Best fit alignment not zero with nominal CAD features

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

  • Best fit alignment not zero with nominal CAD features

    Anyone want to take a stab at why a best fit alignment with all features perfect features somehow has +-0.100mm deviation?

    https://ibb.co/4JLw5cL
    "This is my word... and as such is beyond contestation."

  • #2
    dimensions from AFTER the best fit and the values in the bestfit window are from the previous alignment?
    sigpic
    Originally posted by AndersI
    I've got one from September 2006 (bug ticket) which has finally been fixed in 2013.

    Comment


    • #3
      Originally posted by Matthew D. Hoedeman View Post
      dimensions from AFTER the best fit and the values in the bestfit window are from the previous alignment?
      No. Dimensions are before the best fit. I went through the program several times too manually and with "reset features to nominal".

      The previous alignment is an iterative, and all 5 points actually measured +/- 0.005mm. Now, everything is reset to 0, but the best fit doesn't fit to 0. It doesn't make sense.

      If I make an iterative, everything sets to 0. And if I BF after this iterative, it still shows those odd numbers. With no measured deviation...
      "This is my word... and as such is beyond contestation."

      Comment


      • #4
        I've never been impressed with the BEST FIT performance of Pcdmis. You can best fit a bunch of points, dimension them, then do another best fit, same features, same order, and get different results.
        sigpic
        Originally posted by AndersI
        I've got one from September 2006 (bug ticket) which has finally been fixed in 2013.

        Comment


        • VinniUSMC
          VinniUSMC commented
          Editing a comment
          I'm going back to just doing an iterative without remeasuring.

        • Matthew D. Hoedeman
          Matthew D. Hoedeman commented
          Editing a comment
          well, you have to remember, with an iterative, SNAP=OFF for one thing. For another, if the features are NOT square to the world you will get deviations when you remeasure them. If for 'level' features are 'not square' by 10 degrees, you'll have a vector of 0, 0.173, 0.984. It only slides the part until the Z is correct, it doesn't slide the part per the vector, it only makes the "Z" right (in this example). Yeah, iterative is great, but I think some tweaking to it's function might be nice

        • VinniUSMC
          VinniUSMC commented
          Editing a comment
          Yes, that is why I wanted to use BF. But what BF is doing is 30x worse than what iterative gives me.

          It would be great if there was a way for iterative to take vector into account instead of only single axis, but it is what it is.

      Related Topics

      Collapse

      Working...
      X