Hamiltonian Renormalisation II. Renormalisation Flow of 1+1 dimensional free scalar fields: Derivation

TL;DR

This paper introduces a new Hamiltonian renormalisation flow method for 1+1 dimensional free scalar fields, addressing issues in the traditional approach by proposing a direct flow that better preserves continuum properties.

Contribution

It develops a direct Hamiltonian renormalisation flow that overcomes limitations of the path integral induced flow, demonstrated on the solvable 1+1D free scalar field model.

Findings

The direct flow avoids the projection issues of the path integral induced flow.

Application to the 2D Klein-Gordon field illustrates the differences between the flows.

The method sets the stage for generalisations to higher-dimensional models.

Abstract

In the companion paper we motivated a renormalisation flow on Osterwalder-Schrader data (OS-data) consisting of 1. a Hilbert space, 2. a cyclic vacuum and 3. a Hamiltonian annihilating that vacuum. As the name suggests, the motivation was via the OS reconstruction theorem which allows to reconstruct the OS data from an OS measure satisfying (a subset of) the OS axioms, in particular reflection positivity. The guiding principle was to map the usual Wilsonian path integral renormalisation flow onto a flow of the corresponding OS data. We showed that this induced flow on the OS data has an unwanted feature which disqualifies the associated coarse grained Hamiltonians from being the projections of a continuum Hamiltonian onto vectors in the coarse grained Hilbert space. This motivated the definition of a direct Hamiltonian renormalisation flow which follows the guiding principle but does not suffer from the afore mentioned caveat. In order to test our proposal, we apply it to the only known completely solvable model, namely the case of free scalar quantum fields. In this paper we focus on the Klein Gordon field in two spacetime dimensions and illustrate the difference between the path integral induced and direct Hamiltonian flow. Generalisations to more general models in higher dimensions will be discussed in our companion papers.