Numerical simulations of spiral galaxy formation and recoiling black holes.

摘要

This thesis discusses two major topics in regard to the formation and evolution of galaxies and their central massive black holes (MBH).;Part 1 explores the detectability of recoiling massive back as kinematically and spatially offset active galactic nuclei (AGN). Chapter 3 is devoted to understanding the effect of an aspherical dark matter potential on the trajectories of the MBHs. This is done through collisionless N-body simulations of kicked black holes in the Via Lactea I halo and through a semi-analytical model that accounts for the evolution of the halo's triaxiality as a function of radius over cosmic time. We find that the return time of MBHs that wander through a differentially triaxial halo is significantly extended in comparison with spherical models. This is because their trajectories are become highly non-radial which prevents them from passing near the halo's center, where dynamical friction is most efficient. Chapter 4 puts recoiling MBHs into context. Here we carry out N-body + SPH simulations of recoiling MBHs in high-resolution galaxy mergers with mass ratios 1:1 (Mayer et al. 2007), 1:4, and 1:10 (Callegari et al. 2009). We study not only the trajectories and return times of these black holes, but also their detectability as spatially/kinematically offset AGN. We find that the probability of detection of these MBHs is extremely low. The detection of large kinematic offsets requires that the MBH have relative offset velocities Deltav > 600km s-1 at the time of observation. This is unlikely due to (1) the low probability of large recoils to occur from a general-relativistic viewpoint, and (2) the short time scale during which the MBH can sustain large velocities even if the initial kick is high. The large amounts of gas funneled to the center of the host potential during mergers also prevents MBHs from reaching large apocenter distances, which hampers their detection as spatially offset AGN, especially at high redshift when recoil events are expected to be common.;Part 2 is dedicated to the formation of massive disk galaxies through N-body + SPH simulations. There, I describe the properties of Eris, the highest resolution cosmological simulation to date of the formation of a Milky Way-like galaxy from z = 90 to z = 0. Eris appears to solve the long-standing problems of mass concentration, which traditionally lead to the formation galaxies with large spheroidal components and small disks. A combination of high-resolution and high star formation threshold was the key to the success of Eris, because stars are only allowed to form at the highest density peaks and therefore feedback is more efficient in removing preferentially low angular momentum gas. Previous simulations tended to over-produce stars in low-density regions, where feedback is ineffective. Eris is in agreement with the Tully-Fischer, and M* - Mhalo relations, matches the observed surface brightness breaks in nearby spirals, is consistent with SigmaSFR - Sigma HI observations in spirals, and agrees with constraints on the hot gas mass abundance in the Galaxy. In addition, Eris' baryon fraction is 30% lower than the universal value, due to star formation driven outflows.