Chapter 8.

Partial Dierential Equations

8.1 Partial Dierential Equations
A partial dierential equation (PDE) is an
equation containing an unknown function u(x, y, . . .)
of two or more independent variables x, y, . . . and
its partial derivatives with respect to these variables.
We call u the dependent variable.
8.1.1 Example
(i) u
xy
2x + y = 0
This is a PDE that involves the function u(x, y) with
two independent variables x and y. [Remember that
the subscripts mean that you are taking the partial
derivative, in this case a second order derivative rst
2 MA1506 Chapter 8. Partial Dierential Equations
with respect to x and then with respect to y.]
(ii) w
xy
+ x(w
z
)
2
= yz
This is a PDE that involves the function w(x, y, z)
with three independent variables x, y and z.
8.1.2 Example
The function
u(x, y) = x
2
y
1
2
xy
2
+ F(x) + G(y) (1)
is a solution of the PDE in example 8.1.1 (i). Here
F and G can be any (arbitrary) single variable func-
tions.
Indeed, by taking partial derivatives of (1):
u
x
= 2xy
1
2
y
2
+ F

(x) and
2
3 MA1506 Chapter 8. Partial Dierential Equations
u
xy
= 2x y,
we see that the function (1) satises the PDE. NO-
TICE that, just as the solution of an ordinary dif-
ferential equation involves arbitrary CONSTANTS,
the solution of a PDE will involve arbitrary FUNC-
TIONS!
Suppose we require the PDE to also satisfy the con-
ditions
u(x, 0) = x
3
and u(0, y) = sin(3y).
Then using (1), we have
x
3
= u(x, 0) = F(x) + G(0)
and
sin(3y) = u(0, y) = F(0) + G(y).
3
4 MA1506 Chapter 8. Partial Dierential Equations
In this case, we can simply take F(x) = x
3
and
G(y) = sin(3y) and get the particular solution
u(x, y) = x
2
y
1
2
xy
2
+ x
3
+ sin(3y)
which satises the additional conditions.
8.1.3 Example
In general, the totality of solutions of a PDE is very
large.
The Laplace equation u
xx
+u
yy
= 0 has the follow-
ing solutions
u(x, y) = x
2
y
2
, u(x, y) = e
x
cos y,
u(x, y) = ln (x
2
+ y
2
), etc
which are entirely dierent from each other.
4
5 MA1506 Chapter 8. Partial Dierential Equations
8.1.4 Order of Dierential Equations
The order of the PDE is the order of the highest
derivative present.
Example 8.1.1 (i) is a PDE of order 2 and (ii) is also
a PDE of order 2.
8.1.5 Linearity and Homogeneity
An order 1 linear PDE has the form
Au
x
+ Bu
y
+ Cu = Z
and an order 2 linear PDE has the form
Au
xx
+ Bu
xy
+ Cu
yy
+ Du
x
+ Eu
y
+ Fu = Z
where A, B, C, D, E, F, Z are constants or functions
of x and y but not functions of u.
5
6 MA1506 Chapter 8. Partial Dierential Equations
An order 1 or 2 linear PDE is said to be homogeneous
if the Z term in the above form is 0.
8.1.6 Example
PDE order linear homogeneous
4u
xx
u
t
= 0 2 yes yes
x
2
R
yyy
= y
3
R
xx
3 yes yes
tu
tx
+ 2u
x
= x
2
2 yes no
4u
xx
uu
t
= 0 2 no n.a.
(u
x
)
2
+ (u
y
)
2
= 2 1 no n.a.
8.1.7 Superposition Principle
If u
1
and u
2
are any solutions of a linear homogeneous
dierential equation, then
u = c
1
u
1
+ c
2
u
2
,
where c
1
and c
2
are any constants, is also a solution
of that equation.
6
7 MA1506 Chapter 8. Partial Dierential Equations
8.1.8 Example
Referring to the particular solutions of Laplace equa-
tion u
xx
+u
yy
= 0 in Example 8.1.3, by superposition
principle,
u(x, y) = 3(x
2
y
2
) 7e
x
cos y + 10 ln (x
2
+ y
2
)
is again a solution of the Laplace equation.
8.1.9 Separation of Variables for PDE
This method can be used to solve PDE involving
two independent variables, say x and y, that can be
separated from each other in the PDE There are
similarities between this method and the technique
of separating variables for ODE in Chapter 1.We rst
7
8 MA1506 Chapter 8. Partial Dierential Equations
make an observation:
Suppose u(x, y) = X(x)Y (y).
Then
(i) u
x
(x, y) = X

