Position Concerns

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Position Concerns

    So I have a part that is standing up so I can check both sides. I have shown a drawing of what I'm working with. The position of this hole is called back to ABC, all holes going thru. I have created a Datum B, SAME HOLE as "B" from one side and "B1" from the other side. Datum B and Datum C and the hole involved is .150" in depth and Datum B is .0001 Perpendicular both ways. Now when I choose B from on side A90B90 and Datum B1 A90B-90. I get 2 totally different results. Which has me concerned. I have Calibrated and re-calibrated with same results. I only bring this up because I have some holes that are not thru and results are dramatically different also. Machine is not even a year old. I have recreated alignments to match each callout. and even have used the current alignment, matching alignment, simply put nothing changes. This concerns me for accuracy. Don Ruggieri your thoughts. That's a .004 difference of the same hole from one angle to another angle.

    Capture2.JPG

    Capture.JPG

    Capture3.JPG

    Could a hole only .150 deep actually be off .004?
    Last edited by KIRBSTER269; 10-18-2019, 03:34 PM.
    (In Memory of my Loving wife, "Ronda" who I lost March 7, 2016. I love you baby.)
    They say "Nobody's Perfect." I must be Nobody.

  • #2
    kirbs stick to programming, your artwork sucks.

    0.093" is a relatively small hole, What probe stylus are you using? is there a chance you are shank-hitting on a rolled burr or something?
    is the part fixtured weally weally well? it's not moving on you at all, is it?
    Is your level plane taken using probe Z axis? Can you try it in X or Y of your probe?

    When you are calibrating your probes:
    Have you checked your cal sphere isn't loose, damaged or dirty?
    Do you have a master probe defined? Checking NO cal sphere didnt move?
    Are you calibrating all angles of your probe(s) at the same time?
    Do you have touchspeed the same as in your routine?

    Can we see your cal results for the applicable probes/angles?

    I've had PH10MQ's die on me in very terrible ways recently, but most of the time physical symptoms would present themselves (failure to clamp or unclamp during wrist rotate, sporadic limp-wrist errors, loud grinding/clicking noises, etc).
    Last edited by louisd; 10-18-2019, 03:48 PM.

    Comment


    • KIRBSTER269
      KIRBSTER269 commented
      Editing a comment
      You know me, clean probe checked all that, always clean the Cal. ball and probe before calibration, nothing loose. I have a HP-S-X1H. Everything is less than a year old, taking all good hits you can see the results of the same hole at the bottom shows the hole the same size from each side. Just wondering if .004 difference from one side to the other of a hole .150 deep. maybe the machinist is over clamping the vice? I want to say no, but maybe you never know. and Results on Calibration is 0.0000 both directions

  • #3
    can you put the part on a layout table and double check the hole positions on each side of the part with a height gage?
    sigpic

    Comment


    • KIRBSTER269
      KIRBSTER269 commented
      Editing a comment
      Done that and it is good, but can't really do that on the holes that don't go thru. Just has me concerned being the exact same hole for datum B and such different results
      Last edited by KIRBSTER269; 10-18-2019, 04:03 PM.

  • #4
    Are you able to validate on another piece of equipment?

    Also to validate your CMM axis, are you able to screw in a standoff in a hole on your plate and rotate to A90B90, line up the tip of your probe as best as you can to the center of the standoff. Lock your Y-axis, rotate to A90B-90 and see if you tip is still aligned to the center of the standoff from the opposite side?

    Comment


    • KIRBSTER269
      KIRBSTER269 commented
      Editing a comment
      running another piece, need a little time for the task but will keep you updated even if it's my mistake somehow. but thanks for the suggestion. processing...........................

  • #5
    False alarm Folks, Part is BENT slightly was oversqueezed, I repeat Part is bent!!!!!!, added a Flatness, showed almost .0025 pretty bad for a small part, Ran second Part within .0002 from one side to the other. Amazing Thanks louisd mckenzie bfire85


    Also to add if you noticed in my Picture above PC-DMIS reports the worst case, both shows START POINT, that shows that the worst part being the TOP of each hole, meaning the bottom of each hole is better???? of the same exact hole
    Last edited by KIRBSTER269; 10-18-2019, 04:45 PM.
    (In Memory of my Loving wife, "Ronda" who I lost March 7, 2016. I love you baby.)
    They say "Nobody's Perfect." I must be Nobody.

    Comment


    • louisd
      louisd commented
      Editing a comment
      sounds like someone needs to brush-up on their layout skillz!! :-O

    • KIRBSTER269
      KIRBSTER269 commented
      Editing a comment
      louisd smarta_s_s

  • #6
    I hate those head-scratchers when you know you are doing everything right. Then you start second guessing yourself....but you know you're right. Fun times. Fun fun times!

    Comment


    • KIRBSTER269
      KIRBSTER269 commented
      Editing a comment
      especially on a Friday when I want to go home.

  • #7
    This problem is fixed, but I have some questions...
    Why using "fit ti datums" ON if there's no material condition ? And what does it change in this case ?
    If you measure B as a cylinder, what ever the side, and project all hits on A, then create a max inscribed circle, it should give the right definition of B.
    In this case, if B is different than B1, the cmm is out. In other cases, even if the part is OOT, B and B1 should be the same...
    Just a thought...

    Comment


    • KIRBSTER269
      KIRBSTER269 commented
      Editing a comment
      I actually did use MAX Inscribed and Least square, Besides the size results, only a .0001-.0002 difference. and The "Fit To Datums" doesn't matter nothing shifts or rotates.

  • #8
    And the lesson of this is (among other things): Always check the flatness of your planar datums and the roundness of your circular ones. Without knowing the form, you don't know if it really is a datum, or just a 'datum'...

    (And in the ISO World, planar datums are mandatory tangent planes, circular datums are mandatory max_inscr/min_circ.)
    AndersI
    SW support - Hexagon Metrology Nordic AB

    Comment


    • #9
      Originally posted by AndersI View Post
      And the lesson of this is (among other things): Always check the flatness of your planar datums and the roundness of your circular ones. Without knowing the form, you don't know if it really is a datum, or just a 'datum'...

      (And in the ISO World, planar datums are mandatory tangent planes, circular datums are mandatory max_inscr/min_circ.)
      On that note... why not check the entire part on the granite, whether results are good or bad, to 'double check' everything you've done on your CMM? (when possible of course).
      PcDmis 2015.1 SP10 CAD++
      Global 7-10-7 DC800S

      Comment


      • KIRBSTER269
        KIRBSTER269 commented
        Editing a comment
        are you kidding me?? I always do. That's what gave me the idea to check flatness, was when it wobbled on the plate, then I ran an indicator across it. I'm not criticizing you on this. technically that should have been mentioned in this topic actually when I restrained the part, I got better readings, but still wasn't good
        Last edited by KIRBSTER269; 10-24-2019, 10:01 AM.

      • Kp61dude!
        Kp61dude! commented
        Editing a comment
        Check your parts Kirby, get off your programming chair!

    Related Topics

    Collapse

    Working...
    X