Astron. Astrophys. 324, 137-154 (1997)

Astron. Astrophys. 324, 137-154 (1997)

1. Introduction

Binarity is a very important parameter to consider when dealing with any kind of chemically peculiar stars (hereafter CP stars). It is well known that there is a strong deficiency of close binaries among CP2 (magnetic Ap) stars (Gerbaldi et al. 1985, North 1994). Even among CP1 (Am) stars, whose large majority belong to binary systems, there is a clear deficit of very short orbital periods compared with systems hosting normal stars. Therefore, the number of confirmed CP2 stars in eclipsing systems is vanishingly small, while such systems would be of great interest: very few Ap stars have fundamentally determined parameters; furthermore, if the Ap component turns out to be a spectroscopic variable, its detailed observation when it enters into eclipse may provide strong constraints on the distribution of the abundance anomalies on its surface, complementing the more usual Doppler imaging technique (Vincent et al. 1993).

As yet, there are only a handful of Ap stars in eclipsing binaries, and some of them may well have a spurious classification (North & Burnet 1994, North & Richard 1995). The best studied case is that of a CP3 (HgMn) star, AR Aur (Nordström & Johansen 1994; Johansen & Nordström 1995) - which is non-magnetic - followed by AO Vel, which is the unique Si-type eclipsing binary known (Clausen et al. 1995). The classification of the latter remains to be confirmed, however, even though its lightcurve provides clear evidence for an intrinsic and strictly periodic variation of at least one component. A close examination of other cases is then useful, and is beginning with this work. Am stars are more frequently found in eclipsing systems, and at least ten of them have very precisely determined parameters (Andersen 1991), but Andersen and his coworkers avoided any kind of Ap star in their monumental work on detached systems.

HD 143654 (TV Nor), HD 184035 (HR 7422) and HD 185257 (HR 7464) are three detached eclipsing binaries, which have been observed in the seven-colours Geneva photometric system but have not yet been studied in detail.

The first one, TV Nor, was studied in a preliminary way by North & Burnet (1994), who gave an estimate of its total mass from a spectrum taken at a quadrature, but did not interpret its lightcurves. Here we determine for the first time the solution of the lightcurves, and additional spectroscopic data allow us to propose a complete solution of the system. This binary had been brought to the attention of observers by Renson (1990) as one of the very few eclipsing binaries with an Ap classification. This was the main motivation for observing it.

The two other systems were discovered by Waelkens & Rufener (1983), who just gave their orbital period and V lightcurve. HD 184035 was already known as an SB1 spectroscopic binary and has published radial velocity curve and orbital elements (Buscombe & Morris 1961); however, the mass ratio cannot be directly obtained since the lines of the secondary component have not been observed. Gaspani (1984) proposed a solution of the Geneva V lightcurve of HD 184035, while Giuricin et al. (1984) did the same for HD 185257. Levato (1975) measured [FORMULA] for the primary of HD 184035. On the basis of Geneva photometry alone, Waelkens & Rufener (1983) could assign an A5m photometric spectral type to the primary of HD 185257 (Hauck & Curchod 1980 had also suspected this star to be an Am from its Geneva indices). Their determination was nicely confirmed by Houk (1982), who gives the spectral type A1mA6-F0 to this binary, the quality of the classification spectrum being 1, i.e. the best; here A1 is the type derived from the calcium lines, A6 is derived from the hydrogen lines and therefore should represent best the real effective temperature of the star, and F0 is the type derived from the metallic lines. Therefore the photometric type agrees with the spectroscopic one within one subclass. In fact, the Am peculiarity of HD 185257 was already known from the work of Andersen & Nordström (1977) and even from earlier litterature, since these authors quote an A2m type.

In this sample of three eclipsing binaries, two systems seem a priori to contain one chemically peculiar star, which enhances their astrophysical interest. We shall question below, however, the Ap nature of TV Nor. The third system has been mentioned as showing "unusually sharp lines for an A-type star" by Buscombe & Morris (1961), which may be considered as implicitly designating the primary as a CP star, if the conclusions of Abt & Morrell (1995) hold true.

Here we analyze all seven lightcurves of each system and propose in addition an orbital solution for the primary of HD 185257, which has 9 published radial velocities (Catchpole et al. 1961, Nordström & Andersen 1985).

Online publication: May 26, 1998

helpdesk.link@springer.de