Forum Springer Astron. Astrophys.
Forum Whats New Search Orders

Astron. Astrophys. 320, 500-524 (1997)

Previous Section Next Section Title Page Table of Contents

Appendix: results for individual stars

A.1. Brey 24 (FD23, WN6h)

Brey 24 whose spectral type has recently been revised from WN7 to WN6 (L.F. Smith et al. 1996) is located in the OB association LH45. The derived parameters ([FORMULA] [FORMULA] 36kK, [FORMULA], log ([FORMULA]) [FORMULA] 5.4) from our profile fits (Fig. 11) are fairly consistent with previous determinations as shown in Table 3, while we obtain a slightly higher hydrogen content of H/He [FORMULA] 1.5, relative to L.J. Smith & Willis (1983).

[FIGURE] Fig. 11. Synthetic profile fits (dotted lines) to radial velocity corrected observations of WN6 stars (solid line) for He II [FORMULA] 4686, [FORMULA] 5412, H [FORMULA] - [FORMULA], He I [FORMULA] 5876. Observations are AAT-RGO except where indicated. Spikes observed in profiles of Brey 75 are caused by the poor removal of the 30 Dor nebular component

A.2. Brey 26 (HD 36063, WN6(h))

Brey 26 is another star whose spectral appearance has recently been revised to WN6 from WN7 (L.F. Smith et al. 1996). Although Brey 26 closely resembles Brey 24, its wind is faster ([FORMULA] [FORMULA] 1500 km s-1), denser [FORMULA] and more enhanced in helium (H/He [FORMULA] 0.4) as shown in Fig. 11. A comparison with previous studies is presented in Table 3. Brey 26 shows rapid ([FORMULA] 1.9 days) radial velocity and photometric variations (Moffat 1989; Seggewiss et al. 1991) indicating either a close binary system or significant intrinsic variability. Since no photospheric absorption lines are present, it is possible that this star has a compact companion. In the absence of any quantitative information about the possible companion, we have assumed that the visual spectrum of Brey 26 is entirely due to the Wolf-Rayet star.

A.3. Brey 47 (FD42, WN6h)

Brey 47, whose spectral type has recently been revised from WN8 to WN6h (L.F. Smith et al. 1996) closely resembles Brey 24 both in its visual appearance (see Fig. 11) and derived parameters ([FORMULA] [FORMULA] 35kK, [FORMULA], log ([FORMULA]) [FORMULA] 5.3). From Table 3 our results are broadly consistent with Koesterke et al. (1991) and Vacca (1992).

A.4. Brey 75 (R134, WN6(h))

The spectral type of Brey 75 is revised from WN7 to WN6 in Sect. 3.1. This star, located within the core of 30 Doradus, is highly luminous (log ([FORMULA])=5.9). Moffat (1989) was unable to find evidence for multiplicity in this object. Line profile fits (Fig. 11) result in typical WN6 stellar parameters except for a low hydrogen content of H/He=0.2 [FORMULA] 0.2.

A.5. Brey 89 (R144, WN6h)

Brey 89 is another 30 Doradus member and is visually the brightest single Wolf-Rayet star in the LMC with a spectral classification recently revised from WN7 to WN6. Overall agreement between theoretical and observed profiles is reasonably good (Fig. 11). While the spectrum of Brey 89 does not appear to be unusual (Fig. 2) this star possesses an enormous stellar luminosity (log ([FORMULA]) [FORMULA] 6.25), a factor of two greater than the most luminous Galactic WNL star thus far analysed (WR25, WN6ha; Paper II), and corresponding high mass-loss rate of [FORMULA]. The radial velocity derived for Brey 89 (Table 7) agrees well with that (267 km s-1) determined by Moffat (1989), while H/He [FORMULA] 1.0 is unremarkable relative to other LMC WNL stars. A comparison with the previous study of L.J. Smith & Willis (1983) is shown in Table 3. From our present analysis we obtain H0 and He0 ionizing photon luminosities of log ([FORMULA])=49.85 s-1 and log ([FORMULA])=49.05 s-1, typical of an O5.5 supergiant (Vacca et al. 1996). Incidentally, our synthetic spectrum of Brey 89 also reproduces observed ultraviolet (He II [FORMULA] 1640) and infrared (He I 1.083µm, He II 1.012µm, 1.163µm) line profiles, the latter from I.D. Howarth (private communication).


