We have followed the evolution of a star cluster, to the point of dissolution in the tidal field of the parent galaxy, taking into account both the effects of stellar dynamics and of stellar evolution. Our calculations are based on direct N-body integration, coupled to approximate treatments of stellar evolution.
Our results differ greatly from those obtained with Fokker-Planck calculations, as presented by CW90: their model clusters dissolve after a few times years, whereas our equivalent model clusters live at least ten times longer. As we discussed in the previous section, a number of different reason conspire to produce such a drastic difference.
Our hope was that we would be able to find a way to bridge our N-body results and previous results based on Fokker-Planck approximations. The fact that the GRAPE-4 special-purpose hardware allowed us to model much larger numbers of particles, reaching to within an order of magnitude of that of real globular clusters, seemed to indicate that it would finally be possible to make a firm connection between the two types of simulations. However, our results indicate that no clear process of extrapolation has emerged yet. Even within the different runs we have studied, extrapolation from the smaller to the larger number of stars would have resulted in rather large errors. This suggests that further extrapolation will suffer from the same fate.
In the present paper, we have studied in detail a single model. However, the way the result depends on the time scaling might be different for other models. In a subsequent paper, we plan to carry out a systematic study, similar to the one we have presented here, for a much wider range of initial models.
© European Southern Observatory (ESO) 1998
Online publication: August 17, 1998