How is Concentricity Calculated?

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

  • How is Concentricity Calculated?

    Good Evening,

    I'm stumped on why my CMM is reporting a circle is out of tolerance. My datum is a vertical line. I probe a large circle and a smaller circle to create the line/axis which is the datum. Both circles are machined in the same orientation and same tool in the CNC. Then I probe another circle which is machined in a different orientation and with a different tool in the CNC and report its concenticity to the datum. The CMM reports that the circle is measured 0.0124" and my max tolerance is 0.005".

    So I dimension'd the diamter of the the out of tolernce cirlce and the X and Y axis and it reported 0.006" in X+ and 0.000 in Y.

    I don't understand why the concentricity measured is twice the X deviation nor can I explain it to the operator and manager. So the manager got the part and put it into a large mill to indicate the circles. He found that the circle was off center by only 0.003" so he says the part is good and there is something off in the CMM. concentricity issue.png

  • #2
    The tolerance of concentricity is 0.005". This is the area where the center of the measured circle must be, to be in tolerance. This area is a cylinder with 0.005" diameter. So if your center is 0.006 out of center line is just a radius and must be multiplied by 2. If the operator on the mill measured 0.003" out of center you must multiply this by 2 and the result is 0.006" and the part is not OK.

    Concentricity | GD&T Basics (gdandtbasics.com)

    Comment


    • #3
      Is the alignment OK? The line you constructed (green circles) is a 3D line and the axis line is level to Z+ or any other direction (depend how the part is placed on your CMM). Both green cicles must have X nad Y coordinates 0.000".

      Comment


      • #4
        Concentricity is like TP, but in concentricity the datum and the toleranced features always share a common nominal centerline. Radial deviation from center * 2. Just like TP.

        Your manager must double the value from center.
        PC-DMIS CAD++ 2o22.2 SP1

        Comment


        • #5
          As usual, is it a concentricity, a coaxiality or a true position ???????
          In ISO 1101, the concentricity concerns circles (with or without projection plane), coaxiality concerns cylinders.
          Localizing a circle from a cylinder should be (in ISO world, and in my opinion !) a true position...

          Comment


          • #6
            Just hate this callout. We do a ton of " BORES" from either end and we see this callout used. You have to add a lineal or postional callout anyways as how do you adjust to the reported " concentricity" value. Just makes my blood boil.....

            Comment


            • #7
              Thanks for all the info.

              Yes, 3D line and those circles have XY nominals of 0.000".

              Management is questioning the CMM for this out of tolerance dimension. Everything else in the report is green expect for this. Even other concentricity dimensions are green.

              I forgot to mention that I was able to flip this part over and xy origin on the problem cylinder (third circle). Then I probed the next cylinder with a disk probe and it was off the almost the exact same amount. Deferred by 0.0003".

              Comment


              • #8
                I work with a part almost identical, do your operators deburr the part? just by them cutting to deep will throw your concentricity out of TOL if your alignments are good I would ask for a part before they deburr it. And I only use two features to align my part, the top plane and the large cylinder should do the trick.
                Last edited by GatorActual; 09-23-2022, 11:00 AM.

                Comment


                • #9
                  Yes, the part is deburred. Its actually a very large part. The CMM head goes all way down in there. It almost takes up all the working envelop of my 7107.

                  Comment


                  • #10
                    see if you can have them bring you a part that was lightly deburred and see if there is a change in the variation

                    Comment


                    • GatorActual
                      GatorActual commented
                      Editing a comment
                      Going back and re reading your issue I assume you've verified your nominals on your alignment plane? it being such a large part how many hits are you taking inside that lower circle?

                    • acgarcia
                      acgarcia commented
                      Editing a comment
                      I took 13 hits on each diameter. I'm actually aligned to a constructed cylinder that is perpendicular to red line in my picture. I level to that line, rotate to my constructed cylinder, XY origin to the red line and z origin to the constructed cylinder. I could probe a plane to level to, but thought my alignment was like a cylinder cylinder alignment.

                    • acgarcia
                      acgarcia commented
                      Editing a comment
                      I'm also using a 25mm disk probe for those all 3 of those circles.

                  • #11
                    Have you checked form error, roundness?

                    Comment


                    • #12
                      Yes, the out of tolerance diameter has a roundness of 0.0025".

                      Comment


                      • #13
                        Originally posted by acgarcia View Post

                        I forgot to mention that I was able to flip this part over and xy origin on the problem cylinder (third circle). Then I probed the next cylinder with a disk probe and it was off the almost the exact same amount. Deferred by 0.0003".

                        I would say that by flipping the part over and reproducing the error you have proved that the part is in error and the CMM is reporting correctly. If that doesn't satisfy people, you could also try to verify it manually but that would require some way of holding the part - locating off just the two datum cylinders - and then clocking the bore, but with such a large part and the datums being internal diameters, it would be difficult. If they were external diameters you could sit them in V-blocks, for internal diameters, you'd need some sort of chuck or mandrel.
                        Neil Challinor
                        PC-DMIS Product Owner

                        T: +44 870 446 2667 (Hexagon UK office)
                        E: [email protected]

                        Comment


                        • #14
                          UPDATE:

                          so I’ve been trouble shooting this for a couples of days now. I received another part and I see the same deviation on the same feature. I even inversed it by XY origin as a cylinder in the purple area in the pic above and reported the XY deviation on concentricity of the 2 green circles and they are both show to be -0.007” and -0.0008” on the X axis.

                          I even 3D printed a small test piece over The Weeknd where all the 3 bores where XY 0,0 and another test piece that had 1 bore intentionally 0.010” off and inspected it. Guess what, the CMM reported that the piece with all 3 bore were concentric and the other piece measured 0.0108” off in the X axis. I thought that little experiment was clever and proof that the CMM is measuring properly until this afternoon.

                          A second machinist checked another part in the CNC with a new coaxial indicator and found the problem bore to only be 0.001” at the most. Needless to say is that we are all stumped.

                          I’ll b calling hexagon in the morning. I might be able to get this part on 1 of my Tigos. I will over hang the ends a little bit if I can measure it and repeat the deviation im seeing, the we can say the CMMs are measuring correctly.

                          Comment


                          • #15
                            If you would check the concentricity between the third circle to the small circle and third circle to the large circle, what does these results show?

                            Legacy, Xact or GeoTol?
                            PC-DMIS CAD++ 2o22.2 SP1

                            Comment

                            widgetinstance 190 (Related Topics) skipped due to lack of content & hide_module_if_empty option.
                            Working...
                            X