Table 7. Optical and IR photometry and absolute magnitudes of the programme LMC stars. Since narrow band [FORMULA] (Smith 1968) optical photometry is available for only a subset of our programme stars we utilise wide band B, V Johnson photometry (typically within 0.05 mag of narrow band measurements for WNL stars). Our derived [FORMULA] determination, including a 0.05 mag Galactic component, is compared with the previous values from L.J. Smith & Willis (SW, 1983), Schmutz & Vacca (SV, 1991) and Morris et al. (PWM, 1993a)

A.6. Brey 90 (R145, WN6(h))

Brey 90 located near to Brey 89 in 30 Doradus closely resembles its neighbour and has also been revised to WN6 from WN7 by L.F. Smith et al. (1996). The stellar parameters of Brey 90 based on profile fits (Fig. 12) compare closely with Brey 89 except for a slightly lower luminosity (log ([FORMULA]) [FORMULA] 6.16), mass-loss rate [FORMULA] and hydrogen content (H/He [FORMULA] 0.7), while Table 3 shows a comparison with the previous analysis of L.J. Smith & Willis (1983). Unlike Brey 89, this star shows radial velocity variability. Moffat (1989) tentatively identified a long spectroscopic period of 25 days. Should this be confirmed, the stellar parameters of Brey 90 will require revision although given its visual magnitude ([FORMULA] =-7.43), it is unlikely that the derived values would be significantly affected.

[FIGURE] Fig. 12. Synthetic profile fits (dotted lines) to radial velocity corrected observations of WN6-8 stars (solid line) for He II [FORMULA] 4686, [FORMULA] 5412, H [FORMULA] - [FORMULA], He I [FORMULA] 4471, [FORMULA] 5876. Observations are AAT-RGO except where indicated. Spikes and data gaps observed in profiles of Brey 80 and Brey 90 are caused by the poor removal of the 30 Dor nebular component

A.7. Brey 71 (HDE 269883, WN7h)

Turning now to the WN7 stars, Brey 71 has been little studied because of its location within the core of 30 Doradus. Brey 71 shows the strongest He II signature of our programme stars and therefore has the highest stellar temperature ([FORMULA] [FORMULA] 39kK), and smallest stellar radius, with typical mass-loss rate [FORMULA] and luminosity (log ([FORMULA]) [FORMULA] 5.3). Despite the high stellar temperature, the derived hydrogen content of H/He [FORMULA] 1.0 is typical of our WNL sample. Line profile fits are shown in Fig. 12.

A.8. Brey 80 (R135, WN7h)

Brey 80, also located within the core of 30 Doradus, has a peculiar spectral appearance among the WNL stars with strong, broad non-Gaussian profiles. The derived stellar parameters for Brey 80 are unexceptional except for an enormous mass-loss rate [FORMULA]. In particular, we find a wind performance number of fifty for Brey 80 and a mechanical luminosity [FORMULA] of around 15% that of the radiative luminosity. Such quantities have previously been exclusive to WNE and WC stars (Howarth & Schmutz 1992). Although profile fits are reasonable (see Fig. 12), improved fits to He II and H I (though not He I) can be achieved using a slower velocity law ([FORMULA] [FORMULA] 3) which alters the derived stellar temperature and radius, though not the mass-loss rate or luminosity. The derived hydrogen content of H/He [FORMULA] 0.8 for Brey 80 is therefore less reliable than for other WN6-7 stars due to the difficulty in modelling its Balmer profiles.

