Can satellites deliver substructures and black holes to inner halo by dynamical friction ?

Dynamical friction, or the rate for a satellite to decay its orbit in a host galaxy halo, is often severely overestimated when applying the ChandraSekhar’s formula without correcting for the tidal loss of the satellite and the adiabactic growth of the host galaxy potential over the Hubble time. As a satellite decays to the inner and denser region of the host galaxy, the high ambient density boosts the exchange of energy and angular momentum between the satellite and the host, but on the other hand shrinks the Roche lobe of the satellite by tides. Eventually the processes of orbital decay and tidal stripping hang up altogether once the satellite is light enough (about 10M⊙). These competing processes can be modeled analytically for a satellite if we parametrize the massloss history by an empirical formula. We also take into account the adiabatic contraction of orbits due to moderate evolution of the potential well of the host galaxy in the formation of galaxy disk and bulge. We find a zone in the mass-distance plane forbidding an originally outer halo (high-mass and high-angular momentum) satellite from surviving as a low-mass remnant in the inner galaxy. Observed dwarf galaxies often show a finite density core, which determines how much inwards its remnants could be delivered to the host galaxy. Ghost streams or remnant cores or globular clusters of satellites should populate preferentially the outer halo (e.g., the Magellanic stream and the Ursa Minor dwarf spheroid), rather than the inner halo (e.g., the Sagittarius stream and the ω-Centauri cluster). The fact that strong tides in the inner galaxy can strip off the bulk of the satellite mass and shut off its dynamical friction also makes it problematic for any supermassive black holes in nuclei of satellite galaxies to decay in orbit and merge into the center of the host galaxy. PPARC Advanced Fellowship ([email protected])