Pde Book

Lecture Notes in Mathematics
Arkansas Tech University

Department of Mathematics
A First Course in Quasi-Linear Partial

Differential Equations
for Physical Sciences and Engineering
Marcel B. Finan
Arkansas Tech University
c
All
Rights Reserved
May 1, 2016
Preface
Partial differential equations are often used to construct models of the most
basic theories underlying physics and engineering. The goal of this book is to
develop the most basic ideas from the theory of partial differential equations,
and apply them to the simplest models arising from the above mentioned
fields.
It is not easy to master the theory of partial differential equations. Unlike
the theory of ordinary differential equations, which relies on the fundamental
existence and uniqueness theorem, there is no single theorem which is central
to the subject. Instead, there are separate theories used for each of the major
types of partial differential equations that commonly arise.
It is worth pointing out that the preponderance of differential equations arising in applications, in science, in engineering, and within mathematics itself,
are of either first or second order, with the latter being by far the most prevalent. We will mainly cover these two classes of PDEs.
This book is intended for a first course in partial differential equations at
the advanced undergraduate level for students in engineering and physical
sciences. It is assumed that the student has had the standard three semester
calculus sequence, and a course in ordinary differential equations.
Marcel B Finan
August 2009
ii
PREFACE
Contents
Preface
The Basics of the Theory of Partial Differential Equation

3
1 Basic Concepts . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
2 Solutions to PDEs/PDEs with constraints . . . . . . . . . . . . . 13
Review of Some ODE Results
25
3 The Method of Integrating Factor . . . . . . . . . . . . . . . . . . 26
4 The Method of Separation of Variables for ODEs . . . . . . . . . 31
First Order Partial Differential Equations
5 Classification of First Order PDEs . . . . . . . . . . . . . . . .
6 A Review of Multivariable Calculus . . . . . . . . . . . . . . . .
6.1 Multiplication of Vectors: The Scalar or Dot Product . .
6.2 Directional Derivatives and the Gradient Vector . . . . .
7 Solvability of Semi-linear First Order PDEs . . . . . . . . . . .
8 Linear First Order PDE: The One Dimensional Spatial Transport
Equations . . . . . . . . . . . . . . . . . . . . . . . . . . . .
9 Solving Quasi-Linear First Order PDE via the Method of Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . .
10 The Cauchy Problem for First Order Quasilinear Equations . .
. 71
. 75
Second Order Linear Partial Differential Equations

11 Second Order PDEs in Two Variables . . . . . . . . . . . . . .
12 Hyperbolic Type: The Wave equation . . . . . . . . . . . . . .
13 Parabolic Type: The Heat Equation in One-Dimensional Space
14 Sequences of Functions: Pointwise and Uniform Convergence .
15 An Introduction to Fourier Series . . . . . . . . . . . . . . . .
.
.
.
.
.
.
.
.
.
.
35
36
41
41
50
59
. 65
87
88
93
100
109
120
CONTENTS
16 Fourier Sines Series and Fourier Cosines Series . . . . . . . . .
17 Separation of Variables for PDEs . . . . . . . . . . . . . . . .
17.1 Second Order Linear Homogenous ODE with Constant
Coefficients . . . . . . . . . . . . . . . . . . . . . . .
17.2 The Method of Separation of Variables for PDEs . . . .
18 Solutions of the Heat Equation by the Separation of Variables
Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
19 Elliptic Type: Laplaces Equations in Rectangular Domains . .
20 Laplaces Equations in Circular Regions . . . . . . . . . . . . .
. 133
. 140
. 140
. 141
. 147
. 154
. 165
The Laplace Transform Solutions for PDEs

177
21 Essentials of the Laplace Transform . . . . . . . . . . . . . . . . 178
22 Solving PDEs Using Laplace Transform . . . . . . . . . . . . . . 191
The Fourier Transform Solutions for PDEs
199
23 Complex Version of Fourier Series . . . . . . . . . . . . . . . . . 200
24 An introduction to Fourier Transforms . . . . . . . . . . . . . . 206
25 Applications of Fourier Transforms to PDEs . . . . . . . . . . . 214
Appendix
221
Appendix A: The Method of Undetermined Coefficients . . . . . . . 222
Appendix B: The Method of Variation of Parameters . . . . . . . . 229
Answers and Solutions
233
Index
289
The Basics of the Theory of

Partial Differential Equation
Many fields in engineering and the physical sciences require the study of
ODEs and PDEs. Some of these fields include acoustics, aerodynamics, elasticity, electrodynamics, fluid dynamics, geophysics (seismic wave propagation), heat transfer, meteorology, oceanography, optics, petroleum engineering, plasma physics (ionized liquids and gases), quantum mechanics.
So the study of partial differential equations is of great importance to the
above mentioned fields. The purpose of this chapter is to introduce the reader
to the basic terms of partial differential equations.
4THE BASICS OF THE THEORY OF PARTIAL DIFFERENTIAL EQUATION
1 Basic Concepts
The goal of this section is to introduce the reader to the basic concepts and
notations that will be used in the remainder of this book.
We start this section by reviewing the concept of partial derivatives and the
chain rule of functions in two variables.
Let u(x, y) be a function of the independent variables x and y. The first
derivative of u with respect to x is defined by
u(x + h, y) u(x, y)
u
(x, y) = lim
h0
x
h
provided that the limit exists.
Likewise, the first derivative of u with respect to y is defined by
ux (x, y) =
uy (x, y) =
u
u(x, y + h) u(x, y)
(x, y) = lim
h0
y
h
provided that the limit exists.

We can define higher order derivatives such as
2u
ux (x + h, y) ux (x, y)
uxx (x, y) =
(x, y) = lim
2
h0
x
h
2u
uy (x, y + h) uy (x, y)
(x, y) = lim
2
h0
y
h

2u
u
uy (x + h, y) uy (x, y)
uxy (x, y) =
(x, y) =
= lim
h0
xy
x y
h
uyy (x, y) =
and
2u
ux (x, y + h) ux (x, y)
uyx (x, y) =
(x, y) = lim
h0
yx
h
1
provided that the limits exist.
An important formula of differentiation is the so-called chain rule. If u =
u(x, y), where x = x(s, t) and y = y(s, t), then
u
u x u y
=
+
.
s
x s y s
Likewise,
u
u x u y
=
+
.
t
x t
y t
1
If uxy and uux are continuous then uxy (x, y) = uyx (x, y).
1 BASIC CONCEPTS
Example 1.1
Compute the partial derivatives indicated:
(a) y
(y 2 sin xy)
(b)
2
[ex+y ]2
x2
Solution.
(a) We have
(b) We have
4e2(x+y)
(y 2 sin xy) = sin xy y

(y 2 )+y 2 y
(sin xy) = 2y sin xy+xy 2 cos xy.
y
2(x+y)
2
x+y 2
[ex+y ]2 = x
e
= 2e2(x+y) . Thus, x
] = x
2e2(x+y) =
2 [e
x
Example 1.2
Suppose u(x, y) = sin (x2 + y 2 ), where x = tes and y = s + t. Find us (s, t)
and ut (s, t).
Solution.
We have
us (s, t) =ux xs + uy ys = 2x cos (x2 + y 2 )tes + 2y cos (x2 + y 2 )
=[2t2 e2s + 2(s + t)] cos [t2 e2s + (s + t)2 ].
Likewise,
ut (s, t) =ux xt + uy yt = 2x cos (x2 + y 2 )es + 2y cos (x2 + y 2 )
=[2te2s + 2(s + t)] cos [t2 e2s + (s + t)2 ]
A differential equation is an equation that involves an unknown scalar
function (the dependent variable) and one or more of its derivatives. For
example,
d2 y
dy
5 + 3y = 3
(1.1)
2
dx
dx
or
u 2 u 2 u
2 2 + u = 0.
(1.2)
t
x
y
If the unknown function is a function in one single variable then the differential equation is called an ordinary differential equation, abbreviated by
ODE. An example of an ordinary differential equation is Equation (1.1). In
contrast, when the unknown function is a function of two or more independent variables then the differential equation is called a partial differential
equation, in short PDE. Equation (1.2) is an example of a partial differential
equation. In this book we will be focusing on partial differential equations.

Example 1.3
Identify which variables are dependent and which are independent for the
following differential equations.
d4 y
2
(a) dx
4 x + y = 0.
(b) utt + xutx = 0.
= 4.
(c) x dx
dt
y
(d) u 4 y
= u + 3y.
v
Solution.
(a) Independent variable is x and the dependent variable is y.
(b) Independent variables are x and t and the dependent variable is u.
(c) Independent variable is t and the dependent variable is x.
(d) Independent variables are u and v and the dependent variable is y
Example 1.4
Classify the following as either ODE or PDE.
(a) ut = c2 uxx .
(b) y 00 4y 0 + 5y = 0.
(c) zt + czx = 5.
Solution.
(a) A PDE with dependent variable u and independent variables t and x.
(b) An ODE with dependent variable y and independent variable x.
(c) A PDE with dependent variable z and independent variables t and x
The order of a differential equation is the highest order derivative occurring
in the equation. Thus, (1.1) and (1.2) are second order differential equations.
Example 1.5
Find the order of each of the following partial differential equations:
(a) xux + yuy = x2 + y 2
(b) uux + uy = 2
(c) utt c2 uxx = f (x, t)
(d) ut + uux + uxxx = 0
(e) utt + uxxxx = 0.
Solution.
(a) First order (b) First order (c) Second order (d) Third order (e) Fourth
1 BASIC CONCEPTS
order
A first order partial differential equation is called quasi-linear if it can
be written in the form
a(x, y, u)ux + b(x, y, u)uy = c(x, y, u).
(1.3)
If a(x, y, u) = (x, y) and b(x, y, u) = (x, y) then (1.3) is called semi-linear.

If futhermore, c(x, y, u) = (x, y)u + (x, y) then (1.3) is called linear.
In a similar way, a second order quasi-linear pde has the form
a(x, y, u, ux , uy )uxx +b(x, y, u, ux , uy )uxy +c(x, y, u, ux , uy )uyy = d(x, y, u, ux , uy ).
The semi-linear case has the form

a(x, y)uxx + b(x, y)uxy + c(x, y)uyy = d(x, y, u, ux , uy ).
and the linear case is
a(x, y)uxx + b(x, y)uxy + c(x, y)uyy + d(x, y)ux + e(x, y)uy + f (x, y)u = g(x, y).
Note that linear and semi-linear partial differential equations are special cases
of quasi-linear equations. However, a quasi-linear PDE needs not be linear:
A partial differential equation that is not linear is called non-linear. For
example, u2x + 2uxy = 0 is non-linear. Note that this equation is quasi-linear
and semi-linear.
As for ODEs, linear PDEs are usually simpler to analyze/solve than nonlinear PDEs.
Example 1.6
Determine whether the given PDE is linear, quasi-linear, semi-linear, or nonlinear:
(a) xux + yuy = x2 + y 2 .
(b) uux + uy = 2.
(c) utt c2 uxx = f (x, t).
(d) ut + uux + uxxx = 0.
(e) u2tt + uxx = 0.
Solution.
(a) Linear, quasi-linear, semi-linear.
(b) Quasi-linear, non-linear.
(c) Linear, quasi-linear, semi-linear.

(d) Quasi-linear, semi-linear, non-linear.
(e) Non-linear
A more precise definition of a linear differential equation begins with the concept of a linear differential operator L. The operator L is assembled by
summing the basic partial derivative operators, with coefficients depending
on the independent variables only. The operator acts on sufficiently smooth
functions2 are depending on the relevant independent variables. Linearity
imposes two key requirements:
L[u + v] = L[u] + L[v] and L[u] = L[u],
for any two (sufficiently smooth) functions u, v and any constant (a scalar)
.
Example 1.7
Define a linear differential operator for the PDE
ut = c2 uxx .
Solution.
Let L[u] = ut c2 uxx . Then one can easily check that L[u + v] = L[u] + L[v]
and L[u] = L[u]
A linear partial differential equation is called homogeneous if the dependent variable and/or its derivatives appear in terms with degree exactly one.
A linear partial differential equation that is not homogeneous is called nonhomogeneous. In this case, there is a term in the equation that involves
only the independent variables.
A homogeneous linear partial differential equation has the form
L[u] = 0
where L is a linear differential operator and the non-homogeneous case has
the form
L[u] = f (x, y, ).
2
Smooth functions are functions that are continuously differentiable up to a certain

order.
1 BASIC CONCEPTS
Example 1.8
Determine whether the equation is homogeneous or non-homogeneous:
(a) xux + yuy = x2 + y 2 .
(b) utt = c2 uxx .
(c) y 2 uxx + xuyy = 0.
Solution.
(a) Non-homogeneous because of x2 + y 2 .
(b) Homogeneous.
(c) Homogeneous
Practice Problems
Problem 1.1
Classify the following equations as either ODE or PDE.
(a) (y 000 )4 +
(b)
u
x
t2
(y 0 )2 +4
+ y u
=
y
= 0.
yx
.
y+x
(c) y 00 4y = 0.
Problem 1.2
Write the equation
uxx + 2uxy + uyy = 0
in the coordinates s = x, t = x y.
Problem 1.3
Write the equation
uxx 2uxy + 5uyy = 0
in the coordinates s = x + y, t = 2x.
Problem 1.4
For each of the following PDEs, state its order and whether it is linear
or non-linear. If it is linear, also state whether it is homogeneous or nonhomogeneous:
(a) uux + x2 uyyy + sin x = 0.
2
(b) ux + ex uy = 0.
(c) utt + (sin y)uyy et cos y = 0.
Problem 1.5
For each of the following PDEs, determine its order and whether it is linear
or not. For linear PDEs, state also whether the equation is homogeneous or
not. For non-linear PDEs, circle all term(s) that are not linear.
(a) x2 uxx + ex u = xuxyy .
(b) ey uxxx + ex u = sin y + 10xuy .
(c) y 2 uxx + ex uux = 2xuy + u.
(d) ux uxxy + ex uuy = 5x2 ux .
(e) ut = k 2 (uxx + uyy ) + f (x, y, t).
1 BASIC CONCEPTS
11
Problem 1.6
Which of the following PDEs are linear?
(a) Laplaces equation: uxx + uyy = 0.
(b) Convection (transport) equation: ut + cux = 0.
(c) Minimal surface equation: (1 + Zy2 )Zxx 2Zx Zy Zxy + (1 + Zx2 )Zyy = 0.
(d) Korteweg-Vries equation: ut + 6uux = uxxx .
Problem 1.7
Classify the following differential equations as ODEs or PDEs, linear or
non-linear, and determine their order. For the linear equations, determine
whether or not they are homogeneous.
(a) The diffusion equation for u(x, t) :
ut = kuxx .
(b) The wave equation for w(x, t) :
wtt = c2 wxx .
(c) The thin film equation for h(x, t) :
ht = (hhxxx )x .
(d) The forced harmonic oscillator for y(t) :
ytt + 2 y = F cos (t).
(e) The Poisson Equation for the electric potential (x, y, z) :
xx + yy + zz = 4(x, y, z)
where (x, y, z) is a known charge density.
(f) Burgers equation for h(x, t) :
ht + hhx = hxx .
Problem 1.8
Write down the general form of a linear second order differential equation of
a function in three variables.

Problem 1.9
Give the orders of the following PDEs, and classify them as linear or nonlinear. If the PDE is linear, specify whether it is homogeneous or nonhomogeneous.
(a) x2 uxxy + y 2 uyy log (1 + y 2 )u = 0.
(b) ux + u3 = 1.
(c) uxxyy + ex ux = y.
(d) uuxx + uyy u = 0.
(e) uxx + ut = 3u.
Problem 1.10
Consider the second-order PDE
uxx + 4uxy + 4uyy = 0.
Use the change of variables v(x, y) = y 2x and w(x, y) = x to show that
uww = 0.
Problem 1.11
Write the one dimensional wave equation utt = c2 uxx in the coordinates
v = x + ct and w = x ct.
Problem 1.12
Write the PDE
uxx + 2uxy 3uyy = 0
in the coordinates v(x, y) = y 3x and w(x, y) = x + y.
Problem 1.13
Write the PDE
aux + buy = 0, a 6= 0
in the coordinates s(x, y) = bx ay and t(x, y) = x.
Problem 1.14
Write the PDE
ux + uy = 1
in the coordinates s = x y and t = x.
Problem 1.15
Write the PDE
aut + bux = u, b 6= 0
in the coordinates v = ax bt and w = x.
2 SOLUTIONS TO PDES/PDES WITH CONSTRAINTS
13
2 Solutions to PDEs/PDEs with constraints

By a classical solution or strong solution to a partial differential equation
we mean a function that satisfies the equation. A PDE might have many
classical solutions. To solve a PDE is to find all its classical solutions. In the
case of only two independent variables x and y, a classical solution u(x, y)
is visualized geometrically as a surface, called a solution surface or an
integral surface3 of the PDE in the (x, y, u) space.
A formula that expresses all the solutions of a PDE is called the general
solution of the equation.
Example 2.1
2 2
Show that u(x, t) = e t (cos x sin x) is a solution to the equation
ut 2 uxx = 0.
Solution.
Since
ut 2 uxx = 2 2 e
2 2 t
2 e
2 2 t
(cos x sin x)
(2 cos x + 2 sin x) = 0,
the given function is a classical solution to the given equation

Example 2.2
The function u(x, y) = x2 y 2 is a solution to Laplaces equation
uxx + uyy = 0.
Represent this solution graphically.
Solution.
The given integral surface is the hyperbolic paraboloid shown in Figure 2.1.
3
The idea behind the name is due to the fact that integration is being used to finding
the solution.
Figure 2.1
Example 2.3
Find the general solution of uxy = 0.
Solution.
Integrating first we respect to y we find ux (x, y) = f (x), where f is an
arbitrary Rdifferentiable function. Integrating ux with respect to x we find
u(x, y) = f (x)dx + g(y), where g is an arbitrary differentiable function
Note that the general solution in the previous example involves two arbitrary
functions. In general, the general solution of a partial differential equation
is an expression that involves arbitrary functions. This is in contrast to the
general solution of an ordinary differential equation which involves arbitrary
constants.
Usually, a classical solution enjoys properties such as smootheness (i.e. differentiability) and continuity. However, in the theory of non-linear pdes,
there are solutions that do not require the smoothness property. Such solutions are called weak solutions or generalized solutions. For example,
u(x) = x is a classical solution to the differential equation uu0 = x. In contrast, u(x) = |x| is a generalized solution since it is not differentiable at 0.
In this book, the word solution will refer to a classical solution.
Example 2.4
Show that u(x, t) = t + 12 x2 is a classical solution to the PDE
ut = uxx .
(2.1)
15
Solution.
Assume that the domain of definition of u is D R2 . Since u, ut , ux , utx , uxx
exist and are continuous in D(i.e., u is smooth in D) and u satisfies equation
(2.1), we conclude that u is a classical solution to the given PDE
We next consider the structure of solutions to linear partial differential equations. To this end, consider the linear differential operator L as defined in
the previous section. The defining properties of linearity immediately imply
the key facts concerning homogeneous linear differential equations.
Theorem 2.1
The sum of two solutions to a homogeneous linear differential equation is
again a solution, as is the product of a solution by any constant.
Proof.
Let u1 , u2 be solutions, meaning that L[u1 ] = 0 and L[u2 ] = 0. Then, thanks
to linearity,
L[u1 + u2 ] = L[u1 ] + L[u2 ] = 0,
and hence their sum u1 + u2 is a solution. Similarly, if is any constant, and
u any solution, then
L[u] = L[u] = 0 = 0,
and so the scalar multiple u is also a solution
The following result is known as the superposition principle for homogeneous linear equations. It states that from given solutions to the equation
one can create many more solutions.
Theorem 2.2
If u1 , , un are solutions to a common homogeneous linear partial differential equation L[u] = 0, then the linear combination u = c1 u1 + + cn un is
a solution for any choice of constants c1 , , cn .
Proof.
The key fact is that, thanks to the linearity of L, for any sufficiently smooth
functions u1 , , un and any constants c1 , , cn ,
L[u] =L[c1 u1 + + cn un ] = L[c1 u1 + + cn1 un1 ] + L[cn un ]
= = L[c1 u1 ] + + L[cn un ] = c1 L[u1 ] + + cn L[un ].

In particular, if the functions are solutions, so L[u1 ] = 0, , L[un ] = 0, then
the right hand side of the above equation vanishes, proving that u is also a
solution to the homogeneous equation L[u] = 0
In physical applications, homogeneous linear equations model unforced systems that are subject to their own internal constraints. External forcing
is represented by an additional term that does not involve the dependent
variable. This results in the non-homogeneous equation
L[u] = f
where L is a linear partial differential operator, u is the dependent variable,
and f is a given non-zero function of the independent variables alone.
You already learned the basic philosophy for solving of nonhomogeneous
linear equations in your study of elementary ordinary differential equations.
Step one is to determine the general solution to the homogeneous equation.
Step two is to find a particular solution to the non-homogeneous version.
The general solution to the non-homogeneous equation is then obtained by
adding the two together. Here is the general version of this procedure:
Theorem 2.3
Let up be a particular solution to the non-homogeneous linear equation
L[u] = f. Then the general solution to L[u] = f is given by u = up + uh ,
where uh is the general solution to the corresponding homogeneous equation
L[u] = 0.
Proof.
Let us first show that u = up + uh is also a solution to L[u] = f. By linearity,
L[u] = L[up + uh ] = L[up ] + L[uh ] = f + 0 = f.
To show that every solution to the non-homogeneous equation can be expressed in this manner, suppose u satisfies L[u] = f. Set uh = u up . Then,
by linearity,
L[uh ] = L[u up ] = L[u] L[up ] = 0,
and hence uh is a solution to the homogeneous differential equation. Thus,
u = up + uh has the required form
As you have noticed by the above discussion, one solution of a linear PDE
17
leads to the creation of lots of solutions. In contrast, non-linear equations

are much tougher to deal with, for example, knowledge of several solutions
does not necessarily help in constructing others. Indeed, even finding one
solution to a non-linear partial differential equation can be quite a challenge.
PDEs with Constraints
Also, as observed above, a linear partial differential equation has infinitely
many solutions described by the general solution. In most applications, this
general solution is of little use since it has to satisfy other supplementary
conditions, usually called initial or boundary conditions. These conditions
determine the unique solution of interest.
A boundary value problem is a partial differential equation where either
the unknown function or its derivatives have values assigned on the physical
boundary of the domain in which the problem is specified. These conditions
are called boundary conditions. For example, the domain of the following
problem is the square [0, 1] [0, 1] with boundaries defined by x = 0, x = 1
for all 0 y 1 and y = 0, y = 1 for all 0 x 1.
uxx + uyy =0
u(x, 0) = u(x, 1) =0
ux (0, y) = ux (1, y) =0
if 0 < x, y < 1
if 0 < x < 1
if 0 < y < 1.
There are three types of boundary conditions which arise frequently in formulating physical problems:
1. Dirichlet Boundary Conditions: In this case, the dependent function u is prescribed on the boundary of the bounded domain. For example, if
the bounded domain is the rectangular plate 0 < x < L1 and 0 < y < L2 , the
boundary conditions u(0, y), u(L1 , y), u(x, 0), and u(x, L2 ) are prescribed.
The boundary conditions are called homogeneous if the dependent variable
is zero at any point on the boundary, otherwise the boundary conditions are
called nonhomogeneous.
2. Neumann Boundary Conditions: In this case, first partial derivatives are prescribed on the boundary of the bounded domain. For example,
the Neuman boundary conditions for a rod of length L, where 0 < x < L,
are of the form ux (0, t) = and ux (L, t) = , where and are constants.

3. Robin or mixed Boundary Conditions: This occurs when the dependent variable and its first partial derivatives are prescribed on the boundary
of the bounded domain.
An initial value problem (or Cauchy problem) is a partial differential
equation together with a set of additional conditions on the unknwon function or its derivatives at a point in the given domain of the solution. These
conditions are called initial value conditions. For example, the transport
equation
ut (x, t) + cux (x, t) =0
u(x, 0) =f (x).
It can be shown that initial conditions for a linear PDE are necessary and
sufficient for the existence of a unique solution.
We say that an initial and/or boundary value problem associated with a
PDE is well-posed if it has a solution which is unique and depends continuously on the data given in the problem. The last condition, namely the
continuous dependence is important in physical problems. This condition
means that the solution changes by a small amount when the conditions
change a little. Such solutions are said to be stable.
Example 2.5
For x R and t > 0 we consider the initial value problem
utt uxx =0
u(x, 0) = ut (x, 0) =0.
Clearly, u(x, t) = 0 is a solution to this problem.
(a) Let 0 < <<
1 be a very small number. Show that the function u (x, t) =
t
x
2
sin sin is a solution to the initial value problem
utt uxx =0
u(x, 0) =0
ut (x, 0) = sin
x
.

(b) Show that sup{|u (x, t) u(x, t)| : x R, t > 0} = 2 . Thus, a small
change in the initial data leads to a small change in the solution. Hence, the
initial value problem is well-posed.
19
Solution.
(a) We have
u
t
2
u
t2
u
x
2 u
x2
2
Thus, tu2
(b) We have
2 u
x2

t
= sin
cos

x
t
sin
= sin

x
t
= cos
sin

x
t
= sin
sin
.

x
= 0. Moreover, u (x, 0) = 0 and
u (x, 0)
t
= sin
x

.

t
x
2

sin
: x R, t > 0}
sup{|u (x, t) u(x, t)| : x R, t > 0} = sup{sin

=2
A problem that is not well-posed is referred to as an ill-posed problem. We
illustrate this concept in the next example.
Example 2.6
For x R and t > 0 we consider the initial value problem
utt + uxx =0
u(x, 0) = ut (x, 0) =0.
Clearly, u(x, t) = 0 is a solution to this problem.
(a) Let 0 < << 1 be a very small number. Show that the function u (x, t) =
2 sin x sinh t , where
ex ex
sinh x =
2
is a solution to the problem
utt + uxx =0
u(x, 0) =0
ut (x, 0) = sin
x

(b) Show that sup{| t

u (x, t0)
ut (x, 0)| : x R} = and sup{|u (x, t)
2
u(x, t)| : x R} = sinh .
(c) Find limt sup{|u (x, t) u(x, t)| : x R}.
Solution.
(a) We have
u
t
2 u
t2
u
x
2 u
x2
2
Thus, tu2 +
(b) We have
2 u
x2

t
= sin
cosh

x
t
sinh
= sin

x
t
= cos
sinh

x
t
= sin
sinh
.

x
= 0. Moreover, u (x, 0) = 0 and
u (x, 0)
t
= sin
x

.

x

sup{| u (x, 0) ut (x, 0)| : x R} = sup{ sin
: x R}
t
x

= sup{sin
: x R} =

and

x

t
: x R}

sup{|u (x, t) u(x, t)| : x R} = sup{sinh
sin

t
2
= sinh
.

2
(c) We have

t

lim sup{|u (x, t) u(x, t)| : x R} = lim sinh
= .
t
t

2
Thus, a small change in the initial data leads to a catastrophically change in

the solution. Hence, the given problem is ill-posed
21
Practice Problems
Problem 2.1
Determine a and b so that u(x, y) = eax+by is a solution to the equation
uxxxx + uyyyy + 2uxxyy = 0.
Problem 2.2
Consider the following differential equation
tuxx ut = 0.
Suppose u(t, x) = X(x)T (t). Show that there is a constant such that
X 00 = X and T 0 = tT.
Problem 2.3
Consider the initial value problem
xux + (x + 1)yuy = 0, x, y > 1
u(1, 1) = e.
Show that u(x, y) =
xex
y
is the solution to this problem.
Problem 2.4
Show that u(x, y) = e2y sin (x y) is the solution to the initial value problem

ux + uy + 2u = 0 for x, y > 1
u(x, 0) = sin x.
Problem 2.5
Solve each of the following differential equations:
= 0 where u = u(x).
(a) du
dx
(b) u
= 0 where u = u(x, y).
x
Problem 2.6
Solve each of the following differential equations:
2
(a) ddxu2 = 0 where u = u(x).
2u
(b) xy
= 0 where u = u(x, y).

Problem 2.7
Show that u(x, y) = f (y + 2x) + xg(y + 2x), where f and g are two arbitrary
twice differentiable functions, satisfy the equation
uxx 4uxy + 4uyy = 0.
Problem 2.8
Find the differential equation whose general solution is given by u(x, t) =
f (xct)+g(x+ct), where f and g are arbitrary twice differentiable functions
in one variable.
Problem 2.9
Let p : R R be a differentiable function in one variable. Prove that
ut = p(u)ux
has a solution satisfying u(x, t) = f (x + p(u)t), where f is an arbitrary
differentiable function. Then find the general solution to ut = (sin u)ux .
Problem 2.10
Find the general solution to the pde
uxx + 2uxy + uyy = 0.
Hint: See Problem 1.2.
Problem 2.11
Let u(x, t) be a function such that uxx exists and u(0, t) = u(L, t) = 0 for all
t R. Prove that
Z L
uxx (x, t)u(x, t)dx 0.
0
Problem 2.12
ut + uxx = 0, x R, t > 0
u(x, 0) = 1.
(a) Show that u(x, t) 1 is a solution to this problem.
n2 t
(b) Show that un (x, t) = 1 + e n sin nx is a solution to the initial value
problem
ut + uxx = 0, x R, t > 0

u(x, 0) = 1 +
23
sin nx
.
n
(c) Find sup{|un (x, 0) 1| : x R}.

(d) Find sup{|un (x, t) 1| : x R}.
(e) Show that the problem is ill-posed.
Problem 2.13
Find the general solution of each of the following PDEs by means of direct
integration.
(a) ux = 3x2 + y 2 , u = u(x, y).
(b) uxy = x2 y, u = u(x, y).
(c) uxyz = 0, u = u(x, y, z).
(d) uxtt = e2x+3t , u = u(x, t).
Problem 2.14
uxx + 4uxy + 4uyy = 0.
(a) Use the change of variables v(x, y) = y 2x and w(x, y) = x to show
that uww = 0.
(b) Find the general solution to the given PDE.
Problem 2.15
Derive the general solution to the PDE
utt = c2 uxx
by using the change of variables v = x + ct and w = x ct.
Review of Some ODE Results

Later on in this book, we will encounter problems where a given partial
differential equation is reduced to an ordinary differential equation by means
of a given change of variables. Then techniques from the theory of ODE are
required in solving the transformed ODE. In this chapter, we include some
of the results from ODE theory that will be needed in our future discussions.
25
26
REVIEW OF SOME ODE RESULTS
3 The Method of Integrating Factor

In this section, we discuss a technique for solving the first order linear nonhomogeneous equation
y 0 + p(t)y = g(t)
(3.1)
where p(t) and g(t) are continuous on the open interval aR < t < b.
Since
p(t) is continuous, it has an antiderivative namely p(t)dt. Let (t) =
R
p(t)dt
e
. Multiply Equation (3.1) by (t) and notice that the left hand side of
the resulting equation is the derivative of a product. Indeed,
d
((t)y) = (t)g(t).
dt
Integrate both sides of the last equation with respect to t to obtain
Z
(t)y = (t)g(t)dt + C
Hence,
or
Z
C
1
(t)g(t)dt +
y(t) =
(t)
(t)
Z R
R
R
p(t)dt
y(t) = e
e p(t)dt g(t)dt + Ce p(t)dt
Notice that the second term of the previous expression is just the general
solution for the homogeneous equation
y 0 + p(t)y = 0
whereas the first term is a solution to the nonhomogeneous equation. That
is, the general solution to Equation (3.1) is the sum of a particular solution of
the nonhomogeneous equation and the general solution of the homogeneous
equation.
Example 3.1
Solve the initial value problem
y0
y
= 4t, y(1) = 5.
t
3 THE METHOD OF INTEGRATING FACTOR
27
Solution.
We have p(t) = 1t so that (t) = 1t . Multiplying the given equation by the
integrating factor and using the product rule we notice that
0
1
y = 4.
t
Integrating with respect to t and then solving for y we find that the general
solution is given by
Z
y(t) = t 4dt + Ct = 4t2 + Ct.
Since y(1) = 5, we find C = 1 and hence the unique solution to the IVP is
y(t) = 4t2 + t, 0 < t <
Example 3.2
Find the general solution to the equation
2
y 0 + y = ln t, t > 0.
t
Solution.
R 2
The integrating factor is (t) = e t dt = t2 . Multiplying the given equation
by t2 to obtain
(t2 y)0 = t2 ln t.
Integrating with respect to t we find
Z
2
t y = t2 ln tdt + C.
The integral on the right-hand side is evaluated using integration by parts
3
with u = ln t, dv = t2 dt, du = dtt , v = t3 obtaining
t2 y =
t3
t3
ln t + C
3
9
Thus,
y=
t
t C
ln t + 2
3
9 t
28
Example 3.3
Solve
aux + buy + cu = 0
by using the change of variables s = ax + by and t = bx ay.
Solution.
By the Chain rule for functions of two variables, we have
ux =us sx + ut tx = aus + but
uy =us sy + ut ty = bus aut .
Substituting into the given equation, we find
us +
a2
c
u = 0.
+ b2
Solving this equation using the integrating factor method we find
u(s, t) = f (t)e
cs
a2 +b2
where f is an arbitrary differentiable function of f. Switching back to x and

y we obtain
c (ax+by)
u(x, y) = f (bx ay)e a2 +b2
3 THE METHOD OF INTEGRATING FACTOR
29
Practice Problems
Problem 3.1
Solve the IVP: y 0 + 2ty = t, y(0) = 0.
Problem 3.2
Find the general solution: y 0 + 3y = t + e2t .
Problem 3.3
Find the general solution: y 0 + 1t y = 3 cos t, t > 0.
Problem 3.4
Find the general solution: y 0 + 2y = cos (3t).
Problem 3.5
Find the general solution: y 0 + (cos t)y = 3 cos t.
Problem 3.6
Given that the solution to the IVP ty 0 + 4y = t2 , y(1) = 31 exists on the
interval < t < . What is the value of the constant ?
Problem 3.7
Suppose that y(t) = Ce2t + t + 1 is the general solution to the equation
y 0 + p(t)y = g(t). Determine the functions p(t) and g(t).
Problem 3.8
Suppose that y(t) = 2et + et + sin t is the unique solution to the IVP
y 0 + y = g(t), y(0) = y0 . Determine the constant y0 and the function g(t).
Problem 3.9
Find the value (if any) of the unique solution to the IVP y 0 + (1 + cos t)y =
1 + cos t, y(0) = 3 in the long run?
Problem 3.10
Solve the initial value problem ty 0 = y + t, y(1) = 7.
Problem 3.11
Show that if a and are positive constants, and b is any real number, then
every solution of the equation
y 0 + ay = bet
has the property that y 0 as t . Hint: Consider the cases a = and
a 6= separately.
30
Problem 3.12
Solve the initial-value problem y 0 + y = et y 2 , y(0) = 1 using the substitution
1
u(t) = y(t)
Problem 3.13
Solve the initial-value problem ty 0 + 2y = t2 t + 1, y(1) =
1
2
Problem 3.14
Solve y 0 1t y = sin t, y(1) = 3. Express your answer in terms of the sine
Rt
integral, Si(t) = 0 sins s ds.
4 THE METHOD OF SEPARATION OF VARIABLES FOR ODES
31
4 The Method of Separation of Variables for

ODEs
The method of separation of variables that you have seen in the theory of
ordinary differential equations has an analogue in the theory of partial differential equations (Section 17). In this section, we review the method for
ordinary differentiable equations.
A first order differential equation is separable if it can be written with one
variable only on the left and the other variable only on the right:
f (y)y 0 = g(t)
To solve this equation, we proceed as follows. Let F (t) be an antiderivative
of f (t) and G(t) be an antiderivative of g(t). Then by the Chain Rule
dF dy
d
F (y) =
= f (y)y 0 .
dt
dy dt
Thus,
f (y)y 0 g(t) =
d
d
d
F (y) G(t) = [F (y) G(t)] = 0.
dt
dt
dt
It follows that
F (y) G(t) = C
which is equivalent to
Z
f (y)y dt =
Z
g(t)dt + C.
As you can see, the result is generally an implicit equation involving a function of y and a function of t. It may or may not be possible to solve this to
get y explicitly as a function of t. For an initial value problem, substitute the
values of t and y by t0 and y0 to get the value of C.
Remark 4.1
If F is a differentiable function of y and y is a differentiable function of t and
given by
both F and y are given then the chain rule allows us to find dF
dt
dF
dF dy
=
dt
dy dt
For separable equations, we are given f (y)y 0 = dF
and we are asked to find
dt
F (y). This process is referred to as reversing the chain rule.
32
Example 4.1
Solve the initial value problem y 0 = 6ty 2 , y(1) =
1
.
25
Solution.
Separating the variables and integrating both sides we obtain
Z 0
Z
y
dt = 6tdt
y2
or
d
dt

Z
1
dt = 6tdt.
y
Thus,
1
= 3t2 + C.
y(t)
1
Since y(1) = 25
, we find C = 28. The unique solution to the IVP is then
given explicitly by
1
y(t) =
28 3t2
Example 4.2
Solve the IVP yy 0 = 4 sin (2t), y(0) = 1.
Solution.
This is a separable differential equation. Integrating both sides we find

Z
Z
d y2
dt = 4 sin (2t)dt.
dt 2
Thus,
y 2 = 4 cos (2t) + C.
Since y(0) = 1, we find C = 5. Now, solving explicitly for y(t) we find
y(t) = 4 cos t + 5.
Since y(0) = 1, we have y(t) = 4 cos t + 5. The interval of existence of

the solution is the interval < t <
4 THE METHOD OF SEPARATION OF VARIABLES FOR ODES
Practice Problems
Problem 4.1
Solve the (separable) differential equation
2 ln y 2
y 0 = tet
Problem 4.2
y0 =
t2 y 4y
.
t+2
Problem 4.3
ty 0 = 2(y 4).
Problem 4.4
y 0 = 2y(2 y).
Problem 4.5
Solve the IVP
y0 =
Problem 4.6
Solve the IVP:
4 sin (2t)
, y(0) = 1.
y
yy 0 = sin t, y( ) = 2.
2
Problem 4.7
Solve the IVP:
y 0 + y + 1 = 0, y(1) = 0.
Problem 4.8
Solve the IVP:
y 0 ty 3 = 0, y(0) = 2.
33
34
Problem 4.9
Solve the IVP:
Problem 4.10
Solve the IVP:
y 0 = 1 + y 2 , y( ) = 1.
4
1
y 0 = t ty 2 , y(0) = .
2
Problem 4.11
Solve the equation 3uy + uxy = 0 by using the substitution v = uy .
Problem 4.12
Solve the IVP
(2y sin y)y 0 = sin t t, y(0) = 0.
Problem 4.13
State an initial value problem, with initial condition imposed at t0 = 2,
having implicit solution y 3 + t2 + sin y = 4.
Problem 4.14
Can the differential equation
dy
= x2 xy
dx
be solved by the method of separation of variables? Explain.
First Order Partial Differential

Equations
Many problems in the mathematical, physical, and engineering sciences deal
with the formulation and the solution of first order partial differential equations. Our first task is to understand simple first order equations. In applications, first order partial differential equations are most commonly used
to describe dynamical processes, and so time, t, is one of the independent
variables. Most of our discussion will focus on dynamical models in a single
space dimension, bearing in mind that most of the methods can be readily
extended to higher dimensional situations. First order partial differential
equations and systems model a wide variety of wave phenomena, including
transport of solvents in fluids, flood waves, acoustics, gas dynamics, glacier
motion, traffic flow, and also a variety of biological and ecological systems.
From a mathematical point of view, first order partial differential equations
have the advantage of providing conceptual basis that can be utilized in the
study of higher order partial differential equations.
In this chapter we introduce the basic definitions of first order partial differential equations. We then derive the one dimensional spatial transport
eqution and discuss some methods of solutions. One general method of solvability for quasilinear first order partial differential equation, known as the
method of characteristics, is analyzed.
35
36
FIRST ORDER PARTIAL DIFFERENTIAL EQUATIONS
5 Classification of First Order PDEs

In this section, we present the basic definitions pertained to first order PDE.
By a first order partial differential equation in two variables x and y
we mean any equation of the form
F (x, y, u, ux , uy ) = 0.
(5.1)
In what follows the functions a, b, and c are assumed to be continuously

differentiable functions. If Equation (5.1) can be written in the form
a(x, y, u)ux + b(x, y, u)uy = c(x, y, u)
(5.2)
then we say that the equation is quasi-linear. The following are examples
of quasi-linear equations:
uux + uy + cu2 = 0
x(y 2 + u)ux y(x2 + u)uy = (x2 y 2 )u.
If Equation (5.1) can be written in the form
a(x, y)ux + b(x, y)uy = c(x, y, u)
(5.3)
then we say that the equation is semi-linear. The following are examples
of semi-linear equations:
xux + yuy = u2 + x2
(x + 1)2 ux + (y 1)2 uy = (x + y)u2 .
a(x, y)ux + b(x, y)uy + c(x, y)u = d(x, y)
(5.4)
then we say that the equation is linear. Examples of linear equations are:
xux + yuy = cu
(y z)ux + (z x)uy + (x y)uz = 0.
A first order pde that is not linear is said to be non-linear. Examples of
non-linear equations are:
ux + cu2y = xy
5 CLASSIFICATION OF FIRST ORDER PDES
37
u2x + u2y = c.
First order partial differential equations are classified as either linear or nonlinear. Clearly, linear equations are a special kind of quasi-linear equation
(5.2) if a and b are functions of x and y only and c is a linear function of u.
Likewise, semi-linear equations are quasilinear equations if a and b are functions of x and y only. Also, semi-linear equations (5.3) reduces to a linear
equation if c is linear in u.
A linear first order partial differential equation is called homogeneous if
d(x, y) 0 and non-homogeneous if d(x, y) 6= 0. Examples of linear homogeneous equations are:
xux + yuy = cu
(y z)ux + (z x)uy + (x y)uz = 0.
Examples of non-homogeneous equations are:
ux + (x + y)uy u = ex
yux + xuy = xy.
Recall that for an ordinary linear differential equation, the general solution
depends mainly on arbitrary constants. Unlike ODEs, in linear partial differential equations, the general solution depends on arbitrary functions.
Example 5.1
Solve the equation ut (x, t) = 0.
Solution.
The general solution is given by u(x, t) = f (x) where f is an arbitrary differentiable function of x
Example 5.2
Consider the transport equation
aut (x, t) + bux (x, t) = 0
where a and b are constants. Show that u(x, t) = f (bt ax) is a solution
to the given equation, where f is an arbitrary differentiable function in one
variable.
Solution.
Let v(x, t) = bt ax. Using the chain rule we see that ut (x, t) = bfv (v) and
ux (x, t) = afv (v). Hence, aut (x, t) + bux (x, t) = abfv (v) abfv (v) = 0
38
Practice Problems
Problem 5.1
Classify each of the following PDE as linear, quasi-linear, semi-linear, or nonlinear.
(a) xux + yuy = sin (xy).
(b) ut + uux = 0
(c) u2x + u3 u4y = 0.
(d) (x + 3)ux + xy 2 uy = u3 .
Problem 5.2
Show that u(x, y) = ex f (2x y), where f is a differentiable function of one
variable, is a solution to the equation
ux + 2uy u = 0.
Problem 5.3
Show that u(x, y) = x xy satisfies the equation

xux yuy = u
subject to
u(y, y) = y 2 , y 0.
Problem 5.4
Show that u(x, y) = cos (x2 + y 2 ) satisfies the equation
yux + xuy = 0
subject to
u(0, y) = cos y 2 .
Problem 5.5
Show that u(x, y) = y 21 (x2 y 2 ) satisfies the equation
1
1
1
ux + uy =
x
y
y
subject to u(x, 1) = 21 (3 x2 ).
5 CLASSIFICATION OF FIRST ORDER PDES
39
Problem 5.6
Find a relationship between a and b if u(x, y) = f (ax+by) is a solution to the
equation 3ux 7uy = 0 for any differentiable function f such that f 0 (x) 6= 0
for all x.
Problem 5.7
Reduce the partial differential equation
aux + buy + cu = 0
to a first order ODE by introducing the change of variables s = bx ay and
t = x.
Problem 5.8
Solve the partial differential equation
ux + uy = 1
by introducing the change of variables s = x y and t = x.
Problem 5.9
Show that u(x, y) = e4x f (2x 3y) is a solution to the first-order PDE
3ux + 2uy + 12u = 0.
Problem 5.10
Derive the general solution of the PDE
aut + bux = u, b 6= 0
by using the change of variables v = ax bt and w = x.
Problem 5.11
Derive the general solution of the PDE
aux + buy = 0, a 6= 0
by using the change of variables s = bx ay and t = x.
40
Problem 5.12
Write the equation
ut + cux + u = f (x, t), c 6= 0
in the coordinates v = x ct, w = x.
Problem 5.13
Suppose that u(x, t) = w(x ct) is a solution to the PDE
xux + tut = Au
where A and c are constants. Let v = x ct. Write the differential equation
with unknown function w(v).
6 A REVIEW OF MULTIVARIABLE CALCULUS
41
6 A Review of Multivariable Calculus

In this section, we recall some concepts from vector calculus that we encounter later in the book.
6.1 Multiplication of Vectors: The Scalar or Dot Product

Is there such thing as multiplying a vector by another vector? The answer
is yes. As a matter of fact there are two types of vector multiplication. The
first one is known as scalar or dot product4 and produces a scalar; the
second is known as the vector or cross product and produces a vector. In
this section we will discuss the former one leaving the latter one for the next
section.
One of the motivation for using the dot product is the physical situation to
which it applies, namely that of computing the work done on an object by a
given force over a given distance, as shown in Figure 6.1.1.
Figure 6.1.1
Indeed, the work W is given by the expression
W = ||F~ || ||P Q|| cos
where ||F~ || cos is the component of F~ in the direction of P Q.

