Singularities, trapped sets, and cosmic censorship in asymptotically flat space-times (Q1209242)

The authors study weakly asymptotically simple and empty spacetimes. They introduce the notion of a ``regular partial Cauchy surface'' \(S\). Such a surface is ``partially asymptotically predictable'' [\textit{F. J. Tipler}, Phys. Rev. Lett. 37, 879 (1976)] and satisfies a condition which ensures that the past of the singularity intersects \(S\) in a compact set. They also introduce the ``trapped set condition''. This condition demands that every future inextensible, future incomplete null geodesic which is enclosed in the interior of the domain of dependence of \(S\) has a past which intersects a closed trapped surface or a trapped point. Under this condition and energy and genericity conditions they show that a weakly asymptotically simple and empty spacetime which contains a regular, partial Cauchy surface is future asymptotically predictable. The reader should note that they denote future and past null boundary of spacetime by \(J^ \pm\) instead of \({\mathcal I}^ \pm\). More importantly, the assumptions of their theorem don't harmonise from a physical point of view. Asymptotically simple and empty spacetimes are models for isolated object such as single stars or single black holes. On the other hand, the trapped set conditions seems to imply that the matter concentration is high everywhere. This can only be expected in cosmological models, i.e. the complementary case. I even doubt that it is possible to construct a weakly asymptotically simple and empty spacetime in which the trapped set condition is satisfied.