Field Theory
String dynamics
In this section we consider two closely related problems, transverse oscillations of a stretched loaded string, and of a stretched heavy string. The latter
is is a limiting case of the former. This will provide an introduction to field
theory, in which the dynamical degrees of freedom are not a discrete set but
are defined at each point in space. Later we will discuss more interesting
and involved cases such as the electromagnetic field, where at each point
~ and B
~ as degrees of freedom, though not without conin space we have E
straints. Then we will consider even more interesting fields, transforming
under a nonabelian gauged symmetry group.
The loaded string we will consider is a light string under tension stretched
between two fixed points a distance apart, say at x = 0 and x = . On
the string, at points x = a, 2a, 3a, . . . , na, are fixed n particles each of mass
m, with the first and last a distance a away from the fixed ends. Thus
= (n + 1)a. We will consider only small transverse motion of these masses,
using yi as the transverse displacement of the ith mass, which is at x = ia.
We assume all excursions from the
equilibrium positions yi = 0 are
small, and in particular that the dify
ference in successive displacements
yi+1 yi a. Thus we are asx
a
suming that the angle made by
i 1
i +1
i
each segment of the string, i =
tan1 [(yi+1 yi)/a] 1.
Working to first order in the s in the equations of motion, and second
order for the Lagrangian, we see that restricting our attention to transverse
motions and requiring no horizontal motion forces taking the tension to be
constant along the string. The transverse force on the ith mass is thus
Fi =
yi+1 yi
yi1 yi
+
= (yi+1 2yi + yi1 ).
a
a
a
The potential energy U(y1 , . . . , yn ) then satisfies
U
= (yi+1 2yi + yi1 )
yi
a
1
so
U(y1 , . . . , yi, . . . , yn )
Z yi
dyi (2yi yi+1 yi1 ) + F (y1 , . . . , yi1 , yi+1 , . . . , yn )
=
a
0
2
=
yi (yi+1 + yi1)yi + F (y1, . . . , yi1 , yi+1 , . . . , yn )
a
(yi+1 yi )2 + (yi yi1 )2 + F (y1 , . . . , yi1, yi+1 , . . . , yn )
=
2a
n
X
=
(yi+1 yi )2 + constant.
i=0 2a
The F and F are unspecified functions of all the yj s except yi. In the last
expression we satisfied the condition for all i, and we have used the convenient
definition y0 = yn+1 = 0. We can and will drop the arbitrary constant.
P
The kinetic energy is T = 21 m n1 y i2.
P
The kinetic energy is simply T = 12 m n1 y i2. The potential energy U =
1 T
y A y has a non-diagonal n n matrix
2
2 1
0 0 0
0
1
2 1 0 0
0
1 2 1 0
0
0
A = ..
.
..
..
.. . .
..
..
.
a .
.
.
.
.
.
0
0
0 0 2 1
0
0
0 0 1 2
L
d L
=
=
dt yi
yi
(Ay)i . While this involves an indefinite number of coupled degrees of freedom, it is not hard to find the general solution,
The Lagrangian L = T U and Lagranges equation tells us
y(ja, t) =
Re Bp eip t sin(kp ja),
with kp = p/, p = 1 . . . n, with p = 2 /am sin(kp a/2). We have arbitrary (complex) amplitudes Bp for each mode p. That is interesting for solid
state physics, but we are more interested in the continuum limit, with a view
to understanding how to formulate continuum mechanics.
Consider the limit in which the length is held fixed, but the number of
masses n , a = /(n + 1) 0 with each mass decreasing so that the
2
linear density = m/a is held constant. This constitutes the continuum
limit. The function y(ja) which had been defined only at discrete values of
x = ja will be assumed to become a continuous function of x1 .
What happens to the kinetic and potential energies in this limit? For the
kinetic energy,
1 X 2 1 X 2
1 X
1
T = m
yi =
ay (xi ) =
x y 2 (xi )
2
2 i
2 i
2
i
dx y 2(x),
where the next to last expression is just the definition of a Riemann integral.
