ASCA observations of gamma(2) Velorum (WC8+O9I): The variable X-ray spectrum of colliding winds

We present an analysis of two observations of the nearby Wolf-Rayet (WR) + O-star binary system y(2) Velorum with the Advanced Satellite for Cosmology and Astrophysics (ASCA). These observations were taken at orbital phases phi = 0.4 and 0.5, when the O star is close to being in front of the WR star. At phase phi = 0.5, y(2) Vel is approximately four times more luminous in X-rays than it is at phi = 0.4. We fit the X-ray spectra using simple combinations of absorbed Raymond-Smith emission models. We find evidence of a cool (kT similar to 0.3 keV) and a hot (kT similar to 1.3 keV) component in each spectrum. The hotter component suffers from variable absorption which suggests that it originates in a localized region between the two stars. Between phi = 0.4 and 0.5 the absorbing column for this hotter component drops by a factor of 3 which accounts for most of the corresponding increase in flux. These observations are interpreted in terms of a colliding-wind model. The collision between the winds from the O star and the WR star generates a region of hot X-ray-emitting gas. When the WR star is in front, the dense WR wind effectively absorbs all the X-ray flux from the wind collision. However, when the O star is in front the more diffuse wind of the O star allows substantially more flux from the wind collision to escape. Our ASCA data confirm the colliding-wind origin for the hard X-ray emission from y(2) Vel, first identified from ROSAT data by Willis, Schild & Stevens. At phi = 0.5 the emergent luminosity is L(x) similar to 3.9 x 10(32) erg s(-1). We also use hydrodynamic models of colliding winds to calculate synthetic X-ray spectra for y(2) Vel, which we then fit to the ASCA spectra. This allows us to determine fundamental stellar wind parameters for both stellar components from the X-ray spectra. In particular, we determine a mass loss for the WC8 star of similar to 3 x 10(-5) M. yr(-1), a value 3 times lower than that determined by radio observations. The consequences of this are discussed. We consider that we are definitely seeing colliding-wind emission and also show the great potential for X-ray spectroscopy in determining fundamental wind parameters of WR (and other hot luminous) stars.