hexagon logo
Register Sign in
  • Replies 27 replies
  • Subscribers 225 subscribers
  • Views 890 views
  • Users 0 members are here
Options
Related

CMM Health

acgarcia
Good Afternoon Everyone,

We are starting a CMM health initiative to help trust the CMMs. As of right now, operators can only run parts and the probe calibration program. Management is looking into something like check an artifact of known size weekly or even daily. I have (4) 454 SFs with (3) with manual heads and (1) with an indexing head, (3) Tigos with fixed analog heads, and (1) 7.10.7 Sf with an indexing head.

What kind of artifacts do you all check on a regular basis for CMM confidence? Rings gages? Gage blocks? An actual machine'd part? If its a machine'd part, do you check on 1 CMM to compare to other CMMs? OR does the human check it with hand tools and then looks for correlation? I know Hexagon has the swift check thing but I think its only for indexing heads so it won't work with all my CMMs.

Any advice? Thanks.
  • I do my own ball-bar checks as well as linear checks with a home-made step bar (that was measured and assigned 'nominals' based on a brand new CMM machine machine installed and calibrated and certified by the OEM). Step Bar has no 'certification' that is traceable, however it gets checked right after the annual calibration of the machine, data saved, and used for comparison until the next machine calibration.

    your step bar can be traceable, since you measured it with a calibrated CMM that is traceable to NIST.

    i would measure the step bar and note of the report and serial number of CMM used and calibration status of CMM and then assign a serial number to your homemade step bar.
    bam you have a certified step bar.

    Todo this you should create a procedure of how you're measuring the step bar on the cmm,
    on your report make sure to reference CMM Serial#, Step Bar serial number, procedure used, calibration due date of CMM.
    fill out your calibration record/certificate and include the CMM report. now you have a in house calibrated step bar traceable to NIST or whatever you CMM is traceable to.

    If you get audited about the validity of your step bar, you can show them the results from the cmm and then show them the calibration certificate for the CMM, you shouldn't have any more questions after that.

    I do this all the time in my shop and been audited and haven't had a problem yet, just got to create the right paperwork trail.

    i believe you've got a calibrated step bar missing some paperwork.

    just my .02 cents


  • your step bar can be traceable, since you measured it with a calibrated CMM that is traceable to NIST.

    That is false. There is a lot more to traceability than just the equipment. Also, traceable to NIST is not a thing; it must be traceable to the SI Units. It is possible NIST was the NMI that happened to inspect some ancestor of your equipment but often times that is not true either.


    ...
    If you get audited about the validity of your step bar, you can show them the results from the cmm and then show them the calibration certificate for the CMM, you shouldn't have any more questions after that.
    ...

    That is extremely unlikely. If I was an auditor you would find yourself with a NC *if* that step bar was used for anything involving inspection or treated as a calibration standard for something else. I am assuming your company is not 17025 accredited with step bar measurements listed on your scope.

    does more than most people in his position I believe. He created a step bar for internal use as part of his own interim or sanity check of his equipment. I know another company that does this as well. If you keep it at that level there is no issue.
  • in reply to Rondog
    i send out my spheres to a 17025 accredited calibration lab.

    they simply measure it on their CMM, and state what they measured.

    states its measurements are traceable to NIST, NCR, and NMI/SI

    If i created a step gauge, why cant i do the same thing if my equipment is calibrated?
    why can't i certify it with the correct paper trail / paperwork?
    should be no different than calibrating a OD mic with certified gauge blocks that are traceable, my gauge blocks provide to traceability for the Mic why cant the CMM provide the traceability for the Step gauge?

    aside from the paperwork i agree with you there is a lot more behind the scenes of how the step gauge was programed measured etc could create many different results.

    not trying to argue, this is part of what i do everyday is interpret the requirements for the standards we are bound to, mainly API requirements, i guess what i'm trying to say is i enjoy getting other points of view

    but strictly looking at the paperwork why not?
  • I am currently doing my "is the CMM still happy?" checks on the Global.

