Session 21P. Planetary Formation and Dynamics
Contributed Poster Session, Wednesday, October 14, 1998, 2:00-3:40pm, Hall of Ideas

[21P.13] Direct Simulation of Planet Formation with a Million Planetesimals: First Results

D. C. Richardson, T. Quinn, J. Stadel, G. Lake (U. Washington )

We present the first results from our direct simulations of planet formation using N = 106 rocky particles (see also {\it BAAS} {\bf 30}, 765 and {\it BAAS} {\bf 29}, 1027). Currently ~1000 yr of evolution have been simulated under the assumption of perfect accretion in a 4.7 M\oplus disk that extends from 0.8 to 3.8 AU with a surface density distribution \Sigma \propto r-3/2. The four present-day giant planets are included as perturbers. The disk was started cold, allowing the growth of resonance gaps and spiral density waves due to the giant planets to be seen clearly. We are in the process of testing a scheme to improve the speed of the method by at least an order of magnitude by exploiting the near-Keplerian nature of the planetesimal orbits. Preliminary results from this improvement are presented.

We also present a simulation of the formation of the Galilean satellites using N = 105 icy particles in a 0.065 M\oplus disk extending from 3 \times 105 to 2.3 \times 106 km around Jupiter and having the same r-3/2 density law. Particles are allowed to merge following a collision only if their rebound velocity is less than their mutual escape velocity, and if the spin of the resulting merged body does not exceed the classical breakup limit; otherwise the particles bounce and lose a fixed fraction of their relative energy. Fragmentation has not been implemented at this time. We find that this system evolves very quickly, with ~ 33% of the particles having merged after only ~150 d, or about 80 Io orbits. This simple model begins with a uniform population of 100 km radius planetesimals and ignores gas dynamics. The applicability of this model toward the actual formation of giant planet moons is discussed.