We consider the problem of building inhomogeneous cosmological models in scalar-tensor theories of gravity. This starts by splitting the field equations of these theories into constraint and evolution equations, and then proceeds by identifying exact solutions to the constraints. We find exact, closed form expressions for geometries that correspond to the initial data for cosmological models containing regular arrays of point-like masses. These solutions extend similar methods that have recently been applied to Einstein's equations, and provides sufficient initial conditions to perform numerical integration of the evolution equations. We use our new solutions to study the effects of inhomogeneity in cosmologies governed by scalar-tensor theories of gravity, including the spatial inhomogeneity allowed in Newton's constant. Finally, we compare our solutions to their general relativistic counterparts, and investigate the effect of changing the coupling constant between the scalar and tensor degrees of freedom.

Motivated by the existence of hierarchies of structure in the Universe, we present four new families of exact initial data for inhomogeneous cosmological models at their maximum of expansion. These data generalise existing black hole lattice models to situations that contain clusters of masses, and hence allow the consequences of cosmological structures to be considered in a well-defined and non-perturbative fashion. The degree of clustering is controlled by a parameter $\lambda$, in such a way that for $\lambda \sim 0$ or $1$ we have very tightly clustered masses, whilst for $\lambda \sim 0.5$ all masses are separated by cosmological distance scales. We study the consequences of structure formation on the total net mass in each of our clusters, as well as calculating the cosmological consequences of the interaction energies both within and between clusters. The locations of the shared horizons that appear around groups of black holes, when they are brought sufficiently close together, are also identified and studied. We find that clustering can have surprisingly large effects on the scale of the cosmology, with models that contain thousands of black holes sometimes being as little as 30% of the size of comparable Friedmann models with the same total proper mass. This deficit is comparable to what might be expected to occur from neglecting gravitational interaction energies in Friedmann cosmology, and suggests that these quantities may have a significant influence on the properties of the large-scale cosmology.

We consider the problem of constructing cosmological solutions of the Einstein-Maxwell equations that contain multiple charged black holes. By considering the field equations as a set of constraint and evolution equations, we construct exact initial data for N charged black holes on a hypersphere. This corresponds to the maximum of expansion of a cosmological solution, and provides sufficient information for a unique evolution. We then consider the specific example of a universe that contains eight charged black holes, and show that the existence of non-zero electric charge reduces the scale of the cosmological region of the space. These solutions generalize the Majumdar-Papapetrou solutions away from the extremal limit of charged black holes, and provide what we believe to be some of the first relativistic calculations of the effects of electric charge on cosmological backreaction.

We construct exact initial data for closed cosmological models filled with regularly arranged black holes in the presence of $\Lambda$. The intrinsic geometry of the 3-dimensional space described by this data is a sum of simple closed-form expressions, while the extrinsic curvature is just proportional to $\Lambda$. We determine the mass of each of the black holes in this space by performing a limiting procedure around the location of each of the black holes, and then compare the result to an appropriate slice through the Schwarzschild-de Sitter spacetime. The consequences of the inhomogeneity of this model for the large-scale expansion of space are then found by comparing the lengths of curves in the cosmological region to similar curves in a suitably chosen Friedmann-Lemaitre-Robertson-Walker (FLRW) solution. Finally, we locate the positions of the apparent horizons of the black holes, and determine the extremal values of their mass, for every possible regular arrangement of masses. We find that as the number of black holes is increased, the large-scale expansion of space approaches that of an FLRW model filled with dust and $\Lambda$, and that the extremal values of the black hole masses approaches that of the Schwarzschild-de Sitter solution.