Who else does this?? Qualifying/ Calibration

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

  • Who else does this?? Qualifying/ Calibration

    Right now I have been tasked with finding a better method to track calibration through 3 different shifts and 13+ different inspectors using pcdmis.


    I am a new employee here and right now here is what they do it...

    Every monday morning they calibrate every probe tip/angle they use... and through the week they have a program they run to check the standard and make sure everything is still checking good. This eliminates the need for the inspectors (not properly trained employees to keep their hands out of pcdmis, they use an interface with pcdmis so they never see the code)

    But this process leaves questions if there is a part that is checking bad is it dirt that got on the probe? are the parts previously ran before this job bad?




    My pitch and correct me if you guys know or think anything different,

    To eliminate calibration on monday mornings and start qualifying probe angles used in the program before every lot of parts is ran... saving and documenting the qualification results. We would be looking at giving a little but of training to the inspectors but i think it would be worth it.

    If we do this we can consistently watch the deviations of the probe and know exactly when something went wrong or when somebody crashed the probe even. It would also give us more operation time by eliminating the calibration (5 hour long process every monday)
    And of course always using a master probe to locate the sphere (if it has moved) and then run through the angles used in the program....


    what do you think?
    I will find a center in you...

  • #2
    Yeah that would work.

    If you want to remove the potential 'interface error' (aka cave monkey factor) you could create an auto calibrate section at the beginning of each program.

    You can create parameter sets which specify which tip angles that are used in the prog.

    Then with some flow control you can have a prompt asking if they want to calibrate the tips (which they would do for the first part of a batch).

    This would mean they could still run via the interface (if you don't do it this way they'd have to open the program, find each probe command, mark used and then qualify - which could lead to issues as they're in the PC-Dmis environment they coudld mess with the program, or easily miss a probe.
    Automettech - Automated Metrology Technology

    Comment


    • #3
      A golden part results, can easily deviate .0005-.0008 over time especially if you use multiple probe angles and tips, and you may not catch that something is wrong with the calibration. It maybe good enough for what you're doing, but doesnt mean you calibration is all good and dandy. Having autocal in the beginning of the program is what I would do.
      B&S One
      PC-DMIS CAD v2014

      Romer Infinity

      Comment


      • #4
        We 'MARK USED' and qualify before each program that's being run. We have had some newbies who have gone into the program, marked all the probe angles found, and qualified those. They've also missed a few angles in that process, and the hits made at those probe angles are wonky by quite a bit. (I believe) PC-DMIS uses the most recent calibration for each probe angle, and I have noticed that a current qualification is crucial. An obvious disadvantage to this is that calibration time needs to be built into each run. One BIG advantage is that if you do run into a problem (a crash, for example) you can see where the problem started, and you don't need to throw out the entire day's (or week's) results. For what it's worth, we used to have a separate probe file for each program, and we would simply qualify the entire probe file before each run. We have since seen the light, and have just a few probe files (one for each probe build-up) and we 'MARK USED' before each run. Having done it both ways, if I were starting all over, I would start with only a few probe files, rather than one for each program. We run unique tools on our CMM once a year, though, rather than a (relatively) small number of production parts.

        Comment


        • #5
          Originally posted by WolfMan View Post
          A golden part results, can easily deviate .0005-.0008 over time especially if you use multiple probe angles and tips, and you may not catch that something is wrong with the calibration. It maybe good enough for what you're doing, but doesnt mean you calibration is all good and dandy. Having autocal in the beginning of the program is what I would do.

          +1

          We run primarily for automotive and we calibrate every time we setup parts to run on the CMM's. That typically means that the calibration is done and between 1 to a maximum of 8 (1 per machine cavity/station) is run on a calibration. Many years ago they tried to just qualify all the tips at the beginning of the week but too many dirty hands while changing TP20's, moving the sphere and not using master probe, hard tip hits, etc. led to too many issues trying to get good results.

          We calibrate using 15 hits, 3 levels for our 5 standard probes every time and have gotten most of the programs to auto find themselves on the table with grid plates and fixturing so there is almost no operator manual alignments and this has drastically reduced the issues we've had with poor calibrations and results.

          Using autochangers on the machines has greatly helped with the TP20 dirt/swarf issues too now.

          The little extra time you'll add into the start of programs using autocal or doing this as a normal practice is well worth the peace of mind especially when you have to get setup to make adjustments, etc.


