8 Replies Latest reply on Sep 13, 2012 6:19 PM by Eli Patten

    Inadequate restraints in an assembly with global bonded contact?

    Eli Patten

      Hi,

       

      I'm trying to run a stress simulation on an assembly I put together, but I get an error when I try to run it:

      "The Iterative Solver stopped

      Status code: 9 MATRIX_SINGULAR_OR_INDEFINITE

      May not have adequate restraints.

      No results saved."

       

      For a basic analysis, I have set global bonded contact, fixed some faces on one end of the assembley, and am applying a load on the other end (see photo). I have all the components fully mated and defined, and all the contact surface between the parts register as an interference (coindidence). I'm not sure why it would not be restrained properly - maybe I don't understand how the global contact setting works.

       

      If anyone has come across something similar or knows what the issue is here, I would greatly appreciate if you could please advise me in how to correct it.

       

      thank you very much,

       

      Eli

        • Re: Inadequate restraints in an assembly with global bonded contact?
          Stuart Moore

          Eli, try putting a sliding fixture on the long edges (all 4) of the plates the force is applied to.

          • Re: Inadequate restraints in an assembly with global bonded contact?
            Anthony Botting

            Hi Eli: yes I encounter this frequently. I have found that the message indicates some component is "loose", i.e., bonded contacts may not be bonding where you believe they should be. There could be a tenuous connection at cylindrical surfaces (such as pin to bearing connections, which are "line" contacts and may not actually bond due to differing mesh density across components). As a result of this, I resort to installing my own contact sets.

            A rather quick way to diagnose which component is loose is to sequentially right click on components in the analysis tree, and "exclude from analysis", and attempt a run. When it does run, you've identified the component(s) which is/are loose.

            Stuart's suggestion is a good one, too. You can try to clamp-down on components until you've identified the one that is loose.

            Alternatively, since there is symmetry in geometry, loads and materials, you can cut the assembly in half down the vertical, longitudinal plane. Install roller/slider fixtures on the cut (exposed) faces and that will help tremendously to restrain the model, and reduce solution time.

            Hope that helps - Tony

              • Re: Inadequate restraints in an assembly with global bonded contact?
                Eli Patten

                Thank you for the tips everyone.

                I've tried your suggestions and sometimes it works but usually I just get a different error.

                 

                If I quarter the model, set symmetry conditions at the cut interfaces, and then individually set all contact sets as 'no penetration', I get the following error when it tries to calculate the simulation:

                "The Interative Solver stopped

                Status code:8 FILE_ERROR

                No results saved."

                 

                Sometimes when I run it, I get a windows error:

                "STAR has stopped working

                [check online for a solution... or close the program]"

                 

                Any thoughts on what these errors might be from and how to fix them?

                 

                thanks again,

                 

                Eli

                  • Re: Inadequate restraints in an assembly with global bonded contact?
                    Anthony Botting

                    Hi Eli: those messages appear to be (as was explained to me via tech support) a 'corrupted database'. "How does that happen?...", I inquired. They indicated they really do not know (perhaps the code developers do,  however).

                     

                    Anyway:

                    Try this to (hopefully) clear it up:

                    Shut down and reboot computer.

                    Start up solidworks and turn-on the simulation add-in.

                    Open the assembly, then copy and paste the study you wan to run, to a new one.

                    Delete the old study.

                    Close the model and close solidworks and reopen all that, yet again.

                    Try to run the model at this point.

                    • Re: Inadequate restraints in an assembly with global bonded contact?
                      Stuart Moore

                      Eli, sometimes these errors are associated with the location of your simulation files.  I have had these problems in the past, especially the "STAR has stopped working".  The cause was (I think) my AV software not allowing SW to write to the designated location.  Try setting your simulation file location on your local hard disk - I simply created a 'SolidWorksSimulationResults" folder on my 'C' drive and set Simulation to use that location and all problems went away.  Also, sometimes when the Iterative solver stops, you can restart it and allow the simulation to continue.

                  • Re: Inadequate restraints in an assembly with global bonded contact?
                    Roland Schwarz

                    I've had wretched luck with global bonded contact.  I find things turn out better if I use global=free and set bonds and interferences individually.  I think analyses also go faster because then the model is not "hunting" for changes in contact or bonding beyond what is specified.

                    • Re: Inadequate restraints in an assembly with global bonded contact?
                      Eli Patten

                      Hi everyone,

                       

                      I've found some work arounds that will work, so I wanted to thank everyone who chipped in and post what I found worked.

                       

                      To properly restrain it, a sliding fixture at the side face of the beams worked well.

                      To fix the STAR crash, I tried the suggested tips, and many permeatations, but couldn't get a clean solution. I had two configurations that I wanted to compare. A global 'no penetration' connection worked fine in one configuration, but the STAR error popped up if I tried to do the exact same thing in the other configuration. To get this configuration to run, I had to go through and individually apply the no penetration contact set, and then it ran.

                       

                      thanks again,

                       

                      Eli