Torsion box shelf above garage door

[Cody Jensen]

Does anyone know how thick of skins and webs would be needed for an 96" x ~20"-24" deep shelf supported on only the ends? I don't have an exact weight, but lets say around 300 lbs distributed evenly over the width. I haven't used sagulator much before, but it seems a 1.5" thick plywood shelf is sufficient, so I would think 2-2 1/2" total thickness might work. I have never built one before, but I have seen quite a few build videos and images on here and other sites. The only thing I am not sure of, is does the "frame" of the webs (the boards on the perimeter) need to be thicker than the webs? Thanks.

[David T gray]

why do u need it to be a torsion box seems like a lot of work for a shelf i would just take some plywood and attach some angle steel to it should be stronger that way and wayyyyyy less work.

[Mike Henderson]

The strength of a torsion box is mostly in how thick it is (the depth of the webs). Torsion boxes are a form of an I-beam (https://en.wikipedia.org/wiki/I-beam). Look up what makes an I-beam strong and you'll get a lot of good information on a torsion box.

The biggest disadvantage of skins that are too thin is that they bend between the webs when you put a big load on them. So the skin needs to be strong enough to support the load between the webs.

Mike

[Alan Schaffter]

Are you thinking about building something like this?

This 8' X 18" X 3" torsion box was built as an example of a lightweight torsion box for a two-part article I wrote about my adjustable height, torsion box, assembly table. It was published in American Woodworker several years ago. The skins and longitudinal web pieces were made with 1/8" tempered hardboard. The lateral web pieces were made with 3/16" pegboard (I ran out of hardboard.)

To ensure good fit and alignment and to make it easier to assemble, I made a big jig like a box joint jig, a long fence with indexing peg, to cut perfectly space halved joints in the web pieces (shear web). Spacing was approx. 6". This was much easier than cutting individual pieces like done by David Marks (and later copied by Mark Spagnulo). Everything was glued together with yellow WW glue.

In the photo it is supporting almost 350# of bricks at center span yet it is deflecting less than 1/2"!!! I don't know how much more it would have supported because I did not to load it to failure. Once the article was done, I used the torsion box for a scaffold for awhile until I got it really wet cleaning my porch ceiling- hardboard does not like water!!! It was extremely light- easily lifted and held aloft with one hand. As you can see, you don't need perimeter boards unless you want to dress it up, mount hinges to it, etc.

Pile of bricks is two deep- almost 350#:

http://www.ncwoodworker.net/pp/data/1421/medium/P1080028.JPG

Getting ready to install second skin (being assembled on my original adjustable height torsion box assembly table):

http://www.ncwoodworker.net/pp/data/1421/medium/P1040010.JPG

Assembly table article photo:

http://www.ncwoodworker.net/pp/data/1421/medium/145_AssemTable_spread1.jpg

Cutaway shot:

http://www.ncwoodworker.net/pp/data/1421/medium/Cutaway_Test-1.jpg

As a guest you may still be able to read my NCWW thread ("Torsion boxes sure are strong") that discusses torsion box design and my demonstrator and assembly table torsion boxes.

[Steve Peterson]

What are you planning to put on the shelf? If you just want to store boxes, then 1 or 2 straps from the ceiling would add a tremendous amount of strength. If you want to store something like a kayak, then straps would get in the way but it also doesn't need a lot of strength.

Steve

[Cody Jensen]

So the torsion box idea is not necessary, aluminum or steel angle could work. However, I am just getting started into doing woodworking (at least since high school shop), so I kind of like doing shop projects to just get a better feel for going through the motions. I was hoping to avoid using ceiling supports, mostly just for aesthetics though. A part of me wants to go through the act of building a torsion box just so I know how much of a pain it is if I ever do it again, and another part thinks I should just do whatever is fastest and will work. If I plan on cutting the webs halfway so they interlock, should I expect this to be a 20 hr project? Also, is this pointless without a jointer (I have a TS)?

[John Blazy]

Awesome example of Torsion Box construction Alan!! RIT has a program all about torsion box construction if you're a WW / furniture design major.

Back in the early ninety's as a finish/trim carpenter, I had to take my heavy delta 10" chopsaw (first ones - before Hitachi made the slide compound) to the jobsite every day, so I decided to make a torsion box extension bed / fence system with a flat tape rule built in, zeroed to the blade.

