6 Replies Latest reply on May 8, 2015 1:51 PM by Jared Conway Branched from an earlier discussion.

    Uniform load on extrusions

    Finly Turnier

      This question is branched from my original thread: Static simulation of uniform load on extrusion

       

      I was eventually able to get some results on the stresses and displacement of the extrusions but after letting the solver run for over 20 hours. Currently evaluating whether to get the full simulation package of SolidWorks but there is still some doubts on the capabilities of the software. I am able to get results but after a very long time of simulation and that's if the solver doesn't crash. I am currently running a PC with a I7-4770S CPU @ 3.10 GHz and 12.0 GB memory standard 7200 rpm HD SATA 3. The extrusions that I generally work with are complicated shapes and curves I eventually want work up to larger assemblies of these components. Will SolidWorks be able to run accurate analysis on these type of extrusions?

        • Re: Uniform load on extrusions
          James Riddell

          First, turn off hyperthreading, that will speed up your solve a fair amount.  Second, how many DOF are you going to do?  If you can get a SSHD to run the data/solve on it will help.  What do you consider 'large'?  It isn't so much the number of parts as it is the complexity of the geometry.  If you are using contact surfaces and fasteners then it will take a lot more time.

           

          I've got a M6800, loaded, and even w/ 32GB I have had models with 20M DOF run in fewer than 75 hours.

            • Re: Uniform load on extrusions
              Finly Turnier

              Currently I'm running a simulation that has 1.7M DOF, see attached image, typically the geometries resemble that. Multiple of these type of components to make up wall systems and sliding glass door assemblies. I would like to be able to optimize these type of systems if Solidworks is capable of carrying out FEA on these parts within reasonable amount of time. I am using contact sets no penetration and bonded types, if that's what you mean. Eventually when testing more complete assemblies I will want to implement fasteners.

               

              Another question, I modeled the assembly of the previous post you answered using surface features and simulated using shell thickness of the total solid extrusion thickness and got different results but the simulation completed quicker. Do you think its possible to accurately model these solid extrusions as surfaces and run FEA that produce realistic results?

               

               

              TopView.png

                • Re: Uniform load on extrusions
                  Jared Conway

                  based on your 2 posts, if you have solidworks simulation, pro won't help you

                  have you considered working with someone to optimize the setup of your analysis and get some training?

                  if you're interested, we have something we call mentoring that is basically that

                  you can email me at jared@hawkridgesys.com for more information

                   

                  otherwise I would post your model and get some help here

                   

                  but like others have mentioned, this sounds like some hardware and some setup (mixed meshing) would help a lot.

              • Re: Uniform load on extrusions
                Michael Kalin

                When you say the full package, do you mean you're using Xpress? Or are you thinking of upgrading to professional?

                 

                It's not surprising a simulation with beams run as solids would take that long. Simplifying them to one dimensional beams would greatly alleviate the degrees of freedom and nodal count.

                  • Re: Uniform load on extrusions
                    Finly Turnier

                    Upgrading to simulation professional.

                     

                     

                    Could you check my earlier response to James Riddell regarding modeling as surfaces. Can you expand on your suggestion to modeling as one dimensional beams? How would I do this if I have multiple extrusions connected to each other each having individual inertia? Plus each have different contact interaction with the fixed or anchored component.

                      • Re: Uniform load on extrusions
                        Michael Kalin

                        Surfaces will only work on simple shapes that can be broken down into constant cross sections that don't overlap. Solidworks can however solve beams as a one-dimensional element with the inertial values for the profile saved inside it. This is all covered in the three day training course and I would advise you to call your VAR for more information. Failing that search for tutorials on beam analysis on youtube or on a VAR website.