The dependence of convective core overshooting on stellar mass: reality check, and additional evidence

Abstract

Overshooting from the convective cores of stars more massive than about 1.2 M(Sun) has a profound impact on their subsequent evolution. And yet, the formulation of the overshooting mechanism in current stellar evolution models has a free parameter (f[ov] in the diffusive approximation) that remains poorly constrained by observations, affecting the determination of astrophysically important quantities such as stellar ages. In an earlier series of papers we assembled a sample of 37 well-measured detached eclipsing binaries to calibrate the dependence of f[ov] on stellar mass, showing that it increases sharply up to a mass of roughly 2 M(Sun), and remains constant thereafter out to at least 4.4 M(Sun). Recent claims have challenged the utility of eclipsing binaries for this purpose, on the basis that the uncertainties in f[ov] from the model fits are typically too large to be useful, casting doubt on a dependence of overshooting on mass. Here we reexamine those claims and show them to be too pessimistic, mainly because they did not account for all available constraints --- both observational and theoretical --- in assessing the true uncertainties. We also take the opportunity to add semi-empirical f[ov] determinations for 13 additional binaries to our previous sample, and to update the values for 9 others. All are consistent with, and strengthen our previous conclusions, supporting a dependence of f[ov] on mass that is now based on estimates for a total of 50 binary systems (100 stars).