Cosmological simulations of galaxy formation including hydrodynamics

The formation of galaxies in hierarchical cosmogonies is studied using high resolution N-body simulations. The equations of gravity and hydrodynamics are evolved for two particle fluids: a pressureless dark matter component and a dissipational baryonic component. The collapse of structure is followed self-consistently from Mpc scale filamentary structures to kpc scale galactic objects.; The simulated galaxy population has sizes, masses, and abundances in the range of those for real galaxies. A large fraction of these objects show rotationally supported disk structures. This is the first demonstration that disk formation is a natural consequence of cosmological collapse. Isolated galaxies generally grow through steady mass accretion while objects in clustered regions are more likely to undergo mergers. Galactic disks form via discrete accretion of relatively large gas clouds and viscous transport of angular momentum to the inner regions. The formation era is consistent with high redshift observations and has a peak rate near redshift 3.; Galaxies collapse as segments of filaments and then flow along the filaments toward intersections. The formation of groups and clusters of galaxies is dominated by radial, directed infall which will increase the relaxation timescales as well as the incidence of merging, tidal disruption, and gas stripping. In this specific region, biases in both the spatial and kinematic distributions of galaxies relative to the dark matter are found. Virial mass estimates give results a factor of two to three lower than the true value. However, finer modelling of the stellar fluid within galaxies will be required before these findings can be generalized.; A didactic discussion of smoothed particle hydrodynamics (SPH) is presented along with several computational tests. Density estimation is shown to be sensitive to the resolution of SPH parameter choices and can greatly affect the physical results. Initial work on the self-similar spherical accretion problem indicates the general method is sound, but further work is required to search for optimal parameter sets. Tests of several galaxy identification schemes within simulations show that SPH methods are necessary for tracing galaxies through clustering and that the phenomenon of velocity bias has not yet been convincingly demonstrated.