Learning Curve

Created/Modified: 2025-10-19/2025-10-20

This is how I built a skin-on-frame canoe that's large enough to sleep in and light enough to pick up with one hand, small enough to get on top of a Toyota Matrix and big enough to carry supplies for a week or more.

It wasn't designed so much as grown, slowly, over the course of a year. One influence on the design is the Snowshoe Traveller canoe by Geodesic Aerolite Boats. Another is the style of boat made by Brian Chandler at Dreamcatcher Boats in Vancouver, whom I worked with to build two canoes about ten years ago. A third inspiration is the coracles built by my ancestors along the eastern shore of the Irish Sea, thousands of years ago.

Her name is Learning Curve.

I sold my last sailboat a couple of years ago and since then have been kayaking to get time on the water. My Delta 15.5GT is big enough for an overnight trip or even a couple of days. People have done longer trips in similar boats, but I wanted something more spacious. I've also spent a lot of time under power while sailing, and have long thought that a solar powered canoe would be an interesting variation. Learning Curve is primarily intended to be a platform for experimental propulsion systems, from the mundane to the exotic.

A skin-on-frame boat consists of relatively few basic components: the stringers, which run the length of the boat; the ribs, which are steam-bent and run cross-wise; the stem- and stern-pieces; the gunwales and inwales and breast hook, the latter of which consist of little bits of decking forward and aft that help give the hull stiffness; the keelson, which runs the length of the boat inside the skin; the skin itself, which is ballistic nylon or Dacron; and the cutwaters, keel, and rub-rails, all of which are external to the skin and help hold it on, as well as protecting the boat from rocks and other obstacles. Finally, there are floorboards, which may be removable.

Materials are always a challenge. Stringers, keelson, keel, gunwales, inwales, should be clear (knot-free) fir. Ribs should be maple or oak, as should the cutwaters. The stem and stern pieces may be plywood or fir, as should be the breast hooks.

Beyond wood, there is glue, which in my case has been West System 650 Toughened Epoxy, and lashing: waxed nylon sometimes known as "synthetic gut", as it replaces the material used by Inuit and Aleut on their boats. I also used two sizes of bamboo skewers as pegs, and various size of brass wood screw as well as some stainless steel ones for the keel/rub-strip.

In the Beginning...

There also needs to be a place to build. I live in a modest house in Canada's Gulf Islands, which has as sun porch that overhangs an open area underneath. This was waste ground when we bought the place, but I've since levelled it, put down some gravel and concrete pavers, and built some dry storage bins underneath against the foundation of the house, which still left room for the build... just. It's also just tall enough for me to stand up in . I had no clear idea what I was going to do with this space when I made it. I just had a feeling it would be a good idea: the storage would be useful regardless, and it would add value to the house. But I thought there might be something more, as well. It turns out there was. This is not always the case.

The image shows how limited the space is. I put up a pair of two-by-four cross-pieces at a height of about 2 feet to rest the strongback on. The strongback is a wooden beam designed to give a stable, straight, construction platform. It needs to be about 20 feet long for this boat, and finding 20 foot long lumber is difficult, so I built mine out of various bits of 1x6 board I had left over from other projects. It had a couple of gaps of a foot or two long in the structure but so long at they are in the top and bottom and not the sides they don't allow it to sag, which is what really matters. A bit of twine run from one end to the other ensures a straight line for the build, regardless of any wobble in the wood.

Before setting any of that up, however, there was more work to do. In software develop we sometimes refer to this as "yak shaving", after a story about a CEO who is hiking in the Himalayas and runs across one of his senior software people in a remote Tibetan village. He asks what they are doing there, and the software person describes a series of pre-conditions that have to be fulfilled to build the next feature of the application that'll pay for the CEO's next vacation: to implement it they need to reproduce the behaviour of a library whose source code has long been lost, and the only binaries run on an obscure HP/UX variant that they eventually found the last existing unit on eBay, but it had a bad hard-drive so they couldn't boot it, and having taken apart the hard drive in a clean room found that the problem was with a custom electrical component that had burned out, and was no longer manufactured, but they found the specs for it and tried to build a new one which hadn't worked so they'd tracked down one of the retired engineers from the company that had made it and learned that the windings required for reasons no one understood interlacing with Tibetan yak hair... so here he was.

While the general idea has been around for a long time, some bits of that particular story come directly from my career in technology: fulfilling obscure and indirect dependencies is a real part of any engineering task.

