1. I worked out the general form with sketching and then modeled it Rhino3D. From that I made the parts for the male/female press-mold with a 1/2" allowance between the sections for material.
2. Glued up blocks of MDF to mill. Before milling the blocks we cut pockets out of a sheet of foam to hold the blocks in the correct place on the bed during milling. Milled out all the main parts plus the triangles which support the sides and back of the female portion of the mold, as well as the sides of the middle, wedge shaped keystone box.
3. The layup is 2 layers of 1/4" construction grade plywood, and then a top and bottom layer of paperback veneer which I apply after the plywood comes out of the press mold. Ideally the layup would be all veneer, but I don't have thousands of dollars in disposable income to spend on projects so I use plywood.
In order to get the bend in both directions I had to piece together the layup such that I had the grain going in different ways. I stitched together pieces with waxed thread (first tried dental floss) such that the seems would not directly overlap and create a weak point.
I soak the wood for about 4 to 5 hours before I lay it up in the mold and clamp it down without any glue. I let that sit up over night (or a couple nights) and then unclamp and wait for the wood to dry out enough to glue. I have been told it would help to put some paper towel or cheese cloth in-between the layers of plywood to aid the drying out. I use a urea formaldehyde based glue that comes in a powder that you mix with water. This is important for two reasons, first because it is a water based glue it will suck the remaining moisture out of the wood which I had to soak so drastically in order to achieve compound curves. Second, the bi-product of the chemical reaction that occurs when this glue cures is ammonia, which softens the lignins in the wood, making them more malleable.
Once the plywood is dry enough for gluing I coat each joining surface with a thickened coat of the surupy adhesive, put them back together, and clamp it up in the mold. I put a layer of dropcloth (polypropylene) between the MDF and plywood on both sides to avoid gluing the mold shut, I also wax the hell out of the MDF.
4. I get the form out of the mold and sketch the curves I want to make on one half of the chair, I cut those lines (jigsaw, dremel with wood cut-off wheel) and then use a piece of paper to mirror the lines to the other side. It gets a rough sanding then I apply the paperback veneer, which is not ideal and you can see where it wanted to fold in on itself and I had to cut it with a razor to make it lie flat on the surface.
With strips of course emery cloth I hit the edge shoeshine-style to round it over a little bit, and then oiled it down.
5. The steel base also has digital origins. I first molded the curves, then projected them flat to the c-plane and printed them out on a 3' plotter. From there it was a matter of bending the 1/2" hot rolled steel to match the working drawings, and then MIG welding it together. The rocker is wedged (the front is wider than the back) I think it makes it stronger... The two pieces that make the rocker were rolled. The metal got sanded down with course, medium, fine emery cloth, 0000 steel wool and then sealed with a rustoleum clear enamel (Not many metal pics because I wanted to keep this focused on wood)
6. To get the rubber shock mounts I made mounds in oil clay on the frame and pressed the chair into them, then made a mold of those parts and cast an 80A durometer into that mold, suspending carriage bolts above the mold such that they would be aligned with holes in the frame. The rubber mounts get laminated to the plywood shell, and then secured to the metal frame with nuts.
Thats my Grandpa sittin the chair, he used to be a salesman for Herman Miller... The last image is an 8x10 contact print of my thesis exhibition, taken by my father. I ended up with three bent plywood iterations from the same mold, the last being the rocker with wings.