Thus, we define the dot product of two vectors ~u and ~v to be the number
~u ~v = ||~u|| ||~v || cos , 0
4
Also called inner product.
42
where is the angle between the two vectors as shown in Figure 6.1.2.
Figure 6.1.2
The above definition is the geometric definition of the dot product. We
next provide an algebraic way for computing the dot product. Indeed, let
~u = u1~i + u2~j + u3~k and ~v = v1~i + v2~j + v3~k. Then ~v ~u = (v1 u1 )~i + (v2
u2 )~j + (v3 u3 )~k. Moreover, we have ||~u||2 = u21 + u22 + u23 , ||~v ||2 = v12 + v22 + v32
and
||~v ~u||2 =(v1 u1 )2 + (v2 u2 )2 + (v3 u3 )2
=v12 2v1 u1 + u21 + v22 2v2 u2 + u22 + v32 2v3 u3 + u23 .
Now, applying the Law of Cosines to Figure 6.1.3 we can write
||~v ~u||2 = ||~u||2 + ||~v ||2 2||~u|| ||~v || cos .
Thus, by substitution we obtain
v12 2v1 u1 +u21 +v22 2v2 u2 +u22 +v32 2v3 u3 +u23 = u21 +u22 +u23 +v12 +v22 +v32 2||~u|| ||~v || cos
or
||~u|| ||~v || cos = u1 v1 + u2 v2 + u3 v3
so that we can define the dot product algebraically by
~u ~v = u1 v1 + u2 v2 + u3 v3 .
Figure 6.1.3

Example 6.1.1
Compute the dot product of ~u =
the angle between these vectors.
1 ~
i
2
1 ~
j
2
1 ~
k
2
43
and ~v = 12~i + 12~j + ~k and
Solution.
We have
1 1
1
1
1
1
1 1
~u ~v = + + 1 = + + = 2.
2 2
2 2
2
2 2 2 2
2
We also have
2
2
2
1
1
1
3
||~u|| =
+
+
=
2
2
2
2
2 2
1
3
1
+
+1= .
||~v ||2 =
2
2
2
2
Thus,
~u ~v
2 2
cos =
=
.
||~u|| ||~v ||
3
Hence,
= cos1
!
2 2
0.34 rad 19.5
3
Remark 6.1.1
The algebraic definition of the dot product extends to vectors with any number of components.
Next, we discuss few properties of the dot product.
Theorem 6.1.1
For any vectors ~u, ~v , and w
~ and any scalar we have
(i) Commutative law: ~u ~v = ~v ~u.
(ii) Distributive law: (~u + ~v ) w
~ = ~u w
~ + ~v w.
~
(iii) ~u (~v ) = (~u) ~v = (~u ~v ).
(iv) Magnitude: ||~u||2 = ~u ~u.
(v) Two nonzero vectors ~u and ~v are orthogonal or perpendicular if and
only if ~u ~v = 0.
(vi)) Two nonzero vectors ~u and ~v are parallel if and only if ~u ~v = ||~u|| ||~v ||.
(vii) ~0 ~v = ~0.
44
Proof.
Write ~u = u1~i + u2~j + u3~k, ~v = v1~i + v2~j + v3~k, and w
~ = w1~i + w2~j + w3~k.
Then
(i) ~u ~v = u1 v1 + u2 v2 + u2 v3 = v1 u1 + v2 u2 + v3 u3 = ~v ~u since product of
numbers is commutative.
(ii) (~u + ~v ) w
~ = ((u1 + v1 )~i + (u2 + v2 )~j + (u2 + v3 )~k) (w1~i + w2~j + w3~k) =
(u1 + v1 )w1 + (u2 + v2 )w2 + (u3 + v3 )w3 = u1 w1 + u2 w2 + u3 w3 + v1 w1 + v2 w2 +
v3 w3 = ~u w
~ + ~v w.
~
(iii) ~u (~v ) = (u1~i+u2~j +u3~k)(v1~i+v2~j +v3~k) = u1 v1 +u2 v2 +u3 v3 =
(u1 v1 + u2 v2 + u3 v3 ) = (~u ~v ).
(iv) ||~u||2 = ~u ~u cos 0 = ~u ~u.
(v) If ~u and ~v are perpendicular then the cosine of their angle is zero and so
the dot product is zero. Conversely, if the dot product of the two vectors is
zero then the cosine of their angle is zero and this happens only when the
two vectors are perpendicular.
(vi) If ~u and ~v are parallel then the cosine of their angle is either 1 or 1.
That is, ~u ~v = ||~u|| ||~v ||. Conversely, if ~u ~v = ||~u|| ||~v || then cos = 1
and this implies that either = 0 or = . In either case, the two vectors
are parallel.
(vii) In 3-D, ~0 =< 0, 0, 0 > and ~v =< a, b, c > so that ~0 ~v = (0 a)~i + (0
b)~j + (0 c)~k = ~0
Remark 6.1.2
Note that the unit vectors ~i, ~j, ~k associated with the coordinate axes satisfy
the equalities
~i ~i = ~j ~j = ~k ~k = 1 and ~i ~j = ~j ~k = ~i ~k = 0.
Example 6.1.2
(a) Show that the vectors ~u = 3~i 2~j and ~v = 2~i + 3~j are perpendicular.
(b) Show that the vectors ~u = 2~i + 6~j 4~k and ~v = 3~i 9~j + 6~k are parallel.
Solution.
(a) We have: ~u ~v = 3(2) 2(3) = 0. Hence ~u is perpendicular to ~v .
(b) We have:
cos =
2(3) + (6)(9) 4(6)

~u ~v
p
= p
= 1.
||~u||||~v ||
[ 22 + (6)2 + (4)2 ][ (3)2 + (9)2 + 62 ]
45
Hence, = so that the two vectors are parallel. Another way to see that
the vectors are parallel is to notice that ~u = 23 ~v
Projection of a vector onto a line
The orthogonal projection of a vector along a line is obtained by taking a
vector with same length and direction as the given vector but with its tail on
the line and then dropping a perpendicular onto the line from the tip of the
vector. The resulting vector on the line is the vectors orthogonal projection
or simply its projection. See Figure 6.1.4.
Figure 6.1.4
Now, if ~u is a unit vector along the line of projection and if ~vparallel is the
vector projection of ~v onto ~u then
~vparallel = (||~v || cos )~u = (~v ~u)~u.
See Figure 6.1.5. Also, the component perpendicular to ~u is given by
~vperpendicular = ~v ~vparallel .
Figure 6.1.5
46
We call Comp~u~v = ~v ~u the the scalar projection of ~v onto ~u. We call the
vector Proj~u~v = ~vparallel the vector projection of ~v onto ~u.
It follows that the vector ~v can be written in terms of ~vparallel and ~vperpendicular
~v = ~vparallel + ~vperpendicular .
Example 6.1.3
Write the vector ~v = 3~i + 2~j 6~k as the sum of two vectors, one parallel,
and one perpendicular to w
~ = 2~i 4~j + ~k.
Solution.
~
=
Let ~u = ||w
w||
~
2 ~
i
21
4 ~
j
21

~vparallel = (~v ~u)~u =
1 ~
k.
21
Then,

6
8
6
32
8
16

~u = ~i + ~j ~k.
21
21
21
21
21
21
Also,
~vperpendicular

32 ~
8 ~
16 ~
i+ 2
j + 6 +
k
=~v ~vparallel = 3 +
21
21
21
79
10
118 ~
= ~i + ~j
k.
21
21
21
Hence,
~v = ~vparallel + ~vperpendicular
Example 6.1.4
Find the scalar projection and vector projection of ~u =< 1, 1, 2 > onto
~v =< 2, 3, 1 > .
Solution.
We have
1(2) + (1)(3) + 2(1)
3
~u ~v
= p
=
2
2
2
||~v ||
14
(2) + 3 + 1
~u ~v ~v
Proj~v ~u =
||~v || ||~v ||
3
2
9
3
= ~v = ~i + ~j + ~k
14
7
14
14
comp~v ~u =
Applications
As pointed out earlier in the section, scalar products are used in Physics.
For instance, in finding the work done by a force applied on an object.
47
Example 6.1.5
A wagon is pulled a distance of 100 m along a horizontal path by a constant
force of 70 N. The handle of the wagon is held at an angle of 35 above the
horizontal. Find the work done by the force.
Solution.
The work done is
W = F d cos 35 = 70(100) cos 35 5734 J
48
Practice Problems
Problem 6.1.1
Find ~a ~b where ~a =< 4, 1, 41 > and ~b =< 6, 3, 8 > .
Problem 6.1.2
Find ~a ~b where ||~a|| = 6, ||~b|| = 5 and the angle between the two vectors is
120 .
Problem 6.1.3
If ~u is a unit vector, find ~u ~v and ~u w
~ using the figure below.
Problem 6.1.4
Find the angle between the vectors ~a =< 4, 3 > and ~b =< 2, 1 > .
Problem 6.1.5
Find the angle between the vectors ~a =< 4, 3, 1 > and ~b =< 2, 0, 1 > .
Problem 6.1.6
Determine whether the given vectors are orthogonal, parallel, or neither.
(a) ~a =< 5, 3, 7 > and ~b =< 6, 8, 2 > .
(b) ~a =< 4, 6 > and ~b =< 3, 2 > .
(c) ~a = ~i + 2~j + ~k and ~b = 3~i + 4~j ~k.
(d) ~a = 2~i + 6~j 4~k and ~b = 3~i 9~j + 6~k.
Problem 6.1.7
Use vectors to decide whether the triangle with vertices P (1, 3, 2), Q(2, 0, 4),
and R(6, 2, 5) is right-angled.
Problem 6.1.8
Find a unit vector that is orthogonal to both ~i + ~j and ~i + ~k.
49
Problem 6.1.9
Find the acute angle between the lines 2x y = 3 and 3x + y = 7.
Problem 6.1.10
Find the scalar and vector projections of the vector ~b =< 1, 2, 3 > onto
~a =< 3, 6, 2 > .
Problem 6.1.11
If ~a =< 3, 0, 1 >, find a vector ~b such that comp~a~b = 2.
Problem 6.1.12
Find the work done by a force F~ = 8~i 6~j + 9~k that moves an object from
the point (0, 10, 8) to the point (6, 12, 20) along a straight line. The distance
is measured in meters and the force in newtons.
Problem 6.1.13
A sled is pulled along a level path through snow by a rope. A 30-lb force
acting at an angle of 40 above the horizontal moves the sled 80 ft. Find the
work done by the force.
50
6.2 Directional Derivatives and the Gradient Vector

Given a function z = f (x, y) and let (x0 , y0 ) be in the domain of f. We wish
to find the rate of change of f at (x0 , y0 ) in the direction of a unit vector
~u =< a, b > . To do this, we consider the vertical plane to the graph S of f
that passes through the point P (x0 , y0 , z0 ) in the direction of ~u. This plane
inersects the graph S in a curve C. (See Figure 6.2.1.)
Figure 6.2.1
The slope of the tangent line T to C at the point P is the rate of change
of z in the direction of ~u. Let Q(x, y, z) be an arbitrary point on C and let
P 0 (x0 , y0 , 0) and Q0 (x, y, 0) be the orthogonal projection of P and Q respec
tively onto the xyplane. Then the vectors P 0 Q0 =< xx0 , y y0 > is paral
lel to ~u so that P 0 Q0 = h~u for some scalar h. Hence, x = x0 + ha, y = y0 + hb
and
f (x0 + ha, y0 + hb) f (x0 , y0 )
f (x, y) f (x0 , y0 )
=
.
h
h
If we take the limit of the above average rate as h 0, we obtain the rate
of change of z(with respect to distance) in the direction of ~u, which is called
the directional derivative of f at (x0 , y0 ) in the direction of ~u. We write
f (x0 + ah, y0 + bh) f (x0 , y0 )
.
h0
h
f~u (x0 , y0 ) = lim
51
Notice that if ~u = ~i then a = 1 and b = 0 so that f~u (x0 , y0 ) = fx (x0 , y0 ).

That is, fx is the rate of change of f in the x direction. Likewise, if ~u = ~j
then a = 0 and b = 1 so that f~u (x0 , y0 ) = fy (x0 , y0 ).
The following theorem provides a formula for computing the directional
derivative.
Theorem 6.2.1
If f is a differentiable function of x and y, then f has a directional derivative
in the direction of any unit vector ~u =< a, b > and
f~u (x, y) = fx (x, y)a + fy (x, y)b.
Proof.
Fix a point (x0 , y0 ) in the domain of f and consider the single variable function g(h) = f (x0 + ha, y0 + hb). Then
f (x0 + ah, y0 + bh) f (x, y)
g(h) g(0)
= lim
= f~u (x0 , y0 ).
h0
h0
h
h
g 0 (0) = lim
Let x = x0 + ah and y = y0 + bh. Using the Chain Rule, we find

g 0 (h) =
f dx f dy
+
= fx (x, y)a + fy (x, y)b.
x dh y dh
Letting h = 0 in the above expression, we find

f~u (x0 , y0 ) = g 0 (0) = fx (x0 , y0 )a + fy (x0 , y0 )b
(6.2.1)
Example 6.2.1

2
Find u~v (4, 0) if u(x, y) = x + y and ~v = 12 , 23 .
Solution.
We have

1
u~v (4, 0) = ux (4, 0)
+ uy (4, 0)
2
!
3
1
=
2
2
The Gradient Vector

The gradient is a generalization of the usual concept of derivative of a
function of one variable to functions of several variables. For a function
u(x, y) or u(x, y, z), the gradient are, respectively,
52

u(x, y) = ux~i + uy~j and u(x, y, z) = ux~i + uy~j + uz~k.
Example 6.2.2
Let F (x, y, z) = u(x, y) z. Find F (x, y, z).
Solution.
We have
F (x, y, z) = ux~i + uy~j ~k
Example 6.2.3
Find the gradient vector of f (x, y, z) = (2x 3y + 5z)5 .
Solution.
We have
fx (x, y, z) =10(2x 3y + 5z)4
fy (x, y, z) = 15(2x 3y + 5z)4
fz (x, y, z) =25(2x 3y + 5z)4 .
Thus,
f (x, y, z) = 5(2x 3y + 5z)4 [2~i 3~j + 5~k]
With the notation for the gradient vector, we can rewrite the expression
(6.2.1) for the directional derivative as
f~u (x0 , y0 ) = f (x0 , y0 ) ~u.
This expresses the directional derivative in the direction of ~u as the scalar
projection of the gradient vector onto ~u.
Maximizing the Directional Derivative
Suppose we have a function of two or three variables and we consider all
possible directional derivatives of f at a given point. These give the rates of
change of in all possible directions. We can then ask the questions: In which
of these directions does f change fastest and what is the maximum rate of
change? The answers are provided by the following theorem.
Theorem 6.2.2
The maximum value of the directional derivative of a function f (x, y) or
f (x, y, z) at a point (x, y) or (x, y, z) is ||f || and it occurs in the direction
of the gradient of f at that point.
53
Proof.
We have
f~u (x, y) = f ~u = ||f ||||~u|| cos = ||f || cos ,
where is the angle between f and ~u. The maximum value of cos is 1 and
this occurs when = 0. Therefore the maximum value of f~u is ||f || and it
occurs when = 0, that is, when ~u has the same direction as f
Example 6.2.4
Find the maximum rate of change of the function u(x, y) = 50 x2 2y 2 at
the point (1, 1).
Solution.
The maximum rate of change occurs in the direction of the gradient vector:
u(1, 1) = ux (1, 1)~i + uy (1, 1)~j = 2~i + 4~j.
The maximum rate of change at (1, 1) is
||u(1, 1)|| =
(2)2 + 42 = 2 5
Significance of the Gradient Vector

Suppose that a curve in 3-D is defined parametrically by the equations x =
x(t), y = y(t), z = z(t), where t is a parameter. This curve can be described
by the vector function
~r(t) = x(t)~i + y(t)~j + z(t)~k.
Its derivative is the tangent vector to the curve (See Figure 6.2.2) and is
given by
dx
dy
dz
d
(~r(t)) = ~i + ~j + ~k.
dt
dt
dt
dt
54
Figure 6.2.2
Now, for a function in two variables u(x, y), the equation u(x, y) = C is
called a level curve of u( a level surface of u(x, y, z)). The level curves
u(x, y) = C are just the traces of the graph of u(x, y) in the horizontal plane
z = C projected down to the xyplane.
An important property of the gradient of u is that it is normal to a level
surface of u at every point. To see this, let S be the level surface f (x, y, z) = k
and P0 (x0 , y0 , z0 ) be a point on S. Let C be any curve on S that passes
through P0 . We can describe C in parametric form x = x(t), y = y(t), and
z = z(t). Any point on C satisfies f (x(t), y(t), z(t)) = k. Differentiating both
sides of this equation with respect to t we find by means of the Chain Rule
fx (x, y, z)x0 (t) + fy (x, y, z)y 0 (t) + fz (x, y, z)z 0 (t) = 0
which can be written as f r0~(t) = 0. This means that the gradient is normal
to a level surface (respectively a level curve). See Figure 6.2.3.
Figure 6.2.3
55
If we consider a topographical map of a hill and let f (x, y) represent the

height above sea level at a point with coordinates (x, y), then a curve of
steepest ascent can be drawn as in Figure 6.2.4 by making it perpendicular
to all of the contour lines.
Figure 6.2.4
Vector Fields and Integral Curves
In vector calculus, a vector field is a function F~ (x, y) (or F~ (x, y, z) in 3-D
space) that assigns a vector to each point of its domain as shown in Figure
6.2.5.
Figure 6.2.5
56
Creating vector fields manually is very tedious. Thus, vector fields are generally generated using computer softwares such as Mathematica, Maple, or
Mathlab.
Example 6.2.5
The gradient vector of a function is an example of a vector field called the
gradient vector field. Sketch the gradient vector field of the function
u(x, y) = x2 + y 2 .
Describe the level curves of u(x, y).
Solution.
The gradient vector field of the given function is
u(x, y) = 2x~i + 2y~j.
A level curve is defined by the equation
x2 + y 2 = C, C 0.
Thus, level curves are circles centered at the origin. Figure 6.2.6 shows the
gradient vector field as well as some of the level curves.
Figure 6.2.6
57
For example, at the point (1, 2), the corresponding vector in the vector field
is the vector with tail (1, 2) and tip (2, 4)
An integral curve of a vector field is a smooth curve5 such that F~ (x, y)
assigns a tangent vector at each point of . For example, the integral curves
of the vector field F~ (x, y) = y~i x~j are circles centered at the origin. See
Figure 6.2.7.
Figure 6.2.7
If ~r(t) is a parametrization of then r~0 (t) is continuous and r~0 (t) 6= ~0.
58
Practice Problems
Problem 6.2.1
Find the gradient of the function
F (x, y, z) = exyz + sin (xy).
Problem 6.2.2
Find the gradient of the function
F (x, y, z) = x cos
y
z
Problem 6.2.3
Describe the level surfaces of the function f (x, y, z) = (x 2)2 + (y 3)2 +
(z + 5)2 .
Problem 6.2.4
Find the directional derivative of u(x, y) = 4x2 + y 2 in the direction of ~a =
~i + 2~j at the point (1, 1).
Problem 6.2.5
Find the directional derivative of u(x, y, z) = x2 z +y 3 z 2 xyz in the direction
of ~a = ~i + 3~k at the point (x, y, z).
Problem 6.2.6
Find the maximum rate of change of the function u(x, y) = yexy at the point
(0, 2) and the direction in which this maximum occurs.
Problem 6.2.7
Find the gradient vector field for the function u(x, y, z) = ez ln (x2 + y 2 ).
7 SOLVABILITY OF SEMI-LINEAR FIRST ORDER PDES
59
7 Solvability of Semi-linear First Order PDEs

In this section we discuss the solvability of the semi-linear first order PDE
a(x, y)ux + b(x, y)uy = f (x, y, u)
(7.1)
via the method of characteristics.

To solve (7.1), we proceed as follows. Suppose we have found a solution
u(x, y) to (7.1). This solution may be interpreted geometrically as a surface
in (x, y, z) space called the integral surface where z = u(x, y). This integral
surface can be viewed as the level surface of the function
F (x, y, z) = u(x, y) z = 0.
Then equation (7.1) can be written as the dot product
~v ~n = 0
(7.2)
where ~v = (a, b, f ) is the characteristic direction and ~n = F (x, y, z) =

(ux , uy , 1). Note that ~n is normal to the surface F (x, y, z) = 0 and is pointing downward. Hence, ~n is normal to ~v and this implies that ~v is tangent
to the surface F = 0 at (x, y, z). So our task to finding a solution to (7.1) is
equivalent to finding a surface S such that at every point on the surface the
vector
~v = a~i + b~j + f (x, y, u)~k.
is tangent to the surface. How do we construct such a surface? The idea
is to find the integral curves of the vector field ~v (see Section 6.2) and then
patch all these curves together to obtain the desired surface.
To this end, we start first by constructing a curve parametrized by t such
that at each point of the vector ~v is tangent to . A parametrization of
this curve is given by the vector function
~r(t) = x(t)~i + y(t)~j + u(t)~k.
Then the tangent vector is
dx
dy
du
d
r~0 (t) = (~r(t)) = ~i + ~j + ~k.
dt
dt
dt
dt
60
Hence, the vectors r~0 (t) and ~v are parallel so these two vectors are proportional and this leads to the ODE system
dx
dt
dy
dt
du
dt
f (x, y, u)
(7.3)
or in differential form
dx
dy
du
=
=
.
(7.4)
a
b
f (x, y, u)
By solving the system (7.3) or (7.4), we are assured that the vector ~v is
tangent to the curve which in turn lies in the solution surface S. In our
context, integral curves are called characteristic curves or simply characteristics of the PDE (7.1). We call (7.3) the characteristic equations.
The projection of into the xyplane is called the projected characteristic curve.
Once we have found the characteristic curves, the surface S is the union of
these characteristic curves. In summary, by introducing these characteristic
equations, we have reduced our partial differential equation to a system of
ordinary differential equations. We can use ODE theory to solve the characteristic equations, then piece together these characteristic curves to form a
surface. Such a surface will provide us with a solution to our PDE.
Remark 7.1
dy
= ab one obtains the general solution h(x, y) = k1 where k1
Solving dx
is constant. Likewise, solving du
= fa one obtains the general solution
dx
j(x, y, u) = k2 where k2 is a constant. The constant k2 is a function of
k1 . For the sake of discussion, suppose that h(x, y) = k1 can be expressed as
= fa is being replaced by g(x, k1 ) so that the
y = g(x, k1 ). Then, the y in du
dx
constant in j(x, y, u) = k2 will depend on k1 .
Example 7.1
Find the general solution to aux + buy = 0 where a and b are constants with
a 6= 0.
Solution.
dy
From (7.3) we can write dx
= ab which yields bx ay = k1 for some arbitrary
du
constant k1 . From dx = 0 we find u(x, y) = k2 where k2 is a constant. That
is, u(x, y) is constant on . Since (0, ka1 , k2 ) is on , we have
u(x, y) = u(0,
k1
) = k2
a
61
which shows that k2 is a function of k1 . Hence,

u(x, y) = f (k1 ) = f (bx ay)
where f is a differentiable function in one variable
In the next example, we show how the initial value problem for the PDE
determines the function f.
Example 7.2
Find the unique solution to aux + buy = 0, where a and b are constants with
a 6= 0, with the initial condition u(x, 0) = g(x).
Solution.
From the previous example, we found u(x, y) = f (bx ay) for some differen-
tiable function f. Since u(x, 0) = g(x), we find g(x) = f (bx) or f (x) = g xb
assuming that b 6= 0. Thus,

a
u(x, y) = g x y
b
Example 7.3
2
Find the solution to 3ux + uy = 0, u(x, 0) = ex .
Solution.
2
We have a = 3, b = 1 and g(x) = ex . The unique solution is given by
u(x, y) = e(x+3y)
Example 7.4
Find the general solution of the equation
xux + yuy = xeu , x > 0.
Solution.
We have a(x, y) = x, b(x, y) = y, and f (x, y, u) = xeu . So we have to solve
the system
dy
y du
= ,
= eu .
dx
x dx
62
From the first equation, we can use the separation of variables method to find
y = k1 x for some constant k1 . Solving the second equation by the method of
separation of variables, we find
eu x = k2 .
But k2 = g(k1 ) so that
u
e x = g(k1 ) = g
y
x
where g is a differentiable function of one variable

Example 7.5
Find the general solution of the equation
ux + uy u = y.
Solution.
The characteristic equations are
dy
du
d(u + y + 1)
dx
=
=
=
.
1
1
u+y
u+y+1
dy
Solving the equation dx
= 1 we find y x = k1 . Solving the equation dx =
d(u+y+1)
, we find u + y + 1 = k2 ex = f (y x)ex , where f is a differentiable
u+y+1
function of one variable. Hence,
u = (1 + y) + f (y x)ex
Example 7.6
Find the general solution to x2 ux + y 2 uy = (x + y)u.
Solution.
Using properties of proportions6 we have
dx
dy
du
dx dy
= 2 =
= 2
.
2
x
y
(x + y)u
x y2
6
If
a
b
c
d
then
ab
b
cd
d .
Also,
a
b
c
d
e
f
a+c+e
b+d+f .

Solving
dy
dx
y2
x2
by the method of separation of variables we find
From the equation
du
(x+y)u
d(xy)
x2 y 2
63
1
x
y1 = k1 .
we find
d(x y)
du
=
u
xy
which implies

u = k2 (x y) = f
1 1
x y

(x y)
Example 7.7
Find the solution satisfying yux + xuy = x2 + y 2 subject to the conditions
u(x, 0) = 1 + x2 and u(0, y) = 1 + y 2 .
Solution.
Solving the equation
have
dy
dx
x
y
we find x2 y 2 = k1 . On the other hand, we
du =y 1 (x2 + y 2 )dx
=ydx + x2 y 1 dx
y
=ydx + x2 y 1
dy
x
=ydx + xdy = d(xy).
Hence,
u(x, y) = xy + f (x2 y 2 ).
From u(x, 0) = 1 + x2 we find f (x) = 1 + x, x 0. From u(0, y) = 1 + y 2 we
find f (y) = 1 y, y 0. Hence, f (x) = 1 + |x| and
u(x, y) = xy + |x2 y 2 |
Remark 7.2
The method of characteristics discussed in this section applies as well to any
quasi-linear first order PDE. See Chapter 9.
64
Practice Problems
Problem 7.1
Solve ux + yuy = y 2 with the initial condition u(0, y) = sin y.
Problem 7.2
Solve ux + yuy = u2 with the initial condition u(0, y) = sin y.
Problem 7.3
Find the general solution of yux xuy = 2xyu.
Problem 7.4
Find the integral surface of the IVP: xux + yuy = u, u(x, 1) = 2 + e|x| .
Problem 7.5
Find the unique solution to 4ux + uy = u2 , u(x, 0) =
1
.
1+x2
Problem 7.6
2
Find the unique solution to e2y ux + xuy = xu2 , u(x, 0) = ex .
Problem 7.7
Find the unique solution to xux + uy = 3x u, u(x, 0) = tan1 x.
Problem 7.8
Solve: xux yuy = 0, u(x, x) = x4 .
Problem 7.9
Find the general solution of yux 3x2 yuy = 3x2 u.
Problem 7.10
Find u(x, y) that satisfies yux + xuy = 4xy 3 subject to the boundary conditions u(x, 0) = x4 and u(0, y) = 0.
8 LINEAR FIRST ORDER PDE: THE ONE DIMENSIONAL SPATIAL TRANSPORT EQUATIONS65
8 Linear First Order PDE: The One Dimensional Spatial Transport Equations
Modeling is the process of writing a differential equation to describe a physical situation. In this section we derive the one-dimensional spatial transport
equation and use the method of characteristics to solve it.
Linear Transport Equation for Fluid Flows
We shall describe the transport of a dissolved chemical by water that is traveling with uniform velocity c through a long thin tube G with uniform cross
section A. (The very same discussion applies to the description of the transport of gas by air moving through a pipe.) We assume the velocity c > 0
is in the (rightward) positive direction of the xaxis. We will also assume
that the concentration of the chemical is constant across the cross section
A located at x so that the chemical changes only in the xdirection and
thus the term one-dimensional spatial equation. This condition says that
the quantity of the chemical in a portion of the tube is the same but it is
traveling with time.
Let u(x, t) be a continuously differentiable function denoting the concentration of the chemical (i.e. amount of chemical per unit volume) at position
x and time t. Then at time t, the amount of chemical stored in a section of
the tube between positions a and x (see Figure 8.1) is given by the definite
integral
Z x
Au(s, t)ds.
a
Figure 8.1
66
Since the water is flowing at a constant speed c, so at time t + h the same

quantity of chemical will exist in the portion of the tube between a + ch and
x + ch. That is,
Z x+ch
Z x
Au(s, t + h)ds.
Au(s, t)ds =
a+ch
Taking the derivative of both sides with respect to x and using the Fundamental Theorem of Calculus, we find
u(x, t) = u(x + ch, t + h).
Now, taking the derivative of this last equation with respect to h and using
the chain rule, we find
0 = ut (x + ch, t + h) + cux (x + ch, t + h).
Taking the limit of this last equation as h approaches 0 and using the fact
that ut and ux are continuous, we find
ut (x, t) + cux (x, t) = 0
(8.1)
for all (x, t). This equation is called the transport equation in one-dimensional
space. It is a linear, homogeneous first order partial differential equation.
Note that (8.1) can be written in the form
< 1, c > < ut , ux >= 0
so that the left-hand side of (8.1) is the directional derivative of u(t, x) at
(t, x) in the direction of the vector < 1, c > .
Solvability via the method of characteristics
We will use the method of characteristics discussed in Chapter 7 to solve
(8.1). The characteristic equations are
dt =
dx
du
=
.
c
0
Thus, to solve (8.1), we solve the system of ODEs

dt
1 du
= ,
= 0.
dx
c dx
Solving the first equation, we find x ct = k1 . Solving the second equation
we find
u(x, t) = k2 = f (k1 ) = f (x ct).
One can check that this is indeed a solution to (8.1). Indeed, by using the
chain rule one finds
ut = cf 0 (x ct) and ux = f 0 (x ct).
Hence, by substituting these results into the equation one finds
ut + cux = cf 0 (x ct) + cf 0 (x ct) = 0.
The solution u(x, t) = f (x ct) is called the right traveling wave, since
the graph of the function f (x ct) at a given time t is the graph of f (x)
shifted to the right by the value ct. Thus, with growing time, the function
f (x) is moving without changes to the right at the speed c.
An initial value condition determines a unique solution to the transport equation as shown in the next example.
Example 8.1
2
Find the solution to ut 3ux = 0, u(x, 0) = ex .
Solution.
The characteristic equations lead to the ODEs
1 du
dt
= ,
= 0.
dx
3 dx
Solving the first equation, we find 3t + x = k1 . From the second equation, we
find u(x, t) = k2 = f (k1 ) = f (3t + x). From the initial condition, u(x, 0) =
2
f (x) = ex . Hence,
2
u(x, t) = e(3t+x)
Transport Equation with Decay
Recall from ODE that a function u is an exponential decay function if it
satisfies the equation
du
= u, < 0.
dt
68
A transport equation with decay is an equation given by

ut + cux + u = f (x, t)
(8.2)
where > 0 and c are constants and f is a given function representing

external resources. Note that the decay is characterized by the term u.
Note that (8.2) is a first order linear partial differential equation that can be
solved by the method of characteristics by solving the chracteristic equations
dx
dt
du
=
=
.
c
1
f (x, t) u
Example 8.2
Find the general solution of the transport equation
ut + ux + u = t.
Solution.
dt
du
dx
=
=
.
1
1
tu
From the equation dx = dt we find xt = k1 . Using a property of proportions
we can write
du
dt du
d(1 t + u)
dt
=
=
=
.
1
tu
1t+u
1t+u
Thus, 1 t + u = k2 et = f (x t)et or u(x, t) = t 1 + f (x t)et where
f is a differentiable function of one variable
Practice Problems
Problem 8.1
Find the solution to ut + 3ux = 0, u(x, 0) = sin x.
Problem 8.2
Solve the equation aux + buy + cu = 0.
Problem 8.3
Solve the equation ux +2uy = cos (y 2x) with the initial condition u(0, y) =
f (y), where f : R R is a given function.
Problem 8.4
Show that the initial value problem ut + ux = x, u(x, x) = 1 has no solution.
Problem 8.5
Solve the transport equation ut + 2ux = 3u with initial condition u(x, 0) =
1
.
1+x2
Problem 8.6
Solve ut + ux 3u = t with initial condition u(x, 0) = x2 .
Problem 8.7
Show that the decay term u in the transport equation with decay
ut + cux + u = 0
can be eliminated by the substitution w = uet .
Problem 8.8 (Well-Posed)
Let u be the unique solution to the IVP
ut + cux = 0
u(x, 0) = f (x)
and v be the unique solution to the IVP
ut + cux = 0
u(x, 0) = g(x)
70
where f and g are continuously differentiable functions.

(a) Show that w(x, t) = u(x, t) v(x, t) is the unique solution to the IVP
ut + cux = 0
u(x, 0) = f (x) g(x)
(b) Write an explicit formula for w in terms of f and g.
(c) Use (b) to conclude that the transport problem is well-posed. That is, a
small change in the initial data leads to a small change in the solution.
Problem 8.9
Solve the initial boundary value problem
ut + cux = u, x > 0, t > 0
u(x, 0) = 0, u(0, t) = g(t), t > 0.
Problem 8.10
Solve the first-order equation 2ut +3ux = 0 with the initial condition u(x, 0) =
sin x.
Problem 8.11
Solve the PDE ux + uy = 1.
9 SOLVING QUASI-LINEAR FIRST ORDER PDE VIA THE METHOD OF CHARACTERISTICS71
9 Solving Quasi-Linear First Order PDE via

the Method of Characteristics
In this section we develop a method for finding the general solution of a
quasi-linear first order partial differential equation
(9.1)
This method is called the method of characteristics or Lagranges method.

