Designing an extremely large CNC machine

Hello folks,

I am designing a machine to make high altitude, zero pressure balloons. Each balloon is made up of 50 segments that are all heat fused together. These segments are almost 60 feet long, and since we need so many of them, we are making this machine to do the hard work of fusing them correctly for us. The goal size is roughly 4 feet by 60 feet. The difficulty I'm running into is with the extremely long 60 foot Y-axis.

Pictures of the work in progress are here:

https://drive.google.com/drive/folde...cw?usp=sharing

I am trying to make this machine as cheap as possible. Right now, I have it under $2500 all in (as long as I haven't forgotten anything major). The base is made in segments. Note the close up picture. The angle iron (red) is the bearing surface for the top and bottom bearings (yellow) as well as the horizontal bearing (black). This setup is mirrored on the other side of the base. The motor is not visible, but it consists of a 5/8" drive wheel off McMaster coupled directly to the stepper shaft and running on the angle iron. The segments bolt together and the individual pieces of angle iron come together to form one rail on each side.

The concern of my group mates and I is that the machine will jam if one side of the machine were to move faster than the other, or if we had issues with the tolerances of our base. The original design was a V-rail system so that the gantry could rise up on the V-rails if the separation of the rails were not perfect. However, our machine will be pressing down on the plastic to fuse it, so it might lift the drive wheels off the angle iron, leading me to create the current, more constrained design.

Any help is much appreciated. I know this forum loves a good challenge, I hope I brought one...

I’m a little thin on exactly what you are trying to do here but this seems like an ideal machine for a timing belt drive. Beyond that I suspect that you have this machine under funded.

No matter what your friends are right about driving from both sides of the gantry. This especially with the very loose linear bearing arrangement you have. The bearing solution isn’t bad for what I understand you are doing but they will allow a bit of twisting. Timing belts on both sides of the gantry would solve that to a good enough degree.

However this will not be a trivial timing belt installation. The belts will need support an a good wrap around the drive pinion. We are also talking belts here that are rather wide, maybe 50mm. In any event this is likely the cheapest way to get positive drive for a long distance you need to travel.

Interesting idea with the timing belt. I considered that option, but it seems like it would be a little sketchy (and stretchy) over 50-60 feet, plus it's kind of expensive. This machine also does not do any cutting, so it shouldn't have much loading other than when it accelerates and decelerates, and even then, we can run at low speed.

One question I had was what I can do to make the bearing setup tolerate loose tolerances better. Ideally everything would be snug on the rail, but since the rail is in multiple pieces and we are building it on such a low budget, nothing will be perfect. Is there a controlled way for me to increase the tolerance in the Y axis while keeping the gantry perpendicular to the Y axis? I was thinking about having the side black bearings in slots with springs pressing them onto the rail, but I wasn't sure.

Thanks for your help wizard.

Morning,
To solve the stretchy belt issue, you might consider a design that Bell-Everman uses. In this design only a small portion of the belt is subjected to stretch and it works like a Rack/Pinion drive with teeth meshing to remove the backlash. Use one on each side. Here is the Link. Bell-Everman, Motion Control Technologies You might want to use DOM tubing that is fully supported with a designed bearing assembly that runs on the DOM, something like the thumbnail.
Vic

Any reason you really need to have a 60ft long machine?

Can you not draw the material through a much smaller machine using some rollers to replace the longitudinal axis?

Alternatively, assuming you don't need absolute precision for something like this and that there are no cutting forces, why not drive the gantry just using wheels along a track? A bit like a poor version of rack and pinion but without the expense of 60ft rack... given there are no cutting forces I bet that could be pretty accurate.

Quote:

---

Originally Posted by zeeflyboy

Any reason you really need to have a 60ft long machine?

Can you not draw the material through a much smaller machine using some rollers to replace the longitudinal axis?

Alternatively, assuming you don't need absolute precision for something like this and that there are no cutting forces, why not drive the gantry just using wheels along a track? A bit like a poor version of rack and pinion but without the expense of 60ft rack... given there are no cutting forces I bet that could be pretty accurate.

---

I am actually using wheels to drive the gantry. And the way that these segments need to be put together makes it necessary to have them on a table.

My tolerance on these plastic pieces is +/- 0.125, so wheels should work. Thanks for the sanity check.

I would think that as long as the wheels are suitably grippy and there is enough weight pressing down on them you should be able to hold pretty good accuracy after calibrating. Could perhaps even stick down grit paper on the tracks to reduce the chance of any slipping and aid with acceleration/deceleration.

Given that you have very few forces at play other than inertia and no swarf to deal with I think you can get away with stuff on this machine that wouldn't work well on more traditional machines as long as you keep accelerations down to reasonable levels.

Given the budget constraints, I'd definitely just give it a go and if improvements need to be made take it from there. I don't think there is anything wrong with the theory, just might need design tweaks to get it working reliably in practise.

I would be sure that your Y axis span on the rollers are 4 feet or more.

Don

Calibrate your budget.
Use r&p.

Thank you guys for the suggestions. It sounds like it will come down to making a prototype and seeing the specific kinks that arise.

Thanks again.

If the segments being joined are flexible (like cloth, canvas etc), could you not dispense the segments from a roll/drum, pass them through the heat/fuser (fixed position) and collect the fused segments onto another roll/drum?

https://www.cnczone.com/forums/attac...d=405172&stc=1https://www.cnczone.com/forums/attac...d=405174&stc=1 Hopefully these pictures give you a better idea of what we are doing. Only the profile of the segment needs to be sealed to another segment. In the end, the whole thing will unfold and we can inflate it into a balloon.

Hi,
I would be looking at rack and pinion for your long axis.

discounted automation products: Motion Control, Rack and Pinion

Craig

R&P can be be found cheap in POM (Delrin etc). Driven by an axle it shouldn't rack if you can prevent jump/skip.
This isn't something I've tried and there may be some real practical limits I haven't taken into account.