This article about seagoing rowing boats was written by Nick Newland and appeared on the Duckworks magazine website. Nick’s thoughts were influenced by and included in the design of SwallowBoats’ Teifi Skiff.
Design of Seagoing Rowing Boats
A Guide for users
by Nick Newland – www.swallowboats.com
Seagoing rowing boats of high performance are difficult to design because of the multiplicity of conflicting factors to do with ergonomics and seakeeping, combined with the low power available to drive the boat.
The type of use is also important. A boat designed for 2/3 hour excursions might be quite different in hull form compared to a flat out racer designed to race over a 45 minute course.
The diagram below illustrates some of the factors that go into making a good rowing boat. As can be seen, almost every factor conflicts with another. For example, speed dictates a narrow hull, whilst for stability a wider hull is required. Because of the limited power available the compromises inherent in a good seaboat design are very finely drawn.
It is important therefore to understand the real requirement and then balance the conflicting factors. The computer aids help, but judgement and personal experience are also important in achieving a balanced design.
This note is intended as a non technical aide memoir to help people make the right decisions as to their boat. The note is split into 3 sections. The first discusses some of the contradictory factors in the diagram above, the second looks at hull structure, and the last attempts to summarise what leading particulars should look like. The issues are discussed using a coxed double sculled skiff as a reference, mostly because it represents a frequent requirement from the marketplace.
It is hard to overstate the importance of weight. It fundamentally effects speed, and contributes very much to the difficulty or otherwise of handling the hull ashore. All the calculations that follow assume that each rower can develop a quarter horsepower (about a couple of light bulbs worth !). For short distances this is pessimistic, with powers of as much as half horsepower being developed, but after half an hour lung function begins to dominate and most mortals taper out at about quarter horsepower Typical weights for a 17 foot seagoing double scull with cox might be
Why worry about such a small proportion of the weight? For a 150 lb hull with two people rowing it and some gear, adding 30lbs to the hull might drop the speed by about 0.1 knots. This might not seem much, but in a race it is a lot.
Weight is the place that costs money in production and in design. Minimum weight is a principal reason why aeroplanes cost so much.
The later section on structure looks at typical arrangements in current boats and their effect on weight, durability and other factors. Suffice to say here that weight increases the resistance of a boat by deepening the draft increasing surface area and hence skin friction, and by increasing the size of the waves generated. The latter is more important in general as it is the waves that limit rowing boat speeds.
Length is an important issue. The length of a boat determines it’s maximum speed, so generally speaking the longer a boat the faster it is. However, that is not the whole story, and if you haven’t the power then maximum speed will not be reached. Taking the earlier skiff (including gear) as a starting point, and increasing the length might give the following speed changes:
The weight increases with length, as you can’t increase length without increasing weight in the real world. Length is giving you speed, by reducing the wave making drag at the expense of skin friction drag. The 17 footer has about equal skin friction and wave making, whilst for the 21 footer the wave making is half the skin friction (but the skin friction has gone up due to the length increase).
Whilst skin friction drag increases constantly with speed, wave making drag increases extremely rapidly with speed after a threshold determined by boat length. This threshold begins when the speed of the boat (in knots) equals the square root of the waterline length in feet (root L). For heavy hulls it becomes so punitive that it effectively limits speed to 1.4 x (root L). For lighter hulls (in proportion to their length) slightly higher speeds are possible. This is because the size of the waves produced depend in part on the weight of the boat.
There must be an optimum length for the double scull depending on how fit the crew are. For example looking at smoothwater competition double sculls will set an upper limit as to length of about 27 ft. Seagoing boats need more beam than such boats, thus increasing the wetted area so that the skin friction component will be greater. To reduce the skin friction the length needs to be reduced, but this will increase wavemaking resistance. Thus begin the compromises!
In the end length must also take cognisance of practical things like transport problems sea-keeping (longer hulls need more freeboard) and the Recreational Craft Directive.