In particular, I needed a setup that would allow me to rip really long bits of thin slab fir. There is a timber-frame building school on the island I live on, and around the time I started to get serious about building this boat they advertised on a local buy/sell/trade board that they had a heap of offcuts they'd like to get rid of, and anyone was welcome to come and take some. So I called them and ended up digging through their (huge) pile and found half a dozen boards that were between 3/8 and half an inch thick, all about 20 feet long and around 1 foot wide, with long runs of knot-free wood. I hauled them out of the pile and got them onto my car.

This was great, as the plan called for 20 foot long fir stringers and the thickness was about right in one dimension. Now all I had to do was rip a dozen-odd clear strips out of all this wood. I tried various ideas, with the tools I had, and none of them worked. I didn't really expect them too, but you never know: anyone who tells you they can confidently predict the outcome of something they have never tried themselves has the self-awareness of a turnip, and should not be taken very seriously. Me, I try all kinds of things that don't work because I recognize that--like everyone--I'm a pretty bad judge of what's practical and what isn't. And I've worked in enough fields where the "lore" was all wrong to be somewhat distrustful of "what everyone knows": I like to try stuff myself. This means I go down a lot of rabbit holes. Mostly they're dead ends. Sometimes, I find rabbits.

Having successfully discovered a handful of ways that didn't work, I bit the bullet and bought me a DeWalt portable table saw, something I've been lusting after for quite a while but could never justify the expense.

An important safety note: nothing I am doing here is particularly safe and some of it is considerably less safe than that. I don't consider myself an expert wood-worker, but I am an experienced one. I am also a reasonably expert machinist, and I've been doing this stuff for fifty years, since I was a young teen. The purpose of this write-up is to capture what I did, not what anyone else should do, and for people just getting in to this kind of thing I strongly recommend taking courses at your local community college, trade school, or whatever, and learn the modern and safer way of doing things. You'll notice I don't have the blade guard or riving knife mounted on my saw: this is bad practice, and I do not recommend it.

I have a pathologically high risk tolerance, but even so I do take some safety precautions. In particular, ear, eye, and hand protection are paramount. I always wear ear defenders when working with power tools, always wear appropriate gloves--which means latex or similar for working with glues and solvents--and always wear eye protection. If you wear glasses, modern polycarbonate lenses are a godsend. I also only operate power tools outside, or have good sawdust control in the form of a shopvac that attaches the the sawdust system of the tool. And I use power tools as little as possible, in part because their outdoor use is illegal where I live for most of the summer due to fire hazard.

My indoor shop area--which I rearranged in anticipation of actually building this boat inside, until it became apparent how ridiculous that idea was--is not that large, and especially for ripping 20 foot planks. I needed about 45 feet to do that. The house along the side where the porch is runs about 24 feet, which meant I needed some more work area, as there was just waste ground at one end and a heat pump at the other. That meant more yak-shaving in the form of building a patio area that would allow me to rip the long slabs in relative ease. There was a lot of cutting to do, and setting up to do a job properly is generally worth-while, even if it slows things down for a bit. As well as a boat, I've got a very flexible workspace out of this project that'll be useful into the future.

Ripping the Stringers

With this set-up I was almost ready to start ripping the stringers from the slabs I'd picked up from the timber frame place. All I needed was a way of handling the wood so I could cut 3/8 inch slices off the ~1/2 inch slabs. I sometimes see videos by people with huge shops who have table saws with beds big enough to host a Regency dance party on, and while I don't begrudge those folks their capital investment, I'm obviously a man of more modest means. So I found some 20 foot deck boards at my local lumber yard, and between a bit of finagling to hold them flush with the saw table and a saw horse on one end and an old stool on the other to support them, I found I could lay a whole fir slab down and have reasonably good control. Notice in the image how the partially ripped fir slab sags between the saw and the far saw-horse (it slipped off the deckboard run-out). This is why long solid support was necessary: those slabs are very bendy!

But... I don't like getting my fingers near the saw blade, and the thinness of the wood in both directions proved to be a problem for keeping it pressed flat on the table and flush against the fence, which is the part of the table saw that lines the work up parallel to the blade. This required a bit of innovation, and all innovation requires experimentation. Nothing just springs fully formed from our heads. One of my mantras as a scientist and engineer who has spent most of his life building things that have never existed before is: "My imagination sucks." Yours does too.

What we imagine on our first approach to a problem almost never works as expected. If it did, progress would be no more than a matter of imagining a solution to every problem, and we'd have flying cars by now. Or better public transit. We don't, because our imaginations can only see a few inches into the future. Beyond that is fog. That means to make progress we have to continually inch forward, making a trial, watching it fail, using that failure to see a little further, moving in a new direction... or sometimes realizing we're in a blind gully, and starting all over again in a new direction to begin with. It takes time and it takes persistence and it takes patience.