For the potential energy,
X
X
yi+1 yi
U=
(yi+1 yi )2 =
x
2a i
2 i
x
2
dx
y
x
!2
The equation of motion for yi is
m
yi =
L
U
=
= [(yi+1 yi ) (yi yi1 )],
yi
yi
a
or
([y(x + a) y(x)] [y(x) y( x a)]).
a
We need to be careful about taking the limit
a
y (x) =
y
y(x + a) y(x)
a
x
because we are subtracting two such expressions evaluated at nearby points,
and because we will need to divide by a again to get an equation between
finite quantities. Thus we note that
y(x + a) y(x)
y
+ O(a2 ),
=
a
x x+a/2
so
y(x + a) y(x) y(x) y( x a)
a
a
y (x) =
a
y
y
2y
,
a x x+a/2 x xa/2
x2
This means that the nodes Bp are unrestricted for finite p, but Bn = rightarrow0
for fixed nonzero . The acoustic modes remain but the optical modes dont.
Thus
p
sin(kp a)/a = kp = k/ and we have a nondispersive wave with speed c = /.
and we wind up with the wave equation for transverse waves on a massive
string
2
2y
2 y
c
= 0,
t2
x2
where
s
.
c=
Field theory
We now examine how to formulate the continuum limit directly.
2.1
Lagrangian density
We saw in the last section that the kinetic and potential energies in the
continuum limit can be written as integrals over x of densities, and so we
may also write the Lagrangian as the integral of a Lagrangian density
L(x),
L=T U =
dx L(x),
1
1
L(x) = y 2(x, t)
2
2
y(x, t)
x
!2
This Lagrangian, however, will not be of much use until we figure out what is
meant by varying it with respect to each dynamical degree of freedom or its
corresponding velocity. In the discrete case we have the canonical momenta
Pi = L/ yi , where the derivative requires holding all yj fixed, for j 6= i, as
well as all yk fixed. In the continuum, however, this notion is a bit dubious
how can we vary y(x
0 ) at one point x0 while holding y(x)
fixed at all
other x? In the discrete case, this variation extracts one term from the sum
1 P
ayi2 , and this would appear to vanish in the limit a 0. Instead, we
2
define the canonical momentum as a density, Pi aP (x = ia), so
P (x = ia) = lim
1 X
a L(y(x), y(x),
x)|x=ai .
a yi i
We may think of the last part of this limit,
lim
a0
a L(y(x), y(x),
x)|x=ai =
dx L(y(x), y(x),
x),
if we also define a limiting operation
1
,
a0 a yi
y(x)
lim
1
,
a yi
and similarly for
which act on functionals of y(x) and y(x)
by
y(x1 )
= (x1 x2 ),
y(x2 )
y(x
1)
y(x1 )
=
= 0,
y(x2 )
y(x
2)
y(x
1)
= (x1 x2 ).
y(x
2)
where (x x) is the Dirac delta function2 Thus
P (x) =
y(x)
1
dx y 2 (x , t) =
2
dx y(x
, t)(x x) = y(x,
t).
We also need to evaluate
L=
y(x)
y(x)
dx
y
x
!2
x=x
For this we need3
y(x )
=
(x x) := (x x),
y(x) x
x
Thus
y
2y
dx (x ) (x x) = 2 ,
L=
y(x)
x
x
0
and Lagranges equations give the wave equation
y (x, t)
2y
= 0.
x2
(1)
Rx
The Dirac delta function is defined by its integral, x12 f (x )(x x)dx = f (x) for
any function f (x), provided x (x1 , x2 ).
3
which is again defined by its integral,
Z x2
Z x2
f (x ) (x x)dx
f (x ) (x x)dx =
x
x1
x1
Z x2
f
x
dx (x x)
= f (x )(x x)|x21
x
x1
f
=
(x),
x
2
where after integration by parts the surface term is dropped because (x x ) = 0 for
x 6= x , which it is for x = x1 , x2 if x (x1 , x2 ).
2.2
Lagrangian Mechanics for 3-D Fields
In sections 1 and 2.1 we considered the continuum limit of a chain of point
masses on stretched string. We had a situation in which the potential energy
had interaction terms for particle A which depended only on the relative displacements of particles in the neighborhood of A. If we generalize to motion
of a three-dimensional material, the displacements from equilibrium will be
vectors ~(~r, t), and we expect the potential energy to be integrals over volume
of a function of ~ (~r, t) and its spatial derivatives. More generally, ~ could
be some other fields4 . The dynamics is then determined by a Lagrangian
density
i i i i
L = L(i ,
,
,
,
, x, y, z, t)
x y z t
R
with lagrangian L = dx dy dz L and action I = dx dy dz dt L.
