Applications of PDEs

25.2

Introduction
In this Section we discuss briey some of the most important PDEs that arise in various branches
of science and engineering. We shall see that some equations can be used to describe a variety
of dierent situations.

Prerequisites
Before starting this Section you should . . .

There are no specic prerequisites for this

Section

Learning Outcomes
After completing this Section you should be
able to . . .

recognise the heat conduction equation,

wave equation and have some knowledge
of their spheres of application

The Key Point at the end of Section 1 by no means exhausts the types of PDE which are
important in applications. In this Section we will discuss the three PDEs in the Key Point in
more detail and briey discuss other PDEs over a wide range of applications: We will omit
detailed derivations.

1. Wave Equations
The simplest situation to give rise to the one-dimensional wave equation is the motion of a
stretched string - specically the transverse vibrations of a string such as the string of a
musical instrument. Assume that a string is placed along the xaxis, is stretched and then
xed at ends x = 0 and x = L; it is then deected and at some instant, which we call t = 0, is
released and allowed to vibrate. The quantity of interest is the deection u of the string at any
point x, 0 x L, and at any time t > 0. We write u = u(x, t). The diagram shows a possible
displacement of the string at a xed time t.
u

u= u(x, t)

Subject to various assumptions

1. neglecting damping forces such as air resistance
2. neglecting the weight of the string
3. that the tension in the string is tangential to the curve of the string at any point
4. that the string performs small transverse oscillations i.e. every particle of the string moves
strictly vertically and such that its deection and slope as every point on the string is
small.
it can be shown, by applying Newtons Law of motion to a small segment of the string, that
u satises the PDE
2
2u
2 u
=
c
(1)
t2
x2
T
where c2 = , being the mass per unit length of the string and T being the (constant)

horizontal component of the tension in the string. To determine u(x, t) uniquely, we must also
know
1. the initial denition of the string at the time t = 0 at which it is released
2. the initial velocity of the string.
Thus we must be given initial conditions
u (x, 0) = f (x)

0xL

(Initial position)

u
(x, 0) = g(x)
t

0xL

(Initial velocity)

HELM (VERSION 1: March 18, 2004): Workbook Level 2

25.2: Applications of PDEs

where f (x) and g(x) are known. These two initial conditions are in addition to the two boundary conditions
u(0, t) = u(L, t) = 0

for t 0

which indicate that the string is xed at each end. In the specic example discussed in the
previous Section we had

x 
f (x) = u0 sin

g(x) = 0

(string initially at rest).

The PDE (1) is the (undamped) wave equation. We will discuss solutions of it for various
initial conditions later. More complicated forms of the wave equation would arise if some of the
assumptions were modied. For example:

(a)
(b)

2
2u
2 u
=
c
g if the weight of the string was allowed for,
t2
x2
2
2u
u
2 u
=
c

if a damping force proportional to the velocity of the string

t2
x2
t
(with damping constant ) was included.

(1) is referred to as the one-dimensional wave equation because only one space variable, x, is
present. The two-dimensional (undamped) wave equation is, in Cartesian coordinates,
 2

2

u
u

2u
(2)
= c2
+
t2
x2 y 2
This arises for example when we model the transverse vibrations of a membrane. See Figures
below. Here u(x, y, t) is the denition of a point (x, y) on the membrane at time t. Again, a
boundary condition must be specied: commonly
u=0

t0

on the boundary of the membrane, if this is xed (clamped). Also initial conditions must be
given
u (x, y, 0) = f (x, y)

(initial position)

u
(x, y, 0) = g(x, y)
t

(initial velocity)

For a circular membrane, such as a drumhead, polar coordinates dened by x = r cos , y =

r sin would be more convenient than Cartesian. In this case (2) becomes
 2

u 1 u
1 2u
2u
2
=c
+
0 r R,
0 2
+
t2
r2
r r r2 2
3

HELM (VERSION 1: March 18, 2004): Workbook Level 2

25.2: Applications of PDEs

for a circular membrane of radius R.

y

y
a

2. Heat conduction Equations

Consider a long thin bar, or wire, of constant cross-section and of homogeneous material oriented
along the xaxis.
x
x= 0

x= L

Imagine that the bar is thermally insulated laterally and is suciently thin that heat ows (by
conduction) only in the xdirection. Then the temperature u at any point in the bar depends
only on the xcoordinate of the point and the time t. By applying the principle of conservation
of energy it can be shown that u(x, t) satises the PDE
u
2u
=k 2
t
x

0xL
t>0

(3)

where k is a positive constant. In fact k, sometimes called the thermal diusivity of the bar,
is given by

k=
s
where = thermal conductivity of the material of the bar
s = specic heat capacity of the material of the bar
= density of the material of the bar.
The PDE (3) is called the one-dimensional heat conduction equation (or, in other contexts where
it arises, the diusion equation).

