Capability Development for Modeling Early Planetary System Formation, 15-9447Printer Friendly Version
Inclusive Dates: 01/01/04 06/30/05
Background - Modeling the formation of solar systems is a major interest of astronomical funding agencies, and many groups have undertaken extensive research to study this problem. Traditionally, solar systems have been believed to form predominantly in calm environments, isolated from the effects of other stars. Recent observational discoveries, however, indicate that perhaps as many as 90 percent of stars may form in very hot, violent environments that may make the creation of planets significantly harder. The growth and survival rates for dust grains and icy bodies in these violent environments are strongly controlled by heating and stellar winds from intense ultraviolet (UV) bright massive stars, causing processes that have not yet been considered by planetary formation models. Young disks surrounding these stars appear to be destroyed faster than planets can be formed within them. The formation of planets has been heretofore un-studied in these systems; furthermore, it is not known in what type of region our own solar system formed. The questions posed by these problems have profound significance for unraveling the story of planetary formation throughout the universe.
Approach - We used a commercially-available computer language to develop our model, which contains approximately 4,000 lines of code. It is structured in a way that allows for the easy addition of modules describing new physical processes. We had originally planned to port the code to a higher-performance language such as FORTRAN; this turned out to be unnecessary because our chosen language was of sufficiently high performance and allowed for faster model development.
The code simulates the evolution of a circumstellar disk, from its original state made of small, sub-micron-sized dust grains, to growth to meter-scale 'planetesimals' formed after approximately 1 Myr. The mode's processes were chosen to be representative of those in circumstellar disks inside dense star clusters. We include the following physical processes: a) Photo-evaporation of the disk by EUV and FUV (massive external stars); b) Grain growth by collisional sticking; c) Grain growth by gravitational instability; d) Dust grain settling to midplane; e) Viscous evolution of disk's gas and dust; f) Photo-sputtering of ice grains.
The code tracks abundance in a two-dimensional grid of orbital radius and vertical height. The disk is assumed to be azimuthally symmetric. In each bin, the abundances of gas, ice, and dust are stored; for ice and dust, their sizes are also stored. From an initial condition based on observations, the code progresses by a time-stepping mechanism. At each time step, a transition matrix is calculated that is used to compute the abundance at the next time step. The code uses a variable timestep size to maximize efficiency. It is able to simulate the evolution of a disk from t = 0 to t = 1 Myr in several minutes on a workstation-class machine.
Accomplishments - We used our code to discover a new physical mechanism for planetesimal formation. In this mechanism, the formation of large bodies may be 'triggered' by photo-evaporation of the gas. When photo-evaporation occurs in a settled disk, gas is preferentially removed from the disk, leaving dust grains. These grains are unstable to gravitational collapse (to form kilometer-scale bodies or larger), but only after the gas that inhibits their collapse is removed. This mechanism has not been modeled previously. This model may be of significance in explaining both how the planets in our solar system formed and how those in distant systems may form.
We have also applied our new code to model the formation of organic molecules in circumstellar disks. UV radiation from hot young stars can create organic molecules such as amino acids from shorter molecules via photolysis. Pre-biotic organic molecules created this way may be of significance in explaining the evolution of life on Earth and elsewhere. We demonstrated that the UV flux deposited is sufficient to photolyze nearly the young solar system's entire content of volatiles. The quantity of organic materials produced by UV photolysis exceeds by a factor of 107 that modeled previously, making this process potentially extremely important.