So I used 1/4" luaun for the skins, and 1" blue Styrofoam for the 2-1/2" high webs and made the 96" bed from it, glued with PL Premium polyurethane. Still works in my shop to this day, and you can set it on horses 7 ft apart without sag. I used ash for the outer webs for abuse resistance, but its super lightweight - especially when needed to be carried in and out of houses from my truck.


So Cody - you don't necessarily need the cross webs like alans box if you use 1" Styrofoam - it is twist resistant due to its thickness, but you might want to glue in a few bridges from plywood. Torsion Box construction is really based on keeping the skins parallel without shearing (the job of the long webs - based on high adhesion of the glue), and the cross webs are usually there to prevent the long webs from twisting/buckling under weight, so glue some in based on your judgement.

My saw box has no cross webs and is holding up fine, but if I did it again, it doesn't take hardly any time to glue in cross bridges.

The other advantage to uncut full length webs (even if just Styrofoam) is that they are not cut halfway through to allow the cross lap joint that most people use in making a grid style torsion box webbing. Cutting the long webs reduce their flexural strength a little (although 95% of the effectiveness is from the face skin adhesion).

Making a torsion box is rather fun, due to the cool engineering physics involved, especially the challenge to make it ultra lightweight - so go for it.

You don't need a jointer to do it. Just follow the factory edges of your plywood and whatever you decide for the webbing material. And use PL polyurethane - the greatest all around adhesive in the world - superior for foam to wood.

I lived without a jointer for many years til I could afford one - you'd be amazed at how you can straighten bowed lumber or hand sawn edge on a tablesaw by multiple cuts flipping each edge against the fence, taking off a little at a time before hitting your target dimension, because the fence on a Tsaw kind of functions like the flat face of a jointer. The rough edge of the sawn edge is perfect for PL glue to bond to. Get good with a jack plane, and you may not care to get a jointer for a while.

[Alan Schaffter]

Thanks.

Cody- I just used a tablesaw and rip fence as well. Making the notches using a 2 X 4 mounted to my miter gauge and an indexing peg- no measuring at all! I cut all the laterals and longitudinals for my lightweight table (scaffold) and my other tables (1/2" mdf) in probably 10 - 15 min. tops! You need a dado blade set if your material is other than 1/8" thick, however.

Cutting halved joints (notches) in 1/2" MDF webs (quick, easy, perfect spacing!):

http://www.ncwoodworker.net/pp/data/1421/medium/P10100792.JPG

Note about the shear web- the top halves of the web pieces are under compression, while the bottom halves are under tension, both decreasing to zero stress at the halfway point (neutral axis). The flanges on an I-beam and the skins on a torsion box take care of most of the high stress, however, but you should still make your halved joint cuts in the top edges of the longitudinals. The web is also under compression top to bottom to keep the skins equi-distant.

http://www.ncwoodworker.net/pp/data/1421/medium/Beam_FigA.jpg

One question that always comes up- Why hardboard or MDF instead of solid wood and/or plywood? Both hardboard and MDF have uniform thickness and density, are flat, and are very stable, dimensionally; solid wood and especially plywood are not!!!! You'll notice, however, to protect my assembly table tops I covered them with high pressure laminate.

Styrofoam (brand name) is actually not a good choice- too soft and flexible- to replace the shear web, but (rigid) urethane foam is! It can be purchased in billets and is easy to cut and shape with saw or hotwire saw. I constructed a two place, experimental class aircraft (Rutan design) using foam and fiberglass- the structure (fuselage, frames, and even wings) was actually hand-layed up fiberglass-foam-fiberglass sandwich and very strong! The only issue with using foam only is how to get a good bond between the skins and foam. A good bond is very important and not a lot of adhesives bond well with foams. The urethane foam used for homebuilt aircraft has open cells on the surface which is typically filled with a paste the consistency of peanut butter made from epoxy and filler (micro-balloons). This is one of the potential limitation of using very thin web material- not a lot of glue surface at the edges.

[John Blazy]

Styrofoam (brand name) is actually not a good choice- too soft and flexible- to replace the shear web, but (rigid) urethane foam is!


I agree for obvious reasons, but it sure seems to work for light loads. I suggested Styrofoam because it can be bought cheap at the big box stores, and it worked for me. Would have loved to have had urethane foam panels back in 1989 when I made my chopsaw box.

