Folks,

My plans are to eventually begin construction on a router using a torsion box design. Obviously this is simply a tweak to his design and not an entirely new design of my own. I think that Steve (Spalm) has done an excellent job on making improvements on the JGRO machine to address some possible weak points. But without JGRO, how far back would we be? The successes come from standing on the shoulders of those before us. I already have others' big shoulders to stand on. :)

As much as I like Steve's torsion box design (http://www.cnczone.com/forums/showthread.php?t=12056) , before I began construction on my own machine, I wanted to address the one concern I've seen with the design and that has to do with alignment of the rails. Steve's machine seems to rely soley on the gantry adjustments for alignment. As I understand it, this only affords vertical movement. For obtain better alignment, I'm looking at two alternatives. These will follow in two separate posts.

Rance

This first alternative incorporates both vertical and horizontal adjustments of the rails. The green assy. (in the attached diagram) represents a modified torsion box to which an adjustable rail assembly (red & blue parts) will be bolted. The ends of the angle-iron(AI) will have a short section of pipe attached either by bolts (as shown) or it could be welded. I suppose a solid rod could be used rather than a pipe but weight could be a consideration. I've also shown a modified version of the typical adjustment blocks (Don't ask me why, I'm sure it has something to do with the "not invented here syndrome" :D )

As for the adjusting process... Initially the Adjustable Rail Assy(red) is placed on a flat table and the AI(blue) is screwed to it. It is then checked for level and adjustments can be made via shims between the AI and the the rails to make them parallel to the flat table.

Next, the assembly would be installed onto the torsion box but <b>without</b> the Adjustable Rail Assy. actually 'bolted' to the torsion box. Now adjustments would be made to the Adjustment Blocks to get the two rails parallel with each other as well as parallel to the flat surface of the torsion box.

Once you have what you believe to be correct alignment, you would install shims if necessary between the Adjustable Rail Assy. and the torsion box so when these two units are permanently bolted together, the act of tightening the bolts would not bow the Adjustable Rail Assy.

Although this seems to be a complicated approach, it does afford total adjustment of the rails. I know it seems long and drawn out but I've not figured out a better way to get full adjustment, yet. Any suggestions would be appreciated. Maybe looking at this will spur someone else to come up with a better solution. :)

Rance

This second alternative is simpler but goes back to relying on the gantry bearings alone for adjustment. Similar to the mounting of the angle iron (used for the bearings) on the non-adjustment side of Steve's gantry, I would be using angle iron AS the rail. Although I would not have as much control of the alignment, I believe that between this method of manufacture (the simple cutting of slots) and the use of angle iron for rails (less variation in dimensions than pipe), I would be more assured of a square machine when finished.

As you can see in the picture, this incorporates the use of angle iron (used for rails) mounted into groves in the torsion box. The angle iron rails would not only fit into the groves but could be supported by triangle ribs too. It could be screwed to the ribs & torsion box or bolted (similar to the way the bearing angle iron is bolted to the side of the gantry (being careful not to interfere with the heads of the bolts in the other angle iron).

Steve, I hope you do not mind the fact that I took the liberty of butchering your pics for use here. Based on our earlier correspondence I assumed that you would not mind. Please let me know if I assumed wrong. :)

Rance

PS: Sorry about these image sizes. I'll have to go see how you guys insert just the thumbnails.

Upload the images to this site as attachments, don't link to a web page.

IMO, simple is better. All those adjustments will be a real pita to get set up. BUild an accurate torsion box and everything will fit together fine. Even if it's out of square (parallelogram), as long as the side to side ribs are all the same size, the rails will still be parallel. I'd recommend clamping the angle iron in place during assembly to help ensure this. Imo, building an accurate torsion box is much easier than all those adjustments would be. You'll find that every time you move one thing, something else will move the wrong way.

One other thing, is make sure you're torsion box is FLAT. After the frame is assembled, use a good 4ft level or straightedge and shim it until it is perfectly flat. When I built mine, I couldn't slip a piece of paper under the level at any spot, with the level at any angle or position. Then attach the skin, flip over and repeat. Ideally use a slow setting glue so neither skin is set before both are attached. Once both sides are on, DON'T move it 'til the glue sets. This should give you a perfect base to build on.

