Featured Article » Nonfiction

The Wind is Free: Sailing Ship Design, Part 2, Seaworthiness

Written by Iver P. Cooper

 

The designer of a sailing ship must give it sufficient capacity and speed to carry out its mission, yet without unduly compromising its seaworthiness. And seaworthiness itself is a complex concept, embracing watertightness, buoyancy, stability, hull strength, weatherliness, handiness, and freedom to enter shallow or constricted waters.

 

Capacity and Displacement

 

The ship buyer, be he king or commoner, doesn't specify the hydrodynamic parameters. Instead, he says, "I want a warship of 100 guns" or "I want a merchant ship, capable of voyages to the Indies, which will carry 500 tons of cargo".

Capacity ("burden") is the ability to cram in crew, passengers, provisions, cargo, cannon (and their shot and powder) and miscellaneous supplies (e.g., spare sails). It's limited both in terms of volume (by the dimensions and layout of the ship) and weight (too much, and the ship sinks). Until the nineteenth century, it was probably the single most important desideratum for a ship (other than staying afloat). The different demands on capacity compete with each other; for example, putting on more cannon (and the crews to man them) increases fighting ability but reduces the space for cargo and crew provisions.

The formula developed (1582) by the Elizabethan shipwright Matthew Baker, and one of several formulae in use in the 1630s, was

keel length X maximum beam X depth of hold, all in feet, divided by 100.

The result was a value in tuns burden; a tun was a volume measurement, a container of 252 gallons of wine, about 40 cubic feet, weighing about one English long ton (2240 pounds). Thus, the original meaning of "burden" was the number of tuns of wine that the ship could carry.

There was also something called "tons and tonnage." That added to the burden ("tons") an estimate (typically, one-third of the burden) of the miscellaneous goods ("tonnage") which could be carried. A ship with a burden of 300 tons has a "tons and tonnage" of 400. (BakerCV, 25-6). When "burden" is quoted in the literature it often really means "tons and tonnage," the total cargo capacity (modern "net tonnage").

The largest of the seventeenth-century merchant ships were the Portuguese nao, which were rated as high as 2000 tons burden. (Brigadier 12).

Passenger capacity can be estimated from burden. The tendency was to stuff the ship for maximum profit. A 1534 Spanish ordinance limited New World-bound ships to 60 passengers per 100 tons burden, but some carried almost one per ton (Perez-Mallaina, 130). In the Irish emigration to America, the average was 0.4/ton in 1769-70, and 0.66 in 1771 (Wokeck 185). In 1819, US ships were limited to 0.40/ton. (Blunt 314). Bear in mind that these ships carried cargo, too.

Of course, slave ships were packed even more densely. The average was reportedly 4 slaves/ton for 1600-1650 (Thornton 118). A 1684 Portuguese ordinance limited carriage to 2.5-3.5 slaves/ton (depending on portholes). (Rawley 252).

Displacement is a somewhat slippery concept, as it can be expressed in both weight and volume terms. The sum of the lightship weight (hull, rigging, armament, superstructure) and the deadweight (crew, provisions, stores, and cargo) is the load, which causes the ship to sink until the underwater portion has displaced a volume of water of equal weight. If that comes at a point at which the ship's deck is still above water, then the ship is floating (and if not, you need a new designer). Multiply the burden of a down-time ship by 1.67 (Wikipedia/BOM) or 1.3-5 (warships; Glete 529) to crudely estimate its displacement.

For a ship to be buoyant, the designer has to limit the ratio of its mass to its volume so that its overall density less than that of water. And that means that a steel hulled ship has to have a greater volume than a wooden hulled one of the same surface area, to compensate for the greater density of the hull. Even so, they tend to ride lower in the water. (McCutchan 110).

A battleship by definition must have a large displacement, and that would also be typical of a long-distance trader. There was a tendency to overload long-distance traders to increase profitability. Matters were exacerbated by the nonchalant distribution of weight; heavy cargo often ended up on the upper deck. The Cosmographer Royal said that overloading was one of the reasons the nao Santo Alberto (sunk 1593) "and many others lie buried at the bottom of the sea." (Brigadier 13). Warships also were victims of overloading; excessive armament contributed to the capsizing of the Vasa.

