2     Modelling a Thames Sailing Barge.

 

Contents

 

Scaling problems

 

  The underwater parts

 

  The fin and lead

 

  The rudder

 

  The topmast

 

  Summing up

 

Controlling the model

 

  Radio equipment

 

  Connecting the sheeting system to the rig

 

  Control systems

 

  Choice of system

 

Sheeting systems

 

  The shape of sails

 

  Sheeting arrangements

 

  The main-sail

 

  The foresail.

 

  The staysail and jibs

 

                                            Scroll to search

 

Scaling problems

 

The underwater parts

There are two basic problems that arise from the scaling if we are to make a working model of a Thames sailing barge. The first is that if the model is just made to scale and filled with ballast it is not stable enough to carry its sailing rig in winds that are encountered daily. The second is that the sailing rig works in just the same wind as the full-sized barge but the speed of the model is about ¼ of the speed of the real barge. This mismatch leads to the fact that a scale model of a barge loaded internally to its marks makes excessive leeway and the scale rudder cannot control the course of the barge. We must know why these things happen and what we can do about it.

 

The fin and lead

In practice it is found that the ability to carry sail depends on the stability and this means on the total weight of the barge and cargo multiplied by the metacentric height. In Figure 37 I have shown a cross section of a barge somewhere amidships. On it I have added three all-important points. G is the centre of gravity and most people will know that the weight of the barge acts here. B is the centre of buoyancy and by the very name it is where the buoyant force of the water acts on the barge. M is the metacentre which is a real point that can be found by calculation but which has no special value to us here[1]. What we need is the distance between M and G so that we can see what happens to it when we build a barge to scale. This distance is called the metacentric height.

 

It happens that the metacentric height does not change very much when a full sized barge is loaded, even though all the points move, so the ability of the barge to carry sail increases as the weight of the barge increases.

 

Here we are concerned with a scale model. We would expect that the centre of buoyancy of the model to be in the same relative position as that of the full sized barge when both are loaded to the same marks. In order that the model shall float at a correct looking depth it must carry ballast. Then the value of BM for the model also changes in proportion to the scale. The position of the centre of gravity of the model will then depend on the location of the ballast. Suppose that we arranged for the centre of gravity of the model to be proportionately in more or less the scale position. Then the metacentric height will be to scale. This means that the ability of the model to carry sail will depend on the weight of the model. The ratio of the displaced volume of the model to that of the full sized goes down with the cube of the scale and so the scale weight of the model will be about 1/14000 of the full size. Unfortunately the transverse force on the sailing rig goes down with the square of the scale ie by 1/600 and the model with its centre of gravity in the scale position will not be sufficiently stable to carry the sailing rig by a factor of 1/24. If one is making a scale model it is unlikely that one would accept a vertical distortion of the hull to increase its displaced volume and so increase its weight when floating at its marks. This leaves only one way to increase the stability and that is to increase the metacentric height by lowering the centre of gravity. This means that the ballast must be lowered. Now we need to know whether ballast carried internally will serve our purpose in creating a practical miniature sailing boat. The answer is that we cannot increase the metacentric height sufficiently using internal ballast. The ballast must be carried under the hull and at a significant depth. We are fortunate that this proves to be a practical solution.[2]

 

Now the Thames sailing barge was typified by the very small angle through which it heeled even when it was being pushed to the limit during racing. If we are to create the illusion of the full-size in our model it too must not heel excessively. We have buoyancy provided by the submerged volume of the hull and that is fixed. That buoyancy must support the hull and its contents, the sailing rig and now the underwater weight. The lower that weight goes and the heavier the weight is, the greater the stability of the barge will be and the smaller will be the heel. For practical reasons like avoiding weed and sailing in shallow water we usually limit the draught to about 10 inches. Once that is settled the only way to increase stability is to increase the weight to be suspended under the hull. Given the limit on the buoyant force provided by the hull we must aim to minimise the weight of the hull and the rig and so maximise the under-slung weight. Far too many modellers ignore this rule and make a heavy hull, use a light bulb, and pay the price in every little squall.

 

There is another incentive to maximise the stability. The rigs on most model barges are not designed for sails to be removed easily. The barges have to be sailed as rigged whatever the wind. In heavy weather one must decide whether the model can cope with the conditions fully rigged because, if the barge has too little stability, the heel can become so great that water starts to come over the rails at a rate that beats the freeing ports. Sailing with the deck awash is not the best way to get about.