(x)Y (y)
(ii) u
y
(x, y) = X(x)Y

(y)
(iii) u
xx
(x, y) = X

(x)Y (y)
(iv) u
yy
(x, y) = X(x)Y

(y)
(v) u
xy
(x, y) = X

(x)Y

(y)
Notice that each derivative of u remains separated
as a product of a function of x and a function of y.
We exploit this feature as follows:
8
9 MA1506 Chapter 8. Partial Dierential Equations
8.1.10 Illustration of Separation of Variables
Consider a PDE of the form
u
x
= f(x)g(y)u
y
.
If a solution of the form u(x, y) = X(x)Y (y) exists,
then we obtain
X

(x)Y (y) = f(x)g(y)X(x)Y

(y)
i.e.,
1
f(x)

(x)
X(x)
= g(y)
Y

(y)
Y (y)
.
LHS is a function of x only while RHS is a function
of y only. We conclude that
LHS = RHS = some constant k.
9
10 MA1506 Chapter 8. Partial Dierential Equations
Thus, we obtain two ODEs
1
f(x)

(x)
X(x)
= k X

(x) = kf(x)X(x) (2)

g(y)
Y

(y)
Y (y)
= k Y

(y) =
k
g(y)
Y (y) (3)
Note that (2) is an ODE with independent variable x
and dependent variable X while (3) is an ODE with
independent variable y and dependent variable Y .
By solving (2) and (3) respectively for X(x) and
Y (y), we obtain the solution u(x, y) = X(x)Y (y).
8.1.11 Example
Solve u
x
+ xu
y
= 0.
Solution: If a solution u(x, y) = X(x)Y (y) exists,
10
11 MA1506 Chapter 8. Partial Dierential Equations
then we obtain
X

(x)Y (y) + xX(x)Y

(y) = 0
i.e.,
1
x

(x)
X(x)
=
Y

(y)
Y (y)
(4)
This gives two ODEs :
LHS of (4) = k gives X

= kxX.
This ODE has general solution
X(x) = Ae
kx
2
/2
(a)
Similarly, RHS of (4) = k gives Y

= kY .
This ODE has general solution
Y (y) = Be
ky
(b)
Multiplying (a) and (b), we obtain a solution of the
11
12 MA1506 Chapter 8. Partial Dierential Equations
PDE
u(x, y) = X(x)Y (y) = Ce
k(x
2
/2 y)
.
8.2 The Wave Equation
Suppose you have a very exible string [meaning that
it does not resist bending at all] which lies stretched
tightly along the x axis and has its ends xed at x =
0 and x = . Then you pull it in the y-direction so
that it is stationary and has some specied shape, y
= f(x) at time t = 0 [so that f(0)=0 and f() = 0].
We can assume that f(x) is continuous and bounded,
but we will let it have some sharp corners [but only
a nite number of them.]
What will happen if you now let the string go? Clearly
12
13 MA1506 Chapter 8. Partial Dierential Equations
the string will start to move. We assume that the
only forces acting are those due to the tension in the
string, and that the pieces of the string will only move
in the y-direction.
Now the y-coordinate of any point on the string will
become a function of time as well as a function of
x. So it becomes a function y(t,x) of both t and
x, and we have to use partial derivatives when we
dierentiate it. This function satises
y(t, 0) = 0 y(t, ) = 0
for all t, because the ends are nailed down, also
y(0, x) = f(x)
13
14 MA1506 Chapter 8. Partial Dierential Equations
and
y
t
(0, x) = 0,
because the string is initially stationary. Notice that
we need four pieces of information here, and it is
useful to remember that.
Suppose that the mass per unit length of the string
is constant and equal to . Then the mass of a small
piece of the string is dx. [We assume that we dont
pull the string too far, so it never bends much, hence
the length of the small piece can be approximated by
dx.] So the mass times the acceleration of the small
piece is dx