This method consists of transforming the PDE to a system of ODEs which
can be solved and the found solution is transformed into a solution for the
original PDE.
The method of characteristics relies on a geometrical argument. A visualization of a solution is an integral surface with equation z = u(x, y). An
alternative representation of this integral surface is
F (x, y, z) = u(x, y) z = 0.
That is, an integral surface is a level surface of the function F (x, y, z).
Now, recall from vector calculus that the gradient vector to a level surface
at the point (x, y, z) is a normal vector to the surface at that point. That
is, the gradient is a vector normal to the tangent plane to the surface at the
point (x, y, z). Thus, the normal vector to the surface F (x, y, z) = 0 is given
by
~n = F = Fx~i + Fy~j + Fz~k = ux~i + uy~j ~k.
Because of the negative z component, the vector ~n is pointing downward.
Now, equation (9.1) can be written as the dot product
(a(x, y, u), b(x, y, u), c(x, y, u)) (ux , uy , 1) = 0
or
~v ~n = 0
where ~v = a(x, y, u)~i + b(x, y, u)~j + c(x, y, u)~k. Thus, ~n is normal to ~v . Since
~n is normal to the surface F (x, y, z) = 0, the vector ~v must be tangential
to the surface F (x, y, z) = 0 and hence must lie in the tangent plane to the
surface at every point. Thus, to find a solution to (9.1) we need to find an
integral surface such that the surface is tangent to the vector ~v at each of its
point.
72
The required surface can be found as the union of integral curves, that is,
curves that are tangent to ~v at every point on the curve. If an integral curve
has a parametrization
~r(t) = x(t)~i + y(t)~j + u(t)~k
then the integral curve (i.e. the characteristic) is a solution to the ODE
system
dy
du
dx
= a(x, y, u),
= b(x, y, u),
= c(x, y, u)
(9.2)
dt
dt
dt
or in differential form
dx
dy
du
=
=
.
a(x, y, u)
b(x, y, u)
c(x, y, u)
(9.3)
Equations (9.2) or (9.3) are called characteristic equations. Note that

u(t) = u(x(t), y(t)) gives the values of u along a characteristic. Thus, along
a characteristic, the PDE (9.1) degenerates to an ODE.
Example 9.1
Find the general solution of the PDE yuux + xuuy = xy.
Solution.
dy
dx
= xu
= du
. Using the first two fractions
The characteristic equations are yu
xy
2
2
we find x y = k1 . Using the last two fractions we find u2 y 2 = f (x2 y 2 ).
Hence, the general solution can be written as u2 = y 2 + f (x2 y 2 ), where f
is an arbitrary differentiable single variable function
Example 9.2
Find the general solution of the PDE x(y 2 u2 )ux y(u2 +x2 )yy = (x2 +y 2 )u.
Solution.
The characteristic equations are x(y2dxu2 ) =
property of proportions we can write
dy
y(u2 +x2 )
du
.
(x2 +y 2 )u
xdx + ydy + udu

du
=
.
x2 (y 2 u2 ) y 2 (u2 + x2 ) + u2 (x2 + y 2 )
(x2 + y 2 )u
That is
xdx + ydy + udu

du
= 2
0
(x + y 2 )u
Using a
9 SOLVING QUASI-LINEAR FIRST ORDER PDE VIA THE METHOD OF CHARACTERISTICS73

or
xdx + ydy + udu = 0.
Hence, we find x2 + y 2 + u2 = k1 . Also,
y2
or
Hence, we find
dx
x
u2
dy
y
u2
+ x2
du
(x2 + y 2 )u
dx dy
du
=
.
x
y
u
yu
x
= k2 . The general solution is given by

u(x, y) =
x
f (x2 + y 2 + u2 )
y
where f is an arbitrary differentiable single variable function
74
Practice Problem
Problem 9.1
Find the general solution of the PDE ln (y + u)ux + uy = 1.
Problem 9.2
Find the general solution of the PDE x(y u)ux + y(u x)uy = u(x y).
Problem 9.3
Find the general solution of the PDE u(u2 + xy)(xux yuy ) = x4 .
Problem 9.4
Find the general solution of the PDE (y + xu)ux (x + yu)uy = x2 y 2 .
Problem 9.5
Find the general solution of the PDE (y 2 + u2 )ux xyuy + xu = 0.
Problem 9.6
Find the general solution of the PDE ut + uux = x.
Problem 9.7
Find the general solution of the PDE (y u)ux + (u x)uy = x y.
Problem 9.8
Solve
x(y 2 + u)ux y(x2 + u)uy = (x2 y 2 )u.
Problem 9.9
Solve
1 x2 ux + uy = 0.
Problem 9.10
Solve
u(x + y)ux + u(x y)uy = x2 + y 2 .
10 THE CAUCHY PROBLEM FOR FIRST ORDER QUASILINEAR EQUATIONS75
10 The Cauchy Problem for First Order Quasilinear Equations

When solving a partial differential equation, it is seldom the case that one
tries to study the properties of the general solution of such equations. In
general, one deals with those partial differential equations whose solutions
satisfy certain supplementary conditions. In the case of a first order partial
differential equation, we determine the particular solution by formulating an
initial value porblem also known as a Cauchy problem.
In this section, we discuss the Cauchy problem for the first order quasilinear
partial differential equation
(10.1)
Recall that the initial value problem of a first order ordinary differential
equation asks for a solution of the equation which has a given value at a
given point in R. The Cauchy problem for the PDE (10.1) asks for a solution
of (10.1) which has given values on a given curve in R2 . A precise statement
of the problem is given next.
Initial Value Problem or Cauchy Problem
Let C be a given curve in R2 defined parametrically by the equations
x = x0 (t), y = y0 (t)
where x0 , y0 are continuously differentiable functions on some interval I. Let
u0 (t) be a given continuously differentiable function on I. The Cauchy problem for (10.1) asks for a continuously differentiable function u = u(x, y)
defined in a domain R2 containing the curve C and such that:
(1) u = u(x, y) is a solution of (10.1) in .
(2) On the curve C, u equals the given function u0 (t), i.e.
u(x0 (t), y0 (t)) = u0 (t), t I.
(10.2)
We call C the initial curve of the problem, u0 (t) the initial data, and
(10.2) the initial condition or Cauchy data of the problem. See Figure
10.1.
76
Figure 10.1
If we view a solution u = u(x, y) of (10.1) as an integral surface of (10.1),
we can give a simple geometrical statement of the problem: Find a solution surface of (10.1) containing the curve described parametrically by the
equations
: x = x0 (t), y = y0 (t), u = u0 (t), t I.
Note that the projection of this curve in the xyplane is just the curve C.
The following theorem asserts that under certain conditions the Cauchy problem (10.1) - (10.2) has a unique solution.
Theorem 10.1
Suppose that x0 (t), y0 (t), and u0 (t) are continuously differentiable functions
of t in an interval I, and that a, b, and c are functions of x, y, and u with
continuous first order partial derivatives with respect to their argument in
some domain D of (x, y, u)space containing the initial curve
: x = x0 (t), y = y0 (t), u = u0 (t)
where t I. If (x0 (t), y0 (t), u0 (t)) is a point on that satisfies the condition
a(x0 (t), y0 (t), u0 (t))
dx0
dy0
(t) b(x0 (t), y0 (t), u0 (t))
(t) 6= 0
dt
dt
(10.3)

then by continuity this relation holds in a neighborhood U of (x0 (t), y0 (t), u0 (t))
so that is nowhere characteristic in U. In this case, there exists a unique
solution u = u(x, y) of (10.1) in U such that the initial condition (10.2) is
satisfied for every point on C contained in U. See Figure 10.2. That is, there
is a unique integral surface of (10.1) that contains in a neighborhood of
(x0 (t), y0 (t), u0 (t)).
Figure 10.2
It follows that the Cauchy problem has a unique solution if C is nowhere
characteristic.
The unique solution is found as follows: We solve the PDE by the method
of characteristics to obtain the general solution that involves an unknown
function. This unknown function is determined using the initial condition.
We illustrate this process in the next example.
Example 10.1
Solve the Cauchy problem
ux + uy =1
u(x, 0) =f (x).
78
Solution.
The initial curve in R3 can be given parametrically as
: x0 (t) = t, y0 (t) = 0, u0 (t) = f (t).
We have
dy0
dx0
(t) b(x0 (t), y0 (t), u0 (t))
(t) = 1 6= 0
dt
dt
so by the above theorem the given Cauchy problem has a unique solution.
Next we apply the results of the previous section to find the unique solution.
If we solve the characteristic equations in non-parametric form
a(x0 (t), y0 (t), u0 (t))
dy
du
dx
=
=
1
1
1
we find x y = c1 and u x = c2 . Thus, the general solution of the PDE
is given by u = x + F (x y). Using the Cauchy data u(x, 0) = f (x) we find
f (x) = x + F (x) which implies that F (x) = f (x) x. Hence, the unique
u(x, y) = x + f (x y) (x y) = y + f (x y)
Next, if condition (10.3) is not satisfied and is a characteristic curve, i.e.,
dy0
dx0
=b(x0 , y0 , u0 )
dt
dt
du0
dx0
a(x0 , y0 , u0 )
=c(x0 , y0 , u0 )
dt
dt
dy0
du0
c(x0 , y0 , u0 )
=b(x0 , y0 , u0 )
dt
dt
for all point on then the problem has infinitely many solutions. To see
this, pick an arbitrary point P0 = (x0 , y0 , u0 ) on . Pick a new initial curve
0 passing through P0 which is nowhere characteristic in a neighborhood of
P0 . In this case, condition (10.3) is satisfied and the new Cauchy problem has
a unique solution. Since there are infinitely many ways of selecting 0 , we
obtain infinitely many solutions. We illustrate this case in the next example.
a(x0 , y0 , u0 )
Example 10.2
ux + uy =1
u(x, x) =x.

Solution.
: x0 (t) = t, y0 (t) = t, u0 (t) = t.
We have
dy0
dx0
=b(x0 , y0 , u0 )
=1
dt
dt
dx0
du0
=c(x0 , y0 , u0 )
=1
a(x0 , y0 , u0 )
dt
dt
dy0
du0
c(x0 , y0 , u0 )
=b(x0 , y0 , u0 )
=1
dt
dt
a(x0 , y0 , u0 )
so that is a characteristic curve. As in Example 10.1, the general solution

of the PDE is u(x, y) = y + f (x y) where f is an arbitrary differentiable
function. Using the Cauchy data u(x, x) = x we find f (0) = 0. Thus, the
u(x, y) = y + f (x y)
where f is an arbitrary function such that f (0) = 0. There are infinitely
many choices for f. Hence, the problem has infinitely many solutions
If condition (10.3) is not satisfied and if
c(x0 , y0 , u0 )
du0
dx0
6= a(x0 , y0 , u0 )
dt
dt
for some points on then the Cauchy problem has no solutions. We illustrate
this case next.
Example 10.3
ux + uy =1
u(x, x) =1.
Solution.
: x0 (t) = t, y0 (t) = t, u0 (t) = 1.
80
We have
a(x0 (t), y0 (t), u0 (t))
dy0
dx0
(t) b(x0 (t), y0 (t), u0 (t))
(t) = 0.
dt
dt
and
du0
dx0
6= a(x0 , y0 , u0 )
= 0.
dt
dt
As in Example 10.1, the general solution to the PDE is given by u = y +
f (x y). Using the Cauchy data u(x, x) = 1 we find f (0) = 1 x, which is
not possible since the LHS is a fixed number whereas the RHS is a variable
expression. Hence, the problem has no solutions
1 = c(x0 , y0 , u0 )
Example 10.4
ux uy =1
u(x, 0) =x2 .
Solution.
The initial curve is given parametrically by
: x0 (t) = t, y0 (t) = 0, u0 (t) = t2 .
We have
a(x0 (t), y0 (t), u0 (t))
dx0
dy0
(t) b(x0 (t), y0 (t), u0 (t))
(t) = 1 6= 0
dt
dt
so the Cauchy problem has a unique solution.

The characteristic equations in non-parametric form are
dx
dy
du
=
=
.
1
1
1
Using the first two fractions we find x + y = c1 . Using the first and the third
fractions we find u x = c2 . Thus, the general solution can be represented
by
u = x + f (x + y)
where f is an arbitrary differentiable function. Using the Cauchy data
u(x, 0) = x2 we find x2 x = f (x). Hence, the unique solution is given
by
u = x + (x + y)2 (x + y) = (x + y)2 y

Example 10.5
ut + uux = x, u(x, 0) = 1.
Solution.
The initial curve is given parametrically by
: x0 (t) = t, y0 (t) = 0, u0 (t) = 1.
We have
a(x0 (t), y0 (t), u0 (t))
dx0
dy0
(t) b(x0 (t), y0 (t), u0 (t))
(t) = 1 6= 0
dt
dt

dx
du
dt
=
=
.
1
u
x
Since
dt
d(x + u)
=
1
x+u
t
we find that (x + u)e = c1 . Now, using the last two fractions we find
u2 x2 = k2 = f ((x + u)et ).
Using the Cauchy data u(x, 0) = 1, we find 1 x2 = f (1 + x) or f (1 + x) =
(1 + x)2 2x(1 + x). Thus, f (x) = x2 2x(x 1). The unique solution is
given by
u2 x2 = (x + u)2 e2t 2(x + u)et [(x + u)et 1]
or
u x = (x + u)e2t 2et [(x + u)et 1] = 2et (x + u)e2t .
This can be reduced further as follows: u + ue2t = x + 2et xe2t =
1e2t
2et
2et + x(1 e2t ) = u = 1+e
2t + x 1+e2t = sech(t) + xtanh(t)
Example 10.6
uux + uy = 1
where u(x, y) = 0 on the curve y 2 = 2x.
82
Solution.
A parametrization of is
: x0 (t) = 2t2 , y0 (t) = 2t, u0 (t) = 0, t > 0.
We have
a(x0 (t), y0 (t), u0 (t))
dy0
dx0
(t) b(x0 (t), y0 (t), u0 (t))
(t) = 4t 6= 0, t > 0
dt
dt

dy
du
dx
=
=
.
u
1
1
Using the last two fractions, we find u y = k1 . Using the first and the last
fractions, we find u2 2x = k2 = f (k1 ) = f (u y).
Using the initial condition, we find f (x) = x2 . Hence,
u2 2x = (u y)2
or equivalently
(u y)2 + u2 = 2x.
Solving this quadratic equation in u to find
1
2u = y (4x y 2 ) 2 .
The solution surface satisfying u = 0 on y 2 = 2x is given by
1
2u = y (4x y 2 ) 2 .
This represents a solution surface only when y 2 < 4x. The solution does not
exist for y 2 > 4x
Practice Problems
Problem 10.1
Solve
(y u)ux + (u x)uy = x y

with the condition u x, x1 = 0, x 6= 0, 1.
Problem 10.2
Solve the linear equation
yux + xuy = u
with the Cauchy data u(x, 0) = x3 , x > 0.
Problem 10.3
Solve
x(y 2 + u)ux y(x2 + u)uy = (x2 y 2 )u
with the Cauchy data u(x, x) = 1, x > 0
Problem 10.4
Solve
xux + yuy = xeu
with the Cauchy data u(x, x2 ) = 0, x > 0.
Problem 10.5
xux + uy = 0, u(x, 0) = f (x).
Problem 10.6
ut + aux = 0, u(x, 0) = f (x).
Problem 10.7
aux + uy = u2 , u(x, 0) = cos x
84
Problem 10.8
ux + xuy = u, u(1, y) = h(y).
Problem 10.9
uux + uy = 0, u(x, 0) = f (x)
where f is an invertible function.
Problem 10.10
1 x2 ux + uy = 0, u(0, y) = y.
Problem 10.11
Consider
xux + 2yuy = 0.
(i) Find and sketch the characteristics.
(ii) Find the solution with u(1, y) = ey .
(iii) What happens if you try to find the solution satisfying either u(0, y) =
g(y) or u(x, 0) = h(x) for given functions g and h?
Problem 10.12
Solve the equation ux + uy = u subject to the condition u(x, 0) = cos x.
Problem 10.13
(a) Find the general solution of the equation
ux + yuy = u.
(b) Find the solution satisfying the Cauchy data u(x, 3ex ) = 2.
(c) Find the solution satisfying the Cauchy data u(x, ex ) = ex .
Problem 10.14
ux + 4uy = x(u + 1), u(x, 5x) = 1.

Problem 10.15
ux uy = u, u(x, x) = sin x, x 6=
.
4
Problem 10.16
(a) Find the characteristics of the equation
yux + xuy = 0.
(b) Sketch some of the characteristics.
2
(c) Find the solution satisfying the boundary condition u(0, y) = ey .
Problem 10.17
Consider the equation ux + yuy = 0. Is there a solution satisfying the extra
condition
(a) u(x, 0) = 1
(b) u(x, 0) = x?
If yes, give a formula; if no, explain why.
86
Second Order Linear Partial

Differential Equations
In this chapter we consider the three fundamental second order linear partial
differential equations of parabolic, hyperbolic, and elliptic type. These types
arise in many applications such as the wave equation, the heat equation
and the Laplaces equation. We will study the solvability of each of these
equations.
87
88 SECOND ORDER LINEAR PARTIAL DIFFERENTIAL EQUATIONS
11 Second Order PDEs in Two Variables

In this section we will briefly review second order partial differential equations.
A second order partial differential equation in the variables x and y
is an equation of the form
F (x, y, u, ux , uy , uxx , uyy , uxy ) = 0.
(11.1)

A(x, y, u, ux , uy )uxx +B(x, y, u, ux , uy )uxy +C(x, y, u, ux , uy )uyy = D(x, y, u, ux , uy )
(11.2)
then we say that the equation is quasi-linear.

A(x, y)uxx + B(x, y)uxy + C(x, y)uyy = D(x, y, u, ux , uy )
(11.3)
then we say that the equation is semi-linear.

A(x, y)uxx +B(x, y)uxy +C(x, y)uyy +D(x, y)ux +E(x, y)uy +F (x, y)u = G(x, y)
(11.4)
then we say that the equation is linear.
A linear equation is said to be homogeneous when G(x, y) 0 and nonhomogeneous otherwise.
Equation (11.4) resembles the general equation of a conic section
Ax2 + Bxy + Cy 2 + Dx + Ey + F = 0
which is classified as either parabolic, hyperbolic, or elliptic based on the sign
of the discriminant B 2 4AC. We do the same for a second order linear
partial differential equation:
Hyperbolic: This occurs if B 2 4AC > 0 at a given point in the domain
of u.
Parabolic: This occurs if B 2 4AC = 0 at a given point in the domain
of u.
Elliptic: This occurs if B 2 4AC < 0 at a given point in the domain of
u.
11 SECOND ORDER PDES IN TWO VARIABLES
89
Example 11.1
Determine whether the equation uxx + xuyy = 0 is hyperbolic, parabolic or
elliptic.
Solution.
Here we are given A = 1, B = 0, and C = x. Since B 2 4AC = 4x, the
given equation is hyperbolic if x < 0, parabolic if x = 0 and elliptic if x > 0
Second order partial differential equations arise in many areas of scientific
applications. In what follows we list some of the well-known models that are
of great interest:
1. The heat equation in one-dimensional space is given by
ut = kuxx
where k is a constant.
2. The wave equation in one-dimensional space is given by
utt = c2 uxx
where c is a constant.
3. The Laplace equation is given by
u = uxx + uyy = 0.
Practice Problems
Problem 11.1
Classify each of the following equation as hyperbolic, parabolic, or elliptic:
(a) Wave propagation: utt = c2 uxx , c > 0.
(b) Heat conduction: ut = cuxx , c > 0.
(c) Laplaces equation: u = uxx + uyy = 0.
Problem 11.2
Classify the following linear scalar PDE with constant coefficents as hyperbolic, parabolic or elliptic.
(a) uxx + 4uxy + 5uyy + ux + 2uy = 0.
(b) uxx 4uxy + 4uyy + 3ux + 4u = 0.
(c) uxx + 2uxy 3uyy + 2ux + 6uy = 0.
Problem 11.3
Find the region(s) in the xyplane where the equation
(1 + x)uxx + 2xyuxy y 2 uyy = 0
is elliptic, hyperbolic, or parabolic. Sketch these regions.
Problem 11.4
Show that u(x, t) = cos x sin t is a solution to the problem
utt
u(x, 0)
ut (x, 0)
ux (0, t)
=
=
=
=
uxx
0
cos x
0
for all x, t > 0.

Problem 11.5
Classify each of the following PDE as linear, quasilinear, semi-linear, or nonlinear.
(a) ut + uux = uuxx
(b) xutt + tuyy + u3 u2x = t + 1
(c) utt = c2 uxx
(d) u2tt + ux = 0.
11 SECOND ORDER PDES IN TWO VARIABLES
91
Problem 11.6
Show that, for all (x, y) 6= (0, 0), u(x, y) = ln (x2 + y 2 ) is a solution of
uxx + uyy = 0,
and that, for all (x, y, z) 6= (0, 0, 0), u(x, y, z) =
1
x2 +y 2 +z 2
is a solution of
uxx + uyy + uzz = 0.

Problem 11.7
Consider the eigenvalue problem
uxx = u, 0 < x < L
ux (0) = k0 u(0)
ux (L) = kL u(L)
with Robin boundary conditions, where k0 and kL are given positive numbers
and u = u(x). Can this system have a nontrivial solution u 6 0 for > 0?
Hint: Multiply the first equation by u and integrate over x [0, L].
Problem 11.8
Show that u(x, y) = f (x)g(x), where f and g are arbitrary differentiable
functions, is a solution to the PDE
uuxy = ux uy .
Problem 11.9
Show that for any n N, the function un (x, y) = sin nx sinh ny is a solution
to the Laplace equation
u = uxx + uyy = 0.
Problem 11.10
Solve
uxy = xy.
Problem 11.11
Classify each of the following second-oder PDEs according to whether they
are hyperbolic, parabolic, or elliptic:
(a) 2uxx 4uxy + 7uyy u = 0.
(b) uxx 2 cos xuxy sin2 xuyy = 0.
(c) yuxx + 2(x 1)uxy (y + 2)uyy = 0.

Problem 11.12
Let c > 0. By computing ux , uxx , ut , and utt show that
Z
1
1 x+ct
u(x, t) = (f (x + ct) + f (x ct)) +
g(s)ds
2
2c xct
is a solution to the PDE
utt = c2 uxx
where f is twice differentiable function and g is a differentiable function.
Then compute and simplify u(x, 0) and ut (x, 0).
Problem 11.13
yuxx + uxy x2 uyy ux u = 0.
Determine the region D in R2 , if such a region exists, that makes this PDE:
(a) hyperbolic, (b) parabolic, (c) elliptic.
Problem 11.14
Consider the second-order hyperbolic PDE
uxx + 2uxy 3uyy = 0.
Use the change of variables v(x, y) = y 3x and w(x, y) = x + y to solve the
given equation.
Problem 11.15
uxx + 2uxy 3uyy = 0.
u(x, 2x) = 1, ux (x, 2x) = x.
12 HYPERBOLIC TYPE: THE WAVE EQUATION
93
12 Hyperbolic Type: The Wave equation

The wave equation has many physical applications from sound waves in air
to magnetic waves in the Suns atmosphere. However, the simplest systems
to visualize and describe are waves on a stretched flexible string.
A flexible string of length L and constant mass density (i.e., mass per unit
length) is stretched horizontally along the xaxis with its left end placed at
x = 0 and its right end placed at x = L. From the left end we slightly shake
the string and we notice a small vibrations propagate through the string. We
make the following assumptions:
(a) the (pulling) tension force on the string is the dominant force and all
other forces acting on the string are negligible (no external forces are applied
to the string, the damping forces (resistance) and gravitational forces are
negligible);
(b) clearly a point on the string moves up and down along a curve but since
the horizontal displacement is small compared to the vertical displacement,
we will assume that each point of the string moves only vertically;
(c) at each point, the tension force has constant magnitude and has the
direction of the tangent line to the string.
We denote the vertical displacement from the xaxis of the string by u(x, t)
which is a function of position x and time t. That is, u(x, t) is the vertical
displacement from the equilibrium at position x and time t. Our aim is to
find an equation that is satisfied by u(x, t).
A displacement of a tiny piece of the string between points P and Q is shown
in Figure 12.1
Figure 12.1

where
(i) (x, t) is the angle between T~ (x, t) and ~i at x and time t; for small vibrations, we have 0;
(ii) T~ (x, t) is the (pulling) tension force in the string at position x and time
t and T~ (x + x, t) the tension force at position x + x and t. By (c) above,
we ||T~ (x, t)|| = ||T~ (x + x, t)|| = T.
Remember that the above tiny piece is assumed to be moving vertically.
Now, at P the vertical component of the tension force is T sin (x, t) whereas
at Q the vertical component is T sin (x + x, t). Then Newtons Law of motion
mass acceleration = net applied forces
gives
p
2u
(x)2 + (u)2 2 = T sin (x + x, t) T sin (x, t).
t
Next, dividing by x and letting x 0 we obtain

s
2 2
u u
=T (sin (x, t))
1+
2
x t
x

=T cos (x, t)
.
x
From Figure 12.1, we have
tan [(x, t)] = slope of tangent line to the string =
Applying the chain rule on this last equation, we find
2u
= cos2 (x, t) 2 .
x
x
Hence,
s
Since 0, we have
1+
u
x
2
2u
2u
3
=
T
cos
(x,
t)
.
t2
x2
u
.
x

u
x
95
= tan (x, t) 0 and cos (x, t) 1.
Hence,
2u
2u
=
T
t2
x2
or
utt (x, t) = c2 uxx (x, t)
(12.1)
where c2 = T . We call c the wave speed.

General Solution of (12.1)
Suppose that u(x, t) solves (12.1). Define the function w(x, t) = ut + cux .
Then
wt cwx = (ut + cux )t c(ut + cux )x = utt + cuxt cutx c2 uxx = 0.
Thus, w satisfies the transport equation
wt cwx = 0
which is a first order linear PDE. We will solve this equation by the method
of characteristics.
dt
dw
dx
=
=
.
c
1
0
Solving the equation
dx
dt
=
c
1
we find x + ct = k1 . From the last fraction, we find w(x, t) = k2 = g(x + ct)
where f is a twice differentiable function in one variable. One can easily check
that w(x, t) is a solution to (12.1). Likewise, the function v(x, t) = ut cux
satisfies the equation vt + cvx = 0 which can be solved by the method of
characteristics to obtain v(x, t) = f (x ct) where f is a twice differentiable
function in one variable. Again, one can easily check that v(x, t) is a solution
to (12.1). By the linearity of (12.1), the general solution to (12.1) is given by
u(x, t) = f (x ct) + g(x + ct).
(12.2)
The function f (x ct) repesents wave that are moving to the right at a constant speed c and the function g(x + ct) represents waves that are moving to

the left at the same speed c.
The function u(x, t) in (12.2) involves two arbitrary functions that are determined (normally) by two initial conditions.
Example 12.1
Find the solution to the Cauchy problem
utt =c2 uxx
u(x, 0) =v(x)
ut (x, 0) =w(x).
Solution.
The condition u(x, 0) is the initial position whereas ut (x, 0) is the initial
velocity. We have
u(x, 0) = f (x) + g(x) = v(x)
and
ut (x, 0) = cf 0 (x) + cg 0 (x) = w(x)
which implies that
1
1
f (x) g(x) = W (x) =
c
c
w(s)ds.
0
Therefore,
1
1
f (x) = (v(x) W (x))
2
c
and
1
1
g(x) = (v(x) + W (x)).
2
c
Finally,
1
u(x, t) = [v(x ct) + v(x + ct) +
2
1
= [v(x ct) + v(x + ct) +
2
1
(W (x + ct) W (x ct))]
cZ
1 x+ct
w(s)ds]
c xct
97
Practice Problems
Problem 12.1
Show that if v(x, t) and w(x, t) satisfy equation (12.1) then v + w is also
a solution to (12.1), where and are constants.
Problem 12.2
Show that any linear time independent function u(x, t) = ax + b is a solution
to equation (12.1).
Problem 12.3
Find a solution to (12.1) that satisfies the homogeneous conditions u(x, 0) =
u(0, t) = u(L, t) = 0.
Problem 12.4
utt =9uxx
u(x, 0) = cos x
ut (x, 0) =0.
Problem 12.5
utt =uxx
1
u(x, 0) =
1 + x2
ut (x, 0) =0.
Problem 12.6
utt =4uxx
u(x, 0) =1
ut (x, 0) = cos (2x).

Problem 12.7
utt =25uxx
u(x, 0) =v(x)
ut (x, 0) =0
where

v(x) =
1 if x < 0
0 if x 0.
Problem 12.8
utt =c2 uxx
2
u(x, 0) =ex
ut (x, 0) = cos2 x.
Problem 12.9
Prove that the wave equation, utt = c2 uxx satisfies the following properties,
which are known as invariance properties. If u(x, t) is a solution, then
(i) Any translate, u(x y, t) where y is a fixed constant, is also a solution.
(ii) Any derivative, say ux (x, t), is also a solution.
(iii) Any dilation, u(ax, at), is a solution, for any fixed constant a.
Problem 12.10
Find v(r) if u(r, t) =
v(r)
r
cos nt is a solution to the PDE

2
urr + ur = utt .
r
Problem 12.11
Find the solution of the wave equation on the real line ( < x < +)
with the initial conditions
u(x, 0) = ex , ut (x, 0) = sin x.
99
Problem 12.12
The total energy of the string (the sum of the kinetic and potential energies)
is defined as
Z
1 L 2
(ut + c2 u2x )dx.
E(t) =
2 0
(a) Using the wave equation derive the equation of conservation of energy
dE(t)
= c2 (ut (L, t)ux (L, t) ut (0, t)ux (0, t)).
dt
(b) Assuming fixed ends boundary conditions, that is the ends of the string
are fixed so that u(0, t) = u(L, t) = 0, for all t > 0, show that the energy is
constant.
(c) Assuming free ends boundary conditions for both x = 0 and x = L, that
is both u(0, t) and u(L, t) vary with t, show that the energy is constant.
Problem 12.13
For a wave equation with damping
utt c2 uxx + dut = 0, d > 0, 0 < x < L
with the fixed ends boundary conditions show that the total energy decreases.
Problem 12.14
(a) Verify that for any twice differentiable R(x) the function
u(x, t) = R(x ct)
is a solution of the wave equation utt = c2 uxx . Such solutions are called
traveling waves.
(b) Show that the potential and kinetic energies (see Exercise 12.12) are
equal for the traveling wave solution in (a).
Problem 12.15
Find the solution of the Cauchy wave equation
utt = 4uxx
u(x, 0) = x2 , ut (x, 0) = sin 2x.
Simplify your answer as much as possible.
100SECOND ORDER LINEAR PARTIAL DIFFERENTIAL EQUATIONS
13 Parabolic Type: The Heat Equation in OneDimensional Space

In this section, We will look at a model for describing the distribution of
temperature in a solid material as a function of time and space. More specifically, we will derive the heat equation that models the flow of heat in a rod
that is insulated everywhere except at the two ends.
Before we begin our discussion of the mathematics of the heat equation, we
must first determine what is meant by the term heat? Heat is type of energy
known as thermal energy. Heat travels in waves like other forms of energy,
and can change the matter it touches. It can heat it up and cause chemical
reactions like burning to occur.
Heat can be released through a chemical reaction (such as the nuclear reactions that make the Sun burn) or can be trapped for a limited time by
insulators. It is often released along with other kinds of energy such as light
waves or sound waves. For example, a burning candle releases light and heat
waves. On the other hand, an explosion releases light, heat, and sound waves.
The most common units of heat are BTU (British Thermal Unit), Calorie
and Joule.
Consider now a thin rod made of homogeneous heat conducting material (i.e.,
it is composed of the exact same material and no foreign bodies are in it)
of uniform density and constant cross section A, placed along the xaxis
from x = 0 to x = L as shown in Figure 13.1.
Figure 13.1
Assume the heat flows only in the xdirection, with the lateral sides well
insulated, and the only way heat can enter or leave the rod is at either end.
Also we assume that the temperature of the rod is constant at any point of
the cross section. In other words, temperature will only vary in x and we
can hence consider the rod to be a one spatial dimensional rod. We will also
assume that heat energy in any piece of the rod is conserved. That is, the
13 PARABOLIC TYPE: THE HEAT EQUATION IN ONE-DIMENSIONAL SPACE101

heat gained at one end is equal to the heat lost at the other end.
Let u(x, t) be the temperature of the cross section at the point x and the
time t. Consider a portion U of the rod from x to x + x of length x as
shown in Figure 13.2.
Figure 13.2
Divide the interval [x, x+x] into n subintervals each of length s using the
partition points x = s0 < s1 < < sn = x + x. Consider the portion Ui of
U of height s. The portion Ui is assumed to be thin so that the temperature
is constant throughout the volume. From the theory of heat conduction, the
quantity of heat Qi in Ui at time t is given by
Qi = cmi u(si1 , t) = cu(si1 , t)Vi
where mi is the mass of Ui , Vi is the volume of Ui and c is the specific
heat, that is, the amount of heat that it takes to raise one unit of mass of
the material by one unit of temperature.
But Ui is a cylinder of height s and area of base A so that Vi = As.
Hence,
Qi = cAu(si1 , t)s.
The quantity of heat in the portion U is given by
Q(t) = lim
n
X
i=1
Qi = lim
n
X
i=1
Z
cAu(si1 , t)s =
x+x
cAu(s, t)ds.
x
By differentiating we take the partial of u to find the change in heat with

respect to time.
Z x+x
dQ
=
cAut (s, t)ds.
dt
x

Assuming that u is continuously differentiable, we can apply the mean value
theorem for integrals and find x x + x such that
Z
x+x
ut (s, t)ds = xut (, t).

x
Thus, the rate of change of heat in U is given by

dQ
= cAxut (, t).
dt
On the other hand, by Fourier law of heat transfer, the rate of heat transfer through any cross section is proportional to the area A and the negative
gradient of the temperature normal to the cross section, i.e., KAux (x, t).
Note that if the temperature increases as x increases (i.e., the temperature is
hotter to the right), ux > 0 so that the heat flows to the left. This explains
the minus sign in the formula for Fourier law. Hence, according to this law
heat is transferred from areas of high temperature to areas of low temperature.
Now, the rate of heat flowing in U through the cross section at x is KAux (x, t)
and the rate of heat flowing out of U through the cross section at x + x is
KAux (x + x, t), where K is the thermal conductivity of the rod.
Now, the conservation of energy law states
rate of change of heat in U = rate of heat flowing in rate of heat flowing
out
or mathematically written as,
cAxut (, t) = KAux (x, t) + KAux (x + x, t)
or
cAxut (, t) = KA[ux (x + x, t) ux (x, t)].
Dividing this last equation by cAx and letting x 0 we obtain
ut (x, t) = kuxx (x, t)
(13.1)
K
is called the diffusivity constant.
where k = c
Equation (13.1) is the one dimensional heat equation which is second order,

linear, homogeneous, and of parabolic type.
The non-homogeneous heat equation
ut = kuxx + f (x)
is known as the heat equation with an external heat source f (x). An example of an exterenal heat source is the heat generated from a candle placed
under the bar.
The function
Z L
u(x, t)dx
E(t) =
0
7
is called the total thermal energy at time t of the entire rod.

Example 13.1
The two ends of a uniform rod of length L are insulated. There is a constant source of thermal energy q0 6= 0 and the temperature is initially
u(x, 0) = f (x).
(a) Write the equation and the boundary conditions for this model.
(b) Calculate the total thermal energy of the entire rod.
Solution.
(a) The model is given by the PDE
ut (x, t) = kuxx + q0
with boundary conditions
ux (0, t) = ux (L, t) = 0.
(b) First note that
Z
Z L
Z L
Z L
d L
u(x, t)dx =
ut (x, t)dx =
kuxx dx +
q0 dx
dt 0
0
0
0
= kux |L0 + q0 L = q0 L
since ux (0, t) = ux (L, t) = 0. Integrating with respect to t we find
E(t) = q0 Lt + C.
7
The total internal energy in the rod generated by the rods temeprature.

But C = E(0) =
is given by
RL
0
u(x, 0)dx =
RL
0
f (x)dx. Hence, the total thermal energy
f (x)dx + q0 Lt
E(t) =
0
Initial Boundary Value Problems

In order to solve the heat equation we must give the problem some initial
conditions. If you recall from the theory of ODE, the number of conditions
required for solving initial value problems always matched the highest order
of the derivative in the equation.
In partial differential equations the same idea holds except now we have to
pay attention to the variable we are differentiating with respect to as well.
So, for the heat equation we have got a first order time derivative and so we
will need one initial condition and a second order spatial derivative and so
we will need two boundary conditions.
For the initial condition, we define the temperature of every point along the
rod at time t = 0 by
u(x, 0) = f (x)
where f is a given (prescribed) function of x. This function is known as the
initial temperature distribution.
The boundary conditions will tell us something about what the temperature
is doing at the ends of the bar. The conditions are given by
u(0, t) = T0 and u(L, t) = TL .
and they are called as the Dirichlet conditions. In this case, the general
form of the heat equation initial boundary value problem is to find u(x, t)
satisfying
ut (x, t) =kuxx (x, t), 0 x L, t > 0
u(x, 0) =f (x), 0 x L
u(0, t) =T0 , u(L, t) = TL , t > 0.
In the case of insulated endpoints, i.e., there is no heat flow out of them, we
use the boundary conditions
ux (0, t) = ux (L, t) = 0.

These conditions are examples of what is known as Neumann boundary
conditions. In this case, the general form of the heat equation initial boundary value problem is to find u(x, t) satisfying
ut (x, t) =kuxx (x, t), 0 x L, t > 0
u(x, 0) =f (x), 0 x L
ux (0, t) =ux (L, t) = 0, t > 0.
Practice Problems
Problem 13.1
Show that if u(x, t) and v(x, t) satisfy equation (13.1) then u + v is also a
solution to (13.1), where and are constants.
Problem 13.2
Show that any linear time independent function u(x, t) = ax + b is a solution
to equation (13.1).
Problem 13.3
Find a linear time independent solution u to (13.1) that satisfies u(0, t) = T0
and u(L, T ) = TL .
Problem 13.4
Show that to solve (13.1) with the boundary conditions u(0, t) = T0 and
u(L, t) = TL it suffices to solve (13.1) with the homogeneous boundary
conditions u(0, t) = u(L, t) = 0.
Problem 13.5
Find a solution to (13.1) that satisfies the conditions u(x, 0) = u(0, t) =
u(L, t) = 0.
Problem 13.6
Let (I) denote equation (13.1) together with intial condition u(x, 0) = f (x),
where f is not the zero function, and the homogeneous boundary conditions
u(0, t) = u(L, t) = 0. Suppose a nontrivial solution to (I) can be written in
the form u(x, t) = X(x)T (t). Show that X and T satisfy the ODE
X 00 k X = 0 and T 0 T = 0
for some constant .
Problem 13.7
Consider again the solution u(x, t) = X(x)T (t). Clearly, T (t) = T (0)et .
Suppose that > 0.
(a) Show that X(x) = Aex + Bex , where = k and A and B are
arbitrary constants.
(b) Show that A and B satisfy the two equations A + B = 0 and A(eL
eL ) = 0.
(c) Show that A = 0 leads to a contradiction.
(d) Using (b) and (c) show that eL = eL . Show that this equality leads
to a contradiction. We conclude that < 0.

Problem 13.8
Consider the results of the previous exercise.
(a) Show that X(x) = c1 cos x + c2 sin x where =
.
k
2 2
(b) Show that = n = knL2 , where n is an integer.

Problem 13.9
P
ki2 2
Show that u(x, t) = ni=1 ui (x, t), where ui (x, t) = ci e L2 t sin
fies (13.1) and the homogeneous boundary conditions.
i
x
L
satis-
Problem 13.10
Suppose that a wire is stretched between 0 and a. Describe the boundary
conditions for the temperature u(x, t) when
(i) the left end is kept at 0 degrees and the right end is kept at 100 degrees;
and
(ii) when both ends are insulated.
Problem 13.11
Let ut = uxx for 0 < x < and t > 0 with boundary conditions
R 2 u(0,2t) =
0 = u(, t) and initial condition u(x, 0) = sin x. Let E(t) = 0 (ut + ux )dx.
Show that E 0 (t) < 0.
Problem 13.12
Suppose
ut = uxx + 4, ux (0, t) = 5, ux (L, t) = 6, u(x, 0) = f (x).
Calculate the total thermal energy of the one-dimensional rod (as a function
of time).
Problem 13.13
Consider the heat equation
ut = kuxx
for x (0, 1) and t > 0, with boundary conditions u(0, t) = 2 and u(1, t) = 3
for t > 0 and initial condition u(x, 0) = x for x (0, 1). A function v(x) that
satisfies the equation v 00 (x) = 0, with conditions v(0) = 2 and v(1) = 3 is
called a steady-state solution. That is, the steady-state solutions of the
heat equation are those solutions that dont depend on time. Find v(x).

Problem 13.14
Consider the equation for the one-dimensional rod of length L with given
heat energy source:
ut = uxx + q(x).
Assume that the initial temperature distribution is given by u(x, 0) = f (x).
Find the equilibrium (steady state) temperature distribution in the following
cases.
(a) q(x) = 0, u(0) = 0, u(L) = T.
(b) q(x) = 0, ux (0) = 0, u(L) = T.
(c) q(x) = 0, u(0) = T, ux (L) = .
Problem 13.15
Consider the equation for the one-dimensional rod of length L with insulated
ends:
ut = kuxx , ux (0, t) = ux (L, t) = 0.
(a) Give the expression for the total thermal energy of the rod.
(b) Show using the equation and the boundary conditions that the total
thermal energy is constant.
Problem 13.16
Suppose
ut = uxx + x, u(x, 0) = f (x), ux (0, t) = , ux (L, t) = 7.
(a) Calculate the total thermal energy of the one-dimensional rod (as a function of time).
(b) From part (a) find the value of for which a steady-state solution exist.
(c) For the above value of find the steady state solution.
14 SEQUENCES OF FUNCTIONS: POINTWISE AND UNIFORM CONVERGENCE109
14 Sequences of Functions: Pointwise and Uniform Convergence

In the next section, we will be constructing solutions to PDEs involving
infinite sums of sines and cosines. These infinite sums or series are called
Fourier series. Fourier series are examples of series of functions. Convergence of series of functions is defined in terms of convergence of a sequence of
functions. In this section we study the two types of convergence of sequences
of functions.
Recall that a sequence of numbers {an }
n=1 is said to converge to a number
L if and only if for every given > 0 there is a positive integer N = N ()
such that for all n N we have|an L| < .
What is the analogue concept of convergence when the terms of the sequence
are variables? Let D R and for each n N consider a function fn : D R.
Thus, we obtain a sequence of functions {fn }
n=1 . For such a sequence, there
are two types of convergenve that we consider in this section: pointwise convergence and uniform convergence.
We say that {fn }
n=1 converges pointwise in D to a function f : D R if
and only if
lim fn (x) = f (x)
n
for each x D. Equivalently, for a given x D and > 0 there is a positive

integer N = N (x, ) such that if n N then |fn (x) f (x)| < . It is
important to note that N is a function of both x and .
Example 14.1
nx
Define fn : [0, ) R by fn (x) = 1+n

2 x2 . Show that the sequence {fn }n=1
converges pointwise to the function f (x) = 0 for all x 0.
Solution.
For all x 0,
nx
= 0 = f (x)
n 1 + n2 x2
lim fn (x) = lim
Example 14.2
For each positive integer n let fn : (0, ) (0, ) be given by fn (x) = nx.
Show that {fn }
n=1 does not converge pointwise in D = (0, ).

Solution.
This follows from the fact that lim nx = for all x D
n
One of the weaknesses of this type of convergence is that it does not preserve
some of the properties of the base functions {fn }
n=1 . For example, if each fn
is continuous then the pointwise limit function need not be continuous. (See
Problem 14.1) A stronger type of convergence which preserves most of the
properties of the base functions is the uniform convergence which we define
next.
Let D be a subset of R and let {fn }
n=1 be a sequence of functions defined on
D. We say that {fn }n=1 converges uniformly on D to a function f : D R

if and only if for all > 0 there is a positive integer N = N () such that if
n N then |fn (x) f (x)| < for all x D.
This definition says that the integer N depends only on the given (in contrast to pointwise convergence where N depends on both x and ) so that
for n N , the graph of fn (x) is bounded above by the graph of f (x) + and
below by the graph of f (x) as shown in Figure 14.1.
Figure 14.1
Example 14.3
For each positive integer n let fn : [0, 1] R be given by fn (x) = nx . Show
that {fn }
n=1 converges uniformly to the zero function.
Solution.
Let > 0 be given. Let N be a positive integer such that N > 1 . Then for

n N we have
x
|x|
1
1

|fn (x) f (x)| = 0 =

<
n
n
n
N
for all x [0, 1]
Clearly, uniform convergence implies pointwise convergence to the same limit
function. However, the converse is not true in general. Thus, one way to show
that a sequence of functions does not converge uniformly is to show that it
does not converge pointwise.
Example 14.4
nx
Define fn : [0, ) R by fn (x) = 1+n
2 x2 . By Example 14.1, this sequence
converges pointwise to f (x) = 0. Let = 31 . Show that there is no positive
integer N with the property n N implies |fn (x) f (x)| < for all x 0.
Hence, the given sequence does not converge uniformly to f (x).
Solution.
For any positive integer N and for n N we have

1
1 1
fn
= >
f

n
n 2
Problem 14.1 shows a sequence of continuous functions converging pointwise
to a discontinuous function. That is, pointwise convergence does not preserve the property of continuity. One of the interesting features of uniform
convergence is that it preserves continuity as shown in the next example.
Example 14.5
Suppose that for each n 1 the function fn : D R is continuous in D.
Suppose that {fn }
n=1 converges uniformly to f. Let a D.
(a) Let > 0 be given. Show that there is a positive integer N such that if
n N then |fn (x) f (x)| < 3 for all x D.
(b) Show that there is a > 0 such that for all |x a| < we have |fN (x)
fN (a)| < 3 .
(c) Using (a) and (b) show that for |x a| < we have |f (x) f (a)| < .
Hence, f is continuous in D since a was arbitrary.