In nineteenth-century wooden warships, about half of the load displacement was attributable to the hull. For merchantmen, the hull was only 35-45% (wood) or 30-35% (iron) displacement. (White 384).

The development of integral calculus in the second half of the seventeenth century made possible the calculation of the underwater volume corresponding to various waterlines and thus the calculation of the waterline corresponding to a particular load. (Glete 50ff), and for that matter, the location of the center of buoyancy for a particular angle of heel.

In Weber's "In the Navy" (Ring of Fire), Eddie tells Mike, "I don't have the least idea how to figure displacements or allow for stability requirements, and I know the designers screwed up the displacement calculations big time for a lot of the real ironclads built during the Civil War. There was one class of monitor that would've sunk outright if they'd ever tried to mount their turrets!"

 

Draft and Freeboard

 

The ship's draft (distance from waterline to bottom of the keel), and also the waterline length and breadth, will change depending on how heavily it's loaded, and how salty and warm the water is. Shallowness of draft is desirable if the captain wants to negotiate rivers and coastal waters (perhaps to escape a deep-drafted pursuer which would, if it followed, run aground). The great draft of the Constitution-class ships limited which ports they could use. (ChapelleHASN 130). But a deep draft ship can sail closer to the wind, and is less likely to drift to leeward (Anderson 88; ChapelleHASS 46). And it's less susceptible to wave action (Walton 168).

A ship with high freeboard (distance from waterline to deck) will suffer from windage, and be driven to leeward, but one with low freeboard is also easier for hostiles in small craft to board, and will be more likely to take on water if the sea state is high. (The Egyptians have the colorful term, "sailing with your coffin.") (Hollander 58). Freeboard on early nineteenth century British frigates was usually 6-9', with drafts of 15-20'. (Gardiner 143). Lloyd's rule was to provide 2-3 inches freeboard per foot of depth (White 33).

 

Speed and Resistance

 

The wind exerts a force on the sails, which cause the ship to accelerate. But there are forces which oppose the motion of the ship through the water (and air).

Frictional resistance, which is dominant at low speeds, is the result of friction between the hull and the water it contacts, and is proportional to the "wetted surface" of the hull, and the hull roughness. It increases as the square, or nearly so, of the speed (Baker 19ff). It's usually 80-90% of total resistance for ships at 6-8 knots; 50-60% at twice that speed. (White 448).

Form (pressure) resistance is the result of the hull pushing water out of its way, and the water returning to form a turbulent wake (eddies). It is proportional to the cross-sectional area of the underwater portion of the hull, and affected by the shape of the bow and stern. For ships with easy curves at bow and stern, eddymaking resistance is about 8% of the frictional resistance (White 449).

Air resistance is the result of the ship's above-water structure pushing air out of the way, and thus is akin to form resistance, but much weaker. The resistance is increased if there is a headwind.

Wavemaking resistance is simple in concept but difficult to quantify. As it moves through the water, the ship makes waves, which cost energy. In general, the faster the ship is going, the greater the resistance, roughly as the square of the speed.

However, there is also a periodic fluctuation in resistance, depending on speed. The bow waves and stern wave systems interact, and, depending on the speed, they may reinforce each other or partially cancel each other out. The distance (wavelength) between the waves increases as the square of the ship speed, and when the wavelength is near the waterline length, the waves reinforce each other. The wavelength equals the waterline length when the ship is traveling at "hull speed" (in knots, 1.34 times the square root of the waterline length). At speeds near the hull speed; this reinforcement means that the resistance increases faster than the square of the ship speed—indeed, as the third, fourth or fifth power.

Wave (added) resistance, as the name implies, is the result of ocean wave action. It is roughly proportional to the square of the wave height. (Nabergoj), which in turn depends on the wind speed and fetch. The direction of wave motion is also important, with "head seas" being the worst. Long, heavy ships are less affected. (Prpic-Orsic).