2
y
t
2
. The force acting in the y direction
is just the dierence between the y-components of
14
15 MA1506 Chapter 8. Partial Dierential Equations
the tension at the two ends of the piece, and this
turns out to be d(K
y
x
) where K is a certain positive
constant. Using Newtons second law we get
d
_
K
y
x
_
= dx

2
y
t
2
,
or
c
2

2
y
x
2
=

2
y
t
2
,
where c
2
is a positive constant. This is the famous
WAVE EQUATION. Notice that it involves FOUR
derivatives altogether, two involving x, and two in-
volving t. That matches up with the fact we men-
tioned earlier, that we needed FOUR pieces of infor-
mation to nail down a solution [the endpoints, the
initial position, the initial velocity].
15
16 MA1506 Chapter 8. Partial Dierential Equations
Note that the following function solves the wave equa-
tion [with those four conditions]:
y(t, x) =
1
2
_
f(x + ct) + f(x ct)
_
.
Here f(x) gives the initial shape of the string, as
above. [Verify that this does satisfy the PDE, that
y(0,x) = f(x), and that
y
t
= 0 for all x when t = 0, as
it should be. Initially f(x) was only dened between
0 and , but we can extend it to be an odd, periodic
function of period 2, and then you can also verify
that y(t,0) = y(t,) = 0. See Tutorial 11.] This is
called dAlemberts solution of the wave equation.
You can think about f(x-ct) in the following way.
First, think about f(x): it is a function with some
16
17 MA1506 Chapter 8. Partial Dierential Equations
denite shape. Now what is f(x-1)? It is exactly the
same shape, but shifted to the right by one unit.
Similarly f(x-2) is the same shape shifted 2 units
to the right, and similarly f(x-ct) is f(x) shifted to
the right by ct. But if t is time, then ct is some-
thing which increases linearly with time, at a rate
controlled by c. In other words, f(x-ct) represents
the shape f(x) moving to the right at a constant
speed c. In other words, it represents a WAVE [of ar-
bitrary shape] moving to the right at constant speed
c. Similarly f(x+ct) represents a wave [with the same
shape as f(x)] moving to the left at constant speed
c. So dAlemberts solution says that the solutions of
17
18 MA1506 Chapter 8. Partial Dierential Equations
the wave equation [with the given boundary and ini-
tial conditions] can be found by superimposing two
waves of those forms.
8.2.1 Separation of Variables for the Wave
Equation
We can also solve the wave equation using the method
of separation of variables. We want to solve
c
2
y
xx
= y
tt
,
y(t, 0) = y(t, ) = 0, y(0, x) = f(x), y
t
(0, x) = 0,
where f(x) is a given function which is zero at 0 and
. Remember that f(x) is continuous but it can have
[a nite number of] sharp corners.
18
19 MA1506 Chapter 8. Partial Dierential Equations
We separate the variables:
y(t, x) = u(x)v(t),
and get
u

(x)
u(x)
=
1
c
2
v

(t)
v(t)
.
The usual separation argument now gives, for some
constant ,
u

+ u = 0,
v

+ c
2
v = 0.
Now lets force u(x) to vanish at 0 and , so we can
enforce y(t,0) = u(0)v(t) = 0 and y(t,) = u()v(t)
= 0. NOTICE that this is very dierent from what
we normally do when we try to solve second-order
ODEs. Normally we give some information about
19
20 MA1506 Chapter 8. Partial Dierential Equations
the function at ONE point, for example we might
ask for solutions of u

+u = 0 where u(0) and u

(0)
are given, as in Chapter 1. But here we are giving
the information at TWO dierent points.
If is negative, then u(0) = 0 implies that all solu-
tions of u

