I don’t think it actually requires much. 2-3 Y axis stringers/braces would probably do it. Unless you are using it as an assembly table, maybe add one more. I think most torsion box tables are pretty over built.
4 Likes
We’ll come back to this in 2.5 years. 
I like this table! Looks like it could build itself too. Start with a couple janky ribs laying on the ground and a spoilboard and you can cut out the ribs (and the LR3 strut plate).
During assembly, before adding the top, surface the top edges of the ribs and check for twist before securing spoil board.
(A torsion box with only one face is still a little bit compliant in twisting, right?)
1 Like
I really keep looking at the design of these things. I am not really sure. I think once you glue on a face, or glue in the ribs twist is mostly locked in, and each face is doing more for tension (bowing), than the compression aspect.
I almost forget…a door is still a great prebuilt smaller torsion table. Sometimes you can get them dirt cheap.
2 Likes
It might be fun to simulate one. I don’t think I have any simulation software I can use anymore.
1 Like
I agree. A long time ago I made one with 5 ribs 36" long and only two ribs 60" long. The ribs were all 3/4" ply and the faces were 3/32" hardboard. It was plenty rigid. I also added 3/4" spoil board, screwed into the edges, but it was pretty flat even before that.
This is a pic of my 50lb vise sitting on the table on a tape measure.
It was as flat as the level.
6 Likes
A torsion box gets most of its value from having 2 faces. Those are the biggest contributor to the overall stiffness of the box. The reinforcement comes from having the ribs joined and reinforces on both sides. Opening up one end does allow much more compliance
Yeah, when we are dealing with machines that “need” finishing passes to ensure a cut accurate to ~0.1mm… Also, it’s reasonable to expect that these tables shouldn’t really have anything on them that we can’t reasonably be expected to lift…
They are… A pair of 32" doors from a building supply recycler can be had cheaper than a sheet of plywood
2 Likes
I need to at least add a blurb to the tables page tonight. I don’t think I will have any plans or CAD up anytime soon but I think this info will actually help some people decide. I hate that the first step to building either CNC is so wide open and unguided.
3 Likes
Whoa. These tables are needing even less than I thought.
Especially for an initial janky version that’s just stable enough to cut the real table, maybe it can be super, super quick and cheap. If it sits on the ground on 4 blocks, it doesn’t need to be stiff against twisting.
Fence pickets are cheap (about $3) for 6 feet long x 5.5 inches wide by 3/4" thick. Can I get a decent setup with only five pickets and one spoilboard (2’ x 4’ area) using only a crude reciprocating saw? I have to try it now.
4 Likes
Fence pickets are usually sold still wet and they will not be straight. IDK what tolerances you will get with those. The 3/4" plywood strips I used were 3.5" wide, and as straight as my table saw can cut them. That is a big difference.
Maybe if you found pretty clear pickets and let them dry and then jointed the edges, you could do pretty well.
But they sill still swell with the seasons. Plywood doesn’t, much.
I bet you could build a table out of pickets and then cut the torsion box out of plywood. But you could probably do that with a sheet of plywood on two saw horses.
1 Like
That design you drafted looks amazing to me.
Reminds me of a creeper
1 Like
My craigslist $40 beaten up solid wood dining room table is still flat as can be. Although the door option is interesting.
2 Likes
I. Like. It!!!
2 Likes
I’ve been a bit reluctant to pursue the “floating rails” thing because it adds a bit of complexity to levelling since there are three surfaces to get parallel not just one.
However I’ve just realised that it may serve well in keeping the dust away from the wheels.
Is that a reasonable thought?
I think, in this case, because of the CNC method of cutting this for assembly, having the surfaces parallel should not be an issue. I do also think there could be some benefit regarding keeping some dust/chips away from wheels.
1 Like
Just two, the middle in this case will always be surfaced since you can cut off the edge.
It could keep stuff away from the bearings, there is a gap, or a step. I have never had an issue with that though. This last round of HDPE was the most messy thing I have cut, a few chips get near the bearings but for the most part my Z was not critical to that degree, if it was I would either put a brush there or a longer tape skirt on the dust shoe. I know this seems to bother a lot of people, I just don’t see it as a common issue. Worst case is cutting right next to the bearings. At that point what is the biggest chip your machine throws and how far off could it possibly move the Z? Big stuff gets pushed out of the way. At the other end there is some trig involved but basically zero effect of a chip under the wheels.
2 Likes
I made that design that way for different thickness materials, and hopefully slide the vac hose on the rails side gap.
2 Likes
I would think the twist of the table (making rails not parallel) would be the only degree of freedom, and there would need to be a procedure to true the table in any case.
Parhaps shim the feet until the rails are parallel. Use the cnc to surface the top of the ribs. Attach top and bottom skins to make the table super stiff.
3 Likes
That table is beautiful! I love how you used the slot fittings, kinda figured you would, being a CNC cut CNC table, but wow, that’s impressive! I’d get on board funding that research & development! haha
2 Likes
Hopefully when I get back from Rmrrf and the dust settles a bit, that is on my short list.
2 Likes