 

Stability

 

A ship can heel over as a result of wind or wave action, making a turn, or firing a broadside. What happens next depends on the relative positions of the center of gravity (whose position is dependent on how the ship is loaded, and whether it carries ballast) and the center of buoyancy (which depends on the hull form). The ship can right itself, remain at a "list," or be driven over further until it passes the "angle of vanishing stability" (AVS) and capsizes. The effect of the design parameters on stability can be complex.

Sailing ships typically had a maximum safe heel of 45-65°, depending on hull form, loading, and the possibility of flooding. (ChapelleSSUS 213).

A ship can be stiff, that is, have too high an initial stability. If there is a sudden gust, and it doesn't timely reduce sail, then since the ship doesn't heel much, the sails take the full force of the wind, and "the topsails are often carried away, or the sails torn to shreds." (Walton, 215). Worse, if the ship heels and then rights itself too quickly, it could be dismasted (as happened repeatedly with the 1800 Akbar—Gardiner 137). Stiffness can be reduced by "winging" weights out to the ship's sides or raising the center of gravity. (Walton 168).

For a warship, you want a slow and easy roll, limited in angle, to make it easy to aim.(ChapelleHASN 24).

Stability predictions are inherently more difficult to make for wooden ships because of the great variation in the specific gravity of wood. (Reed 360).

The Swedish crown had an unpleasant reminder that even kings are subject to the laws of physics. The pride of the Swedish navy, the Vasa, sank in 1628, on its maiden voyage, blown over by a gentle wind gust estimated as being just eight knots. Fairley says that according to modern calculations, four knots would have been enough to capsize it. Its maximum angle of heel was just ten degrees.

The basic problem was that the Vasa was top-heavy. It was the first Swedish warship with two enclosed gundecks. This was not part of the original plans, but rather a last minute development in the Swedish-Danish arms race. There were also several upward revisions, during construction, of the number and weight of the cannon. All this meant that the ship was not only taller, but wider. Since the Vasa's keel had already been laid, the width had to be added mostly in the upper part of the hull, which further raised the center of gravity. The keel was found to be a bit thin for supporting all the added weight so additional braces were added in the hold. With space reduced, Vasa could only carry about 120 tons of ballast, and Fairley says it would have needed more than twice that amount to be stable. (But it was impossible to add more since the gunports were already only 3.5 feet above the waterline—Franzen 19)

It is interesting to note that the Vasa underwent a crude stability ("lurch") test. Thirty men ran side to side three times. The result? The ship rocked back and forth like crazy. The outcome was not reported to the shipyard or the king. Curiously, the admiral who witnessed the experiment concluded that the ship was carrying too much ballast because the gunports were close to the water.

 

Hull Strength

 

The hull of a ship has to be strong enough to withstand the stresses imposed by the opposing forces of gravity and buoyancy, as well as those added, once it ventures out of harbor, by wind and wave. It is obvious that a warship must also be able to endure enemy gunfire. But the warship's own broadside can deliver a considerable recoil shock. (Glete 35).

The resistance to these stresses is the compound effect of the ship's frames, decks, deck supports (beams and knees) and longitudinal or diagonal stiffeners.

The hull bends as a result of variations in the local ratio of weight to buoyancy along the length of the hull. When a ship hogs, the center droops; when it sags, its ends droop. In the hogged state, the main deck is compressed and the bottom is stretched; sagging has the opposite effect.

Hogging and sagging can occur even in still water because of the narrowing of the ends (reducing buoyancy) and the non-uniform loading of the ship. Hogging and sagging is even more pronounced when a ship encounters waves, because buoyancy is increased at the crests and reduced at the troughs. The worst situation is when the waves have a wavelength equal to the hull length. (Thearle 312). Moreover, as the center of the ship passes from crest to trough, its state changes from sagging to hogging, at a frequency of perhaps a few seconds (315). Obviously, this challenges hull integrity. If the bending is too great, the ship snaps. Goodbye ship. Even if the stress isn't catastrophic, the strains tend to reduce the speed and increase leakage. (Glete 36).

Once the relationship of hull length to speed potential was recognized, there was an incentive to build longer hulls. (ChapelleSSUS 412). But hogging and sagging stresses are typically proportional to the square of the hull length.

