1 Reply Latest reply on Oct 16, 2018 11:53 AM by Matt Cullin

    Submodel clarification

    Matt Cullin

      I have a few questions that I am hoping someone can answer concerning some "odd" behavior I am noticing in SOLIDWORKS 2018 when running Static submodels with No Penetration contact sets:

      (1) When in a submodel, you can turn on h-adaptive meshing, but then when you run it doesn't work. If you return to the adaptive tab, it is set back to "none". Is adaptive mesh refinement just not supported in submodels?

      (2) Since adaptive mesh refinement is not available, I am using a Design Study to manually refine my mesh in areas with large stress gradients. I notice that scenario 1 takes a very long time to solve, but then subsequent scenarios run very quickly. Is there a way to show the normal solver status while in a Design Study, or, do you just have to look at the Design Study progress dialog? I believe Simulation is using an iterative solver to deal with the No Penetration contact. The best explanation I could come up with is that it seeds the subsequent scenarios with the results from the previous scenario. Is this what is happening. It is just odd because when I run scenario 1 outside the Design study, it takes a lot less time than it appears to take when the Design Study runs a scenario with the same parameter values.

      (3) In the submodel, I am unable to delete or modify contact sets. Is this normal? I thought you could change almost anything about the submodel except the boundaries where displacements are specified.  The only explanation I could come up with is that since the full model displacements were calculated with those contact set parameters, the submodel needs to preserve them. Am I missing something?

      (4) Is it inefficient/inappropriate to run No Penetration contact models as a Static study? It seems like the solver procedure is a less robust version of a non-linear study. Seems like the non-linear solver takes longer, but works in smaller time steps. This leads me to believe that perhaps the results are a bit more reliable.

       

      Thank you all in advance for any insight you may offer.