4 Replies Latest reply on May 15, 2014 7:18 AM by Roy Dean

    Non-linear Dynamic Impact oddities....

    Roy Dean

      I'm trying to simulate an impact.   The model is simple: a hollow rectangular beam with a cylindrical "projectile" hitting it.

       

      Initial velocity is define for the cylinder.   Other boundary conditions are set, and the simulation runs.   "NSTAR.EXE has stopped working" is generally the result, but I get about a tenth of a second into the impact (significant deflection), prior to it stopping.   The odd things are:

       

      1)  it appears as though the cylinder never slows down.   Is there a way to show a plot of the velocity of the cylinder to see if it's slowing down as it hits the beam?

      2)  the beam appears to be "bouncing" off the cylinder (again, as if the cylinder isn't slowing down at all, and the beams elastic/inertia effects are causing it to bounce off the cylinder which could be traveling at a constant velocity... ie, the "unstoppable object").

      3)  Is NSTAR quitting right at the "reverse loading" point (ie, where the cylinder velocity has reached zero)?

       

      I've got bogus numbers defined for rayleigh damping (no other damping option available, .01 for both alpha and beta), which seems to have affected the results greatly.... prior to that, the cylinder would "load up" the beam, and then fly right through it (no penetration contacts are defined).

       

      Any casual observations?  Jared?????