3 Replies Latest reply on Jan 9, 2008 12:58 PM by Andrew Thompson

    rocking chair simulation

    Andrew Thompson
      Hello,

      Absolute newbie to cosmotion here. I'd like to use simulation to test the movement of various rocking chair designs, in particular I'd like to find out how each design will sit at rest after rocking (that is how leaned-forward or leaned back)--no weight in the chair.

      So far I've set up an assembly that includes my rocking chair (or substitute arc) and a platform acting as the floor.

      I've set a tangent mate between the face of the floor and the bottom face of the rocker. I begin the simulation with the chair tilted back (or forward) and let gravity and various 3dcontact setups do their thing.

      Problem is the system does not seem to come to a rest. The rocking movement slows quickly at first, but then seems top plateau-out and maintain a sort of consistent wiggle from there on.

      Any ideas about how I might better set this up to get more realistic results?

      Thanks.
        • rocking chair simulation
          Pete Yodis
          Andrew,

          I would think you would need some damping characeristics in your motion model, otherwise your system will remain in movement (conservation of energy). A damping characeristic would provide a means of energy dissipation. Instead of a tangent mate between the floor and rocker, you could choose a 3D contact interaction and define the contact properties for the bodies. You may be able to enable friction in your tangency joint as well - without having to delete the tangnecy joint and create a 3D contact parameter.
            • rocking chair simulation
              Ian Hogg
              Andrew,

              I concur with Peter.

              By default there is no energy loss without contact damping or friction. You may however see a little numeric damping/instability just due to accuracy (normally by having a large time step or low accuracy setting for the solver)

              If you have a mate between the chair and ground and 3-D contact, the mate will superscede the contact. It's all about load paths. A mate is infinitely stiff whereas contact defines a stiffness and damping between the bodies contacting. Hence the stiffest path to ground is through the mate.

              Remember also that you will be ignoring other forms of energy loss (air resistance, structural damping, noise) and these will have some effect.

              If you are just interested in seeing how the chairs will sit at equilibrium, physical simulation should be sufficient as it does not account for inertia and achieves the steady state solution (in 2007 at least, it does do a little more than that in 2008). in other words, based on the geometry, and the cg location, it will show the orientation of the chair.

              You can sort of do the same thing in COSMOSMotion by adding a damper to the system to quickly eliminate the movement due to starting at a leant forward or backward position and depending on the value, it will come to rest in a quick time.

              Hope this helps you out.

              Cheers,

              Ian
                • rocking chair simulation
                  Andrew Thompson
                  Thanks for the replies. This makes sense. Let's assume I do this with cosmosmotion, since I'd like to learn more about how to use it.

                  I've started by eliminating the tangent mate altogether. The simulation seems to run exactly as before.

                  You've both noted that the system will not come to rest without some sort of dampener. I thought I had done that with 3dContact, but I surely don't fully understand how it works.

                  I'm currently applying 3dContact between the chair and the floor. I've tried it by assiging specific materials+friction and by entering various friction values alone. These seem to work in that the simulation does rapidly slow, and the range of rocking motion is reduced, but it eventually hits a plateau, where the motion and speed stay constant. How else should I apply a dampener so that it actually stops?

                  EDIT: I just applied a damper to the faces that contact. That slowed it to a rest. I'll play with that for a while. Thanks.

                  Many thanks in advance.