Author Topic: 0.001" end mill  (Read 2943 times)

Offline steamer

  • Global Moderator
  • Full Member
  • *****
  • Posts: 10162
  • Central Massachusetts, USA
Re: 0.001" end mill
« Reply #45 on: February 14, 2019, 10:47:08 PM »
My understanding is that to get the lowest possible runout, the spindle needs to have air bearings for both axial and length positions.
The most accurate collets I know of, are hydraulic collets that distort to hold the cutter.
Grade 3 ceramic balls are to or better than 0.00008mm for roundness and size variation from the smallest to the largest. Places like Cerbec can supply balls tested to 0.000026mm for roundness and variation at a cost . There may be better and more accurate balls, but then the housing and the outer and inner races have to be made super accurate as well. A lot of steels are just not stable enough to make things from to be that accurate. Which is the reason for testing of gauge blocks and micrometers on a yearly or so basis.  I know that the concentricity of the spindle bearings in  the cylindrical grinder is 0.0003mm aproximate , it may be 0.0004mm but is less than 0.0005mm total run out. The housing bearing is ground in situation on the work head, and requires to be returned for the factory to refurbish it.
I would think that a concentricity under 0.0005mm is about all you could ever expect.
When you get to measuring things less than 0.005mm , it starts to get real expensive and takes time for all sorts of reasons. But mostly because of the materials stability and the temperature it is being processed at.
Neil

Yes...All true.   The Tormach is a machine built to a price point, unlike the Diamond Turning lathes I used to work on for Moore tool and the precision grinders for Heald some 30 years ago.  Its a duplex pair angular contact bearing spindle...Probably Grade ABEC 5 or maybe ABEC 7...but I'm guessing ABEC 5.   To get the cutter run out down to the lowest point, and to get around stick slip from the dovetail ways, I had to pull some tricks out of the bag.   I'll post up my results sometime tonight....I and the machine did far better than I expected frankly! and although I don't think we hit the mark..we got closer than I expected by a long shot!....

....but it is Valentines day....and Mama wants some attention..20 years this year....and it's not the years...it's the mileage....   8)

« Last Edit: February 14, 2019, 11:51:06 PM by steamer »
"Mister M'Andrew, don't you think steam spoils romance at sea?"
Damned ijjit!

Offline Tennessee Whiskey

  • Full Member
  • *****
  • Posts: 3770
  • Springfield, Tennessee. USA
Re: 0.001" end mill
« Reply #46 on: February 14, 2019, 10:53:07 PM »
Good Lord, thatís four times as long as Iíve been putting up with you: sheís got to be a good woman  :lolb:

Cletus

Offline steamer

  • Global Moderator
  • Full Member
  • *****
  • Posts: 10162
  • Central Massachusetts, USA
Re: 0.001" end mill
« Reply #47 on: February 14, 2019, 11:39:56 PM »
I've attached a picture of the test cut I did with a 0.001" diameter ( 25 micron) end mill.   I think the boss wanted to give me something fun to do and perhaps he wanted to see what I'd do with it....

There was no doubt in my mind this machine wasn't going to deliver all we wanted from this, but he asked if I would try...so I did...very hard actually.

The machine is a Tormach PCNC -440 vertical mill.   About 1.5 HP....smaller than a bridgeport, and of Chinese Origin.
https://www.tormach.com/pcnc-440/?utm_source=adwords&utm_term=tormach&utm_medium=1530712282&gclid=EAIaIQobChMIv66coam84AIVj6GzCh3fmAa3EAAYASACEgJdAPD_BwE

To see if the geometry we were after was even close to being attainable, I configured a test that would challenge all three axes.   A simple profile pocket.    It would show all three axis errors and not take long to cut..   I have things to do, so I didn't want to spend a lot of time programming this.   As it was it was programmed with the machine conversational language.  Easy peasy.

The pocket shows a center square that is supposed to be 1mm square.  The slotted "Mote"  all around are supposed to be 0.050 mm wide, and 0.025 mm deep.   The program called for 0.006 mm DOC  so several passes were required to get to full depth.

The machine spindle is a grease pack angular contact affair.  As such it's limited to 10,000 rpm maximum.    That is extremely slow for a cutter of this size....but it is what it is.

I began by seeing what kind of run out would the shank of the tool have in an ER 25 collet holder.   These holders are Bespoke Tormach quick change items.    First test showed 25 micron run out!   I think rotated the holder in the spindle until I found a minimum runout of 0.007 mm.  I marked the spindle and holder with a sharpy.   If I installed the holder in that position, the runout stayed right there.    Now 0.007mm runout  ( 0.00025") isn't very good with a 25 micron end mill......but ....it is what it is.

Next I did some checks on the repeatability of the axes, specifically the Z axis as I felt that errors there were going to break end mills....and it was a good thing I did.
the best I could manage was 0.050 mm repeatability.    That's terrible..even with "full size" end mills....so I decided to try giving the machine a few pumps of oil from it's one shot oiler while I did the measuring.    That resulted in 0.005 mm repeatability!!!  Much better!!!!   Uses a LOT of oil to do this......but .....it is what it is

Next I wrote a "do nothing" program that had the tool plane 4" above the table, and just let it run for 10 minutes at speed..to get everything up to temperature.  that would have affects on everything , so I figured I would set home and zero's after this.  and then cut.

