9 Replies Latest reply on Jun 5, 2017 7:03 PM by Jordan Kamga

    Nonlinear simulation

    Jordan Kamga

      Error message:   " PCGLSS0139:Matrix is Singular or Indefinite***************** "

       

      I am trying to do a nonlinear simulation in Solidworks using silicon rubber and I get this error message. Can anyone help me out here?  I believe there isn't any errors in my part but the simulation always stops at about 3% with the following error message :

      "Solution failure in a Step>1, it could be due to:

      The solution may be at a buckling or limit point, i.e.,

      displacements grow large under constant forces. If so,

      for force control or contact problems, this is the end

      of solution (check out the response graphs).

      The solver numerical difficulties:

      1. Reduce the Singularity-elimination-factor (.5 or 0.)"  

       

      What does it mean and how can I solve it. please and thanks. Any help will be greatly appreciated !

      Jordan

        • Re: Nonlinear simulation
          Ryan Dark

          Hi Jordan,

          It means that the solver was not able to find a stable location for this iteration of the nonlinear solve.  Sometimes setting the Singularity Elimination factor to zero or specifying a smaller maximum auto-time step can help it solve further.  Are you able to attach the file set to this thread for us to look through?

            • Re: Nonlinear simulation
              Jordan Kamga

              Hi Ryan,

              Thank you for your reply! Unfortunately I won't be able to attach the file here. Forgive my ignorance but I don't understand lot of what you said because i'm not that good at this yet. I will be trying to figure out how to do what you just described to me. Just to make sure I got you right, do I need to changed anything on my design to fix this issue? Also I should mention when I run the linear static simulation I have no problem. Kindly

              Jordan

                • Re: Nonlinear simulation
                  Ryan Dark

                  Hi Jordan,

                  Without being able to have your model in my hands I could not tell you whether it requires model changes to help the nonlinear solver work.  The nonlinear solver is a little more complex than a text/picture explanation allows for.

                   

                  Probably the easiest approach is to put all your boundaries for this study into a Static type study instead and see if the setup you have created is stable.  This may mean reducing loading or prescribed displacements to fractions of their normal amounts but could help you find instabilities in your setup faster than you would in a nonlinear analysis.

                  • Re: Nonlinear simulation
                    Bill McEachern

                    Are you using a remote load by chance?

                    Essentially something has zero stiffness. Think along those lines. Is anything free to move? Are you using a hyper-elastic model? Which one? Have you proved it works on some test case with your inputs? Do you have contacts constraints in the model? I doubt it is buckling if it failed in the first step.

                      • Re: Nonlinear simulation
                        Jordan Kamga

                        Hi,

                        No I am not using a remote load. I just applied an external force like I always do without any further complication. I have just one part not an assembly. I did fix a portion of it(you can't simulate if you don't fix something anyways). My material is silicon rubber Durometer 60. I have no problem with this model when I run linear elastic model instead, in fact it completes a lot faster but that is not what I want.

                          • Re: Nonlinear simulation
                            Bill McEachern

                            What hyper-elastic model r u using - Mooney -Rivlin? How many coefficients and how did you get them? Did you put in curves? How many? Where they modified to be in stress ratio form? Maybe post the data for the model. It sounds like a material model problem. Do you have the large strain option turned on? If so turn it off. It is for plasticity problems. The problem description sounds pretty simple so I would not pursue Siavash's comments till you confirm your material model is realistic. Well, that's my 2 cents on this one. Sounds like a model problem as the problem runs no issue with a linear elastic model of very low stiffness (I assume) was used.

                    • Re: Nonlinear simulation
                      Siavash Khajehhasani

                      There would be more that one step for troubleshooting for that error. Since you cannot provide the model, we can only guess and give you some hints, then you can do your homework to see which would resolve the issue. I am going to mention most of them even the least effective ones.

                      1)  change time steps/increment

                      2) change parameters according to advises by SW (read help) in Step/tolerance options section; most importantly, reduce the singularity factor.

                      3) switch to MNR

                      4) please make sure using curvature based mesh - increase the mesh density to a fair resolution

                      5) make sure contacts are defined properly (most likely they are, as you ran static analysis with no issue)

                      the last 3 are the least effective ones.

                      very much hoping it helps.

                      • Re: Nonlinear simulation
                        Jordan Kamga

                        Thank you everyone who gave me an idea o how to solve this. I completely redrew my design from scratch and it works now. Thank you very much.