Draw It, Measure It, Compare It

A yacht begins taking shape before any material reaches the building floor. Drawings locate its surface and connections; calculations describe how it should float and move; a design brief gives those choices a purpose. We can now bring the physical relationships together to interpret that proposed boat and compare it with another.

Start with the builder’s question: how do you recover a curved hull from marks on a flat sheet?

Locate one point on the hull

Looking at a hull from the side conceals its width. Looking down conceals its depth. A lines plan overcomes that loss by combining related views of the same surface. Its curves show where imaginary cutting planes intersect the hull. The cuts let us recover three-dimensional shape from distances that can be drawn and checked on flat paper.

First establish references fixed to the boat. The baseline is a declared straight reference for heights, usually drawn longitudinally below or near the bottom of the hull. It need not be the water surface or a physical timber. The centreline provides the reference for sideways distances. A station is a specified position along the hull at which a transverse section is taken. Stations may be regularly spaced, with extra ones where the ends change quickly; they need not coincide with built frames.

At one station, imagine the transverse cutting plane passing through bottom and sides. The outline of that section appears in the body plan, the view looking along the boat. Several sections can share this view, labelled by station. On a symmetrical hull, showing forward sections on one side of the centreline and after sections on the other saves overlap without implying that the real port and starboard sides have different shapes.

The half-breadth plan looks down on one half of the hull. Its longitudinal reference is the centreline; distances across the page give half-widths. Horizontal cuts at several specified heights produce the curves called waterlines in this drawing. Only the one corresponding to a particular flotation condition is that boat’s actual waterline. The others remain useful cuts through its shape.

The sheer plan, or profile, looks from the side. Length runs along it and height above the baseline runs upwards. The familiar sheerline, following the deck edge, appears here with its rise towards the ends. Sheer describes that longitudinal rise; neither the sheer nor its line supplies the transverse flare of a section. The body plan must agree with the profile at the deck edge while supplying the width the profile conceals.

Choose station 2 and a particular horizontal waterline. Their intersection with the hull has one height and one half-breadth. In the body plan, find the height and read sideways to the station-2 section. In the half-breadth plan, find station 2 along the length and read sideways to that waterline curve. Both readings must give the same half-breadth. The profile supplies the waterline’s height and the station’s longitudinal position. Three views are now describing one point rather than three vaguely similar boats.

The Gougeon brothers’ boatbuilding text gives a concrete notation in its table of offsets. At station 2, the entry for waterline 2A is 1-8-5: one foot, eight inches and five eighths of an inch from the centreline. It is a half-breadth, so the corresponding full width of a symmetrical section at that level is twice the distance. The entry does not give the height of waterline 2A; that comes from its separate reference on the drawing. Reading the number without the row, column and units loses the point it locates. Gougeon, fifth edition, chapter 16, Figures 16-2 and 16-3.

Such a reference-to-surface distance is an offset. A table of offsets organises many of them by station and cutting line, allowing a builder to transfer coordinates more accurately than by measuring every point from a small printed curve. An offset is one coordinate distance; a complete point also needs its other references.

Two more cuts help check the surface between those points. A buttock line is the intersection with a vertical longitudinal plane at a stated distance from the centreplane. It appears as a length-and-height curve in profile. A diagonal uses an oblique longitudinal plane, cutting through the section at a slant. Its intersections can expose a hollow or hump that is difficult to see in the other views. The diagonal is a geometrical check through the hull, not a diagonal plank or brace fitted to it.

A point shifted to smooth one curve must still agree with the others. This is why the lines are a connected description: correcting a buttock while leaving contradictory waterline and body-plan points would not define a buildable surface.

Carry the lines into the workshop

On a small drawing, the thickness of a pencil line can hide a consequential full-size error. Traditional lofting transfers the lines and offsets to full size, checks their agreement and fairs the curves before parts are made. The mould loft is the working space used for this layout. It is a place in the building process, not another moulded component aboard.

The builder marks the baseline, station grid and other references, transfers the offsets, then bends flexible battens through the resulting points. Nails placed beside a batten or weights hold its curve without requiring every imperfect point to dictate a kink. Heights and breadths are checked between views; corrections are carried through the connected geometry. A fair line follows the intended changing curvature without an accidental local bump. The operation resolves the shape, rather than merely enlarging an attractive sketch.

The checked outline can then be transferred to a mould, in the timber-template sense encountered during construction. A frame mould gives the form needed to mark or assemble that frame. It does not become part of the finished hull simply because a structural timber was cut to it. In composite production, the female mould has another job: its receiving surface supports the laminate while the shell is formed. That surface also has to embody the intended geometry, but a production mould receiving resin and reinforcement is not the flat template used to mark a timber frame.

Computer-generated geometry and cut templates can carry the design through a different route, avoiding much traditional floor layout. The need for consistent coordinates, a fair intended surface and correct allowances survives. Gougeon’s fifth edition acknowledges this newer practice while explaining the manual method in detail. Both routes depend on the same agreement between views. Gougeon, chapter 16.

We also have to know which surface the dimensions describe. A moulded dimension refers to a specified construction surface, often excluding outer shell thickness. Curtis’s 1919 timber-ship account measures moulded beam to the outside of the frames, whereas extreme breadth includes the planking outside them. The metal-hull convention illustrated in Aalto’s naval-architecture notes uses the inside of the plating. Neither label can be substituted for the maximum outside width until the construction reference is known. Curtis, glossary, pages 26–27; Aalto, dimensions and forms.

The same drawing also specifies structural members. On the deck beam already examined, moulding gives its depth and the sided dimension its fore-and-aft width. These belong to its scantlings, chosen for its loads. A lofted outline supplies the shape to make; the scantlings supply the material dimensions within it. MCA, wood construction, Figure 7.36.14.

