Torsion box... Not flat?? (OSB? MDF? or..?)

[Nil]

So I wanted to make a miter saw stand, and I needed to start with a torsion box that was completely flat. I had some scrap OSB around, so I went with that for a 4 foot practice torsion box. The ribs were 2.25" thick. After I was done it was dead flat.

Now on to the 6-foot one out of MDF. I ripped a sheet of MDF in half for the top & bottom, and then cut lots & lots of 2" strips for the internal ribs. I didn't adjust the fence once while cutting the ribs, so they should all be the exact same thickness. Then I glued & clamped it all up... And it wasn't flat in the end.

I bought more clamps. Repeated the process, but was a little more observant this time. I knew my sheets of MDF that I was starting with weren't dead flat, but I *assumed* that with enough clamps it would pull it flat. It turns out that a 2" by 1/2" thick rib of mdf can bend up to almost 1/4" over a 6 foot span in the 2" thick orientation. And so I ended up with another torsion box that wasn't flat even though I clamped it with almost 30 bar clamps. No my surface I was working on wasn't 100% flat, but I thought that shouldn't matter that much with the design of a torsion box...

I'm considering going back to OSB (I've got plenty of scrap of it, and my first prototype ended up dead flat) and then just gluing a MDF top to it afterwards (to make it nice & smooth). I know I could do plywood as well, but it seems like a waste of an expensive material if OSB will work fine... Thoughts? Has anyone used OSB like this?

I've also considered ripping a bunch of 2.75" strips of OSB, flipping them on edge and gluing them together, and then jointing/planing them down into 2.5" boards that I could use to build a top. Thoughts?

Other options are using solid lumber & planing/jointing it flat and face gluing it together, or using thicker ribs of MDF, or using plywood...

[LarryG]

Your problem is that the 2" rib height is insufficient. Using OSB or plywood is unlikely to help. The ribs need to be taller, so they'll be stiffer.

My 42" x 60" torsion box assembly table is 4" thick; 3" ribs with a 1/2" skin top and bottom. Still flat after approximately five years.

[natausch]

The biggest challenge I can think of would be moisture->swelling, but it should be fine with MDF on top.

[LCHIEN]

the biggest challenge i see is getting a flat surface to build your torsion box on.
If you build it on a table with a 1/2" gradual rise over the center, then the MDF will conform to that, when you clamp the ribs to the MDF they will take the rise, and then clamp the top MDF over the whole thing it will not be flat. You have to make sure the whole thing is flat when you assemble it.

More clamps don't help if its not flat to begin with.

[phi1l]

After studying everything I could find on how to make a torsion box, I lucked out & found 5 50 yr old hollow core doors at a garage sale. Those old doors are stiff enough that no problem from a "box" made by gluing a pair of those together.

But From what I recall:
1)It is important to have a flat reference surface to work off of.
2) The thicker the box the, the stiffer the box.
3) All examples I recall did not use clamps, they were all screwed or nailed (although one guy did use clamps to temporarily hold down grid members to keep the m straight). I guess that makes it it easier to assemble on your flat reference surface.
4) Use a grid size that is nominally twice the thickness of the ribs.

[Nil]

My guess was that the 2" ribs weren't thick enough--I was wondering how thick they needed to be...

I thought because of the design of a torsion box, if the ribs didn't bend, it wouldn't matter whether the reference surface wasn't flat because the bendable top & bottom would conform to the flatness of the ribs...

Any reason not to use OSB?

[LCHIEN]

My guess was that the 2" ribs weren't thick enough--I was wondering how thick they needed to be...

I thought because of the design of a torsion box, if the ribs didn't bend, it wouldn't matter whether the reference surface wasn't flat because the bendable top & bottom would conform to the flatness of the ribs...

Any reason not to use OSB?

I think the rib thickness is the least of your problems, 2" ribs will give you the equivalent stiffness of nearly 3" thick top - very stiff. But build it on a non-flat reference top, it'll never be flat.

You need to invest in a 6-foot level, its stiff enough it won't sag in the middle and can tell you if your top is flat when you start gluing it up.

OSB is OK, the surface can chip and it won't hold nails real well in the edges. I would think MDF would be better.

[Mr__Bill]

I would suggest that you take one of the boxes and level it with floor leveling. If necessary put edges on the box to hold the leveling. Now build the new one on the flat surface or just screw a top on the now flat surface.

I have built several and they never turn out flatter than the surface that they are built on. You can also use sawhorses and jointed 2x4s for the surface and shim everything to flat and then build on that surface.


