True Position Problems

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

  • True Position Problems

    Occasionally I get a couple of problems when reporting true positions (mainly holes)

    1) If reporting TP using polar radius and polar angle and using 3 datum features in the dimension sometimes the nominal polar angle gets rotated by exactly 90°. This obviously results in the feature being shown way out of position. The quick fix is simple, just "re focus" the dimension back to the feature.
    I don't know if this is just a problem when using polar co-ordinates, I can't ever remember seeing it with rectangular co-ordinates

    2) Occasionally a dimension as above (i.e. polar co-ordinates, using 3 datum’s) shows the measured polar angle to be exactly the same as the nominal. We are quite good at machining, but not that good!
    "Re-focusing" the dimension back to the feature does not alter anything but creating a new dimension using the same parameters as the dodgy one fixes the problem, i.e. measured and nominal now differ.

    Using PC-Dmis 3.2063 on a Ghibli 15 20 14 with a CW43L

    Hope someone can help

    Bob Fairbrother

  • #2
    Dimensioning TP using PA/PR is not practical even though the drawing states angle, because the nature of angle call out, I converted to Cartesian on all my TP report.
    But if that what you choose to do, then go to Edit/Preference/SetUp Options/Dimensional Tab/check 0º-360º at lower right corner of 'Angle Degrees'.

    0º starting from positive pole of first name axis toward positive pole of second name axis (Xy,Yz,Zx planes, the first letter of the plane is the first name axis), if you want to report angle from hole to hole then you have to rotate the coordinate for each individual hole do not use the actual holes but using basic angle from the original coordinate, another word you should be using one set of coordinate for the entire part. For v's 3.6,3.7,4.0,4.1, I don't know v3.2063

    Comment


    • #3
      It has been a long time since I used 3.2. I am using 4.2 now. I don't realy know what you are doing or what your part looks like. I hope some of this might help.

      True Position menu option calculates the true position
      from the feature to the X, Y or Z origin parallel to its respective axis. The feature's
      diameter, angle and vector are also part of the calculation.
      This section relates only to 'true position' dimensions. 'Position' dimensioning is
      discussed in the "Dimensioning Location" section on page 866.
      Important: If datums are used, the measured and deviation XYZ values are
      calculated with respect to the alignment of the datums, but are displayed in the
      current alignment in order to interpret the values. This means that a feature reported with one true position dimension can have different measured and deviation valuesthan another true position dimension if the dimensions have different or no datums
      defined, even if they have the same nominal values.



      Using Datums
      Although there are many different combinations of features that may be used as
      datums for True Position dimensions, there are some general principles to follow
      when selecting datum features.
      One common scenario for designs that utilize True Position dimensions is to use a
      circle or cylinder as the sole datum feature.
      Another accepted practice is to select a set of datum features that follow 3-2-1
      alignment principles. (Remember, the minimum definition for a datum is 3 datum
      points to describe the first datum, 2 datum points to describe the second datum and 1
      datum point to describe the third datum.) This means that the selected features would
      be a plane, a line, and then a single point.
      However, 3-2-1 alignment principles can also be used with plane/line/line,
      plane/line/circle, plane/cylinder/cylinder, and many other combinations.
      When multiple circular features have been used with the Maximum Material
      Condition (MMC) or Least Material Condition (LMC),
      · The bonus tolerance from the TP axis will not likely be a simple
      addition of other bonus tolerances, as a best-fit calculation will occur to
      determine the correct bonus tolerance.
      · Any circular or slot features may be measured using MMC or LMC.
      · All non-circular features are measured Regardless of Feature Size
      (RFS).
      Note: When datum features have been selected, the true position X,Y,Z, PA, and PR axes are calculated with respect to the alignment of the datums, but are displayed in the current alignment in order to interpret the values. Also, the datum features must be selected using the general steps discussed in this topic or else the dimension is likely to give unexpected results.
      sigpichttp://npmichaels.com/ (My sons book)
      THANKS (CAD IS ALWAYS MASTER)
      4.3 & 4.2MR2/GLOBAL/PH10MQ/WINDOWS XP

