3 Replies Latest reply on Jul 16, 2012 7:10 PM by Kevin Corr

    simple static analysis comparing local and global contacts

    lorenzo rivelli

      I have tried to do a simple static analysis taking into account only two parts.

      simple test.PNG

      There is only a constraint at the base of the bigger cilinder and only a force of 200N applied on the little cilinder.

      As you can check in the attached files, firstly I have used global bounded contact and secondly global contact with no penetration. In these cases I obtain the results, but my first question is: What is the more realistic analysis? The first one with bounded contacts or the second one?

       

      After I have tried local contact with no penetration but in this case the solver failed and I really don't understand the reason. Why local contacts didn't work?

       

      Also for the last test in which I have used global contact with no penetration and a force of 1000N the running stopped and the following error appeared: Automatic global contact search. What does this error mean? And how can I work out the problem?

        • Re: simple static analysis comparing local and global contacts
          Kevin Corr

          this pin fits in a hole sized for sliding fit or for interference fit? If sliding, is it welded in place? are you trying to size the pin diameter with the study results? Are you trying to size the wall thickness of the tubing? these things are important in determining the relationship between nodes in the pin and tube. If you are trying to size the tube thickness your solid mesh element size needs to be such that the thickness is at least two elements thick.

            • Re: simple static analysis comparing local and global contacts
              lorenzo rivelli

              the pin fits in a hole sized for sliding fit and it is not welded. Then it is just a pin inserted, without interference, in the holes of the tube. On the pin a vertical force has been applied. I would like to see how the force is transferred to the tube and depending on these results increase or decrease the thickness of the tube. At the same time I would like to verify pin's displacements in order to decide if adopt a different material for the pin or increase his diameter.

                • Re: simple static analysis comparing local and global contacts
                  Kevin Corr

                  For stress accuracy in the contact region, I strongly suggest controlling the number of elements in the radial direction of the tube. One possibility is to extrude the tube in two concentric bodies that share a cylindrical face. That way the mesher automatically recognizes each body's need for at least one element in the thickness direction.

                   

                  For the local vs global problem it seems like you are using a very diverse face pair that may be causing local contact to fail. Try to regionalize the pin cyl. surface into sub surfaces, two(four, if you follow the above) of which are to contact the pin. Use these two(four) in the local contact.  One possibility for regionalizing is a split line feature on the pin that uses a sketch converted from the tube ID and OD.

                   

                  Related tip: in the contact property, the target should be the stiffer face of the pair.

                  Other related tip: You don't have to create contact sets for each Source and Target.  But all Source entities should belong to one component and all target entities should belong to the other component.