Bill

[LarryG]

If you build it on a table with a 1/2" gradual rise over the center, then the MDF will conform to that, when you clamp the ribs to the MDF they will take the rise, and then clamp the top MDF over the whole thing it will not be flat.

This is why the ribs need to be taller: so they will be less likely to bend if the reference surface is not perfectly flat -- something that can be very difficult to achieve if one does not already have something like a torsion box on which to build (the classic chicken or egg dilemma).

Consider a 2x4 laid flat vs the same 2x4 stood on edge. Big difference in stiffness. The latter will be far less likely to bend when skins are applied to both sides and drawn down tight with clamps or mechanical fasteners. As phi1l said, I put mine together without using clamps other than a few to hold the various parts in alignment just until I could get them nailed together. If clamps are used, the components being assembled should only be clamped to each other -- NOT to the surface being built upon (unless it's something very stiff, like a torsion box). This help eliminates problems with the assembly surface not being perfectly flat.

But, yes: the surface on which one builds does need to be as flat as possible, mainly to prevent a twist from forming in the torsion box as it goes together. I used the technique Mr Bill describes, of jointing 2x4s and shimming them carefully until they were flat.

[cabinetman]

My guess was that the 2" ribs weren't thick enough--I was wondering how thick they needed to be...

I thought because of the design of a torsion box, if the ribs didn't bend, it wouldn't matter whether the reference surface wasn't flat because the bendable top & bottom would conform to the flatness of the ribs...

Any reason not to use OSB?

Thickness does matter. For assembly tables, or saw extension tables a torsion table clad both top and bottom is unnecessary. Sheet goods fabricators, like cabinetmakers make tables with a grid and a top only.

For the grid, using 3/4" plywood cut @ 5 7/8", a 4'x8' sheet yields 8 lengths (using an 1/8" kerf blade). The grid arrangements are at a maximum of about 16" square or less. Top that with a minimum of 5/8" melamine, makes for a good work/assembly surface. Legs for that table get inserted in the corners and at the middle and center. Adjustable feet can account for an uneven floor.

The surface stays flat and for a 4'x8' table it is strong enough to walk on, without creating any deflection.
.

[phi1l]

Theoretically, if for some reason you can't use taller ribs, you should be able to compensate by increasing the rib density. for example David Marks uses 3" ribs & a grid spacing of about 6". If you can only use 2" ribs the same strength with a 4" grid ... theoretically, anyway...

[pecker]

At the beginning of this article he shows how to establish a flat surface onto which you can then build your box.
http://www.finewoodworking.com/Proje....aspx?id=28855

[JR]

At the beginning of this article he shows how to establish a flat surface onto which you can then build your box.

That's a great article. I really want that assembly table! Of course, what I want more than that is the shop where I can put that assembly table so I can walk all the way around it...

JR

[Nil]

It seems like there are two different ways to make a torsion box... It looks like whether you need to have a dead flat surface to start with is dependent on how you build it.

I can totally see why one would need a dead flat surface ff you did it like in the video. There Mark glued & nailed all the ribs together, with them just resting on the bottom surface, and then attached the top & bottom faces--he fixed the ribs' position relative the bottom surface. There wasn't any clamping to force the bottom surface to conform to the lengthwise ribs.

However, the way I tried making it both times was to apply glue to all the surfaces & then clamp it like crazy (just to the top & bottom faces, not to the assembly surface), all hopefully while the glue was still wet... Theoretically it seems like if the ribs were rigid enough, they'd pull the bottom & top faces flat. Based on how the 2" MDF ribs bent down and the bottom face board didn't bend up, I don't think that I had rigid enough ribs because they weren't tall enough...

[Kristofor]

Theoretically it seems like if the ribs were rigid enough, they'd pull the bottom & top faces flat. Based on how the 2" MDF ribs bent down and the bottom face board didn't bend up, I don't think that I had rigid enough ribs because they weren't tall enough...

Well, I guess I'd agree with the theory, but that's going to be a much harder way to skin your particular cat...

You'd have to work carefully making sure that the top and bottom skin slippage will be aligned when you're done while trying to force things into alignment while doing a pretty large scale glue up... Contrast that with the effort needed to joint and level $6 worth of 2x material which allows you to start with a flat surface and work on one side of the glue-up at a time. There's a reason why this approach is so popular

Now, for full disclosure, we have a clothes sorting/folding table made from my first attempt at a torsion box for a work/assembly table. This was one of my first projects and I didn't quite grasp the importance of a flat surface to work on when building a flat surface so I'm speaking from earned knowledge on this one