When I began this project, side bending was the process that most fueled my imagination. Until about 20 years ago, bending the wood for guitar sides had been accomplished the same way for centuries: the wood was moistend with water and carefully bent around a heated pipe until it conformed to the desired shape. Clearly, this proceedure required a great deal of practice to achieve the proper curvature without breaking or scorching the wood. This time-honored method is still used by thousands of quality luthiers around the globe. Indeed, Martin Guitars still uses this method for some of its special-order instruments.
In the '70's, luthiers began experimenting with a number of alternatives designed to ease and speed this tedious, time-consuming process. One of those methods was to laminate very thin sheets of wood together in a mold using the new miracle glues then becoming axailable. This "cold molding" technique suits the demands of large-scale poduction shops where price point is the objective. Some very good guitars are made this way, but the instruments lack the tone and resonance of solid wood, especially when the back or even the top are also made from laminates.
Another breakthrough was the use of heated forms to which the
moistened sides were clamped and held until the wood assumed their
shape. This technique allowed the use of solid wood and could
be accomplished by semi-skilled workers in very short order. In
the '80's Luthier's Mercantile International, http://www.lmii.com/
another very respected merchant of guitamaking supplies, began
selling the Fox Universal Side Bender, a wooden-framed version
of this device in both kit and ready to use versions. It consists
of a box containing regular houshold lightbulbs that supply the
heat and a form that sits on top of the box. This form is shaped
like half of a guitar body.The wood for the sides, which has been
soaked in or sprayed with water, is layered between two slats
of stainless steel and slowly wrapped around the 
heated
form using built-in clamps that screw down on top of the piece,
pressing it against the form. The heat is kept on for about 15
minutes and then shut off. the wood remains on the machine until
completely cool. When removed, it retains the shape of the form.
This machine is beautifully functional and reasonably priced (
about $500) for the luthier who builds and sells a good number
guitars a year.
But for those who are just learning the craft, or do not intend
to produce many guitars, that's a lot of money. Naturally, I began
working on a simpler, more cost effective design for my uke. Using
some 3/4" MDF that I had left over from a bookcase project,
I used my template half for the body shape and traced it onto
the wood. I cut two halves and joined them with MDF cross pieces
for a width of about 4''. I searched all over town for small sheets
of stainless steel, but to no avail. I considered alumium, which
might have worked, and galvanized tin, but I began thinking about
even heat distribution and remembered that 
copper
was revered as the ultimate metal for cooking vessels and brewing
equipment for its great heat transfer properties. I remembered
seeing some 4" wide strips of copper at a local Hobby Lobby
store and bought two.
Instead of sandwiching the wood between pieces of metal, I decided to try mounting the copper on the form and pressing the wood down onto the hot metal with clamps similar to the Fox machine. I built a box equipped with two 150 watt halogen bulbs as my heating unit and lined both the box and the form with aluminum foil to avoid scorching the wood inside. Here's the result.
The clamps are just 1/4" eye bolts with washers and wing nuts. The cauls are hardwood dowling pieces cut off by customers in the moulding section of Home Depot. If they are less than a foot long, they will give them to you!
The form is removable, exposing the lightbulb fixtures below.
The side pieces had previously been cut to shape. One side was straight, as the top plate would be flat, but the side of the piece on which the back plate will rest is angled so that the lower bout of the uke will be deeper than the upper bout. However, rather than a continuous ramp, it must level back out before reaching the "top" or "neck" end. This prevents the back plate from "dishing" where the sides meet. Though that sounds confusing, it should become apparent when you test fit the back. In my excitment to start bending, I neglected to photograph the sides before proceeding. Here's a diagram:
I preheated the form for about
10 minutes and then sprayed the side piece with distlled water
(both sides) until it was thoroughly wet. Our local water supply
is very high in minerals, and tap water might stain the wood when
the water evaporates, hence distilled water.
Raising the cauls all the way. I slid the piece onto the form and let it start to heat. I had penciled registration lies on the wood and the form so that could place the "waist" curve exactly where I wanted it. Remenber to allow for movement of the piece as it is clamped down.