One thing you really want to avoid if you spend your life doing stuff like this is anyone who is even a tiny bit negative about the stuff that doesn't work. Anyone who says, "Ha! I knew it wouldn't work!" is worse than useless. They are also an idiot. For any new idea there is always going to be someone who says, "It'll never fly, Orville." They will be right nine times out of ten, by chance alone: most new ideas don't work. So it isn't some deep insight that allows them to see what will work and what won't. It's simply the laws of chance and their moronic insistence that nothing will work, ever. Ditch them.

What I ended up with to solve the problem of keeping the workpiece tight against the saw table and the fence is shown in the picture below: a couple of blocks of scrap wood, screwed together and screwed on to the deckboard on the input side, with holes drilled in them to take two pieces of stiff plastic tube that I had left over from another project, which had just enough elasticity and strength to keep things in place.

From the number of left-over bits I've used in this project it's clear that keeping your scrap is almost always a good idea, and when you start out with your first projects you shouldn't be shy about buying stuff you may or may not need: extras and errors are how you build up a useful collection of odds and ends, although keeping them all in some sort of order is also quite helpful. I have a lot of projects that were more educationally rather than practically useful, and that means I'm well-equipped with odd ends. Again: don't be afraid to do stuff that doesn't work out as you'd hoped. Learn from it, don't invest too much in it--emotionally or financially--and enjoy the process of doing it.

The outcome of all this was a dozen or so 1/2x1/8 fir stringers, 20 feet or more long, all more or less knot-free. As it turned out I did have to scarf a few of them to get rid of knots, and to reinforce a couple more in places where the grain wasn't quite straight, which I'll talk about more below.

A scarf joint is used to join two pieces of wood of the same size. You cut it on a high angle--about 8:1 in this case--and then glue the pieces together. Cutting was a challenging process, and I actually had a different jig for that that I didn't get a picture of. It was a piece of wood attached to the mitre gauge to form a highly angled fence that I could clamp the wood against. This was by far the trickiest part of the cutting process, and would have been better done with two people. As it was, I ended up pulling the work through from behind the saw, as its extremely flexible nature made it a pain to push. This also kept me well away from the blade and meant that if the workpiece did get kicked out at high speed--which is one of the many risks working with table saws--it would be firing off away from me.

Reflecting on it, I'm not totally sure why I didn't do the scarf cuts with my circular saw. It would have been simpler, easier, and safer. The answer is that I probably just didn't think of it, and I didn't think of it because I was totally focused on the problem of how to do it with the table saw, and like all human beings my attention is only capable of holding about three things at once: by focusing one way of doing it that's complicated enough to use up those three open slots, we become literally blind to alternatives.

This is why it's important to step back and consider the bigger picture now and then: our view will be more coarse-grained, but we might see completely new alternatives that were not visible when we were in the weeds. We can't, of course ever have an actually wholistic view of anything, because that three-slot limit would reduce such a view to a cartoon, but we can use our limited attention more effectively by being aware of both higher (less detailed) and lower (more detailed) conceptual and perceptual levels than whatever level we happen to be focused on at the moment.

The Best Laid Plans

The Snowshoe Traveller plan is something I used as a source of ideas and a jumping off point. At this point in the build, almost six months in, I still had no idea how far I'd be jumping off. As well as build instructions, the plan includes outlines for the stem and stern--which I did use more-or-less as provided--and station moulds, which I modified to raise the gunwale and add an extra stringer to the sides before I even cut them out.

I have a tendency to treat plans more as guidelines, really: I was looking to the plan to get a general sense of how I might build a boat like the one I wanted, not to build the boat described in the plan. And I want to be clear: the plan is good and if you're interested in reliably building a boat with known characteristics that you can be confident in, you should stick to it. Me, I was and am interested in experimentation and exploration. It's what I do.

I had raised the gunwale and added the extra stringer to give the hull a flatter profile. Many traditional canoes have a lot of rocker, which is the name for the upward sweep of the stem and the stern. There are fair reasons for this: it gives the boat more resistance to surf while giving it less side to catch the wind. Me, I've generally preferred canoes with low rocker and lower freeboard (the distance between the gunwale and the waterline) but in this case, because I'm planning to use it to sleep in, I wanted a little higher gunwale, and I wanted the flatter profile because I'm planning a Dacron foredeck that will be covered in solar panels.

The stringers, I have to say, went on very nicely, and in a very short time I had something that was starting to look like a boat. I did find when I inspected the stringers that there were a couple of places where the grain was week. They weren't knots, but places where the wood bent sharply rather than smoothly because the grain wasn't quite straight. I adjusted the position of those stringers so the affected location would be between ribs, and glued on additional bits of stringer material to reinforce things. This seems to have worked pretty well.

