Questions about Stitch-and-Glue

Robert has left a couple of comments on the blog which bring up some interesting points: -

My proposed Swampscott-style dory.

I'd be interested to hear Ross's comments about stitch and glue construction in general, and perhaps when he writes more about his new dory design he will. Was the caveat more the puzzle joints themselves or the stitch and glue process itself? Seems the boat went together well once the planks were correctly shaped. I've not yet completed a hull from either method but am eager to learn. Seems a lot of great stitch and glue boats are on the water, and it always seemed to me that for a one off boat the strongback doubles the workload

A photo of Scram Pram, copied from Jim Michalak's website

Ross, looking forward to hearing more about the Scram Pram. Hopefully a sailing report will accompany a future writeup. As regards your Swampscott ideas, when writing more about the design could you elaborate more about the strengths and weaknesses of stitch-and-glue and glued lap? Looking forward to more about that design.

Kits have a lot of appeal, especially to people who have limited time for marking and cutting, or for whom the idea of marking and cutting is intimidating. The instinctive reaction of many is that a kit of cut panels will reduce construction time enormously, and that the troublesome business of searching around for materials will be dealt with by the kit manufacturer, who in a sense becomes a one-stop-shop.

The problem is that the marking and cutting process makes up only a tiny proportion of the time required to complete the building of a boat. It generally takes me about four months to build a 15 foot dinghy, including painting and making the mast, spars, and rigging. This is when I'm working full-time at building and generally with two boats going at a time. But the marking and cutting takes me only two or three days, and I find it to be relaxing work. Once those couple of days are gone - even allowing a week for a first-timer - the remaining work is going to take the same time regardless of whether it is a kit or a scratch-built boat.

Marking-out panels for a boat using a batten made from a pine offcut and homemade lead 'ducks'. You don't need ducks - just use anything heavy, or hammer in small tacks to locate the batten.

So in my opinion, there is not much of a time saving, in percentage terms, gained by using a kit. Not only that, but the place where time-savings may be achieved is the easiest part of the whole construction anyway!

Kit manufacturers face plenty of challenges in getting good-quality buildable kits to their customers. One perennial problem involves getting long panels into an easily transportable 8-foot flat-pack. In the last photo you can see that I'm marking a panel shape onto a 16-foot long panel made up of two sheets of plywood joined end-to-end using a scarph joint. Once that part of the boat is cut out the two ends will stay in the correct relationship to each other because they are joined by a rigid scarph joint. By that I mean that the marking of the curvaceous panel shape is done after the sheets of plywood have been joined together.

Using CAD and CNC cutting processes, kit manufacturers do a great job of cutting out accurate panel shapes, but mostly they do it on 8-foot long sheets. That means that the panels have to be joined after being cut. Because the person making the boat is sighting along the curved edge of a narrow plank or hull panel, it is very difficult to make sure that one end of the panel is in exactly the relationship to the other end that the designer intended.

In the above sketch you can see two topside panels for my stitch-and-glue design Flint. The line in the middle of each panel represents a scarph joint or a 'jigsaw puzzle' joint as used by kit manufacturers. The two panels appear to be identical, but the lower one is misaligned at the joint by a tiny amount for illustration purposes (actually one half of one degree, or, 0.5 of a degree). Although the hull panels look the same, the difference is substantial.

In the picture above, I have superimposed the correctly shaped panel (shown in dotted red) over the one with the half degree error in the joint. You will probably have to click on the picture to get an enlarged view, but the right-hand ends of the panels are out of alignment by 18mm or 3/4". If one tried to build a stitch-and-glue hull from these panels, the boat would inevitably come out with a twist and other longitudinal asymmetry.
 My experience with stitch-and-glue hulls has shown me that the building method is wonderfully versatile, and that the time required to produce a light, strong and clean hull is vastly reduced over that needed to build conventionally over a station mold. BUT, the system can only work (without a mold, that is) if the marking out and cutting process results in port and starboard panels which are exact mirror images.

I try for an accuracy of less than half a millimetre (two hundredths of an inch). That may sound like boasting, but it is actually quite easy to achieve if care is taken. It does not refer to the absolute accuracy of the shape, but rather relates to the similarity in shape of the mirror-image panels. I mark and cut one panel from pre-scarphed plywood sheets and then lay it on top of the material for the next panel and trace around with a sharp pencil or a ballpoint. After cutting out the second hull panel I place the two mirror images one on top of the other and run a sharp block plane around the edges of both to trim them to the required accuracy.

In the photo above you can see a kit 'jigsaw puzzle' joint after being glued with epoxy. It should be obvious that the required angular alignment cannot be relied on from the puzzle joints themselves. There are ways around it. One way would be to glue one plank together, then trace around it onto the scarphing bench or floor, and them lay-up the mirror image for gluing over the tracing. Another way, which I think is better, is to glue-up the two planks at the same time, one on-top of the other, with waxed paper or thin plastic between them at the joints. What is important is to understand the danger of planks and panels which are cut prior to being joined being glued-up out of alignment.

The alignment issue is a problem, but one which is easily overcome with planning. What I consider to be the far more concerning issue with these 'jigsaw puzzle' joints is that the hard glue lines intersect the soft surface of the plywood at 90 degrees. This means that there is a stress-riser at the surface and there is a danger that repeated expansion and contraction over time will result in a fine jigsaw puzzle-shaped crack appearing through the paint. The only effective solution to that is to have a fabric sheathing over the surface of the joint - say 4oz glass in epoxy, or maybe even lighter fabric - I'm not sure.

It is important that you understand that I am not against kits. For some people they provide an answer to problems, and the manufacturers go to great lengths to improve the breed all the time. From a purely personal point-of-view, I prefer to build from scratch - but if you do wish to build from a kit it is very important that you are aware of the pitfalls. As for stitch-and-glue, I believe that it is the best way of utilising plywood for hull construction from an engineering perspective, even though I prefer cutting bevels and working timber, to the epoxy and glass work required by stitch-and-glue. My Swampscott design should allow me to indulge myself with both stitch-and-glue and properly beveled glued-lapstrake in the one boat.

So, after a long and wordy response you have my answer to one part of Robert's question. I'll try to get around to the Swampscott method in another post.