 
* | ||||||||||
|
Beta Lyrae (also named Sheliak, HD 174639, and others), is an eclipsing binary star in the constellation Lyra. The two stars cannot be resolved with current telescopes, but their combined brightness varies by 0.86 magnitudes (over 50%) during the course of a two-week cycle, as they alternately eclipse each other. The light curve of beta Lyr has a characteristic shape with no flat regions (Figure 1). From this and other observations, such as the fact that no spectral lines of the secondary star appear in the combined spectrum, astronomers surmise that the system consists of a large, evolved, distorted primary star losing mass to a thick edge-on accretion disk surrounding the secondary star. Some mass may also leave the system in a process called nonconservative evolution. My thesis research uses polarimetry and modeling to locate and describe the circumstellar material in beta Lyr; understanding the mass flow in this prototypical system can lead to important insights to our understanding of stellar evolution and the enrichment of the interstellar medium.
Ken Nordsieck and I conducted spectropolarimetric observations of beta Lyr with the WUPPE instrument and with the HPOL spectropolarimeter at the University of Wisconsin's Pine Bluff Observatory which showed that the UV radiation from beta Lyr is polarized at 90° to the visible-wavelength light (Figure 2). Since we believe the accretion disk polarizes the visible light, this means that another component exists in the system: a jet or outflow of gas oriented perpendicular to the disk. We have recently obtained data from STIS on the Hubble Space Telescope that should reveal further information about this jet.
We are also working on modeling this system with a Monte Carlo model that simulates the polarization behavior of binary-disk systems. We represent the primary star with a prolate ellipsoid and the disk with a spherical wedge with a constant density of electrons, and use the eclipse depths in the light curve to constrain the relative sizes of the two. Since the secondary star is not seen in the spectrum, we assume it's hidden by the disk. A typical model looks something like Figure 3. We then vary the optical depth and albedo (relative importance of scattering and absorption) within the disk in order to match the observed polarization behavior.
We have found, however, that this simple model cannot reproduce the V-band polarized flux curve of beta Lyr (Figure 4). The size of the "bump" at phases 0.3 and 0.7, which is produced by primary light scattering off the disk, is easy to match; however, the self-scattered disk cannot produce enough "DC" polarization at the other phases while still hiding the secondary star. We conclude that the uniform disk is too simple an approximation. The secondary star must contribute some polarized flux (increasing the DC polarization) without contributing any direct flux, so the disk must block its direct light while allowing some scattered light to emerge. We are testing a two-component disk model — an optically thick midplane surrounded by an optically thin disk "atmosphere" — that should help resolve this issue. A paper in preparation will discuss these and other results related to our models for beta Lyr.
|
