Abstract. Dynamical models of star clusters are maturing in the sense that effects other than simple point particle dynamics are taken into account. We summarize the relevance of and prospects for this new generation of N -body models.
Summary and IntroductionRealistic star clusters are not born instantly nor are they isolated from external perturbations nor do they consist of single time-independent equal-mass points. On the contrary a realistic star cluster is born gradually from a contracting gas cloud, which is embedded in the external potential of a galaxy, and consists of evolving single stars as well as multiple systems. Owing to lack of computer power and partially to lack of software, computations, in which all these effects are accounted for, have not yet been performed.However numerical models of star clusters are beginning to come of age and incorporate deviations from the ideal star cluster. Some models already include mutual interaction between stellar evolution and star cluster dynamics (see Chernoff & Weinberg 1990;Fukushige & Heggie 1995;de la Fuente Marcos 1996;Spurzem & Aarseth 1996;Einsel & Spurzem 1997;Portegies Zwart et al. 1997;Tout et al. 1997, and the reviews Hut et al. 1992 andHeggie 1997). Here we study the arguments for combining stellar evolution and stellar dynamics in hybrid models, the advantages and disadvantages of performing such model computations and outline the future of dynamical models of star clusters.
From zero to first order modellingPerturbations of the evolution of dynamical star clusters can be subdivided into two classes: those which affect the cluster's evolution from outside and are not significantly affected by the evolution of the star cluster, and those which affect it from inside and actually live in symbiosis with it undergoing mutual interaction. This mutual interaction can be thought of as the ecology of the star cluster (see Heggie 1992).
THE TIMESCALE ARGUMENTThe internal evolution of a stellar system is governed by two fundamental time scales: these are the crossing time t crss = r vir / v , the ratio of the cluster virial radius to the mean velocity of its components, and the relaxation time t rlx . The ratio of these t rlx /t crss ≡ n rlx is roughly proportional to N . In real globular clusters N ≈ 10 5 and both timescales are usually well separated by more than three orders of magnitude.Stimulated evolution of the star cluster such as stellar evolution or the external potential of the Galaxy introduces new time scales. It is generally the shortest timescale which drives the stimulated