Surface or Vector Points?

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

  • Surface or Vector Points?

    Alright I'm having a discussion with a counter part. We are using CAD models to profile T values. They are saying they get .002-.003" different when switching to surface points from vector points. Any idea why? What do you use?
    I used to be high on life but I built up a tolerance.

    Brown & Sharpe Global Advantage
    PCDMIS CAD++ v2011mr2
    PH10MQ/SP600M


    sigpic

  • #2
    Vector points drive to the surface at the vectors defined by the Cad model.
    Surface points sample the surface (with the number of points you define when creating the point) then the final point drives to the surface at the vector defined by the sample points.

    If the actual surface deviates significantly from the CAD defininition (angularity) you will get slightly different values for T.

    .002-.003 sounds like a lot......

    I always use Vector points unless I am doing points for a very tight profile tolerance (like profile .002)
    Lately, it occurs to me
    What a long, strange trip it's been.

    2017 R1 (Offline programming)

    Comment


    • #3
      I was using vector points and now they are telling me to use surface points because of the better results. We are not using sample hits. I can't say I'm too confident in the results now.
      I used to be high on life but I built up a tolerance.

      Brown & Sharpe Global Advantage
      PCDMIS CAD++ v2011mr2
      PH10MQ/SP600M


      sigpic

      Comment


      • #4
        I am not sure what you get if you are not using sample points.
        Why are they different?
        What makes one more accurate than the other?

        From the HELP files

        Last edited by John Riggins; 04-20-2009, 03:43 PM.
        Lately, it occurs to me
        What a long, strange trip it's been.

        2017 R1 (Offline programming)

        Comment


        • #5
          like John R. said vetor points is the ideal if you have a cad, if no cad I use surface poinnts with sample hits.

          Comment


          • #6
            I would highly question them on their observations. There is NO difference in the calculation of a surface over a vector point. They are either : 1/High 2/ Seeing things 3/ Delusional, or its possible ( But I doubt it) that there is a glitch in the software causing what they are seeing.
            Jim Jewell

            Comment


            • #7
              unles he's got sample hits instea of the single hit.

              Comment


              • #8
                No sample hits. I can't see how it can make a huge difference seeing that I'm using CAD anyway.
                I used to be high on life but I built up a tolerance.

                Brown & Sharpe Global Advantage
                PCDMIS CAD++ v2011mr2
                PH10MQ/SP600M


                sigpic

                Comment


                • #9
                  ompare xyz and vector values from both points? they shoul be the same

                  Comment


                  • #10
                    Vector points are great, IF you have refined and refined the alignment already. (yes, two times at least). Otherwise they're meaningless.
                    sigpic
                    Global Advantage 12-22-10
                    TESASTAR M SP25 4.3mr2

                    Comment


                    • #11
                      what do you mean refined ?

                      Comment


                      • #12
                        All the points and nominals are coming off the CAD model I don't see how you can refine that.
                        I used to be high on life but I built up a tolerance.

                        Brown & Sharpe Global Advantage
                        PCDMIS CAD++ v2011mr2
                        PH10MQ/SP600M


                        sigpic

                        Comment


                        • #13
                          I would ASSUME Keego is telling you to either loop your alignment or measure the alignment twice to make sure the part is aligned as near perfect as possible. If you are using plane-line-line or plane-line-point to align your part once thru may not square the part to absolute zero/zero. If your part is not square on the CMM you may see some cosine error in the results.

                          .002-.003" is an awful lot of variation to see between the 2 measurement techniques. Run your program the way it is then change several of your vector points to surface and see what difference it makes. If between techniques is within tenths something else is wrong.

                          Just my .02

                          Duane
                          Xcel & MicroVal Pfx & Global 37mr4 thru 2012mr1sp3
                          Contura Calypso 5.4

                          Lord, keep Your arm around my shoulder and Your hand over my mouth. Amen.

                          Comment


                          • #14
                            Originally posted by Underspec View Post
                            Alright I'm having a discussion with a counter part. We are using CAD models to profile T values. They are saying they get .002-.003" different when switching to surface points from vector points. Any idea why? What do you use?
                            The T value is calculated using the I J K values. So the surface and vector points if they have a different vector will change the T value.

                            T Cal.zip

                            This will show you How it is calculated.
                            sigpichttp://npmichaels.com/ (My sons book)
                            THANKS (CAD IS ALWAYS MASTER)
                            4.3 & 4.2MR2/GLOBAL/PH10MQ/WINDOWS XP

                            Comment


                            • #15
                              That also depends on the radius of curvature and the shape of the part. A point of surface gives good performances for spherical radius, less on cones or cylinders, nor on small radius.
                              That comes to the fact that the micro-plan of the point of surface is or is not tangent on the surface locally. On a cone, for example, the surface point creates a plan with 3 points. The plan is a triangle whose base cuts the surface, so the vector is not perpendicular to the surface. At all events, the difference should be small !
                              You can compare the vector values between the 2 methods, to check if the deviation comes from the vector or from another parameter (a defect of the CAD, for example).

                              Comment

                              Related Topics

                              Collapse

                              Working...
                              X