Solution.
(a) This follows from the definition of uniform convergence.
(b) This follows from the fact that fN is continuous at a D.
(c) For |x a| < we have |f (x) f (a)| = |f (a) fN (a) + fN (a) fN (x) +
fN (x)f (x)| |fN (a)f (a)|+|fN (a)fN (x)|+|fN (x)f (x)| < 3 + 3 + 3 =

From this example, we can write
lim lim fn (x) = lim lim fn (x).
xa n
n xa
Indeed,
lim lim fn (x) = lim f (x)
xa n
xa
=f (a) = lim fn (a)

n
= lim lim fn (x).

n xa
Does pointwise convergence allow the interchange of limits and integration?

The answer is no as shown in the next example.
Example 14.6
The sequence of function fn : (0, ) R defined by fn (x) =
pointwise to the zero function. Show that
Z
Z
lim fn (x)dx.
lim
fn (x)dx 6=
n
Solution.
We have
Z
1

x
x2
dx =
= .
n
2n 1
Hence,
Z
fn (x)dx =
lim
whereas
lim fn (x)dx = 0
x
n
converges

Contrary to pointwise convergence, uniform convergence preserves integration. That is, if {fn }
n=1 converges uniformly to f on a closed interval [a, b]
then
Z
Z
b
lim
fn (x)dx =
a
lim fn (x)dx.
a n
Theorem 14.1
Suppose that fn : [a, b] R is a sequence of continuous functions that
converges uniformly to f : [a, b] R. Then
Z b
Z b
Z b
lim
fn (x)dx =
lim fn (x)dx =
f (x)dx.
n
a n
Proof.
From Example 14.5, we have that f is continuous and hence integrable. Let
> 0 be given. By uniform convergence, we can find a positive integer N

for all x in [a, b] and n N. Thus, for n N ,
such that |fn (x) f (x) < ba
we have
Z b
Z b
Z b

|fn (x) f (x)|dx < .
f (x)dx
fn (x)dx

a
This completes the proof of the theorem

Now, what about differentiability? Again, pointwise convergence fails in
general to conserve the differentiability property. See Problem 14.1. Does
uniform convergence preserve differentiability? The answer is still no as
shown in the next example.
Example 14.7
Consider the family of functions fn : [1, 1] given by fn (x) =
(a) Show that fn is differentiable for each n 1.
(b) Show that for all x [1, 1] we have
x2 + n1 .
1
|fn (x) f (x)|
n
q
where f (x) = |x|. Hint: Note that x2 + n1 + x2 1n .

(c) Let > 0 be given. Show that there is a positive integer N such that for
n N we have

|fn (x) f (x)| < for all x [1, 1].
Thus, {fn }
n=1 converges uniformly to the non-differentiable function f (x) =
|x|.
Solution.
(a) fn is the composition of two differentiable functions so it is differentiable
with derivative

1
1 2
2
0
fn (x) = x x +
.
n
(b) We have
r
q 2 1 2 q 2 1 2

1 ( x + n x )( x + n + x )

q
|fn (x) f (x)| = x2 + x2 =

n

x2 + n1 + x2
1
n
=q
x2 + n1 + x2
1
n
1
n
1
= .
n
(c) Let > 0 be given. Since limn 1n = 0 we can find a positive integer
N such that for all n N we have 1n < . Now the answer to the question
follows from this and part (b)
Even when uniform convergence occurs, the process of interchanging limits and differentiation may fail as shown in the next example.
Example 14.8
Consider the functions fn : R R defined by fn (x) = sinnnx .
(a) Show that {fn }
n=1 converges uniformly to the function f (x) = 0.
(b) Note that {fn }
n=1 and f are differentiable functions. Show that
h
i0
lim fn0 (x) 6= f 0 (x) = lim fn (x) .
That is, one cannot, in general, interchange limits and derivatives.

Solution.
(a) Let > 0 be given. Let N be a positive integer such that N > 1 . Then
for n N we have

sin nx 1
<
|fn (x) f (x)| =
n n
and this is true for all x R. Hence, {fn }
n=1 converges uniformly to the
function f (x) = 0.
(b) We have limn fn0 () = limn cos n = limn (1)n which does not
converge. However, f 0 () = 0
Pointwise convergence was not enough to preserve differentiability, and neither was uniform convergence by itself. Even with uniform convergence the
process of interchanging limits with derivatives is not true in general. However, if we combine pointwise convergence with uniform convergence we can
indeed preserve differentiability and also switch the limit process with the
process of differentiation.
Theorem 14.2
Let {fn }
n=1 be a sequence of differentiable functions on [a, b] that converges
pointwise to some function f defined on [a, b]. If {fn0 }
n=1 converges uniformly
on [a, b] to a function g, then the function f is differentiable with derivative
equals to g. Thus,
h
i0
0
0
lim fn (x) = g(x) = f (x) = lim fn (x) .
n
Proof.
First, note that the function g is continuous in [a, b] since uniform convergence
preserves continuity. Let c be an arbitrary point in [a, b]. Then
Z x
fn0 (t)dt = fn (x) fn (c), x [a, b].
c
Taking the limit of both sides and using the facts that fn0 converges uniformly
to g and fn converges pointwise to f , we can write
Z x
g(t)dt = f (x) f (c).
c

Taking the derivative of both sides of the last equation yields g(x) = f 0 (x)
Finally, we conclude this section with the following important result that
is useful in testing uniform convergence.
Theorem 14.3
Consider a sequence fn : D R. Then this sequence converges uniformly to
f : D R if and only if
lim sup{|fn (x) f (x)| : x D} = 0.
Proof.
Suppose that fn converges uniformly to f. Let > 0 be given. Then there
is a positive integer N such that |fn (x) f (x)| < 2 for all n N and all
x D. Thus, for n N, we have
sup{|fn (x) f (x)| : x D}

< .
2
This shows that

lim sup{|fn (x) f (x)| : x D} = 0.
Conversely, suppose that

lim sup{|fn (x) f (x)| : x D} = 0.
Let > 0 be given. Then there is a positive interger N such that

sup{|fn (x) f (x)| : x D} <
for all n N. But this implies that
|fn (x) f (x)| <
for all x D. Hence, fn converges uniformly to f in D
Example 14.9
Show that the sequence defined by fn (x) =
zero function.
cos x
n
converges uniformly to the

Solution.
We have
cos x
1
| : x R} .
n
n
8
Now apply the squeeze rule for sequences we find that
0 sup{|
lim sup{|
cos x
| : x R} = 0
n
which implies that the given sequence converges uniformly to the zero function on R
L.
If an bn cn for all n N and if limn an = limn cn = L then limn bn =
Practice Problems
Problem 14.1
Define fn : [0, 1] R by fn (x) = xn . Define f : [0, 1] R by

0 if 0 x < 1
f (x) =
1
if x = 1.
(a) Show that the sequence {fn }
n=1 converges pointwise to f.
(b) Show that the sequence {fn }
n=1 does not converge uniformly to f. Hint:
Suppose otherwise. Let = 0.5 and get a contradiction by using a point
1
(0.5) N < x < 1.
Problem 14.2
Consider the sequence of functions
fn (x) =
nx + x2
n2
defined for all x in R. Show that this sequence converges pointwise to a

function f to be determined.
Problem 14.3
Consider the sequence of functions
fn (x) =
sin (nx + 3)
n+1
defined for all x in R. Show that this sequence converges pointwise to a

function f to be determined.
Problem 14.4
Consider the sequence of functions defined by fn (x) = n2 xn for all 0 x 1.
Show that this sequence does not converge pointwise to any function.
Problem 14.5
Consider the sequence of functions defined by fn (x) = (cos x)n for all 2
x 2 . Show that this sequence converges pointwise to a noncontinuous
function to be determined.

Problem 14.6
n
Consider the sequence of functions fn (x) = x xn defined on [0, 1).
(a) Does {fn }
n=1 converge to some limit function? If so, find the limit function and show whether the convergence is pointwise or uniform.
(b) Does {fn0 }
n=1 converge to some limit function? If so, find the limit function and show whether the convergence is pointwise or uniform.
Problem 14.7
xn
Let fn (x) = 1+x
n for x [0, 2].
(a) Find the pointwise limit f (x) = limn fn (x) on [0, 2].
(b) Does fn f uniformly on [0, 2]?
Problem 14.8
For each n N define fn : R R by fn (x) =
(a) Show that fn 21 uniformly.
R7
(b) Find limn 2 fn (x)dx.
n+cos x
.
2n+sin2 x
Problem 14.9
Show that the sequence defined by fn (x) = (cos x)n does not converge uniformly on [ 2 , 2 ].
Problem 14.10
Let {fn }
n=1 be a sequence of functions such that
sup{|fn (x)| : 2 x 5}
2n
.
1 + 4n
(a) Show that this sequence converges uniformly

to a function f to be found.
R5
(b) What is the value of the limit limn 2 fn (x)dx?
15 An Introduction to Fourier Series

In this and the next section we will have a brief look to the subject of Fourier
series. The point here is to do just enough to allow us to do some basic solutions to partial differential equations later in the book.
Motivation: In Calculus we have seen that certain functions may be represented as power series by means of the Taylor expansions. These functions
must have infinitely many derivatives, and the series provide a good approximation only in some (often small) vicinity of a reference point.
Fourier series constructed of trigonometric rather than power functions, and
can be used for functions not only not differentiable, but even discontinuous
at some points. The main limitation of Fourier series is that the underlying
function should be periodic.
Recall from calculus that a function series is a series where the summands
are functions. Examples of function series include power series, Laurent series, Fourier series, etc.
Unlike series of numbers, there exist many types of convergence of series of
functions,
namely, pointwise, uniform, etc. We say that a series of functions
P
n=1 fn (x) converges pointwise to a function f if and only if the sequence
of partial sums
Sn (x) = f1 (x) + f2 (x) + + fn (x)
converges pointwise to f. We write
fn (x) = lim Sn (x) = f (x).
n=1
Example 15.1
P
n
Show that
n=0 x converges pointwise to a function to be determined for
all all 1 < x < 1.
Solution.
The nth term of the sequence of partial sums is given by
Sn (x) = 1 + x + x2 + + xn =
1 xn+1
.
1x
Since
lim xn+1 = 0 1 < x < 1,
15 AN INTRODUCTION TO FOURIER SERIES

the partial sums converge pointwise to the function
X
n=0
xn =
121
1
.
1x
Thus,
1
1x
P
Likewise, we say that a series of functions
n=1 fn (x) converges uniformly
to a function f if and only if the sequence of partial sums {Sn }
n=1 converges
uniformly to f.
The following theorem provide a tool for uniform convergence.
Theorem 15.1 (Weierstrass M-test)
P
Suppose that for each x in an interval I the series
n=1 fn (x) is well-defined.
Suppose further that
|fn (x)| Mn , x I.
P
P
If n=1 Mn (a scalar series) is convergent then the series
n=1 fn (x) is uniformly convergent.
Example 15.2
P
Show that
n=1
sin (nx)
n2
is uniformly convergent.
Solution.
For all x R, we have

sin (nx) | sin (nx)|
1

2.
n2
2
n
n
P
1
The series
n=1 n2 is convergent being a pseries with p = 2 > 1. Thus, by
Weierstrass M-test the given series is uniformly convergent
In this section we introduce a type of series of functions known as Fourier
series. They are given by
n
n i
a0 X h
f (x) =
+
an cos
x + bn sin
x , LxL
2
L
L
n=1
(15.1)
where an and bn are called the Fourier coefficients. The expression on the
right is called a trigonometric series. Note that we begin the series with
a0
as opposed to simply a0 to simplify the coefficient formula for an that we
2

will derive later in this section.
The main questions we want to consider next are the questions of determining which functions can be represented by Fourier series and if so how to
compute the coefficients an and bn .
Before answering these questions, we look at some of the properties of Fourier
series.
Periodicity Property
Recall that a function f is said to be periodic with period T > 0 if
f (x + T ) = f (x) for all x, x + T in the domain of f. The smallest value
of T for which f is periodic is called the fundamental period. A graph of
a T periodic function is shown in Figure 15.1.
Figure 15.1
For a T periodic function we have
f (x) = f (x + T ) = f (x + 2T ) = .
Note that the definite integral of a T periodic function is the same over any
interval of length T. By Problem 15.1 below, if f and g are two periodic functions with common period T, then the product f g and an arbitrary linear
combination c1 f + c2 g are also periodic with period T. It is an easy exercise
to show that the Fourier series (15.1) is periodic with fundamental period 2L.
Orthogonality Property
Recall from Calculus that for each pair of vectors ~u and ~v we associate a
scalar quantity ~u ~v called the dot product of ~u and ~v . We say that ~u and ~v
are orthogonal if and only if ~u ~v = 0. We want to define a similar concept
for functions.
Let f and g be two functions with domain the closed interval [a, b]. We define
123
a function that takes a pair of functions to a scalar. Symbolically, we write

Z b
f (x)g(x)dx.
< f, g >=
a
We call < f, g > the inner product of f and g. We say that f and g are
orthogonal if and only if < f, g >= 0. A set of functions is said to be mutually orthogonal if each distinct pair of functions in the set is orthogonal.
Before we proceed any further into computations, we would like to remind
the reader of the following two facts from calculus:
RL
If f (x) is an odd function defined on the interval [L, L] then L f (x)dx =
0.
RL
If f (x) is an even function defined on the interval [L, L] then L f (x)dx =
RL
2 0 f (x)dx.
Example 15.3

Show that the set 1, cos n
x , sin
L
ally orthogonal in [L, L].
n
x
L

: n N , where m 6= n, is mutu-
Solution.
Since the cosine function is even, we have
Z L
Z L
n
n
2L h n iL
x dx = 2
cos
x dx =
sin
x
= 0.
1 cos
L
L
n
L
0
0
L
Since the sine function is odd, we have
Z L
n
1 sin
x dx = 0.
L
L
Now, for n 6= m we have

Z L
Z
m
n
1 L
(m + n)
(m n)
cos
x cos
x dx =
cos
x + cos
x dx
L
L
2 L
L
L
L

1
L
(m + n)
=
sin
x
2 (m + n)
L

L
L
(m n)
+
sin
x
=0
(m n)
L
L

where we used the trigonometric identity
1
cos a cos b = [cos (a + b) + cos (a b)].
2
We can also show (see Problem 15.2):
Z
sin
L
and
Z
m
n
x sin
x dx = 0
L
L
n
m
x sin
x dx = 0
cos
L
L
L
The reason we care about these functions being orthogonal is because we will
exploit this fact to develop a formula for the coefficients in our Fourier series.
Now, in order to answer the first question mentioned earlier, that is, which
functions can be expressed as a Fourier series expansion, we need to introduce some mathematical concepts.
A function f (x) is said to be piecewise continuous on [a, b] if it is continuous in [a, b] except possibly at finitely many points of discontinuity within
the interval [a, b], and at each point of discontinuity, the right- and lefthanded limits of f exist. An example of a piecewise continuous function is
the function

x
0x<1
f (x) =
2
x x 1 x 2.
We will say that f is piecewise smooth in [a, b] if and only if f (x) as well
as its derivatives are piecewise continuous.
The following theorem, proven in more advanced books, ensures that a
Fourier decomposition can be found for any function which is piecewise
smooth.
Theorem 15.2
Let f be a 2L-periodic function. If f is a piecewise smooth on [L, L] then
for all points of discontinuity x [L, L] we have
n
n i
a0 X h
f (x ) + f (x+ )
=
+
an cos
x + bn sin
x
2
2
L
L
n=1
125
where as for points of continuity x [L, L] we have
n
n i
a0 X h
f (x) =
+
an cos
x + bn sin
x .
2
L
L
n=1
Remark 15.1
(1) Almost all functions occurring in practice are piecewise smooth functions.
(2) Given a piecewise smooth function f on [L, L]. The above theorem
applies to the periodic extension F of f where F (x + 2nL) = f (x) (n Z)
and F (x) = f (x) on (L, L). Note that if f (L) = f (L) then F (x) = f (x)
on [L, L]. Otherwise, the end points of f (x) may be jump discontinuities
of F (x).
Convergence Results of Fourier Series
We list few of the results regarding the convergence of Fourier series:
(1) The type of convergence in the above theorem is pointwise convergence.
(2) The convergence is uniform for a continuous function f on [L, L] such
that f (L) = f (L).
P
2
2
(3) The convergence is uniform whenever
n=1 (|an | + |bn | ) is convergent.
(4) If f (x) is periodic, continuous, and has a piecewise continuous derivative,
then the Fourier Series corresponding to f converges uniformly to f (x) for
the entire real line.
(5) The convergence is uniform on any closed interval that does not contain
a point of discontinuity.
Euler-Fourier Formulas
Next, we will answer the second question mentioned earlier, that is, the question of finding formulas for the coefficients an and bn . These formulas for an
and bn are called Euler-Fourier formulas which we derive next. We will assume that the RHS in (15.1) converges uniformly to f (x) on the interval
[L, L]. Integrating both sides of (15.1) we obtain
Z LX
Z L
Z L
h
n
n i
a0
dx +
an cos
x + bn sin
x dx.
f (x)dx =
L
L
L n=1
L
L 2
Since the trigonometric series is assumed to be uniformly convergent, from
Theorem 14.2, we can interchange the order of integration and summation
to obtain
Z L
Z L
Z L h
n
n i
X
a0
f (x)dx =
dx +
an cos
x + bn sin
x dx.
L
L
L
L 2
n=1 L

But
Z
n
n iL
L
x dx =
sin
x
=0
L
n
L
L
cos
L
and likewise
Z
sin
L
n iL
n
L
x dx =
cos
x
= 0.
L
n
L
L
Thus,
Z
1 L
f (x)dx.
a0 =
L L
To find the other Fourier coefficients, we recall the results of Problems 15.2
- 15.3 below.

Z L
n
m
L if m = n
cos
x cos
x dx =
0 if m 6= n
L
L
L

Z L
n
m
L if m = n
sin
x sin
x dx =
0 if m 6= n
L
L
L
Z L
n
m
sin
x cos
x dx = 0, m, n.
L
L
L
Now, to find the formula for the Fourier
coefficients am for m > 0, we multiply

both sides of (15.1) by cos m
x
and
integrate
from L to L to otbain
L
Z L
Z L

m Z L a
m
n
m
X
0
f (x) cos
x =
cos
x dx +
an
cos
x cos
x dx
L
L
L
L
L 2
L
L
n=1
Z L
n
m
+ bn
sin
x cos
x dx .
L
L
L
Hence,
Z
f (x) cos
L
and therefore
m
x dx = am L
L
1
am =
L
Likewise, we can show that
1
bm =
L
f (x) cos
m
x dx.
L
f (x) sin
m
x dx.
L

Example 15.4
Find the Fourier series expansion of

0, x 0
f (x) =
x, x > 0
on the interval [, ].
Solution.
We have
Z
Z
1
1
f (x)dx =
xdx =
a0 =

0
2

Z
1
1 x sin nx cos nx
(1)n 1
an =
x cos nxdx =
+
=
0
n
n2 0
n2

Z
1
x cos nx sin nx
(1)n+1
1
x sin nxdx =
+
=
.
bn =
0
n
n2 0
n
Hence,

X (1)n 1
(1)n+1
f (x) = +
cos (nx) +
sin (nx) < x <
4 n=1
n2
n
Example 15.5
Apply Theorem 15.2 to the function in Example 15.4.
Solution.
Let F be a periodic extension of f of period 2. See Figure 152.
Figure 152
127

Thus, f (x) = F (x) on the interval (, ). Note that for x = , the Fourier
series coverges to
F ( ) + F ( + )
= .
2
2
Similar result for x = . Clearly, F is a piecewise smooth function so that
by the previous thereom we can write
,

if x =
n
n+1
X
2
(1) 1
(1)
f
(x),
if
<x<
+
sin
(nx)
=
cos
(nx)
+

4 n=1
n2
n
,
if x = .
2
Taking x = we have the identity
X (1)n 1
+
(1)n =
2
4 n=1
n
2
which can be simplified to
X
n=1
1
2
=
.
(2n 1)2
8
This provides a method for computing an approximate value of

Remark 15.2
An example of a function that does not have a Fourier series representation
is the function f (x) = x12 on [L, L]. For example, the coefficient a0 for this
function does not exist. Thus, not every function can be written as a Fourier
series expansion.
The final topic of discussion here is the topic of differentiation and integration
of Fourier series. In particular we want to know if we can differentiate a
Fourier series term by term and have the result be the Fourier series of the
derivative of the function. Likewise we want to know if we can integrate a
Fourier series term by term and arrive at the Fourier series of the integral of
the function. Answers to these questions are provided next.
Theorem 15.3
A Fourier series of a piecewise smooth function f can always be integrated
term by term
R L and the result is a convergent infinite series that always converges to L f (x)dx even if the original series has jumps.
129
Theorem 15.4
A Fourier series of a continuous function f (x) can be differentiated term by
term if f 0 (x) is piecewise smooth. The result of the differentiation is the
Fourier series of f 0 (x).
Practice Problems
Problem 15.1
Let f and g be two functions with common domain D and common period
T. Show that
(a) f g is periodic of period T.
(b) c1 f + c2 g is periodic of period T, where c1 and c2 are real numbers.
Problem 15.2
Show
have

R Lthat for m6= n we
n
x
sin
x
dx = 0 and
(a) L sin m
L
L

RL
(b) L cos m
x sin n
x dx = 0.
L
L
Problem 15.3
Compute
the following
integrals:

RL
2 n
(a) L cos L x dx.

RL
(b) L sin2 n
x dx.
L

RL
n
(c) L cos n
x
sin
x
dx.
L
L
Problem 15.4
Find the Fourier coefficients of
, x < 0
,
0<x<
f (x) =
0,
x = 0,
on the interval [, ].
Problem 15.5
Find the Fourier series of f (x) = x2
1
2
on the interval [1, 1].
Problem 15.6
Find the Fourier series of the function
1, 2 < x <
0,
< x <
f (x) =
1,
< x < 2.
131
Problem 15.7
Find the Fourier series of the function

1 + x, 2 x 0
f (x) =
1 x, 0 < x 2.
Problem 15.8
Show that f (x) =
1
x
is not piecewise continuous on [1, 1].
Problem 15.9
Assume that f (x) is continuous and has period 2L. Prove that
Z L+a
Z L
f (x)dx
f (x)dx =
L
L+a
is independent of a R. In particular, it does not matter over which interval

the Fourier coefficients are computed as long as the interval length is 2L.
[Remark: This result is also true for piecewise continuous functions].
Problem 15.10
Consider the function f (x) defined by

1 0x<1
f (x) =
2 1x<3
and extended periodically with period 3 to R so that f (x + 3) = f (x) for all
x.
(i) Find the Fourier series of f (x).
(ii) Discuss its limit: In particular, does the Fourier series converge pointwise
or uniformly to its limit, and what is this limit?
(iii) Plot the graphs of f (x) and its extension F (x) on the interval [0, 3].
Problem 15.11
For the following functions f (x) on the interval L < x < L, determine the
coefficients an , n = 0, 1, 2, and bn , n N of the Fourier series expansion.
(a) f (x) = 1.

.
(b) f (x) =
2 + sin x
L
1 x0
(c) f (x) =
0 x > 0.
(d) f (x) = x.

Problem 15.12
Let f (t) be the function with period 2 defined as

2 if 0 x 2
f (t) =
0 if 2 < x 2
f (t) has a Fourier series and that series is equal to
a0 X
+
(an cos nt + bn sin nt), 0 < x < 2.
2
n=1
Find
a3
.
b3
Problem 15.13
Let f (x) = x3 on [, ], extended periodically to all of R. Find the Fourier
coefficients an , n = 1, 2, 3, .
Problem 15.14
Let f (x) be the square wave function

x < 0
f (x) =
0x
extended periodically to all of R. To what value does the Fourier series of
f (x) converge when x = 0?
Problem 15.15
(a) Find the Fourier series of

f (x) =
1 x < 0
2 0x
extended periodically to all of R. Simplify your coefficients as much as possible.

P
(1)n+1
(b) Use (a) to evaluate the series
n=1 (2n1) . Hint: Evaluate the Fourier
series at x = 2 .
16 FOURIER SINES SERIES AND FOURIER COSINES SERIES
133
16 Fourier Sines Series and Fourier Cosines Series

In this section we discuss some important properties of Fourier series when
the underlying function f is either even or odd.
A function f is odd if it satisfies f (x) = f (x) for all x in the domain of
f whereas f is even if it satisfies f (x) = f (x) for all x in the domain of f.
Example 16.1
Show the following
(a) If f and g are either both even or both odd then f g is even.
(b) If f is odd and g is even then f g is odd.
Solution.
(a) Suppose that both f and g are even. Then (f g)(x) = f (x)g(x) =
f (x)g(x) = (f g)(x). That is, f g is even. Now, suppose that both f and g
are odd. Then (f g)(x) = f (x)g(x) = [f (x)][g(x)] = (f g)(x). That
is, f g is even.
(b) f is odd and g is even. Then (f g)(x) = f (x)g(x) = f (x)g(x) =
(f g)(x). That is, f g is odd
Example 16.2
(a) Show that for any even function f (x) defined on the interval [L, L], we
have
Z L
Z L
f (x)dx = 2
f (x)dx.
L
(b) Show that for any odd function f (x) defined on the interval [L, L], we
have
Z L
f (x)dx = 0.
L
Solution.
(a) Since f (x) is even, we have f (x) = f (x) for all x in [L, L]. Thus,
Z 0
Z 0
Z 0
Z L
f (x)dx =
f (x)dx =
f (u)du =
f (x)dx.
L
For this, it follows that

Z L
Z
f (x)dx =
L
Z
f (x)dx +
Z
f (x)dx = 2
f (x)dx.
0

(b) Since f (x) is odd, we have f (x) = f (x) for all x in [L, L]. Thus,
Z L
Z 0
Z 0
Z 0
f (x)dx.
f (u)du =
[f (x)]dx =
f (x)dx =
L
Hence,
Z
f (x)dx
f (x)dx =
f (x)dx +
0=
Example 16.3
RL
(a) Find the value of the integral L f (x) sin
RL
(b) Find the value of the integral L f (x) cos
n
x dx when f
L

n
x
dx when f
L
is even.
is odd.
Solution.

x is odd and f is even, we have that f (x) sin n
x
(a) Since the function sin n
L
L
is odd so that
Z L
n
f (x) sin
x dx = 0
L
L

n
x
is
even
and
f
is
odd,
we
have
that
f
(x)
cos
x
(b) Since the function cos n
L
L
is odd so that
Z L
n
f (x) cos
x dx = 0
L
L
Even and Odd Extensions
Let f : (0, L) R be a piecewise smooth function. We define the odd
extension of this function on the interval L < x < L by

f (x)
0<x<L
fodd (x) =
f (x) L < x < 0.
This function will be odd on the interval (L, L), and will be equal to f (x)
on the interval (0, L). We can then further extend this function to the entire
real line by defining it to be 2L periodic. Let f odd denote this extension. We
note that f odd is an odd function and piecewise smooth so that by Theorem
15.2 it possesses a Fourier series expansion, and from the fact that it is odd
all of the a0n s are zero. Moreover, in the interval (0, L) we have
n
f (x) =
bn sin
x .
L
n=1
(16.1)
135
We call (16.1) the Fourier sine series of f.

The coefficients bn are given by the formula
Z
Z
n
n
2 L
1 L
f sin
x dx =
f sin
x dx
bn =
L L odd
L
L 0 odd
L
Z
n
2 L
=
f (x) sin
x dx
L 0
L

since f odd sin n
x is an even function.
L
Likewise, we can define the even extension of f on the interval L < x < L
by

f (x)
0<x<L
feven (x) =
f (x) L < x < 0.
We can then further extend this function to the entire real line by defining
it to be 2L periodic. Let f even denote this extension. Again, we note that
f even is equal to the original function f (x) on the interval upon which f (x)
is defined. Since f even is piecewise smooth, by Theorem 15.2 it possesses a
Fourier series expansion, and from the fact that it is even all of the b0n s are
zero. Moreover, in the interval (0, L) we have
f (x) =
n
a0 X
+
an cos
x .
2
L
n=1
(16.2)
We call (16.2) the Fourier cosine series of f. The coefficients an are given
by
Z
n
2 L
f (x) cos
x dx, n = 0, 1, 2, .
an =
L 0
L
Example 16.4
Graph the odd and even extensions of the function f (x) = x, 0 x 1.
Solution.
We have fodd (x) = x for 1 x 1. The odd extension of f is shown in
Figure 16.1(a). Likewise,

x
0x1
feven (x) =
x 1 x < 0.

The even extension is shown in Figure 16.1(b)
Figure 16.1
Example 16.5
Find the Fourier sine series of the function

f (x) =
0 x 2
x .
2
x,
x,
Solution.
We have
2
bn =
"Z
( x) sin nxdx .
x sin nxdx +
0
Using integration by parts we find
Z
0
Z
h x
i 2
1 2
x sin nxdx = cos nx +
cos nxdx
n
n 0
0
cos (n/2)
1
=
+ 2 [sin nx]02
2n
n
cos (n/2) sin (n/2)
=
+
2n
n2

while
Z
( x)
( x) sin nxdx =
cos nx
n
cos nxdx
cos (n/2)
1
2 [sin nx]
2
2n
n
cos (n/2) sin (n/2)
=
+
2n
n2
=
Thus,
4 sin (n/2)
,
n2
and the Fourier sine series of f (x) is
bn =
f (x) =
X
4 sin (n/2)
n=1
n2
X
4(1)n+1
sin nx =
sin (2n 1)x
(2n 1)2
n=1
137
Practice Problems
Problem 16.1
Give an example of a function that is both even and odd.
Problem 16.2
Graph the odd and even extensions of the function f (x) = 1, 0 x 1.
Problem 16.3
Graph the odd and even extensions of the function f (x) = L x for 0 x
L.
Problem 16.4
Graph the odd and even extensions of the function f (x) = 1 + x2 for 0
x L.
Problem 16.5
Find the Fourier cosine series of the function

x,
0 x 2
f (x) =
x, 2 x .
Problem 16.6
Find the Fourier cosine series of f (x) = x on the interval [0, ].
Problem 16.7
Find the Fourier sine series of f (x) = 1 on the interval [0, ].
Problem 16.8
Find the Fourier sine series of f (x) = cos x on the interval [0, ].
Problem 16.9
Find the Fourier cosine series of f (x) = e2x on the interval [0, 1].
Problem 16.10
For the following functions on the interval [0, L], find the coefficients bn of
the Fourier sine expansion.

(a) f (x) = sin 2
x
.
L
(b) f (x) = 1

(c) f (x) = cos L x .
139
Problem 16.11
For the following functions on the interval [0, L], find the coefficients an of
the Fourier cosine expansion.

(a) f (x) = 5 + cos L x .
(b) f (x) = x
(c)

1 0 < x L2
f (x) =
0 L2 < x L.
Problem 16.12
Consider a function f (x), defined on 0 x L, which is even (symmetric)
around x = L2 . Show that the even coefficients (n even) of the Fourier sine
series are zero.
Problem 16.13
Consider a function f (x), defined on 0 x L, which is odd around x = L2 .
Show that the even coefficients (n even) of the Fourier cosine series are zero.
Problem 16.14
The Fourier sine series of f (x) = cos
cos
x
L
bn sin
n=1
where
b1 = 0, bn =
x
L
(n2
for 0 x L is given by
nx
L
, nN
2n
[1 + (1)n ].
1)
Using term-by-term integration, find the Fourier cosine series of sin
x
L

.
Problem 16.15
Consider the function

f (x) =
1 0x<1
2 1 x < 2.
(a) Sketch the even extension of f.

(b) Find a0 in the Fourier series for the even extension of f.
(c) Find an (n = 1, 2, ) in the Fourier series for the even extension of f.
(d) Find bn in the Fourier series for the even extension of f.
(e) Write the Fourier series for the even extension of f.
17 Separation of Variables for PDEs

Finding analytic solutions to PDEs is essentially impossible. Most of the
PDE techniques involve a mixture of analytic, qualitative and numeric approaches. Of course, there are some easy PDEs too. If you are lucky your
PDE has a solution with separable variables. In this chapter we discuss the
application of the method of separation of variables in the solution of PDEs.
17.1 Second Order Linear Homogenous ODE with Constant Coefficients

In this section, we review the basics of finding the general solution to the
ODE
ay 00 + by 0 + cy = 0
(17.1)
where a, b, and c are constants. The process starts by solving the characteristic equation
ar2 + br + c = 0
which is a quadratic equation with roots
b b2 4ac
.
r1,2 =
2a
We consider the following three cases:
If b2 4ac > 0 then the general solution to (17.1) is given by

y(t) = Ae
b2 4ac
2a

t
b+
+ Be
b2 4ac
2a

t
If b2 4ac = 0 then the general solution to (17.1) is given by

b
y(t) = Ae 2a t + Bte 2a t .
If b2 4ac < 0 then
b
4ac b2
r1,2 = i
2a
2a
and the general solution to (17.1) is given by

b
b
4ac b2
4ac b2
2a
t
2a
t
y(t) = Ae
cos
t + Be
sin
t.
2a
2a
17 SEPARATION OF VARIABLES FOR PDES
141
17.2 The Method of Separation of Variables for PDEs

In developing a solution to a partial differential equation by separation of
variables, one assumes that it is possible to separate the contributions of
the independent variables into separate functions that each involve only one
independent variable. Thus, the method consists of the following steps
1. Factorize the (unknown) dependent variable of the PDE into a product of
functions, each of the factors being a function of one independent variable.
That is,
u(x, y) = X(x)Y (y).
2. Substitute into the PDE, and divide the resulting equation by X(x)Y (y).
3. Then the problem turns into a set of separated ODEs (one for X(x) and
one for Y (y).)
4. The general solution of the ODEs is found, and boundary initial conditions are imposed.
5. u(x, y) is formed by multiplying together X(x) and Y (y).
We illustrate these steps in the next three examples.
Example 17.1
Find all the solutions of the form u(x, t) = X(x)T (t) of the equation
uxx ux = ut .
Solution.
It is very easy to find the derivatives of a separable function:
ux = X 0 (x)T (t), ut = X(x)T 0 (t) and uxx = X 00 (x)T (t),
which is basically a consequence of the fact that differentiation with respect
to x sees t as a constant, and vice versa. Now, the equation uxx ux = ut
becomes
X 00 (x)T (t) X 0 (x)T (t) = X(x)T 0 (t).
We can separate variables further. Division by X(x)T (t) gives
T 0 (t)
X 00 (x) X 0 (x)
=
.
X(x)
T (t)

The expression on the LHS is a function of x whereas the one on the RHS is
a function of t only. They both have to be constant. That is,
X 00 (x) X 0 (x)
T 0 (t)
=
= .
X(x)
T (t)
Thus, we have the following ODEs:
X 00 X 0 X = 0 and T 0 = T.
The second equation is easy to solve: T (t) = Cet . The first equation is
solved via the characteristic equation 2 = 0, whose solutions are
1 1 + 4
=
.
2
If > 14 then
X(x) = Ae
In this case,
u(x, t) = De
If = 41 then
1+ 1+4
x
2
1+ 1+4
x
2
+ Be
1 1+4
x
2
e + Ee
x
1 1+4
x
2
et .
X(x) = Ae 2 + Bxe 2
and in this case
u(x, t) = (D + Ex)e 2 4 .
If < 14 then
x
2
X(x) = Ae cos
!
p
(1 + 4)
x
x + Be 2 sin
2
!
p
(1 + 4)
x .
2
In this case,
u(x, t) = D0 e
x
+t
2
cos
!
p
(1 + 4)
x
x + B 0 e 2 +t sin
2
(1 + 4)
x
2
Example 17.2
Solve Laplaces equation using the separation of variables method
u = uxx + uyy = 0.
143
Solution.
We look for a solution of the form u(x, y) = X(x)Y (y). Substituting in the
Laplaces equation, we obtain
X 00 (x)Y (y) + X(x)Y 00 (y) = 0.
Assuming X(x)Y (y) is nonzero, dividing for X(x)Y (y) and subtracting
from both sides, we find:
Y 00 (y)
Y (y)
X 00 (x)
Y 00 (y)
=
.
X(x)
Y (y)
The left hand side is a function of x while the right hand side is a function
of y. This says that they must equal to a constant. That is,
Y 00 (y)
X 00 (x)
=
=
X(x)
Y (y)
where is a constant. This results in the following two ODEs
X 00 X = 0 and Y 00 + Y = 0.
The solutions of these equations depend on the sign of .
If > 0 then the solutions are given
+ Be x
Y (y) =C cos y + D sin y
X(x) =Ae
where A, B, C, and D are constants. In this case,
u(x, t) =k1 e x cos y + k2 e x sin y
+k3 e x cos y + k4 e x sin y.

If = 0 then
X(x) =Ax + B
Y (y) =Cy + D
where A, B, and C are arbitrary constants. In this case,
u(x, y) = k1 xy + k2 x + k3 y + k4 .

If < 0 then
X(x) =A cos x + B sin x
Y (y) =Ce
+ De
where A, B, C, and D are arbitrary constants. In this case,
u(x, y) =k1 cos xe y + k2 cos xe y
+k3 sin xe y + k4 sin xe y

Example 17.3
Solve using the separation of variables method.
yux xuy = 0.
Solution.
Substitute u(x, t) = X(x)Y (y) into the given equation we find
yX 0 Y xXY 0 = 0.
This can be separated into
Y0
X0
=
.
xX
yY
X0
Y0
=
=
xX
yY
X 0 xX = 0 and Y 0 yY = 0.
Solving these equations using the method of separation of variable for ODEs
y 2
x2
we find X(x) = Ae 2 and Y (y) = Be 2 . Thus, the general solution is given
by
u(x, y) = Ce
(x2 +y 2 )
2
145
Practice Problems
Problem 17.1
Solve using the separation of variables method
u + u = 0.
Problem 17.2
Solve using the separation of variables method
ut = kuxx .
Problem 17.3
Derive the system of ordinary differential equations for R(r) and () that
is satisfied by solutions to
1
1
urr + ur + 2 u = 0.
r
r
Problem 17.4
Derive the system of ordinary differential equations and boundary conditions
for X(x) and T (t) that is satisfied by solutions to
utt = uxx 2u, 0 < x < 1, t > 0
u(0, t) = 0 = ux (1, t) t > 0
of the form u(x, t) = X(x)T (t). (Note: you do not need to solve for X and
T .)
Problem 17.5
Derive the system of ordinary differential equations and boundary conditions
for X(x) and T (t) that is satisfied by solutions to
ut = kuxx , 0 < x < L, t > 0
u(x, 0) = f (x), u(0, t) = 0 = ux (L, t) t > 0
of the form u(x, t) = X(x)T (t). (Note: you do not need to solve for X and
T .)
Problem 17.6
Find all product solutions of the PDE ux + ut = 0.

Problem 17.7
Derive the system of ordinary differential equations for X(x) and Y (y) that
is satisfied by solutions to
3uyy 5uxxxy + 7uxxy = 0.
of the form u(x, y) = X(x)Y (y).
Problem 17.8
Find the general solution by the method of separation of variables.
uxy + u = 0.
Problem 17.9
ux yuy = 0.
Problem 17.10
utt uxx = 0.
Problem 17.11
For the following PDEs find the ODEs implied by the method of separation
of variables.
(a) ut = kr(rur )r
(b) ut = kuxx u
(c) ut = kuxx aux
(d) uxx + uyy = 0
(e) ut = kuxxxx .
Problem 17.12
Find all solutions to the following partial differential equation that can be
obtained via the separation of variables.
ux uy = 0.
Problem 17.13
Separate the PDE uxx uy + uyy = u into two ODEs with a parameter. You
do not need to solve the ODEs.
18 SOLUTIONS OF THE HEAT EQUATION BY THE SEPARATION OF VARIABLES METHOD147
18 Solutions of the Heat Equation by the Separation of Variables Method

In this section we apply the method of separation of variables in solving the
one spatial dimension of the heat equation.
The Heat Equation with Dirichlet Boundary Conditions
Consider the problem of finding all nontrivial solutions to the heat equation
ut = kuxx that satisfies the initial time condition u(x, 0) = f (x) and the
Dirichlet boundary conditions u(0, t) = T0 and u(L, t) = TL .
From Exercise 13.4, it suffices to solve the problem with the Dirichlet boundary conditions being replaced by the homogeneous boundary conditions u(0, t) =
u(L, t) = 0 (that is, the endpoints are assumed to be at zero temperature)
with u not the trivial solution. Lets assume that the solution can be written in the form u(x, t) = X(x)T (t). Substituting into the heat equation we
obtain
T0
X 00
=
.
X
kT
Since the LHS only depends on x and the RHS only depends on t, there must
be a constant such that
X 00
X
= and
T0
kT
= .
This gives the two ordinary differential equations

X 00 X = 0 and T 0 kT = 0.
As far as the boundary conditions, we have
u(0, t) = 0 = X(0)T (t) = X(0) = 0
and
u(L, t) = 0 = X(L)T (t) = X(L) = 0.
Note that T is not the zero function for otherwise u 0 and this contradicts
our assumption that u is the non-trivial solution.
Next, we consider the three cases of the sign of .

