4 Replies Latest reply on Mar 12, 2015 11:42 PM by Jared Conway

    few question about goal convergence

    C. Bryant
    C. Bryant Mar 11, 2015 12:31 AM

      Hello,

       

      I'm doing a steady-state simulation,

      擷取1231323.PNG

      in"solver" window, the progress column shows the goal was achieved

       

      however, it seems that calculation still continues and goal temperature seems stable

       

      should I just stop calculation and save current result ?

       

       

      Another question is about transient flow simulation

       

      I set physical time to 270 s

       

      few of volume goals achieved early (about 30 s physical time )

       

      does it mean the steady state have already been achieved?

      • 1903 Views
      • Categories: Flow Simulation
      • Tags:
        • Re: few question about goal convergence
          Amit Katz
          Amit Katz Mar 11, 2015 8:30 AM (in response to C. Bryant)

          If you set the time to 270s, your simulation will continue until that time. "Steady state" can be a subjective term. If you see that your critical parameters have settled into a more or less constant value, then you could consider that steady state.

          • Re: few question about goal convergence
            Jared Conway
            Jared Conway Mar 12, 2015 1:07 AM (in response to C. Bryant)

            you'll notice in your screenshot that it is still checking the criteria. if you let it finish that, the calculation will finish if all of your calculation completion parameters are complete. convergence is not the only thing that can stop the calc, check the calc control options to see what you have set.

             

            this is similar for your transient question. the simulation will stop based on the criteria you choose. if you have convergence, it may stop before 270s. if you have time, then it would stop when your time is set.

             

            however like amit said, the solver doesn't stop at steady state. it doesn't understand that concept. you need to monitor the parameters to determine if that happened. and this requires you to run the analysis basically indefinitely. in 2015 there are some improvements in this to give you more options.

              • Re: few question about goal convergence
                C. Bryant
                C. Bryant Mar 12, 2015 8:25 AM (in response to Jared Conway)

                thanks for reply

                 

                I let the calculation continue for few days (steady-state simulation ), the goal temperature values seem no difference to the last time I check

                 

                I did not change setting of calc control option, so the flow internal convergence criteria was used.(default)

                 

                the default criteria are too strict in my case I guess

                 

                 

                for transient analysis part, now I understand calculation would continue until specific physical time

                 

                I still  let calculation continues

                 

                the goal values change slightly though the progress column shows "achieved "

                 

                so I think I should let it continue until 270 s.

                  • Re: few question about goal convergence
                    Jared Conway
                    Jared Conway Mar 12, 2015 11:42 PM (in response to C. Bryant)

                    "the default criteria are too strict" i'm not sure what you mean here. if it converged it will stop. that is what you want. but if you used default, it may stop at 4 travels which may be before convergence occurs.

                     

                    for transient, convergence should not be used as a calculation stopping condition. either run till it reaches steady state or your value.