Physlib

PhyslibAlpha.ClassicalFieldTheory.Local.Action

Local action functionals

i. Overview

This module defines the local action functional associated with a local Lagrangian, together with the first notions needed to talk about variational criticality.

For the first implementation pass, the action is defined directly as the integral of a local Lagrangian evaluated along jets of a field. The integrability conditions needed for this action and for its variations are kept explicit in the API.

ii. Key results

- `ClassicalFieldTheory.Local.actionDensity` : the density associated with a field. - `ClassicalFieldTheory.Local.action` : the action of a field. - `ClassicalFieldTheory.Local.HasFiniteAction` : finiteness of the action integral. - `ClassicalFieldTheory.Local.IsAdmissibleForAction` : symmetric admissibility of a lagrangian and field pair for the action functional. - `ClassicalFieldTheory.Local.actionVariation` : the action under an admissible variation. - `ClassicalFieldTheory.Local.IsCritical` : vanishing first derivative of the varied action.

iii. Table of contents

  • A. Action densities and action
  • B. Action under variation
  • C. Critical fields

iv. References

- J. Cortés and A. Haupt, *Lecture Notes on Mathematical Methods of Classical Physics*, Chapter 5.

A. Action densities and action

B. Action under variation

C. Critical fields

18 declarations

definition

Action density of a field ff for a Lagrangian LL

Given a local Lagrangian LL of order kk and a field f:RdRmf: \mathbb{R}^d \to \mathbb{R}^m, the action density is a scalar function on the base space RdR\mathbb{R}^d \to \mathbb{R}. For any point xRdx \in \mathbb{R}^d, the value of the function is L(jkf(x))L(j^k f(x)), where jkf(x)j^k f(x) denotes the kk-jet of the field ff at xx.

definition

Integrability of the action density L(jkf)L(j^k f)

For a local Lagrangian LL of order kk and a field f:RdRmf: \mathbb{R}^d \to \mathbb{R}^m, the property `HasFiniteAction L f` holds if the action density function xL(jkf(x))x \mapsto L(j^k f(x)) is integrable over the space Rd\mathbb{R}^d with respect to the Lebesgue measure.

definition

Admissibility of a Lagrangian LL and field ff for the action functional

For a local Lagrangian LL of order kk and a field f:RdRmf: \mathbb{R}^d \to \mathbb{R}^m, the pair (L,f)(L, f) is admissible for the action functional if the field ff is smooth (CC^\infty) and the action density function xL(jkf(x))x \mapsto L(j^k f(x)) is integrable over Rd\mathbb{R}^d (i.e., the field has a finite action).

definition

Action S(L,f)=L(jkf(x))dxS(L, f) = \int L(j^k f(x)) dx

Given a local Lagrangian LL of order kk and a field f:Space dRmf: \text{Space } d \to \mathbb{R}^m, the action is the real number defined as the integral of the action density over the base space Space d\text{Space } d: S(L,f)=Space dL(jkf(x))dx S(L, f) = \int_{\text{Space } d} L(j^k f(x)) \, dx where jkf(x)j^k f(x) is the kk-jet of the field ff at point xx.

theorem

Action density at xx equals L(jkf(x))L(j^k f(x))

Let LL be a local Lagrangian of order kk, f:RdRmf: \mathbb{R}^d \to \mathbb{R}^m be a field, and xRdx \in \mathbb{R}^d be a point in the base space. The action density of the field ff for the Lagrangian LL evaluated at xx is equal to the Lagrangian evaluated at the kk-jet of the field at xx, denoted by jkf(x)j^k f(x). That is, actionDensity(L,f,x)=L(jkf(x))\text{actionDensity}(L, f, x) = L(j^k f(x)).

theorem

Action Equals the Integral of the Action Density

For a local Lagrangian LL of order kk and a field f:Space dRmf: \text{Space } d \to \mathbb{R}^m, the action action(L,f)\text{action}(L, f) is defined as the integral of the action density over the base space Space d\text{Space } d: action(L,f)=Space dactionDensity(L,f,x)dx \text{action}(L, f) = \int_{\text{Space } d} \text{actionDensity}(L, f, x) \, dx where actionDensity(L,f,x)\text{actionDensity}(L, f, x) is the Lagrangian evaluated at the kk-jet of the field ff at xx.

definition

Varied field f+sηf + s\eta

Given a field f ⁣:Space dRmf \colon \text{Space } d \to \mathbb{R}^m, an admissible variation η ⁣:Space dRm\eta \colon \text{Space } d \to \mathbb{R}^m, and a scalar parameter sRs \in \mathbb{R}, the **varied field** is the function mapping each point xx in the domain to f(x)+sη(x)f(x) + s\eta(x).

