Wind-wind collision in the eta Carinae binary system - III. The HeII 4686 line profile

We modeled the HeII 4686 line profiles observed in the eta Carinae binary system close to the 2003.5 spectroscopic event, assuming that they were formed in the shocked gas that flows at both sides of the contact surface formed by wind-wind collision. We used a constant flow velocity and added turbulence in the form of a gaussian velocity distribution. We allowed emission from both the primary and secondary shocks but introduced infinite opacity at the contact surface, implying that only the side of the contact cone visible to the observer contributed to the line profile. Using the orbital parameters of the binary system derived from the 7 mm light curve during the last spectroscopic event (Paper II) we were able to reproduce the line profiles obtained with the HST at different epochs, as well as the line mean velocities obtained with ground based telescopes. A very important feature of our model is that the line profile depends on the inclination of the orbital plane; we found that to explain the latitude dependent mean velocity of the line, scattered into the line of sight by the Homunculus, the orbit inclination should be close to 90 degrees, meaning that it does not lie in the Homunculus equatorial plane, as usually assumed. This inclination, together with the relative position of the stars during the spectroscopic events, allowed us to explain most of the observational features, like the variation of the Purple Haze with the orbital phase, and to conciliate the X-ray absorption with the postulated shell effect used to explain the optical and UV light curves