Feeding supermassive black holes through supersonic turbulence and ballistic accretion

It has long been recognized that the main obstacle to the accretion of gas on to supermassive black holes (SMBHs) is a large specific angular momentum. It is feared that the gas settles in a large-scale disc, and that accretion would then proceed too inefficiently to explain the masses of the observed SMBHs. Here we point out that, while the mean angular momentum in the bulge is very likely to be large, the deviations from the mean can also be significant. Indeed, cosmological simulations show that velocity and angular momentum fields of gas flows on to galaxies are very complex. Furthermore, inside bulges the gas velocity distribution can be further randomized by the velocity kicks due to feedback from star formation. We perform hydrodynamical simulations of gaseous rotating shells infalling on to an SMBH, attempting to quantify the importance of velocity dispersion in the gas at relatively large distances from the black hole. We implement this dispersion by means of a supersonic turbulent velocity spectrum. We find that, while in the purely rotating case the circularization process leads to efficient mixing of gases with different angular momenta, resulting in a low accretion rate, the inclusion of turbulence increases this accretion rate by up to several orders of magnitude. We show that this can be understood based on the notion of 'ballistic' accretion, whereby dense filaments, created by convergent turbulent flows, travel through the ambient gas largely unaffected by hydrodynamical drag. This prevents the efficient gas mixing that was found in the simulations without turbulence, and allows a fraction of gas to impact the innermost boundary of the simulations directly. Using the ballistic approximation, we derive a simple analytical formula that captures the numerical results to within a factor of a few. Rescaling our results to astrophysical bulges, we argue that this 'ballistic' mode of accretion could provide the SMBHs with sufficient fuel without the need to channel the gas via large-scale discs or bars. We therefore argue that star formation in bulges can be a strong catalyst for SMBH accretion.