Phase shifts and nonellipsoidal light curves: Challenges from mass determinations in x-ray binary stars

We consider two types of anomalous observations which have arisen from efforts to measure dynamical masses of X-ray binary stars: (1) Radial velocity curves which seemingly show the primary and the secondary out of antiphase in most systems, and (2) The observation of double-waved light curves which deviate significantly from the ellipsoidal modulations expected for a Roche lobe filling star. We consider both problems with the joint goals of understanding the physical origins of the anomalous observations, and using this understanding to allow robust dynamical determinations of mass in X-ray binary systems.;In our analysis of phase-shifted radial velocity curves, we discuss a comprehensive sample of X-ray binaries with published phase-shifted radial velocity curves. We show that the most commonly adopted explanation for phase shifts is contradicted by many observations, and consider instead a generalized form of a model proposed by Smak in 1970. We show that this model is well supported by a range of observations, including some systems which had previously been considered anomalous. We lay the groundwork for the derivation of mass ratios based on our explanation for phase shifts, and we discuss the work necessary to produce more detailed physical models of the phase shift.;In our analysis of non-ellipsoidal light curves, we focus on the very well-studied system A0620-00. We present new VIH SMARTS photometry spanning 1999-2007, and supplement this with a comprehensive collection of archival data obtained since 1981. We show that A0620-00 undergoes optical state changes within X-ray quiescence and argue that not all quiescent data should be used for determinations of the inclination. We identify twelve light curves which may reliably be used for determining the inclination. We show that the accretion disk contributes significantly to all twelve curves and is the dominant source of nonellipsoidal variations. We derive the disk fraction for each of the twelve curves and show that, after correcting for the disk component, these twelve curves point to a consistent inclination. Finally, we consider the very different binary system V4641 Sgr and show that it has some qualitative similarities to A0620-00, suggesting that the phenomena we find in A0620-00 are likely to be widespread.