Profile Point Results

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

  • Profile Point Results

    V4.1
    I am trying to interpet the"point info" or "texual analysis"deviation results from a profile dimension.
    For example if I compare the deviation results of a point within the profile dimension of a constructed set and the same point as a unique stand alone feature the deviation results will be different. When looking at the point coordinates in the edit window the deviation of any single axis will not agree with anything. Can anyone confirm what is taking place? Is a 3D true position calculation taking place when looking at the point as a unique feature? How about the deviation when looked at within the profile dimension?
    All comparisons have been made under the same alignment.
    Thanks
    ["Paper is poverty,... it is only the ghost of money, and not money itself." --Thomas Jefferson ][/SIGPIC]

  • #2
    Originally posted by Bob Pierce View Post
    V4.1
    I am trying to interpet the"point info" or "texual analysis"deviation results from a profile dimension.
    For example if I compare the deviation results of a point within the profile dimension of a constructed set and the same point as a unique stand alone feature the deviation results will be different. When looking at the point coordinates in the edit window the deviation of any single axis will not agree with anything. Can anyone confirm what is taking place? Is a 3D true position calculation taking place when looking at the point as a unique feature? How about the deviation when looked at within the profile dimension?
    All comparisons have been made under the same alignment.
    Thanks
    You have three possible outcomes.

    1. Report the point out using a "T" deviation.

    2. Report the profile using "form only".

    3. Report the profile using "form and location"

    Each one could you give a different result.

    The "form only" will report the form error.

    The "form and location" profile could report the location if that has the highest deviation or it will report out the form error if that is the largest deviator.

    "T deviation will report out what it reports out all the time.


    And that is as far as I'm am willing to discuss profile. It has been beaten to death on the forum.
    sigpic

    James Mannes

    Comment


    • #3
      Originally posted by JamesMannes View Post
      You have three possible outcomes.

      1. Report the point out using a "T" deviation.

      2. Report the profile using "form only".

      3. Report the profile using "form and location"

      Each one could you give a different result.

      The "form only" will report the form error.

      The "form and location" profile could report the location if that has the highest deviation or it will report out the form error if that is the largest deviator.

      "T deviation will report out what it reports out all the time.


      And that is as far as I'm am willing to discuss profile. It has been beaten to death on the forum.
      FORM ONLY will allow Pcdmis to 'best-fit' the data to get it as balances as possible, basically, it makes up it's own alignment.
      sigpic
      Originally posted by AndersI
      I've got one from September 2006 (bug ticket) which has finally been fixed in 2013.

      Comment


      • #4
        Originally posted by Matthew D. Hoedeman View Post
        FORM ONLY will allow Pcdmis to 'best-fit' the data to get it as balances as possible, basically, it makes up it's own alignment.
        Yeah I guess that's right. I don't use form only too much.
        sigpic

        James Mannes

        Comment


        • #5
          You have three possible outcomes.

          1. Report the point out using a "T" deviation.

          2. Report the profile using "form only".

          3. Report the profile using "form and location"

          Each one could you give a different result.
          I actually find this alarming!
          Not so much as in number one, but the others should not deviate in reported values.
          How can reporting "Form" only deviate from reporting "Form and Location"!?
          In most (95%) cases customers want to know the Basic value of the reported form, hence form and location is important for our shop.

          Here is a question:
          When testing form with a single point entity, report the results as "Form" only and "Form and Location" then using the same entity report it with a True Position (which will give, essentially, form and location). After all, Profile and True Position share the same formula.
          2*Sqrt(X^2 + Y^2)

          You should not get different results between Profile (Form) and True Position!

          Please post some results here as I would like to see them.

          Comment


          • #6
            Originally posted by QC Rick View Post
            I actually find this alarming!
            Not so much as in number one, but the others should not deviate in reported values.
            How can reporting "Form" only deviate from reporting "Form and Location"!?
            In most (95%) cases customers want to know the Basic value of the reported form, hence form and location is important for our shop.

