5. A hierarchical triple system
At a projected separation of only 3" (1400 AU), VLA 1 and 2 are likely to be somehow related. Yet, the two sources are rather different, VLA 1 (which appears to be a very close binary) being very reddened and having significant outflow, while VLA 2 shows no optical evidence of outflow and is much less reddened. Most puzzling is the location of both stars outside a dark ridge which in the cases of other young stars normally suggest the presence of a flattened dense envelope (e.g. Padgett et al. 1999).
We can explain these features if we assume that the three stars were born together and until recently formed a non-hierarchical triple system located inside the dark lane. It is well known from the statistical theory of three-body interactions (e.g. Monaghan 1976, Anosova 1986, Valtonen & Mikkola 1991) that such systems are inherently unstable, and within about a hundred crossing-times will eject a member, while the remaining components contract to form a closer binary. The ejected member may or may not escape entirely, depending on the details of the interaction. If the crossing time is , where R is a characteristic length scale for the system in AU, M is the total system mass in and is in yr (Anosova 1986), and we assume the non-hierarchical triple members were formed within 150 AU of each other with a total mass of 2 (e.g. 1.0, 0.6, 0.4 ), then statistically a member is ejected after 22,000 yr. This is comparable with the dynamical age of the giant HH 111 outflow of about 25,000 yr, if one assumes a mean flow velocity of 150 km s-1 (Reipurth et al. 1997a).
Assuming VLA 2 is moving with a mean transverse velocity of 2 km s-1, then the ejection took place 2700 yr ago. Since the VLA 1 binary has only travelled one quarter of the distance of VLA 2, it follows that it must be 4 times more massive, consistent with statistical studies which show that usually the component of lowest mass is ejected. It seems probable that VLA 2 will eventually escape, since with a minimum separation of 1400 AU the gravitational binding must be feeble, especially in view of the influence of large nearby cloud structures.
That the above numbers fit together so well is obviously fortuitous given the stochastic nature of the process, and they merely serve to show that a plausible choice of parameters could be consistent with the observed facts. Astrometry a few centuries from now could test this conjecture.
It is interesting to note that VLA 2/Star B is now located above the flattened envelope around its supposed birthplace, whereas VLA 1/Star A is below and presumably slowly moving out behind it. The difference in extinction is obvious from Figs. 1 and 2. While the stars almost certainly carry their inner circumstellar disks with them, the large scale envelope, which has served as a reservoir to feed the disks, is left behind. Consequently, the supply of material for both the build-up of the stars and the great outflow activity will gradually be choked off. Formation of a hierarchical triple system, especially if the ejected member eventually escapes, may thus help to terminate the main accretion phase of its members .
© European Southern Observatory (ESO) 1999
Online publication: December 2, 1999