Long - term evolution of a dipole type magnetosphere interacting with an accretion disk . II . Transition

The evolution of an initially stellar dipole type magnetosphere interacting with an accretion disk is investigated numerically using the ideal MHD ZEUS-3D code in the 2Daxisymmetry option. Depending mainly on the magnetic field strength, our simulations may last several thousands of Keplerian periods of the inner disk. A Keplerian disk is assumed as a boundary condition prescribing a mass inflow into the corona. Additionally, a stellar wind from a rotating central star is prescribed. We compute the innermost region around the stellar object applying a non-smoothed gravitational potential. Our major result is that the initially dipole type field develops into a spherically radial outflow pattern with two main components, a disk wind and a stellar wind component. These components evolve into a quasi-stationary final state. The poloidal field lines follow a conical distribution. As a consequence of the initial dipole, the field direction in the stellar wind is opposite to that in the disk wind. The half opening angle of the stellar wind cone varies from 30 to 55 depending on the ratio of the mass flow rates of disk wind and stellar wind. The maximum speed of the outflow is about the Keplerian speed at the inner disk radius. An expanding bubble of hot, low density gas together with the winding-up process due to differential rotation between star and disk disrupts the initial dipole type field structure. An axial jet forms during the first tens of disk/star rotations, however, this feature does not survive on the very long time scale. A neutral field line divides the stellar wind from the disk wind. Depending on the numerical resolution, small plasmoids are ejected in irregular time intervals along this field line. Within a cone of 15 along the axis the formation of small knots can be observed if only a weak stellar wind is present. With the chosen mass flow rates and field strength we see almost no indication for a flow self-collimation. This is due to the small net poloidal electric current in the (reversed field) magnetosphere which is in difference to typical jet models.