A He3 driven instability near the fully convective boundary

TL;DR

This paper discovers a new He3-driven instability in low-mass stars near 0.35 solar masses, causing periodic convective episodes that affect stellar properties and may explain observed phenomena in binary systems.

Contribution

It introduces the convective kissing instability caused by non-equilibrium He3 burning in stars just above the fully convective threshold, a phenomenon not previously described.

Findings

The instability causes radius and luminosity variations of a few percent.

It occurs over Myr to Gyr timescales during stellar evolution.

Potential links to observed features in cataclysmic variable systems.

Abstract

We report on the discovery of an instability in low mass stars just above the threshold ($\sim 0.35 \textrm{M}_{\odot}$) where they are expected to be fully convective on the main sequence. Non-equilibrium He3 burning creates a convective core, which is separated from a deep convective envelope by a small radiative zone. The steady increase in central He3 causes the core to grow until it touches the surface convection zone, which triggers fully convective episodes in what we call the "convective kissing instability". These episodes lower the central abundance and cause the star to return to a state in which is has a separate convective core and envelope. These periodic events eventually cease when the He3 abundance throughout the star is sufficiently high that the star is fully convective, and remains so for the rest of its main sequence lifetime. The episodes correspond to few percent changes in radius and luminosity, over Myr to Gyr timescales. We discuss the physics of the instability, as well as prospects for detecting its signatures in open clusters and wide binaries. Secondary stars in cataclysmic variables (CVs) will pass through this mass range, and this instability could be related to the observed paucity of such systems for periods between two and three hours. We demonstrate that the instability can be generated for CV secondaries with mass-loss rates of interest for such systems, and discuss potential implications.