IX. MODELS OF INNER ENGINES

The Fireball model tells us how GRBs operate. However, it does not answer the most interesting astrophysical question: what produces them? which astrophysical process generates the energetic ultrarelativistic flows needed for the Fireball model? Several observational clues help us answer these questions.

Energy: The total energy involved is large ~ 10⁵¹ ergs, a significant fraction of the binding energy of a stellar compact object. The "inner engine" must be able to generate this energy and accelerate ~ 10^-5 M (or the equivalent in terms of Poynting flux) to relativistic velocities.

Collimation: Most GRBs are collimated with typical opening angles 1° < < 20°. The "inner engine" must be able to collimate the relativistic flow.

Long and Short Bursts: The bursts are divided to two groups according to their overall duration. Long bursts with T > 2 sec and short ones with T < 2 sec. As the duration is determined by the inner engine this may imply that there are two different inner engines.

Rates: GRBs take place once per 3^. 10⁵ yr per galaxy. GRBs are very rare at about 1/3000 the rate of supernovae.

Time Scales: The variability time scale, t, is at times as short as 1ms. The overall duration (of long GRBs), T, is of the order of 50 sec. According to the internal shocks model these time scales are determined by the activity of the "inner engine". t ~ 1 ms suggests a compact object. T ~ 50 sec is much longer than the dynamical time scale, suggesting a prolonged activity. ¹². This requires two (or possibly three [270, 330]) different time scales operating within the "inner engine". This rules out any "explosive" model that release the energy in a single explosion.

These clues, most specifically the last one suggest that GRBs arise due to accretion of a massive (~ 0.1 m) disk onto a compact object, most likely a newborn black hole. A compact object is required because of the short time scales. Accretion is needed to produce the two different time scales, and in particular the prolonged activity. A massive (~ 0.1 m) disk is required because of the energetics. Such a massive disk can form only simultaneously with the formation of the compact object. This leads to the conclusions that GRBs accompany the formation of black holes. This model is supported by the observations of relativistic (but not as relativistic as in GRBs) jets in AGNs, which are powered by accretion onto black holes.

An important alternative to accretion is Usov's model [412, 413] in which the relativistic flow is mostly Poynting flux and it is driven by the magnetic and rotational energies of a newborn rapidly rotating neutron star.

¹² The ratio t / T << 1 for short bursts as well [269]. Back.