Hollow core door for torsion box?

[dlminehart]

Anyone used a hollow core door as the core of a torsion box for a shop mobile cabinet/workbench, sandwiched (glued) between a couple pieces of MDF? I'm thinking that it would be quite easy to do this, compared to cutting and gluing all those cross-pieces of MDF for one of the more standard torsion boxes. But I wonder if it would be strong enough to hold a heavy saw, some drawers of tools, a router, etc. The corners, particularly, would seem to require some kind of reinforcing to hold swivel locking casters.

[BobSch]

The hollow core doors I've seen were filled with loose cardboard tubes. I'm not sure how much support it would provide.

[eccentrictinkerer]

IMHO, a hollow core door would be best suited for building model airplanes. Small ones.

In my experience HC doors are usually made with 1.5" x 1" rails and stiles skinned with 1/8" 3-ply plywood.

They stay pretty flat, but can't tolerate any load when lying horizontal.

One of my handyman client's 4 year old put a toy hammer through one.

[dlminehart]

I realize that a hammer could go through one . . . but this involves a fairly concentrated application of force. A modest swing would put quite a few pounds per square inch of hammer head. My 400 pound saw would be spread over about 28x28 inches of MDF atop the core, or say roughly 800 square inches. So only 2 pounds per square inch . . .

[chopnhack]

I wouldn't risk your safety or your saw on a HC door, better off to build a true torsion box, that way when your cutting some big piece of lumber you won't hear your saw support collapsing !

[lkazista]

Why not upgrade to a solid core door, still less than $40, and nobody's kid can put a hammer through it.

I think that makes complete sense if you sandwich it with plywood.

Lee

[pecker]

I have an old hollow core door in my house, and it is rather heavy and quite flat. So I thought I'd buy one at the BORG, and use it as an assembly table.

The new one isn't even in the same league. It has a bulge about an 1/8" high in one spot, so it's not all that flat. And I accidentally dropped a large C-clamp on it, and the clamp punched a hole right through the skin. It appears to be much more "hollow" than the older ones.

[LarryG]

David, I think people are reading your post and arriving at two different conclusions about what you're asking. I know I did, in that I caught something the second time around that didn't quite register the first time.

If you took a good, straight, hollow-core door and skinned it with MDF on both sides, and placed it atop some sort of leg structure ... that would make a fairly decent general-purpose work surface. It wouldn't be substantial enough to make a good traditional-style woodworker's bench on which you'd do a lot of planing and other hand tool work. It could, however, make a suitable base for benchtop power tools as long as they didn't weigh too much, or had legs directly under the tool. For example, you could mount a bench grinder or disc/belt sander at one end; but you wouldn't want to place that 400 pound saw out in the middle. The weight distribution example you give is only valid for a situation in which those 800 sq in are themselves fully supported; i.e., like a concrete slab on grade. When you have a heavy point load out in the middle of a suspended span like a workbench top, the load has to transfer to the four legs ... which means the top will sag under that heavy weight out in the middle.

If you're talking about using the door as the bottom-most component in the kind of mobile bases that many members here have built -- no. A hollow core door is much too lightly-built and flexible for that. (Given your closing sentence regarding casters, I think this is your intent, but I'm not certain.)

[jackellis]

I have a little different take on this. The skin of the door is only relevant if you plan to set something directly on it, but you're actually going to reinforce the skin with MDF. Even if the inside of the door is filled with cardboard tubes, they provide exactly the kind of stiffening a torsion box needs, provided there are enough of them. Stringers in the wing of my airplane are only about 25 thousandths of an inch thick, but they help support a load of almost two tons while airborne.

Here's what you might want to do. Buy an inexpensive hollow core door. CL always has some that may or may not be cheaper than the ones at the Borg. Lay up the MDF-door-MDF sandwich with short pieces of 2x4 about where the casters would be, flat side down for stability. It would be better if you could clamp the sandwich in a few places around the edges so the MDF won't slide on the door, both for safety and so that they whole thing will act as a single structural member. Stand in the middle and see how much deflection you get.

If the deflection is small, you should be fine. Once you glue the MDF to the door (use LOTS) and band the whole thing with 1x trim, it will be even stiffer because loads will be more equally distributed across all those little beams in the door.

Another way to think about this is to envision how much effort it takes to crush the cardboard spacers that separate bottles in a box of bottles (as with a case of wine) if you stand it up on end. Surprisingly difficult, even though it's just cardboard.

OTOH, while building a torsion box piece-by-piece is tedious (I did mine with OSB), you do learn a few things from it. Make sure you pin the grid to the top and bottom plates. That's the big secret to making the whole thing as stiff as possible.

[LarryG]

Jack, everything you write is true, but a critical flaw remains: a standard hollow core door is only 1-3/8" thick. Most mobile-base torsion boxes are more like 4" thick, and must be in order to resist deflecting under load. It may be nigh unto impossible to crush the cardboard honeycomb inside an HC door at any given point, but when the door is placed flat and supported at each end, it becomes a beam.

Place a 2x4 on edge atop a couple bricks and stand on it out in the middle. Then turn the 2x4 flat and stand on it that way. Note the difference in stiffness. Same deal with a door that's only 1-3/8" thick.

[Carlos]

I think the door is hardly relevant. If you want the cheap/fast bench, just use the edges of the door (IE, some 1x2) in between sheets of MDF. I don't believe the rest of the door offers any structural value after you add the MDF.

Larry explained it best; a real benchtop is 4" thick for a very good reason. I recommend the Wood Whisperer episode on building a torsion box; he explains it well and offers some nice cheats.

[jackellis]

Larry,

Ahem. Please pass the crow.

I ran the numbers through a little spreadsheet I developed to help me understand how joists of varying designs would deflect under load (housebuilding project). One door would clearly not be adequate. Three might if the wood frame was thick enough. A wood beam 4" wide by 4" high by 8' long would support a 640 load (20 pounds per square foot over a 4'x8' piece of sheet stock) distributed uniformly along its length with .1" of deflection in the center. MDF would deflect by twice as much. I'd have to do a little work to figure out how much deflection there would be for a load that's largely concentrated in the center.

After looking at this again, I'd consider using something other than MDF for the internals and the plates if it is going to hold more than about 300 pounds. According to the formulas, compared with pine you'd need a structure twice as tall or with 8 times as many lengthwise internal pieces to keep the deflection at .1".

[dlminehart]

So, maybe I need to use angle iron, welded at the corners! Does your spreadsheet include parameters for that, Jack? Perhaps you could email me the spreadsheet, so I can tinker with different materials?

Like, how about 6 casters instead of 4? So the span is 3' instead of the 6' I'm likely to have? A couple extra casters is probably not much different in cost and time than doubling the thickness of the box. On the other hand, I don't know if my garage floor is level enough that all six would be on it simultaneously. An issue for even 4 of them, perhaps? I'd been thinking of doing a pivot arm at the light end, but that wouldn't help with 6 casters.