LumberJocks Woodworking Forum banner

Reduced weight torsion box

Tags
question
6K views 15 replies 9 participants last post by  clin 
#1 · (Edited by Moderator)
I'm considering a torsion box as the base for a 9' X 11' Table Saw out-feed table / CNC machine.

I'm using the rules of thumb from Ken Horner's "More Woodworkers Essentials" book for 1/4" skins that calls for a 4" x 4" core and I plan to make the table 3" thick. I also plan to use a 1/8 hardboard on the top as a sacrificial surface. Core will be 1/4 MDF.

My concern is the weight. I'm getting about 565 lbs when I work it out based on 1.15lb per square foot for the MDF. I have a small shop and will need to flip the table on its side when not in use.

I'm thinking about boring 2" holes on all the core members and on each grid of the top and bottom skin. This would reduce the weight by about 35%. What I'm not clear on is what this would do to the strength of the box.

Will adding the holes impact the structural integrity of the torsion box?
 
#5 · (Edited by Moderator)
There are a lot of questions that will come
up in seriously considering such a build.
Not only will you invest considerable time
into it, do you have the resources to attach
the skins and maintain flatness to the standard
you're expecting?

Even doing it with a vacuum bag would be
a relatively heroic effort for a one-time glue-up.

Torsion boxes can be made with lighter cores.
Perhaps a layered series of torsion boxes could
allow weight reduction by spacing the bottom
layer mdf interior core much wider, then laminating
paper-core material on top.

http://www.woodworkersjournal.com/wp-content/uploads/ian_kirby_torsion_boxes.pdf

While making the foundation is 2 or 4 sections
would complicate matters in some ways, it would
be a lot easier to get each smaller section closer
to flat than doing the whole thing at once.

Try googling "paper core torsion box" for more
ideas on materials and methods.

I also think a welded steel frame may offer the
same rigidity at a lower weight.
 
#6 ·
I think a 9' x 11' foot table is much too large to expect to be able to be able to flip it on it's side for storage. Short of using a block and tackle or some type of hoist system.

I'm also curious why you need such a large outfeed table. You'd have to be ripping stock 20 ft or longer to need that much outfeed table. or perhaps dealing with very flimsy stock that can't be allowed to hang off the back of the outfeed.

While torsion boxes are much lighter than the equivalent solid table, in your case that would be in comparison to a 3" MDF table. A 9' x 11' 3" solid MDF would weigh about 1,100 lbs. With 99 sq-ft of table, at 1.15 lbs/sq-ft of 1/.4" MDF, the skins would be 114 lbs each, 228 total. I come up with 128 sq-ft of webs (what I think you are calling the core). That's another 147 lbs for a total of 375 lbs. Not sure about the hardboard weight, but I'm not seeing how you are coming up with 565 lbs. Maybe I'm misunderstanding your construction.

I assume the flatness and rigidity you want is for the CNC operations. Because generally you wouldn't need that for an outfeed table.

Regardless of weight, I too would look at making it in sections. If you build flat sections, you should have little trouble lining them up.

As to this flatness, it will be very hard to make a table that large that flat unless you already have a very large flat reference. Keep in mind a concrete slab is no where near flat. So you'd have to got to a lot of work to create that surface. Much easier to create a smaller flat surface. Build the first section, then it becomes the reference surface for the remaining sections.

I also would not remove 2" holes from the webs. If the table is 3" with 1/4" skins on each side, the webs are 2.5" tall, and you'd only leave 1/4" in the middle of each side.

I would never put holes that large in the skin. The strength of a torsion box comes from the skins. Removing that much material would reduce the strength substantially. I think it would be close to the point that it would hardly function as a torsion box anymore, no matter what the material, and with MDF, I think it might concentrate so much stress that the material could fail.

And by my rough count that's something like 3,600 holes in skins and webs. That only saves about 78 sq-ft of 1/4" MDF for a weight savings of about 90 lbs. Which is 24% of my weight estimate. Not insignificant, but it's not really going to make it so much lighter. And 3,600 holes!

