As it became apparent that my first strip-built kayak, the Geyrfugl, was going to be more successful (and look better) than I had imagined, I developed the confidence to start to plan a kayak of my "own" design. I put the "own" in inverted commas because inevitably it would be evolved from some other design rather than pulled out of thin air.

Clearly, if I want another Greenland style kayak much like my existing North Shore Mistral (which is an excellent boat) there are plenty of such designs on the market. But the whole point of strip building (to me) is to make something that you can't buy. The Geyrfugl is a sea kayak for a paddler so small that commercial kayak designers just don't find a big enough market. My second boat - a hybrid Cormorant, is also scaled to a smaller volume than most commercially built boats. For longer trips, I wanted a kayak of a particular style - one which can't be bought off-the-shelf, at least in the UK.

Inspired by Rob Mack's North Star design, I decided that there must be something to be said for the Aleut style baidarka. But even after reading George Dyson's book, I didn't feel that I wanted to build a skin-on-frame kayak, (yet). Strip-built boats are just so attractive that I wanted to build another one, and I appreciate the benefits of watertight cargo compartments and hatches. So I returned to sources and looked at the design of several original Aleut baidarkas in various museums, documented by David Zimmerly. The Atka Island kayak (LM 2/14886), collected in 1934 and now in the Hearst Museum at the University of California at Berkeley, is the most modern of these Aleut designs, and seems to have been the starting point for a number of people's modern designs, particularly skin on frame (wood or aluminium) designs. I read Rob Mack's description of the sorts of changes he made to an Aleut design to come up with the North Star. Some of those ideas seemed to agree with mine, and others perhaps weren't what I wanted. Rob rounded out the keel of his design because this seemed better suited to a recreational boat which would most often be used with very little load. I wanted a boat for multi-day camping trips and I'm prepared to do as the Aleuts did and carry some ballast to improve handling with no cargo, so I left the keel somewhat angled.

The Atka Island kayak is shown with a waterline based on a displacement of 176.5 lb. The first thing I did was rebase the table of offsets to that waterline (which is not parallel with the datum line used by Zimmerly). The offsets given in Zimmerly's pdf file, when lofted, don't give a fair shape - perhaps because these are measurements from a kayak which was not in perfect condition when measured. So the next thing was to redraw a kayak based on these offsets, but faired out to give smooth chines. Then I re-scaled these offsets to give a length overall of 5.5m (I am probably bigger than an Aleut and a displacement of ~ 80 kg would be a bit on the low side even for just me and the boat). I now had some rather widely spaced forms - not ideal for strip building, so I wrote a little program to fit splines through the existing offsets and give me a table of offsets at new form positions. This was not quite perfect, but the few odd anomalous points were easily enough fixed. Then I tweaked and fiddled with the hull, with the eventual result that I had a set of offsets giving a displacement of 98kg, perhaps still a little low, but I didn't want a longer boat than 5.5m. In fact, I suspect that an Aleut paddler with a hull containing the results of his day's hunting would be displacing more than that 176lb figure, and the boat would ride a little lower than Zimmerly's postulated waterline. By adjusting the offsets so the hull rides 2 cm lower in the water, I got a displacement of 130kg, which should be OK for several days camping.

To keep the skin-on-frame boat light, and shed water, the original deck has a sharp peaked section, with just one stringer along the ridge. The strip-construction technique doesn't need stringers, and I wanted a little more room under the foredeck, and perhaps a slightly flatter aft deck, at least just behind the cockpit. I also prefer a more rounded sheer both to make rolling easier, and because I find a sharp corner quite hard to build. So I replaced the deck with a half-elliptical section of about the same height as the original, except in front of the cockpit, where I raised it a little. This cants the cockpit back to ease entry, so I can keep a fairly small cockpit. I didn't want a huge keyhole affair where a breaking wave could perhaps implode my spray deck ! Originally the cockpit was drawn 65 cm long, but I built the cockpit in my Cormorant 59cm long, and this is a really comfortable boat, so I am inclined to go for the same size on the baidarka.

Having sat and thought all summer about the design, I added to my collection of code by adding routines to calculate centres of buoyancy, lateral resistance and wind resistance. This confirmed my "eyeball" suspicion that the centre of lateral resistance was some way behind that of wind resistance of the hull alone. A paddler would move it back, but I still worried about getting a boat with lee helm. So I deepened the keel over the forward 2 metres, which somewhat increased the waterline length and reduced the prismatic coefficient, which should have the additional benefit of making the boat faster at higher speeds, possibly at the expense of making it slightly harder work at low speeds. Lee helm is to be avoided at all costs, even if I overcorrect and get weather helm - the latter can be corrected by adding skeg. By October 2002, I was getting confident enough to build a prototype, but 5.5m is too big for the biggest place I can heat overwinter. Hence the project was delayed whilst I built a smaller boat for additional experience.

