Professional Documents
Culture Documents
This article is about the Lagrangian function in equivalent Lagrangians will give the same equations of
Lagrangian mechanics. For other uses, see Lagrangian motion.[3][4]
(disambiguation).
The Lagrangian is named after Italian-French mathematician and astronomer Joseph Louis Lagrange. The
concept of a Lagrangian was introduced in a reformulation of classical mechanics introduced by Lagrange
known as Lagrangian mechanics in 1788. This reformulation was needed in order to explore mechanics
in alternative systems to Cartesian coordinates such as
Polar, Cylindrical and Spherical coordinates for which
Newtonian mechanics was not suitable.[1]
If Nature has dened the mechanics problem of the thrown ball in so elegant a fashion,
might She have dened other problems similarly. So it seems now. Indeed, at the present
time it appears that we can describe all the
fundamental forces in terms of a Lagrangian.
The search for Natures One Equation, which
rules all of the universe, has been largely a
search for an adequate Lagrangian.
Robert Adair, The Great Design: Particles,
Fields, and Creation[5]
Using only the principle of least action and the Lagrangian we can deduce the correct trajectory, by trial
and error or the calculus of variations.
Denition
Physical action and quantum-mechanical phase are related via Plancks constant, and the principle of stationary
action can be understood in terms of constructive interference of wave functions.
1
3 EXPLANATION
2.3
2.4
systems. This formalism was generalized further to handle eld theory. In eld theory, the independent variable
is replaced by an event in spacetime (x, y, z, t), or more
generally still by a point s on a manifold. The dependent
variables (q) are replaced by the value of a eld at that
point in spacetime (x,y,z,t) so that the equations of motion are obtained by means of an action principle, written
as:
S
An important property of the Lagrangian is that conser- i = 0,
vation laws can easily be read o from it. For example,
if the Lagrangian L does not depend on qi itself, then the where the action, S , is a functional of the dependent variables i(s) with their derivatives and s itself
generalized momentum ( pi ), given by:
L
pi =
,
qi
S [i ] =
)
(
i (s)
,
s
dn s
L i (s),
s
is a conserved quantity, because of Lagranges equations: and where s = { s } denotes the set of n independent variables of the system, indexed by = 1, 2, 3,..., n. Notice
L is used in the case of one independent variable (t) and
L
L is used in the case of multiple independent variables
pi =
= 0.
(usually four: x, y, z, t).
qi
L
d L
=
.
dt i
i
4.2
In spherical coordinates
Dynamical systems whose equations of motion are obtainable by means of an action principle on a suitably
chosen Lagrangian are known as Lagrangian dynamical systems. Examples of Lagrangian dynamical systems
range from the classical version of the Standard Model,
to Newtons equations, to purely mathematical problems
such as geodesic equations and Plateaus problem.
L=
m 2
(r + r2 2 + r2 sin2 2 ) V (r).
2
4.1
In Cartesian coordinates
=
L(x, x)
1
mx 2 V (x).
2
The equations of motion for the particle are found by applying the EulerLagrange equation
m
r mr(2 + sin2 2 ) + V = 0,
d
mr2 sin cos 2 = 0,
(mr2 )
dt
d
(mr2 sin2 )
= 0.
dt
Here the set of parameters si is just the time t, and the
dynamical variables i(s) are the trajectories x(t) of the
particle.
Despite the use of standard variables such as x, the Lagrangian allows the use of any coordinates, which do not
need to be orthogonal. These are "generalized coordinates".
Then
L
V
=
,
xi
xi
d
dt
L
x i
= 0,
xi
where i = 1, 2, 3.
=
x i
x i
1
mx 2
2
and
d
dt
L
x i
)
=
1 ( 2)
m
x = mx i ,
2 x i i
)
= m
xi .
T (t) = 12 m |x(t)|
Thus
m
x + V = 0,
which is Newtons second law of motion for a particle Then its Lagrangian is L joules, where
subject to a conservative force. Here the time derivative
is written conventionally as a dot above the quantity being
2
m(x(t), t).
L(t) = T (t) V (t) = 12 m |x(t)|
dierentiated, and is the del operator.
L(t) dt =
0=
L(t) dt
d mx(t)
0=
t)q x(t)A(x(t),
t)
dt
v 2 (t)
1 c2
which is
x(t)
(mx(t)
m(x(t), t) x(t)) dt.
