Outflow from super-Eddington flow: where it originates from and how much impact it gives?

It is widely believed that super-Eddington accretion flow can produce powerful outflow, but where it originates from and how much mass and energy are carried away to which directions? To answer to these questions, we newly perform a large-box, two-dimensional radiation hydrodynamic simulation, paying special attention lest the results should depend on adopted initial and boundary conditions. We could achieve a quasi-steady state in an unprecedentedly large range, $r=2~r_{\rm S}$-$600~r_{\rm S}$ (with $r_{\rm S}$ being the Schwarzschild radius) from the black hole. The accretion rate onto the central $10 ~M_{\odot}$ black hole is $\dot{M}_{\rm BH} \sim 180 ~L_{\rm Edd}/c^{2}$, whereas the mass outflow rate is ${\dot M}_{\rm outflow} \sim 24 ~L_{\rm Edd}/c^2$ (where $L_{\rm Edd}$ and $c$ are the Eddington luminosity and the speed of light, respectively). The ratio (${\dot M}_{\rm outflow}/{\dot M}_{\rm BH} \sim 0.14$) is much less than those reported previously. By careful inspection we find that most of outflowing gas which reach the outer boundary originates from the region at $R\lesssim140~r_{\rm S}$, while gas at $140~r_{\rm S}$-$230 ~r_{\rm S}$ forms failed outflow. Therefore, significant outflow occurs inside the trapping radius $\sim 450 ~r_{\rm S}$. The mechanical energy flux (or mass flux) reaches its maximum in the direction of $\sim 15^\circ$ ($\sim 80^\circ$) from the rotation axis. The total mechanical luminosity is $L_{\rm mec}\sim 0.16~L_{\rm Edd}$, while the isotropic X-ray luminosity varies from $L_{\rm X}^{\rm ISO}\sim 2.9~L_{\rm Edd}$, (for a face-on observer) to $\sim 2.1~L_{\rm Edd}$ (for a nearly edge-on observer). The power ratio is $L_{\rm mec}/L_{\rm X}^{\rm ISO}\sim 0.05$-$0.08$, in good agreement with the observations of Ultra-Luminous X-ray sources surrounded by optical nebulae.