Delrin Surface Finish (again)

I know that Delrin is sh--ty material for getting a nice finish. Found a discussion from seven years back regarding the problem:

http://www.practicalmachinist.com/vb/general/shining-up-delrin-205013/

Maybe it's a fruitless effort but I wonder if - since then - anyone has devised a process to improve surface finish, such as on the following part, particularly the tapered length:

Three parts, about five inches long with a 4 degree taper starting at .50 diameter at tailstock end (supported with live center) increasing and terminating at around .75 diameter with a fairly large radius fillet blending to the one inch diameter end (held in a collet).

Tolerances not critical +/- .010 but appearance is important. (FWIW there will be three or four subsequent operations in a vertical mill but the resulting surfaces will not be visible.)

Ignorantly, I tried to improve the turned surface on the first part with a couple of grades of paper and made the finish worse. Draw filing helped to a certain extent but basically, what I turned was the best that I got.

I honed and honed the tool and the tapers on the other two parts looked quite a bit better. The job has been put aside for the moment. Fortunately these were rough-turned and there is a bit of allowance on the diameter, left for the magic cutting tool angles, end radius, speed/feed da-da-da-da, if an optimum combination even exists !

Oh yeah, the only lathe with a taper attachment is limited to a top speed of 1200 RPM.

Thanks

I think the only way is to cut with very sharp positive rake tooling. Acetal surface finish will not improve by any abrading process, IME.

Regards.

Mike

Yep, you want sharp positive rake tooling. I always used ground and honed HSS. The tip grind I used and found very good success with was similar to one I use for long reach carbide boring bars. Basically grind to a sharp point and almost a 90° dub on the point to produce a small flat about 1/16" - 3/32" wide. Align the tool so the small flat produced by the dub is almost but not quite parallel to the central axis of the work, so it acts like a wiper. You can get the feed up pretty good with this type of grind (with tweaks to the flat width and angle to the work) and still get a great finish. Sort of similar to broadnosing but a bit different.

No radius on tool ?

The "pile-up" of the material when cutting is irritating, as well, especially when nearing a shoulder or other feature that requires a bit of precision. All of that Deltrin sh--t in front of the cutting tool makes me worry (with reason) that my travel indicator (no DRO on this lathe) isn't necessarily telling the truth because of tool deflection !

That is just whining, doubt if anyone has a good idea for preventing the pile-up. I really hate this material, I'd rather do 1018 all day long.

Any straight turn I make on Delrin looks like glass to the naked eye, and I don't think I do anything special.I just use insert tooling with no more than a .010 radius and a feed around .008 IPR. As for shining it up try steel wool.

Dualkit said:

Any straight turn I make on Delrin looks like glass to the naked eye, and I don't think I do anything special.I just use insert tooling with no more than a .010 radius and a feed around .008 IPR. As for shining it up try steel wool.

Click to expand...

Hey Dualkit, tried steel wool. Not disputing your experience but the PM thread pretty much duplicates my observations.

http://www.practicalmachinist.com/vb/general/shining-up-delrin-205013/

Regarding your insert, any details (I'm surprised that you don't use HSS) ? Orientation to the work ? Speed ?

Thanks for the response, it's good to hear that SOMEONE can make it work !

randyc said:

Hey Dualkit, tried steel wool. Not disputing your experience but the PM thread pretty much duplicates my observations.

http://www.practicalmachinist.com/vb/general/shining-up-delrin-205013/

Regarding your insert, any details (I'm surprised that you don't use HSS) ? Orientation to the work ? Speed ?

Thanks for the response, it's good to hear that SOMEONE can make it work !

Click to expand...

How about a blow torch to heat the plastic on the surface?

Sent from my Nexus 10 using Tapatalk

In the thread that I mentioned, some say yea, some say nay about torches. Since I only have three parts and I'm short on material, experimenting is risky. But maybe I could try it on a drop-off of the remaining material - will wait and see other responses. Thanks

I always used a shop vac to suck up chips (chip? Always a long continuous string) as I turned.

randyc said:

Hey Dualkit, tried steel wool. Not disputing your experience but the PM thread pretty much duplicates my observations.

http://www.practicalmachinist.com/vb/general/shining-up-delrin-205013/

Regarding your insert, any details (I'm surprised that you don't use HSS) ? Orientation to the work ? Speed ?

Thanks for the response, it's good to hear that SOMEONE can make it work !

Click to expand...

