Simulations of common envelope interactions with a late phase asymptotic giant branch donor star

<p dir="ltr">The common envelope interaction is an evolutionary phase of binary stellar systems in which one of the components overflows its Roche lobe via radial expansion and/or orbital shrinkage, which induces mass transfer from its envelope onto the companion. Typically, mass transfer is unstable and leads to the companion being engulfed by the envelope of the donor star, with the subsequent drastic and dynamic reduction of the orbit by gravitational friction. If the envelope is ejected, the stellar cores may emerge as a compact binary. If the envelope is not ejected, a merger ensues. Common envelope interactions have grown in interest in the past few years as they are channels for other astrophysical processes such as gravitational wave sources caused by merging neutron stars and back holes, cataclysmic variables or planetary nebulae. At least one in five of all planetary nebulae are the product of a common envelope interaction; moreover for low mass envelope donor stars ( 1-4 M_⊙) they are the most likely explanation for asymmetrical planetary nebulae. </p><p dir="ltr">In this work, we perform a set of common envelope simulations with the smoothed particle hydrodynamics code phantom. We first created a 2 M_⊙ star using the stellar evolution grid code mesa and evolved it until it reached the thermally pulsating asymptotic giant branch phase. We then mapped this stellar profile into the computational domain of phantom and stabilized the resulting isolated star using a series of numerical relaxation techniques. We set up this model alongside a point mass companion to simulate the binary interaction during one of the thermal pulses. We carried out a set of simulations with three different equations of state in order to understand the contribution of different components of the internal energy of the envelope. We found that all simulations led to a common envelope interaction before the end of the thermal pulse. We also discover that full unbind of the envelope may be possible using recombination energy at the cost of larger final orbital separations. </p><p dir="ltr">Furthermore, we have extended these simulations to study the effect of the mass ratio on the common envelope outcome, as well as to determine the influence of resolution and the point masses softening lengths (both numerical features) on the results. Simulation outcomes indicate that the unbinding efficiency and morphology is affected by mass ratio. We also described a stage in the CE simulations in which strong polar outflows of gas unbind the envelope at later times. </p><p dir="ltr">Finally, we implemented dust-driven acceleration in these simulations using a robust, first approximation for the dust opacity sensitive to the gas temperature. We found that the main contribution of dust-driven acceleration is during the early stages of the common envelope and in the post-common envelope system, especially in simulations with an ideal gas equation of state. Conversely, it is clear that if simulations with recombination energy are more reliable, then dust driving is not a major contributor to envelope unbinding. These simulations created the starting point for additional work on dust nucleation in common envelope interactions. These simulations rise further questions related to radiation transport in the envelope, which are being tackled next. </p>