At this point I was still six months away from finishing.

Ripping the Ribs

Ok, this should really be called "Bending the Ribs" but "mere alliteration, natural to a poet" has long been my motto. Although I did rip the ribs, from some 8 foot long 3/4 by 3 1/2 maple stock I got at the local Finishing Store franchise, which are a great place to go for your boat-building needs. I understand they also do some supply for house builders.

I ripped the ribs are to about 1/4 inch thick, but the more interesting part of the process was bending of them. There is a claim sometimes made that you can't steam-bend kiln-dried wood. This is false, as I learned from Brian Chandler at Dreamcatcher Boats. Soaking the wood--not hard to do in the climate of the Wet Coast--will bring it back to life. Bending it is still harder than it would be if the wood was still green or just air dried, but it is by no means impossible.

I had at this point relatively little experience in steam bending, and for some reason found it a bit more intimidating than was entirely reasonable. My parents were both surgeons, and had a low tolerance for error, especially in their youngest child, who was, in the vernacular of the day, "a retard". My father was an excellent wood-worker and skilled boat-builder, and although I've done extensive rebuilds on boats--including boats my father built that didn't work out too well--I had at this point in my life never built one from scratch on my own before, especially not from what was becoming increasingly my own design. So I might have on occasion finished an important step on the build, and then sat quietly on a sawhorse on the patio I'd built while waves of panic and anxiety flowed over me, pouring out of the helpless child I once was, still afraid of my father's judgment and disapproval.

This is a shitty way to live, and if you live this way, I strongly recommend some time in therapy--I did, long ago--and then take a mindful and stoic approach to the surprise parties your subconscious will still continue to throw you. Part of this build process was reclaiming my soul. Reclaiming my manhood, really, because for me at least the idea of manhood is very bound up in being able to do stuff myself, although being able to ask for help--which is what going to therapy is--matters too. Every man should know how to fish, hunt, cook, sew, change a diaper, comfort a crying child, teach, learn, fight, stand up for what he believes without getting into a fight unless the other party brings it, read a challenging book, write a moving poem, start a fire with nothing but dry wood, and travel across the water in a boat of his own making, whatever concrete shape that metaphor may take.

For me, it takes the shape of a boat of my own making. I've been told that I'm very literal-minded.

So after the panic and anxiety passed, and I had done what I could to reminded that emotionally abandoned child that he'd grown up strong and secure, I went on to bend the ribs.

Water is a very nearly miraculous substance, and wood is not particularly far behind. Wood consists mostly of cellulose and lignin, which together form a composite material that is both light and strong. Until it gets wet.

I've done a bit of reading on the subject but I'm still not clear exactly what the role of water is in bending wood. I've seen claims that it allows the various fibres to move against each other, which seems unlikely, and more plausible claims that it affects the strength of the various fibres, although it isn't clear in any detail why that should be so. Cellulistic fibres are somewhat hydrophilic, which means they absorb water. Lignins are somewhat hydrophobic, so they tend not to. Regardless of the specific mechanism, when wood is saturated with water it gets bendier.

Heat allows wood to absorb water more quickly, so steaming is an efficient way to get water deep into the wood structure. I let the ribs sit out in the wet for a week or two in the hopes of rehydrating them somewhat, and then steamed them in a very simple steamer made from an old kettle and a couple of bits of four inch stove-pipe wrapped in cast-off towels to insulate them a bit.

To actually bend the ribs I came up with the idea of a "bending board", which I expect others have thought of before me. I hung up a piece of plywood on the fence surrounding the area where I had the steamer set up, and used a drywaller's square to measure out positions for screws I would bend the ribs around. I could have used the station moulds--except they were already attached to the strongback--or the plan itself to do this, and interpolated between them, but it was already becoming apparent that the boat wanted to belly out beyond the confines of the plan, so instead I measured the rib positions with a tape measure, taking the rib-to-rib distance and dividing by two to get each screw's position relative to a centre-line. I also had the idea that with a quantitative sense of the actual shape of the boat maybe I could reduce this new pattern to a plan, although I later realized this was neither desirable nor necessary, for reasons given below.

It was around this time that I realized I'd cut the stringers a little long when I'd glued the ends to the stem and stern, which I shouldn't have done at all until the ribs were all on. Or maybe not: the boat needed some structure to give it shape while I fitted the ribs. Regardless, the belly was fatter than the plan, so while I didn't remove the station moulds completely, I did take the zip ties off a bunch of them to give the stringers more freedom to negotiate with the ribs.

