Investigating lack of accretion disk eccentricity growth in a global 3D MHD simulation of a superhump system

TL;DR

This study uses 3D MHD simulations to investigate why accretion disk eccentricity growth is absent in a superhump system, finding magnetic forces damp eccentricity unlike in 2D alpha models where tidal forces promote growth.

Contribution

It demonstrates that magnetic forces in 3D MHD simulations suppress eccentricity growth, contrasting with 2D alpha models where viscosity and tidal forces promote it.

Findings

3D MHD simulation shows no significant eccentricity growth.

In 2D alpha simulations, high alpha values lead to eccentricity growth.

Magnetic forces damp eccentricity in the 3D MHD simulation.

Abstract

We present the results of a 3D global magnetohydrodynamic (MHD) simulation of an AM CVn system that was aimed at exploring eccentricity growth in the accretion disc self-consistently from a first principles treatment of the MHD turbulence. No significant eccentricity growth occurs in the simulation. In order to investigate the reasons why, we ran 2D alpha disc simulations with alpha values of 0.01, 0.1, and 0.2, and found that only the latter two exhibit significant eccentricity growth. We present an equation expressing global eccentricity evolution in terms of contributing forces and use it to analyze the simulations. As expected, we find that the dominant term contributing to the growth of eccentricity is the tidal gravity of the companion star. In the 2D simulations, the alpha viscosity directly contributes to eccentricity growth. In contrast, the overall magnetic forces in the 3D simulation damp eccentricity. We also analyzed the mode-coupling mechanism of Lubow, and confirmed that the spiral wave excited by the 3:1 resonance was the dominant contributor to eccentricity growth in the 2D $\alpha=0.1$ simulations, but other waves also contribute significantly. We found that the $\alpha=0.1$ and 0.2 simulations had more relative mass at larger radii compared to the $\alpha=0.01$ and 3D MHD simulation, which also had an effective $\alpha$ of 0.01. This suggests that in 3D MHD simulations without sufficient poloidal magnetic flux, MRI turbulence does not saturate at a high enough $\alpha$ to spread the disc to large enough radii to reproduce the superhumps observed in real systems.