3. Formation of an Accretion Disk.

Bonnet-Bidaud and van der Klis (1979) showed that even if the accretion is from the stellar wind, the mass accretion rate will be large enough to require an accretion disk to remove angular momentum from infalling material. Tjemkes et al. (1986) used a simple geometric model to analyze the optical light curve and found that ellipsoidal variations alone were unable to account for the observed amplitude. They obtained a satisfactory description of the light curves by including the effects of X-ray heating of the companion and optical emission from a Roche lobe filling accretion disk ( R_disk $\approx$ 2.4 x 10¹¹ cm). The average V band light curve is shown in Figure 3. Khruzina and Cherepaschchuk (1986) similarly found that an accretion disk with R_disk = 2.44$\pm$0.84 x 10¹¹ cm was required and that the companion almost filled its Roche lobe, having a filling factor $\mu$ = 0.995$\pm$0.005. Equation 1.8 gives

$$\approx$$ (83)

while the distance from the neutron star to the L₁ point is

$$\approx$$ (84)

which is larger than R_circ as expected. This is because the material in the disk will spread towards both larger and smaller radii than R_circ as the angular momentum shed by the accreting matter is transported to the outer parts of the disk.

The existence of an accretion disk is also implied by the presence of 40 mHz quasi-periodic oscillations (QPO). These were discovered by Tennant (1988) in EXOSAT ME data. They were also present in Ginga data (Takeshima et al. (1991)). These are believed to be due to the interaction between the accretion disk and the magnetosphere.

Figure 1: Average V band light curve of Cen X-3 (Tjemkes et al. (1986)). The solid curve includes the effects of X-ray heating and an accretion disk while the dashed curve only includes ellipsoidal variations.