![]() A sink interacts with the remaining gas through gravity and accretion only. Instead of artificially stopping the collapse at a chosen scale, sink particles approximate the unresolved small-scale evolution by an immediate collapse on to a point mass. Another way to deal with limited resolution in simulations of gravitational collapse are sink particles. A problem of this approach is that objects are kept artificially big and therefore more vulnerable to disruption through shocks and tidal stripping. ( 2010) named this approach ‘Jeans heating’. This is usually achieved by implementing a barotropic equation of state (EOS) that strongly heats the gas once a certain density is exceeded. A possible way to avoid this is changing the physical model in a way that will artificially stop the gravitational collapse at a scale that can still be resolved. ( 1997) have shown, not resolving the Jeans length and Jeans mass in regions of gravitational collapse can lead to artificial fragmentation of the gas. Introducing a maximum resolution raises another problem: as Truelove et al. It is therefore inevitable to define a maximum resolution at which one does not follow the ongoing collapse any further. In addition to the problem of time-scales, following the collapsing regions to higher densities requires an ever increasing spatial and mass resolution which increases the necessary number of resolution elements in the simulation. Adaptive time stepping that allows for different resolution elements to be updated with different time steps (see Bate, Bonnell & Price 1995 for a description in SPH, Teyssier 2002 for AMR) increases the computationally achievable dynamic range in time-scales, but long-term evolution of systems hosting sites of gravitational collapse is still not possible in many cases. Advancing the whole simulation at the smallest time step therefore lets the large-scale motions appear completely frozen. For example, a density contrast of 10 10 observed in giant molecular clouds from the entire cloud down to the first hydrostatic core (Stahler & Palla 2005) translates into a factor 10 5 between the smallest and the largest time-scale of the problem. The local free-fall time |$t_$| is a good estimate for the relevant time-scales of the dynamics at a given density. Resolving those collapses while still following the large-scale evolution of the gas therefore requires a huge dynamic range in the density. Hydrodynamics, methods: numerical, stars: formation 1 INTRODUCTIONĪstrophysical simulations of self-gravitating gas often involve regions of gravitational collapse. Statistical properties such as the sink mass function, the average sink mass and the sink multiplicity function are used to evaluate the impact that our new scheme has on accurately predicting fundamental quantities such as the stellar initial mass function or the stellar multiplicity function. We compare our new recipe for sink formation to other popular implementations. We detail all the necessary steps to follow the evolution of sink particles in turbulent molecular cloud simulations, such as sink production, their trajectory integration, sink merging and finally the gas accretion rate on to an existing sink. This is achieved using a general integral form of the virial theorem, where we use the curvature in the gravitational potential to correctly account for the background potential. Furthermore, we develop a new scheme to decide if the gas in which a sink could potentially form, is indeed gravitationally bound and rapidly collapsing. ![]() This allows us to unambiguously define a discrete set of dense molecular cores as potential sites for sink particle formation. Our main addition is the use of a clump finder to identify density peaks and their associated regions (the peak patches). We present a new sink particle algorithm developed for the adaptive mesh refinement code ramses. ![]()
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |