: i am also trying to build a sweet dream. at the same stage as you, but my ply
: cracked when the sides were half closed. the ply is french- made okume,
: seems of good quality with no voids. should i have wetted the ply first?
: the whole project is lost as far as i can see, even if i glass tape the
: cracks. anyone know pettingills e-mail address? regards, tony

hi tony,
i'm not building a tortured s&g, but mine has some severe twists, quite beyond the limits of the untouched ply i was using. so I had to play around a bit first with diff methods to test limits and possibilities etc.

these probably won't help on your project as you are getting failure real early(early in the design - don't know early for your ply), but here goes. you got nothing much to lose.

1/ mentioned before, but wet wood, even ply, will be more bendable than very dry wood. The moisture content in ply is low when bought, so in bendy areas its worth pre wetting the general area both sides for a significant pd of time(day or 2?) to get thru to the core. And the core seems to be a real important driver in how ply will bend. If you have radiant heaters or some way to keep the ply warm or hot(during the bending), even this will help a bit. Keeping a large area very very hot is probably too hard as such large surfaces cool quickly and you don’t want uneven heating etc.

2/ when bending ply from the edge, it is critical to bend it as a whole and not from any one point. For if there is an edge defect(holes, cracks,core gap, or edge profile inflection point), or the converse, a change in stiffness due to the end of a backer pc,joiner, or doubler - failure can propogate. (you have holes, have failure)(others get failure at ends of stiffeners)

As you have major access to the interior of the canoe, what about supporting this whole edge first by wiring in several semicurcular(stem profile) (say 4" wide) layers of ply from the outside on each side to support the edge so no kinks are allowed even at the holes and distributing the pressure over a larger area. Use steel wire not copper, more holes, anything that you can successively increase pressure over a large perimeter. And go real slow tightening evenly , with lots of support all around.

3/in extreme bending of ordinary wood, it is very important to not allow wood fibres to be in tension-they break. There are wood bending straps(lee valley) out there to restrain the outside surfaces of narrow wood when doing complex bends. but ANYthing that restrains the outside from expanding will increase the bending or achievablility of the bending.
For narrow strips of wood like strippers or skin on frame ribs, this is easy by using leevalley straps,metal packing straps, or anytype of strap that doesn't expand.
But what the heck to use on large areas like plywood? The simple boneass answer is what you’re going to do anyway. Glass it, but glass only the outside( use a ltwt glass to low glass/resin ratio to minimize additional stiffness – no fill coats at this pt). (Glass past the curvature areas for some distance to not introduce sharp differences) Even if the total assembly now is marginally stiffer, it’s ability to bend is INCreased because the outside is restrained
a lot more. Certainly worth trying as there is nothing to lose.

4/ the above points will help the existing situation, but if they don’t seem like will actually work to make the final curves:

take the pieces apart, and on the back of the preglassed outside areas to be prebent, route off in a square say ¾” grid ½ the way thru the ply removing ½ core thickness.
This also will greatly increase the ability of the ply to bend, but will retain just enough stiffness to act like a panel.
(This is what I eventually did. Took the ply, prestain and presealed, then panel cut, then exterior glass just past to be bent areas, route off back plies, and end up w/ floppy, thin, cardboard but very tough – can still drill holes and stitch)
this other helpful aspect of preglassing is that the edge where the wires go is now way stronger and resistant to wire pullout than before.
You have to be very very careful to remove a constant amt of material - the problem is that it is very difficult to get it to behave uniformly – may take some tiny core removal here and there – use lights to check curvature. A longish process, but will allow you SOME form of success.
The sacrifice is that now the interior will have to be dealt with aeshetically some way, and you’ll have to put in some extra layers of glass to account for some of the wood removed. i.e. fill in the router rebates w/ v. low density filler for core analog, then some tinted/coloured glass layers.

5/and if none of this works, take everything apart and trace outlines on bendable ply and use that, cause that’s what bendable ply is made to do. Seal real well as poplar not good w/ water - but you might end up w/ a lighter boat as I think poplar is lower density than most ply.

Anyway some experience, some ideas
-mick