Simulations of radiative turbulent mixing layers

TLDR: The authors performed 3D magnetohydrohynamics simulations with non-equilibrium (NEI) and photoionization modeling, with an eye towards testing simple analytic models, and found that column densities are not independent of density or metallicity as analytic scalings predict, and show surprisingly weak dependence on shear velocity and density contrast.

Abstract: Radiative turbulent mixing layers should be ubiquitous in multi-phase gas with shear flow. They are a potentially attractive explanation for the high ions such as O VI seen in high-velocity clouds and the circumgalactic medium (CGM) of galaxies. We perform 3D magnetohydrohynamics (MHD) simulations with non-equilibrium (NEI) and photoionization modelling, with an eye towards testing simple analytic models. Even purely hydrodynamic collisional ionization equilibrium (CIE) calculations have column densities much lower than observations. Characteristic inflow and turbulent velocities are much less than the shear velocity, and the layer width h∝t^(1/2)_(cool) rather than h ∝ t_(coo)l. Column densities are not independent of density or metallicity as analytic scalings predict, and show surprisingly weak dependence on shear velocity and density contrast. Radiative cooling, rather than Kelvin–Helmholtz instability, appears paramount in determining the saturated state. Low pressure due to fast cooling both seeds turbulence and sets the entrainment rate of hot gas, whose enthalpy flux, along with turbulent dissipation, energizes the layer. Regardless of initial geometry, magnetic fields are amplified and stabilize the mixing layer via magnetic tension, producing almost laminar flow and depressing column densities. NEI effects can boost column densities by factors of a few. Suppression of cooling by NEI or photoionization can, in principle, also increase O VI column densities, but, in practice, is unimportant for CGM conditions. To explain observations, sightlines must pierce hundreds or thousands of mixing layers, which may be plausible if the CGM exists as a ‘fog’ of tiny cloudlets.