And negotiate they did. Although the bending board worked pretty well, the ribs were still dry enough that they wanted to spring back a bit. If I had been more expert I would have probably either soaked them longer in cold water to rehydrate the lignins and celluloses, or steamed them longer (they were in for over half an hour, which should have been lots for such thin members) or over-bent them so the spring-back would take them to the correct shape. Or something else. But doing things badly is how you gain experience. There is no other way. So I'm happy they worked at all, and have some ideas on how to improve things in future. My next attempt at this will still probably have some things I could improve on, but so long as it's better than this time I'll be happy. And if it isn't better, hopefully it'll go wrong in ways I can learn from.

With the ribs bent, putting them on became an interesting exercise. There are 29 all together, and this was the point where the number of clamps I own started to rise dramatically. With 12 stringers plus gunwales and keelson crossing 29 ribs, I had 15*29 = 435 crossings. I only clamped about 10% of them, but that's still getting on for 50 clamps.

I used to see woodworker's shops with endless racks of clamps on the wall and wondered how they accumulated them all. Now I know.

The kind of spring clamps I used were reasonably cheap and very effective. Having two sizes was helpful: the smaller ones were stronger and fit into some awkward spaces, but the larger ones could span bigger members. The plastic alternatives to these metal ones were not useful here, although a tried a few: they were too weak for this kind of work.

Problems

This is the point where a lot of problems with the way I was doing things started to show up. Most of them were consequences of me working things out as I went, so I did step 1 based on some idea of how I was going to do step 2, and then did step 2 differently. This is OK. The point was to learn, and get a working boat out of it if I could. And doing things this way was certainly educational, if for no other reason than I had to learn new techniques to carry out effective remediation.

As well as the stringers really wanting to come off the station moulds, the keelson was starting to lift up in the middle of the boat as the ribs were trying to bow out. I could have maybe pulled them back in by tying them together across the width of the boat, but instead I took an indirect approach and lay a bit of fir 1x2 down the keelson and clamped it on, which forced it to straighten out. This was what I wanted, after all: the keelson should be flat. That's a strong constraint on the structure of the boat I wanted to fulfill.

As a consequence of this, the sides started to splay out more, giving her a broader, flatter, more coracle-like appearance, albeit still with pointy ends.

And about those ends... By this time the fact that some of the stringers were too long was abundantly evident. The problem was that now they were off the station moulds the overall shape of the boat had changed in several dimensions, so the original lengths were no longer correct for some of the intermediate members. The gunwales and keelson hadn't changed much, but some of the intermediate stringers were no longer consistent with a smooth continuous curve of the hull. Oops.

At the same time, it became apparent that the plywood I'd used for the stem and stern was not quite up to scratch: some of the too-long stringers were exerting a shearing force on the outer lamination of the ply, and starting to strip it off the layers underneath. I ended up doing three kinds of remediation for this. First off, I shortened the stringers that were causing the problem by putting an extra scarf joint in them, cutting them in place with a hand-saw and gluing them back together again. The kerf of the blade plus maybe a little shift along the scarf took out enough material to correct the length defect: they really weren't off by much, but that little made a big difference.

Second, I pegged each stringer through the stem and stern pieces with bamboo dowels, in two places where the contact area was high enough. This is how I had attached the keelson as well, and it seemed to work pretty well.

Finally, I made custom bits of fir that were about 1/4 by 1 and a foot or so long that I cut notches into one of the flat sides to accommodate most of the stringers, glued in place, while the other side was glued flush against the stem- or stern-post, and pegged through with bamboo as well. This gave a lot of support. It also taught me a lot about getting a really good edge on my chisels, as that was necessary to shaping the slots. I use a fairly coarse double-sided whetstone for most sharpening: one side very coarse, one side a good deal finer but still short of the finest stone I have, which I use for knife edges and similar. I've found this more than adequate for working with softwoods.

Lashings of Lashings

The build document for the Snowshoe Traveller strongly recommends not lashing the ribs to the stringers because there are so very many of them, but I had already deviated from the plan in so many ways that I was effectively building a completely different boat, so I wasn't going to let that stop me. One feature of the Geodesic Aerolite boats is the use of Kevlar roving to add tension to the structure, so when the hull goes out of shape the wood provides compressive strength and the Kevlar tensile strength. It's a very clever way to make a strong, light, structure.

In engineering design there are two very basic tricks: you can separate two functions that used to be together, and you can put together two functions that used to be separate. The former usually increases robustness, and is a good move if things are a bit flakey. The latter usually increases efficiency, however defined, and is a good move if you have some robustness to burn. The GA design is already very robust, and I was planning to put on a tougher skin: ballistic nylon rather than Dacron. So instead of using Kevlar as a separate element in the design, I decided to combine the idea of lashing--which I had used on Dreamcatcher boats--with the idea of diagonal roving, which came from Geodesic Aerolite.