The actual motion of the system will be given by a particular set of
functions i (x, y, z, t), which are functions over the volume in question and
of t [tI , tf ]. The function will be determined by the laws of dynamics of
the system, together with boundary conditions which depend on the initial
configuration i (x, y, z, tI ) and perhaps a final configuration. Generally there
are some boundary conditions on the spatial boundaries as well. For example,
our stretched string required y = 0 at x = 0 and x = L, for all values of t.
Before taking the continuum limit we say that the configuration of the
system at a given t was a point in a large N dimensional configuration space,
and the motion of the system is a path (t) in this space. In the continuum
limit N , so we might think of the path as a path in an infinite dimensional space. But we can also think of this path as a mapping t (, , , t)
of time into the (infinite dimensional) space of functions on ordinary space.
Hamiltons principal says that the actual path is an extremum of the
action. If we consider small variations i (x, y, z, t) which vanish on the
boundaries, then
Z
I = dx dy dz dt L = 0
determines the equations of motion.
Note that what is varied here are the functions i , not the coordinates
(x, y, z, t). x, y, z do not represent the position of some atom they represent
a label which tells us which atom it is that we are talking about. Often they
are chosen to be the equilibrium position of that atom, but they are fixed
4
In the physicists definition, a function of space and time, not the mathematicians.
labels independent of the motion. It is the i (~x), for each ~x, which are the
dynamical degrees of freedom, specifying the configuration of the system. In
our discussion of section 2 i specified the displacement from equilibrium,
but here we generalize to an arbitrary set of dynamical fields.
The variation of the Lagrangian density is
i i i i
,
,
,
, x, y, z, t)
x y z t
X L
X
L
i X
L
i
=
i +
i i
i (i /x) x
i (i /y) y
X
X
L
i
i
L
.
+
i (i /t) t
i (i /z) z
L(i ,
Notice there is no variation of x, y, z, and t, as we discussed.
The notation is getting awkward, so we need to reintroduce the notation
A,j = A/rj , for rj = (x, y, z). In fact, we see that /t enters in the same
way as /x, so it is time to introduce notation which will become crucial
when we consider relativistic dynamics, even though we are not doing so
here. So we will consider time to be an additional component of the position,
called the zeroth rather than the fourth component. We will also change our
notation for coordinates to anticipate needs from relativity, by writing the
indices of coordinates as superscripts rather than subscripts. Thus we write
x0 = ct, where c will eventually be taken as the speed of light, but for the
moment is an arbitrary scaling factor. Until we get to special relativity,
one should consider whether an index is raised or lowered as irrelevant, but
they are written here in the place which will be correct once we make the
distinction between them. In particular the Kronecker delta is now written
. For the partial derivatives we now have
:= =
x
,
, , ,
ct x y z
for = 0, 1, 2, 3, and write , := . If there are several fields i , then
i = i, . The comma represents the beginning of differentiation, so we
must not use one to separate different ordinary indices.
In this notation, we have
L =
X
i
3
XX
L
L
i +
i, ,
i
i =0 i,
and
I =
X
i
3
XX
L
L
i +
i, d4 x,
i
i,
i =0
where5 d4 x = cdx dy dz dt. Except for the first term, we integrate by parts,
I =
X
i
3
L X X
L
i d4 x,
i
i,
i =0
where we have thrown away the boundary terms which involve i evaluated
on the boundary, which we assume to be zero. Inside the region of integration,
the i are independent, so requiring I = 0 for all functions i (x ) implies
X
d L
L
= 0.
dx i, i
(2)
We have written the equations of motion (which are now partial differential equations rather than coupled ordinary differential equations), in a
form which looks like we are dealing with a relativistic problem, because t
and spatial coordinates are entering in the same way. We have not made
any assumption of relativity, however, and our problem will not be relativistically invariant unless the Lagrangian density is invariant under Lorentz
transformations (as well as translations).
I have given you this introduction to continuum mechanics chiefly so we
can discuss gauge field theories, so I am not pursuing very useful ideas such
as the energy-momentum (or stress-energy) tensor and Noethers theorem6 ,
and the actual description of motion of solid bodies or fluids.
We have also multiplied I by c, which does no harm in finding the extrema.
You can continue this discussion at http://www.physics.rutgers.edu/shapiro/507/book9.pdf
page 236, which is where this leaves off.
6
