Centrifugal disc finish machine

What's the advantage of the rotary unit over a traditional vibratory option? It appears to generate the same tumbling action, and other than component longevity I'm not sure I see the purpose of the higher-complexity option- it's much easier to put a bucket on a crankshaft than it is to make a relatively high tolerance rotary union of two matching compound curved surfaces.

 

Ok, just looked into it a little bit. Not sure the finishes are necessarily any better, depending on media, but the "high-energy" thing for very short cycle times is intriguing. Looks a bit messier than vibratory, but we deal with that with resin printers vs FDM printers and live to tell the tale just fine.

So basically you have to build a spindle that can take fairly high speeds, I'm guessing they run at about 3000-5000rpm (can't find a definitive speed range from manufacturers on my cursory search)- a piece of chrome shaft and a couple pillow block bearings *might* work, depending on tolerances and ratings, and have that belted to a motor at the bottom (none of them seem to be direct drive, for some reason?) and a flange to take the platter at the top. The frame to hold the bearings and the top barrel section has to be one part, both very accurate and with zero flex; maybe a weldment, if you built a jig to hold all the pieces of steel tubing. Looks like commercial machines use castings, which is obviously pretty difficult for DIY.

The top barrel seems to just be a urethane liner on a sheet steel ring, which is relatively straightforward- if I were doing the "jagged" liner with the teeth around the outside, I'd print a bunch of pieces to glue together and make a silicone mold, which is the easiest way of doing accurate and repeatable urethane parts. I'd do the same thing for the curved bottom piece with some low-profile "vanes" to help move media to the outside.

With a slightly flexible urethane choice, you may have a little more leeway on the mating surfaces- it might be possible to continuously lubricate the interface between the two urethane parts with the tumbling fluid, I'm not sure though. I'd guess it would probably be possible to add some kind of removable and adjustable lathe-like cutter on both the edge of the rotating platter and the fixed frame, so that you can cut both urethane components to a very precise diameter and finish using the machine itself.

Of course if you were building a very small one, the whole thing could be 3D printed, but I don't think it would be as effective as larger diameters and would probably need a much higher speed, which is itself a rabbit hole of tolerances, bearings, heat, radial forces, etc.

Then the whole thing needs to be able to pivot- barrel, platter, shaft, motor and all- to dump out everything to be screened, doesn't seem as convenient to just pick parts out of while running like a vibratory unit. That's a pretty straightforward part though, just another couple of pillow block bearings and a simple frame, could even be wood as long as it's fairly accurate and solid/heavy enough to take the running forces.

Personally I'd probably try to run it with a VFD and a 3-phase 3600rpm motor, maybe 2-3HP should be sufficient? That would be perfect for the soft starts and low-speed operations like insert trimming.

 

That seems very slow for a "high energy" machine, even at a foot or two diameter, but at least it explains the gearing/belting then! If you used a 1700rpm motor you could probably just timing belt at a straight 8:1 reduction and use a VFD for the rest of it (I assume the potentiometer is connected to a VFD, somewhere, because basic chopper controls cut a lot of power away from the motor).

(At only 100-300rpm, that would definitely work- hydraulic tube would work as well for a hollow version)

They flow the fluid backwards against spring pressure and use a hollow spindle shaft instead as the drain. This allows both positive pressure against the abrasive particles and a continuous hydrostatic bearing surface for the disc itself. It makes sense and would be relatively straightforward to DIY, but it would need some experimentation with the amount of spring pressure as well as the pump flow and pressure ratings.