As such, I lashed in a diamond pattern, working out from the keelson, where I put two diagonal clove-hitchs around the rib and the keelson in two pieces of waxed nylon, which were each about 40 feet long. I would reel off the nylon on the 16 foot bench under the porch, running it from one end to the other and back, and then half way back again. Then I'd find the middle and knot it on the rib I was working on, diagonally over the keelson. Then do the whole process again but with the second diagonal knot crossing the first one, to form a "X" across the keelson, and giving me four free ends that ran off the diagonals. I would then zig-zag those ends down the ribs on either side, coming back to the central rib at every other stringer, lashing as I went, and keeping the waxed nylon on the outside for the diagonal runs, so when the skin went on it would be flush against it.

Wearing reinforced nitrile gloves was a good idea for all this: the nylon is hard on the fingernails and the wax dries the skin. The reinforcing was duct tape over one thumb and along the inside of both index fingers, where the thumb rests against them. I don't do a tonne of work with my hands these days and have found in the last decade I prefer to protect them rather than let callouses build up. My father hated gloves, and I never wore them growing up, which was kind of silly. Why not use technology to protect yourself when it's cheap and easily available?

The diamond pattern meant every joint had two lashings on it, upping the total number of knots to just shy of a thousand. It took a long time. I worked a few hours a day, most days, in the evenings, starting in June and running into early September. I figured out a technique that I was fairly happy with early on and mostly stuck with it, although I experimented wiht a few variations here and there because of course I did.

One thing I did change was that I started out gluing the rib to the gunwale just before I did the final lashing. This made it harder to negotiate the shape of the boat as I went along, working outward from the central rib, to the extent that I had to break a couple of the joints and reglue them before I could make the ribs further down the hull happy. Lesson learned, and this is the only way of actually learning lessons without an expert or teacher to guide you: do something new, see if it works (it probably won't), figure out why it might not have, revise it on that basis, and iterate. If you're willing to do things badly it's amazing how quickly you'll learn to do things well. If you're not willing to do things badly, you'll likely never learn to do them well.

By early September, the lashing was done and the hull got out from under the porch. I had actually planned to do more work out in the open air, which would have been quite pleasant on the long summer evenings, but a robin took up residence in a nook in the pumphouse wall, and I wasn't about to endanger her chicks, so I did all the work down below.

After all the lashing was done, with the rib ends clamped in place, I did a final adjustment on the hull and glued the ribs at the gunwale and then did the final lashing on each rib there. This was done in such a way as to keep the inner rib clear for the inwale to be glued on, although in theory I could have left some waxed nylon over and lashed that as well. I did add extra lashing to the inwale amidships after completing everything else, and may yet lash the whole thing on top of the glue. It adds very little weight and no small amount of strength.

Gluing the inwale, which has to be cut and formed pretty carefully to length, used a lot of clamps.

There's More Than One Way to Skin a Boat

With the inwales in place and the breast hooks glued and screwed, it was time to start with the skin. I had ordered a 21 foot length of 840D 9 oz/sq yd "Xtra Tuff" ballistic nylon from Corey's shop in Anacortes, not long before he passed away, and not long before, at the national level, Canada's former friend and ally went insane. I've since found alternative sources in Poland and Sweden, and know where I'll be sending my patriotic business in future.

The GA plan calls for the use of Heat'n'Bond double-sided tape to hold the skin in place, and I wasn't totally enthused about it. I'm sure it works well on Dacron, but I was using nylon, not Dacron, so the material has a coarser weave and maybe other differences that might, well, make a difference. The wood I was sticking it to was pretty rough as well, so I really wondered how well it would bond. Working with Brian at Dreamcatcher we had used staples to hold the skin on the gunwales before adding rub strips with piping to finish the job, and that seemed like a good idea to me: something flexible, familiar, and conventional. Given how not those things most of the other techniques I'd been using were, something a little tamer at this point appealed.

There are different theories of nylon skinning. Some people advise putting it on wet, and I don't think this is a bad idea. Brian Chandler recommends putting it on dry and then soaking it, which is what I did. If I were to do it all over again I'd probably do it wet, as while the skin is fairly tight it could be tighter. Nylon is hydroscopic--it absorbs water--and expands by a few percent when it does so. Even varnished it'll draw some water into the plastic, so the argument for putting it on wet--and therefore expanded--seems like a good one to me. I didn't do it this time because I was nervous about the whole skinning process, and didn't want to change things too dramatically from the way I'd been taught to do them at Dreamcatcher. Even I can be cautious sometimes.

