Astron. Astrophys. 341, 499-526 (1999)

1. Introduction

The question of the origin of Gamma Ray Bursts (GRBs) - whether they are galactic or cosmological - has been controversial during the three decades since the announcement of their detection (Klebesadel et al. 1973). In spite of the fact that the Burst and Transient Source Experiment (BATSE) detected bursts at a rate of around one per day, no clear indication of a distance scale could be found. It was only recently that observations of the afterglow of Gamma Ray Bursts in the optical (van Paradijs et al. 1997; Sahu et al. 1997; Galama et al. 1997; Djorgovski et al. 1997), X-ray (Costa et al. 1997) and radio (Frail et al. 1997) part of the spectrum seem to have settled their cosmological origin. The identification of Fe and Mg absorption lines and the determination of the cosmological redshift (; Metzger et al. 1997) seem to have definitely ruled out galactic sources of GRBs. The most popular model for the central engine to power cosmological GRBs is the merger of two neutron stars (or a neutron star and a black hole; see e.g. Paczynski 1986, 1991, 1992; Eichler et al. 1989; Rees & Mesczáros 1992; Narayan et al. 1992).

This scenario is also attractive for other reasons. By now five such neutron star binary systems are known (Thorsett 1996) and no mechanism is known that could prevent the inspiral and final coalescence. During the last 15 minutes of its inspiral the binary system will emit gravitational waves with a frequency ranging from 10 Hz to 1000 Hz. The inspiral is thus one of the prime candidates to be detected by earth bound gravitational wave detector facilities such as LIGO (Abramovici et al. 1992), VIRGO (Bradaschia et al. 1990) or GEO600 (Luck 1997). Another reason to study this scenario - and this will be our main focus here - is the question of nucleosynthesis, namely the production site of r-process nuclei. Despite considerable efforts, it has up to now not been possible to identify the corresponding astrophysical event unambiguously. Most recent nucleosynthesis studies (see Freiburghaus et al. 1997, 1998; Meyer & Brown 1997) raise questions concerning the ability of high entropy neutrino wind scenarios in type II supernovae to produce r-process nuclei for in correct amounts. In addition, it remains an open question whether the entropies required for the nuclei with can actually be attained in type II supernova events. Thus, an alternative or supplementary low entropy, low r-process environment seems to be needed (decompression of neutron star material).

The decompression of neutron star matter as a possible source of r-process nuclei was first discussed by Lattimer & Schramm (1974, 1976) in their study of the tidal disruption of a neutron star by a black hole. The coalescence of two neutron stars has in this context been examined by Symbalisty & Schramm (1982) and Eichler et al. (1989). Detailed studies of the nuclear decompression process have been performed by Meyer (1989).

Hydrodynamic simulations of coalescing neutron star binaries have been performed by various groups. The first calculations were done by Nakamura and Oohara (see Shibata et al. (1993) and references therein) using polytropic equations of state and focusing mainly on the emitted gravitational waves.

Similar calculations, but using SPH, have been performed by Davies et al. (1994) who discussed a variety of physical effects related with the merging event. Zhuge et al. (1994, 1996) focussed in their work on the gravitational wave energy spectra .

In a long series of papers Lai, Rasio and Shapiro examined close binary systems. They developed an approximate analytical energy variational method and applied it to analyze the stability properties of binary systems and rotating stars (Lai et al. 1993a, 1993b, 1994a, 1994b, 1994c, Lai 1994). Two of them (Rasio and Shapiro) performed complementary SPH-simulations, where apart from stability questions the emission of gravitational waves was investigated (Rasio & Shapiro 1992, 1994, 1995).

Ruffert et al. (1996, 1997) performed PPM-simulations of neutron star mergers, using the Lattimer/Swesty EOS and accounting for neutrino emission by means of an elaborate leakage scheme. They discussed gravitational wave and neutrino emission and made an attempt to address the question of mass ejection by looking at the material they lost from their grid.

Rosswog et al. (1998a,b) discussed preliminary results of their SPH calculations concerning mass ejection and its implications for nucleosynthesis.

Fully relativistic hydrodynamical calculations are beginning to yield results (see Wilson & Mathews 1995, Wilson et al. 1996, Mathews & Wilson 1997, Baumgarte et al. 1997). However, these are still controversial (see e.g. Lai 1996, Thorne 1997, Lombardi & Rasio 1997, Wiseman 1997, Shibata et al. 1998).

Our focus in this paper will be on the matter that becomes unbound in an equal mass neutron star binary merger. In Sect. 2 we describe the numerical method, we discuss our chosen binary system parameters and the appropriateness of the approximations of our model. The results are discussed in Sect. 3. In 3.1 we describe the morphology of the mergers, in 3.2 the mass distribution is discussed, 3.3 deals with the thermodynamic properties of the merged configuration, in 3.4 we discuss our neutrino treatment and 3.5 deals with mass ejection and related questions of nucleosynthesis. The results are summarized in Sect. 4. Details of the initial SPH-particle setup, the equation of state and the neutrino emission are given in the appendices.

© European Southern Observatory (ESO) 1999

Online publication: December 4, 1998
helpdesk.link@springer.de