    STEP 1: Using the most common probe build, I calibrate 9 angles, then check the cal-sphere using those same 9 angles, 10 times, and plot it on a run chart. (My build is roughly 6" from pivot of head to probe tip)
    STEP 2: Using a home-made ball-bar, I check 29 positions on the machine (I have a map of the locations that I use)
    STEP 3: Using a home-made step bar, I check the entire length in X & Y (using a cumulative method, not 24" over and over and over) as well as 24" in Z, every inch
    STEP 4: I check the actual table of the machine, 2" grid (almost 1900 points) and report out the Z deviation as well as over-all flatness.

    You may think that the table check is a waste of time, but it is not. If the table and ways were all "perfect" to begin with, it will all check "zero", but, if you your table isn't properly supported and correctly supported, that rock will change shape over time. This change WILL show up with this check, and the deviations will be greater for a taller (more Z travel) machine than for a shorter machine.

    I actually "fixed" a small amount of error on the Validator when it got moved to the new lab and set back up by getting the support cans placed "more better" than they were before the move. The Validator is a 4-point support for the bridge, so even easier for table issues to show up in a 'flatness' check. I reduced roughly 0.0008" of droop in the right-rear corner of the machine down to 0.0004" over the course of 3 years of the new support. I don't know if it will get any better over the next few years or not. Yes, it takes years for the rock to change shape.

    PH10MQ, TP20 STD FORCE, 6" total length results:

  • in reply to Matthew D. Hoedeman
    Table flatness: +0.00051 to -0.00071 (0.00122 'flatness')
    Ball Bar: +0.00020 to -0.00024
    (all values are in inch)
    starting the step bar runs now
  • in reply to Matthew D. Hoedeman
    pretty sure it has a good, clean bill of health. Yes, it takes a LONG time to do this (almost 7 hours today) but I have a very high level of confidence in the machine. These tests get run (usually) within the week after the machine is calibrated, then any time I think there may be an issue (VERY rare!). BUT, they all get run if the machine gets whacked in any way (usually due to some floor monkey doing something he shouldn't), but also in case of operator error (had a 2nd shifter run the column into a fixture).

  • Another way to "increase trust" is to correlate your CMM programs to your previously established inspection techniques.

    Due to the level of our work, I had to create a documented procedure for this. We call it our "CMM Program Correlation Procedure".

    If you want, PM me your email & I'll send you the form we use.

    Our process:
    -I receive request for CMM program
    -CMM program gets made. Is saved in a "WORKING DIRECTORY" (Only I have access to this folder)
    -I run CMM program in real life & collect the data on all required dims
    -I measure the part with hand tools and collect the data on all of those same dims
    -Input dim data into the worksheet. The sheet compares the CMM results vs the hand tool results..it also compares that dimensional data against the allowable tolerance zone for each dim..once its done thinking about alllll that stuff it spits out a PERCENTAGE score for that row. Score less than or equal to 10% is acceptable. Score greater than 10% while still less than 25% you must "explain". Score greater than or equal to 25% is unacceptable.
    -If the worksheet is acceptable, we keep it on file as objective evidence that the program is accurate. CMM program itself now gets moved to a "VALIDATED PROGRAM DIRECTORY" folder (we only use programs from this folder to measure products).
    -Inspection paperwork gets updated to allow for the inspector to now use CMM PRG# XXXXX AS WELL AS the hand tools that their inspection sheet originally called out for. This does not remove a "micrometer" from the paperwork....this adds a "/CMM" to the paperwork to give them another option.

    {"alt":"Click image for larger version Name:\tCapture.jpg Views:\t0 Size:\t44.8 KB ID:\t515623","data-align":"none","data-attachmentid":"515623","data-size":"full"}

    I think it is good for hole size only. How about TP? It is hard to verify TP with other tools.


  • I think it is good for hole size only. How about TP? It is hard to verify TP with other tools.

    You can verify true position "with other tools" if you have the correct equipment and know what you're doing, we do it every day Slight smile If you ever have any questions about gaging/datum interpretation/whatever...feel free to reach out. Will help if I can.

    For big/funky stuff we'll run the program on two CMMs and have one correlate the other
© 2024 Hexagon AB and/or its subsidiaries. Privacy Policy | Cloud Services Agreement