Case 1: = 0
In this case, X 00 = 0. Solving this equation we find X(x) = ax + b. Since
X(0) = 0 we find b = 0. Since X(L) = 0 we find a = 0. Hence, X 0 and
u(x, t) 0. That is, u is the trivial solution.
Case 2: > 0
In this case, X(x) = Ae x + Be x . Again, the conditions X(0) = X(L) =

0 imply A = B = 0 and hence the solution is the trivial solution.
Case 3: < 0
In this case, X(x) = A cos x + B sin x. Thecondition X(0) = 0

implies A = 0. The condition X(L) = 0 implies B sin L = 0. We must
have B 6= 0 otherwiseX(x) = 0 andthis leads to the trivial solution. Since
B 6= 0, we obtain sin L = 0 or L = n where n Z. Solving for
2 2
we find = nL2 . Thus, we obtain infinitely many solutions given by
Xn (x) = An sin
n
x, n N.
L
Now, solving the equation

T 0 kT = 0
by the method of separation of variables we obtain
Tn (t) = Bn e
n2 2
kt
L2
, n N.
Hence, the functions

n n2 2
un (x, t) = Cn sin
x e L2 kt , n N
L
satisfy ut = kuxx and the boundary conditions u(0, t) = u(L, t) = 0.
Now, in order for these solutions to satisfy the initial value condition u(x, 0) =
f (x), we invoke the superposition principle of linear PDE to write
u(x, t) =
Cn sin
n=1
n n2 2
x e L2 kt .
L
(18.1)
To determine the unknown constants Cn we use the initial condition u(x, 0) =

f (x) in (18.1) to obtain
n
f (x) =
Cn sin
x .
L
n=1

Since the right-hand side is the Fourier sine series of f on the interval [0, L],
the coefficients Cn are given by
Z
n
2 L
Cn =
x dx.
(18.2)
f (x) sin
L 0
L
Thus, the solution to the heat equation is given by (18.1) with the Cn0 s
calculated from (18.2).
Remark 18.1
According to Exercise 13.4, the solution to the heat equation with nonhomogeneous condition u(0, t) = T0 and u(L, t) = TL is given by
n n2 2
TL T0
Cn sin
x e L2 kt + T0 +
x.
u(x, t) =
L
L
n=1
The Heat Equation with Neumann Boundary Conditions
When both ends of the bar are insulated, that is, there is no heat flow out
of them, we use the boundary conditions
ux (0, t) = ux (L, t) = 0.
In this case, the general form of the heat equation initial boundary value
problem is to find u(x, t) satisfying
ut (x, t) =kuxx (x, t), 0 x L, t > 0
u(x, 0) =f (x), 0 x L
ux (0, t) =ux (L, t) = 0, t > 0.
Since 0 = ux (0, t) = X 0 (0)T (t) we obtain X 0 (0) = 0. Likewise, 0 = ux (L, t) =
X 0 (L)T (t) implies X 0 (L) = 0. We again consider the following three cases:
If = 0 then X(x) = A + Bx. Since X 0 (0) = 0, we find B = 0. Thus,
X(x) = A and T (t) = constant so that u(x, t) = constant which is impossible
if f (x) is not the constant function.
If > 0 then a simple calculation shows that u(x, t) is the trivial solution.
Again, because of the condition
this solution is discarded.
u(x, 0) = f (x),
If < 0 then X(x) = A cos x + B sin x and upon differentiation
with respect to x we find
X 0 (x) = A sin x + B cos x.
The conditions X 0 (0) = X 0 (L) = 0 imply B = 0 and A sin L =

2 2
0. Hence, B = 0 and = nL2 and
n
Xn (x) = An cos
x , n = 0, 1, 2,
L
and
n n2 2
x e L2 kt .
L
By the superposition principle, the required solution to the heat equation
with Neumann boundary conditions is given by
un (x, t) = Cn cos
u(x, t) =
where
2
Cn =
L
n n2 2
C0 X
Cn cos
+
x e L2 kt
2
L
n=1
f (x) cos
0
n
x dx, n = 0, 1, 2, .
L
Practice Problems
Problem 18.1
Find the temperature in a bar of length 2 whose ends are kept at zero

and lateral
surface insulated if the initial temperature is f (x) = sin 2 x +

x .
3 sin 5
2
Problem 18.2
Find the temperature in a homogeneous bar of heat conducting material of
length L with its end points kept at zero and initial temperature distribution
given by f (x) = Lxd2 (L x), 0 x L.
Problem 18.3
Find the temperature in a thin metal rod of length L, with both ends insulated (so that there is no passage of heat
through the ends) and with initial
temperature in the rod f (x) = sin L x .

Problem 18.4
Solve the following heat equation with Dirichlet boundary conditions
ut = kuxx
u(0, t) = u(L, t) = 0

1 0 x < L2
u(x, 0) =
2 L2 x L.
Problem 18.5
Solve
ut = kuxx
u(0, t) = u(L, t) = 0

9
u(x, 0) = 6 sin
x .
L
Problem 18.6
Solve
ut = kuxx
subject to
ux (0, t) = ux (L, t) = 0

0 0 x < L2
u(x, 0) =
1 L2 x L.

Problem 18.7
Solve
ut = kuxx
subject to
ux (0, t) = ux (L, t) = 0

3
x .
u(x, 0) = 6 + 4 cos
L
Problem 18.8
Solve
ut = kuxx
subject to
ux (0, t) = ux (L, t) = 0

8
u(x, 0) = 3 cos
x .
L
Problem 18.9
Find the general solution u(x, t) of
ut = uxx u, 0 < x < L, t > 0
ux (0, t) = 0 = ux (L, t), t > 0.
Briefly describe its behavior as t .
Problem 18.10 (Energy method)
Let u1 and u2 be two solutions to the Neumann boundary value problem
ut = uxx u, 0 < x < 1, t > 0
ux (0, t) = ux (1, t) = 0, t > 0
u(x, 0) = g(x), 0 < x < 1
Define w(x, t) = u1 (x, t) u2 (x, t).
(a) Show that w satisfies the initial value problem
wt = wxx w, 0 < x < 1, t > 0
wx (0, t) = wx (1, t) = w(x, 0) = 0, 0 < x < 1, t > 0
R1
(b) Define E(t) = 0 w2 (x, t)dx 0 for all t 0. Show that E 0 (t) 0.
Hence, 0 E(t) E(0) for all t > 0.
(c) Show that E(t) = 0, w(x, t) = 0. Hence, conclude that u1 = u2 .

Problem 18.11
Consider the heat induction in a bar where the left end temperature is maintained at 0, and the right end is perfectly insulated. We assume k = 1 and
L = 1.
(a) Derive the boundary conditions of the temperature at the endpoints.
(b) Following the separation of variables approach, derive the ODEs for X
and T.
(c) Consider the equation in X(x). What arethe values of X(0) and X 0 (1)?
Show that solutions of the form X(x) = sin x satisfy the ODE and one
of the boundary conditions. Can you choose a value of so that the other
boundary condition is also satisfied?
Problem 18.12
Using the method of separation of variables find the solution of the heat
equation
ut = kuxx
satisfying the following boundary and initial conditions:
(a) u(0, t) = u(L, t) = 0, u(x, 0) = 6 sin 9x
L

3x
(b) u(0, t) = u(L, t) = 0, u(x, 0) = 3 sin x
sin
L
L
Problem 18.13
Using the method of separation of variables find the solution of the heat
equation
ut = kuxx
satisfying the following boundary and initial conditions:
+ 4 cos 5x
.
(a) ux (0, t) = ux (L, t) = 0, u(x, 0) = cos x
L
L
(b) ux (0, t) = ux (L, t) = 0, u(x, 0) = 5.
Problem 18.14
Find the solution of the following heat conduction partial differential equation
ut = 8uxx , 0 < x < 4, t > 0
u(0, t) = u(4, t) = 0, t > 0
u(x, 0) = 6 sin x, 0 < x < 4.
19 Elliptic Type: Laplaces Equations in Rectangular Domains

Boundary value problems are of great importance in physical applications.
Mathematically, a boundary-value problem consists of finding a function
which satisfies a given partial differential equation and particular boundary conditions. Physically speaking, the problem is independent of time,
involving only space coordinates.
Just as initial-value problems are associated with hyperbolic PDE, boundary value problems are associated with PDE of elliptic type. In contrast to
initial-value problems, boundary-value problems are considerably more difficult to solve.
The main model example of an elliptic type PDE is the Laplace equation
u = uxx + uyy = 0
(19.1)
where the symbol is referred to as the Laplacian. Solutions of this equation are called harmonic functions.
Example 19.1
Show that, for all (x, y) 6= (0, 0), u(x, y) = ax2 ay 2 + cx + dy + e is a
harmonic function, where a, b, c, d, and e are constants.
Solution.
We have
ux
uxx
uy
uyy
=2ax + c
=2a
= 2ay + d
= 2a.
Plugging these expressions into the equation we find uxx + uyy = 0. Hence,
u(x, y) is harmonic
The Laplace equation is arguably the most important differential equation in
all of applied mathematics. It arises in an astonishing variety of mathematical and physical systems, ranging through fluid mechanics, electromagnetism,
potential theory, solid mechanics, heat conduction, geometry, probability,
19 ELLIPTIC TYPE: LAPLACES EQUATIONS IN RECTANGULAR DOMAINS155

number theory, and on and on.
There are two main modifications of the Laplace equation: the Poisson
equation (a non-homogeneous Laplace equation):
u = f (x, y)
and the eigenvalue problem (the Helmholtz equation):
u = u, R.
Solving Laplaces Equation (19.1)
Note first that both independent variables are spatial variables and each
variable occurs in a 2nd order derivative and so we will need two boundary
conditions for each variable a total of four boundary conditions.
Consider (19.1) in the rectangle
= {(x, y) : 0 x a, 0 y b}
with the Dirichlet boundary conditions
u(0, y) = f1 (y), u(a, y) = f2 (y), u(x, 0) = g1 (x), u(x, b) = g2 (x)
where 0 x a and 0 y b.
The separation of variables method is most successful when the boundary
conditions are homogeneous. Thus, solving the Laplaces equation in requires solving four initial boundary conditions problems, where in each problem three of the four conditions are homogeneous. The four problems to be
solved are
uxx + uyy = 0
uxx + uyy = 0
u(0, y) = f1 (y),
u(a, y) = f2 (y),
(I)
(II)
u(a, y) = u(x, 0) = u(x, b) = 0
u(0, y) = u(x, 0) = u(x, b) = 0
uxx + uyy = 0
u(x, 0) = g1 (x),
(III)
u(0, y) = u(a, y) = u(x, b) = 0
(IV )
uxx + uyy = 0
u(x, b) = g2 (x),
u(0, y) = u(a, y) = u(x, 0) = 0.
If we let ui (x, y), i = 1, 2, 3, 4, denote the solution of each of the above

problems, then the solution to our original system will be
u(x, y) = u1 (x, y) + u2 (x, y) + u3 (x, y) + u4 (x, y).

In each of the above problems, we will apply separation of variables to (19.1)
and find a product solution that will satisfy the differential equation and the
three homogeneous boundary conditions. Using the Principle of Superposition we will find a solution to the problem and then apply the final boundary
condition to determine the value of the constant(s) that are left in the problem. The process is nearly identical in many ways to what we did when we
were solving the heat equation.
We will illustrate how to find u(x, y) = u4 (x, y). So lets assume that the solution can be written in the form u(x, y) = X(x)Y (y). Substituting in (19.1),
we obtain
X 00 (x)Y (y) + X(x)Y 00 (y) = 0.
Assuming X(x)Y (y) is nonzero, that is u is the non-trivial solution. Dividing
00 (y)
for X(x)Y (y) and subtracting YY (y)
X 00 (x)
Y 00 (y)
=
.
X(x)
Y (y)
X 00 (x)
Y 00 (y)
=
=
X(x)
Y (y)
X 00 X = 0 and Y 00 + Y = 0.
As far as the boundary conditions, we have for all 0 x a and 0 y b
u(0, y) = 0 = X(0)Y (y) = X(0) = 0
u(a, y) = 0 = X(a)Y (y) = X(a) = 0
u(x, 0) = 0 = X(x)Y (0) = Y (0) = 0
u(x, b) = g2 (x) = X(x)Y (b).
Note that X and Y are not the zero functions for otherwise u 0 and this
contradicts our assumption that u is the non-trivial solution.
Consider the first equation: since X 00 X = 0 the solution depends on the
sign of . If = 0 then X(x) = Ax+B. Now, the conditions X(0) = X(a) = 0

imply A = B = 0 and
so u 0. So assume that 6= 0. If > 0 then
x
x
X(x) = Ae
+ Be
. Now, the conditions X(0) = X(a) = 0, 6= 0
imply A = B = 0 and hence the solution is the trivial solution. Hence, in
order to have a nontrivial solution we must have < 0. In this case,
X(x) = A cos x + B sin x.

The condition
X(0) = 0 implies A = 0. The condition X(a) = 0 implies
X(x) = 0 and
this leads to
B sin a = 0. We must have B 6= 0 otherwise
the trivial solution. Since B 6= 0, we obtain sin a = 0 or a = n
2 2
where n Z. Solving for we find n = na2 . Thus, we obtain infinitely
many solutions given by
Xn (x) = sin
n
x, n N.
a

Y 00 + Y = 0
we obtain
Yn (y) = an e
n y
+ bn e
n y
= An cosh
p
p
n y + Bn sinh n y, n N.
Using the boundary condition Y (0) = 0 we obtain An = 0 for all n N.

n
n
un (x, y) = Bn sin
x sinh
y , nN
a
a
satisfy (19.1) and the boundary conditions u(0, y) = u(a, y) = u(x, 0) = 0.
Now, in order for these solutions to satisfy the boundary value condition
u(x, b) = g2 (x), we invoke the superposition principle of linear PDE to write
u(x, y) =
X
n=1
Bn sin
n
n
x sinh
y .
a
a
(19.2)
To determine the unknown constants Bn we use the boundary condition

u(x, b) = g2 (x) in (19.2) to obtain
g2 (x) =

X
n=1
n
n
Bn sinh
b sin
x .
a
a

Since the right-hand side is the Fourier sine series of f on the interval [0, a],
the coefficients Bn are given by
Z a
n
n
2
Bn =
g2 (x) sin
x dx [sinh
b ]1 .
(19.3)
a 0
a
a
Thus, the solution to the Laplaces equation is given by (19.2) with the Bn0 s
calculated from (19.3).
Example 19.2
Solve
uxx + uyy = 0
u(0, y) = f1 (y),
u(a, y) = u(x, 0) = u(x, b) = 0.

Solution.
Assume that the solution can be written in the form u(x, y) = X(x)Y (y).
Substituting in (19.1), we obtain
X 00 (x)Y (y) + X(x)Y 00 (y) = 0.
Assuming X(x)Y (y) is nonzero, dividing for X(x)Y (y) and subtracting
Y 00 (y)
Y (y)
X 00 (x)
Y 00 (y)
=
.
X(x)
Y (y)
X 00 (x)
Y 00 (y)
=
=
X(x)
Y (y)
X 00 X = 0 and Y 00 + Y = 0.
As far as the boundary conditions, we have for all 0 x a and 0 y b
u(0, y) = f1 (y) = X(0)Y (y)
u(a, y) = 0 = X(a)Y (y) = X(a) = 0

u(x, 0) = 0 = X(x)Y (0) = Y (0) = 0
u(x, b) = 0 = X(x)Y (b) = Y (b) = 0.
Note that X and Y are not the zero functions for otherwise u 0 and this
contradicts our assumption that u is the non-trivial solution.
Consider the second equation: since Y 00 +Y = 0 the solution depends on the
sign of . If = 0 then Y (y) = Ay +B. Now, the conditions Y (0) = Y (b) = 0
imply A = B = 0 and so u 0. So assume that 6= 0. If < 0 then
Y (y) = Ae y + Be y . Now, the condition Y (0) = Y (b) = 0 imply
A = B = 0 and hence the solution is the trivial solution. Hence, in order to
have a nontrivial solution we must have > 0. In this case,
Y (y) = A cos y + B sin y.

The condition
Y (0) = 0 implies A = 0. The condition Y (b) = 0 implies
B sin b = 0. We must have B 6= 0 otherwiseY (y) = 0 and

this leads to
the trivial solution. Since B 6= 0, we obtain sin b = 0 or b = n where
2 2
n Z. Solving for we find n = nb2 . Thus, we obtain infinitely many
solutions given by
n
y , n N.
Yn (y) = sin
b
X 00 X = 0, > 0
we obtain
Xn (x) = an e
n x
+ bn e
n x
= An cosh
n
n
x + Bn sinh
x , n N.
b
b
However, this is not really suited for dealing with the boundary condition
X(a) = 0. So, lets also notice that the following is also a solution.
n

n

Xn (x) = An cosh
(x a) + Bn sinh
(x a) , n N.
b
b
Now, using the boundary condition X(a) = 0 we obtain An = 0 for all n N.
n

n
un (x, y) = Bn sin
y sinh
(x a) , n N
b
b

satisfy (19.1) and the boundary conditions u(a, y) = u(x, 0) = u(x, b) = 0.
Now, in order for these solutions to satisfy the boundary value condition
u(0, y) = f1 (y), we invoke the superposition principle of linear PDE to write
u(x, y) =
Bn sin
n=1
n

n
y sinh
(x a) .
b
b
(19.4)
To determine the unknown constants Bn we use the boundary condition

u(0, y) = f1 (y) in (19.4) to obtain
f1 (y) =

X
n=1
n
n
Bn sinh a sin
y .
b
b
Since the right-hand side is the Fourier sine series of f1 on the interval [0, b],
the coefficients Bn are given by
Z b
n i1
n h
2
f1 (y) sin
y dy sinh a
.
(19.5)
Bn =
b 0
b
b
Thus, the solution to the Laplaces equation is given by (19.4) with the Bn0 s
calculated from (19.5)
Example 19.3
Solve
uxx + uyy = 0, 0 < x < L, 0 < y < H
u(0, y) = u(L, y) = 0, 0 < y < H
u(x, 0) = uy (x, 0), u(x, H) = f (x), 0 < x < L.
Solution.
Using separation of variables we find
Y 00
X 00
=
= .
X
Y
We first solve
X 00 X = 0
0<x<L
X(0) = X(L) = 0.
2 2
We find n = nL2 and

Xn (x) = sin
n
x, n N.
L

Next we need to solve

Y 00 + Y = 0
0<y<H
0
Y (0) Y (0) = 0.
The solution of the ODE is

n
n
y + Bn sinh
y , n N.
L
L
The boundary condition Y (0) Y 0 (0) = 0 implies
n
An Bn
= 0.
L
Hence,
n
n
n
Yn = Bn
cosh
y + Bn sinh
y , n N.
L
L
L
Using the superposition principle and the results above we have
n
n i
X
n h n
x
cosh
y + sinh
y .
u(x, y) =
Bn sin
L
L
L
L
n=1
Yn (y) = An cosh
Substituting in the condition u(x, H) = f (x) we find
n
n i
X
n h n
f (x) =
Bn sin
x
cosh
H + sinh
H .
L
L
L
L
n=1
Recall the Fourier sine series of f on [0, L] given by
f (x) =
X
n=1
An sin
n
x
L
where
Z
n
2 L
f (x) sin
x dx.
An =
L 0
L
Thus, the general solution is given by
n
n i
X
n h n
u(x, y) =
Bn sin
x
cosh
y + sinh
y .
L
L
L
L
n=1
with the Bn satisfying
Z
n
n i
n
h n
2 L
Bn
cosh
H + sinh
H =
f (x) sin
x dx
L
L
L
L 0
L
Practice Problems
Problem 19.1
Solve
uxx + uyy = 0
u(a, y) = f2 (y),
u(0, y) = u(x, 0) = u(x, b) = 0.

Problem 19.2
Solve
uxx + uyy = 0
u(x, 0) = g1 (x),
u(0, y) = u(a, y) = u(x, b) = 0.

Problem 19.3
Solve
uxx + uyy = 0
u(x, 0) = u(0, y) = 0,
u(1, y) = 2y, u(x, 1) = 3 sin x + 2x

where 0 x 1 and 0 y 1. Hint: Define U (x, y) = u(x, y) 2xy.
Problem 19.4
Show that u(x, y) = x2 y 2 and u(x, y) = 2xy are harmonic functions.
Problem 19.5
Solve
uxx + uyy = 0, 0 x L,
H
H
y
2
2
subject to
u(0, y) = u(L, y) = 0,
u(x,
H
H
<y<
2
2
H
H
) = f1 (x), u(x, ) = f2 (x), 0 x L.
2
2
Problem 19.6
Consider a complex valued function f (z) = u(x, y) + iv(x, y) where i = 1.

We say that f is holomorphic or analytic if and only if f can be expressed
as a power series in z, i.e.
u(x, y) + iv(x, y) =
an z n .
n=0
(a) By differentiating with respect to x and y show that

ux = vy and uy = vx
These are known as the Cauchy-Riemann equations.
(b) Show that u = 0 and v = 0.
Problem 19.7
Show that Laplaces equation in polar coordinates is given by
1
1
urr + ur + 2 u = 0.
r
r
Problem 19.8
Solve
uxx + uyy = 0, 0 x 2, 0 y 3
subject to
u(x, 0) = 0, u(x, 3) =

u(0, y) = sin
x
2

4
y , u(2, y) = 7.
3
Problem 19.9
Solve
uxx + uyy = 0, 0 x L, 0 y H
subject to
uy (x, 0) = 0, u(x, H) = 0
y
u(0, y) = u(L, y) = 4 cos
.
2H
Problem 19.10
Solve
uxx + uyy = 0, x > 0, 0 y H
subject to
u(0, y) = f (y), |u(x, 0)| <
uy (x, 0) = uy (x, H) = 0.

Problem 19.11
Consider Laplaces equation inside a rectangle
uxx + uyy = 0, 0 x L, 0 y H
subject to the boundary conditions
u(0, y) = 0, u(L, y) = 0, u(x, 0)uy (x, 0) = 0, u(x, H) = 20 sin
x
L

5 sin
Find the solution u(x, y).

Problem 19.12
Solve Laplacee equation uxx + uyy = 0 in the rectangle 0 < x, y < 1 subject
to the conditions
u(0, y) = u(1, y) = 0, 0 < y < 1
u(x, 0) = sin (2x), uy (x, 0) = 2 sin (2x), 0 < x < 1.
Problem 19.13
Find the solution to Laplaces equation on the rectangle 0 < x < 1, 0 < y < 1
with boundary conditions
u(x, 0) = 0, u(x, 1) = 1
ux (0, y) = ux (1, y) = 0.
Problem 19.14
Solve Laplaces equation on the rectangle 0 < x < a, 0 < y < b with the
boundary conditions
ux (0, y) = a, ux (a, y) = 0
uy (x, 0) = b, uy (x, b) = 0.
Problem 19.15
Solve Laplaces equation on the rectangle 0 < x < , 0 < y < 2 with the
boundary conditions
u(0, y) = u(, y) = 0
uy (x, 0) = 0, uy (x, 2) = 2 sin 3x 5 sin 10x.

3x
.
L
20 LAPLACES EQUATIONS IN CIRCULAR REGIONS
165
20 Laplaces Equations in Circular Regions

In the previous section we solved the Dirichlet problem for Laplaces equation
on a rectangular region. However, if the domain of the solution is a disc,
an annulus, or a circular wedge, it is useful to study the two-dimensional
Laplaces equation in polar coordinates.
It is well known in calculus that the cartesian coordinates (x, y) and the polar
coordinates (r, ) of a point are related by the formulas
x = r cos and y = r sin
1
2
where r = (x2 + y 2 ) and tan = xy . Using the chain rule we obtain

ux =ur rx + u x = cos ur
sin
u
r
uxx =uxr rx + ux x

sin
sin
= cos urr + 2 u
ur cos
r
r

sin
sin
cos
u
u
+ sin ur + cos ur
r
r
r
cos
uy =ur ry + u y = sin ur +
u
r
uyy =uyr ry + uy y

cos
cos
= sin urr 2 u +
ur sin
r
r

cos
cos
sin
u +
u
.
+ cos ur + sin ur
r
r
r
Substituting these equations into u = 0 we obtain
1
1
urr + ur + 2 u = 0.
r
r
Example 20.1
Find the solution to
u = 0, x2 + y 2 < a2
subject to
(i) Boundary condition: u(a, ) = f (), 0 2.
(ii) Boundedness at the origin: |u(0, )| < .
(iii) Periodicity: u(r, + 2) = u(r, ), 0 2.
(20.1)

Solution.
First, note that (iii) implies that u(r, 0) = u(r, 2) and u (r, 0) = u (r, 2).
Next, we will apply the method of separation of variables to (20.1). Suppose
that a solution u(r, ) of (20.1) can be written in the form u(r, ) = R(r)().
Substituting in (20.1) we obtain
1
1
R00 (r)() + R0 (r)() + 2 R(r)00 () = 0.
r
r
Dividing by R (under the assumption that R 6= 0) we obtain
00 ()
R00 (r)
R0 (r)
= r2
r
.
()
R(r)
R(r)
The left-hand side is independent of r whereas the right-hand side is independent of so that there is a constant such that
R0 (r)
00 ()
R00 (r)
= r2
+r
= .
()
R(r)
R(r)
This results in the following ODEs

00 () + () = 0
(20.2)
r2 R00 (r) + rR0 (r) R(r) = 0.
(20.3)
and
The second equation is known as Eulers equation. Both of these equations
are easily solvable. To solve (20.2), We only have to add the appropriate
boundary conditions. We have (0) = (2) and 0 (0) = 0 (2). The
periodicity of implies that = n2 and must be of the form
n () = A0n cos n + Bn0 sin n, n = 0, 1, 2
The equation in R is of Euler type and its solution must be of the form
R(r) = r . Since = n2 , the corresponding characteristic equation is
( 1)r + r n2 r = 0.
Solving this equation we find = n. Hence, we let
Rn (r) = Cn rn + Dn rn , n N.
167
For n = 0, R = 1 is a solution. To find a second solution, we solve the

equation
r2 R00 + rR0 = 0.
This can be done by dividing through by r and using the substitution S = R0
to obtain rS 0 + S = 0. Solving this by noting that the left-hand side is just
(rS)0 we find S = rc . Hence, R0 = rc and this implies R(r) = C ln r. Thus,
R = 1 and R = ln r form a fundamental set of solutions of (20.3) and so a
general solution is given by
R0 (r) = C0 + D0 ln r.
By assumption (ii), u(r, ) must be bounded at r = 0, and so does Rn . Since
rn and ln r are unbounded at r = 0, we must set D0 = Dn = 0. In this case,
the solutions to Eulers equation are given by
Rn (r) = Cn rn , n = 0, 1, 2, .
Using the superposition principle, and combining the results obtained above,
we find
X
rn (An cos n + Bn sin n).
u(r, ) = C0 +
n=1
Now, using the boundary condition u(a, ) = f () we can write
X
f () = C0 +
(an An cos n + an Bn sin n)
n=1
which is usually written in a more convenient equivalent form by
f () =
a0 X
+
(an cos n + bn sin n).
2
n=1
It is obvious that an and bn are the Fourier coefficients, and therefore can be
determined by the formulas
Z
1 2
an =
f () cos nd, n = 0, 1,
0
and
1
bn =
f () sin nd, n = 1, 2, .
0

Finally, the general solution to our problem is given by
u(r, ) = C0 +
rn (An cos n + Bn sin n)
n=1
where
Z 2
a0
1
C0 = =
f ()d
2
2 0
Z 2
1
an
f () cos nd, n = 1, 2,
An = n = n
a
a 0
Z 2
bn
1
f () sin nd, n = 1, 2,
Bn = n = n
a
a 0
Example 20.2
Solve
u = 0, 0 < 2, 1 r 2
subject to
u(1, ) = u(2, ) = sin , 0 < 2.
Solution.
Use separation of variables. First, solving for (), we see that in order to
ensure that the solution is 2periodic in , the eigenvalues are = n2 .
When solving the equation for R(r), we do NOT need to throw out solutions
which are not bounded as r 0. This is because we are working in the
annulus where r is bounded away from 0 and . Therefore, we obtain the
general solution
u(r, ) = (C0 + C1 ln r) +
[(Cn rn + Dn rn ) cos n + (An rn + Bn rn ) sin n].
n=1
But
C0 +
X
[(Cn + Dn ) cos n + (An + Bn ) sin n] = sin
n=1
and
C0 +
X
n=1
[(Cn 2n + Dn 2n ) cos n + (An 2n + Bn 2n ) sin n] = sin
169
Hence, comparing coefficients we must have

C0
Cn + Dn
An + Bn
A1 + B1
n
Cn 2 + Dn 2n
An 2n + Bn 2n
2A1 + 21 B1
=0
=0
=0 (n 6= 1)
=1
=0
=0 (n =
6 1)
=1.
Solving these equations we find C0 = Cn = Dn = 0, A1 = 31 , B1 = 23 , and

An = Bn = 0 for n 6= 1. Hence, the solution to the problem is

1
2
u(r, ) =
r+
sin
3
r
Example 20.3
Solve Laplaces equation inside a 60 wedge of radius a subject to the boundary conditions
1
1
u (r, ) = 0, u (r, ) = 0, u(a, ) = cos 9 cos 3.

3
3
9
You may assume that the solution remains bounded as r 0.
Solution.
Separating the variables we obtain the eigenvalue problem
00 () + () = 0

= 0.
0 (0) = 0
3
As above, because of periodicity we expect the solution to be of the form
() = A cos + B sin .

The condition 0 (0) = 0 implies B = 0. The condition 0 3 = 0 implies
n = (3n)2 , n = 0, 1, 2, . Thus, the angular solution is
n () = An cos 3n, n = 0, 1, 2,

The corresponding solutions of the radial problem are
Rn (r) = Bn r3n + Cn r3n , n = 0, 1, .
To obtain a solution that remains bounded as r 0 we take Cn = 0. Hence,
u(r, ) =
Dn r3n cos 3n, n = 0, 1, 2,
n=0
Using the boundary condition

u(a, ) =
1
1
cos 9 cos 3
3
9
we obtain D1 a3 = 19 and D3 a9 =
u(a, ) =
1
3
and 0 otherwise. Thus,
1 r 9
1 r 3
cos 9
cos 3
3 a
9 a
171
Practice Problems
Problem 20.1
Solve the Laplaces equation as in Example 20.1 in the unit disk with u(1, ) =
3 sin 5.
Problem 20.2
Solve the Laplaces equation in the upper half of the unit disk with u(1, ) =
.
Problem 20.3
Solve the Laplaces equation in the unit disk with ur (1, ) = 2 cos 2.
Problem 20.4
Consider
u(r, ) = C0 +
rn (An cos n + Bn sin n)
n=1
with
Z 2
1
a0
f ()d
C0 = =
2
2 0
Z 2
an
1
An = n = n
f () cos nd, n = 1, 2,
a
a 0
Z 2
bn
1
Bn = n = n
f () sin nd, n = 1, 2,
a
a 0
Using the trigonometric identity
cos a cos b + sin a sin b = cos (a b)
show that
1
u(r, ) =
2
"
f () 1 + 2

X
r n
n=1
#
cos n( ) d.
Problem 20.5
(a) Using Eulers formula from complex analysis eit = cos t + i sin t show that
1
cos t = (eit + eit ),
2
where i = 1.
(b) Show that
1+2

X
r n
n=1
cos n( ) = 1 +

X
r n
n=1
in()

X
r n
n=1
ein() .
(c) Let q1 = ar ei() and q2 = ar ei() . It is defined in complex analysis that

1
the absolute value of a complex number z = x+iy is given by |z| = (x2 +y 2 ) 2 .
Using these concepts, show that |q1 | < 1 and |q2 | < 1.
Problem 20.6
(a)Show that

X
r n

X
r n
n=1
rei()
a rei()
ein() =
rei()
a rei()
n=1
and
ein() =
Hint: Each sum is a geoemtric series with a ratio less than 1 in absolute
value so that these series converges.
(b) Show that
1+2

X
r n
n=1
cos n( ) =
a2 r 2
.
a2 2ar cos ( ) + r2
Problem 20.7
Show that
a2 r 2
u(r, ) =
2
Z
0
a2
f ()
d.
2ar cos ( ) + r2
This is known as the Poisson formula in polar coordinates.

Problem 20.8
Solve
uxx + uyy = 0, x2 + y 2 < 1
subject to
u(1, ) = ,
173
Problem 20.9
The vibrations of a symmetric circular membrane where the displacement
u(r, t) depends on r and t only can be describe by the one-dimensional wave
equation in polar coordinates
1
utt = c2 (urr + ur ), 0 < r < a, t > 0
r
with initial condition
u(a, t) = 0, t > 0
and boundary conditions
u(r, 0) = f (r), ut (r, 0) = g(r), 0 < r < a.
(a) Show that the assumption u(r, t) = R(r)T (t) leads to the equation
1 00 1 R0
1 T 00
=
R +
= .
c2 T
R
rR
(b) Show that < 0.
Problem 20.10
Cartesian coordinates and cylindrical coordinates are shown in Figure 20.1
below.
Figure 20.1

(a) Show that x = r cos , y = r sin , z = z.
(b) Show that
1
1
uxx + uyy + uzz = urr + ur + 2 u + uzz .
r
r
Problem 20.11
An important result about harmonic functions is the so-called the maximum
principle which states: Any harmonic function u(x, y) defined in a domain
satisfies the inequality
min u(x, y) u(x, y) max u(x, y), (x, y)
(x,y)
(x,y)
where denotes the boundary of .

Let u be harmonic in = {(x, y) : x2 + y 2 < 1} and satisfies u(x, y) = 2 x
for all (x, y) . Show that u(x, y) > 0 for all (x, y) .
Problem 20.12
Let u be harmonic in = {(x, y) : x2 + y 2 < 1} and satisfies u(x, y) = 1 + 3x
for all (x, y) . Determine
(i) max(x,y) u(x, y)
(ii) min(x,y) u(x, y)
without solving u = 0.
Problem 20.13
Let u1 (x, y) and u2 (x, y) be harmonic functions on a smooth domain such
that
u1 | = g1 (x, y) and u2 | = g3 (x, y)
where g1 and g2 are continuous functions satisfying
max g1 (x, y) < min g1 (x, y).
(x,y)
(x,y)
Prove that u1 (x, y) < u2 (x, y) for all (x, y) .

Problem 20.14
Show that rn cos (n) and rn sin (n) satisfy Laplaces equation in polar coordinates.

Problem 20.15
Solve the Dirichlet problem
u = 0, 0 r < a,
u(a, ) = sin2 .
Problem 20.16
Solve Laplaces equation
uxx + uyy = 0
outside a circular disk (r a) subject to the boundary condition
u(a, ) = ln 2 + 4 cos 3.
You may assume that the solution remains bounded as r .
175
The Laplace Transform

Solutions for PDEs
If in a partial differential equation the time t is one of the independent variables of the searched-for function, we say that the PDE is an evolution
equation. Examples of evolutions equations are the heat equation and the
wave equation. In contrast, when the equation involves only spatial independent variables then the equation is called a stationary equation. Examples
of stationary equations are the Laplaces equations and Poisson equations.
There are classes of methods that can be used for solving the initial value or
initial boundary problems for evolution equations. We refer to these methods as the methods of integral transforms. The fundamental ones are the
Laplace and the Fourier transforms. In this chapter we will just consider the
Laplace transform.
177
178
THE LAPLACE TRANSFORM SOLUTIONS FOR PDES
21 Essentials of the Laplace Transform

Laplace transform has been introduced in an ODE course, and is used especially to solve linear ODEs with constant coefficients, where the equations
are transformed to algebraic equations. This idea can be easily extended
to PDEs, where the transformation leads to the decrease of the number of
independent variables. PDEs in two variables are thus reduced to ODEs. In
this section we review the Laplace transform and its properties.
Laplace transform is yet another operational tool for solving constant coefficients linear differential equations. The process of solution consists of three
main steps:
The given hard problem is transformed into a simple equation.
This simple equation is solved by purely algebraic manipulations.
The solution of the simple equation is transformed back to obtain the solution of the given problem.
In this way the Laplace transformation reduces the problem of solving a differential equation to an algebraic problem. The third step is made easier by
tables, whose role is similar to that of integral tables in integration.
The above procedure can be summarized by Figure 21.1
Figure 21.1
In this section we introduce the concept of Laplace transform and discuss
some of its properties.
The Laplace transform is defined in the following way. Let f (t) be defined
for t 0. Then the Laplace transform of f, which is denoted by L[f (t)]
or by F (s), is defined by the following equation
Z
L[f (t)] = F (s) = lim
T
st
f (t)e
0
Z
dt =
f (t)est dt
The integral which defines a Laplace transform is an improper integral. An

improper integral may converge or diverge, depending on the integrand.
When the improper integral is convergent then we say that the function f (t)
possesses a Laplace transform. So what types of functions possess Laplace
21 ESSENTIALS OF THE LAPLACE TRANSFORM
179
transforms, that is, what type of functions guarantees a convergent improper

integral.
Example 21.1
Find the Laplace transform, if it exists, of each of the following functions
2
(a) f (t) = eat
(b) f (t) = 1 (c) f (t) = t (d) f (t) = et
Solution.
(a) Using the definition of Laplace transform we see that
Z
Z T
(sa)t
at
e
dt = lim
e(sa)t dt.
L[e ] =
T
But
(sa)t

dt =
T
1e(sa)T
sa
if s = a
if s 6= a.
For the improper integral to converge we need s > a. In this case,

L[eat ] = F (s) =
1
, s > a.
sa
(b) In a similar way to what was done in part (a), we find

Z T
Z
1
st
est dt = , s > 0.
e dt = lim
L[1] =
T 0
s
0
(c) We have
Z
L[t] =
st
te
0
test est
2
dt =
s
s

=
0
1
, s > 0.
s2
(d) Again using the definition of Laplace transform we find

Z
2
t2
L[e ] =
et st dt.
0
R 2
2
t2 st
If
1 and this implies that 0 et st dt
R s 0 then t st 0 so that e
dt. Since the integral on the right is divergent, by the comparison theo0
rem of improper
integrals,Rthe integral on the left is also divergent. Now, if
R t(ts)
s > 0 then 0 e
dt s dt. By the same reasoning the integral on the
180

2
left is divergent. This shows that the function f (t) = et does not possess a
Laplace transform
The above example raises the question of what class or classes of functions
possess a Laplace transform. To answer this question we introduce few mathematical concepts.
A function f that satisfies
|f (t)| M eat ,
tC
(21.1)
is said to be a function with an exponential order at infinity. A function

f is called piecewise continuous on an interval if the interval can be broken into a finite number of subintervals on which the function is continuous
on each open subinterval (i.e. the subinterval without its endpoints) and
has a finite limit at the endpoints (jump discontinuities and no vertical
asymptotes) of each subinterval. Below is a sketch of a piecewise continuous
function.
Note that a piecewise continuous function is a function that has a finite

number of breaks in it and doesnt blow up to infinity anywhere. A function defined for t 0 is said to be piecewise continuous on the infinite
interval if it is piecewise continuous on 0 t T for all T > 0.
Example 21.2
Show that the following functions are piecewise continuous and of exponential
order at infinity for t 0
(a) f (t) = tn
(b) f (t) = tn sin at
Solution.
P
tn
tn
n
t
n
(a) Since et =
n=0 n! n! , t n!e . Hence, t is piecewise continuous and
of exponential order at infinity.
(b) Since |tn sin at| n!et , tn sin at is piecewise continuous and of exponential
order at infinity
181
The following is an existence result of Laplace transform.

Theorem 21.1
Suppose that f (t) is piecewise continuous on t 0 and has an exponential
order at infinity with |f (t)| M eat for t C. Then the Laplace transform
Z
f (t)est dt
F (s) =
0
exists as long as s > a. Note that the two conditions above are sufficient, but
not necessary, for F (s) to exist.
In what follows, we will denote the class of all piecewise continuous functions
with exponential order at infinity by PE. The next theorem shows that any
linear combination of functions in PE is also in PE. The same is true for the
product of two functions in PE.
Theorem 21.2
Suppose that f (t) and g(t) are two elements of PE with
|f (t)| M1 ea1 t ,
t C1
and
|g(t)| M2 ea1 t ,
t C2 .
(i) For any constants and the function f (t) + g(t) is also a member of
PE. Moreover
L[f (t) + g(t)] = L[f (t)] + L[g(t)].
(ii) The function h(t) = f (t)g(t) is an element of PE.
We next discuss the problem of how to determine the function f (t) if F (s)
is given. That is, how do we invert the transform. The following result on
uniqueness provides a possible answer. This result establishes a one-to-one
correspondence between the set PE and its Laplace transforms. Alternatively, the following theorem asserts that the Laplace transform of a member
in PE is unique.
Theorem 21.3
Let f (t) and g(t) be two elements in PE with Laplace transforms F (s) and
G(s) such that F (s) = G(s) for some s > a. Then f (t) = g(t) for all t 0
where both functions are continuous.
182
With the above theorem, we can now officially define the inverse Laplace
transform as follows: For a piecewise continuous function f of exponential
order at infinity whose Laplace transform is F, we call f the inverse Laplace
transform of F and write f = L1 [F (s)]. Symbolically
f (t) = L1 [F (s)] F (s) = L[f (t)].
Example 21.3

1
, s > 1.
Find L1 s1
Solution.
1
From Example 21.1(a), we have that L[eat ] = sa
, s > a. In particular, for

1
1
t
1
a = 1 we find that L[e ] = s1 , s > 1. Hence, L
= et , t 0 .
s1
The above theorem states that if f (t) is continuous and has a Laplace transform F (s), then there is no other function that has the same Laplace transform. To find L1 [F (s)], we can inspect tables of Laplace transforms of
known functions to find a particular f (t) that yields the given F (s).
When the function f (t) is not continuous, the uniqueness of the inverse
Laplace transform is not assured. The following example addresses the
uniqueness issue.
Example 21.4
Consider the two functions f (t) = H(t)H(3 t) and g(t) = H(t) H(t 3),
where H is the Heaviside function defined by

1, t 0
H(t) =
0, t < 0
(a) Are the two functions identical?
(b) Show that L[f (t)] = L[g(t).
Solution.
(a) We have

1, 0 t 3
0,
t>3
1, 0 t < 3
0,
t3
f (t) =
and
g(t) =
183
Since f (3) = 1 and g(3) = 0, f and g are not identical.

