4. Stellar parameters
4.1. estimates based on photometry
Strömgren and Geneva photometry (if available) has been used to derive effective temperatures and surface gravities for our program stars. The calibrations provided by Castelli (1991), Napiwotzki (1992) and Kunzli et al. (1997) were applied. The photometric data for NGC 6475, NGC 6633 and IC 2602 were taken from the SIMBAD data base. For NGC 3496 the situation with the photometric data appeared to be more complicated. Fortunately, Balona & Laney (1995) have carried out CCD Strömgren photometry of NGC 3496. From their finding chart we have identified our program stars S 43 and S 115 as SAAO 175 and SAAO 30, respectively. It should be noted that Balona & Laney (1995) found that E varies within the cluster field from 0.25 to 0.55. Therefore we did not use the mean reddening value for NGC 3496 (E = 0.39), but determined individual reddenings for each program star. For S 43 we have obtained E = 0.48, while for S 115, E = 0.33. Unfortunately, S 117, which is situated in the direct vicinity of S115 (see finding chart from Sher 1965), appeared to be out of the region observed by Balona & Laney. Thus, for S 117 we have adopted the same reddening as for S 115 and used it to make a rough estimate of the effective temperature of S 117 from the calibration of Castelli (1999), and we obtained =13000 K and =11000 K. All determinations are listed in Table 2.
Table 2. Results of the determination for program stars.
4.2. UV spectra
As a check on the temperature for some stars (HD 170054, HD 162374 and HD 93194), we used the large-aperture IUE spectra provided by SIMBAD. The original spectra in the short and long wavelength regions for HD 170054 (images SWP21337 and LWR02120, exposed on 23 October 1983), for HD 162374 (images SWP14048 and LWR10696 exposed on 24 May 1981 and SWP14085 and LWR10722 on 26 May 1981), and for HD 93194 (images SWP290891 and LWP89701, exposed on 30 August 1986) were combined and then dereddened using the Seaton (1979) law for selective interstellar extinction in the UV. The dereddened UV spectrum for each star has then been converted into emergent fluxes using the stellar angular diameter. Assuming that it is not dependent upon wavelength, the latter was found for the effective wavelength of the V band using the absolute flux calibration given by Heber et al. (1984) and the corresponding theoretical flux calculated by means of the SYNSPEC code. All of these calculations were performed using the photometric values V = 8.20 and E = 0.124 (HD 170054), V = 5.90 and E = 0.058 (HD 162374), V = 4.83 and E=0.02 (HD 93194).
The synthetic spectra in the region 1000 Å - 3500 Å are smoothed to a resolution comparable to the observed one (about 6 Å), and compared with the observations for blue straggler stars in Figs. 1-2. As one can see, in each case the temperature estimated by photometry gives good agreement.
4.3. Surface gravity determination
Gravity values were estimated from the photometric data (using the same calibrations as for the determination) and also independently by interpolating in a two-dimensional - grid of theoretical H and H line profiles based on Kurucz's model atmospheres. The results are given in Table 3, and individual fits for some of the stars are presented in Figs. 3-6. Note that in these cases the typical error of the gravity determination is about 0.1-0.2 dex.
Our finally adopted atmospheric parameters for the program stars, those which were used for atmosphere model interpolation in the Kurucz's (1992) grid, are collected in Table 4. The adopted gravities are weighted averages (weight 3 has been given for gravities based on the hydrogen profiles, and 1 for photometric estimates).
Table 4. Adopted parameters for our program stars.
4.4. Microturbulence and projected rotational velocities
Note that for all stars (except CMa) we adopted a microturbulent velocity of = 3 km s-1, a value that is most appropriate for late B - early A stars. For CMa, the number of lines in the spectrum is large enough to determine this value using the iron lines. We have obtained 2.5 km s-1 by requiring that there should be no dependence of the iron abundance on equivalent width. Our value is 0.5 km s-1 higher than that determined by Sadakane & Ueta (1989).
Projected rotational velocities for our program stars were derived by matching observed and calculated profiles for specified spectral lines (for fast rotators it was the 4481 Å line only). These results are also given in Table 4. Note that our determination for HD 162374 and HD 162586 are in the complete accordance with Abt's (1975) measurements (he gives for both stars approximately 40 km s-1 or slightly less). Our rotational velocity of HD 170054 agrees fairly well with that obtained by Andersen & Nordström (1983), = 40 km s-1, but their value is marked as being uncertain.
Levato (1975) determined the rotational velocities for some stars from IC 2602. In particular, his results on HD 92837 (220 km s-1), HD 93194 (310 km s-1) and HD 93540 (305 km s-1) generally agree with our values. The only exception is HD 92385. For this star Levato gives = 215 km s-1, while we found 135 km s-1. The origin of such strong disaccord is unknown, but in Fig. 7 we show for this star the fit between observed and synthetic spectra which supports our estimate.
Here we have to note that high rotation may affect the spectral classification (Gray & Garrison, 1987). The photometric indices ( and , in particular) of a rapidly rotating star can mimic those of a slowly rotating one with a lower temperature. Fortunately, as it was shown by Gray & Garrison (1987), this effect is small. Up to 100-200 km s-1, the correction is not greater than -0.02, and it was therefore not taken into account.
© European Southern Observatory (ESO) 2000
Online publication: April 10, 2000