However, lengthening the hull has compensations. In the nineteenth century, designers took advantage of longer hulls by repositioning the foremast further aft, reducing the bow load and thus reducing hogging. (McCuchan 36). Also, a very long hull might not often encounter waves whose wavelength equals the hull length.

Increasing breadth and depth increases the weight, and therefore the tendency to bend, but also the ability of the hull to resist the bending forces.

The usual antidote to bending was reinforcement. The thickness of the main deck and the keel could be increased (McCutchan 37). The French frigate L'Oiseau (1772) had diagonal planking (ChapelleSSUS 207), and the USS Constitution (1797) had diagonal risers (Otton) , both to inhibit hogging. This became common in early nineteenth century. (ChapelleHASN 365).

Since hogging was feared more than sagging, from time to time, builders experimented with laying the keel with a slight sag in the middle. This expedient was recommended by Griffith in the 1850s. (ChappelleSSUS 366).

Handiness

 

The longer the ship, the slower it turns (Laing 32) and the larger its turning radius. A ship half the length is probably about four times as maneuverable. It also helps to have fine ends, and weight concentrated amidships. (Atwood).

 

Weatherliness

 

When the force of the wind upon the sail is not parallel to the keel, the ship will be pushed, not just in the direction its bow is pointing, but also laterally. This undesirable lateral motion is called leeway and a ship with minimal leeway when traveling upwind is said to be weatherly. The resistance to leeway increases with the draft and underwater length of the hull.

Windage (the force of the wind other than on the sails) is also important, and it is strongest when the wind is on the beam. In general, the ship with the higher freeboard or greater superstructure is going to suffer more leeway. If a single and a double decker were both making five knots close-hauled, in an hour the former might be pushed two miles to leeward, and the latter three. (Laing 75).

Weatherliness is especially important for fighting ships because it determines who obtains the weather gage. (ChapelleHASS 47).

 

Hull Dimensions.

 

The basic dimensions of the hull are its length, breadth (beam), and depth, all of which vary as a result of the curves of the hull.

Length. Seventeenth-century sources generally quote the keel length. The length of the gun deck limits the number of guns which can be carried upon it. (MurrayS 6). A late seventeenth-century naval regulation required a minimum spacing of 6.5 feet between gunports to accommodate the gun crew. (Grieco 110). Based on data for nineteenth-century British warships (Creuze 53), the length on the gun deck is perhaps 20% greater than the keel length. For the waterline length, which affects wavemaking resistance and pitch stability, I would split the difference.

Raleigh advised against building ships much longer than 100 feet (Creuze 17). For large warships launched between 1600 and 1640, a typical keel length would be 100-130 feet (Temmu). A first rate in Nelson's navy might have a 175 foot keel. (Longridge 7).

Breadth. For large warships launched between 1600 and 1640, a typical maximum beam (B) would be 35-45 feet, usually closer to 35.

Length/Breadth Ratio. This ratio determines the overall shape. Chapelle says that a ship with too wide a hull was slow, and one too narrow was an unsteady gun platform and couldn't carry sail well, presumably because of lack of stability. (ChapelleHASS 46). Vasa had a ratio of about 4:1. (Franzen 74ff).

The Portuguese nao had a length:beam ratio of 3:1 (Brigadier 11; Konstam 7). William Burroughs, the controller of the British Royal Navy, around 1596 suggested relative proportions for three "orders" of ships: (1) pure merchant ship, keel length twice breadth amidships, depth in hold half breadth; (2) all-purpose, length 2-2.25 times breadth, depth 11/24ths breadth; and (3) warship, length three times breadth, depth 0.4 times breadth. Looking at seven merchantmen (1582-1627) of 130-200 tons "burden" (see Capacity, below), these had length/breadth ratios of 2.14-2.92, and depth/breadth of 0.42-0.5. (BakerNM 8-9; Myers 106). The anonymous Treatise on Shipbuilding (c1620) called for length to be 2-3 times breadth, and depth 0.33-0.5 times breadth, and said the ideal warship was KL/B 2.78, D/B 0.43 (107).

