Abstract
The orbital evolution of small satellite galaxies or globular clusters sinking into triaxial parent systems is calculated for parent galaxies with Stackel (separable) potentials. Analytic expressions for the instantaneous decay rates of change of the orbital turning points are derived for arbitrary anisotropic drag forces. Orbit averaging is then used to compute evolutionary paths in turning-point space. While such a path strictly represents the average evolution of an ensemble of satellites distributed at random phase within the same initial orbit, it is demonstrated that it is often a good approximation for a single satellite if the satellite's mass is sufficiently small. Under the assumption that the drag forces can be approximated by the anisotropic version of Chandrasekhar's local dynamical friction formula, the effects of various sorts of velocity anisotropy and ordered motion (internal streaming) in the parent galaxy are estimated. In particular, the circumstances that favor orbital circularization, planarization, or margin crossing (i.e., change of family) are identified. Determining the validity of Chandrasekhar's formula in this application, either by modal analysis or through N-body simulations, is at the limit of present feasibility; however, a set of simulations using a restricted (non-selfgravitating) N-body code is presented, the results of which are consistent with the predictions of orbit-averaged local drag. Finally, some of the consequences of the sinking process, including the further evolution of stars and gas tidally stripped from satellites and the formation of "X-structures," boxy ellipticals, and gaseous disks or rings, are discussed.

This publication has 0 references indexed in Scilit: