Lorenz gauge and the Poisson bracket in canonical electromagnetism

D. Jaffino Stargen · 2024

Read Lorenz gauge and the Poisson bracket in canonical electromagnetism by D. Jaffino Stargen (2024) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2412-05350.txt.

Introduction

Stargen examines a subtle point in electromagnetism in a 2024 physics paper. Textbooks often assume that the field's vector potential and its partner momenta obey the standard bracket rules of classical mechanics. Yet the parts of that potential are tied down by gauge conditions, so the assumption looks shaky.

The author works through the Lorenz gauge, a standard choice that fixes those ties. After the gauge is applied, only some field variables remain free. Imposing the bracket rules on those free variables alone turns out to give the same result as the naive textbook assumption. The match, the author argues, comes from a lucky alignment between the gauge conditions and the principle of relativity. The note clears up a point that many treatments pass over in silence.

Catalog record

CREATORD. Jaffino Stargen
YEAR2024
COLLECTIONPAPERS
FORMATTXT
FILE SIZE17 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2412-05350

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Lorenz gauge and the Poisson bracket in canonical electromagnetism D. Jaffino Stargen Abstract In treatments of electromagnetism, it is often tacitly assumed that the vector potentials of the field and their conjugate momenta satisfy the canonical Poisson bracket relations, despite the fact that the components of the vector potential are constrained by gauge conditions. Here I explicate how this comes about by imposing Poisson bracket relations on the independent field variables remaining after the Lorenz gauge constraint is accounted for. The naively assumed Poisson brackets happens to be correct even after gauge fixing, owing to a conspiracy between the gauge conditions and the principle of relativity. I Introduction In the Hamiltonian view of classical systems [ 1 , 2 , 3 , 4 ] , the state of a system is specified by a set of n generalized coordinates \mathbf{q}=\{q_{i}\} and their conjugate momenta \mathbf{p}=\{p_{i}\} , assumed to be unconstrained and satisfying the Poisson bracket relation \{q_{i},p_{j}\}=\delta_{ij} ; time evolution of the state is described by Hamilton’s equations: \dot{q}_{i}=\{q_{i},H\} and \dot{p}_{i}=\{p_{i},H\} . Thus, in this scheme, the Poisson bracket between the n independent pairs of variables (“degrees of freedom”) plays a crucial role. If the degrees of freedom are not independent, i.e. if they are constrained, \phi(\mathbf{q},\mathbf{p})=0 , then the Hamiltonian formalism has to be amended [ 5 , 6 , 7 , 8 , 9 ] : either the Poisson bracket has to be modified, or, if they are retained by treating the variables as independent, then the constraints \phi=0 need to be imposed as auxiliary conditions. The classical theory of the electromagnetic field naturally presents a constrained problem. If the system is to be treated in an explicitly Lorentz covariant manner, then the appropriate field coordinates are the vector potentials \{A^{\mu}\} . However, not all configurations of the potentials are physically distinct, i.e. a gauge constraint of the form \phi(A^{\mu})=0 appears between them. Paralleling the discussion above, one of two possibilities can address these constraints. The Poisson bracket suitable, had the field variables been independent, \{A^{\mu}(t,{\bf x}),\pi^{\nu}(t,{\bf x}^{\prime})\}=\eta^{\mu\nu}\delta({\bf x}-{\bf x}^{\prime}), (1) needs to be modified to account for the gauge constraints, where the signature of the Minkowski metric is \eta=(-1,1,1,1) [ 9 , 10 , 11 , 12 ] . Or, if they are to be retained