Fine Points of Torsion Boxes
A torsion box is a clever way to create a strong, lightweight, flat surface for shelves or worktables. Here's some info on materials and assembly. September 23, 2006
Question
I am experimenting with torsion boxes. I cannot see how using no joinery makes the assembly easier and it is emphasized that it is the lamination of the skins that gives the torsion box its strength. Does anyone have guidelines of grid sizes and skin thicknesses? Is there a problem making the grid squares more rectangular that square? Would shallow dadoes (1/16) to assist with alignment during assembly compromise the strength of the grid at all?
Forum Responses
(Cabinetmaking Forum)
From contributor K:
I have built a number of torsion boxes using thin lumber, MDF and paper honeycomb cores and plywood and MDF skins. I don't know of any printed guidelines for scantlings. I think the box will be most efficient if the grid openings are square. Ian Kirby wrote an article in Fine Woodworking on the subject back around issue #50 or thereabouts. I don't feel in general that you need to overdo it on the internal framework or the skins either, but I don't know how to quantify it beyond gut feeling and common sense.
My first foray used 5/16" x 3 1/2" basswood strips about 3" o.c. with 1/4" luan skins and p-lam over that to make a 4'x8' assembly table that I have used for 20 years. Light and strong enough, though I did crack the surface once with an errant hammer blow. I have made others that size with 1/2" skins and 1/2" cores on 5-6" centers; heavy but bombproof. For a lumber or sheetgoods core, I found the easiest method was to cut strips the length and width of the box and slot them together with halflap joints. A vacuum press at low pressure is good to put it all together, although clamps and cauls also work. The hardest part is finding a truly flat large surface to generate a flat assembly.
Most of my recent torsion boxes have used paper honeycomb available from Darryl Keil at Vacuum Pressing Solutions. He has a lot of experience with this technique and will respond to questions on the forum at his website.
I made one desktop about 3'x5' with 3/4" honeycomb and 3/8" skins, and a library desk with 3 sections about 26" x 72" with 3/4" core and 1/2" skins, plenty stiff and strong. If you look at the inside of a hollow core door you will see it's mostly air. You just need enough thickness and glue surface to let the skins do the work.
From contributor R:
"MORE Woodworkers' Essential Facts, Formulas & Short-Cuts: Hundreds of All New, No-Math Rules of Thumb Help You Figure it Out" (Paperback) by Ken Horner is supposed to have tech info on torsion boxes.
I think Ian Kirby who first wrote about torsion boxes many years ago in FWW had the simplest way to join the interior frame of a torsion box. Simply use a stapler to bridge the mating pieces. You only have to hold it together long enough to apply the skin. Making halflap joints adds extra time to the box.
From the original questioner:
Built a dado sled this week and set out making another torsion box with half laps. It was much faster to assemble and much easier to get all the pieces in the same plane. I used shallow dadoes only in the outer perimeter but am considering using half laps here too. The half laps drastically reduce the number of pieces you are handling and you can gang cut the joints.
From the original questioner:
More experimentation: Half laps are 100 percent the way to go. Tried dadoes in the perimeter but switched these to half laps too. Gang cutting the dadoes speeds production. Tried a 2 inch grid with 1/8 inch masonite tops and was quite pleased. Tried 1/2 MDF tops and could not believe the weight. Next experiment will be a 1/4 MDF grid of 4 inches and a 1/4 MDF top and bottom or possibly 1/4 plywood. I really cannot see using 1/2 inch or greater sheet material unless the surface is meant to be pounded on or you are attaching casters through the base. Right now we have many benches and assembly tables that must be levelled and flattened in position with their levellers but if they are moved they are no longer flat. I want to see if making a torsion box shelf on the bottom will keep the bench from racking even if it is moved. I really could care less if our benches are always perfectly level, but flat, I do care about.
From contributor E:
I've done them with rectangular and square grids and can't imagine it would make a difference. The torsion box is based on aircraft technology, so somewhere in the aeronautics engineering field would be technical info on torsion boxes. Use the butt joint method and crown staple to assemble your framework. Dados and half laps only take more time and do nothing to speed alignment up. Been following Kirby's original technique for over twenty years with perfect results. Try and make a curved torsion box with all those half lap joints. With the butt joint method I'm done before you even get started.
From contributor R:
Just got a copy of Ken Horner's "More Essential Facts, Formulas and Short Cuts." Most informative reading on torsion boxes since Kirby's original article in FWW. Great calculations on strength in relation to grid sizing and skin. Also another in favor of staples to hold the frames together.