      Comment


      • #4
        Originally posted by lambo View Post
        Dimensioning TP using PA/PR is not practical even though the drawing states angle, because the nature of angle call out
        Why not? Enlighten me.
        PC-DMIS CAD++ 2o19 R1 SP11

        Comment


        • #5
          Originally posted by vpt.se View Post
          Why not? Enlighten me.
          It's time comsuming, too much unnecessary set up to print a report, let's say you have 60 holes 6º basic 60 places don't you have to rotate part coordinate 60 times to report each individual hole? You can build a loop but that will become a different issue.
          If the inspector performing a F/A and the holes are out of position, the operator would ask which way and how much then you have to convert it to cartesian not all machinist knows trig, it's much easier to print out Cartesian coordinate that makes both world happy.

          Comment


          • #6
            I concurr with lambo. It's fine to dimension using pa & pr, but not practical to anyone having to produce the parts.

            Readouts on cnc type of machine equipment move in x, y, z axis; hence the cartesian coordinates make more sense.

            ZydecoPete
            sigpic

            Comment


            • #7
              Originally posted by lambo View Post
              It's time comsuming, too much unnecessary set up to print a report, let's say you have 60 holes 6º basic 60 places don't you have to rotate part coordinate 60 times to report each individual hole?
              Uhm, no? I might be misunderstanding you? Unless you want the report to show the deviation from 0° and radius there is no need to rotate 60x.

              Originally posted by lambo
              If the inspector performing a F/A and the holes are out of position, the operator would ask which way and how much then you have to convert it to cartesian not all machinist knows trig, it's much easier to print out Cartesian coordinate that makes both world happy.
              Originally posted by Peter Warcholyk View Post
              I concurr with lambo. It's fine to dimension using pa & pr, but not practical to anyone having to produce the parts.

              Readouts on cnc type of machine equipment move in x, y, z axis; hence the cartesian coordinates make more sense.
              I totally agree with these reasons. However, I am in the 'other end' of this chain - not making the part. but 'buying it' - so I check it according to drawing. I totally concur with the fact that it is easier for the machining operator to adjust using the cartesian coordinates than the polar.
              PC-DMIS CAD++ 2o19 R1 SP11

              Comment


              • #8
                Originally posted by vpt.se View Post
                Uhm, no? I might be misunderstanding you? Unless you want the report to show the deviation from 0° and radius there is no need to rotate 60x.

                I totally agree with these reasons. However, I am in the 'other end' of this chain - not making the part. but 'buying it' - so I check it according to drawing. I totally concur with the fact that it is easier for the machining operator to adjust using the cartesian coordinates than the polar.

                Like I said, if you are picky then you have to report each individual hole as B/P called out, it could be 'equally spaced" or the angle for each individual hole from one axis or it could be an angle from hole to hole typ. or equally spaced AND cartesian BUT NOT from the datum but from one hole to another, it's all depends on the stupidity level of the designer, you've been around this field for a long time and I'm very sure you had seen many unexplainable call out even you combined all ANSI Y14.5 dating back to stone age when first wheel was invented and whatever they are using in China can't even explain it. The example from prev. post for picky type of inspector, it said 6º apart 60 places rotate coordinate 60x is exaggerating to make my point, may be 15x but if you are really really picky and you want to report each hole in sequence then yes, 60x.( recall an alignment is included)


                What's wrong with buying good parts with different report format, both are for sale?
                Last edited by lambo; 06-05-2008, 04:06 PM.

                Comment


                • #9
                  I agree with VPT.SE, I also use PR most of the time unles I need to make adjustments then yes you need to know XYZ otherwise PR is good enough.

                  Comment


                  • #10
                    Hey BOb are you using CAD? sometime cad can be off.

                    What I usually do is this: I create a perfect circle at the normal angle ex.
                    x0 y1.2 z-.04 This way I know that this circle is good (make sure all ijk are good) then i copy and paste with pattern. This way it iluminates all the errors.

                    Otherwise you have to calculate XY value for each circle.

                    Comment

                    Related Topics

                    Collapse

                    Working...
                    X