Forum Springer Astron. Astrophys.
Forum Whats New Search Orders

Astron. Astrophys. 319, 481-486 (1997)

Previous Section Next Section Title Page Table of Contents

1. Introduction

Open clusters present a large variety of spectroscopic binaries among their red giant members, with periods covering a wide interval from 40 to more than 7000 days (the longest orbital period we have determined so far) and a wide range of secondary masses. Several composite red giants with a marked effect on their colours due to the presence of a late B- or A-type star on the upper main sequence (Mermilliod & Mayor 1989, Paper I) have been found, but in many systems the secondary is not visible, neither photometrically nor spectroscopically. In such cases we suppose that the secondary star is either a main sequence star at least five magnitudes fainter than the primary, or a white dwarf. From the present data it is not possible to estimate the ratio of main sequence- to white dwarf secondaries, but the fact that the observed orbital eccentricities are generally normal (e [FORMULA] 0.15), in contrast to the case of barium stars which show predominantly small eccentricities (Jorissen & Boffin 1992), suggests that the fraction of white dwarf secondaries is small. The abundance of red giant binaries in open clusters contrasts with the paucity of spectroscopic binaries discovered so far among red giants in globular clusters.

Analysis of the distribution of orbital elements for the first sample of 100 orbits (Mermilliod 1996) has shown that the shortest period at any primary mass is determined by the Roche limit at the red giant tip. A star evolving along the ascending red giant branch reaches its largest radius at the red giant tip. If it fills its Roche lobe, mass will be transferred to the secondary and its evolution is interrupted. Thus, the red giant binary will survive beyond the red giant tip phase only if the orbital period is longer than a critical period defined by the condition that the size of the Roche lobe is equal to the maximum radius of the red giant. Computations of these critical periods for various masses from 1.5 to 6 [FORMULA] show good agreement with the shortest periods observed (Mermilliod & Mayor 1996). The shortest period determined by us so far (41.5 days for NGC 2477 #1025, unpublished orbit) is in good agreement with the mass ([FORMULA] 2 [FORMULA]) and maximum radius of this star as deduced from the evolutionary tracks of Schaller et al. (1992). In this mass domain, stars with about 2 [FORMULA] show the shortest ascending path before the onset of helium burning.

The period-eccentricity diagram for the red giants shows a well-marked transition from circular to eccentric orbits (Mermilliod & Mayor 1992). A mass-period diagram plotted for the same sample of 100 orbits also shows clearly that the cut-off period for circular orbits is a function of the primary mass and increases with mass for stars more massive than about 2 [FORMULA] (Mermilliod & Mayor 1996).

Raboud & Mermilliod (1994) have looked for effects of mass segregation in a study of 88 binary red giants, which were compared with 260 single red giants in 14 clusters of different ages. No statistically significant effect has been found, except for the well known case of M67 (Mathieu & Latham 1986). The explanation may either lie in incomplete survey of the cluster members in terms of radial distance from the cluster centre, as shown by simulations, or result from an age effect, most clusters of the sample being too young. Or the clusters may have lost their most massive short period central binaries if their periods were shorter than the Roche critical period mentioned above.

Orbital elements for 43 spectroscopic binaries in clusters have been published in earlier papers in this series. The present paper presents five additional orbits for red giant spectroscopic binaries in three clusters which have already been analysed in previous papers: NGC 5822 (Mermilliod et al. 1990), NGC 6134 (Claria & Mermilliod 1992) and NGC 6664 (Mermilliod et al. 1987), and five orbits for new binaries in NGC 2489, 2567, 3033 and IC 2488.

As in Paper II (Mermilliod et al. 1989), to which the reader is referred for a more detailed scientific justification of this programme, we shall discuss here only the binary stars. Open cluster results not presented in previous papers will be discussed later in this series.

Previous Section Next Section Title Page Table of Contents

© European Southern Observatory (ESO) 1997

Online publication: July 3, 1998