Construction - part ten

Mike Griffiths (guest blogger) is he shaping up?


My posts telling the story of my stitch and glue lapstrake canoe construction are likely to slow as the main stages in the construction come to and end and I start in on the sanding and finishing. Before we get to that stage though there is still quite a lot of work to do preparing for (just about) the final addition to the structure - the gunwales. I still had to add some epoxy to the joints that had not been completely filled but first I deviated from the instructions. I knew that the European oak inwales and outwales were probably going to be a little bit difficult to bend and that this would stress the top plank in particular as they were bent to the boat's final curves. I decided to add a small fillet to both the top two plank joints behind the bulkheads - I also added a narrow glass cloth reinforcing strip. Why the addition? Well, as a first time builder, I could not be sure that those rather tight joints had been completely filled with glue and the last thing I wanted to happen was a joint failure.




With these additions in place, I then topped up the plank joints that were not completely filled the first time. The construction book suggests that you should use a slightly thicker mix than the glue used for the first shot but to apply it in the same way - using a syringe. I decided on a different tack and thickened my epoxy with lots of colloidal silica. I then used a pallet knife to push the thickened mix into the joints - using the knife to level off to what I hoped should be the correct hull lines.




Perhaps foolishly,I did all of this epoxy work in the lowest ambient temperature so far - which had an interesting effect. The epoxy cured nicely in the usual 24 hours or so but a few days later I noticed that there was an 'amine blush' on the surface. This waxy layer needs to be cleaned off with detergent and water. It is strange stuff though and I will try and avoid a re-occurrence by working at higher temperatures in future.




I took advantage of generally higher outdoor temperatures of the next couple of days to move the canoe hull back to the barn while I first of all cleaned my office and then started work on the inwales and outwales. First off, I sanded off any epoxy that had extruded from the scarfed joints and then I inspected each length of oak to decide which piece was going to do which duty. I then used the router to round over the top and bottom edges of the outwales. I found it was best to clamp two lengths side by side to give me a good platform to rest the router base on. After routing, I sanded the top and bottom edge to remove any saw marks left over from ripping the lengths from the original plank.

While I want to finish the corners of the inwales, I could not round them at this stage as the pieces needed to be fitted to the boat before I could determine the locations of the decks and the seats. I want to leave the inwales square under the decks for maximum adhesion and I will also need to fit "doublers" to increase the width of the inwales where the seats will be fitted. It is not clear yet if I will leave this finishing until after the inwales are glued in place or if I will try and sort them out before that.




I then set about clamping one of the outwales to the outer edge of the canoe hull so that I could cut it to length. First off, this looked somewhat tricky as the oak lath had to bend in three dimensions to follow the curve of the top plank as well as the twist of the hull. In practice it turned out to be fairly steady although a couple of extra arms would have been helpful from time to time. When the outwale was clamped and cut to length I could admire the improved curve of the hull - not quite perfect yet but definitely smoother.

Offering up the first outwale

I then clamped the second outwale in position and cut that to length. The next task was to stretch a string along the length of the hull to measure the mitre angle that would need to be cut between the inwales where they met at the stem and stern. With this measure I could cut the first mitre on the first inwale and then start working it into the boat. The involved removing each clamp holding the outwale in turn and the adjusting it to hold the inwale in place as well. One I neared the other end of the canoe I could mark and make the cut to complete the inwale and to set it snugly in place.

With the inwale in place, the top edge of the hull on that beam took on a very pleasing curve from end to end. Then it was just a case of following the same routine with the other inwale. Wow do you need a lot of clamps to build a canoe. It is very clear that I will need all that I have just to glue one gunwale into position if I want a good fit.

After that I was out of time and could not start in on gluing the gunwales into place. In any case, it was clear from my efforts so far that I was going to need an assistant for this particular process. Next weekend I am visiting the UK which means no work on the canoe but it will allow me the opportunity to pick up some materials difficult to obtain locally - some good yacht varnish and some coloured epoxy top coat for the outside of the hull. In the mean time the gunwales can stay clamped in place which might reduce some of the tension in the wood and make the gluing less of a fight - well maybe.

