suez, while any refinement of pointing ability, however small, might enable new routes to become viable.



Practical tests (2014) with a very basic model in natural wind indicate that this close-windedness could be further enhanced some 2.5º by the use of squaresails which are split vertically, thereby creating a slot effect in each sail unit. This slot might  also possibly reduce the amplitude of the turbulence affecting sails on the mast immediately astern when hard on the wind.



The rudimentary nature of these trials means they need to be confirmed and their usefulness  evaluated by wind tunnel tests. These would enable the propulsive thrust developed by one single stack to be quantified, and then an apparatus with 3,4 or 5 stacks could be tested at progressive mast spacngs until full thrust by all masts was achieved.


Despite wind tunnel tests, it makes no sense to build a full-sized ship without first having tested a significantly sized replica in open water, say 4 metres in length or one-fiftieth scale. The apparatus involved here need not be very sophisticated since a model cam be steered by windvane perhaps even more usefully than by radio control. The anemometer, which should be very accurate at low windspeeds since scale breezes are very light, can be carried in the accompanying tender and heading relative to wind is shown by windvanes.

The only truly useful gadget the model herself could carry is some means of recording ship velocity. Such tests are the only sure way to confirm hull/rig balance when on the wind. In practice, the 226 GRT “Pelican” was found to behave in every respect as her 1:25 scale model had done.


One being that in progressive reefing dictated by a rising wind, the leeward halves of sails would be furled leaving the leading halves as an uninterrupted airfoil free of the widening gaps which result from conventional reefing. It would really justify the split rig´s duplicated sheets and tacks (which would be self-tending, see later) if it was found that it effectively reduced turbulence on the following masts when pointing high.


The elimination of this interference is a goal very much worth striving for.

Could down-wind performance be affected by the slots? Both spinnakers and parachuts have been found to benefit from holes or slots, and this might apply here too.


The rig is a substantially more economical version of the generally advocated Dyna-rig, in that the masts are fixed and stayed, the moving parts, which are free of standing-rigging load, being a rotating mastlet to which the yards are fixed so that they may be rotated together.

As alternatives to the mastlet, the yards could be slung directly from the forward pillar of the tripod tower,  which would be heavier than the other two for this purpose. Each yard would be   actuated either by an individual (and synchronized)   hydraulic ram ,  or else by a chain drive going from one side of the yard to the other around the back of the lattice tower,  also to operate all yards synchronously. If these latter alternatives were employed, the “fan” of upper yards to accommodate wind gradient could be built in.

Masts would be stepped on deck over main bulkheads. One reason for then being lattice towers is the potential for stiffness of these strucures. A thrsee sided tower will not only allow full yard rotation more easily but it will have less windage and weight than one with square section. A combination of tubular  elements (in compression; in tension they could be flat) might offer the best  stiffness figures, the connections between these parts being gusset plates in a combination of bolting and welding.

In the interests of lowering the centre of gravity, the lattice structure could be given some taper. Access aloft would be by Jacob´s ladders inside the masts with working “landings” at yard maintenance points, and also to contain hydraulic pipes and items of running rigging.


As explained above, they could be simultaneously braced in several ways.

It would be normal that wear should at some point induce rotational play in the actuating mechanism. For this reason, and also to take the load off the gear in prolonged passages (Roaring Forties) or heavy weather, when shortening sail would result in asymmetrical yard loading (if the split squaresails are used), preventer braces would be rigged to two out of the four yards, not necessarily to be employed all at the same time. When not in use these falls would be brought out of the way to the mast´s heel againt the aft side of yards. When in use they would be rigged for best effect to snatch-blocks (then also manually to self-tailing winches) on either side of the deck situated for most effective lead according to which tack the ship were on, this being a requirement imposed by the extreme bracing angle when making into the wind.

To counter the downward pull of these braces the yards would be fitted with permanent fixed-length lifts to the yardarms as shown in the thwartship elevation in the sail plan. As a general design suggestion, number of yards should be kept to a minimum; although in this particular case, in view of the sail sizes involved,  it might pay to consider one additional  yard. Their undersides having a slot along their entire length to take the split sails means that they would be of larger diameter than a traditional square-rigger man would like.  As regards the combined weight of each mast and its yards and mastlet (if employed) , a comparison with “Preussen” (cubic scale factor 3.95) gives 224 tons per mast, including its rigging but ex sails.

The weight of a three-sided tower mast alone being in the order of 50 tons, this suggests that the combined weight could be comparable to traditional steel masts with their inner stiffening, their doublings, tops and  standing and running rigging. Extrapolating further from F.L.Middendorf¨s epic 1903 work “Mastung und Takelund der Schiffe”, the weight of wire rigging, chain and blocks for a conventional mast this size would have been 64 tons.


This is intended to be as minimal as a seagoing structure will permit, benefitting from the lattice masts´ high resistance to bending and in order to minimize interference with cargo handling equipment at terminals. The mastheads are secured by three main wires, i.e. two backstays and one forestay. These in turn are linked to the mast at about its mid-height by short wires which take the place of lower stays, on the “Sky Link” principle. The three sky links thus oppose each other at the mast.

The unusually large aft drift of the backstays, which permits the yards to rotate through 143º, is copied from “Pelican”. Despite this, wire sizes involved are reasonable, even if the theoretical scenario on which these are based may pose a bit of a quandary. In this case, a 30-knot wind perpendicular to the full rig hard-braced, striking the upright ship before she had time to heel , calling for a breaking strain (with F. of S. = 2), of 205 tons. A suggestion for each of the three main stays would be: a parallel 4-part bundle formed by two twinned loops of 40mm galvanized steel wire, each loop end-spliced and served over thimbles at each bight. Possibly they might be more stable with a core wire in the middle. Compaed to using a single, thicker wire this would reduce risk due to cargo handling damage or to the insidious effects, particularly in the jigger mast, of engine vibration. The forward sky link runs perpendicularly to the mast, passing between the topsail yard and the foot (“feet”?) of the topgallant sail/s without touching either, which allows unhampered yard rotation.

