New Machine Build- Greybeard's bootstrap Z axis

[greybeard]

While the snow and ice are being kept at bay by the log fire, my mind turned back to considering a small cnc machine

It's going to be the sum total of most of the crazy thoughts experiments I've had.
Possibly a last desperate throw of the dice, but then who knows what lies ahead.
With that somewhat wistful bit of philosophy, to work.
Please remember, every detail is experimental, and this is a test bed for greybeard.

The X and Y ways will be based on air bearings, HDPE porous pads running over glass bedded onto epoxy/quartz beams, spun cast of course. They will carry safety slides in case of air loss.
The drives will be my own take on Mike Everman's "Best belt drive ever"(see zone).

Now to the Z drive. This will be the "Greybeard Bootstrap Drive", details now follow.

First the power source is a small air grinder, cheap and cheerful, and easily replaced - ~£10 on ebay. Comes with a 1/4" collet chuck, screw in 1/4" bsp air connecter, and has a grip switch to power on/off. Will swap the air connector for a right-angled bend for compactness, and I aim to link the air supply to the grinder via a relay control to keep the air demand to a minimum. Only on when the grinder is actually cutting. This allows a smaller air supply, ie the one I've got.

The grinder will be mounted in an epoxy block which will have bronze bushes for the vertical slides cast in situ.
In another cast block will be a horizontal mounting for the Z stepping motor also with embedded bronze bushes, and this slides on the same rods as the grinder carriage.

The stepper motor has a thin rubber coating bonded to the shaft, and this presses against a similarly coated angle strip supported by twin bearings, bolted to the stepper block, on the other side. This angle strip is bolted to the grinder mounting block.

The arangement that follows is my favorite.
The two cast blocks are connected by wires over pulleys, so that as the stepper motor moves upwards, the grinder block moves down.
Thus the linear movement is halved for a given rotational movement of the stepper shaft, and at the same time the stepper acts as a counterweight to the grinder. This will produce an estimated movement of 0.025mm/step

If and when I get to building it, photos will follow.

John

[Jesse B]

Very interesting ideas you've got there. Subscribed

[greybeard]

Essentially the same idea and design, but rehashed to become more compact, and therefore stiffer, I hope.

I've assumed that ~ 1" of Z movement is sufficient for the prototype, as it may well finish up as little more than a 2.5D machine, cutting and piercing wood of veneer thickness, ~ 1/32" - 1/16".
Being able to do bas relief carving on the surface of 1/8" thick wood will be a bonus.
Tool changing will be done at a specific location away from the working area, so no extra depth is needed for that.
John

[epineh]

Count me in...watching to see where the go faster stripes go

Russell.

[epineh]

Oh and I am assuming that the machine will be red to maximize speed ?

[greybeard]

Over here that would have to be infra red due to the Doppler shift, so I think it will need a source of stealth paint.
Hi Russell, glad to see your keeping an eye on the old man
John

See your having a bit of a warm time at the moment. Hope it gets cooler for you.

[epineh]

Heh, well somebody has to keep an eye on you...

I live in a wet tropics area so it is either hot, or hot and raining, that is about our four season's in a nutshell...

The guys in the southern parts of the country have had a heat wave the last couple of weeks so they are feeling it, also huge fire risk as it is so dry for them at the moment, talk about a land of extremes.

Any Hoo, back to the build, does this mean we will see some Epoxy Granite action ? I followed a couple of the E.G. threads but got lost in all the tech stuff (it was almost as bad as those crazy printer threads ) I am currently doing a lathe retrofit and need to make a really large overhead bar/gantry to make a milling attachment, am thinking of lots of steel or maybe E.G. not too sure yet, need to get the lathe going first, but I would be interested in checking it out...

Cheers.

Russell.

[mlabruyere]

This is amazing....I can't wait to see what this becomes.

I love the cadence of your writing style. Very easy to read.

Thank you.

[greybeard]

Hi gentle readers, and welcome.

Some time ago I made an epoxy/granite based carriage to carry Z and Y air pads, based on porous carbon blocks.
It was a useful exercise in spin casting a variety of shapes, and I learnt a lot from the experience which I hope to put to could use here.
As this part of the machine is crucial to its success, I'll spell out the detail here.

