Horizontal shear instabilities at low Prandtl number

TL;DR

This paper investigates horizontal shear instabilities in stars through direct numerical simulations, identifying different dynamical regimes and providing scaling laws to better understand stellar mixing processes, especially in the solar tachocline.

Contribution

It introduces a novel numerical approach to quantify stellar mixing by horizontal shear instabilities and proposes new scaling laws for turbulent regimes based on stratification and thermal diffusion.

Findings

Identified distinct dynamical regimes depending on stratification and thermal diffusion.

Predicted substantial vertical mixing in the solar tachocline.

Found that current models like Spiegel & Zahn are incompatible with new turbulence predictions.

Abstract

Turbulent mixing in the radiative regions of stars is usually either ignored or crudely accounted for in most stellar evolution models. However, there is growing theoretical and observational evidence that such mixing is present and can affect various aspects of a star's life. In this work, we present a first attempt at quantifying mixing by horizontal shear instabilities in stars using Direct Numerical Simulations. The shear is driven by a body force, and rapidly becomes unstable. At saturation, we find that several distinct dynamical regimes exist, depending on the relative importance of stratification and thermal diffusion (viscosity can in principle also matter, but is usually negligible in most stellar applications). In each of the regimes identified, we put forward a certain number of theoretically motivated scaling laws for the turbulent vertical eddy scale, the typical turbulent diffusion coefficient, and the typical amplitude of temperature fluctuations (among other quantities). Based on our findings, we predict that the majority of stars should fall into one of two categories: high P\'eclet number stratified turbulence, and low P\'eclet number stratified turbulence. The latter is presented in detail in a related paper by Cope et al. (2020), while the former is discussed here. Applying our results to the best-known stellar shear layer, namely the solar tachocline, we find that it should lie in the high P\'eclet number stratified turbulence regime, and predict a substantial amount of vertical mixing for temperature, momentum and composition. Taken as is, the new turbulence model predictions are incompatible with the Spiegel & Zahn (1992) model of the solar tachocline. However, we also show that rotation and magnetic fields are likely to affect the turbulence, and need to be taken into account in future studies.