Where planking enters the backbone, the drawing must also locate the actual joint. A rabbet line marks a boundary of the recess receiving the plank edges or ends. Curtis distinguishes inner and outer rabbet lines from the bearding line, where the inner planking surface meets the backbone’s surface beyond the bevelled region. Their separation describes a joint with thickness and changing angle, not several names for the same decorative stripe. Transferring the appropriate lines helps the plank fit against the supporting wood while preserving its outside form. A different construction, including Gougeon’s added outer laminated stem, can avoid that particular rabbet arrangement. Curtis, page 27 and Figures 28–30.

Spiling takes the transfer problem in the other direction. Instead of taking a frame shape from the loft, a batten records the required plank edges on the partly built hull. Marks or compass arcs preserve their positions relative to the batten; laid on fresh stock, it lets the builder recover the curved cutting line. Curtis’s garboard example transfers the upper edge and the inner-rabbet fit separately, with bevel allowances still required. Lofted frame geometry, an actual plank-fitting surface and the finished plank must agree, but they are different stages of the work. Curtis, pages 172–176, Figures 159–160.

Add the immersed slices

The same section that locates a frame can tell us how much of the hull is underwater. Stay at one station and move an imaginary horizontal water level upwards across its body-plan outline. At each level, measure the area enclosed below the water and inside the selected hull surface. A wider region adds more area for the next small rise than a narrow region does.

A Bonjean curve records that station’s immersed sectional area against water level. It belongs to one transverse slice. In the arrangement shown in the Aalto notes, height is read vertically and area horizontally from the station’s reference. A chosen water level intersects the curve and gives the area of that section below it. Repeating the reading at a higher level gives a different area. Neither reading is yet the whole hull’s volume.

To reach volume, move along the hull. Choose a waterline, including any stated trim, and find its height at each station. Read the corresponding area from each station’s Bonjean curve. Now consider a short length of hull around one section. Sectional area multiplied by that short length gives an approximate slice volume. Adding the slices accumulates the immersed body. More closely spaced readings, especially where the ends change quickly, can represent the varying shape more faithfully.

For a deliberately simple illustration, a one-metre slice with a representative immersed area of two square metres contributes about two cubic metres. Successive areas and their spacing capture the changing ends as well as the broad middle. Numerical integration refines this accumulation. Aalto, sectional areas, volume integration and Figure 5-9.

Repeat the whole-hull calculation at several water levels. A displacement curve then relates draught to displaced volume, or to displacement mass using the stated water density. Read its axes to identify which. The relationship lets us find the flotation level for a particular load.

Hydrostatic curves gather several such still-water properties against a common draught or displacement reference. Displacement, waterplane area and the positions of buoyancy-related centres each have their own scale and units. The shared reference makes these connected calculations convenient to consult.

The promised connection to stability now follows. At a specified displacement and held heel angle, the immersed geometry determines both the required volume and its centre of buoyancy, B. The buoyant force acts vertically through that B. Repeat the calculation for different displacements while retaining the same heel angle, and record the arm to that force line from a declared reference for G. Repeating at other angles creates cross curves of stability.

The US Naval Academy example places displacement horizontally and reference arm vertically, with one curve for each fixed heel angle. Its assumed G is at the declared base-plane reference K. To construct a righting-arm curve for a real loading condition, select that displacement across the family, read the reference arms, then correct for the actual G.

With G on the centreplane above K, the correction subtracts the sideways separation introduced by KG as the hull heels: GZ = KN − KG × sin(heel angle). Here KN is the reference arm to the buoyancy vertical, and the sine factor converts the tilted KG distance into its horizontal component.

Plot the corrected GZ values against heel angle and we have the righting-arm curve for the selected loading. Raising G reduces those arms, as the earlier stability lesson showed. This calculation retains the stated density, trim, intact buoyant volume and quasi-static condition; it does not describe a rolling boat in waves. The cross curves let us reuse the hull’s geometry while changing the assumed mass distribution. US Naval Academy, chapter 4, pages 4-4–4-13.

Give fullness a denominator

The complete lines contain more shape information than a comparison often needs. A coefficient can retain one useful proportion, provided we say what was divided by what. Use a hypothetical immersed hull body with a waterline length of 10 metres, reference waterline beam of 3 metres and hull-body draught of 1 metre. Exclude a separate fin keel. Let its volume be 14.4 cubic metres, its immersed midship area 2.4 square metres and its waterplane area 21 square metres. In this example the maximum immersed transverse area and the widest part of the waterplane are both at midships.

The block coefficient, Cb, compares the hull’s displaced volume with the enclosing reference box. That box is 10 × 3 × 1 = 30 cubic metres; 14.4 ÷ 30 = 0.48. The hull fills 48 per cent of that volume. The reference length, beam and depth are all needed. Substituting overall length or the tip depth of a separate keel would create a different box.

The midship coefficient, Cm, compares the immersed midship section with its beam-by-draught rectangle. Here the rectangle is 3 × 1 = 3 square metres; 2.4 ÷ 3 = 0.80. It describes how fully that section occupies its rectangle. It says nothing by itself about how far the section continues towards either end.

The prismatic coefficient, Cp, adds that longitudinal question. Imagine carrying the maximum immersed section unchanged along the reference length to make a prism. Its volume is 2.4 × 10 = 24 cubic metres; the actual hull’s 14.4 ÷ 24 = 0.60. This coefficient describes the hull’s volume relative to that constant-section body. Some references use the midship section in the denominator. Our maximum is amidships, so the two conventions coincide here; they would need distinguishing on another hull.

