Concept Design for a new 6-axis router - take your best shot!

[Jim Dawson]

That's an interesting offer, Jim -"free" is a very attractive price. What does your system consist of? Does it require a particular model of Galil control, or can I find an old one someplace that will work? No problem about it not being quite ready yet, since this thing will take me a while to construct. And since I'm a VisualMill distributor, the six-axis CAM is something I have access to, and really should become more familiar with.

It's just software that will run on any Win7 - Win10 computer. I'll have a demo version pretty quick that has the Galil comm stuff stripped out so that it will load on a computer without the comm drivers installed. It will work with many of the Galil controllers, PCI or Ethernet. There are a number of them on EBay from about $250 and up. 21x0, 22x0, 18x0, 18x6, 41x3, 40x0 would be the models you want to look for. The ''x'' = number of axes. Depending on what you buy, may also require cabling and breakouts, some of the used ones have the needed hardware included. If you decide to buy a Galil controller from EBay, contact me and I'll be able to advise you of any needed hardware. The Galil controllers are an industrial strength motion controller and PLC in one package.
I've installed about 70 of them in industrial applications over the last 20 years. Only had one failure when an ''electrician'' managed to connect 240AC to the 5V DC buss.

When you get around to creating 6 axis G code, let me know. I'll need a sample to make sure my software is compatible and to debug everything. So far, I've only tested 4 axis G code.

[johncarpenter]

It would, I'm afraid - thanks for checking that. The gantry probably will weigh about 250 lbs with the Z axis, heavy ball screw, spindle and the 5th/6th axis assemblies. But don't I get to divide the weight by two, since it's supported on two sides? I suppose when all that stuff moves to one side or another, it starts weighing a lot more on that side, though. Fortunately, I haven't ordered the steel yet. What if I went up to 4"x 10" - would that improve things significantly?

With your non-standard beam end conditions it is probably time to do some simplified FEA if you have that capability. Your 10 foot 4x8 beams weigh 275 lbs each. Making them taller definitely helps (by the cube of the height). But also linearly heavier. Maybe diagonal braces or a middle set of leg stands would be better. Decide what deflection (sag) you can stand and build the lightest structure that gives you that.

[awerby]

There was an interesting suggestion in the carbon fiber gantry thread, about pre-tensioning the gantry with interior cables running from the top corners to the middle of the bottom. I've seen that done on some cranes, and I wonder how much that would help with the sag problem pointed out above, also how to implement it without having it get in the way of the moving parts. Perhaps it should have that plus another steel cable that goes from the bottom to the top, so the beam can be pre-tensioned without distorting it. Has anyone done that before? I have decided to go with 10" high beams instead of the 8" ones in the original design, and will be looking for a Galil controller so I can try Jim's new system.

[1Jumper10]

I think that's called a torsion box??? It would probably help with sag but it's doubtful it would help with torsional rigidity. I tried using an 8x10"x3/16 wall rectangular tube for a gantry on my machine. The long side of the tube had a noticeable bow in it. It was kinda egg shaped but not that severe. My steel supplier that this was typical. We welded cross braces in it to give it some strength. I eventually set it aside and went with a 6x8x1/4" wall tube for the gantry. The thicker wall and shorter height made a more dimensionally consistent tube. A 10" tube with pre-tensioned interior cables for sag and welded in cross braces for torsional rigidity might work good. But from a fabrication standpoint, it sounds difficult to achieve without inducing twist or distortion. Have you considered using an 8" I-beam or, a pair of 8" I-beams? You could use 2 I-beams and have your Z axis carriage span both and ride in between them.

[awerby]

Actually, the tubes I was considering doing that to were the fixed beams that form the X axis, not the gantry. The idea was that they'd go inside the tube from one end on top to a bar fixed onto the bottom, and then to the other end, in a shallow "V" shape, with a second pair doing the same thing from bottom to top. Thinking about it some more, I was looking at fabricating some steel cables with threaded ends that would protrude through some angled end caps, so I could tension them with nuts. Using 4 of them, I was hoping I could control or remediate some of the twisting and sagging that hot-rolled tubing normally has. I was wondering, though, if this would induce weird harmonic vibrations - after all, these would be a lot like piano wires or guitar strings. I was also thinking about cross braces, but welding them in would probably warp the tubes to an unacceptable level - at least I'm not sure how to stop that from happening.

