Modelling fully convective stars in eclipsing binaries: KOI-126 and CM Draconis

TL;DR

This study models fully convective stars in eclipsing binaries KOI-126 and CM Draconis, successfully fitting KOI-126 but highlighting discrepancies in CM Draconis that may be explained by metallicity and convection efficiency variations.

Contribution

It provides the first detailed models of fully convective stars in these systems, exploring the effects of metallicity and convection parameters on stellar properties.

Findings

KOI-126 models match observations with solar-like convection parameters.

CM Draconis models require higher metallicity and reduced convection efficiency to fit observed radii.

Reduced mixing length parameter partially explains radius discrepancies in CM Draconis.

Abstract

We present models of the components of the systems KOI-126 and CM Draconis, the two eclipsing binary systems known to date to contain stars with masses low enough to have fully convective interiors. We are able to model satisfactorily the system KOI-126, finding consistent solutions for the radii and surface temperatures of all three components, using a solar-like value of the mixing-length parameter \alpha in the convection zone, and PHOENIX NextGen 1D model atmospheres for the surface boundary conditions. Depending on the chemical composition, we estimate the age of the system to be in the range 3-5 Gyr. For CM Draconis, on the other hand, we cannot reconcile our models with the observed radii and T_eff using the current metal-poor composition estimate based on kinematics. Higher metallicities lessen but do not remove the discrepancy. We then explore the effect of varying the mixing length parameter \alpha. As previously noted in the literature, a reduced \alpha can be used as a simple measure of the lower convective efficiency due to rotation and induced magnetic fields. Our models show a sensitivity to \alpha (for \alpha < 1.0) sufficient to partially account for the radius discrepancies. It is, however, impossible to reconcile the models with the observations on the basis of the effect of the reduced \alpha alone. We therefore suggest that the combined effects of high metallicity and \alpha reduction could explain the observations of CM Draconis. For example, increasing the metallicity of the system towards super-solar values (i.e. Z = 2 Z_sun) yields an agreement within 2 \sigma with \alpha = 1.0.