The waterplane coefficient, Cwp, compares the waterplane area with its length-by-beam rectangle. Here 21 ÷ (10 × 3) = 0.70. It describes the horizontal cut’s fullness, not the fullness of the immersed midship section or the whole volume. The four coefficients are dimensionless because compatible areas or volumes divide out, but each has kept a different piece of the shape. Aalto, form coefficients, pages 23–25.

Different hulls can share all these numbers while placing breadth, curvature and volume differently. Move waterplane area from one part of the length to another while preserving its total and Cwp does not notice. Alter the section’s curvature while retaining its area and Cm does not notice. The lines and the load cases remain necessary when the question becomes resistance, motion, structure or control.

Measure the space for the sails

The rig has an equivalent temptation: replace the whole sail plan with four letters. They are useful dimensions of a defined arrangement, once their endpoints have been established.

Take a conventional mainsail and foretriangle with visible measurement bands. Under PHRF Southern California’s stated convention, P runs from the top of the boom to the lower edge of the upper mast band. E runs from the mast’s after surface to the inner edge of the outer boom band; an external mast track changes that starting surface. These measure available rig space, not the length of the cloth’s luff and foot.

For the forward triangle, I is vertical from the local sheerline level to whichever is higher: the forestay’s intersection with the mast face or the top edge of the highest headsail-halyard sheave. J is horizontal from the forestay’s intersection with the sheerline to the mast face. That authority changes the forward reference when a headsail is tacked beyond the stay. Thus I is not the sloping forestay length, and J is not necessarily the full projecting bowsprit. PHRF Southern California, rig measurements.

These are named conventions rather than universal abbreviations carved into the boat. The Equipment Rules of Sailing specify upper, lower and outer points and the relevant edges of limit marks; IMS applies its own rig definitions and symbols. A certificate must be read under the rule that produced it. A tape stretched along a different route can give an accurate distance with the wrong meaning. World Sailing, Equipment Rules 2025–2028, sections F2, F3 and F6; IMS 2026, rig measurement.

For a geometrical example, express the chosen endpoints in metres: P is 12 and E is 4. A right triangle filling that height and base has an area of 12 × 4 ÷ 2 = 24 square metres. A foretriangle with I of 13 and J of 4 similarly gives 26 square metres. These are reference triangles. A mainsail’s roach, head shape and attachment clearances prevent the first number from being an exact cloth measurement.

The earlier luff perpendicular, LP, measures the shortest distance from the headsail clew to its luff. If J is 4 metres and LP is 4.8 metres, LP/J is 1.20, or 120 per cent. If that sail were an exact flat triangle with a 13-metre luff, its area would be 13 × 4.8 ÷ 2 = 31.2 square metres. That calculation uses luff length and LP, not I and J. Curved edges and the rule’s sail-measurement method still determine the actual reported area. World Sailing, headsail dimensions, section G7.

Cloth area, a measured flattened-sail area, a chosen geometric reference area and the area projected towards a particular view or flow are therefore distinct. Rated area adds another possibility: a rule can calculate the quantity it chooses to charge, rather than directly count every square metre of exposed cloth. That choice will matter when we reach racing rules.

The letters also leave the rig classification unfinished. A mainsail with the same P and E could accompany one ordinary headsail or two simultaneously working forward sails. I and J describe one forward reference triangle; they do not tell us whether a second stay supplies a cutter’s inner working sail, whether alternative headsails are used one at a time, or whether another mast stands aft. The earlier sloop, cutter and divided-rig distinctions still require actual supports and sail use.

Put the boat into its measurement condition

Before comparing two sets of particulars, return to the hull drawing and declare the water surface. A load waterline belongs to a specified load condition; a boat’s present loaded waterline is the actual intersection for what it is carrying now. A design waterline represents the designer’s selected condition. Added stores can move the actual surface away from that line, changing immersion and perhaps trim.

In the sheer plan, follow the selected waterline to its forward and after intersections with the hull. Their longitudinal separation gives LWL, waterline length. The forward and after overhangs extend from those intersections towards the respective hull ends. They can change when the water level changes even though the boat has acquired no new timber. A bowsprit projects beyond the bow for another purpose; it is not automatically included in the hull’s forward overhang.

LOA, overall length, needs a statement about whether projecting fittings and spars are included. LOD, length on deck, follows its stated hull/deck endpoints instead. LBP, length between perpendiculars, uses the prescribed forward and after reference planes, which must be identified for that drawing or rule. None can silently replace LWL in a calculation. IMS, for example, excludes removable projections from its LOA except when they contribute to the hydrostatic length. A marina’s occupied length can still include the projecting gear. IMS 2026, hull dimensions, section B6.

Use the half-breadth plan to read the width of that waterline. Its largest width gives waterline beam, which need not equal maximum beam higher on the hull. Check the corresponding station in the body plan to see the difference. Hull draught runs down from the water surface to the declared lowest boundary, with keel or board state stated. Air draught runs upwards from the water to the declared highest point, including the specified equipment. The baseline helps locate their endpoints on the drawing, but a height above the baseline is not already a clearance above water.

For berth width, use the finished outside surface. The moulded breadth and offsets locate their declared construction surface; shell thickness and any corrections made in lofting complete the route to the actual boat. A production mould repeats its chosen geometry, so its reference must agree with the dimensions we are comparing.

Loading needs comparable precision. IMS2026 defines a measurement condition containing specified fixed equipment while excluding such items as crew, sails, anchor and rode, and portable stores. Liquid conditions are also specified or recorded. A yacht prepared for a passage carries many of those omitted things. Freeboards measured at designated hull positions, together with the recorded geometry and water density, let the measurement calculation establish its flotation condition. They do not make that state a universal cruising load. IMS 2026, sections B3–B5.

