2 Replies Latest reply on Feb 15, 2016 5:34 PM by Alessandro Tornincasa

    The "Spark" to push MBD adoption

    Alessandro Tornincasa

      During my customer visits I've more and more found out that, in the country where I live, companies are relying mainly on "cordinate" dimensioning and just plus/minus tolerancing.

       

      If you dimension a part with DimXpert by using plus/minus tolerances and show tolerance status it will show the part as undefined.

      Take the example of this shaft: these are used datums:

      DimXpert datums.png

       

      And this is the tolerance status when using Plus/Minus dimensioning scheme:

      DimXpert plus-minus tolerance status.png

      Cylindrical surfaces are undefined.

       

      This is not a fault in the software, it is correct, because there no tolerances to define form, location, and orientation of the surfaces.

       

      When using GD&T dimensioning scheme the shaft is completely defined:

       

      DimXper geometric tolerance status.png

       

      When a designer uses DimXpert the software will take care of all the hard work of completely defining the part and check consistency of dimensioning respect to GD&T rules.

      This will ease desginers' life.

      A more detailed explanation of why GD&T is better than coordinate dimensioning is here:

       

       

       

      So we should push more the importance of using GD&T.

       

      Alex

        • Re: The "Spark" to push MBD adoption
          Kevin Chandler

          Hello,

           

          <diatribe>

          DimXpert can be handy for populating GD&T into the model data.

          But being for parts and not linked in with mating assemblies, it's still reliant upon the human for the required GD&T data.

          Having only a single part, adding GD&T things (by SW or it's driver) doesn't take advantage of GD&T's optimizing for mating.

           

          You show MMC & MMB for A on all of the other ODs.

          How does SW know that either or both of these conditions are required and at the tolerance indicated?

          How does SW know to use positional instead of profile, runout or concentricity?

          It doesn't know any of this. And w/o the mating part, it never will.

           

          Why aren't the faces GD&T? One end is held square to 0.10, but all other faces' profiles are allowed to float to size: +/- 0.50. That's 1000% of the one toleranced face. I don't know, and SW doesn't know, the reasons for this relative "looseness".

          Why isn't there a cylindricity applied to datum A? Its profile is currently held to its limits of size, which is twice the tolerance of its dependents at LMC. (If SW did have even rudimentary GD&T knowledge, it should have, at a minimum, prompted you for this.)

          Why is datum B created? And why there? (I don't know and neither does SW.)

           

          GD&T is supposed to allow for tightening of tolerance where needed while loosening them where not, but a lot of shops have see GD&T parts so overtoleranced across the board that they're hesitant about working on these jobs since the cost for compliant parts is high and arguments generally ensue. Some will even no quote a GD&T job (especially one-offs) unless you've worked with them before.

           

          Having GD&T in SW (& every other CAD program) is a hindrance, not boon, for GD&T's health as well, since it permits clicking without thinking.

          I'm not picking on GD&T or anyone applying it, but just like all of the other features SW is loaded with, just because you can model it, doesn't mean it can be made, that someone will want to make it or that they'll do so in your timeframe and budget.

           

          I think GD&T's adoption is being slowed and hampered by its misuse and misapplication.

           

          GD&T is not software dependent. It's human dependent.

          GD&T advances when its humans advance. (Or we reach reach singularity, then it's all for naught.)

           

          "The difference between what I want and what I get is the difference between what I mean and what I say."

          This truism applies when you're speaking GD&T.

          </diatribe>

           

          Cheers,

           

          Kevin

            • Re: The "Spark" to push MBD adoption
              Alessandro Tornincasa

              Kevin,

              thanks for replying to this discussion, I agree with you say about GD&T, the software cannot replace decisions taken by a designer.

              What I like of MBD and DimXpert is that they opened me to the world of GD&T, infact I'm in my work in progress of studying them, and I think it won't take me weeks, but it will take me several months to get more in-depth knowledge of this discipline.

               

              I think that with auto-dimension scheme, MBD asks the designer "human" intervention, because those 3 references are fundamental and deeply tied to the function of the part or how it will be inspected.

              For what concerns how does DimXpert know what tolerances to put, here's an explanation of the logic it uses for geometric tolerances:

              2016 SOLIDWORKS Help - How Auto Dimension Scheme Works

              I don't always agree with its choiches (for example I tend to use profile surface much more than DimXpert does) but I can alway delete tolerances and put my own.

               

              For what concerns geoemtric tolerance values, they are defaults, and it's up to the user to change them: obviously there are some rules and pratices you use according to the component function in the assembly, and the software cannot know them.

              For example it surely cannot know that positional tolerance values are different in the case you have bolts, or in the case you have threaded holes.

               

              So I would say it's good it gives you some guidance and avoid macro-errors to start putting the GD&T, but the guesswork is still left to the user because obviously GD&T is a difficult science and you need to be an experienced designer to master it.

               

              Thanks for your contribution to this discussion.

               

              Alex