12 Replies Latest reply on Nov 26, 2014 4:23 PM by Pierre Saby

    virtual wall

    Pierre Saby

      Hi,

      I want to ask the simulation.

      I want to simulate such an example. What type of simulation is the most resemblance to reality?

      1. do not use the part_2 and replace it with a virtual wall and fasten with screws part_1

      2. Do not use screws and fixings part_1 fixed in the screw holes

      3. Use the original part_2 and fasten part_1 on part_2 screws

       

      thank you very much for advice

      0_.jpg

       

      1.1_.jpg

      2.2_.jpg

      3.3_.jpg

        • Re: virtual wall
          Shaun Densberger

          Between the three options you've given, #3 is technically the most realistic, but it should be very close to #2.

          • Re: virtual wall
            Jared Conway

            it looks like you made one part fixed/rigid, to be "realistic" you'd want that one to be deformable

             

            in the end, all of them should yield similar results, comes down to what assumptions/approximations you are ok with and what results you want

            if you want bolt forces, no choice but to use the bolts for example

            if you want to see interaction with the components, no choice but to use contact between them

            • Re: virtual wall
              Pierre Saby

              1, So if you want to know the tension in the beam, so I use option 3?

              2, If you do not know the reason znejakého part_2, so I use a virtual wall?

               

              3, Why resulting tensions in the 2nd and 3rd simulation is not comparable to the same? When the 2nd and 3rd simulations all the same, the difference is only in part_2 a virtual wall.

               

              4, When to use a virtual wall?

               

              5, 1rd is completely wrong for use in practice?

                • Re: virtual wall
                  Jared Conway

                  1, So if you want to know the tension in the beam, so I use option 3? > what do you mean by tension in the beam?

                  2, If you do not know the reason znejakého part_2, so I use a virtual wall? > i don't understand this question?

                   

                  3, Why resulting tensions in the 2nd and 3rd simulation is not comparable to the same? When the 2nd and 3rd simulations all the same, the difference is only in part_2 a virtual wall. > what do you mean by tensions? are you talking about stress? probably best to post all your examples

                   

                  4, When to use a virtual wall? > when you don't want to model a wall because you don't care about the effect on the wall part.

                   

                  5, 1rd is completely wrong for use in practice? > i'm not sure what you are asking here

                    • Re: virtual wall
                      Pierre Saby

                      I thought. general stress, for example stress Von Mises.

                      I assumed that when using virtual wall instead of component (part_2), the resulting stress - (Von Mises) comparably the same as if I did the simulation with the original component.

                      Why the difference between the simulation with virtual wall and original components, virtual wall = 130,062MPa and original components = 182,753MPa?

                        • Re: virtual wall
                          Jared Conway

                          I would recommend looking at displacements first

                            • Re: virtual wall
                              Pierre Saby

                              So we recommend to use simulation with virtual wall?

                              Do not you tell me why the difference 130 and 180 MPa?

                               

                              Still, I would like to ask:

                              Diameter clearance hole is 6.4mm. Set nominal shank diameter to 6 mm in the simulations?

                              Or let 6.4mm, as set solidworks automatically?

                               

                              4x.jpg

                                • Re: virtual wall
                                  Mikael Martinsson

                                  Dear Pierre.

                                  All 3 will yield similar results, but 2 and 3 is more realistic since 1 fixes the holes in all degrees of freedom and prevents any rotation or deformation.

                                   

                                  The maximum stress, that you're concerned about, is meaningless in all 3 examples and occurs around the holes.

                                  Fixed restraint (1) creates singularities (infinite high stress) and bolt connectors (2&3) creates unrealistic stress results in the vicinity of the bolt. (see help file that explains this).

                                   

                                  The color scale is set automatic by the maximum stress result, so if you manually set them to the same max/min as in first simulation, you will see that the bending stress in part 1 is similar for all 3 simulations. You can also probe the top surface to see that they will be similar.

                                   

                                  If you're concerned about the stress and deformations in the holes, you need to model a pin or bolt to get more realistic result in these areas.

                                  • Re: virtual wall
                                    Jared Conway

                                    see mikael's response

                                    i would also recommend you look at some articles on convergence and also singularities, that will more appropriately explain why stress is not the first thing you look at

                                    • Re: virtual wall
                                      Pierre Saby

                                      Dear Mikael and Jared,

                                      I do not worry. I just thought that when using a virtual wall as an alternative to the original parts, so that the results are comparable. I can not explain it. What have I to think, that the maximum stress is 130 MPa and 180 MPa? You look so good to me on simulation, whether somewhere I was not mistaken, I give files of the simulation.

                                        • Re: virtual wall
                                          Jared Conway

                                          are the displacement results good?

                                           

                                          remember in fea displacements are calculated > then strains > then stress

                                           

                                          sometimes there are things that affect stress, they are called singularities, these singularities can causes stresses to go up and down and not be realistic directly where they exist. as such the results may not be comparable if they exist. BUT, displacements don't have this issue. so it is best to always check displacements first when you are comparing solutions. worry about stress later.

                                           

                                          but yes, they should be comparable