2.3 Beam and Stability
The next most important factor after length and weight is beam. Beam has a crucial effect on drag, as any beam over the minimum will increase surface area, and hence drag. So for speed you want minimum beam.
For stability however you want beam and a low centre of gravity. Since two crew and a cox is probably the predominant weight, the height of the crew is very important in assessing stability, and as will be seen later this has an important effect on seakeeping since it directly affects freeboard.
The diagram below illustrates the effect on circumference and hence surface area of different boat beams. The minimum value is for a perfectly circular bottom as is adopted for racing sculls. It is unstable, and requires the oars and body weight to stabilise the boat.
We have built a 24″ beam boat (12″ ellipse in above diagram) and it feels unstable (tippy). The Teifi Skiff at 36″ waterline beam feels fine so somewhere in between 24″ and 36″ is probably the ideal waterline beam for a practical rowing boat.
Looking through boat design references in relation to rowing boats you will find that recreational sculls seem to have waterline beams of about 23″ to 28″, whilst more practical rowing boats have waterline beams of at least 2 ft 10 ins, and generally much more.The American Naval Architect Phil Bolger has designed many rowing boats and we can’t find a design of his below 3 ft waterline beam.
The next diagram illustrates the effect of different cross section shapes on surface area. Again for a given beam the minimum is an arc of circle, but the other shapes are remarkably close, and even the hard chine is only 3% worse than the best.
It is easy to be simplistic about beam and section shapes. Stability is determined by beam and shape, and shapes with a flatter bottom are much more stable than those with a high rise of floor. The pure vee bottom in the diagram for example would probably be unstable.
For a given weight and length of boat, an increase in beam of 20% might increase the surface area (drag) by 13% and the stability by more than 60%. The stability numbers can therefore change radically with quite small differences in beam.
This compromise between beam and stability is at the heart of good rowing boat design. To get it right is difficult. Much depends on the skill of the crew, and the kind of conditions they habitually row in. Experience and a good set of references to extant designs is essential to reaching this compromise, since no one has managed to create a computer model for rowing skill as yet !
2.4 Ergnomics – Layout of the works
The layout of the seats, gates (oarlocks), and footrests, is extremely important. Experience is a hard won prize, and most people with experience will have their own views on what parameters constitute a good layout. A minimum list would be
Height from heel to seat
Height from gate to seat
Distance from gate to rear edge of seat
Distance from rear edge of seat to foot stretcher
Angle of foot stretcher to the vertical
Distance between gates fore and aft
In relation to seating positions different areas have their own different traditions. Here in West Wales we use central seating, in Cornish gigs they use offset seating to get a longer oar for the same beam. In Australian surf boats they use offset seats with a semi-sliding seat. All the arrangements have advantages and disadvantages.
Stroke rate has a major influence on the layout as it determines for a given length of oar what sort of speed can be achieved. The length of oar in turn determines the beam required at the oarlocks or gates. Or to put it another way the beam determines the length of oar which determines the speed via stoke rate.
The diagram below shows a representation of the oar stroke. The oars are rotating through 90 degrees in time t, with an oar length of r1 outboard of the gates.
The distance swept approx = 90 x 2 pi x r1 /360 = pi x r1 /2
Speed approximately = Distance swept/ time = pi r1/(2 x t)
All oars have about 70% of their length outboard give or take a few percent. Thus knowing the beam of the boat (and allowing a little for overlap) the oarlength can be calculated and thus the outboard length r1 above.
Rowers aim to achieve a cycle of power stroke to recovery of about one third power to two thirds recovery. Hence given the rate R (usually in strokes per minute) the time for the power stroke t can calculated. Rates vary from 20 to 40 plus depending how close you approach Olympic standards !!