            Here is a question:
            When testing form with a single point entity, report the results as "Form" only and "Form and Location" then using the same entity report it with a True Position (which will give, essentially, form and location). After all, Profile and True Position share the same formula.
            2*Sqrt(X^2 + Y^2)

            You should not get different results between Profile (Form) and True Position!

            Please post some results here as I would like to see them.
            Lets start with something simple:

            Your profile is a close approx to a circle and that is the profile tou are checking. Lets say the entire profile is shifted 5mm in the Y axis.

            FORM ONLY will compare the form ONLY to itself. So, as long as the SHAPE is right, it will show in tolerance, no matter how far off location it is (thus the NO LOCATION in the dimension, form only)

            FORM AND LOCATION will show it being off. Depending on which version of Pcdmis you use, it will either be DOULBE the maximum deviaiton of an individual point OR it will be the MAX to MIN deviation difference. Wilcox/B&S waffle back and forth between version on which method is right, meanwhile claiming that what ever method is used in the most current version is the correct way.

            USING individual points and the T axis, this will give you ONLY the deviation of a single point, NOT a dimension for the entire peofile.

            Now, you CAN program single points instead of a scan and then do the following:

            Dinemsion the individual points using 1/2 of the total tolerance (ie tol=0.5mm, use +/-0.25)

            Construct a feature set of the individual points, then dimension it twice, once form and location, once form only. The form and location WILL give you the same results as the individual points, AS LONG AS YOU turn on MIN/MAX, those values will match exactly with the min and max from teh individual point dimensions. The OVER-ALL profile deviation will either be DOUBLE the greatest deviation (which simulates TP) or it will give you the MAX to MIN (depends on Pcdmis version). When you dimension the feature set FORM ONLY, Pcdmis WILL 'shift' teh alignmetn around to get the greatest number of points closest to the nominals, then report DOUBLE the max dev or the MAX-MIN differnece, same as above).

            However, you can see that you can get 3 different results from the EXACT SAME POINTS measured.

            Your best bet (IMO) is to forget the scan or feature set and measure/dimension individual points. This will give you a CONCREAT answer as to WHAT is wrong and WHERE exactly it is wrong. If you use the set and profile method, you will have additional work to do to get this information.

            Myself, I have NEVER seen a FORM ONLY dimension that I had to report, all of mine have been to ABC datums.

            (EDIT) Oh, and PROFILE will be a 3-D calculion, SQRT(X dev ^2 + Y dev ^2 + Z dev ^2) or SQRT ((X dev * X Vect)^2 + (Y dev * Y Vect)^2 + (Z dev * Z Vect)^2) depending on if you use SNAP points or not.
            Last edited by Matthew D. Hoedeman; 08-30-2007, 04:39 PM.
            sigpic
            Originally posted by AndersI
            I've got one from September 2006 (bug ticket) which has finally been fixed in 2013.

            Comment


            • #7
              Geesh!! Hey Matt,
              Where do you guys get these equations from? When I think I know something, I see a posting and feel like a newbie,,,,LMAO!!!!!

              A.Gore
              sigpicA.Gore

              Comment


              • #8
                Originally posted by ag162 View Post
                Geesh!! Hey Matt,
                Where do you guys get these equations from? When I think I know something, I see a posting and feel like a newbie,,,,LMAO!!!!!

                A.Gore
                Usually I just make them up, but in this case, they are correct. It is actually the formula for the radius of a sphere, which is what you need to get the 3-D vector deviation of a point.

                SQRT(X^2 + Y^2 + Z^2) = RADIUS of Sphere where X is the distance in the X axis and Y the distance in the Y axis and Z the distance in the Z axis from the center point of the sphere to a point on the surface of the sphere.
                sigpic
                Originally posted by AndersI
                I've got one from September 2006 (bug ticket) which has finally been fixed in 2013.

                Comment

                Related Topics

                Collapse

                Working...
                X