+u = 0 are proportional to a hyperbolic

sine function [sinh(x)], and such a function cannot
cut the x axis twice. So cannot be negative. Sim-
ilarly, if = 0, then u(x) is a straight line function
which cannot cut the x axis twice. So has to be pos-
itive, and we can write it as = n
2
for some constant
n, and u(x) = C cos(nx) + D sin(nx) for arbitrary
constants C and D. Since u(0) = 0, we have C = 0
20
21 MA1506 Chapter 8. Partial Dierential Equations
and so u(x) = D sin(nx). If we want u() = 0, this
means sin(n) = 0 and that is correct PROVIDED
that n is an integer [whole number]. So we have to
enforce that.
Solving the other equation for v(t), we get now
v(t) = Acos(nct) + Bsin(nct)
for arbitrary constants A and B. We force v(t) to
satisfy v(0) = 0 because we want y
t
(0, x) = u(x)v(0)
= 0. So now B = 0 and we are left with v(t) =
Acos(nct). So our complete solution is
y(t, x) = bsin(nx)cos(nct),
where b is an arbitrary constant and n is any integer.
This satises all of our conditions... except one! We
21
22 MA1506 Chapter 8. Partial Dierential Equations
still have not satised y(0,x) = f(x).
8.2.2 Fourier Series
This discussion reminds us of Fourier series. Remem-
ber that we have extended the function f(x) to be an
odd function of period 2. So it has a Fourier sine
series, and since f(x) is continuous and has only a
nite number of sharp corners, we have
f(x) =

n=1
b
n
sin(nx).
Actually this is often a good way to describe a func-
tion with sharp corners.
If we are given f(x), then we know how to compute
22
23 MA1506 Chapter 8. Partial Dierential Equations
the coecients b
n
of a Fourier sine series:
b
n
=
2

_

0
f(x)sin(nx)dx.
So we know all of these numbers if we are given f(x).
Now consider the series

n=1
b
n
sin(nx)cos(nct),
where the numbers b
n
are the coecients in the Fourier
series of f(x). Then we notice two things:
[1] If we put t = 0 into this series, then we get exactly
f(x), expressed as its Fourier sine series; and
[2] Since the wave equation is linear, and each term
in this series is a solution of the wave equation, then
this series is also a solution of the wave equation, as
above.
23
24 MA1506 Chapter 8. Partial Dierential Equations
So we have solved the problem: this series is just
exactly what we want, y(t,x)!
Summary: the solution of
c
2
y
xx
= y
tt
,
y(t, 0) = y(t, ) = 0, y(0, x) = f(x), y
t
(0, x) = 0,
is just
y(t, x) =

n=1
b
n
sin(nx)cos(nct),
where the b
n
are just the Fourier sine coecients of
f(x) [regarded as an odd periodic function.]
8.2.3 Example
Consider the function
f(x) = sin(Kx),
24
25 MA1506 Chapter 8. Partial Dierential Equations
where K is some xed integer. You can easily work
out the Fourier sine coecients,
b
n
=
2

