9 Replies Latest reply on Jul 27, 2018 1:51 PM by Rich Masel

    Weird result in simulation express

    Rich Masel

      Hi:

       

      I am using simulation express in solidworks professional 2018 SP3.  I am doing a simulation of deflections in a part that I need to brace.  The part is symmetric but the simulation gives asymmetric results.  (i.e. large displacements along one axis, small displacements along another equivalent axis).   I talked to a support person in teh company that sold me the software and he ran the same simulation with the same parameters in simulation and got a reasonable result.    How do I get simulationexpress to get the same result as simulation.

       

      Here is a screenshot of the  setup - this is a fixed plate with some holes.  All four corners are the same

       

      I fixed a point at the center of the part and applied pressure at bolt holes.

       

      Capture.PNG

      Here is the result

       

      Capture2.PNG

      Notice that the projected displacement is different in the corners even though the corners are equivalent in the model.

       

      Mesh parameters 6 mm and 0.3 mm.

      Part size 457x508 mm

      Solidworks 2018 SP3

       

      Message was edited by: Rich Masel

        • Re: Weird result in simulation express
          J. Mather

          I don't have 2018, so I won't be able to answer your question, but for anyone to answer your question I think you will have to Attach your *.sldprt file here.

          • Re: Weird result in simulation express
            James Riddell

            Your deflection appears to be non-linear (more than 10%).  What results were you expecting?  Does your stress look reasonable?  Need to see what you put for constraints and loads.  Some of your 'bolt' pressures aren't showing the same as the others.

            • Re: Weird result in simulation express
              Ryan Navarro

              It's hinging like a seesaw over that extremely small area that you have fixed in the center which is essentially acting like a pivot. I bet the stresses reported there are extreme

               

              There is always some noise in the solution so internally if the applied pressure loads fluctuate (by even a fraction of a percent)  your current model has inadequate stiffness to stop the plate from rotating.

               

              Short version is, you need to fix a bigger area in the center if you want to prevent the rotation. I would use a split line command to break out an area maybe the size of one of those square cutouts to start, placed dead center, and fix that. On the other end of the spectrum if you make the fixed area too large, it would overly stiffen the plate.

                • Re: Weird result in simulation express
                  Rich Masel

                  I tried making the section bigger.   It made things more symmetric but still not right, and the absolute displacement changed by two orders of magnitude.   This does not give me confidence in the result.

                   

                  Capture.PNG

                    • Re: Weird result in simulation express
                      Josh Brady

                      Once your displacements start to get really large, you're going to get into the issue of the plate starting to fold.  Even if your force is evenly distributed, at some point you're going to start trying to "fold" your plate.  It can only "fold" in one axis at a time.  What are you expecting the result to look like?

                      • Re: Weird result in simulation express
                        Ryan Navarro

                        Don't worry about the old results being orders of magnitude higher. The old results are junk, because of the seesaw tilting that was going on. Your new results are much closer to reality (assuming your boundary conditions are correct - I know nothing about your application)

                         

                        I had time to look at your files, here is what was happening in the old results

                        seesaw.PNG

                         

                        Your new problem of the slight asymmetry in results is because your geometry is in fact not symmetric. I'm sure this also contributed (or caused) the previous problem as well.

                         

                        It is not symmetric top to bottom because of the square hole cutouts being offset but that appears to be intended.


                        It is not symmetric corner to corner because at least one of the bosses is much smaller than the others. Because the load is applied as a pressure, this means less force is being applied to that corner. In fact, that is the corner that is lifting up.

                         

                        I'm not sure why that sketch for the boss-extrudes is the only one that is underdefined, but this illustrates the danger of under-defined sketches. Note the smaller boss below.

                        boss.PNG

                         

                        It would be interesting  to go back and see if the old fixture approach works once this geometry problem is corrected.My natural tendency though would still be to fix a larger area as you've done in the revised version. Whatever is closest to reality.

                         

                        Also, are those bosses real or are you just using them to apply the load? If that's the case, use Split Lines instead. It will make your life easier when it comes to meshing.

                          • Re: Weird result in simulation express
                            J. Mather

                            Ryan Navarro wrote:

                            ... your geometry is in fact not symmetric.

                            one of the bosses is much smaller than the others.

                            ... this illustrates the danger of under-defined sketches.

                            Also, are those bosses real or are you just using them to apply the load? If that's the case, use Split Lines instead.

                            I don't have SolidWorks 2018, but I made the changes that you found/suggest and these are the results that I get in another analysis software, before and after the changes.

                             

                            FEA Analysis.png

                            • Re: Weird result in simulation express
                              Rich Masel

                              Thanks.   The change to the boss fixed the problem

                               

                              I am using the bosses to simulate the forces I put on the bolts.   I do not know how to use split lines for a round force.