GD&T for Assembly

  • Filter
  • Time
  • Show
Clear All
new posts

  • GD&T for Assembly

    I need to check the surface profile of a box assembly to see if it is within tolerance. The problem is that I can't measure all of the surfaces while the assembly is put together and one of the diatoms is on a different piece (the lid) of the assembly.

    What I want to know is if I can measure the surface profile to the plain that will be connecting the lid (Dat-A) Then measure the flatness of both connecting surfaces to see how far off my number could be.
    Is there some kind of algorithm for figuring this out?
    I know I won't get exact numbers this way, but if both surfaces are REALLY flat then I should be able to know to within a high level of certainty.

    I have a link to some pictures in my drive if you think they will help. I know there are certain probe tips and attachments I could buy in order to measure it all at the same time, but this isn't a part we are making. We were asked to measure it as a one-time thing.

  • #2
    This is going to sound stupid and snarky but if it assembles it should be OK. You really can't tell anything about X if you are measuring Y. You'd have to do massive correllation studies and even then the data would be highly suspect.


    • Calvin.Korver
      Calvin.Korver commented
      Editing a comment
      I know the assembly fits together, they have actually been using it in production for some time now. I'm not sure why they want it measured now maybe they are having a problem with it.
      But there should be a way to measure both parts independently to see how the surface profile reflects on both pieces. Otherwise, how could engineers design and dimension press fixtures

  • #3
    Possibly, yes, you use a datum simulator.

    Before CMM's (and in a lot of shops that don't own a CMM), you bolt/clamp parts to an angle plate or the granite surface plate and measure from that datum simulator (unless you have to do tooling points).

    If -A- sits on the surface you want to use, AND the inside of this lid isn't functional, you can easily get away with making a plane on the surface. If something else is sitting on -A- inside this container that could interfere with the container and -A- isn't flat, you have a bigger problem.

    If you are trouble shooting, put the part on a granite, and lightly move it back and forth. This makes shiny spots.

    Those are the actual spots where contact is made and where -A- actually, functionally interfaces with the rest of the assembly. Make your plane there.

    If the surface you are mating to on the assembly is really out of flat (up to design based on usage) you will have a lot of problems, go for the min-sep math method and a lot of points.


    • #4
      Can you separate both parts ?
      If yes , I would create an alignment on the lid, then measure a lot of poins on it.
      Then I would measure without the lid the same coordinates on the other part.
      I would then measure the lid alone, with the same alignment than the first, but allowing measuring both sides, with the same coordinates. Export the inside points in a cop, or write coordinates in a file.
      Then place the lid, and measure only the alignment.
      Import the file as generic points and measure the distances between hits of the other part and genric points.

      You can also export all hits, and calculate it with Excel...


      • #5
        Thanks, guys! For this job, I talked with the company's fixturing engineer and he said it would be fine to switch the datums up (it's actually better now). But I will defiantly keep this in mind, I dought it will be the last time I run into this we make a lot of fixtures over here.


        Related Topics