Choked accretion: from radial infall to bipolar outflows by breaking spherical symmetry

TL;DR

This paper investigates how small deviations from spherical symmetry in accretion flows can lead to bipolar outflows, with a maximum accretion rate close to the Bondi solution, and excess material being ejected.

Contribution

It demonstrates through analytic models and hydrodynamical simulations that slight asymmetries cause bipolar outflows and limit the accretion rate to near the Bondi value.

Findings

Small asymmetries induce bipolar outflows.

Accretion rate is capped near the Bondi rate.

Excess material is ejected at high speeds.

Abstract

Steady state, spherically symmetric accretion flows are well understood in terms of the Bondi solution. Spherical symmetry however, is necessarily an idealized approximation to reality. Here we explore the consequences of deviations away from spherical symmetry, first through a simple analytic model to motivate the physical processes involved, and then through hydrodynamical, numerical simulations of an ideal fluid accreting onto a Newtonian gravitating object. Specifically, we consider axisymmetric, large-scale, small amplitude deviations in the density field such that the equatorial plane is over dense as compared to the polar regions. We find that the resulting polar density gradient dramatically alters the Bondi result and gives rise to steady state solutions presenting bipolar outflows. As the density contrast increases, more and more material is ejected from the system, attaining speeds larger than the local escape velocities for even modest density contrasts. Interestingly, interior to the outflow region, the flow tends locally towards the Bondi solution, with a resulting total mass accretion rate through the inner boundary $choking$ at a value very close to the corresponding Bondi one. Thus, the numerical experiments performed suggest the appearance of a maximum achievable accretion rate, with any extra material being ejected, even for very small departures from spherical symmetry.