Differential galois theory and mechanics

TL;DR

This paper extends classical Galois theory to systems of algebraic partial differential equations using new methods from differential algebra, geometry, and algebraic geometry, with applications to mechanics and classical physics problems.

Contribution

It develops a general differential Galois theory for algebraic pseudogroups, combining differential algebra, geometry, and invariant derivations, and demonstrates its usefulness in mechanics.

Findings

Revisits shell and chain theories in mechanics.

Provides new methods for integrating Hamilton-Jacobi equations.

Connects differential Galois theory with classical mechanics applications.

Abstract

The classical Galois theory deals with certain finite algebraic extensions and establishes a bijective order reversing correspondence between the intermediate fields and the subgroups of a group of permutations called the Galois group of the extension. It has been the dream of many mathematicians at the end of the nineteenth century to generalize these results to systems of algebraic partial differential (PD) equations and the corresponding finitely generated differential extensions, in order to be able to add the word differential in front of any classical statement. The achievement of the Picard-Vessiot theory by E. Kolchin between 1950 and 1970 is now well known. The purpose of this paper is to sketch the general theory for such differential extensions and algebraic pseudogroups by means of new methods mixing differential algebra, differential geometry and algebraic geometry. As already discovered by E. Vessiot in 1904 through the use of automorphic systems, a concept never acknowledged, the main point is to notice that the Galois theory (old and new) is mainly a study of principal homogeneous spaces (PHS) for algebraic groups or pseudogroups. Hence, all the formal theory of PD equations developped by D.C. Spencer around 1970 must be used together with modern algebraic geometry, in particular tensor products of rings and fields. However, the combination of these new tools is not sufficient and we have to create the analogue for Lie pseudogroups of the so-called invariant derivations introduced by A. Bialynicki-Birula after 1960 in the study of algebraic groups and fields with derivations. We shall finally prove the usefulness of the resulting differential Galois theory through striking applications to mechanics, revisiting shell theory, chain theory, the Frenet-Serret formulas and the integration of Hamilton-Jacobi equations.