I will look tomorrow on those inserts. I don't use HSS turn tools on anything. I have used dead sharp brazed carbide also, pretty much everything I have thrown in there works. You are using Delrin, correct? I have ordered Delrin from some places and was given acetal copolymer which doesn't machine as nice. As for steel wool I used a very fine grade that was similar to Brillo pads. Nothing else worked, also maybe my finish was pretty good to begin with.

What exactly is the finish requirement? I mean Profilometer? Or visual? Is a Matte finish okay, so long as it's uniform? Or do you need both shiny and surface finish registration?

I found recently, I can get a pretty nice Matte finish using Window cleaner and red ®Scotchbrite pads with small circular motion. The Ammonia in the window cleaner helps blend tool marks very well and makes it all uniform and probably better for a Profilometer check, BUT NOT shiny Rinse complete with dish soap and warm water to get the Ammonia off.

R

We have gotten good finishes with sharp tools, hand ground high speed or Arno high positive carbide inserts. We always flood with cutting oil using moderate spindle speeds and fine feeds, if we ever got a bad finish it was machining dry, the chips get hot wrap around the work and weld up to the finished surface. It does not happen with cutting oil. Water based coolant could work but we never use the stuff to know for sure.

no need to rough them out. start with 1" material and turn 1 pass (actual better chance of holding tolerances). I use the iscar AS insert with .032tnr ground for aluminum and 1000 rpm with about .003 feed. To fast and the chip just likes to wrap around the part.

WRT chip wrapping, if you really get after it with the roughing pass(es), you can break a chip in acetal. I rough @ .02" feed and leave .003", then the finish pass swarf is usually wispy enough to not mess things up.

Regards.

Mike

I have always used aluminum specific inserts (whatever mcmaster carries) and had good results.

Insert numbers I know have worked in the past the first 3 I pulled out of the drawer were all CNMG431 3 different brands and 3 different coatings, I think they all have .015 radius. I pulled some programs and they had .008 feed, 3,000 rpm on .500 to 1.00 diameters with .1 to .250 DOC.

randyc said:

I know that Delrin is sh--ty material for getting a nice finish.

Click to expand...

Since when?
Delrin is easy to get a nice finish, provided you have the right tooling to work with.

Mtndew said:

Since when?
Delrin is easy to get a nice finish, provided you have the right tooling to work with.

Click to expand...

Apparently all of the other members of PM in the thread that I posted don't know that it's easy. Also you forgot to mention what the "right tooling to work with" happens to be. My request was for suggestions; since you find Delrin easy to work, I'd appreciate specifics like most of the above provided.

I'll have a break in an hour or so and will try some of the previous suggestions on a piece of scrap.

Regarding the quantitative surface finish, as Littlerob1 brought up, there is no specification, the parts need to be visually attractive. At this point, a matte finish would be fine because unless I have a breakthrough in the next couple of hours, I'm going to discard the desire for a semi-gloss finish.

Re-capping the suggestions:

*Although I have seen better results with a finely-honed HSS tool run at very coarse feed, perhaps the tool needs to be even sharper.

* MAYBE * try torch de-fuzzing but this would prob be a last resort (not very controllable process)

*Try the window cleaner idea if the semi-glossy finish continues to evade me

*See if I have an insert sim to CNMG431

*Speed/feed suggestions range from fairly slow to fairly fast, feed suggestions range from very fine to very coarse. The point that slowing down the speed might ease the chip pile-up problem gets my attention. Ditto with roughing at high feed.

*Coolant - I did try cutting oil with no positive results but this was early on so the tool geometry was probably not optimum. (BTW, the tool that I'm now using looks sort of like a wood-turning chisel, .125 radius with generous pos rake and reliefs.)

Cheers
.

Mtndew said:

Since when?
Delrin is easy to get a nice finish, provided you have the right tooling to work with.

Click to expand...

When I saw this thread this morning I was wondering if I was living in an alternate universe, or
the delrin I've machined wasn't delrin.

That's the best thing about delrin, it always looks good. Especially when its black..

You didn't say what color you are running, but if its white, see if your customer would like
black (or just buy a piece and show them).. Black Delrin always looks good.

Edit: tools, SHARP!! The ground and polished upsharp inserts for aluminum work really well.

I've also been lazy and run just whatever happened to be in the machine and it still looked good.

Delrin is so easy to get any finish you want on, I can't even believe this thread exists. Are you sure it's really Delrin. Even the POM subs usually machine fine. Is there something wrong with your machine? Sharp tooling of any type, with a decent radius, should do the trick.