Integrate the rst term by parts and discard the total integral. Then divide out the variation to get
d mx(t)
= qE(x(t), t) + q x(t)
B(x(t), t)
2
dt
1 v c(t)
2
which is the equation for the Lorentz force, where:
0 = m
x(t) m(x(t), t)
and thus
B(x, t) = A(x, t)
are the elds and potentials.
is the equation of motion two dierent expressions for
5.3
the force.
5.2
Special relativistic test particle with both the classical kinetic energy and the interaction with
the gravitational eld. It becomes:[8][9]
electromagnetism
dx (t) dx (t)
energy) of a free test particle is
2 d (t)
mc
= mc g (x(t))
.
dt
dt
dt
The Lagrangian of a general relativistic test particle in an
dt
mc2
3 v 4 (t) electromagnetic eld is:
2 1
2
= +mc + mv (t)+ m 2 +. . . .
mc
=
2
d
2
8
c
1 v c(t)
2
dx (t) dx (t)
dx (t)
However, the term in the Lagrangian that gives rise to
L(t) = mc g (x(t))
+q
A (x(t)).
the derivative of the momentum is no longer the kinetic
dt
dt
dt
energy. It must be changed to
If the four spacetime coordinates x are given in arbitrary
units (i.e. unitless), then g in m2 is the rank 2 symmetric metric tensor which is also the gravitational potential.
1 v 4A
(t) Vs is the electromagnetic 4-vector potential.
v 2 (t)
2 d (t)
2
2 1
2
mc
= mc 1 2 = mc + mv (t)+ Also,
m 2 in
+. . .
dt
c
2
8
c that a factor of c has been absorbed into the square
Notice
root because it is the equivalent of
where c is the vacuum speed of light in ms1 , is the
proper time in seconds (i.e. time measured by a clock
v 2 (t)
L(t) = mc2
7.1
Newtonian gravity
made between the Lagrangian L, of which the action is substituted back in equation (1), the Lagrangian equation
the time integral:
for the test particle in a Newtonian gravitational eld, provides the information needed to calculate the acceleration
of the particle.
S=
L dt
mass density in kgm3 . This is necessary because using a point source for a eld would result in mathematiS[i ] = L(i (x)) d4 x.
cal diculties. The resulting Lagrangian for the classical
gravitational eld is:
General form of Lagrangian density: L =
i
L(i , i, ) [10] where i, x
i (see
4-gradient)
1
L(x, t) = (x, t)(x, t)
((x, t))2
n1
8G
The relationship between L and L : L = L d
x
[10]
, where
similar to
n is the space-time dimension
where G in m3 kg1 s2 is the gravitational constant.
q = dV .
Variation of the integral with respect to gives:
In eld theory, the independent variable t was replaced by an event in spacetime (x, y, z, t) or still
2
more generally by a point s on a manifold.
L(x, t) = (x, t)(x, t)
((x, t))((x, t)).
8G
The Lagrangian is then the spatial integral of the La- Integrate by parts and discard the total integral. Then
grangian density. However, L is also frequently sim- divide out by to get:
ply called the Lagrangian, especially in modern use; it
is far more useful in relativistic theories since it is a
locally dened, Lorentz scalar eld. Both denitions of
1
the Lagrangian can be seen as special cases of the gen- 0 = (x, t) + 4G (x, t)
eral form, depending on whether the spatial variable x
is incorporated into the index i or the parameters s in and thus
i(s). Quantum eld theories in particle physics, such as
quantum electrodynamics, are usually described in terms
of L , and the terms in this form of the Lagrangian trans- 4G(x, t) = 2 (x, t)
late quickly to the rules used in evaluating Feynman diagrams.
which yields Gausss law for gravity.
Notice that, in the presence of gravity or when using general curvilinear coordinates,
the Lagrangian density L 7.2 Einstein Gravity
will include a factor of |g| or its equivalent to ensure
that it is a scalar density so that the integral will be in- Main article: EinsteinHilbert action
variant.
Selected elds
c4
(R 2) + Lmatter
16G
R is the curvature scalar, which is the Ricci tensor contracted with the metric tensor, and the Ricci tensor is
the Riemann tensor contracted with a Kronecker delta.
