A systematic study of the structure and interaction of winds in selected hot close binary stars

The connection of stellar winds to many astrophysical problems ranges from star formation, through stellar evolution, to the dynamics of the interstellar medium in galaxies. In order to understand the winds from hot O and B type stars, we must determine and explain the distribution of such physical variables as density, velocity, and temperature throughout the wind volume. This thesis presents an innovative technique similar to that of tomography in medical imaging, and its implementation in determining very accurate wind parameters for spectroscopic binary stars. It is based on the observation that certain aspects of the morphological variation of the observed wind-line profiles can be attributed to purely geometrical effects: as the stars revolve around their common center of mass, the geometry of a binary presented to the observer changes with phase. Therefore, at one time the stars will align so that the front star/wind occults/shadows the companion star and its wind, and at a later time, they will appear side-by-side and the shadowing effects will be minimal. These effects depend upon the inclination of the binary orbital plane to the observer\u27s line of sight. The evidence for binary wind line variability which the present study analyzes exists in sequences of ultraviolet spectra secured by the International Ultraviolet Explorer satellite at various phases of the orbit, and which sample different geometrical aspects. The presence of winds is initially quantified by deriving wind light curves and by comparing these to photospheric light curves. Subsequently, the wind lines are modeled so as to determine the structure of each wind, and to locate any wind-interaction effects between the two such as density enhancements and/or shocks. This dissertation reports the results of studying the wind structure and interaction for four binaries. As an additional 20 binaries have a sufficient number of images to be analyzed, it can be seen that the work reported here commences a much larger scale effort which will eventually claim the most accurate wind parameters derived from wind profile modeling