The difference between machining accuracy and positioning accuracy - Page 2
Close
Login to Your Account
Page 2 of 3 FirstFirst 123 LastLast
Results 21 to 40 of 41
  1. #21
    Join Date
    Aug 2006
    Location
    Wisconsin
    Posts
    1,471
    Post Thanks / Like
    Likes (Given)
    618
    Likes (Received)
    546

    Default

    I meant to say three to five microns. .003-.005mm in my first post.

  2. #22
    Join Date
    Jun 2017
    Country
    UNITED STATES
    State/Province
    Illinois
    Posts
    29
    Post Thanks / Like
    Likes (Given)
    1
    Likes (Received)
    2

    Default

    Quote Originally Posted by 5 axis Fidia guy View Post
    This will be an interesting read. I will tell you this. We just purchased a Hermle, C42 5x machine. The utmost in positioning accuracy. When I check the kenematics with the ball over the entire 800mm range of travel, I get an error or around .003-.005 microns, very impressive to say the least. BUT, trying to do a simultanious cut and hold tight tolerances, that is another story. Machining accuracy consists of, programming software, toolpath tolerance, tool tolerance, holder accuracy, shop temperature, tool runout, the list goes on and on. Take a ball endmill with a .0004 tolerance, then add a .0002 toolpath tolerance, then add holder runout of .0001 or so, now the tool is hanging out a way and you have some deflection, the material is heat treated so add some more deflection, now warm up the spindle of the machine a while, now the shop heats up.. you are out .0005-.0006 before you even hit a button. IF I am very, very careful, I can hit a .0005 range, +-.00025. But that is taking everything into account. To just set up a tool, run it thru the laser and hit the button, maybe .0007 if your lucky. And this is with premium tolerance tooling.

    I am very curious as to how 5 axis fidia guy is checking across that 800mm.....I don't believe it, SORRY!? I've been around Moore Nanotecnology, Precitech and Toyoda machines turning lens and other components using diamond tools, and I don't believe that .003/.005 micron for a millisecond.
    That machine, and any other 5-Axis is not as accurate producing a "machined". The 5 axis trunion with rotary only adds to amount of variables that make for lost tolerances. I was at a customer in Texas who had a Hermle, and was told that they dont use tooling over 1" cutting the turbines/impellers....the 1.500" tool I gave them ran 3 minutes because of the lack of rigidity and material they were cutting offered a large amount of resistance and deflection. What the spec's say and what you get are two different things.

    Now the rest of his information regarding the tool, holder material and other variables do stand up to what you will get when you machine something.
    It's like saying this reamer has a +.0002/-0.0 tolerance, but getting that to repeat in the machine, on the material you have is another story.....I've met dozens of people who never ever have indicated a reamer in a holder or used a floating holder to help with that.


    OOOOHHHHHH, 3-5 microns....that I think I can believe, not .003-.005 microns. I didn't read on to see that one!!!!
    LOL!!!!!

  3. #23
    Join Date
    Jun 2017
    Country
    UNITED STATES
    State/Province
    Illinois
    Posts
    29
    Post Thanks / Like
    Likes (Given)
    1
    Likes (Received)
    2

    Default

    any machine will move the .0001" it says in the manual, and And being the skeptic I am, I don't believe the control on the machine to "eliminate" the backlash or error in its movement is going to be as accurate as a third party system used to measure it.

  4. #24
    Join Date
    Nov 2013
    Country
    UNITED STATES
    State/Province
    Utah
    Posts
    2,047
    Post Thanks / Like
    Likes (Given)
    573
    Likes (Received)
    1064

    Default

    For the sake of the advertisement they are meant contextually as the same thing. Positioning would be the ability to travel away from a point and return to that same point, basically. Using "Machining accuracy" as a hot word or flash word--to sell machinery. As it were, "Positioning accuracy" and "Machine accuracy" being seperate would be redundant, let alone stupid.