As the wood dried where it touched the form, I re-wet with my spray bottle. When fragrant steam began to rise, I began to tighten the wing nuts on the waist (middle) caul a few turns at a time. I could actually feel the wood relax and bend as I proceeded. Checking my registration marks now and then, I tightend a few turns, waited 30 seconds, and tightend again. The waist bend took about 5 minutes total.
I re-wetted the ends of the piece and began tightening those cauls alternately. On these, I would tighten, then slide the caul down the curve an inch or so, wait, and begin again.
Here you can see how I proceeded.
"Easy Does It" is the way to go here. Listen carefully
for any signs of protest from the wood. Push down gently on the
end of the piece with your fingers and feel how it "gives"
as the heat and moisture combine to make it pliable. After doing
this, I began to feel that I could better understand how a side
could be bent using the old "hot pipe" method
When the wood was snug against
the form all around, I wet it again and let the heating continue
for another 10 minutes. Using a meat thermometer, I stuck the
probe between the center caul and the copper and took a reading.
With about 10 minutes preheating, 15 minutes spent gettiing the
wood tightened down over the form and another 10 "cooking
the wood", the thermomter read 152 degrees, and I turned
off the heat.
Nowhere in any of the various articles that I had read about the Fox Bender had there been a discussion of temperatures. I had no idea what the optimum bending temp was for 3/32" mahogany, so I just went by sight and smell. At no time had I smelled buring wood, so I assumed that my temperature had been adequate to bend the wood, but not too high.
I allowed the wood to cool on the form for several hours and then released the cauls. There was considerable "springback" on both the waist and the ends. I decided to clamp the piece back on the machine, re-wet and "cook" it again. I did this two times over the course of the day, allowing it to cool completely between each cycle. When I released the wood this time, the bend was set very well, with minimal springback.
I can't tell you how satisfied I am with my first attempt at side bending. I don't mind revealing that I had serious doubts that my homemade bender would work at all. I estimate that total cost of the device to be about $25, including the wood, which I had on hand. Though I would need to build a larger unit for a full-sized guitar, I have no doubt that I could bend the sides for any size guitar with a similar machine. I would imagine it to be more time-consuming to bend wood this way, compared to LMI's Fox Bender, but it can't be too much more, and at a 400% savings, I'll be doing it this way, I believe.
I want to offer a couple of alternatives and variations for this process. LMI also offers a silicone heating blanket for side bending that could be used as a heat source for my mold, rather than the light bulb method. Kathy Matsushita uses a similar set-up for her very fine instruments.
Also, one might consider using a heat gun as a heating source, either under the mold, or playing it across the top of the wood as you tighten the cauls. I have not heard of anyone using this method, but it seems reasonable to me.
The bending machine described in the above step is a great way to bend unfigured "semi harwoods" like mahogany and plain koa, but I have since discovered that bending denser hardwoods like rosewood and any wood with curl, flame or quilt to it is much more difficult with this apparatus alone. Since going pro, I have been using forms just like the one described above, but with a silicone blanket as my heat source, These are available through several of the lutherie supply houses on my LINKS page (I got mine from Ukulele Supply of Hawaii) and they make bending almost any wood much easier. I wet the wood thoroughly as described above and sandwich it between two metal strips ( I use hardware store tin. but any thin, flexible metal will do). I place this sandwich on the bending form and lay the heating blanket on top it. I plug in the blanket and, when I see steam rising from the wood, I tighten down the cauls the same way as described above. The blanket gets very hot, very fast, so you must unplug it periodically to prevent scorching. NEVER leave it unattended while plugged in - I have a once-gorgeous piece of flamed Oregon Myrtle that is charred almost beyond recognition to help me remember!
The metal strips perform two functions: they help distribute heat evenly over the wood, and they provide backing reinforcement on both sides of the wood that prevents cracking on the outside of the sharper bends - especially helpful with highly figured wood.
With this heating method, you do not need the halogen light base for the form - making it simpler still.
