The good, the bad and the really ugly: composite-spectrum binaries

Composite-spectrum binaries (containing a cool giant primary and a hot dwarf secondary) should be an answer to a theoretician’s prayer. Because of their luminosity difference, both spectra are visible in the near UV, a region that includes several valuable luminosity and temperature indicators. If we just measure the radial velocity (RV) of the secondary at different dates, and construct an SB2 orbit, we immediate get the mass ratio of the component stars. A guess of the mass of the dwarf, and one thence obtains the mass of the cool giant – a unique and immensely valuable method. But just measure a hot dwarf’s RV?? It raises many problems, mostly caused by the nature of the star. A hot (late-B or early-A) dwarf has few lines, most are weak, they can be very blurred by rotation, and they get hidden by the crowded spectra of the giant. Spectral subtraction works a treat in separating the two spectra, but the residue is inevitably rather noisy. Nevertheless, results from well over half of the 45 brightest northern composite-spectrum binaries have been published; about 1 in 6 are triple systems, and a few have characteristics that defy any theoretician to explain. Several are ‘bad’ because their giant primaries have high luminosity and finding a matching standard for the subtraction process is troublesome, while the ‘really ugly’ include an Am star in a simple SB2 binary with a period of 75 years, a pair of early-A dwarfs in a 3-day orbit with amplitudes of 100 km s−1 and which show absolutely no rotational broadening at all, and another whose secondary (apparently a single star) is more that twice as massive as its primary giant. But 8 of the sample are also eclipsing, and manifest the hugely important phenomenon of chromospheric absorption – those are the ‘really good’ systems. The talk showed examples of each kind.