 

The problem of the leeway and the rudder is more complicated than just the rudder. The full size barge and the model barge sail in the same wind. The full size may make 6 mph in some wind in which a model could also sail but when the model sails in the same wind on the calm surface of a boating pool it does less than 2 mph. We know that, for 1/24 scale, the forces on the model sails will be about 1/600 of the force on the full size in the same wind. But the underwater parts that have scale areas are quite unable to produce matching forces to limit leeway and give steering because the model is moving at less than one third of the speed of the full sized barge. In fact the forces are too small by a ratio of 2²/6² =1/9. There is a serious mismatch here. A model of a barge made to scale has insufficient underwater area to permit the model to have an acceptable leeway and for it to be steered. If the model barge is to be propelled by its sails we must do something about it. We must provide some additional area under the water by way of a keel and increase the area of the rudder.

 

The other limiting factor is the loss of control. Most barge rigs, and racing rigs as well, are set up to be balanced. This means that the net effect of the sails when beating is to just point up into the wind. The course is then controlled by a small rudder offset. At higher wind speeds this balance may alter and require greater rudder offset and then, when a tack is necessary, the further rudder movement required may lead to a stalled rudder especially with the crudely shaped rudders that are often used on model barges.

 

So what precisely are we to do? There is an aesthetic problem to solve first. The problem is, what can be added underwater to a Thames sailing barge without jarring too much with the whole concept of the barge? That concept of the barge is rooted in cut and try and no-one had to design and make a fin and keel weight by cut and try methods because any fixed underwater parts were out of the question and, as it happened, the provision of a leeboard of a practical size was more or less adequate.[3] So one might argue that, as the barge is crudely made, the addition should be crudely made. I do not share this view. The fully developed leeboard had a very acceptable aerofoil shape so, at least in shape, it was not crude. It was made of shaped wooden planks held together at the top by plates and rivets and bound with protective iron at the lower edge. In between it was smooth. I think that the only change that we might make would be to make them thicker. The construction of the leeboard was right for its job[4]. However we are contemplating a device that must function efficiently at model size and speed and this means that we should make our addition as accurately and as smooth as we possibly can. The second is that if the object of the model is to look like a full size on the water and to sail like the full size we must use the best thing available to us provided that its presence is not so apparent as to destroy the illusion.

 

Clearly the problems of the keel and the weight are interconnected because if we have to hang a weight under the model and also provide an area under water to produce a sideways force a single addition comprising a fin and a weight can be a solution.

 

Before we go on to design our underwater parts there are other factors to be considered. It is desirable for the barge to be capable of going through head to wind and not be forced to bear away to turn. A look at the Figure 33 shows how much distance is lost in bearing away when compared with going through head to wind and how much space is needed. If the barge is being sailed, as distinct from being “paddled”, it relies on momentum to go through head to wind and considerable momentum is required at the start of the tack because all the cordage etc produces a wind drag that is quite high and slows the barge during the tack. If the barge is not heavy enough it starts off with too little momentum to go round and falls back on to the same tack.[5] If the barge is fitted with a large unbalanced rudder (see later) this problem can be avoided by paddling round a tack. But the fin will still have to generate a windward force.

 

Figure 34 shows what experience in engineering in this field tells me is required for my boom sail barge. It is a fin having a symmetrical section such as is used for an aeroplane wing. This provides the area. It has a weight that has the same shape as the tip tanks that are sometimes fitted to aeroplanes. It weighs 9 pounds. Let me say why I think that it is right.

The dominant requirement is that we need these parts to offer the minimum resistance to motion through the water. The lead has its least resistance when it is circular in section and of aerofoil shape when seen from the side.

 

There are other shapes offering low drag but we have to recognise that some of them like the delta lead in Figure 35 and the very long thin lead have been designed for use on racing yachts where there is a draught restriction which, if exceeded, leads to disqualification or severe penalty. This leads to the design of leads that are thin top to bottom and either long or wide to get the required weight with a low centre of gravity of the lead and so of the whole yacht. (This shape does not produce any hydrodynamic stabilising force. It is just a weight.) We have no such draught restriction except the practical one of about 10 inches. So we can use a weight with a shape that offers the lowest drag. Of course the modern shape of the delta bulb has an attraction simply because of its appearance. Fortunately the increase in drag of the delta over the tip tank shape is not big enough to be easily discernable and it can be driven through weed.

