Iterative Alignment; Assist with explanation

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

  • Iterative Alignment; Assist with explanation

    Good Day Folks,

    I would like to ask for your assistance in trying to explain to the "engineer" types here the issue with my iterative alignment.

    Running 3.7mr3. In my alignment (manual then Dcc), I have 3 points on different planes for leveling, and 2 constructed points for rotating and my orign.

    I am required to report out the T values for my levelling points. The reported values are not "zero", and I am being asked why they are not.

    I don't know how to explain this and could use your assistance.

    Thank you in advance for your time.
    sigpic

  • #2
    Originally posted by Peter Warcholyk View Post
    Good Day Folks,

    I would like to ask for your assistance in trying to explain to the "engineer" types here the issue with my iterative alignment.

    Running 3.7mr3. In my alignment (manual then Dcc), I have 3 points on different planes for leveling, and 2 constructed points for rotating and my orign.

    I am required to report out the T values for my levelling points. The reported values are not "zero", and I am being asked why they are not.

    I don't know how to explain this and could use your assistance.

    Thank you in advance for your time.
    Report the "level" axis instead of the "T" axis. The reason is that the iterative alignment works to XYZ axis, and NOT to the "T" axis. If the points are NOT square to an axis, then the "T" axis will not be 'right'. You might 'try' (but I do NOT suggest it as a good thing) using SNAP points for the iterative points.

    OR, you could re-measure them AFTER the alignment, you might like what you see a little better, but hardware variation will show in the results.
    sigpic
    Originally posted by AndersI
    I've got one from September 2006 (bug ticket) which has finally been fixed in 2013.

    Comment


    • #3
      Matt,

      Thank you for your quick reply. I'm sorry, but what / how do you mean by reporting the "level" instead of the"T"?

      If I re-measure the points after the iterative alignment, I would assume that it would then be ok to use the "snap"?
      sigpic

      Comment


      • #4
        Yes, after the alignment, snap=on is perfect.

        As for the level axis.... Pcdmis will take the 3 "level" points to level the alignment (DUH!) but it will only use the axis with the biggest vector for the level points. So if your vector for the level points are 0.183,0.284,0.827, it will make the Z axis the level axis and make those Z values 'perfect'.
        sigpic
        Originally posted by AndersI
        I've got one from September 2006 (bug ticket) which has finally been fixed in 2013.

        Comment


        • #5
          Originally posted by Matthew D. Hoedeman View Post
          Yes, after the alignment, snap=on is perfect.

          it will only use the axis with the biggest vector for the level points. So if your vector for the level points are 0.183,0.284,0.827, it will make the Z axis the level axis and make those Z values 'perfect'.
          Matt,

          Fully Understood. Thanks!

          (Lack of coffee, sleep and Scotch)
          sigpic

          Comment


          • #6
            Hi Peter.
            One way you can create zero errors, or as close to zero as is, is to create an offset point off the surface of the job at the centre of where the ruby will theoretically hit.
            Use this as your intended contact point, turn probe comp off, and you can vector in at a vector of 1 in which ever axis you are measuring in. I use this when "gauge" is stated on the drawing.
            It probably sounds complicated but if you want to see the code I could post it if required?
            Bristol Citysigpic
            Home of the Reds

            Comment


            • #7
              You could just construct an offset plane using nominal print dimensions and then your three datum points would be perfect.
              Scott Staral
              Layout Tech.
              Wisconsin Aluminum Foundry

              Comment


              • #8
                Virgil,

                Can you expand on your suggestion a bit? I think I understand what you are suggesting and I understand that you eliminate some potential for error by removing the probe comp. calculation. Won't there still be many situations where it is not possible to achieve a nice fit due to part deviations? How much improvement do you expect to see in the fit by taking this approach? I work with parts that are difficult to fit on a regular basis and I think this concept could be very helpful to me. I just want to make sure I understand fully what you are doing.

                Thanks for your time.
                2013MR1 SP6
                Global Frames, Tesastar-M Heads, LSP-X1M/H Probes

                Comment


                • #9
                  Please send email address, it will be easier to explain that way.
                  Bristol Citysigpic
                  Home of the Reds

                  Comment

                  Related Topics

                  Collapse

                  Working...
                  X