RECURRENT SUPERSOFT X-RAY SOURCES

An analytical approach is presented to estimate the mass in the envelope of a condensed white dwarf (WD) and henceforth the mass of the WD itself as both are directly correlated and to estimate the constant mass accretion rate from a simple repeating on-off shape of the X-ray lightcurve for recurrent supersoft X-ray sources. As supersoft X-ray sources have been detected in substantial number by Einstein and ROSAT in the Magellanic Clouds (similar to 11), the Galaxy (similar to 7) and M31 (>15) and these objects have been found in part to be of a transient or even recurrent nature it is interesting to relate observational parameters to system parameters in order to explore these unique galactic and extragalactic laboratories. It is shown, that in this approach the critical envelope mass M(env)(crit) is related to the accretion rate for stable hydrogen burning M(stable), the decay time (or the X-ray on-time) t(decay) and the time of recurrence of subsequent outbursts t(recur) by M(env)(crit) = M(stable) (t(decay) - ((t(decay)(2)\t(recur)) and that the (constant) mass accretion rate M(accr) is given by M(accr) = M(stable) (t(decay)\t(recur)). Applying this approach to the four transient or recurrent supersoff; X-ray sources RXJ0513.96951, RXJ0527.8-6954, RXJ0019.8+2156 and RXJ0045.44154 I deduce constraints for the WD mass and mass accretion rates for these systems (using the relation between envelope mass and WD mass calculated by Fujimoto 1982). This favors high mass WDs (M(WD) similar to 1.3, 1.1-1.3, 1.0-1.3 and 1.35-1.40 M(.)) for these systems. Accretion rates of 7.9 x 10(-8) less than or equal to 1.2 x 10(-7), >8.9 x 10(-8) and 3 x 10(-9) -4 x 10(-7) M(.) yr(-1) are found. This makes RXJO513.9-6951 and RXJ0045.4-4154 strong candidates for the Accretion Induced Collapse, which may happen after an accretion phase of 10(5) - 10(6) and 2 x 10(4) - 2 x 10(7) years, respectively. An alternative possibility is that these are systems with considerable mass accretion (of the order a few x 10(-1) M(.)) which already had occurred over evolutionary timescales. RXJ0527.8-6954 and RXJ0019.8+2156 are more likely candidates for type Ia supernovae which may happen in a few 10(6) years. The deduced system parameters have to be modified in case a significant fraction of the envelope of the WD is ejected during the outburst. Especially a lower WD mass is achieved. For GQ Mus (Nova Muscae 1983) consideration of the ejected mass (similar to 3 X 10(-5) M(.)) leads to a correction of the WD mass from less than or equal to 1.0 - 1.2 M(.) to less than or equal to 0.84 M(.). In the supersoft transient and recurrent sources mass ejection (during the outburst as a consequence of the thermonuclear runaway) is not expected to be significant due to the large WD mass and the high accretion rates involved. Such a mass ejection has also not been observed. A constant mass loss due to a wind reducing the effective mass accretion rate may well be present. But such a wind loss will not affect the physical parameters deduced for the condensed WD in supersoft systems. Recurrent supersoft sources may be similar to recurrent novae (in terms of high accretion rates >10(-8) M(.) yr(-1) and large WD masses >1 M(.)) but may differ due to lack of significant mass ejection during the outburst. This may be due to higher accretion rates close to the value of steady-state nuclear burning of a few x 10(-7) M(.) yr(-1).