The skin went on pretty easily. I laid it over the boat, stapled along the gunwales on either side, working out to the ends bit by bit and side by side, so it was all more-or-less symmetrical all the time. I spent a lot of time getting from one side of the boat to the other, stepping high over the crossbars that held up the strong-back like someone from the Ministry for Silly Walks. I kept the skin under tension with one hand and stapled with the other, using fairly short staples that didn't go all the way through the gunwale, and stapling every six inches or so. I could easily have put them in more closely together, and maybe should have.

Many people sew the stem and stern, but I wanted wooden cutwaters fastened over top the skin. I don't have any real objection to sewing except that it's not very hydrodynamic.

I cut the skin up to the stem using scissors--more on cutting later--and sealed the cut edge with a mini-torch, then folded it over the stem-piece with some epoxy underneath, and stapled it in place. Then did the same with the other side, leaving an inch or so over that I glued down onto the nylon skin itself. Then I folded a sheet of wax paper over everything and used a bit of wood on either side to clamp it all into place.

I had bent the cutwaters out of bits of rib material, and they didn't fit quite as well as they might, leaving a gap the sharpest part of the curve that I filled with a custom wooden shim. It is pretty? Not really. Does it work? Pretty well.

In keeping with what I'd learned long ago at Dreamcatcher, I then soaked the skin with a hose to get it well saturated. It did tighten up more after that. And it sure looked pretty drying in the sunshine.

Before varnishing I put on rub-strips along the gunwales and a short keel, about 1/2 inch deep from stem to stern. Punching holes in a perfectly good hull below the water line always bugs me, but it's worth doing to improve tracking and protect her from rocks.

The rub-strips improve the connection between the skin and the frame: I folded the skin over a bit of 1/8 inch plastic twine to act as piping, and put the rub-strip on top of the doubled layer, using clamps to keep everything in place as I snugged it up, drilled, and screwed. The screws were all through ribs, so they could be longer. Because the skin transmits forces through the boat as waves and whatnot try to deform the frame, this is a significant aspect of the overall structural strength.

Cutting the skin along the rub-strip with scissors or shears proved to be difficult, so I decided to experiment with heat. I'd already used a mini-torch to seal the cut edges at the stem and stern, melting the loose fibres together to ensure the weave stayed intact, and wondered if the same might be possible as a cutting tool. It didn't really work on a piece of scrap cloth--always test on scrap first!--so instead I tried a 40 W soldering iron, which worked brilliantly, along with a metal ruler that I slipped in between the folded-back edge of the skin that I was cutting off, and the skin proper. The metal protected the skin from the heat, and I could just rest the tip of the iron on it and slide them both along. Practice on scrap first, though.

The iron was the most powerful one I could find at my local hardware store, and if I could find a higher power one I'd use it. This one was hot enough to cut and cauterize the cloth at a reasonable rate, but it didn't want to be in contact with the metal ruler too long. Still, it created a beautiful clean edge along the under-side of the rub-strip. There were a few loose fibres that I was able to clean up with a mini-torch, still using the steel ruler to protect the skin: it would be embarrassing to burn a hole in it at this point, even well above the waterline.

The nylon weave is open, so has to be sealed with varnish or similar. Once again I went with what I know: Helmsman Spar Varnish, semi-gloss, three coats over about a week, with a very light sand using 600 grit in between, just enough to rough the previous coat up a bit, and very light around the lashing, which makes little bumps in the skin that are easy to over-sand if you're not careful. Gentle as she goes.

Proof

At some point in any experiment you have to power up the apparatus and see if it works. I always get a bit nervous doing that. By far the most nervous I've ever been is the first time a system I designed that was built by a team I led and which contained a great deal of my own code was used to guide a surgeon operating on a human being. Putting a boat in the water is relatively low-stakes by comparison, but I still felt a bit of a flutter. Would it simply fold up and sink when I got in, collapsing like one of those old-timey airplanes with too many wings? One way to find out!

The fit on the car was perfect, exactly as I had intended. At 47 inches wide she fit snugly on the 48 inch roof racks, which are a custom job I built a few years ago out of a couple of cross-bars from real roof-racks I got from a local with an eclectic collection of hardware, and some scrap 2x4's from a shed on the property I took down around the same time. The tie-down straps I use for my kayak were just long enough to span the hull properly. The vital forward tie-down rope, which resists aerodynamic lifting forces, kept the bow secure, and a bit of red ribbon on the stern warned tail-gaters to be careful. I kept to the speed limit, and would be cautious about taking it up to highway speeds, but it was very solid on the car.