Construction - part nine

Mike Griffiths (guest blogger) sticks to his canoe building project

I forgot to post my tips on bulkhead fitting. The first is that you could consider using plastic ties for this part of the construction as it may be hard to remove the bulkhead ties after the epoxy fillets fixing them in place have cured. However my key piece of advice is – go up a drill size or two. The holes will be covered by the epoxy fillets on the inside and can be filled and sanded on the outside of the hull. This means that a larger size should not detract from the final finish but those bigger holes make threading the bulkhead ties simple rather than a nightmare struggle (you can tell I got to this solution after a bit of a fight). Otherwise – just concentrate on getting the bulkheads square to the line of the keel and nice and central.

Now to the sticky part. The first step in gluing the hull is to create epoxy and glass cloth fillets joining the planks at the bow and stern. My canoe construction bible specifies the use of wood flour as a filler for the epoxy – trust me on this – go with colloidal silica. This mix is simpler to manage and control – you will get better results. Having said that you will want to use wood flour on the fillets on the front (visible) side of the bulkheads to get a good colour match. Perhaps there is something to be said for getting in some practice before creating your first visible joint. I used masking tape to limit the width of the visible fillets although I think they will benefit from some additional sanding and shaping when I get a few minutes.

I made a few filleting tools from some scrap pieces of plywood picking my shapes and sizes at random so I was gratified to find that my hit and miss approach provided me with some pretty useful aids for spreading and shaping the epoxy mix.

I strayed from the instructions a little and used glass tape over the hidden bulkhead fillets as well as on the stem/stern fillets. The additional weight is trivial and I liked the idea of the extra adhesion following the fight to adjust the twist in a couple of the planks to get the bulkheads properly fitted. I also brushed some un-filled epoxy mix onto the plywood surfaces that were going to be joined by all of the fillets to soak into the top ply and so to ensure that the bond to the wood was optimal.

Once the fillets were cured I could turn the canoe upside down and set the trestles square and aligned – it was now time to glue the joints between the planks. Some of these joints are quite “open” on the underside while the last few inches at each end of the planks are pretty tightly closed with no visible gap to insert the epoxy mix. The tools of choice here are a syringe minus the sharp needle point and a pallet knife to slightly open the tighter joints for the epoxy glue to be pumped in. The glue is well mixed epoxy with a little filler. The specified consistency was “thin gravy”. Like the previous “jam” specification I was facing a potential transatlantic cultural gap here. I went with just a very small portion of filler in the mix and that seemed to work well. You need to work down the seams reasonably quickly and I found that mixing frequent lots of small batches of epoxy was best after my second batch went off at such speed that I lost the use of my first syringe. It would probably be a great idea to have an assistant available to mop up small runs but reality probably means that you are (like me) going to have to sand some of them out later.

Once the epoxy in the seams had cured I could assess the results. There were a good number of places where some additional epoxy would be required to completely fill the seams but I decided that this would be an easier task with the wire ties removed. So now was the time to pull out the 350 odd ties I had so laboriously threaded into the boat just a few days before. Like all the other tasks a few trials soon got me into a routine that sped up the work. You need three tools for this job – the needle nosed pliers used to put the ties in, a small pair of side cut wire cutters and (the surprise ingredient) an electrical soldering iron.

If you untwist a tie and then bend the two ends so they are once again vertical to the plank surface at the location of the hole they are easy to clip off on the outside of the hull more or less flush with the surface. Certainly without any “hooks” in their ends. Then it is a simple task to give the tie a yank with the pliers on the inside of the hull – and most will come straight out. One or two (more at the ends of the hull behind the bulkheads) will be held in place by epoxy – perhaps where they were inserted very close to a seam or there has been an unnoticed “dribble”. Here you use the soldering iron to warm up the wire to the point where the epoxy surrounding it starts to soften. Just a (very) few seconds of heating and out they come. I picked up this idea from reading around the stitch and glue boat building subject and in particular from Samuel Devlin's Boat Building book. Mr Devlin actually uses a car battery and a pair of jump leads rather than a namby pamby soldering iron but then he does generally work on a larger scale than my little lapstrake canoe. The same heating technique worked for most of the wires running through the stem and stern fillets as well – but not all, as the copper wire has a low tensile strength. It got me wondering about using steel wire at those points on the hull in any future project.