"By 1634 it was very difficult to find a [British war]ship with a keel/beam ratio of less than 2.90, and there were several higher than 3.00." (Myers) In 1841, the typical length/breadth ratio for an English warship was 3.15, whereas in America, even merchant ships averaged 4.6. The extreme clippers which developed later in the nineteenth century reached a ratio of 5.7. (Laing 54).

Depth/breadth ratio. The depth determines the number of decks (Glete 52), the height of the gun deck, and the amount of freeboard. Duhamel (1764) says that the depth of a warship is usually 0.5B; but in the list of warships he provides, it is usually a bit less than that (4-6). In nineteenth-century British ships, depth was 0.55B for sloops and smacks, 0.58-0.75B for schooners and brigs, and 0.66-0.75B for large schooners and ships. This evolved under tonnage rules which penalized breadth and ignored depth. (Creuze 36).

 

Body

 

The body of the ship consists of the hull, deck, and any superstructures other than rigging. The hull is the backbone of the ship. It's partly out of sight, beneath the waves, but it should never be out of mind. Without a hull, you're swimming!

 

Frame vs. Monocoque Construction

 

In the classic truss frame construction, an internal skeleton carries the load, and the skin of the structure just keeps out wind and water. In a monocoque ("single shell") construction, it is the skin that bears all or most of the load.

Generally speaking, the truss frame is superior (on weight and cost basis) for resisting compression and bending, and the monocoque shell for resisting shear and torsion. In the Thirties, airplanes got large enough and fast enough for the monocoque strategy to prevail. (Gordon 311-3).

While classical ships were monocoque, their construction was labor- and timber-intensive, and hence this construction strategy was gradually abandoned. By the seventeenth century, the transition to frame construction was almost complete. The principal exception was that the Dutch used a hybrid process in building flutes; a few bottom strakes were attached to the keel before doing any framing. (Unger 124).

Chinese junks have been characterized as monocoques because they lack the keel, stem and sternpost, but their strength is not attributable just to their skin; they are reinforced by transverse bulkheads. (Thomas). Modern European monocoques are mostly open boats made of plywood or fiberglass, but there are also some monocoque minesweepers.

The backbone of the framed ship was formed by the keel, stem and sternpost. The length of these pieces determined the length and depth of the ship. Beginning in the early eighteenth century, there was a false keel under the true one. The idea was that if the ship ran aground, the false keel would absorb the impact, like the bumper on an automobile. (Mondfeld 74).

If a ship has a strongly tapered stern profile, it may have "deadwood," a vertical extension of the keel, to connect the aft end of the keel to the buttock of the ship. (After 1860, the bow was tapered enough so deadwood was needed in front, too.) The sternpost is connected to the aft end of the deadwood and the rudder mechanism is attached to the sternpost. Both the stem and sternpost are likely to be made of a single log of first class oak. (Longridge 11). Deadwood reduces leeway but increases frictional resistance. (Winters).

 

Transverse vs. Longitudinal Framing

 

There are two basic framing methods. In transverse framing, curved ribs run up from the keel, forming the load-bearing elements of the sides of the ship, and then deck beams bridge the tops of the ribs. It is called transverse because the ribs are (viewed from above) perpendicular to the keel.

The alternative is longitudinal framing, in which the sides of the ship are established by longitudinal (parallel to the keel) stiffeners. This was introduced in the nineteenth century (Young).

Bear in mind that a ship would not be purely transversely or purely longitudinally framed. Even if a ship has transverse ribs, they are attached to a longitudinal keel. And if a ship has longitudinal girders, then these must be linked by transverse "webs."

In wooden sailing ships, and early steel ships, transverse framing predominated. However, that meant that most of the structure did not offer any resistance to longitudinal bending. (NavArchWeb). That's fine for an accordion, but not good for a ship.

 

Bulkheads

 

 

These divide the ship into watertight compartments, and also increase its hull strength and its stability after being damaged. However, they also make it more time-consuming to load and unload cargo. Both transverse and longitudinal bulkheads were used regularly in Chinese junks since at least the second century (Temple 190), but prior to the nineteenth century, their use in European ships appears to have been spottier. Bulkheads were initially made of wood, but iron ones were introduced in the 1830s. (Young; Gould 79).