The integral of LEH is known as the Einstein-Hilbert action. The Riemann tensor is the tidal force tensor, and
is constructed out of Christoel symbols and derivatives
of Christoel symbols, which are the gravitational force
eld. Plugging this Lagrangian into the Euler-Lagrange
7 SELECTED FIELDS
equation and taking the metric tensor g as the eld, we We can then write the interaction term as
obtain the Einstein eld equations
1
8G
Rg + g = 4 T
2
c
+ j A = j A
F = A A
2 (Lmatter g)
Lmatter
T
= 2
+g Lmatter .
g
g
g
The term we are looking out for turns out to be
g is the determinant of the metric tensor when regarded
as a matrix. is the Cosmological constant. Generally,
in general relativity, theintegration measure of the action 0 E 2 1 B 2 = 1 F F = 1 F F
20
40
40
of Lagrange density is gd4 x . This makes the integral 2
coordinate independent, as the root of the metric deterWe have made use of the Minkowski metric to raise the
minant is equivalent to the Jacobian determinant. The
indices on the EMF tensor. In this notation, Maxwells
minus sign is a consequence of the metric signature (the
equations are
determinant by itself is negative).[11]
7.3
F = 0 j
and F = 0
A(x(t), t)
q(x(t), t) + q x(t)
are replaced by terms involving a continuous charge den- L(x) = j (x)A (x) 1 F (x)F (x)
40
sity in Asm3 and current density j in Am2 . The
resulting Lagrangian for the electromagnetic eld is:
In this notation it is apparent that classical electromagnetism is a Lorentz-invariant theory. By the equivalence
principle,
it becomes simple to extend the notion of elec0
1 2
L(x, t) = (x, t)(x, t)+j(x, t)A(x, t)+ E 2 (x, t) tromagnetism
B (x, t). to curved spacetime.[12][13]
2
20
Varying this with respect to , we get
0 = (x, t) + 0 E(x, t)
which yields Gauss law.
Varying instead with respect to A , we get
1
0 = j(x, t) + 0 E(x,
t)
B(x, t)
0
which yields Ampres law.
Using tensor notation, we can write all this more compactly. The term (x, t) + j A is actually the inner
product of two Lorentz vectors. We package the charge
density into the current 4-vector and the potential into the
potential 4-vector. These two new vectors are
j = (, j) and A = (, A)
1
c4
F (x)F (x)g (x)g (x)+
R(x) = L
40
16G
1
2
SMaxwell =
T (x) =
0
g(x) g (x)
1
F (x)F (x) g (x)F
4
7.6
Dirac Lagrangian
It can be shown that this energy momentum tensor is is not necessary because forms have coordinate dierentraceless, i.e. that
tials built in. Variation of the action leads to
dF = J.
T = g T = 0
8G
T
c4
dF = 0
So the tracelessness of the energy momentum tensor im- because F is an exact form.
plies that the curvature scalar in an electromagnetic eld
vanishes. The Einstein equations are then
R
g (x)F (x)F (x)
4
/ mc2
L = ic
Additionally, Maxwells equations are
0
8G 1
= 4
c 0
F (x)F (x)
7.5
7.8 Quantum
grangian
chromodynamic
La-
S[A] =
(
)
1
F F + A J .
2
M
11 NOTES
8.1
9 Uses in Engineering
Examples
In classical mechanics, in the Hamiltonian formalism, M is the one-dimensional manifold R , representing time and the target space is the cotangent
bundle of space of generalized positions.
In eld theory, M is the spacetime manifold and the
target space is the set of values the elds can take at
any given point. For example, if there are m realvalued scalar elds, 1 , ..., m, then the target manifold is Rm . If the eld is a real vector eld, then the
target manifold is isomorphic to Rn . There is actually a much more elegant way using tangent bundles
over M, but we will just stick to this version.
8.2
Mathematical development
Consider a functional,
S:CR
called the action. Physical considerations require it be a
mapping to R (the set of all real numbers), not C (the set
of all complex numbers).
In order for the action to be local, we need additional restrictions on the action. If C , we assume S[] is the
integral over M of a function of , its derivatives and the
position called the Lagrangian, L(, , , ..., x) .
In other words,
C, S[]
50 years ago Lagrangians were a general part of the engineering curriculum, but quarter of a century later, even
with the ascendency of dynamical systems, they were
dropped as requirements from the majority of engineering programs, and considered to be the domain of physics.
A decade ago this changed dramatically, and Lagrangians
are not only a required part of many ME and EE curricula,
but are now seen as far more than the province of physics.
This is true of pure and applied engineering, as well as
the more physics-related aspects of engineering, or engineering optimization, which itself is more the province of
Lagrange multipliers.
