Logo Computer scientist,
engineer, and educator
• Articles • Articles about science and technology

How to build an inexpensive, study plywood boat — maybe

Should we be going round in circles?''
My kids have been reading Arthur Ransome's Swallows and Amazons lately. For those who don't know the story — which will be just about anybody under 50, I guess — it concerns a bunch of children aged seven to fifteen, who sail off to a small island and have lots of marvellous adventures. Between adventures what they mostly do is to eat Pemmican (like Spam, but with added fat), so far as I can tell. It's a gentle and good-humoured story for a gentle and good-humoured age, when nobody would have thought it odd to have a central character called `Titty'. Yes, really — no sniggering at the back, please.

I imagine that Ransome's idea of adequate parental supervision (Better drowned than duffers!'') would result in a prison sentence these days. Nevertheless, my kids' evident enjoyment of Swallows had the effect of bringing forward a plan which had been lurking in the cobwebs of my mental closet for some time: since we live almost at the water's edge, it would be nice to have a boat of some sort.

The problem is, I can't afford a boat. I can't even afford a small boat. In fact, I can't even afford a toy boat. I did look at a few second-hand ones within my meagre budget, but none of them looked particularly seaworthy. So it eventually occured to me that if I can build a computer, I ought really to be able to build a small boat. How hard can it be?

I approached my boat-building project with the same meticulous attention to detail and planning that characterizes all my construction activities: I spent ten minutes looking at Web sites, then laid into the timber stack with a power saw. At the start of the project, I new nothing about boats, other than that they float (hopefully) and tend to be pointy at the front. Now, I know even less — but at least I know how little I know, which has to be a step in the right direction. And I've learned a whole new vocabulary — just about every part of a boat has a name you couldn't easily guess. Who would have thought that a painter was a kind of rope?

Design criteria

My requirements of a boat are very simple:

1. It must not sink
2. It must be cheap
3. It must be able to carry the whole family, at a pinch
4. It must be able to be rowed, motored and — eventually — sailed
5. It must not sink
6. All parts of the boat — mast, spars, oars, motor, etc — must be capable of being stored inside it
7. It must be possible to haul it down to the water's edge without motor power
8. It really, really must not sink

I also knew from my extensive research that boats tended to be at least a bit pointy in the front, and that a boat that has a completely flat bottom will tend to cope with shallow waters better than one with a roundy bottom.

So I hit upon the idea of building a plywood box that tapered slightly at the stern and slightly more at the bow. It would be about 8' x 4' because that's how big a plywood sheet is, and about the largest object I can fit in my garage and still have room to work around it. The amount of taper was set in the most pragmatic way imaginable — how much I could bend the longitudinal timbers without needing to steam them.

During the construction, I discovered that what I was building was very similar to a class of small sailing dinghy called a `puddleduck racer' (`PDR' for short). Since these boats tend to have duck-y names, and since we had originally intended to name our boat `Death or Glory' (after one of the small boats in Swallows), we eventually hit on the name `Duck or Glory' for our boat. The main difference between our boat and a real PDR is that proper PDRs are required to have parallel sides. But it was reassuring to find out about PDRs because, until then, I assumed that I was the only person on the planet daft enough to make a plywood box boat in his garage. Apparently not.

Hull design


Testing the ingenious transportation system, and the patience of local motorists
Boat hulls come in basically two shapes — planing and displacement. A displacement hull is the traditional boat shape: point at both ends, but more pointy at the bow. A displacement hull cuts through the water, and is relatively tolerant of weight distribution (because however you load the boat, it's always pointy at the front).

A planing hull is designed to ride on top of the water. Most displacement hulls will plane if you push them through the water with enough gusto, but a planing hull has a large flat surface in contact with the water to encourage this effect.

Building a decent displacement hull is a skilled job. A planing hull can, at a pinch, be a shoebox shape, but with one end slightly taller than the other. A plain shoebox won't perform very well in the water, because the square bow will create a huge drag. However, if the bow is less tall, and therefore sits just out of the water, that friction is much reduced.

The real problem with the planing hull design is that it is very sensitive to loading, unless you're going really, really fast. If there is too much weight in the bow, then the bow is pushed down into the water, and essentially you're sailing a shoebox again.

