I have been mulling over scarf joints quite a bit based on what I know about the properties of joinery and glue-ery, and stuff I vaguely remember from my Principles of Mechanics class about ten years ago.

[we'll see how this goes... might get kinda long and technical... and just bug out now if you flailed on that part of the SAT where you had to mentally fold up flat figures into cubes and tetrahedrons! :) ]

There are three properties of gluing that I'm thinking of in particular. One is that gluing into endgrain is basically worthless -- the small shredded fibers at the end of the grain tear out easily. The second is that a glued joint is highly resistant to shear forces parallel to the plane of the joint. The third is that a glued joint is not resistant to twisting forces where the axis of twist is perpendicular to the plane of the joint. (There is a fourth case which I do not know the properties of -- resistance to tensive forces perpendicular to the plane of the joint.)

To exemplify: imagine xyz axes such that X points out of the screen at you, Y points to the right, and Z points up. Now imagine two pieces of wood arranged horizontally such that their horizontal surfaces (XY plane) are glued together. Principle 2 means that this joint is resistant to forces applied in the +X/-X or +Y/-Y directions. Principle 3 means that this joint is not resistant to twisting applied to the top board clockwise (around the +Z axis) and to the bottom board counterclockwise. Principle 4, which I do not know the strength of, would be where you apply force +Z to the top board and -Z to the bottom board.

Now, to apply this to building a kayak... imagine that you have two pieces of hull section that you wish to join together. These are arranged left-to-right along the Y axis, with the plane of the boards in the Y-Z plane. You're standing on the +X axis.

First, take a butt joint. The two boards are simply butted together and glued. (a) This joint will be very resistant to compressive force applied +Y with your left hand and -Y with your right hand -- the boards get in each others' way and would have to buckle to get out of the way. (b) It will be very unresistant to tensive force applied -Y with the left hand and +Y with the right hand -- the end-grain fibers will tear out. (c) It will be reasonably resistant to bending moment applied CCW with the left hand and CW with the right hand -- the fibers at the top of the joint will be in tension and those at the bottom in compression. (d) It will be fairly resistant to shear force applied +X with the left hand and -X with the right hand. (e) Finally, (the most common complaint about butt joints, from what I see) the joint itself will not bend evenly if you apply a -X force with both hands -- it will leave a little flat spot in the center of the joint.

Not a very spectacular repertoire.

But, for a scarf joint... my first and primary question is, isn't it also ALL end-grain? For the same cases I described above, and assuming the end-grain problem isn't really a problem, the scarf joint will be very resistant to both the tensive and compressive forces in (a) and (b), since they are nearly in the plane of the joint. It will be fairly unresistive to the bending moment of (c), since this is twisting perpendicular to the joint. It will either be very resistive, or not very resistive (point 4 above??) to the perpendicular tensive forces in (d) -- depending on whether you are trying to shove the top scarf into the bottom scarf or trying to shove the bottom scarf away from the top. And with the bending in (e), ... I seem to intuit that the scarfed section won't bend very easily, but spreads the "flat" area over a longer section so that you don't notice it as much.

(Anybody still with me? hello? hello? )

So from this it sounds like a scarf joint is generally better all around -- however, what are the primary forces that a hull will experience? Seems to me like bending moment -- your weight pushing down in the center and the boat's buoyancy pushing up semi-uniformly along its length, meaning some forces pushing up on both ends; also, you'll get dynamic bending moments if you take a wave on the nose. These twisting forces are the ones that the scarf joint is the poorest at resisting.

One answer would be that the hull is not a single layer... it's a composite made up of the layers of glass as well. If you try to twist the joint along its plane, the glass fibers at the bottom of the joint will be put in tension, which they are very, very, very, very good at resisting. The wood layer is good at resisting compression, the glass at resisting tension.

(Well, if nothing else, this little monologue has really served to clarify my thinking...)

Anybody want to critique my thinking, or comment on other merits of the two joints?