New infrared spectra of Wolf-Rayet stars are analysed to derive C/He abundance ratios and ionization ratios. These ratios are discussed in relation with subtype, terminal velocity and local metallicity. They are compared with the predictions of evolutionary models.

We derive the dynamics of a steady-state Wolf-Rayet wind using a non-isotropic diffusion approximation applied to the transfer between strongly overlapping spectral lines. Following the approach of Friend and Castor (1983), the line list is assumed to approximate a statistically parametrized Poisson distribution in frequency, so that photon transport is controlled by an angle-dependent, effectively gray opacity. We find that multiple radiative momentum deposition is greatest when photons undergo a nearly local diffusion, e.g. through scattering between many lines closely spaced in frequency. Our results reiterate the view that the so-called “momentum problem” of WR winds is better characterized as an “opacity problem” of simply identifying enough driving lines. One way of increasing the number of thick lines in WR winds is to transfer opacity from saturated to unsaturated lines, yielding a steeper opacity distribution than that found in OB winds. We also consider the growth of instabilities, and find that WR winds are up to an order of magnitude less unstable than OB winds if both are driven radiatively. However, initially small perturbations can still be multiplied by many orders of magnitude as they advect through WR winds.

There are many indications for the existence of dense cold clumps in Wolf-Rayet (WR) atmospheres, expanding outwards from the stellar core (e.g., Cherepaschuck et al. 1984, St-Louis et al. 1993). Based on this, we consider a clump model for WR model atmospheres. The main features of the model are described. The mechanism of clump structure formation is outlined. The clump model is used for calculating emission line profiles.

The formation of the lines of ions He, H, N and C in the WR stars of different spectral subtypes has been studied using clumped wind models. Our new results agree well with the previous results obtained with smooth wind codes which make use of highly excited lines of different ions.

Synthetic line-profiles from unified model atmospheres are presented. The main characteristics of the code are also shown concerning both the calculation of NLTE populations (in a consistent way with the radiation field), and the formal solution to obtain emergent profiles. Stark-broadening is considered in this last step. Emergent Hα profiles for a grid of models are shown for Teff = 50 000 K, covering a wide range of stellar radii and mass-loss.

We recently showed that Alfvén waves acting jointly with radiation pressure in Wolf-Rayet winds can possibly solve the momentum problem (dos Santos, Jatenco-Pereira and Opher 1993a, 1993b). In these analyses, however, constant radiation luminosity as a function of the distance to the star was assumed. We impose here conservation of the energy flux and examine the effect on the radiation luminosity.

We present a Wolf-Rayet wind model including clumping, which is assumed to be due to full-scale supersonic compressible turbulence as suggested by Moffat & Robert (1994). It is shown that clumping leads to a larger transfer of radiative to mechanical luminosity, which corresponds to a momentum ratio of (Ṁv∞)/(L/c) = 12 for a typical WN5 star. We also show how clumping reduces the effects of line blanketing in spectral analyses.

The possibility of driving the dense winds of Wolf-Rayet stars via multiple scattering of photons in resonance lines is examined. For the case of many overlapping lines it is found that the diffuse radiation arising from the multiple scattering process leads to a radiative acceleration grad even larger than the commonly used unattenuated flux models where each line is assumed to interact separately with the unattenuated photospheric flux. The radial dependence of grad in the outer parts of the wind is now weaker than r−2, which in turn leads to a steeper rise in the velocity field.

When the effective opacity becomes sufficiently great, a qualitatively different type of radiatively driven wind solution appears, in which a high wind thickness and trapping of radiation cause each other in a bootstrap fashion. It is proposed that the thick winds from Wolf-Rayet stars can be understood in this manner. Opacity computations using the method of super transition arrays (STA) are presented that show that outflows from WR stars can attain the needed opacities.

Results from pointed ROSAT PSPC observations of nine single WN-type Wolf-Rayet stars are presented. Spectra of sufficient quality were obtained for two of them (WR1, WR110). The long exposure (35.5 ksec) X-ray spectrum of WR1 is more closely investigated with a semi-empirical model developed by Baum et al. (1992).

Motivated by the question of what drives the winds of Wolf-Rayet (WR) stars, I have been investigating the structure of WR winds. Observational evidence has been accumulated which demonstrates that the winds of some (but not all) WR stars are not spherically symmetric or homogeneous. I review results of work in the literature pertaining to the detection of such wind asymmetries in single WR stars and WR binary components. Particular emphasis is placed on the technique of spectropolarimetry, and how it has contributed to our knowledge of the wind-geometry of WR stars.

