Model RL is extreme in two aspects: firstly the radius of the accretor is one accretion radius in size and thus very large, and secondly the flow speed upstream of the accretor is subsonic, Mach 0.6. Additionally the gradient was chosen to be large (0.2). These conditions, although extreme, might, however, be present in stellar binary systems with a wind from a nova (K. Schenker, personal communication), thus the model is not totally academic, but one has to take care in relating the physical size of the stars to the radius of the accretor used in the simulations. The contour plot showing the density distribution together with the instantaneous velocities is shown in Fig. 10b, while the corresponding accretion rates are shown in the bottom panels of Fig. 11. The large scale of the contour plot shows the velocity gradient clearly.
After time units model RL has practically reached a stationary state in which mass is accreted at a constant rate (bottom left panel in Fig. 11. This behaviour as well as the value of the mass accretion rate is very similar to model SL in Ruffert (1994b; top left panel of Fig. 2). The small fluctuations of the angular momentum visible in model SL are initial transients due to a small (3%) perturbation of the initial density distribution. I did not perturb the density in any model presented in this paper. This model RL is the only model presented in this work that reaches a constant state, which is probably due to the subsonic flow, sa has been reported elsewhere for models without gradients (e.g. Ruffert 1995).
The main difference concerning accreted quantities between model RL and all other models in this paper is the sign of the accreted angular momentum: it is positive (cf. Fig. 11), while the z -component of the angular momentum accreted of all other models is negative (see also Table 1). This is probably due to the fact that an accretor with such a large radius accretes mainly via its large geometric cross section, and not following the classic Bondi-Hoyle-Lyttleton scenario involving gravitational focussing, accretion mainly from the downstream side, etc. (the low Mach number might also play a role). Thus the matter is directly absorbed from the incoming stream without first being "processed" though a shock, taking with it the angular momentum it had upstream: this produces a anti-clockwise rotating vortex around the accretor (contrary to the clockwise vortex mentioned in Sect. 5).
© European Southern Observatory (ESO) 1997
Online publication: July 8, 1998