5 Replies Latest reply on Mar 8, 2016 10:51 AM by Willem Van Opstal

    Extremely simple, but unstable model?

    Willem Van Opstal

      Hi all, I''m having some trouble with a fairly simple model. A sort of bookshelf: a shelf on one rigid body. I actually don't want to use inertia relief because I don't know the consequences...

      The supports are rigid. And I would like a No Penetration contact. When a bonded contact is used, no problems. As soon as I go to No Penetration-contact SW15 gives me the error of an unstable model.


      I attached the assembly with its studies. The problems are occuring with the study: Shelf - No Screws.

      Are there any advices?? Thanks!



      I think the No penetration contact is the problem. Check the 2nd screencapture. Where it's said 'no penetration' it just penetrates the full 100%. I don't get anything of what's happening here..

        • Re: Extremely simple, but unstable model?
          Prasad Bhonsule

          Hi Willem, have you tried adding a local contact set as no penetration rather than setting the Global Component Contact to No Penetration?





          • Re: Extremely simple, but unstable model?
            Shaodun Lin



            It seems nothing wrong with No Penetration contact set.


            • Re: Extremely simple, but unstable model?
              Siavash Khajehhasani

              Hello Willem,

              I would like to add one more point to what the other two gentlemen mentioned. Adding the local contact set of No Penetration for the simulation study of Shelf - No Screws with the support body treated as rigid, does not solve the stability issue of your simulation. In order to fix the stability issue, it is recommended to use softspring option to stabilize your model. To put further explanation on why do you get the instability error and should use the soft spring, please notice this: You may think that the Shelf body is placed at the top of support and the only load is vertical, why this part is going to be unstable? This argue makes sense in reality, not in numerical simulation. The reason for that could be the type of mesh used here; the tetrahedral elements are used in SW FEA. Having said that, a "very" small force might be places in X or Z direction as transferring the load between all mesh elements. This could be resulted in a "very" small displacement of the component in either of those directions and since no constraint stop the body to move, you will receive the unstable error. Those minor forces will be damped in those directions through those soft springs and helps your model to become stabilized.

              Some screenshots are added here for clarification.

              Please let me know if it solves your issue and answered your question.


              - Siavash

              soft spring.jpg