Question about Leveling to a Primary Datum -A-

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

  • Question about Leveling to a Primary Datum -A-

    Hello everyone...

    We are currently having a discussion about leveling to a primary plane, (Datum A). The discussion concern is whether we should be leveling to a "high point plane" or if we should be leveling to a "best-fit" plane using every point that is probed.

    I am currently probing a planar surface that is approximately 150mm diametrically in 3 seperate rings that are 5 degrees apart so there are a total of 216 points. I have an overall flatness of approximately 0.3mm with a total tolerance of 0.8mm.

    As I stated before the disussion is whether we should be finding the high points of the plane and then be leveling to that plane or not (high point plane). I guess my feeling is that you would not level to the high point plane but in certain applications you may want to set the "Z" origin to that plane.

    What are your most common approaches of leveling to the primary Datum plane? Thanks for your replies!

    Mike

  • #2
    The high point plane is not repeatable. just like the max insc or min circ. diameters.

    Also, it would be difficult to set origin on a plane without leveling to it first.

    I think by definition the plane should be the high point plane but it is very difficult to achieve on a CMM and much less repeat.

    Most commonly used is best-fit. If you guys think that "high-point" is required then a datum simulator is needed.
    Last edited by cmmguy; 01-08-2007, 08:39 AM.
    Links to my utilities for PCDMIS

    Comment


    • #3
      Well, if you level to the 'high-points' of the plane, then you have just cut your tolerance zone in half, if it is bi-lateral, you are forcing all the 'non-plane' points to one side of your origin. Bad.
      sigpic
      Originally posted by AndersI
      I've got one from September 2006 (bug ticket) which has finally been fixed in 2013.

      Comment


      • #4
        It would certainly take an enourmous ammount of points to find the actual high points.

        I can see applications for both depending on the function of the part. I would say...

        I would think if datum targets are defined then use them. If no datum targets are specified, I would say average unless otherwise specified.

        I always do the average and have never had anyone ask about high points.

        Comment


        • #5
          I couldn't agree more cmmguy... Any more thoughts. I need to gather as much information as possible so that I can provide facts and logic to this conversation because these guys are talking in a theoretical world where every point would be analyzed and all that conceptual stuff. I'm dealing with a plastic part that is 6 inches in diameter and do not have the luxury of analyzing every point on the part nor could I do that in the first place.

          Comment


          • #6
            I think that so far all of us are on the same page here. The methods that you have mentioned are how I was "taught" orginally but my methods are being questioned so I need to explain why I do the things I do. Thanks a bunch for your replies I really appreciate it!!!

            Mike

            Comment


            • #7
              Originally posted by Matthew D. Hoedeman View Post
              Well, if you level to the 'high-points' of the plane, then you have just cut your tolerance zone in half, if it is bi-lateral, you are forcing all the 'non-plane' points to one side of your origin. Bad.
              ???
              The y14 (pg60) definition of the datum plane is exactly that - to the high points. It is just too difficult to achieve on a CMM. How are you cutting your tolerance in half?
              Links to my utilities for PCDMIS

              Comment


              • #8
                What everyone else says -

                There's only one way to repeat the high point plane and that's if you use a fixture, probing the fixture to simulate the contact points which would be the highest points on the datum surface.

                If the surface stays within the form tolerance then it's probably good enough to level and assign as origin.
                PC-DMIS CAD++ 3.7 from 4.2 MR1

                Comment


                • #9
                  Originally posted by cmmguy View Post
                  ???
                  The y14 (pg60) definition of the datum plane is exactly that - to the high points. It is just too difficult to achieve on a CMM. How are you cutting your tolerance in half?
                  Everything has a tolerance. Everything. In sheet metal (and plastic too, I would imagine) every surface has a tolerance. I would imagine his print calls out a tolerance for that surface, either flatness or position or both. Now, do you want to shoot yourself in the foot by turning your 1mm tolerance band into 0.5 by choosing the highest points of the surface? If the surface is allowed (toleranced) to move 0.5mm in each direction, shouldn't you use the average plane to place both + and - points on the plane?
                  sigpic
                  Originally posted by AndersI
                  I've got one from September 2006 (bug ticket) which has finally been fixed in 2013.

                  Comment


                  • #10
                    Originally posted by Matthew D. Hoedeman View Post
                    Everything has a tolerance. Everything. In sheet metal (and plastic too, I would imagine) every surface has a tolerance. I would imagine his print calls out a tolerance for that surface, either flatness or position or both. Now, do you want to shoot yourself in the foot by turning your 1mm tolerance band into 0.5 by choosing the highest points of the surface? If the surface is allowed (toleranced) to move 0.5mm in each direction, shouldn't you use the average plane to place both + and - points on the plane?
                    Flatness is a moving band - so that doesnt apply
                    Dont think position would apply either since it is the primary datum. Why would you apply anything beyond flatness to a primary datum anyway?

                    So I still dont see where you are robbing yourself of any tolerance - sheetmetal, plastic, or machined parts.
                    Links to my utilities for PCDMIS

                    Comment


                    • #11
                      Originally posted by cmmguy View Post
                      Flatness is a moving band - so that doesnt apply
                      Dont think position would apply either since it is the primary datum. Why would you apply anything beyond flatness to a primary datum anyway?

                      So I still dont see where you are robbing yourself of any tolerance - sheetmetal, plastic, or machined parts.
                      I think the robbing part is features dimensioned from said datum in this case. At least that is what I think would be a scenario
                      <internet bumper sticker goes here>

                      Comment


                      • #12
                        Originally posted by cmmguy View Post
                        Why would you apply anything beyond flatness to a primary datum anyway?
                        Haahaa ha ha ha ha ha ha. Is that the way it is where you work?

                        I have gained the opinion that ALL design engineers should be forced to work in inspection prior to actually designing anything.

                        Comment


                        • #13
                          I couldn't have said it any better....

                          You hit the nail on the head... In my opinion it is there decision on how to handle the form of the part. If it really concerns them then squish the flatness callout to 0.1mm and then develop a process to make it flat.

                          Comment


                          • #14
                            Originally posted by craiger_ny View Post
                            I think the robbing part is features dimensioned from said datum in this case. At least that is what I think would be a scenario
                            That is the definition in Y14. I use best fit as most everyone else does. I just dont get the argument when it is in compliance with the standard, that is all.


                            Originally posted by Goodluck View Post
                            Haahaa ha ha ha ha ha ha. Is that the way it is where you work?

                            I have gained the opinion that ALL design engineers should be forced to work in inspection prior to actually designing anything.
                            What other tolerance would you apply to the primary datum that is a plane, other than flatness?? (Even if you were a designer that did not know what you were doing?)
                            Links to my utilities for PCDMIS

                            Comment


                            • #15
                              Hi eveyone. I hope you don't mind I butt into this conversation. I believe fit and function are key, hence capturing the higest points is important on a leveling datum. I agree that a fixture is one way to capture the hight points on a leveling datum, but when there is no fixture, nor a flatness tolerance, nor datum targets, then what do you do? Do you use the high point plane feature on PC_DMIS to try to approximate this?
                              Thanks

                              Comment

                              Working...
                              X