A.9. Brey 13 (HD 33133, WN8h)

Brey 13, the standard WN8 star of the LMC, is probably single (Moffat 1989) and has an ejecta ring nebula (Garnett & Chu 1994). The derived stellar parameters ([FORMULA] [FORMULA] 34kK, [FORMULA], log ([FORMULA]) [FORMULA] 5.6), chemistry (H/He [FORMULA] 1.4 by number) and spectral appearance of Brey 13 are remarkably similar to the Galactic WN8 star WR40 (HD 96548, Paper II), as is its ejecta plus wind blown nebula, enriched in nitrogen (L.J. Smith et al. 1985; Garnett & Chu 1994). Profile fits are shown in Fig. 12 and show good agreement with observation, except for the weak P Cygni absorption component observed at He II [FORMULA] 5412. Our results show reasonable agreement with previous determinations (Table 3).

A.10. Brey 36 (FD32, WN8h)

Brey 36 is another well known LMC WN8 star. Relative to Brey 13, the emission spectrum of Brey 36 is significantly weaker and narrower, resulting in a low wind velocity (650 km s-1) and mass-loss rate [FORMULA], although a similar stellar temperature and helium content (H/He [FORMULA] 1.7) are derived (see Table 2). Our derived parameters are fairly consistent with Vacca (1992) while we obtain a high radial velocity of [FORMULA] 350 km s-1 from our line profile fits (Fig. 12).

A.11. Brey 81 (AB11, WN8(h))

Brey 81 represents the third (and final) LMC WN8 star. Another Wolf-Rayet star close to the core of 30 Doradus, Brey 81 was found to be multiple by Parker (1993), although our observed line strengths for this star do not indicate significant contamination by its (unseen) companion. Relative to the other LMC WN8 stars, Brey 81 shows a high luminosity (log ([FORMULA])=5.9) and very low hydrogen content of H/He [FORMULA] 0.2 ([FORMULA] 5% by mass), and is thus similar to the hydrogen-poor Galactic WN8 star WR123 (HD 177230, Paper II).

A.12. AB18 (Brey 44a, WN9h)

This star, discovered by Azzopardi & Breysacher (1985) and originally classified WN8-9, marginally falls into the WN9h category. Relative to other LMC WN9 stars, this object shows a weak emission line spectrum (see Fig. 13 for profile fits). The derived stellar parameters for AB18 resemble Sk- [FORMULA]  249c from Paper I except for a lower luminosity (log ([FORMULA]) [FORMULA] 5.5), and mass-loss rate [FORMULA]. The absence of ultraviolet observations means that the reddening is poorly constrained, and thus a rigorous [FORMULA] determination is not possible.

[FIGURE] Fig. 13. Synthetic profile fits (dotted lines) to radial velocity corrected observations of WN8-11 stars (solid line) for He II [FORMULA] 4686, [FORMULA] 5412, H [FORMULA] - [FORMULA], He I [FORMULA] 4471, [FORMULA] 5876. Observations are AAT-RGO except where indicated

A.13. BE294 (HDE 269582, WN10h)

BE294 has previously been classified as Ofpe/WN9 (Bohannan & Walborn 1989) and identified as a LBV (Bohannan 1989), and is now re-classified as WN10h. The optical morphology of this star closely resembles the other LMC WN10 star S9 (Sk- [FORMULA]  40, Paper I). While the stellar temperature ([FORMULA] 29kK) and wind velocity ([FORMULA] 300 km s-1) of BE294 are almost identical to S9 (Table 2), it is more luminous and exhibits a more chemically processed surface composition (H/He [FORMULA] 1.9). McGregor et al. (1988) estimated H/He=0.5-2 for BE294 from its K-band spectrum. Profile fits for BE294 are shown in Fig. 13. Our present model slightly underestimates the observed emission strength of He I 2.058µm relative to Br [FORMULA] (Blum et al. 1995), as was previously found for S9 in Paper I. Incidentally, BE294 showed significant variability in the IR He I /Br [FORMULA] flux ratio between 1987 January and 1993 September, when McGregor et al. (1988) recorded [FORMULA] 3 and Blum et al. (1995) obtained [FORMULA] 0.3.

