Forum Springer Astron. Astrophys.
Forum Whats New Search Orders

Astron. Astrophys. 322, 533-544 (1997)

Previous Section Next Section Title Page Table of Contents

1. Introduction

The excellent agreement of the observed rate of decrease of the orbital period of PSR 1913+16 system with the theoretical rate predicted by general relativity (Taylor & Weisberg 1989) is considered today as evidence, although indirect, for the existence of gravitational waves. Therefore, the merger of a NS-NS binary appears to be not only one of the most promising sources for the planned gravitational wave detectors Ligo/Virgo, but above all the most important one due to the quality of information which could be extracted from a direct detection (cf. Schutz 1986, Chernoff & Finn 1993).

Much uncertainty remains about the merger rate of NS-NS systems. The first observational estimates (Narayan et al. 1991; Phinney 1991) reported a Galactic disk merger rate [FORMULA] of [FORMULA] yr-1. To date, no further close NS-NS systems have been found by the latest sensitive millisecond pulsar surveys. Using these results and a more recent distance model for pulsars, Curran & Lorimer (1995) estimated a merger rate of [FORMULA] yr-1. However, taking account of the beaming factor and of pulsars with low radio luminosity, they suggested a more realistic value of [FORMULA] yr-1. In addition, Van den Heuvel & Lorimer (1996) revised the lifetimes [FORMULA] of observed NS-NS systems and arrived at a Galactic merger rate of [FORMULA] yr-1. Numerical simulations based on evolutionary scenarios in which neutron stars do not acquire a kick velocity at birth (Tutukov & Yungelson 1993) suggest a larger rate, of up to [FORMULA] yr-1 (see also Clark et al. 1979), which departs from the Bailes' upper limit of [FORMULA] yr-1 for the NS-NS birthrate (Bailes 1996). On the other hand, Monte-Carlo simulations performed with a high kick velocity and with a different treatment of mass and angular momentum losses lead to a merger rate of [FORMULA] yr-1 (Portegies Zwart & Spreeuw 1996), more consistent with observational estimates.

The high sensitivity of the formation rate of NS-NS systems to the late stages of evolution is mainly responsible of the uncertainty in theoretical merger rates. Unfortunately, very few observations of binaries that have evolved beyond the high mass X-ray binary (HMXB) phase are available to constrain evolutionary models for massive close binaries. The only known systems are SS 433, a peculiar binary in which a massive evolved donor transfers matter to the X-ray source at a rate much higher than the Eddington limit (Margon 1984), Cyg X-3 a short-period system composed of a WR star and a compact object (Van Kerkwijk et al. 1992) and the four NS-NS systems in the disk of the Galaxy. It is worth noting that the nature of the compact object (NS/BH) is still unknown in SS 433 and in Cyg X-3. The aim of this paper is to rediscuss the physical processes used to describe the late stages of evolution (Sect. 2) and to constrain these models from the observed parameters of NS-NS systems (Sect. 3). In Sect. 4, we shall discuss the particular case of binary pulsars with a wide orbit for which Van den Heuvel et al. (1994) have proposed an alternative model of formation.

Previous Section Next Section Title Page Table of Contents

© European Southern Observatory (ESO) 1997

Online publication: June 5, 1998