[ Home | Research | Papers | Thesis | Interests | Fun And Games ]

The Manifestly Gauge Invariant
Exact Renormalisation Group

Abstract

We construct a manifestly gauge invariant Exact Renormalisation Group (ERG) whose form is suitable for computation in SU(N) Yang-Mills theory, beyond one-loop. An effective cutoff is implemented by embedding the physical SU(N) theory in a spontaneously broken SU(N|N) Yang-Mills theory.

To facilitate computations within this scheme, which proceed at every step without fixing the gauge, we develop a set of diagrammatic techniques. As an initial test of the formalism, the one-loop SU(N) Yang-Mills beta-function, beta_1, is computed, and the standard, universal answer is reproduced.

It is recognised that the computational technique can be greatly simplified. Using these simplifications, a partial proof is given that, to all orders in perturbation theory, the explicit dependence of perturbative beta-function coefficients, beta_n, on certain non-universal elements of the manifestly gauge invariant ERG cancels out. This partial proof yields an extremely compact, diagrammatic form for the surviving contributions to arbitrary beta_n, up to a set of terms which are yet to be dealt with. The validity of the compact expression is reliant on an unproven assertion at the third loop order and above.

Starting from the compact expression for beta_n, we specialise to beta_2 and explicitly construct the set of terms yet to be dealt with. From the resulting diagrammatic expression for beta_2, we extract a numerical coefficient which, in the limit that the coupling of one of the unphysical regulator fields is tuned to zero, yields the standard, universal answer. Thus, we have performed the very first two-loop, continuum calculation in Yang-Mills theory, without fixing the gauge.

Thesis pdf (3178kB)


Last modified: Thu Aug 11 11:35:48 BST 2005