In General > s.a. [classical mechanics];
Euler-Lagrange; lagrangian
systems; quantum mechanics [need
for action].
* Idea: A formulation
of dynamics based on the principle of least action, in turn based on variational
calculus.
* History: The term "action",
and the concept that physical laws are a consequence of "economy of means" were
introduced by Maupertuis in 1744, with W =
p dq,
and improved by Euler and Lagrange; In 1832 Hamilton proposed the action S =
L(q, q·)
dt.
* Motivation: (1) Elegant
way of expressing the dynamical content of a physical theory; (2) Study of
symmetries of the theory; (3) Important
for the pathy integration quantization program (> but see path
integrals);
(4) For a field theory, it is a "spacetime
covariant" formulation,
and for relativistic quantum field theory it is preferable to a Hamiltonian
formulation.
* Action: (1) Choose
a configuration space, with coordinates q depending
on x;
(2) If the equations of motion are known, the problem is to find, given a region X of
the x's,
a Lagrangian (density)
or
an action functional S[q,
q, x],
written
S[q,
q, x]
=
M
(q,
q, x)
dx ,
from which equations of interest are derived using variational principles.
* Formal setting for a field
theory:
A Lagrangian of order k is a horizontal
form L: JkY →
nM,
where (Y,
, M)
is the configuration bundle for the theory, (JkY,
k, M)
its k-jet
prolongation, and M the
spacetime manifold; A
variation of the fields is a vertical vector field X =
yi (
/
yi),
under which the Lagrangian varies by
XL = e(L, X)
+ dF(L, X) , with e(L, X)
= Euler-Lagrange form , and F(L, X)
= Poincaré-Cartan form
.
* Normalization constants:
They do not affect the equations of motion, but they do appear in the conjugate
momenta.
* Interacting systems: To
get the Lagrangian, just add the ones for each
system and (if they are needed) the extra interaction terms, but we have
to
be careful with the relative coefficients for each term.
And Canonical Formalism > s.a. constrained
systems.
* Legendre transformation: The map F: TQ → T*Q
(Q is
configuration space) defined, for a Lagrangian L, by F(w)v:=
dL(w+tv)/dt|t=0.
@ General references: Kastrup PRP(87);
Geyer et al JPA(03)ht/02 [and
gauge symmetries]; Bertin et al MPLA(05)ht [first-order
actions, Hamilton-Jacobi analysis]; Compère a0708-PhD
[covariant Hamiltonian formulation, symmetries, and black holes].
@ For field theories: Echeverría-Enríquez et al IJMMS(02)mp/01, JMP(04)
[unified formalism with Hamiltonian].
Choice of Action and Ambiguities > s.a. variational
principles.
* Idea: In practice,
in particular for field theories, we get the invariances from nature (e.g.,
gauge symmetries), and this often is very restrictive; However,
classically
one can always add a boundary term to the action, or a div to
,
that depends on the fields; This has the effect of changing the definitions of
the
canonical momenta and energy but not the equations of motion – the Lagrangians
are
in the same variational cohomology class; In the quantum theory, it
is
not so simple.
* Field redefinitions: For example, by conformal transformations.
@ Ambiguity in Lagrangian: Hojman & Shepley in(88), Lunev TMP(91),
TMP(92)
[for field theories];
in
Matzner & Shepley 92; Faraoni CQG(94);
Cislo & Lopuszanski
JMP(01)mp/00 [and
quantization].
@ Lagrangian cohomology classes: in Marathe & Martucci 92; in Kolár
et al 93.
@ Related topics: Frieden 98 [L's from Fisher information]; Buniy
& Kephart PLA(08) [higher-order topological actions and quantum theory].
Symmetries and Conserved Quantities > s.a. conservation
laws; noether
theorem; symmetries.
* Symmetric variations:
The issue is whether the Principle of Symmetric Criticality holds; It states
that for any group-invariant lagrangian the equations
obtained by restriction of Euler-Lagrange equations to group-invariant
fields are equivalent to the Euler-Lagrange equations of a canonically defined,
symmetry-reduced lagrangian.
* Conserved currents and Lie
algebras: One may introduce at least three different Lie algebras in any
Lagrangian field theory, (i) the Lie algebra of local BRST cohomology classes
with the odd Batalin-Vilkovisky antibracket; (ii) the Lie algebra of local
conserved currents with the Dickey bracket; and (iii) the Lie algebra
of conserved, integrated
charges with the Poisson bracket.
* Question: Are there known
examples of systems for which symmetric criticality fails?
@ General references: Gràcia & Pons JPA(92);
Grigore JPA(95)ht/94;
Barnich & Henneaux JMP(96) [Batalin-Vilkovisky antibracket and Poisson
bracket]; Banerjee et al PLB(99)ht,
PLB(00)ht/99,
Banerjee
ht/00 [and
Hamiltonian].
@ Conservation laws: Lutzky JPA(79), JPA(82), JPA(95), JPA(98)
[non-Noether]; Sardanashvily mp/03
[Noether].
@ Symmetric variations: Palais CMP(79); Christodoulakis & Korfiatis
NCB(94); Davis GRG(98)gq/96,
Fels & Torre CQG(02)gq/01 [gravity].
@ Lagangians from Lie transformation groups: Paal & Virkepu a0706;
Nucci & Leach JMP(07)-a0706 [Jacobi
last multiplier].
References
@ Textbooks: Lagrange 1788; in Sudarshan & Mukunda 75; Brizard 08 [II].
@ General: Bailey FP(81)
["more precise statement"]; Cariñena
et al PRP(95)
[need for lagrangian in physics]; Stöltzner SHPMP(03)
[principle of least action]; Evans AJP(03)may
[interpretation of particle action]; Toffoli IJTP(03)
[meaning?]; Butterfield phy/04 [ontology];
Gray & Taylor AJP(07)may
[not minimized in general]; Cuell RPMP(07)
[as a skew critical problem].
@ And equations of motion: Grigore NCB(96)ht [conditions
for first-order Lagrangian]; Gitman & Kupriyanov JPA(07)-a0710;
Nucci & Tamizhmani a0809 [finding L by the method of Jacobi Last Multiplier].
@ Geometrical: Cariñena et al JMP(88);
del Olmo & Santander
JPA(89),
JGP(90);
Sharipov
m.DG/01 [in
Riemannian manifolds].
@ And spacetime transformations: Whiston IJTP(72).
@ Inverse variational principle: Santilli 78; Okubo PRD(80) [non-unique];
Cariñena et al PRP(95).
@ Generalizations: Grigore IJMPA(92);
Núñez-Yépez
et al mp/01-in;
Arizmendi
et al CSF(03)mp/04;
Gravanis & Willison a0901 [distributional
fields].
@ Related topics: Gocksch PLB(88),
Anagnostopoulos & Nishimura PRD(02)
[complex
action];
Ferraris
et
al
JMP(00)
[dual
Lagrangians]; Muñoz Díaz a0801 [and
time].
main page – abbreviations – journals – comments – other
sites – acknowledgements
send feedback and suggestions to bombelli at olemiss.edu – modified 28
jun 2009