Low Angular Momentum Accretion in the Collapsar: How Long Can a Long GRB Be?

The collapsar model is the most promising scenario to explain the huge release of energy associated with long-duration gamma-ray bursts (GRBs). Within this scenario GRBs are believed to be powered by accretion through a rotationally supported torus or by fast rotation of a compact object. In both cases rotation of the progenitor star is a key property, because it must be high enough for the torus to form, the compact object to rotate very fast, or both. Here, we check what rotational properties a progenitor star must have in order to sustain torus accretion over relatively long activity periods such as observed in most GRBs. We show that simple, often cited, estimates of the total mass available for torus formation and consequently the duration of a GRB are only upper limits. We revise these estimates by taking into account the long-term effect that as the compact object accretes, the minimum specific angular momentum needed for torus formation increases. This in turn leads to a smaller fraction of the stellar envelope that can form a torus. We demonstrate that this effect can lead to a significant (an order of magnitude) reduction of the total energy and overall duration of a GRB event. This of course can be mitigated by assuming that the progenitor star rotates faster then we assumed. However, our assumed rotation is already high compared to observational and theoretical constraints. We estimate the GRB duration times, first by assuming a constant accretion rate, an also by explicitly calculating the free-fall time of the gas during the collapse. We discuss the implications of our results.