Type of Best-Fit Contruction

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

  • Type of Best-Fit Contruction

    We are measuring ring gages and I was looking for input on the type of Best-Fit contruction to use. We currently use LEAST_SQR. Not sure if that's they way to go because I'm not too familiar with what the differences are between the types.

    Thanks in advance.

  • #2
    The following was posted originally by James Mannes. It should help you understand better. HTH


    ************************************************** ******************


    Keep in mind the different methodologies that PC-DMIS has to offer for algorithmic analysis...
    (doesn't that sound fancy?)

    LEAST_SQR
    Least Squares ?This calculation type provides a method of fitting in which the average squared radial distance from the data points to the circle is minimized. The square root of this quantity is the Root Mean Square (RMS) distance. Since the RMS distance is based on an average, some points may be further than the RMS distance from the computed circle.

    MIN_SEP
    Minimum Separation ?This calculation type generates a circle that is halfway between two concentric circles containing the data points, with the difference of their radii as small as possible. The Min/Max math used by the MIN_SEP calculation minimizes the maximum error, or deviation, from the input data to the circle. The Min/Max error is one-half of the minimal separation. No input data points (or input features) lie farther than the Min/Max error from the Min/Max circle. This calculation determines whether or not all the input data (or input features) are within the given tolerances.

    MAX_INSC
    Maximum Inscribed ?This calculation type generates an empty circle with the largest possible diameter that lies within the data. PC-DMIS first computes a Minimum Circumscribed circle and requires that the center of the Maximum Inscribed circle lies within it. This option could be used for a circular feature that requires a mating stud. For example, if the input data represents a hole, then this calculation returns a circle with the diameter of the largest stud that will fit inside the hole.

    MIN_CIRCSC
    Minimum Circumscribed ?This calculation type generates a circle with the smallest possible diameter that encloses the input data (or input features). This option could be used when measuring a stud that would fit into a mating circular feature. The resulting feature would be the smallest hole into which the stud would fit.

    FIXED_RAD
    Fixed Radius ?This calculation type creates a circle of a given diameter, positioned so that the maximal radial distance from the data points to the circle is minimized. It is similar to the Max/Min math used in the MIN_SEP calculation except that,since the diameter is known in advance, the radius cannot vary. Only the circle's position is allowed to vary.
    __________________
    James Mannes
    I am IRON MAN!!!

    sigpic"Hated by Many, Loved by Few" _ A.B. - Stone brewery

    Comment


    • #3
      Good morning John,
      What kind of tolerances are you trying to hold with regards to the ring gages? We send ours out for calibration. There is a help file regarding the best fit types...


      LEAST_SQRLeast Squares – This calculation type provides a method of fitting in which the average squared radial distance from the data points to the circle is minimized. The square root of this quantity is the Root Mean Square (RMS) distance. Since the RMS distance is based on an average, some points may be further than the RMS distance from the computed circle.

      MIN_SEPMinimum Separation – This calculation type generates a circle that is halfway between two concentric circles containing the data points, with the difference of their radii as small as possible. The Min/Max math used by the MIN_SEP calculation minimizes the maximum error, or deviation, from the input data to the circle. The Min/Max error is one-half of the minimal separation. No input data points (or input features) lie farther than the Min/Max error from the Min/Max circle. This calculation determines whether or not all the input data (or input features) are within the given tolerances.

      MAX_INSCMaximum Inscribed – This calculation type generates an empty circle with the largest possible diameter that lies within the data. PC-DMIS first computes a Minimum Circumscribed circle and requires that the center of the Maximum Inscribed circle lies within it. This option could be used for a circular feature that requires a mating stud. For example, if the input data represents a hole, then this calculation returns a circle with the diameter of the largest stud that will fit inside the hole.

      MIN_CIRCSCMinimum Circumscribed – This calculation type generates a circle with the smallest possible diameter that encloses the input data (or input features). This option could be used when measuring a stud that would fit into a mating circular feature. The resulting feature would be the smallest hole into which the stud would fit.FIXED_RADFixed Radius – This calculation type creates a circle of a given diameter, positioned so that the maximal radial distance from the data points to the circle is minimized. It is similar to the Max/Min math used in the MIN_SEP calculation except that,since the diameter is known in advance, the radius cannot vary. Only the circle's position is allowed to vary.
      sigpic

      James Mannes

      Comment


      • #4
        Well I guess my guns aren't as quick as Wes'

        I've been hit!! That's gonna leave a mark!!
        sigpic

        James Mannes

        Comment


        • #5
          Originally posted by JamesMannes
          Well I guess my guns aren't as quick as Wes'

          I've been hit!! That's gonna leave a mark!!
          Sorry about that James. Good thing you are Ironman!
          sigpic"Hated by Many, Loved by Few" _ A.B. - Stone brewery

          Comment


          • #6
            Well, we were shipping them all out, but getting crazy results. QM decided that we can check them just as well as anybody so we are now checking them all. Our machine can handle it, we are consistently holding our cal. sphere deviation to .000006". I have a program that checks them with the operator inputing the gage #, nominal size and class. The program then inputs all that into the dimensions with the correct nominal and correct tolerances for the class of gage.

            I just got to thinking as which was the best way accurately construct the Auto Circles. I think MIN_SEP may be the way to go, any thoughts?

            Comment

            Related Topics

            Collapse

            Working...
            X