No abstract

We have obtained high-resolution spectra of four eclipsing binary systems (FM Leo, NN Del, V963 Cen and AI Phe) with the view to gaining insight into the relative orientation of their stellar spin axes and orbital axes. The so called Rossiter-McLaughlin (RM) effect, i.e. the fact that the broadening and the amount of blue-or redshift in the spectra during an eclipse depend on the tilt of the spin axis of the background star, has the potential of reconciling observations and theoretical models if such a tilt is found. We analyse the RM effect by disentangling the spectra, removing the front component and measuring the remaining, distorted lines with a broadening function (BF) obtained from single value decomposition (SVD), weighting by the intensity centre of the BF in the eclipse. All but one of our objects show no significant misalignment, suggesting that aligned systems are dominant. We provide stellar as well as orbital parameters for our systems. With five measured spin-orbit angles we significantly increase (from 9 to 14) the number of stars for which it has been measured. The spin-orbit angle β calculated for AI Phe's secondary component shows a misalignment of 87±17 degrees. NN Del, with a large separation of components and a long dynamical timescale for circularisation and synchronisation, is an example of a close to primordial spin-orbit angle measurement.

We derive the absolute physical and orbital parameters for a sample of 18 detached eclipsing binaries from the All-Sky Automated Survey (ASAS) data base based on the available photometry and our own radial velocity (RV) measurements. The RVs are computed using spectra we collected with the 3.9-m Anglo-Australian Telescope (AAT) and its University College London Echelle Spectrograph (UCLES), and the 1.9-m Radcliffe telescope and its Grating Instrument for Radiation Analysis with a Fibre-Fed Echelle (GIRAFFE) at the South African Astronomical Observatory (SAAO). In order to obtain as precise RVs as possible, most of the systems were observed with an iodine cell available at the AAT/UCLES and/or analysed using the two-dimensional cross-correlation technique (TODCOR). The RVs were measured with TODCOR using synthetic template spectra as references. However, for two objects we used our own approach to the tomographic disentangling of the binary spectra to provide observed template spectra for the RV measurements and to improve the RV precision even more. For one of these binaries, AI Phe, we were able to the obtain an orbital solution with an RV rms of 62 and 24 m s −1 for the primary and secondary, respectively. For this system, the precision in M sin 3 i is 0.08 per cent.For the analysis, we used the photometry available in the ASAS data base. We combined the RV and light curves using PHOEBE and JKTEBOP codes to obtain the absolute physical parameters of the systems. Having precise RVs, we were able to reach ∼0.2 per cent precision (or better) in masses in several cases but in radii, due to the limited precision of the ASAS photometry, we were able to reach a precision of only 1 per cent in one case and 3-5 per cent in a few more cases. For the majority of our objects, the orbital and physical analysis is presented for the first time.