Common Envelope: The Dynamical Aspects

Abstract

Interactions between stars are crucial for diverse areas of astrophysics, but they are not yet fully understood. These interactions come in forms as diverse as a direct physical collision, matter transfer from one star to another, and a "common envelope" (CE). The CE phase is a short period of time when the two stellar cores share their outer layers. During a CE phase, the rotating binary shrinks its orbit by transferring the orbital angular momentum and mechanical energy of the stars to the envelope. Depending on the amount of energy deposited in the envelope, the CE phase will lead to either ejection of the envelope and formation of a tight binary or, by merging the binary to form a single, rapidly-rotating star. During a CE phase, various physical processes take place, each operating on a different time-scale. In this work I consider events occurring on the dynamical time scale of the system and employ smoothed particle hydrodynamics (SPH) to simulate the interactions. This class of numerical methods is a natural choice for simulating interacting binaries without imposing boundary conditions. Compared to grid based codes, SPH also has the advantage of conserving energy and angular momentum of the system during the CE. Previously, it was expected that during a CE event, the binary will not strongly boost its luminosity compared to its pre-CE state. Combined with the short duration of the event, and the expected CE rates, it was thought that a CE event would not likely ever be observed. However, the recent discovery of a new type of object, the so called Luminous Red Novae (LRNe), gave us an insight that the CE events have already been observed. We proposed that instead of just boosted surface luminosity, a CE event will be accompanied by an outburst powered by a recombination wave in the ejected material, similar to Type II supernovae. The luminosity and the duration of this CE outburst depend on the mass and velocity of the ejected material. This model allows us to match all observed characteristics of LRN light curves. We tested this idea by modelling in detail the recent transient event, V1309 Sco, famous for being identified as a binary merger. We continued a detailed study of the outburst of V1309 Sco, reproducing its merger while aiming to constrain the initial conditions. In particular, we attempted to find out whether the companion star was a low-mass main sequence star or a low-mass white dwarf, and whether the subgiant was synchronized with the orbit. We found that the observed orbital period decay of V1309 Sco could be reproduced by a synchronized subgiant donor with a low-mass main sequence star companion. The formation of a double white dwarf (DWD) binary system is widely accepted to require two mass exchanges, the latter of which has to be a CE event between a low-mass red giant and the first-formed white dwarf. Low-mass red giants have a well-defined relationship between their core masses and radii, which provides the tightest theoretical constraint on the orbital separations prior to a CE event. Not surprisingly, there have been several attempts to model the latter mass exchange with hydrodynamic methods. However, all these attempts have failed to fully eject the envelope of the red giant . Although it has been acknowledged in the past that recombination energy of atoms could help CE ejection, it was unclear whether recombination takes place while this energy can still be used to drive the ejection. We incorporated in our SPH code the equation of state that includes recombination, and found that the red giant's envelope is fully ejected when we take into account recombination. We applied our method to model the system WD 1101+364. We found that for DWD systems, synchronization at the start of the CE event does not play a role. We also discussed the modified energy budget formalism.