STATISTICS OF N-BODY SIMULATIONS .2. EQUAL MASSES AFTER CORE COLLAPSE

This paper presents and analyses statistical results from a large numb er of N-body simulations of isolated systems with equal masses, in whi ch 250 less than or equal to N less than or equal to 2000. It concentr ates on the phase starting around the end of core collapse. Binaries p lay a crucial role, and we find that the total energy of bound pairs i s in line with theoretical expectations. Interpretation of the total n umber is complicated by the presence of a number of binaries on the ha rd/soft threshold. Interactions of hard binaries are consistent with t he Spitzer cross-section. The spatial evolution of the half-mass radiu s after core collapse nearly follows classical theory, and, by compari son with Fokker-Planck and gas models, allows a redetermination of the effective thermal conductivity and the argument of the Coulomb logari thm in the expression for the relaxation time. The evolution of the in ner parts of the system around the time of core bounce is consistent w ith these simplified models, provided that the continuous production o f energy, as is usually assumed, is replaced by a model of stochastic energy production. Similarly, postcollapse evolution of the core requi res a modest recalibration of the coefficient of energy generation, es pecially for small N. These remarks refer to the behaviour averaged ov er many models; individual cases show alternate and irregular phases o f expansion and recollapse. The distributions of velocity dispersion a nd anisotropy become remarkably homologous soon after core bounce. The bound mass of the systems very nearly follows a power law with time. A small number of escapers, presumed to be those associated with binar y activity, dominate the energy that is carried off: the distribution of energies of escapers changes abruptly at the end of core collapse. The 'internal' energy of escaping binaries is consistent with theoreti cal expectations, and again supports Spitzer's reaction cross-section for hard binaries.