Thin shells of dust in a compact universe

I present the first analytical study of gravitational collapse in a compact CMC foliation with $S^3$ spatial topology. The solutions I find, in this context, will be both solutions of Shape Dynamics and General Relativity. The aim is to describe a system undergoing gravitational collapse in Shape Dynamics, so a well-justified and useful simplification is to assume spherical symmetry. This kills all the local gravitational degrees of freedom, but some nontrivial degrees of freedom are recovered by introducing matter. The simplest form of matter is infinitely thin spherical shells of dust, of which I need at least two in order to have a nontrivial dynamics. With a single shell the system is dynamically trivial, but it nevertheless admits a solution which represents a frozen' shell at equilibrium in a globally de Sitter universe. Such a solution is, to my knowledge, new. I am able to solve analytically also the case with two shells, which has a nontrivial dynamics. When the rest mass of one shell is much smaller than the other, the system is suitable to model a compact universe in which one subsystem (the light’ shell) undergoes gravitational collapse while the rest of the matter (the heavy' shell) plays the role of spectator. It turns out that, if the cosmological constant is zero or positive but small, and the rest mass of the two shells are sufficiently different, when the light’ shells collapses the ADM equations become ill-defined and cease to admit a solution. The shape-dynamical description, however, seems still well defined and can be continued past this point, possibly signalling a departure of Shape Dynamics from exact equivalence with General Relativity.