12 Replies Latest reply on Mar 24, 2017 12:39 PM by Chris Clouser

    Zero Thickness Geometry: another excellent example...

    Chris Clouser

      In the past, I and others have produced many excellent examples of why a modern, high-end CAD system should be able to handle ZTG. 

       

      All other reasons aside, the simple fact is that many come from importing files from other CAD system that can do this.  Not being able to cleanly import or manipulate this CAD data leaves the tree with a mess of errors.

       

      Some people say that ZTG is a physical impossibility (bunk), but here is a good example I ran across when importing a file.

       

      When creating a sheet metal part with a punch/tear like this, there is a ZTG condition in real life.  We can argue all day about the molecules if they are touching or not, but these are just silly arguments.  The fact remains, the attached part is just ONE MORE example of legitimate, real-life parts that need ZTG capability.

        • Re: Zero Thickness Geometry: another excellent example...
          Umberto Zanola

          SW is not an HIGH END system to begin with...your expectations are a bit too much IMHO

          • Re: Zero Thickness Geometry: another excellent example...
            John Lhuillier

            Well NX must not be "high end" enough either. It doesn't do ZTG either. If I put a hole thru a part & then create a rectangle that has one edge tangent it errors out as well.

            One question on your sheet metal part though. In real world metal working would there not be some gap between the tab that's formed from being torn even if it is only a tenth (.0001) or so. I'm no sheet metal expert so just asking. What kind of issue would be created on models if you either added .00001 or removed .00001 to be able to generate the geometry in MOST instances that we work on. The machines that these parts get made on can't hold that tight of a tolerance anyhow. Yeah the model is not going to be exact but when working in most real world conditions in manufacturing isn't going to hurt a thing.

              • Re: Zero Thickness Geometry: another excellent example...
                Chris Clouser

                John Lhuillier wrote:

                 

                Well NX must not be "high end" enough either. It doesn't do ZTG either. If I put a hole thru a part & then create a rectangle that has one edge tangent it errors out as well.

                One question on your sheet metal part though. In real world metal working would there not be some gap between the tab that's formed from being torn even if it is only a tenth (.0001) or so. I'm no sheet metal expert so just asking. What kind of issue would be created on models if you either added .00001 or removed .00001 to be able to generate the geometry in MOST instances that we work on. The machines that these parts get made on can't hold that tight of a tolerance anyhow. Yeah the model is not going to be exact but when working in most real world conditions in manufacturing isn't going to hurt a thing.

                Is it reasonable to talk about the "real world"?  Some people assume that the software is here to assist us from making mistakes.  It's not.

                 

                I can draw infinite examples of something on CAD that can never happen in the "real world".  How about 8 decimal places on a machined part just for starters.

                 

                When I interview new engineers and designers, I have a document full of crappy CAD models (mostly from GrabCad) and I ask the interviewee what's wrong with each one.  So, no, let's not rely on software to dictate our designs, rather human intelligence.

                 

                On a sheetmetal part, like this example, in the real world, the deformation would vary around the tear.  Around the bend between the tear, the outside of the radius the material would shrink transversely causing a small gap, but on the inside radius of the bend, the opposite would happen.  The material would expand causing some minuscule interference.

                 

                So what we do in SolidWorks is put a tiny gap there so it doesn't fail.  Then, that gap causes problems due to the fact that the laser will want to cut it.  So some operator down the line who is programming the laser will have to know to eliminate that feature, or engineering will have to remember to remove it before sending out the cut pattern.

                 

                Basically, extra housekeeping because of the software dictating the design due to it's limitations.

              • Re: Zero Thickness Geometry: another excellent example...
                Steve Calvert

                Chris, can you post a pic or attach a parasolid?  I'm on SW2015 and wanted to see this.

                 

                I understand what ZTG means but I don't know of a real life situation where it is used.  Doesn't there have to be something there which means it's not Zero?

                 

                Steve C

                • Re: Zero Thickness Geometry: another excellent example...
                  Paul Risley

                  Chris Clouser

                   

                  I will agree w/ your opinion on ZTG on import. I have ran into this scenario on imported models.Capture.JPG

                  The red parts are "welds" on a customers part. We work on big structure weldments for an outdoor equipment manufacturer. Solidworks creates issues with every weld bead. Which leads to ID errors of imported bodies. Mate to a surface close the model and re-open it and the mates are blown. Our Var traced it to the welds and the fact that they are ZTG and may have "surface" issues because Solidworks cannot interpret them properly. (Which is retarded since we don't mate to weld beads.)

                   

                  So we have to create a configuration for our weld fixtures with no-weld beads, then one with them for the customer to see the welds in the fixture. Released production no problem, but we start around rev 4 models which go as high as rev 20,after that their models get released with alpha numerical revisions. On average our fixtures have 10-15 imported versions of the customers files because of the revisions applied on their end.Which means we have to maintain 2 configurations of these weldments everytime we have to update their models in our fixtures.

                   

                  But hey I like job security so I don't bitch.(TOO MUCH ANYWAYS.)

                  • Re: Zero Thickness Geometry: another excellent example...
                    Craig Schultz

                    Actually, this part has problems other than ZTG if you want to convert it to sheet metal.

                     

                    1)  The tip end, isn't a uniform thickness.  It's faces are at an 8 degree angle to each other.

                     

                    2) The chamfers around the edges.

                     

                    3) The top part of this transition if flat.  The bottom face is a radius.

                    Capture.PNG

                    • Re: Zero Thickness Geometry: another excellent example...
                      Paul Salvador

                      I hear you Chris.. and for me,.. what is extremely frurstrating with this type or problem.. there are NO good or easy ways to FIX them.. here I'm showing how to fix it using a copy/delete/trim surfaces and another example/way.. but it really should NOT be this hard!...

                      personally,.. Trim Surface has LIMITATIONS..  (and they REALLY mucked it up in 2016).. with recognizing and trimming boundaries... imho.. his could be EASILY closed as is.. ... BUT NO.. we have to go thru a rabbit hole to resolve this simple stuff..

                      (btw,.. one way to get rid if the "red" error.. is export as parasolid and open.. note: the errors are still there though)

                      ztg-z.png

                        • Re: Zero Thickness Geometry: another excellent example...
                          Chris Clouser

                          Exactly Paul.

                           

                          I should be more clear for those still trying to understand all this.

                           

                          There are issues with MODELING ZTG in SolidWorks.  that's one discussion in and of itself.

                           

                          Then there are issues with IMPORTING ZTG in SolidWorks.  And that's a completely different discussion.

                           

                          Both discussions overlap, but if you are one of the people who for some reason don't want to be able to model ZTG, the question still remains, should we be allowed to import it?

                           

                          Most of the ZTG comes from other cad systems that allow sheetmetal like the above and bending edge flanges that actually touch along the edge.

                           

                          BUT ALSO, multibody parts and assemblies SAVED AS ONE BODY in another CAD system will result often in ZTG, if you think about it, there are a bunch of parts touching in a variety of ways, you are bound to create ZTG (in places where SWX gives you the infamous "invalid body" error.)

                           

                          As you mention above, Paul, I would like to be able to quickly cut small slots on this part and then it is happy again, but SWX has NO idea what to do with this unless you do some unholy surfacing operations as you have shown.  And when you import an assembly that may have hundreds or thousands of errors, then what?  If the ZTG error disappeared, half or more of my import headaches would go away.  Many others, the software can fix on it's own.

                           

                          I have an ER that says "OPTION TO IGNORE IMPORTED PART ERRORS IN THE TREE."  This may be a temporary bandaid.  I can't work with errors in the tree.