MDF is great material for a torsion box becasue it is very flat and it is plenty strong enough unless you plan to beat on it with mallets and hammers. But it is a bit heavy. Plywood will be about 2/3 the weight, but won't be as flat. But still possible flat enough.

I also second Rick's suggestion that you look at the Ron Paulk table design. It is specifically designed to break down. Though his design I believe is 2'x8' sections, but can of course be modified to whatever dimensions suit the need.

I'm not sure what you are doing with the CNC, I know a guy that has a 6'x10' or so professional CNC bed. The top appears to be particle board. It is solid with no holes. No idea what the frameworks is. But it must form the top of some sort of box. He hooks up a vacuum to it, and it draws vacuum through the top surface. Apparently the particle board (or whatever) is porous enough. He can lay a sheet of material on this, turn on the vacuum and that sheet will not budge.

As long as the parts he is cutting are not too small, he doesn't need to hold the material down in any other way. Not sure if that might apply to your situation, but something to consider.

One other comment on torsion boxes. There's a lot of misunderstanding about how these work and many people go to a lot more work than needed. The internal webs or core serve one purpose, and that is to hold the skins apart at an equal distance. The webs do NOT need to join to each other. This doesn't add strength. The webs do NOT need to be continuous. They could just as well be random lengths placed randomly around.

Again, they only serve to hold the skins apart. The important joint in a torsion box is the bond between the web and the skin. This needs to be strong. Fortunately this joint is in shear which is where a glue joint is strongest.
 
#7 ·
Look into restroom partitions. They are honeycomb construction and come in many different sizes. There is also a steel honeycomb panel system that is used in marine and aircraft construction. You could combine smaller panels to get the overall size you want.
 
#8 ·
I don't want to get into a protracted discussion but it isn't as simple as you make it out to be. The regularity and spacing of the grid patternl and the strength of material is most certainly critical. I agree that the components of the grid pattern do not have to be glued together but they must be glued to the skins at every point of contact. There are many factors that go into the design of a torsion box and hand calculations can be complex.

I have a utility room behind my garage that is 10 feet tall and 8 feet wide. I wanted to build a torsion box overhead shelf that was 32 inches deep and spanned the whole 8 feet with no support except on the edges. I wanted it to be light enough that I could lift and and drop it into place by myself. I needed for it to support at least 200 pounds of dead weight in the middle. The final design was constructed of B/C grade 1/2 inch plywood for both skins and the internal grid was 3/4" MDF in 1.5" strips. I can't remember the grid pattern but it was somewhere around 8 inches square. The total thickness of the assembly was 2-1/2". I was able to install it myself and I quit testing it for deflection at 300 pounds. You can't get that kind of strength and rigidity by randomly placing a bunch of uniformly thick spacers between two flimsy skins.

As I understand it, the OP wants to build a long outfeed table with no supports in the middle. Making it in sections seems to me to defeat the purpose of using torsion box in the first place.

I would also like to know where the 565 pound projected weight came from.

One other comment on torsion boxes. There s a lot of misunderstanding about how these work and many people go to a lot more work than needed. The internal webs or core serve one purpose, and that is to hold the skins apart at an equal distance. The webs do NOT need to join to each other. This doesn t add strength. The webs do NOT need to be continuous. They could just as well be random lengths placed randomly around.
- clin
 
#9 ·
I don t want to get into a protracted discussion but it isn t as simple as you make it out to be. The regularity and spacing of the grid patternl and the strength of material is most certainly critical.
- ArtMann
Art is correct- the primary purpose of the core in a torsion box is to prevent buckling of the skin that is under compression (usually the top in a typical table type arrangement).
 
#10 ·
You can t get that kind of strength and rigidity by randomly placing a bunch of uniformly thick spacers between two flimsy skins.

