Jacobi Geometry and Hamiltonian Mechanics: the Unit-Free Approach

TL;DR

This paper develops a unit-free categorical framework for Jacobi and Poisson geometries, extending Hamiltonian mechanics to a dimensionless setting, with implications for physical quantities and dimensional analysis.

Contribution

It introduces a novel categorical approach to Jacobi and Poisson geometries, enabling a unit-free formulation of Hamiltonian mechanics and related structures.

Findings

Jacobi geometry is a natural unit-free extension of Poisson geometry.

Hamiltonian mechanics can be reformulated in a unit-free categorical framework.

The approach facilitates a formal treatment of physical quantities and dimensional analysis.

Abstract

We present a systematic treatment of line bundle geometry and Jacobi manifolds with an application to geometric mechanics that has not been noted in the literature. We precisely identify categories that generalise the ordinary categories of smooth manifolds and vector bundles to account for a lack of choice of a preferred unit, which in standard differential geometry is always given by the global constant function $1$. This is what we call the `unit-free' approach. After giving a characterisation of local Lie brackets via their symbol maps we apply our novel categorical language to review Jacobi manifolds and related notions such as Lichnerowicz brackets and Jacobi algebroids. The main advantage of our approach is that Jacobi geometry is recovered as the direct unit-free generalisation of Poisson geometry, with all the familiar notions translating in a straightforward manner. We then apply this formalism to the question of whether there is a unit-free generalisation of Hamiltonian mechanics. We identify the basic categorical structure of ordinary Hamiltonian mechanics to argue that it is indeed possible to find a unit-free analogue. This work serves as a prelude to the investigation of dimensioned structures, an attempt at a general mathematical framework for the formal treatment of physical quantities and dimensional analysis.