For my next (and, unfortunately, final) piece while here at College of the Redwoods, I wanted to make a small-ish wall cabinet using both veneer and glass. I had a rough idea of the size I wanted to make it, but I first needed to figure out what sort of curve (either concave or convex) the front of the cabinet would have. Shown here is a plan view drawing of the door, with lines for the glass panes and mullions drawn in.
Once I had a good idea of the way the door was going to curve, I was able to make a full-scale mockup. This would (hopefully) tell me two things: 1) How the various dimensions relate to the overall volume and scale of the piece (Taller? Wider?); and 2) How the curve of the door relates to the rest of the cabinet (Should it flatten more in the middle, or should it curve more?)
Once overall dimensions have been firmed up, full-scale templates are made from the mockup. Shown in the foreground are the three templates used to make the bending form (shown behind): one template that is an exact copy of the curve from the mockup, one template with an 1/8" more curve on one side to account for the springback of the laminates (more on that later), and the final, "master" template, which was made by mirroring the springback template across the its centerline.
Closeup shot of the old-growth plank of teak I bought/begged (mostly begged) from my friend Andrew. The plan is to slice it up to make the door frame and exterior veneers. This is the sort of board that you go to sleep thinking about, and wake up excited about...
One thing I have learned here at school is that when making a cabinet, you always build the door first, then the case. The idea is that it is easier to correct for any errors in the door by fixing the case, rather than the other way around. Although it seems counterintuitive, constructing a piece this way actually gives you much more control over the final product.
Anyway, because the door is the first thing to get built, the laminates of the curved stiles of the door are the first parts to get sawn out. In this case, one small block will be resawn into six thin strips, then glued back together over a bending form to make one curved, 5/8"-thick part.
The laminates being glued up over the form. I used urea-formaldehyde glue, which is apparently very stiff, and and has very little "creep" over time. One down, two to go...