Astron. Astrophys. 317, 793-814 (1997) 6. Results of models with index 4/3As a check for the numerics two models were run with the same physical initial and boundary conditions, but differing in resolution of the accretor: model SS used 9 grids, model ST used 10 grids (cf. Table 1 for the other parameters). The two left panels of Fig. 10 show contour plots of the density distribution, while the acretion rates of mass and angular momentum can be found in Fig. 11. Especially the temporal evolution of the accretion rates confirm that also for using grids to 9 levels deep is sufficient: within the time in which the two models SS and ST overlap, the mass accretion rate and the angular momentum accretion agree to within a few percent.
Models SS and ST (Fig. 11) are equivalent to models IS and IT, (Fig. 4) so a comparison should show what effects are due to the different adiabatic index . Additionally, model SS can be compared to model CS shown in Fig. 6 of Ruffert (1995). Both the mean and the amplitude of the mass accretion rate of model SS are similar to the ones of model CS, and even the angular momentum accretion fluctuations do not show any striking difference. When comparing model SS to model IS one notices that the mass accretion rate proceeds in a much more unstable way in model SS. The rate of mass accretion as well as the specific angular momentum of the z -component are larger in model SS. So much so, that in a short burst model ST actually reaches the analytically estimated value (middle right panel of Fig. 11) contrary to model IS or IT. Thus we conclude that the models with accrete in a less stable way than their equivalents, which is the opposite from what had been observed in the simulations without gradients (Ruffert 1995 and references therein). More models are necessary before any more systematic statements can be made. © European Southern Observatory (ESO) 1997 Online publication: July 8, 1998 |