Computer simulations of planetary accretion dynamics: Sensitivity to initial conditions

Abstract

We have tested the implications and limitations of Program ACRETE, a scheme based merely on Newtonian physics and accretion with unit sticking efficiency, devised by Dole in 1970 to simulate the origin of the planets. The dependence of the results on a variety of radial and vertical density distribution laws, on the ratio of gas to dust in the solar nebula, on the total nebula mass, and on the orbital eccentricity, ϵ, of the accreting grains are explored. Only for a small subset of conceivable cases are planetary systems closely like our own generated. Many models have tendencies toward one of two preferred configurations: multiple-star systems, or planetary systems in which Jovian planets either have substantially smaller masses than in our system or are absent altogether. But for a wide range of cases recognizable planetary systems are generated, ranging from multiple-star systems with accompanying planets, to systems with Jovian planets at several hundred astronomical units, to single stars surrounded only by asteroids. Many systems exhibit planets like Pluto and objects of asteroidal mass, in addition to usual terrestrial and Jovian planets. No terrestrial planets were generated more massive than five Earth masses. The number of planets per system is for most cases of order 10, and, roughly, inversely proportional to ϵ. All systems generated obey a relation of the Titius-Bode variety for relative planetary spacing. The case with which planetary systems are generated using such elementary and incomplete physical assumptions supports the idea of abundant and morphologically diverse planetary systems throughout the Galaxy.