theorem

Smoothness of the Varied Field f+sηf + s\eta

Let f:Space dRmf: \text{Space } d \to \mathbb{R}^m be an infinitely differentiable (CC^\infty) field. For any admissible variation η:Space dRm\eta: \text{Space } d \to \mathbb{R}^m and any scalar sRs \in \mathbb{R}, the varied field xf(x)+sη(x)x \mapsto f(x) + s\eta(x) is also infinitely differentiable (CC^\infty).

definition

Action under variation sS(L,f+sη)s \mapsto S(L, f + s\eta)

Given a local Lagrangian LL of order kk, a field f ⁣:Space dRmf \colon \text{Space } d \to \mathbb{R}^m, and an admissible variation η ⁣:Space dRm\eta \colon \text{Space } d \to \mathbb{R}^m, the **action variation** is the function that maps a scalar parameter sRs \in \mathbb{R} to the action of the varied field f+sηf + s\eta. It is defined as: sS(L,f+sη)=Space dL(jk(f+sη)(x))dx s \mapsto S(L, f + s\eta) = \int_{\text{Space } d} L(j^k(f + s\eta)(x)) \, dx where jk(f+sη)(x)j^k(f + s\eta)(x) denotes the kk-jet of the varied field at point xx.

definition

Integrability of the action density for the varied field f+sηf + s\eta for all sRs \in \mathbb{R}

Let LL be a local Lagrangian of order kk, f:Space dRmf: \text{Space } d \to \mathbb{R}^m be a field, and η\eta be an admissible variation. The property `HasFiniteActionVariation` states that for every scalar parameter sRs \in \mathbb{R}, the action density associated with the varied field f+sηf + s\eta is integrable over Space d\text{Space } d. Mathematically, this means that for all sRs \in \mathbb{R}, the function xL(jk(f+sη)(x))x \mapsto L(j^k(f + s\eta)(x)) is integrable.

definition

Compactly supported difference in action density under variation ff+sηf \to f + s\eta

For a local Lagrangian LL, a field f:RdRmf: \mathbb{R}^d \to \mathbb{R}^m, and an admissible variation η\eta, the property `HasCompactlySupportedActionVariationDifference` holds if, for every scalar parameter sRs \in \mathbb{R}, the difference between the action density of the varied field f+sηf + s\eta and the action density of the original field ff, given by xL(jk(f+sη)(x))L(jkf(x))x \mapsto \mathcal{L}(j^k(f + s\eta)(x)) - \mathcal{L}(j^k f(x)) is a continuous function with compact support on Rd\mathbb{R}^d. Here, L(jkf(x))\mathcal{L}(j^k f(x)) denotes the action density evaluated at the kk-jet of the field ff at point xx.

theorem

Finite action and compactly supported density difference imply finite action of the varied field

Let LL be a local Lagrangian of order kk, f ⁣:RdRmf \colon \mathbb{R}^d \to \mathbb{R}^m be a field, and η\eta be an admissible variation. Suppose that the field ff has finite action, meaning the action density xL(jkf(x))x \mapsto \mathcal{L}(j^k f(x)) is integrable over Rd\mathbb{R}^d. Furthermore, suppose that for every scalar parameter sRs \in \mathbb{R}, the difference between the action density of the varied field and the original field, xL(jk(f+sη)(x))L(jkf(x))x \mapsto \mathcal{L}(j^k(f + s\eta)(x)) - \mathcal{L}(j^k f(x)), is a continuous function with compact support on Rd\mathbb{R}^d. Then the varied field f+sηf + s\eta also has finite action for every sRs \in \mathbb{R}, i.e., the function xL(jk(f+sη)(x))x \mapsto \mathcal{L}(j^k(f + s\eta)(x)) is integrable over Rd\mathbb{R}^d.

theorem

Pointwise evaluation of the varied field f+sηf + s\eta

Given a field f ⁣:Space dRmf \colon \text{Space } d \to \mathbb{R}^m, an admissible variation η ⁣:Space dRm\eta \colon \text{Space } d \to \mathbb{R}^m, and a scalar parameter sRs \in \mathbb{R}, the value of the varied field evaluated at a point xSpace dx \in \text{Space } d is given by f(x)+sη(x)f(x) + s \eta(x).

