Three routes to jet collimation by the Balbus-Hawley magnetorotational instability

Abstract

Three completely different lines of work have recently led to the conclusion that the magnetorotational instability (MRI) may create a hoop-stress that collimates jets. One argument is based upon consideration that magnetohydrodynamic turbulence, in general, and turbulence driven by the MRI, in particular, is more nearly viscoelastic than it is viscous. Another argument is based upon the dispersion relation for the MRI in the context of 1D simulations of core collapse. Yet a third argument rests in the results of direct numerical magnetohydrodynamic simulations of collapsars and thick accretion flows. I elaborate on my previous work regarding the first argument above and I briefly discuss how these three sets of results are all related. I also discuss the different roles played by the magnetic tension and the magnetic pressure within the context of this work. I point out that this leads to consideration of the normal stress difference between the hoop stress and the radial stress, in preference to a focus on just the hoop stress itself. Additionally, I argue that simulations of thick accretion flows and collapsars are not self-consistent if they include a phenomenological model for an MRI-induced viscous stress but disregard these other MRI-induced stress components. I comment briefly on the RHESSI observation of polarization in the gamma-ray burst GRB0212206. I argue that this polarization is consistent with a tangled field, and does not require a large-scale organized field. Finally I suggest that the role of magnetic fields in creating jets, as described here, should be understood not to work within the confines of magnetocentrifugal models of jets, but rather as an alternative to them.