 

The fin is just an underwater aerofoil. If made to a high standard it will have a very low drag and, when the barge is skewed to its course the fin will produce the essential sideways force. This fin is a small hydrofoil that is lifting sideways. On submarines, hydrofoils control the depth by lifting against negative buoyancy. They do just what we want our fin to do. They use a section that is just like the NACA 0012-64 shown in Figure 36. For barges it is best to scale down its thickness to 7.5% of the chord length and then it will perform as well as seems to be possible[6]. Its area should be about 55 square inches.

 

The importance of this shape, as compared to a flat plate for instance, is that it can maintain its transverse lift to as high an angle of attack as seems to be possible for a symmetrical section. This is important because it lets the rudder work for smaller turning circles without seriously slowing the boat.

 

The engineering solution to the mix of problems of the spritsail barge is to make the hull and its equipment as light as possible and fit a deep fin of the correct area with a streamlined weight. This arrangement lowers the centre of gravity and so increases the stability. If you plan to sail round the turns, the weight to be attached to the keel should be only just adequate to go through head to wind in light winds to keep the overall weight down. It is worth noting that a reduction of 1 pound in the weight increases the freeboard by only about 1/16² so we need a reduction of 8 pounds to give a useful 1/2². My best guess is that the barge should have two keels, a short keel with a light lead for light conditions and a deep keel with a heavy lead for heavy weather and it should carry no excess weight in either the hull or the rig. If there is to be only one keel fit the heavy one.

 

The rudder

The rudder of a Thames sailing barge is small because the main control of the steering is by adjustment of the rig. When the rudder of a barge is scaled down for a model it is just too small to be effective for the reasons I have given above. We have seen that the mizzen could be used to help the rudder on the full size but the automatic system cannot be recreated to scale and still work because the chain will be too light. It does not seem to be too difficult to make the little mizzen operational by radio. However it does add another channel to be controlled whilst the barge is sailing. Some modellers choose to have the mizzen rigged for appearance and let it be slack so that it cannot interfere with rudder action. Then the oversize rudder replaces the mizzen. Others use a taut sail and attach is to the rudder blade so that the mizzen turns with the rudder to help in the tack but not to help drive the barge.

 

So the area of the rudder has to be increased substantially in some way. However the fact that some of the rudder is visible above the water line and is part of the image of the barge means that the additional area must either be under the water or unobtrusive.

 

Now we must consider “paddling”. Paddling is the use of the rudder to change course by repeatedly swishing the rudder to one side and returning it slowly. A barge or a yacht can be turned through 360° in a series of jerks quite easily without any forward motion. It is not sailing but being paddled by use of the power of the radio battery. This is possible in a model but seems to me to be unlikely in the full size where only muscle power is available. For models paddling is most effective with big flat rudder extensions fitted aft of the scale rudder. Paddling can achieve the same result as the mizzen when tacking.[7]

 

The common method of adding area to the rudder is to add a piece of Perspex as shown in Figure 42. This is not easily seen when submerged and is effective. Some add a downward extension to the rudder so that the extension is not visible. Usually that extension is flat. In both cases the rudder travel is often more than 45° each way and possibly as much as 70°, which makes fine control for sailing difficult, and the drag produced by the rudder can be seen to slow the barge. These all suffer from stalling at very low angles of attack.

 

Figure 43 shows the option that I use on my sprit-sailed barge Jasper. It is an aerofoil shaped extension that moves only 30°. It works well but paddling with it is impossible. This puts Jasper at a disadvantage in light winds and in swirling winds but it can be sailed properly in a steady wind. One must choose between having a rudder that will permit paddling in light winds and in swirling winds and little control in mid position for sailing or having a balanced rudder that cannot be paddled but gives very good control for sailing in steady winds but is poor in light or swirling winds.

 

 

The topmast

We have seen that what I see as a weakness in the design of the full-sized rig is the fact that the topmast is not supported to aft. As, in the model, it is still necessary to have tension in the top forestay we have to think of a way of providing it. We might start by seeing whether the methods used on the full sized barge can be used.

 

We have seen that in the full sized barge we can put tension on the top forestay using the topsail and the weight of the sprit but the top-sail is not always set.  As the top-sail will always be set on the model we might think using it to put tension on the top forestay but it is doubtful whether, mechanically, it is really an option. We must look elsewhere.

 

The model sails work in the same wind as the full size although the model would not work up to the same wind speeds as the full size. In any given wind the force on a sail is proportional to the area of the sail. The ratio of the area of a model sail to the area of its equivalent full size sail is equal to the square of the scale ie to 1/24² = 1/576 or about 1/600. It follows that the tension in the top forestay will need to be about 1/600 of the tension in the full size.

