Side Bending--Sorta

To bend the sides I use a contraption I built myself using some plans. Though it's called a Fox Side Bender, after luthier Charles Fox, who invented it, George Morris, the luthier I studied under, was working for Charles at the time and collaborated with him on the design. Here's George:


Great teacher. Very patient and easy-going. He has a pretty amazing life story that I'll have to share with you some time.

And here's his side-bending machine that I built. Could have bought one for $600; I built it instead for about $100, and most of that was the hardware. The cutaway bending form is in the machine.


This is a front view with the non-cutaway form next to it.

In the center, attached to a press screw, are a series of 1/4" thick slotted leaves. I loosen the bolts in the slots and press the leaves down until they contact the waist of the form. They then take the form of the waist, and by tightening the bolts, they retain the exact shape.

The wood is shaped by heat. It used to be common to soak the sides in water prior to bending, the idea being that this would soften the wood and make it more pliable. However, this also weakens the fibers of the wood, and can lead to cracking in highly figured woods like curly maple. It's not water that allows the side to bend, it's heat that softens the natural resins in the wood. Water just acts as a conductor, but soaking is not needed. When the wood cools, the resins harden again and the wood become rigid.

So here's the procedure. I lay out some aluminum foil the length of the side, and lay the side on the foil. I've attached the rosewood bindings to the side so that they can be bent at the same time. I only need two pieces of binding, but these are easily cracked or broken on the cutaway side, so I hedge my bet by taping three pieces to the side, there at the bottom of the picture. At the top are two pieces of the kerfed lining that I want to bend as well, but this makes the whole thing too wide for the stainless steel slats and I take them out. Because of the kerfing they are pretty flexible, but pre-bending them helps them conform to the tight curves of the cutaway without breaking. Breaking isn't a big deal, however, so it doesn't really matter if I pre-bend them or not. You'll see what I mean when we get to that stage.


I spray all of this on both sides with a liberal amount of distilled water, then wrap it all up tight in the foil.

Then it's over to the bender. The side is sandwiched between two stainless steel slats which support the sides and help keep the fibers from lifting up around the tight curves. So a slat goes on the form first:
Then the side:
Then the other slat:
Then I top the sandwich off with an industrial heating blanket:


This is plugged into a variable controller which will allow me to increase or decrease the heat. I clamp the sandwich on either end and plug it in. I have a digital candy thermometer that I stick in the sandwich so I can see when it's time to start bending. When it reaches around 250 degrees I start to slowly press the waist screw down into the waist. The extra clamps are to make sure that the areas that is going to form the cutaway gets plenty of heat.

The temp continues to rise to 300 degrees as I slowly slide the lower bout caul down:

Then the upper bout caul, which can't go very far because of the cutaway ram:


This is the tightest bend, however, so I want good pressure here. Then I very slowly start to screw the cutaway caul down into the form. I take a good ten minutes to go the three inches or so, but this is the critical juncture. I listen very closely for any sounds of cracking, and am grateful that I don't hear any.

When everything is in place I unplug the heating blanket and let this cool overnight.


-------------------------

The next morning I undo everything and open up the foil to see how it all turned out. One of the pieces of binding has cracked a little bit, but nothing major. It is still usable but the other two pieces come out perfect. Good to hedge your bets.

There's a decent amount of fibers that have lifted up around the cutaway curves, but this can often be solved by using superglue and pressing the fibers back down. After sanding everything usually looks good. However, there is a lot of springback, and when I try to press the side into the cutaway form, I find that some of the fibers haven't just lift up--there is actually some cracking that has occurred.

Crap.

Here are some pictures. I'm flexing it some so you can see the cracks:

The cracks have occurred both across the grain and with the grain. There's no amount of superglue that's going to fix this. I'm going to have to get some new sides.

This is the first thing that has gone wrong with this build, so I shouldn't be too discouraged, but I decide to quit for the day. Not a good idea to be shaping braces when I'm a little frustrated. Frustration can only lead to more mistakes, so I just walk away.

I need some time to think about what might have gone wrong in the process, and what my next steps are going to be. Is it the zebrawood? Some woods are more brittle than others and don't bend as well. If this is the case with zebrawood, I'm better off paying LMI to bend the cutaway. Then it doesn't matter how many sides break, I'm only paying for one more set.

But if it's something I can fix myself, that's better. I'm saving money not paying someone else to do my work, and then I can truly say I did it all myself, which is the point, after all.
The thing that gets me is that the cracking didn't occur in the tightest bend--what I call the horn of the cutaway. It occurred in what I intentionally designed to be a more gentle bend than on the last guitar. This area was tighter on the last guitar, and that caused some problem--more with the binding than the side, but I still wanted to avoid this. Why would the tight bend hold and the gentler bend crack?

Since I can only order sides in sets, I have no other use for the other side, so I decide to bend the cutaway on this one. If it too breaks or cracks, I'll pay LMI to do it; but if it doesn't, then I'll just have to pay for another set of sides.

There are two things that I think about that need to be different. First, it didn't look to me like the cutaway caul was positioned so that it mated perfectly with the form. The ram swings up and down so it can be out of the way (or even removed) when bending non-cutaway sides, so I press the caul into the form and mark on the sides of the machine the exact position.

As I look back on it, however, I think the problem was not enough heat. I started dialing down the heat when it hit 300, whereas before I let it get up to 350 degrees. I really don't know why I was more cautious this time. It could be that because the zebrawood is overall lighter in color than rosewood I was afraid of scorching the wood and not being able to sand it out. Or maybe I was just chicken.

This time, however, I crank the heat up. I don't start bending the waist until it hits 300 degrees, and I let it get up to 350 before working on the upper bout and cutaway. I actually work a little quicker in pressing the cutaway caul. I hear no cracking, and it's a good, tight fit with the form. Unplug it and leave it alone for the night.

This morning I unwrapped it all and guess what? It's perfect. Not even a single raised grain. It fits in the mold with no gaps.


I don't know why I doubted that heat was my friend. So I'll call my friend Chris at LMI and ask him to pick out a nice set of sides that match the back that he sent me, and I'll be back on track. A new set of sides will cost me about $40, but that's the price of a good learning experience--and the cost for doubting myself.