If a stationary yacht gains 1,000 kilograms, it must displace water containing that additional mass. At a stated density of 1,025 kilograms per cubic metre, the extra volume is about 0.976 cubic metres. The displacement curve connects that increase to the hull’s new draught; the distribution of the added load also affects trim.

Count the sails carried, then those set

A sail inventory is the collection of sails carried or specified for the boat. It can include alternative headsails, reefable working sails, heavy-weather sails and flying sails for more open angles. Adding their areas does not give the sail area used at one instant. Some replace one another; some work together; some require extra sheets, poles, stays or handling work before they can be used.

The brochure for the 2010 cutter-ketch Twizzle lists mainsail and mizzen, several forward sails, a gennaker and a mizzen staysail separately. The two principal masts supply a starting sail distribution, while the extra inventory gives choices within it. The crew does not gain a single enormous set by summing every listed area. Selection changes the cloth exposed, its support routes and the work of hoisting, trimming, reducing and stowing it. Twizzle brochure, particulars on page 21.

For a comparison of two boats, state whether “sail area” means main plus ordinary headsail, includes an overlapping genoa, includes both cutter headsails, or describes a flying-sail set. Also state which reefs or alternatives are being considered. A sail carried in a bag still contributes mass and storage demand while contributing no aerodynamic drive.

Separate tonnage from displacement

The same Twizzle particulars put 550 tonnes displacement at the design waterline beside 498 gross tonnage. The first is a mass associated with a specified flotation level. The second follows a volume-based measurement convention. It is not 498 tonnes of yacht and does not contradict the displacement figure.

Under the international gross-tonnage convention, enclosed volume is measured with defined boundaries and adjusted by a coefficient that varies with its size. In the formulation reproduced in Ireland’s 2007 regulations, GT = K1 × V, with V in cubic metres and K1 calculated from that volume. The resulting gross tonnage is a dimensionless index, rather than either a mass or an unadjusted cubic-metre count. The convention has a stated international application; other vessels and periods can use other tonnage rules. IMO, 1969 Tonnage Convention; 2007 tonnage regulations, regulations 6–7.

The brochure pair is a useful tonnage-versus-displacement comparison because its two meanings are stated. Loading stores increases displacement mass and changes flotation. Enlarging a qualifying enclosed space changes the volume entering gross tonnage; the structural alteration may also add mass, but through a separate physical effect. Changing a measurement rule can change a tonnage figure without changing the boat. Neither a single empty-space addition nor a new rule tells us a new displacement automatically.

Historical tonnage could influence the hull itself. Michael Jarvis’s study of Bermudian shipbuilding reports an eighteenth-century formula which took extreme length minus breadth, multiplied that result by breadth and half-breadth, then divided by 94, using feet. Half-breadth supplied an assumed depth in this calculation. It was a capacity-related measurement convention used in the economic system, not a weighing of the vessel afloat.

Holding length constant in that formula, breadth affects several factors. Builders therefore faced a measurement incentive alongside the physical advantages and costs of a particular hull. The rule did not dictate the finished boat: cargo, materials, handling and service still mattered. The account shows why a reported tonnage must be read with its date and method, rather than converted into floating mass by attaching a modern unit. Jarvis reports the earlier manual; his thesis is the source for this historical formula here. Jarvis, 1992, pages 31–33.

Recalculate the same boat with more aboard

Ratios can now compare stated quantities without pretending to replace the boat. Use a hypothetical yacht with a 30-foot waterline and a selected main-and-headsail area of 600 square feet. Compare two declared loading states: 11,200 pounds and 13,440 pounds of displacement. To isolate the arithmetic, keep the reference waterline length and selected sail area fixed. An actual boat’s LWL may also change as it sinks and trims, requiring another calculation.

The customary displacement-length ratio, D/L, used in PHRF Southern California’s parameter table divides displacement in long tons by the cube of one hundredth of LWL in feet. A long ton here is 2,240 pounds: that conversion belongs to this formula’s convention. The first displacement is therefore 5 long tons. One hundredth of the 30-foot LWL is 0.30; its cube is 0.027. The ratio is 5 ÷ 0.027, about 185.

The second displacement is 6 long tons, giving 6 ÷ 0.027, about 222. We have changed the numerator while preserving the chosen length reference. This is why a D/L quoted without its load state can mislead even when the formula has been applied correctly. Substituting 30 metres, short tons or LOA would not be the same convention.

The sail-area/displacement ratio, SA/D, in the same table compares sail area with an area scale derived from displaced volume. In its feet-and-pounds convention, divide displacement by 64 pounds per cubic foot to obtain the reference salt-water volume. Taking the cube root gives a length scale; squaring that length gives an area scale. The notation for those two operations is volume raised to the power two-thirds.

For the first state, 11,200 ÷ 64 = 175 cubic feet; 600 ÷ 175^(2/3) is about 19.2. For the second, 13,440 ÷ 64 = 210 cubic feet; 600 ÷ 210^(2/3) is about 17.0. The sails have not shrunk. Their selected area has become smaller relative to the displacement-derived scale. The density and the definition of sail area are part of the comparison. PHRF Southern California, calculated parameters, DLR-LT and SA/D.

These numbers can organise a comparison or expose inconsistent inputs. They do not calculate a passage speed, a sheet load or power. D/L omits the detailed distribution of the immersed body; SA/D omits the sails’ shape, force direction, height and working airflow. Neither states what the crew can carry, control or maintain. To investigate those consequences, we need evidence of the boat working.