A.14. S119 (HDE 269687, WN11h)

S119, whose spectral classification is here revised from Ofpe/WN9 to WN11h, is another LBV candidate (Nota et al. 1994). As described in Sect. 3.3, the stellar Balmer lines in S119 are contaminated by emission from the surrounding nebula (Nota et al. 1994). We have carried out two analyses: Model A uses our AAT-RGO spectra including Balmer nebular contamination; and Model B is based on our AAT-UCLES and MSO-coudé spectra, with Balmer nebular contamination removed (see Fig. 4a and b). The derived stellar parameters for both models of S119 are shown in Table 2, and profile fits are presented in Fig. 14. Overall, the narrow, weak emission line signature of S119 results in a low stellar temperature ([FORMULA] 28kK), mass-loss rate [FORMULA] and wind velocity ([FORMULA] 230 km s-1) for both models with a wind momentum that does not exceed the single scattering limit. We are unable to simultaneously reproduce the Balmer profiles in Model A since the relative stellar and nebular components behave differently for different lines. Using H [FORMULA], H/He [FORMULA] 4 is obtained for Model A, while a reliable value of H/He [FORMULA] 1.50 is obtained for Model B. Finally, while Model B predicts marginally stronger He I 2.058µm emission than Br [FORMULA], the observed He I /Br [FORMULA] flux ratio is 3 (Blum et al. 1995) which presumably includes a nebular contribution for Br [FORMULA].

[FIGURE] Fig. 14. Synthetic profile fits (dotted lines) to radial velocity corrected observations of S119 (solid lines). Model A fits are to observations (AAT-RGO, MSO-coudé) including Balmer nebula contamination while Model B fits are to observations (AAT-UCLES, MSO-coudé) excluding nebula contamination. Data are AAT-RGO except where indicated

A.15. S142 (BE470, WN11h)

S142 is another star revised from Ofpe/WN9 to WN11h with somewhat stronger profiles than S119 (see Fig. 13) but very similar stellar parameters, except for a lower luminosity (Table 2). While our observations do show weak [N II ] [FORMULA] 6548-6583 emission (Fig. 5), contamination of the stellar Balmer features is negligible so that the derived chemical composition (H/He [FORMULA] 3) should be accurate. While we are unaware of existing K-band spectroscopy of S142, our model predicts He I 2.058µm [FORMULA] Br [FORMULA] emission.

A.16. S61 (BE153, WN11h)

Finally, S61 is another star revised to WN11h from Ofpe/WN9 whose spectrum and physical parameters (Fig. 13) closely mimic S119, except for marginally stronger He II [FORMULA] 4686. The nebula surrounding S61 contributes [FORMULA] 75% of the H [FORMULA] line flux (Sect. 3.3; Fig. 5). After nebular subtraction, we obtain a relatively low hydrogen content of H/He [FORMULA] 1.2. Indeed, as for Model A for S119, assuming that the Balmer line strengths result purely from stellar emission would imply a huge hydrogen content (H/He [FORMULA] 6). McGregor et al. (1988) obtained K-band spectroscopy of S61 showing He I 2.058µm and Br [FORMULA] emission of comparable strength, and estimated H/He [FORMULA] 5. In common with other WN10-11 stars studied here the predicted He I 2.058µm emission is somewhat weaker than that observed, which we attribute to the neglect of line blanketing (Paper I; Najarro et al. 1994).



Table 4. Journal of high dispersion UV (IUE -HIRES, HST -FOS) spectroscopic observations


Table 5. Wind velocities (in km s-1) measured from UV P Cygni resonance lines in high dispersion spectra following Prinja et al. (1990), UV low dispersion spectra following Prinja (1994, given in parenthesis) or optical He I P Cygni profiles. Previous determinations by Koesterke et al. (1991, KHSW) using optical He I profiles and Rochowicz & Niedzielski (1995, RN) using UV low dispersion spectra are also shown for comparison


Table 6. Selected optical emission equivalent widths ([FORMULA], in Å), FWHM (in Å) and radial velocities (in km s-1) for the programme LMC WN6-11 stars based on our new AAT and MSO observations ([FORMULA] indicates H [FORMULA] data from MSO). Measurements for WN9-10 stars from Paper I are also shown (including new H [FORMULA] MSO observations). The Balmer line strengths for S119 and S61 include a strong nebular contribution

Previous Section Next Section Title Page Table of Contents

© European Southern Observatory (ESO) 1997

Online publication: June 30, 1998