Вы находитесь на странице: 1из 7

3.

3: WAVE EQUATION AND THE METHOD OF


SEPARATION OF VARIABLES
KIAM HEONG KWA

The goal of this section is to present the method of separation of


variables as an approach to solve the one-dimensional wave equation
2u
2 u
=
c
, 0 < x < L, t > 0,
t2
x
where c > 0, subject to the boundary conditions

(1.1)

(1.2)

u(0, t) = u(L, t) = 0 for all t > 0

and the initial conditions


(1.3)

u(x, 0) = f (x) and

u
(x, 0) = g(x) for 0 < x < L.
t

Step 1: Separating Variables in (1.1) and (1.2). We begin by


searching for nonzero product solutions1 of (1.1) of the form
(1.4)

u(x, t) = X(x)T (t),

where X is a function of x alone and T is a function of t alone. It is


worth noting that
(1.5)

X(x) 6= 0 and T (t) 6= 0

because we have assumed that such a product solution u = XT is


nonzero. Differentiating (1.4) with respect to t and x yields
2u
2u
00
=
X(x)T
(t)
and
= X 00 (x)T (t).
t2
x2
Plugging these into (1.1) and dividing the resulting equation by c2 XT
gives
(1.6)

(1.7)

T 00 (t)
X 00 (x)
=
.
c2 T (t)
X(x)

Date: April 14, 2011.


1As we will see soon, these product solutions are building blocks of more general
solutions. We cannot really build any solutions from the trivial one, whence the
assumption that the product solutions are nonzero.
1

KIAM HEONG KWA

This way we have separated the variables in the sense that the left
side of (1.7) is a function of t alone, while the right side is a function
of x alone. Since t and x are variables independent of each other, this
can occur only if both sides of (1.7) are constant and equal2:
(1.8)

T 00 (t)
X 00 (x)
=
= k.
c2 T (t)
X(x)

Such a constant k is called a separation constant. Note that (1.8)


holds if and only if so do
X 00 kX = 0

(1.9)
and

T 00 kc2 T = 0.

(1.10)

So far we have separated the variables t and x in (1.1).


We can also separate the variables in the boundary conditions (1.2).
By (1.2) and (1.4), we have
(1.11)

X(0)T (t) = X(L)T (t) = 0 for all t > 0.

If X(0) 6= 0 or X(L) 6= 0, then T (t) = 0 for all t > 0. This will not
consist with (1.5). Hence we set
(1.12)

X(0) = X(L) = 0.

Step 2: Solving the Separated Equations (1.9) and (1.10). Combining (1.9) and (1.12), we arrive at a boundary value problem in X:
(1.13)

X 00 kX = 0, X(0) = X(L) = 0.

It should be noted that not all values of the separation constant k yield
nontrivial solutions for (1.13).
If k > 0, so that k = 2 with > 0, then
X 00 2 X = 0,
from which it follows that
X(x) = c1 cosh x + c2 sinh x,
2Let

f (t) = g(x), where f is a function of t alone and g is a function of x alone.


df
f
g
dg
g
f
Then
=
=
= 0 and
=
=
= 0, from which it follows that f
dt
t
t
dx
x
x
and g are constant and equal.

3.3: WAVE EQUATION AND THE METHOD OF SEPARATION OF VARIABLES3

where c1 and c2 are integration constants. Then the boundary conditions X(0) = X(L) = 0 imply that
c1 = c1 cosh 0 + c2 sinh 0 = X(0) = 0,
c2 sinh L = c1 cosh L + c2 sinh L = X(L) = 0.
This in turn implies that c1 = c2 = 0, so that X = 0 if k > 0.
If k = 0, then
X 00 = 0,
whence it follows that
X(x) = c1 x + c2 ,
where c1 and c2 are integration constants. The boundary conditions
X(0) = X(L) = 0 imply that
c2 = c1 0 + c2 = X(0) = 0,
c1 = c1 L + c2 = X(L) = 0.
So, X = 0 if k = 0.
Finally, suppose that k < 0, so that k = 2 with > 0. Then
X 00 + 2 = 0.
The general solution of this equation is
X(c) = c1 cos x + c2 sin x,
where c1 and c2 are integration constants. The boundary conditions
X(0) = X(L) = 0 imply that
c1 = c1 cos 0 + c2 sin 0 = X(0) = 0,
c2 sin L = c1 cos L + c2 sin L = X(L) = 0.
Clearly, c1 = 0, so that X(x) = c2 sin x. Also, from the boundary
condition X(L) = 0, we have either c2 = 0 or sin L = 0. If c2 = 0,
then X = 0. To avoid trivializing X, we take c2 = 1 (or any nonzero
value of c2 ) and sin L = 0, so that
L = n L = n, n = 1, 2, ,
or equivalently
= n =

n
, n = 1, 2, ,
L

from which it follows that


(1.14)

k = kn = 2n =

n2 2
, n = 1, 2, ,
L2

KIAM HEONG KWA

are the only values of the separation constant for which (1.13) admits
nontrivial solutions. In addition, for each of these values k = kn ,
nx
(1.15)
X(x) = Xn (x) = sin n x = sin
L
is such a nontrivial solution.
For each n, n = 1, 2, , substituting k by kn = 2n in (1.10) yields
T 00 + c2 2n T = 0,
the general solution of which has the form
(1.16)

Tn (t) = bn cos cn t + bn sin cn t = bn cos

cnt
cnt
+ bn sin
,
L
L

where bn and bn are integration constants.


