The dynamics of tidal tails from massive satellites

(Abbreviated) We investigate the dynamical mechanisms responsible for producing tidal tails from dwarf satellites using N-body simulations. We identify two important dynamical co-conspirators: 1) the points where the attractive force of the host halo and satellite are balanced do not occur at equal distances from the satellite centre or at the same equipotential value for massive satellites, breaking the morphological symmetry of the leading and trailing tails; and 2) the escaped ejecta in the leading (trailing) tail continues to be decelerated (accelerated) by the satellite's gravity leading to large offsets of the ejecta orbits from the satellite orbit. The effect of the satellite's self gravity decreases only weakly with a decreasing ratio of satellite mass to host halo mass, demonstrating the importance of these effects over a wide range of subhalo masses. Not only will the morphology of the leading and trailing tails for massive satellites be different, but the observed radial velocities of the tails will be displaced from that of the satellite orbit; both the displacement and the peak radial velocity is proportional to satellite mass. If the tails are assumed to follow the progenitor satellite orbits, the tails from satellites with masses greater than 0.0001 of the host halo virial mass in a spherical halo will appear to indicate a flattened halo. Therefore, a constraint on the Milky Way halo shape using tidal streams requires mass-dependent modelling. Similarly, we compute the the distribution of tail orbits both in E_{r}-r^{-2} space and in E-L_{z} space, advocated for identifying satellite stream relics. The acceleration of ejecta by a massive satellite during escape spreads the velocity distribution and obscures the signature of a well-defined ``moving group'' in phase space.