concentricities of cyl's vs cir's

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

  • concentricities of cyl's vs cir's

    I have noticed that if I take three rows of points to contruct 3 circles on each side bore of a part and call each circle from one side to the other side to report concentricities everything turns out great, but when I construct a cylinder out of all the points, my concentricities all go to ****.

    best practices? Do most of you report cir to cir or would you trust the cyl to cyl as the most accurate data?
    Feed us-Fetus-Fajitas:monkey::monkey:

  • #2
    Originally posted by KaSaBiS View Post
    I have noticed that if I take three rows of points to contruct 3 circles on each side bore of a part and call each circle from one side to the other side to report concentricities everything turns out great, but when I construct a cylinder out of all the points, my concentricities all go to ****.

    best practices? Do most of you report cir to cir or would you trust the cyl to cyl as the most accurate data?

    Allow me to drag out my favorite passages from GeoTol Pro by Scott & Al Neumann:

    "Concentricity" and "Symmetry" are often misapplied and confusing for the entry level user of geometric tolerancing. The words "concentricity" and "symmetry" are very generic sounding terms, but in geometric tolerancing per the ASME Y14.5-2009 standard they are very complex requirements. The dilemma is that often untrained personnel will apply these controls but really want to apply position, runout, or profile tolerances.

    The application for concentricity is rare and commonly misused. Irregularities in the form can make it difficult to find the opposing median points. Therefore, unless there is a definite need to control the opposing median points, consider position, runout, or profile instead."



    I attended the Pc-Dmis GD&T class last month. This was discussed at some length. We were told that:

    1) Pc-Dmis uses the center to calculate "concentricity", not opposing points. Pc-Dmis will happily report a "concentricity" from a circle with an odd number of hits.

    2) Rob Jensen who co-wrote the manual for that class & who is on one of the ASME Y14.5 committees, replied to a question from our instructor that, 'he would suggest measuring an adequate number of circles along the length of the feature and reporting concentricity for each circle rather than measuring or constructing a cylinder.'

    3) the above comments apply to both Xactmeasure and legacy dimensions.

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

    Comment


    • #3
      +1

      Comment


      • #4
        I agree. In my experience, I realised that, in most cases, measures should be as 'local' as possible. If, for example, I measure circle A at a height of 1mm and circle B at a height of 15mm, I can evaluate the value of their concentricity with accuracy. And this value is a locale value, so its meaningful for analysis and evaluations. Later, I can measure another section for example at a height of 7mm, that should be useful for evaluating if hole has a linear axis deviation or is a 'banana-shaped' hole (they still exists, unfortunately). Conversely, a cylinder is likely to be a misleading straightened banana.

        Comment

        widgetinstance 190 (Related Topics) skipped due to lack of content & hide_module_if_empty option.
        Working...
        X