Abstract
Observations have shown that supermassive black holes exist at the nuclei of almost all galaxies. Unlike stellar-mass black holes, which are formed by gravitational collapse of stars, the origin of these supermassive objects is not well understood. In this talk, I will use radiation hydrodynamic simulations to investigate whether supermassive black holes can grow purely by accretion. I will first give a review on some recent results in accretion physics. Because of our new understandings of magnetorotational instability, we can now build more realistic models that go beyond the standard Shakura-Sunyaev alpha-prescription. I will then describe how to incorporate these models into simulations in order to study super-Eddington accretion flows. Our results show that black holes can accrete at hundreds of the Eddington rates. Finally, I will show our recent effort in developing a new radiation hydrodynamic code that runs on graphics hardware. The large performance gain of this new code allows us to place limit on the growth of black holes in the early Universe.
Coffee and tea will be served 20 minutes prior to the seminar.