Measure progress towards something

Point the bow closer to the wind and watch the speed fall. The boat may now look more purposeful while reaching the windward destination more slowly. To compare the two states, choose the direction in which progress matters and take the component of actual velocity along it. That component is velocity made good, or VMG.

For a water-relative windward comparison, use speed through water, STW, and the actual track through water. Suppose that velocity is 6 knots at 60 degrees from the desired windward direction. Its component towards windward is 3 knots: the six-knot movement has only half its length along the chosen axis. At the same speed on a track directly across that axis, windward VMG is zero. The boat is still sailing, but it is not gaining in that direction. A knot means one nautical mile per hour; VMG is another speed, not a distance or a force.

The heading is where the boat points. Leeway can put its water track farther from the wind than its heading suggests. For a destination fixed to the ground, use ground-relative motion instead: speed over ground, SOG, and course over ground, COG. Current can change that progress without changing STW. An intended course is the direction we mean to follow; it is not proof of the track achieved, and it has no connection here with a square rig’s lowest course sail. A wake may help reveal recent water-relative movement, but it cannot by itself supply the complete ground track and current correction.

Pointing ability therefore concerns useful sailing close to the wind, including speed and leeway. Holding a heading close to the wind with almost no forward progress is not strong evidence of it. The best angle for progress can be farther off than the closest heading at which the sails remain filled.

A polar diagram organises speed by wind angle and wind strength. In a common sailing polar, distance outwards from the centre represents boat speed, while angle around the centre represents a stated wind angle. Separate curves represent different wind strengths or sail sets. Choose the curve first, follow the required angle outwards and read the speed where they meet. A true-wind-angle plot and an apparent-wind-angle plot require different readings even when they concern the same boat.

The ORC’s published 2020 sample Speed Guide makes the comparison concrete. In its headsail table at 10 knots true wind, a 39-degree true-wind angle gives about 6.7 knots boat speed and 5.2 knots windward VMG. At 52 degrees, boat speed rises to about 7.5 knots while VMG falls to about 4.6. At 90 degrees, the table gives about 7.8 knots boat speed and zero windward VMG. These are rounded values from a prediction example, not three observed trials of a boat carrying our hypothetical 600-square-foot sail plan. ORC sample, headsail table, page 10.

The sample’s other plots and tables describe different sails, including symmetric and asymmetric spinnakers. They let us ask which selected configuration is predicted to improve a particular leg, rather than treat the fastest point on any curve as the boat’s universal speed. The ORC guide also explains that actual achievement depends on such things as sail condition, crew, instruments and waves. The curve is a stated target whose origins must be known. ORC, Speed Guide explanation.

Separate an outline from its force

To understand where a prediction comes from, return to the geometry of a sail or foil. Its planform is its projected outline in a specified view: for example, the span-and-chord outline of a foil, rather than the curved section seen edge-on. That outline lets us compare how area is distributed along the span. It does not include both physical faces as two separate planform areas, and it does not specify the section’s camber, thickness or working angle. NASA, wing geometry.

Pressure varies over the loaded surface. A region carrying greater pressure difference contributes more force than an equal-sized lightly loaded region. Combining those contributions gives a resultant pressure force and an effective location through which its line acts: the centre of pressure. Change the surface shape or flow and that location can move. Measuring or computing the pressure distribution therefore gives information that the outline alone cannot supply. NASA, centre of pressure.

This is the reason for the earlier caution about the sail plan’s centre of effort. A geometric centre drawn from areas is a useful design reference; an effective working sail-force location depends on the loading. The pressure centre is not a new name for every use of centre of effort, nor a fitting fixed at the centroid. Several interacting sails and their rigging add further loads and moments to the whole rig.

The lift-to-drag ratio also needs working forces. If a selected foil condition gives 600 newtons of lift and 100 newtons of drag, L/D is 6. Both forces use the same relative-flow reference. It is not the geometric aspect ratio and not a yacht’s drive divided by displacement. A finite span creates a trailing flow and induced drag associated with producing lift; surface friction and other pressure losses also contribute. A change that reduces one contribution need not reduce total drag by the same amount. NASA, lift-to-drag ratio.

Those distinctions set the questions for a test. We may need total tow resistance, sail forces and moments, local pressure, or the complete yacht’s speed. No one instrument measures all of them simply by being accurate.

Make a smaller boat, then control the test

A scale model reproduces chosen geometry at a different size. It can preserve the proportions of hull, keel and rudder, while its mass and loading are set for the intended experiment. A museum display model may establish a represented arrangement; a test model adds a declared measurement purpose. Reducing every length does not automatically reduce every physical effect in the same proportion.

In a tow-tank test, a carriage moves the model through water at a controlled speed. A force-measuring instrument in the towing connection records the force needed to pull it. The arrangement must allow the motions being studied and avoid adding unwanted restraint or force through guides and cables. Model loading, trim, water temperature, speed, sinkage and running trim are recorded with the result. The ITTC resistance procedure describes those controls for displacement-vessel tests. Its measured resistance is the result for that model, fitted state and test condition. ITTC, resistance-test procedure, revision 05, 2021.

Scaling then matters. Wave-making similarity uses speed relative to the square root of length and gravity, while viscous flow depends on a different combination of speed, length and fluid properties. A small model in the same water cannot normally match both merely by choosing one towing speed. Frictional treatment, roughness and other corrections enter the move from model result to full-size prediction. A force read from the carriage is measured; the full-size result derived from it contains a model of those differences.