Then I brought the tool holder and actual tool over to our OGP optical comparator, to get a tool height....cause you can't see the ))@(()#  thing!!!    It maxed the machine out, but we got a number   68.332mm.

Back to the machine, I had nearly everything I needed.   Tool length, diameter.  Now I needed Z zero.   The position of the work with respect to the tool....not an easy thing to figure if you can't see it and cant' touch it!   I set up my USB video microscope and put a gage block on the part to give me room to get the camera in.  I offset the tool by the 3 " gage block, and compared the images....Now.   I can see everything, but paralax was killing me.   At some point, you take your best guess and go with it!  ( see attached picture   at 225 X maginfication...)    But it is what it is

OK set up an air jet on the cut zone.   

Now....hit cycle start while pumping the 1 shot pump until Z stopped.   I then watched the G code scroll and when it got to a z axis change, I would pump some more!.

Interestingly, the machine moved 0.005 mm per step and not 0.006 as instructed....

In any case, on the last cut the pump was pulling air!...so the last cut was different than the previous...somewhere in the the end mill broke..I think later than earlier..but it's hard to say.

It was finishing the outside ring when it broke...the remnant of the broken endmill is still there in the part...that was when the oil supply dropped, so I'm calling it the cause.

You will notice the slots are a good deal wider than 0.05.   Between runout, and stick slip in the X and Y axis, the width got wider.    ..

My CEO was impressed!....and that was worth it....but I'm glad it's done...on to more productive endeavors!!!

The last couple of pictures show the glue chuck I made to hold the lexan, and the "chicklet" cut out of the picture frame with the little feature in it, and the 1 shop pump with the feed lines full of air...

Dave
"Mister M'Andrew, don't you think steam spoils romance at sea?"
Damned ijjit!

Offline steamer

  • Global Moderator
  • Full Member
  • *****
  • Posts: 10162
  • Central Massachusetts, USA
Re: 0.001" end mill
« Reply #48 on: February 14, 2019, 11:46:05 PM »
One last picture of the tool in front of a 3" gage block...trying to compare the two....it got a little better...but not a lot better .....

Dave
"Mister M'Andrew, don't you think steam spoils romance at sea?"
Damned ijjit!

Online crueby

  • Full Member
  • *****
  • Posts: 8641
  • Rochester NY
Re: 0.001" end mill
« Reply #49 on: February 15, 2019, 01:12:44 AM »
Cameras have a hard time focussing on a cutter that is less than a pixel across! Wow.

Offline steamer

  • Global Moderator
  • Full Member
  • *****
  • Posts: 10162
  • Central Massachusetts, USA
Re: 0.001" end mill
« Reply #50 on: February 15, 2019, 01:18:47 AM »
Cameras have a hard time focussing on a cutter that is less than a pixel across! Wow.

Well if you give it a chance...its not that bad   the paralax error was killing me
"Mister M'Andrew, don't you think steam spoils romance at sea?"
Damned ijjit!

Offline dieselpilot

  • Full Member
  • ****
  • Posts: 277
Re: 0.001" end mill
« Reply #51 on: February 15, 2019, 01:48:45 AM »
That's not a terrible result.

but maybe one of these would fair better if they want to see much of that kind of work.

https://www.ebay.com/itm/202594606134

Offline steamer

  • Global Moderator
  • Full Member
  • *****
  • Posts: 10162
  • Central Massachusetts, USA
Re: 0.001" end mill
« Reply #52 on: February 15, 2019, 02:03:44 AM »
Couldnt agree more.   They had a drill press and they gave me a few bucks to put a shop together....thats what you get for a few bucks....
"Mister M'Andrew, don't you think steam spoils romance at sea?"
Damned ijjit!

Offline 10KPete

  • Full Member
  • ****
  • Posts: 1560
  • Nordland, WA, USA
Re: 0.001" end mill
« Reply #53 on: February 15, 2019, 02:53:27 AM »
A most interesting experiment, Dave! I know that fluid logic stuff can be very tiny but that's just pushing the edges of mechanical machining right to the extreme!

Maybe get an electric pump for the oil?   :ROFL:

Pete
Craftsman, Tinkerer, Curious Person.
Retired, finally!
SB 10K lathe, Benchmaster mill. And stuff.

Offline steamer

  • Global Moderator
  • Full Member
  • *****
  • Posts: 10162
  • Central Massachusetts, USA
Re: 0.001" end mill
« Reply #54 on: February 15, 2019, 07:11:52 AM »
A most interesting experiment, Dave! I know that fluid logic stuff can be very tiny but that's just pushing the edges of mechanical machining right to the extreme!

Maybe get an electric pump for the oil?   :ROFL:

Pete
Nah    get the right machine, or make the part another way...both are options....but in R&D...sometimes people just need to be shown.
"Mister M'Andrew, don't you think steam spoils romance at sea?"
Damned ijjit!

Online crueby

  • Full Member
  • *****
  • Posts: 8641
  • Rochester NY
Re: 0.001" end mill
« Reply #55 on: February 15, 2019, 12:20:22 PM »
We used to do passages for tiny fluidics as parts of our chip manufacturing process, could go very tiny, but it did require a silicon wafer facility. Doing the same with a mill is impressive.