 

We also know that the sprit can weigh as much as 1000 lb. The weight of the sprit varies with the cube of the scale ratio ie with 1/13,824 or about 1/14,000. 1,000 lb scaled down by 14,000 is about  ounces which is just what model sprits weigh. The gravitational force on the sprit is out of scale with the sail forces.

 

If we are to attempt to get some tension in the top forestay of our model barge by supporting the topmast to aft using the topsail we have only a sprit that is far too light, a mainsail that will also be light and whatever force there may be in the vang. It is all too light.

 

We have to think of a way round this difficulty. One might argue for a weighted sprit but, as any weight aloft will reduce the ability of the model to carry sail, we must think how to add weight so that it causes least trouble. If we simply made a sprit of lead the weight at the lower end would have little purchase on the topmast so we really want the weight all to be at the top. Having the top half of lead and the lower half of wood comes to mind but, this is not practical because, even if we could maintain stability with the required weight aloft, we cannot make a scale sized sprit using lead to weight it. We would need something of greater density than lead.

 

If it is made to scale a wooden topmast is very slender at model size and would bend excessively. Some make a topmast with a much-oversized diameter. Some modellers fit the two running backstays to stabilise the main-mast and the other two running backstays to stabilise the top-mast but of course in the full size only the windward running back-stays would be rigged. The running back-stays become fixed back stays. They all tend to slacken.

 

I make the top-mast of 6 mm carbon fibre tubing and cover it with timber so that it looks like wood. When it is fitted securely to the mainmast it is stiff enough to take the tension in the top forestay without the running backstays. It can be bent forwards to look like the full size by putting tension on the top forestay and so give a good tight luff for the jib sail but it does put a load on the vulnerable bob stay.

 

Summing up

The additions required in the scale model are a fin and lead for certain. I do not think that the area of the fin is critical either in its role of providing a resistance to the transverse force produced by the sailing rig or in terms of its drag. It should have an area of about 55 square inches. The area of the addition to the rudder should be about 25% of the area of the fin. The hull and rig should be as light as possible and the available buoyancy utilised to carry a heavy lead as deep as is seen to be practical. The goal must be to carry as much weight in the lead as is needed to make the barge stiff and so help the barge through head to wind provided that this is consistent with a barge that does not ship too much water over the bow gunwales. Some enthusiastic modellers may think that two sets of fin and weight are probably desirable.

 

Controlling the model

 

Radio equipment

Before a model barge is built some decisions about the control of the barge by radio must be made. Whilst it is obvious that we must have rudder control, decisions have to be made on the extent to which the sails are to be controlled. The possibilities range from no control at all to having all the sails operated accurately by radio.

 

I have already argued for the use of rudder control only for sailing on a sheltered pond. There is an advantage to be gained from the fact that all the sails of the rig can be set properly whereas it is difficult to use radio equipment to set the sails equally well for all points of sailing. All the sails of the sprit sailed barge with the exception of the main topsail are loose footed. Loose-footed sails really need two sheets to set independently the camber and the angle to the wind and this is particularly troublesome to achieve using one sheet operated by simple radio control when the sails must also switch properly and automatically from tack to tack. The only tricky one to pre-set is the staysail because there is no “natural” place to fix its sheets. If the sails can be set up so that the barge sails well when following a course just off the best setting for beating the barge can be sailed very well in a steady wind on most other points of sailing.

 

However a barge with only rudder control misses half the challenge of the model and most modellers will eventually want to look into sail control. For the staysail barge there are three sails to control but they require at least four sheets[8]. They are the mainsail and main topsail as one, the foresail and the staysail. As the mainsail really requires two controls, one for the vangs and one for the clew as it moves across the horse with the traveller this really means four sheets and the rudder.

 

Let me look at the possible control equipment that might be used for this task.

 

The basic sheeting systems

We really have only two actuators to use to operate the sails, winches or lever arm servos. Let me look at winches first.

 

The winch is really a servo that has been adapted by fitting it with a drum and making internal changes so that it can be rotated through up to five turns in response to the full travel of the transmitter stick. Usually the actual number of turns corresponding to full throw can be adjusted at the winch. Some system is needed to convert the rotary motion of the winch into linear motion to operate the sails. The most simple system for maintaining a tension is shown in Figure 44. The winch cord is secured to only one sheave of the drum and is attached at its free end to a ring. A length of elastic then joins this ring to an anchor point. When the winch cord is wound in, the elastic stretches to maintain the tension and, provided that the winch cord is aligned with the drum, it will wind on reliably. One or more sheeting cords can then be attached to the ring.

