laminated mahogany springback

[galleywench]

Hi All,
I'm starting the process of laminating toerails for my Alberg 35 and I did a test piece over the weekend to make sure I'm on the right track. For full details goto http://www.alberg35.com, but in a nutshell, I am bending and epoxying 3 - 1 inch mahogany strips onto a form that mimics the curve of my bulwarks. The idea is to laminate the curve to avoid bending a big 3 inch piece of mahogany on and risk breaking the wood. The test piece (~2') I did over the weekend bent onto the form fine and when I released it the next day it seemed to hold the curve fine... for a while. Eventually it sprung back to it's original form and I suspect that it is because I failed to let the epoxy cure completely (I released it from the form after about 18 hours).

So the question is... Will the wood spring back regardless of how long I let it cure?

Thanks

[Hirilondë]

First you mention toe rail, then bulwark. The difference might seem moot, but it isn't. From looking at your blog it is apparent that you are making a new caprail for your bulwark.

After looking at the orientation of the boards, the dimensions, etc., I am not surprised they sprung back. For laminated pieces to hold a curve the curved joints (seams) must offer far more resistance to spring back than the memory of its components. You have 2 joints resisting 3 boards each with substantial memory and desire to spring back. You were doomed before you started. Most curved laminated components are made from many thin (fractions of an inch) pieces glued together. The result is quite stable and very strong.

For cap rail I have always used boards that were wide enough to become the cap all by themselves and cut them to the curve and pieced sections together with decorative joints. Unlike a toe rail the cap rail is merely a decorative top for the bulwark and does not need to be bent to maintain strength like the toe rail does.

I suggest cutting the cap rail to the curve out of solid wood or previously laminated straight pieces. I think the solid will look better as the seams will not follow the curve of the cut.

[Quetzalsailor]

Springback is always an issue when bending wood, by steaming, laminating, whichever. One way to combat this is by overbending, that is, judiciously build your form with more bend than the final result will require. True for metal shapes, too. It's the 'judicious' part that's hard.

Jack already mentioned another way to combat the problem, by using thinner lamina. I would suggest that what's at work here is not the thinness of the lamina but the percentage of total thickness of the piece that the lamina comprise. Think of the bent glued-laminated timbers that used to be popular for churches, etc. The lamina are 2" x stock, but the members are usually quite thick at the bend, say 8-10 thicknesses.

What's going on here can be understood by taking a stack of paper or a paperback book and bending it while allowing the pages to slip over each other. The glue, which you did not put between the pages of the book - I trust, has to resist the shear force that the lamina want to apply while trying to straighten back out. If you instead, took some thicknesses of rubber and glued them, you could see that the individual lamina bent readily and when they were taken out of the form, the extreme fiber - the rubber lamina at the inside and the outside of the lamination - would both want to spring back but would also be able to change length to allow the springback.

Since your glue joints are roughly thirds of the total thickness of the lamination, the amount of slip they're locking in is very small compared to the change in length that the surfaces of each lamina had to have for the shape you bent. Presuming you like the appearance of the three stripes of wood and you've bought the wood, experiment by overbending the lamination in the form. Most materials creep under load. Even concrete. Glass. Wood. Fiberglass. Combat your springback by leaving the object in the form until the wood creeps a bit, or get it into position.

[galleywench]

Thanks for the replies, I stand corrected on the toerail/caprail... it is now firmly planted in my head as a caprail. Looking at the shear force of the lamination vs the shear strength of the green epoxy it's now clear that I would lose badly. I'm glad I did a test first. I'm going to go back and retest but leave the lamination in the form for a few weeks to ensure a full cure and see how much better it fares (if any). Ultimately, I will have to get it onto the bulwark faster than I had hoped. This means that I'll have to wait until spring (warmth) so I can clean up the bulwark and fill the holes left from the original caprail.

I originally planned on cutting the curve out of a solid piece as was suggested (in fact that was the way the original caprail was cut), but I never cared for the look of grain run out that occurs when wood is cut in that fashion and didn't want to waste so much wood. A friend of milled a 2.5 inch solid mahogany caprail that we bent onto the bulwark during his restoration. The look is absolutely stunning, but it was not for the faint of heart. We used a come-along and pulled it onto the bulwark over the course of 2 days. It could have easily split and ruined a lot of expensive wood and I'm worried that my 3 inch rail will just not take to the curve.

I'm not completely sold on this method yet, but I'm still not deterred.

[Sailordave]

While there is a simplistic formula there are a number of variables at play here. How much curve, what type of wood, how readily it accepts a bend and most important the number and thickness of plies.

http://www.woodweb.com/knowledge_base/C ... gback.html
is a link that might help for future use. As you can see the more plies you have for a given thickness the less springback.
And if you are using a species of wood that steams easily you can prebend it by steaming, let dry and THEN glue up.

[Sailordave]

I'd still use a wide board and cut out the caprail though...