The wind structure diagnostic possibilities offered by combined study of photo-, spectro-, and polari-metric variability are discussed using data from the Montreal group. These demand the presence in the wind of localised density inhomogeneities and in particular of large ‘blobs’ denser than the ‘mean’ wind and put bounds on the size, mass, and density of individual blobs and their distributions, with implications for theories of blob formation. Blobs responsible for polarimetric variability must be present at the stellar surface rather than forming in the wind, and the relation of this to spectrometric indications that blobs are only detectable at several stellar radii is discussed.

Theoretical models for Wolf-Rayet winds that include the effects of rotation and magnetic fields are discussed. These are of two types: magnetic rotator models and wind compression models. The magnetic rotator models are examples of equatorial expulsion models, typically requiring large rotation rates to produce significant equatorial density enhancements. Wind- compression models are introduced here as an application of the two-dimensional Wind-Compressed Disk (WCD) model that has recently been developed for Be stars. An equatorial disk forms because of the supersonic confluence of the flow from the upper and lower hemispheres of the star. In the application to WR stars, we suggest that disk formation is not required, and instead a less extreme example that we call a wind-compressed zone (WCZ) may suffice. Because the winds of WR stars have a geometrically extended acceleration region, or a “slow” velocity law, there can be compressions by an order of magnitude, even with moderate stellar rotation rates of about 16% critical. Magnetic flux conservation is then used to estimate the enhancement of the field strength that occurs because of the equatorial wind-compression. We consider an application of the model to explain the occurrence of polarization and dust formation among only some WR stars. Finally, we consider the WR + compact companion system WR 147, and suggest that the enhanced accretion in a WCZ could help explain the large X-ray emission.

We discuss the theory of optically thick winds from Novae, how they are treated numerically, and the effect of the mass loss on the long term evolution of the WD.

It is argued that, with few exceptions, WR stars in binaries are essentially indistinguishable from single WR stars of the same subtype. In particular, mass transfer from one star to the other can be neglected, being completely dominated by the winds and their interaction. Empirical masses and mass-loss rates derived for WR stars in binaries are then examined as typifying WR stars in general; they are used to derive a plausible scenario for the evolution of WR stars.

Revised orbits are presented for the double-lined binary WR+O systems HD 63099 (WR9) and HDE 320102 (WR97). The orbit derived for WR9 from OVI emission and HeII absorptions is considerably different from that previously published based on the CIVΛ4658 emission line velocities. WR97 is found to have a period of 12.6 days and small minimum masses, probably due to a low orbital inclination. Two stars in our Galaxy, WR30a and WR70, and two in the Magellanic Clouds, Brey44 and AB7, show radial velocity variations of emission and absorption lines in opposite sense. Although no periods are yet available, preliminary values of the mass ratios can be estimated from the velocity variations.

New medium—resolution spectroscopic observations of two long—period WR binaries have been obtained using the 1.5 m telescope at the Tartu Astrophysical Observatory. The binarity of these systems is confirmed and orbital parameters are determined.

We report here the results, to date, of our monitoring of northern WR stars. New orbits have been published for the short period binaries HD 193576 and HD 214419. Of particular interest are two long period binaries with fairly eccentric orbits, HD 190918 and HD 193793. A new orbit has been determined for HD 190918. We are currently analysing spectra obtained over one cycle of the 7.9 year period of HD 193793. Other stars on our primary target list include HD 192163, HD 192641, and HD 193077.

Using a simple absorption line model the phase dependent effects of ‘wind eclipses’ have been studied for the N IV 1718Â line in the WN4+05 binary, HD 90657. The varying strength of the absorption, as measured by the equivalent width, has been compared with the model for a range of wind parameters and assumptions. It is found that the fraction of absorbing atoms drops quickly, to the first or second power of the radius, and that a significant fraction persists beyond 60 R⊙.

New spectroscopic and photometric observations of WR22 are presented. High signal-to-noise digital spectra with high dispersion are used to improve the parameters of the orbit (P = 80.31 ±0.01 d), formerly determined on the basis of photographic spectra only. The intrinsic photometric variability of WR22 and recent observations of the eclipse are also discussed.