My design for Duck or Glory was for a planing hull, but with the bow slightly tapered in the width direction as well as the height. That is, it would be like shoebox that was squeezed at one end, both in width and height. My intention here was to provide some displacement properties, in case the bow transome could not be kept out of the water. Of the many design errors I made, this one was the most significant, as discussed later.

Building the hull

The hull of Duck or Glory is made from two 8' x 4' sheets of 6mm exterior plywood and a half sheet of 19mm exterior plywood. The thick plywood is used for the bow and stern transoms (the ends, to us landlubbers) of the boat. I'm not at all sure that they need to be this thick for seaworthiness, but there are handles on the transoms for dragging the boat around, so probably thicker is better. In addition, the stern transom needs to be stout enough to take the weight of the outboard motor. But wood this thick does add considerably to the overall weight, and might not actually be necessary.

I did think a lot about the thickness of the bottom of the boat. 6mm plywood is not very thick if four people are standing on it. I appreciate that it is, to some extent, the water that provides the support for the passengers' weight rather than the hull; nevertheless, I don't have sufficient marine engineering skill (actually, none at all) to work out how much I could rely on this factor. So I decided to make the hull strong enough to take the passengers' weight out of the water, reasoning that if it was stong enough out of water, it would be strong enough in water. To that end, I put two sets of timber ribs (cross-wise supports) inside the boat, and two timber skids (lengthwise supports) under it. The skids are probably necessary in any case, because the boat will have to be dragged around, and it won't take much dragging to wear a hole in 6mm plywood. With hindsight, I could probably have done without the ribs and still had a strong enough hull.

The hull is also strengthend by the hard chines. That's boat-speak for the construction technique where the hull panels are joined by lengthwise timbers, rather than directly to each other. This, it turns out, is a rather old-fashioned (1950s) method of construction. A more modern approach is to stich and glue the panels — to join them temporarily with wire or ties, and then make a permanent joint with fibreglass tape and epoxy resin. Had I known about this technique I might have used it; but marine epoxy is a specialist material and very expensive in the UK. As it happens, I joined my plywood panels to the chines with screws and Wickes construction glue. That was probably a mistake, but one I seem to have got away with.

The reason it's probably a mistake is that, however carefully you plane the chines smooth, you can't — with only mortal skill — make them so perfectly smooth that the plywood will butt up to them without even the tiniest gap. And water can get through the tiniest gap. And a box full of water, it has to be said, is not really a boat in any meaningful sense. Wood glue on hard chines was a standard construction technique in the 1950's, and you'll find lots of reprints of old woodworking magazines that suggest building this way. What those old magazines won't tell you is that boats built this way only work if they're kept wet. They rely on the tendency of wood to swell when wet to form a watertight seal between the panels. For a boat that will mostly stand in a warm, dry garage, it's a no-no.

What I should have done, I think, was to dispense with the screws altogether, and join the panels to the chines with a gap-filling adhesive, like very slightly thickened epoxy resin. The problem with screws is that as the wood dries and shrinks, the screws become lose in their holes, and water can get into the timeber, which will eventually rot it away. So I had to fill all the screw holes (I used car bodywork putty), which was a drag.

For transport, I installed on the bottom of the hull a pair of small plastic wheels (trolley wheels from Homebase) mounted on axles made from long steel carriage bolts, that just happened to be the right size. The axles are glued into wooden blocks which are themselves glued onto the hull with Araldite epoxy glue. Because the wheels don't have proper bearings, and the boat is quite heavy compared to the load that I assume they are designed to carry, the plastic of the wheel hubs tended to melt and, occasionally, stick the wheels to the axles. They're OK for the short distances I have to travel, but better wheels would be necessary for a tow of any distance.

Mast placement

Most boats of the size of Duck or Glory have their masts right at the bow of the boat, and there are good technical reasons for this. But the problem with putting the mast so far forward is that the supporting structure for the mast won't add that much extra rigidity to the hull. How much of a problem that is in practice, I don't really know — other people seem to get away with it. I put my mast about 24 inches back from the bow, and the supporting boxwork makes the hull absolutely rigid. I can pick up the boat by one corner and it does not flex or twist at all. This was certainly not the case before the boxwork was put in place. The mast boxwork also serves as a buoyancy chamber (more on this later). Moreover, putting the mast back from bow provides a bow seating position for the Ship's Boy — in proper Swallows and Amazons style. The mast can also be used as a handhold while climbing in and out of the boat, which is useful.

