Annu. Rev. Astron. Astrophys. 1988. 36: 539-598
Copyright © 1998 by . All rights reserved

Next Contents Previous


Observations show that an AGN jet undergoes a huge expansion at the exit from the inner core. In a few parsecs, its radius multiplies by a thousand or more. Afterwards it recollimates to a conical structure with a small opening angle, without apparently going through any dramatic event, and travels megaparsec distances while maintaining its directionality. Instead, laboratory experiments with fluids indicate that both dense and light flows (with respect to the external medium) expand at the sonic speed and suffer various types of fluid instabilities, developing vortices and internal shocks. After a few scale lengths, collimation is lost, velocities become subsonic, and flows are disrupted (van Dyke 1982). For light jets, disruption is due primarily to matter entrainment and mixing, and for dense jets, to thermal expansion.

This morphological difference initially prompted the idea that extragalactic jets were highly supersonic, freely expanding flows in underdense atmospheres with a negative density gradient. As shown in laboratory experiments (Thompson 1972), after exiting the nozzle, the flow expands with a bending of streamlines fixed by the local Mach number Mj; for Mj >> 1, the opening angle theta approx (Mj)-1. On the other hand, the jet radial evolution emerging from high-resolution observations is much more complex than expected in freely expanding jets. Jets must be confined, but at the same time the interaction with the external gas is nondestructive. Incidentally, we mention that "ballistic confinement" of a stream of dense aligned bullets has been discarded on the basis of global energetics, as it would require kinetic luminosities that are too large and would imply fast deceleration (Pacholczyk 1977).

"Thermal pressure confinement" by interstellar and intergalactic matter is the most natural possibility, as already mentioned in the Introduction. Minimum jet pressure is estimated from the equipartition argument in the synchrotron model, while external thermal pressures can be evaluated from X-ray observations. As discussed by Feretti et al (1995), internal pressure in FR I jets is below external pressure. The situation is more complicated for FR II powerful jets. For instance, knots and filaments in the jet of M87 are out of pressure equilibrium and overpressured by a factor of ~ 10 with respect to the external medium, although these may simply be transient structures, given the relatively short lifetime of this jet (Bicknell & Begelman 1996). Also, the projection effect and filling factor may affect the estimates of internal pressure when applying the equipartition argument to jets that point to the observer. Finally, pressure confinement models require a well-tuned external pressure profile to maintain small opening angles over several decades of length scales. A related possibility is "inertial confinement" by a cold ambient medium, in which ram pressure is used to oppose jet expansion (Icke et al 1991).

External magnetic field confinement can be provided by a large-scale intergalactic field frozen in the inflowing plasma. This field corresponds to the poloidal fields assumed in all MHD models of jet acceleration from disks: The plasma collapsing into the central black hole deforms the fields into an hourglass shape that reaches an asymptotic cylindrical geometry outside the core (Heyvaerts & Norman 1989). However, laboratory plasma experiments do not allow much confidence in the stability of this field structure. A virial theorem argument shows they are highly unstable with regard to interchange and diffusion.

A more satisfactory approach seems to be self-confinement by internal magnetic fields. As noted in the previous section, pinching toroidal magnetic fields are consistently generated by differential rotation of the disk at the base of the jet and advected along the rotation axis. We discuss below how a backflow is formed at the working surface, where the advancing supersonic head impinges upon the confining medium. A shear layer is formed around the jet where instabilities and turbulent field amplification and/or dissipation play a major role in the dynamics, leading to mass entrainment and mixing (De Young 1996). A magnetized overpressured region around the jet is then formed that may in fact be the real confining agent.

Next Contents Previous