It did not seem a practical proposition to completely remove the ties holding the bulkheads in place although again I wondered if in future steel wire might work the trick here as well. However I threaded the bulkhead ties through larger holes than the rest of the ties and this made it simpler to cut them off flush with the outside of the hull. I could also bend some of them back into their holes below the hull surface. I found that I did not need to worry about the odd copper wire end poking up through the planks though as the copper is pretty soft and sands off easily with 80 grit paper.

After a couple of work sessions I had all of the wire ties removed. I was now the proud possessor of a canoe hull with nearly seven hundred holes in it – not all below the water line though.

Construction - part eight

Mike Griffiths (guest blogger) bends it like Beckham

Balls, you ask? No canoe planks! With the last set of planks stitched in, it was time to (I quote the book) “pull the planks together at bow and stern”. The book further suggests that this could be a fight and that a spring clamp or two might be needed to position things. A few practice flexes of the hull halves suggested that all this was true – and then some. Plus I should point out that the ends of the copper ties are sharp – this adds some pain and the odd shallow but painful gash to the process.

Inside and outside shots before stitching bow and stern

Slacking off the first three ties along the keel helped the first two planks to flex enough to allow me to start drawing the bow and stern together – but (of course) I need these ties in place to help get the bottom two planks to twist into shape. Note – spring clips immediately slide off stressed planks and fly across the room with a “ping” sound. Did I mention that the ends of copper wires are sharp? Oh yes I did – and they are. If you decide to build a stitch and glue canoe, recruit an assistant for this stage – and get someone with strong hands who is reasonably impervious to pain. You can do it by yourself though – I can't even begin to describe how – but you can. Working slowly from the keel line and allowing the wood to rest at a couple of stages seemed to work. Whoopee! I have something that looks quite like a canoe – well ish. The stem and stern were a bit wobbly and the overall shape was a little bit odd but it was a good start.

Placing a spreader stick measured to just short of the canoe's designed beam at the centre line helped with the overall shape but it then needed some adjustment with a mallet to bring each pair of planks properly into line and to fair the stems overall – well in one dimension. The other dimension needed a couple of lengths of wood screwed and clamped to the canoe to bring them exactly into line. One thing you learn is that even dimensionally matched pairs of planks do not bend equally without some encouragement – the minor differences in the internal plies in the board affect the rate of (and resistance to) bending. The “gains” are a weak point as well and do not help bring the ends of the boat naturally into line.

The continued sub-zero day time temperatures and a couple of days of snow falls meant that I ended up lofting the lines of the bulkheads and decks on the kitchen table. Sawing them out had to be done outside in the big barn but the physical effort generated some heat to counteract the cold. The Japanese style pull saw came into it's own here and the resulting pieces needed almost no finishing. I am really getting into this tool – it is just about the saw of choice now for close and accurate work.

Ready to start tightening all of those wires.

The good book says that fitting the bulkheads is “one of the hardest chores in boat building”. Amen to that. The holy writ also suggests that the bulkheads could move an inch or two either way to get a good fit. In my youth I spent me leisure hours with a fun group where most of the guys worked in the (then still vibrant) Chatham dockyards. We had a little mantra that went something like this:

“An electrician works to the neatest quarter of an inch. A joiner to the nearest sixteenth. A millwright to the nearest thousandth and a shipwright to the nearest ship.”

I have probably missed one or two trades but you get the drift I am sure. I suspect that the canoe project has just shifted from the domain of the joiner to the domain of the shipwright.

Next weekend seems to be dominated by some sort of winter solstice celebration – named after some chap called Noel. There was a hint that I was not supposed to spend most of my free time outside boat building so there might be a short gap before I report on the process of gluing the hull.

Construction - part seven

Mike Griffiths (guest blogger) is routed and then stitched up.

My canoe construction project had got to the point where I was gluing the half planks together to form the full length strakes used to construct the canoe body.

These two planks being glued side by side give a better view of the curvature that needs to be accurately maintained right through the joint

All of the planks were glued before the weekend and I was looking forward to cutting the rebates in the bottom edges to manage the overlaps that create the lapstrake effect. Both days were cold – the thermometer showed 2C and there was a strong North Easterly wind – no idea what the temperature was taking the chill factor into account but it was most unpleasant in our big barn where my workbench was set up ready to start work. The doors needless to say faced North and I needed them open for the light. Still – this was my opportunity to advance the project, so with gritted teeth...

