18 Replies Latest reply on Oct 12, 2018 10:51 AM by Brian Brazeau

    Avoiding Over-Dimensioning

    Uchenna Offor

      Hello all,

       

      So I am starting to apply GD&T to my drawings since I wasn't too much of a fan of it in the past, so I am currently reading some books about it. I am not knowledgeable yet, but I am trying to apply the concepts as quick as I can. My issue with it is that it can make the drawing pretty crowded and cause over-dimensioning which is a no-no for you manufacturers out there. My question is how do you guys balance GD&T with dimensions especially with complex objects?

       

      All these cute videos on YouTube work with simple objects which is not my reality. For example, the below is what I am trying to practice on and it's not a simple object where you can easily find your 3 datums if you aren't well versed in GD&T (end mini rant).

        • Re: Avoiding Over-Dimensioning
          James Riddell

          What (IMHO) you need to understand is that GD&T allows you to minimize dimensioning and draw attention to the functional aspects of the part/assembly, i.e. what it really needs to make things work properly. This allows you to open up tolerances on dimensions which are not critical to the function. This in turn, allows manufacturing cost reduction.

           

          Edit: you don't always need 3 datums and for some applications of GD&T.

          • Re: Avoiding Over-Dimensioning
            Shawn Stugard

            Please post a screenshot (JPEG or similar) of the drawing, some of us (ahem...me!) can't DL zip files right now... for various reasons.

            • Re: Avoiding Over-Dimensioning
              Brian Brazeau

              I would say.... Your Top Plane is Datum A, The Axis of your 1.85 diameter is Datum B, Your Plane1 is Datum C

                • Re: Avoiding Over-Dimensioning
                  Brian Brazeau

                  Opps! Right Plane is Datum C

                  • Re: Avoiding Over-Dimensioning
                    Uchenna Offor

                    Wow. I got close then, but I used a width feature between the top plane and the plane at the top of the chassis as my datum A. Does that make sense to you? My thought was that I was supposed to limit translation with the primary datum.

                      • Re: Avoiding Over-Dimensioning
                        Brian Brazeau

                        Using the Top Plane constrains 2 axis of rotation and 1 axis of translation, Using the Axis of the 1.85 diameter constrains 2 axis of trans lation, Using the right plane constrains the last axis of rotation. I Believe.

                        • Re: Avoiding Over-Dimensioning
                          James Riddell

                          Uchenna Offor wrote:

                           

                          Wow. I got close then, but I used a width feature between the top plane and the plane at the top of the chassis as my datum A. Does that make sense to you? My thought was that I was supposed to limit translation with the primary datum.

                          Whoa there, ALL datums must have some basis in reality. You will have QC screaming at you 24/7 if you choose arbitrary planes as datums if there isn't a darn GREAT reason.

                           

                          Again - not all GD&T controls require datums. Since this appears that it might be a tripod joint, unless it is supposed to be extremely accurate, general tolerances would probably convey the intent.

                            • Re: Avoiding Over-Dimensioning
                              Frederick Law

                              James Riddell wrote:

                              ALL datums must have some basis in reality. You will have QC screaming at you 24/7 if you choose arbitrary planes as datums if there isn't a darn GREAT reason.

                               

                              I got lots of drawings from automotive with datums at middle of nowhere.  Some even put datum where I'll machine it out.  Try to support the datum point and machine it at the same time.

                                • Re: Avoiding Over-Dimensioning
                                  James Riddell

                                  Frederick Law wrote:

                                   

                                  James Riddell wrote:

                                  ALL datums must have some basis in reality. You will have QC screaming at you 24/7 if you choose arbitrary planes as datums if there isn't a darn GREAT reason.

                                   

                                  I got lots of drawings from automotive with datums at middle of nowhere. Some even put datum where I'll machine it out. Try to support the datum point and machine it at the same time.

                                  Haha, yup, if you are a practitioner in almost any industry that produces real parts you will see some doozies. Now a projected point (i.e. using a gage pin to verify the CL is as close to perpendicular as possible) can be a very viable GD&T feature.

                                • Re: Avoiding Over-Dimensioning
                                  Uchenna Offor

                                  Ok thanks, especially the part about imaginary datums. Only a few parts of the tripod are supposed to be accurate: the middle hole and the three areas where the legs will be so I'll focus on those.

                                    • Re: Avoiding Over-Dimensioning
                                      Brian Brazeau

                                      If this is to be an injection molded part or even a machined part you might want to consider redesign of the 3 rods that connect one wall to the next on the underside of the part. They are not readily producible by those means.(although they could be pressed in as separate components)If the rods play into the dimensional scheme of this part (i.e. they interface with other parts in the assembly and therefore must be accurate) you will probably have to set this part on the C.M.M. with the top plane facing up. Then pick up would be: 3 points on the top plane face, which is now exposed, to establish the plane (primary datum). 2 points, one as high as reachable and one as low as reachable on the 1.85 diameter to establish the axis. (secondary datum) Note the 2 points will each be established as the center of 2 circles made by touching off on a min of 5 points (but always an odd number) on the inside diameter. Finally 1 point at the midpoint of 2 points by touching off on either side of the inside wall of a leg (because I assume a male leg fits in there) to establish the (tertiary datum). Then use basic dimensions for the 120 degree angle between legs and other dimensions to define the required tolerances. Be careful when dimensioning things like say the inside of the legs with just a tolerance as the planes formed between the 2 inside walls could be the correct distance apart from each other but if geometric tolerance does not constrain their perpendicularity to say the axis (secondary datum) the legs could end up slanted with respect to the main shaft and still meet the print. This is the kind of thing geometric tolerances ensure.

                                        • Re: Avoiding Over-Dimensioning
                                          Uchenna Offor

                                          Wow that was very specific and informative! The rods on the underbelly are there to stop the legs from moving further back and don't need to be very accurate, since I designed the legs in such a way that the back of the legs will hit the rods when someone is opening all the legs of the tripod.

                                           

                                          Will the same datums that we established in the previous comment still work for this setting?

                                            • Re: Avoiding Over-Dimensioning
                                              Brian Brazeau

                                              I would say yes. Your still going to want to geometrically dimension the inside walls of the leg cutouts to maintain some degree of parallelism to the main axis. I wouldn't get too strict with it as it is probably not a mechanical strength issue as much as it would be a cosmetic issue if they were out a great degree. I don't think there's any rule that says you can't put a standard tolerance on the width of the slot and a geometric tolerance on it with regard to a datum you would just need to use max or min material condition modifier. (I think.)

                                  • Re: Avoiding Over-Dimensioning
                                    Rick Becker

                                    Think about...

                                    Form, Fit, and Function.

                                     

                                    What are the features or surfaces will be touching other surfaces on other parts when this is fully assembled?. How accurate and how perfect must the form (shape, clearance, geometry (position/straightness/circularity/profile/flatness/angularity/runout/etc.)) be between these adjoining parts?

                                     

                                    Tolerancing of dimensions works side-by-side with geometric tolerances.