_

0
sin(Kx)sin(nx)dx;
using the trigonometric identity for sin(A) sin(B)
you can show that it is zero for all n = K, and you
can also show that it equals 1 when n = K. [Actu-
ally this is just revision of Fourier series theory.] So
b
K
= 1 and we see that for this f(x) the whole Fourier
sine series is just sin(Kx) itself, and so the solution
of the wave equation [with our standard initial and
boundary conditions] is just
y(t, x) = sin(Kx)cos(Kct).
25
26 MA1506 Chapter 8. Partial Dierential Equations
So y(t,x) is zero at the places where sin(Kx) is zero,
and this is true at ALL times. For example, sin(3x)
is zero not just at x = 0 and x = , but also at x =
/3 and x = 2/3, and the wave is always zero at
those points [called NODES]. The pattern that you
get in this way is called a STANDING WAVE.
8.3 Heat Equation
Consider the temperature in a long thin bar or wire
of constant cross section and homogeneous material,
which is oriented along the x-axis and is perfectly
insulated laterally, so that heat only ows in x-
direction. Then the temperature u depends only on
x and t and is given by the one-dimensional heat
26
27 MA1506 Chapter 8. Partial Dierential Equations
equation
u
t
= c
2
u
xx
, (5)
where c
2
is a positive constant called the thermal dif-
fusivity [units length
2
/time]. It measures how quickly
heat moves through the bar, and depends on what it
is made of.
8.3.1 Zero temperature at ends of rod
Lets assume that the ends x = 0 and x = L of the
bar are kept at temperature zero, so that we have the
boundary conditions
u(0, t) = 0, u(L, t) = 0 for all t (6)
and the initial temperature of the bar is f(x), so that
27
28 MA1506 Chapter 8. Partial Dierential Equations
we have the initial condition
u(x, 0) = f(x). (7)
Here we will assume that, when f(x) is extended to
be an odd function, it is then equal to its Fourier sine
series everywhere. [Remember that this can happen,
even if f(x) is discontinuous at some points.] We
call the p.d.e (5) together with the conditions (6)
and (7) a boundary value heat equation problem.
Notice that, unlike the wave equation, which needs
four pieces of data, here we only need three, which
matches the fact that the heat equation only involves
a total of three derivatives [two in the spatial direc-
tion, but only one in the time direction].
28
29 MA1506 Chapter 8. Partial Dierential Equations
8.3.2 Example
Solve
u
t
= 2u
xx
, 0 < x < 3, t > 0,
given boundary conditions u(0, t) = 0, u(3, t) = 0,
and initial condition u(x, 0) = 5 sin 4x.
Solution:
We use the method of separation of variables. Let
u(x, t) = X(x)T(t). Then u
t
= 2u
xx
gives
XT

= 2X

T,
or equivalently
X

X
=
T

2T
.
29
30 MA1506 Chapter 8. Partial Dierential Equations
Each side must be a constant k. So
X

kX = 0 (A)
T

2kT = 0 (B)
The solutions of (A) are, as in the cases we studied
earlier, of three types; but clearly we want X(x) to
vanish at TWO values of x [the two ends of the bar].
Of course, hyperbolic sine [k positive] and straight-
line [k = 0] functions cannot do that. So we have to
choose k to be negative, so as to get trigonometric
functions which can vanish at two values of x. So we
have
X(x) = a cos

kx + b sin

kx. (8)
30
31 MA1506 Chapter 8. Partial Dierential Equations
The solutions of (B) are
T(t) = de
2kt
for the same negative value of k as in (A).
So now we have a simple solution of the heat equa-
tion, by just multiplying X(x) into T(t).
Now we just have to use the boundary conditions.
They can be written as
X(0)T(t) = 0 and X(3)T(t) = 0.
Now since T(t) = de
2kt
= 0 for any t, we conclude
that X(0) = 0 and X(3) = 0.
Substituting x = 0 and 3 separately into our expres-
31
32 MA1506 Chapter 8. Partial Dierential Equations
sion for X(x), we get
X(0) = a = 0
X(3) = a cos 3

k + b sin 3

k = 0
Solving these two equations, we get
a = 0 and b sin 3

k = 0.
Since we do not want a and b both zero, this implies
sin 3

k = 0 which gives

k =
n
3
or k =
n
2

2
9
where n = 0, 1, 2, . . .
Putting this back into our general solution and ab-
sorbing d into b [since the product of arbitrary con-
stants is another arbitrary constant] we have
u
n
(x, t) = b
n
e
2n
2

