Computational Studies of MHD Accretion Flows

Princeton University, Princeton NJ

Investigators

Abstract

AST 0098625 Stone A wide variety of objects, ranging from new stars in formation (protostars), to objects which are the cinders of burned out stars (white dwarfs, neutron stars, and black holes), to active galactic nuclei (AGN, such as quasars) are thought to have accretion disks surrounding them. Inflows of gas from these disks onto the central object seem to account for some of the most dramatic components of the Universe. They emit prodigious amounts of power and radiate it over a tremendous swath of the electromagnetic spectrum (from the mid-infrared to hard X-rays and gamma rays). Our theoretical understanding of these accretion flows, however, is still limited by the complexities involved in developing the necessary magnetohydrodynamics (MHD) and radiation hydrodynam-ics computer models. Our understanding of the local physics that control such flows has progressed rapidly in the last few years. It is now important to examine how local processes such as the magnetorotational instability (MRI) determine global disk structure and evolution, especially since only global disk models can be directly compared to high spatial-, spectral-, and time-resolution observations of accretion flows around protostars, white dwarfs, neutron stars, and black holes. Using computational methods, this project will develop the first time-dependent, three-dimensional MHD models of the interaction of an accretion disk with a magnetized and rotating central star. These calculations will allow quantitative measurement of the mixing rate of the stellar field into the disk, the size of the interaction region, the time-averaged torque exerted on the star, and the geometry and kinematics of any polar cap accretion flows that might form. Such quantities are fundamental to the theory of how magnetized stars interact with accretion disks, yet to date they have yet to be calculated from first principles. Direct comparison of the simulations to a large and varied set of observations will be undertaken, including spectroscopic observations of magnetospheric funnel flows and accretion shocks in T Tauri stars, and the observed distribution of the rotation rates in T Tauri stars. Synthetic spectra of the models will be compared to that observed for accreting black holes at the center of early type galaxies and the galactic center, while fluctuations in the mass accretion rate can be compared to X-ray variabil-ity observed by RXTE in X-ray binaries. These global calculations are the first step towards star-disk interaction models which span many decades in radius. The calculations will all be performed with a variety of 2D and 3D MHD computer codes, using large allocations of supercomputer time on massively parallel machines at the national supercomputer centers. Funding for this project was provided by the NSF program for Extragalactic Astronomy & Cosmology (AST/EXC). ***