next previous
Up: Search for reference A0 parallaxes


9 Conclusion

Our aim was to define a set of "reference" dwarf A0 stars, having accurate parallaxes, in order to be combined with Vega in the role of "standard" stars.

The selection of non binary and non peculiar objects has been obtained from a selected sample of dwarf A0 extracted from Hipparcos Input Catalogue and observed spectroscopically. By combining photometric and spectroscopic informations with astrometric data and stellar evolutionary models, the original sample had to be reduced by 35%.

The inclusion of a Balmer line in the analysis of stellar spectra of A-type stars is not trivial, owing the small size of the CCD detectors, but the informations that can be extracted are often underestimated. In a binary system, the contamination detected on the combined spectrum, when the two components belong to similar spectral types and have a different radial velocity, is expected first on Balmer lines since these are the strongest features over a large temperature interval. The second key line is the Mg II doublet at 4481 Å; its intensity is barely dependent on temperature and gravity for stars from late B to early A stars. A broadening of this line different from that expected by rotation can be detected with the resolution $R=28\,000$ in the spectra of stars with $v\,\sin\,i$ higher than 15 $\mathrm {km\, s^{-1}}$.

In the original sample of non evolved A-type stars, we have detected a fairly large number of new binary stars, as well as suspected ones (Sects. 5 and 6); we also have detected peculiar objects such as a shell star, hot Am stars and a misclassified object.

The masses and luminosities derived for the reduced sample of stars which can be considered as "normal" has allowed to fill up the gap of the classical $ \cal M$-L relation obtained from eclipsing binaries and to point out the important role of accurate logg values to define this relation (Sect. 8). The agreement found between the evolutionary based surface gravity log$g_{\rm ev}$ and the value of log$g_{\rm ph}$ obtained from photometric data does not confirm the discrepancies found by Balona (1994) at least in our gravity range.

Independently of logg, the masses can be derived directly from the $L-T_{\mathrm{eff}}$ HR diagram, but in this case the reliability of the evolutionary tracks becomes crucial.

In conclusion the two methods remain valid and complementary tools to determine the stellar masses.

Acknowledgements


The implementation of the Kurucz programs and the friendly help in using them by F. Castelli at Trieste has been highly appreciated. We would also thank P. Bonifacio for helpful suggestions and advices. We acknowledge A. Sellier for her technical assistance. Use was made of the SIMBAD data base, operated at the CDS, Strasbourg, France. Grants from MURST 40% and 60% are acknowledged.



next previous
Up: Search for reference A0 parallaxes

Copyright The European Southern Observatory (ESO)