 

Planking

 

 

The framing must be covered with wooden planks or iron plates. There are two basic construction methods. In both, the planks run fore-and-aft. Carvel construction, invented in the ancient Mediterranean, fitted the planks or plates so they met edge to edge, forming a smooth surface. In clinker (lapstrake) building, used in northern Europe and in China, the planks or plates overlap their edges. (Svensson, 8). Clinker is not as streamlined as carvel, but it is stronger, and hence the planks can be made thinner.

Gould contends that the adoption of gunports favored the adoption of carvel planking (Gould 215). In the seventeenth century, the English used carvel planking everywhere, but the Dutch used clinker for their upper works. (Anderson 153).

John Smith's Sea Grammar (1627) says that a ship of 400 tuns requires four inch planking; one of 300, three inch; and smaller ships two inch.

 

Waterproofing

 

 

Wooden ships are made watertight by caulking them. Traditionally European ships were caulked by filling the seams with oakum (fibers from old ropes) soaked in pitch. The pitch can be distilled from pine resin, or from asphalt. Lime was sometimes used in place or in addition to pitch. The Chinese instead used tung oil or fish oil. Modern sealants include silicone and polyurethane.

 

Deck

 

 

In profile, the deck of a ship may be flush (horizontal), or it may have a sheer (upward curve) toward either or both ends. Judging from contemporary illustrations, seventeenth-century vessels will most likely have a very pronounced stern sheer. The sheer was gradually flattened out over the course of the eighteenth century. (Anderson 176).

Viewed from the front, the deck will usually be either flat or slightly cambered (convex upward). Camber slightly increases the structural strength, and reduces the recoil distance of the cannon (Dodds 89). Sheer and camber both permit water to drain away.

 

Hull Material

 

 

Wood has the advantage of being naturally buoyant; wrought iron and steel, that of considerably greater tensile strength and stiffness, both absolutely and in proportion to weight. Wood is vulnerable to biological attack; metal, to corrosion.

Wood. The deck could be made of any of variety of woods, such as pine (BakerCV 95). The materials requirements for the hull were more stringent. The official march of the eighteenth century British Navy proclaimed, "Heart of oak are our ships. . . ." Oak was the principal hull wood for European navies in the seventeenth century, too. Pine was used, especially in the cost-conscious Dutch flutes, when oak was unavailable or deemed too pricey. (Unger), but it was definitely inferior.

However, beginning in the sixteenth century, Portuguese ships built in Goa shipyards used teak. The teak hulls lasted a decade longer than those made of oak or pine, perhaps because of its resistance to teredo worms. (Brigadier). Moreover, oak contains tannic acid, which corrodes iron, and teak doesn't (Jordan). Teak requires less seasoning than oak, it doesn't expand significantly when heated, and it is extremely durable. (Bowen 143). The British displaced the Portuguese, but it wasn't until the early nineteenth century that the British permitted warship construction in India. Soon thereafter, the British use of teak surpassed that of oak. (Schlich, 578).

Mahogany was used by the Spanish in the New World, it being readily available in Cuba and the Honduras. It is more buoyant than oak, easy to bend and carve, and resistant to dry rot. It also doesn't burn or splinter as easily as oak. (Glete 31; Fine Woodworking 25).

As commemorated by a Bermuda postage stamp, native Red cedar (Juniperus bermudiana) was used to build the Deliverance and Patience, which went to the rescue of the Jamestown colony in 1610. Many sloops for the West Indian and African trade were constructed from this wood. (ChapelleSSUS 65).

Long, straight timbers are used for the keel, and are also sawn to make planking. For large ships, several pieces had to be "scarfed" together; HMS Thunderer needed seven baulks, each 26' (Dodds 58).