The problem with this choice of mast position is that it will put the centre of effort of the sail rather too far back. This will have implications for the sail and keel design — more later.

First test in the water

"Don't tell me it leaks — I know it leaks!"
So, on a bitterly cold January day we dragged the boat hull the 200 yards or so to the water. One thing we learned immediately was that it was too heavy to drag that distance ever again. I'm certainly glad I didn't use thicker plywood for the hull.

From what I've said before, you probably won't be surprised to find that it leaked. A lot. But I was able to climb in and walk around without the boat tipping over or coming apart, which was a result. Also, the hull did not twist and there was no evidence of `oil canning' — the tendency of the bottom of the hull to distort upwards under load when on the water. But the hull is very heavily braced for a boat this size, so I'm not altogether surprised.

Waterproofing

In sub-zero January temperatures, it took about three days in the garage for the Duck or Glory to dry out sufficiently for the next step. Painting would have made it more waterproof, but I didn't want to paint until I was sure I was not going to have to glue anything else to the hull for waterproofing.

The way I made the hull waterproof, and which seems to be working so far, was to mix up a very small pot (200ml — and even that much was expensive) of marine epoxy resin with a glass filament thickener until it was the consistency of bricklaying mortar, and then ram it into all the seams with a wooden spatula. I also made a fillet of the same material along all the chine joints inside the boat, just to be on the safe side. When cured, epoxy resin, even thickened, is incredibly hard. If you put it on carelessly and end up with it in the wrong places, you need to grind it off — sandpaper won't shift it. But it does take a long, long time to cure fully when the temperature is around freezing. In fact, I had to bring an electric heater into the garage to get the temperature up to a few degrees, because the epoxy was still tacky after a week. Like most things, I guess boat-building is best done in summer.

But the worst problem with the low temperatures wasn't the epoxy filler, but the epoxy-based battleship paint I got from a military surplus dealer. This paint uses a xylene solvent and it is foul. Unbelievably foul. There's absolutely no way to apply it safely in an enclosed space unless you have breathing aparatus. So I had to paint with the garage doors fully open, when the temperature was below freezing in the daytime. At these temperatures the epoxy paint has the consistency of treacle, making it very difficult to apply. So the paint finish is far from satisfactory, and if the boat is still in one piece when the weather improves enough to work outdoors, I will need to paint it again properly.

Less than ideal towing point placement
Still, for all these cosmetic faults, on its second test the boat appeared to be watertight. All four of us got on board and paddled about a bit, and it didn't let any water in. Result!

Testing was made much easier by the realization that I could tow the Duck or Glory behind my bicycle. Initially I tied a piece of rope between the bow handles and looped this over a hook on my bicycle rack. This was fine uphill, but downhill the boat bumped against the rear tyre, which as a bit disconcerting. So eventually I made a proper towing point on the bow. This is better for cycling, but for dragging the boat around the tow point is really in the way, and pokes you in the belly. There's probably a better way, but I haven't figured out what it is yet.


Oars

Not a bad attempt at rowing, considering that the oars are twice as tall as the rower
The oars, like the rest of the boat, are cheap and cheerful. The handles are two 6-foot hardwood tree-planting stakes from the local garden centre, and the paddles are cut from a piece of floorboard. I planed the stakes roughly cylinderical, and planed a blade shape into the floorboard paddles. The paddles and handles are attached using Araldite, and the wood is treated with a few coats of pure tung oil. Tung oil is a very expensive way of waterproofing a pair of hack oars, but I happened to have some leftover from another job.

The total cost of my oars was therefore only a few pounds. But you get what you pay for — the tree stakes are not perfectly straight (they don't have to be for that application) so the oars tend to rotate in use, such that the paddles are not vertical. That's OK for me, but is a problem for the kids, who can only just handle the weight of the oars as it is. I did manage to straighten the oars somewhat by leaving them with heavy weights stood on them for a while. But choosing better wood to start with would probably be better.

Buoyancy and seating

I rather like the idea of a boat that is unsinkable although, of course, making such a claim is tempting fate, as the designers of the Titanic found out. Still, a small boat can be made difficult to sink, if not actually impossible, by adding bouyancy compartments. If these are full of air, then they need be no larger than the volume of water whose weight is equal to the loaded weight of the boat. In my case that's about 50kg, so I need 50 litres of buoyancy compartment. In fact, wood is only slightly more dense than water, and it doesn't need anywhere near that amount of buoyancy to keep the boat afloat (for the same reason that a lifejacket with a volume of 30 litres can keep a 200-pound man afloat). However, a subsidiary reason for adding buoyancy is to prevent the inside of the boat filling with water in a capsize. In that case, you need as much buoyancy around the sides as you can get.

