12 Replies Latest reply on Feb 20, 2017 12:02 PM by Mark Biasotti

    Master Model Best Practices

    Mark Biasotti

      Thank you all who attended my "Master Model Best Practices" session at SWW2017 on Monday at 4:30.  I've come across a new discovery this week that validates my conclusion that the Save-Bodies technique is more robust than Insert-Part method. Don't get me wrong, Insert-Part is necessary when you want to bring along other geo besides solids, but while working on a client model this week, that is using the Insert-part method - all my derived parts (14 of them) were suddenly failing big-time. Upon further discovery, every insert part feature was failing to bring in the solid body. I did some debugging by creating a new blank part and inserting the master into it; SW let me select the file and I checked all the options I wanted, including Solids, but when okaying the feature, it failed to insert nothing.

       

      What turned out to be the issue was that earlier that day I created some features (you guessed it - a fillet in just the right place with just the right value) which cause a general Fault in one of the bodies. This general fault (even with VOB on) was not evident until I did a model check.

       

      What I learned from this is if you do a Save-Bodies, the derived part/s do not fail like they do when you use the Insert-Part Method even with a General fault in the the master part. What is deviating about this is that it was only one feature of one body of the 14 that failed all 14 derived parts (because insert part brings in all bodies even if that one is not used for the derived part.)  First glance at my assembly was quite a panic.

       

      This experience shows that the Save-Bodies method of inserting just the Stock Feature in the derive is a lot safer than Insert Part. Now, I know what your saying "Mark, the Insert Part method showed you a general fault that you would not have seen if you'd used Save-Bodies" - yes, there is merit in that, but I probably would have discovered that if I had continued modeling that any simple modeling op on that body would have not worked (which is something that you should know if you don't - that if you are doing a cut or extrude that is failing, and you are scratching your head as to why on earth this doesn't work? - go check the model.)

       

      I've attached a reproduced example file of the master (sorry can't show actual client data) and you can play with it yourself by suppressing and unsuppressing the labeled fillet feature and try to insert the master into a new part. This is a SW2015 part/s (since that is what my client is using) I haven't verified this on SW2016 or SW2017.

       

      Mark

        • Re: Master Model Best Practices
          Paul Salvador

          Thanks Mark for the debug on these two methods.. I'm interested in understanding how/why?...

          So.. let me get this right,.. VOB (in 2016 and 2017 and/or past versions?) was not catching the fillet fault with a Insert-Part  (and it corrupts the shared geo data?) but the fillet fault is present using Save-Bodies and transfers ok (and it's acceptable)?..

          Is there a SPR# for this?

           

           

          ..ok, now I'm confused,.. I opened this in 2015/2016/2017 and see the fillet general fault in each version,.. have VOB on with each version.... do a ctrl-q and general fault is gone?   Why did this fillet ever have a general fault?

            • Re: Master Model Best Practices
              Mark Biasotti

              Hi Paul,

               

              - VOB (in 2016 and 2017 and/or past versions?) was not catching the fillet fault with a Insert-Part  (and it corrupts the shared geo data?)  Yes in 2015 (have not tried 2016 or 2017)

              - but the fillet fault is present using Save-Bodies and transfers ok (and it's acceptable)?.. MAB - Yes

               

              do a ctrl-q and general fault is gone? -  MAB again, I have not tried that with 2016 or 2017. I know for me that in 2015 it general faults whether VOB is on or off.

               

              I think the point here is not the general fault - although I agree with you that they could do better at notifying; it is that the two master-model methods vary from each other in this regard and a point I was making it SWW was that Save Bodies is more robust  (perhaps for the wrong reasons) than the Insert-Part method.

               

              Mark

            • Re: Master Model Best Practices
              Roland Schwarz

              Still disagree. Inserting a part does not place any dependent features in the master part the way Save Bodies does. And then there's still that "other entities" thing...

               

              The issue with the model fault may mean Save Bodies is a micro-tad more robust, but it seems like an outlier and doesn't outweigh the advantages of Insert Part.

                • Re: Master Model Best Practices
                  Roland Schwarz

                  I think a major part of our non-agreement stems from a fundamental difference in master modelling approaches. I keep my master model detail to a bare minimum. For me, master models are for planning overarching features like A-surfaces, master sketches and key datums. I almost certainly would not have included a funky little fillet in a master model. That sort of detail belongs at the part level.

                   

                  With a more minimalist approach to master models, I have never seen the issues with Insert Part that you describe.

                   

                  I cringe when I see "Save Bodies" used in a master model because it always means that there is too much detail in the master model, resulting in far more difficult part editing as the project matures. It's a time-suck. Not questioning the validity of the method. It's a perfectly valid feature, but usually indicative of impending trouble later in the project.

                    • Re: Master Model Best Practices
                      Mark Biasotti

                      Hi Roland,

                       

                      ..."Save Bodies" used in a master model because it always means that there is too much detail in the master model

                       

                      Of course, this is not always true, it is, more to your point, how to properly use Master Model Technique. I certainly agree with what I call the Master Model Golden Rule. I think where the decision to use the Insert-Part vs. Save-Bodies approach comes down to is how many designers will be involved in the overall product design and how to product's parts will be structured and divided up.

                      As far as the funky little fillet, this was only done as an example feature to get the part to a general fault state.

                       

                      If you are interested in more, you can download my presentation from SWW here

                       

                      Mark

                        • Re: Master Model Best Practices
                          Paul Salvador

                          BTW,.. I wonder how much of a difference between the two would be "if" insert-part (w/all geometry refs, ala,.. copy geom (pro/e)) was more automated or more like save-bodies? (w/options to save name for prt/asm/.. templates, properties,.. solids/surfaces/planes/sketches/...)

                           

                           

                          (re: the fillet/fault.. sorry Mark it was not my intent to make that a distraction/issue.. more of a question,.. of why it existed?.. appreciate/understand it was a example for what can go wrong)

                          • Re: Master Model Best Practices
                            Roland Schwarz

                            Thanks much for posting that, Mark. Good stuff. I agree with nearly all of it in principle.

                            "In theory, theory and practice are the same. In practice, they are not." --unknown, attributed to Yogi Berra

                            My SW weltanschauung is colored by decades of rescue missions. Over-detailing of master models is a very common theme of foundering master models. Save Bodies is a central character.

                             

                            Can't blame the feature for the operator. No doubt in your capable, nigh-magical hands this works great. If more people followed your level of discipline on projects this conversation probably would never have happened.

                              • Re: Master Model Best Practices
                                Mark Biasotti

                                Hi Roland,

                                 

                                I knew going into this that presenting this would be controversial - but that's a good thing right? We all work in such vastly different industries with a common tool. Like many things in SW, it is trial and error and in the end what works best for your situation. 

                                 

                                Now you need to present next year at SWW2018!

                                 

                                 

                                Mark

                        • Re: Master Model Best Practices
                          Chris Saller

                          I have had issues with insert part into part lately where the custom properties data transfers to the part it's inserted into.

                          I submitted the problems a while back.

                          • Re: Master Model Best Practices
                            Rob Edwards

                            I had problems with Insert Part, with custom properties not being inserted as Chris says.  I had even more serious problems when I tried to use a design table to control configurations of the inserted part.  I sent my file to the VAR, they told me Solidworks said that this was broken in 2015 and was fixed in 16.  I am unable to confirm this, it's a real shame because the part in part arrangement is just what I need for my work