There is also a demand for "compass" (curved) timber for use in framing. Foresters would survey forests and mark the trees which had branches of a particular desired curve. In like manner, they looked for "knee" (angled) timber, taking it from the junction of branch and trunk. The knee timbers secured the deck beams to the frame (BakerCV 95). Warships had a particular need for crooked timber for reinforcement (Glete 52).

Wooden planking can be curved, but the curves must be gentle. If the curve is too sharp, the wood will break rather than bend. (Henderson 118). In the seventeenth century, "green planks were often scorched or heated in wet sand to render them pliable enough to be fitted around the customary bluff bows. . . ." (BakerCV 31).

The natural supply of compass and knee timber was gradually depleted, and hence means were sought to produce it artificially. Unfortunately, the heavier the timber, the harder it is to bend. According to Baker, "the steam bending of frames was unknown" in the seventeenth century, and Griffiths (16) says, "from time immemorial shipbuilders have bent their planks by a due application of heat and moisture but it is not . . . until the present [nineteenth] century, that any of them discovered how to bend frame timbers and knees."

Another issue was how to attach the planking to the frame, and indeed the various frame elements to each other. Baker says that in the early seventeenth century, there was extensive use of "treenails" (wooden pegs), with iron nails being used mainly in fastening down the planking of the superstructure. (BakerCV 33). However, the Portuguese apparently just used iron nails in the hull, and the Spanish San Martin (sunk 1618) used both kinds on the same planking (Crisman). Treenails were cheaper than iron ones even in the mid-eighteenth century (Dodds 24).

The great enemy of the wooden ship was dry rot. (Dodds 13). Three expedients were used to minimize it. First, shipbuilders selected resistant woods. Teak and greenheart are good for perhaps twelve years but weren't used by Europeans (except as noted) in the early seventeenth century. Oak is durable if seasoned (English practice was three years), lasting perhaps a decade, but unseasoned oak can be destroyed in a few months. (Which didn't stop the American colonists from using unseasoned wood for smugglers and privateers, ChapelleSSUS 13.) The heartwood was preferred even though that meant that the tree had to be allowed to grow longer to get enough of it. Elm doesn't rot if it's constantly immersed in water and hence it was used for the keel and the lowest planks. (Dodds 18; Murray 72). The Sparrowhawk (wrecked 1626) had an elm keel (Riess 71).

Secondly, at least by the nineteenth century, they experimented with various preservatives. Metallic salts didn't work well, but cresoted timber was resistant to both rot and marine worms. (161).

Finally, they grudgingly recognized that they needed to hold down the moisture level in the wood by forced ventilation. (162). In the nineteenth century, steam fans were available (Lewis 112).

****

Iron. Iron use in seventeenth-century ships was mostly in cannon, bolts, hinges, chainplates, hooks and the like. (ChapelleSSUS 14). Iron knees were used in the first rate Royal James in the 1670s, but weren't routinely used in England until after 1800. At first the knees were a hybrid of iron and wood. The complete iron knee appeared in the Unicorn (1824). (Goodwin) By the end of the eighteenth century, iron had also been used in the cross-bracings of warships. (Dodds 7).

An iron-hulled pleasure boat was built as early as 1777, but little is known about it. Wilkinson's Trial (1787) weighed eight tons yet drew only eight inches empty. Unfortunately, it and the three additional barges he constructed in 1788 cost at least three times as much as its wooden counterparts. (Barker)

In 1810, Sir Samuel Bentham unsuccessfully urged the Admiralty to switch to iron-hulled warships, in view of the shortage of timber. (It took 2-3 loads, each fifty cubic feet or one "standard" oak tree, per ton of ship, to build a warship, and 1-1.5 for a merchant ship, and the cost even in the 1750s was almost 10£/load. Dodds 13)..

Nonetheless, iron hulls started popping up a few decades later. The Aaron Manby, a wrought iron steamship, was built in 1820, and the 218 ton bark Ironsides in 1838 (McCutchan 111; Young). Ma Roberts (1858) was the first steel paddle steamer, and the 1271 ton Formby (1863) the first steel square-rigger. It cost twenty pounds to the ton. (McCutchan 36). The first iron warship was HMS Warrior (1860).