- ArtMann
Yes you can. For example, webs randomly spaced could be stronger. If simplistically, the largest space between the random webs was smaller than the spacing between evenly spaced webs.

My point was not that throwing spacers anywhere you want is an ideal design. But rather that, some sort of uniform grid or pattern is not a requirement for strength. Though certainly a very specific pattern will maximize strength. But even then, this pattern may not be evenly spaced. For example, closer together where loads are higher.

Art is correct- the primary purpose of the core in a torsion box is to prevent buckling of the skin that is under compression (usually the top in a typical table type arrangement).

- TungOil
Preventing bending or deflection might be a more accurate term than buckling.

Though what is really going on, is the webs keep the skins parallel to each other. But it actually does this by preventing the skins from sliding relative to each other. This is the shear stress put on the web to skin joint.

This is why in some applications these webs are called "shear webs".

If you simply had two skins laid, but not glued to a core. For example, two 1/4" plywood skins with a foam board core, place this on some saw horses and I think we would all agree it wouldn't be very strong. Assuming the foam was pretty weak, it would pretty much just behave like two independent pieces of 1/4" plywood. Not even like a single 1/2" piece. So even though the skins are held evenly spaced apart, there is very little strength. Nothing to prevent deflection or bending of the plywood.

But glue the plywood to the foam core and it would become much stronger. The difference is the skin to core joint now resists shearing. I.E., the skin cannot move or slide relative to the core and opposite skin. And to a large degree it will be almost as stiff and strong as a solid piece of plywood the same thickness as the entire structure.

Bottom line the webs or core keep the skins parallel to each other which requires that the skin spacing be maintained AND that the skins not be allowed to "slide" relative to each other.
 
#11 ·
Preventing bending or deflection might be a more accurate term than buckling.
- clin
I meant buckling, in the engineering sense. I think we are talking about two different things. What you say above is true regarding deflection of the structure under load, but it is not the first form of failure. The top skin will always fail under buckling before the bottom skin will fail under tension (assuming equal thickness & materials are used in the skin). Revisit Euler's formula if you need more convincing, or look here:

https://en.wikipedia.org/wiki/Torsion_box
 
#13 ·
Thank you for the feedback. The extra 200 LBS is actually the table saw weight. I planned to mount the saw in a pocket vs. hanging it on the end as done in the Paulk design.

I've found a couple papers on stress skin panels and the effects of holes in laminated beams that I'm reading over this week. Looks like this may not the way to trim weight but I see a lot of tables with holes for clamping in the top skin at least.

I may just build up two samples to see how they break. :>
 
#15 ·
Wood will always be a compromise. I think what you really want, and possibly need, is a metal frame. I also think your question would get more productive answers if asked in the CNC sub with a title something like, How to build an ultralight but rigid CNC table?
 
#16 ·
Preventing bending or deflection might be a more accurate term than buckling.
- clin

I meant buckling, in the engineering sense. I think we are talking about two different things. What you say above is true regarding deflection of the structure under load, but it is not the first form of failure. The top skin will always fail under buckling before the bottom skin will fail under tension (assuming equal thickness & materials are used in the skin). Revisit Euler s formula if you need more convincing, or look here:

https://en.wikipedia.org/wiki/Torsion_box

- TungOil
I agree with TungOil. I think "buckling" is a more appropriate engineering term than "deflection" in this circumstance.

- ArtMann
Bucking is an unstable state, that in this case would occur at the onset of catastrophic failure. Essentially, the torsion box would start to fold. That of course is to be avoided, but the rigidity and ability of the structure to resist deflection is the primary purpose of torsion box construction.

In other words, we want our table to stay nice and flat.

But it's clear, we are just debating the term bucking vs deflecting, and even in the engineering world, there's confusion about these terms. Therefore, I think it's pointless to debate the terminology any further.

Bottom line, I think we can all agree that the webs or core, keep the skins flat.
 
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top