I still like the idea of machine the rib portion in place, It does require a jig to hold a router @ 45 deg. but the finish product is fully aligned to the torsion box.' One could even just cut the angles in a table saw pretty acurrate. I don't like using screws to attach the rail either. If the ribs of the torsion box had a hole in them then machine screws, washers and nuts could be used. A simple shim could be then placed behind the rial if some miss alignment was detect after assembly.

The only thing is that Gerry is absolutly correct, you must have a flat torsion box to even have a chance at it.

Jeez wish I had those Cad skills

The arrangement looks really nice, I suspect you would wind up pulling the angle in at the site of the screws. You would probably get less screw effect getting closer to the 90 degree point. Of course you would have to allow for the bearing placement.

I would vote with Gerry on making the box square and the rest will take care of itself. I would build the box by gluing the struts to one of the skins instead of making the rib frame first, but hey. Something that I must report when I did some cutting this weekend is that my flat box had bowed up in the center a little bit. I suppose that constant pressure had pressed it that way. I have never put the bottom skin on, which is a real no no.

IMHO I would stay away from all those adjustments. Maybe if you were making a 6 inch by 6 inch machine to do jewelry, but these wood machines are more for 2.5 D stuff with the occasional 3D sign or rosette.

But if you want adjustment for the rails parallel to the table, I would vote for option #2, only make the groves in a second strip of wood that is bolted directly to the side of the table. I don’t think (?) that the little wood triangles will support the angle iron much so delete them. Just try to flatten a piece of angle iron by laying it on the floor and stepping on it. Nor do I believe that the angle needs any attachment except at the ends. Make the strip attachment bolt holes oversized and use Tnuts inside the box. (Screws stink in MDF and ply when no glue is used.) Extend the rails out the ends and tie together with threaded rod.

Wish I had Photoshop in the shop to change designs like that.

Steve

In the above sketch just consider the brown as one cross section of the torsion box cross ribs.

You guys are a great help to me. I appreciate it. Gerry, I like KISS too, I think that's the key. Steve, it seems like I was so close but just couldn't see to "make the groves in a second strip of wood". So here's an acceptable method IMO. I could live with this, it gives me adjustability both vertically via the second strip of wood (bolts & T-nuts), and horizontally (if need be) using shims. I believe that glue alone is enough to keep the angle iron held to the strip of wood.

Oh, and one other thing, I think I'll make the adjustable side of the gantry bearings spring-loaded. Thanks again guys.

Rance

Edit: Image adjustments.

Something that I must report when I did some cutting this weekend is that my flat box had bowed up in the center a little bit. I suppose that constant pressure had pressed it that way. I have never put the bottom skin on, which is a real no no.


It's not a torsion box without BOTH skins attached. It's just a panel with some ribs sort of supporting it. ;)

Gerry, I know. Just never got a round to it. I wanted access from the bottom so I could install some hold down nuts before I closed it up.

Rance, those grooves will require some thought. You can see in your cad that the rail doesn’t really mate too well. You will have to worry about depth, width, and spread of the grooves to get it just right. Of course it will be easy to replace if you need to do it again. Two thoughts I had were to make the strip a bit thicker so these grooves can be a bit more robust. The other is that the rail can be attached to the strip with bolts just like on the gantry. Make oversized holes in the box for the nuts to pass through. You might even consider this eye bolt attachment technique to keep the bolt heads from running into each other.

If you go with it, and get it adjusted and tightened to your liking, I would think that a couple of randomly placed 1/4” holes with unglued dowels would keep it from slipping.

Steve

> You can see in your cad...

CAD? What's a CAD? I'm still using MS Paint, really! :D I've GOT to get started on learning an affordable, REAL drawing pkg.

> the rail doesn’t really mate too well...
> make the strip a bit thicker so these grooves can be a bit more robust.

Yep, yep. The drawing was meant more to get a general idea. :) Thanks for noticing a possible problem though. So would alum. be acceptable for mounting the bearings? I'm guessing it would be too soft for this. How bout for the rail? This seems reasonable. Other than for strength, is there any reasons (structural, dimensional, etc.) that I'd want to go with steel over alum. (or visa versa)?

Rance

Aluminum would be strong enough for the bearing holders. Never use it as a rail as bits will get permanently imbedded into it and the bearings will have a rough ride.

Steve