Nonperturbative Studies of Quantum Gravity

Abstract

One of several possibilities to construct a quantum theory of gravity is employing the Feynman path integral. This approach is plagued by some problems: the integration measure is not uniquely defined, the Einstein-Hilbert action unbounded, and perturbation theory nonrenormalizable. To make the path integral tractable one can approximate the continuous geometry of spacetime by a simplicial complex. The edge lengths of this lattice are considered as the dynamical degrees of freedom and Regge calculus is applied. In this work, numerical simulations using the Regge-Einstein action and a "compact" action show the occurence of a phase transition. The strength of this transition, separating a well-defined phase with finite expectation values from an ill-defined phase, is weaker for the compact action, which might be important for the continuum limit. To analyze the interaction mechanism of this formulation of quantum gravity, correlation functions between geometrical quantities like edge lengths, volume elements, and local curvatures have been computed. Our results for the two-point functions seem to prefer exchange particles with an effective mass. To ease treatment of quantum gravity a new approach is proposed consisting in a transformation of the path integral to the partition function of a spin system. This facilitates analytical and numerical calculations considerably. First results for the phase structure in two as in four dimensions are presented and indicate promising similarities to the original Regge theory.