Combining (1.15) and (1.16) yields the following infinite set of product solutions of (1.1) subject to the boundary conditions (1.2):
(1.17)


nx
cnt
cnt

un (x, t) = Xn (x)Tn (t) = sin


bn cos
+ bn sin
, n = 1, 2, .
L
L
L
These are called the normal modes of the wave equation.
Step 3: Fourier Series Solution of the Entire Problem. In order
to take care of the general initial conditions (1.3), we form the series
solution
(1.18)



X
X
nx
cnt
cnt

u(x, t) =
un (x, t) =
sin
+ bn sin
bn cos
L
L
L
n=1
n=1
from the normal modes (1.17). Differentiating this series solution term
by term with respect to t yields



nx
cnt
cnt
u X cn

=
sin
bn sin
+ bn cos
.
(1.19)
t
L
L
L
L
n=1
Then the initial conditions (1.3) demand that

X
nx
(1.20)
f (x) =
bn sin
L
n=1
and
(1.21)

g(x) =

X
cnb

n=1

sin

nx
.
L

3.3: WAVE EQUATION AND THE METHOD OF SEPARATION OF VARIABLES5

These equations can be fulfilled by setting


Z
2 L
nx
(1.22)
bn =
f (x) sin
dx
L 0
L
and
(1.23)

bn

2
=
cn

g(x) sin
0

nx
dx
L

cnbn
for n = 1, 2, ; that is, by setting bn and
as the coefficients in
L
the Fourier sine series of f and g respectively.
1
Example 1 (Exercise 3.3.2(a) in the text). With L = 1, c = , and

the initial data


1
1
f (x) = sin x cos x = sin 2x, g(x) = 0, and c = ,
2

the series solution (1.18) of the wave equation needs to satisfy the
boundary conditions
u(x, 0) =

bn sin nx =

n=1

1
sin 2x
2

and

X
u
(x, 0) =
nbn sin nx = 0.
t
n=1
These can be achieved by setting
(
1
if n = 2,
and bn = 0 for all n.
bn = 2
0 otherwise;
So the solution of the wave equation is
u(x, t) =

1
sin 2x cos 2t.
2

Example 2 (Exercise 3.3.4(a) in the text). With L = 1, c = 1, and


the initial data
1
f (x) = sin x + sin 3x + 3 sin 7x, g(x) = sin 2x,
2

KIAM HEONG KWA

the series solution (1.18) of the wave solution needs to satisfy the boundary conditions
u(x, 0) =

bn sin nx = sin x +

n=1

1
sin 3x + 3 sin 7x
2

and

X
u
(x, 0) =
nbn sin nx = sin 2x.
t
n=1
This can be achieved by setting

1 if n = 1,

1 if n = 3,
1 if n = 2,

bn = 2
and nbn =

3 if n = 7,
0 otherwise.

0 otherwise;
So the solution of the wave equation is
u(x, t) = sin x cos t+

1
1
sin 2x sin 2t+ sin 3x cos 3t+3 sin 7x cos 7t.
2
2

Example 3 (Exercise 3.3.6(a)


the initial data

0

1
f (x) = 30
x 31

1 (1 x)
30

in the text). With L = 1, c =

1
, and

if 0 x 31 ,
if 13 x 23 , and g(x) = 2,
if 32 < x 1;

the series solution (1.18) of the wave solution needs to satisfy the boundary conditions

if 0 x 13 ,

0
X

1
u(x, 0) =
bn sin nx = f (x) = 30
x 13
if 13 x 32 ,

1 (1 x) if 2 < x 1
n=1
30
3
and

X
u
(x, 0) =
nbn sin nx = g(x) = 2.
t
n=1

3.3: WAVE EQUATION AND THE METHOD OF SEPARATION OF VARIABLES7

This can be achieved by setting


Z 1
bn = 2
f (x) sin nx dx
0

1/3

2/3

0 sin nx dx + 2

=2

1/3

1
30



1
x
sin nx dx
3

1
(1 x) sin nx dx
2/3 30


2/3
1
cos nx
sin nx
2
x

1 2 2
=
30
3
n
n
1/3

1
cos nx
sin nx
2
(1 x)
(1) 2 2
+
30
n
n
2/3


1
n
2n
=
sin
2 sin
2
2
15n
3
3
+2

and
bn

Z
2 1
=
g(x) sin nx dx
n 0
4
= 2 [1 (1)n ]
n

8
if n is odd,
= n2
0
if n is even

for n = 1, 2, . So the solution of the wave equation is






X
1
n
2n
u(x, t) =
sin nx
sin
cos nt
2 sin
22
15n
3
3
n=1

4
n
+ 2 [1 (1) ] sin nt
n



X sin nx  1 
2n
n
n
=
2 sin
sin
cos nt + 4[1 (1) ] sin nt .
2
n
15
3
3
n=1

Вам также может понравиться