The final orbital separation in common envelope evolution

Abstract

In the majority of population synthesis calculations of close binary stars, the common envelope (CE) phase is modeled using a standard prescription based upon the conservation of energy, known as the alpha prescription. In this prescription, the orbital separation of the secondary and giant core at the end of the CE phase is taken to be the orbital separation when the envelope becomes unbound. However, recent observations of planetary nebulae with binary cores (BPNe), believed to be the immediate products of CE evolution, indicate orbital periods that are significantly shorter than predicted by population synthesis models using the alpha prescription. We argue that unbinding the envelope provides a necessary, but not sufficient, condition to escape a merger during CE evolution. The spiral-in of the secondary must also be halted. This requires the additional dynamical constraint that the frictional torque on the secondary be reduced to approximately zero. In this paper, we undertake a preliminary examination of the effect of adding this dynamical constraint in population synthesis calculations of BPNe. We assume that the frictional torque will be sufficiently reduced when the secondary enters a region within the giant where the mass-radius profile is flat. We crudely estimate the location of this region as a function of the core mass based upon existing stellar models of AGB stars between 1 and 7 M⊙. We calculate a theoretical orbital period distribution of BPNe using a population synthesis code that incorporates this dynamical constraint along with the alpha prescription.