Over the past few years, much attention has been given to the environment and evolution of young stellar objects. The presence of outflows emanating from the vicinity of the young stellar object (Snell et al. 1980), of optical jets (e.g. Mundt 1985) and of accretion disks (Adams et al. 1990, Sargent & Beckwith 1991, Cabrit et al. 1995) have raised new theoretical problems and provide a rich literature in this field. The numerous observations suggesting the existence of disks around young stars (e.g. Beckwith et al. 1990) reinforce the idea that they naturally follow from the star formation process.
It is commonly accepted that star formation results from the gravitational contraction of an interstellar cloud. Numerical calculations first performed by Larson (1969) indicate that collapse is distinctly non-homologous. The region of highest density near the center first collapses, while the outer layers remain temporarily static. As material reaches the center, it is stopped in a strongly radiating shock front. Inside the shock, the material settles slowly onto a growing hydrostatic core, which will in time become a pre-main sequence star (Winkler & Newman 1980a, b, Stahler et al. 1980 a, b).
Larson's original simulations assumed spherical symmetry and absence of initial rotation. Recent computations including these effects have been performed by several groups (e.g. Bodenheimer et al. 1990; Yorke et al. 1993, 1995). Their results show that the initial angular momentum of the interstellar molecular cloud will cause the centrifugal forces to increase during the collapse. Consequently a disk will form, surrounding a central low-mass core. Thereafter, the remaining material in the outer regions rains predominantly onto the surface of the disk (Terebey et al. 1984, Cassen et al. 1985, Lin & Pringle 1990). Indeed, the small cross-sectional area of a star in comparison with its disk leads to expect that most of the mass from a collapsing cloud must first fall into the disk.
Some works however have gauged the effect of accretion geometry. Mercer-Smith et al. (1984) handle accretion from a disk by specifying as boundary conditions a mass addition rate and an accretion luminosity. For their study of accretion onto main sequence stars, Shaviv & Starfield (1988) included the internal energy of the infalling material in the surface layers. Palla & Stahler (1992), in their computations of protostars, changed the stellar boundary conditions from those of a shock to those of an ordinary photosphere. However none of them considered that accretion could also involve deeper layers. In this work, we treat the accretion process not only as a surface condition but as a global phenomenon that may also concern the central part of the star. To model the mass redistribution of accreted matter inside the star, we use the formalism developed by Siess & Forestini (1996a, hereafter SF).
In the next section we discuss how we incorporate the accretion process in our stellar evolution code, giving special attention to the nucleosynthesis treatment. In Sect.3, the structure and evolution of low-mass accreting stars is analyzed in detail. More specifically, we focus on the effect of accretion rate, the influence of chemical composition and the consequences of different mass accretion distributions. In Sect.4 we compare our results with standard pre-main sequence (PMS) evolution, and in Sect.5 we examine how accreting stars relax once accretion stops. Finally we summarize and discuss our results in Sect.6.
© European Southern Observatory (ESO) 1997
Online publication: April 8, 1998