Wednesday, November 30, 2011

Twist Table

This was a project I did in conjunction with Al Pierce at Taubman College. It's a dining table made from laminated bending plywood. Here's the computer rendering:

Design / Joinery

This project uses six identical pieces in radial symmetry. The surface is double-curvature, although it is meant to be “gentle” (read semi-sorta-kinda-maybe-sane).

The materials are bending plywood, white oak veneer and glass.

The parts are held together with through-mortise and tenon joints. The edges which mate are compound curves. The tenon and its corresponding mortise are curved as well as tapered. This joinery would be very difficult to fabricate without the use of CNC tools.

The goal was to try to use a single fixture to support all operations – laminating, trimming, and mortising.

Materials / Layout

Fixture: (2) 4’x8’x3/4” Standard MDF
Core: (3) 4’x8’x1/8” Bending Birch Plywood (Column)
Veneer: (3) Flexwood Wood-On-Wood, White Oak, Plain-sawn
Adhesive: Polyurethane Glue

The fixture was 3-axis routed from MDF, glued-up with Titebond, then 5-axis routed to smooth the form and cut the gasket and vacuum grid.Here are some progress pictures:

The parts having been cut on the 3-axis: 

Al Pierce applying glue and aligning each piece with dowels:

Done laminating the pieces... over the 5-axis router...

Smoothing the top of the form: 

The finished fixture with the vacuum grid cut into the surface: 

Testing / Dry Bending / Vacuum Forming

The first test we tried was vacuum forming a dry piece over the form. The Smash command was used to flatten the double-curved surface to a plane. Readers - don't ever use the Smash command! Worthless piece of shit. I should have know better...

 Here's the dry piece over the form ready to test with vacuum pressure:

 The piece cracked in two places, but mostly on the inside corner. Gaussian curvature analysis shows that this area has negative curvature – it is most highly saddle shaped in that region.

We wet a single piece of bending plywood for 5 minutes then vacuum formed it alone on the fixture. It bent without cracking.

We then attempted three sheets of plywood and two layers of veneer. These were soaked in water for 30 minutes and glued with polyurethane glue.

 Results / Future Work

There was no cracking but lots of puckering in the veneer and quite a bit of spring back relative to the true surface of the form. 

The puckering may be partially related to the use of tape to hold everything together during the vacuuming (the glue isn’t tacky and the parts want to slip relative to one another. 
We could possibly eliminate the tape by securing the work on the form in some other fashion (dowel pins perhaps). Then again may be not… The spring back could possibly be eliminated by using epoxy as the adhesive although the wet material may be an issue. 

The best approach is to change materials. There is a board which bends without spring back when heat is applied.