next up previous
Next: 5 Observational tests of Up: Abstract Previous: 3 Constraints upon the

4 Eigenmodes confined to the inner part of the disc

In order to examine the characteristics of the m=1 eigenmodes confined to the inner part of the disc, we have numerically solved Eqs. (16) and (17) with boundary conditions for a wide range of parameter values: We surveyed eigenmodes for a deformation factor of , radiative parameters of and , disc temperature of , density gradient index of , and central stars with spectral types of B0V and B5V. Since we found that the obtained eigenmodes are roughly similar for different values of disc parameters and , we restrict our attention to cases with , , and .

 

 


Fig.2: Distribution of the period and the propagation-region width of the fundamental mode confined to the inner part of the disc around a a B0V star and b a B5V star. Thick contours denote the oscillation period in units of years and thin contours denote . The interval of contours is 0.5dex. Scales for are also shown for a and b . The ragged boundary denotes the critical curve below which no confined eigenmode is found

Figure 2 shows the distribution of the two characteristic quantities, the period of the oscillation and the size of the propagation region, of the fundamental m=1 eigenmodes confined to the inner part of the disc. Figure 2(a) is for a B0V star and Fig. 2(b) for a B5V star. Other values of parameters are annotated in the figure. In each figure thick contours denote the oscillation period in units of years and thin contours denote , where is the radius of the inner Lindblad resonance. Recall that is a good measure of the width of the propagation region of the eigenmode. The interval of contours is 0.5dex. At the top of each panel we show the scale for for (a) and (b) . The ragged boundary in each figure denotes the critical curve discussed in the previous section; we found no confined eigenmode for the values of and below this line. As mentioned earlier, the raggedness of the critical curve reflects that we surveyed the (, )-space with a resolution of 0.02dex.

 

 


Fig.3: Fundamental mode for a and b . The disc temperature is . The values of the other parameters are indicated in the upper panel. In the upper panel, the solid, dashed, and dash-dotted lines denote , , and , respectively. The location of the inner Lindblad resonance is shown by the vertical bar. In the lower panel, the gray-scale plot denotes the density perturbation in the -plane, while arrows denote the perturbed-velocity vectors (, )

From Fig. 2, we note that the period of the eigenmode is sensitive to the parameters characterizing the rotational deformation and the weak line force. It rapidly decreases with increasing values of or . The dependence of the period on these parameters is similar for different values of and (or equivalently ), except that, as the pressure effect decreases, the critical curve in the (, )-plane moves to the lower left and modes with longer periods appear. Needless to say, the period of the confined mode is independent of the disc outer radius as long as it is located far outside the propagation region of the mode. Consequently, the range of observed periods of the V/R variations places a rather narrow constraint on and . Suppose that the observed periods range from a year to a decade both for B0- and for B5-type Be stars. Suppose also that the radiative effect is dominant in discs of B0-type Be stars, while the rotational effect is dominant in discs of B5-type Be stars. Then, the range of for B0-type Be stars and the range of for B5-type Be stars are, respectively, restricted to and . If a star exhibits a V/R variation with a period longer than a decade, it indicates that the pressure effect in the disc would also be weaker than that adopted in Fig. 2, i.e., and/or .

Now we discuss the characteristics of the eigenmodes confined to the inner part of Be-star discs. Figure 3 presents two examples of the fundamental m=1 modes. Figure 3(a) is for the radiation-effect dominant case for a B0V star, while Fig. 3(b) is for the rotation-effect dominant case for a B5V star. In this example, both modes have the same period of 5 years. The adopted disc temperature is . The values of the other parameters are indicated in the upper panels. In each upper panel, the solid, dashed, and dash-dotted lines denote , , and , respectively. They are normalized such that the maximum value of is unity. The vertical bar denotes the location of the inner Lindblad resonance. Note that and vary as , while varies as . The lower panels show the distribution of these quantities in the inner in the -plane. The circle at the center denotes the location of the star/disc interface. The perturbation pattern as well as the unperturbed disc rotates counterclockwise. The density perturbation is denoted by a gray-scale representation. Arrows superposed on the gray-scale plot are the perturbed-velocity vectors (, ). The length of each arrow is proportional to the strength of the perturbation.

From the upper panels of Fig. 3, we note that in both cases the fundamental m=1 modes are well confined to the inner part of the disc. We also note that the global features of the eigenfunctions are very similar in both cases, except that the region in which the amplitude of the perturbation is large is slightly narrower in the rotation-effect dominant case than in the radiation-effect dominant case. The azimuthal component of the velocity perturbation anticorrelates with the density perturbation , except in the innermost narrow part of the disc. Note that this property, which was also seen in the modes studied in Okazaki (1991), has been shown to cause the observed line-profile variability of the V/R variations (Hummel & Hanuschik 1994, 1996; Hanuschik et al. 1995; Okazaki 1996).



next up previous
Next: 5 Observational tests of Up: Abstract Previous: 3 Constraints upon the



Atsuo Okazaki
平成9年3月15日 (土), 午後 1時19分3秒