It might seem odd for me to say that it was a bad thing that my first run with a router along a piece of test plywood went well. In fact it went swimmingly – like a hot knife through butter and with a perfect 9mm wide rebate just 1.5mm deep. So that made that decision – the router it was for this job even if this was the first time I had used it. The reason why the test run was not great was that the first real plank I rebated fought tooth and nail. I was not ready for the variability in the grain of the plywood layers as I advanced down the plank. I learned though and learned how to deal with the feedback from the router. Like everything else this was skill to be acquired.

I also had to cut the “gains” at the end of each rebate – allowing the planks to be drawn together at the two stems. I found that a combination of a small rebate plane and a sanding block was best for this final step in the shaping of the planks.

I had to drill 1.5mm holes every six inches along every edge of the bottom planks and along the top edges of plank sets two to four. I used the little jig advised by the book to assist in placing the holes accurately and drilled through each pair of planks together.

My simple drilling jig and the routed rebate along on plank.

The next step was to start stitching the planks together – but first I needed at least 300 short pieces of copper wire. I could not locate bare copper wire anywhere in my locality so set about stripping the plastic insulation off metre after metre of 1.5mm electrical wire. My younger daughter helped in clipping the lengths to around 10cm while I risked injury stripping the plastic coating using a Stanley knife. She watched “X Factor” while she clipped and I watched my thumb.

The weather forecast for the next week showed temperatures dipping below (and staying below) zero. This persuaded me that I would have to manage the next phase of the canoe construction in my office – which I could heat enough to keep from shivering too violently – and even warm up to a comfortable level when I next need to apply some epoxy resin. Of course, that makes lofting the bulkheads and decks problematic – perhaps I can borrow the kitchen table for that task – well maybe.

I extended the width of the two metal trestles I had to a tad more than the planned width of the canoe by drilling through the top bars and screwing some softwood to the top edge. I set these up and draped the first two planks (those that will form the bottom of the canoe) over them. I say draped because all five pairs of planks were very flexible and difficult to handle even allowing for my caution when moving them - I wanted to avoid (even more) accidental damage that might be difficult to repair.

I used the copper wire to stitch the two bottom planks together along the keel line. I then opened out the two planks and flexed them a little to get the two edges to butt together. I then fetched the next two planks to stitch to the first two. Offering them up got me very worried as the next two planks seemed too long. I hastily re-checked the lengths of both plank sets against the plan offsets – but they appeared to be in order. Even allowing for the fact that the next two planks were going to have to follow the curved edge of the bottom plank I could not see how I would not end up with too much plank at one end of the canoe. I re-read my bible (“The Canoe Shop” by Chris Kulczycki) – and there was a hint that this inequality would be evident at this stage. Also on close inspection, one of the pictures in the book showed an overlap between the first and second planks during the stitching process. Relieved? Well a bit – I decided to reserve judgement.

Once I had got part way through stitching the first of these planks to the bottom pair I felt the project change. The planks started to stiffen up as they took on the curves and the whole thing began to feel like it might become a boat. It might sound silly to say it but the canoe was starting to take on a life of its own.

A number 3 plank waiting to be stitched into the canoe.

Stitching on a plank involved drilling a hole just above the rebate opposite each of the pre-drilled holes in the adjoining plank. With a little twist the 1.5mm copper wire passed through the 1.5mm hole. I could then use some long nosed pliers to pull the wire through and to apply an initial twist or two to complete each stitch. I pressed a small piece of plywood to the back of the plank where I was drilling the holes to minimise any damage to the outer ply as the drill broke through – added the possibility of drilling into my fingers I suppose but (so far) no spectacular blood letting.

Construction - part six

Mike Griffiths (guest blogger) gets stuck into gluing.

I left off my tale of building a lapstrake canoe at the point where the scarf joints needed to turn the half planks into full length planks had been cut. Now all I needed to do was glue those joints. This was another scary stage as the planks are all curved and that curve needs to run smoothly and regularly through the glued joint. The canoe plans provided offsets from a line drawn between the end of the planks and the edge of the plank at the point where the joint should be made. However ensuring that the planks actually end up being glued in that position looked problematical at first.