2
t/9
sin
nx
3
(HS
n
)
32
33 MA1506 Chapter 8. Partial Dierential Equations
is a solution for each n = 1, 2, 3, . . ..
We have used up two of our pieces of information.
But one remains: we have to satisfy
u(x, 0) = 5 sin 4x.
We want to construct a solution from among (HS
n
)
that satises this initial condition.
Now substituting t = 0 into (HS
n
) for any n,
u
n
(x, 0) = b
n
sin
nx
3
.
If we take n = 12 and b
12
= 5, we have
u
12
(x, 0) = 5sin
12x
3
= 5sin4x
.
Hence, the particular solution that will also satisfy
33
34 MA1506 Chapter 8. Partial Dierential Equations
the initial condition is
u
12
(x, t) = 5e
32
2
t
sin 4x.
8.3.3 Example
Find the temperature u(x, t) of the bar in Example
8.3.2 if the initial temperature of the rod is (i) given
by the function sin
5
x; (ii) 25

C throughout the rod

EXCEPT the two end points, where it is zero so
in other words this function is discontinuous. That
is not actually a problem, because even though it is
discontinuous, the odd extension of this function has
a Fourier sine series which is equal to it everywhere,
even at the discontinuities. [As usual, the Fourier
series converges to the average value at any jump
34
35 MA1506 Chapter 8. Partial Dierential Equations
discontinuity.]
Solution: Since the p.d.e and the boundary con-
dition [the fact that the temperatures at the ends are
both zero] is the same as in Example 8.3.2, the cal-
culations are identical. The only thing we need to
change is the initial condition.
(i) u(x, 0) = sin
5
x. Using trigonometric identities,
this function can be rewritten as
sin
5
x =
5
8
sin x
5
16
sin 3x +
1
16
sin 5x.
This time, to get a solution that satises the initial
condition, it is insucient to use only a single u
n
(x, t)
from (HS
n
). Instead, we need to take three appro-
priate u
n
(x, t) and take their linear combination. By
35
36 MA1506 Chapter 8. Partial Dierential Equations
the superposition principle, this combination will still
satisfy the p.d.e and the boundary conditions.
If we choose n = 3, n = 9, n = 15 separately, then
we get the three terms in the initial condition.
Therefore, the particular solution is of the form
u(x, t) = u
3
(x, t) + u
9
(x, t) + u
15
(x, t)
= b
3
e
2
2
t
sin x + b
9
e
18
2
t
sin 3x + b
15
e
50
2
t
sin 5x
By comparing coecients, we need to choose b
3
=
5
8
, b
9
=
5
16
, b
15
=
1
16
in order to satisfy the initial
condition. That is, the particular solution in this case
is
u(x, t) =
5
8
e
2
2
t
sin x
5
16
e
18
2
t
sin 3x+
1
16
e
50
2
t
sin 5x.
(ii) u(x, 0) = 25, except at the ends, where it is
36
37 MA1506 Chapter 8. Partial Dierential Equations
zero. The best way to think about this discontinuous
function is to take the constant function y = 25 on
the open interval 0 < x < 3 and extend it to an
odd periodic function of period 6, so the function is
zero [like every odd function] at x = 0. Lets call
this function f(x). It has a Fourier sine series [of the
form

n=1
b
n
sin
nx
3
, because the period is 6], and
this series converges to f(x) everywhere [including at
the jump points!] The Fourier sine coecients for
f(x) are
b
n
=
2
L
_
L
0
f(x) sin
nx
L
dx
=
2
3
_
3
0
25 sin
nx
3
dx =
50
n
(1 cos n);
of course, the fact that f(x) vanishes at the end points
37
38 MA1506 Chapter 8. Partial Dierential Equations
has no eect on the integral.
Now from our previous examples you should be able
to see that the solution is
u(x, t) =

n=1
b
n
e
2n
2

2
t/9
sin
nx
3
. (9)
We know this because its clear that each term in
this series separately is a solution which is zero at
the endpoints, and of course when t = 0, equation
(8) yields
f(x) = u(x, 0) =

n=1
b
n
sin
nx
3
.
38
39 MA1506 Chapter 8. Partial Dierential Equations
We can write out the solution in this case as
u(x, t) =

n=1
50
n
(1 cos n)e
2n
2

2
t/9
sin
nx
3
=
100

_
e
2
2
t/9
sin
x
3
+
1
3
e
2
2
t
sin x +
_
.
39