Close

Mechanical 'design'*

A project log for Computer Numerical Considerations

Not the full-fledged CNC project had.io deserves but the brain-dump @esot.eric asked for

stefan-lochbrunnerStefan Lochbrunner 11/11/2015 at 16:541 Comment

*not really design, more like doing something and seeing what works/sticks. Problem: Metal doesn't stick.

First attempt:

I think you see what I was going for here:

Square aluminium profiles as structural parts with the axis sliding on brass rods/tubes, all of which is driven by stepper motors salvaged from old printers and M3 threaded rods. If machined better this probably would have worked for applications with a very low load, like a plotter or maybe a 3D printer, but for a mill I think this was not a very good approach since the brass rods are nod as rigid as the square profiles.

The problem at the time was that my drill stand wasn't too precise anymore meaning right-angled holes ended up skewed. This was a major issue since in order to get identical spacings on pieces that belong together I would stack them up and drill them in one go. With a skewed drill-hole you can imagine that the holes on the pieces didn't end up having identical spacings. To make up for this I drilled out some holes such that the brass rods, for example, would be parallel which would have been fine for the lowest axis but not for the gantry above it. Also the shaft couplings were terrible.

1st redesign

Initially I wanted to remake the parts from above but after my local hardware store changed management and stopped carrying the 8mm square aluminium profiles and I had to switch to the 10mm ones:

The previous z-axis had turned out pretty good so I decided to keep using it. The coupling between x and z axis shown above, however, didn't work as I had hoped so I scrapped that idea in favor of something more modular. You can barely make it out in this image which is the current state of the machine:

Close-up of partially assembled x-axis stage:

The stage of the x-axis consists of two pieces of acrylic that are held together by four screws. Also attached to these screws are four T aluminium profiles on which the stage slides on the bottom and top. The space between the T profiles/pieces of acrylic can be adjusted by nuts holding the profiles to the acrylic which should allow minimizing the slop of the stage. The same was planned for the z-axis such that this axis assembly could be attached to the machine by mounting the z stage to the x stage. At this point I thought about the friction between the aluminium profiles but decided to just see how it worked out.

The y-axis was built in such a way that the y stage is aligned in x direction to only one of the rails while the other rail only provides stability in z direction and has a pulley that pushes it against the first y rail.

The 'good' thing about this 'design' was that the distance between the rails and the threaded rod driving the stages was adjustable. By fastening the nuts at the ends of the threaded rods going through the rails they are clamped in place. Also the shaft couplings worked better.

The problem, however, was that I was unable to build the x stage precise enough. It kind of worked but was just frustrating so I stopped and thought about how I could improve my fabrication processes.

Redesign 1.1

I initially went with the 'aluminium sliding on aluminium' option because the only alternative I saw at the time was to add a whole bunch of bearings, for the x stage alone 10-12. Only the other day I had the idea of turning the rails of the x (and z) axis by 45° and use v-groove bearings (4 per stage) that should stabilize the stage(s) in the remaining two directions. Something like this:

If you're familiar with my 'work' you probably know that I went straight to eBay and ordered some bearings, so unfortunately I'll have to wait for those to arrive to see if this idea is any good.


Update 2015-11-23: Added an image to redesign 1.1


Update 2016-03-03:

Redesign 1.1 cont'd

Well, those v-groove bearings I ordered back in November/December never arrived so I had to reorder them. To my surprise the new ones actually shipped quite quickly so now I can finally do some more work on the machine.

Basically I implemented my ideas outlined above and in the spindle log so here's an image of the current state of the machine:

The main change is the new coupling of the z-axis to the x-stage:

The left image shows the coupling in the z-axis and the right image shows it (and the z-axis) in the x-axis.

The coupling is still missing something to be driven by the threaded rods but the general concept seems to work. As I mentioned above the x-axis rails are simply clamped between the vertical posts of the gantry by threaded rods running through the posts and the rails which allows the rails to be adjusted. But since the spacing of the z-axis rails is adapted from the mounting holes on the stepper motor adjustments is not possible here. This makes the z-axis quite loose and I still have to come up with a solution to this.

Initially I planned on having the bearings between two pieces of acrylic and maybe even a third one in the middle, as seen in the 3D model above. However, the space between the spindle mounting plate and the z-axis bearings was too small to fit one of those pieces so in order to test this concept I just assembled it this way for now. Also these M4 screws were too short to fit another acrylic piece on the back of the coupling (right image) but that should be easily fixed.

Discussions

Eric Hertz wrote 03/04/2016 at 04:21 point

Aws!

  Are you sure? yes | no