Iron had advantages other than availability, of course. It was stronger than wood, and hence could be used to build longer (hence faster) ships. While iron was more costly per unit volume, its strength meant that less was needed, so 19c iron ships were 10-25% cheaper. Iron hulls also lasted two or three times longer than wood ones (White 412).

Iron ships usually came in two basic flavors, the all-iron ship, and the composite, which had an iron frame and wooden planking and decking. (Svensson 62). Composites were favored for tropical waters, where copper sheathing was necessary to protect against marine borers; iron set against copper would experience bimetallic corrosion (Lewis 117). Another variation was seen in the Great Republic (1853), 335' long; it was mainly wood, but its hull was reinforced with diagonal strips of iron.

It is important to recognize that there wasn't a rapid transition from wooden to iron ships; the two types co-existed for decades. Iron ships were not only more expensive to construct than unsheathed wooden ships, they had nasty effects on ship's compasses. The bottom of the all-iron ship was a haven for barnacles and seaweed, increasing skin drag if not cleaned frequently. So their maintenance cost was higher than for sheathed wooden ships. And iron corroded three times as fast as copper. (McCutchan 110, Atwood 299, White 415).

An advantage of iron plating, over wood planking and decking, was that the iron plates could be bent easily. But of course the iron added more weight to the hull.

Mild steel was 25-30% stronger than iron, allowing a saving of 20% in weight of scantling and 13-15% overall, but in 1880 it was 50% more expensive (White 429ff).

 

Hull Form

 

As the ship moves, water is parted by the bow, and passes around and under the hull, rejoining at the stern. If there is separation of the flow from the hull, eddies will form where the water returns. The energy to form these eddies comes from the ship's propulsion, so these flow disturbances are felt as "form resistance". Separation tends to occur where there is an abrupt change in underwater hull form. (ChapelleSSUS 49).

The choice of hull shape isn't easy. For example, the 3D shape which would yield the minimum wetted surface for a given volume, and thus the least skin drag for its displacement, would be a hemisphere. However, the fluid flow at the fore and aft "ends" would be kinked, creating significant form drag. (Gougeon 32).

To reduce form drag, water must be moved out of the way and back again more gently, i.e., the ship needs a streamlined shape. Fish bodies have offered inspiration to ship designers for centuries (and there are old plans which actually depict a fish body beneath the hull diagram).

Tapered shapes reduce form drag and frictional resistance, but also reduce both capacity and the ratio of capacity to resistance.

Midship Section Position. Imagine that the ship is sliced into vertical sections, like a loaf of bread. The section with the largest beam is called, somewhat misleadingly, the "midship section."

A good ship, old salts said, should have a "cod's head and a mackerel's tail" (a teardrop shape, with the midship section forward). In keeping with the adage, the seventeenth-century midship section was actually located about one-third keel length from the fore end of the keel. That yielded a short full (broad) bow and a long fine (narrow) stern. (BakerCV 20-21). A Dutch merchantman shown in Furttenbach's Architectura Navalis (1629) exemplifies this shape. I would estimate that quarter-length from the bow, it is a third broader than a quarter-length from the stern. (Landstrom 146). On the Mayflower replica, based on Matthew Baker's manuscript, the midship section was placed 21 feet from the forward end of the 58 foot keel. (BakerNM 80).

In later centuries, the midship section was moved aft, to true midships, or even somewhat aft (the "wedge" shape), the latter being touted by EB11 "Yachting." For our purposes, the key point is that the position of the midship section is something that the designers are going to argue about.

Midship Section Shape. On a ship plan, the sectional view of the ship is the view from the front. We need to consider both the underwater portion (the bottom) and the abovewater lines (the sides).

Bottom. A semicircular underwater section requires the least "wetted area" (which determines frictional drag) for a given capacity, and this was recognized by Georges Fournier (1595-1652) in his treatise Hydrographie (1643)(Laing 162). Unfortunately, it provides no stability, and hence is practical only on a multihull or if there is substantial ballast. Rounding makes the hull "tender"; a small degree of tilt produces only a small righting tendency so the hull heels easily and recovers slowly. However, if deep-ballasted (see below), ...