Further simulations of merging galaxies

Galaxy collisions and the structure of the resulting merger remnants are studied using a large number of numerical simulations. These experiments extend earlier calculations of mergers between pairs of similar ‘galaxies’ (White 1978). The tidal coupling in collisions is found to depend strongly on the rotational properties of the ‘galaxies‘ involved. It is greatly enhanced if their spin vectors are aligned with that of their orbit, and it is suppressed if this alignment is reversed. The structure of a merger product depends only weakly on that of its progenitors. Such remnants are typically axisymmetric oblate systems with radially decreasing velocity dispersions and density profiles which have near power-law form over two decades in radius. This density structure is reasonably well described by de Vaucouleurs' empirical formula for the surface brightness distribution of elliptical galaxies. The flattening of merger remnants may be partly supported by an anisotropic pressure distribution, but the systems studied here nevertheless rotate considerably more rapidly than most observed elliptical galaxies, and a natural preference for nearly head-on collisions must be invoked if all ellipticals are to be identified as merger remnants. Mass and energy losses are found to be very small for mergers between bound or marginally unbound ‘galaxies’. Escapers can, however, carry away a significant amount of angular momentum.