A significant part of the work that Josh is doing is lab experiments where he looks at impacts into granular material in a vacuum. We will try to duplicate the behavior seen in the experiments in simulations. This page shows what has been done so far.
We begin with the early stages. My code was originally developed to simulate particles in planetary rings. The simulation code is a hybrid code that has fixed time steps, but models collisions as discrete events that can happen between time steps. Root finders are used to find exact times of collisions and handle them. In planetry rings, the only collision events are between particles. One of my students, Cameron Swords, convinced me this summer that it wouldn't be as hard as I had always thought to put in other types of events. With that addition, it is possible to make systems with various surfaces in them for particles to collide with.
The first round of simulations were simply to verify the new methods. These began with an infinite plane and a collection of dropping spheres that bounce on it. After that, a cylinder was added to confine the particles and mirror the actual experiment. Nothing here is scaled properly, but it was used to verify the methods used.
With the methods verified, the next step was to create a system similar in
scale to the actual experiment. The first tests of the experimental setup
were done with glass spheres ~1mm across in a basin 14.9cm in diameter.
Once particles had been allowed to settle, a larger particle was dropped onto
them.