``HeII $\lambda$4686 Line Profile Variations

of HD 191765 (WR134)''

A. Flores¹, O. Cardona¹ and G. Koenigsberger²

1: Instituto Nacional de Astrofisica Optica y Electronica, Apodo. Postal 51, Puebla, Pue. 7200, Mexico

2: Instituto de Astronomia. UNAM, Apodo. Postal 70-264, Mexico D. F. 04510, Mexico

ABSTRAC.

HD 191765 is a Wolf-Rayet star with a high degree of variablity in its emission lines, is a controversial star because some models which based in the results of monitoring with different observational techniques have been proposed for the description of the physical mechanisms that induce the variability. Assuming the star to be a binary star, we have been able to make a direct comparison between high resolution observational HeII 4686 line profiles and synthetic profiles calculated with the code for "atmospheric eclipses" of Auer & Koenigsberger (1994).

INTRODUCTION.

HD 191765 (WR134) is a Wolf-Rayet star (WRs) which based on their emission line profile variations and from the analysis of their photometric data, was proposed to be a WR+cc system (Antokhin et al. 1982) . However, recent intensive monitoring with different observational techniques, has given rise to alternative explanations for the variability in these WRs include the presence of a disk (Shulte-Ladbeck et al. 1992), rotating jet-like structures (Vreux et al. 1992), and corotating interaction regions (Morel et al. 1998). In order to resolve whether or not binary interaction effects can be responsible for the line profile variability being observed, numerical simulations generated under the assumption of this interaction must be generated and compared with the observations. Koenigsberger (1995) showed that the line profiles produced under the binary assumption were qualitatively similar to those observed in the WR+collapsed companion candidate HD 50896. In this poster we present a comparison between the line profiles computed assuming orbital motion and atmospheric eclipses with observations of HD 191765 (WR134). The effect on the line profiles due to orbital motion and atmospheric eclipses is computed as in Auer and Koenigsberger (1994). The intrinsic line profile arising in the WR wind is calculated using the Sobolev approximation in those portions of the wind in which this approximation is valid. In the regions of slower acceleration, the equations of radiative transfer are solved exactly. For the profiles presented here, a linear velocity law was employed. The WR emission line is assumed to remain constant throughout the orbital cycle, and the wind is assumed to be spherically symmetric. The companion is assumed to have an associated Gaussian emission line which follows its orbital motion. The input parameters are described in Table 1, which also contains the values we have used for these parameters in this test case. The calculated profiles are produced on a trial-and-error basis, by systematically modifying the input parameters until a reasonably good fit to a reference observed profile is obtained. Because the basic assumption is that the variations are due to the presence of a companion, an assumption must be made regarding which of the observed line profiles most resembles one which we could assign to a specific orbital phase. Once this reference spectrum is identified, and the input parameters of the code are defined, synthetic spectra are generated for the other orbital phases, corresponding to the available observational data for comparison (in Figure 3, is seems synthetic spectrum superposed on the real spectrum) . Phase zero corresponds to the companion "in front" of the WR star (see Figure 1). Differences between the observed and the computed line profiles can be attributed to effects which are not considered in the computer code, such as line formation in a shock cone, non-spherically symmetric mass loss from the WR star, and intrinsic variability (including clumping) in the WR wind. The observations were obtained at the 2.1m "Guillermo Haro Observatory" (OGH) in Cananea Sonora, during various observing runs during 1997-1998. The high S/N ($\approx$ 300) and reciprocal dispersion (0.31 $\rm\AA \;$/pix) at the HeII 4686$\rm\AA \;$allow a good comparison with the computed models, (in Figure 2, He II 4686 line profiles are ordered in phase).

Code input parameters

 

PARAMETER	NUMERICAL VALUE	DESCRIPTION
Rmax	40	maximum extent of line-emitting region in stellar radii units
Vmax	50	maximum expansion velocity, in thermal velocity units
Ra	40	acceleration radius in stellar radii units
F(r)	2.5e4	monocromatic opacity function
PRIM	35	continuum intensity of WR star emerging	from $r=1R_{\star}$
SEC	5	intensity of companion's radiation
Rorb	40	orbital separation
Rc	1	radius of the companion
Vel	15	orbital velocity, in thermal velocity units
DEPTH	-10	intensity of line arising in companion
WIDTH	2.5	width of line arising in companion

Table 1: Parameters describing the WRs and its wind structure ( F_wr,R_max v(r),F(r), and $V_{\infty }$), the binary parameters (orbital separation, speed), and the characteristics of the companion (relative intensity, radius, width of its emission).

  

Figure 1: Geometric representation of the simplified WR+cc Model.

  

Figure 2: HeII 4686 ordered in phase with period 2.27 d, and ephemerides of Mc Candliss et al. 1994.

  

Figure 3: Phase dependent synthetic line profiles superposed on the HeII 4686 of the WR+cc candidate HD 191765 (WR134).

CONCLUSIONS.

1.- The simple model of a binary system in which both components have associated emission arising in spherically symmetric line-forming regions is capable of reproducing some of the observed profiles, but not all. The differences between the observed and predicted line profiles indicate that:

a) Mc Candliss period is incongruent with the real period of variability in the line profiles

b) line emission arising in a shock cone surrounding the companion needs to be incorporated;

c) non-spherically symmetric line forming regions are required;

d) wind instabilities play a major role in the shape of the upper portion of the line HeII $\lambda$4686 profile and are not the result of the presence of a possible binary companion.

2.- This star shows a greater degree of variability on the upper portions of the line profile on timescales shorter than 2.5 hrs.

3.- Is necessary to establish a well defined period with our set of spectral data.

REFERENCES.

Auer. L. H., Koenigsberger, G. 1994, Apj, 436, 859

Antokhin, I. I., Asianov, A. A., Cherepashchuk, A. M. 1982, Soviet Astronomy Letters, vol 8, 156

Koenigsberger, G., 1995, in: K. A. van der Hucht & P. M. Williams (eds.), Wolf-Rayet Stars: Binaries, Colliding Winds, Evolution, Proc. IAU Symp, No. 139, p. 538

Mc Candliss, S. R., Bohannan, B., Robert, C., & Moffat, A. F. J. 1994, Ap & SS, 221, 155

Morel, T. 1999, Apj, submitted

Shulte-Ladbeck, R. E., Meade, M. R. 1992, in: L. Drissen, L., C. Leitherer, & Nota, A. (eds.), Nonisotropic and Variable Outflows from Stars, ASP Conf. Series 22, 118

Vreux, J. M., Gosset, Bohannan, B., Conti, P. 1992, A & A, 256, 148