5 Replies Latest reply on Jan 30, 2015 7:51 PM by Jared Conway

    1) SolidWorks simulation and shell Shell elements. 2) Intentional gap between contacts

    Nikhil Phatak

      Dear Friends,

      I am trying to run static simulation on a pipeline with shell elements with some flanges ,clamps and supports that are treated as solids.

       

      1) My problem starts with a simple bonded contact between two connected pipes(both shell)  which are initially touching without a gap which i think should be bonded with a global bonded contact.But even i add a local contact set between them (By selecting the outer edges of two pipes / or tangent outer faces..because shell is defined on outer face) the contact fails.(which i figured out by using soft springs and inertia relief)...

      ......somehow for same set of pre-processing even if i slightly modify these contacts at one pipe joint (edge to edge, edge to face or face to face) contacts at different pipe joint fails and the contact which was failing previously is now seen as bonded. 

       

      2) Another thing is , I define a shell by outer face of a Pipe with certain thickness and by aligning this outer face as a shell's top face which also shows me a correct colored preview (aligned with a thickness). But when i actually mesh the model exactly reverse happens and i manually flip the shell elements.( i am wondering why i need to flip the shell elements after meshing  if i am defining them correctly)

       

      3) I am trying to simulate a gap between clamp's inner face and a pipe's outer face(Both are concentric initially)Please refer image attached. I have added no penetration contact between them and clamp is having some imposed displacements on it which will try to make them off center.My intention is there will be no stresses on pipe till clamp touches the pipe and after that whatever of imposed displacement left will actually produce a stress on the pipe.( I thought of a bearing connector but in reality it is free to move).....so the real question is compare to pipes overall length this gap is small so what i am thinking is static solver can solve it. when i solve it separately with a single pipe it gives me good results .But as soon as i combine it with the assembly equilibrium is not achieved.

      (** gap is not too small such that it will be ignored ..it is greater than 0.5% of global element size that's what i got from the help.but not too large to go in large displacement mode compared to overall pipes length and deflections.

       

      I think i typed too much and sorry for the rough images. Kindly advice and thanks in advance.

       

      Regards,

      Nikhil

        • Re: 1) SolidWorks simulation and shell Shell elements. 2) Intentional gap between contacts
          Jared Conway

          1) is this a different analysis all together? i'm confused?

          2) please post some screenshots, really flipping the shell is only to change the thickness direction and what is output as top vs bottom

          3) why can't they start touching? what is the importance of having the gap to start. are you using a nonlinear or linear static analysis?

            • Re: 1) SolidWorks simulation and shell Shell elements. 2) Intentional gap between contacts
              Nikhil Phatak

              HI Jared Conway,

               

              1) It is same analysis and the same study, Lets say i have Gravity as the only activated force all fixtures and boundary conditions along with shell to shell /shell to solid contacts are added then i mesh it with draft quality mesh.Now solver starts and while computing contacts it shows me "Equilibrium is not achieved" message, So to recheck my contacts I select "edit definition" on each of the contacts and just de-select and re-select the same set of edges nothing else, everything remains the same even contacts are same.and to my surprise solver starts and study is complete with expected results.(This happens randomly or in fact frequently)

               

              2) I will try to get some screen shots and will post it here as internet is not accessible.But my pipe is lying on the 'rigid virtual wall' and shell interferes with that virtual wall after meshing that's why i flip it.(But i define it correctly) so i was thinking of avoiding that flip every time i re-mesh.

               

              3) I am using solid works premium linear static analysis. Gap in the pipe and clamp allows free movement of pipe when clamp is deflected by certain amount (In real situation there is very soft rubber like material inside the gap) but to solve it with the static solver I am neglecting it. (I don't know the exact stiffness of that material that's why i am not defining a bearing connector also this clamp do not share any vertical reaction it will only contribute in lateral reaction forces perpendicular to the pipe's axis)

               

              Regards,

              Nikhil Phatak

                • Re: 1) SolidWorks simulation and shell Shell elements. 2) Intentional gap between contacts
                  Jared Conway

                  1 > sounds like a model issue, might also be an explanation for the need to flip the shells

                  2 > you can turn off shell alignment so it doesn't do this but this is part of using shells

                  3 > sounds like you need to model the component

                    • Re: 1) SolidWorks simulation and shell Shell elements. 2) Intentional gap between contacts
                      Nikhil Phatak

                      Hi Jared Conway,

                       

                      I rechecked the model , maybe it was a model issue.So i have re-drawn it and this time made all pipe joints as a single piece of pipe like a continuous sweep instead of multiple pipes mating together as parts and multi-bodies. Then everything looked OK. But the moment i try to split a pipe using a plane(Split operation),....shell to shell contact is lost even after manually adding a contact set. when i right click the study properties and click ignore clearance for surface contact results are as expected.

                       

                      My question is why split operation should create a clearance between two bodies if it doesn't create a clearance some how simulation is treating it as a clearance.?? Is contact set dependent on mesh type and mesh size.(Since pipe joint is very small compared to pipe length) and i am using a default set draft quality curvature based mesh. I will do some more trials to find out ...thanks in advance.

                       

                      Another thing about virtual wall .After defining a Rigid virtual wall my pipe is still sinking in it by 2mm. i expected zero displacement in the direction normal to the wall towards the wall.( i will re-run it once to check if i am defining it correctly)

                       

                      and sorry i will try to upload at least screen shots or if possible my model if permitted.

                       

                       

                      Regards,

                      Nikhil Phatak