(b) We have
Z 3
1 e3s
est dt =
L[f (t)] = L[g(t)] =
, s > 0.
s
0
Thus, both functions f (t) and g(t) have the same Laplace transform even
though they are not identical. However, they are equal on the interval(s)
where they are both continuous
The inverse Laplace transform possesses a linear property as indicated in
the following result.
Theorem 21.4
Given two Laplace transforms F (s) and G(s) then
L1 [aF (s) + bG(s)] = aL1 [F (s)] + bL1 [G(s)]
for any constants a and b.
Convolution integrals are useful when finding the inverse Laplace transform
of products. They are defined as follows: The convolution of two scalar
piecewise continuous functions f (t) and g(t) defined for t 0 is the integral
Z t
(f g)(t) =
f (t s)g(s)ds.
0
Example 21.5
Find f g where f (t) = et and g(t) = sin t.
Solution.
Using integration by parts twice we arrive at
Z t
(f g)(t) =
e(ts) sin sds
0
t
1
= e(ts) (sin s cos s) 0
2
et 1
=
+ (sin t cos t)
2
2
(21.2)
Next, we state several properties of convolution product, which resemble

those of ordinary product.
184
Theorem 21.5
Let f (t), g(t), and k(t) be three piecewise continuous scalar functions defined
for t 0 and c1 and c2 are arbitrary constants. Then
(i) f g = g f (Commutative Law)
(ii) (f g) k = f (g k) (Associative Law)
(iii) f (c1 g + c2 k) = c1 f g + c2 f k (Distributive Law)
Example 21.6
Express the solution to the initial value problem y 0 + y = g(t), y(0) = y0
in terms of a convolution integral.
Solution.
Solving this initial value problem by the method of integrating factor we find
Z t
t
e(ts) g(s)ds = et y0 + (et g)(t)
y(t) = e y0 +
0
The following theorem, known as the Convolution Theorem, provides a way

for finding the Laplace transform of a convolution integral and also finding
the inverse Laplace transform of a product.
Theorem 21.6
If f (t) and g(t) are piecewise continuous for t 0, and of exponential order
at infinity then
L[(f g)(t)] = L[f (t)]L[g(t)] = F (s)G(s).
Thus, (f g)(t) = L1 [F (s)G(s)].
Example 21.7
Use the convolution theorem to find the inverse Laplace transform of
1
.
P (s) = 2
(s + a2 )2
Solution.
Note that

P (s) =
1
2
s + a2

1
2
s + a2

.
1
1
So, in this case we have, F (s) = G(s) = s2 +a
2 so that f (t) = g(t) = a sin (at).
Thus,
Z
1 t
1
(f g)(t) = 2
sin (at as) sin (as)ds = 3 (sin (at) at cos (at))
a 0
2a
185
Example 21.8
4y 00 + y = g(t), y(0) = 3, y 0 (0) = 7
Solution.
Take the Laplace transform of all the terms and plug in the initial conditions
to obtain
4(s2 Y (s) 3s + 7) + Y (s) = G(s)
or
(4s2 + 1)Y (s) 12s + 28 = G(s).
Solving for Y (s) we find
G(s)
12s 28

+
4 s2 + 14
4 s2 + 41

1 2
1 2
3s
1
2
2
=
7
+ G(s)
2
1 2
1 2
2
2
2
4
s + (2
s + 2
s + 21
Y (s) =
Hence,

Z
s
t
t
1 t
sin
g(t s)ds.
y(t) = 3 cos
7 sin
+
2
2
2 0
2
So, once we decide on a g(t) all we need to do is to evaluate the integral and
well have the solution
We conclude this section with the following table of Laplace transform pairs
where H is the Heaviside function defined by H(t) = 1 for t 0 and 0
otherwise.
186

f(t)
F(s)

H(t) =
1, t 0
0, t < 0
1
s,
tn , n = 1, 2,
et
sin (t)
cos (t)
sinh (t)
cosh (t)
et f (t), with |f (t)| M eat
et H(t)
et tn , n = 1, 2,
et sin (t)
et cos (t)
f (t )H(t ), 0
with |f (t)| M eat
s>0
n!
, s>0
sn+1
1
,
s s >
, s>0
s2 + 2
s
, s>0
s2 + 2
, s > ||
s2 2
s
, s > ||
s2 2
F (s ), s > + a
1
s , s >
n!
, s>
(s)n+1
, s>
(s)2 + 2
s
, s>
(s)2 + 2
s
e
F (s), s > a
H(t ), 0
tf (t)
es
s , s
-F 0 (s)
t
2
s
,
(s2 + 2 )2
s>0
1
,
(s2 + 2 )2
s>0
sin t
1
[sin t
2 3
t cos t]
>0
f 0 (t), with f (t) continuous

and |f 0 (t)| M eat
sF (s) f (0)
s > max{a, 0} + 1
f 00 (t), with f 0 (t) continuous

and |f 00 (t)| M eat
s2 F (s) sf (0) f 0 (0)

s > max{a, 0} + 1
f (n) (t), with f (n1) (t) continuous

and |f (n) (t)| M eat
sn F (s) sn1 f (0)

-sf (n2) (0) f (n1) (0)
s > max{a, 0} + 1
Rt
F (s)
s ,
f (u)du, with |f (t)| M eat
Table L
s > max{a, 0} + 1
187
Practice Problems
Problem 21.1
R
Determine whether the integral 0
verges, give its value.
1
dt
1+t2
converges. If the integral con-
Problem 21.2
R
Determine whether the integral 0
verges, give its value.
t
dt
1+t2
converges. If the integral con-
Problem 21.3
R
Determine whether the integral 0 et cos (et )dt converges. If the integral
converges, give its value.
Problem 21.4
Using the definition, find L[e3t ], if it exists. If the Laplace transform exists
then find the domain of F (s).
Problem 21.5
Using the definition, find L[t 5], if it exists. If the Laplace transform exists
Problem 21.6
2
Using the definition, find L[e(t1) ], if it exists. If the Laplace transform
exists then find the domain of F (s).
Problem 21.7
Using the definition, find L[(t 2)2 ], if it exists. If the Laplace transform
exists then find the domain of F (s).
Problem 21.8
Using the definition, find L[f (t)], if it exists. If the Laplace transform exists

f (t) =
0,
0t<1
t 1,
t1
188
Problem 21.9
Using the definition, find L[f (t)], if it exists. If the Laplace transform exists
0t<1
0,
t 1, 1 t < 2
f (t) =
0,
t 2.
Problem 21.10
Let n be a positive integer. Using integration by parts establish the reduction
formula
Z
Z
tn est n
n st
+
tn1 est dt, s > 0.
t e dt =
s
s
Problem 21.11
For s > 0 and n a positive integer evaluate the limits
(a) limt0 tn est
(b) limt tn est
Problem 21.12
Use the linearity property of Laplace transform to find L[5e7t + t + 2e2t ].
Find the domain of F (s).
Problem 21.13

3
.
Find L1 s2
Problem 21.14
Find L1 s22 +
1
s+1
Problem 21.15

2
2
Find L1 s+2
+ s2
.
Problem 21.16
Use Table L to find L[2et + 5].
Problem 21.17
Use Table L to find L[e3t3 H(t 1)].
Problem 21.18
Use Table L to find L[sin2 t].

Problem 21.19
Use Table L to find L[sin 3t cos 3t].
Problem 21.20
Use Table L to find L[e2t cos 3t].
Problem 21.21
Use Table L to find L[e4t (t2 + 3t + 5)].
Problem 21.22
+
Use Table L to find L1 [ s210
+25
4
].
s3
Problem 21.23
5
Use Table L to find L1 [ (s3)
4 ].
Problem 21.24
2s
Use Table L to find L1 [ es9 ].
Problem 21.25
h
i
12
.
Using the partial fraction decomposition find L1 (s3)(s+1)
Problem 21.26
h
i
1 24e5s
Using the partial fraction decomposition find L
.
s2 9
Problem 21.27
Use Laplace transform technique to solve the initial value problem
y 0 + 4y = g(t), y(0) = 2
where
0, 0 t < 1
12, 1 t < 3
g(t) =
0,
t3
Problem 21.28
Use Laplace transform technique to solve the initial value problem
y 00 4y = e3t , y(0) = 0, y 0 (0) = 0.
189
190
Problem 21.29
Consider the functions f (t) = et and g(t) = e2t , t 0. Compute f g in
two different ways.
(a) By directly evaluating the integral.
(b) By computing L1 [F (s)G(s)] where F (s) = L[f (t)] and G(s) = L[g(t)].
Problem 21.30
Consider the functions f (t) = sin t and g(t) = cos t, t 0. Compute f g in
two different ways.
(a) By directly evaluating the integral.
(b) By computing L1 [F (s)G(s)] where F (s) = L[f (t)] and G(s) = L[g(t)].
Problem 21.31
Compute t t t.
Problem 21.32
Compute H(t) et e2t .
Problem 21.33
Compute t et et .
22 SOLVING PDES USING LAPLACE TRANSFORM
191
22 Solving PDEs Using Laplace Transform

The same idea for solving linear ODEs using Laplace transform can be exploited when solving PDEs for functions in two variables u = u(x, t). The
transformation will be done with respect to the time variable t 0, the spatial variable x will be treated as a parameter unaffected by this transform.
In particular we define the Laplace transform of u(x, t) by the formula
Z
u(x, t)est dt.
L(u(x, t)) = U (x, s) =
0
The time derivatives are transformed in the same way as in the case of
functions in one variable, that is, for example
L(ut )(x, t) = sU (x, s) u(x, 0)
and
L(utt )(x, s) = s2 U (x, s) su(x, 0) ut (x, 0).
The spatial derivatives remain unchanged, for example,
Z
Z
s
ux (x, )e d =
u(x, )es d = Ux (x, s).
Lux (x, t) =
x 0
0
Likewise, we have
Luxx (x, t) = Uxx (x, s).
Thus, applying the Laplace transform to a PDE in two variables x and t we
obtain an ODE in the variable x and with the parameter s.
Example 22.1
Let u(x, t) be the concentration of a chemical contaminant dissolved in a
liquid on a half-infinte domain x > 0. Let us assume that at time t = 0 the
concentration is 0 and on the boundary x = 0, constant unit concentration of
the contaminant is kept for t > 0. The behaviour of this problem is described
by the following mathematical model
ut uxx = 0 , x > 0, t > 0
u(x, 0) = 0,
u(0, t) = 1,
|u(x, t)| < .

Find u(x, t).
192
Solution.
Applying Laplace transform to both sides of the equation we obtain
sU (x, s) u(x, 0) Uxx (x, s) = 0
or
Uxx (x, s) sU (x, s) = 0.
This is a second order linear ODE in the variable x and positive parameter
s. Its general solution is
U (x, s) = A(s)e
sx
+ B(s)e
sx
Since U (x, s) is bounded in both variables, we must have B(s) = 0 and in

this case we obtain
U (x, s) = A(s)e sx .
Next, we apply Laplace transform to the boundary condition obtaining
1
U (0, s) = L(1) = .
s
This leads to A(s) =
1
s
and the transformed solution becomes

1
U (x, s) = e sx .
s
Thus,
u(x, t) = L
1 sx
e
s
2
=
ew dw
x
2 t
Example 22.2
Solve the following initial boundary value problem
ut uxx = 0 , x > 0, t > 0
u(x, 0) = 0,
u(0,
t) = f (t),
|u(x, t)| < .

Solution.
Following the argument of the previous example we find
U (x, s) = F (s)e
sx
, F (s) = Lf (t).
193
Thus, using Theorem 21.6 we can write

1
sx
u(x, t) = L
F (s)e
= f L1 (e sx ).
It can be shown that
L1 (e
sx
x2
x
)=
e 4t .
4t3
Hence,
t
Z
u(x, t) =
0
x2
x
p
e 4(ts) f (s)ds
4(t s)3
Example 22.3
Solve the wave equation
utt c2 uxx = 0
, x > 0, t > 0
u(x, 0) = ut (x, 0) = 0,
u(0, t) = f (t),
|u(x, t)| < .

Solution.
Applying Laplace transform to both sides of the equation we obtain
s2 U (x, s) su(x, 0) ut (x, 0) c2 Uxx (x, s) = 0
or
c2 Uxx (x, s) s2 U (x, s) = 0.
This is a second order linear ODE in the variable x and positive parameter
s. Its general solution is
s
U (x, s) = A(s)e c x + B(s)e c x .

Since U (x, s) is bounded, we must have B(s) = 0 and in this case we obtain
s
U (x, s) = A(s)e c x .
Next, we apply Laplace transform to the boundary condition obtaining
U (0, s) = L(f (t)) = F (s).
194
This leads to A(s) = F (s) and the transformed solution becomes

s
U (x, s) = F (s)e c x .
Thus,

x
x
x
f t
u(x, t) = L1 F (s)e c s = H t
c
c
Remark 22.1
Laplace transforms are useful in solving parabolic and some hyperbolic PDEs.
They are not in general useful in solving elliptic PDEs.
195
Practice Problems
Problem 22.1
Solve by Laplace transform
, x > 0, t > 0
ut + ux = 0
u(x, 0) = sin x,
u(0, t) = 0
Hint: Method of integrating factor of ODEs.
Problem 22.2
ut + ux = u , x > 0, t > 0
u(x, 0) = sin x,
u(0, t) = 0
Problem 22.3
Solve
ut = 4uxx
u(0, t) = u(1, t) = 0
u(x, 0) = 2 sin x + 6 sin 2x.
Hint: A particular solution of a second order ODE must be found using the
method of variation of parameters.
Problem 22.4
ut ux = u , x > 0, t > 0
u(x, 0) = e5x ,
|u(x, t)| <

Problem 22.5
ut + ux = t , x > 0, t > 0
u(x, 0) = 0,
u(0, t) = t2
196
Problem 22.6
xut + ux = 0 , x > 0, t > 0

u(x, 0) = 0,
u(0, t) = t
Problem 22.7
utt c2 uxx = 0
, x > 0, t > 0
u(x, 0) = ut (x, 0) = 0,
u(0, t) = sin t,
|u(x, t)| <

Problem 22.8
utt 9uxx = 0, 0 x , t > 0
u(0, t) = u(, t) = 0,
ut (x, 0) = 0, u(x, 0) = 2 sin x.
Problem 22.9
uxy = 1
, x > 0, y > 0
u(x, 0) = 1,
u(0, y) = y + 1.
Problem 22.10
utt = c2 uxx
, x > 0, t > 0
u(x, 0) = ut (x, 0) = 0,
ux (0, t) = f (t),
|u(x, t)| < .

Problem 22.11
, x > 0, t > 0
ut + ux = u
u(x, 0) = sin x,
u(0, t) = 0
Problem 22.12
ut c2 uxx = 0 , x > 0, t > 0
u(x, 0) = T,
u(0, t) = 0,
|u(x, t)| <

Problem 22.13
ut 3uxx = 0, 0 x 2, t > 0
u(0, t) = u(2, t) = 0,
u(x, 0) = 5 sin (x)
Problem 22.14
ut 4uxx = 0, 0 x , t > 0
ux (0, t) = u(, t) = 0,
x
u(x, 0) = 40 cos
2
Problem 22.15
utt 4uxx = 0, 0 x 2, t > 0
u(0, t) = u(2, t) = 0,
ut (x, 0) = 0, u(x, 0) = 3 sin x.
197
198
The Fourier Transform

Solutions for PDEs
In the previous chapter we discussed one class of integral transform methods, the Laplace transfom. In this chapter, we consider a second fundamental
class of integral transform methods, the so-called Fourier transform.
Fourier series are designed to solve boundary value problems on bounded
intervals. The extension of Fourier methods to the entire real line leads naturally to the Fourier transform, an extremely powerful mathematical tool for
the analysis of non-periodic functions. The Fourier transform is of fundamental importance in a broad range of applications, including both ordinary and
partial differential equations, quantum mechanics, signal processing, control
theory, and probability, to name but a few.
199
200
THE FOURIER TRANSFORM SOLUTIONS FOR PDES
23 Complex Version of Fourier Series

We have seen in Section 15 that a 2Lperiodic function f : R R that is
piecewise smooth on [L, L] can be expanded in a Fourier series
n
n
a0 X
f (x) =
+
an cos
x + bn sin
x
2
L
L
n=1
at all points of continuity of f. In the context of Fourier analysis, this is
referred to as the real form of the Fourier series. It is often convenient to
recast this series in complex form by means of Euler formula
eix = cos x + i sin x.
It follows from this formula that
eix + eix = 2 cos x and eix eix = 2i sin x
or
cos x =
eix +eix
2
and sin x =
eix eix
.
2i
Hence the Fourier expansion of f can be rewritten as

"
!
inx
inx
a0 X
e L + e L
an
f (x) = +
2
2
n=1
!#
inx
inx
e L e L
+bn
2i
f (x) =
cn e
inx
L
n=
where c0 =
a0
2
and for n N we have

cn =
cn =
an ibn
2
an + ibn
.
2
(23.1)
23 COMPLEX VERSION OF FOURIER SERIES
201
It follows that if n N then

an = cn + cn
and bn = i(cn cn ).
(23.2)
That is, an and bn can be easily found once we have formulas for cn . In order
to find these formulas, we need to evaluate the following integral
Z
inx
L
imx
L
Z
dx =
i(nm)x
L
dx
i
i(nm)x L
L
L
e
=
i(n m)
L
iL
=
[cos [(n m)] + i sin [(n m)]
(n m)
cos [(n m)] i sin [(n m)]]
=0
if n 6= m. If n = m then
Z
inx
L
inx
L
dx = 2L.
L
inx
Now, if we multiply (23.1) by e L and integrate from L to L and apply

the last result we find
Z L
inx
f (x)e L dx = 2Lcn
L
which yields the formula for coefficients of the complex form of the Fourier
series:
Z L
inx
1
cn =
f (x)e L dx, n = 0, 1, 2, .
2L L
Example 23.1
Find the complex Fourier coefficients of the function
f (x) = x,
extended to be periodic of period 2.
202
Solution.
Using integration by parts and the fact that ei = ei = 1 we find
Z
1
xeinx dx
cn =
2

Z
i
ix inx
1
inx
e
e
dx
=

2
n
n

1
i in
i in
=
e
+
e
2
n
n

1 in
1 1 in
e
2e
+
2 n2
n
h
i
1
1
(1)n i
=
2i (1)n +
(0) =
2
n
2
n
for n N and for n = 0, we have
1
c0 =
2
xdx = 0
Remark 23.1
It is often the case that the complex form of the Fourier series is far simpler
to calculate than the real form. One can then use (23.2) to find the real form
of the Fourier series. For example, the Fourier coefficients of the real form of
the previous function are given by
an = (cn + cn ) = 0 and bn = i(cn cn ) = n2 (1)n+1 , n N
203
Practice Problems
Problem 23.1
Find the complex Fourier coefficients of the function
f (x) = x,
1x1
extended to be periodic of period 2.

Problem 23.2
Let
0 < x <
2
1
<
x
<
f (x) =
2
2
0
<x<
be 2periodic. Find its complex series representation.

Problem 23.3
Find the complex Fourier series of the 2periodic function f (x) = eax over
the interval (, ).
Problem 23.4
Find the complex Fourier series of the 2periodic function f (x) = sin x
over the interval (, ).
Problem 23.5
Find the complex Fourier series of the 2periodic function defined

1 0<x<T
f (x) =
0 T < x < 2
Problem 23.6
Let f (x) = x2 , < x < , be 2periodic.
(a) Calculate the complex Fourier series representation of f.
(b) Using the complex Fourier series found in (a), recover the real Fourier
series representation of f.
Problem 23.7
Let f (x) = sin x, 21 < x < 21 , be of period 1.
(a) Calculate the coefficients an , bn and cn .
(b) Find the complex Fourier series representation of f.
204
Problem 23.8
Let f (x) = 2 x, 2 < x < 2, be of period 4.
(a) Calculate the coefficients an , bn and cn .
(b) Find the complex Fourier series representation of f.
Problem 23.9
Suppose that the coefficients cn of the complex Fourier series are given by
2
if |n| is odd
in
cn =
0 if |n| is even.
Find an , n = 0, 1, 2, and bn , n = 1, 2, .
Problem 23.10
Recall that any complex number z can be written as z = Re(z) + iIm(z)
where Re(z) is called the real part of z and Im(z) is called the imaginary
part. The complex conjugate of z is the complex number z = Re(z)
iIm(z). Using these definitions show that an = 2Re(cn ) and bn = 2Im(cn ).
Problem 23.11
Suppose that

cn =
i
[einT
2n
T
2
1] if n 6= 0
if n = 0.
Find an and bn .
Problem 23.12
Find the complex Fourier series of the function f (x) = ex on [2, 2].
Problem 23.13
Consider the wave form
205
(a) Write f (x) explicitly. What is the period of f.

(b) Determine a0 and an for n N.
(c) Determine bn for n N.
(d) Determine c0 and cn for n N.
Problem 23.14
If z is a complex number we define sin z = 21 (eiz eiz ). Find the complex
form of the Fourier series for sin 3x without evaluating any integrals.
206
24 An introduction to Fourier Transforms

One of the problems with the theory of Fourier series discussed so far is that
it applies only to periodic functions. There are many times when one would
like to divide a function which is not periodic into a superposition of sines
and cosines. The Fourier transform is the tool often used for this purpose.
Like the Laplace transform, the Fourier transform is often an effective tool
in finding explicit solutions to partial differential equations.
We will introduce the Fourier transform of f (x) as a limiting case of a Fourier
series. This requires a tedious discussion which we omit and rather explain
the underlying ideas. More specifically, the approach we introduce is to
construct Fourier series of f (x) on progressively longer and longer intervals,
and then take the limit as their lengths go to infinity. This limiting process
converts the Fourier sums into integrals, and the resulting representation of
a function is renamed the Fourier transform.
To start with, let f : R R be Ra piecewise continuous function with the
properties limx f (x) = 0 and 0 |f (x)|dx < . Define the function fL

which is equal to f in an interval of the form [L, L] and vanishes outside
this interval. Note that f (x) = limL fL (x). This function can be extended
to a periodic function, denoted by fe , of period 2T with T > L and where
fe (x) = f (x) for |x| L and 0 for T x L and L x T.
Note that f (x) = limL fL (x) = limL fe (x). From the previous section
we can find the complex Fourier series of fe to be
fe (x) =
cn e
inx
T
(24.1)
n=
where
1
cn =
2T
fe (x)e
inx
T
dx.
Let R. Multiply both sides of (24.1) by eix and then integrate both sides
from T to T. Assuming integration and summation can be interchanged
we find
Z T
Z T
X
inx
ix
fe (x)e
dx =
cn
eix e T dx.
T
n=
24 AN INTRODUCTION TO FOURIER TRANSFORMS
207
It can be shown that the RHS converges, say to f(), as L (and

T ) Hence, we find
Z
f (x)eix dx.
(24.2)
f () =
The function f is called the Fourier transform of f. We will use the notation
F[f (x)] = f().
Next, it can be shown that
n
f
= 2T cn
T
so that
1 X n inx
f
eT .
fe (x) =
2T n=
T
It can be shown that as L , we have

Z
1 X n inx
T
f
e =
f()eix d
lim
T T
T
n=
so that
1
f (x) =
2
f()eix d
(24.3)
Equation (24.3) is called the Fourier inversion formula and we use the
notation F 1 [f()]. Now, if we make use of Eulers formula, we can write the
Fourier inversion formula in terms of sines and cosines,
Z
Z
i
1
f (x) =
f() cos xd +
f() sin xd
2
2
a superposition of sines and cosines of various frequencies.
Example 24.1
Find the Fourier transform of the function f (x) defined by
ax
e
if x 0
f (x) =
0
if x < 0
for some a > 0.
208
Solution.
We have
f() =
ix
eax eix dx
0

Z
x(a+i)
e
axix

=
e
dx =
(a + i) 0
0
1
=
a + i
f (x)e
dx =
The following theorem lists the basic properties of the Fourier transform
Theorem 24.1
Let f, g, be piecewise continuous functions. Then we have the following
properties:
(1) Linearity: F[f (x) + g(x)] = F[f (x)] + F[g(x)], where and are
arbitrary numbers.
(2) Shifting: F[f (x )] = ei F[f (x)].
(3) Scaling: F[f x ] = F[f (x)].
R
(4) Continuity: If |f (x)|dx < then f is continuous in .
n
(5) Differentiation:RF[f (n) (x)]
F[f (x)].
= (i)
x
1
(6) Integration: F 0 f (s)ds = i F[f (x)].
R
R
1
|f()|2 d.
(7) Parsevals Relation: |f (x)|2 dx = 2
(8) Duality: F[F[f (x)]] = 2f (x).

(9) Multiplication by xn : F[xn f (x)] = in f(n) ().
p
2
2
(10) Gaussians: F[ex ] = e 4 .
1
(11) Product: F[(f (x)g(x)] = 2
F[f (x)] F[g(x)].
(12) Convolution: F[(f g)(x)] = F[f (x)] F[g(x)].
Example 24.2
2
Determine the Fourier transform of the Gaussian u(x) = ex , > 0.
Solution.
We have
u() =
ex eix dx.
If we differentiate this relation with respect to the variable and then integrate by parts we obtain
209
xex eix dx
u () = i
Z
i
d x2 ix
(e
)e
dx
=
2 dx

Z

i
x2 ix
x2 ix
(e
)e
dx
=
+ i
e

2
Z
i2 x2 ix
=
(e
)e
dx = u()
2
2
0
Thus, we have arrived at the ODE u0 () = 2

u() whose general solution
has the form
2
u() = Ce 4 .
Using a result from real analysis which states that

Z
2
ex dx = ,
we can write
Z
u(0) =
x2
r
dx =
and therefore
r
u() =
= C,
2
e 4
Example 24.3
Prove
F[f (x)] = f().
Solution.
Using a change of variables we find
Z
F[f (x)] =
f (x)e
ix
dx =
f (x)eix dx = f()
Example 24.4
Prove
F[F[f (x)]] = 2f (x).
210
Solution.
We have
1
f (x) =
2
Thus,
Z
f()eix d
f()eix d = F[f()] = F[F[f (x)]]
2f (x) =
The following theorem lists the properties of inverse Fourier transform

Theorem 24.2
Let f and g be piecewise continuous functions.
(10 ) Linearity: F 1 [f() +
g ()] = F 1 [f()] + F 1 [
g ()].
0
1 (n)
(2 ) Derivatives: F [f ()] = (ix)n f (x).
(30 ) Multiplication by n : F 1 [ n f()] = (i)n f (n) (x).
(40 ) Multiplication by ei : F 1 [ei f()] = f (x ).
2
x2
1
e 4 .
(50 ) Gaussians: F 1 [e ] = 4
(60 ) Product: F 1 [f()
g ()] = f (x) g(x).
(70 ) Convolution: F 1 [f g()] = 2(f g)(x).
Remark 24.1
It is important to mention that there exists no established convention of how
to define the Fourier transform. In the literature, we can meet an equivalent
1
in front of the integral.
definition of (24.3) with the constant 12 or 2
There also exist definitions with positive sign in the exponent. The reader
should keep this fact in mind while working with various sources or using the
transformation tables.
211
Practice Problems
Problem 24.1
Find the Fourier transform of the function

1 if 1 x 1
f (x) =
0
otherwise.
Problem 24.2
Obtain the transformed problem when applying the Fourier transform with
respect to the spatial variable to the equation and initial condition
ut + cux = 0
u(x, 0) = f (x).
Problem 24.3
respect to the spatial variable to the equation and both initial conditions
utt = c2 uxx , x R, t > 0
u(x, 0) = f (x)
ut (x, 0) = g(x).
Problem 24.4
respect to the spatial variable to the equation and both initial conditions
u = uxx + uyy = 0, x R, 0 < y < L
u(x, 0) = 0

1 if a < x < a
u(x, L) =
0
otherwise
Problem 24.5
Find the Fourier transform of f (x) = e|x| , where > 0.
212
Problem 24.6
Prove that
F[ex H(x)] =
where

H(x) =
Problem 24.7
Prove that
1
1 + i
1 if x 0
0 otherwise.

1
= 2e H().
F
1 + ix

Problem 24.8
Prove
F[f (x )] = ei f().
Problem 24.9
Prove
F[eix f (x)] = f(x ).
Problem 24.10
Prove the following
1
F[cos (x)f (x)] = [f( + ) + f( )]
2
1
F[sin (x)f (x)] = [f( + ) f( )]
2
Problem 24.11
Prove
F[f 0 (x)] = (i)f().
Problem 24.12
Find the Fourier transform of f (x) = 1 |x| for 1 x 1 and 0 otherwise.
Problem 24.13
Find, using the definition, the Fourier transform of
1 a < x < 0
1
0<x<a
f (x) =
0
otherwise

Problem 24.14
2
Find the inverse Fourier transform of f() = e 2 .
Problem 24.15

1
1
Find F
.
a+i
213
214
25 Applications of Fourier Transforms to PDEs

Fourier transform is a useful tool for solving differential equations. In this
section, we apply Fourier transforms in solving various PDE problems. Contrary to Laplace transform, which usually uses the time variable, the Fourier
transform is applied to the spatial variable on the whole real line.
The Fourier transform will be applied to the spatial variable x while the variable t remains fixed. The PDE in the two variables x and t passes under the
Fourier transform to an ODE in the tvariable. We solve this ODE to obtain
the transformed solution u which can be converted to the original solution u
by means of the inverse Fourier transform. We illustrate these ideas in the
examples below.
First Order Transport Equation
ut + cux = 0
u(x, 0) = f (x).
Let u(, t) be the Fourier transform of u in x. Performing the Fourier transform on both the PDE and the initial condition, we reduce the PDE into an
ODE in t
u
+ ic
u=0
t
u(, 0) = f().
Solution of the ODE gives
u(, t) = f()eict .
Thus,
u(x, t) = F 1 [
u(, t)] = f (x ct)
which is exactly the same as we obtained by using the method of characteristics.( Section 8)
Second Order Wave Equation
Consider the two dimensional wave equation
utt = c2 uxx , x R, t > 0
25 APPLICATIONS OF FOURIER TRANSFORMS TO PDES
215
u(x, 0) = f (x)
ut (x, 0) = g(x).
Again, by performing the Fourier transform of u in x, we reduce the PDE
problem into an ODE problem in the variable t:
2 u
= c2 2 u
2
t
u(, 0) = f()
ut (, 0) = g().
General solution to the ODE is
u(, t) = ()eict + ()eict
where and are two arbitrary functions of . Performing the inverse
transformation and making use of the translation theorem, we get the general
solution
u(x, t) = (x ct) + (x + ct)
where F() = and F() = . But

1
1
f ()
g()
() =
2
ic

1
1
f () +
g() .
() =
2
ic
By using the integration property, we find the inverse transforms of and

Z
1
1 x
(x) =
f (x) +
g(s)ds
2
c 0

Z
1
1 x
(x) =
g(s)ds .
f (x)
2
c 0
Application of the translation property then yields directly the DAlambert
solution
Z
1
1 x+ct
u(x, t) = [f (x ct) + f (x + ct)] +
g(s)ds.
2
2c xct
216
Second Order Heat Equation

Next, we consider the heat equation
ut = kuxx , x R, t > 0
u(x, 0) = f (x).
Performing Fourier Transform in x for the PDE and the initial condition, we
obtain
u
= k 2 u
t
u(, 0) = f().
Treating as a parameter, we obtain the solution to the above ODE problem
2
u(, t) = f()ek t .
Application of the convolution theorem yields

2
u(x, t) =f (x) F 1 ]ek t ]

x2
1
4kt
=f (x)
e
4kt
Z
(xs)2
1
=
f (s)e 4kt ds.
4kt
Laplaces Equation in 2D
Consider the problem
u = uxx + uyy = 0, x R, 0 < y < L
u(x, 0) = 0

1 if a < x < a
u(x, L) =
0
otherwise.
Performing Fourier Transform in x for the PDE we obtain the second order
ODE in y
uyy = 2 u.
The general solution is given by
u(, y) = A() sinh (y) + B() cosh (y).
217
Using the boundary condition u(, 0) = 0 we find B() = 0. Using the second
boundary condition we find
Z
u(x, L)eix dx
u(, L) =
Z
Z a
a
ix
=
e
dx =
cos (x)dx
a
2 sin (a)
.
=
Hence,
A() sinh (L) =
2 sin (a)
and this implies

A() =
2 sin (a)
.
sinh (L)
Thus,
u(, y) =
2 sin (a)
sinh (y).
sinh (L)
Taking inverse Fourier transform we find

Z
1
2 sin (a)
u(x, y) =
sinh (y)eix d.
2 sinh (L)
Noting that the integrand is an even function in , we can simplify a little to
to obtain
Z
2 sin (a)
1
u(x, y) =
sinh (y) cos (x)d
2 sinh (L)
218
Practice Problems
Problem 25.1
Solve, by using Fourier transform
ut + cux = 0
x2
u(x, 0) = e 4 .
Problem 25.2
ut = kuxx u, x R
x2
u(x, 0) = e .
Problem 25.3
Solve the heat equation
ut = kuxx
subject to the initial condition

u(x, 0) =
1 if x 0
0 otherwise.
Problem 25.4
Use Fourier transform to solve the heat equation
ut = uxx + u,
<x<< t>0
u(x, 0) = f (x).
Problem 25.5
Prove that
e||y eix d =
x2
2y
.
+ y2
Problem 25.6
Solve the Laplaces equation in the half plane
uxx + uyy = 0,
< x < , 0 < y <
subject to the boundary condition

u(x, 0) = f (x), |u(x, y)| < .

Problem 25.7
Use Fourier transform to find the transformed equation of
utt + ( + )ut + u = c2 uxx
where , > 0.
Problem 25.8
ut + 3ux = 0
u(x, 0) = ex
using the Fourier transform.
Problem 25.9
ut = kuxx
u(x, 0) = ex
Problem 25.10
ut = kuxx
u(x, 0) = ex

Problem 25.11
ut + cux = 0
u(x, 0) = x2
Problem 25.12
u = 0
uy (x, 0) = f (x)
lim
x2 +y 2
u(x, y) = 0.
219
220
Appendix
221
222
APPENDIX
Appendix A: The Method of Undetermined

Coefficients
The general solution to the nonhomogeneous differential equation
y 00 + p(t)y 0 + q(t)y = g(t), a < t < b
(26.1)
has the structure

y(t) = c1 y1 (t) + c2 y2 (t) + yp (t)
where yp (t) is a particular solution to the nonhomogeneous equation. We
will write y(t) = yh (t) + yp (t) where yh (t) = c1 y1 (t) + c2 y2 (t).
In this and the next section we discuss methods for determining yp (t). The
techinque we discuss in this section is known as the method of undetermined coefficients.
This method requires that we make an initial assumption about the form of
the particular solution yp (t), but with the coefficients left unspecified, thus
the name of the method. We then substitute the assumed expression into
equation (26.1) and attempt to determine the coefficients as to satisfy that
equation.
The main advantage of this method is that it is straightforward to execute
once the assumption is made as to the form of yp (t). Its major limitation is
that it is useful only for equations with constant coefficients and the nonhomogeneous term g(t) is restricted to a very small class of functions, namely
functions of the form et Pn (t) cos t or et Pn (t) sin t where Pn (t) is a polynomial of degree n.
We next illustrate the method of undetermined coefficients by several simple
examples.
Example 26.1
Find the general solution of the nonhomogeneous equation
y 00 2y 0 3y = 36e5t .
Solution.
We seek a function where the combination yp00 2yp0 3yp is equal to 36e5t .
Since the exponential function reproduces itself through differentiation, the
most plausible guessing function will be yp (t) = Ae5t where A is a constant
to be determined. Inserting this into the given equation we arrive at
25Ae5t 10Ae5t 3Ae5t = 36e5t .
APPENDIX A: THE METHOD OF UNDETERMINED COEFFICIENTS223

Simplifying this last equation we find 12Ae5t = 36e5t . Solving for A we find
A = 3. Thus, yp (t) = 3e5t is a particular solution to the differential equation.
Next, the characteristic equation r2 2r 3 = 0 has the roots r1 = 1
and r2 = 3. Hence, the general solution to the differential equation is y(t) =
c1 et + c2 e3t + 3e5t
Example 26.2
Find the general solution of
y 00 y 0 + y = 2 sin 3t.
Solution.
The combination yp00 yp0 + yp must be equal to 2 sin 3t. Lets try with the
guess yp (t) = A sin 3t. Inserting this into the given differential equation leads
to
(2 16A) sin 3t 6A cos 3t = 0
(26.2)
and this is valid for all t. Letting t = 0 we find 6A = 0 or A = 0. Letting

t = 6 we find 2 16A = 0 or 2 = 0 which is impossible. This means
there is no choice of A that makes equation (26.2) true. Hence, our choice is
inadequate. The appearance of sine and cosine in equation (26.2) suggests a
guessing of the form yp (t) = A cos 3t + B sin 3t. Inserting this into the given
differential equation leads to
(8A 3B) cos 3t + (3A 8B) sin 3t = 2 sin 3t.
Setting 8A 3B = 0 and 3A 8B = 2 and solving for A and B we find
6
and B = 16
. Thus, a particular solution is
A = 73
73
yp (t) =
6
16
cos 3t
sin 3t.
73
73
Next, the
characteristic equation r2 r + 1 = 0 has roots r1 = 21 i 23 and
r2 = 12 + i 23 . Thus, the general solution to the homogeneous equation is
1
3
3
yh (t) = e 2 t (c1 cos
t + c2 sin
t).
2
2
The general solution to the differential equation is
1
3
3
6
16
y(t) = yh (t) + yp (t) = e 2 t (c1 cos
t + c2 sin
t) +
cos 3t
sin 3t
2
2
73
73
224
APPENDIX
Example 26.3
y 00 + 4y 0 2y = 2t2 3t + 6.
Solution.
We see from the previous two examples that the trial function has usually
the appearance of the nonhomogeneous term g(t). Since g(t) is a quadratic
function, we are going to try yp (t) = At2 + Bt + C. Inserting this into the
differential equation leads to
2At2 + (8A 2b)t + (2A + 4B 2C) = 2t2 3t + 6.
Equating coefficients of like powers of t we find A = 1, B = 25 , and
C = 9. Thus, a particular solution is
5
yp (t) = t2 t 9.
2
We next solve the homogeneous equation.
The characteristic
equation r2 +
4r 2 = 0 has the roots r1 = 2 6 and r2 = 2 + 6. Thus,
yh (t) = c1 e(2
6)t
+ c2 e(2+
6)t
The general solution of the given equation is
y(t) = yh (t) + yp (t) = c1 e(2
6)t
+ c2 e(2+
6)t
5
t2 t 9
2
Remark 26.1
The same principle used in the previous three examples extends to the case
where g(t) is a product of any two or all three of the three types of functions
discussed above, as the next example illustrates.
Example 26.4
Find a particular solution of
y 00 3y 0 4y = 8et cos 2t.