The first thing I did was to buy three sheets of cheap composite board that looks like it is made from wood shavings. The ones I ended up with were narrow but had tongues and grooves along their edges making it simple to join them into a continuous strip. I nailed the joints to hold the sheets in register as I wanted to be able to draw a continuous line down their length. I was not going to rely upon this line but it helped get the planks roughly in position – I then re-checked that positioning using a line of string. I placed the first pair of half planks from one of the sets into position with the scarfs carefully overlapped and (after double checking all of the measurements) then drove pins into the board tightly alongside the ends and in a couple of positions along the top edge of each plank. This gave me a jig for setting up the second plank. After checking that the fit still worked for the second plank I mixed up a batch of epoxy resin – my first.

Mixing the resin was a new experience but the resin pumps supplied along with the West resin and hardener ensured that the mix was correctly proportioned. I had to thicken the mix using some colloidal silica – another first that needed some experimentation to get the thickness right. The book said to make the mix the same consistence as “jam” but I have met a lot of different jams in my time. In the end I went with apricot conserve.

I applied the epoxy to both surfaces of the joints of both planks. The planks were set in position within my mini jig with pieces of polythene sheet under and between the sheets. The joints were clamped using a short section of timber screwed into the boarding just outside the edges of the plank joints again separated with a strip of polythene.

With plenty of epoxy left over I then glued up two of the oak lengths destined to form inwales or outwales. These I clamped using plastic spring clamps as these would not stick to the epoxy and clearly had sufficient clamping strength to hold the scarf joint tightly closed. Three clamps per joint seemed optimal.

I had pre-heated the gluing area to about 16C (roughly 60F in old money) and ensured that the epoxy and hardener were also warmed through so they flowed properly – it was a bit chillier outside so a small fan heater did the trick. Now all I had to do was wait for the epoxy to cure.

The first canoe plank clamped in my gluing jig plus two oak lengths made from three 2 metre lengths clamped and curing.

Next come the other four plank sets and two more oak lengths. I also have to play around with the router to determine if I should use that or a rebate plane to cut the plank rebates – but that is getting ahead of myself.

Construction - part five

Mike Griffiths (guest blogger) continues his canoe construction story with plane tales.

I finished off sawing out the remaining sets of planks rather more quickly than the first couple of sets. By the time I had cut the fifth set I was reasonably happy that I was on top of the task. I was feeling so happy with progress that i went back and re-lofted and then re-cut one of the sets that I rather felt I had made a “pigs ear” of (now that's a phrase I have not heard in a long time - appropriate though).

With five plank sets now sawn out I had next to plane them to their lines. Planing a curved line seems just a bit disingenuous before you start and planing the inside of a curve even more so. However with a sharp block plane it can be done quite quickly. This was my final chance to ensure that my curves were fair (the canoe's you understand and not any personal love handles) and that the cut lines were square – giving me four half planks from each set that were as near as possible identical. By the time I had completed the last set I was really getting a feel for the tool. I suspect that a constant feature of this project will be the feeling that I am just achieving some measure of competence as I complete each phase. Is this why Paul described boat building as addictive? Not so much an addition perhaps but a desire to see what you can do starting out with better honed skills.

I now had to cut the scarf joints to make each pair of half planks a single full length plank. The scarfs should form a 1 in 8 ramp for optimal strength so, as each sheet of plywood is six millimetres thick, the ramps should be about forty eight millimetres long. The plans made provision for a one inch overlap on each half plank so I went for a five centimetre ramp – just about two inches long. Most builders cut their scarfs with a block plane but after a practice run on the discarded plank set I decided to try using a belt sander.

It is best to cut all four ramps for a plank set together as the staircase formed by aligning one plank end with the start of the ramp for the next automatically gives you the correct cutting angle. How cool is that?

Nearly half way through cutting one set of scarf joints. Just a bit wobbly but that can be corrected

Just checking that a finished scarf ramp is level

I made up a simple jig to help me cut the scarf joints on the oak lengths that will make up the inwales and outwales. This allowed me to quickly and accurately cut the 16 joints required in around 10 minutes (perhaps 20 if you want to include making the jig) – a lot less work than planing them or even using the belt sander. Now can anyone thing of a use for all those oak wedges?