Substituting and simplifying a bit then:
Speed approximately equals 0.1 x R x B
Taking the 17 ft boat and fitting in the values of 4 ft for the beam and say 30 for the rate gives a speed of about 12 ft/sec or 7 knots with oars of about 7.5 ft. This matches the theoretical hull speed for the 17ft double scull quite well at about 6.5 knots. The point to note is that narrow hulls, whilst capable of being driven fast, cannot achieve the speed because the oars will be too short, necessitating unrealistic stroke rates.
Some boats do overcome this problem by rowing with hands crossed to the elbow. The most notable example of this layout is probably the American guide boat used in the Aidirondak lakes, but all accounts suggest that this is an ergonomic nightmare that is an acquired art.
The other obvious solution is outriggers. Outriggers are a good solution if you have the need for a narrow hull. The 24 inch beam scull mentioned earlier was fitted with stubby outriggers and it works fine. Rowing it at sea required some skill with the oars as stabilisers and it would probably have benefited from more waterline beam, which would have very likely removed the need for outriggers So in real seagoing boats outriggers may not be of much advantage, the speed being limited by the length of the boat, the beam being dictated by stability, and the beam being adequate to support oars of adequate length. Certainly outriggers are a pain when coming alongside, putting it on the roof rack etc.
For boats used at sea this is an important aspect. It is dominated by stability and freeboard (height of the sides of the boat above the water) considerations. Stability is required to keep the boat upright and freeboard to stop the waves coming inboard. Seagoing boats have to have more stability than smooth water boats. In rough water too little stability gives problems in handling the boat, the uncertain motion giving rowers difficulty in positioning their oars.
On the other hand too much stability is also bad news in that it leads to a jerky motion, throwing the crew around and creating almost as much of a problem as too little stability. Too much flare above the waterline can have similar results as too much stability. Stability is largely determined by waterline beam, and the position of the centre of gravity and needs good judgement if a reasonable performance is to be achieved since beam adds surface area and hence drag.
In small rowing boats with light hulls the centre of gravity is largely determined by the position of the crew. Ergonomics put a minimum height to the crew centre of gravity and this in turn determines the position of the oarlocks or gates and hence the freeboard. Seakeeping requirements might then increase the freeboard depending on use. Additional freeboard drives the rowlock/gates up and hence drives the crew centre of gravity higher and requires more beam for the same stability penalising speed.
Freeboard also has a major effect on wind drag and particularly ease of steering in cross winds. For speed in windy conditions, a very low freeboard is required. The place where freeboard is most useful is near the bow and stern, so low freeboard at the rowing position can to some extent be offset by a strong sheer and good arrangements to lift the bow and stern quickly to meet oncoming or following waves. Such arrangements might include a strong rake to the bow/stern in profile, and additional flare at the ends.
Crew skill also comes into the equation since very skilled crews can handle an otherwise difficult hull using the oars and their body weight to overcome the hull’s limitations. Heavier hulls also help in giving a more predictable motion and a lower centre of gravity, but always at the expense of speed in smoother waters.
Whatever is done to improve seakeeping therefore is almost invariably in conflict with the parameters to give speed. Experience and a good database of designs is an essential requirement to get the compromises right.
After the hull shape, the important features that are necessary to a good rowing boat are:
Light weight – because weight is a major component of resistance
Stiffness at the pin- because this wastes effort in flexing the gunwale
Stiffness at the foot stretchers – because this flexes the hull locally increasing resistance
Stiffness in torsion of the whole hull – open boats are particularly vulnerable to twist
A recent examination of materials by the American naval architect Dave Gerr (The Nature of Boats ) gives a ranking for material efficiency as follows (1.00 is best)
It can be seen that wood is an excellent choice for structures like gunwales where bending and column strength predominate. If it is so good why are racing boats built with the other higher tech materilas such as graphite and Kevlar ? The reason is that these materials use a double skin structure or deep stiffeners to effectively increase the thickness. The minimum weight hull is probably a double skin hand laid up glass/carbon structure with a high tech core and utilising systems like resin transfer moulding or vaccuum bagging to minimise resin.
