Global magnetohydrodynamic simulations of cylindrical Keplerian disks

Authors
Citation
Jf. Hawley, Global magnetohydrodynamic simulations of cylindrical Keplerian disks, ASTROPHYS J, 554(1), 2001, pp. 534-547
Citations number
38
Categorie Soggetti
Space Sciences
Journal title
ASTROPHYSICAL JOURNAL
ISSN journal
0004637X → ACNP
Volume
554
Issue
1
Year of publication
2001
Part
1
Pages
534 - 547
Database
ISI
SICI code
0004-637X(20010610)554:1<534:GMSOCK>2.0.ZU;2-#
Abstract
This paper presents a series of global three-dimensional accretion disk sim ulations carried out in the cylindrical limit in which the vertical compone nt of the gravitational field is neglected. The simulations use a cylindric al pseudo-Newtonian potential, proportional to 1/(R - R-g), to model the ma in dynamical properties of the Schwarzschild metric, The radial grid domain runs out to 60R(g) to minimize the influence of the outer boundary on the inner disk evolution. The disks are initially constant density with a Keple rian angular momentum distribution and contain a weak toroidal or vertical held that serves as the seed for the magnetorotational instability. These s imulations reaffirm many of the conclusions of previous local simulations. The magnetorotational instability (MRI) grows rapidly and produces MHD turb ulence with a significant Maxwell stress that drives accretion. Tightly wra pped low-m spiral waves are prominent. In some simulations radial variation s in Maxwell stress concentrate gas into rings, creating substantial spatia l inhomogeneities. As in previous global simulations, there is a nonzero st ress at the marginally stable orbit. The stress is smaller than seen in str atified torus simulations but nevertheless produces a small decline in spec ific angular momentum inside the last stable orbit. Detailed comparisons be tween simulations are used to examine the effects of various choices in com putational setup. Because the driving instability is local, a reduction in the azimuthal computational domain to some fraction of 2 pi does not create large qualitative differences. Similarly, the choice of either an isotherm al or adiabatic equation of state has little impact on the initial evolutio n. Simulations that begin with vertical fields have greater field amplifica tion and higher ratios of stress to magnetic pressure compared with those b eginning with toroidal fields. In contrast to MHD, hydrodynamics alone neit her creates nor sustains turbulence.