 

The more common system involves winding a cord on to and off a double drum. I shall call this cord the winch cord. It is essential then that the winch cord should run reliably on to the drum (it will always wind off if it is not tangled). There are two requirements for this. The first is that the drum should have substantial cheeks to act as guides and the second is that the winch cord should always be under tension. I have drawn the common arrangement of a drum that permits one cord to wind off one drum whilst the other cord is wound on.

 

The next most simple system is shown in Figure 45. The single winch cord is fed through the middle cheek of the drum and about three turns wound in opposite directions on each of the two halves of the drum. One free end goes round the pulley and then the ends are joined by a spring. When the drum rotates the winch cord winds off one half of the drum and on to the other half and the joint moves along. The spring serves two purposes; to keep the cord under tension so that it feeds easily on to the drum and to accommodate changes in length resulting from overlapping of the cord on the drum.

 

This system is not always the most convenient because it puts a constraint on the location of the winch. This can be avoided by using a second pulley as shown in Figure 46. The picture is for a deck-mounted winch on a racing boat. It is clear that this arrangement permits the winch to be placed in a convenient position to suit the intended arrangement of the radio gear.

 

All three of these systems convert the rotary motion of the winch to a linear motion. They all have to be connected in some way to the sheeting

Figure 47 shows three lever arm servos that operate a foresail and two jibs as installed in a barge. The levers are sheet aluminium alloy bolted to the ordinary output arm. The servos have metal gears and are ball-raced.

 

Lever arms are handy things because they can have two or more holes to operate two or more sails. The travels can be adjusted by the location of the holes.

 

 

 

 

Connecting the sheeting system to the rig

In the first system shown above the sheeting cord could be attached to the ring and pulled from right to left to sheet in the rig. For the closed-loop systems shown diagrammatically above the sheeting cord could be attached to the winch cord on the right hand end of the spring and be used to pull from left to right. They, and the lever arm servo, are mechanically sound actuating mechanisms.

 

Both the rotary winch and the lever arm servo run freely but this is not always true for the sheeting systems that are attached to them. So, before we consider how to connect the sheeting system to the rig we must have a few guidelines. A rig is seldom a problem in a steady wind. Problems arise when the wind is light. Then the rig must move very freely if it is to set properly under the very small forces. Further to this the sheeting cord or cords must be very flexible and they must move freely in their fairleads and pulleys.

 

We know that somehow four sheets have to be adjusted by radio. All that happens to the sheets is that they are either pulled in or let go. The letting out and taking in is done by either a winch, or a winch and servos, or just by servos depending on the size of the sails. The problem that must be faced is that it is unlikely that the travel required by each sheet will be the same because the travels depend on the requirements of the sails. They are likely all to be different. We should examine the possible ways of providing the four different sheeting travels.

 

Control systems

The first, one servo for the rudder and one winch for the sails, is the most simple from the point of view of radio equipment. The winch will produce one travel (that may be adjustable) and somehow this has to be reduced or increased as required for each sail that is to be controlled. There is scope here for ingenious systems of levers and pulleys and I am sure that some modellers get it to work.

 

The problem of matching sheeting travel to the sails can be solved by driving every sail from a separate winch or servo. Then, instead of joining all the sheets to one cord, all the servos or winches operate from one channel. This channel could have a four-way “Y” lead and all the sails would then be controlled by one stick. The servos can be fitted with extension arms of suitable lengths. Unfortunately servo reverse is not then an option nor is the use of travel adjustment (ATV). It is more costly in equipment than the first system but it is much more likely to work easily for light wind conditions.

 

The best way is to use a computer-controlled transmitter that is capable of mixing four channels. It is no longer an expensive option and should be considered seriously. Setting up these transmitters is no more difficult than setting up a mobile telephone. Each winch or servo has a channel. Each channel is adjustable for direction, travel and trim. Provided that the servos are giving more or less the right travel they can be adjusted accurately through the transmitter and not by fiddling in the barge. The four sheeting channels can be operated by one transmitter stick by “mixing” them. Then the barge is controlled by two sticks even though five channels are in use. A further facility that comes with this system is that sail trimming can be done by radio when sailing.

