7 Replies Latest reply on Jul 16, 2014 8:03 PM by Jared Conway

    NonLinear Dynamic simulation

    Roshan Santhosh
    Roshan Santhosh Jul 14, 2014 5:21 AM

      I am conducting a non-linear dynamic simulation to understand the deformation of a fixed a object when a rigid body moves over it. My simulation consists of a fixed semi-sphere and a rectangular object which is supposed to move over this sphere.

       

      I placed the rectangular block near the sphere and gave two extrenal inputs.

      1. A downward force to make sure that block continues exerting pressure on the sphere and that it doesnt fly off.

      2. A advanced fixtures of a displacement of 100mm forward so that the block neccessarily moves a 100mm in the forward direction.

       

      But the simulation ends abruptly irrespective of whatever adjustments are made.

       

      What is the correct procedure for going about this simulation

        • Re: NonLinear Dynamic simulation
          Peter MacDonald
          Peter MacDonald Jul 14, 2014 9:55 AM (in response to Roshan Santhosh)

          What error message do you get?

          Could you use a downwards displacement rather than a force? If that works then you know where the problem is.

          • Re: NonLinear Dynamic simulation
            Nicholas Luyster
            Nicholas Luyster Jul 14, 2014 11:33 AM (in response to Roshan Santhosh)

            Hi Roshan,

             

            I've done similar simulations.  These sorts of simulations aren't easy to run but I can give you a few pointers.  First, start out making sure that the two bodies are initially kissing (meaning they are at the initial point of contact).  Then, instead of applying a displacement, apply an initial velocity to the block.  Then, instead of worrying about the block bouncing back, only run the simulation up until the point in time where the block reaches a velocity of zero. 

             

            If you need to add more time to the simulation, create a time ramping force which will start after the block reaches a 0 velocity. Timing is critical so you will probably need to run several simulations leading up to the final simulation.  Expect that this will take a little bit of work to lead up to the final simulation so to save time, make sure that the mesh is course when you're doing your initial simulations.

             

            There are a few things to note in addition to what I discussed above.

             

            Make sure to use appropriate time steps

             

            Make sure you have appropriate contact

             

            All the best,

             

            Nick Luyster

            Gosimulation.com

            Online SolidWorks Simulation Training

            • Re: NonLinear Dynamic simulation
              Jared Conway
              Jared Conway Jul 14, 2014 10:28 PM (in response to Roshan Santhosh)

              what holds the sphere?

               

              what holds the block in the other directions?

               

              what type of contact do you have defined?

               

              have you already broken this simulation down to its simplest parts to make sure they work on their own? for example a force on the sphere and moving the block on its own?

               

              have you considered not using a pressure on the block but just setting that direction to 0.

               

              also i personalyl have not had a lot of success with any contacts and rigid components. i've had best success with real material properties.

              • Re: NonLinear Dynamic simulation
                Nicholas Luyster
                Nicholas Luyster Jul 15, 2014 9:39 PM (in response to Roshan Santhosh)

                Hi Rosham,

                 

                I created a video which should get you pointed in the right direction.

                 

                http://gosimulation.com/non-linear-dynamic-simulations-with-solidworks-simulation/

                 

                All the best,

                 

                Nick Luyster

                Gosimulation.com

                Online SolidWorks Simulation Training

                  • Re: NonLinear Dynamic simulation
                    Roshan Santhosh
                    Roshan Santhosh Jul 16, 2014 7:27 AM (in response to Nicholas Luyster)

                    Thank you for the video. What changes would be required if instead of falling straight on top of the semi-sphere , the block moves towards the semi-sphere from the side. As if the block is running over it?

                     

                    In the problem that Im trying to solve the flat surface of the semi-sphere is horizontal,lying on the ground. The block comes in contact with the semi-sphere  and moves over it , deforming the semi-sphere in the process.

                     

                    I am interestd in understanding the deformation of the semi-sphere during this movement.

                     

                    So since the block moves all over the sphere I believe that defining a contact with a small portion of the semi-sphere wouldnt work.

                     

                    What method should be employed in this case?

                      • Re: NonLinear Dynamic simulation
                        Nicholas Luyster
                        Nicholas Luyster Jul 16, 2014 3:30 PM (in response to Roshan Santhosh)

                        Hi Roshan,

                         

                        The problem which you are stating sounds very tricky. It will be a challenge to converge and will take some time to complete. 

                         

                        With this said, it's important to make sure the fixtures are defined in such a way so that you can maximize stability.  If possible, you want to restrict the way the block bounces off or end the simulation before that happens. 

                         

                        The other thing to keep in mind is that the half sphere should be made out of a deformable non linear material. However, you should start with a linear material first and then add a non linear material later.

                         

                        Keep in mind that the best approach will be to keep making new simulations.  In each simulation, only change a little bit and then see how that affects the model.

                         

                        If after this, the simulation is still too difficult to perform, it may be best to hire someone to assist.  You're VAR would normally be your first point of contact.  If they are unable, I can assist as well. (I do charge for time but I am well worth it )

                         

                        All the best,

                         

                        Nick Luyster

                        Gosimulation.com

                        Online SolidWorks Simulation Training

                        • Re: NonLinear Dynamic simulation
                          Jared Conway
                          Jared Conway Jul 16, 2014 8:03 PM (in response to Roshan Santhosh)

                          is the dynamic part important? if not, go with just a nonlinear static analysis, in the video you'll see nick make an assumption about damping, that part will go away.

                           

                          that will make things a "bit easier".

                           

                          then just use a different kind of condition to move the block or the sphere. just make sure they are restrained in all directions and that the moving part only moves in the one direction that you want it to.

                           

                          but like nick said, make sure you know what you want to get out of this simulation. it will be challenging to setup and get to converge.