The dynamical evolution of substructure

The evolution of substructure embedded in non-dissipative dark haloes is studied through N -body simulations of isolated systems, both in and out of initial equilibrium, complementing cosmological simulations of the growth of structure. We determine by both analytic calculations and direct analysis of the N -body simulations the relative importance of various dynamical processes acting on the clumps, such as the removal of material by global tides, clump–clump heating, clump–clump merging and dynamical friction. The ratio of the internal clump velocity dispersion to that of the dark halo is an important parameter; as this ratio approaches a value of unity, heating by close encounters between clumps becomes less important, while the other dynamical processes continue to increase in importance. Our comparison between merging and disruption processes implies that spiral galaxies cannot be formed in a protosystem that contains a few large clumps, but can be formed through the accretion of many small clumps; elliptical galaxies form in a more clumpy environment than do spiral galaxies. Our results support the idea that the central cusp in the density profiles of dark haloes is the consequence of self-limiting merging of small, dense haloes. This implies that the collapse of a system of clumps/substructure is not sufficient to form a cD galaxy, with an extended envelope; plausibly, subsequent accretion of large galaxies is required. The post-collapse system is in general triaxial, with rounder systems resulting from fewer, but more massive, clumps. Persistent streams of material from disrupted clumps can be found in the outer regions of the final system, and at an overdensity of around 0.75, can cover 10 to 30 per cent of the sky.