theorem

actionVariation(L,f,η)(s)=S(L,f+sη)\text{actionVariation}(L, f, \eta)(s) = S(L, f + s\eta)

Given a local Lagrangian LL of order kk, a field f ⁣:Space dRmf \colon \text{Space } d \to \mathbb{R}^m, an admissible variation η ⁣:Space dRm\eta \colon \text{Space } d \to \mathbb{R}^m, and a scalar parameter sRs \in \mathbb{R}, the value of the action variation at ss is equal to the action of the varied field f+sηf + s\eta: actionVariation(L,f,η)(s)=S(L,f+sη) \text{actionVariation}(L, f, \eta)(s) = S(L, f + s\eta)

theorem

jk(f+sη)(x)=jkf(x)j^k (f + s\eta)(x) = j^k f(x) for xsupp(η)x \notin \text{supp}(\eta)

Let kNk \in \mathbb{N} be a derivative order, f:Space dRmf : \text{Space } d \to \mathbb{R}^m be an infinitely differentiable (CC^\infty) field, and η:Space dRm\eta : \text{Space } d \to \mathbb{R}^m be an admissible variation. For any scalar sRs \in \mathbb{R} and any point xSpace dx \in \text{Space } d, if xx is not in the topological support of η\eta (denoted supp(η)\text{supp}(\eta)), then the kk-jet of the varied field f+sηf + s\eta at xx is equal to the kk-jet of the field ff at xx: jk(f+sη)(x)=jkf(x) j^k (f + s\eta)(x) = j^k f(x)

theorem

actionDensity(L,f+sη,x)=actionDensity(L,f,x)\text{actionDensity}(L, f + s\eta, x) = \text{actionDensity}(L, f, x) for xsupp(η)x \notin \text{supp}(\eta)

Let LL be a local Lagrangian of order kk for fields mapping from Rd\mathbb{R}^d to Rm\mathbb{R}^m. Let f:RdRmf: \mathbb{R}^d \to \mathbb{R}^m be an infinitely differentiable (CC^\infty) field and η:RdRm\eta: \mathbb{R}^d \to \mathbb{R}^m be an admissible variation. For any scalar sRs \in \mathbb{R} and any point xRdx \in \mathbb{R}^d, if xx is not in the topological support of η\eta (denoted supp(η)\text{supp}(\eta)), then the action density of the varied field f+sηf + s\eta at the point xx is equal to the action density of the original field ff at xx: actionDensity(L,f+sη,x)=actionDensity(L,f,x) \text{actionDensity}(L, f + s\eta, x) = \text{actionDensity}(L, f, x)

theorem

Continuity of LL in coordinates implies compactly supported action variation difference for smooth fields ff

Let LL be a kk-th order local Lagrangian and f:RdRmf: \mathbb{R}^d \to \mathbb{R}^m be an infinitely differentiable (CC^\infty) field. If LL is continuous in its local jet coordinates (x,uIa)(x, u^a_I), then for any admissible variation η\eta, the difference between the action density of the varied field f+sηf + s\eta and the original field ff has compact support. Specifically, for every scalar sRs \in \mathbb{R}, the function xL(jk(f+sη)(x))L(jkf(x))x \mapsto \mathcal{L}(j^k(f + s\eta)(x)) - \mathcal{L}(j^k f(x)) is a continuous function with compact support on Rd\mathbb{R}^d, where L(jkf(x))\mathcal{L}(j^k f(x)) denotes the action density evaluated at the kk-jet of the field ff at the point xx.

definition

Field ff is critical for the action: ddss=0S(L,f+sη)=0\left. \frac{d}{ds} \right|_{s=0} S(L, f + s\eta) = 0

Let LL be a local Lagrangian of order kk and f ⁣:Space dRmf \colon \text{Space } d \to \mathbb{R}^m be a field. The field ff is **critical** for the action functional if, for every admissible variation η\eta such that the action of the varied field f+sηf + s\eta is finite for all sRs \in \mathbb{R} (i.e., `HasFiniteActionVariation L f η`), the derivative of the action variation with respect to the parameter ss at s=0s = 0 is zero. This is expressed by the condition: ddss=0S(L,f+sη)=0 \left. \frac{d}{ds} \right|_{s=0} S(L, f + s\eta) = 0 where S(L,f+sη)=Space dL(jk(f+sη)(x))dxS(L, f + s\eta) = \int_{\text{Space } d} L(j^k(f + s\eta)(x)) \, dx is the action functional and jk(f+sη)(x)j^k(f + s\eta)(x) denotes the kk-jet of the varied field at point xx.