Metallicity-dependent wind parameter predictions for OB stars

TL;DR

This paper updates wind parameter predictions for OB stars, emphasizing metallicity effects on mass-loss rates and wind velocities, with implications for stellar evolution and black hole formation.

Contribution

It introduces a new Monte Carlo recipe for wind parameters, including a detailed analysis of metallicity dependence and the weak-wind dip phenomenon.

Findings

Weak-wind dip identified around 35,000 K.

Terminal velocity independent of Z for B supergiants below bi-stability jump.

Mass-loss rate scales as Z^{0.85} for B supergiants and Z^{0.42} for O-type stars above the jump.

Abstract

Mass-loss rates and terminal wind velocities are key parameters that determine the kinetic wind energy and momenta of massive stars. Furthermore, accurate mass-loss rates determine the mass and rotational velocity evolution of mass stars, and their fates as neutron stars and black holes in function of metallicity (Z). Here we update our Monte Carlo mass-loss Recipe with new dynamically-consistent computations of the terminal wind velocity -- as a function of Z. These predictions are particularly timely as the HST ULLYSES project will observe ultraviolet spectra with blue-shifted P Cygni lines of hundreds of massive stars in the low-Z Large and Small Magellanic Clouds, as well as sub-SMC metallicity hosts. Around 35 000 K, we uncover a weak-wind "dip" and we present diagnostics to investigate its physics with ULLYSES and X-Shooter data. We discuss how the dip may provide important information on wind-driving physics, and how this is of key relevance towards finding a new gold-standard for OB star mass-loss rates. For B supergiants below the Fe IV to III bi-stability jump, the terminal velocity is found to be independent of Z and M, while the mass-loss rate still varies as $\dot{M} \propto Z^{0.85}$. For O-type stars above the bi-stability jump we find a terminal-velocity dependence of $v_{\infty} \propto Z^{0.19}$ and the Z-dependence of the mass-loss rate is found to be as shallow as $\dot{M} \propto Z^{0.42}$, implying that to reproduce the `heavy' black holes from LIGO/VIRGO, the `low Z' requirement becomes even more stringent than was previously anticipated.