Solution.
We are going to try yp (t) = Aet cos 2t + Bet sin 2t. Inserting into the differential equation we find
(10A 2B)et cos 2t + (2A 10B)et sin 2t = 8et cos 2t.
Thus, A and B satisfy the equations 10A + 2B = 8 and 2A 10B = 0.
2
10
and B = 13
. Therefore, a particular solution is given
Solving we find A = 13
by
10
2
y(t) = et cos 2t + et sin 2t
13
13
The following example illustrates the use of Theorem 15.2.
Example 26.5
y 00 2y 0 3y = 4t 5 + 6te2t .
Solution.
The characteristic equation of the homogeneous equation is r2 2r 3 = 0
with roots r1 = 1 and r2 = 3. Thus,
yh (t) = c1 et + c2 e3t .
By Theorem 15.2, a guess for the particular solution is yp (t) = At + B +
Cte2t + De2t . Inserting this into the differential equation leads to
3At 2A 3B 3Ce2t + (2C 3D)e2t = 4t 5 + 6te2t .
From this identity we obtain 3A = 4 so that A = 34 . Also, 2A3B = 5
so that B = 23
. Since 3C = 6 we find C = 2. From 2C 3D = 0 we find
9
4
D = 3 . It follows that
y(t) = c1 e

4
23
4 2t
+ c2 e t +
2t +
e
3
9
3
3t
Although the method of undetermined coefficients provides a nice general

method for finding a particular solution, some difficulty arise as illustrated
in the following example.
226
APPENDIX
Example 26.6
y 00 y 0 2y = 4et .
Solution.
Lets try with yp (t) = Aet . Substituting this into the differential equation
leads to 0Aet = 4et . Thus, A does not exist. Why did the procedure of the
previous examples fail here? The reason is that the function et that appears
in yp is a solution to the homogeneous equation and so cannot possibly be
a solution to the nonhomogeneous equation at the same time. Then comes
the question of how to find a correct form for the particular solution.
We will try to solve a simpler equation with the same difficulty and to use
its general solution to suggest how to proceed with our given equation. The
simpler equation we consider is y 0 + y = 4et . By the method of integrating
factor we find the general solution y(t) = 4tet +cet . The second term is the
solution to the homogeneous equation whereas the first one is the solution
to the nonhomogeneous equation. We conclude from this discussion that a
good guess for the original equation would be yp (t) = Atet . If we insert
this into the differential equation we end up with 3Aet = 4et . Solving
for A we find A = 34 . Thus, yp (t) = 43 tet and the general solution to the
differential equation is y(t) = c1 et + c2 e2t 43 tet
Example 26.7
y 00 + 2y 0 + y = 2et .
Solution.
The characteristic equation is r2 +2r +1 = 0 with double roots r1 = r2 = 1.
Thus, yh (t) = c1 et + c2 tet . Our trial function can not contain either et or
tet since both are solutions to the homogeneous equation. Thus, a proper
guess is yp (t) = At2 et . Finding derivatives up to order 2 we find yp0 (t) =
2Atet At2 et and yp00 (t) = 2Aet 4Atet + At2 et . Substituting this in
the original equation and collecting like terms we find
2Aet = 2et .
Solving for A we find A = 1 so that yp (t) = t2 et . Hence, the general solution
is given by
y(t) = c1 et + c2 tet + t2 et

In the following table we list examples of g(t) along with the corresponding
form of the particular solution.
Form of g(t)
Pn (t) = an tn + an1 tn1 + + a0
Pn (t)et
Pn (t)et cos t or Pn (t)et sin t
Form of yp (t)
tr [An tn + An1 tn1 + + A1 t + A0 ]
tr [An tn + An1 tn1 + + A1 t + A0 ]et
tr et [(An tn + An1 tn1 + + A1 t + A0 ) cos t
+(Bn tn + Bn1 tn1 + + B1 t + B0 ) sin t]
The number r is chosen to be the smallest nonnegative integer such that

no term in the assumed form is a solution of the homogeneous equation
ay 00 + by 0 + cy = 0. The value of r will be 0, 1, or 2.
Example 26.8
Find the general solution of y 00 y = t + tet .
Solution.
The characteristic equation r2 1 = 0 has roots r = 1. Thus, the homogeneous solution is yh (t) = c1 et + c2 et . A trial function for the particular
solution is A0 + A1 t + t(B0 + B1 t)et since et is a solution of the homogeneous
equation. Inserting into the differential equation we find
2B1 et + 2(B0 + 2B1 t)et + (tB0 + t2 B1 )et A0 A1 t t(B0 + B1 t)et = t + tet
or
A0 A1 t + (2B1 + 2B0 + 4B1 t)et = t + tet
From this we obtain, A0 = 0, A1 = 1, B1 + B0 = 0, 4B1 = 1. Hence,
A0 = 0, A1 = 1, B0 = 41 , B1 = 14 . So
1
yp (t) = t + t(t 1)et
4
and the general solution is
1
y(t) = c1 et + c2 et t + t(t 1)et
4
Example 26.9
Solve using the method of undetermined coefficients:
y 00 + y = et + t3 , y(0) = 2, y 0 (0) = 0.
228
APPENDIX
Solution.
First, the characteristic equation is r2 + 1 = 0, with roots r = i, so the
homogeneous solution is yh (t) = c1 sin t + c2 cos t. The trial function for the
particular solution is yp (t) = Aet + Bt3 + Ct2 + Dt + E. Plugging into the
differential equation, we see
Aet + 6Bt + 2C + Aet + Bt3 + Ct2 + Dt + E = et + t3 .
Matching coefficients, we see:
2A = 1, B = 1, C = 0, 6B + D = 0, E = 0
The particular solution is
1
yp (t) = et + t3 6t,
2
and so the general solution is
1
y(t) = c1 sin t + c2 cos t + et + t3 6t
2
When t = 0, this is y(0) = c2 + 21 = 2, so c2 = 32 . The first derivative of the
general solution is y 0 (t) = c1 cos t 23 sin t + 21 et + 3t2 6. At t = 0, y 0 (0) =
c1 + 21 6 = 0, so c1 = 11
. We thus have solution y(t) = 11
sin t + 23 cos t +
2
2
1 t
e + t3 6t
2
APPENDIX B: THE METHOD OF VARIATION OF PARAMETERS 229
Appendix B: The Method of Variation of Parameters

In this section, we discuss a second method for finding a particular solution
to a nonhomogeneous differential equation
y 00 + p(t)y 0 + q(t)y = g(t), a < t < b.
(27.1)
This method has no prior conditions to be satisfied by either p(t), q(t), or

g(t). Therefore, it may sound more general than the method of undetermined
coefficients. We will see that this method depends on integration while the
previous one is purely algebraic which, for some at least, is an advantage.
To use this method, we first find the general solution to the homogeneous
equation
y(t) = c1 y1 (t) + c2 y2 (t).
Then we replace the parameters c1 and c2 by two functions u1 (t) and u2 (t)
to be determined. From this the method got its name. Thus, obtaining
yp (t) = u1 (t)y1 (t) + u2 (t)y2 (t).
Observe that if u1 and u2 are constant functions then the above y is just the
homogeneous solution to the differential equation.
In order to determine the two functions one has to impose two constraints.
Finding the derivative of yp we obtain
yp0 = (y10 u1 + y20 u2 ) + (y1 u01 + y2 u02 ).
Finding the second derivative to obtain
yp00 = y100 u1 + y10 u01 + y200 u2 + y20 u02 + (y1 u01 + y2 u02 )0 .
Since it is up to us to choose u1 and u2 we decide to do that in such a way to
make our computation simple. One way to achieving that is to impose the
condition
y1 u01 + y2 u02 = 0.
Under such a constraint yp0 and yp00 are simplified to
yp0 = y10 u1 + y20 u2
(27.2)
230
APPENDIX
and
yp00 = y100 u1 + y10 u01 + y200 u2 + y20 u02 .
In particular, yp00 does not involve u001 and u002 .
Inserting yp , yp0 , and yp00 into equation (27.1) to obtain
[y100 u1 + y10 u01 + y200 u2 + y20 u02 ] + p(t)(y10 u1 + y20 u2 ) + q(t)(u1 y1 + u2 y2 ) = g(t).
Rearranging terms,
[y100 + p(t)y10 + q(t)y1 ]u1 + [y200 + p(t)y20 + q(t)y2 ]u2 + [u01 y10 + u02 y20 ] = g(t).
Since y1 and y2 are solutions to the homogeneous equation, the previous
equation yields our second constraint
u01 y10 + u02 y20 = g(t).
(27.3)
Combining equation (27.2) and (27.3) we find the system of two equations
in the unknowns u01 and u02
y1 u01 + y2 u02 =0
u01 y10 + u02 y20 =g(t).
Since {y1 , y2 } is a fundamental set, the expression W (t) = y1 y20 y10 y2 is
nonzero so that one can find unique u01 and u02 . Using the method of elimination, these functions are given by
(t)g(t)
u01 (t) = y2W
and u02 (t) =
(t)
y1 (t)g(t)
.
W (t)
Computing antiderivatives to obtain

R
R
(t)g(t)
u1 (t) = y2W
dt and u2 (t) =
(t)
Example 27.1
y 00 y 0 2y = 2et
using the method of variation of parameters.
y1 (t)g(t)
dt.
W (t)
APPENDIX B: THE METHOD OF VARIATION OF PARAMETERS 231

Solution.
The characteristic equation r2 r 2 = 0 has roots r1 = 1 and r2 = 2.
Thus, y1 (t) = et , y2 (t) = e2t and W (t) = 3et . Hence,
Z 2t
e 2et
2
u1 (t) =
dt = t
t
3e
3
and
Z
u2 (t) =
et 2et
2
dt = e3t .
t
3e
9
The particular solution is

2
2
yp (t) = tet et .
3
9
The general solution is then given by
2
2
y(t) = c1 et + c2 e2t tet et
3
9
Example 27.2
Find the general solution to (2t 1)y 00 4ty 0 + 4y = (2t 1)2 et if y1 (t) = t
and y2 (t) = e2t form a fundamental set of solutions to the equation.
Solution.
First we rewrite the equation in standard form
y 00
4
4t 0
y +
y = (2t 1)et .
2t 1
2t 1
Since W (t) = (2t 1)e2t we find

Z 2t
e (2t 1)et
dt = et
u1 (t) =
(2t 1)e2t
and
Z
u2 (t) =
t (2t 1)et
1 3t 1 3t
dt
=
te e .
(2t 1)e2t
3
9
Thus,
1
1
2
1
yp (t) = tet tet et = tet et .
3
9
3
9
The general solution is
2
1
y(t) = c1 t + c2 e2t + tet et
3
9
232
APPENDIX
Example 27.3
Find the general solution to the differential equation y 00 + y 0 = ln t, t > 0.
Solution.
The characterisitc equation r2 + r = 0 has roots r1 = 0 and r2 = 1 so that
y1 (t) = 1, y2 (t) = et , and W (t) = et . Hence,
Z t
Z
e ln t
u1 (t) =
dt = ln tdt = t ln t t
et
Z
Z
Z t
ln t
e
t
t
dt
u2 (t) =
dt = e ln tdt = e ln t +
t
e
t
Thus,
Z t
e
t
dt
yp (t) = t ln t t ln t + e
t
and
Z t
e
t
t
y(t) = c1 + c2 e + t ln t t ln t + e
dt
t
Example 27.4
1
y 00 + y =
.
2 + sin t
Solution.
Since the characteristic equation r2 + 1 = 0 has roots r = i, the general
solution of the corresponding homogeneous equation y 00 + y = 0 is given by
yh (t) = c1 cos t + c2 sin t
Since W (t) = 1 we find
Z
sin t
2
dt = t +
dt
u1 (t) =
2 + sin t
2 + sin t
Z
cos t
u2 (t) =
dt = ln (2 + sin t)
2 + sin t
Hence, the particular solution is
Z
2
yp (t) = sin t ln (2 + sin t) + cos t(
dt t)
2 + sin t
and the general solution is
Z
y(t) = c1 cos t + c2 sin t + yp (t)
Answers and Solutions

Section 1
1.1 (a) ODE (b) PDE (c) ODE.
1.2 uss = 0.
1.3 uss + utt = 0.
1.4 (a) Order 3, nonlinear (b) Order 1, linear, homogeneous (c) Order 2,
linear, nonhomogeneous.
1.5 (a) Linear, homogeneous, order 3.
(b) Linear, non-homogeneous, order 3. The inhomogeneity is sin y.
(c) Nonlinear, order 2. The non-linear term is ex uux .
(d) Nonlinear, order 3. The non-linear terms are ux uxxy and ex uuy .
(e) Linear, non-homogeneous, order 2. The inhomogeneity is f (x, y, t).
1.6 (a) Linear. (b) Linear. (c) Nonlinear. (d) Nonlinear.
1.7 (a) PDE, linear, second order, homogeneous.
(b) PDE, linear, second order, homogeneous.
(c) PDE, nonlinear, fourth order.
(d) ODE, linear, second order, nonhomogeneous.
(e) PDE, linear, second order, nonhomogeneous.
(f) PDE, quasilinear, second order.
1.8 A(x, y, z)uxx +B(x, y, z)uxy +C(x, y, z)uyy +E(x, y, z)uxz +F (x, y, z)uyz +
G(x, y, z)uzz +H(x, y, z)ux +I(x, y, z)uy +J(x, y, z)uz +K(x, y, z)u = L(x, y, z).
233
234
ANSWERS AND SOLUTIONS
1.9 (a) Order 3, linear, homogeneous.

(b) Order 1, nonlinear.
(c) Order 4, linear, nonhomogeneous
(d) Order 2, nonlinear.
(e) Order 2, linear, homogeneous.
1.10 uww = 0.
1.11 uvw = 0.
1.12 uvw = 0.
1.13 ut = 0.
1.14 ut = 1.
1.15 uw = 1b u.
Section 2
2.1 a = b = 0.
2.2 Substituting into the differential equation we find
tX 00 T XT 0 = 0
or
T0
X 00
=
.
X
tT
The LHS is a function of x only whereas the RHS is a function of t only.
This is true only when both sides are constant. That is, there is such that
X 00
T0
=
=
X
tT
and this leads to the two ODEs X 00 = X and T 0 = tT.
2.3 We have xux + (x + 1)yuy =
x x
(e
y
+ xe ) + (x + 1)y
x
xe
y2
= 0 and
235
u(1, 1) = e.
2.4 We have ux +uy +2u = e2y cos (x y)2e2y sin (x y)e2y cos (x y)+
2e2y sin (x y) = 0 and u(x, 0) = sin x.
2.5 (a) The general solution to this equation is u(x) = C where C is an
arbitrary constant.
(b) The general solution is u(x, y) = f (y) where f is an arbitrary diferentiable function of y.
2.6 (a) The general solution to this equation is u(x) = C1 x + C2 where
C1 and C2 are arbitrary constants.
(b)
R We have uy = f (y) where f is an arbitrary function of y. Hence, u(x, y) =
f (y)dy + g(x).
2.7 Let v(x, y) = y + 2x. Then
ux
uxx
uy
uyy
uxy
=2fv (v) + g(v) + 2xgv (v)

=4fvv (v) + 4gv (v) + 4xgvv (v)
=fv (v) + xgv (v)
=fvv (v) + xgvv (v)
=2fvv (v) + gv (v) + 2xgvv (v)
Hence,
uxx 4uxy + 4uyy =4fvv (v) + 4gv (v) + 4xgvv (v)
8fvv (v) 4gv (v) 8xgvv (v)
+4fvv (v) + 4xgvv (v) = 0.
2.8 utt = c2 uxx .
2.9 Let v = x + p(u)t. Using the chain rule we find
ut = fv vt = fv (p(u) + pu ut t).
Thus
(1 tfv pu )ut = fv p.
If 1 tfv pu 0 on any tinterval I then fv p 0 on I which implies that
fv 0 or p 0 on I. But either condition will imply that tfv pu 0 and
236
this will imply that 1 = 1 tfv pu = 0, a contradiction. Hence, we must have

1 tfv pu 6= 0. In this case,
ut =
fv p
.
1 tfv pu
Likewise,
ux = fv (1 + pu ux t)
or
ux =
fv
.
1 tfv pu
It follows that ut = p(u)ux .

If ut = (sin u)ux then p(u) = sin u so that the general solution is given by
u(x, t) = f (x + t sin u)
where f is an arbitrary differentiable function in one variable.
2.10 u(x, y) = xf (x y) + g(x y).
2.11 Using integration by parts, we compute
Z
Z L
L
uxx (x, t)u(x, t)dx = ux (x, t)u(x, t)|x=0
0
u2x (x, t)dx

0
Z L
u2x (x, t)dx
=ux (L, t)u(L, t) ux (0, t)u(0, t)
0
Z L
u2x (x, t)dx 0
=
0
Note that we have used the boundary conditions u(0, t) = u(L, t) = 0 and
the fact that u2x (x, t) 0 for all x [0, L].
2.12 (a) This can be done by plugging in the equations.
(b) Plug in.
(c) We have sup{|un (x, 0) 1| : x R} = n1 sup{| sin nx| : x R} = n1 .
(d) We have sup{|un (x, t) 1| : x R} =
en t
.
n
(e) We have limt sup{|un (x, t) 1| : x R, t > 0} = limt

Hence, the solution is unstable and thus the problem is ill-posed.
en t
n
= .
237
2.13 (a) u(x, y) = x3 + xy 2 + f (y), where f is an arbitrary function.

R
3 2
(b) u(x, y) = x 6y + F (x)R+ g(y), where
F (x) = f (x)dx.
R
1 2x+3t
(c) u(x, t) = 18
e
+ t f1 (x)dx + f2 (x)dx + g(t).
2.14 (b) u(x, y) = xf (y 2x) + g(y 2x).
2.15 We have
ut
utt
ux
uxx
=cuv cuw
=c2 uvv 2c2 uwv + c2 uww
=uv + uw
=uvv + 2uvw + uww
Substituting we find uvw = 0 and solving Rthis equation we find uv = f (v)

and u(v, w) = F (v) + G(w) where F (v) = f (v)dv.
Finally, using the fact that v = x + ct and w = x ct; we get dAlemberts
solution to the one-dimensional wave equation:
u(x, t) = F (x + ct) + G(x ct)
where F and G are arbitrary differentiable functions.
Section 3
2
3.1 y = 21 (1 et ).
3.2 y(t) =
3t1
9
+ e2t + Ce3t .
3.3 y(t) = 3 sin t +

3.4 y(t) =
3 cos t
t
1
(3 sin (3t)
13
+ Ct .
+ 2 cos (3t)) + Ce2t .
3.5 y(t) = Ce sin t 3.

3.6 = 2.
3.7 p(t) = 2 and g(t) = 2t + 3.
238
3.8 y0 = y(0) = 1 and g(t) = 2et + cos t + sin t.

3.9 1.
3.10 y(t) = t ln |t| + 7t.
3.11 Since p(t) = a we find (t) = eat . Suppose first that a = . Then
y 0 + ay = beat
and the corresponding general solution is
y(t) = bteat + Ceat
Thus,
limt y(t) = limt ( ebtat + eCat )
= limt aebat = 0
Now, suppose that a 6= then
y(t) =
b t
e + Ceat
a
Thus,
lim y(t) = 0.
3.12 y(t) = (tet + et )1 .

3.13 y(t) =
t2
4
3t + 12 +
1
.
12t2
3.14 y(t) = tSi(t) + (3 Si(1))t.
Section 4
4.1 y(t) =
3 t2
e
2
+C
t2
4.2 y(t) = Ce 2 2t .
4.3 y(t) = Ct2 + 4.
31
239
4.4 y(t) =
2Ce4t
.
1+Ce4t
4.5 y(t) =
p
5 4 cos (2t).
p
4.6 y(t) = (2 cos t + 4).
4.7 y(t) = e1t 1.
4.8 y(t) =
2
.
4t2 +1
4.9 y(t) = tan (t + ) = cot t.

2
4.10 y(t) =
3et
.
3+et2
4.11 u(x, y) = F (y)e3x + G(x) where F (y) =

4.12 y 2 + cos y + cos t +
t2
2
f (y)dy.
= 2.
4.13 3y 2 y 0 + cos y + 2t = 0, y(2) = 0.

4.14 The ODE is not separable.
Section 5
5.1 (a) Linear (b) Quasi-linear, nonlinear (c) Nonlinear (d) Semi-linear, nonlinear.
5.2 Let w = 2x y. Then ux + 2uy u = ex f (w) + 2ex fw (w) 2ex fw (w)
ex f (w) = 0.
3 1
1 1
5.3 We have xux yuy = x 32 x 2 y 2 y 12 x 2 y 2 = x xy = u. Also,

u(y, y) = y 2 .
5.4 We have yux + xuy = 2xy sin (x2 + y 2 ) + 2xy sin (x2 + y 2 ) = 0. Moreover, u(0, y) = cos y 2 .
240
5.5 We have x1 ux + y1 uy = x1 (x)+ y1 (1+y) = y1 . Moreover, u(x, 1) = 12 (3x2 ).

5.6 3a 7b = 0.
5.7 aut + cu = 0.
5.8 u(x, y) = x + f (x y).
5.9 We have
ux = 4e4x f (2x 3y) + 2e4x f 0 (2x 3y)
uy = 3e4x f 0 (2x 3y)
Thus,
3ux + 2uy + 12u = 12e4x f (2x 3y) + 6e4x f 0 (2x 3y)
6e4x f 0 (2x 3y) + 12e4x f (2x 3y) = 0.
x
5.10 u(x, t) = f (ax bt)e b .

5.11 u(x, y) = f (bx ay).

5.12 cuw + u(v, w) = f w, wv
.
c
5.13 vwv (v) = Aw(v).
Section 6.1
6.1.1 19.
6.1.2 15.
6.1.3 ~u ~v =
6.1.4 63 .
6.1.5 52 .
1
2
and ~u w
~ = 12 .
241
6.1.6 (a) Neither (b) Orthogonal (c) Orthogonal (d) Parallel.
6.1.7 P QR is a right triangle at Q.
6.1.8 ~u =< 13 , 13 , 13 > .
6.1.9 45 .
6.1.10 Comp~a~b =
9
7
and Proj~a~b =<
27 54
18
, , 49
49 49
>.
6.1.11 ~b =< x, y, 3x 2 10 > where y and z are arbitrary numbers.

6.1.12 144 J.
6.1.13 1839 ft lb = 1839 slug.
Section 6.2
6.2.1 F (x, y, z) = (yzexyz + y cos (xy))~i + (xzexyz + x cos (xy))~j + xyexyz~k.
6.2.2 F (x, y, z) = cos
y
z

~i x sin
z
y
z

~j +
xy
z2
sin
y
z

~k.
6.2.3
The level surfaces are spheres centered at (2, 3, 5) and with radius
C, C 0.
6.2.4
12
.
5
6.2.5
1 (3x2
10
+ 6y 3 z 3xy 2xz + yz).
6.2.6 The maximum rate of change is 17 and the maximum occurs in

the direction of
u(0, 2)
1
4
= ~i + ~j.
||u(0, 2)||
17
17
~i
6.2.7 u(x, y, z) = x22x
+y 2
Section 7
2y ~
j
x2 +y 2
+ ez~k.
242
7.1 u(x, y) = 12 y 2 21 y 2 e2x + sin (yex ).

7.2 u(x, y) =
1
.
csc (yex )x
2
7.3 u(x, y) = ex f (x2 + y 2 ).

7.4 u(x, y) = y(2 + e| y | ).
x
7.5 u(x, y) =
1
.
(x4y)2 +1y
7.6 u(x, y) =
1
.
2y
ex2 +e 1 y
7.7 u(x, y) = 32 x 32 xe2y + ey tan1 (xey ).

7.8 u(x, y) = x2 y 2 , xy 0.
7.9 uy = k2 = f (k1 ) = f (y + x3 ).
7.10 u(x, y) = y 4 (y 2 x2 )2 = 2x2 y 2 x4 .
Section 8
8.1 u(x, t) = sin (x 3t).
c
8.2 u(x, y) = k2 e a x = f (bx ay)e a x .

8.3 u(x, y) = x cos (y 2x) + f (y 2x).
dy
8.4 Solving the equation dx
= 1 we find x y = k1 . Solving the equation
du
1 2
=
x
we
find
u(x,
y)
=
x
+ f (x y) where f is a differentiable function
dx
2
2
of one variable. Since u(x, x) = 1 we find 1 = 21 x2 + f (0) or f (0) = 1 x2
which is impossible since f (0) is a constant. Hence, the given initial value
problem has no solution.
8.5 u(x, t) =
e3t
.
1+(x2t)2

8.6 u(x, t) = e3t (x t)2 + 19 13 t 91 .
243
8.7 Using the chain rule we find wt = ut et + uet and wx = ux et . Substituting these equations into the original equation we find
wt et u + cwx et + u = 0
or
wt + cwx = 0
8.8 (a) w(x, t) is a solution to the equation follows from the principle of
superposition. Moreover, w(x, 0) = u(x, 0) v(x, 0) = f (x) g(x).
(b) w(x, t) = f (x ct) g(x ct).
(c) From (b) we see that
sup{|u(x, t) v(x, t)|} = sup{|f (x) g(x)|}.
x,t
Thus, small changes in the initial data produces small changes in the solution. Hence, the problem is a well-posed problem.
8.9

u(x, t) =
8.10 u(x, t) = sin
2x3t
2
g t
x
c
e c x
if x < ct
if x ct.

.
8.11 u(x, y) = x + f (x y).
Section 9
9.1 u = y+f (y ln (y + u)x) where f is an arbitrary differentiable function.
9.2 u =
f (x+y+u)
xy
where f is an arbitrary differentiable function.
9.3 x4 u4 2xyu2 = f (xy) where f is an arbitrary differentiable function.

9.4 xy + u = f (x2 + y 2 u2 ) where f is an arbitrary differentiable function.
244
9.5 x2 + y 2 + u2 = f
y
u
where f is an arbitrary differentiable function.
9.6 x+u = k2 et = et f (u2 x2 ) where f is an arbitrary differentiable function.

9.7 x2 + y 2 + u2 = f (x + y + u) where f is an arbitrary differentiable function.
9.8 x2 + y 2 2u = f (xyu) where f is an arbitrary differentiable function.
9.9 y = sin1 x + f (u) where f is a differentiable function.
9.10 21 (x2 y 2 u2 ) = f (xy
u2
2
where f is a differentiable function.
Section 10
10.1 u(x, y) =
1xy
,
x+y
x + y 6= 0.
10.2 u(x, y) = (x + y)(x2 y 2 ).

10.3 2xyu + x2 + y 2 2u + 2 = 0.
10.4 u(x, y) = ln x + 1
y
x

.
10.5 u(x, y) = f (xey ).

10.6 u(t, x) = f (x at).
10.7 u(x, y) =
1
.
sec (xay)y
10.8 u(x, y) = h y
(x2
2
1
2
ex1 .
10.9 u(x, y) = f (x uy).

10.10 u(x, y) = y sin1 x.
10.11 (i) y = Cx2 . The characteristics are parobolas in the plane centered
at the origin. See figure below.
245
(ii) u(x, y) = eyx .

(iii) In the first case, we cannot substitute x = 0 into yx2 (the argument
of the function f, above) because x2 is not defined at 0. Similarly, in the
second case, wed need to find a function f so that f (0) = h(x). If h is not
constant, it is not possible to satisfy this condition for all x R.
10.12 u(x, y) = ey cos (x y).
10.13 (a) u = ex f (yex ) where f is an arbitrary differential function.
(b) We want 2 = u(x, 3x) = ex f (3ex ex ) = ex f (3). This equation is impossible so this Cauchy problem has no solutions.
(c) We want ex = ex f (ex ex ) = f (1) = 1. In this case, there are infinitely
many solutions to this Cauchy problem, namely, u(x, y) = ex f (yex ) where
f is an arbitrary function satisfying f (1) = 1.
x2
10.14 u(x, y) = 1 + 2e 2 e
(4xy)2
2
10.15 The Cauchy problem has no solutions.

dy
10.16 (a) The characteristics satisfy the ODE dx
= xy . Solving this equation we find x2 y 2 = C. Thus, the characteristics are hyperbolas.
(b)
246
(c) The general solution to the PDE is u(x, y) = f (x2 y 2 ) where f is

2
an arbitrary differentiable function. Since u(0, y) = ey we find f (y) = ey .
2
2
Hence, u(x, y) = ex y .
10.17 (a) infinitely many (b) no solutions.
Section 11
11.1 (a) Hyperbolic (b) Parabolic (c) Elliptic.
11.2 (a) Ellitpic (b) Parabolic (c) Hyperbolic.
11.3 The PDE is of hyperbolic type if 4y 2 (x2 + x + 1) > 0. This is true for
all y 6= 0. Graphically, this is the xyplane with the xaxis removed,
The PDE is of parabolic type if 4y 2 (x2 + x + 1) = 0. Since x2 + x + 1 > 0
for all x R, we must have y = 0. Graphically, this is xaxis.
The PDE is of elliptic type if 4y 2 (x2 + x + 1) < 0 which can not happen.
247
11.4 We have
ux (x, t)
uxx (x, t)
ut (x, t)
utt (x, t)
=
=
=
=
sin x sin t,
cos x sin t,
cos x cos t,
cos x sin t.
Thus,
uxx (x, t)
u(x, 0)
ut (x, 0)
ux (0, t)
=
=
=
=
cos x sin t = utt (x, t),

cos x sin 0 = 0,
cos x cos 0 = cos x,
sin 0 sin t = 0.
11.5 (a) Quasi-linear (b) Semi-linear (c) Linear (d) Nonlinear.

11.6 We have
2x
x2 + y 2
2y 2 2x2
uxx = 2
(x + y 2 )2
2y
uy = 2
x + y2
2x2 2y 2
uyy = 2
(x + y 2 )2
ux =
Plugging these expressions into the equation we find uxx + uyy = 0. Similar
argument holds for the second part of the problem.
11.7 Multiplying the equation by u and integrating, we obtain
Z L
Z L
2
u (x)dx =
uuxx (x)dx
0
0
Z L
=[u(L)ux (L) u(0)ux (0)]
u2x (x)dx
0

Z L
2
2
2
= kL u(L) + k0 u(0) +
ux (x)dx
0
248
For > 0, because k0 , kL > 0, the right-hand side is nonpositive and the
left-hand side is nonnegative. Therefore, both sides must be zero, and there
can be no solution other than u 0, which is the trivial solution.
11.8 Substitute u(x, y) = f (x)g(y) into the left side of the equation to obtain
f (x)g(y)(f (x)g(y))xy = f (x)g(y)f 0 (x)g 0 (y). Now, substitute the same thing
into the right side to obtain (f (x)g(y))x (f (x)g(y))y = f 0 (x)g(y)f (x)g 0 (y) =
f (x)g(y)f 0 (x)g 0 (y). So the sides are equal, which means f (x)g(y) is a solution.
11.9 We have
(un )xx = n2 sin nx sinh ny and (un )yy = n2 sin nx sinh ny
Hence, un = 0.
11.10 u(x, y) =
x2 y 2
4
+ F (x) + G(y), where F (x) =
f (x)dx.
11.11 (a) We have A = 2, B = 4, C = 7 so B 2 4AC = 1656 = 40 < 0.

So this equation is elliptic everywhere in R2 .
(b) We have A = 1, B = 2 cos x, C = sin2 x so B 2 4AC = 4 cos2 x +
4 sin2 x = 4 > 0. So this equation is hyperbolic everywhere in R2 .
(c) We have A = y, B = 2(x 1), C = (y + 2) so B 2 4AC =
4(x 1)2 + 4y(y + 2) = 4[(x 1)2 + (y + 1)2 1]. The equation is parabolic if
(x 1)2 + (y + 1)2 = 1. It is hyperbolic if (x 1)2 + (y + 1)2 > 1 and elliptic
if (x 1)2 + (y + 1)2 < 1.
11.12 Using the chain rule we find
1
1
ut (x, t) = (cf 0 (x + ct) cf 0 (x ct)) + [g(x + ct)(c) g(x ct)(c))
2
2c
c 0
1
0
= (f (x + ct) f (x ct)) + (g(x + ct) + g(x ct))
2
2
c
c2 00
utt = (f (x + ct) + f 00 (x ct)) + (g 0 (x + ct) g 0 (c xt))
2
2
1 0
1
ux (x, t) = (f (x + ct) + f 0 (x ct)) + [g(x + ct) g(x ct)]
2
2c
1 00
1
uxx (x, t) = (f (x + ct) + f 00 (x ct)) + [g 0 (x + ct) g 0 (x ct)]
2
2c
249
By substitutition we see that c2 uxx = utt . Moreover,
Z
1
1 x
u(x, 0) = (f (x) + f (x)) +
g(s)ds = f (x)
2
2c x
and
ut (x, 0) = g(x).
11.13 (a) 1 + 4x2 y > 0, (b) 1 + 4x2 y = 0, (c) 1 + 4x2 y < 0.
11.14 u(x, y) = f (y 3x) + g(x + y).
11.15 u(x, y) = f (y 3x) + g(x + y) =
10x2 +y 2 7xy+6
.
6
Section 12
12.1 Let z(x, t) = v(x, t) + w(x, t). Then we have
c2 zxx =c2 vxx + c2 wxx
=vtt + vtt
=ztt .
12.2 Indeed we have c2 uxx (x, t) = 0 = utt (x, t).
12.3 u(x, t) = 0.
12.4 u(x, t) = 12 (cos (x 3t) + cos (x + 3t)).
h
i
1
1
.
12.5 u(x, t) = 21 1+(x+t)
2 + 1+(xt)2
12.6 u(x, t) = 1 +
12.7
1
[sin (2x
8
+ 4t) sin (2x 4t)].
if x 5t < 0 and x + 5t < 0

if x 5t < 0 and x + 5t > 0
2
u(x, t) =
1
2 if x 5t > 0 and x + 5t < 0
0 if x 5t > 0 and x + 5t > 0

2
12.8 u(x, t) = 12 [e(x+ct) + e(xct) ] + 2t +
1
4c
cos (2x) sin (2ct).
250
12.9 Just plug the translated/differentiated/dialated solution into the wave

equation and check that it is a solution.
12.10 v(r) = A cos (nr) + B sin (nr).
12.11 u(x, t) = 12 [exct + ex+ct + 1c (cos (x ct) cos (x + ct))].
12.12 (a) We have
dE
(t) =
dt
Z
ut utt dx +
c2 ux uxt dx
0
L
2
ut utt dx + c ut (L, t)ux (L, t) c ut (0, t)ux (0, t) c

ut uxx dx
0
0
Z L
2
2
=c ut (L, t)ux (L, t) c ut (0, t)ux (0, t) +
ut (utt c2 uxx )dx
0
2
=c (ut (L, t)ux (L, t) ut (0, t)ux (0, t))

since utt c2 uxx = 0.
(b) Since the ends are fixed, we have ut (0, t) = ut (L, t) = 0. From (a) we
have
dE
(t) = c2 (ut (L, t)ux (L, t) ut (0, t)ux (0, t)) = 0.
dt
(c) Assuming free ends boundary conditions, that is ux (0, t) = ux (L, t) = 0,
(t) = 0.
we find dE
dt
12.13 Using the previous exercise, we find
dE
(t) = d
dt
(ut )2 dx.
The right-hand side is nonpositive, so the energy either decreases or is constant. The latter case can occur only if ut (x, t) is identically zero, which
means that the string is at rest.
12.14 (a) By the chain rule we have ut (x, t) = cR0 (x ct) and utt (x, t) =
c2 R00 (x ct). Likewise, ux (x, t) = R0 (x ct) and uxx = R00 (x ct). Thus,
utt = c2 uxx .
251
(b) We have
1
2
L
2
(ut ) dx =
0
c2 0
[R (x ct)]2 dx =
2
c2
(ux )2 dx.
2
12.15 u(x, t) = x2 + 4t2 + 14 sin 2x sin 4t.
Section 13
13.1 Let z(x, t) = u(x, t) + v(x, t). Then we have
kzxx =kuxx + kvxx
=ut + vt
=zt .
13.2 Indeed we have kuxx (x, t) = 0 = ut (x, t).
13.3 u(x, t) = T0 +
TL T0
x.
L
13.4 Let u be the solution to (13.1) that satisfies u(0, t) = u(L, t) = 0. Let
w(x, t) be the time independent solution to (13.1) that satisfies w(0, t) = T0
and w(L, t) = TL . That is, w(x, t) = T0 + TLLT0 x. From Exercise 13.1,
the function u(x, t) = u(x, t) + w(x, t) is a solution to (13.1) that satisfies u(0, t) = T0 and u(L, t) = TL .
13.5 u(x, t) = 0.
13.6 Substituting u(x, t) = X(x)T (t) into (13.1) we obtain
k
X 00
T0
= .
X
T
Since X only depends on x and T only depends on t, we must have that

there is a constant such that
00
k XX = and
T0
T
= .

X 00 k X = 0 and T 0 T = 0.
252
the ODE
X 00 X = 0 whose general
13.7 (a) Letting = k > 0 we obtain
solution is given by X(x) = Aex + Bex for some constants A and B.

(b) The condition u(0, t) = 0 implies that X(0) = 0 which
in turn implies
L
+ BeL = 0.
A + B = 0. Likewise,
the
condition
u(L,
t)
=
0
implies
Ae
Hence, A(eL eL ) = 0.
(c) If A = 0 then B = 0 and u(x, t) is the trivial solution which contradicts
the assumption that u is non-trivial.Hence, we
must have
A 6= 0.
L
2L
L
= e
or e
= 1. This equa(d) Using (b) and (c) we obtain
e
>
0.
Hence,
we
must
have
< 0 so that
tion is impossible
since
2L
X(x) = A cos (x ) + B sin (x ).

q
13.8 (a)Now, write = k . Then we obtain the equation X 00 + 2 X = 0
whose general solution is given by
X(x) = c1 cos x + c2 sin x.
(b) Using X(0) = 0 we obtain c1 = 0. Since c2 6= 0 we must have sin L = 0
2 2
which implies L = n]pi where n is an integer. Thus, = knL2 , where n
is an integer.

ki2 2
13.9 For each integer i 0 we have ui (x, t) = ci e L2 t sin i
x is a solution
L
to (13.1). By superposition, u(x, t) is also a solution to (13.1). Moreover,
u(0, t) = u(L, t) = 0 since ui (0, t) = ui (L, t) = 0 for i = 1, , n.
13.10 (i) u(0, t) = 0 and u(a, t) = 100 for t > 0.
(ii) ux (0, t) = ux (a, t) = 0 for t > 0.
13.11 Solving this problem we find u(x, t) = et sin x. We have
Z
Z
2t
2
2t
2
E(t) =
[e sin x + e cos x]dx =
e2t dx = e2t .
0
Thus, E 0 (t) = 2e2t < 0 for all t > 0.

13.12 E(t) =
RL
0
f (x)dx + (1 + 4L)t.
13.13 v(x) = x + 2.
253
13.14 (a) v(x) = TL x.
(b) v(x) = T.
(c) v(x) = x + T.
RL
13.15 (a) E(t) = 0 u(x, t)dx.
(b) We integrate the equation in x from 0 to L :
Z
kuxx dx = kux (x, t)|L0 = 0,
ut (x, t)dx =
0
since ux (0, t) = ux (L, t) = 0. The left-hand side can also be written as

d
dt
u(x, t)dx = E 0 (t).
Thus, we have shown that E 0 (t) = 0 so that E(t) is constant.

13.16 (a) The total thermal energy is
Z
u(x, t)dx.
E(t) =
0
We have
dE
=
dt
ut (x, t)dx =
ux |L0
xdx = (7 ) +
+
0
L2
.
2
Hence,
Z
E(t) =
0

L2
f (x)dx + (7 ) +
t.
2

(b) The steady solution (equilibrium) is possible if the right-hand side vanishes:
L2
(7 ) +
=0
2
2
Solving this equation for we find = 7 + L2 .

(c) By integrating the equation uxx + x = 0 we find the steady solution
u(x) =
x3
+ C1 x + C2
6
254
From the condition ux (0) = we find C1 = . The steady solution should

also have the same value of the total energy as the initial condition. This
means

Z L
Z L 3
x
f (x)dx = E(0).
+ x + C2 dx =
6
0
0
Performing the integration and then solving for C2 we find
Z
L
1 L
L3
.
C2 =
f (x)dx +
L 0
24
2
Therefore, the steady-state solution is
Z
1 L
L3
L
x3
u(x) =
f (x)dx +
+ x .
L 0
24
2
6
Section 14
14.1 (a) For all 0 x < 1 we have limn fn (x) = limn xn = 0. Also,
limn fn (1) = 1. Hence, the sequence {fn }
n=1 converges pointwise to f.
(b) Suppose the contrary. Let = 21 . Then there exists a positive integer N
such that for all n N we have
|fn (x) f (x)| <
1
2
for all x [0, 1]. In particular, we have

|fN (x) f (x)| <
1
2
for all x [0, 1]. Choose (0.5) N < x < 1. Then |fN (x)f (x)| = xN > 0.5 =
which is a contradiction. Hence, the given sequence does not converge uniformly.
14.2 For every real number x, we have
nx + x2
x
x2
=
lim
+
lim
=0
n
n n
n n2
n2
lim fn (x) = lim
Thus, {fn }
n=1 converges pointwise to the zero function on R.
255
14.3 For every real number x, we have
1
1
fn (x)
.
n+1
n+1
Moreover,
1
= 0.
lim
n
n+1
Applying the squeeze rule for sequences, we obtain
lim fn (x) = 0
for all x in R. Thus, {fn }

n=1 converges pointwise to the zero function on R.
14.4 First of all, observe that fn (0) = 0 for every n in N. So the sequence
{fn (0)}
n=1 is constant and converges to zero. Now suppose 0 < x < 1 then
n2 xn = n2 en ln x . But ln x < 0 when 0 < x < 1, it follows that
limn fn (x) = 0 for 0 < x < 1
Finally, fn (1) = n2 for all n. So,
lim fn (1) = .
Therefore, {fn }
n=1 is not pointwise convergent on [0, 1].
14.5 For 2 x < 0 and 0 < x
we have
lim (cos x)n = 0.
For x = 0 we have fn (0) = 1 for all n in N. Therefore, {fn }

n=1 converges
pointwise to

0 if 2 x < 0 and 0 < x 2
f (x) =
1
if x = 0.
14.6 (a) Let > 0 be given. Let N be a positive integer such that N > 1 .
Then for n N

n

|x|n
x
1
1
x
=
x
<
< .

n
n
n
N
256
Thus, the given sequence converges uniformly (and pointwise) to the function
f (x) = x.
(b) Since limn fn0 (x) = 1 for all x [0, 1), the sequence {fn0 }
n=1 converges
pointwise to f 0 (x) = 1. However, the convergence is not uniform. To see
this, let = 12 and suppose that the convergence is uniform. Then there is a
positive integer N such that for n N we have
1
|1 xn1 1| = |x|n1 < .
2
In particular, if we let n = N + 1 we must have xN < 21 for all x [0, 1).
1
But x = 12 N [0, 1) and xN = 12 which contradicts xN < 12 . Hence, the
convergence is not uniform.
14.7 (a) The pointwise limit is
0 if 0 x < 1
1
if x = 1
f (x) =
2
1 if 1 < x 2
(b) The convergence cannot be uniform because if it were f would have to
be continuous.
14.8 (a) Let > 0 be given. Note that

2 cos x sin2 x
1
3 .

|fn (x) | =
2
2
2(2n + sin x) 4n
3
= 0 we can find a positive integer N such that if n N
Since limn 4n
3
then 4n < . Thus, for n N and all x R we have
1
3
|fn (x) |
< .
2
4n
This shows that fn
(b) We have
Z
lim
1
2
uniformly on R and also on [2, 7].
Z
fn xdx =
Z
lim fn xdx =
2 n
1
5
dx = .
2
2
257
14.9 We have proved earlier that this sequence converges pointwise to the
discontinuous function

0 if 2 x < 0 and 0 < x 2
f (x) =
1
if x = 0
Therefore, uniform convergence cannot occur for this given sequence.
14.10 (a) Using the squeeze rule we find
lim sup{|fn (x)| : 2 x 5} = 0.
Thus, {fn }
n=1 converges uniformly to the zero function.
(b) We have
Z 5
Z 5
0dx = 0.
fn (x)dx =
lim
n
Section 15.
15.1 (a) We have (f g)(x + T ) = f (x + T )g(x + T ) = f (x)g(x) = (f g)(x).
(b) We have (c1 f +c2 g)(x+T ) = c1 f (x+T )+c2 g(x+T ) = c1 f (x)+c2 g(x) =
(c1 f + c2 g)(x).
15.2 (a) For n 6= m we have

Z
m
n
1 L
(m n)
(m + n)
sin
x sin
x dx =
x cos
x dx
cos
L
L
2 L
L
L
L

1
L
(m + n)
=
sin
x
2 (m + n)
L

L
L
(m n)
sin
x
(m n)
L
L
=0
where we used the trigonometric identiy
1
sin a sin b = [ cos (a + b) + cos (a b)].
2
258
(b) For n 6= m we have

Z L
Z
m
n
1 L
(m + n)
(m n)
cos
x sin
x dx =
sin
x sin
x dx
L
L
2 L
L
L
L

L
(m + n)
1
cos
x
=
2
(m + n)
L

L
L
(m n)
+
cos
x
(m n)
L
L
=0
where we used the trigonometric identiy
1
cos a sin b = [sin (a + b) sin (a b)].
2
15.3 (a) L (b) L (c) 0.
15.4
Z
1
f (x)dx = 0
a0 =

Z
1
an =
f (x) cos nxdx

Z
Z 0
cos nxdx = 0
cos nxdx +
=
0
Z
1
bn =
f (x) sin nxdx

Z
Z 0
=
sin nxdx +
sin nxdx
2
= [1 (1)n ]
n
15.5 f (x) = 61 +
15.6 f (x) =
15.7 f (x) =
2
n=1 n
4
n
n=1 (n)2 (1)

cos
4
n=1 (n)2 [1
n
2
cos (nx).

