Аңдатпа
In this paper we use two suites of ultra-high resolution N-body simulations Phoenix and Aquarius Projects to study the assembly history of sub-halos and its dependence on host halo mass. We found that more massive haloes have more progenitors, which is in contrast with former works because they counted dynamical progenitors repeatedly. Less massive halos have larger fraction of dynamical progenitors than more massive ones. The typical accretion time depends strongly on host halo mass. Progenitors of galactic halos are accreted at higher redshift than that of cluster halos. Once these progenitors orbit their primary systems, they rapidly lose their original mass but not their identifiers. Most of the progenitors are able to survive to present day. At given redshift, the survival fraction of accreted sub-halos is independent of host halo mass, while sub-halos in high mass halos lost more mass. In the second part, we use a semi-analytical galaxy formation model compiled on a Millennium Simulation to study the size evolution of massive early-type galaxies from redshift z = 2 to present days. We find that the model we used is able to well reproduce the amplitude and slope of size-mass relation, as well as its evolution. The amplitude of this relation reflects the typical compactness of dark matter halos at the time when most stars are formed. This link between size and star formation epoch is propagated in through galaxy combinations. Minor combinations are increasingly important with increasing present day stellar mass for galaxies more massive than 1011.4M⊙. At lower masses, major combinations are more important. In situ star formation contributes more to the size growth than it does to stellar mass growth. Similar to former works, we find that minor combinations dominate the subsequent growth both in stellar mass and in size for early formed early-type galaxies.