Using X-ray observations to explore the binary interaction in Eta Carinae

We study the usage of the X-ray light curve, column density towards the hard X-ray source, and emission measure (density square times volume), of the massive binary system eta Carinae to determine the orientation of its semimajor axis. The source of the hard X-ray emission is the shocked secondary wind. We argue that, by itself, the observed X-ray flux cannot teach us much about the orientation of the semimajor axis. Minor adjustment of some unknown parameters of the binary system allows to fit the X-ray light curve with almost any inclination angle and orientation. The column density and X-ray emission measure, on the other hand, impose strong constrains on the orientation. We improve our previous calculations and show that the column density is more compatible with an orientation where for most of the time the secondary - the hotter, less massive star - is behind the primary star. The secondary comes closer to the observer only for a short time near periastron passage. The 10-week X-ray deep minimum, which results from a large decrease in the emission measure, implies that the regular secondary wind is substantially suppressed during that period. This suppression is most likely resulted by accretion of mass from the dense wind of the primary luminous blue variable star. The accretion from the equatorial plane might lead to the formation of a polar outflow. We suggest that the polar outflow contributes to the soft X-ray emission during the X-ray minimum; the other source is the shocked secondary wind in the tail. The conclusion that accretion occurs at each periastron passage, every five and a half years, implies that accretion had occurred at a much higher rate during the Great Eruption of eta Car in the 19th century. This has far reaching implications for major eruptions of luminous blue variable stars.