It’s expensive. It’s vulnerable to damage due to core/skin separation and minute leaks into the core through grounding on hard jettys etc. Any such damage is difficult to repair.
The next lowest weight is probably cold moulded wood. For a one off it is competitive on costs, but not for a production run. The most competitive system in wood is probably plywood in either single or multi chine construction using epoxy resins and taped seams or glued lapstrake construction. Modern plywood protected by modern paints is an excellent material with low maintenance costs, and high durability.
Whatever the merits of other materials glass Reinforced Plastic (GRP) has become the industry standard in terms of cost. Options possible are double hull GRP, single skin GRP and some form of composite construction.GRP is not inherently light, it’s strength/weight ratio being only marginally better than steel.
The most common structures are the conventional inner and outer moulding (double hull) that turn up in boats all over the country. The double hull is there to provide the consumer with a clean modern looking hull interior with in-built buoyancy. It is hopeless when it comes to weight, since it essentially adds another skin for little structural benefit. This is because the core between the inner and outer moulding is not strong in shear or compression and so cannot provide much in the way of support for the outer skin without difficult and expensive bonding operations between skins.
A rough comparison can be made by dividing the boat weight by the (length x beam) to give a factor for weight per unit area Some typical values for open boats of double hull construction give a weight sq ft of plan area of about 4.5 to 5.
By omitting the interior skin, a single skin GRP structure with GRP stiffening reduces the weight per unit area to about 2.5.
By adopting a more sophisticated approach to hull design and material selection, it is possible to get robust GRP hulls down to a weight per unit area of less than 1.8 using wood, aluminium, and plastics in a composite shell, selecting each material to be best for it’s application.
The other point to bear in mind is that modern materials do allow designers to move away from some features of traditional hulls that were there because of the material rather than because it was needed from an a hydrodynamic reason. Long vestigial keels, and deadwood aft were needed to attach planking to. In terms of hydrodynamics they add drag and though they achieve something (directional stability for one) there might now be better ways of achieving the objectives. Keep an open mind.
Torsional rigidity is particularly problematic in open boats. The ideal approach is to deck it all in and make it a closed boat! Lacking this way forward, stiff gunwales, well connected Port to Starboard through the seat structure and with adequate knees at the bow and stern have been the commonest solution. Real care is needed in designing such boats to address this issue.
Details such as the strength and stiffness of the seat knees are important. The feature that has come to dominate material selection for the leisure market is ease of maintenance. GRP is perceived to be lower maintenance than other materials, but modern paints and glues have changed the position of wood though not yet the public’s perception. If timber is used, correct choice of timber (only durable species) and its treatment is important.
Where does all this lead? Speed can be determined easily enough by calculation, the oar length for that speed can be estimated and hence a shot at the beam overall. The needs of stability and seakeeping will depend on users appetite for excitement and the areas they expect to use the boat in. Outriggers might be an option if extreme performance is required.
Looking at the coxed double sculled example. The power of two rowers in a long race will limit the speed to around 7 knots even for the longest practicable boat, and this speed can be attained with 7.5 to 8 ft oars with a reasonable stroke rate. Such oars can be handled in a 4 ft beam boat without recourse to outriggers. So the length might be 18 to 24 ft and the beam about 4 ft.
The waterline beam is more problematical. Three feet seems on the higher side of our experience, and maybe some reduction might be possible here. Not much since remember that the 20% change in beam gave a 60% change in stability
If the longer length’s are aimed at then more freeboard will be necessary to allow for the longer length and practical considerations of storage, transport and handling need to be carefully considered before the final decision is made.Not surprisingly if designs in the 15 to 20 ft region are considered there are many boats not far from these numbers. In particular a beam of 4 ft seems a very common feature.
This points to a truism of sorts, in that the boats that have evolved over time will generally show the way to what works. Modern structure and a better understanding of hydrodynamics will refine the solution but the evolutionary ghosts will probably shine through in many places.