A wind-tunnel test instead passes controlled air around a model rig. The model’s supporting arrangement transmits loads to a balance that measures forces and moments. The way it is supported determines which loads reach that instrument. If we want one sail’s force separately from the rest of the rig, we must separate its measurement route as well as name it separately.

Richards and colleagues described such an arrangement in 2006. Their model, at 1:12.5 scale, normally loaded a balance beneath the hull. They could instead support the spinnaker head from a crossbeam and its lower corners from independent floor-mounted supports, taking that sail’s attachment loads away from the hull balance. Comparing the two arrangements helped isolate its contribution. Controlled flow above the model represented the chosen wind profile; the apparatus did not become a full sea state because the sails looked realistic. Richards and colleagues, independent supports, sections 2.1–2.2.

The force route is the familiar one through sail corners and supports, now arranged so an instrument can distinguish contributions. A number attributed to “the spinnaker” needs that route or an equivalent measurement method. A balance under the complete boat would otherwise receive the combined transmitted loads.

Compute a flow, then predict a sailing state

Computational fluid dynamics, usually shortened to CFD, calculates an approximation to fluid flow around specified geometry. The calculation divides the flow region into many small parts and solves a chosen set of flow relationships across them. Geometry, fluid properties, speed, attitude, free-surface treatment and modelling choices affect the result. Its calculated pressure field can show how loading varies over that specified surface.

MARIN’s 2018 account of sailing-yacht design methods distinguishes a detailed viscous-flow calculation from faster methods suited to exploring more candidate shapes. It connects both with test data used to assess prediction quality. A useful calculation can examine a shape before a test model is built; comparison with measured results then checks whether its assumptions are adequate for that purpose. MARIN Report 124, pages 12–13.

A velocity prediction program, or VPP, asks a connected question: what steady sailing state satisfies its chosen models of air forces, water forces and stability? Give it hull and appendage geometry, rig and sail data, loading and mass distribution, and a wind condition. It tries a boat speed and attitude, obtains the apparent wind, calculates sail drive and heeling demand, and compares them with hydrodynamic resistance and righting response. It changes its trial values until the selected balances agree. The models may use empirical test relationships, computed flow results or both.

The ORC2026 VPP provides a declared example. It balances drive and resistance along the track and heeling and righting moments, with a model of the side force needed to oppose the sails. Its hydrodynamic calculation associates that force with leeway. It also explores permitted sail reduction, flattening and crew position to find the modelled performance. Changing loading, foil geometry or the assumed sail set can therefore change the predicted speed, heel and leeway. Close-hauled sailing is one case; the same program evaluates other wind angles to build a polar. ORC, VPP Documentation 2026, sections 1–3.2.

That documentation explicitly omits yaw-moment balance. It does not solve every steering demand merely because its forward force and heel equations converge. The real yacht must also carry the rudder forces and turning moments already examined. Likewise, a steady smooth-water result cannot establish response to every wave, gust or manoeuvre. A VPP is an estimate of a defined state, not a complete simulated voyage hidden behind a speed number.

These methods recover information our earlier ratios discarded. Two hulls with the same Cb can have different ends and resistance; equal Cp can conceal different distributions of the volume. D/L omits where the mass lies, while SA/D leaves out the sail loading and its induced losses. Use the detailed geometry and force model when those differences matter to the proposed choice.

Conditions remain part of the evidence. Sea state describes the surface-wave conditions in which the boat works; swell can arrive from weather elsewhere and add another direction and period to local waves. They affect motion, resistance and the ability to keep sails and appendages working. A rapid trial in flat water does not establish general seaworthiness, which also depends on structure, stability, water exclusion, handling and crew in the intended service. The earlier stability curve and the new speed polar each help answer a particular part of that question.

Independent scrutiny supplies another kind of evidence. A classification society publishes technical rules and checks the relevant design and construction against them. Its surveys examine the built structure and equipment, with further surveys during service. A class certificate records compliance with the applicable rules within its scope. This is independent technical assessment of a vessel, distinct from the racing classifications considered below. IACS, classification and surveys.

The sailing yacht Pink Gin VI, launched in 2017, supplies a concrete rig example. DNV GL reviewed the rig’s materials, dimensions and loads, assessed its physical components and monitored mast stepping. It then compared sensor measurements from specified sea-trial manoeuvres with calculated loads and performed a final inspection before issuing the Rig Design Certificate. The evidence connected design, built structure and observed sailing conditions. DNV, Pushing the limits, rig-design account.

The examination’s scope remains part of the result: establish which structures, equipment and conditions it covers. A hull or machinery assessment alone does not demonstrate that every rig component or operating choice was examined. A class certificate is not a guarantee of seaworthiness or every response to failure; operation and maintenance remain the owner’s and crew’s responsibilities. IACS, scope of classification, pages 3 and 5.

Test an interpretation of the past

Historical evidence introduces another boundary. An experimental reconstruction is a newly made object intended to test an interpretation of earlier evidence. Its useful question is what the reconstructed arrangement can do under recorded conditions. Missing original material, substituted materials and choices made by the builders remain part of the answer.

Ma‘agan Mikhael II supplies a concrete example. Its wreck-based hull carried a reconstructed square sail and paired quarter rudders, but the complete original rig had not survived. The 2021 trial account describes modern practice cloth and changes to gear as the crew learned what the new arrangement required. The tests therefore concern a particular physical interpretation, with real sail, yard, brace and steering loads, rather than a recovered ancient performance certificate.

Its published polar used speed through water and true-wind angle relative to heading. The authors explicitly noted that leeway was not included in that plot. We cannot read an angle from it, call that the water track and calculate windward VMG as though the distinction had disappeared. Measured sailing can be valuable while answering a narrower question than the one a later reader wants to ask. Palzur and Cvikel, Sailing Ma‘agan Mikhael II, 2021.