as above, the gauge constraint needs to be imposed separately, and further, it must be ensured that once imposed, they remain valid for all times, i.e. \{\phi(A^{\mu}),H\}=0 . Both these routes can be explored by using Dirac’s theory of constrained Hamiltonian mechanics [ 7 , 8 , 13 , 14 , 15 ] . A less sophisticated, and arguably more transparent, approach is conceivable: eliminate the gauge constraint explicitly in order to identify the true independent degrees of freedom of the field, and impose Poisson brackets on those. Although not apriori obvious, such a procedure, if performed consistently, agrees with eq. 1 . The primary purpose of our exposition is precisely to demonstrate how this comes about, i.e., how the Poisson bracket structure of electromagnetism [ eq. 1 ] is upheld after the gauge constraint is accounted for. As it happens, this relies on the principle of relativity in an essential manner. The exposition is organized as follows: in Section II we describe the appropriate degrees of freedom corresponding to the electromagnetic field system; Section III explicates how the Poisson bracket corresponding to the field variables, A^{\mu} and \pi^{\mu} , is modified due to the gauge constraint, and indeed how the Poisson bracket preserves its structure even after the gauge constraint is taken into consideration; and we conclude in Section IV . II Canonical electromagnetism We start with the Lagrangian density of the electromagnetic field as [ 16 , 10 ] {\cal L}=-\frac{1}{2}{\partial}_{\mu}A_{\nu}{\partial}^{\mu}A^{\nu}, (2) where the four-potential, A^{\mu} , are taken to be the (generalized) coordinates, whose canonically conjugate momentum, \pi_{\mu} , is \pi_{\mu}\equiv\frac{{\partial}{\cal L}}{{\partial}({\partial}_{0}A^{\mu})}={\partial}_{0}A_{\mu}. (3) However, due to the gauge constraints involved, the field coordinates, A^{\mu} , are not independent of each other. The gauge constraint of our choice is of the Lorenz form, i.e. , \partial_{\mu}A^{\mu}=0. (4) In order to render the subsequent discussion as simple as possible, we decompose the field coordinate, A^{\mu} , as A^{\mu}(\mathbf{x},t)=\sum_{\lambda=0}^{3}\int{\rm d}^{3}{\bf k}\varepsilon^{\mu}_{(\lambda)}q^{(\lambda)}_{{\bf k}}(t)e^{-{\mathrm{i}}{\bf k}\cdot{\bf x}}, (5) where \varepsilon^{\mu}_{(\lambda)} are polarization vectors, with \mu and \lambda signifying the spacetime and polarization indices, respectively; the variables, q^{(\lambda)}_{{\bf k}} , capture the dynamical degrees of freedom corresponding to the field, and are complex variables in general. Since the vector potential, A^{\mu} , is real, the dynamical variables, q^{(\lambda)}_{{\bf k}} , and its complex conjugate, q^{*(\lambda)}_{{\bf k}} , are related by q^{*(\lambda)}_{{\bf k}}=q^{(\lambda)}_{-{\bf k}} . Therefore, the field Lagrangian, L , in terms of the dynamical variables, q^{(\lambda)}_{{\bf k}} , can be obtained from the Lagrangian density, {\cal L} , of the field [ eq. 2 ] as L\equiv\int{\rm d}^{3}{\bf x}~{\cal L}(t,{\bf x})=\int{\rm d}^{3}{\bf k}~L_{{\bf k}}, (6) where L_{{\bf k}}\equiv\eta_{\lambda\lambda^{\prime}}\biggl({\dot{q}}_{{\bf k}}^{(\lambda)}{\dot{q}}_{-{\bf k}}^{(\lambda^{\prime})}-|{\bf k}|^{2}q_{{\bf k}}^{(\lambda)}q_{-{\bf k}}^{(\lambda^{\prime})}\biggr). (7) Since the dynamical variables, q^{(\lambda)}_{{\bf k}} , are independent of each other prior to taking gauge constraint into consideration, the Lagrangian, L_{{\bf k}} , corresponding to a field mode, labelled by {\bf k} , is composed of four independent dynamical systems corresponding to the dynamical variables, q^{(\lambda)}_{{\bf k}} . The momentum variables canonically conjugate to the variables, q^{(\lambda)}_{{\bf k}} , can be found to be p_{{\bf