Scarf jig for the table saw – the same 1 in 8 ramp as for the plywood planks.

Construction - part four

Mike Griffiths (guest blogger) gets to grips with his saw.

Cutting out the planks

This is the stage in canoe construction where you turn four sheets of expensive marine plywood into 20 well formed half planks of a canoe or, perhaps, just a small pile of sawdust and a stack of “off-cuts”. So much depends upon the accuracy of the preceding lofting process as well as the 'saw action'.

The half planks are cut from four sheets of marine ply sandwiched together. The idea is to ensure that both ends of the canoe are identical and (more importantly) that both sides of the canoe are identical. The alternative would be to end up with a canoe that did not fit together or only paddled in circles. It is important to ensure that the four sheets stay “in register” while each half plank set is cut out. The best approach is to nail or screw the sheets together at various points between the marked out planks – avoiding the cutting lines and larger waste areas that might come in useful at some later date.
Then, as each plank is cut from the sheet, the section hanging over the edge of the working surface can be clamped to keep everything tightly controlled as it is cut away.

My plan was to cut the planks out by hand. I had modified a Stanley 1-15-215 floorboard saw by grinding off the teeth on the upper/front curved face of the blade to stop these wider teeth jamming in the kerf when I was sawing at a low angle through the boards. I had also been reading that many expert woodworkers favour a Japanese ‘pull’ saw (in particular a ‘ryoba noko giri’) as they cut on the pull rather than push stroke and this allows the blade to be very thin with a very narrow kerf – ideal for this sort of project. Quality Japanese woodworking tools might be expensive but Irwin do a very reasonably priced version with two cutting edges. So a “head to head” between these two saws was in the offing.

Irwin version of ‘ryoba nook giri’ pull saw and the less exotic Stanley floorboard saw – prior to my modification

I elected to screw the four sheets of plywood together at intervals along the waste areas. I decided that screws would be easier than nails  to remove if they proved to be in the way. After one last look to check that my plank outlines matched the curves and lines of the sketch in the book I set to.

The Stanley saw was first up. I am a frequent enough saw user to have some confidence in what I am doing and this saw was spot on for accuracy – it followed the plank curves well and effortlessly held a pretty constant distance of less than a millimetre from the pencil lines. The snag was that progress was slow. At 13 teeth to the inch (now there is a non-metric standard measure) I suppose I should have expected this but progress was frustrating.

The Irwin “pull” saw made much faster progress – but – you knew there was a but – this was my first attempt to work with this new style of saw. I found it difficult to keep my saw cut close enough to the lines on a long cut – thus increasing the amount of work required to complete the cutting and shaping of each set of planks. Having said that, the advantages of this style of saw were becoming apparent – I just needed more practice. I suspect that I am going to become a frequent user of this saw and I can’t wait to use it on some jobs where it’s fine cut is going to be a real benefit.

So which saw is the best for this task then? Remember that I am cutting roughly twenty five linear metres through 2.4 centimetres of African mahogany. The winner is – “save up and buy a decent jig saw!”.

No seriously, (homespun philosophy moment) – once you have achieved proficiency with a hand tool then you will find the equivalent power tool finished the job faster – but no more easily. Power tools are not always better. This is just a longish job that is worth taking time over. I had far more fun ripping a couple of oak planks on the table saw to extract some lengths 2.5 centimetres wide to scarf together to form the inwales and outwales of the canoe. Given that I ended up buying two metre planks then I will have to cut 16 scarf joints to achieve the (roughly) five linear metres required – so one of the next tasks will be to make a simple jig to cut the scarfs as the time will be repaid in accuracy and efficiency.

Why were my planks only two metres long when this part of France is virtually a continuous oak forest? Well, most consumers of oak timber are artisan carpenters. They all have planer/thicknesers and band saws and can buy the pretty crude pieces of timber available from the saw mills. I could have easily purchased a slice through a tree trunk but would have had extreme difficulty in getting the required lengths out of that. So it was buy what was available at the DIY level or involve a local artisan to prepare the lengths of oak for me.

More from me when I have finished sawing – which might be a while as a combination of family commitments and entertaining look like “taking out” the coming weekend one way and another.