 

Now we have to try to weigh up these options.

 

Choice of system

This is especially troublesome for skippers intending to race at barge meetings because they go to many venues. A factor that I think weighs heavily in this choice is the venue where the barge will generally be sailed. Many lakes are surrounded by trees and buildings and are far from ideal for sailing vessels. The wind blows over and round the obstructions producing all manner of swirls and eddies of widely different sizes. Some of these eddies are small when compared with model boats, some are quite large. It is nothing to see a racing yacht change tack twice in a few seconds without a change in course in response to swirl. On such a lake sophisticated control systems are no better than simple systems because the element of chance is so high. On an open lake the model with fully-controlled sails will perform better than one under simple control and, of course, looks more realistic

 

Some choose to fit an oversize rig to offset this difficulty and it appears to be a practical option although it runs counter to the basic idea of racing scale models of barges.

 

No doubt cost will also be a factor and many modellers regard computer-controlled equipment as simply too complicated to consider[9] even where price is no object. This narrows the choice to some combination of the first two systems that I have outlined above. Most opt for one servo and one winch and use them to control selected sails and not the whole rig.

 

Sheeting systems

The shape of sails

Once a decision is made we need to have some idea of the actual systems of sail control that can be employed on a barge. This involves having a clear idea of the shape of sails and how this shape can be controlled.

 

I do not think that it is possible to overemphasise the importance of being able to recognise the best shape for sails. Forget all about the shapes shown in cartoon films and look at real yachts doing it properly. Whenever you get the opportunity to look at photographs of top class yachts under way study the shape of the sails carefully. A great deal of money will have been spent to get those shapes. I have included a picture of modern yachts sailing in a steady wind with shading and shadows that give us an idea of their shape. As this picture was published in 2001 one must suppose that the shape is pretty well as good as it ever will be for sails that can be furled that is un-battened sails. For a triangular sail such as the mainsail on these racing yachts the sail is hollow (concave) within the three edges (the luff, the leech and the foot) all of which are taut. That hollow is the result of very careful sail making. The sail is twisted slightly but the object seems to be to have no twist beyond that which is unavoidable. It is clear that the sails form a smooth curve from luff to leech. It is hard to know what the best shape for this curve will be but there is no doubt that sails do not work well with too much belly. For barges we need to create this shape using two sails, the main and the top-sail, joined together and shaped by the location of the clew, the sprit and the outhaul of the top sail. It is not easy.

 

This curve of the cross section of a sail is shown in Figure 53. The curve is not intended to be an arc of a circle but more like the upper surface of a thick aerofoil. The actual shape depends first on the cut of the sail and then on the precise distance between the luff and the leech. As this distance is reduced the hollow of the sail increases and the curvature seems to be much more important than the actual profile. We need a way of measuring this hollow. The common way is first to imagine a line joining the luff and the leech and call this the chord and then to measure the maximum offset of the sail profile from this chord. The ratio of the offset to the length of the chord is expressed as a percentage and called the camber. The camber in the diagram is 10%. For good sail performance the camber should be not more than about 10% for light weather and a bit less for heavier weather.

 

The foresails of the yachts in the figure 52 are genoas and we can see that they are set at a greater angle to the centreline of the hull than the main. We cannot see all of the foresails but it looks as though the leech is curved (as one might expect) and we can see that the camber varies with the largest camber at about half height. These modern shapes are only possible because of monofilament materials and the use of materials like carbon fibre and Kevlar. Furthermore a modern rig requires substantial crew numbers to set and operate it. Nevertheless this is what the makers of sails for barges and the skippers were trying to achieve using natural fibres and we have the same goal.

 

The sails used on sailing barges were made of cloth. The red colour of the sails was the result of applying a stodgy preservative to the sail that also filled the inevitable holes in the weave. This preservative gave the sailcloth no increased resistance to diagonal forces and it was the behaviour of sailcloth that determined the final shapes of the sails of a barge. We must also recognise that the sailing rig of a barge evolved to be operated by only two men and that the rig has to satisfy other practical requirements like being up out of the way when the barge is being loaded and unloaded. These requirements led to the use of loose-footed sails instead of booms.

 

When we look at photographs of barges under way such as Figure 54 we can see what is possible when the sails are made of sailcloth. Can one doubt that the sailing barge has evolved into an efficient sailing machine? The sails are beautifully set and it is clear that we must accept that the camber of the sail is no less important for the sails of a barge than it is for a modern racing yacht. The photographs of barges show us how far the sail-makers had progressed towards a modern sail despite the fact that the sails are all loose-footed.