A separate Polynesian-canoe study combined several methods. It tested two-metre model sails based on museum examples, compared three reconstructed hull forms through CFD and one-tenth-scale tank tests, then joined the sail and hull results in a VPP. The paired V-shaped hulls could contribute side force together. In an outrigger combination, changing the second immersed body’s contribution changed the leeway needed from the package. The result belongs to the combined sail, hull and loading assumptions; counting hulls alone does not predict pointing ability. The objects tested were model sails and hulls, not intact ancient craft or their original cloth. Irwin and colleagues, 2023, sail and hull performance sections.

Dates locate these kinds of evidence. A wreck date belongs to surviving material; a reconstruction’s launch date belongs to the new object; a trial date belongs to its observed state. A first surviving description is a first attestation, not automatically the moment of invention. The earliest surviving use of a word is narrower still. None of those dates, alone, shows how the original rig performed or proves descent between two similar-looking vessels.

Give two cruisers different jobs

We can now make the design brief explicit: the intended waters and passages, people and payload, acceptable workload, access, serviceability, cost and performance priorities. Those requirements determine which measurements and tests matter. Improving an answer means improving it for that brief, while keeping its other necessary conditions satisfied.

Cruising priorities vary within that category. Pogo Structures presents the Pogo 36 as a cruising yacht shaped by offshore-racing experience. Its published particulars combine a 10.86-metre overall hull, 4-metre beam, about 3.8 tonnes light displacement and a swinging keel giving stated draughts of 1.18 and 2.95 metres. Its cruising accommodation shares the platform with a substantial sailing plan and a moving appendage that changes access. Light construction and that keel are parts of a chosen package, with the structural and operating duties already examined. Pogo Structures, Pogo 36.

Rustler describes the Rustler 37 around offshore passage-making, accommodation and manageable work for a small crew. Its particulars give 11.28 metres overall, 3.76 metres beam, 1.91 metres draught and 8,845 kilograms displacement. The maker’s separate measurement explanation identifies its published displacement convention as lightship. Its standard sail and control specification includes a furling genoa and a reefable fully battened mainsail. The fixed keel, hull and handling arrangement are another cruising answer, rather than the same boat with ballast added to it. Rustler 37 particulars; Rustler’s displacement convention.

These are two cruising briefs, not a racer and a cruiser. They show different allocations of access, sailing response, accommodation and machinery. Their builder figures are not measurements of equally loaded individual boats, and neither selected model page supplies LWL for our D/L calculation. A claim that their mass difference alone proves one is faster, more comfortable or more seaworthy would discard the conditions we have just established.

A design trade-off is the connected gain and cost of a choice. Shallow access gained by retracting a keel comes with a housing, bearings, restraint and movement system; useful sail area brings attachment loads and handling work; more carried stores need both volume and displacement allowance. Those mechanisms let us compare benefits without treating a compromise as a failure to design properly.

Change the objective without changing the question’s scale

Now make a bounded racing-versus-cruising comparison. Consider two hypothetical briefs for an offshore monohull within about 12.2 metres overall, the same waters, two people and the same declared people-and-stores payload. Both require adequate structure, stability, water exclusion and workable sail handling. These shared assumptions are a design exercise based on the documented kinds of priorities, not invented test conditions for the Pogo and Rustler.

For the passage brief, use Rustler’s stated small-crew cruising priorities: reliable progress, usable accommodation and a workload that leaves capacity for watches, maintenance and the rest of life aboard. For the racing brief, use Class40’s documented offshore competition objective within its dimensional, mass, material and sail constraints. The result being sought is performance around the specified race, under those rules. The same two people and payload allowance do not make the objectives identical.

Take sail inventory as the constrained choice. An additional specialised sail could improve a predicted race leg but consumes storage, handling effort and allowance within the chosen rules. The racing brief may accept more changes when they produce a worthwhile result. The cruising brief may prefer a more adaptable working set when reduced sail-changing effort is worth more than the predicted gain. Neither choice abolishes the need for enough crew control in the same weather. This is cruising versus racing optimisation: selecting useful performance against the priorities of each brief.

Mixed use is a separate, documented possibility. In March 1891, the Herreshoff firm’s account of Gloriana combined racing in the 46-foot class with fitted cruising accommodation, including a saloon and separate sleeping spaces. The contemporary intention joined a competitive objective to life aboard one yacht. It does not prove that racing was a stage on the way to cruising or that every modern compromise descends from her. Herreshoff Marine Museum Chronicle 21, reproduced company statement.

A yacht is a vessel considered in its pleasure or sporting use; a sailing yacht identifies wind propulsion within that use. Vessel remains the broader object. The labels neither determine its rig nor prescribe its tonnage measure. Pogo, Rustler and the mixed-brief Gloriana can all be sailing yachts while answering different questions.

Read the working name beside the arrangement

The working vessel families already encountered show why a brief is useful alongside a rig name. Cutty Sark’s fine, composite-built hull and tall square rig served rapid tea passages with cargo aboard. The question was not simply how fast a narrow hull could sail, but how it could carry its paying load along the required route. A lines plan or trial becomes evidence for that carrying-and-speed problem. Royal Museums Greenwich, Cutty Sark’s tea trade.

Alma’s scow-schooner package answered another freight problem: shallow routes, a broad cargo space and sails manageable by a small crew. Her centreboard trunk occupied some of that space, while cargo height could interfere with the booms. Comparing the boats through those constraints explains more than ranking their rigs or quoting one dimension. Alma, HAER CA-60.

