hexagon logo
Register Sign in
  • State Verified Answer
  • Replies 10 replies
  • Subscribers 226 subscribers
  • Views 95 views
  • Users 0 members are here
Options
Related

Here's a fun brain teaser regarding Profiles AND Datums!

BIGWIG7

ok. I have a simple bracket type part with some flat parallel and perp surfaces. There is a PROFILE callout to -A,B-. When reporting using both -A,B-, profile is OOT. When reporting using just -B-, results are conforming. 

ALL surfaces are perp and parallel to each other within .001. 

Why does -A- have such a massive impact on profile assuming -A- is 1) flat within .0005''(inches)  2) perp to -D- within .0005'' (inches)

Results shown are reported in metric. 

when using -A,B- 

when using -B- only



CLEARER DATUM PIC
[edited by: BIGWIG7 at 3:03 PM (GMT -5) on Jun 17, 2025]
Parents
  • 0

    -D- PROFILE to -A-  only measures .0005''

    -D- PROFILE to -B- measures (as noted above) .0044'' (you can see the excess material towards the tip of -D- and we can verify the extra thickness by hand)

    but combining -A,B- throws it off.    

    I'm trying to understand how the combined datum structure sends it over the edge. 

  • 0 in reply to BIGWIG7

    Its good to A alone and good to B alone right? 

    Is A twisted a bit creating projection error?

  • 0 in reply to ArkansasBeeper

    1)YES GOOD TO A, B ALONE

    2)possibly. 

    i think this is just a case of staring at the part for so long these past few weeks (few hundred thousand parts production) and not coming to terms that the two datums combined simply push the profile readings  OOT. 

  • +1 in reply to BIGWIG7

    When you do the profile to A, you are really only capturing the perpendicularity because A can only control two rotations (think the LEVEL) and translation along it's axis (which will have no effect on this particular profile anyway).
    .

    When you do the profile to B, you are capturing parallelism and the distance between B & your surface.

    .

    When you do the profile to A|B (as it's supposed to be done), you are introducing additional constraints to the DRF because it first rotates and translates to A, then rotates and translates to B whilst maintaining it's orientational relationship to A - imagine sitting the base of an angle block on the A surface and then sliding and rotating it until the side face is making maximum permissible contact with B.  From this DRF you are then capturing the perpendicularity, parallelism, form and location of your profile surface.

  • +1 in reply to BIGWIG7

    Profile to B is already near max. By combining the two, you are now wrapping in perpendicularity, parallelism, and form and location. Seeing you are already near max with just to B. It makes sense. 

    (And I see Neil chimed in with pretty much the same thing with more eloquent language.) 

Reply Children
© 2025 Hexagon AB and/or its subsidiaries. Privacy Policy | Cloud Services Agreement