Proper way to identify datums using auto features

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

  • Proper way to identify datums using auto features

    Attached below, the left side of the part has three diameters that are the same dimension the true position of these 3 is supposed to be datum B, and datum A is the top surface i take as a plane. So I measured the three diameters with auto circle, then I constructed a circle through the 3 diameters and used the constructed circle to creat datum B in my program. My question is would that be the correct way to do that? Because when I ran my program I was getting some weird measurements with the FCF callouts. For datum A i just did a grid scan and constructed a plane from the grid scan and identified this at datum A in the datum definition. In general I am curious what other people do on here as far as datuming.
    Attached Files

  • #2
    Depends on your version, the plane should be a L2 constrained, or a minimax, not a Least square plane.
    Then the circlles should be measured and constructed in that plane.

    Comment


    • #3
      Can''t see the datum definition of that drawing.

      I assume you mean there's a pattern of holes (i..e 3 x Øx.xx) with a position callout attached to them?
      Automettech - Automated Metrology Technology

      Comment


      • #4

        Time to up your Blue Steel game, Zoolander.

        Originally posted by zoolander303 View Post
        ... the left side of the part has three diameters that are the same dimension the true position of these 3 is supposed to be datum B,...
        I am guessing that you mean to say, "The Pattern of 3 identical holes is Datum B", because what you wrote makes little to no sense.
        ---> The term "true position" means "the requirement" or "the measured results" depending how it's used.
        ---> The term "supposed to be datum B" means "this is the datum feature that controls some degrees of freedom so we can see where the measured results are compared to it."

        Saying "the true positon is datum B" is exactly like saying "the flatness result is where you hold the hook end of the tape measure". The two things are fundamentally different.

        We could easily tell if you had included the callout and datum definition flags in your screenshot.

        So going on my guess that the datum is the pattern of 3 holes, then you should probably Construct a Set of the 3 individual round features and define that Set as Datum B.
        Sometimes the callout demands that the members of a pattern of holes (which is the datum) be positioned correctly to each other, so you would be reporting the position of each individual B hole.

        Originally posted by zoolander303 View Post
        ... datum A is the top surface i take as a plane. So I measured the three diameters with auto circle, then I constructed a circle through the 3 diameters and used the constructed circle to creat datum B in my program. My question is would that be the correct way to do that? Because when I ran my program I was getting some weird measurements with the FCF callouts. For datum A i just did a grid scan and constructed a plane from the grid scan and identified this at datum A in the datum definition. In general I am curious what other people do on here as far as datuming.
        I hope you know the 6 Degrees of Freedom. If not, you need more training.

        Datum Scheme defines which features control which degrees of freedom, and in which order.

        When you are dealing with holes or posts, you need to know what degrees of freedom they are supposed to control.
        If they have to control rotation-around-axes degrees, then you must use cylinders (or 3D lines constructed between separate circles along hole/post).
        If they only control translation-along-axes, then circles are fine as they are purely 2D in PC-DMIS.


        Comment


        • #5
          Show us a better picture that clearly shows the Datums and we could help better.
          With the picture and what you wrote only lets us guess what you mean and probably wont help you.

          Comment


          • KIRBSTER269
            KIRBSTER269 commented
            Editing a comment
            says the guy, that only shows up, once every 9 months or so.

          • Mikeg
            Mikeg commented
            Editing a comment
            don't yell at me
            I'm just getting settled in at the new job

        • #6
          The material thickness is over the high limit tolerance due to this being a painted part the painter went to thick on the coating, so since I have material modifiers I am wondering if thats throwing everything off? Even trying to align this thing is a pain, then I try and take the radius of the bend as a cylinder and the probe cant find it.
          Attached Files

          Comment


          • #7
            Not sure why so many changes to this print and why they would be changed to the way they are.
            All dimensions are from the original B (D) and should be measured as it was.
            Datum B should be located to outside edges (or other critical features) and then all other holes located off that (as shown originally except the location of B (D).

            Comment


            • #8
              Similar to what Mikeg said, I interpret this to say that Datum B is not all 3 holes, but only the bottom right hole in the top view, originally noted as Datum D. The other B in the top right of the top view is not a Datum callout, but a true position callout aligned to Datum B, a correction from it originally being labeled D.

              Datum C is along the edge of the bent section, and A is as you originally stated, the top surface. So you have a plane, a line and a point to align to.

              That 3x shouldn't mean all 3 holes are Datum B, only that all 3 share the diameter and position nominals and tolerances.

              Comment


              • #9
                Originally posted by zoolander303 View Post
                The material thickness is over the high limit tolerance due to this being a painted part the painter went to thick on the coating, so since I have material modifiers I am wondering if thats throwing everything off? Even trying to align this thing is a pain, then I try and take the radius of the bend as a cylinder and the probe cant find it.
                Nice.

                FYI, the radius of the bend is a "CMM no no zone", stay away from stuff like that as it will never be a real round partial cylinder. Note that it's nothing on the print, not even a radius callout.

                Don't worry about the paint, it will only make the hole sizes smaller and the 5mm holes further away from the effective A plane.

                6 Degrees of Freedom:
                A plane (facing Z+) controls Rotation Around X, Rotation Around Y, and Translation in Z
                B pattern of 3 holes (as a team, according to redline edits) controls Rotation around Z, Translation in X, Translation in Y.
                C edge doesn't control jack ****, it's useless as all 6 degrees are spoken for. It would only be necessary if only 1 of the 6mm holes was the datum B.

                I would definitely go speak with whomever red-lined that print and ask them How Does This Bracket Fit It's Mating Part In Real Life?
                I bet you 1000 internets that the C edge doesn't mate anything, that the pattern of three 6mm holes really does control rotational orientation and the "real" datum scheme is AB, and they will agree they should have wiped out all C references with their mighty red pen.

                Next, ask them why the heck they left the wacky BAC datum scheme for the two 5mm holes on the lower flange. If the part fits AB, the datum scheme should be AB. It's not like the two flanges fit different mating parts, obviously this is a little corner strength bracket.

                How to do AB in the alignment dialog:
                Level A to Z+
                Click Best Fit button, in that dialog pick all 3 circles, choose 2D best fit to Z+, make sure Least Squares is selected on the right, click Compute, OK out, Ok to finish the alignment.
                Note that this really only works seamlessly if you've created all your manual alignment features and all your DCC alignment features by clicking on CAD so they have correct nominals.







                Comment


                • zoolander303
                  zoolander303 commented
                  Editing a comment
                  Worked like a charm when I dont use C as a datum, therefore I am having the customer fix the print. Thanks I appreciate it

              • #10
                thanks guys, yeah the print got red lined and this part just has been throwing me for a loop, I appreciate all of the knowledge you guys have given me thank you, Ill try what you mentioned and see what happens.

                Comment


                • #11
                  I will have my staff invoice you for the 1000 internets.

                  Comment


                  • zoolander303
                    zoolander303 commented
                    Editing a comment
                    Lol thanks murph

                Related Topics

                Collapse

                Working...
                X