Service names add another layer. The pilot cutter Alpha, revenue cutter Active and sloop of war Constellation had the different sail plans already described. Their working titles identify an occupation or service classification; drawings and contemporary records identify the arrangement that performed it. The same discipline applies to the earlier cog, caravel, carrack and galleon examples: hull evidence, reconstructed rig and documented work must retain their own sources. A comparison can use that evidence without asking the vessel name to supply missing measurements.

We can transfer equipment between briefs too. The battened junk sail on Poppy’s Westerly hull retains its support and control jobs without acquiring the vessel history of Keying. A traditional sail system and a present cruising brief can coexist. The useful comparison follows what the chosen arrangement asks the hull and crew to do.

Let the rule change what is rewarded

Racing adds another stated environment. A one-design class controls a common design so that the permitted equipment choices, preparation and crew work operate within a constrained physical package. The February 2026 J/70 rules specify licensed hull and rig production and controlled sails, with their own certification and permitted variations. One-design does not mean zero manufacturing tolerance or identical results from every crew. J/70 Class Rules 2026.

A development class leaves deliberate design freedom inside stated boundaries. Class40’s August 2026 rules combine that freedom with, among other constraints, maximum hull length of 12.19 metres, maximum beam of 4.50 metres, maximum draught of 3 metres, a minimum measurement-condition displacement and material restrictions. Designers can choose different shapes within those limits; the class is not an unrestricted experiment. Its offshore, competitive and budget aims explain why freedom and restraint appear together. Class40, 2026 version 2.

The forward-shape rule applies the drawing language directly. At a station 2 metres aft of the hull’s forward length reference, the 2026 maximum breadth is 3.15 metres; another boundary controls the shape closer to the bow. The maximum does not prescribe every section between those stations. The class’s 2025 explanation of Sharp used a different test: less than 2.30 metres beam 2 metres aft of the bow defined that subgroup. A subgroup threshold and a whole-class envelope do different jobs. The historical working sharpie remains a different name and form. Class40 2025 guide, page 47.

A rating rule offers another route to comparing different boats. It uses selected physical measurements to estimate a performance allowance or class value. An empirical handicap instead derives its adjustment from earlier race results, which contain crew and condition effects as well as boat performance. Both require a scoring convention. Neither a numerical allowance nor corrected time removes every consequence of weather, decisions and handling. World Sailing, class rules and ratings systems.

Every rated feature can acquire two consequences: its physical effect and the cost the rule assigns to it. A change worthwhile on the water may be too expensive in rating; a modest physical gain may be attractive if the rule charges little for it. Legal design effort responds to that difference. The measurements have begun to shape the boat they measure.

The contemporary 1973 account of Jerry Milgram’s Cascade gives a bounded example under the International Offshore Rule, IOR. Its headsailless cat-ketch arrangement received 325 square feet of rated sail area and an initial rating of 22.00, according to the report. That was rated area, not a securely measured total of all its cloth. The article attributed the advantage to how the rule treated the divided sail plan, described strong reaching but weaker beating and direct running, and reported three race wins and fifth overall in that year’s SORC series.

Before that SORC series, the authority had applied an exceptional ten-per-cent increase to 24.2. Milgram then shortened the rig and made smaller changes to reduce the rating to 22.8 for the series. The example connects the measurement proxy, design choices, earlier competitive results and changed treatment. It does not show that the unconventional rig was fastest on every course, that the rating was a linear speed measure, or that one penalty ended development. Underwood’s March 1973 report, reproduced by Scuttlebutt.

ORC’s later VPP-based route uses more of the physical model, but it still selects measurement conditions, models and scoring assumptions. Better-supported estimates can improve a defined comparison; they do not cease to be estimates. One-design, development classes and measured or empirical allowances put the limits in different places rather than removing the design problem.

Put causes beside the dates

The dated sequence is a chronology: what is recorded before or after what. Technical history asks how a working problem, available materials and skills, alternatives and incentives contributed to a choice. The two belong together, but a chronology alone cannot explain why the boat changed.

Return to Bermudian shipbuilding. Jarvis’s account connects local cedar hull construction with imported materials for other parts of the vessel and with an economy of cargo, markets and measurement. A light local hull material did not supply every spar, fastening and rope. Builders needed material access, skills and viable construction arrangements as well as a promising shape. The tonnage formula affected one economic constraint while the actual vessel still had to carry its trade. These are causes and limits to examine, not a deduction that every narrow hull was created by taxation. Jarvis, pages 31–33 and 50–51.

Working craft and leisure craft also exchanged solutions. John Leather’s historical introduction describes yacht improvements being imitated in some commercial craft, alongside later yacht use of earlier working practice, experiment and rediscovery. The earlier gaff and triangular-sail lessons showed why a changed support could alter mast height, mass aloft and handling without winning every comparison. Dates help locate the documented arrangements; resemblance and a later name do not fill gaps in their descent. Leather, The Gaff Rig Handbook, historical introduction, pages 9–10.

Voyages and trials then feed back into repair, alteration, borrowing and replacement. Two communities can face similar water and choose differently because their materials, markets, labour and institutions differ. A racing result may expose an incentive in a rule; a difficult hoist may expose a weakness in an attachment; a construction method may make a previously awkward shape practical. Technical history follows the supported connection rather than promoting the newest silhouette into the destination of all the others.

We now have ways to recover the intended boat, describe its measured condition and compare predicted or observed results with a brief. The next complication is inside the working machine itself. The crew may ask a powered system to move the load, hold a target or make repeated corrections. That saves some work and introduces another chain to understand: where its energy comes from, what controls its action and what happens when the normal route is unavailable.