k}(\lambda)}\equiv\frac{{\partial}L_{{\bf k}}}{{\partial}{\dot{q}}^{(\lambda)}_{{\bf k}}}={\dot{q}}_{-{\bf k}(\lambda)}, (8) p_{-{\bf k}(\lambda)}\equiv\frac{{\partial}L_{{\bf k}}}{{\partial}{\dot{q}}^{(\lambda)}_{-{\bf k}}}={\dot{q}}_{{\bf k}(\lambda)}. (9) Therefore, the field Hamiltonian is \displaystyle H \displaystyle\equiv\int{\rm d}^{3}{\bf k}~\eta_{\lambda\lambda^{\prime}}\biggl(p_{{\bf k}}^{(\lambda)}{\dot{q}}_{{\bf k}}^{(\lambda^{\prime})}+p_{-{\bf k}}^{(\lambda)}{\dot{q}}_{-{\bf k}}^{(\lambda^{\prime})}-L_{{\bf k}}\biggr) (10) \displaystyle=\int{\rm d}^{3}{\bf k}~\eta_{\lambda\lambda^{\prime}}\biggl(p_{{\bf k}}^{(\lambda)}p_{-{\bf k}}^{(\lambda^{\prime})}+|{\bf k}|^{2}q_{{\bf k}}^{(\lambda)}q_{-{\bf k}}^{(\lambda^{\prime})}\biggr). (11) Using this canonical setup of classical electromagnetism, we eliminate the redundant variables due to the gauge constraint [ eq. 4 ], and analyze the effects of incorporating this redundancy in the functional form of the Poisson bracket corresponding to the field variables, (A^{\mu},\pi^{\mu}) . III Lorenz gauge and Poisson bracket The Lorenz gauge condition in eq. 4 , when written in terms of the dynamical variables, (q_{\pm\mathbf{k}}^{(\lambda)},p_{\pm\mathbf{k}}^{(\lambda)}) , becomes \left({\rm k}_{i}\varepsilon^{i}_{(\lambda)}\right)q^{(\lambda)}_{\pm{\bf k}}\pm i\varepsilon^{0}_{(\lambda)}p^{(\lambda)}_{\mp{\bf k}}=0, (12) which are a set of algebraic constraints between the variables q_{\pm\mathbf{k}}^{(\lambda)} and p_{\pm\mathbf{k}}^{(\lambda)} . In the Hamiltonian view the gauge constraints in eq. 12 are imposed at one instant of time; it is then necessary to ensure that they remain true under the time evolution generated by the Hamiltonian H . That is, one needs to demand that \{{\partial}_{\mu}A^{\mu},H\}=0, (13) with the expression for Poisson bracket (see appendix A ) as \displaystyle\{f_{{\bf k}},g_{{\bf k}^{\prime}}\} \displaystyle\equiv\eta^{\lambda\lambda^{\prime}}\biggl(\frac{{\partial}f_{{\bf k}}}{{\partial}q^{(\lambda)}_{{\bf k}}}\frac{{\partial}g_{{\bf k}^{\prime}}}{{\partial}p^{(\lambda^{\prime})}_{{\bf k}}}-\frac{{\partial}g_{{\bf k}^{\prime}}}{{\partial}q^{(\lambda)}_{{\bf k}}}\frac{{\partial}f_{{\bf k}}}{{\partial}p^{(\lambda^{\prime})}_{{\bf k}}} \displaystyle\hskip-7.0pt+\frac{{\partial}f_{{\bf k}}}{{\partial}q^{(\lambda)}_{-{\bf k}}}\frac{{\partial}g_{{\bf k}^{\prime}}}{{\partial}p^{(\lambda^{\prime})}_{-{\bf k}}}-\frac{{\partial}g_{{\bf k}^{\prime}}}{{\partial}q^{(\lambda)}_{-{\bf k}}}\frac{{\partial}f_{{\bf k}}}{{\partial}p^{(\lambda^{\prime})}_{-{\bf k}}}\biggr)\delta({\bf k}-{\bf k}^{\prime}), (14) and it will generate an independent set of algebraic constraints in terms of the dynamical variables, (q_{\pm\mathbf{k}}^{(\lambda)},p_{\pm\mathbf{k}}^{(\lambda)}) . Using these expressions, and the field Hamiltonian, H , [ eq. 11 ], to compute the Poisson bracket in eq. 13 implies the algebraic constraints |{\bf k}|^{2}\varepsilon^{0}_{(\lambda)}q^{(\lambda)}_{\pm{\bf k}}\pm i\left({\rm k}_{i}\varepsilon^{i}_{(\lambda)}\right)p^{(\lambda)}_{\mp{\bf k}}=0. (15) One can further check whether the condition imposed before, i.e., \{{\partial}_{\mu}A^{\mu},H\}=0 , too remains enforced in time, which means demanding \{\{{\partial}_{\mu}A^{\mu},H\},H\}=0 . Direct computation yields \begin{split}\{\{{\partial}_{\mu}A^{\mu},H\},H\}&=\int{\rm d}^{3}{\bf k}~e^{-i\mathbf{k}\cdot\mathbf{x}}\\ &\times\left[\left({\rm k}_{i}\varepsilon^{i}_{(\lambda)}\right)q^{(\lambda)}_{\pm{\bf k}}\pm i\varepsilon^{0}_{(\lambda)}p^{(\lambda)}_{\mp{\bf k}}\right].\end{split} (16) Comparing the expression in the square bracket with the algebraic equation in eq. 12 , it is evident that the condition, \{\{{\partial}_{\mu}A^{\mu},H\},H\}=0 , does

Read the complete text