 

No doubt the crew of real barges could add ropes and set them to get things right but in our models we are going to have to make compromises none more so than for the foresail and the staysail or the jib. It is all too easy to have much too much camber. We also have little control of the twist of the sails because large tensions are required to keep a sail reasonably flat and we cannot provide such tension.

 

Model makers have to choose a cloth to use for the sails and there is a wider choice. We could use a modern fabric such as Dacron as used on light aeroplanes. Such materials have a plastic coating on them that seals the weave and this gives some diagonal stiffness. This limits the deformation and it is almost impossible to make a model sail that will “belly” as the full sized sails do. Sails for a model barge made in such material simply do not look right and we are trying to create an illusion. Only sails woven from natural fibres[10] can be made to look right and although they are not sealed they seem to work all right on models although it makes sense to use a cloth with a tight weave.

 

One might think that there is some black art to cutting model sails. There may be to making full sized sails but when we come down to the size of model sails it is unlikely that the shape of the sails can be affected predictably by subtle shaping of the edges or seams. However I think that there are simple ways of making sails set better that anyone can use.

 

I think that for triangular sails like jibs and foresails the luffs should be straight but the other two edges should have a curve of say ¼ inch for short edges and 3/8 inch for long edges. This reduces the belly and makes the corners look right when the sail is filled. However it is the main that needs some attention if it is to look right.

 

The main was a heavy sail; it took six men to carry it. The head of the sail was not fitted with a bolt rope but had a tunnel through which a wire called the head-rope was passed to link the top of the sprit and the “muzzle” round the top of the mast. This head-rope supported the head of the sail. There was no provision for adjustment of its length. However the topping lift could be adjusted and if it were to be released the whole weight of the main sail and the sprit would be taken by the head-rope. So it was possible to share the weight of the sail and sprit between the topping lift and the head-rope. This means that, if the head of the main was cut on a curve, the sail could be lifted by increasing the tension in the head rope and this could shape the sail.

 

In my opinion there is a case for putting lead weighting like curtain weighting in the foot of the model main-sail. I think that there is also a case for stiffening both the foot of the main-sail and the foot of the foresail. I think that the main must be the first sail to be made and fitted because, until it is properly set up, the dimensions for the top sail are not known.

 

There is one aspect of sail making that makes a difference in the end and gets no real mention. I do not know where people get the measurements for the sails for their barges. For my boomie barge Pearl and for my bowsprit barge there are sail plans so I suppose that there are similar sail plans for other barges. I cut the sails for Pearl  to correspond to the sail plan and, whilst they work quite well, I now realise that they needed to be fitted very carefully to the rig so that adequate space is left for the ropes to actually pull on the sails. I did not leave quite enough space yet leaving too much space spoils the appearance. It is a fine balance that I think I got right for Jasper. It applies especially to the jibs where the sheeting system is not going to work properly if the clew of a jib is too close to the stay that the sheets must cross. Then I think that one must think about fitting a concealed boom.

 

Sheeting arrangements

By sheeting arrangements I mean the systems of ropes etcetera that will be used to set the sails and to adjust the position of the sails relative to the barge to suit different points of sailing. We have seen some of the sheeting arrangements on full sized barges and now we must see what can be done on a model for remote control.

 

We have one other dominant requirement. A barge like all other boats is very easily driven at very low speed and if the sails can be set properly they can make the barge go in almost no wind. The trouble is that the wind that can drive the barge must first set the sails. If the sheeting does not move freely sailing in very light winds becomes almost impossible. So this must be kept in mind when choosing and constructing the sheeting arrangements.

 

The simplest form of sail control will be to control only the mainsail through the vangs. The clews of the mainsail and the foresail can be allowed to run to the ends of their horses, and the staysail have some equivalent system fitted perhaps to the cross trees. If a decision is made to go use a computer-controlled Tx we can pick from the following control systems for the various sails

 

Let me deal with the sails in turn.

 

The main-sail

The first sail is really the main sail and the main topsail working together as one. They are joined at the top of the sprit and it seems to me that the most important point to control is the position of the top of the sprit. The first requirement is to set up the balance between the topping lift and the head-rope but, once that is done and the top-sail correctly set, the travel of the clew on the horse and the swing of the sprit must be controlled.