Since I'm making the gantry with dual tubes, so I have a place for my big ball-screw nut, I didn't think the tensioning cables were going to be feasible there, although I'm considering some braces in between the tubes for stiffness.

I thought about using I-beams originally (I've even got some lying around) but everybody said they weren't as rigid or straight as tubing of equivalent weight.

[1Jumper10]

When we welded in our cross braces it didn't noticeably warp the tube. We were careful and welded things up in a set sequence. I'm sure there was some but it was machined after welding. If you could use it I'd give it to you. Too expensive to ship I'm sure.

I've heard the same thing about I beams not being as straight as tubes. And being a little flimsy. I saw mine flex a little as I moved them around building my router, hoisting them and such. But it can work to your benefit in that you can place them straight and then lock them down. Once they are fastened in place they are no longer flimsy. There's also 2 distinct types of profile's: H beam and I beam. Depending on the application one is probably better than the other.

Sent from my Pixel XL using Tapatalk

[awerby]

When we welded in our cross braces it didn't noticeably warp the tube. We were careful and welded things up in a set sequence. I'm sure there was some but it was machined after welding.

[That would definitely help, but I got a quote for blanchard grinding to true up my tubes, and it didn't make sense for this project (and my budget). I've decided to use epoxy to provide flat level beds for my rails. Did you take any pictures of the cross-bracing in progress? I'd like to see how you went about it.]

If you could use it I'd give it to you. Too expensive to ship I'm sure.

[How much does it weigh? And where are you located?]

I've heard the same thing about I beams not being as straight as tubes. And being a little flimsy. I saw mine flex a little as I moved them around building my router, hoisting them and such. But it can work to your benefit in that you can place them straight and then lock them down. Once they are fastened in place they are no longer flimsy. There's also 2 distinct types of profile's: H beam and I beam. Depending on the application one is probably better than the other.

Sent from my Pixel XL using Tapatalk

I'll have to look at the beams I've got again, and try to figure out which they are.

[1Jumper10]

http://www.cnczone.com/forums/cnc-wo...are-forum.html
See posts #8 - 13 for how we welded in the braces. This was on an 6"x10"x3/16" wall rec tube.
I'm in Michigan and the tube is heavy. I doubt UPS or FedEx would handle something like it. It would probably have to ship via freight. I'm sure its not worth it but if your ever in the area its yours.

[awerby]

You certainly went the extra mile to ensure rigidity in that beam. It would probably hold up a house - in canteliever. But you say it went "egg-shaped" with the heat from all that welding? Or was that another one? How long is it?

[1Jumper10]

No, it was a little "egg shaped" when we got it from the steel supplier. Not severely, but noticeable to the naked eye if you looked at the end of it. It might be more correct to say the 10" sides bowed out a little in the middle. It was just over 6ft long.

If you look at the middle pic in post #12, you'll see how we welded a pad on the bottom and closed off the end. In your design, if you're going to drill through the top and bottom of the tube and then bolt it down to the concrete pillars, if you put the nuts on the top of the tube and dont somehow support the inside of the tube you will deform it with bolt tension. That was part of why I suggested you weld a pad to the bottom and bolt the pad to the pillar. I had forgotten that particular reason until I reviewed my pics. You might want to take anything I say with a large measure of salt

[CutMasterCNC]

Look at mine


I could add 2 more axes and go to 20' or more. I have placed 500 lbs on xy axis. had 210lbs on z axis. I am completing the documentation on he build. Should have it sone.

[awerby]

No, it was a little "egg shaped" when we got it from the steel supplier. Not severely, but noticeable to the naked eye if you looked at the end of it. It might be more correct to say the 10" sides bowed out a little in the middle. It was just over 6ft long.

[That's about the length I need, but I decided against a single beam for the gantry, since adding the ball nut would push the whole Z axis assembly way out, adding leverage in the wrong direction. I figure using two smaller beams connected with a stout plate would give about the same degree of rigidity while forming a recess that the nut could travel in.]