          4/1/96

          Comment


          • #6
            My operator interface does all this for you, including verifying qualification results.

            You need to create a section using flow control and parameter sets in a qualification program for each part program though, so it's not perfect, but it does give control to the programmer and removes the cave monkey factor!

            RunDmis_Qualification.pdf
            Attached Files
            Last edited by NinjaBadger; 08-05-2015, 05:30 AM.
            Automettech - Automated Metrology Technology

            Comment


            • #7
              awesome guys thanks for the input I would really like to pitch the idea of inserting autocal in the beginning of each program, bad part is... im not sure how to program that in....
              I will find a center in you...

              Comment


              • #8
                Open program

                Find first Loadprobe command > F9

                Click Mark Used (this selects all tips/angles used in program)

                Click Measure

                In the Parameter Sets section type a name (Part Number is a good option here) and click Save

                This in effect creates a stored setting of the current settings (number of hits/ touchspeed etc) for the currently selected Tips.

                Cancel (to close the Measure Probe window and return to the Probe Utilities window)

                Click OK - (if prompted to save changes to current probe say Yes)

                Now Insert > Calibrate > Auto Calibrate Probe to insert and auto calibrate command

                Click on the command in the edit window and F9 - in here set your settings (Ref sphere moved: Yes for Master probe, No for all others) and from the drop down list select the parameter set name you just created.

                Now when you run the program it will calibrate the required tips

                Repeat for all probes used in program


                You can then use an YesNo type comment and in IF_GOTO command to skip the Auto Qual commands so it gives an option of whether to qualify or not
                Automettech - Automated Metrology Technology

                Comment


                • #9
                  Awesome! thanks ninja man! that aint to bad.... so what if the sphere had been moved? , i think you know what im dealing with we dont want the inspectors to mess with anything besides 'yes, no, go, stop' so would the sphere need to be in a fixed location for this to consistently work?
                  I will find a center in you...

                  Comment


                  • #10
                    Originally posted by himmel_dustin View Post
                    Awesome! thanks ninja man! that aint to bad.... so what if the sphere had been moved? , i think you know what im dealing with we dont want the inspectors to mess with anything besides 'yes, no, go, stop' so would the sphere need to be in a fixed location for this to consistently work?
                    what version are you working with? the new 2014 and up allows to do this in DCC complete, older versions prompt you to take the first hit on the sphere if you have selected " ref sphere moved" option.

                    What Ninja was talking about is having an additional Master probe as a part of your calibration for every program. You don't have to, but I would.

                    another thing I would suggest is for you to read this http://www.pcdmisforum.com/showthrea...ht=calibration
                    Last edited by WolfMan; 08-05-2015, 08:23 AM.
                    B&S One
                    PC-DMIS CAD v2014

                    Romer Infinity

                    Comment


                    • #11
                      2013mr1
                      I will find a center in you...

                      Comment


                      • #12
                        Originally posted by himmel_dustin View Post
                        2013mr1
                        not sure if they added that option in v2013, maybe someone can verify that
                        B&S One
                        PC-DMIS CAD v2014

                        Romer Infinity

                        Comment


                        • #13
                          sorry im just tryin to figure this out... so in the beginning of every program I should ...

                          Insert a loadprobe:master probe and a autocal....
                          followed by all of the other probes/tips used in the program and autocal for each one.

                          and if I dont want the opeators to take manual hit then I need to have a fixed qual sphere?
                          I will find a center in you...

                          Comment


                          • #14
                            Originally posted by himmel_dustin View Post
                            Awesome! thanks ninja man! that aint to bad.... so what if the sphere had been moved? , i think you know what im dealing with we dont want the inspectors to mess with anything besides 'yes, no, go, stop' so would the sphere need to be in a fixed location for this to consistently work?

                            Nope.

                            Does your ref sphere go back in the same location each time (within a couple of mm?), or could it be anywhere on your table?
                            Automettech - Automated Metrology Technology

                            Comment


                            • #15
                              all of our parts are for transmissions so they are all cylindrical sooooo they are set up on fixtures (chucks) so we have built in chucks on tracks on our cmm table..... every monday morning the sphere gets mounted on the chuck fixture so its always in the same spot. then it gets moved to the side. so i guess before each batch of parts is ran the operators could mount the sphere in the fixture and say that it hasnt moved.
                              I will find a center in you...

                              Comment

                              Related Topics

                              Collapse

                              Working...
                              X