I present results from three theoretical numerical studies relating to destructive events in the lives of outer solar system satellites and smaller bodies. The first project is a study of the implications that a Late Heavy Bombardment in the outer solar system, such as predicted by the Nice model, might have had for the mid sized moons of Jupiter, Saturn, and Uranus. A Monte Carlo calculation shows that Mimas, Enceladus, Tethys, and Miranda each would almost certainly have experienced at least one catastrophic collision after formation. If true, these bodies would have disrupted and then reaccreted as scrambled mixtures of rock and ice -- potentially preserving this signature in their present day structure. Conversely, if these satellites are fully differentiated today then they would have required a heat source sufficient for melting and differentiation in the absence of short half life radioactive elements. Tidal heating may have been sufficient for Tethys, Enceladus, and Miranda, but a differentiated Mimas would present a difficulty to either the Nice model or to the classical formation model of the Saturn system.

The second study is a numerical investigation of the expected outcome of destructive collisions between gravity-dominated bodies; in particular of the conditions required for a collision to be catastrophic, defined as one that leaves behind a surviving body with less than half the total colliding mass. In this study I focus on bodies with radii between 100 and 1000 km, a previously neglected size range, and derive a simple scaling law for the threshold impact energy required for disruption in this size range. This scaling law is expected to hold for all projectile-to-target size ratios and is independent of material, so long as elastic strength may be ignored. Compared with scaling laws existing in the literature the newly derived scaling generally predicts lower threshold energy for disruption, except for highly oblique impacts by projectiles much smaller than the target.

The third project is a study of the tidal breakup of rubble piles by modeling the breakup of comet Shoemaker-Levy 9 in a rigid body code that, for the first time, treats non-spherical rubble pile elements. This introduces dilatation and grain locking as the major forces acting against gravity tides during the comet's close approach to Jupiter and changes the outcome of tidal encounters compared with that predicted by models using spherical elements. By comparing simulation results to the well-studied post-breakup morphology of comet SL9 we were able to constrain the progenitor's bulk density at 300--400 kg/m^3, half that of previous estimates.