Getting her in the water was easy, and I christened her with a generous splash of 18 year old Bowmore. The sea needs to be appreciated, the boat needs to be loved. Her name is "Learning Curve", for what should now be obvious reasons.

The tide happened to be high, and I launched at a local park with easy access from the parking lot to a sandy beach. A guy in a truck was pulling out as I was taking her off the car. "You build that?" "Yup." "Nice." These are the small gifts men give to each other.

The centre thwart seen in the picture below was a late addition, made and varnished a few days before the launch, and put on the morning of with some two and half inch #8 brass screws through the gunwales.

Getting in and out was easy, standing in about a foot of water. I was wearing my drysuit, as the ocean is getting cold and I don't fancy a plunge. I've swum in water cold enough to make my diaphragm go into spasm, and on one memorable occasion commando-crawled half-naked across water-covered ice to rescue my dog, who had managed to fall between the spring floes while chasing some ducks. I am acutely aware of how dangerous cold water can be: after all of about thirty seconds on the ice, getting out there, helping the dog up onto the ice by lifting it by its collar, and crawling back to dry land, I couldn't feel my arms below the elbow. Cold water saps strength much faster than it kills, but because of that it means that without protection from the cold your odds of saving yourself diminish very rapidly.

The floorboards are removable, made of some cedar lathe that would normally be used for fencing. They are light, but every ounce counts when hauling a boat around. I may in future put in some more permanent floorboards, but for now these ones worked just fine.

I'm pretty sure she doesn't leak, but as I brought water aboard when I got in and wasn't exactly paddling dry, it's hard to tell for sure. I'll fill her up a bit at some point and see if anything comes out. Those holes to hold the keel in place have glue around them, but you never know.

In terms of handling, she's pretty dreamy with no wind. Tracks well, turns nicely, and very stable. Kayakers talk about "primary" and "secondary" stability, but from my point of view as an engineering physicist what they are getting at is the shape of the righting moment with respect to angle. A "moment" in physics-speak is a torque, and when the boat is displaced from the level the "righting moment" is the torque due to differential buoyancy forces on either side that tends to push it back to the level. One of the experimental designs I built with Dreamcatcher had very nearly zero righting moment, regardless of angle, and could only be paddled by me, my kids, and a half-Polynesian guy whose people have been traversing the oceans for sixty-thousand years. It was fun, but the longest paddle I ever made in it was from the foot of False Creek out to the Kits Yacht Club, maybe 3 km.

Her primary stability is moderate, and her secondary stability is good: she tends to tip over onto the chine and then stay there. A lot of traditional canoes are like this, so you can move to one side and then paddle with deep vertical stroke, rather than reaching long over the gunwale with a sweep. I've not come close to exploring her performance envelope yet, but I like what I've seen so far, and don't expect any major surprises.

The only entirely expected feature I'm going to have to work with is her tendency to be blown around in the wind. There was no wind to speak of in the Pilot Bay, but a good 25 knots running out in the straight, and when I took her out far enough to catch a little of it she happily danced along with it. I was able to keep her under control, but it was work, and the breeze was really modest. So I'm thinking she's going to be really good for exploring the Gulf Islands, which are extremely sheltered, and probably not so good for going further afield. We'll see.

What's Next?

I've already figured out how to mount a seat for rowing on a couple of bits of rib material that will be lashed behind the centre thwart--which is actually a little bit forward of centre--and upon which will rest a removable seat. Then it just needs oarlocks attached to the gunwale somehow, probably via simple wooden clamps. At that point she'll be ready fro some more serious work-outs, and I have plans for a couple of along-shore local trips before taking her on an overnight to one of the marine parks in the area.

My longer-term ambition is to add an electric outboard--which I have from a previous boat--and solar panels, at which point I'll be able to travel in (lazy) style. Very few projects in my world are ever truly finished. My books are an exception, because the publication process draws a line under them, but authors have been known to revise published works, so even there nothing is quite final while we still draw breath.

This build has good for me. I've enjoyed it, fought a few demons, made friends with a few others, and created something I'm proud of, however many rough edges she may have. The state of the world frequently gets me down these days, and apparently I'm unable to do much of anything about it--I write a lot of letters to my MP, MLA, the PM, the Premier, and so on, to no great effect, and contribute what I can to various pushes for clean air and science-driven policy--so being able to create something, something of my own, something new, and something I can use to get out on the water is a considerable comfort. This may not be the last boat I build, but even if it is, I'm happy to be able to set out in a boat of my own making, across a sunlit sea.