Evolution and Regularisation of Vacuum Brill Gravitational Waves in Spherical Polar Coordinates

TL;DR

This thesis demonstrates that vacuum Brill gravitational waves universally collapse into singularities or black holes, using numerical and analytical methods, and introduces techniques for stable evolution in spherical coordinates.

Contribution

It provides a comprehensive numerical and analytical framework for regularising and evolving vacuum Brill waves, showing their inevitable collapse and instability of Minkowski space under such perturbations.

Findings

All vacuum Brill waves collapse to black holes.

No critical black hole mass scaling observed in vacuum systems.

Stable numerical evolution achieved through advanced discretisation and regularity conditions.

Abstract

In this thesis the universal collapse of vacuum Brill waves is demonstrated numerically and analytically. This thesis presents the mathematical and numerical methods necessary to regularise and evolve Brill Gravitational Waves in spherical polar coordinates. A Cauchy ADM formulation is used for the time evolution. We find strong evidence that all IVP formulations of pure vacuum Brill gravitational waves collapse to form singularities/black holes, and we do not observe critical black hole mass scaling phenomena in the IVP parameter phase space that has been characterised in non-vacuum systems. A theoretical framework to prove this result analytically is presented. We discuss the meaning of Brill metric variables, the topology of trapped surfaces for various scenarios, and verify other results in the field related to critical values of initial value parameters and black hole formation approaching spatial infinity. The instability of Minkowski (flat) space under Brill wave and more general perturbations is demonstrated. The main numerical tools employed to achieve a stable evolution code are (1) derivation of appropriate regularity conditions on the lapse function and metric function q, (2) the move to a 4th order correct discretisation scheme with appropriate boundary conditions, (3) the use of exponential metric terms, (4) an understanding of the right mix of free versus constrained evolution and (5) the development of appropriate numerical techniques for discretisation and differencing to reduce numerical error, along with a characterisation of condition numbers.