Radiatively Inefficient Mhd Accretion-ejection Structures

We present magnetohydrodynamic simulations of a resistive accretion disk continuously launching transmagnetosonic, collimated jets. We time-evolve the full set of magnetohydrodynamic equations, but neglect radiative losses in the energetics (radiatively inefficient). Our calculations demonstrate that a jet is self-consistently produced by the interaction of an accretion disk with an open, initially bent large-scale magnetic field. A constant fraction of heated disk material is launched in the inner equipartition disk regions, leading to the formation of a hot corona and a bright collimated, superfastmagnetosonic jet. We illustrate the complete dynamics of the “hot” near steady-state outflow (where thermal pressure ≃ magnetic pressure) by showing force balance, energy budget and current circuits. The evolution to this near stationary state is analyzed in terms of the temporal variation of energy fluxes controlling the energetics of the accretion disk. We find that unlike advection-dominated accretion flow, the energy released by accretion is mainly sent into the jet rather than transformed into disk enthalpy. These magnetized, radiatively inefficient accretion-ejection structures can account for under-luminous thin disks supporting bright fast collimated jets as seen in many systems displaying jets (for instance M87). Subject headings: Accretion, accretion disks — galaxies: jets — ISM:jets and outflows — MHD 1. ACCRETION DISKS AND JETS 1.1. Accretion-Ejection models Astrophysical jets are quite common phenomena across our visible universe. They are typically observed in association with accreting objects such as low-mass young stellar objects (YSO), X-ray binaries (XRB) or active galactic nuclei (AGN, e.g. Livio (1997) and references therein). In all these systems, the mass outflows exhibit very good collimation at large distances from the central object as well as high velocities along the jet axis. Although operative at widely disparate lenghtscales, these accretion-ejection systems have other features in common, in particular, observational links have been established in all cases between accretion disk luminosity and jet emission: for YSO see e.g. Hartigan et al. (1995), for XRB see Mirabel et al. (1998) and for AGN Serjeant et al. (1998). The most promising unifying model relies on a scenario where an accretion disk interacts with a largescale magnetic field in order to give birth to bipolar collimated jets (for the specific case of early protostars see also Lery et al. (1999)). Since the seminal work by Blandford & Payne (1982), it is known that the action of an open magnetic field configuration threading a disk can brake rotating matter in order to transfer angular momentum into the jet and provide energy for acceleration of jet matter. This magnetohydrodynamic (MHD) model describes the interaction of the accretion flow with a magnetic field whose origin can be due to advection of interstellar magnetic field (Mouschovias 1976) and/or dynamo produced (Rekowski et al. 2003). The collimation of the flow is self-consistently achieved by the electric current produced by the flow itself (Heyvaerts & Norman 1989). This current provokes a radial pinching of the plasma that can balance both magnetic and thermal pressure gradients (Sauty et al. 2002). The analytical work by Blandford & Payne (1982) did make the simplifying assumption of a cold plasma. Numerous studies have dealt with these Magnetized Accretion-Ejection Structures (MAES) in the last two decades. Sophisticated semi-analytical models deal with stationary self-similar investigations gradually extending the Blandford & Payne (1982) model with more physical effects, starting from simple vertical mass flux prescriptions (Wardle & Königl 1993) to realistic disk equilibria where ambipolar diffusion (Li 1996), resistivity (Ferreira 1997) or both viscosity and resistivity are adequately incorporated (Casse & Ferreira 2000). These studies bring deep insight in the physical conditions prevailing at the disk surface required to launch jets, but fail to give realistic jet topologies in the superAlfvénic region. The latter has spurred a variety of numerical MHD studies, aiming at a more realistic description of the trans-Alfvénic flows. The complexity of the dynamics of the accretion-ejection flow has forced many authors to either focus on jet dynamics alone (Krasnopolski et al. 1999; Ouyed & Pudritz 1997; Ustyugova et al. 1995, 1999) or to study disk-outflow dynamics over very short timescales (Kato et al. 2002; Matsumoto et al. 1996; Ushida & Shibata 1985). The latter class of studies were done using ideal MHD framework which is inconsistent with long-term jet production since the frozen-in magnetic structure is advected with the disk material. This leads to a rapid destabilization of the system due to a magnetic flux accumulation in the inner part of the disk. Moreover, in most of the previous studies, the accretion disk is modeled as a non-Keplerian thick torus without any outer mass inflow which would mimic the mass reservoir of the disk outer