    Like saying "if it's not making chips it repeats really good, but when your in the cut then repetition falls of $#^%*R^&" Dumb.

    R

    ON EDIT, The actual sales pitch would be the "positioning accuracy" not the "Machining accuracy" same point though.

  5. #25
    Join Date
    May 2007
    Location
    Central Texas, West and North of Austin
    Posts
    1,628
    Post Thanks / Like
    Likes (Given)
    178
    Likes (Received)
    523

    Default

    I have NOTHING to add. This has been one of the more educational threads I have followed, and has reinforced my opinion that ,even after 35+ years in this, there is a whole lotta stuff regarding it that I do not know yet. And why I follow threads that I know very little about on this Forum. I genuflect at the feet of some very smart sumbitches here abouts. You guys amaze me.

  6. #26
    Join Date
    Mar 2011
    Location
    NY USA
    Posts
    776
    Post Thanks / Like
    Likes (Given)
    142
    Likes (Received)
    491

    Default

    One of my nerdy habits is calling up the thermal comp screen on the Okuma and watching as the control actively comps out thermal growth in the various parts of the machine that it monitors. It's neat to watch it change things by parts of a micron. I didn't really know if it worked or if it was just something shiny for nerds like me to watch, but this past week I had to shoot some shallow bores in aluminum, half thou tolerance so I calibrated up the fancy bore gage and 100% checked them. Granted 6061 isn't exactly hard work for a 560 but I was floored by the consistency. The digital bore gage calibrated to a master ring, has a resolution of .00005" and it never changed. Now maybe those interpolated bores were wandering for the 50 millionths and the gage wasn't showing it..or maybe they weren't but still...1.3 micron over a 5 hour run at aluminum feeds and speeds is pretty damn impressive. This is only 3 ax, so I get that 5ax is exponentially harder to pull off but it's pretty amazing what they can do with real machining accuracy these days.

  7. #27
    Join Date
    May 2014
    Location
    Ann Arbor, MI
    Posts
    5
    Post Thanks / Like
    Likes (Given)
    0
    Likes (Received)
    0

    Default

    Somewhat OT, but what the heck...

    The ATLAS experiment at the Large Hadron Collider is about the size of my house, and there are many, many thousands of fine wires in the detectors that track the paths of the collision products. To turn the raw data into useful information they need to know the position of every wire to within something like 20 microns. There are lasers and retroreflectors all of the instrument so they can measure everything interferometrically - they don't just worry about thermal expansion, but also tidal effects. I doubt that any 5-axis machines pay attention to the phase of the moon...

  8. #28
    Join Date
    May 2009
    Country
    UNITED STATES
    State/Province
    Wyoming
    Posts
    2,842
    Post Thanks / Like
    Likes (Given)
    6165
    Likes (Received)
    4319

    Default

    A given machine today will very likely repeat position to within .0001, but that's only half the battle. Imagine a half-inch end mill pocketing 3 inches deep. How square is the wall going to be, assuming you don't have time to sneak up on it for two hours? The results in a case like that simply cannot be guaranteed and have a lot more to do with operator experience and the stiffness of the tool and setup than the accuracy of the ballscrews...

  9. #29
    Join Date
    Nov 2013
    Country
    UNITED STATES
    State/Province
    Utah
    Posts
    2,047
    Post Thanks / Like
    Likes (Given)
    573
    Likes (Received)
    1064

    Default

    Quote Originally Posted by khillig View Post
    Somewhat OT, but what the heck...

    The ATLAS experiment at the Large Hadron Collider is about the size of my house, and there are many, many thousands of fine wires in the detectors that track the paths of the collision products. To turn the raw data into useful information they need to know the position of every wire to within something like 20 microns. There are lasers and retroreflectors all of the instrument so they can measure everything interferometrically - they don't just worry about thermal expansion, but also tidal effects. I doubt that any 5-axis machines pay attention to the phase of the moon...
    OT?? meaning Other Team? Or maybe O.S.--Other Sport

  10. #30
    Join Date
    Oct 2013
    Location
    Rochester, NY, USA
    Posts
    93
    Post Thanks / Like
    Likes (Given)
    4
    Likes (Received)
    30

    Default

    It sounds like they really mean:
    Positioning repeatability
    and
    Position accuracy

    The stage may move and come back to the same position within 0.0001" (repeatability) but may travel crooked or curved so the accuracy (of the expected straight line travel) is much worse.

