Controlling UHMW swarf

[MechanoMan]

I tried lathing a piece of UHMW, 1.5" OD down to as low as 0.75" on the Sherline.

Now getting smooth, coiling swarf wasn't the problem. The thing is that it's all wrapped around the stock and clogged the job up unless I kept stopping it and manually cutting it free. If I didn't do that it'd bulk it up enough to start rubbing into the crosslide and get pushed off the end into the tailstock where it'd half be wrapped on the spinning work and half on the fixed tailstock so it would bind, snap, melt, and load down the motor.

Well stopping often to cut it off worked of course, but if I'm gonna CNC this job- which I ultimately want to do- then this is not gonna fly. It won't be automated really.

Is there a way to stop this? Should I try to take shallower cuts? I often used the full width of the ground cutting edge, like 1/4" maybe, but the overall problem didn't seem lessened much with shallower cuts. The swarf was easier to cut off but I'd have to do it more often with shallower cuts on more passes.

I'm pretty sure by tool is not ground as sharp as it could be, but I don't see what difference it makes, the swarf won't break or leave in a different direction (where it won't wrap around the stock) due to that. Would adjusting the tool angle make it leave differently?

[blades]

I think that's pretty much an inherent problem with UHMW. It machines great, but it is known for doing what you are describing. I've seen this many times, even with very light cuts. It's almost like hair wrapped up on the workpiece.
You may have to try some creative programming, such as starting the cut towards the chuck and cutting away from the chuck. Might also have to put a stop point in the program (if you're running some sort of production) so you can manually clear the mess.

[ibuildstuff4u]

You can also try changing the feed rate as well as the spindle speed. Some times this can help. Also it sounds like your using high speed steel tooling that you ground your self? If so you may need to play around with the geometry of your tool and try to make a relief in the top to cause the chips to break instead of coiling around.

Good luck~

Dale P.

[HuFlungDung]

Depending on the shape of the part, you might try milling a flat, a groove or a slitting saw cut on one side of the part. Don't go to full depth, of course, with this interruption, but it should break the chip

It may be worth experimenting with something like Iscar cutgrip tooling, and plunge rough using radial infeed. It is tough to predict what the chips will do, but you can plunge fast, and for the final cut, you can feed at a very coarse rate and still get smooth flat surfaces with this style of tooling. What I like about the cutgrip is that there is lots of clearance around the toolpoint, so there is somewhat less piling up of chips being wedged in behind the tool.

[Geof]

If/When you start doing it on a CNC machine you can use G74 and G75 to peck your way through and breake the chips.

[MechanoMan]

I'm actually considering doing it on the rotary table on the Taig mill instead and sick the 1/4" micrograin carbide endmill on it, just to avoid the uber-swarf. Does that still sound reasonable? I've got to chuck it in the Sherline to do a center bore first. Re-chucking it in the mill is kind of a PITA, especially since it's pretty low tolerance on end diameter so I've got to setup that table carefully so the rotary axis is totally parallel to X, but it sounds like it'll do a better job.

play around with the geometry of your tool and try to make a relief in the top to cause the chips to break instead of coiling around.

Depending on the shape of the part, you might try milling a flat, a groove or a slitting saw cut on one side of the part. Don't go to full depth, of course, with this interruption, but it should break the chip

Pictures/links/whatever?

Actually I'm using some tools that the previous owner had made up. My grinder was taken apart for other reasons, I just got it back together but the grinding surface evenness doesn't look "great" and I'll probably get an uneven tool edge as a result. I figured I just need a dressing stone, like the one I use with my Dremel grinding stones but larger, but neither Home Depot nor Lowes carries that, Harbor Freight does but only by mail, not in stores.

There's that dressing WHEEL thing, but that's more for picking out bits of aluminum gumming up the wheel, isn't it? It's not gonna be able to reshape the wheel flat, can it?

[MechanoMan]

Originally Posted by Geof If/When you start doing it on a CNC machine you can use G74 and G75 to peck your way through and breake the chips.

What motion would this perform on a lathe?
I could see if we stopped progress periodically, the swarf would get sucked between the work and the tool and would hopefully break. But that UHMW stuff is tough! And the finish on the work would be affected. I'd say do a roughing and finishing pass, but from what I saw even that fine "hair" will tangle on the work.

[Geof]

On both G74 and G75 the tool backs up a little but to break the chip. You control the amount of cut depth between back-ups and also the amount the tool retracts. You are correct that it affects the finish and you have to do a straight finish cut. You may never be abe to get around the problem of the swatf winding on the part during the finish cut.

[blades]

Problem is, UHMW doesn't produce chips that can be broken off with tool geometry different feed rates, etc. It's just too soft and tough at the same time. It's simply the nature of the beast when machining UHMW.

HuFlungDung has suggested what I would consider the only really sure-fire way to control the chips, which is to create an interrupted cut. Only down-side is that it creates an additional machining operation on the part. Other than that, Geof's suggestion will certainly control the length of the mess that gets wound up on the work. Not the amount, just the length of the individual chips. Or like I already suggested, try cutting it from the chuck out so the chips are forced off the part and not wound up into the chuck, or simply put a couple of stops in the program so you can manually clear the mess.