The epoxy/granite thread, of great renown(and length) is aiming at a diy recipe and method that will enable users to get the maximum strength structures for what ever needs they have. Fundamentally this means getting the quartz/granite grains to pack as tightly as possible into the mould space, and replace all the air with epoxy, bound tightly to the grains' surface.
The combination of different grain sizes along with different proportions of them in the mix is the first indeterminate, but a general principle emerges that acts as a guide to a recipe. Once that has been chosen, treating the grains chemically to improve the bonding of the epoxy to the quartz is desirable, and suitable reagents have been identified in the thread. The remaining problems centre on the choice of resin and how to get it to replace all the air that would be trapped in the remaining spaces between the grains.
Vibration is being currently discussed as a means to pack the grains together, and at the same time remove any air bubbles.

My own take on the whole process gets me going off in a different direction, namely round in circles
I've decided that for my own needs, generally of much smaller size and accuracy, the high spec recipe can be bypassed in favour of the means of achieving the casting.
By adopting rotational casting at relatively high G force, I can get the maximum grain packing density at the surface of any component I need to make, along with any air being driven towards the centre of the said component.
This is highly desirable on account of the stress at the centre of a loaded beam is a minimum.
I shall still use the grain size distribution alluded to above, but I can use more resin than theory suggests is desirable for the simple reason that any excess is also displaced by the grains and moves towards the centre.
This was borne out by my castings last year.
One problem did emerge, and that was that even though the moulds I made were strong enough to withstand the forces on them, because I spun at about 150 - 200 G for the whole curing cycle - about 90 minutes, there was a swelling in the central area.
If I'd had a variable speed motor, I'd have spun at full speed to disperse the grains etc for about 10 minutes, then dropped the speed to produce about 5G, just to prevent any slumping of the mix during cure.

The first thing to go is the use of carbon. It's easy to cut and shape, but incredibly dirty work. The source was a replacement water filter cartridge with a quoted pore size of 5 micron, so would have been good for the project.
I have since found a company close by that produce HPDE porous sheets, and who were happy to sell me an A4 size "sample", more than enough for present needs.
Having now got a good idea of the size of the unit, I shall start by producing a "cell" to contain the pad, and set up some kind of test bed to determine the working conditions I'll need to support the carriage.

As I write the mail has just delivered one air grinder, weighing in at 350 gms, so I now have something to measure, and adjust the drawings.

[greybeard]

Regarding the movement of the Z drive, I've been inspired by Mike's belt drive.
Needing only a small (<2 inch) movement, and with the mill head being counterbalanced by the stepper, I've probably come up with my best idea yet( or greybeard has finally lost it - again ).
Sellotape.
See diagram attached, showing the general principle. The non-contact lengths are very short, so stretching isn't an obvious problem. With the small size of the unit, the whole could be enclosed in a positive pressure air chamber to reduce/remove contamination of the adhesive layer by dust.
If it proved to have a short life, replacement is hardly a major expense.
I have pondered on some sort of continual feed of new tape into the set up, but that negates the simplicity of the concept.
The tape is stuck to the mill block, then around the stepper motor shaft, and back to the block. The motor mounting applies the tension to the tape between the two bearings running on the back surface of the tape. They are mounted onto the stepper block.
The bootstraps connecting the two blocks run around bearings mounted on the Z frame. I might consider those to be made from steel tape, rather than wire.

OK, there it is.
Comments please.
John

[greybeard]

A single layer of 1" selotape can be stretched about 1% , so with 2 lengths in parallel of ~18mm, equivalent to one length of 9mm, I could stretch it ~ 0.1mm. If I doubled the layers, that would reduce to 0.05mm, which I think is tolerable.
Perhaps I'll look at laminating double sided tape to a cloth based measuring tape.
Actually double sided tape would stick to the bearings as well, and this would reduce the "free" lengths considerably.
Problem then is contamination of the outer layer by dust, so the extra complication of an air jacket would be needed.
Mmmm.


A layer of double sided tape around the bearings, acting on the outer face of the tape - much better, and would leave a "free" length of only about 4mm each side.

[greybeard]

I've now settled on a maximum Z movement of 1", a wire cable as bootstrap, and reverted to a single layer of the toughest adhesive tape I can find.
The latest iteration of the Z carriage has a free tape length of about 4mm either side of the drive shaft, reducing any possible stretch from such an apparent flimsy material to a minimum.
Well, that's the theory.