Gone with the wind: the impact of wind mass transfer on the orbital evolution of AGB binary systems

In low-mass binary systems, mass transfer is likely to occur via a slow and dense stellar wind when one of the stars is in the asymptotic giant branch (AGB) phase. Observations show that many binaries that have undergone AGB mass transfer have orbital periods of 1-10 yr, at odds with the predictions of binary population synthesis models. In this paper we investigate the mass-accretion efficiency and angular-momentum loss via wind mass transfer in AGB binary systems and we use these quantities to predict the evolution of the orbit. To do so, we perform 3D hydrodynamical simulations of the stellar wind lost by an AGB star in the time-dependent gravitational potential of a binary system, using the AMUSE framework. We approximate the thermal evolution of the gas by imposing a simple effective cooling balance and we vary the orbital separation and the velocity of the stellar wind. We find that for wind velocities higher than the relative orbital velocity of the system the flow is described by the Bondi-Hoyle-Lyttleton approximation and the angular-momentum loss is modest, which leads to an expansion of the orbit. On the other hand, for low wind velocities an accretion disk is formed around the companion and the accretion efficiency as well as the angular-momentum loss are enhanced, implying that the orbit will shrink. We find that the transfer of angular momentum from the binary orbit to the outflowing gas occurs within a few orbital separations from the centre of mass of the binary. Our results suggest that the orbital evolution of AGB binaries can be predicted as a function of the ratio of the terminal wind velocity to the relative orbital velocity of the system, v(infinity)/v(orb). Our results can provide insight into the puzzling orbital periods of post-AGB binaries. The results also suggest that the number of stars entering into the common-envelope phase will increase, which can have significant implications for the expected formation rates of the end products of low-mass binary evolution, such as cataclysmic binaries, type Ia supernovae, and double white-dwarf mergers.