Can We Trust Models for Adiabatic Mass Loss?

TL;DR

This paper challenges the traditional view that giant stars with deep convective envelopes inevitably undergo catastrophic mass transfer in binaries, showing that thermal relaxation can stabilize mass loss if it occurs on appropriate timescales.

Contribution

It demonstrates that the local thermal timescale of the star's outer layer influences mass transfer stability, revising the canonical adiabatic response model for giants in binary systems.

Findings

Thermal relaxation can prevent catastrophic expansion during rapid mass loss.

Polytropic models overestimate the likelihood of instability.

Donor response depends on radius and core mass, not just core properties.

Abstract

In interacting binaries, comparison of a donor star's radial response to mass loss with the response of its Roche radius determines whether mass loss persists and, if so, determines the timescale and stability of the ensuing evolutionary phase. For giants with deep convective envelopes, the canonical description holds that once mass transfer begins it typically proceeds catastrophically on the dynamical timescale, as the star cannot lose sufficient heat in order to avoid expansion. However, we demonstrate that the local thermal timescale of the envelope's superadiabatic outer surface layer remains comparable to that of mass loss in most cases of "dynamical" mass loss. We argue therefore that if mass loss proceeds on a timescale longer than this, then even a deep convective envelope will not dramatically expand, as the surface layer will have time to relax thermally and reconstitute itself. We demonstrate that in general the polytropic approximation gives much too strict a criterion for stability, and discuss the dependence of the donor's response on its radius in addition to its core mass. In general, we find that the effective response of the donor on rapid timescales cannot be determined accurately without detailed evolutionary calculations.