What is the obvious dierence between the wave equation (1) and the heat
conduction equation (3)?
Your solution

HELM (VERSION 1: March 18, 2004): Workbook Level 2

25.2: Applications of PDEs

Both equations involve second derivatives in the space variable x but whereas the wave equation
has a second derivative in the time variable t the heat conduction equation has only a rst
derivative in t. This means that the solutions of (3) are quite dierent in form from those of
(1) and we shall study them separately later.
The fact that (3) is rst order in t means that only one initial condition at t = 0 is needed,
together with two boundary conditions, to obtain a unique solution. The usual initial condition
species the initial temperature distribution in the bar
u(x, 0) = f (x)
where f (x) is known. Various types of boundary conditions at x = 0 and x = L are possible.
For example:
(a) u(0, t) = T1 and u(L, T ) = T2 (ends of the bar are at constant temperatures T1 and
T2 ).
u
u
(0, t) =
(L, t) = 0 which are insulation conditions since they tell us that there
(b)
x
x
is no heat ow through the ends of the bar.
As you would expect, there are two-dimensional and three-dimensional forms of the heat conduction equation. The two dimensional version of (3) is

 2
u
u 2u
(4)
+
=k
t
x2 y 2
where u(x, y, t) is the temperature in a at plate. The plate is assumed to be thin and insulated
on its top and bottom surface so that no heat ow occurs other than in the Oxy plane. Boundary
conditions and an initial condition are needed to give unique solutions of (4). For example if
the plate is rectangular:
y
y=b

x= a

typical boundary conditions might be

u(x, 0) = T1

0 x a (bottom side at xed temperature)

u
(a, y) = 0
x
u(x, b) = T2

0 y b (right hand side insulated)

u(0, y) = 0

0 x a (top side at xed temperature)

0 y b (left hand side at zero xed temperature).

An initial condition would have the form u(x, y, 0) = f (x, y), where f is a given function.

HELM (VERSION 1: March 18, 2004): Workbook Level 2

25.2: Applications of PDEs

3. Transmission Line Equations

In a long electrical cable or a telephone wire both the current and voltage depend upon position
along the wire as well as the time.
i(x, t)
v(x, t)
i(x, t)

It is possible to show, using basic laws of electrical circuit theory, that the electrical current
i(x, t) satises the PDE
2i
i
2i
=
LC
+ (RC + GL) + RGi
(5)
2
2
x
t
t
where the constants R, L, C and G are, for unit length of cable, respectively the resistance,
inductance, capacitance and leakage conductance. The voltage v(x, t) also satises (5). Special
cases of (5) arise in particular situations. For a submarine cable G is negligible and frequencies
are low so inductive eects can also be neglected. In this case (5) becomes
2i
i
= RC
(6)
2
x
t
which is called the submarine equation or telegraph equation. For high frequency alternating currents, again with negligible leakage, (5) can be approximated by
2i
2i
=
LC
x2
t2
which is called the high frequency line equation.

(7)

What PDEs, already discussed, have the same form as equation (6) or equation
(7)?

Your solution

(6) has the same form as the one-dimensional heat conduction equation. (7) has the same form
as the one-dimensional wave equation.
HELM (VERSION 1: March 18, 2004): Workbook Level 2
25.2: Applications of PDEs

4. Laplaces Equation
If you look back at the two-dimensional heat conduction equation (4)

 2
u 2u
u
+
=k
t
x2 y 2
it is clear that if the heat ow is steady, i.e. time independent, then
u(x, y) is a solution of

u
= 0 so the temperature
t

2u 2u
+
=0
x2 y 2

(8)

(8) is the two-dimensional Laplace equation. Both this, and its three-dimensional counterpart
2u 2u 2u
+
+
=0
x2 y 2 z 2

(9)

arise in a wide variety of applications, quite apart from steady state heat conduction theory.
Since time does not arise in (8) or (9) it is evident that Laplaces equation is always a model
for equilibrium situations. In any problem involving Laplaces equation we are interested in
solving it in a specic region R for given boundary conditions. Since conditions may involve
(a) u specied on the boundary curve C (two dimensions) or boundary surface S (three
dimensions) of the region R. Such boundary conditions are called Dirichlet conditions.
u
(b) The derivative of u normal to the boundary, written
, specied on C or S. These
n
are referred to as Neumann boundary conditions.
(c) A mixture of (a) and (b).
Some areas in which Laplaces equation arises are
(a) electrostatics (u being the electrostatic potential in a charge free region)
(b) gravitation (u being the gravitational potential in free space)
(c) steady state ow of inviscid uids
(d) steady state heat conduction (as already discussed.)

Other important PDEs in science and engineering

1. Poissons equation
2u 2u
+
= f (x, y)
x2 y 2

(two-dimensional form)

where f (x, y) is a given function. This equation arises in electrostatics, elasticity theory
and elsewhere.
7

HELM (VERSION 1: March 18, 2004): Workbook Level 2

25.2: Applications of PDEs

2. Helmholtzs equation
2u 2u
+
+ k2u = 0
x2 y 2

(two dimensional form)

which arises in wave theory.

3. Schrodingers equation

 2
2 2
h2
+ 2 + 2 = E
2
8 m x2
y
z
which arises in quantum mechanics. (h is Plancks constant.)
4. Transverse vibrations in a homogeneous rod
a2

4u 2u
+ 2 =0
x4
t

where u(x, t) is the displacement at time t of the cross-section through x.

All the PDEs we have discussed are second order (because the highest order derivatives that
arise are second order) apart from the last example which is fourth order.

HELM (VERSION 1: March 18, 2004): Workbook Level 2

25.2: Applications of PDEs