Today, Lagrangians nd their way into hundreds of direct engineering solutions, including robotics, turbulent
ow analysis (Lagrangian and Eulerian specication of
the ow eld), signal processing, microscopic component contact and nanotechnology (superlinear convergent
augmented Lagrangians), gyroscopic forcing and dissipation, semi-innite supercomputing (which also involve
Lagrange multipliers in the subeld of semi-innite programming), chemical engineering (specic heat linear
Lagrangian interpolation in reaction planning), civil engineering (dynamic analysis of trac ows), optics engineering and design (Lagrangian and Hamiltonian optics)
aerospace (Lagrangian interpolation), force stepping integrators, and even airbag deployment (coupled EulerianLagrangians as well as SELMthe stochastic Eulerian
Lagrangian method).[19]
10 See also
(
)
dn xL (x), (x), (x), ..., x . 11
Notes
It is assumed below, in addition, that the Lagrangian depends on only the eld value and its rst derivative but not
the higher derivatives.
Given boundary conditions, basically a specication of
the value of at the boundary if M is compact or some
limit on as x (this will help in doing integration
by parts), the subspace of C consisting of functions, ,
such that all functional derivatives of S at are zero and
satises the given boundary conditions is the subspace
of on shell solutions.
The solution is given by the EulerLagrange equations
(thanks to the boundary conditions),
S
=
L
( )
)
+
L
= 0.
12
References
10
13
13
13.1
Lagrangian Source: http://en.wikipedia.org/wiki/Lagrangian?oldid=639997735 Contributors: CYD, Zundark, The Anome, Tarquin, Awaterl, Andre Engels, Roadrunner, Peterlin, Stevertigo, Michael Hardy, Anders Feder, AugPi, Andres, Schneelocke, Charles Matthews, Dysprosia, Rednblu, Patrick0Moran, Cameronc, Phys, Bevo, Robbot, Pps, Aetheling, Rho, Jheise, Ahadley, Giftlite, BenFrantzDale, Tom harrison, Art Carlson, Wwoods, Dratman, FrYGuY, Jason Quinn, DefLog, Zeimusu, Karol Langner, Balcer, AmarChandra, Zowie, Jim Fraser,
Masudr, Mal, Bender235, ChristophDemmer, Laurascudder, Army1987, Wisdom89, Matt McIrvin, Physicistjedi, Ardric47, Jrme, Diego
Moya, Wtmitchell, Tbsmith, Linas, Ae-a, Mpatel, SeventyThree, Salix alba, Mathbot, Srleer, Chobot, ChrisChiasson, YurikBot, RussBot,
RL0919, Voidxor, E2mb0t, Larsobrien, Arthur Rubin, Reyk, RG2, Zvika, Eigenlambda, SmackBot, Melchoir, Unyoyega, Papa November, Complexica, Hongooi, Berland, Twalton, Rajkishan211990, Jgates, DJIndica, Lambiam, WhiteHatLurker, Makyen, Dr.K., JRSpriggs,
NormHardy, Cydebot, Xxanthippe, Dchristle, Hugozam, Mbell, LeBofSportif, JAnDbot, MER-C, WolfmanSF, Jpod2, User A1, Sprevrha,
R'n'B, Lilac Soul, DrKiernan, Maurice Carbonaro, KIAaze, Tarotcards, Plasticup, Lseixas, Red Act, Michael H 34, Geometry guy, Lejarrag, Ketyner, Antixt, Austinstkong, SieBot, Gerakibot, MikeGogulski, Tugjob, Correogsk, MenoBot, SchreiberBike, 1ForTheMoney,
Crowsnest, Addbot, CUSENZA Mario, Erik Streb, Luckas-bot, Yobot, AnomieBOT, Xqbot, Gsard, FrescoBot, Craig Pemberton, Biniamin, Kiefer.Wolfowitz, 124Nick, Puzl bustr, Fermat618, Amirhdrz 91, EmausBot, Rami radwan, GoingBatty, HolyCookie, Quondum,
Zueignung, Davidaedwards, ChuispastonBot, Ebehn, Gilderien, Diogenes2000, Helpful Pixie Bot, CedricMC, Bart van den Broek, Odysei, F=q(E+v^B), Pratyush Sarkar, Andyhowlett, Saloon.cat, Pdecalculus, Msurajit, MutluMan, L Minh Nht, Ndeine, Susilehtola, Cap'n
Squid, Anthul, Dierential 0celo7 and Anonymous: 118
13.2
Images
13.3
Content license