I have built quite a few TB's so the process and build details are not a problem but now I want to build some more that will be long shelves carrying quite a bit of weight and I wonder if anyone knows of any way to calculate the depth if I specify the materials to be used so I don't over build them. To give you an idea of a past project I built a 3.6 metre TB that is unsupported except at the ends, it is loaded with 6 sewing machines and has never bent just deflected slightly. That one had 6mm MDF top and bottom skins with a 3mm MDF grid 50mm deep and it is very light as TB's should be. Are there any engineers here or can a specification even be calculated for a TB built in MDF? I have done the search but can't find anything that I might use and even if I did I would not be able to do the maths anyway so any help would be appreciated. Thanks.

There's a whole chapter on Torsion Boxes including every calculation you could possibly ever want to do in "More Woodworkers' Essential Facts, Formulas & Short-Cuts" by Ken Horner (ISBN 1892836211). It's an amazing reference book, as is the first book in the series without "More" in its title (ISBN 1892836157).

Hope this helps.

From 1970 to 1974, I was the star salesman for Craft-Bilt in Philadelphia.  They produced SIPs (Structural Insulated Panels) before the term was invented.

It consisted of a pre-painted aluminum skin (about 0.030” thick), a resin impregnated honeycomb core, and another aluminum skin.  These were bonded under pressure with an adhesive.  The result was a panel that was rated for 50 pounds per lineal foot.  The panels were 3 feet wide.  It would easily support my weight.

They sell resin impregnated honeycomb panels.  I have not seen them being offered to retail.  You would have to make some contacts.  If you are able to apply uniform pressure across the entire panel, you can build some incredibly light and strong shelves. 

I can assure you that no one from Craft-Bilt had any engineering background.  They worked trial and error. 

Note:  The company went out of business shortly after I left.  (They should have tried harder to keep me.)

Thanks for the replies, I will have a look for the book and see what that tells me. I have found there is a general misunderstanding of how to build TB's as a lot of people overspec the size of materials and what should be a light stiff structure finishes up very heavy with no benefit to the stiffness of the finished TB.

There does not appear to be any copies of the book available in Oz, Amazon want $53 + plus freight which will work out to the thick end of $100 AUD and I did find a copy for $24 USD but can't complete the transaction online because the site won't let me and god only knows what the freight costs will be if they have a copy. The link goes to the site which won't allow me to order.

https://www.alibris.com/Woodworkers-Essential-Facts-Formulas-Short-Cuts-Figure-It-Out-with-or-Without-Math-Ken-Horner/book/29753122

The link above is the wrong one for the edition with the chapter on Torsion Boxes.

https://www.fishpond.com.au/Books/MORE-Woodworkers-Essential-Facts-Formulas-and-Short-Cuts-Ken-Horner/9781892836212

I've looked at buying in the UK and shipping to you but it's heading towards the cost that Fishpond is offering.

It's not something I can really post under fair use as the chapter is 22 pages long!

Thanks for the offer Gary it is much appreciated. I just gave the problem to my local library and they think there is a copy in another library and if so I can borrow that.

Great news!

You can 'borrow' this book for free as an online version from the web archive:

Great find @Bernmc - I signed up for a free account and was able to access the whole book! All the calculations you should ever need @Mini Me 

It sounds as thought you have a great method of building torsion boxes.  Please can you share some details, pictures, drawings etc.  I'd be very interested to know the secrets.
Regards
Bob

I think the biggest difference in deflection is made by the height of the ribs not the thickness of material. Similar to wood I joists they dont really increase the thickness of the osb web for longer spans they do increase the height of the joist and the top and bottom flange to increase span/load capability.  The larger they are the thinner the material can be. At some point glue surface area is going to be an issue. I have made some similar stuff from 5mm ish ply and is cut on co2 laser with hundreds of tabs and mortises and can say that even though the material is very flimsy on its own once assembled it becomes amazingly strong and stiff and light.  If you plan to use thin material I would mill a groove dado into top and bottom skins to increase glue surface unless you have access to a co2 laser and can tab and mortise everything. I would stay away from mdf for this project that stuff is pretty flexible and sucks up glue so I would worry about bad glue joints also. That also brings up glue type some glues are more pliable when fully dry and "might" allow some creep. Notice I said "might" Im no glue scientist but everything starts to come into question when you are trying to maximize results while minimizing weight and material.

I think the biggest difference in deflection is made by the height of the ribs not the thickness of material. Similar to wood I joists they dont really increase the thickness of the osb web for longer spans they do increase the height of the joist and the top and bottom flange to increase span/load capability.  The larger they are the thinner the material can be. At some point glue surface area is going to be an issue. I have made some similar stuff from 5mm ish ply and is cut on co2 laser with hundreds of tabs and mortises and can say that even though the material is very flimsy on its own once assembled it becomes amazingly strong and stiff and light.  If you plan to use thin material I would mill a groove dado into top and bottom skins to increase glue surface unless you have access to a co2 laser and can tab and mortise everything. I would stay away from mdf for this project that stuff is pretty flexible and sucks up glue so I would worry about bad glue joints also. That also brings up glue type some glues are more pliable when fully dry and "might" allow some creep. Notice I said "might" Im no glue scientist but everything starts to come into question when you are trying to maximize results while minimizing weight and material.

You are correct, the little technical detail I know is that the deeper the internal ribs are the stiffer the TB becomes.

I have found MDF is an excellent material as it generally does not twist and warp and glues well with PVA. Putting grooves into the skins would make assembly a bit of a challenge but I think it should add stiffness if it was possible to assemble it before the glue went off. Smaller TB's might make it possible to assemble fairly easily but bigger ones would be a real problem. Epoxy glue might be better because of the longer open time but I have not tried it.