Once I had paddled this new boat, which is really quite low volume, but a very comfortable fit, I went back to the baidarka design. More additions to my code produced an overall volume figure of almost 500 litres, and I had already been worrying that the boat was a lot deeper (and therefore a lot higher in the water) than other baidarka-inspired designs I had looked at. British waters are windy, and I didn't want something that bobbed about on the waves at the slightest whim of the breeze. Another consideration was that I had thought of building a stitch-and-glue hull as my prototype, and at 5.5m, the boat was fractionally too long to get out of two plywood sheets. Making it a few centimetres shorter was going to be a lot more practical than scarfing on a whole extra sheet.

Unfolding the panels for stitch and glue showed the chines were still not as fair as would be needed for this type of construction. So I printed the cross-sections full sized, redrew the chines and measured off new offsets. This made almost no change to the hydrostatics, but should make fitting ply panels together a whole lot easier. It will probably be something of an aesthetic improvement, even if I build it as a stripper.

In 2004, having had a season of paddling the Cormorant, I had a fair idea of what I wanted in terms of stability and fit, and I could compare the sections of the proposed Borealis design with those of the Cormorant. It was apparent that Borealis had a lot more freeboard that would be comfortable to paddle, and a lot more knee room than was required. The overall volume was also very high, even for an expedition boat. Over several iterations, this resulted in a revised design scaled to 5.4m long, beam almost unchanged and depth reduced to 84%. The lower volume design is still 375 litres, with a beam of 55 cm (51 cm at waterline) and a displacement of 120 kg at the waterline as now drawn, with enough freeboard to load it a lot more than that. I now believe this is a boat I can build and paddle, as an expedition single.

"Borealis" is my code name for the project - not the name of the boat, which has yet to be decided. Current thinking is to build a stitch and glue prototype, followed by a strip-built boat with whatever adjustments seem necessary. You can download my offsets as revised to 2004-02-27 in plain text (Unix newlines) and see my study drawings on this page. The rather odd form spacing is the result of starting off with a neat 0.3m spacing on the 5.5m design and scaling it down to 5.4m.

Having revised and refined for over 2 years, the vital statistics are:

I've now paddled both a commercial fibreglass kayak (4.9 m x 55 cm) and my new hybrid Cormorant (4.66 m x 52 cm) in quite rough conditions for a full season. I've also paddled Geyrfugl (a boat designed for a much lighter paddler, with a 49 cm beam) in the sea, though not in rough conditions except very close to a sheltered estuary. This suggests that I want a boat with a reserve of secondary stability and a low windage for longer trips. The extra length and flared sections fore and aft should meet these criteria with the current depth and a beam of 55 cm. If I feel that the design needs modification and a second build, then I'll consider going down to a narrower beam and, for a stripper, perhaps back up to a longer length.

My original drawings showed the boat with two aft bulkheads giving a main hatch and a day hatch. In practice, I've not found much benefit from having a day hatch in my fibreglass boat, and haven't built one into either of my wooden boats. I think a knee tube would be a better solution for keeping small things handy and should work well with the small "Ocean-style" cockpit (59x47 cm). This is the same cockpit size I used on my second boat, which is really comfortable, as one's knees are right under the deck where the knee tube will provide excellent bracing. This gives me the option of having a single larger aft compartment, with access through a hatch in the bulkhead, which should be more seaworthy and make the aft deck much neater. This was exactly the approach I took in my second boat, and I have been quite happy with it so far - access to the cargo is easy, without having to fiddle about removing split paddles and other things carried on the deck. These revised ideas are now reflected in the study drawings.

Having read Norm Sanders' article in New South Wales Sea Kayaker, I plan to add a mast step and some deck reinforcement to support his design for a downwind sail rig to this boat.

Among a number of other ideas I considered for this boat was a take-apart bulkhead so the boat could be taken into sections small enough to transport by air. An obvious point to split the boat is at the forward bulkhead, whose position is determined by the cockpit position and the length of my legs. This is 2.25m from the stem, leaving a 3.25m section aft. Another split at a bulkhead behind the cockpit would give a rather small centre section, but a day hatch giving access to a small compartment aft of the cockpit would put another bulkhead at perhaps 4m from the stem, where a split would give a 1.75m cockpit section and a 1.5m stern section. My current thoughts are that this will have too much of a weight penalty for use in the UK, and, as mentioned above, my thoughts are turning against a day hatch anyway. I have also been building with sloping and curved bulkheads to make my feet more comfortable whilst keeping the cockpit volume as low as possible - take-apart bulkheads really need to be completely flat to be feasible to build. Maybe a splittable boat will be a future project. If the prototype turns out to need a drop skeg, then there will be no chance of adding a split aft of the cockpit.

Unfortunately, I have been told by the one in charge of household affairs that she cannot take me building more than one boat a year, and since this is too big for my heatable garage, it could be spring 2004 before work starts in earnest. Or maybe I'll spend my summer building a heatable interior space big enough for a fast double in our barn.... there are lots more boats to build yet :-)

• Geyrfugl building page
• Hybrid Cormorant pages
• Andy Waddington's kayaking home page
• Andy's main home page