Compound Datum Alignment Question

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

  • Compound Datum Alignment Question

    So I have a few questions about how to align this part using a compound datum. So from what the datum reference frame tells me is that the Primary is A-D, Secondary is B, and Tertiary datum is C. Now from my understanding of 321 alignment I would A-D Level, A-D Rotate to B, and C Origin. But I am not sure how to get A and D to be one datum. I asked a CMM programmer from a different plant how he would go about aligning the part. He said he would Level B Rotate A to C and Origin AB. He bounced make and forth with Datum D. First he said rotate to it then he said level it. Finally he said level all the features then rotate A to C. Now I am confused and trying to clear my head. It could be my understanding but I thought the Datum Reference Frame told you the order of the datums as in first letter in the frame is Primary and second is secondary and so on. So you always Level the Primary and rotate the secondary and so on. The confusion is probably on my end as I am using the word "always". Could someone explain to me what I need to do to align this part and maybe have a small conversation so that I better understand the what and why. Thanks in advance to anyone who responds. If the print is too small to read I apologize, let me know and I will try to upload a better version. Thanks again.
    Attached Files

  • #2
    Can't see the image but what you think is right. Datums work in order of precedence as they appear in the control frame. Each datum constrains all degrees of freedom it can.

    Think Can, May, Must!

    If a dutum can constrain a degree of freedom, and it may (i.e. that degree of freedom hasn't already been constrained by a datum with higher precedence), then it must.



    Automettech - Automated Metrology Technology

    Comment


    • #3
      No image to be seen here either.

      Ninja has the rule of thumb correct, but it is not explicitly stated as such in the standard and many people applying GD&T to drawings do not understand or follow the rule of thumb.
      This is a weak spot in ASME GD&T circa 1994. 2009 is slightly better, providing controls if the designer chooses to use them to specify exactly which degrees of freedom each datum constrains.

      The compound Datum A-D will have to be a construction. I have most often seen this as two diameters on either end of a cylinder. Measure each as a cylinder or circle, (cylinder is often the more correct choice by the standard, but if either is long or the part is less than straight, one circle at the extreme end of each datum surface can be the best way to construct the compound datum.), then construct a 3D best fit line between them.

      Level to that 3D line and set it as the origin of the two axis it can constrain. The secondary will pick up whichever rotation or origin it can and the tertiary gets whatever is left.

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

      Comment


      • #4
        I've been reviewing the drawing................Alas, I can't make 'heads nor tails' of this one. Datum A is a cylinder (Radius) and Datum D is a point on that cylinder; which would tell me to 'clock; the part in such orientation. But also you have Datum B, which is the surface of the cylinder-on the same vector of Datum A. Then you have Datum C, radial cutout, which would constrain the final Degree of rotational freedom. But to define A-D datum alignment would be far fetched; I think this part is overly defined (Constrained)--Conflicting Datum Structuring. Just my thoughts though.

        Comment


        • #5
          Here is another upload of the print. I can see it in my post and click on it to get it, but this is my first time uploading an image to this forum so I might of made a mistake. @Brain.freeman, Yes the A and B Datum are the same vector or surface. This caused me some problems in my train of thought, Level Rotate Origin. It's like A-B should be a Datum not A-D. But I am not an expert in blueprints or PCDMIS...yet. Thanks for the replies, I look forward to more. I hope the new upload of the print works, let me know if it doesn't.
          Attached Files

          Comment


          • VinniUSMC
            VinniUSMC commented
            Editing a comment
            That's not any better. You should try hosting the image somewhere, like TinyPic, imgur, or photobucket, and then link to it.

        • #6
          http://i65.tinypic.com/2agvryb.jpg


          Hopefully this will be better for everyone.

          Comment


          • #7
            http://i65.tinypic.com/2agvryb.jpg

            Comment


            • #8
              -A- is the 93.05/94.55 diameter

              -B- is the "Bottom" surface

              -C- is the "width" of the |R0.5| (BASIC 0.5 Radius)

              -D- is the Datum Point at |70.4| |17.8|

              I have no idea what A-D is.

              B&S CHAMELEON/PCDMIS CAD++ V2011

              There are no bugs, only "UNDOCUMENTED ENHANCEMENTS!"

              sigpic

              Comment


              • #9
                'D' is something to do with a point on the flat area in side view reference boxed dimensions 17.8 & 70.4 it's strange as boxed dimensions normally have a boxed tolerance linked to them.

                I would level and datum 'B' datum in 'A' in 2 axis and set 'C' on the centrelines of axis. If I was you I would go back to engineering and ask them to explain what D actually is

                Comment


                • #10
                  D is a "supplemental" form datum (at least, according to Chrysler). On a fixture, you would clamp the part onto A, B, C, and then one last clamp on D to constrain the part. How to really accomplish this in PC-DMIS? I'm having my own issues with customers on this.

                  I have a part with a nice ABC datum structure, and then D surfaces at 90 degrees to A, and parallel to B-C. The callout is A|B|C-D... Physically, it is easy to do. Making PC-DMIS do it though, not fun.

                  I would Align to ABC, then measure D. Give the point at D a very small tolerance (+/- 0.05mm). If the point is in tolerance without clamping, then go. If it's not in tolerance, modify your setup to physically constrain the part until ABC is good and D is simultaneously good.
                  "This is my word... and as such is beyond contestation."

                  Comment


                  • #11
                    ' D ' is on the inside face see left view at a distance of 50.8 (section 'A'). This would appear to be for the rotation alignment. Imagine a hard gauge were the part was free to rotate about and a 'A' stop pin at the
                    positions 70.4, 17.8, 50.8.

                    Comment


                    • VinniUSMC
                      VinniUSMC commented
                      Editing a comment
                      C already stops rotation. D is to constrain that flat surface parallel to the axis of A.

                    • UKCMM
                      UKCMM commented
                      Editing a comment
                      Not sure I agree with that, the OP needs to have clarification from the design authority.

                  • #12
                    Measure A as a cylinder (good luck with 95mm X 10mm) and do A/B/C alignment. Level A, rotate A to C, two axis origin at A and one at B
                    Now measure datum point D at basics, offset by basics. Offset point should be on same centerline as datum A, create line from datum A and offset point D and use it as your primary datum.
                    sigpic...engineering

                    Comment


                    • #13
                      Thanks for all the input from everyone. I came from sheet metal using only fixtures to measure a part to this new place which doesn't use fixtures to measure the part (part is put on setup blocks) and it's pipe (mufflers , exhaust). The pipe datum schemes are throwing me for a loop. Obelix, this is what I was looking for. I am new to compound datums but from my understanding, it's a single datum created by two or more features. I just wasn't sure how to get both features to represent a single feature through construction. UKCMM, yes I do need to contact the design engineer as this is what caused my confusion. I have a senior CMM person within corporate that I discuss prints with and he wasn't sure if it D was for the rotation or not either. Thanks to everyone who commented! I really appreciate everyone's insight and advice.

                      Comment

                      Related Topics

                      Collapse

                      Working...
                      X