(1)n sin
(1)n ] cos
n
x
2

.
nx
2
259
15.8 Since the sided limits at the point of discontinuity x = 0 do not exist,
the function is not piecewise continuous in [1, 1].
15.9 Define the function
Z
L+a
g(a) =
f (x)dx.
L+a
Using the fundamental theorem of calculus, we have

Z L+a
dg
d
=
f (x)dx
da da L+a
=f (L + a) f (L + a) = f (L + a + 2L) f (L + a)
=f (L + a) f (L + a) = 0
Hence, g is a constant function, and in particular we can write g(a) = g(0)
for all a R which gives the desired result.

P 1
2n
2nx
2n
1
+
sin
cos
cos
+
1
sin
15.10 (i) f (x) = 10
n=1
3
n
3
3
n
3
(ii) Using the theorem discussed in class, because this function and its derivative are piecewise continuous, the Fourier series will converge to the function
at each point of continuity. At any point of discontinuity, the Fourier series
will converge to the average of the left and right limits.
(iii)
.
15.11 (a) a0 = 2, an = bn = 0 for n N.
(b) a0 = 4, an = 0, b1 = 1, and bn = 0.
1
(c) a0 = 1, an = 0, bn = n
[(1)n 1], n N.
2L
(d) a0 = an = 0, bn = n
(1)n+1 , n N.
2nx
3
260
15.12 1
15.13 an = 0 for all n N.
15.14
f (0 )+f (0+ )
2
+
2
15.15 (a) f (x) = 32 + 2

P
(1)n+1
(b)
n=1 2n1 = 4 .
Section 16
16.1 f (x) = 0.
16.2
16.3
= 0.
n=1
sin (2n1)x
.
2n1
261
16.4
16.5 f (x) =
16.6 f (x) =
16.7 f (x) =
2
n=1 n2 [2 cos (n/2)
2
n
n=1 n2 [(1)
2
n=1 n [1
1 (1)n ] cos nx.
1] cos nx.
(1)n ] sin nx.
262
16.8 f (x) =
n=1 n
1+(1)n
n2 1
16.9 f (x) = 21 (e2 1) +
16.10 (a) If f (x) = sin

(b) If f (x) = 1 then
2
x
L
n=1
sin nx.
4[(1)n e2 1]
4+n2 2
cos (nx).
then bn = 0 if n 6= 2 and b2 = 1.
Z
n
2 L
2
bn =
sin
x dx =
[1 (1)n ].
L 0
L
n

(c) If f (x) = cos L x then
Z

2 L
cos
x sin
x dx = 0
b1 =
L 0
L
L
and for n 6= 1 we have
Z

n
2 L
bn =
cos
x sin
x dx
L 0
L
L
Z
x
i
1 2 L h x
=
sin
(1 + n) sin
(1 n) dx
2L 0
L
L

L
x
x
1
L
L
=
cos
(1 + n) +
cos
(1 n)
L
(1 + n)
L
(1 n)
L
0
2n
= 2
[1 + (1)n ].
(n 1)
16.11 (a) a0 = 10 and a1 = 1, and an = 0 for n 6= 1.
2L
n
(b) a0 = L and an = (n)
n N.
2 [(1) 1],

n
2
(c) a0 = 1 and an = n sin 2 , n N.
16.12 By definition of Fourier sine coefficients,
Z
n
2 L
bn =
f (x) sin
x dx
L 0
L
The symmetry around x = L2 can be written as

L
L
f
+x =f
x
2
2
263
for all x R. To use this symmetry it is convenient to make the change of
variable x L2 = u in the above integral to obtain
L
2
Z
bn =

f
L
2

L
n L
+ u sin
+ u du.
2
L 2

L
nu
Since f L2 + u is even in u and for n even sin n
+
u
=
sin
is
L
2
L
odd in u, the integrand of the above integral is odd in u for n even. Since
the intergral is from L2 to L2 we must have b2n = 0 for n = 0, 1, 2,
16.13 By definition of Fourier cosine coefficients,
2
an =
L
The anti-symmetry around x =

Z
0
n
f (x) cos
x dx
L
L
2
can be written as
L
y
2

= f
L
+y
2
for all y R. To use this symmetry it is convenient to make the change of

variable x = L2 + y in the above integral to obtain
Z
L
2
an =

f
L
2

L
n L
+ y cos
+ y dy.
2
L 2

L
Since f L2 + y is odd in y and for n even cos n
is
+ y = cos ny
L
2
L
even in y, the integrand of the above integral is odd in y for n even. Since
the intergral is from L2 to L2 we must have a2n = 0 for all n = 0, 1, 2, .
16.14 sin
16.15 (a)
x
L
1+(1)n
n=2 n2 1
cos
nx
L

.
264
R2
(b) a0 = 22 0 f (x)dx = 3.
(c) We have
2
an =
2
Z
=
f (x) cos
0
1
nx
nx
dx
Z 2
nx
2 cos
dx +
dx
2
2
0
1
nx 1
nx 2
2
2

sin
+2
sin
=
n
2 0
n
2 1
n
2
=
sin
.
n
2
(d) bn = 0 since f (x) sin

(e)
cos
nx
2
is odd in2 x 2.

n
nx
3 X
2
f (x) = +
sin
cos
.
2 n=1
n
2
2
Section 17
17.1 We look for a solution of the form u(x, y) = X(x)Y (y). Substituting in
the given equation, we obtain
X 00 Y + XY 00 + XY = 0.
265
Assuming X(x)Y (y) is nonzero, dividing for X(x)Y (y) and subtract both
00 (x)
sides for XX(x)
, we find:
X 00 (x)
Y 00 (y)
=
+ .
X(x)
Y (y)
Y 00 (y)
X 00 (x)
=
+ = .
X(x)
Y (y)

X 00 + X = 0 and Y 00 + ( )Y = 0.
If > 0 and > 0 then

p
p
Y (y) =C cos ( )y + D sin ( )y
If > 0 and < 0 then
Y (y) =Ce ()y + De ()y

If = > 0 then

Y (y) =Cy + D
If = < 0 then
X(x) =Ae x + Be
Y (y) =Cy + D
If < 0 and > 0 then
X(x) =Ae
+ Be
p
p
Y (y) =C cos ( )y + D sin ( )y
266
If < 0 and < 0 then
X(x) =Ae
Y (y) =Ce
+ Be
()y
+ De
()y
17.2 Lets assume that the solution can be written in the form u(x, t) =
X(x)T (t). Substituting into the heat equation we obtain
X 00
T0
=
.
X
kT
X 00
X
= and
T0
kT
= .

X 00 X = 0 and T 0 kT = 0.
Case 1: = 0
In this case, X 00 = 0 and T 0 = 0. Solving these equations we find X(x) =
ax + b and T (t) = c.
Case 2: > 0
In this case, X(x) = Ae x + Be x and T (t) = Cekt .

Case 3: < 0
In this case, X(x) = A cos x + B sin x and and T (t) = Cekt .

17.3 r2 R00 (r) + rR0 (r) R(r) = 0 and 00 () + () = 0.
17.4 X 00 = (2 + )X, T 00 = T, X(0) = 0, X 0 (1) = 0.
17.5 X 00 X = 0, T 0 = kT, X(0) = 0 = X 0 (L).
17.6 u(x, t) = Ce(xt) .
17.7 5X 000 7X 00 X = 0 and 3Y 00 Y 0 = 0.
267
17.8 u(x, y) = Cex .

17.9 u(x, y) = Cex y .
17.10 We look for a solution of the form u(x, y) = X(x)T (t). Substituting in the wave equation, we obtain
X 00 (x)T (t) X(x)T 00 (t) = 0.
Assuming X(x)T (t) is nonzero, dividing for X(x)T (t) we find:
T 00 (t)
X 00 (x)
=
.
X(x)
T (t)
of t. This says that they must equal to a constant. That is,
T 00 (t)
X 00 (x)
=
=
X(x)
T (t)
X 00 X = 0 and T 00 T = 0.
The solutions of these equations depend on the sign of .
If > 0 then the solutions are given
X(x) =Ae
T (t) =Ce
+ Be
+ De
where A, B, C, and D are constants. In this case,
u(x, t) = k1 e
(x+t)
+ k2 e
(xt)
+ k3 e
If = 0 then
X(x) =Ax + B
T (t) =Ct + D
(x+t)
+ k4 e
(xt)
268
where A, B, and C are arbitrary constants. In this case,

u(x, t) = k1 xt + k2 x + k3 t + k4 .
If < 0 then
T (t) =A cos t + B sin t

where A, B, C, and D are arbitrary constants. In this case,
u(x, t) =k1 cos x cos t + k2 cos x sin t
+k3 sin x cos t + k4 sin x sin t.

17.11 (a) u(r, t) = R(r)T (t), T 0 (t) = kT, r(rR0 )0 = R.
(b) u(x, t) = X(x)T (t), T 0 = T, kX 00 ( + )X = 0.
(c) u(x, t) = X(x)T (t), T 0 = T, kX 00 aX 0 = X.
(d) u(x, t) = X(x)Y (y), X 00 = X, Y 00 = Y.
(e) u(x, t) = X(x)T (t), T 0 = kT, X 0000 = X.
17.12 u(x, y) = Ce(x+y) .
17.13 X 00 = X, Y 0 Y 00 + Y = Y.
Section 18
18.1 u(x, t) = sin
x
2
2 k t
e 4 + 3 sin
18.2 u(x, t) =
8d
3
18.3 u(x, t) =
18.4 u(x, t) =
1
n=1 (2n1)3
n=1
n=1
Cn sin
sin
1
(4n2 1)
n
x
L
5
x
2
252 k t
e 4 .
(2n1)
x
L
cos
2n
x
L
k(2n1)2 2
t
L2
k 4n2 2 t
L2
e
.
kn2 2 t
e L2 where
4
n = 2, 6, 10,
n
0
n = 4, 8, 12,
Cn =
6
n is odd.
n
269
9
x
L
18.5 u(x, t) = 6 sin

18.6 u(x, t) =
1
2
81k2 t
e L2 .
n=1 Cn cos
Cn =
18.8 u(x, t) = 3 cos

18.9
n = 1, 5, 9,
n = 3, 7, 11,
n is even
2
n
3
x
L
8
x
L
kn2 2 t
e L2 where
2
n
18.7 u(x, t) = 6 + 4 cos
n
x
L
0
9k 2
t
L2
64k2 t
e L2 .
n n2 2
1+ 2 t
L
u(x, t) =
an cos
x e
.
L
n=0
As t , e

2 2
1+ n 2 t
L
0 for each n N. Hence, u(x, t) 0.
18.10 (b) We have

Z
E (t) =2
w(x, t)wt (x, t)dx

0
w(x, t)[wxx (x, t) w(x, t)]dx

Z 1

Z 1
1
2
2
wx (x, t)dx +
w (x, t)dx
= 2w(x, t)wx (x, t)|0 2
0
0

Z 1
Z 1
2
2
=2
wx (x, t)dx +
w (x, t)dx 0
=2
Hence, E is decreasing, and 0 E(t) E(0) for all t > 0.

(c) Since w(x, 0) = 0, we must have E(0) = 0. Hence, E(t) = 0 for all t 0.
This implies that w(x, t) = 0 for all t > 0 and all 0 < x < 1. Therefore
u1 (x, t) = u2 (x, t). This means that the given problem has a unique solution.
18.11 (a) u(0, t) = 0 and ux (1, t) = 0.
270
(b) Lets assume that the solution can be written in the form u(x, t) =
X 00
T0
= .
X
T
X 00
X
= and
T0
T
= .

X 00 X = 0 and T 0 T = 0.
u(0, t) = 0 = X(0)T (t) = X(0) = 0
and
ux (1, t) = 0 = X 0 (1)T (t) = X 0 (1) = 0.
our assumption that
u is the
non-trivial00solution.
0
00
(c) We have X = cos x and X = sin x.
= 0.
Thus, X X
0
Moreover X(0) = 0. Now, X (1) = 0 implies cos = 0 or =

2
n 21 , n N. Hence, = n 12 2 .
18.12 (a) Lets assume that the solution can be written in the form u(x, t) =
X 00
T0
=
.
X
kT
Since the LHS only depends on x and the RHS only depends on t, there must
be a constant such that
X 00
X
= and
T0
kT
= .

X 00 X = 0 and T 0 kT = 0.
271
u(0, t) = 0 = X(0)T (t) = X(0) = 0
and
u(L, t) = 0 = X(L)T (t) = X(L) = 0.
our assumption that u is the non-trivial solution.
Case 1: = 0
In this case, X 00 = 0. Solving this equation we find X(x) = ax + b. Since
X(0) = 0 we find b = 0. Since X(L) = 0 we find a = 0. Hence, X 0 and
u(x, t) 0. That is, u is the trivial solution.
Case 2: > 0
In this case, X(x) = Ae x + Be x . Again, the conditions X(0) = X(L) =

0 imply A = B = 0 and hence the solution is the trivial solution.
Case 3: < 0
In this case, X(x) = A cos x + B sin x. Thecondition X(0) = 0

implies A = 0. The condition X(L) = 0 implies B sin L = 0. We must
have B 6= 0 otherwiseX(x) = 0 andthis leads to the trivial solution. Since
B 6= 0, we obtain sin L = 0 or L = n where n N. Solving for
2 2
we find = nL2 . Thus, we obtain infinitely many solutions given by
Xn (x) = An sin
n
x, n N.
L

T 0 kT = 0
by the method of separation of variables we obtain
Tn (t) = Bn e
n2 2
kt
L2
, n N.

un (x, t) = Cn sin
n n2 2
x e L2 kt , n N
L
272
satisfy ut = kuxx and the boundary conditions u(0, t) = u(L, t) = 0.

Now, in order
for these solutions to satisfy the initial value condition u(x, 0) =

9x
6 sin L , we invoke the superposition principle of linear PDE to write
u(x, t) =
Cn sin
n=1
n n2 2
x e L2 kt .
L
(13.4)
To determine
the unknown constants Cn we use the initial condition u(x, 0) =

6 sin 9x
in
(13.4)
to obtain
L
X

n
9x
Cn sin
=
x .
6 sin
L
L
n=1
By equating coefficients we find C9 = 6 and Cn = 0 if n 6= 9. Hence, the
solution to the problem is given by

9x 8122 kt
u(x, t) = 6 sin
e L .
L
(b) Similar to (a), we find

2 kt
9 2 kt
3
2
u(x, t) = 3 sin
x e L sin
x e L2
L
L
18.13 u(x, t) = cos
(b) u(x, t) = 5.
x
L
2 kt
e L2 + 4 cos
5x
L
252 kt
e L2 .
18.14 u(x, t) = 6 sin xe8t .
Section 19
P

n
n
y
sinh
x
where
B
sin
n
n=1
b
b
Z b
n h
n i1
2
Bn =
f2 (y) sin
y dy sinh
a
.
b 0
b
b
19.1 u(x, y) =

n
n
B
sin
x
sinh
(y
b)
where
n
n=1
a
a
Z a

n
n
2
Bn =
g1 (x) sin
x dx [sinh b ]1 .
a 0
a
a
19.2 u(x, y) =
273
19.3 u(x, y) = 2xy +
3
sinh
sin x sinh y.
19.4 If u(x, y) = x2 y 2 then uxx = 2 and uyy = 2 so that u = 0.

If u(x, y) = 2xy then uxx = uyy = 0 so that u = 0.
19.5
u(x, y) =
h
n
n i
X
n
An cosh
y + Bn sinh
y sin
x.
L
L
L
n=1
where

2
An =
L
Z
0
and

2
Bn =
L
Z
0
n
(f1 (x) + f2 (x)) sin
xdx
L

1
nH
cosh
2L
n
(f2 (x) f1 (x)) sin
xdx
L

1
nH
sinh
2L
19.6 (a) Differentiating term by term with respect to x we find

ux + ivx =
nan (x + iy)n1 .
n=0
Likewise, differentiating term by term with respect to y we find

uy + ivy =
nan i(x + iy)n1 .
n=0
Multiply this equation by i we find

iuy + vy =
nan (x + iy)n1 .
n=0
Hence, ux + ivx = vy iuy which implies ux = vy and vx = uy .

(b) We have uxx = (vy )x = (vx )y = uyy so that u = 0. Similar argument
for v = 0.
19.7 Polar and Cartesian coordinates are related by the expressions x =
274

1
r cos and y = r sin where r = (x2 + y 2 ) 2 and tan = xy . Using the chain
rule we obtain
ux =ur rx + u x = cos ur
sin
u
r
uxx =uxr rx + ux x

sin
sin
= cos urr + 2 u
ur cos
r
r

sin
sin
cos
u
u
+ sin ur + cos ur
r
r
r
cos
uy =ur ry + u y = sin ur +
u
r
uyy =uyr ry + uy y

cos
cos
= sin urr 2 u +
ur sin
r
r

sin
cos
cos
+ cos ur + sin ur
u +
u
r
r
r
Substituting these equations into (21.1) we obtain the dersired equation.
19.8 u(x, y) = u1 (x, y) + u2 (x, y) + u3 (x, y) + u4 (x, y) where
u1 (x, y) = 0
#
n
(1)n
n
2
sin
x
sinh
y
u2 (x, y) =
n sinh 3n
2
2
2
n=1

1
4(x 2)
4
sinh
u3 (x, y) =
sin
y
3
3
sinh 8
3
"
n
n
X
14(1 (1)n )

u4 (x, y) =
sin
y sinh
x .
3
3
n sinh 2n
3
n=1
19.9
u(x, y) =
4
sinh
L
2H
sinh
x
2H

sinh
(x L)
2H

cos
y
.
2H
275
19.10 u(x, t) =
A0
2
n x
A
e
cos n y where
n
n=1
Z
2 H
f (y)dy
A0 =
H 0
Z
2 H
n
An =
ydy.
f (y) cos
H 0
H
19.11
x
20

sin
L
cosh L H + sinh L H
L

3x
5

sin
3
3
3
L
cosh
H
+
sinh
H
L
L
L
u(x, y) =
19.12 u(x, y) = sin (2x)e2y .

19.13 u(x, y) = y.
19.14 u(x, y) =
stant.
19.15 u(x, y) =
1 2
x
2
12 y 2 ax + by + C where C is an arbitrary con-
2 cosh 3y sin 3x
3 sinh 6
5 cosh 10y sin 10x

.
10 sinh 20
Section 20
20.1 u(r, ) = 3r5 sin 5.
20.2 u(r, ) =
n
n=1 r
1(1)n
n2
cos n +
sin n
n
20.3 u(r, ) = C0 + r2 cos 2.

20.4 Substituting C0 , An , and Bn into the right-hand side of u(r, ) we find
Z 2
Z 2
X
1
rn
u(r, ) =
f ()d +
f () [cos n cos n + sin n sin n] d
n
2 0
a
0
n=1
"
#
Z 2

X
1
r n
=
f () 1 + 2
cos n( ) d.
2 0
a
n=1
276
20.5 (a) We have eit = cos t + i sin t and eit = cos t i sin t. The result
follows by adding these two equalities and dividing by 2.
(b) This follows from the fact that
1
cos n( ) = (ein() + ein() ).
2
q
(c) We have |q1 | = ar cos ( )2 + sin ( )2 = ar < 1 since 0 < r < a. A
similar argument shows that |q2 | < 1.
20.6 (a) The first sum is a convergent geometric series with ratio q1 and
sum

X
r n in() ar ei()
e
=
a
1 q1
n=1
=
rei()
a rei()
Similar argument for the second sum.

(b) We have
1+2

X
r n
n=1
cos n( ) =1 +
rei()
a rei()
rei()
a rei()
r
r
+ i()
=1 + i()
ae
r ae
r
r
=1 +
a cos ( ) r ai sin ( )
r
+
a cos ( ) r + ai sin ( )
r[a cos ( ) r + ai sin ( )]
=1 +
a2 + 2ar cos ( ) + r2
r[a cos ( ) r ai sin ( )]
+
a2 2ar cos ( ) + r2
a2 r 2
= 2
.
a 2ar cos ( ) + r2
+
277
20.7 We have
1
u(r, ) =
2
"
f () 1 + 2

X
r n
n=1
2
#
cos n( ) d
1
a r2
=
f () 2
d
2 0
a 2ar cos ( ) + r2
Z
f ()
a2 r2 2
=
d.
2
2
a 2ar cos ( ) + r2
0
Z
20.8 u(r, ) = 2
n+1 n sin n
r n .
n=1 (1)
20.9 (a) Differentiating u(r, t) = R(r)T (t) with respect to r and t we find
utt = RT 00 and ur = R0 T and urr = R00 T.
Substituting these into the given PDE we find

1 0
00
2
00
RT = c R T + R T
r
Dividing both sides by c2 RT we find
R00 1 R0
1 T 00
=
+
.
c2 T
R
rR
Since the RHS of the above equation depends on r only, and the LHS depends
on t only, they must equal to a constant .
(b) The given boundary conditions imply
u(a, t) = 0 = R(a)T (t) = R(a) = 0
u(r, 0) = f (r) = R(r)T (0)
ut (r, 0) = g(r) = R(r)T 0 (0).
If = 0 then R00 + 1r R0 = 0 and this implies R(r) = C ln r. Using the condition
R(a) = 0 we find C = 0 so that R(r) = 0 and hence u 0. If > 0 then
T 00 c2 T = 0. This equation has the solution
T (t) = A cos (c t) + B sin (c t).
278
The condition u(r, 0) = f (r) implies that A = f (r) which is not possible.
Hence, < 0
20.10 (a) Follows from the figure and the definitions of trigonometric functions in a right triangle.
(b) The result follows from equation (20.1).
20.11 By the maximum principle we have
min u(x, y) u(x, y) max u(x, y), (x, y)
(x,y)
(x,y)
But min(x,y) u(x, y) = u(1, 0) = 1 and max(x,y) u(x, y) = u(1, 0) = 3.

Hence,
1 u(x, y) 3
and this implies that u(x, y) > 0 for all (x, y) .
20.12 (i) u(1, 0) = 4 (ii) u(1, 0) = 2.
20.13 Using the maximum principle and the hypothesis on g1 and g2 , for
all (x, y) we have
min u1 (x, y) = min g1 (x, y)
(x,y)
(x,y)
u1 (x, y) max u1 (x, y)

(x,y)
= max g1 (x, y) < min g2 (x, y)

(x,y)
(x,y)
= min g2 (x, y) = min u2 (x, y)

(x,y)
(x,y)
u2 (x, y) max u2 (x, y) = max g2 (x, y)

(x,y)
(x,y)
20.14 We have
2 n
1 n
1 2 n
(r
cos
(n))
+
(r
cos
(n))
+
(r cos (n))
r2
r r
r2 2
=n(n 1)rn2 cos (n) + nrn2 cos (n) rn2 n2 cos (n) = 0
(rn cos (n)) =
Likewise, (rn sin (n)) = 0.
279
1
2
20.15 u(r, ) =
r2
2a2
cos 2.
20.16 u(r, ) = ln 2 + 4

a 3
r
cos 3.
Section 21
21.1 Convergent.
21.2 Divergent.
21.3 Convergent.
21.4
1
,
s3
21.5
1
s2
s > 3.
5s , s > 0.
2
21.6 f (t) = e(t1) does not have a Laplace transform.

21.7
4
s
21.8
es
,
s2
4
s2
2
,
s3
s > 0.
s > 0.
2s
21.9 e s +
n est
21.10 t
1
(es
s2
n
s
e2s ), s 6= 0.
tn1 est dt, s > 0.
21.11 (a) 0 (b) 0.

21.12
5
s+7
1
s2
2
,
s2
s > 2.
21.13 3e2t , t 0.
21.14 2t + et , t 0.
21.15 2(e2t + e2t ), t 0.
21.16
2
s1
+ 5s , s > 1.
280
21.17
es
,
s3
21.18
1
2
21.19
3
,
s2 +36
21.20
s2
,
(s2)2 +9
21.21
2
(s4)3
1
s
s > 3.
s
s2 +4 2
, s > 0.
s > 0.
s > 3.
3
(s4)2
5
,
s4
s > 4.
21.22 2 sin 5t + 4e3t , t 0.

21.23 56 e3t t3 , t 0.
21.24

0,
9(t2)
0t<2
,
t 2.
21.25 3e3t 3et , t 0.

21.26 4[e3(t5) e3(t5) ]H(t 5), t 0.
21.27 y(t) = 2e4t +3[H(t1)H(t3)]3[e4(t1) H(t1)e4(t3) H(t
3)], t 0.
21.28 51 e3t +
1 2t
e
20
21.29
et e2t
.
3
21.30
t
2
21.31
t5
.
120
21.32
1
2
14 e2t , t 0.
sin t.
et + 21 e2t .
21.33 t +
et
2
Section 22
et
.
2
281
22.1 u(x, t) = sin (x t) H(t x) sin (x t).

22.2 u(x, t) = [sin (x t) H(t x) sin (x t)]et .
2
22.3 u(x, t) = 2e4 t sin x + 6e16 t sin 2x.

22.4 u(x, t) = [sin (x t) H(t x) sin (x t)]et .
22.5 u(x, t) = 12 t2 + 21 H(t x)(t x)2 .

22.6 u(x, t) = t 12 x2 H t 21 x2 .
sx

c
22.7 u(x, t) = L1 es2 +1
= H t xc sin t xc .
22.8 u(x, t) = 2 sin x cos 3t.
22.9 u(x, y) = y(x + 1) + 1.
22.10 u(x, t) = c
Rt
0
f (t )H( xc )d.
22.11 u(x, t) = e5x e4t .

s
22.12 u(x, t) = L1 Ts e c x + Ts .
2
22.13 u(x, t) = 5e3 t sin (x).

22.14 u(x, t) = 40et cos x2 .
22.15 u(x, t) = 3 sin x cos 2t.
Section 23
23.1
(1)n i
.
n
23.2 f (x) =
1
2
1
n=1 n
sin
n
2
(einx + einx ).
282

P
23.3 f (x) =
sinh a
23.4 f (x) =
eix eix
.
2i
23.5 f (x) =
1
2
n=
T+
(1)n (a+in) inx

e .
(a2 +n2 )
P1
i inT
n= n [e
1]einx +
P
P
2
2
n inx
23.6 (a) f (x) = 3 + 1
+
n= n2 (1) e
n=1
P
2
4
n
(b) f (x) = 3 +
(1)
cos
nx.
n=1 n2
i inT
n=1 n [e
2
(1)n einx .
n2
23.7 (a)
Z
1
2
a0 =2
12
2
sin xdx = [cos cos ] = 0
2
2
1
2
an =2
sin x cos 2nxdx = 0

12
1
2
bn =2
sin x sin 2nxdx =

12

1]einx .
8n
4n2
c0 =0
4(1)n n
i( 4n2 )
4i(1)n n
cn =
.
4n2
P
(1)n n 2nix
(b) f (x) = 4
.
n= i(14n2 ) e
cn =
23.8 (a)
Z
1 2
(2 x)dx = 4
a0 =
2 2
Z
n
1 2
an =
(2 x) cos
x dx = 0
2 2
2
Z
n
1 2
4(1)n
bn =
(2 x) sin
x dx =
2 2
2
n
P
P
2(1)n+1 i ( in
2(1)n+1 i ( in
(b) f (x) = 2
e 2 x) +
e 2 x) .
n=1
n=1
n
n
283
4
23.9 an = cn + cn = 0. We have for |n| odd bn = i in
=
|n| even bn = 0.
4
n
and for
23.10 Note that for any complex number z we have z + z = 2Re(z) and
z z = 2iIm(z). Thus,
c n + c n = an
which means that an = 2Re(cn ). Likewise, we have
cn cn = ibn
That is ibn = 2iIm(cn ). Hence, bn = 2Im(cn ).
23.11 an = 2Re(cn ) =
23.12 f (x) = i
n=
1
n
sin (nT ) and bn =
1cos (nT )
.
n
i sin (2in) in x
e 2 .
2in
23.13 (a) We have

f (t) =
1 0<t<1
0 1<t<2
and f (t + 2) = f (t) for all t R.

(b) We have
Z
Z 2
Z 1
2 L
f (x)dx =
dx =
dx = 1
a0 =
L 0
0
0
Z 1
sin n
an =
cos nxdx =
= 0.
n
0
(c) We have
Z
bn =
sin nxdx =
0
1 cos n
1 (1)n
=
.
n
n
Hence,

bn =
(d) We have c0 =
a0
2
1
2
2
n
if n is odd
if n is even
and for n N we have

i
an ibn
n
if n is odd
cn =
=
0
if
n is even
2
=
284
23.14 sin 3x = 21 (e3ix e3ix ).
Section 24
24.1
f() =
24.2
2 sin if 6= 0
2
if = 0.
u
+ ic
u=0
t
u(, 0) = f().
24.3
2 u
= c2 2 u
2
t
u(, 0) = f()
ut (, 0) = g().
24.4
uyy = 2 u
u(, 0) = 0, u(, L) =
24.5
1
i
1
+i
2 sin a
.
2
.
2 + 2
24.6 We have
Z
F[e H(x)] =
ex H(x)eix dx
Z
=
e
0
x(1+i)

ex(1+i)
1
=
.
dx =

1 + i 0
1 + i
24.7 Using the duality property, we have

1
F
= F[F[e H()]] = 2e H().
1 + ix
285
24.8 We have
Z
f (x )eix dx
Z
i
f (u)eiu du
=e
F[f (x )] =
=e
f()
where u = x .
24.9 We have
ix
F[e
ix
f (x)] =
ix
f (x)e
dx =
f (x)ei()x dx = f( ).
24.10 We will just prove the first one. We have

F[cos (x)f (x)] =F[
f (x)eix
eix
+ f (x)
2
2
1
= [F[f (x)eix ] + F[f (x)eix ]]
2
1
= [f ( ) + f( + )].
2
24.11 Using the definition and integration by parts we find
Z
0
f 0 (x)eix dx
F[f (x)] =
Z

ix
= f (x)e
+ (i)
f (x)eix dx
=(i)f()
where we used the fact that limx f (x) = 0.
24.12
2
(1
2
cos ).
24.13
2
(1
i
cos a).
286

2
x
24.14 F 1 [f()] = 12 e 2 .

1
= eax , x 0.
24.15 F 1 a+i
Section 25
25.1 u(x, t) = F 1 [u(, t)] = e
(xct)2
4
25.2
r
t 1 2 (kt+ )
4 ]
u(x, t) =
e F [e
4
r
r
x2
t
=
e
e 4(kt+/4)
4
kt + /4
r
x2
e 4kt+ et .
=
4kt +
25.3 u(x, t) =
1
4kt
R
0
(xs)2
4kt
ds.
25.4
2
u(x, t) =et F 1 [e t ]
1 x2
=et
e 4t .
4t
25.5 We have
Z
||y ix
Z
d =
0
y ix
d +
0

1
(y+ix)
=
e

y + ix
e e
ey eix d

1
(y+ix)
e

y ix
0
1
1
2y
=
+
= 2
.
y + ix y ix
x + y2
287
25.6
Z
1
u(x, y) =
f()e||y eix d
2

2y
1
= f (x) 2
2
x + y2
Z
1
2y
=
f (x)
d.
2
(x )2 + y 2
25.7 utt + ( + )
ut + u = c2 2 u.
25.8 u(x, t) = e(x3t) .
25.9 u(x, t) = e(xkt) .
25.10 u(x, t) =
1
4kt
2
2 (xs)
4kt
es
ds.
25.11 u(x, t) = (x ct)2 .

1 ||y
25.12 u(x, t) = f (x) F 1 [ ||
e
].
288
Index
Boundary value problem, 17
Burgers equation, 11
Cauchy data, 75
Cauchy problem, 75
Characteristic curve, 60
Characteristic direction, 59
Characteristic equation, 140
Characteristic equations, 60, 72
Characteristics, 60
Classical solution, 13
Convection (transport) equation, 11
Convolution, 183
Descriminant, 88
Differential equation, 5
Diffusion equation, 11
Diffusivity constant, 102
Directional derivative, 50
Dirichlet boundary conditions, 17
Dirichlet conditions, 104
Dot product, 41
Eigenvalue problem, 155
Elliptic, 88
Euler equation, 166
Euler-Fourier Formulas, 125
Even extension, 135
Even function, 133
Evolution equation, 177
Exponential order at infinity, 180
First derivative, 4
First order PDE, 36
Forced harmonic oscillator, 11
Fourier coefficients, 121
Fourier cosine series, 135
Fourier inversion formula, 207
Fourier law, 102
Fourier series, 109, 121
Fourier sine series, 135
Fourier transform, 207
Function series, 120
Fundamental period, 122
General solution, 13
Generalized solution, 14
Gradient, 51
Gradient vector field, 56
Harmonic function, 154
Heat equation, 89, 100
Heat source, 103
Helmholtz equation, 155
Homogeneous, 37, 88
Homogeneous linear PDE, 8
Hyperbolic, 88
Ill-posed, 19
Initial curve, 75
Initial data, 75
Initial temperature distribution, 104
Initial value conditions, 18
289
290
Initial value problem, 18
Inner product, 123
integral curve, 57
Integral surface, 13, 59
integral transforms, 177
Integrating factor, 26
Inverse Laplace transform, 182
Korteweg-Vries equation, 11
Lagranges method, 71
Laplace equation, 89, 154
Laplace transform, 178
Laplaces equation, 11
Laplacian, 154
Level curve, 54
Level surface, 54
Linear, 7, 36
linear, 88
Linear differential operator, 8
Linear operator, 8
INDEX
Order, 6
Ordinary differential equation, 5
Orthogonal, 123
Orthogonal projection, 45
Parabolic, 88
Partial differential equation, 5
Piecewise continuous, 124, 180
Piecewise smooth, 124
Pointwise convergence, 109, 120
Poisson Equation, 11
Poisson equation, 155
Projected characteristic curve, 60
Quasi-linear, 7, 36, 88
Right traveling wave, 67
Scalar projection, 46
Semi-linear, 7, 36, 88
Separable, 31
Separation of variables, 31
Method of characteristics, 59, 66, 71 Smooth functions, 8
Method of undetermined coefficients, Solution surface, 13
Specific heat, 101
222
Method of Variation of Parameters, Squeeze rule, 117
Stable solution, 18
229
stationary equation, 177
Minimal surface equation, 11
Strong solution, 13
Mixed boundary condition, 18
Superposition principle, 15
Mutually orthogonal, 123
Neumann boundary conditions, 17, 105 Thermal conductivity, 102
Thermal energy, 100
Non-homogeneous, 37, 88
Thin film equation, 11
Non-homogeneous PDE, 8
Total thermal energy, 103
Non-linear, 7, 36
Transport equation, 18
Nowhere characteristic, 77
Transport equation with decay, 68
Odd extension, 134
Transport equationin 1-D space, 66
Odd function, 133
Trigonometric series, 121
INDEX
Uniform convergence, 110, 121
Vector field, 55
Vector function, 53
Wave equation, 89, 93
wave equation, 11
Weak solution, 14
Weierstrass M-test, 121
Well-posed, 18
291

Pde Book

Uploaded by

Copyright:

Available Formats

Pde Book

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Pde Book

Uploaded by

Copyright:

Available Formats

Lecture Notes in Mathematics

Arkansas Tech University

A First Course in Quasi-Linear Partial

The Basics of the Theory of Partial Differential Equation

Second Order Linear Partial Differential Equations

The Laplace Transform Solutions for PDEs

The Basics of the Theory of

4THE BASICS OF THE THEORY OF PARTIAL DIFFERENTIAL EQUATION

provided that the limit exists.

(y 2 sin xy) = sin xy y

6THE BASICS OF THE THEORY OF PARTIAL DIFFERENTIAL EQUATION

If a(x, y, u) = (x, y) and b(x, y, u) = (x, y) then (1.3) is called semi-linear.

The semi-linear case has the form

8THE BASICS OF THE THEORY OF PARTIAL DIFFERENTIAL EQUATION

Smooth functions are functions that are continuously differentiable up to a certain

10THE BASICS OF THE THEORY OF PARTIAL DIFFERENTIAL EQUATION

12THE BASICS OF THE THEORY OF PARTIAL DIFFERENTIAL EQUATION

2 SOLUTIONS TO PDES/PDES WITH CONSTRAINTS

2 Solutions to PDEs/PDEs with constraints

the given function is a classical solution to the given equation

14THE BASICS OF THE THEORY OF PARTIAL DIFFERENTIAL EQUATION

2 SOLUTIONS TO PDES/PDES WITH CONSTRAINTS

16THE BASICS OF THE THEORY OF PARTIAL DIFFERENTIAL EQUATION

2 SOLUTIONS TO PDES/PDES WITH CONSTRAINTS

leads to the creation of lots of solutions. In contrast, non-linear equations

18THE BASICS OF THE THEORY OF PARTIAL DIFFERENTIAL EQUATION

2 SOLUTIONS TO PDES/PDES WITH CONSTRAINTS

= 0. Moreover, u (x, 0) = 0 and

20THE BASICS OF THE THEORY OF PARTIAL DIFFERENTIAL EQUATION

(b) Show that sup{| t

= 0. Moreover, u (x, 0) = 0 and

Thus, a small change in the initial data leads to a catastrophically change in

2 SOLUTIONS TO PDES/PDES WITH CONSTRAINTS

is the solution to this problem.

22THE BASICS OF THE THEORY OF PARTIAL DIFFERENTIAL EQUATION

2 SOLUTIONS TO PDES/PDES WITH CONSTRAINTS

(c) Find sup{|un (x, 0) 1| : x R}.

24THE BASICS OF THE THEORY OF PARTIAL DIFFERENTIAL EQUATION

Review of Some ODE Results

REVIEW OF SOME ODE RESULTS

3 The Method of Integrating Factor

3 THE METHOD OF INTEGRATING FACTOR

REVIEW OF SOME ODE RESULTS

Solving this equation using the integrating factor method we find

where f is an arbitrary differentiable function of f. Switching back to x and

3 THE METHOD OF INTEGRATING FACTOR

REVIEW OF SOME ODE RESULTS

4 THE METHOD OF SEPARATION OF VARIABLES FOR ODES

4 The Method of Separation of Variables for

REVIEW OF SOME ODE RESULTS

Since y(0) = 1, we have y(t) = 4 cos t + 5. The interval of existence of

4 THE METHOD OF SEPARATION OF VARIABLES FOR ODES

REVIEW OF SOME ODE RESULTS

First Order Partial Differential

FIRST ORDER PARTIAL DIFFERENTIAL EQUATIONS

5 Classification of First Order PDEs

In what follows the functions a, b, and c are assumed to be continuously

5 CLASSIFICATION OF FIRST ORDER PDES

FIRST ORDER PARTIAL DIFFERENTIAL EQUATIONS

= 0. Moreover, u (x, 0) = 0 and

= 0. Moreover, u (x, 0) = 0 and