 

It is evident from inspection of existing barges that the clew of the main-sail was not allowed to run much beyond the rails. There is not sufficient rope to permit this given the multi-rope system that is fitted to the foot of the main-sail. The principal control is the vang but no doubt the clew is brought in for beating.

 

The clew is controlled by the vangs as in Figure 55. In both the model and the full size the sprit is allowed to swing to leeward by letting go the windward vang. The sprit will swing around the mast to leeward until the windward vang is tight and the leeward vang slack.

 

We have a choice when we set about the control of a model barge. The vangs may be operated by letting out both vangs and having only the windward one tight as in Figure 55. Then the two vangs must be attached to the head of the sprit as is done on the full size. The alternative as shown in Figure 56 is to attach a cord to a gunwale, run it through a pulley at the head of the sprit and then through a fairlead that is placed at deck level and near to the gunwale to a sheeting mechanism operated by a winch or a servo. Letting the cord out will increase the effective length of both vangs provided that the cord can be made to move freely through the pulley. This latter arrangement has the disadvantage that both vangs are under tension and this creates an unnecessary force on the muzzle and creates unnecessary friction at the pulley at the top of the sprit. It also requires a large travel at the winch.

 

Once a choice has been made the main problem for the modeller is to decide how much swing to have and from what sheeted-in position. We do not need to bring the sprit in to the centreline of the hull but perhaps to bring the top of the sprit to two or so inches from the centre line. The furthest out we can sheet is until the sprit touches the running back stay. This turns out to be quite a long way because of the angles involved. This is the position for running although some modellers do not go this far. By and large it is a simple and practical arrangement but whilst it settles the control of the topsail and the top point of the mainsail the position of the clew of the mainsail is by no means settled. This can be achieved with a lever arm servo.

 

The mainsail is loose footed and the distance from tack to clew is not fixed as it would be on a boom. The sail tends to fill out far beyond the best camber and, in order to control the camber, the clew must be pulled aft some way and held at some desired distance from the tack. The simple system of traveller and horse is not really adequate to do this. On page 12 I pointed out that on full sized barges the link between the traveller working on the horse and the clew of the mainsail is not just a simple rope. It is a multi rope pulley system. We cannot hope to use such a system on a model but we can model the multi rope system to give the right impression and give it a fixed length between the traveller and the clew and provide a simple system to control the position of the traveller. The horse is either curved to follow the curve of the deck and straight in plan view or it is curved in plan and straight in end elevation. If the horse is straight across the deck it will lead to a decrease in camber as the traveller moves out. If the horse were to be curved in an arc around the mast as centre the camber would not change. Some full-sized barges had curved horses but not with the centre at the mast so, in practice, some variation of the camber for different sail positions might have been found to be desirable or just inevitable. A simple system is shown in Figure 57.

The traveller is of wire and can move freely across the horse. The sheet is attached to the dummy block and is knotted to a loop in the traveller. The distance between the knot and the traveller must be adjusted so that the vangs can go out to their designated limit without straining the rig when the traveller is at it limits.

 

 

If the top sail and the main sail are to look like and behave like a single sail with extra twist the control of the vangs and of the traveller will have to be synchronised.  This requires two lever arm winches. These two servos were set up using a computer controlled TX. It was quite straightforward.

 

 

 

 

The foresail.

I do not think that the modeller has much choice in how this sail will operate. It was normal for the clew to be shackled to the horse by a loop of chain through the clew and round the horse. Then the sail will set at the end of the horse according to the tack. This seems to be contrary to the best way to operate this sail but, when the barge is running, the foresail does not contribute much to the drive because it is blanketed by the mainsail, and when the barge is beating and it might be better to set this sail more accurately, this simple arrangement is as much as one might expect with so few in the crew

 

The staysail and jibs

The usual arrangement is shown in Figure 61. Two sheets run up from the gunwales and over the foresail or the forestay. The sail is set for angle to the hull by letting out the windward sheet. The camber is set by changing the leeward sheet. Changing tack involves resetting both sheets for both sails.

 

This is not practical in a model and the simple and fairly effective system is shown in the next diagram. The sheeting is a continuous rope from a cleat on the gunwale up and over the forestay and running freely through the clew of the staysail. The sheet passes through a fairlead near to the other gunwale to the servo.

Sheeting in gives quite a good version of the full size and it can be adjusted at the cleat. The use of “bone” stiffening in the cross seam of the sail prevents excessive camber.

 

The mizzen sail and the top sail

I will deal with these in section 3 on modelling.