Australia Telescope National Facility

The Active Algol Binary KZ Pavonis
E. Budding,
S. C. Marsden ,
O. B. Slee
, PASA, 18 (2), in press.

Next Section: Conclusions
Title/Abstract Page: The Active Algol Binary
Previous Section: Observations
Contents Page: Volume 18, Number 2

Discussion

For the previously cited case of CF Tuc (Budding et al., 1999) spot radii were found to be $\sim$ 16 $^\circ$ . Even if such spots were present on the active component of KZ Pav, which is not established, they would not be optically distinct, as argued above. In the radio region, however, the amplitude of the main variation for CF Tuc was $\sim$ 2 mJy, i.e. an order of magnitude greater than for KZ Pav, resulting in the corresponding difference in radio brightness temperatures of the two systems. While we thus cannot rule out a low level of microwave flux from the cool secondary of KZ Pav due to RS CVn-type gyrosynchrotron emission, there may be other effects at play in the special context of Algols (Gunn et al., 1999). The previously demonstrated Algol status of KZ Pav (Walker & Budding, 1996) is confirmed in the light curve solution of Table 4, which essentially repeats the main results of Walker & Budding (1996). The irregularities of the second half of the light cycle may perhaps be associated with mass transfer. Walker & Budding interpreted thus the delayed minima of the binary, and, using reasonable estimates for key quantities, obtained a value of

$\Delta P/P \sim10^{-10}$ for the current rate of period variation. With Walker & Budding's epoch of HJD 2447666.1028, some 3573 cycles have elapsed before the epoch of the present data set, which, using the zero phase correction of Table 4, occurs at 2451060.0247. The corresponding mean period for the interval in question is then 0.949880 days. In some 13490 periods, between the epoch at which Mallama's period was representative and that of the foregoing period, therefore, the period has increased by 0.0000032 days, or

$\Delta P/P \approx10^{-10}$ for the current rate of period variation. This variation ties in reasonably well with the conservative Case B model considered previously by Walker & Budding (1996). In such a regime one may write, for the mass transfer rate

$\begin{displaymath} \dot{m_1}/m_1 = - 3 \eta s/\left( R_1 \left[ 1 + 6 \eta (1-2x)/(1-x) -0.27 \eta/f(x)\right] \right) , \end{displaymath}$

(2)

where $\dot{m_1}$ is the rate of mass loss of the mass-losing star, of mass m₁, $\eta$ is the density of the surface layer of the mass-losing component as a fraction of its mean density, R₁ is the mean radius of this star, x is the value of m₁ expressed in terms of the mass of the entire system, f(x) is the formula for the mean relative radius of the Roche lobe (cf. Plavec, 1968), and s is the rate of expansion of the mass-losing component. If we substitute in the appropriate numbers, as in Walker & Budding, we find a representative mass loss rate of

$1.9\times10^{-8} M_{\hbox{$\odot$}}$ per year.

**Figure 7:** Difference between observations and optimal model fit. Note the systematic enhancement around phase 300 $^\circ$ ( $\approx$ 0.8).
$\begin{figure} \centerline{ \psfig{file=kzdif.ps,height=8cm,width=11cm,angle=-90} } \end{figure}$

Considering the potential energy that the transferred mass picks up in dropping from the inner Lagrangian point L₁(

$\sim2.7R_{\hbox{$\odot$}}$ above the primary's centre) to the surface of the primary, $\sim$ 0.3

$G\dot{m_1}m_2/r_2$ , we would thence infer a figure of about 0.15

$L_{\hbox{$\odot$}}$ , or

5.5 x 10²⁵W. The systematic enhancement in Fig. 7 at around phase 300 $^\circ$ ( $\sim$ 1% of the 5

$L_{\hbox{$\odot$}}$ of system light, c.f. Table 1), that can be associated with this transferred energy in the visible range of the spectrum, would thus be about 30% of the available power. The optical data can thus be seen as reasonably self-consistent if, instead of appearing as as a direct, black-body-like emission from the photospheric impact region (`hot spot'), an appreciable fraction of the transferred mechanical energy of the stream is transferred into dissipative processes in plasma accreting around the hot spot's vicinity. It is feasible that the small peak in the radio emission around phase 0.8 is associated with energy release processes related to such mass transfer, but, in view of the low significance of the radio flux variation with phase, it is unrealistic to develop details from the present evidence.

Next Section: Conclusions
Title/Abstract Page: The Active Algol Binary
Previous Section: Observations
Contents Page: Volume 18, Number 2

ASKAP

Public