  11. #31
    Join Date
    Feb 2004
    Location
    Napa, CA
    Posts
    2,562
    Post Thanks / Like
    Likes (Given)
    43
    Likes (Received)
    673

    Default

    Quote Originally Posted by john-san View Post
    any machine will move the .0001" it says in the manual, and And being the skeptic I am, I don't believe the control on the machine to "eliminate" the backlash or error in its movement is going to be as accurate as a third party system used to measure it.
    On machines with direct position measurement (like most Hermle) the position will be as accurate at the measurement system, which can be much better than 0.0001". Even if the belts are loose and the screws are screwed. The position servo loop will compensate.

    Quote Originally Posted by greif1 View Post
    It sounds like they really mean:
    Positioning repeatability
    and
    Position accuracy

    The stage may move and come back to the same position within 0.0001" (repeatability) but may travel crooked or curved so the accuracy (of the expected straight line travel) is much worse.
    Interpolation accuracy or motion accuracy is a whole different thing from either position accuracy or repeatability. The latter two are typically measured statically, the former dynamically. A ball bar test does measure interpolation accuracy for some circumstances. Interpolation accuracy depends on matching the dynamics of motion between the axis exactly, or moving so slowly that the position servo loop can compensate for mismatched dynamics.

  12. #32
    Join Date
    Feb 2017
    Country
    UNITED STATES
    State/Province
    Tennessee
    Posts
    6
    Post Thanks / Like
    Likes (Given)
    2
    Likes (Received)
    0

    Default

    Quote Originally Posted by 5 axis Fidia guy View Post
    No idea who that is. But I will take a picture of it next time I do it.
    picture it in a machining cycle with tooling//.

  13. #33
    Join Date
    Nov 2013
    Country
    UNITED STATES
    State/Province
    Utah
    Posts
    2,047
    Post Thanks / Like
    Likes (Given)
    573
    Likes (Received)
    1064

    Default

    Quote Originally Posted by one horse View Post
    picture it in a machining cycle with tooling//.
    ?? Gmatov in a Machining cycle? That would be interesting, most cycles are too new to match that personality. Maybe the wrong cycle, like using G84 before G83, very explosive.

    R

  14. #34
    Join Date
    Sep 2012
    Country
    SWEDEN
    Posts
    60
    Post Thanks / Like
    Likes (Given)
    22
    Likes (Received)
    16

    Default

    Good question.

    I have driven machines with machine accuracy claimed around or less than 0.0002'' (0.005mm) when new, and been running parts around that tolerance. Good all the time? nope! Timing, pace of working, not talk about temperature, day of time, or tools wear out, etc made measurements go just mad. The tolerances I had to do was just that, 0.005mm. 30% of the workpieces was scrap, but still it was manageable and somewhat doable. Me sux? I buy that too.
    So from my point of view the machine was able to do (my best guess) half of claimed 'machine accuracy'. So conclusion machining accuracy and machine repeateably is NOT the same thing. I have a good advice tho. Dont go there, that land of precision is evil. Even a bad look makes it go haywire and kick your butt sore. =)

  15. #35
    Join Date
    Jan 2011
    Location
    telford, pa
    Posts
    353
    Post Thanks / Like
    Likes (Given)
    31
    Likes (Received)
    102

    Default

    Okuma!! Repeat all day long!

  16. #36
    Join Date
    May 2005
    Location
    CA
    Posts
    886
    Post Thanks / Like
    Likes (Given)
    1
    Likes (Received)
    146

    Default

    I don't believe there is much new in physics (which is ultimately what determines how consistently and accurately a parts can be machined).