At sea the rig is beefed up by conventional lower backstays, shown dotted, which would be half the cross-section of the main stays. The (principal) masthead forestays themselves are also not led to the deck but here, the reason is to get them out of the way of cargo handling equipment. Instead, they go to a strop on the lower part of the mast ahead. In exchange for compressive load on this section of the mast, use of this strop reduces the aft-ward pull of the stay on the mast ahead by a factor of 3.


Aside from the preventer braces referred to before, this would consist almost exclusively of tacks and sheets, the yardarm lifts being fixed-length. If split squaresails were used, there would be four of these lines to each yard. However, it would not be difficult to make them completely self-tending if the line from each inboard clew was led to a lead block on the opposite side of the mast.

Then, with the falls belayed at prearranged marks, the slackening of the weather clew, permitting the slot to form, and the tightening of the lee luff would be automatic as the ship went about.


The relative sail area of this ship is very low compared to that of the 1902 five-master “Preussen”, at 124 nm the largest sailing bulk carrier ever built. The proposed ship has a ratio of sail area to two-thirds power of displacement 35% less than that of “Preussen” under full rig, being closer to that of the older ship under working sails only.

If the new ship had five masts instead of four the areas would be more comparable, but it is not a question of how much sail is set but of how efficient it is. One reason for this reduced rig is that the 1902 ship carried 40 able seamen to handle her sails whereas the new ship could be run with 6, even though in terms of cubic capacity she is nearly four times larger. Quite possibly the proposed rig would reach full propulsive efficiency without the overkill obvious in any of the breathtaking shots of “Preussen” under full sail, or at least get close enough to this point, to be a good economic proposition. For a common speed of 12 knots the new ship would be operating at a speed/length ratio 15% lower than the older ship´s.

Down-wind, despite the new ship´s relatively narrow yards and generous mast spacing, these very factors might help her sail a more direct route because her sails have less tendency to blanket one another. It can be argued that “Preussen”s cloud of sail was the only way to get her through belts of calms, whereas the new ship would have modest auxiliary power.
Ghosting staysails in very light cloth couild be set flying and recovered y threading the downhauls through rings in the luff, although with only six A.B.s it would make more sense to turn on the engine and head for where there was more wind. These are matters which can only be evaluated in the light of sailing-model test results, targeted routes and financial considerations.


Length of lower yards, instead of being twice the beam as was traditional, is only 1.33 times the beam. Besides giving the stack of sails a more efficient aspect ratio this makes for less overhang if the yards are squared off at the terminal.

The courses are not split sails but single ones. The reason they are cut down narrow at the foot is because when the sail is furled up into the main yard, the club at the foot of the course rides up the mast along a track or guide so that it is right out of the way, allowing a 17-metre clearance above the hatch coamings. Because the mechanics of this track or guide would inevitably lead to large play if the yard were full-length, and also because of the weight of sucha yard (or “Bentinck boom”), it is drastically reduced in length, when the angle of leech and luff also help to give the small club axial stability.

The lower backstays of fore, man and mizzen masts are shown dotted because they would only be rigged at sea in order to take some of the load off the main backstays. In cargo operations they would be slacked off hydraulically, the shoreside ones being released altogether and moved out of the way. Following on this principle of adapting staying to cargo handling conditions, it would not be impossible to envisage getting all the lateral rigging of main and mizzen masts out of the way. Also shown dotted are the triatic stays linking the mastheads.

These only have tension put on them as a safety precaution during terminal operations, when an accident with loading/unloading machinery might damage one of the main backstays. At sea they would be eased off to avoid hull flexure straining the mastheads.



The attached figure shows a  suggestion for a sailing ship destined to carry a cargo with a dangerous moisture content such as coal slurry, biomass,  organic waste, etc., in the form of a hopper-shaped cargo hold to accommodate 25,000 tons of coal, allowing a GM of 0.61 m with full load. The side tanks could carry ballast water when necessary, on the “full or empty” principle  . For use in delivery or distribution of non-urgent non-perishable containmerized goods, the side bulkheads would be upright and the coamings wider-  800 TEUs could be accommodated below decks in six tiers and the lattice masts would make good bases for  cranes for self-unloading.


Whatever information thou gettest from computers and wind tunnels, thou shalt put to practical test in a large sized sailing model, this being an insignificant expense compared to its useful results. The model´s scale hull, rig and cargo weights thou shalt apportion accurately, confirming GM by an inclining experiment.

Hull: the underwater profile shall be cut away fore and aft to reduce wetted surface and yawing. Length of Parallel Middlebody optimized to 0.7 Prismatic coefficient and 12 knots operating speed. Avoid low freeboard simply for economy´s sake, this should be adequate to humane working conditions in heavy weather, and superstructure adequate to a weatherly sailing vessel. Mast spacing: any mast which does not deliver the same thrust as the foremast when close hauled is not worth its capital investment.

Electronic controls: seek not automatic or touchpad rig control, nor to show off just how convenient electronics can be, but return to simple electrics for robust reliability in keeping with prolonged heavy duty and misuse in a hostile environment.

If the split sail option was abandoned, remember that flat cut square sauils are still an impressive windward proposition.