Curious about linear dimensions and alignments

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

  • Curious about linear dimensions and alignments

    Hey guys
    I've got a question about linear dimension and how to align correctly to them to get the most "correct" way measure them.
    Ive got two examples:

    Example one:
    You got a square block, lets say 100x100x100 (L x H x W) and you want to measure the lenght 100 (direction of Y+ in machine/program).
    I see two ways to measure it in CMM.

    1. Either measure one side (plane) of it (Y+/Y-) and align it as primary. Then evaluate the 2D lenght between the Y+ and Y- planes.
    This would simulate you measuring it with a height gauge. (since you place it on the primary plane and then place the tip of the height gauge on the other plane.

    2. You measure the top plane (Z+), align it as primary, then measure a line in Y+ direction and rotate Y+ around Z+. And then evaluate 2D distance between Y+ and Y- planes.
    I cant find a way this would simulate any manual measurement equipment, but a lot of people tend to use this method.

    Example two:
    You want to measure the distance between two inner holes. They are both supposed to have the same center in example X.

    1. You do as the first example. Primary on a plane in the direction you want to measure, and then evalute the distance between the holes(circles).
    2. construct a line between the circles, and use this line as rotation. Then evaluate the distance between the circles.

    First example simulates a height guage again. You rest the plane on primary, Zero on the lowest point on the height guage, and measure the distance to the lowest point of the other circle, then add/remove radius.

    Second example simulates a caliper, since you use the direction of the circles (wether they are in center or not) and then add/remove radius.

    I really dont like linear dimensions, they confuse me. Alignment, type of element (plane, point etc).
    Any one got some good sources of information for this? (besides expensive courses)



  • #2
    The fact you're thinking about these dimensions and the numerous ways of checking them tells me you'll be okay.

    Programming a well dimension part IMHO take about 50% of the time it takes to measure a rubbish one.

    This is why I love GD&T, if done properly then these sort of ambiguities don't exist.


    For linear dimensions as you've discussed above, you can use either of the methods and argue that you're checking to drawing.
    Applications Engineer
    Hexagon UK

    Comment


    • pcdmisstudent
      pcdmisstudent commented
      Editing a comment


      GD&T is great, but I feel like 50% of the time I get GD&T callout its a bogus one.
      Like you kinda know what they re trying to say.

      Like if you take that 100 mm block, and drill thru 3 holes. Do a callout for true position without a datum but but a boxed dimensions between each one.
      Either I have to make my own datums or I just try to simulate whatever the operators else would measure with. (they dont like cmm showing different numbers than ex. a micrometre.

  • #3
    That's the drawing which should give you the method...
    Let's say you have a double arrow with 100±0.1, then you should use size dimensions between hits, maybe aligning the dimensions along the vector of a mid plane...
    If you have a datum and a boxed dimension with a loc, then you should construct a tangent plane (L2 constraint in ASME world, minimax (L1 in ISO)), and check if all the hits are in the tolerance zone.
    There are many solutions if the draw is bas, there should be only one if the GD&T is correctly used...

    And don't forget that a dimension is often the distance between the centroid of the first feature to the centroid of the second, projected on the vector of the second.
    (Dist PL1/PL2 = ABS(DOT(PL21.XYZ-PL2.XYZ,PL2.IJK)))
    Last edited by JEFMAN; 02-24-2020, 02:20 PM.

    Comment


    • #4
      Originally posted by JEFMAN View Post
      That's the drawing which should give you the method...
      Let's say you have a double arrow with 100±0.1, then you should use size dimensions between hits, maybe aligning the dimensions along the vector of a mid plane...
      What vector would that get? Since both vectors are opposite?

      Originally posted by JEFMAN View Post
      If you have a datum and a boxed dimension with a loc, then you should construct a tangent plane (L2 constraint in ASME world, minimax (L1 in ISO)), and check if all the hits are in the tolerance zone.
      I understand that it would be easier with GD&T. But I dont know what you mean about that L2 and L1 thing..

      Originally posted by JEFMAN View Post
      And don't forget that a dimension is often the distance between the centroid of the first feature to the centroid of the second, projected on the vector of the second.
      (Dist PL1/PL2 = ABS(DOT(PL21.XYZ-PL2.XYZ,PL2.IJK)))
      Is there a difference between a centroid of a plane and a gauss plane? So the centroids vectors gets the same vectors as the 2nd one?
      Also i dont know that code, is that how you would evaluate a 100mm <-----> linear dimension?

      Comment

      Related Topics

      Collapse

      Working...
      X