Conventionally, small boats that have buoyancy chambers have them under the seats, which is why these two issues — buoyancy and seating — have to be considered together. A simple and elegant approach is to make the whole of each side of the boat a combined seat and buoyancy chamber, and many PDR boats are, in fact, built this way. It is, perhaps, an approach I should have adopted for Duck or Glory. The reason I did not do so was primarily that I thought it would be difficult for a small child to step over the seats into the boat, and that stepping on the seat itself would be unpleasantly unstable. I've since found from experience that the boat is stable enough to enter by stepping on seats at the side although this does, of course, put a horrible muddy footprint on the seat. But there is the additional factor that, unlike a PDR, Duck or Glory has curved sides, and forming (curved) seats along the whole side would be constructionally tricky.

At present, the seating in Duck or Glory consists of two 3-foot stern benches, which are for passengers or for driving with the outboard motor, and a moveable cross-wise seat for rowing. There are seat supports that run the whole length of the boat about 4" below the gunwale, so there's a fair amount of flexibility in the position of the rowing seat. It's also possible for children to sit on the mast box, facing forward or astern. The spaces under the benches and the mast box provide about 200l capacity for buoyancy material — making airtight buoyancy chambers requires a fair degree of engineering skill, so my bouyancy chambers are filled with polystyrene.

Motoring

This close to the water it seems faster than it probably is
Duck or Glory pootles along surpringly well when powered by a 40-year-old British Seagull 30cc motor. Of course, this close to the water you don't have to be going all that fast before it seems fast. I believe the Seagull makes about 1 hp on full throttle and, so far, I haven't had the throttle more than half open.

I mounted the outboard so that the exhaust was about an inch below the water level with the boat loaded, which puts the top of the propellor about four inches below the water level and about an inch below the hull. To get this mounting I had to make a cut-out in the stern transom about three inches deep and ten inches wide. I also put an extra thickness of 3/4" plywood under the cutout to strenghthen the transom where the motor clamps on, but this probably wasn't necessary.

The motor is mounted slightly off-centre, so that I can hang the rudder centrally on the stern transom. I don't envisage using the rudder and the motor together — the tiller on the motor provides adequate manoevring capacity — but the rudder will require mounting hardware on the transom.


Engineering observations

I never did get around to fitting Duck or Glory with sails. This would have meant installing a centreboard (because you can't sail upwind even slightly without one), and that would have meant cutting into the hull, and all sorts of other structural changes. However, I have rowed and motored around quite a bit, and I'm reasonably pleased with how the project turned out.

So what did I learn?

1. With hindsight, the hull probably does not need to be quite as sturdy as it is, and the boat could probably have dropped a lot of weight with only a modest reduction in strength. Making the transomes from thick ply was probably the right thing to do, although I could have just strengthened the area around the motor mount. I have since seen similar boats made without hard chines, and the seem reasonably solid. On the other hand, I've not seen any of them carrying four people, so I'm not really sure.

2. The sides of the boat are probably too high. The height is directly related to the expected loading — as more load is added, the boat will sink into the water in line with Archimedes principle. I sized the sides to give good clearance with a load of 700lb (that's a family of four, more or less). However, with only one person, or a couple of children, the hull sits a long way out of the water. That's not a problem in itself but, again, it does add to the weight. If I were doing this again, I'd make the hull 3-4 inches less tall.

3. If you're going to sail, plan for a centreboard or keel from the start. It's a hard thing to retro-fit.

4. By far the biggest design mistake I made was to make the hull side converge towards the bow, rather than parallel as they are in a puddle-duck racer. When I started, I could not get over the idea that a boat should be at least slightly pointy at the front. Since the design is for a planing hull, not a displacement hull, there is simply no need for the bow to be less wide than the stern. The bow must be less high that the stern, but that's a different matter. Making the hull curve in two dimensions, rather than just one, complicated every aspect of the design and construction. It made the cutting more complex, and meant that I had to bend substantial chimes to make them fit.

Concluision

It floats. It was cheap. Result.
Copyright © 1994-2013 Kevin Boone. Updated Feb 08 2013