Astron. Astrophys. 344, 277-281 (1999)

4. Discussion

Despite the large observational efforts made during last years and mainly after the 1991-1992 outburst, the Hen 3-640/A1118-61 system is still poorly understood.

Actually, the orbital period of the system is unknown. Corbet's pulse period/orbital period diagram (Corbet, 1986) gives an orbital period estimate of 350 days.

Recently, Reig et al. (1997) have found another correlation between the orbital period of Be/X-ray binaries and the H EW of the optical companion. However in their diagram, the representative point of A1118-61 (assuming the period deduced from the P_spin-P_orb diagram and the maximum H EW) lies definitely out of the best fit of the other points. On the contrary, if we consider the average H EW (70 Å), A1118-61 could have an orbital period of 350 days, in agreement with the Corbet's diagram. If the maximum measured value of the H EW (107 Å) is used, the A1118-61 orbital period deduced by the Reig et al. (1997) relationship is 585 days.

However, none of these periodicities is clearly present both in the X-ray and the H EW time history, that are now relatively well known due to the CGRO-BATSE X-ray monitoring (Bildsten et al., 1997) and our long term optical survey. On the other hand, our data seem to confirm the correlation between the hard X-ray flux and the H EW, suggested by Coe et al. (1994). However, the long time interval ( 16 years) between the two recorded outbursts of A1118-61 suggests that the neutron star is usually weakly interacting with Hen 3-640. Thus, apart from the outburst phase, the long term spectral variability of the optical star is most probably intrinsic and not connected with the presence of the compact companion. Motch et al. (1988) and Coe et al. (1994) already suggested that the neutron star is orbiting just at the boundary of a large extended envelope surrounding Hen 3-640 and is continuously accreting matter at low rate from this environment.

The long term H EW variability is compatible with a smooth rise of the star activity from a quiescent state to the outburst, while the post-outburst data demonstrate a short time-scale passage to the quiescent state. The H increase can be due either to increasing UV radiation flux of Hen 3-640 or to higher UV absorption efficiency in a larger envelope surrounding the optical star. In the framework of the experimental evidence discussed in this paper, the second mechanism seems the most probable. Actually, the variability of the amount of circumstellar matter is suggested by the dectected variability of the H/H flux ratio (see Fig. 5): the 1985 value (6) indicates a strong residual circumstellar extinction, while the post-outburst value (3) is compatible, inside the uncertainties, with a standard "case A" nebular emission, indicating that we are looking the star only through the interstellar matter. The 1997 value (5) shows that the circumstellar extinction is again increasing.

Fig. 5. The H/H flux ratio of Hen 3-640 from 1985 to 1997.

Many mechanisms of matter accumulation around Be stars have been proposed in the past in order to explain the secular variability of the emission line fluxes (see, e.g., Underhill & Doazan, 1982; Slettebak et al., 1992, and references therein). Recently, Marlborough (1997) has suggested that a non-linear amplification of non-radial pulsation should occur in Be stars, resulting in the ejection of a significant amount of matter in the line of sight. On the other hand, Kakouris & Moussas (1998) have studied a physical mechanism which is able to generate in a short time a steady-state shell around hot stars.

Whatever the process of matter accumulation around Hen 3-640 may be, the compact object is orbiting in an increasing density environment and the X-ray flux gradually increases due to the higher mass transfer rate. But, when density around the compact object reaches a value high enough to generate a steady accretion disk, the X-ray flux sharply increases. This higher energy input should be able to transfer to the circumstellar envelope an energy equal to the gravitational binding energy, and all the matter is swept out in a very short time. Actually, our data taken immediately after the outburst shown the presence of a significant matter flux leaving the system at high velocity (Villada et al., 1992; Polcaro et al., 1993).

If this scenario is proven true, we cannot expect a real periodicity in the X-ray outburst, that will be not linked in any way to the system orbital parameters. We can thus foresee that the next outburst will happen after an interval connected to the matter accumulation time-scale: a 16 year interval between the two recorded outbursts is fully compatible with this hypothesis. If this scenario is valid, it is not surprising that the periodicity deduced from the pulse-period/orbital-period relationship has not been detected. Actually, we cannot obtain information about the orbital period from Corbet's diagram, when the outburst mechanism is not connected with the periastron passage of the compact object.

Therefore we could have recognized a new sub-class of "atypical" X-ray/Be systems in which X-ray outbursts are not triggered by periastron passage mechanism like for instance in the so-called "typical" X-ray/Be system (A0535+26/ HDE245770-like). One possibility of the formation of "atypical" or "typical" X-ray/Be systems could be related to the explosion of a supernova in a progenitor medium-mass (10-20 ) early type binary system: the "atypical" X-ray/Be system would be the remnant of a symmetric SN explosion (low eccentricity: e0.3); the "typical" X-ray/Be system would be the remnant of an asymmetric SN explosion (0.3e0.8). The possibility of such events has been discussed and demonstrated by Giovannelli et al. (1994b; 1996) and in references therein.

Of course, the lack of continous monitoring before of the 1991-1992 outburst does not allow us to completely rule out the possibility that the Hen 3-640/A1118-61 system has a very long orbital period. However, such a period will not satisfy Corbet's relationship.

The final answer will come only after a long monitoring of this interesting system: we suggest that the next X-ray outburst of A1118-61 will start when the Hen 3-640 H EW will reach again values 100 Å.

© European Southern Observatory (ESO) 1999

Online publication: March 10, 1999
helpdesk.link@springer.de