If you look at the middle pic in post #12, you'll see how we welded a pad on the bottom and closed off the end. In your design, if you're going to drill through the top and bottom of the tube and then bolt it down to the concrete pillars, if you put the nuts on the top of the tube and dont somehow support the inside of the tube you will deform it with bolt tension. That was part of why I suggested you weld a pad to the bottom and bolt the pad to the pillar. I had forgotten that particular reason until I reviewed my pics. You might want to take anything I say with a large measure of salt

I thought about that, but decided I could just slip a length of heavy pipe over the threaded rod as I stuck it up through the tube, which would keep it from deforming when I cranked down on the nuts. But I was thinking about end-caps, if only to have something to tension my cables against.

[awerby]

Look at mine

[That's a pretty nice build, but have you used it as a router as well as a plasma cutter? Plasma cutters can be built a lot lighter, since they don't have any counter-forces to resist, and the action of cutting doesn't cause any vibrations that need damping. Routing, especially of hard materials, requires a much stiffer machine that's less prone to resonance.]


I could add 2 more axes and go to 20' or more. I have placed 500 lbs on xy axis. had 210lbs on z axis. I am completing the documentation on he build. Should have it sone.

Keep it coming!

[routalot]

I don't entirely understand the manner in which the spindle is mounted to the fifth axis gearbox.I would have expected the motor to be mounted by it's side and getting all the alignments correct is not the work of a few minutes.

[CutMasterCNC]

Yes I have used it with a 2 hp router and a 5 hp spindle. That is why I had to legthen the plates from 6" wide to 12". With the six inch plates it tried to lift a little under heavy pressure from the router and spindle. Now even when I get to aggressive with a cut something else gives ..tool bit .

[awerby]

I don't entirely understand the manner in which the spindle is mounted to the fifth axis gearbox.I would have expected the motor to be mounted by it's side and getting all the alignments correct is not the work of a few minutes.

[This is just a "concept design" - I haven't precisely worked out all the details yet. But the spindle motor will be attached by screws to a back plate that comes off the pivoting arm of the 5th/6th axis assembly. I will endeavor to get the tool concentric with the swivel and in line with the pivot; that will take some careful measurement and machining which I haven't yet done.]

[awerby]

Yes I have used it with a 2 hp router and a 5 hp spindle. That is why I had to legthen the plates from 6" wide to 12". With the six inch plates it tried to lift a little under heavy pressure from the router and spindle. Now even when I get to aggressive with a cut something else gives ..tool bit .

I would expect that, since tool bits - especially small carbide ones - are pretty sensitive to chatter, which a machine with less than optimum rigidity will exhibit when the cutting forces build up. Your design looks good for a plasma cutter, but if you want a router too, that might require another build.

[NIC 77]

I like the simplicity of your design.

I've never seen stainless used for a CNC. I have to think it's not needed.

It's not the way I would do it, simply because of ground heave.

The house I live in shifts a bit every now and then, a couple doors are hard to close when the weather shifts, and then mysteriously fix themselves a few days later. Probably only a few mm difference, but it does happen. Not sure if it's a huge issue.

You know what, I was looking around online the other day, and I did see some used industrial robot arms for 5K. I can't remember where.

I think I just realized that I should have gotten myself a used industrial robot arm and vertical rotary axis instead of building my current 4 axis build that I am working on. Ouch. I hate those kinds of realizations.

Robot arm and rotary would be pretty cool.

[louieatienza]

I like the simplicity of your design.

I've never seen stainless used for a CNC. I have to think it's not needed.

It's not the way I would do it, simply because of ground heave.

The house I live in shifts a bit every now and then, a couple doors are hard to close when the weather shifts, and then mysteriously fix themselves a few days later. Probably only a few mm difference, but it does happen. Not sure if it's a huge issue.

You know what, I was looking around online the other day, and I did see some used industrial robot arms for 5K. I can't remember where.

I think I just realized that I should have gotten myself a used industrial robot arm and vertical rotary axis instead of building my current 4 axis build that I am working on. Ouch. I hate those kinds of realizations.

Robot arm and rotary would be pretty cool.

Yes you can use this to run it...

https://www.rapcam.eu/

[louieatienza]