Formation of Plasmoids in Accretion Disks

Arguably the most extensively studied (supermassive) black hole in our universe is Sgr A*, which is found at the center of our galaxy and a prime target of the Event Horizon Telescope (EHT): The culmination of decade-long quest to photograph the black hole shadow. Due to the availability of high resolution, high cadence observational data across the electromagnetic spectrum, it forms a critical benchmark for numerical simulations. One open mystery is the origin of daily/weekly flares observed in the NIR and X-Ray which originate very close to the black hole, and were not found in general relativistic magnetohydrodynamics (GRMHD) simulations. The absence of any self-consistent model or simulation producing these flares and/or hotspots illustrates that our understanding of the microphysics in accretion disks is incomplete.

Several models have been proposed to address this discrepancy, including magnetic reconnection in current sheets due to the formation of plasmoids. Plasmoids are bubbles of hot, magnetized gas formed in elongated current sheets that can locally speed up magnetic reconnection, which leads to hotspots. While plasmoids were observed in 2-dimensional GRMHD simulations it was, until this work, unclear if current sheets subject to 3-dimensional turbulence will remain sufficiently elongated and stable to form plasmoids. Namely, plasmoids can only form when the current sheet reaches a critical length to thickness (aspect) ratio. Because the thickness of a sheet is limited by the numerical resolution, numerical simulations were unable to reach the plasmoid dominated regime.

The video below shows a transverse slice of the plasma Beta (ratio of gas and magnetic pressures) of a black hole accretion disk in the largest ever GRMHD simulation (22 billion cells) performed on OLCF Summit using 7200 NVIDIA V100 GPUs. The black hole spin (a=0.9375) axis points along the positive z-axis. The extreme resolution (5400x2304x2304) on Summit allows the central current sheet to become thin enough such that it can break apart into plasmoids (e.g. the current sheet in the right hemisphere forms a chain of plasmoids), which subsequently form hotspots. Simulations at resolutions considered state of the art in the GRMHD field (560x256x512) do not allow the central current sheet to become thin enough for plasmoids to form. A better understanding of the micro-physical processes driving magnetic reconnection has the potential of revolutionizing our understanding of radiative emission processes in black hole systems.

The image below shows the dimensionless temperature T=p/ρ, plasma-β, and density ρ (from left to right) in the poloidal plane before a large flare (top row), during the flare (middle row) in the inner 10 r_g and after the flare (bottom row) in the inner 40rg. During the flare, the accretion disk is ejected and the broad accretion inflow is reduced to a thin current sheet with plasmoids, indicated by magnetic bubbles squeezed between antiparallel in-plane field lines (in green, left panel). The hot exhaust of the reconnection layer in the current sheet heats the jet sheath. Reconnection transforms the horizontal field in the current sheet to vertical field that is ejected in the form of hot coherent flux tubes (bottom-left panel) at low β and density (bottom-right and middle).

Formation of Poloidal Magnetic Fields in Accretion Disks

There exists a scientific consensus that large scale poloidal magnetic field loops are necessary to launch astrophysical jets. Understanding the formation of such jets is crucial for understanding structure formation in the universe. However, the origin of these poloidal field loops is a mystery and previous simulations that did not include poloidal magnetic fields failed to produce any jets.

In my recent work I, however, showed that an accretion disk threaded with a toroidal magnetic field is able to generate large scale poloidal magnetic flux after a relatively short time of 20.000 rg/c (black hole time units). When this poloidal magnetic flux accretes on the black hole, a powerful astrophysical jet is launched along the poles (blue regions). This is the first demonstration of a large scale poloidal magnetic flux dynamo in (GR)MHD. The colors in the movie below indicate the logarithm of the rest mass density, while the black lines show the generated poloidal flux (solid lines show positive flux, dashed lines show negative flux). The axes are shown in units of black hole radii (rg). The left panel is a zoomed in version of the right panel.

Magnetic Prandtl Number Study

Future work in the H-AMR collaboration will address the effects of non-unity resistivity and viscosity coefficients. Namely, practically all GRMHD simulations to date rely on numerical dissipation to account for viscous and resistive effects in the plasma. This is done by imposing certain boundary conditions in the MHD equations which themselves do not include explicit viscosity and resistivity terms. This causes the ratio of the viscosity and resistivity terms to become of order unity. However, the true value of this so called magnetic Prandtl number is unknown while it has been suggested that a non-unity magnetic Prandtl number can significantly change the strength of accretion disk turbulence. Our collaboration has implemented a resistivity scheme (Ripperda et al 2019) into H-AMR and is actively working to also implement a viscosity scheme. Future 3-dimensional simulations combining explicit resistivity and viscosity terms will thus provide a sanity check on practically all previous GRMHD simulations performed inside and outside our collaboration. Below you see a snapshot of the plasma beta from a simulation with an explicit resistivity term in H-AMR.

My Publications

1) Large-Scale Poloidal Magnetic Field Dynamo Leads to Powerful Jets in GRMHD Simulations of Black Hole Accretion with Toroidal Field

2) Nozzle Shocks, Disk Tearing and Streamers Drive Rapid Accretion in 3D GRMHD Simulations of Warped Thin Disks

3) Radiation Transport Two-Temperature GRMHD Simulations of Warped Accretion Disks

4) Millimeter Observational Signatures of Flares in Magnetically Arrested Black Hole Accretion Models