A good bond is very important and not a lot of adhesives bond well with foams. The urethane foam used for homebuilt aircraft has open cells on the surface which is typically filled with a paste the consistency of peanut butter made from epoxy and filler (micro-balloons). This is one of the potential limitation of using very thin web material- not a lot of glue surface at the edges.

The bummer about foams is that they have bad cohesion. Adhesion can be great with the right glue, but the foam can crumble off - destroying the bond due to CO-hesion. Fillet it well and your golden :).


It would be a fun test to challenge all the torsion box nerds out there like me and you Alan, to make a torsion box with the highest strength to weight ratio. Im picturing an aluminum skin over plywood ribs.

[Myk Rian]

After replacing all the doors in the house, I had a bunch of hollow core doors. They make great shelving. I have them in the garage and shed.

[Bill Space]

Alan said:

Both hardboard and MDF have uniform thickness and density, are flat, and are very stable, dimensionally; solid wood and especially plywood are not!!!!

This comment caught me by surprise. I always thought plywood with its 90 degree oriented layers was very dementionally stable.

Am I wrong again?

[Alan Schaffter]

It would be a fun test to challenge all the torsion box nerds out there like me and you Alan, to make a torsion box with the highest strength to weight ratio. Im picturing an aluminum skin over plywood ribs.

How about aluminum over divincel or nomex honeycomb? That is what is used in many commercial aircraft partitions.

The bulkheads I made for my experimental airplane were made from 1/4" thick small cell urethane foam (Clark Foam) covered with 2 - 3 laminations of fairly lightweight bi-directional e-glass (pre-carbon fiber days) and epoxy. To save weight, once the glass was fully wetted out it was aggressively squeegied to remove all excess resin.

I saved a 10" by 30" piece of it. When skeptics would ask if I was "really going to fly a plane made from foam" I set the bulkhead piece between two cinderblocks and then stood on and gently bounced up and down on it. It would flex a little but it wouldn't break- pretty high strength to weight ratio!

[Alan Schaffter]

Alan said:

Both hardboard and MDF have uniform thickness and density, are flat, and are very stable, dimensionally; solid wood and especially plywood are not!!!!

This comment caught me by surprise. I always thought plywood with its 90 degree oriented layers was very dementionally stable.

Am I wrong again?

Unlike high-grade marine ply, most plywood has some voids, lamination overlaps, and will warp like a bandit with little to no moisture. Just go to your local big box store and look at the ply, especially the thin stuff you might consider for the skins and web of a lightweight torsion box- try to find a flat sheet! If you find one and take it home, see how long before it warps.

[Bill Space]

Honestly that does not seem to be my overall experience with plywood, and I wonder if others feel like you do? Of course with any product there is the occasional bad apple...not arguing that. But overall...not my experience in general.

[Alan Schaffter]

Honestly that does not seem to be my overall experience with plywood, and I wonder if others feel like you do? Of course with any product there is the occasional bad apple...not arguing that. But overall...not my experience in general.

You must put it in context of a building a torsion box. Generally, people build torsion boxes because they want a flat, lightweight, yet strong structure, that will retain these characteristics and is easy to build. There is no doubt that ply can be used to build a torsion box, but I've made enough of them to know that ply is a less than ideal material.

[John Blazy]

How about aluminum over divincel or nomex honeycomb? That is what is used in many commercial aircraft partitions.

The bulkheads I made for my experimental airplane were made from 1/4" thick small cell urethane foam (Clark Foam) covered with 2 - 3 laminations of fairly lightweight bi-directional e-glass (pre-carbon fiber days) and epoxy. To save weight, once the glass was fully wetted out it was aggressively squeegied to remove all excess resin.

I saved a 10" by 30" piece of it. When skeptics would ask if I was "really going to fly a plane made from foam" I set the bulkhead piece between two cinderblocks and then stood on and gently bounced up and down on it. It would flex a little but it wouldn't break- pretty high strength to weight ratio!

I always loved high tech composite honeybcomb construction. The honeycomb by itself has little rigidity, which proves that the skin tension/compression is really what makes it. And epoxy is the best in these cases - very little elasticity. I would have no reservations about flying in your plane, Alan!


I totally get what you mean about HDF and MDF vs wood and plywood - uniformity is its strength. However, I would love to see Baltic birch come in 4 x 8 sheets. Pretty uniform stuff. Marine grade Okoume is nice too. The neat thing, is that plywood is a miniaturized version of torsion box construction.