    Spindle length comp for temp
    Coolant temperature
    Shop temperature
    Overall machine temperature (Mikrons will run 3 chillers -- coolant, spindle, and ways (linear drives)
    Stiffness of holder
    Runout of holder (*everything* has runout)
    Stiffness of spindle
    Stiffness of cutting tool
    Or how about the natural frequency of each of those pieces... is there any undesirable "stacking" of harmonics?
    etc. The list goes on and on...

    And then finally... on a particular machine (whether a Bridgeport, an Okuma, a Haas, or in this case a C-42): how far is it from the part... ALL THE WAY back around to the cutting tool? e.g. on a C-42: Part, to table, through C-axis, through B-axis trunnion, into the main casting of the machine (much of it "poured" on a Hermle), then through the guideways for the X axis and Y axis, through the guideways of the Z axis, then through all the bearings and back down the spindle to the cutting tool. It's a *long* ways from part to cutter! (though possibly a shorter distance and stiffer on a Hermle than most any other 5x machine of the same capacity).

    The fact that these machines can even *position* (much less cut!) to microns is just amazing.


    PM
    Last edited by precisionmetal; 02-11-2018 at 01:14 AM. Reason: corrected spelling

  17. #37
    Join Date
    Jul 2015
    Country
    UNITED STATES
    State/Province
    North Carolina
    Posts
    61
    Post Thanks / Like
    Likes (Given)
    15
    Likes (Received)
    15

    Default

    I think it's unambiguous in this context, but I don't doubt their are people available to misunderstand it. What's the machine do? It machines. How? By positioning itself. Since there is a direct relationship between machine positioning and machining accuracy I think the claim is sound.

    However, I did have a supervisor that ran everything in metric because he thought the positioning was more accurate. I disagreed; I think the higher resolution of the adjustments allowed the operator to be more "accurate", but the machine was most likely indifferent to SAE vs Metric.

    He also took a manufactures claim that a certain end mill could be used for roughing and finishing and eliminated the finishing tool we were using. And basically ended up going through a lot of tooling unnecessarily because the accuracy suffered quickly from tool wear.

    2 cents.

  18. #38
    Join Date
    Nov 2013
    Country
    UNITED STATES
    State/Province
    Utah
    Posts
    2,047
    Post Thanks / Like
    Likes (Given)
    573
    Likes (Received)
    1064

    Default

    Quote Originally Posted by Cheshire View Post
    I think it's unambiguous in this context, but I don't doubt their are people available to misunderstand it.
    Ever considered, not saying what you think it may not mean, or don't mean in double negatives?

    R

  19. Likes Cheshire liked this post
  20. #39
    Join Date
    Jul 2015
    Country
    UNITED STATES
    State/Province
    North Carolina
    Posts
    61
    Post Thanks / Like
    Likes (Given)
    15
    Likes (Received)
    15

    Default

    Quote Originally Posted by litlerob1 View Post
    Ever considered, not saying what you think it may not mean, or don't mean in double negatives?

    R
    I haven't not done that or not.

  21. Likes litlerob1 liked this post
  22. #40
    Join Date
    Jan 2003
    Location
    Pacific Northwest
    Posts
    1,260
    Post Thanks / Like
    Likes (Given)
    0
    Likes (Received)
    64

    Default

    My Haas TM1 has small issues. When the spindle heats up, the Y grows - and the Z grows down. When new it was about -.0005 in Y and -.0008 in Z. Now that the spindle has (moved the grease to where it does not heat up as much?), these numbers are much less. There is a little wobble in X when moving in Y, about .0001. I notice this on the sides of a surface. A slightly bent Y screw? This is only noticeable when a Y surface is swiped on a surface plate and examined as it matches the pitch of the Y screw.

    Still like the TM1.


Tags for this Thread

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •